Citation
The Use of information and budget decision : how administrative capacity and information play a role in local budgeting in Thailand

Material Information

Title:
The Use of information and budget decision : how administrative capacity and information play a role in local budgeting in Thailand
Creator:
Chuayprakong, Pawinee
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Doctorate ( Doctor of philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Public Affairs, CU Denver
Degree Disciplines:
Public affairs
Committee Chair:
Martell, Christine
Committee Members:
Guy, Mary E.
Ely, Todd
Suwanmala, Charas

Notes

Abstract:
The current budget reform aims to move a budget process, in which incrementalism prevails, towards a rational decision-making. This study investigates such reform in Thailand’s local governments with mixed methods research design. The results show that, despite budgetary incrementalism, administrative capacity—skills in participatory budgeting, communication, financial management and accounting, and strategic planning—and spending preferences among local frontline public administrators lead local budgeting to an information-based decision-making process. During this information-based decision-making process, the levels of information utilization, the strategic of information use (i.e., the storytelling styles), and the professional judgment among frontline public administrators can drive the municipalities’ budgetary decisions to be more efficient and effective. These findings confirms that administrative capacity and information are indispensable for a rational decision-making process as well as for efficient and effective budgetary decisions. As a result, these findings can be used to promote the use of information for the developing countries.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Pawinee Chuayprakong. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
THE USE OF INFORMATION AND BUDGET DECISION: HOW ADMINISTRATIVE
CAPACITY AND INFORMATION PLAY A ROLE IN LOCAL BUDGETING IN
THAILAND
By
Pawinee Chuayprakong B.A., Chulalongkorn University, 2005 MPP, Graduate Institute for Policy Studies, 2008 MPA, Chulalongkorn University, 2009
A thesis submitted to the Faculty of the Graduate School of the University of Colorado Denver in partial fulfillment Of the requirements for the degree of Doctor of Philosophy Public Administration 2019
1


©2019
PAWINEE CHUAYPRAKONG
ALL RIGHTS RESERVED


This thesis for the Doctor of Philosophy degree by
Pawinee Chuayprakong has been approved for the Public Affairs Program by
Christine R. Martell, Chair Mary E. Guy Todd Ely
Charas Suwanmala
Date: August 3, 2019


Chuayprakong, Pawinee (PhD, Public Affairs)
The Use of Information and Budget Decision: How Administrative Capacity and Information Play a Role in Local Budgeting in Thailand Dissertation directed by Professor Christine R. Martell
ABSTRACT
The current budget reform aims to move a budget process, in which incrementalism prevails, towards a rational decision-making. This study investigates such reform in Thailand’s local governments with mixed methods research design. The results show that, despite budgetary incrementalism, administrative capacity—skills in participatory budgeting, communication, financial management and accounting, and strategic planning—and spending preferences among local frontline public administrators lead local budgeting to an information-based decisionmaking process. During this information-based decision-making process, the levels of information utilization, the strategic of information use (i.e., the storytelling styles), and the professional judgment among frontline public administrators can drive the municipalities’ budgetary decisions to be more efficient and effective. These findings confirms that administrative capacity and information are indispensable for a rational decision-making process as well as for efficient and effective budgetary decisions. As a result, these findings can be used to promote the use of information for the developing countries.
The form and content of this abstract are approved. I recommend its publication.
Approved: Christine R. Martell
IV


This dissertation is dedicated to my parents—Kanueng Chuayprakong and Nongchawee Chuayprakong—who are uneducated but always give their all to my education. Every day, your unconditional love and support inspire me to be a better person and convince me that I can make
a difference.


ACKNOWLEDGEMENTS
I would like to express the deepest gratitude and appreciation to my dissertation advisor, Prof. Christine R. Martell. Her academic advice, life guidance, and incredible patience are absolutely an integral part of my graduation. Her faith in me always helped make me get out of every single struggle alive throughout my eight years in this PhD program.
Next, I highly appreciate all help and suggestions from my dissertation committee members. Prof. Mary E. Guy helped and reminded me look at a big picture of the field of public administration and public management. She made sure that I sensibly applied it to my dissertation’s overarching structure. Meanwhile, Prof. Todd Ely gave several methodological suggestions to improve my analytical components. Also, I would like to thank Prof. Charas Suwanmala who offered insightful comments regarding Thailand’s decentralization policy and local public finance. Wisdom and kindness from my dissertation committee will definitely shape my scholarship as well as my career in academia.
Finally, I would like to thank my sister who set the high bar since the first grade until the PhD level. I followed her school path and could not succeed it without her support. Also, I highly appreciate the everlasting help and guidance from Prof. Weerask Kreuathep who talked me into this amazing PhD journey and introduced me to a life of academic adventure. Moreover, I would like to thank my friend in Denver—Plaa, Scott, and Pi New—for their food and comfort especially during the lonely, cold days. Also, I appreciate the moral support from my friends in Bangkok—Amm, Bee, Pi Den, Om, Nhua, Pi Phueng, Tong, Pailin, Pi Wasin, Whan, June, and Anne who were there when I occasionally wanted to give up. Your listening and beverage matter!
VI


TABLE OF CONTENTS
I. INTRODUCTION........................................................................1
Budget Decision, a Product of Rationality?...........................................2
Politics-administration dichotomy in budgeting....................................2
Progress on Information Utilization and Its Impact on Budget Decision Making.........3
Information utilization in a budget process.......................................3
Information utilization and budget decisions......................................4
Performance information...........................................................5
Developed countries and administrative capacity...................................5
Thailand, the least-likely case...................................................6
Research Questions...................................................................7
Methodology.......................................................................8
Contributions to Theory and Practice..............................................8
II. LITERATURE REVIEW.................................................................11
Theories of Information Economics...................................................12
Information Asymmetry............................................................12
Principal-Agent Theory...........................................................15
Empirical Studies on Information Asymmetry, Information Use, and Decision-Makings... 21
Budget Rationality Model............................................................26
Administrative Capacity and Information Utilization..............................27
Budget Orientations and Information Utilization..................................29
Findings from the Theoretical Review................................................32
Research Questions and Hypotheses...................................................34
Rival Hypotheses....................................................................35
Theoretical Framework...............................................................37
III. CASE JUSTIFICATION AND THAILAND’S LOCAL PUBLIC FINANCE AND
BUDGETING.............................................................................39
The Case Justification for Thailand’s Local Governments.............................39
Representativeness...............................................................39
Causal leverage..................................................................40
Thailand as a least likely case..................................................42
Thailand’s Local Public Finance and Budgeting.......................................45
vii


Thailand’s Decentralization and Local Governments...................................45
Local Governments’ Revenue and Expenditure Structure................................48
Local Budget Process................................................................51
Information Use in Local Budgeting in Thailand......................................56
IV. RESEARCH DESIGN AND METHODOLOGY....................................................58
The quantitative mixed methods design.................................................59
The Quantitative Approach.............................................................60
Sampling Design.....................................................................60
Operationalizing Concepts of Interest.................................................65
Pilot Study.........................................................................65
Quantitative Data Analysis..........................................................78
The Qualitative Approach..............................................................80
Sampling Design.....................................................................80
Qualitative Analysis..................................................................84
Tri angulation........................................................................85
V. QUANTITATIVE RESULTS...............................................................87
The Overall Descriptive Statistics of the Survey Participants and the Study Variables.87
Demographic data....................................................................87
The Study Variables.................................................................90
Information, Administrative Capacity, Budget Rationality, and Information Utilization... 106
Administrative Capacity, Information Utilization, and Budget Decisions.............120
Actual Budget Allocation...........................................................135
VI. THE QUALITATIVE RESULTS........................................................... 152
The Determinants of Information Utilization..........................................152
Information Utilization............................................................152
Factors Affecting Information Utilization..........................................153
The Determinants of Budget Allocation..............................................161
VII. DISCUSSION ON INFORMATION UTILIZATION.............................................167
Overall and Function-Based Information Utilization...................................167
Overall information utilization....................................................167
Information Utilization by the Education Department Heads..........................168
Information Utilization by the Public Works Department Heads.......................169
viii


Discussion and Findings...........................................................170
Administrative Capacity, Relevant Information, and Information Utilization......170
Organization Cultures and Information Utilization...............................177
Theoretical Contributions.........................................................182
Empirical Contributions to the Current Literature and Empirical Studies...........184
Practical Contributions...........................................................187
VIII. DISCUSSION ON BUDGET ALLOCATION...............................................189
Actual and Perceptional Budget Allocation.........................................189
Actual Budget Allocation........................................................189
Perceptional Budget Allocation..................................................190
Discussion and Findings...........................................................191
Budgetary Incrementalism and Municipal Finance..................................191
How Management Matters in Local Budgeting.......................................193
Intergovernmental Relations: The Indomitable Centralization.....................198
Theoretical Contributions.........................................................199
Contributions to the Current Literature...........................................199
Practical Contributions...........................................................201
IX. CONCLUSION....................................................................203
Research Questions................................................................203
Theoretical Relevancy.............................................................204
Policy Implications...............................................................204
Threats to Legitimation and Future Research.......................................205
REFERENCES..........................................................................209
APPENDIX
A. The Ordinal Regression Model for Information Utilization.......................226
_B. The Regression Model for Perceptional Budget Allocation.........................230
C. The Regression Model for Impact of Administrative Capacity and Information Utilization on Actual Budget Allocation.........................................................233
IX


LIST OF TABLES
Table 2.1 Relevant Theories to Understand the Interrelationship between Capacity, Information,
and Budget Decisions....................................................................12
Table 2.2 Budget Orientations, Relevant and Important Information Considered, and Norms ... 31 Table 3.1 Types of Local Governments in Thailand and Their Differences 46
Table 3.2 Source of Local Governments’ Revenue (Unit: Million Baht)....................49
Table 3.3 Local Governments’ Expenditure in 2012 (Unit: Million Baht)...................51
Table 4.1 The Stratified Sampling and Sample Size for the Survey........................62
Table 4.2 Sampling Design...............................................................65
Table 4.3 Interview Protocol Based on the Literature....................................68
Table 4.4 Interview Sites for the Subgroup Design.......................................82
Table 4.5 Interview Protocol............................................................83
Table 5.1 Demographic Data of the Participating Municipalities..........................89
Table 5.2 Demographic Data of Profile of Participants in the Survey the Survey Respondents.. 90
Table 5.3 Frequency for Information Pieces Used During the Budget Process...............92
Table 5.4 Descriptive Statistics for Reliability of Each Information Piece..............94
Table 5.5 Descriptive Statistics for Relevance of Each Information Piece................95
Table 5.6 Frequency and Descriptive Statistics for Administrative Capacity..............97
Table 5.7 Frequency of Budget Rationality and Spending Preference.......................98
Table 5.8 Frequency and Descriptive Statistics for Information Utilization for Budget Discussion
.......................................................................................100
Table 5.9 Frequency for Strategies of Information Utilization the Department Heads Always
Apply..................................................................................101
Table 5.10 Frequency and Descriptive Statistics for Perceptional Budget Allocation.....103
Table 5.11 ANOVA for a Comparison of Spending Efficiency and Effectiveness according to
Municipality Size......................................................................104
Table 5.12 ANOVA for a Comparison of Actual Budget Allocation (BTH1,000 per capita) for
FY2019.................................................................................105
Table 5.13 Frequency and Descriptive Statistics for Information Use by Department Heads... 107
Table 5.14 Regression Results of Information Utilization for Budget Discussion among all
Department Heads.......................................................................108
Table 5.15 Frequency and Descriptive Statistics for Information Use by the Education
Department Heads.......................................................................Ill
Table 5.16 Regression Results of Information Utilization for Budget Discussion among all
Education Department Heads.............................................................112
Table 5.17 Frequency and Descriptive Statistics for Information Use by Public Works
Department Heads.......................................................................115
Table 5.18 Regression Results of Information Utilization for Budget Discussion among Public
Works Department Heads.................................................................117
Table 5.19 The Variables Predicting the Perceptional Budget Allocation for the Education and
Public Works Departments...............................................................121
Table 5.20 Spending Efficiency and Effectiveness.......................................123
Table 5.21 The Variables Predicting the Perceptional Budget Allocation in the Education
Department.............................................................................127
Table 5.22 Spending Efficiency and Effectiveness in the Education Departments..........128


Table 5.23 The Variables Forecasting the Perceptional Budget Allocation in the Public Works
Departments...........................................................................132
Table 5.24 Spending Efficiency and Effectiveness in the Public Works Departments......133
Table 5.25 The Frequency and Descriptive Statistics of Variables Forecasting the Total
Education Expenditure..................................................................137
Table 5.26 The Variables Forecasting the Education Expenditure (Decrease).............138
Table 5.27 The Frequency and Descriptive Statistics of Variables Predicting the Education
Expenditure............................................................................140
Table 5.28 The Regression Results for the Education Expenditure (Increase)............142
Table 5.29 The Frequency and Descriptive Statistics of Variables Predicting the Public Works
Expenditure............................................................................144
Table 5.30 The Regression Results for the Public Works Expenditure (Decrease).........145
Table 5.31 The Frequency and Descriptive Statistics of Variables Predicting the Public Works
Expenditure............................................................................147
Table 5.32 The Regression Results for the Public Works Expenditure (Increase).........148
Table 5.33 Summary of Regression Results for the Information Utilization...............150
Table 5.34 Summary of Regression Results for the Spending Efficiency and Effectiveness .... 151 Table 5.35 Summary of Regression Results for the Spending Efficiency and Effectiveness .... 151
Table 6.1 Summary of Determinants of Information Utilization...........................161
Table 6.2 Determinants of Budget Allocation............................................165
XI


LIST OF FIGURES
Figure 2.1 The causal graph of theories of information economics under the context of
information utilization and budget decisions..............................................21
Figure 2.2 Budget Rationality and Information Utilization.................................27
Figure 2.3 Theoretical Framework of Administrative Capacity, the Use of Information, and Budget Decisions...........................................................................37
xii


CHAPTER I
INTRODUCTION
Information can resolve uncertainty and reduce ambiguity, thereby assisting in decisionmaking (Faibisoff & Ely, 1974; Feldman, 1989). Information always has economic value and anyone who possesses it has advantage over others who do not (Arrow, 1962). Lack of information or false information can lead to both market and government failures (Akerlof, 1970; Niskanen, 1971; Salamon, 1987; Seldon, 2002; Tullock, 2002). Hence, information is a key element in an administrative decision-making process as it is used in formulating goals, searching for alternatives, and calculating consequences of each alternative (Downs, 1957; Lindblom, 1959; Simon, 1957). Nonetheless, how information is used for administrative accomplishment is not well understood in administrative science (Simon, 1946; Van De Walle & Van Dooren, 2008). Its effects during budget process and on budget allocation are also still inconclusive in research (Joyce, 2008; Kroll, 2015; Moynihan, 2015; Pollitt, 2006) evidenced by the fact that continues to be the debate on whether budgeting is a product of incremental politics or a product of rationality.
This study considers budgeting a product of rationality and attempts to disentangle how related information has an impact on the information utilization during a budget process and on budget decision making. As an introduction, this chapter first shows that governments gear their budget reforms toward rationality despite the fact that budget process is also of incremental politics. After that, it presents the evolution of information utilization during a budget process and how such utilization eventually influences budget decision making. Since such overview leaves some inadequately explored areas, this chapter then sets the research question
l


accordingly. This research question directs this study to relevant literature and appropriate methodology discussed in the following chapters.
Budget Decision, a Product of Rationality?
Politics-administration dichotomy in budgeting
Somewhat like public administration, the study of public budgeting has revealed an ongoing struggle between politics (i.e., social, political, and legal aspects) and administration or management (i.e., economic and technical aspects) (Goodnow, 1900; Rubin, 2015; Thurmaier, 2001; Wilson, 1887). On one hand, budget decisions are a product of bureaucratic inertia and budget-maximizing behaviors (Niskanen, 1971; Wildavsky, 1978), negotiation between politicians and pressure groups or public service clientele (Smith, 2003), and also logrolling and pork-barrel legislation (Buchanan & Tullock, 1962). On the other hand, they are the product of a rational decision-making process which reflects the comprehensive cost/benefit analysis (Key, 1940), the organization’s planned objectives (Schick, 1990), and fiscally adaptive strategies (Levine, 1978).
Nonetheless, from an administrative standpoint, the latter approach is normatively preferred. As a result, budget reforms always gear a budget process toward managerial improvement, program effectiveness, and progressive roles of budget participants (Caiden, 2010; Schick, 1966). Budget reforms require budget participants to possess both capacity to budget and adequate information in order for them to instrumentally utilize information—to conduct budget analysis and then make better budget decisions (Weiss, 1977). Even though there is still a long way to go for the instrumental use of information, the literature demonstrates that it is making a progress (Pandey, 2015). The following section discusses this progress.
2


Progress on Information Utilization and Its Impact on Budget Decision Making Information utilization in a budget process
No matter how or whether it is used, organizations always gather, store, and communicate information, especially the information pieces that selectively signal the structures of their environments and their functioning in relation to the environment (Feldman & March, 1981; Katz & Kahn, 1966). In the public sector, this informational behavior has been promoted after the advent of information technology and the New Public Management (NPM) which aims to enhance performance through utilizing such information as public services’ output, outcome, and results for monitoring and evaluation purposes (Dahler-Larsen, 2007; Hood, 1991, 2007; Lynn, 2006). Though related information can improve performance via changes in resource allocation that can expand, modify, or terminate public programs according to their performances (Pollitt, 2006; Julnes, 2008), the use of information has actually been inadequately applied into the realm of budgeting.
A budget process is highly complex and theoretically unsettled because it is not exclusively social, political, legal, economic, or technical in its rationality (Key, 1940;
Thurmaier, 2001). On one hand, a budgeting process is, and should be, more rational-comprehensive, relying on economic and technical rationality in order to reach informed and thoughtful decisions (Lindblom, 1959; Thurmaier, 2001). They can be influenced by a statement of objectives and by weighing costs and benefits of relevant alternatives (Schick, 1966). On the other hand, a budgeting process is of incremental politics. Except for very special circumstances, decision makers usually rely on simple calculations or heuristic rules and eventually only find satisfactory solutions for policy problems (Davis, Dempster, & Wildavsky, 1966; True, Jones, & Baumgartner, 2007). Not only does incremental politics not require complex calculation, but also
3


it limits negotiations and conflicts during the budget process, allowing for political acceptance (Nielsen & Baekgaard, 2015). Hence, incremental politics of budget decision process reigns supreme virtually everywhere (Wildavsky, 1978). This incremental politics also significantly contributes to the limited use of information to improve a budgetary process and appropriations, despite that fact that the information technology and NPM movements promote the information gathering and utilization (Joyce, 2011; Melitski & Manoharan, 2014; Moynihan, 2006). Accordingly, many scholars believe that to use information in the budget process and for budget decisions is truthfully the next great challenge for governments (Ingraham, Joyce, & Donahue, 2003; Laegreid, Roness, & Rubecksen, 2008).
Information utilization and budget decisions
There are four examples for the positive associations between information utilization and budget decisions. First, at the program level and for the small and medium size programs, thinking more logically and strategically about how resources can be used and about what results might be accomplished, the department managers can influence the policymakers to restructure activity planning and reallocate resources (Gilmour & Lewis, 2006a; Ho, 2011; Joyce, 2008; Melkers & Willoughby, 2005). Second, when no clear political signals emerge, budget actors make decisions based on technical and economic rationalities, falling back on analytical information (Goodman & Clynch, 2004; Thurmaier & Willoughby, 2001). Third, the time and efforts public mangers invest in shaping information interpretation as well as their communication skills can persuade the policymakers to make changes in budget allocation (Moynihan, 2015). Fourth, organizations that develop performance reporting practices as a part of budget transparency (i.e., the use of information and communication technology (ICT)-based system for reporting performance and the reports of how current outputs and outcomes are to be
4


used for the future resource allocation, to be rewarded for good results, and to be sanctioned for poor results) trigger the policymakers to make actual budget changes accordingly (Laegreid et al., 2008).
The studies attempting to explore the linkage between information and budget decisions are not only limited but also offer inconclusive results (Ho, 2011). Several studies find that information such as public services’ outputs and outcomes exists in a budget decision process (Goodman & Clynch, 2004; Melitski & Manoharan, 2014), encourage learning and deliberation among different departments (Moynihan, 2008a, 2008b), but is not taken into serious consideration in terms of budget allocations. However, other studies confirm the linkage between the use of information and the government’s spending levels (Ho, 201 l;Moynihan, 2015; Zaltsman, 2009).
Performance information
The related studies are based heavily on the use of performance information. Outputs and outcomes of public service programs are at the center stage of the linkage between the information use and budget allocations (Laegreid et al., 2008; Melkers & Willoughby, 2005; Moynihan, 2008a, 2015; Pollitt, 2006). How other types of relevant information such as local knowledge (Petts, 1997; Johnson, 2009), political cues (Goodman & Clynch, 2004; Thurmaier & Willoughby, 2001), and even fiscal reports and guidance (Alijarde, 1997; Lee, 1991, 1992) that are used (alone or together with performance information) have not been adequately explored. Developed countries and administrative capacity
In addition, the current literature focuses mainly on the developed countries’ contexts— especially the U.S. (Ho, 2011; Lee, 1991; Moynihan, 2008a, 2015; Thurmaier & Willoughby, 2001) and the European countries (Laegreid et al., 2008; Bouckaert & Halligan, 2008; de Kool,
5


2008) despite the fact that large differences of how information is used exist between countries in patterns of its use (Pollitt, 2005, 2006). Performance information, for instance, has been used prevalently among large cities and countries which undoubtedly have more administrative capacity so that they can successfully handle such tasks as strategic planning, performance analysis, cost analysis, program budget analysis, and communication with the publics (Ho, 2011; Van Dooren, 2008). Nonetheless, the availability of information does not guarantee its use and in fact administrative capacity of policymakers is a key to the use of information as information is always selected, perceived, and presented by people (Moynihan, 2008a; Stone, 1997).
The administrative capacity, both qualification and attitudes, are key factors to determine information utilization. It cognitively shapes several activities during a budget process such as to find data sources, to conduct field visits, to relate data gathered to budget problems, to read political cues that condition fiscal alternatives’ means and ends, to apply information technology to integrating fiscal information into budgetary deliberation, and to develop and testify budgetary recommendation at executive and legislative hearings (Coe, 2003; Goodman & Clynch, 2004; Thurmaier, 1992; Willoughby & Finn, 1994). Nonetheless, even though many studies concentrate on the relationship between administrative capacity and whether and how the information is used, these studies do not further link their results to budget allocations (Coe, 2003; Lee, 1991, 1992).
Thailand, the least-likely case
For the case of developing countries, stagnant economics and inefficient central bureaucracies have led them to public-sector reforms. Among these reforms has been decentralization (Kettl, 2005; Schneider, 2003), which has occurred to various degrees across countries. Note that among the East Asian countries, Thailand has demonstrated slow progress
6


on fiscal decentralization. For instance, its subnational expenditure as a share of total public spending is approximately 10 to the average of 37 percent, the lowest of East Asian countries (White & Smoke, 2005). This limited local expenditure results in the lack of budget officials’ administrative capacity (Kittipanyasiri, 2015; Krueathep, 2014; Nagai, Funatsu, & Kagoya,
2008) as well as the lack of related information for decision making (Wongpredee & Sudhipongpracha, 2014). Hence, Thailand’s local governments can serve as the least-likely case (George & Bennet, 2004; Gerring, 2007) to unfold the linkage between administrative capacity, information utilization, and budget decision making.
Research Questions
It can be said that the current literature has put forward the comprehensive rationality of using information in a budget process, suggesting the existence of the linkage between the information utilization and budget decisions. However, there are still certain areas that have been inadequately explored. Those areas include the utilization of information during a budget process, the inclusion of administrative capacity as the key determinant of information utilization, the impact of information other than performance information on budget decisions, and the relationship between information utilization and budget decisions in order to confirm and possibly strengthen the argument that a budget process is of comprehensive rationality through its information utilization. Consequently, this study asks: How do policymakers with different administrative capacities utilize information during a budget process in order to have an impact on budget decisions? More specifically, it first asks: How do policymakers with different administrative capacities utilize relevant and reliable information during a budget process?
Then it asks: How does the information utilization influence budget decisions?
7


Methodology
To understand this complex and multifaceted budgetary phenomenon, this study conducts mixed methods research based on quantitative and qualitative data from a survey and interviews respectively (Creswell, 2013). The data will be analyzed based on the theories of information economics—information asymmetry (Akerlof, 1970), the principal-agent theory (Eisenhardt, 1989; Moe, 1984), and the budget rationality model (Thurmaier, 2001; Thurmaier &
Willoughby, 2001) against their competing theories—structural isomorphism (Meyer & Rowan, 1977) and incremental politics (Davis et al., 1966; Wildavsky, 1964).
Contributions to Theory and Practice
Theoretically, since this study attempts to (dis)confirm the relationships among the concepts of administrative capacity, the use of information, and budget decision making in Thailand, it serve as a theory testing case study which assesses the validity of competing theories (George & Bennet, 2004). Particularly, given the context of a developing country, the results from this study can increase the external validity or generalizability of the aforementioned theories to settings other than the developed countries (Onwuegbuzie & Leech, 2006). Additionally, information other than performance measures are empirically examined so that this study can increase the construct validity of all types of related information to the budget decision making (Van de Ven, 2007).
For policy implications, the results of this study can (dis)confirm the extent to which the current direction of the Thai budget reform which focuses mostly on whether the competency development and the provision of information significantly impact the information utilization during a budget process and the budgetary decision making. For example, the issues regarding which group of policymakers (i.e., specialized in education or public works) should be targeted


for the more comprehensive use of performance information and why some information are gathered but not being utilized. This research intends to inform policymakers of budget reforms which are of present-day problems (Gerring, 2012).
After this introduction, Chapter 2 reviews the theories and the current studies pertaining to information utilization and budget decisions. Based on these theories and studies, it offers a theoretical framework and sets up the hypotheses responding to the research questions. Next, Chapter 3 provides the reasons why this study focuses on Thailand’s local governments. It also describes the country’s progress on fiscal decentralization and how the local budget actors prepare and approve their budgets. Then, Chapter 4 elaborates the research design and methodology. Particularly, this study applies the mixed methods design which emphasizes the quantitative component and complements the quantitative results with the qualitative data. Later in Chapter 5, the quantitative and qualitative data and analyses are provided in order to answer the research question of how policymakers with different administrative capacities utilize relevant and reliable information during a budget process. Note that these analyses concentrate on all survey participants as well as on their groups according to the policy areas they serve (i.e., educational and public works services). Meanwhile, Chapter 6 provides the answers to the research question of how the information utilization influence budget decisions. Similar to Chapter 5, Chapter 6 also offers the quantitative and qualitative analyses of the overall survey participants and those based on policy areas (i.e., educational and public works services). Later in Chapter 7, the study discusses the results from Chapter 5 underlining information utilization. It shows how administrative capacity, information types, and budget orientations have an impact on information utilization during a local budgeting process. Next, Chapter 8 discusses the results from Chapter 6, explaining the influence of the information use can balance the budgetary
9


incrementalism in municipal budgets. Then in the final chapter, Chapter 9 summarizes the study. It provides the overall findings and the policy implications. The study’s limitations which lead to future research are also provided.
10


CHAPTER II
LITERATURE REVIEW
As this study focuses on how rationality plays a role through information utilization and administrative capacity, it comprehensively reviews the theories of information economics (information asymmetry and principal agent theory) and the administrative characteristics especially in terms of budget rationality model which concerns information, informational behaviors, and decision making.
Since the budget reforms gear a budget process toward managerial improvement, program effectiveness, and progressive roles of budget participants (Caiden, 2010; Schick,
1966), this study is based on the rational choice theory and consequently assumes the instrumentalist’s view—budget actors search and use information in order to make budget decisions that benefit their organizations the most (Hindmoor, 2006; Weiss, 1989). Therefore, this chapter delineates the relevant theories by explaining how individuals approach information, utilize it during a budget process, and eventually let it impact their budget decisions. Table 2.1 below briefly explains how these theories relate to information, information behaviors, and budgetary behaviors.
ll


Table 2.1 Relevant Theories to Understand the Interrelationship between Capacity,
Information, and Budget Decisions
Theory Approach to Information Informational Behaviors Budgetary Behaviors
Theories of information economics Information as a key to solve information problems (adverse selection and moral hazard) How public managers and organization leaders gather information and utilize it during the budget process How relevant and reliable information determines its utilization and leads to budget decisions
Budget rationality model Information types and information use as the representation of budget norms and spending preferences How public managers choose and use only particular types of information during the budget process How budget norms and spending preferences influence the information utilization and determine budget decisions
Theories of Information Economics
In most cases, there is always limited information for any goods or services so that additional credible information can resolve the problems from such limitation. In market transactions, there are two parties—buyers and sellers. According to the assumption of rationality (both comprehensive and limited or bounded), both parties attempt to gain and use related information to make their best decisions against a set of alternatives in order to receive the highest level of utility or satisfaction (Hindmoor, 2006; Lindblom, 1959; Simon, 1957). Unfortunately, information is not distributed evenly between buyers and sellers. This results in the situation where sellers can take advantage of possessing complete information and disseminate deceiving information to their buyers. This informational gap leads to two relevant theories—information asymmetry and principal-agent theory.
Information Asymmetry
Generally, there exist two parties in a transaction who have different levels of information and different values regarding such transaction (Moe, 1984). In markets it is possible for buyers to canvas various sellers by searching and evaluating market statistics (Stigler, 1961).
12


However, in reality, buyers can only rely on the market statistic and sellers remain possessing the informational advantages—they know the true values of their goods and services and can disseminate deceptive information to the buyers (Spence, 1973). Here exists the information asymmetry. Even a small amount of this informational gap has a significant effect on a competitive market, resulting in the reduction in the average quality of goods and also in the size of such market (Akerlof, 1970a; Rothschild & Stiglitz, 1976). The following paragraphs describe this informational gap in terms of the information problems it causes as well as possible resolving mechanisms to relieve such problems.
The first information problem relates to the public-good characteristic of information itself. Even though sellers can sell information to buyers, buyers (or the purchasers of the information) can share or resell such piece of information to others without diminishing its usefulness to themselves (Leland & Pyle, 1977). On this ground, sellers are not willing to sell their information, resulting in the under-provision of it. However, there are economies of scale in the process of gathering and selling the economic information. Due to the economies of scale, third parties can profit from producing and disseminating information. Specifically, a financial intermediary, generally a private firm, collects the data regarding the risks of poor-quality goods and services, sorts the risk level, and sells such specialized and confidential information to buyers (Campbell & Kracaw, 1980). A financial intermediary can carry out these activities in a specialized market that is small enough to control the transferability of information and, meanwhile, in an amount of scale that is large enough to cover fixed costs (Leland & Pyle, 1977; Spence, 1973).
The second problem is called adverse selection which involves the situation where sellers may claim that their product or service is of high quality while buyers cannot completely verify
13


such quality either at the time of purchasing or while sellers are delivering such product or service (Eisenhardt, 1989). This problem intensifies when many sellers have incentive to exaggerate quality of their merchandise due to the fact that the returns for good quality accrue to the whole group whose statistic is affected rather than to the individual sellers (Akerlof, 1970; Hirshleifer, 1973). With limited and unauthentic information, buyers are likely to choose poor quality merchandise at the higher price than its true quality warrants. Since buyers cannot know the quality of goods and services due to information problems, they rely on signaling—to gather the information and check other characteristics of the good or services in order to predict the quality prior to market transactions (Riley, 1975; Spence, 1973). Signaling includes searching for sellers’ reputations and qualifications. A sellers’ reputation or credibility is considered a valuable good which signals their expected performance in terms of the quality of goods and services they sell (A1 chian & Demsetz, 1972). When signals exist, sellers, especially ones with a sufficient stake in the market, have no incentive to misrepresent their information since their initial endowment acts as a barrier to exiting the market (Campbell & Kracaw, 1980; Heinkel, 1982). Additionally, their qualifications (i.e., profiles and capability) can, to some extent, signal their expected performance. For example, buyers can use the sellers’ investment in capacity building as a proxy for product quality (Spence, 1973).
The third information problem is moral hazard, which arises from unobservability of the sellers’ actions after buyers purchase goods or services from sellers (Wood, 2010). Since sellers’ actions cannot be observed and hence contracted upon, they have incentive to behave dysfunctionally, not doing their best for the interest of the buyers (Holmstrom, 1979;
Williamson, 1971). A subset of the moral hazard problem is the principal-agent problem. Particularly in the context of public organizations, it is called the principal-agent problems.
14


These problems exist in a relationship between two utility-maximizing parties when one (designated as an agent) acts for, on behalf of, or as a representative of the other (designated as a principal) in particular decision problems where goals or values of both parties conflict and when it is costly for a principal to closely observe his agent’s behaviors (Ross, 1973). A natural remedy to this problem is to invest in monitoring and evaluation of action and use this information in the contract design (Holmstrom, 1979).
Regarding resolving mechanisms to these three problems, additional relevant and reliable information, and use of it, can alleviate the information problems above. Specifically, information production or the systems that provide additional information to the market can alleviate information problems. This generation and provision of economic information, which renders the true value as well as the riskiness of the assets to be known to the market, can minimize or resolve the problems from aforementioned information asymmetries (Blazenko, 1987; Heinkel, 1982; Riley, 1975; Spence, 1973).
Principal-Agent Theory
Principal-agent theory explains and attempts to solve the problems rooted in an agency relationship—a relationship between two utility-maximizing parties when one (designated as an agent) acts for, on behalf of, or as a representative of the other (designated as a principal) in particular decision problems where goals or values of both parties conflict (Ross, 1973). Additionally, it is so difficult or expensive for the principal to verify the agent’s actions (Eisenhardt, 1989; Moe, 1984). The principal-agent theory has seven main components as follows.
First, the key parties for the principal-agent theory in the public organizations or bureaucracies are bureaucratic superiors and their subordinates, as principals and agents
15


respectively. This relationship is “therefore structured by a principal-agent chain, for instance, from citizens to politician to bureaucratic superior to bureaucratic subordinate and on down the hierarchy of government to the lowest-level bureaucrats who actually deliver services directly to citizens” (Moe, 1984, p. 775-776).
Second, at the individual level, there exists the preference asymmetry. According to the rational choice theory, both parties are rational, utility-maximizing individuals but with asymmetry in preferences—they have different utility functions or, in other words, different interests, values, or goals (Hindmoor, 2006; Miller, 2005; Moe, 1984). Policymakers concentrate on politics so that they would like to satisfy their citizens or constituencies in order to solve public problems, make public policies, and be reelected or to stay on in their political positions (Kingdon, 2003; Moe, 1984; Wood, 2010). Meanwhile, bureaucrats are budget maximizers— they strive to make budget (organization’s resources including salary, perquisites, power, prestige, patronage, public reputation, and agency output) as large as possible as a way to seek autonomy as well as to shirk work (Frederickson, Smith, Larimer, & Licari, 2012; Niskanen, 1971).
Third, at the organizational level, there is an information asymmetry—the situation where the relevant and reliable information is distributed asymmetrically throughout a bureaucracy. With technical knowledge and specialization, bureaucrats have and enjoy close to monopolistic informational power over their executives because they are the only party that possess information on both production costs and production process (Frederickson et al., 2012; Niskanen, 1971). As self-interested budget maximizers, bureaucrats are motivated to satisfy their policymakers by handpicking and highlighting information that reflects favorably upon themselves and repressing or even suppressing information that does not (Downs, 1967; Tullock,
16


1965). This monopolistic informational power leaves the executives informationally disadvantaged or lacking accurate and/or complete information. Policymakers can acquire only limited or even distorted information regarding production costs, production process, and the bureaucrats’ actions so that they usually form skewed expectations and decisions about the bureaucrats’ performance and capabilities (Frederickson et al., 2012; Moe, 1984).
Fourth, there are risk and uncertainty as a result of both preference and information asymmetries. Risk means the measurable, unknown uncertainty where one can use additional information as a mechanism to counter such risk while uncertainty connotes the absolute unknown mechanism and therefore the outcomes cannot be predicted (Ellsberg, 1961; Georgescu-Rogen, 1954). Hence, under the circumstances of risk and uncertainty, utility theory cannot be fully applied (Ellsberg, 1961; Friedman & Savage, 1948). Risk and uncertainty here imply the high possibility that policymakers fall prey to bureaucrats’ opportunism and suffer from the problems of adverse selection and moral hazard due to the agents’ shrouding of information—the concealment, distortion, and falsification of information (Akerlof, 1970; Breton, 2009; Spence, 1973). Additionally, given that the production of public goods and services are of technical knowledge and specialization, even though it is theoretically feasible for policymakers to monitor the agent’s actions, it is prohibitively expensive or economically unviable for executives to gather complete information in order to conduct a thorough monitoring (Moe, 1984; Ross, 1973). On this ground, the executives are likely to allocate budget for inferior public goods and services offered at the higher price than their true quality (Niskanen, 1971).
Fifth, policymakers as risk-averse individuals seek resolving mechanisms to minimize or cope with information problems. Though informational advantage is located on the bureaucrats’
17


side, authority is located on the policymakers’ side (Miller, 2005). Policymakers can seek relevant and reliable information through two main mechanisms—information intermediaries and performance contracts and also can use their professional judgment when making decisions.
1. Information intermediary: An information intermediary refers to organizational units or members who gather and champion the use of knowledge (Corburn, 2007), simplify scientific information, and bridge the gap between the science and technology and the end users (i.e., executives, legislators, and citizens) (Faibisoff & Ely, 1974). Even though the question of which information a public organization should gather and utilize in decision making cannot be answered, such organizations should at least possess a basic set of information in order to show its constituents that they are fiscally transparent and accountable and that they promote fiscal sustainability, good governance, and overall fiscal rectitude (Alt, Lassen, & Skilling, 2002b; Kopits & Craig, 1998). This basic set of fiscal information includes such information as economic projection, statement of future plan, outside opinion and evaluation, input, output, service quality, efficiency, and outcomes (Kopits & Craig, 1998a; Seifert, Carlitz, & Mondo, 2013).
2. Performance contract design: Since it is prohibitively expensive for policymakers to monitor their bureaucrats’ actions, they invest in contract design which is based mainly on performance information (Holmstrom, 1979; Moe, 1984; Wood, 2010). Performance information reflects actionable and quantifiable production process and products or services, encompassing outputs (public goods and services produced or delivered), efficiency (a result from a cost-benefit analysis), effectiveness (the evaluation of missed and achieved targets in terms of both quantity and standards),
18


and outcome (economic and social implications of such public goods and services) (Ammons & Roenigk, 2015; Talbot, 2007). Given the fact that this performance information is easily observed and evaluated, the shirking and budget-maximizing behaviors among bureaucrats can be controlled reasonably well without substantial monitoring costs (Moe, 1984).
3. Professional judgments: Despite their attempts to gather all the accurate and relevant information through the previous mechanisms above, the perfect or complete information is not guaranteed for principals to solve or even avoid the problems resulting from the preference and information asymmetries (Ellsberg, 1961). Therefore, policymakers significantly depend on their professional judgments, which are indispensable for taking actions to effectively formulate and implement public policies (Bertelli, 2007; Smith, 2003). Indeed, governmental decisions are products of a complex collective judgment (Appleby, 1949). Professional judgments represent technical competence, knowledge, and specialization and are significantly accumulated through training and experiences (Karl, 1976). Under the context of performance management and budgeting, public administrators have flexibility combining expertise with professional judgment in order to improve their organizations’ technical efficiency and effectiveness (Moynihan, 2008b).
Sixth, according to rational choice theory, policymakers would utilize all types of relevant information instrumentally they largely use it for budget analysis and eventually for making budget decisions (Knorr, 1977; Niskanen, 1971; Weiss, 1977). In other words, relevant and reliable information is possibly taken into account in various stages of information
19


utilization during a budget process—reception, cognition, discussion, reference, scrutiny, effort, and influence (Knott & Wildavsky, 1980; Rejean Landry, Lamari, & Amara, 2003).
Seventh, eventually, budget decisions in form of budget allocations, which constitute payments to bureaucrats for their delivery of public goods and services, can act as incentives to spur agents more toward the policymakers’ policy directives. For example, an increase in budget allocation means the promotion of bureaucrats’ high-performing activities and a decline in budgets can be cast as sanctions to curtail certain actions of the agents (Eisenhardt, 1989; Wood, 2010).
To sum up, information asymmetry and principal-agent theory suggest that there is an informational gap between policymakers and bureaucrats as principals and agents, respectively. Though informational advantage is located on the bureaucrats’ side, authority is located on the policymakers’ side. Policymakers can acquire relevant and reliable information by creating an information intermediary. It can create its own information or take advantage of the characteristics of information as a public good, gathering available information offered by other organizations. This information can keep policymakers so informed of the quality and cost of service delivery that they can make budget decisions appropriately. Therefore, the problem of adverse selection can be diminished. In addition, policymakers also require bureaucrats to act on their behalf by designing performance contracts and performance measures in order to make sure that the shirking and budget-maximizing behaviors among bureaucrats can be controlled without substantial monitoring costs. Accordingly, the moral hazard can be reduced. Policymakers, therefore, utilize information from both information intermediaries and performance measures to make budget decisions, specifically budget allocation and budget effect. Figure 2.1 below
20


graphically elaborates this relationship between the theories of information economics and budget decision making.
Sources of Information
Information intermediary and performance contracts (A)
Fiscal Information
Relevant and Reliable Information and its characteristics: (X1)
Administrative Capacity
Personal judgment: experience and intuition
(X2)
â–¼
Information Utilization
The use of information throughout a budget process CY1)
Budget Decisions
Budget allocation (Y2)
Of theoretical interest Background factors General Feature
X = Causal factor A = Antecedent -* = Causal
Y = Outcome - = Possibly causal
Figure 2.1 The causal graph of theories of information economics under the context of information utilization and budget decisions
Empirical Studies on Information Asymmetry, Information Use, and Decision-Makings
As suggested in Figure 2.1, there are many causal relationships in the principal-agent theory. In the context of information use and budget decision making, the related empirical studies add several perspectives and nuances into this theory. The following parts explain such empirical studies through each causal step shown in Figure 2.1. Some empirical conclusions as well as research gaps are also provided.
Sources of Information and Fiscal Information
Contracts and their performance information is automatically available if a public organization adopts any form of performance measures into decision making process (Kong,
2005). The performance information encompasses input, output, efficiency, performance targets, benchmarking data, and service quality that prevails in public organizations such as the U.S.’s state and local governments (Bourdeaux, Chikoto, Bourdeaux, & Chi, 2008; Julnes & Holzer, 2001), Chinese local governments (Zhang, Van de Walle, & Zhuo, 2016), Norwegian central
21


government’s agencies (Askim, Johnsen, & Christophersen, 2008; Laegreid et al., 2008), and Danish municipalities (Nielsen & Baekgaard, 2015).
Meanwhile, other types of information are gathered and transmitted to policymakers by an information intermediary—individuals (policymakers themselves also included) or a unit responsible for retrieving all the information available concerning the policy areas of concern to their policymakers (Rich, 1975). For instance, in Australian coastal cities (Shaw, Danese, & Stocker, 2013) and some communities in Pacific islands (Cash, Borck, & Patt, 2006) public organizations and outside information producers create and transfer information from local people, scientists, and stakeholders to policymakers while ensuring that local knowledge, which also is trusted and legitimate, makes it into a policy process. Boundary organizations’ informational role sometimes has an impact on decision making. The case in point is that the U.S. Environmental Protection Agency has had a great policy influence on stakeholders through its roles of information dissemination despite being definitely liberal during the period in which Republicans were dominant in the Congress (Lee, 2012).
These boundary organizations or information disseminators produce information that is relevant to the policy areas of concern. Even though not specifying the information elements like the case of performance information, they offer two main types of information—expert knowledge and local knowledge. Expert knowledge is the systematically gathered information presented in research or evaluation reports which highlights analysis that meets standards of coherence and honesty and uses generally accepted professional, scientific, and technical methods (Lundin & Oberg, 2014; Weible, 2008). Meanwhile, local knowledge is the information about a local context or setting, including empirical knowledge of specific characteristics, circumstances, events, and relationships, as well as the normative understandings of their
22


meanings; is usually held by members of a community that can be both geographically located and contextual to specific identity groups; and owes its status to causal empiricism, thoughtful reflection, and common sense (Corburn, 2003b; Fischer, 2000). In the case of environmental policy, the Ontario Ministry of Natural Resources requires local knowledge to be included in forestry research for formulating its policies (Klenk & Hickey, 2011). At the same time, local knowledge has been used in terms of the environmental risk in the water use planning in Canada (Failing, Gregory, & Harstone, 2007); how a low-income, Latino immigrant neighborhood in New York City organized its knowledge to participate in and significantly change a U.S. Environmental Protection Agency exposure assessment (Corburn, 2007); and in local plans for the communities facing the most serious environmental and health risk (Corburn, 2003).
To sum up, the literature that mostly explores information use in developed countries specifies both the information source (i.e., performance contracts) and the performance information piece (i.e., input, output, efficiency, performance targets, and benchmarking data, and service quality). It also identifies information sources (i.e., information intermediary) for other relevant information, namely local knowledge. Accordingly, this implies that there are two areas the literature has not adequately explored—the information production in the context of developing countries and how local knowledge is used for budget decision making.
Fiscal Information and Information Utilization
Generally, policymakers can use information in two different fashions (Weiss, 1977). They can symbolically employ information only to legitimate political or bureaucratic activities decided upon before the research was undertaken—activities necessary to maintain the postindustrial welfare state—and to manipulate the consent of the citizen. However, sometimes information can permeate decision making in subtle ways and bring about the changes in
23


political or bureaucratic thinking. As the budget reforms aims to promote the latter type of information use, there are several related studies exploring how to promote the latter type of information utilization.
The relationship between information and its utilization for decision making rests much upon the availability and nature of the information itself (Pandey, 2015). In other words, information producers may offer a variety of information but not all of it is utilized (Caplan, 1979; Dunn, 1980; Ottoson, 2009) and some of it is utilized more frequently than others (Wolman & Page, 2002). According to the principal-agent theory, relevant and reliable information is used for decision making (Akerlof, 1970; Moe, 1984). Relevance of information elements is measured in terms of usefulness, significance, and frequency of use (Melkers & Willoughby, 2005; Wolman & Page, 2002) and reliability depends on sources of information (Boggs, 1992; Wolman & Page, 2002). The utilization can be divided into two different levels, understanding and deliberation. The information in terms of understanding involves the situation where information has been considered but not used for decision making (Knorr, 1977; Rich, 1975). Meanwhile, information utilization in terms of deliberation refers to sequential stages of utilization encompassing such cognitive and management activities as scrutiny, effort, and influence (Knott & Wildavsky, 1980).
There are also some examples of the relationship between the characteristics of information and the utilization. Regarding reliability, the findings from related empirical studies state that policymakers comprehensively utilize information that is generated by experts or researchers. For instance, the trust in information sources (i.e., perceived ability, integrity, and benevolence of boundary organizations or information disseminators) leads to the active use of performance information among the Chinese central government agencies (Zhang et al., 2016)
24


and the involvement of policymakers (specifically public managers in the supervisor level in the U.S. federal government agencies) in terms of preparing for or gathering any Performance Assessment Rating Tool (PART) assessment has an impact on the extent to which such information is utilized (Moynihan, Lavertu, & Kamensky, 2012).
To conclude, the empirical studies suggest that information—fiscal, performance, and other—that is reliable and relevant is likely to be utilized comprehensively and actively by policymakers. Note that even though policymakers receive, reference, and discuss such information with each other, they do not always deliberate it, potentially limiting their effect to understand the information by not asking questions, scrutinizing the information, and preparing to change their opinions or letting the information influence their decisions. In this vein, the literature has not adequately explored how performance information together with other information types are utilized especially under the budgeting’s context.
Information Utilization and Budget Decisions
Generally, decisionmakers use information in two fashions. First, according to the concept of
Based on performance management doctrine (Jakobsen & Mortensen, 2016; Moynihan, 2008b), several studies assume that performance information is fully utilized and it eventually has an influence on budget decisions. The literature studies these decisions through both actual and perceptional budget allocation. For example of actual budget allocation, PART scores play a role in the determination of program budgets suggested by the U.S. federal government’s Office of Management and Budget (Gilmour & Lewis, 2006a); government agencies in Chile that comply more with performance targets have seen their aggregate budgets decrease relative to those with weaker compliance records (Zaltsman, 2009); and the number of outcome-related
25


performance measures used by city officials accounts for the change in annual budget allocation of major departments and operations of municipalities and counties in Indianapolis (Ho, 2011). For example of perceptional budget allocation, if administrators and budget analysts in the U.S. local government use performance measures at various stages of the budget process, they report the impact of such utilization on cost saving, appropriation levels, and reduction or elimination of ineffective or duplicative services or programs (Melkers & Willoughby, 2005). Furthermore, Norwegian public managers who thoroughly utilize performance information (i.e., formulating goals and performance indicators and making evaluation reports) report that such information utilization has impacts on resource allocation, rewards for good results, and sanctions for poor results (Laegreid et al., 2008). Therefore, it can be said that the comprehensive use of performance information can have an influence on the budget decisions. However, information other than performance information or the combination of performance information and other information still have not been adequately explored in the context of budget decisions. Professional Judgment and Budget Decisions
Even though, theoretically speaking, governmental decisions are products of a complex judgment (Appleby, 1949), the literature exploring relationships between professional judgment and budget decisions is limited. The related study applies administrative capacity in terms of professional training and educational background as a proxy of professional judgment and finds that professional judgment of the U.S. government budget officers are not associated with their choices of resource allocation (Reck, 2000).
Budget Rationality Model
The budget rationality model theorizes that decision makers use the information necessary to advance their agendas. Briefly, this budget rationality model states that with
26


different administrative capacity (i.e., spending attitudes) policymakers possess differing orientations toward budgetary as well as spending preferences and thus only certain types of information are given priority in order to guide their budget decision making (Moynihan, 2006; Nielsen & Baekgaard, 2015). The current literature develops this model though grounded theory, it theoretically progresses as its empirical findings accumulate (Goodman & Clynch, 2004; McCue, 1999). Based on those findings, the Figure 2.2 below graphically portrays these relationships.
Administrative Capacity: Tenure, bureaucratic ranks, and academic discipline (X1) , Fiscal Information: Information Utilization:
Budget Rationality: Types of information Stages or levels of
Budget orientations * and characteristics information utilization
CX2) CM) {Y1)
Of theoretical interest General Feature
X = Causal factor ---â–º = Causal relationship
Y = Outcome M = Mechanism
Figure 2.2 Budget Rationality and Information Utilization Administrative Capacity and Information Utilization
In general, budget actors’ spending preferences varies according to their seniorities (Hildreth, Yeager, & Miller, 2011; Thurmaier, 2001). In particular, seasoned analysts with more years in service should be able to anticipate political cues, thereby weighing political factors more heavily than non-seasoned analysts (Goodman & Clynch, 2004). For example, the tenure of budget analysts in the U.S. state governments impacts their spending preferences because they tend to respond to political cues quicker than less senior decision makers who tend to rely on economic cost-benefit decision criteria (Willoughby, 1991). Especially the senior analysts in Kansas State Budget Office acknowledge a larger role for political rationality in their budget recommendations (Thurmaier, 2001). However, a more current study in the thirteen U.S. western states does not find the statistically significant association between hierarchical position or the
27


number of years each budget analyst has held in her/his current position and the propensity to weigh political factors more heavily than junior analysts (Goodman & Clynch, 2004).
In addition, administrative capacity is actually associated with the type of fiscal information one finds relevant or important and usually utilizes. For example, the U.S. state governments’ budget analysts with business administration education are more likely to exercise the information of financial accounting control than those with social science education (Lee, 1991). Meanwhile, the budget analysts with social science education are more likely to use performance information than those with business administration education (Lee, 1991).
Overall, the administrative capacity of policymakers obviously plays a role in the use of information during a budget process. Administrative capacity is an umbrella concept which covers such qualities and conditions as education, knowledge, skills, training, experiences, orientations, attitudes, and motivation (Christensen & Gazley, 2008; Sun & Gargan, 1993). How administrative capacity specifically influences the information utilization depends largely on which dimension one chooses to study. Consequently, in general, it can be said that if an individual possesses higher administrative capacity on one particular dimension, he or she is more likely to utilize information directly related to such dimension during the budget process relative to other individuals. Nonetheless, the current literature focuses mostly on the administrative capacity in terms of qualification (i.e., experiences and educational background), leaving other dimensions (i.e., skills and attitudes) unexplored. Additionally, these studies only focus on the types of information policymakers use and not on how policymaker use it during the budget process.
28


Budget Orientations and Information Utilization
There always appears a struggle among three traditional distinct budget norms or budget decision-making patterns shared among budget actors—control (fiscal accountability accuracy), management (efficiency and effectiveness), and policy (strategic planning and advocacy for and responsiveness to policy outcome) (David, 1998; Smith, 2003; Willoughby, 2002). Relevant studies that concentrate on the budget roles played by the U.S. state budget analysts find that there are five budgeting orientations, ranging from the control to the management and to the policy orientations as follows (Smith, 2003; Thurmaier & Willoughby, 2001; Willoughby, 1993; Willoughby, 2002).
(1) Conduit: Policymakers act as conduit in the budget process—only transmitting information between executives and agency and back again and usually presenting a wide range of information to reflect all the spending preferences.
(2) Accounting: Policymakers act as fiscal guardians who serve traditional guardian functions and are oriented toward control (number crunching), accounting principles, integrity, fiscal responsibility, and accountability. They think that the agency workload and agency efficiency are the most important work when making budget decisions.
(3) Techno-rational: Budgeteers act as both strategic planners and budgeting professionals. Not only do they enlighten agencies of executive priorities and funding strategies, they also promote such budgeting techniques as Program Planning Budgeting System (PPBS), Management by Objectives, Zero-Based Budgeting while alert other budget actors of legal and other requirements that must be met.
29


(4) Policymaking: Budget actors are policy analysts who take into account a comprehensive view of the overall budget and policy outcomes or how the budgetary decisions impact the government’s overall fiscal condition and then push forward their policy initiatives based on such analysis. The budget actors in this group apply weight to both political and analytical forces.
(5) Advocating: Concentrating on policy responsiveness and current political and economic agendas, budgeteers act as policy advocates who seize the window of opportunities to push forward their policy initiatives into the budget process. The budget actors apply more weight to political factors such as executive agendas and public support. Budget actors’ spending preference are based on individual decision modes—how they attend to, encode, and process particular information (McCue, 1999).
From the budget orientations above, it implies that only particular information is carefully or purposefully considered and used during a budget process. As illustration, based on a basic set of fiscal information that a transparent government should make available and the concept of budget rationality, Table 2.2 shows how policymakers with different budget orientations utilize information during the budget process.
30


Table 2.2 Budget Orientations, Relevant and Important Information Considered, and Norms
Budget Orientation Possibly Relevant Information Information Type Budget Norms
Conduit Unspecified Unspecified Unspecified
Accounting Audit report and financial ratios Financial and accounting data (Analytical information) Control
Techno- rational Possibly all types of budget information All types (Analytical information) Control, management, and Policy
Policymaking Economic projections and policy results or outcomes All types (Analytical information) Control and Policy
Advocating Economic projections and policy platforms Social and political data (Political cues) Policy
Source: Adapted from Willoughby’s model of decision environment of the CBO Analysts (2002); Goodman and
Clynch’s information attributes (2004); and the examples of information for budget transparency from De Renzio & Masud (2011) and Kopits & Craig (1998).
Based on Table 2.2, policymakers with conduit orientation plausibly show the very limited use of information during the budget process and do not possess a particular spending preference. Meanwhile, ones with accounting orientation are assumed, on average, to utilize higher levels of information especially from the comprehensive use of the financial and accounting data in order to keep the budget allocation efficient and transparent. Policymakers with techno-rational orientation in general are likely to use information to a higher extent than other groups of policymakers. This is because they contemplate all types of information as the budget process proceeds in order to ensure the efficiency, effectiveness, and policy goals. For ones with a policymaking orientation, similar to the techno-rational policymakers, they generally use information more comprehensively than other groups. However, the policymaking budget actors use information in order to ensure efficiency and transparency for particular policy areas with which they are concerned. Finally, for the policymakers with advocating orientation, they are likely to use information more comprehensively than those with conduit orientation but lower than those with techno-rational and policymaking orientations. This is due to the fact that they
31


only concentrate on information regarding political cues in order to make sure that the budget allocation follows the policy platforms initiated by politicians or organization’s executives.
Nonetheless, the empirical studies which thoroughly review the relationship between budget orientations and the utilization of information are limited. A study on budgetary decision making by budget analysts in the thirteen U.S. western states confirms that not only one type of information (i.e., performance information), but a bundle of varied information is contemplated during the budget process according to analysts’ budget orientations (Goodman & Clynch, 2004). The study neither gives insight on the level of information use nor on the ways policymakers utilize the information. These issues also warrant the comprehensive study.
Findings from the Theoretical Review
From the theories of information economics (information asymmetry and principal agent theory) and budget rationality model, there are four main findings. First, sources of information and quality of the information itself play a crucial role in how policymakers utilize such information. Particularly the source of information determines the reliability of the information while the information itself determines its relevance. These two characteristics of information influence how the policy makers use information during the budget process. The current literature, which heavily focuses on performance information, assumes the high levels of its reliability and relevance without actually measuring them. Additionally, individuals indeed do utilize performance information together with other information. In other words, they use a combination or a bundle of information consisted of performance information, financial and accounting information, and political and social information. As a result, it is necessary for the related studies to investigate these two information characteristics of information bundle
32


policymakers utilize while initiating, analyzing, discussing, and defending their budget proposals.
Second, administrative capacity and budget orientations also have an impact on how policymakers perceive available information in terms of its relevance and reliability and utilize that relevant and reliable information during a budget process according to the instrumental use of information (Weiss, 1977). Policymakers with higher levels of administrative capacity are more likely to comprehensively utilize information for drafting their budget proposals. Additionally, policymakers with different budget orientations utilize different pieces of fiscal information for their budget decisions.
Third, administrative capacity and information utilization are possibly the determinants of budget decisions. The empirical studies confirm existence of the relationship between information utilization and budget decisions. However, the studies regarding whether the administrative capacity, in terms of skills and knowledge and how information is specifically used during the budget process to impact on budget allocations, are limited. Furthermore, the literature also finds the impact of information utilization on budget allocation in both perceptional and actual terms for major spending programs. Nonetheless, these studies do not consider both types of budget allocations simultaneously.
Fourth, except for China (Zhang et al., 2016) and Chile (Zaltsman, 2009), the public budgeting literature regarding the use of information and budget decisions does not pay attention to the situations in developing countries. Particularly, the recent studies are heavily based on the experiences of such countries as the U.S. and Canada. On this ground, the use of information and how it affects the budget allocation under the context of developing counties should be advanced.
33


Research Questions and Hypotheses
Considering the research areas the extant literature has inadequately explored, this study asks the following overarching question: How do policymakers with different administrative capacities utilize information during a budget process in order to have an impact on budget decisions? Specifically, there are two sub-questions supporting the main research question above. This section presents the sub-questions and offers hypotheses regarding each.
First, how do policymakers with different administrative capacities utilize relevant and reliable information during a budget process?
HI A: The more policymakers perceive information to be relevant and reliable, the more likely they are to utilize it during the budget process.
H1B: Policymakers with higher levels of administrative capacity are more likely to use relevant and reliable information during the budget process relative to those with lower levels of administrative capacity.
H1C: Policymakers with techno-rational and policymaking orientations are more likely to utilize information than those with conduit, accounting, and advocating orientations.
Second, how does the information utilization influence budget decisions?
H2A: The more policymakers use information during a budget process, the higher more such use impacts the budget decisions.
H2B: How policymakers combine different types of information during the budget process affects their budget decisions.
H2C: Policymakers also base their budget decisions on their administrative capacity (technical skills and knowledge) to frame their budget decisions.
34


Rival Hypotheses
There are some factors that confound the relationship between administrative capacities, information utilization, and budget decisions. Two theoretical explanations account for such confounding effects.
This first is structural isomorphism. For a certain distinguishable environmental configuration, there is one organizational structure (i.e., a set of values, norms, standard of procedures, management techniques, products, services, programs, or policies) that is widely accepted as the most appropriate way to do things so that other organizations attempt to resemble it (DiMaggio & Powell, 1983; Hannan & Freeman, 1977; Selznick, 1948). This resembling behavior, or structural isomorphism, possibly stems from normative and/or coercive pressure (Dimaggio & Powell, 1983; Meyer & Rowan, 1977). Normative pressure exists when formal education or the growth and elaboration of a professional network is considered the source of values, norms, and standards of procedures with which organizational leaders or managers comply. Meanwhile, coercive isomorphism stems from political influence exerted on organizations by other organization upon which they are dependent. However, due to this pressure instead of the necessity inside the organization, it is possible that despite the existence of formal structures, it does not have an effect on a behavioral change. This phenomenon is called formalism and seen as a dilemma of the administrative reformer in developing countries including Thailand (Riggs, 2006).
Under this study’s context, to appear as a highly performing organization, local governments are normatively required to adopt performance budgeting in which budgetary actors produce or gather and utilize performance information when making budgetary decision (Hou, Lunsford, Sides, & Jones, 2011; Kong, 2005; Willoughby & Melkers, 2000). However,
35


budgetary actors also do not tie performance information to incentive and disincentive systems in term of pay, status promotion, and budget cut (Ammons & Roenigk, 2015; Hou et al., 2011; Laegreid, Roness, & Rubecksen, 2008). For instance, state government agencies in Florida do not actively implement the performance-based budgeting (PBB) as a replacement of old budget systems (Andrews & Hill, 2003). Australian government agencies also use performance information in a symbolic fashion, not for making decisions on budget allocation but for elaborating plans and reports in order to project an image of good governance (Yetano, 2013). U.S. local governments extensively utilize input and output measures while outcome measures are less developed and utilized when making budgetary decisions (Julnes & Holzer, 2001; Melkers & Willoughby, 2005). Also, the main driver for the development of Thai public sector’s performance management system is the Office of Public Sector Development Commission, acting as the institutional entrepreneur or a source of normative pressure (Sutheewasinnon, Hoque, & Nyamori, 2016). On this ground, it can be said that due to the coercive and normative pressures on adopting performance management, even though policymakers find relevant and credible information produced by boundary organization relevant, they do not utilize it for deliberation during a budget process.
Second, even though information matters in the market realm, the theory of incrementalism holds that the addition of related information will not transform previous pattern of budget decision making (Wildavsky, 1964). Budget information is almost never actively reviewed as a whole in the sense of considering at once the value of all existing programs as compared to all possible alternatives so that the current year’s budget is based on last year’s budget, with special attention given to a narrow range of increase or decrease (Davis et al., 1966; Wildavsky, 1978).
36


For instance, though the U.S. local governments make effort to integrate performance measure into budget decisions, the application of performance information for budget decisions progresses incrementally (Melkers & Willoughby, 2005). Also, the factor indicating incrementalism such as age of spending program has an impact on the change in budget allocation (Gilmour & Lewis, 2006). In conclusion, the incremental view holds that, despite of the availability of relevant and credible information, budget decision makers do not comprehensively utilize it and thus it does not have an impact on budget decisions.
Theoretical Framework
Based on the findings from the theoretical review and the related empirical studies previously discussed in this chapter and the rival hypotheses, this study proposes the theoretical framework on administrative capacity, the use of information, and budget decision as elaborated in Figure 2.3.
Figure 2.3 Theoretical Framework of Administrative Capacity, the Use of Information, and
Budget Decisions
37


It is worth noting that even though the related theories discuss all of the relationships in Figure 2.3, there are still some causal relationships which are not empirically confirmed. These relationships are information sources and fiscal information (types and characteristics), administrative capacity and information utilization, budget rationality and information utilization, and administrative capacity and budget decisions. This study aims to test these theoretical associations through the local governments in Thailand with mixed methods research, presented in Chapter 3 and Chapter 4.
38


CHAPTER III
CASE JUSTIFICATION AND THAILAND’S LOCAL PUBLIC FINANCE AND
BUDGETING
This chapter first describes how the study selects the case, considering representativeness and causal leverage. As the previous chapter asserts that the public budgeting literature regarding the use of information and budget decisions does not pay attention to the situations in developing countries, this study proposes Thailand’s local government as the case studies. Accordingly, after the case justification, this chapter reviews the country’s progress on decentralization, the revenue and expenditure’s structure, and the local budgeting, including a process of budget decision making and roles of budget actors.
The Case Justification for Thailand’s Local Governments
Both large-N cross-case analysis and case study analysis aim to reproduce the causal features pertaining to a larger universe (representativeness) and provide variation along the dimensions of theoretical interest (causal leverage) (Gerring, 2007). To promote representativeness and causal leverage, this study selects Thailand’s local governments for case study analysis. The following details explicate how these cases can contribute in terms of representativeness and causal leverage to the study of how administrative capacity and the use of information impact budget allocation. Finally, based on these contributions, Thailand’s local governments can serve as the least likely case to (dis)confirm the related theories. Representativeness
The research on the relationship among capacity, the use of information, and budget decisions is mostly based on the experiences of developed countries such as the U.S. (Laegreid, Roness, & Rubecksen, 2008; Melkers & Willoughby, 2005; Moynihan, 2008a, 2015; Pollitt,
39


2006) and the European countries such as the Netherlands (Buylen & Christiaens, 2016) and Denmark (Nielsen & Baekgaard, 2015). Accordingly, in order to leverage the representativeness of such theoretical relationships, there should be research that incorporates new subtypes of the research settings (George & Bennett, 2004). Specifically, in this case, there should be studies explicating the relationship among capacity, the use of information, and budget allocation in the developing countries. However, in order to identify the potential country, it is required to consider the second case selection criteria—the causal leverage—which is discussed in details as follows.
Causal leverage
The causal linkages this study aims to confirm involve two relationships: administrative capacity and the use of information and the use of information and budget allocation. The following theoretical explanations clarify these two linkages.
Administrative capacity and the use of information. Overall, all else being equal, if public organizations have high administrative capacity (i.e., quality managers and good management systems) or the potential to obtain desired program results and policy outcomes, we can assume that such organizations are likely to be effective performers (Ingraham et al., 2003). Since the productivity or performance of bureaucrats is unknown to public organizations until after the critical decisions regarding job, wage, salary, or promotion offers have been made, administrative capacity (i.e., educational backgrounds and past experiences) serves as a signal or a key information element that is an observable attribute upon which public organizations make judgements about individual bureaucrats’ performance (Riley, 1975; Spence, 1973).
Regarding the relationship between administrative capacity and use of information, public organizations with sufficient administrative capacity specifically in terms of public
40


managers capable to handle informational tasks such as performance analysis, cost analysis, program budget analysis, and communication with public, are more likely to succeed with the quality and the effective use of information (Ho, 2011; Melkers & Willoughby, 2005). This is because the freedom, consistency, and speed with which these managers make information flow throughout an organization; the attention they give to specific information or knowledge; and their willingness to share such information or knowledge converge to facilitate (or thwart) the extent to which a government’s management systems take into account and utilize such information or knowledge (Ingraham et al., 2003). As a result, administrative capacity is a key determinant of and positively associated with the success in the use of information (Ho, 2011; Lee, 2012; Pollitt, 2006; Van Dooren, 2008).
The use of information and budget allocation. Information in general can resolve uncertainty and reduce ambiguity so that it enhances economic decision-makings (Faibisoff & Ely, 1974; Feldman, 1989). As the rational choice theory suggests, information is a key element for an organization’s logical decision-making because it is used extensively in formulating goals, searching for alternatives, and calculating consequences of each alternative (Downs, 1957; Lindblom, 1959). Particularly in a budgeting process, information such as public service outputs and outcomes encourages learning and deliberation among different departments in state and local governments (Goodman & Clynch, 2004; Melitski & Manoharan, 2014; Moynihan, 2008a, 2008b). Such information utilization can have budget effects in form of cost savings, appropriation levels, elimination of ineffective programs, and duplicative services especially among state governments (Melkers & Willoughby, 2001, 2005; Willoughby & Melkers, 2000).
Additionally, even from the bounded rationality perspective, information impacts decision making significantly. Different types or aspects of information have different levels of
41


accessibility and possess particularly different percepts (Kahneman, 2003). Such informational elements serve as an incentive for individuals to focus their attention and prolong their deliberation only on a particular issue and in a particular fashion so that they give consideration only to a particular set of alternatives with the exclusion of others during the decision process (Simon, 1985; Tversky & Kahneman, 1986). For instance, if provided with a variety of budget scenarios as well as different amounts and types of information, individuals make their decisions on public spending dependent on how performance information is presented to them in terms of goal ambiguity, expectancy disconfirmation, and advocacy (Moynihan, 2015).
Thailand as a least likely case
Though providing inconclusive results, the current studies in public policy, public management, and public budgeting have theoretically explicated the causal relationships between administrative capacity, the use of information, and budget allocation through the information asymmetry and the rational choice theories. So as to promote the causal leverage or causal inference of these relationships, a diagnostic case study approach is considered helpful given the fact that it is intended to assess whether the broadly understood assumptions based on the information asymmetry and rational choice theories hold (Gerring, 2017). In addition, since the research results on how the use of information has an impact on budget allocation and how administrative capacity as a background factor may affect the use of information are inclusive, influence or the probability of the aforementioned assumptions being true may be understood in a counterfactual fashion, specifically via the influential case (Gerring, 2007a, 2017). The influential case appears at first glance to nullify a theory, or at least to cast doubt on a theory but upon a closer inspection, it does not and it ends up confirming the validation and reidentifying the scope conditions of such theory (George & Bennett, 2004; Gerring, 2007).
42


Regarding the relationships among administrative capacity, the use of information, and budget allocation in the subnational governments, it can be said that if those organizations possess relatively low administrative capacity, have limited access to relevant information, and constrainedly utilize such limited information, they are unlikely to see the change in their budget allocations as a result of administrative capacity and the use of information.
Many studies in public budgeting measure the concept of administrative capacity via expenditure level of local authorities (Brace, 1991; Jeong, 2007). An expenditure level serves as a blunt but plausible and theoretically justifiable measure of overall organizational capacity (Brace, 1991). Even though the high level of expenditure may represent waste and inefficiency (Niskanen, 1971), it is also possible that high local expenditure can bring about capacity building activities, creating quality local bureaucrats. In addition, the higher local expenditure level is, the more resources local authorities possess to promote activities related to information utilization such as access, deliberation, and analysis of information (Terman & Feiock, 2015).
Low level of local public expenditure. Countries whose subnational government expenditure is considered relatively very low (between 8 percent of GDP and 20 percent) includes most African countries, most Latin America countries, several Euro-Asia countries, and some Asia-Pacific countries (United City and Local Government, 2016). However, since decentralization in East Asia began later than in other parts of the world, countries there are still redesigning the framework of and addressing the details of civil service management at a local level (Green, 2005; White & Smoke, 2005). Among East Asian countries, Thailand has the lowest subnational expenditures as a share of total public spending (approximately 10 relatively to the average of 37 percent) (White & Smoke, 2005). Additionally, notwithstanding a middle-income country with a bustling capital city, Thailand’s local governments possess low
43


administrative capacity—in general staff lack the technical skills and institutional resources to deliver even basic services (Green 2005) and specifically they lack quality budget officers to enhance the budgetary decision making process (Kittipanyasiri, 2015; Krueathep, 2014; Nagai et al., 2008). This lack of administrative capacity at the local level in Thailand prevails due to the fact that the competing central agencies do not coordinate well to plan the local governments’ capacity building and most Thai bureaucrats perceive the transferring from the central government agencies to the local government as a demotion (NESCB & Thamasat University, 2009; World Bank, 2012).
Low level of the use of information. Although efforts have been placed on data collection regarding performance monitoring and evaluation through such tools as electronic local administrative accounting system (e-LAAS) and delivery local quality management (LQM), robust use of information is not widespread. In terms of quantity, only 1,200 out of 7,853 local governments are implementing e-LAAS because of a combination of connectivity, capacity, and enforceability problems (World Bank, 2012). In terms of quality, these tools are not able to provide appropriate information on the application of public finances or service delivery quality and quantity because they focus largely on ensuring compliance with processes rather than on quality of services and outcomes (World Bank, 2012). Somewhat related to the relatively low administrative capacity discussed above, local governments in Thailand have limited relevant information available for a budgeting process and their budgetary decision making (Wongpredee & Sudhipongpracha, 2014).
In conclusion, according to the information asymmetry and the rational choice theories, with relatively low administrative capacity and limited use of information during a budgeting process, Thailand’s local governments at first glance are unlikely to be able to leverage their
44


administrative capacity and use their available information to enhance their budget allocation decisions. However, it is important to test these theories through the closer inspection of these least likely cases because if the results uphold these theoretical linkage, there is strong, robust empirical evidence to validate such theories; or, in other words, to reassure that the probability of such theoretical assumptions being true is very high (George & Bennett, 2004; Gerring, 2017).
Thailand’s Local Public Finance and Budgeting Thailand’s Decentralization and Local Governments
Subnational government administration in Thailand is structurally straightforward but politically complex. Beside the special localities (Bangkok Metropolitan Administration and Pattaya City), there are two main types of local governments in each province. First, for each town (often equivalent to a sub-district size and consisted of a group of communities), there exist a local government, which can be either a municipality or a Tambon Administrative Organization (TAO) depending on such town’s economic development and political importance. Second, for each province, there is a Provincial Administrative Organization (PAO), administering the local affairs beyond the authority of municipalities and TAOs. There are 7,853 local governments in Thailand. Note that PAOs are mainly responsible for financially assisting the other local governments in their public service delivery and TAOs do not carry out the key functional programs (i.e., park and recreation, sanitation, education, health, and transportation) according to the PAOs Act 1997 and the TAOs Act 1994 respectively. Table 3.1 gives additional details on level, economic and political importance, and number of each type of local governments in Thailand.
45


Table 3.1 Types of Local Governments in Thailand and Their Differences
Type Level1 Political Significance Number2
Provincial Administrative Organization (PAO) Province (several districts together) Managing inter-jurisdictional public services within a province 76
Metropolitan Municipality Sub-district Metropolis of a certain region (i.e., population is over 50,000) 30
City Municipality Sub-district Provincial capital or populated sub-district (i.e., population is between 10,000-50,000) 178
Tambon Municipality Sub-district Populated sub-district in a certain district (i.e., population is between 7,000-10,000) 2,233
Tambon Administrative Organization (TAO) Sub-district Rural community 5,333
Special Localities [Bangkok Metropolitan Administration (BMA), and Pattaya Cityl Ranging from a sub-district to a province Fast-growing metropolitans (i.e. the country’s capital and the cities of tourism) 2
Total 7,852
Note: 1. Two or more sub-districts constitute a district and two districts or more constitute a province. There is no district-level local government.
2. Number of each type of local governments was updated on March 24, 2018 by the Research and Development Unit, Department of Local Administration, Ministry of Interior (through http://www.dla.go.tli/work/abt/index.isr)).
Politically speaking, decentralization in Thailand is unique due to the country’s indomitable centralization. Local governments in Thailand are significantly under the supervision of the Department of Local Administration (DOLA), Ministry of Interior (MOI). According to MOI’s ministerial regulation 2008, DOLA is responsible for promoting and supporting the work of local governments through the development and series of advices on the local development plans, personnel administration, finance, and administration in order to strengthen the capacity and efficiency of the local governments on public service provision. As a part of its responsibility regarding giving advice on finance and fiscal management, DOLA encourages the local governments to adopt the performance management and budgeting. Every year, DOLA evaluates all the local administrative governments in four key administrative aspects—management, human resource management and council affairs, financial management,
46


and public services in order to learn about their administrative strengths and weaknesses of local administrative organizations and encourage them to design the local development plan accordingly.
In addition to DOLA’s supervision, deconcentration of power as well as limited administrative and fiscal decentralization are the main obstacles to local autonomy. The central government manages local government affairs by maintaining a powerful delegation mechanism in a deconcentrated system. Provincial governors and district government officials (both of whom are MOI officials) have the statutory authority to direct and order government officials from other ministries and departments at the provincial and district levels: the vertical relationship between ministries and department outweighs horizontal coordination among central governments offices at these subnational levels (Nagai, Funatsu, & Kagoya, 2008). This powerful deconcentrated structure prevails even after the promulgation of decentralization policy in 1997. In every locality, there exists both deconcentrated administration of central government and local administration. This dual-system of administration is considered a major characteristic of local administration in Thailand (Nagai, Funatsu, & Kagoya, 2008).
Local people elect their local representatives who later draft the four-year local development plans—identifying local needs, framing local policies, and providing local public goods and services. However, these representatives do not have full administrative and fiscal autonomy. In term of administrative aspect, there were originally 245 items of local goods and services that the central ministries and departments planned to transfer to be local governments’ responsibilities by 2003. However, the central authorities only transferred 181 items to the local governments by 2011 (ThaiPublica, 2011). Note that there is no progress on transferring local personnel management to local governments. Recruitment, promotion, and reassignment of high-
47


level local government officials (especially city managers) are still subject to DOLA’s review. Fundamentally, local governments’ four-year local plans and annual budgets are subject to provincial governors’ approval.
Local Governments’ Revenue and Expenditure Structure
The Decentralization Plan and Process Act of 1999 commands that local governments share at least 20 and 35 percent of the central government’s total revenue in 2001 and 2005 respectively. The central government successfully gave such shares to local government in 2001 but this share grew very slowly and did not reach the 2005 goal. In subsequent years, as the central government realized that the goal in 2005 was unattainable, it revised the Decentralization Plan and Process Act of 1999, stating that it will be sharing 35 percent of its total revenue with local governments by 2016. However, the recent trend of the share of revenue between the local governments and the central government still suggests the very slow growth of the share of local government’s revenue. From 2011 to 2015, the average share of local government revenue to total revenue of all government is equal to 21.3 percent. Table 3.2 details the local governments’ revenue sources and compares the local governments’ total revenue to that of the central government.
48


Table 3.2 Source of Local Governments’ Revenue (Unit: Million Baht)
Source of Revenue Fiscal Year
2011 2012 2013 2014 2015
Own-source revenue 38,745.96 46,529.72 50,281.54 56,306.25 61,458.00
Percent of local revenue 8.98 8.78 8.78 9.04 9.51
Shared tax (collected by the central government) 148,109.04 175,457.28 187,988.46 203,818.75 218,222.00
Percent of local revenue 34.34 33.11 32.83 32.74 33.76
Local tax (collected by the central government) 70,500.00 86,900.00 97,900.00 109,000.00 109,000.00
Percent of local revenue 16.35 16.40 17.10 17.51 16.86
Intergovernmental fiscal transfers 173,900.00 221,091.79 236,500.00 253,500.00 257,663.78
Percent of local revenue 40.32 41.72 41.30 40.71 39.86
Total Revenue of the Local Governments 431,255.00 529,978.79 572,670.00 622,625.00 646,343.78
Total Revenue of the Central Government 1,650,000.00 1,980,000.00 2,100,000.00 2,275,000.00 2,325,000.00
Percentage of local governments' revenue to total revenues of all governments 20.72 21.11 21.43 21.49 21.75
Source: Fiscal Policy Office, Ministry of Finance (though
http://www.fpo.go.th/FPO/index2.php?mod=Content&file=contentview&contentID=CNT00068Q3&categorvID=C AT0001183).
Note: The percentage of local revenue from shared tax is specifically fixed by such local government acts as the Provincial Administrative Organization Act 1997, the Municipality Act 1954, and Bangkok Metropolitan Administrative Organization Act 1985. Meanwhile, the intergovernmental fiscal transfers are dependent on the central government's intergovernmental policy, which varies from year to year.
It is also worth noting that the local governments’ own-source revenue is only around 9
percent. Meanwhile, local governments depend on general grants from the central government,
which consist of approximately 40 percent of the local governments’ total revenue between 2011
and 2015. In addition, a strong political connection with the central government is allegedly the
main factor making a certain local government likely to receive relatively large and consistent
grants from the central governments (Wongpredee & Sudhipongpracha, 2014).
49


On the expenditure side, the largest share of the local governments’ expenditure is general administration. The programs with the relatively high spending are education, which under the supervision of the education department, and housing and manufacturing and infrastructure, which are under the supervision of the public works department. It is also worth noting that even though social security policy is the central government’s responsibility, it is carried out by the local governments. The central government calculates the number of disadvantaged people (the elderly, HIV-AIDS patients, and the disabled) living in each jurisdiction and financially supports these groups of people through the special transfers to local governments. Then the local governments distribute such benefits to the disadvantaged. Table 3.3 details all programs of local governments’ expenditure.
50


Table 3.3 Local Governments’ Expenditure in 2012 (Unit: Million Baht)1
Expenditure Programs Provincial Administrativ e Organizations Municipalities Tambon Administrativ e Organizations Total Expenditure by Program
General administration 6,203.31 (20.44%) 14,925.01 (27.02%) 12,845.15 (33.61%) 33,973.47 (27.44%)
Security 126.68 (0.42%) 1,896.91 (3.43%) 506.93 (1.33%) 2,530.52 (2.04%)
Education 4,174.79 (13.76%) 9,863.14 (17.86%) 5,223.14 (2.08%) 19,261.07 (15.56%)
Public health 2,734.33 (9.01%) 2,733.81 (4.95%) 796.81 (1.79%) 6,264.95 (5.06%)
Social welfare 391.70 (1.29%) 1,141.05 (2.07%) 684.29 (13.76%) 2,217.04 (1.79%)
Civil Engineering 4,095.06 (13.49%) 11,370.10 (20.59%) 5,257.74 (13.76%) 20,722.90 (16.74%)
Community building 1,191.73 (3.93%) 1,226.89 (2.22%) 535.61 (1.40%) 2,954.23 (2.39%)
Religion, culture, and recreation 1,810.66 (5.97%) 1,190.53 (2.16%) 852.26 (2.23%) 3,853.45 (3.11%)
Manufacturing and infrastructure 4,541.40 (14.96%) 2,003.19 (3.63%) 4,969.82 (13.00%) 11,514.42 (9.30%)
Agriculture 1,256.83 (4.14%) 145.10 (0.26%) 363.84 (0.95%) 1,765.77 (1.43%)
Commerce 1,240.92 (4.09%) 266.95 (0.48%) 355.38 (0.93%) 1,863.25 (1.51%)
Central budget2 2,580.68 (8.50%) 8,466.07 (15.33%) 5,830.64 (15.25%) 16,877.39 (13.63%)
Total Expenditure 30,348.12 55,228.74 38,221.60 123,798.46
Source: Center for Local Initiatives and Governance, Chulalongkom University (Forthcoming). Local Governments’ Fiscal Data in Thailand: How the Central Government’s Policy Affect Their Local Governments’ Fiscal Conditions. Bangkok: KingPrajadipok’s Institute.
Note: 1. The fiscal data included in this table are from 2,560 local governments (47 Provincial Administrative Organizations, 1,101 Municipalities, and 1,512 Tambon Administrative Organizations) or approximately 32.60 percent of all local governments in Thailand.
2. Central budget includes such expense as debt service, emergency funds, infrastructure maintenance, and social security for the disadvantaged especially according to the central government’s policy.
Local Budget Process
A particular local budget system emphatically reflects its locality’s historical traditions and diversity in culture, capacity, and political and administrative institutions (Mullins, 2007). Thailand’s local governments are no exception. Their local budget system, called the strategic
51


performance-based budget system, depicts the country’s distinctive assemblage of fiscal decentralization, performance management, civic participation and intergovernmental relations. Broadly, their medium-term expenditure frameworks and annual budgets are well-established within the local governments by local councils and local administrators. The annual budget calendar of the local governments follows that of the central government, running from the beginning of October to the end of September. This study delineates this local budget process via the budget development cycle which is divided into three stages—sequential budget planning, budget formation, and budget approval. Each of these possesses numerous policy and administrative activities as follows.
Budget planning. Regarding budget planning or budget pre-preparation (Mullins, 2007), firstly, the local governments start their budget planning as early as January and continue through approximately June, soliciting and gathering community needs and problems, policy priorities and expectations as well as challenges and opportunities of the current local public service provisions. During this stage, views and opinions from any parties are highly welcome through many approaches of civic participation. Three approaches widely adopted include focus group consultation, civic forum, and citizen surveys (Lorsuwannarat, 2017; Suwanmala, 2007).
Focus group consultation begins with the mayors and local administrators introducing a particular plan for a certain policy issue to their community members (especially the concerned target groups). Via a series of focus group meetings, local individuals are given opportunity not only to make comments or recommendations but also to determine whether the plan corresponds to their needs and preferences. It is expected that this approach of focus group consultation can promote the citizen’s sense of ownership in community development.
52


Civic forums are particularly popular among relatively small local governments in the rural areas. This civic participation approach features a civic forum committee which formally comprises several groups of stakeholders including village heads, community leaders, representatives of occupational groups, women and youth leaders, spiritual and religious leaders, school teachers, and provincial or local senior administrators from the central government’s agencies. Regularly, their meetings involve identifying and prioritizing community problems, needs, and preferences and then initiating local development programs or activities. Note that both policy initiations and functional programs from civic forums are usually highly respected by the local councils.
Citizen surveys inquire about the satisfaction levels of local public services and the provision of traditional or religious activities (i.e., long-holiday festivals and celebration events). Local public service users and festival or event participants are actively asked about their satisfaction and suggestions, which are indispensable for making such activities more effective— culturally attractive and cost-saving. It is worth noting that this is one of the crucial parts of performance management extensively promoted by DOLA following the national agenda of budget reform (Sutheewasinnon et al., 2016).
Secondly, the mayors together with the local administrators (i.e., municipal clerks, finance directors as well as directors of spending agencies) formulate the local development schemes based largely on the community needs and problems, policy priorities and expectations, and challenges and opportunities from the aforementioned focus group consultation, civic forum, and citizen surveys. Specifically, each local government sets overall policy platforms by identifying broad goals and policy outcomes or results; developing preliminary programmatic,
53


operating, and capital plans; and contemplating the realistic budget ceilings for functional allocations, carefully keeping in mind their revenue and expenditure statistics and forecasts.
Thirdly, progressing their established policy platform to a strategic performance-based plan, the local executives, the local administrators, and the local development plan committee (whose members are mostly from the civic forum committee) thoroughly analyze the whole community’s strengths, weaknesses, opportunity, and threats. Broad goals and policy outcomes or results and developing preliminary programmatic, operating, and capital plans are prioritized, sequenced, and packaged according to the four-year timeframe. Simultaneously, they also often ponder the national policy agendas, the regional development scheme, and the provincial plans in order to ensure policy harmonization across levels of governments. Eventually they officially offer the four-year local strategic plan which reveals the locality’s policy outlook, overall development goals, programmatic priorities, and fiscal environment (either fiscal restraints and expectations for budget reductions or opportunities for an expansion of tax and nontax revenue sources).
Budget formation. The budget formation stage of local budget process takes three months, specifically from June to August. It generally begins when the local executives and the local administrators prepare a draft of the annual budget. They painstakingly review the estimated revenue and expenditure, financial statements, agency budget requests, and performance indicators as required by DOLA’s guidelines on managerial and service delivery standards.
Note that, in order to gain efficiency and effectiveness, they are advised that they review those documents along programmatic, technical, and managerial dimensions (Mullins, 2007). For programmatic dimension, the budget proposal must be consistent with the policy platform and
54


priority highlighted in the four-year local strategic plan. Meanwhile, the technical review ensures that all of the required estimates are accurate and consistent with historic spending as well as spending envelopes. Managerial review guarantees that local administrators adhere to administrative and legal requirements, following the designated mechanisms, systems, process, procedures, and guidelines for conducting administrative and programmatic operations.
After that, there exists the final review of the annual budget proposals, which involves an executive budget development process (Mullins, 2007). A mayor and his or her advisory board deliberate spending requests in terms of the trade-offs of relative costs and outcomes across spending units and program objectives and the trade-offs of spending priorities and managerial operations. Then, a sharp or mild reduction may possibly apply to some spending requests. Nevertheless, this review results in compilation of spending requests into a package of the locality’s working budget proposal. Prior to the middle of August, the mayor submits this budget proposal to his or her local council.
Budget approval. This local budget stage of legislative review and approval proceeds from the middle of August to the end of September. Regulatorily, dominating the budget approval stage, a local council (an elected legislative body) is designated to exercise its amendatory power—an authority to substantially or moderately alter the budget proposals. The council first examines the overall policy agendas and key fiscal aspects. Next, the comprehensive review is conducted usually by the legislative budget committee—a group of budget specialists who provide assistance to the council especially in such tasks as deliberation of policy outlook and priority, reviews aggregate revenues and expenditures, a comparison of programmatic alternatives, and a consideration of spending at the functional and organizational levels. Due to the fact that informed review is inevitably information-intensive, the council as well as this
55


legislative budget committee have authority to collect information and to request testimony of both mayors and administrative officials on the merits of all elements of the proposed budgets (Mullins, 2007). After that the council votes to approve the budget proposals according to majority rule. Lastly, provincial governors or chiefs of district administration finalize and verify these budget proposals.
Information Use in Local Budgeting in Thailand
Based on the local budgeting process described in the last section, it can be said that local budget actors use information for their decision making. Particularly, after gathering such information as community needs and problems, policy priorities and expectations, and challenges and opportunities of the current local public service provisions from focus group consultation, civic forum, and citizen surveys, local budget actors can deliberate such information during the budget formation and the budget approval. Throughout the budget formation stage, a chief executive and local administrators prepare the proposal by considering this information against technical and managerial conditions in order to formulate a four-year local strategic plan. Eventually, during the budget approval stage, local legislators and possibly budget specialists review such proposed plan by inquiring more information from local administrators. They mainly depend on information regarding a projection and a profile of previous-year spending of each programmatic budget so as to approve the budget proposals with the majority rule.
In conclusion, Thailand’s local budgeting process offers the policy arena which local budget actors can deliberate information in order to make budgetary decisions. However, with limited administrative capacity, at first glance it is unlikely that local administrators have enough access to information and thereby they cannot use such information to enhance their budget
56


allocation. Accordingly, it is important to test this theoretical linkage based on information economics and budget rationality because if the result uphold it, there is a strong empirical evidence to validate related theories. In addition, due to the fact that there are few studies on use of information and budget decisions in developing countries, this study applies mixed methods to capture complexity of a developing country’s local budgeting as well as to test the aforementioned theoretical linkage. Chapter 4 provides the methodological details of this study highlighting the pilot study, the quantitative component, the qualitative component, and finally the triangulation of research results.
57


CHAPTER IV
RESEARCH DESIGN AND METHODOLOGY
As formulated in Chapter 2, the overarching research question for this study is: How do policymakers with different administrative capacities utilize fiscal information during a budget process in order to have an impact on budget decisions? The focus of this study is to understand the use of information and its influence on budget decisions—the complex phenomenon which is fundamentally subjective, political, multiple, and constructive through administrative interactions during a budgeting process (Feldman & March, 1981; Wildavsky, 1978). This study, thus, applies a mixed methods design to explore and explain the complex interrelationship between the administrative capacity, information characteristics, information utilization, and budget decisions (Newman, Ridenour, Newman, & DeMarco, 2003). Conducting a mixed methods research project involves combining quantitative and qualitative approaches in the same study or, in other words, collecting, analyzing, and interpreting both quantitative and qualitative data in a single study or in a series of studies that explore the same phenomenon (Collins, Onwuegbuzie, & Sutton, 2006; Onwuegbuzie & Leech, 2006). The logic behind the mixed methods research includes the use of induction (or discovery of patterns), deduction (testing of theories and hypotheses), and abduction (uncovering and relying on the best of a set of explanations for understanding one’s results) (Johnson & Onweugbuzie, 2004).
To explore the nature of how information is used in the context of developing counties which is still understudied, this study first introduces a pilot study in order to guide and develop the research plan (Kim, 2011; Teijlingen & Hundley, 2002). This pilot study is of qualitative approach so as to provide a causal description on the interrelation of concepts of interest (Miles, Huberman, & Saldana, 2014). Specifically, the pilot study aims to improve content validity by
58


assessing whether the study’s questionnaire items (the survey instrument) as well as interview questions to ensure that the research measurements truly and adequately represent the study’s concepts and constructs discussed in Chapter 2. Once the pilot study verifies the questionnaire items and the interview questions, this study introduces the full-scale research with the quantitative mixed methods research design (Morse, 1997). It first implements the quantitative approach with survey as the mode of data collection. After analyzing the quantitative results, this study implements its qualitative approach with interview as the mode of data collection primarily in order to complement the quantitative results.
To elucidate this research design, this chapter first elaborates the methods design. Then, it details the sampling design, instrumentations, and data analyses for the quantitative and qualitative approaches, respectively. Note that both of the quantitative and qualitative approaches are largely based on the results from the pilot study (i.e., operationalization, data collection instruments, and analytical techniques). After that, this chapter offers how the results of these two approaches are triangulated. Finally, it presents threats to legitimation given the type of mixed method design proposed by this study.
The quantitative mixed methods design
As Chapter 2 demonstrates, this study locates substantial and relevant literature on the topic, creates a theoretical framework, and identifies the testable hypotheses. At the core of this study based on the philosophy of science is, therefore, post-positivism which features a hypothetico-deductive model and a confirmatory research (Morse, 1997; Teddlie & Tashakkori, 2009). Fundamentally, this study warrants a quantitative inquiry. However, in order to provide a holistic description of the information use and how it influences the budget allocation, a qualitative approach is also necessary for elaborating the phenomenon of interest, especially
59


descriptively complementing the key findings based on the precedentially conducted quantitative phase. In short, as a result, this study applies the partially mixed sequential dominant status design with the greater emphasis on the quantitative phase which is for short written quantitative mixed methods (Leech & Onwuegbuzie, 2007; Newman et al., 2003). The following sections delineate the methodological components of quantitative and qualitative phases, respectively.
The Quantitative Approach
In order to collect data and then empirically describe and explain several concepts and the interrelationships among them graphically depicted in the theoretical framework in Chapter 2, this quantitative phase applies a survey research with a systematic questionnaire asking prescribed questions of respondents (Singleton & Straits, 2010). The responders’ answers are later numerically coded and analyzed via multiple regression analyses to estimate, ceteris paribus, the effects of causal factors (i.e., administrative capacity and reliability and relevance of information) on outcome variables (i.e., the use of information and budget allocations) (Wooldridge, 2010). The following parts details this survey research which constitutes the quantitative approach.
Sampling Design
Population. As mentioned, local governments in Thailand represent the less-likely cases, so the population in this research is Thailand’s local governments (Gerring, 2007b, 2012). However, this research first excludes from the sample pool Bangkok Metropolitan Authority (BMA), Pattaya City, the Provincial Administrative Organizations (PAOs), and the Tambon Administrative Organizations (TAOs). This is because BMA and Pattaya City are special local governments which possess different organizational and political structures from other local governments. Meanwhile, PAOs are mainly responsible for financially assisting the other local
60


governments in their public service delivery and TAOs do not carry out the key functional programs (i.e., park and recreation, sanitation, education, health, and transportation) according to the PAOs Act 1997 and the TAOs Act 1994 respectively. After eliminating local governments based on the extraneous variables (the organizational structure and the scope of functional programs), 2,441 municipalities are this study’s sample population—the evidence that will be subjected to direct examination (Gerring, 2012).
Sampling scheme. Like other quantitative studies, this study aims to make statistical generalizations—generalizing findings and inferences from a representative statistical sample (i.e., municipalities in Thailand) to the population from which the sample was drawn (all the municipalities in Thailand) (Onwuegbuzie, Onwuegbuzie, Collins, & Collins, 2007).
Sample size. Given that the unknown percentage of the policymakers in Thai municipalities who deliberate and actually utilize relevant and reliable information for their budget proposals and discussion as well as the unknown standard deviation of such information utilization, this study applies the Cochran formula to calculating the sample size. Specifically, this study sets the confidence level at 95% (1.96 for its Z values) with the level of precision at least at 5 percent. Additionally, it assumes that half of policymakers use performance information during the budget process. As a result, after adjusting for the target population of 2,441 municipalities, the calculation suggests that the sample size of 333 municipalities be adequate to give the confidence levels the study requires for the budget allocation. In order to maximize a response rate, a voluntary response sample is applied.
It is possible that the size of municipalities in Thailand (small, medium, or large) determines the extraneous factors—for instance, the organizational structure, the scope of functional programs and legal mandates (World Bank, 2012), and long-term debt level and debt
61


service ratio (Krueathep, Nudam, & Termtawewut, 2015)—which may affect the relationships of interest. This study, therefore, adopts the proportionate stratified sampling method in order to ensure that each of three groups of these municipalities is adequately represented within the whole sample population of the study (Teddlie & Yu, 2007). Table 4.1 calculates the sample size for each stratum, resulting in the sample selection of three large, twenty five medium, and 305 small municipalities.
Table 4.1 The Stratified Sampling and Sample Size for the Survey
Municipality’s Type (Size) Number of Municipalities in Stratum Stratum Sample Size
Nakon or Metropolitan (Large) 30 3
Meung or City (Medium) 178 25
Tambon or Sub-District (Small) 2,233 305
Total 333
Cases. Cases are a restricted range of units that are bounded in terms of settings (e.g., a particular place and point in time that the research is conducted) and constructs (e.g., a few possible variables from the theories on which the operational research model is based) (Gerring, 2012; Van de Ven, 2007). From the temporal perspective, the literature asserts that it takes at least a few years for the use of performance information to become established and eventually have an impact on budget allocations (Gilmour & Lewis, 2006; Melkers & Willoughby, 2001; Zaltsman, 2009). Furthermore, the impact of performance information on the budget allocation is more pronounced during the boom than the bust years (Hou, Lunsford, Sides, & Jones, 2011b). Since the Department of Local Administration (DOLA), Ministry of Interior has required, since 2006, local budget officials to collect and present related fiscal information and statistics to the local policymakers (Article 3, Section 22 of Ministry of Interior’s Regulation on Local Budgetary Process, 2000) (NESCB & Thamasat University, 2009; World Bank, 2012), this research can expect the see such impact anytime approximately after 2003. This study uses Thai
62


local government data from the most recent budget year of 2019—from October 2018 to September 2019—because the budget actors are likely to accurately recall their experience on information use throughout the budget process and such period of time represents the country’s economic growth trend, approximately 2.64 percent per year (Asian Development Bank, 2018).
From the construct-bounded perspective, the principal-agent theory suggests researchers capture the change of annual budgets in specific programmatic functions—a designated balance, the portion of the government expenditure that is restricted, committed, assigned, reserved, or designated under existing laws for specific uses and purposes by the current administration (Hou, 2003; Tomes, Berger, & Bassett, 2011). Empirically, the performance budgeting scholars confirm that there are certain situations when the policymakers or budget actors can drive changes in their programmatic budgets and improve the operation and outcome of such programs in general through the use of performance information (Hou et al., 2011b). Under the Thai local governments’ context, the key functional programs based on the composition of spending levels in Table 3.3 include civil engineering (16.74%) and education (15.56%). This study therefore concentrates on the budget process of these two functional programs. Note that the department of education is responsible for education programs meanwhile the department of public works are responsible for civil engineering programs.
Unit of analysis and observation. The unit of analysis, the entities which are being described and generalized (Singleton & Straits, 2010), in this study is both the individual and the organization.
As information utilization is a result of a cognitive process of each policymaker, the subquestion of how policymakers or budget actors with different administrative capacities utilize relevant and reliable information during a budget process uses the individual as a unit of
63


analysis. However, there are three main groups of budget actors in the municipalities in Thailand who are directly and routinely involved in budget preparation, budget debates, and budget decision making—city mayors, municipal clerks, and the department heads. The key budget actors this study is interested are the budget actors who (1) possess a variety of information, (2) utilize such information throughout the budget process, and (3) have impact on budget allocation. Even though both city mayors and municipal clerks officially have decision-making authority, they exercise such authority only on strategic plans. The functional plans are largely under the department’s consideration. The departments are legally responsible for thoroughly preparing budget proposals, presenting, and defending them during the budget debates. This study, therefore, considers the department heads as the key budget actors and individuals of interest.
Meanwhile, budget decisions or budget allocations for the functional programs are a result of a collective process where budget actors altogether distribute their financial resources to spending programs according to their organization’s policy directions and priorities. Therefore, the sub-question of how information utilization influences budget decisions uses the organization as a unit of analysis. This study also asks the department heads to provide their organizations’ budget allocation.
Table 4.2 sums up the sampling design—population, sample, cases, and observations— and the example of principal concepts according to the theoretical framework operationalized in the following part.
64


Table 4.2 Sampling Design
Population1 Sample Case Observation2
Local governments in Thailand (7,852 Units) Municipalities (2,441 Units) Municipality 1 Xl.1,1 X2.1.1 Xl.1,2 X2.1.2 yi.i.i y2,u y 1,1,2 y 2,1,2
Municipality 2 Xl.2,1 X2.2.1 Xl.2,2 X2.2.2 yu.i y2,2,i yu,2 y2,2,2
Municipality 3 Xl.3,1 X2.3.1 Xl.3,2 X2.3.2 yu.i y2,3,i yu,2 y2,3,2



Municipality 333 Xl,333,1 x2,333,1 Xl,333,2 X2,333,2 y 1,333,1 Y2,333,1 yi,333,2 y2,333,2
1 Number of total local governments and municipality in Thailand is from the DOLA, Ministry of Interior retrieved from http://www.dla.go.tli/work/abt/index.isp on March 24, 2018.
2 For each observation: x is variable of interest; y is dependent variable; i is variable lor 2; j is municipality 1 through 333; and k represents the department when 1 is the education department and 2 is the public works department.
Note that concepts of interest in this study include administrative capacity, budget rationality, information use, and actual and perceptional budget allocations. Some of these concepts are operationalized and verified by the pilot study which is the center of discussion in the next section.
Operationalizing Concepts of Interest
This section reconceptualizes the constructs introduced in Chapter 2 and provides their operational measures. The key concepts for this study cover administrative capacity, information production, information utilization, and budget decisions. Note that the pilot study is conducted in order to operationalize these key concepts and specify their measures which fit the context of Thailand’s local budgeting. The following parts explain why and how the pilot study is administered.
Pilot Study
A pilot study is a miniature version of full-scaled research which features planned
methods in order to identify and resolve epistemological and methodological issues as well as to guide and develop the research plan (Kim, 2011; Teijlingen & Hundley, 2002). Even though
65


conducting a pilot study does not guarantee success in the full-scaled research, it does increase the likelihood of such success via enhancing the research’s validity and reliability (Barriball & While, 1994; Teijlingen & Hundley, 2002). The following parts describe goals, how to administer, and the results of the pilot study.
Goals
Reliability. This pilot study improves content validity by assessing whether the measures of the related concepts or the specific survey and interview questions as the research measurements truly and adequately represent the study’s concepts. In addition, a pilot study can increase internal validity, or specifically population validity, since it helps recruit key participants from the whole population. (Collins et al., 2006; Onwuegbuzie, 2003). In other words, with condition-seeking methods, the study seeks to discover which group of local budget actors both comprehensively utilize information and potentially influence their organization budget decisions.
Validity. This pilot study promotes instrument fidelity (Collins et al., 2006;
Onwuegbuzie et al., 2007). It provides the study with an opportunity to assess the appropriateness and utility of existing survey questions and interview protocol with a view to making modifications or creating and improving new instruments so that any differences in the interview answers are due to differences among the interviewees rather than in the questions asked (Barriball & While, 1994; Collins et al., 2006). Particularly, all questions should conform to the principles of specification (the focus of each question), division (the questions’ appropriate sequence and wording) and tacit assumption (the true meanings behind interviewee’s answers) (Lazarsfeld, 1935).
66


In short, in order to enhance construct validity, the pilot study tests whether each survey and interview question as a measurement actually focuses on and reflect related true meaning of concepts or constructs through the interviewees’ viewpoint. It also verifies whether interview protocol’s and questionnaire’s sequences and wordings are appropriate so as to increase the instruments’ reliability. Then it identifies the key budget actors from a variety of groups of individuals working in local governments in order to promote population validity.
How to Administer the Pilot Study
The pilot study is consisted of three sequential phases. The first phase aims at understanding the roles of each group of budget actors on information collection and utilization and operationalizing the key concepts of administrative capacity, information production, and information use. Three municipalities were selected on the basis of convenience and chain or snowball sampling strategies in order to identify cases that offer rich information enough to operationalize the concepts of interest (Miles & Huberman, 1994). This study thus interviewed key budgetary actors (mayors, municipal clerks, and department heads) from these three municipalities. Totally, there were twenty individuals for the pilot interviews (three mayors, three municipal clerks, and fourteen department heads).
The second phase of the pilot study focused on testing and verifying the interview protocol. The participants are the same as the first phase. The interview questions which had been derived from the literature are used and then revised based on the participants’ understanding of those questions as well as their recommendations. Table 4.3 shows the list of interview protocol being verified in this pilot study.
67


Table 4.3 Interview Protocol Based on the Literature
Interview Approach Construct of Interest Question
1. Unstructured Tenure and years in What is your educational and professional background?
conversational supervisor position How long have you been working in this organization?
interview Budgetary orientation Are you engaged in budgeting? What is your budgetary
approach responsibilities? What is your role during the budget process? How do you interact with other groups of budget actors?
2. Interview Budgetary orientation Can you prioritize these following competing spending
guide approach preferences: financial control, efficiency, management techniques, policy issues, and executive’s plan? Why?
Structural Which budgeting system does your organization adopt?
isomorphism Why such system is adopted? What are managerial structures, standard of procedures, and information such system requires? How has your organization’s budget process changed over
Budget the last 4 years or until the recent budget system has been in
incrementalism place? Currently, how does a budget process in your organization
Information usually proceed? What are the sequence and the key
utilization activities in such process? On which activity the budget committee spend quite a long time and why?
3. Highly Information provider What kind of information do you usually receive from the
structured departments?
interview Does your organization have an official unit providing
approach information for policymakers? What kind of information do they provide? How do they get such information?
Information Which types of information you find useful (important or
characteristics adaptable)? Why? Which types of information you find reliable? Why? Which piece of information do you usually think of and
Information bring up during the budget process?
utilization Which piece of information do you base your budget decisions on?
Budget decisions What factors carry the most weight in the definition of your organization’s budget decisions? How do such factors impact the spending priorities, eliminating the duplicative programs, or cost-saving? If there are differences in allocation priorities how does the committee resolve them? Which spending program or functions usually brings such differences?
4. Non- Information What are your plans on information management system?
controversial Production Which information should be additionally collected?
wrap-up How does this additional information enhance the current
approach budget process? How can it ultimately benefit local people?
68


The third phase involved the survey questions, which were partially based on the results of the first phase. For each concept, the questionnaire had multiple-item measures with Likert-type scales. The study applies the convenience sampling strategy to select forty department heads to participate in the pretest this questionnaire. In order to measure the internal consistency or reliability of this survey instrument, the pilot study calculated the Cronbach’s Alpha. Since this study is an exploratory research based on its new measures proposed by the results of the first phase above, a Cronbach’s Alpha value of 0.70 is acceptable (Campbell & Fiske, 1959; Leech, Barrett, & Morgan, 2011). Note that the values of Cronbach’s Alpha are reported in the descriptive statistics section in Chapter 5.
The Pilot Study’s Results
Participants. Even though the chief executives are obviously key budget decision makers, they do not possess a variety of types of information. Their budget decisions are often based on communal problems and demands. Somewhat like city mayors, municipal clerks only utilize communal problems and demands and local strategic plans (i.e., policy and priority) as their key information for formulating budget proposals.
However, during the budget debate, a chief executive and a municipal clerk have department heads present, explain, or even defend the budget proposals. This is because department heads from each department possess technical information specific to their functions or policy areas. This is necessary when the study examines the relationship between the information and the budget allocation. In addition, there are a couple of functional departments (i.e., departments of education and public works) which possibly receive information from several information sources with different levels of relevance and reliability. The more relevant
69


and reliable the information perceived by policymakers, the more likely it is utilized during a budget process.
Administrative capacity. The literature operationalizes administrative capacity not by capacity itself but by its proxy of professional qualifications such as educational background and work experiences (Potter & Brough, 2004; Tamkin, Hillage, & Willison, 2002). There are also other components of administrative capacity including, for instance, skills and knowledge that are less investigated in the literature. This study, thus, chooses to explore these components of administrative capacity. Particularly in this context, skills and knowledge are necessary for budget actors to initiate, draft, and defend budget, or called capacity to budget in this study. The budget actors in the pilot study from the three municipalities—three mayors, three municipal clerks, and fourteen department heads, in total twenty individuals—identify key components of capacity to budget. After applying the constant comparative method, this study finds that there are six kinds of administrative capacities necessary for budget actors as follows.
(1) Participatory budgeting—ability to solicit citizen opinions, promote local participation and model preferences.
(2) Communication skills—ability to prepare data presentation and express and discuss opinions.
(3) Intergovernmental relations—ability to network and collaborate with other local governments, the central government, and its provincial agencies.
(4) Strategic management—ability to structure problems and strategies based on the community’s strengths, weaknesses, opportunities, threats.
(5) Project management—ability to assess project feasibility, monitor and evaluate project, and conduct stakeholder analysis.
70


(6) Accounting and financial management—ability to apply budgetary techniques,
conduct cost-benefit analysis, calculate and interpret financial ratios, and apply risk management.
Information production. Information production encompasses available information and information characteristics (reliability and relevance). The budget actors in the pilot study identify two main types of information—performance information and non-performance information.
Performance information. Performance information includes the value of money, service quality, and policy results.
(1) Budget actors see the value of money for a specific project or program through the unit costs (based on outputs and a standard price or a surveyed price) and the service quality. Output of local public goods and services are, for example, the length of local roads constructed, the number of patients treated by the municipal health centers, the primary school graduation rate, the amount of sewage treated, and the number of participants in each of local festivities. The budget actors compare such outputs with cost, either their total costs or unit costs based largely on the standard price given by the Ministry of Commerce or the price survey of that each department conducts at least three local stores or merchandisers.
(2) Service quality, for instant, includes the rate of middle school students successfully accepted into the leading high schools, the sports trophies the municipalities’ school win throughout the year, and the reduced rate of rabies fatalities compared to the previous years.
71


(3) Policy results are positive impacts initiated or stimulated by the municipalities’ projects. The case in point is that local roads can improve the road transportation
between different communities so that it is easier for the residents to have more social and political gatherings as well as to have more economic and trading activities, such as market fairs, farmer markets, and banking mobile services.
Non-performance information. The non-performance information includes spending ceilings, executives’ policies and priorities, community’s problems and demands, and professional standards and knowledge.
(1) Spending ceilings are the programmatical budgetary limits each department receives for each fiscal year. These ceilings are based mainly on revenue forecasts or the previous-year amounts and the change in municipal revenue which is the estimated summation of intergovernmental transfers (i.e., grants, shared taxes, and local taxes collected by the central government’s agencies) and own-source revenues (i.e., property taxes).
(2) The executives’ policies and priorities include the flagship policies, town initiatives, as well as development directions promoted through their electoral campaigns. For example, some municipalities have a city-wide plan to push their communities to be tourist destinations while others have a smaller plan to reduce use of the plastic and increase the recycling activities. Note that the executives as politicians consider these policies and priorities the promises they give and hold themselves responsible. Consequently, these polities and priorities are politically sensitive especially if they were left unimplemented.
72


(3) Communities’ problems and demands are usually local issues or agendas raised directly by the residents or sometimes by the community leaders through town meetings. For example, the residents request the municipalities to install a lighting safety system in their communities, to construct or renovate the local market buildings and facilities, or to financially support the local craft groups in order that their handmade goods can attract more customers especially from outside the communities. However, communities’ problems also mean unfortunate events or emergencies (for example, floods, droughts, fires, and epidemics) to which municipalities must react fast.
(4) Professional standard and knowledge are technical or specialized information important or necessary for initiating as well as implementing a certain project. This type of information becomes prominent when local residents can only identify the problematic situations but cannot suggest any possible solutions. This is because certain problems require sophisticated technology or specific technicality to analyze and come up with technical recommendations. For example, in order to have a sustainable solution for flooded areas, the municipality’s department of infrastructure must design a watershed flood control dam and present it in form of a special project to the town meeting.
Information characteristics. Before deliberating how to use the information mentioned above, the budget actors prefer reliable and relevant information. For reliability, the information should be verified, valid, and updated (Lundin & Oberg, 2014). Specifically, the information should be from the source that should be of integrity and benevolence, have good command of data gathering, processing, and analysis. The budget actors usually gauge the reliability of
73


information by its provider or source, which includes self-observation, performance contracts and reports, the municipal information centers, the central government’s agencies, other local governments, the municipality’s contractors or service providers, and other professional and academic sources such as research institutes, universities, and think tanks. Meanwhile, the budget actors measure information relevance in terms of its usefulness (Wolman & Page, 2002) and applicability (Landry, Lamari, & Amara, 2003).
Information utilization. Based on the pilot study, there are two perspectives into the information utilization. First, the pilot study results point to the information utilization for decision making, which reflects the active or purposeful information utilization during the budget process (Donald P. Moynihan, Lavertu, et al., 2012; Taylor, 2007). The information use reflects the degree to which the budget actors use the information to (1) influence their budget decisions especially in terms of program prioritization; (2) combine with other data to reach budget decisions; (3) prompt further budgetary questions or enquires during budget discussion or debate; (4) use those information to support a budget decision reached by other means (e.g., professional experience); and (5) argue that a certain budget choice may not be desirable.
The second perspective of information utilization that the municipal budget actors widely discussed during the pilot study relate to the information presentation. They stated that they bundle some or all the information mentioned in the previous section and present them through story-telling strategies. Even though the empirical studies in public policy widely discusses these strategies, the local budgeting literature especially in Thailand limitedly investigates them. This study groups these strategies of information use under the context of Thailand’s local budgeting into five story-telling schemes as follows.
74


(1) “Doctrine of performance management”: Budget actors try to prepare a budget proposal with a rational decision-making process. They specify an area of inefficiency or ineffectiveness, then offer the projects as a solving mechanism. This mechanism through budget allocation does not only fix those inefficiencies and ineffectiveness but also benefits a variety of target groups (both citizens and local government alike) in terms of better quality of service, outcomes, efficiency, and effectiveness (Donald P. Moynihan, 2008b; Mucha, 2009). For example, an education department head observes the decreasing trend of average mathematics score of the primary school students. The mathematics tutor and related activities are bundled into the mathematical achievement program. This department head shows the budget committee how these tutor activities can generally enhance the students’ mathematical skill and possibly increase the pass rate. Meanwhile, some students with a flair for mathematics also can benefit from this project in the way that they may have a better chance in winning prizes at the provincial or regional mathematics competitions. In addition, the parents are probably more satisfied with the educational policy directions of the school which is led by the municipality’s executives.
(2) “We are the experts”: A budget proposal is possibly a complex collection of ideas and possibilities (i.e., technical feasibility, affordability, and value acceptability) so that budget actors present it by comparing its components with professional or expert standards (Jones et al., 2015; Mikesell, 2007; Zahariadis, 2014). This expert-led scheme of information utilization emphasizes the important consideration based on the information collection methods, standards of evidence, and analytical techniques of scientific knowledge to guide local development initiatives or projects (Corbum, 2003; Failing et al., 2007). For example, an infrastructure department head proposes the bridge construction to his or her municipal budget committee by introducing the
75


design plan including such details as width, length, height, materials, construction areas, and costs. Demographics (i.e., population and local economic activities), spatial conditions, disaster speculation and statistics are also considered. The department heads may also mention its pros and cons in comparison to other designs.
(3) “Framing”: Based largely on a particular community’s problems, some budget actors predict and present the focusing events (events emphasizing the occurrence of natural or manmade crises or disasters) and their dire consequences in terms of indicators and conditions which already happened or would possibly happen if the proposed projects were not funded (Liu, Lindquist, Vedlitz, & Vincent, 2010; Mikesell, 2007; Zahariadis, 2014). Indicators are used to assess the existence and magnitude of a condition and the scope of change made by a community’s problem itself and sometimes they are used politically to measure the magnitude of change in the hope of catching officials’ attention (Zahariadis, 2014). Conditions of a focusing event direct officials’ attention to specific evaluative dimensions of a particular community’s problems (Zahariadis, 2014). This utilization of problems’ indicators, conditions, and focusing events can at least create awareness or trigger widespread attention, a shared concern that some action is required, or a shared perception that the matter falls within the bounds of municipality’s authority (Cobb & Elder, 1972). For example, the watercourse maintenance project has been postponed for a few years. Yet, heavy rainfalls during monsoon season make a river overflow its banks and cause prolonged flooding in some areas in the municipality. The municipal clerk and the infrastructure department head survey the affected neighborhoods and estimate the flood damage. While proposing the watercourse maintenance project, they exhibit the detrimental impact on property (i.e., amount of lost residential buildings, industrial buildings, flooded crops, and increase in animal and livestock health problems) and on people (i.e., the number of missing,
76


injured, and drowned). They also emphasize that this calamity requires a quick response from the municipality, which is directly and wholly responsible for the well-being of residents in the jurisdiction.
(4) “Mandates”: Budget actors state that the projects they propose are a part of or required by mandates directly or indirectly enforced by the state statutes and the central government (Mikesell, 2007). Responsibilities for making allocation decisions and actually delivering services as a part of implementing the central government’s policies have been delegated to local governments even though those services are only partially or even not at all funded by the central government’s budgets (Radin, 2000). In order to fulfil these responsibilities, policymakers at the local level usually follow the central government’s policy directions. Budget actors, while discussing their budget proposals, elaborate how the services and policy results of such proposals can serve or promote the policies or the statutes initiated by the central government and its agencies. For example, DOLA issues the official letter to request the municipalities to actively take good care of the elderly as the central government prepares the country for an aging society and economy. In order for a budget proposal of setting up a school for the senior citizens to get approved, the social welfare department heads convince their policymakers by showing that the school’s curriculum is in line with the Second National Plan for Older Person 2002-2021 and that the policy results of senior schools match with the central government’s aging society and economy platform—the enhancement of recognition of the elderly, the dignity of old people, and the improvement of their quality of life.
(5) “Incrementalism”: Budget actors present and defend their budget proposals by arguing that their proposal is not significantly different from the previous year or insisting that “it is an old stuff’ or “it’s so small” (Mikesell, 2007). As long as the spending ceilings allows,
77


the department heads show their policymakers that there is no significant change. This can keep the current beneficiaries satisfied and the political climate unchanged. The current executives— the incumbents in the next local election—still have electoral advantage. However, some small changes are possible. For example, since the Thai new year festivity this year draws more local attendants than the previous year, the education department head may ask for additional budget to accommodate more people.
Budget Decision. The second outcome of interest for this study is budget decision, which is operationalized as perception and actual change of annual budget. The operationalization of such perception is the budget actors’ view of the impact of information utilization on budget allocation. Based on the studies on performance management, perceptional budget allocation is measured through the questions of how effective has the development and the use of relevant information been in the respondents’ public organization regarding the following issues: affecting cost savings, changing appropriation levels, reducing duplicative services or programs, eliminating ineffective services or programs, and improving service quality (Bourdeaux & Chikoto, 2008; Melkers & Willoughby, 2005; Willoughby & Melkers, 2000).
For actual budget allocation, the study retrieves the constant-baht figures of 2018 and 2019 expenditures in the programmatic functions under the departments of education and public works (Zaltsman, 2009). The spending levels on education and public infrastructures of the fiscal year 2019 as well as their increases and decreases compared to that of the year 2018 are statistically analyzed as the next section details.
Quantitative Data Analysis
The results of the pilot study allow formulation of testable regression models of how policymakers use information and how the use of information affect budget allocation.
78


Regression Models. The first model regresses administrative capacity and information characteristics on information utilization (see Appendix A for the complete list of variables in the model and its measures). This study applies the ordered logistic regression analyses because of the ordinal nature of the degree of information utilization. It also adds into these analytical models organizational factors (i.e., leadership and organizational culture) and political factors (i.e., outsiders’ use of information against the organization) (Landry et al., 2001; Moynihan, Lavertu, & Kamensky, 2012). The form of model that estimates the level of information utilization is:
Information Use = Po + Pilnformation Characteristics + P2Administrative Capacity
+ PsOrganizational Factors + P4 Political Factors + error
For the perceptional allocation, this study regresses administrative capacity, information characteristics, and information utilization on perceptions on municipal budget decisions (see Appendix B for the complete model and measures). This study applies the ordered logistic regression analyses because it asks the department heads to rate effectiveness and efficiency (four levels from low to high) of the functional programs for which the department heads are responsible. The models control for community characteristics and political factors (Melkers & Willoughby, 2005). The form of model that estimates the perceptions among department head on spending efficiency and effectiveness is:
Budget Decision = Po + Pi Administrative Capacity + P2lnformation Utilization + PsBudget Factors + P4Political Factors + errors
79


In case of actual budget allocation, this study regresses administrative capacity and information utilization on the budget levels (see Appendix C for the complete model and measures). Budgetary factors (population, areas, number of full-time employees, budget size of the previous year and own-source revenue) and other control variables are in the model (Lewis, 2013; Zaltsman, 2009). The functional form that estimates the spending levels is:
2018-2019 Budget Change = Po + Pi Administrative Capacity + P2lnformation
Utilization + PiOrganizational Factors + P4Political factors + PsBudget Factors + error The Qualitative Approach
As the second phase of the quantitative mixed methods design, the qualitative component of this study, involves conducting sixteen case study analyses about the department heads’ use of information and the impact of such information use on budget allocation in order to complement and enrich the quantitative findings. Interviews form the basis of data collection. Meanwhile, keyword-in-context (KWIC), constant comparison analysis, and classical constant analysis constitute the analyses. The following section provides the details regarding the sampling design, cases, instrumentation, and analytical techniques of this qualitative approach.
Sampling Design
Population and sample. The population of this qualitative approach is the same as that of the quantitative approach. However, this phase relies on the case studies, selecting information-rich cases for in-depth study in order to document the diverse theoretical concepts as well as to identify important common patterns among such diverse cases (Miles & Huberman, 1994; Onwuegbuzie & Leech, 2007).
80


Sampling scheme. Searching for the common patterns of interrelated concepts and maximizing the variation among the cases, this study applies the subgroup sampling design to the qualitative phase (Onwuegbuzie & Leech, 2007). There are three attributes considered in order to select cases or municipalities for interviews—type of functional program, municipal size, and budget behavior. Taking into account type of functional program and municipal size for sampling design, the qualitative component of this study can control for extraneous variables, thereby reducing threats to internal validity (specifically causal validity) (Newman et al., 2003; Singleton & Straits, 2010). Regarding the budget behavior for each functional program, this study calculates the inter-annual budget changes of years 2018-2019 of the education and public works departments and then the Z-score of those changes in order to statistically divide the municipalities into three different subgroups—increased, unchanged, and decreased budgets (Gerring, 2007).
Cases and observations. Within each subgroups of increase, unchanged, and decreased budgets for education and public works spending, three cases at the mean of Z-score are selected in order to include small, medium, and large municipalities (Onwuegbuzie & Leech, 2007). However, note that there are limited numbers of the large-sized municipalities and most of them are not at the mean of Z-score of each subgroup. Consequently, there are only three department heads from the large-sized municipalities for the interviews. Table 4.4 summarizes these case studies for education and public works departments according to their sizes and the inter-annual budget changes.
81


Table 4.4 Interview Sites for the Subgroup Design
Municipal Size Inter-annual Budget Chan ge
Decrease No-Change Increase
Small E,PW E,PW E,PW
Medium E,PW E,PW 2Es,PW
Large PW E,PW -
Note: E and PW refers to the department of education and the department of public works, respectively.
Instrumentation. There are two sources of data collection for this study—available documents and materials, and participant responses from field interviews. The numerical data (i.e., annual budgets for programmatic spending) are secondary data deriving from the quantitative approach. Meanwhile, the more subjective and primary data involving motive, interpretation, and explanation of budgetary information and a budget process which include administrative capacity, information characteristics, information utilization, and perceived budget allocation (i.e., efficiency and effectiveness) are gathered through field interviews of the education and public works department heads (Singleton & Straits, 2010; Teddlie & Tashakkori, 2009).
Interview protocol. In order to allow interviewees to express their own understanding in their own terms and at the same time to test the hypotheses discussed in Chapter 2, this study applies the standardized open-ended interview in which the exact wording and sequence of an interview guide are mostly predetermined, structured, and based on the theoretical framework and its hypotheses but in a completely open-ended format (Singleton & Straits, 2010; Teddlie & Tashakkori, 2009). Table 4.5 lists the interview questions according to the concepts and constructs of interest.
82


Table 4.5 Interview Protocol
Concept Constructs of Interest Questions
1. Structural isomorphism Coercive and normative isomorphism Which budgeting system does your organization adopt? Why is such system adopted? What are managerial structures, standard of procedures, and fiscal information such system requires?
2. Incremental! sm Budget incrementalism How has your organization’s budget process changed over the last 4 years or until the recent budget system has been in place? Currently, how does a budget process in your organization usually proceed? What are the sequence and the key activities in such process? On which activity does the budget committee spend quite a long time and why?
3. Budget rationality Budgetary orientations Are you engaged in budgeting? What are your budgetary responsibilities? What is your role during the budget process? How do you interact with other groups of budget actors? Can you prioritize these following competing spending preferences: financial control, efficiency, management techniques, policy issues, and executive’s plan? Why and how so?
4. Capacity to budget Knowledge and skills What are skills, knowledge, or technical practices necessary for you to initiative, present, and implement budget proposals? Why?
5.Bugetary information Information piece Information characteristics Information provider Which piece of information do you usually think of and bring up during the budget process? Which information do you base your budget decisions on? Which types of information do you find useful (important or adaptable)? Why? Which types of information do you find reliable? Why? What kind of information do you usually receive from the departments? Does your organization have an official unit providing information for policymakers? What kind of information do they provide? How do they get such information?
6. Information utilization Decision making Presentation strategies How does that information have a role during your budget drafting, presentation, debate, and discussion? How do you present your projects to the municipal budget committee? How does the committee react to such presentation style?
7. Budget decisions Perceptional budget allocation What factors carry the most weight in the definition of your organization’s budget decisions? How do such factors impact the spending priorities, eliminating the duplicative programs, or costsaving? If those factors involve information and information use, how was information linked to the differences between the 2017-2018 budgets? In what other ways did each type of information affect allocation at the following steps of the budget cycle: formation of the agency’s budget requests, budget debate, and budget approval?
83


Qualitative Analysis
With a priori causal relationship and hypotheses formulated in Chapter 2, this study depends on a confirmatory nature of data analysis. In order to elaborate those relationships among the key concepts, this qualitative analytical phase begins with the keyword-in-context (KWIC) technique so as to identify the meanings of such keywords as budget process, efficiency, effectiveness, key information, and budget allocation. Next, it forms those meanings into themes through the constant comparison analysis. Then, this study groups and counts such themes by the classical content analysis. The results from these three analytical techniques are triangulated and then supplemented and/or complemented with the quantitative findings in the next chapter.
KWIC. KWIC can reveal the budgeting norms and practices shared among key budget actors via the meaning of this research’s priori theoretical concepts (Leech & Onwuegbuzie, 2008)—for example, knowledge, supervisors, roles, performance, performance information, information sources, relevance, adaptability, importance, reliability, budget process, decisions, spending levels, and also others as the interview results suggest. Comparing texts that appear before and after such keywords, KWIC helps clarify such technical terms and how they can interrelate under the context of Thailand’s local governments (Leech & Onwuegbuzie, 2007b). The KWIC helps to provide meaning of those key concepts as well as identify plausible themes which can clarify the interrelationships among them.
Constant comparison analysis. To identify and understand the interrelationships among those meanings of key concepts identified by KWIC, the constant comparison analysis first conducts an open coding where the study chunks the interview data into smaller but meaningful segments in order to assign labels to summarize in a word or short phrase of the research’s key concepts (i.e., descriptive coding), to record such concepts in the interviewees’ own language
84


(i.e., in Vivo coding), and to connote observable and conceptual sequential practices in the data (i.e., process coding) (Leech & Onwuegbuzie, 2007b; Miles et al., 2014; Strauss & Corbin, 1998). Then, the study applies axial coding, which groups the segments from open coding into similar categories—administrative capacity, the use of information, budget decisions, and others—as antecedent, independent, dependent, and extraneous variables respectively. Finally, the study conducts selective coding, creating themes out of data of each subgroups of programmatic spending by interrelating the segments from the axial coding (i.e., causation coding) and comparing and contrasting with the hypotheses (i.e., hypothesis coding).
Classical content analysis. The researcher records how the frequent themes are used to determine which key concepts are most discussed throughout the interview data (Leech & Onwuegbuzie, 2007, 2008). For instance, the researcher finds the number of department heads who identify themselves with a particular budget orientation, explains how they utilize certain information during the local budget process, and assesses whether such information use affects the budget decisions in terms of both actual allocation and their perceptions on cost-saving, eliminating duplicative programs, and changing spending priorities.
Triangulation
As a final analytical step, this research triangulates the results from quantitative and qualitative approaches mainly by allowing qualitative results to confirm, cross-validate, or corroborate the quantitative findings so as to build a coherent justification for the themes on how administrative capacity, and performance and non-performance information and its characteristics, have an impact on information utilization during a budget process and eventually on budget allocation (Creswell, 2013; Teddlie & Yu, 2007).
85


In conclusion, this study applies the quantitative mixed methods design in order to investigate the relationships between administrative capacity, budget rationality, information utilization, and budget allocations. These concepts were operationalized through the pilot study in order for the study to have the related constructs and their measures fit under the context of a developing country’s local budgeting. After that, the study designs the survey instruments so as to collect the data from all 2,441 municipalities in Thailand. Then, the study calculates the interannual budgets in order for it to serve as a sampling frame from which cases for interviews are chosen. Finally, the study analyzes the quantitative data from the survey instrument and the qualitative data from the interviews. The results from these quantitative and qualitative analyses are discussed in detail in Chapter 5 and Chapter 6, respectively.
86


CHAPTER V
QUANTITATIVE RESULTS
Applying the quantitative mixed methods design, this chapter presents the quantitative results especially through how they answer the research questions and (dis)confirm the hypotheses. The quantitative results here are from the ordinal and the multiple regression analyses. The ordinal regression analyses aim at answering the first research question: how do policymakers with different administrative capacities utilize relevant and reliable information during a budget process? The ordinal regression analyses also explore the first component of budget allocation—the perceptional budget allocation or the efficiency and effectiveness of the municipal spending—which responds to the second research question: how does the information utilization influence budget decisions? Additionally, the multiple regression analysis strives to complement the perceptional budget allocation by investigating the actual budget allocation. For each research question, this study investigates two main local public services for which the education and public works departments are responsible. For the research question about the information utilization, this study disaggregates the analyses by groups that use performance information and those that use non-performance. Prior to proceeding to the regression analyses and answering the research questions, this chapter first provides the overall descriptive statistics of the study variables that operationalize the theoretical concepts of this study.
The Overall Descriptive Statistics of the Survey Participants and the Study Variables Demographic data
This study collects the data regarding administrative capacity, budget rationality, the use of information and budget allocation through the survey instrument. This survey is conducted during October 2018, immediately after the municipal budget preparation and approval for the
87


start of the 2019 fiscal year, running from October 1, 2018 to September 30, 2019. This study maximized the response rate by sending out the questionnaires to all 2,441 municipalities in Thailand. The response rate is 28.23 percent on average, and 27.82 and 28.64 percent for the departments of education and public works, respectively. Additionally, most are by far from the sub-district municipalities, representing the agricultural and urban/rural fringe neighborhoods with the population density of 920 residents per squared kilometer. Table 5.1 details the demographic data of the municipalities participating in the survey.


Full Text

PAGE 1

THE USE OF INFORMATION AND BUDGET DECISION: HOW ADMINISTRATIVE CAPACITY AND INFORM A TION PLAY A ROLE IN LOCAL BUDGETING IN THAILAND By Pawinee Chuayprakong B.A., Chulalongkorn University, 2005 MPP, Graduate Insti tute for Policy Studies, 2008 MPA, Chu lalongkorn University, 2009 A thesis submitted to the Faculty of the Graduate School of the University of Colorado Denver in partial fulfillment Of the requirements for the degree of Doctor of Philosophy Public Administration 2019 i

PAGE 2

PAWINEE CHUAYPRAKONG ALL RIGHTS RESERVED ii

PAGE 3

This thes is for the Doctor of Philosophy de gree by Pawinee Chuayprakong has been approved for the Public Affairs Program by Christine R. Martell, Chair Mary E. Guy Todd Ely Charas Suwanmala Date: August 3, 2019 iii

PAGE 4

Chuayprakong, Paw inee (PhD, Public Affairs ) The Use of Information and Budget Decision: How Administrative Capacity and Information Play a Role in Local Budgeting in Thailand Dissertation directed by Professor Christine R. Martell ABSTRACT The current budget reform aims to move a budget process, in which incrementalism prevails, towards a rational decision making. This study investigates such reform in T hailand’s local governments with mixed methods research design. The results show that , despite budgetary incrementalism, administrative capacity —skills in participatory budgeting, communication, financial management and accounting, and strategic planning —a nd spending preferences among local frontline public administrators lead local budgeting to an informationbased decision making process. During this information based decision making process, the levels of information utilization, the strategic of informa tion use (i.e., the storytelling styles), and the professional judgment among frontline public administrators can drive the municipalities’ budgetary decisions to be more efficient and effective. Th ese findings confirms that administrative capacity and information are indispensable for a rational decision making process as well as for efficient and effective budgetary decisions. As a result, these findings can be used to promote the use of information for the developing countries. The form and content of this abstract are approved. I recommend its publication. Approved: Christine R. Martell iv

PAGE 5

This dissertation is dedicated to my parents —Kanueng Chuayprakong and Nongchawee Chuayprakong—who are uneducated but always give their all to my education. Every day, your unconditional love and support inspire me to be a better person and convince me that I can make a difference. v

PAGE 6

ACKNOWLEDGEMENTS I would like to express the deepest gratitude and appreciation to my dissertation advisor, Prof . Christine R. Ma rtell. Her academic advice, life guidance, and in credible patience are absolutely an integral part of my graduation. Her faith in me always helped make me get out of every single struggle alive throughout my eight years in this PhD program. Next , I highly appreciate al l help and suggestions from my dissertation committee members. Prof. Mary E. Guy helped and reminded me look at a big picture of the field of public administration and public management. She made sure that I sensibly applied i t to my dissertation’s overarching structure. Meanwhile, Prof. Todd Ely gave several methodological suggestions to improve my analytical components. Also, I would like to thank Prof. Charas Suwanmala who offered insightful comments regarding Thailand’s dec entralization policy and local public finance. Wisdom and kindness from my dissertation committee will definitely shape my scholarship as well as my career in academia. Finally, I would like to thank my sister who set the high bar since the first grade until the PhD level. I followed her school path and could not succeed it without her support. Also, I highly appreciate the everlasting help and guidance from Prof. Weerask Kreuathep who talked me in to this amazing PhD journey and introduced me to a life of academic adventure. Moreover, I would like to thank my friend in Denver —Plaa, Scott, and Pi New —for their food and comfort especially during the lonely, cold days. Also , I appreciate the moral support from my friends in Bang kok—Amm, Bee, Pi Den, Orn, Nhua, Pi Phue ng, Tong, Pailin, Pi Wasin, Whan, June, and Anne who were there when I occasionally wanted to give up. Your listening and beverage matter! vi

PAGE 7

TABLE OF CONTENTS I . INTRODUCTION ...................................................................................................................... 1 Budget Decision, a Product of Rationality? ................................................................................ 2 Politics administration dichotomy in budgeting ...................................................................... 2 Progress on Information Utilization and Its Impact on Budget Decision Making ...................... 3 Information utilization in a budget process ............................................................................. 3 Information utilization and budget decisions .......................................................................... 4 Performance information ......................................................................................................... 5 Developed countries and administrative capacity ................................................................... 5 Thailand, the least likely case ................................................................................................. 6 Research Questions ..................................................................................................................... 7 Methodology ............................................................................................................................ 8 Contributions to Theory and Practice ...................................................................................... 8 II. LITERATURE REVIEW ......................................................................................................... 11 Theories of Information Economics .......................................................................................... 12 Information Asymmetry ........................................................................................................ 12 PrincipalAgent Theory ......................................................................................................... 15 Empirical Studies on Information Asymmetry, Information Use, and DecisionMakings ... 21 Budget Rationality Model ......................................................................................................... 26 Administrative Capacity and Information Utilization ........................................................... 27 Budget Orientations and Information Utilization .................................................................. 29 Findings from the Theoretical Review ...................................................................................... 32 Research Questions and Hypotheses ......................................................................................... 34 Rival Hypotheses ....................................................................................................................... 35 Theoretical Framework ............................................................................................................. 37 III. CASE JUSTIFICATION AND THAILAND’S LOCAL PUBLIC FINANCE AND BUDGETING ............................................................................................................................... 39 The Case Justification for Thailand’s Local Governments ....................................................... 39 Representativeness ................................................................................................................. 39 Causal leverag e ...................................................................................................................... 40 Thailand as a least likely case ................................................................................................ 42 Thailand’s Local Public Finance and Budgeting ...................................................................... 45 vii

PAGE 8

Thailand’s Decentralization and Local Governments ........................................................... 45 Local Governments’ Revenue and Expenditure Structure .................................................... 48 Local Budget Process ............................................................................................................ 51 Information Use in Local Budgeting in Thailand .................................................................. 56 IV . RESEARCH DESIGN AND METHODOLOGY .................................................................. 58 The quantitative mixed methods design .................................................................................... 59 The Quantitative Approach ....................................................................................................... 60 Sampling Design .................................................................................................................... 60 Operationalizing C oncepts of Interest ....................................................................................... 65 Pilot Study ............................................................................................................................. 65 Quantitative Data Analysis .................................................................................................... 78 The Qualitative Approach ......................................................................................................... 80 Sampling Design .................................................................................................................... 80 Qualitative Analys is .................................................................................................................. 84 Triangulation ............................................................................................................................. 85 V . QUANTITATIVE RESULTS .................................................................................................. 87 The Overal l Descriptive Statistics of the Survey Participants and the Study Variables ........... 87 Demographic data .................................................................................................................. 87 The Study Variables .............................................................................................................. 90 Information, Administrative Capacity, Budget Rationality, and Information Utilization ... 106 Administrative Capacity, Information Utilization, and Bu dget Decisions .......................... 120 Actual Budget Allocation .................................................................................................... 135 VI . THE QUALITATIVE RESULTS ........................................................................................ 152 The Determinants of Information Utilization .......................................................................... 152 Information Utilization ........................................................................................................ 152 Factors Affecting Information Utilization ........................................................................... 153 The Determinants of Budget Allocation .............................................................................. 161 VII . DISCUSSION ON INFORMATION UTILIZATION ...................................................... 167 Overall and FunctionBased Information Utilization .............................................................. 167 Overall information utilization ............................................................................................ 167 Information Utilization by the Education Department Heads ............................................. 168 Information Utilization by the Public Works Department Heads ....................................... 169 viii

PAGE 9

Discussion and Findings .......................................................................................................... 170 Administrative Capacity, Relevant Information, and Information Utilization .................... 170 Organization Cultures and Information Utilization ............................................................. 177 Theoretical Contributions ........................................................................................................ 182 Empirical Contributions to the Current Literature and Empirical Studies .............................. 184 Practical Contributions ............................................................................................................ 187 VIII. DISCUSSION ON BUDGET ALLOCATION ................................................................. 189 Actual and Perceptional Budget Allocation ............................................................................ 189 Actual Budget Allocation .................................................................................................... 189 Percepti onal Budget Allocation ........................................................................................... 190 Discussion and Findings .......................................................................................................... 191 Budgetary Incrementalism and Municipal Finance ............................................................. 191 How Management Matters in Local Budgeting .................................................................. 193 Intergovernmental Relations: The Indomitable Centralization ........................................... 198 Theoretical Contributions ........................................................................................................ 199 Contributions to the Current Literature ................................................................................... 199 Practical Contributions ............................................................................................................ 201 IX. CONCLUSION ..................................................................................................................... 203 Research Questions ................................................................................................................. 203 Theoretical Relevancy ............................................................................................................. 204 Policy Imp lications .................................................................................................................. 204 Threats to Legitimation and Future Research ......................................................................... 205 REFERENCES ........................................................................................................................... 209 APPENDIX A . The Ordinal Regression Model for Information Utilization ............................................... 226 B . The Regression Model for Perceptional Budget Allocation ............................................... 230 C . The Regression Model for Impact of Administrative Capacity and Information Utilization on Actual Budget Allocation ...................................................................................................... 233 ix

PAGE 10

LIST OF TABLES Table 2.1 Relevant Theories to Understand the Interrelationship between Capacity, Information, and Budget Decisions ................................................................................................................... 12 Table 2.2 Budget Orientations, Relevant and Important Information Considered, and Norms ... 31 Table 3.1 Types of Local Governments in Thailand and Their Differences 46 Table 3.2 Source of Local Governments’ Revenue (Unit: Million Baht) ..................................... 49 Table 3.3 Local Governments’ Expenditure in 2012 (Unit: Million Baht) .................................. 51 Table 4.1 The Stratified Sampling and Sample Size for the Survey ............................................ 62 Table 4.2 Sampling Design ........................................................................................................... 65 Table 4.3 Interview Protocol Based on the Literature .................................................................. 68 Table 4.4 Interview Sites for the Subgroup Design ...................................................................... 82 Table 4.5 Interview Protocol ......................................................................................................... 83 Table 5.1 Demographic Data of the Participating Municipalities ................................................ 89 Table 5.2 Demographic Data of Profile of Participants in the Survey the Survey Respondents .. 90 Table 5.3 Frequency for Information Pieces Used During the Budget Process ........................... 92 Table 5.4 Descriptive Statistics for Reliability of Each Information Piece .................................. 94 Table 5.5 Descriptive Statistics for Relevance of Each Information Piece .................................. 95 Table 5.6 Frequency and Descriptive Statistics for Administrative Capacity .............................. 97 Table 5.7 Frequency of Budget Rationality and Spending Preferenc e ......................................... 98 Table 5.8 Frequency and Descriptive Statistics for Information Utilization for Budget Discussion ..................................................................................................................................................... 100 Table 5.9 Frequency for Strategies of Information Utilization the Department Heads Always Apply........................................................................................................................................... 101 Table 5.10 Frequency and Descriptive Statistics for Perceptional Budget Allocation ............... 103 Table 5.11 ANOVA for a Comparison of Spending Efficiency and Effectiveness according to Municipality Size ........................................................................................................................ 104 Table 5.12 ANOVA for a Comparison of Actual Budget Allocation (BTH1,000 per capita) for FY2019 ....................................................................................................................................... 105 Table 5.13 Frequency and Descriptive Statistics for Information Use by Department Heads ... 107 Table 5.14 Regression Results of Information Utilization for Budget Discussion among all Department Heads ....................................................................................................................... 108 Table 5.15 Frequency and Descriptive Statistics for Information Use by the Education Department Heads ....................................................................................................................... 111 Table 5.16 Regression Results of Information Utilization for Budget Discussion among all Education Department Heads ..................................................................................................... 112 Table 5.17 Frequency and Descriptive Statistics for Information Use by Public Works Department Heads ....................................................................................................................... 115 Table 5.18 Regression Results of Information Utilization for Budget Discussion among Public Works Department Heads ........................................................................................................... 117 Table 5.19 The Variables Predicting the Perceptional Budget Allocation for the Education and Public Works Departments ......................................................................................................... 121 Table 5.20 Spending Efficiency and Effectiveness .................................................................... 123 Table 5.21 The Variables Predicting the Perceptional Budget Allocation in the Education Department .................................................................................................................................. 127 Table 5.22 Spending Efficiency and Effectiveness in the Education Departments .................... 128 x

PAGE 11

Table 5.23 The Variables Forecasting the Perceptional Budget Allocation in the Public Works Departments ................................................................................................................................ 132 Table 5.24 Spending Efficiency and Effectiveness in the Public Works Departments .............. 133 Table 5.25 The Frequency and Descriptive Statistics of Variables Forecasting the Total Education Expenditure ................................................................................................................ 137 Table 5.26 The Variables Forecasting the Education Expenditure (Decrease) .......................... 138 Table 5.27 The Frequency and Descriptive Statistics of Variables Predicting the Education Expenditure ................................................................................................................................. 140 Table 5.28 The Regression Results for the Education Expenditure (Increase) .......................... 142 Table 5.29 The Frequency and Descriptive Statistics of Variables Predicting the Public Works Expenditure ................................................................................................................................. 144 Table 5.30 The Regression Results for the Public Works Expenditure (Decrease) ................... 145 Table 5.31 The Frequency and Descriptive Statistics of Variables Predicting the Public Works Expenditure ................................................................................................................................. 147 Table 5.32 The Regression Results for the Public Works Expenditure (Increase) ..................... 148 Table 5.33 Summary of Regression Results for the Information Utilization ............................. 150 Table 5.34 Summary of Regression Results for the Spending Efficiency and Effectiveness .... 151 Table 5.35 Summary of Regression Results for the Spending Efficiency and Effectiveness .... 151 Table 6.1 Summary of Determinants of Information Utilization ............................................... 161 Table 6.2 Determinants of Budget Alloc ation ............................................................................ 165 xi

PAGE 12

LIST OF FIGURES Figure 2.1 The causal graph of theories of information economics under the context of information utilization and budget decisions ................................................................................ 21 Figure 2.2 Budget Rationality and Information Utiliz ation .......................................................... 27 Figure 2.3 Theoretical Framework of Administrative Capacity, the Use of Information, and Budget Decisions .......................................................................................................................... 37 xii

PAGE 13

CHAPTER I INTRODUCTION Information can resolve uncertainty and reduce ambiguity, thereby assisting in decisionmaking (Faibisoff & Ely, 1974; Feldman, 1989) . Information always has economic value and anyone who possesses it has advantage over others who do not (Arrow, 1962) . Lack of information or false information can lead to both market and government failures (Akerlof, 1970; Niskanen, 1971; Salamon, 1987; Seldon, 2002; Tullock, 2002) . Hence, information is a key element in an administrative decision making process as i t is used in formulating goals, searching for alternatives, and calculating consequences of each alternative (Downs, 1957; Lindblom, 1959; Simon, 1957) . Nonetheless, how information is used for administrative accomplishment is not well understood in administrative science (Simon, 1946; Van De Walle & Van Dooren, 2008) . Its effects during budget process and on budget allocation are also still inconclusive in research (Joyce, 2008; Kroll, 2015; Moynihan, 2015; Pollitt, 2006) evidenced by the fact that continues to be the debate on whether budgeting is a product of incremental politics or a product of ra tionality. This study considers budgeting a product of rationality and attempts to disentangle how related information has an impact on the information utilization during a budget process and on budget decision making. As an introduction, this chapter fir st shows that governments gear their budget reforms toward rationality despite the fact that budget process is also of incremental politics. After that, it presents the evolution of information utilization during a budget process and how such utiliz ation e ventually influences budget decision making. Since such overview leaves some inadequately explored areas, this chapter then sets the research question 1

PAGE 14

accordingly. This research question directs this study to relevant literature and appropriate methodology discussed in the following chapters. Budget Decision, a Product of Rationality? Politics administration dichotomy in budgeting Somewhat like public administration, the study of public budgeting has revealed an ongoing struggle between politics (i.e., soc ial, political, and legal aspects) and administration or management (i.e., economic and technical aspects) (Goodnow, 1900; Rubin, 2015; Thurmaier, 2001; Wilson, 1887) . On one hand, budget decisions are a product of bureaucratic inertia and budget maximizing behaviors (Niskanen, 1971; Wildavsky, 1978) , negotiation between politicians and pressure groups or public service clientele (Smith, 2003) , and also logrolling and porkbarrel legislation (Buchanan & Tullock, 1962) . On the other hand, they are the product of a rational decisionmaking process which reflects the comprehensive cost/benefit analysis (Key, 1940) , the organization’s planned objectives (Schick, 1990) , and fiscally adaptive strategies (Levine, 1978) . Nonetheless, from an administrative standpoint, the latter approach is normatively preferred. As a result, budget reforms always gear a budget process toward managerial improvement, program effectiveness, and progressive roles of budget participants (Caiden, 2010; Schick, 1966) . Budget reforms require budget participants to possess both capacity to budget and adequate information in order for them to instrumentally utilize information —to conduct budget analysis and then make better budget decisions (Weiss, 1977) . Even though there is still a long way to go for the instrumental use of inf ormation, the literature demonstrates that it is making a progress (Pandey, 2015) . The following section discusses this progress. 2

PAGE 15

Progress on Information Utilization and Its Impact on Budget Decision Making I nformation utilization in a budget process No matter how or whether it is used, organizations always gather, store, and communicate information, especially the information pieces that selectively signal the structures of their environments and their functioning in relation to the environment (Feldman & March, 1981; Katz & Kahn, 1966) . In the public sector, this informational behavior has been promoted after the advent of information technology and the New Public Management (N PM) which aims to enhance performance through utilizing such information as public services’ output, outcome, and results for monitoring and evaluation purposes (Dahler Larsen, 2007; Hood, 1991, 2007; Lynn, 2006) . Though related information can improve performance via changes in resource allocation that can expand, modify, or terminate public programs according to their performances (Pollitt, 2006; Julnes, 2008) , the use of information has actually been inadequately applied into the realm of budgeting. A budget process is highly complex and theoretically unsettled because it is not exclusively social, political, legal, economic, or technical in its rationality (Key, 1940; Thurmaier, 2001) . On one hand, a budgeting process is, and should be, more rational comprehensive, relying on economic and technical rationality in order to reach informed and thoughtful decisions (Lindblom, 1959; Thurmaier, 2001) . They can be influenced by a statement of objectives and by weighing costs and benefits of relevant alternatives (Schick, 1966) . On the other hand, a budgeting process is of incremental politics. Except for very special circumstances, decision makers usually rely on simple calculations or heuristic rules and eventually only find satisfactory solutions for policy problems (Davis, Dempster, & Wildavsky, 1966; True, Jones, & Baumgartner, 2007) . Not only does incremental politics not require complex calculation, but also 3

PAGE 16

it limits negotiations and conflicts during the budget process , allowing for political acceptance (Nielsen & Baekgaard, 2015) . Hence, incremental politics of budget decision process reigns supreme virtually everywhere (Wildavsky, 1978) . This incremental pol itics also significantly contributes to the limited use of information to improve a budgetary process and appropriations, despite that fact that the information technology and NPM movements promote the information gathering and utilization (Joyce, 2011; Melitski & Manoharan, 2014; Moynihan, 2006) . Accordingly, many scholars believe that to use information in the budget process and for budget decisions is truthfully the next great challenge for governments (Ingraham, Joyce, & Donahue, 2003; Laegreid, Roness, & Rubecksen, 2008) . Information utilization and budget decisions There are four example s for the positive associations between information utilization and budget decisions. First, at the program level and for the small and medium size programs, thinking more logically and strategically about how resources can be used and about what results might be accomplished, the department managers can influence the policymakers to restructure activity plannin g and reallocate resources (Gilmou r & Lewis, 2006a; Ho, 2011; Joyce, 2008; Melkers & Willoughby, 2005) . Second, when no clear political signals emerge, budget actors make decisions based on technical and economic rationalities, falling back on analytical information (Goodman & Clynch, 2004; Thurmaier & Willoughby, 2001) . Third, the time and ef forts public mangers invest in shaping information interpretation as well as their communication skills can persuade the policymakers to make changes in budget allocation (Moynihan, 2015) . Fourth, organizations that develop performance reporting practices as a part of budget transparency (i.e., the use of information and communication technology (ICT) based system for reporting performance and the reports of how current outputs and outcomes are to be 4

PAGE 17

used for the future resource allocation, to be rewarded for good results, and to be sanctioned for poor results) trigger the policymakers to make actual budget changes accordingly (Laegreid et al., 2008) . The studies attempting to explore the linkage between information and budget decisions are not only limited but also offer inconclusive results (Ho, 2011) . Several studies find that information such as public services’ outputs and outcomes exists in a budget decision process (Goodman & Clynch, 2004; Melitski & Manoharan, 2014) , encourage learning and deliberation among different departments (Moynihan, 2008a, 2008b) , but is not taken into serious consideration in terms of budget allocations. However, other studies confirm the linkage between the use of information and the government’s spending levels (Ho, 2011; Moynihan, 2015; Zaltsman, 2009) . Performance information The related studies are based heavily on the use of performance information . Outputs and outcomes of public service programs are at the center stage of the linkage between the information use and budget allocations (Laegreid et al., 2008; Melkers & Willoughby, 2005; Moynihan, 2008a, 2015; Pollitt, 2006) . How other types of relevant information such as local knowledge (Petts, 1997; Johnson, 2009), political cues (Goodman & Clynch, 2004; Thurmaier & Willoughby, 2001) , and even fiscal reports and guidance (Alijarde, 1997; Lee, 1991, 1992) that are used (alone or together with performance information) have not been adequately explored. Developed countries and administrative capacity In addition, the current literature focuses mainly on the developed countries’ contexts — especially the U.S. (Ho, 2011; Lee, 1991; Moynihan, 2008a, 2015; Thurmaier & Willoughby, 2001) and the European countries (Laegreid et al., 2008; Bouckaert & Halligan, 2008; de Kool, 5

PAGE 18

2008) despite the fact that large differences of how inf ormation is used exist between countries in patterns of its use (Pollitt, 2005, 2006) . Performance information, for instance, has been used prevalently among large cities and countries which undoubtedly have more administrative capacity so that they can successfully handle such tasks as strategic planning, performance analysis, cost analysis, program budget analysis, and communication with the publics (Ho, 2011; Van Dooren, 2008) . Nonetheless, the availability of information does not guarantee its use and in fact administrative capacity of policymakers is a key to the use of information as information is always selected, perceived, and presented by people (Moynihan, 2008a; Stone, 1997) . The administrative capacity, both qualification and attitudes, are key factors to determine information utilization. It cognitively shapes several activities during a budget process such as to find data sources, to conduct field visits, to relate data gathered to budget problems, to read political cues that condition fiscal alternatives’ means and ends, to apply information technology to integrating fiscal information into budgetary deliberation, and to develop and testify budgetary recommendation at executive and legislat ive hearings (Coe, 2003; Goodman & Clynch, 2004; Thurmaier, 1992; Willoughby & Finn, 1994) . Nonetheless, even though many studies concentrate on the rel ationship between administrative capacity and whether and how the information is used, these studies do not further link their results to budget allocations (Coe, 2003; Lee , 1991, 1992) . Thailand, the least likely case For the case of developing countries, stagnant economics and inefficient central bureaucracies have led them to public sector reforms. Among these reforms has been decentralization (Kettl, 2005; Schneide r, 2003) , which has occurred to various degrees across countries. Note that a mong the East Asian countrie s, Thailand has demonstrated slow progress 6

PAGE 19

on fiscal decentralization. For instance, its subnational expenditure as a share of total public spending is approximately 10 to the aver age of 37 percent, the lowest of East Asian countries (White & Smoke, 2005) . This limited local expenditure results in the lack of budget officials’ administrative capacity (Kittipanyasiri, 2015; Krueathep, 2014; Nagai, Funatsu, & Kagoya, 2008) as well as the lack of related information for decision ma king (Wongpredee & Sudhipongpracha, 2014) . Hence, Thailand’s local governments can serve as the least likely case (George & Bennet, 2004; Gerring, 2007) to unfold the linkage between administrative capacity, in formation utilization , and budget decision making. Research Question s It can be said that the current literature has put forward the comprehensive rationality of using information in a budget process, suggesting the existence of the linkage between the information utilization and budget decisions. However, there are still ce rtain areas that have been inadequately explored. Those areas include the utilization of information during a budget process, the inclusion of administrative capacity as the key determinant of information utilization, the impact of information other than p erformance information on budget decisions, and the relationship between information utilization and budget decisions in order to confirm and possibly strengthen the argument that a budget process is of comprehensive rationality through its information utilization. Consequently, this study asks: How do policymakers with different administrative capacities utilize information during a budget process in order to have an impact on budget decisions? More specifical ly, it first asks: How do policymakers with dif ferent administrative capacities utilize relevant and reliable information during a budget process? Then it asks: How does the information utilization influence budget decisions? 7

PAGE 20

Methodology To understand this complex and multifaceted budgetary ph enomenon, this study conducts mix ed methods research based on quantitative and qualitative data from a survey and interview s respectively (Creswell, 2013) . The data will be analyzed based on the the ories of information economics —information asymmetry (Akerlof, 1970) , the principal agent theory (Eisenhardt, 1989; Moe, 1984) , and the budget rationality model (Thurmaier, 2001; Thurmaier & Willoughby, 2001) against their competing theories —structural isomorphism (Meyer & Rowan, 1977) and incremental politics (Davis et al., 1966; Wildavsky, 1964) . Contributions to T heory and P ractice Theoretically, since this study attempts to (dis)confirm the relationships among the concepts of administrative capacity, the use of information, and budget decision making in Thailand, it serve as a theory testing case study which assesses the validity of competing theories (George & Bennet, 2004) . Particularly, given the context of a developing country, the results from this study can increase the external validity or generalizability of the aforementioned theories to settings other than the developed countries ( Onwuegbuzie & Leech, 2006) . Additional ly, information other than performance measures are empirically examined so that this study can increase the construct validity of all types of related information to the budget decision making (Van de Ven, 2007) . For policy implications, the results of this study can (dis)confirm the extent to which the current direction of the Thai budget reform which focuses mostly on whether the competency development and the provision of information significantly impact the information utilization during a budget process and the budgetary decision making. For example, the issues regarding which group of policymakers (i.e., specialized in education or public works ) should be targeted 8

PAGE 21

for the more comprehensive use of performance information and why some information are gathered but not being utilized. This research inte nds to inform policymakers of budget reforms which are of present day problems (Gerring, 2012) . After this introduction , Chapter 2 reviews the theories and the current studies pertaining to information utilization and budget decisions. Based on these theories and studies, it offers a theoretical framework and sets up the hypotheses responding to the research questions. Next, Cha pter 3 provides the reasons why this study focuses on Thailand’s local governments. It also describes the country’s progress on fiscal decentralization and how the local budget actors prepare and approve their budgets. Then, Chapter 4 elaborates the research design and methodology. Particularly, this study applies the mixed methods design which emphasizes the quantitative component and complements the quantitative results with the qualitative data. Later in Chapter 5, the quantitative a nd qualitative data and analyses are provided in order to answer the research question of how policymakers with different administrative capacities utilize relevant and reliable information during a budget process . Note that the se analyses concentrate on all survey participants as well as on their groups according to the policy area s they serve (i.e., educational and public works services). Meanwhile, Chapter 6 provides the answers to the research question of how t he information utilization influence budget decisions. Similar to Chapter 5, Chapter 6 also offers the quantitative a nd qualitative analyses of the overall survey participants and those based on policy area s (i.e., educational and public works services) . Later in Chapter 7, the study discusses the results from Chapter 5 underlining information utilization. It shows how administrative capacity, information types, and budget orientations have an impact on information utilization during a local budgeting proc ess. Next, Chapter 8 discusses the results from Chapter 6, explaining the influence of the information use can balance the budgetary 9

PAGE 22

incrementalism in municipal budgets. Then in the final chapter, Chapter 9 summarizes the study. It provides the overall fin dings and the policy implications. The study’s limitations which lead to future research are also provided. 10

PAGE 23

CHAPTER II LITERATURE REVIEW As this study focuses on how rationality plays a role through information utilization and administrative capacity, it comprehensively reviews the theories of information economics (information asymmetry and principal agent theory) and the administrative characteristics especially in terms of budget rationality model which concerns information, informational behaviors, and decision making. Since the budget reforms gear a budget process toward managerial improvement, program effectiveness, and progressive roles of budget participants (Caiden, 2010; Schick, 1966) , this study is based on the rational choice theory and consequently assumes the instrumentalist’s view — budget actors search and use information in order to make budget decisions that benefit their organizations the most (Hindmoor, 2006; Weiss, 1989) . Therefore, this chapter delineates the relevant theories by explain ing how individuals approach information, utilize it during a budget process, and eventual ly let it impact their budget decisions. Table 2.1 below briefly explains how these theories relate to information, information behaviors, and budgetary behaviors. 11

PAGE 24

Table 2.1 Relevant Theories to Understand the Interrelationship between Capacity, Information, and Budget Decisions Theory Approach to Information Informational Behaviors Budgetary Behaviors Theories of information economics Information as a key to solve informati on problems (adverse selection and moral hazard) How public managers and organization leaders gather information and utilize it during the budget process How relevant and reliable information determines its utilization and leads to budget decisions Budget rationality model Information types and information use as the representation of budget norms and spending preferences How public managers choose and use only particular types of information during the budget process How budget norms and spending preferences influence the information utilization and determine budget decisions Theories of Information Economics In most cases, there is always limited information for any goods or services so that additional credible information can resolve the proble ms from such limitation. In market transactions, there are two parties—buyers and sellers. According to the assumption of rationality (both comprehensive and limited or bounded), both parties attempt to gain and use related information to make their best d ecisions against a set of alternatives in order to receive the highest level of utility or satisfaction (Hindmoor, 2006; Lindblom, 1959; Simon, 1957) . Unfortunately, information is not distributed evenly between buyers and sellers. This results in the situation where sell ers can take advantage of possessing complete information and disseminate deceiving information to their buyers. This informational gap leads to two relevant theories —information asymmetry and principal agent theory. Information Asymmetry Generally, there exist two parties in a transaction who have different levels of information and different values regarding such transaction (Moe, 1984) . In markets it is possible for buyers to canvas various sellers by searching and evaluating market statistics (Stigler, 19 61) . 12

PAGE 25

However, in reality, buyers can only rely on the market statistic and sellers remain possessing the informational advantages —they know the true values of their goods and services and can disseminate deceptive information to the buyers (Spence, 1973) . Here exists the information asymmetry . Even a small amount of this informational gap has a significant effect on a competitive market, resulting in the reduction in the average quality of goods and also in the size of such market (Akerlof, 1970a; Rothschild & Stiglitz, 1976) . The following paragraphs describe thi s informational gap in terms of the information problems it causes as well as possible resolving mechanisms to relieve such problems. The first information problem relates to the public good characteristic of information itself. Even though sellers can s ell information to buyers, buyers (or the purchasers of the information) can share or resell such piece of information to others without diminishing its usefulness to themselves (Leland & Pyle, 1977) . On this ground, sellers are not willing to sell their information, resulting in the under provision of it. However, there are economies of scale in the process of gathering and selling the economic information. Due to the economies of scale, third parties can profit from producing and disseminating information. Specifically, a financial intermediary, generally a private firm, collects the d ata regarding the risks of poor quality goods and services, sorts the risk level, and sells such specialized and con fidential information to buyers ( Campbell & Kracaw, 1980) . A financial interme diary can carry out these activities in a specialized market that is small enough to control the transferability of information and, meanwhile, in an amount of scale that is large enough to cover fixed costs (Leland & Pyle, 1977; Spence, 1973) . The second problem is called adverse selection which involves the situation where sellers may claim that their product or service is of high quality while buyers cannot completely verify 13

PAGE 26

such quality either at the time o f purchasing or while sellers are delivering such product or service (Eisenhardt, 1989) . This problem intensifies when many sellers have incentive to exaggerate quality of their merchandise due to the fact that the returns for good quality accrue to the whole group whose stati stic is affected rather than to the individual sellers (Akerlof, 1970; Hirshleifer, 1973) . With limited and unauthentic information, buyers are likely to choose poor quality merc handise at the higher price than its true quality warrants . Since buyers cannot know the quality of goods and services due to information problems, they rely on signaling —to gather the information and check other characteristics of the good or services in order to predict the quality prior to market transactions (Riley, 1975; Spence, 1973) . Si gnaling includes searching for sellers’ reputations and qualifications. A sellers’ reputation or credibility is considered a valuable good which signals their expected performance in terms of the quality of goods and services they sell (Alchian & Demsetz, 1972) . When signals exist, selle rs, especially ones with a sufficient stake in the market, have no incentive to misrepresent their information since their initial endowment acts as a barrier to exiting the market ( Campbell & Kracaw, 1980; Heinkel, 1982) . Additionally, their qualifications (i.e., profiles and capability) can, to some extent, signal their expected performance. For example, buyers can use the sellers’ investment in capacity buil ding as a proxy for product quality (Spence, 1973) . The third information problem is moral hazard, which arises from unobservability of the sellers’ actions after buyers purchase goods or services from sellers (Wood, 2010) . Since sellers’ actions cannot be observed and hence contracted upon, they have incentive to behave dysfunctionally, not doing their best for the interest of the buyers (Hlmstrom, 1979; Williamson, 1971) . A subset of the moral hazard problem is the principal agent problem. Particularly in the context of public organizations, it is c alle d the principal agent problems. 14

PAGE 27

These problems exist in a relationship between two utility maximizing parties when one (designated as an agent) acts for, on behalf of, or as a representative of the other (designated as a principal) in particular decision problems where goals or values of both parties conflict and when it is costly for a principal to closely observe his agent’s behaviors (Ross, 1973) . A natural remedy to this problem is to invest in monitoring and evaluation of action and use this information in the contract design (Hlmstrom, 1979) . Regarding resolving mechanisms to these three problems, additional relevant and reliable information, and use of it, can alleviate the information problems above. Specifically, information production or the systems that provide additional information to the market can alleviate information problems . This generation and provision of economic information, which renders the true value as well as the riskiness of the assets to be known to the market , can minimize or resolve the problems from aforementioned information asymmetries (Blazenko, 1987; Heinkel, 1982; Riley, 1975; Spence, 1973) . Principal Agent Theory Principalagent theory explains and attempts to solve the problems root ed in an agency relationship —a relationship between two utility maximizi ng parties when one (designated as an agent) acts for, on behalf of, or as a representative of the other (designated as a principal) in particular decision problems where goals or values of both parties conflict (Ross, 1973) . Additionally, it is so difficult or expensive for the principal to verify the agent’s actions (Eisen hardt, 1989; Moe, 1984) . The principal agent theory has seven main components as follows. First, the key parties for the principal agent theory in the public organizations or bureaucracies are bureaucratic superiors and their subordinates , as principals and agents 15

PAGE 28

respectively. This relationship is “therefore structured by a principal agent chain, for instance, from citizens to politician to bureaucratic superior to bureaucratic subordinate and on down the hierarchy of government to the lowest level bureaucrats who actually deliver services directly to citizens” (Moe, 1984, p. 775776) . Second, at the individual level, there exists the preference asymmetry. According to the rational choice theory, both parties are rat ional, utility maximizing individuals but with asymmetry in preferences—they have different utility functions or, in other words, different interests, values, or goals (Hindmoor, 2006; Miller, 2005; Moe, 1984) . Policymakers concentrate on politics so that they would like to satisfy their citizens or constituencies in order to solve public problems, make public policies, and be reelected or to stay on in their political positions (Kingdon, 2003; Moe, 1984; Wood, 2010) . Meanwhile, bureaucrats are budget maxim izers— they strive to make budget (organization’s resources including salary, perquisites, power, prestige, patronage, public reputation, and agency output) as large as possible as a way to seek autonomy as well as to shirk work (Frederickson, Smith, Larimer, & Licari, 2012; Niskanen, 1971) . Third, at the organizational level, there is an information asymmetry —the situation where the relevant and reliable information is distributed asymmetrically throughout a bureaucracy. With technical knowledge and specialization, bureaucrats have and enjoy close to monopolistic informational power over their executives because they are the only party that possess information on both production costs and production process (Frederickson et al., 2012; Niskanen, 1971) . As self interested budget maximizers, bureaucrats are motivated to satisfy their policymakers by handpicking and highlighting information that reflects favorably upon themselves and repressing or even suppressing informat ion that does not ( Downs, 1967; Tullock, 16

PAGE 29

1965) . This monopolistic informational power leaves the executives informationally disadvantaged or lacking accurate and/or complete information. Policymakers can acquire only limited or even distorted information regarding production costs, production process, and t he bureaucrats’ actions so that they usually form skewed expectations and decisions about the bureaucrats’ performance and capabilities (Frederickson et al., 2012; Moe, 1984) . Fourth, there are risk and uncertainty as a result of both preference and information asymmetries. Risk means the measurable, unknown uncertainty where one can use additional information as a mech anism to counter such risk while uncertainty connotes the absolute unknown mechanism and therefore the outcomes cannot be predicted (Ellsberg, 1961; Georgescu Rogen, 1954) . Hence, under the circumstances of risk and uncertainty, utility theory cannot be fully applied (Ellsberg, 1961; Friedman & Savage, 1948) . Risk and uncertainty here imply the high possibility that policymakers fall prey to bureaucrats’ opportunism and suffer from the problems of adverse selection and moral hazard due to the agents’ shrouding of information —the concealment, distortion, and fal sification of information (Akerlof, 1970; Breton, 2009; Spence, 1973) . Additionally, given that the production of public goods and services are of technical knowledge and specialization, even though it is theoretically feasible for policymakers to monitor the agent’s actions, it is prohibitively expensive or economically unviable for executives to gather complete information in order to conduct a thorough monitoring (Moe, 1984; Ross, 1973) . On this ground, the executives are likely to allocate budget for inferior public goods and services offered at the higher price than their true quality (Niskanen, 1971) . Fifth , policymakers as risk averse individuals seek resolving mechanisms to minimize or cope with information problems. Though informational advantage is located on the bureaucrats’ 17

PAGE 30

side, authority is located on the policymakers’ side (Miller, 2005) . Policymakers can seek relevant and reliable information through two main mechanisms —information intermediaries and performance contracts and also can use their professional judgment when making decisions. 1. Information intermediary: An information intermediary refers to organizational units or members who gather and champion the use of knowledge (Corburn, 2007) , simplify scientific information, and bridge the gap between the science and technology and the end users (i.e., executives, legislators, and citizens) (Faibisoff & Ely, 1974) . Even though the question of which information a public organization should gather and utilize in decision making cannot be answered, such organizations should at least possess a basic set of information in order to show its co nstituents that they are fiscally transparent and accountable and that they promote fiscal sustainability, good governance, and overall fiscal rectitude (Alt, Lassen, & Skilling, 2002b; Kopits & Craig, 1998) . This basic set of fiscal information includes such information as economic projection, statement of future plan, outside opinion and evaluation, input, output, service quality, efficiency, and outcomes (Kopits & C raig, 1998a; Seifert, Carlitz, & Mondo, 2013) . 2. Performance contract design: Since it is prohibitively expensive for policymakers to monitor their bureaucrats’ actions, they invest in contract design which is based mainly on performance information (Hlmstrom, 1979; Moe, 1984; Wood, 2010) . Performance information reflects actionable and quantifiable production pr ocess and products or services, encompassing outputs (public goods and services produced or delivered), efficiency ( a result from a cost benefit analysis), effectiveness (the evaluation of missed and achieved targets in terms of both quantity and standards ), 18

PAGE 31

and outcome (economic and social implications of such public goods and services) (Ammons & Roenigk, 2015; Talbot, 2007) . Given the fact that this performance information is easily observed and evaluated, the shirking and budget maximizing behaviors among bureaucrats can be controlled reasonably well without substantial monitoring costs (Moe, 1984) . 3. P rofessional judgments: Despite their attempts to gather all the accurate and relevant information through the previous mechanisms above, the perfect or complete information is not guaranteed for principals to solve or even avoid the problems resulting from the preference and information asymmetries (Ellsberg, 1961) . Therefore, policymaker s significantly depend on their professional judgments , which are indispensable for taking actions to effectively formulate and implement public policies (Bertelli, 2007; Smith, 2003) . Indeed, governmental decisions are products of a complex collective judgment (Appleby, 1949) . Professional judgment s represent technical competence, knowledge, and specialization and are significantly accumulated through training and experiences (Karl, 1976) . Under the context of performance management and budgeting, public administrators have flexibility combining expertise with professional judgment in order to improve their organizations’ technic al efficiency and effectiveness ( Moynihan, 2008b) . Sixth, according to rational choice theory, policymakers would utilize all types of relevant information instrumentally they largely use it for budget analysis and eventually for making budget decisions (Knorr, 1977; Niskanen, 1971; Weiss, 1977) . In other words, relevant and reliable information is possibly taken into account in various stages of information 19

PAGE 32

utilization during a budget process — reception, cognition, discussion, reference, scrutiny, effort, and influence (Knott & Wildavsky, 1980; Rejean Landry, Lamari, & Amara, 2003) . Seventh, eventually, budget decisions in form of budget allocations, which constitute payments to bureaucrats for their delivery of public goods and services , can act as incentives to spur agents more toward the policymakers’ policy directives. For example, an increase in budget allocation means the promotion of bureaucrats’ highperforming activities and a decline in budgets can be cast as sanctions to curtail certain actions of the agents (Eisenhardt, 1989; Wood, 2010) . To sum up, information asymmetry and principal agent theory suggest that there is an informational g ap between policymakers and bureaucrats as principals and agents, respectively. Though informational advantage is located on the bureaucrats’ side, authority is located on the policymakers’ side. Policymakers can acquire relevant and reliable information b y creating an information intermediary. It can create its own information or take advantage of the characteristics of information as a public good, gathering available information offered by other organizations . This information can keep policymakers so informed of the quality and cost of service delivery that they can make budget decisions appropriately. Therefore, the problem of adverse selection can be diminished. In addition, policymakers also require bureaucrats to act on their behalf by designing perf ormance contracts and performance measures in order to make sure that the shirking and budget maximizing behaviors among bureaucrats can be controlled without substantial monitoring costs. Accordingly, the moral hazard can be reduced. Policymakers, therefo re, utilize information from both information intermediaries and performance measures to make budget decisions, specifically budget allocation and budget effect. Figure 2.1 below 20

PAGE 33

graphically elaborates this relationship between the theories of information economics and budget decision making. Figure 2.1 The causal graph of theories of information economics under the context of information u tilization and budget decisions Empirical Studies on Information Asymmetry, Information Use, and Decision Makings As suggested in Figure 2.1, there are many causal relationships in the principal agent theory. In the context of information use and budget decision making, the related empirical studies add several perspectives and nuances into t his theory. The following parts explain such empirical studies through each causal step shown in Figure 2.1. Some empirical conclusions as well as research gaps are also provided. Sources of I nformation and F iscal I nformation Contracts and their performan ce information is automatically available if a public organization adopts any form of performance measures into decision making process (Kong, 2005) . The performance information encompass es input, output, efficiency, performance targets, benchmarking data, and service quality that prevails in public organizations such as the U.S.’s state and local governments (Bourdeaux, Chikoto, Bourdeaux, & Chi, 2008; Julnes & Holzer, 2001) , Chinese local governme nts (Zhang, Van de Walle, & Zhuo, 2016) , Norwegian central 21

PAGE 34

government’s agencies (Askim, Johns en, & Christophersen, 2008; Laegreid et al., 2008) , and Danish municipalities (Nielsen & Baekgaard, 20 15) . Meanwhile, other types of information are gathered and transmitted to policymakers by an information intermediary —individuals (policymakers themselves also included) or a unit responsible for retrieving all the information available concerning the policy a reas of concern to their policymakers (Rich, 1975) . For instance, in Australian coastal cities (Shaw, Danese, & Stocker, 2013) and some communities in Pacific islands (Cash, Borck, & Patt, 2006) public organizations and outside information producers create and transfer information from local people, scientists, and stakeholders to policymakers while ensur ing that local knowledge, which also is trusted and legitimate, make s it into a policy process. Boundary organizations’ informational role sometimes has an impact on decision making. The case in point is that the U.S. Environmental Protection Agency has had a great policy influence on stakeholders through its roles of information dissemination despite bei ng definitely liberal during the period in which Republicans were dominant in the Congress (Lee, 2012) . These boundary organizations or information disseminators produce information that is relevant to the policy areas of concern. Even though not specifying the information elements like the case of performance information, they offer two main types of information—expert knowledge and local knowledge. Expert knowledge is the systematically gathered information presented in research or evaluation reports which highlights analysis that meets standards of coherence and honesty and uses generally accepted professional, scientific, and technic al methods (Lundin & berg, 2014; Weible, 2008) . Meanwhile, local knowledge is the information about a local context or setting, including empirical knowledge of specific characteristics, circumstances, events, and relationships, as well as the normative understandings of their 22

PAGE 35

meanings; is usually hel d by members of a community that can be both geographically located and contextual to specific identity groups; and owes its status to causal empiricism, thoughtful reflection, and common sense (Corburn, 2003b; Fischer, 2000) . In the case of environmental policy, the Ontario Ministry of Natural Resources requires local knowledge to be included in forestry research for formulating its policies (Klenk & Hickey, 2011) . At the same time, local knowledge has been used in terms of the environmental risk in the water use planning in Canada (Failing, Gregory, & Harstone, 2007) ; how a low income, Latino immigrant neighborhood in New York City organized its knowledge to participate in and significantly change a U.S. Environmental Protection Agen cy exposure assessment (Corburn, 2007) ; and in local plans for the communities facing the most serious environmental and health risk (Corburn, 2003) . To sum up, the literature that mostly explores information use in developed countries specifies both the information source (i.e., performance contracts) and the performance information piece (i.e., input, output, effi ciency, performance targets, and benchmarking data, and service quality) . It also identifies information sources (i.e., information intermediary) for other relevant information, namely local knowledge. Accordingly, this implies that there are two areas the literature has not adequately explored —the information production in the context of developing countries and how local knowledge is used for budget decision making. Fiscal I nformation and I nformation U tilization Generally, policymakers can use information in two different fashions (Weiss, 1977). They can symbolically employ information only to legitimate political or bureaucratic activities decided upon before the research was undertaken —activities necessary to maintain the postindustrial welfare state —and to manipulate the consent of the citizen. However, sometimes information can permeate decision making in subtle ways and bring about the changes in 23

PAGE 36

political or bureaucratic thinking. As the budget reforms aims to promote the latter type of information use, there are several related studies exploring how to promote the latter type of information utilization. T he relationship between information and its utilization for decision making rests much upon the availabilit y and nature of the information itself (Pandey, 2015) . In other words, information producers may offer a variety of information but not all of it is utilized (Caplan , 1979; Dunn, 1980; Ottoson, 2009) and some of it is utilized more frequently than others (Wolman & Page, 2002) . According to the principal agent theory, relevant and reliable information is used for decision making (Akerlof, 1970; Moe, 1984) . Relevance of information elements is measured in terms of usefulness, significance, and frequen c y of use (Melkers & Willoughby, 2005; Wolman & Page, 2002) and reliability depends on sources of information (Boggs, 1992; Wolman & Page, 2002) . The utilization can be divided into two different levels, understanding and deliberation. The information in terms of understanding involves the situation where information has been considered but not used for decision making (Knorr, 1977; Rich, 1975) . Meanwhile, information utilization in terms of deliberation refers to sequential stages of utilization encompassing such cognitive and management activities as scrutiny, effort, and influence (Knott & Wildavsky, 1980) . There are also some examples of the relationship between the characteristics of information and the utilization. Regarding reliability, the findings from related empirical studies state that policymakers comprehensively utilize information that is generated by experts or researchers . For instance, the trust in information sources (i.e., perceived abi lity, integrity, and benevolence of boundary organizations or information disseminators) leads to the active use of performance information among the Chinese central government agencies (Zhang et al., 2016) 24

PAGE 37

and the involvement of policymakers (specifically public mana gers in the supervisor level in the U.S. federal government agencies) in terms of preparing for or gathering any Performance Assessment Rating Tool (PART) assessment has an impact on the extent to which such information is utilized ( Moynihan, Lavertu, & Kamensky, 2012) . To conclude, the empirical studies suggest that information —fiscal, performance, and other — that is reliable and relevant is likely to be utilized comprehensively and actively by policymakers. Note that even though policymakers receive, reference, and discuss such information with each other, they do not always deliberate it, potentially limitin g their effect to understand the information by not asking questions, scrutinizing the information, and preparing to change their opinions or letting the information influence their decisions . In this vein , the literature has not adequately explored how pe rformance information together with other information types are utilized especially under the budgeting’s context. Information U tilization and B udget D ecisions Generally, decisionmakers use information in two fashions. First, according to the concept of Based on performance management doctrine (Jakobsen & Mortensen, 2016; Moynihan, 2008b) , several studies assume that performance information is fully utilized and it eventually has an influence on budget decisions. The literature studies these decisions through both actual and perceptional budget allocation. For example of actual budget allocation, PART scores play a role in the determination of program budgets suggested by the U.S. federal government’s O ffice of Management and Budget (Gilmour & Lewis, 2006a); government agencies in Chile that comply more with performance targets have seen their aggregate budgets decrease relative to those with weaker compliance records (Zaltsman, 2009) ; and the number of outcome rela ted 25

PAGE 38

performance measures used by city officials accounts for the change in annual budget allocation of major departments and operations of municipalities and counties in Indianapolis (Ho, 2011) . For example of perceptional budget allocation, if administrators and budget analysts in the U.S. local government use performance measures at various stages of t he budget process, they report the impact of such utilization on cost saving, appropriation levels, and reduction or elimination of ineffective or duplicative services or programs (Melkers & Willoughby, 2005) . Furthermore, Norwegian public managers who thoroughly utilize performance information (i.e., formulating goals and performance indicators and making evaluation reports) report that such information utilization has impact s on resource allocation, rewards for good results, and sanctions for poor results (Laegreid et al., 2008) . Therefore, it can be said that the comprehensive use of performance information can have an influence on the budget decisi ons. However, information other than performance information or the combination of performance information and other information still have not been adequately explored in the context of budget decisions. Professional Judgment and B udget D ecisions Even though, theoretically speaking, governmental decisions are products of a complex judgment (Appleby, 1949) , t he literature exploring relationships between professional judgment and budget decisions is limited. The related study applies administrative capacity in terms of professional training and educational background as a proxy of professional judgment and finds that professional judgment of the U.S. government budget officers are not associated with their choices of resource allocation (Reck, 2000) . Budget Rationality Model The budget rationality model theorizes that decision makers use the information necessary to advance their agendas. Briefly, this bu dget rationality model states that with 26

PAGE 39

different administrative capacity (i.e., spending attitudes) policymakers possess differing orientations toward budgetary as well as spending preferences and thus only certain types of information are given priority in order to guide their budget decision making (Moynihan, 2006; Nielsen & Baekgaard, 2015) . The current literature develops this model though grounded t heory, it theoretically progresses as its empirical findings accumulate (Goodman & Clynch, 2004; McCue, 1999) . Based on those findings, the Figure 2.2 below graphically portrays these relationships. Figure 2.2 Budget Rational ity and Information Utilization Administrative C apacity and I nformation U tilization In general, budget actors’ spending preferences varies according to their seniorities (Hildreth, Yeager, & Miller, 2011; Thurmaier, 2001) . In particular, seasoned analysts with more years in service should be able to anticipate political cues, thereby weighing political factors more heavily than nonseasoned analysts (Goodman & Clynch, 2004) . For example, the tenure of budget analysts in the U.S. state governments impacts their spending preferences because they tend to respond to political cues quicker than less senior decision makers who tend to rely on economic cost benefit decision criteria (Willoughby, 1991). Especially the senior analysts in Kansas State Budget Office acknowledge a larger role for political rationality in their budget recommendations (Thurmaier, 2001) . However, a more current study in the thirteen U.S. western states does not find the statistically significant association between hierarchical position or the 27

PAGE 40

number of years each budge t analyst has held in her/his current position and the propensity to weigh political factors more heavily than junior analysts (Goodman & Clynch, 2004) . In addition, administrative capacity is actually associated with the type of fiscal information one finds relevant or importa nt and usually utilize s . For example, the U.S. state governments’ budget analysts with business administration education are more likely to exercise the information of financial accounting control than those with social science education (Lee, 1991) . Meanwhile, the budget analysts with social science education are more likely to use performance information than those with business administration education (Lee, 1991) . Overall, the administrative capacity of policymakers obviously plays a role in the use of information during a budget process. Administrative capacity is an umbrella concept which covers such qualities and conditions as education, knowledge, skills, traini ng, experiences, orientations, attitudes, and motivation (Christensen & Gazley, 2008; Sun & Gargan, 1993) . How administrative capacity specifically influences the information utilization depend s largely on which dimension one chooses to study. Consequently, in general, it can be said that if an individual possesses higher administrative capacity on one par ticular dimension, he or she is more likely to utilize information directly related to such dimensio n during the budget process relative to other individuals. Nonetheless, the current literature focuses mostly on the administrative capacity in terms of qualification (i.e., experiences and educational background), leaving other dimensions (i.e., skills and attitudes) unexplored. Additionally, these studies only focus on the types of information policymakers use and not on how policymaker use it during the budget process. 28

PAGE 41

Budget O rientations and I nformation U tilization There always appears a struggle among three traditional distinct budget norms or budget decisionmaking patterns shared among budget actors — control (fiscal accountability accuracy), management (efficiency and effectiveness), and policy (strategic planning and advocacy for and responsiveness to policy outcome) (David, 1998; Smith, 2003; Willoughby, 2002) . Relevant studies that concentrate on the budget roles played by the U.S. state budget analysts find that there are five budgeting orientations, ranging from the control to the management and to the policy orientations as follows (Smith, 2003; Thurmaier & Willoughby, 2001; Willoughby, 1993; Willoughby, 2002) . (1) Conduit: Policymakers act as conduit in the budget process — only transmitting information between executives and agency and back again and usually presenting a wide range of information to reflect all the spending preferences. (2) Accounting: Policymakers act as fiscal guardians who serve traditional guardian functions and are oriented toward control (num ber crunching), accounting principles, integrity, fiscal responsibility, and accountability. They think that the agency workload and agency efficiency are the most important work when making budget decisions. (3) Technorational: Budgeteers act as both strateg ic planners and budgeting professionals. Not only do they enlighten agencies of executive priorities and funding strategies, they also promote such budgeting techniques as Program Planning Budgeting System ( PPBS) , Management by Objectives, Zero Based Budgeting while alert other budget actors of legal and other requirements that must be met. 29

PAGE 42

(4) Policymaking: Budget actors are policy analysts who take into account a comprehensive view of the overall budget and policy outcomes or how the budgetary decisions impact the government’s overall fiscal condition and then push forward their policy initiatives based on such analysis. The budget actors in this group apply weight to both political and analytical forces. (5) Advocating: Concentrating on policy responsiveness and current political and economic agendas, budgeteers act as policy advocates who seize the window of opportunities to push forward their policy initiatives into the budget process. The budget actors apply more wei ght to political factors such as executive agendas and public support. Budget actors’ spending preference are based on individual decision modes —how they attend to, encode, and process particular information (McCue, 1999) . From the budget orient ations above, it implies that only particular information is carefully or purposefully considered and used during a budget process. As illustration, based on a basic set of fiscal information that a transparent government should make available and the conc ept of budget rationality, Table 2.2 shows how policymakers with different budget orientations utilize information during the budget process. 30

PAGE 43

Table 2.2 Budget O rientations, R elevant and I mportant I n formation C onsidered, and N orms Budget Orientation Possibly Relevant Information Information Type Budget Norms Conduit Unspecified Unspecified Unspecified Accounting Audit report and financial ratios Financial and accounting data (Analytical information) Control Technorational Possibly all types of budget information All types (Analytical information) Control, management, and Policy Policymaking Economic projections and policy results or outcomes All types (Analytical information) Control and Policy Advocating Economic projections and policy platforms Social and political data (Political cues) Policy Source: Adapted from Willoughby’s model of decision environment of the CBO Analysts (2002) ; Goodman and Clynch’s information attributes (2004) ; and the examples of information for budget transparency from De Renzio & Masud (2011) and Kopits & Craig (1998) . Based on Table 2.2, policymakers with conduit orientation plausibly show the very limited use of information during the budget process and do not possess a particular spending preference. Meanwhile, ones with accounting orientation are assumed, on a verage, to utilize higher level s of information especially from the comprehensive use of the financial and accounting data in order to keep the budget allocation efficient and transparent. Policymakers with techno rational orientation in general are likely to use informat ion to a higher extent than other groups of policymakers. This is because they contemplate all types of information as the budget process proceeds in order to ensure the efficiency, effectiveness, and policy goals. For ones with a policymaking orientation, similar to the techno rational policymakers, they generally use information more comprehensively than other groups. However, the policymaking budget actors use information in order to ensure efficiency and transparency for particular polic y areas with which they are concerned. Finally, for the policymakers with advocating orientation, they are likely to use information more comprehensively than those with conduit orientation but lower than those with technorational and policymaking orienta tions. This is due to the fact that they 31

PAGE 44

only concentrate on information regarding political cues in order to make sure that the budget allocation follows the policy platforms initiated by politicians or organization’s executives. Nonetheless, the empiric al studie s which thoroughly review the relationship between budget orientations and the utilization of information are limited. A study on budgetary decision making by budget analysts in the thirteen U.S. western states confirms that not only one type of i nformation (i.e., performance information) , but a bundle of varied information is contemplated during the b udget process according to analysts’ budget orientations (Goodman & Clynch, 2004) . The study neithe r give s insight on the level of information use nor on t he ways policymakers utilize the information. These issues also warrant the comprehensive study. Findings from the Theoretical Review From the theories of information economics (information asymmetry and principal agent theory) and budget rationality model, there are four main findings. First, sources of information and quality of the information itself play a crucial role in how policymakers utilize such information. Particularly the source of infor mation determines the reliability of the information while the information itself determines its relevance. These two character istics of information influence how the policy makers use information during the budget process. The current literature , which he avily foc uses on performance information, assumes the high level s of its reliability and relevance without actually measuring them. Additionally, individuals indeed do utilize performance information together with other information. In other words, they use a combination or a bundle of information consisted of performance information, financial and accounting information, and political and social information. As a result, it is necessary for the related studies to investigate these two information charact eristics of information bundle 32

PAGE 45

policymakers utilize while initiating, analyzing, discussing, and defending their budget proposals. Second, administrative capacity and budget orientations also have an impact on how policymakers perceive available informati on in terms of its relevance and reliability and utilize that relevant and reliable information during a budget process according to the instrumental use of information (Weiss, 1977) . Policymakers with higher level s of administrative capacity are more likely to comprehensively utilize information for drafting their budget proposals. Additionally, policymakers with different budget orientations utilize different pieces of fiscal information for their budget decisions. Third, administrative capacity and information utilization are possibly the determinants of budget decisions. The empirical studies confirm existence of the relationship between information utilization and budget decisions. However, the studies regarding whether the administrative capacity , in terms of skills and knowledge and how information is specifically used during the budget process to impact on budget allocations , are limited. Furthermore, the literature also finds the impact of information utilization on budget allocation in both perceptional and actual terms for major spending programs. Nonetheless, these studies do not consider both types of budget allocations simultaneously. Fo urth, except for China (Zhang et al., 2016) and Chile (Zaltsman, 2009) , the public budgeting literature regarding the use of information and budget decisions does not pay attention to the situations in developing countries. Particularly, the recent studies are heavily based on the experiences of such countries as the U.S. and Canada. On this ground, the use of information and how it affects the budget allocation under the context of developing counties should be advanced. 33

PAGE 46

Research Questions and Hypotheses Considering the research areas the extant literature has inadequately explored, this study asks the following overarching question: How do policymakers with different administrative capacities utilize information during a budget process in order to have an impact on budget decisions ? Specifically, there are two sub questions supporting the main research question above. This section presents t he sub questions and offers hypotheses regarding each. First, how do policymakers with different administrative capacities utilize relevant and reliable information during a budget process? H1A: The more policymakers perceiv e information to be relevant and reliable, the more likely they are to utilize it during the budget process. H1B: Policymakers with higher level s of administrative capacity are more likely to use relevant and reliable information during the budget process relative to those with low er level s of administrative capacity . H1C: Policymakers with techno rational and policymaking orientations are more likely to utilize information than those with conduit, accounting, and advocating orientations. Second, how does the information utilization influence budget decisions? H2A: The more policymakers use information during a budget process, the higher more such use impact s the budget decisions. H2B: How policymakers combine different types of information dur ing the budget process affects their budget decisions. H2C: Policymakers also base their budget decisions on their administrative capacity (technical skills and knowledge ) to frame their budget decisions. 34

PAGE 47

Rival Hypotheses There are some factors that co nfound the relationship between administrative capacities, information utilization, and budget decisions. Two the oretical explanations account for such confounding effect s . This first is structural isomorphism. For a certain distinguishable environmental configuration, there is one organizational structure (i.e., a set of values, norms, standard of procedures, management techniques, products, servic es, programs, or policies) that is widely accepted as the most appropriate way to do things so that other organizations attempt to resemble it (DiMaggio & Powell, 1983; Hannan & Freeman, 1977; Selznick, 1948) . This resembling behavior, or structural isomorphism , possibly stems from normative and/or coercive pressure (Dimaggio & Powell, 1983; Meyer & Rowan, 1977) . Normative pressure exists when formal education or the growth and elaboration of a professional network is considered the source of values, norms, and standards of procedures with which organizational leaders or managers comply. Meanwhile, coercive isomorphism stems from political influence exerted on organizations by other organization upon which they are dependent. However, due to this pressure instead of the n ecessity inside the organization, it is possible that despite the existence of formal structures, it does not have an effect on a behavioral change. This phenomenon is called formalism and seen as a dilemma of the administrative reformer in developing coun tries including Thailand (Riggs, 2006) . Under this study’s context, to appear as a highly performing organization, local governments are normatively required to adopt performance budgeting in which budgetary actors produce or gather and utilize performance information when making budge tary decision (Hou, Lunsford, Sides, & Jones, 2011; Kong, 2005; Willoughby & Melkers, 2000) . However, 35

PAGE 48

budgetary actors also do not tie performance information to incentive and disincentive systems in term of pay, status promotion, and budget cut (Ammons & Roenigk, 2015; Hou et al., 2011; Laegreid, Roness, & Rubecksen, 2008) . For instance , state government agencies in Florida do not actively implement the performance based budgeting (PBB) as a replacement of old budget systems (Andrews & Hill, 2003) . Australian government agencies also use performance information in a symbolic fashion, not for making decisions on budget allocation but for elaborating plans and reports in order to project an image of good governance (Yetano, 2013) . U.S. local governments extensively utilize input and output measures while outcome measures are less developed and utilized when making budgetary decisions ( Julnes & Holzer, 2001; Melkers & Willoughby, 2005) . Also, the main driver for the development of Thai public sector’s performance management system is the Office of Public Sector Development Commission, acting as the institutional entrepreneur or a source of normative pressure (Sutheewasinnon, Hoque, & Nyamori, 2016) . On this ground, it can be said that due to the coercive and normative pressures on adopting performance management, even though policymakers find relevant and credible information produced by boundary organization relevant, they do not utilize it for deliberation duri ng a budget process. Second, even though information matters in the market realm, the theory of incrementalism holds that the addition of related information will not transform previous pattern of budget decision making (Wildavsky, 1964) . Budget information is almost never actively reviewed as a whole in the sense of considering at once the value of all existing programs as compared to all possible alternatives so that the current year’s budget i s based on last year’s budget, with special attention given to a narrow range of increase or decrease (Davis et al., 1966; Wildavsky, 1978) . 36

PAGE 49

For instance, though the U.S. local governments make effort to integrate performance measure into budget decisions, the application of performance information for budget decisions progresses incrementally (Melkers & Willoughby, 2005) . Also, the factor i ndicating incrementalism such as age of spending program has an impact on the change in budget allocation (Gilmour & Lewis, 2006) . In conclusion, the incremental view holds that, despite of the availability of relevant and credible information, budget decision makers do not comprehensively utilize it and thus it does not have an impact on budget decisions. Theoretical Framework Based on the findings from the theoretical review and the related empirical studies previously discussed in this chapter and the rival hypotheses, this study proposes the theoretical framew ork on administrative capacity, the use of information, and budget decision as elaborated in Figure 2.3. Figure 2.3 Theoretical Framework of Administrative Capacity, the Use of Information, and Budget Decisions 37

PAGE 50

It is worth noting that even though the related theories discuss all of the relationships in Figure 2.3, there are still some causal relationships which are not empirically confirmed. These relationships are information sources and fiscal information (types and charac teristics), administrative capacity and information utilization, budget rationality and information utilization, and administrative capacity and budget decisions. This study aims to test these theoretical associations through the local governments in Thail and with mixed methods research , presented in Chapter 3 and Chapter 4. 38

PAGE 51

CHAPTER III CASE JUSTIFICATION AND THAILAND’S LOCAL PUBLIC FINANCE AND BUDGETING This chapter first describes how the study selects the case, considering representativeness and causal leverage. As the previous chapter asserts that the public budgeting literature regarding the use of information and budget decisions does not pay attention to the situations in developing countries, this study proposes Thailand’s local government as the case studies. Accordingly, after the case justification, this chapter reviews the country’s progress on decentralization, the revenue and expenditure’s structure, and the local budgeting, including a pr ocess of budget decision making and roles of budget actors . The Case Justification for Thailand’s Local Governments Both large N crosscase analysis and case study analysis aim to reproduce the causal features pertaining to a larger universe (representativeness) and provide varia tion along the dimensions of theoretical interest (causal leverage) (Gerring, 2007) . To promote representativeness and causal leverage, this study selects Thailand’s local governments for case study analysis . The following details explica te how these cases can contribute in terms of representativeness and causal leverage to the study of how administrative capacity and the use of information impact budget allocation. Finally, based on these contributions, Thailand’ s local governments can se rve as the least likely case to (dis)confirm the related theories. Representativeness The research on the relationship among capacity, the use of information, and budget decisions is mostly based on the experiences of developed countries such as the U.S. (Laegreid, Roness, & Rubecksen, 2008; Melkers & Willoughby, 2005; Moynihan, 2008a, 2015; Pollitt, 39

PAGE 52

2006) and the European countries such as the Netherlands (Buylen & Christiaens, 2016) and Denmark (Nielsen & Baekgaard, 2015) . Accordingly, in order to leverage the representativeness of such theoretical relationships, there should be research that incorporates new subtypes of the research settings (George & Bennett, 2004) . Specifically , in this case, there should be studies explicating the relationship among capacity, the use of information, and budget allocation in the developing countries. However, in order to identify the potential country, it is required to consider the second case selection criteria—the causal l everage —which is discussed in details as follows. Causal leverage The causal linkages this study aims to confirm involve two relationships: administrative capacity and the use of information and the use of information and budget allocation. The following theoretical explanations clarify these two linkages. Administrative capacity and the use of information. Overall, all else being equal, if public organizations have high administrative capacity (i.e., quality managers and good management systems) or the po tential to obtain desired program results and policy outcomes, we can assume that such organizations are likely to be effective performers (Ingraham et al., 2003) . Since the productivity or performance of bureaucrats is unknown to public organizations until after the critical decisions regarding job, wage, salary, or p romotion offers have been made, administrative capacity (i.e., educational backgrounds and past experiences) serves as a signal or a key information element that is an observable attribute upon which public organizations make judgements about individual bureaucrats’ performance (Riley, 1975; Spence, 1973) . Regarding the relationship between administrative capacity and use of information, public organizations with sufficient administrative capacity specifically in terms of public 40

PAGE 53

managers capable to handle informational tasks such as perfo rmance analysis, cost analysis, program budget analysis, and communication with public, are more likely to succeed with the quality and the effective use of information (Ho, 2011; Melkers & Willoughby, 2005) . This is because the freedom, consistency, and speed with which these managers make information flow throughout an organization; the attention th ey give to specific information or knowledge; and their willingness to share such information or knowledge converge to facilitate (or thwart) the extent to which a government’s management systems take into account and utilize such information or knowledge (Ingraham e t al., 2003) . As a result, administrative capacity is a key determinant of and positively associated with the success in the use of information (Ho, 2011; Lee, 2012; Pollitt, 2006; Van Dooren, 2008) . The use of information and budget allocation. Informatio n in general can resolve uncertainty and reduce ambiguity so that it enhances economic decisionmakings (Faibisoff & Ely, 1974; Feldman, 1989) . As the rational choice theory suggests, information is a key element for an organiz ation’s logical decision making because it is used extensively in formulating goals, searching for alternatives, and calculating consequences of each alternative (Downs, 1957; Lindblom, 1959) . Particularly in a budgeting process, information such as public service outputs and outcomes encourage s learning and deliberation among different departments in state and local governments (Goodman & Clynch, 2004; Melitski & Manoharan, 2014; Moynihan, 2008a, 2008b) . Such information utilization can have budget effects in form of cost savings, appropriation levels, elimination of ineffective programs, and duplicative services especially among state governments (Melkers & Willoughby, 2001, 2005; Willoughby & Melkers, 2000) . Additionally, even from the bounded rationality perspective, information impacts dec ision making significantly. Different types or aspects of information have different levels of 41

PAGE 54

accessibility and possess particularly different percepts (Kahneman, 2003) . Such informational elements serve as an incentive for individuals to focus their attention and prolong t heir deliberation only on a particular issue and in a particular fashion so that they give consideration only to a particular set of alternatives with the exclusion of others during the decision process (Simon, 1985; Tversky & Kahneman, 1986) . For instance, if provided with a variety of budget scenarios as well as different amounts and types of information, individuals make their decisions on public spending dependent on how performance information is presented to them in terms of goal ambiguity, expectancy disconfirmation, and advocacy ( Moynihan, 2015) . Thailand as a least likely case Though providing inconclusive r esults, the current studies in public policy, public management, and public budgeting have theoretically explicated the causal relationships b etween administrative capacity, the use of information, and budget allocation through the information asymmetry an d the rational choice theories. So as to promote the causal leverage or causal inference of these relationships, a diagnostic case study approach is co nsidered helpful given the fact that it is intended to assess whether the broadly understood assumptions based on the information asymmetry and rational choice theories hold (Gerring, 2017) . In addition, since the research results on how the use of information has an impact on budget allocation and how administrative capacity as a background factor may affect the use of information are inclusive, influence or the probability of the aforementioned assumptions being true may be understood in a counterfactual fashion, specifically via the influential case (Gerring, 2007a, 2017) . The influential case appears at first glance to nullify a theory, or at least to cast doubt on a theory but upon a closer inspection, it does not and it ends up confirming the validation and re identifying the scope conditions of such theory (George & Bennett, 2004; Gerring, 2007) . 42

PAGE 55

Regarding the relationships among administrative capacity, the use of information, and budget allocation in the subnational governments, it can be said that if those organizations possess relatively low administrative capacity, have limited access to relevant information, and constrainedly utilize such limited informatio n, they are unlikely to see the change in their budget allocations as a result of administrative capacity and the use of information. Many studies in public b udgeting measure the concept of administrative capacity via expenditure level of local authoritie s (Brace, 1991; Jeong, 2007) . An expenditure level serves as a blunt but plausible and theoretically justifiable measure of overall organizational capacity (Brace, 1991) . Even though the high level of expenditure may represent waste and inefficiency (Niskanen, 1971) , it is also possible that high local expenditure can bring about capacity building activities, creating quality local bureaucrats. In addition, the higher local expenditure level is, the more resources local authorities possess to promote activities related to information utilization such as access, deliberation, and analysis of information (Terman & Feiock, 2015) . Low level of local public expenditure. C ount r ies whose subnational government expenditure is considered relatively very low (between 8 percent of GDP and 20 percent) includes most African countries, most Latin America countries, several Euro Asia countries, and some AsiaPaci fic countries (United City and Local Government, 2016) . However, since decentralization in East Asia began later than in other parts of the world, countries there are still redesigning the framework of and addressing the details of civil se rvice management at a local level (Green, 2005; White & Smoke, 2005) . Among East Asian countries, Thailand has the lowest subnational expenditures as a sha re of total public spending (approximately 10 relatively to the average of 37 percent) (White & Smoke, 2005) . Additionally, notwithstanding a middle income country with a bustling capital city, Thailand’s local governments possess low 43

PAGE 56

administ rative capacity —in general staff lack the technical skills and institutional resources to deliver even basic services (Gr een 2005) and specifically they lack quality budget officers to enhance the budgetary decision making process (Kittipanyasiri, 2015; Krueathep, 2014; Nagai et al., 2008) . This lack of administrative capacity at the local level in Thailand prevails due to the fact that the competing central agencies do not coordinate well to plan the local governments’ capacity building and most Thai bureaucrats perceive the trans ferring from the central government agencies to the local government as a demotion (NESCB & Thamasat University, 2009; World Bank, 2012) . Low level of the use of information. Although efforts have been placed on data collection regarding performance monitoring and evaluation through such tools as electronic local administrative acc ounting system (e LAAS) and delivery local quality management (LQM), robust use of information is not widespread. In terms of quantity, only 1,200 out of 7,853 local governments are implementing e LAAS because of a combination of connectivity, capacity, an d enforceability problems (World Bank, 2012) . In terms of quality, these tools are not able to provide appropriate information on the application of public finances or service delivery quality and quantity because they focus largely on ensuring compliance with pr ocesses rather than on quality of services and outcomes (World Bank, 2012) . Somewhat related to the relatively low administrative capacity discussed above, local governments in Thailand have limited relevant information available for a budgeting process and their budgetary decision making (Wongpredee & Sudhipongpracha, 2014) . In conclusion, according to the information asymmetry and the rational choice theories, with relatively low administrative capacity and limited use of information during a budgeting process, Thailand’s local governments at first glance are unlikely to be able to leverage their 44

PAGE 57

administrative capacity and use their available information to enhance their budget allocation decisions. However, it is important to test these theories through the closer inspection of these least likely cases because if the results uphold these theoretical linkage, there is strong, robust empirical evidence to validate such theories; or , in other words , to reassure that the probability of such theoretical assumptions being true is very high (George & Bennett, 2004; Gerring, 2017) . Thailand’s Local Public Finance and Budgeting Thailand’s Decentralization and Local Governments Subnational government administration in Thailand is structurally straightforward but politically complex. Beside the special localities (Bangkok Metropolitan Administration and Pattaya City), there are two main types of local governments in each province. First, for each town (often equivalent to a subdistrict size and consisted of a group of communities), there exist a local government, which can be either a municipality or a Tambon Administrative Organization (TAO) depending on such town’s economic development and political importance. Second, for each province, there is a Provincial Ad ministrative Organization (P AO), administering the local affairs beyond the authority of municipalities and T AOs. T here are 7,853 local governments in Thailand. Note that PAOs are mainly responsible for financially assisting the other local governments in their public service delivery and TAOs do not carry out the key functional programs (i.e., park and recreation, sanitation, education, health, and transportation) according to the PAOs Act 1997 and the TAOs Act 1994 respectively. Table 3.1 gives additional details on level, economic and political importance, and number of each type of local governments in Thailand. 45

PAGE 58

Table 3.1 Types of Local Governments in Thailand and Their Differences Type Level 1 Political Significance Number 2 Provincial Administrative Organization (PAO) Province (several districts together) Managing inter jurisdictional public services within a province 76 Metropolitan Municipality Sub district Metropolis of a certain region (i.e., population is over 50,000) 30 City Municipality Sub district Provincial capital or populated subdistrict (i.e., population is between 10,000 50,000) 178 Tambon Municipality Sub district Populated sub district in a certain district (i.e., population is between 7,000 10,000) 2,23 3 Tambon Administrative Organization (TAO) Sub district Rural community 5,33 3 Special Localities [ Bangk ok Metropolitan Administration (BMA), and Pat taya City] Ranging from a subdistrict to a province Fast growing metropolitans (i.e. the country’s capital and the cities of tourism) 2 Total 7,85 2 Note: 1. Two or more subdistricts constitute a district and two districts or more constitute a province. There is no district level local government. 2. Number of e ach type of local governments was updated on March 24, 2018 by the Research and Development Unit, Department of Local Administration, Ministry of Interior (through http://www.dla.go.th/work/abt/index.jsp ). Politically speaking, decentralization in Thailand is unique due to the country’s indomitable centralization. Local governments in Thailand are significantly under the supervision of the Department of Local Administration (DOL A), Ministry of Interior (MOI). According to MOI’s ministerial regulation 2008, DOLA is responsible for promoting and supporting the work of local government s through the development and series of advices on the local development plans, personnel administration, finance, and administration in order to strengthen the capacity and efficiency of the local government s on public service provision. As a part of its responsibility regarding giving advice on finance and fiscal management, DOLA encourage s the local governments to adopt the performance management and budgeting. Every year, DOLA evaluates all the local administrative government s in four key administrative aspects—management, human resource management and council affairs, financial management, 46

PAGE 59

and public services in order to learn about the ir administrative strengths and weaknesses of local administrative organizations and encourage them to design the local development plan accordingly. In addition to DO LA’s supervision, deconcentration of power as well as limited administrative and fiscal decentralization are the main obstacles to local autonomy. The central government manages local government affairs by maintaining a powerful delegation mechanism in a deconcentrated system. Provincial governors and district government officials (both of whom are MOI officials) have the statutory authority to direct and order government officials from other ministries and departments at the provincial and district levels: the vertical relationship between ministries and department outweighs horizontal coordination among central governments offices at these subnational levels (Nagai, Funatsu, & Kagoya, 2008) . This powerful deconcentrated structure prevails even after the promulgation of decentralization policy in 1997. In every locality, there exists both deconcentrated administration of central government and local administration. This dual system of administration is considered a major c haracteristic of local administration in Thailand (Nagai, Funatsu, & Kagoya, 2008) . Local people elect their local representatives who later draft the four year local development plans — identifying local needs, framing local policies, and providing local public goods and services. However, these representatives do not have full administrative and fiscal autonomy. In term of administrative aspect, there were originally 245 items of local goods and services that the central ministries and departments planned to transfer to be local governments’ responsibilities by 2003. However, the central a uthorities only transferred 181 items to the local governments by 2011 (ThaiPublica, 2011) . Note that there is no progress on transferring local personnel management to local governments. Recruitment, promotion, and reassignment of high-47

PAGE 60

level local government officials (especially city managers) are still subject to DOLA’s review. Fundamentally, local governments’ four year local plans and annual budgets are subject to provincial governors’ approval. Local Governments’ Revenue and Expenditure Structure The Decentralization Plan and Process Act of 1999 commands that local governments share at least 20 and 35 percent of the central government’s total revenue in 2001 and 2005 respectively. The central government successfully gave such share s to local government in 2001 but this share grew very slowly and did not reach the 2005 goal. In subsequent years, as the central government realized that the goal in 2005 was unattainable, it revised the Decentralization Plan and Process Act of 1999, stating that it will be sharing 35 percent of its total revenue with local governments by 2016. However, the recent trend of the share of revenue between the local governments and the central government still suggests the very slow growth of the sha re of local government’s revenue. From 2011 to 2015, the average share of local government revenue to total revenue of all government is equal to 21.3 percent. Table 3.2 details the local governments’ revenue sources and compares the local governments’ tot al revenue to that of the central government. 48

PAGE 61

Table 3.2 Source of Local Governments’ Revenue (Unit: Million Baht) Source of Revenue Fiscal Year 2011 2012 2013 2014 2015 Own source revenue 38,745.96 46,529.72 50,281.54 56,306.25 61,458.00 Percent of local revenue 8.98 8.78 8.78 9.04 9.51 Shared tax (collected by the central government) 148,109.04 175,457.28 187,988.46 203,818.75 218,222.00 Percent of local revenue 34.34 33.11 32.83 32.74 33.76 Local tax (collected by the central government) 70,500.00 86,900.00 97,900.00 109,000.00 109,000.00 Percent of local revenue 16.35 16.40 17.10 17.51 16.86 Intergovernmental fiscal transfers 173,900.00 221,091.79 236,500.00 253,500.00 257,663.78 Percent of local revenue 40.32 41.72 41.30 40.71 39.86 Total Revenue of the Local Governments 431,255.00 529,978.79 572,670.00 622,625.00 646,343.78 Total Revenue of the Central Government 1,650,000.00 1,980,000.00 2,100,000.00 2,275,000.00 2,325,000.00 Percentage of local governments' revenue to total revenues of all governments 20.72 21.11 21.43 21.49 21.75 Source: Fiscal Policy Office, Ministry of Finance (though http://www.fpo.go.th/FPO/index2.php?mod=Content&file=contentview& contentID=CNT0006803&categoryID=C AT0001183 ). Note: The percentage of local revenue from shared tax is specifically fixed by such local government acts as the Provincial Administrative Organization Act 1997, the Municipality Act 1954, and Bangkok Metropolitan Administrative Organization Act 1985. Meanwhile, the intergovernmental fiscal transfers are dependent on the central government’s intergovernmental policy, which varies from year to year. It is also worth noting that the local governments’ own source revenue is only around 9 percent. Meanwhile, local governments depend on general grants from the central government, which consist of approximately 40 percent of the local governments’ total revenue between 2011 and 2015. In addition, a strong political connection with the central government is allegedly the main factor making a certain local government likely to receive relatively large and consistent grants from the central governments (Wongpredee & Sudhipongpracha, 2014) . 49

PAGE 62

On the expenditure side, the largest share of the local governments’ expenditure is general administration. The programs with the relatively high spending are education, which under the supervision of the education department , and housing and manufacturing and infrastructure, which are under the supervision of the public works department. It is also worth noting that even though social security policy is the central government’ s responsibility, it is carried out by the local governments. The central government calculates the number of disadvantaged people (the elderly, HIV AIDS patients, and the disabled) living in each jurisdiction and financially supports these groups of peopl e through the special transfers to local governments. Then the local governments distribute such benefits to the disadvantaged. Table 3.3 details all programs of local governments’ expenditure. 50

PAGE 63

Table 3.3 Local Governments’ Expenditure in 2012 (Unit: Million Baht)1 Expenditure Programs Provincial Administrativ e Organizations Municipalities Tambon Administrativ e Organizations Total Expenditure by Program General administration 6,203.31 (20.44%) 14,925.01 (27.02%) 12,845.15 (33.61%) 33,973.47 (27.44%) Security 126.68 (0.42%) 1,896.91 (3.43%) 506.93 (1.33%) 2,530.52 (2.04%) Education 4,174.79 (13.76%) 9,863.14 (17.86%) 5,223.14 (2.08%) 19,261.07 (15.56%) Public health 2,734.33 (9.01%) 2,733.81 (4.95%) 796.81 (1.79%) 6,264.95 (5.06%) Social welfare 391.70 (1.29%) 1,141.05 (2.07%) 684.29 (13.76%) 2,217.04 (1.79%) Civil Engineering 4,095.06 (13.49%) 11,370.10 (20.59%) 5,257.74 (13.76%) 20,722.90 (16.74%) Community building 1,191.73 (3.93%) 1,226.89 (2.22%) 535.61 (1.40%) 2,954.23 (2.39%) Religion, culture, and recreation 1,810.66 (5.97%) 1,190.53 (2.16%) 852.26 (2.23%) 3,853.45 (3.11%) Manufacturing and infrastructure 4,541.40 (14.96%) 2,003.19 (3.63%) 4,969.82 (13.00%) 11,514.42 (9.30%) Agriculture 1,256.83 (4.14%) 145.10 (0.26%) 363.84 (0.95%) 1,765.77 (1.43%) Commerce 1,240.92 (4.09%) 266.95 (0.48%) 355.38 (0.93%) 1,863.25 (1.51%) Central budget2 2,580.68 (8.50%) 8,466.07 (15.33%) 5,830.64 (15.25%) 16,877.39 (13.63%) Total Expenditure 30,348.12 55,228.74 38,221.60 123,798.46 Source: Center for Local Initiatives and Governance, Chulalongkorn University (Forthcoming). Local Governments’ Fiscal Data in Thailand: How the Central Government’s Policy Affect Their Local Governments’ Fiscal Conditions. Bangkok: King Prajadipok’s Institute. Note: 1. The fiscal data included in this table are from 2,560 local governments (47 Provinc ial Administrative Organizations, 1,101 Municipalities, and 1,512 Tambon Administrative Organizations) or approximately 32.60 percent of all local governments in Thailand. 2. Central budget includes such expense as debt service, emergency funds, infrastructure maintenance, and social security for the disadvantaged especially according to the central government’s policy. Local Budget Process A particular local budget system emphatically reflects its locality’s historical traditions and diversity in culture, capacity, and political and administrative institutions (Mullins, 2007) . Thail and’s local governments are no excep tion. Their local budget system, called the strategic 51

PAGE 64

performance based budget system , depicts the country’s distinctive assemblage of fiscal decentralization, performance management, civic participation and intergovernm ental relations. Broadly, their medium term expenditure frameworks and annual budgets are well established within the local governments by local councils and local administrators. The annual budget calendar of the local governments follows that of the cent ral government, running from the beginning of October to the end of September. This study delineates this local budget process via the budget development cycle which is divide d into three stages—sequential budget planning, budget formation, and budget a ppr oval. Each of these possesses numerous policy and administrative activities as follows. Budget planning. Regarding budget planning or budget pre preparation (Mullins, 2007) , firstly, the local governments start their budget planning as early as January and continue through approximately June, soliciting and gathering community needs and problems, policy priorities and expectations as well as challenges and opportunities of the current local public service provisions. During this stage, views and opinions from any parties are highly welcome through many approaches of civic participation. Three approaches widely adopted include focus group consultation, civic forum, and citizen surveys (Lorsuwannarat, 2017; Suwanmala, 2007) . Focus group consultation begins with the mayors and local administrators introducing a particular plan for a certain policy issue to their community members (especially the concerned target groups). Via a series of focus group meetings, local indi viduals are given opportunity not only to make comments or recommendations but also to determine whether the plan corresponds to their needs and preferences. It is expected that this approach of focus group consultation can promote the citizen’s sense of o wnership in community development. 52

PAGE 65

Civic forums are particularly popular among relatively small local governments in the rural areas. This civic participation approach features a civic forum committee which formally comprises several groups of stakeholder s including village heads, community leaders, representatives of occupational groups, women and youth leaders, spiritual and religious leaders, school teachers, and provincial or local senior administrators from the central government’s agencies. Regularly , their meetings involve identifying and prioritizing community problems, needs, and preferences and then initiating local development programs or activities. Note that both policy initiations and functional programs from civic forums are usually highly re spected by the local councils. Citizen surveys inquire about the satisfaction levels of local public services and the provision of traditional or religious activities (i.e., long holiday festivals and celebration events). Local public service users and fe stival or event participants are actively asked about their satisfaction and suggestions , which are indispensable for making such activities more effective— culturally attractive and cost saving. It is worth noting that this is one of the crucial parts of performance management extensively promoted by DOLA following the national agenda of budget reform (Sutheewasinnon et al., 2016) . Secondly, the mayors together with the local administrators (i.e., municipal clerks, finance directors as well as directors of spending agencies) formulate the local development schemes based largely on the community needs and problems, policy priorities and expectations, and challenges and opportunities from the aforementioned focus group consultation, civic forum, and citizen surveys. Specifically, each local government sets overall policy platforms by identifying broad goals and policy outcomes or results; developing preliminary programmatic, 53

PAGE 66

operating, and capital plans; and contemplating the realistic budget ceilings for functional allocations, carefully keep ing in mind their revenue and expenditure statistics and forecasts. Thirdly, progressing their es tablished policy platform to a strategic performancebased plan, the local executives, the local administrators, and the local development plan committee (whose members are mostly fro m the civic forum committee) thoroughly analyze the whole community’s strengths, weaknesses, opportunity, and threats. Broad goals and policy outcomes or results and developing preliminary programmatic, operating, and capital plans are prioritized, sequenc ed, and packaged according to the four year timeframe. Simultaneously, they also often ponder the national policy agendas, the regional development scheme, and the provincial plans in order to ensure policy harmonization across levels of governments. Event ually they officially offer the four year local strategic plan which reveals the locality’s policy outlook, overall development goals, programmatic priorities, and fiscal environment (either fiscal restraints and expectations for budget reductions or oppor tunities for an expansion of tax and nontax revenue sources). Budget formation. The budget formation stage of local budget process takes three months, specifically from June to August. It generally begins when the local executives and the local administrators prepare a draft of the annual budget . They painstakingly review the estimated revenue and expendi ture, financial statements, agency budget requests, and performance indicators as required by DOLA’s guidelines on managerial and service delivery standards. Note that, in order to gain efficiency and effectiveness, they are advised that they review those documents along programmatic, technical, and managerial dimensions (Mullins, 2007) . For programmatic dimension, the budget proposal must be consistent with the policy platform and 54

PAGE 67

priority highlighted in the four year local strategic plan. Meanwhile, the technical review ensures that all of the required estimates are accurate and consistent with historic spending as well as sp ending envelopes. Managerial review guarantees that local administrators adhere to administrative and legal requirements, following the designated mechanisms, systems, process, procedures, and guidelines for conducting administrative and programmatic opera tions. After that, there exists the final review of the annual budget proposals , which involves an executive budget development process (Mullins, 2007) . A mayor and his or her advisory board deliberate spending requests in terms of the trade offs of relative costs and outcomes across spending units and program objectives and the trade offs of spending priorities and managerial operations. Then, a sharp or mild reduction may possibly apply to some spending requests. Nevertheless, this review results in compilation of spending requests into a package of the locality’s working budget proposal. Prior to the middle of August, the ma yor submits this budget proposal to his or her local council. Budget approval. This local budget stage of legislative review and approval proceeds from the middle of August to the end of September. Regulatorily, dominating the budget approval stage, a local council (an elected legislative body) is designated to exercise its amendatory power —an authority to substantially or moderately alter the budget proposals. The council first examines the overall policy agendas and key fiscal aspects. Next, the comprehensive review is conducted usually by the legislative budget committee—a group of budget specialists who provide assistance to the council especiall y in such tasks as deliberation of policy outlook and priority, reviews aggregate revenues and expenditures, a comparison of programmatic alternatives, and a consideration of spending at the functional and organizational levels. Due to the fact that informed review is inevitably information intensive, the council as well as this 55

PAGE 68

legislative budget committee have a uthority to collect information and to request testimony of both mayors and administrative officials on the merits of all elements of the proposed budgets (Mullins, 2007) . After that the council votes to approve the budget proposals according to majority rule. Lastly, provincial governors or chiefs of district administration finalize and verify these budget proposals. Information Use in Local Budgeting in Thailand Based on the local budgeting process described in the last section, it can be said that local budget actors use information for their decision making. Particularly, after gathering such information as communit y needs and problems, policy priorities and expectations, and challenges and opportunities of the current local public service provisions from focus group consultation, civic forum, and citizen surveys, local budget actors can deliberate such information during the budget formation and the budget approval. Throughout the budget formation stage, a chief executive and local administrators prepare the proposal by considering this information against technical and managerial conditions in order to formulate a f our year local strategic plan. Eventually, during the budget approval stage, local legislators and possibly budget specialists review such proposed plan by inquiring more information from local administrators . They mainly depend on information regarding a projection and a profile of previous year spending of each programmatic budget so as to approve the budget proposals with the majority rule. In conclusion, Thailand’s local budgeting process offers the poli cy arena which local budget actors can deliberate information in order to make budgetary decisions. However, with limited administrative capacity , at first glance it is unlikely that local administrators have enough access to information and thereby they c annot use such information to enhance their budget 56

PAGE 69

allocation. Accordingly, it is important to test this theoretical linkage based on information economics and budget rationality because if the result uphold it, there is a strong empirical evidence to vali date related theories. In addition, due to the fact that there are few studies on use of information and budget decisions in developing countries, this study applies mixed methods to capture complexity of a developing country’s local budgeting a s well as t o test the aforementioned theoretical linkage . Chapter 4 provides the methodological details of this study highlighting the pilot study, the quantitative component, the qualitative component, and finally the triangulation of research results. 57

PAGE 70

CHAPTER IV RESEARCH DESIGN AND METHODOLOGY As formulated in C hapter 2, the overarching research question for this study is: How do policymakers with different administrative capacities utilize fiscal information during a budget process in order to have an impact on budget decisions? The focus of this study is to understand the use of information and its influence on budget decisions —the complex phenomenon which is fundamentally subjective, political, multiple, and constructive through administrative inte ractions during a budgeting process (Feldman & March, 1981; Wildavsky, 1978) . This study, thus, applies a mixed methods design to explore and explain the complex interrelationship between the administrative capacity, information characteristics, information utilization, and budget decisions (Newman, Ridenour, Newman, & DeMarco, 2003) . Conducting a mixed methods research project involves combining quantitative and qualitative approaches in the same study or, in other words, collecting, analyzing, and interpreting both quantitative and qualitative data in a single study or in a series of studies that explore the same phenomenon (Collins, Onwuegbuzie, & Sutton, 2006; Onwuegbuzie & Leech, 2006) . The logic behind the mixed methods research includes the use of induction (or discovery of patterns), deduction (testing of theories and hypotheses), and abduction (uncovering and relying on the best of a set of explanations for understanding one’s results) (Johnson & Onweugbuzie, 2004) . To explore the nature of how information is used in the context of developing counties which is still understudied, this study first introduces a pilot study in order to guide and develop the research plan (Kim, 2011; Teijlingen & Hundley, 2002) . This pilot study is of qualitative approach so as to provide a causal description on the interrelation of concepts of interest (Miles, Huberman, & Saldana, 2014) . Specifically, the pilot study aims to improve content validity by 58

PAGE 71

assessing whether the study’s questionnaire items (the survey instrument) as well as interview questions to ensure that the research measurements truly and adequately represent the study ’s concepts and constructs discussed in Chapter 2. Once the pilot study verifies the questionnaire items and the interview questions, this study introduces the full scale research with the quantitative mixed methods research design (Morse, 1 997) . It first implements the quantitative approach with survey as the mode of data collection. After analyzing the quantitative results, this study implements its qualitative approach with interview as the mode of data collection primarily in order to complement the quantitative results. To elucidate this research design, this chapter first elaborates the methods design. Then, it details the sampling design, instrumentations, and data analyses for the quantitative and qualitative approaches, respectiv ely. Note that both of the quantitative and qualitative approaches are largely based on the results from the pilot study (i.e., operationalization, data collection instruments, and analytical techniques). After that, this chapter offers how the results of these two approaches are triangulated. Finally, it presents threats to legitimation given the type of mixed method design proposed by this study. The quantitative mixed methods design As Chapter 2 demonstrates, this study locates substantial and relevant literature on the topic, creates a theoretical framework, and identifies the testable hypotheses. At the core of this study based on the philosophy of science is, therefore, post positivism which features a hypotheticodeductive model and a confirmatory r esearch (Morse, 1997; Teddlie & Tas hakkori, 2009) . Fundamentally, this study warrants a quantitative inquiry. However, in order to provide a holistic description of the information use and how it influences the budget allocation, a qualitative approach is also necessary for elaborating t he phenomenon of interest, especially 59

PAGE 72

descriptively complementing the key findings based on the precedentially conducted quantitative phase. In short, as a result, this study applies the partially mixed sequential dominant status design with the greater em phasis on the quantitative phase which is for short written quantitative mixed methods (Leech & Onwuegbuzie, 2007; Newman et al., 2003) . The following sections delineate the methodological components of quantitative and qualitative phases, respectively. The Quantitative Approach In order to collect data and then empirically describe and explain several concepts and the interrelationships among them graphically depicted in the theoretical framework in Chapter 2, this quantitative phase applies a survey research with a systematic questionnaire asking prescribed questions of respondents (Singleton & Straits, 2010) . The responders’ answers are later numerically coded and analyzed via multiple regression analyses to estimate, ceteris paribus , the effects of causal factors (i.e., administra tive capacity and reliability and relevance of information) on outcome variables (i.e., the use of information and budget allocations) (Wooldridge, 2010) . The following parts details this survey research which constitutes the quantitative approach. Sampling Design Population. As mentioned, local governments in Thailand represent the lesslikely cases, so the population in this research is Thailand’s local governments (Gerring, 2007b, 2012) . However, this research first excludes from the sample pool Bangkok Metropolitan Authority (BMA), Pattaya City, the Provincial Administrative Organizations (PAOs), and the Tambon Administrative Organizations (TAOs). This is because BMA and Pattaya Cit y are special local governments which possess different organizational and political structures from other local governments. Meanwhile, PAOs are mainly responsible for financially assisting the other local 60

PAGE 73

governments in their public service delivery and TAOs do not carry out the key functional programs (i.e., park and recreation, sanitation, education, health, and transportation) according to the PAOs Act 1997 and the TAOs Act 1994 respectively. After eliminating local governments based on the extraneous variables (the organizational structure and the scope of functional programs), 2,441 municipalities are this study’s sample population— the evidence that will be subjected to direct examination (Gerring, 2012) . Sampling scheme. Like other quantitative studies, this study aims to make statistical generalizations —generalizing findings and inferences from a representative statistical sample (i.e., municipalities in Thailand) to the population from which the sample was drawn (all the municipalities in Thailand) ( Onwuegbuzie, Onwuegbuzie, Collins, & Collins, 2007) . Sample size. Given that the unknown percentage of the policymakers in Thai municipalities who deliberate and actually utilize relevant and reliable information for their budget proposals and discussion as well as the unknown standard deviation of such information util ization, this study applies the Cochran formula to calculating the sample size. Specifically, this study sets the confidence level at 95% (1.96 for its Z values) with the level of precision at least at 5 percent. Additionally, it assumes that half of polic ymakers use performance information during the budget process. As a result, after adjusting for the target population of 2,441 municipalities, the calculation suggests that the sample size of 333 municipal i ties be adequate to give the confidence levels the study requires for the budget allocation. In order to maximize a response rate, a voluntary response sample is applied. It is possible that the size of municipalities in Thailand (small, medium, or large) deter mines the extraneous factors— fo r instance, the organizational structure, the scope of functional programs and legal mandates (World Bank, 2012) , and longterm debt level and debt 61

PAGE 74

service ratio (Krueathep, Nudam, & Te rmtawewut, 2015) — which may affect the relationships of interest. This study, therefore, adopts the proportionate stratified sampling method in order to ensure that each of three groups of these municipalities is adequately represented within the whole s ample population of the study (Teddlie & Yu, 2007) . Table 4.1 ca lculates the sample size for each stratum , resulting in the sample selection of three large, twenty five medium, and 305 small municipalities . Table 4.1 The Stratified Samp ling and Sample Size for the Survey Municipality’s Type (Size) Number of Municipalities in Stratum Stratum Sample Size Nakon or Metropolitan (Large) 30 3 Meung or City (Medium) 178 2 5 Tambon or Sub District (Small) 2,233 305 Total 333 Cases. Cases are a restricted range of units that are bounded in terms of settings (e.g., a particular place and point in time that the research is conducted) and constructs (e.g., a few possible variables from the theories on which the operational research model is based) (Gerring, 2012; Van de Ven, 2007) . Fro m the temporal perspective, the literature asserts that it takes at least a few years for the use of performance information to become established and eventually have an impact on budget allocations (Gilmour & Lewis, 2006; Melkers & Willoughby, 2001; Zaltsman, 2009) . Furthermore, the impact of performance information on the budget allocation is more pronounced during the boom than the bust years (Hou, Lunsford, Sides, & Jones, 2011b) . Since the Department of Local Administration (DOLA), Ministry of Interior has required, since 2006, local budget officials to collect and present related fiscal information and statistics to the local policymakers (Article 3, Section 22 of Ministry of Interior’s Regulation on Local Budgetary Process, 2000) (NESCB & Thamasat University, 2009; World Bank, 2012) , this research can expect the see such impact anytime approximately after 2003. T his study uses Thai 62

PAGE 75

local government dat a from the most recent budget year of 2019—from O ctober 2018 to September 2019—because the budget actors are likely to accurately recall their experience on information use throughout the budget process and such period of time represents the country’s economic growth trend, approximately 2.64 percent per year (Asian Development Bank, 2018) . From the construct bounded perspective, the principal agent theory suggests researchers capture the change of annual budgets in specific programmatic functions —a designated balance, the portion of the government expenditure that is restricted, committed, assigned, reserved, or designated under existing laws for specific uses and purposes by the current administration (Hou, 2003; Tomes, Berger, & Bassett, 2011) . Empirically, the performance budgeting scholars confirm that there are certain situations when the policymakers or budget actors can drive changes in their programmatic budgets and improve the operation and outcome of such programs in general through the use of performance information (Hou et al., 2011b) . Under the Thai local governments’ context, the key functional programs based on the composi tion of spending levels in Table 3.3 include civil engineering (16.74%) and education (15.56%) . This study therefore concentrates on the budget process of these two functional programs . Note that the department of education is responsible for education programs meanwhile the department of public works are responsible for civil engineering programs . Unit of analysis and observation. The unit of analysis, the entities which are being described and generalized (Singleton & Straits, 2010), in this study is both the individual and the organization. As information utilization is a result of a cognitive process of each policymaker, the sub question of how policymakers or budget actors with different administrative capacities utilize relevant and reliable information during a budget process uses the individual as a uni t of 63

PAGE 76

analysis. However, there are three main groups of budget actors in the municipalities in Thailand who are directly and routinely involve d in budget preparation, budget debates, and budget decision making—city mayors, municipal clerks, and the departme nt heads. The key budget actors this study is interested are the budget actors who (1) possess a variety of information, (2) utilize such information throughout the budget process, and (3) have impact on budget allocation. Even though both city mayors and municipal clerks officially have decision making authority, they exercise such aut hority only on strategic plans. The functional plans are largely under the department’s consideration. The departments are legally responsible for thoroughly preparing budget proposals, presenting, and defending them during the budget debates. This study, therefore, consider s the department heads as the key budget actors and individuals of interest. Meanwhile, budget dec isions or budget allocations for the functional programs are a result of a collective process where budget actors altogether distribute their financial resources to spending programs according to their organization’s policy directions and priorities. There fore, the subquestion of how information utilization influence s budget decision s uses the organization as a unit of analysis. This study also asks the department heads to provide their organizations’ budget allocation. Table 4.2 sums up the sampling de sign —population, sample, cases, and observations — and the example of principal concepts according to the theoretical framework operationalized in the following part. 64

PAGE 77

Table 4.2 Sampling Design Population 1 Sample Case Observat ion 2 Local governments in Thailand (7,852 Units) Municipalities (2,441 Units) Municipality 1 x 1,1,1 x 2,1,1 x 1,1,2 x 2,1,2 y 1,1,1 y 2,1,1 y 1,1,2 y 2,1,2 Municipality 2 x 1,2,1 x 2,2,1 x 1,2,2 x 2,2,2 y 1,2,1 y 2,2,1 y 1,2,2 y 2,2,2 Municipality 3 x 1,3,1 x 2,3,1 x 1,3,2 x 2, 3 ,2 y 1,3,1 y 2,3,1 y 1,3,2 y 2, 3 ,2 Municipality 333 x 1,333,1 x 2,333,1 x 1,333,2 x 2,333,2 y 1,333,1 y 2,333,1 y 1,333,2 y 2,333,2 1 Number of total local governments and municipality in Thailand is from the DOLA, Ministry of Interior retrieved from http://www.dla.go.th/work/abt/index.jsp on March 24, 2018. 2 For each observation: x is variable of interest ; y is dependent variable ; i is variable 1 or 2; j is municipality 1 through 333; and k represents the department when 1 is the education department and 2 is the public works department . Note that c oncepts of interest in this study include administrative capacity , budget rationality, information use , and actual and perceptional budget allocations. Some of these concepts are operationalized and verified by the pilot study which is the center of discussion in the next se ction. Operationalizing Concepts of Interest This section reconceptualizes the constructs introduced in Chapter 2 and provides their operational measures. The key concepts for this study cover administrative capacity, information production, information utilization, and budget decisions . Note that the pilot study is conducted in order to operationalize these key concepts and specify their measures which fit the context of Thailand’s local budgeting. The following parts explain why and how the pilot study is administered. Pilot Study A pilot study is a miniature version of full scaled research which features planned methods in order to identify and resolve epistemological and methodological issues as well as to guide and develop the research plan (Kim, 2011; Teijlingen & Hundley, 2002) . Even though 65

PAGE 78

conducting a pilot study does not guarantee success in the full scaled research, it does increase the likelihood of such success via enhancing the research’s validity and reliability (Barriball & While, 1994; Teijlingen & Hundley, 2002) . The following parts describe goals, how to administer, and the results of the pilot study. Goals Reliability. T his pilot study impr oves content validity by assessing whether the measures of the related concepts or the specific survey and interview questions as the research measurements truly and adequately represent the study’s concepts. In addition, a pilot study can increase interna l validity, or specifically population validity, since it helps recruit key participants from the whole population. (Collins et al., 2006; Onwuegbuzie, 2003) . In other words, with conditionseeking methods, the study seeks to discover which group of local budget actors both comprehensively utilize information and potentially influence their organization budget decisions. Validity. This pilot study promotes instrument fidelity (Collins et al., 2006; Onwuegbuzie et al., 2007) . It provides the study with an opportunity to assess the appropriateness and utility of existing survey questions and interview protocol with a view to making modifications or creating and improving new instruments so that any differences in the interview answers are due to differences among the interviewees rather than in the questions asked (Barriball & While, 1994; Collins et al., 2006) . Particularly, all questions should conform to the principles of specification (the focus of each question), division (the questions’ appropriate sequence and wording) and tacit assumption (the true meanings behind interviewee’s answers) (Lazarsfeld, 1935) . 66

PAGE 79

In short, in order to enhance construct validity, th e pilot study tests whether each survey and interview question as a measurement actually focuses on and reflect related true meaning of concepts or constructs through the int erviewees’ viewpoint . It also verifies whether interview protocol’ s and questionnaire’s sequences and wordings are appropriate so as to increase the instruments’ reliability. Then it identifies the key budget actors from a variety of groups of individuals working in local governments in order to promote population validity. How to Administer the Pilot Study The pilot study is consisted of t hree sequential phases. The first phase aims at understanding the roles of each group of budget actors on information collection and utilization and operationalizing the key concepts of administrative capacity, information production, and information use. Three municipalities were selected on the basis of convenience and chain or snowball sampling strategies in order to identify cases that offer rich information enough to operationalize the concepts of interest (Miles & Huberman, 1994) . This study thus interviewed key budgetary actors (mayors, municipal clerks, and department heads) from these three municipalities. Totally, there we re twenty individuals for the pilot interview s (three mayors, three municipal clerks, and fourteen department heads). The second phase of the pilot study focuse d on testing and verifying the interview protocol. The participants are the same as the first phase. The interview questions which had been derived from the literature are used and then revised based on the participants’ understanding of those questions as well as their recommendations. Table 4.3 shows the list of interview protocol being verified in this pilot study. 67

PAGE 80

Table 4.3 Interview Protocol Based on the Literature Interview Approach Construct of Interest Question 1. Unstructured conversational interview approach Tenure and years in supervisor position Budgetary orientation What is your educational and professional background? How long have you been working in this organization? Are you engaged in budgeting? What is your budgetary responsibilities? What is your role during the budget process? How do yo u interact with other groups of budget actors? 2. Interview guide approach Budgetary orientation Structural isomorphism Budget incrementalism Information utilization Can you prioritize these following competing spending preferences: financial control, efficiency, management techniques, policy issues, and executive’s plan? Why? Which budgeting system does your organization adopt? Why such system is adopted? What are managerial structures , standard of procedures, and information such system requi res? How has your organization’s budget process changed over the last 4 years or until the recent budget system has been in place? Currently, how does a budget process in your organization usually proceed? What are the sequence and the key activities in s uch process? On which activity the budget committee spend quite a long time and why? 3. Highly structured interview approach Information provider Information characteristics Information utilization Budget decisions What kind of information do you usually receive from the departments? Does your organization have an official unit providing information for policymakers? What kind of information do they provide? How do they get such information? Which types of information you find useful (important or adaptable)? Why? Which types of information you find reliable? Why? Which piece of information do you usually think of and bring up during the budget process? Which piece of information do you base your budget decisions on? What factors carry the mos t weight in the definition of your organization’s budget decisions? How do such factors impact the spending priorities, eliminating the duplicative programs, or cost -saving? If there are differences in allocation priorities how does the committee resolve t hem? Which spending program or functions usually brings such differences? 4. Non controversial wrap -up approach Information Production What are your plans on information management system? Which information should be additionally collected? How does this additional information enhance the current budget process? How can it ultimately benefit local people? 68

PAGE 81

The third phase involved the survey questions, which were partially based on the results of the first phase. For each concept, t he questionnaire had multiple item measures with Likert type scales. The study applies the convenience sampling strategy to select forty department heads to participat e in the pretest this questionnaire. In order to measure the internal consistency or reliability of this survey instrument, the pilot study calculated the Cronbach’s Alpha. Since this study is an exploratory research based on its new measures proposed by the results of the first phase above, a Cronbach’s Alpha value of 0.70 is acceptable (Campbell & Fiske, 1959; Leech, Barr ett, & Morgan, 2011) . Note that the values of Cronbach’s Alpha are reported in the descriptive statistics section in Chapter 5. The Pilot Study’s Results Participants. Even though the chief executives are obviously key budget decision makers, they do not possess a variety of types of information. Their budget decisions are often based on communal problems and demands. Somewhat like city mayors, municipal clerks only util ize communal problems and demands and local strategic plans (i.e., policy and priority) as their key information for formulating budget proposals. However, during the budget debate, a chief executive and a municipal clerk have department heads present, explain, or even defend the budget proposals. This is because department heads from each department possess technical information specific to their functions or policy areas. This is necessary when the study examines the relationship between the informatio n and the budget allocation. In addition, there are a couple of functional departments (i.e., departments of education and public works) which possibly receive information from several information sources with different levels of relevance and reliability. The more relevant 69

PAGE 82

and reliable the information perceived by policymakers, the more likely it is utilized during a budget process. Administrative capacity. The literature operationalizes administrative capacity not by capacity itself but by its proxy of p rofessional qualifications such as educational background and work experiences (Potter & Brough, 2004; Tamkin, Hillage, & Willison, 2002) . There are also other components of administrative capacity including, for instance, skill s and knowledge that are less investigated in the literature . This study, thus, chooses to explore these components of administrative capacity. Particularly in this context , skills and knowledge are necessary for budget actors to initiate, draft, and defend budget , or ca lled capacity to budget in this study. The budget actors in the pilot study from the three municipalities —three mayors, three municipal clerks, and fourteen department heads , in total twenty individuals—identify key components of capacity to budget . After applying the constant comparative method, this study finds that there are six kinds of administrative capacities necessary for budget actors as follows. (1) Participatory budgeting —ability to solicit citizen opinions, promot e local participation and model pre ference s . (2) Communication skills — ability to prepare data presentation and express and discuss opinions . (3) Intergovernmental relations —ability to network and collaborat e with other local governments, the central government, and its provincial agencies . (4) Strategic management —ability to structure problem s and strateg ies based on the community’s strengths, weaknesses, opportunities, threats . (5) Project managemen t —ability to assess project feasibility, monitor and evaluat e project , and conduct stakeholder analy sis. 70

PAGE 83

(6) Accounting and financial management —ability to apply budgetary techniques, conduct cost benefit analysis, calculate and interpret financial ratios, and apply risk management . Information production. Information production encompasses available inform ation and information characteristics (reliability and relevance). The budget actors in the pilot study identify two main types of information—performance information and nonperformance information. Performance information. Performance information includes the value of money, service quality, and policy results. (1) Budget actors see the value of money for a specific project or program through the unit costs (based on outputs and a standard price or a surveyed price) and the service quality. Output of local public goods and services are, for example, the length of local roads constructed, the number of patients treated by the municipal health centers, the primary school graduation rate, the amount of sewage treated, and the number of participants in each of local festivities. The budget actors compare such outputs with cost, either their total costs or unit costs based largely on the standard price given by the Ministry of Commerce or the price survey of that each department conducts at least three local stores or merchandisers. (2) Service quality, for instant, includes the rate of middle school students successfully accepted into the leading high schools, the sports trophies the municipalities’ school win throughout the year, and the reduced rate of rabie s fatalities compared to the previous years. 71

PAGE 84

(3) Policy results are positive impacts initiated or stimulated by the municipalities’ projects. The case in point is that local roads can improve the road transportation between different communities so that it is easier for the residents to have more social and political gatherings as well as to have more economic and trading activities , such as market fai rs, farmer markets, and banking mobile services. Non performance information. The nonperformance information includes spending ceilings, executives’ policies and priorities, community’s problems and demands, and professional standards and knowledge. (1) Spending ceilings are the programmatical budgetary limits each department receives for each fiscal year. These ceilings are based mainly on revenue forecasts or the previous year amounts and the change in municipal revenue which is the estimated summation of intergovernmental transfers (i.e., grants, shared taxes, and local taxes collected by the central government’s agencies) and ownsource revenues (i.e., property taxes). (2) The executives’ policies and priorities include the flagship policies, town initiatives, as well as development directions promoted through their electoral campaigns. For example, some municipalit ies have a city wide plan to push their communities to be tourist destinations while others have a smaller plan to reduce use of the plastic and increase the recycling activities. Note that the executives as politicians consider these policies and prioriti es the promises they give and hold themselves responsible . Consequently, these polities and priorities are politically sensitive especially if they were left unimplemented. 72

PAGE 85

(3) Communities’ problems and demands are usually local issues or agendas raised direc tly by the residents or sometimes by the community leaders through town meetings. For example, the residents request the municipalities to install a lighting safety system in their communities, to construct or renovate the local market buildings and facili ties, or to financially support the local craft groups in order that their handmade goods can attract more customers especially from outside the communities. However, communities’ problems also mean unfortunate events or emergencies (for example, floods, droughts, fires, and epidemics) to which municipalities must react fast. (4) Professional standard and knowledge are technical or specialized information important or necessary for initiating as well as implementing a certain project. This type of information becomes prominent when local residents can only identify the problematic situations but cannot suggest any possible solutions. This is because certain problems require sophisticated technology or specific technicality to analyze and come up with technical recommendations. For example, in order to have a sustainable solution for flooded areas, the municipality’s department of infrastructure must design a watershed flood control dam and present it in form of a special project to the town meeting. Informati on characteristics. Before deliberating how to use the information mentioned above, the budget actors prefer reliable and releva nt information. For reliability, the information should be verified, valid, and updated (Lundin & berg, 2014) . Specifically, the information should be from the source that should be of integrity and benevolence, have good command of data gathering, processing, and analysis. The budget actors usually gauge the reliability of 73

PAGE 86

information by its provider or source , which include s self observation, performance contracts and reports, the municipal information centers, the central government’s agencies, other local governments, the municipality’s contractors or service providers, and other professional and academic sour ces such as research institutes, universities, and think tanks. Meanwhile, the budget actors measure information relevance in terms of its usefulness (Wolman & Page, 2002) and applicability (Landry, Lamari, & Amara, 2003) . Information utilization . Based on the pilot study, there are two perspectives into the informat ion utilization. First, the pilot study results point to the information utilization for decision making, which reflects the active or purposeful information utilization during the budget process (Donald P. Moynihan, Lavertu, et al., 2012; Taylor, 2007) . The information use reflects the degree to which the budget actors use the information to (1) influence their budget decisions especially in terms of program prioritization ; (2) combine with other data to reach budget decisions ; (3) prompt further budgetary questions or enquires during budget discussion or debate ; (4) use those information to support a budget decision reached by other means (e.g., professional experience) ; and (5) argue that a certain budget choice may not be desirable. The second perspective of informa tion utilization that the municipal budget actors widely discussed during the pilot study relate to the information presentation. They state d that they bundle some or all the information mentioned in the previous section and present them through story telling strategies. Even though the empirical studies in public policy widely discusses these strategies, the local budgeting literature especially in Thailand limitedly investigates them. This study groups these strategies of information use under the context of Thailand’s local budgeting into five story telling schemes as follows. 74

PAGE 87

(1) “ Doctrine of performance management ” : Budget actors try to prepare a budget proposal with a rational decisionmaking process. They specify an area of inefficienc y or ineffect iveness, then offer the projects as a solving mechanism. This mechanism th r ough budget allocation does not only fix those inefficiencies and ineffectiveness but also benefits a variety of target groups (both citizens and local government alike) in term s of better quality of service, outcomes, efficiency, and effectiveness (Donald P. Moynihan, 2008b; Mucha, 2009) . For example, an education department head observes the decreasing trend of average mathematics score of the primary school students. The mathematics tutor and related activities are bundled into the mathematical achievement program. This department head shows the budget committee how these tutor activities can generally enhance the students’ mathematical skill and possibly increase the pass rate. Meanwhile, some students with a flair for mathematics also can benefit from this project in the way that they may have a better chance in winning prize s at the provincial or regional mathematics competitions. In addition, the parents are probably mo re satisfied with the educational policy directions of the school which is led by the municipality’s executives. (2) “We are the experts”: A budget proposal is possibly a complex collection of ideas and possibilities (i.e., technical feasibility, affordab ility, and value acceptability) so that budget actors present it by comparing its components with professional or expert standards (Jones et al., 2015; Mikesell, 2007; Zahariadis, 2014) . This expert led scheme of information utilization emphasizes the important consideration based on the information collection methods, standards of evidence, and analytical techniques of scientific knowledge to guide local development initiatives or projects (Cor burn, 2003; Failing et al., 2007) . For example, an infrastructure department head proposes the bridge construction to his or her municipal budget committee by introducing the 75

PAGE 88

design plan including such details as width, length, height, materials, constr uction areas, and costs. Demographics (i.e., population and local economic activities), spatial conditions, disaster speculation and statistics are also considered. T he department heads may also mention its pros and cons in comparison to other designs. (3) “Framing”: Based largely on a particular community’s problems, some budget actors predict and present the focusing events (events emphasizing the occurrence of natural or manmade crises or disasters) and their dire consequences in terms of indicators and conditions which already happened or would possibly happen if the proposed projects were not funded (Liu, Lindquist, Vedlitz, & Vincent, 2010; Mikesell, 2007; Zahariadis, 2014) . Indicators are used to assess the existence and magnitude of a condition and the scope of change made by a community’s problem itself and sometimes they are used politically to measure the magnitude of change in the hope of catching official s’ attention (Zahariadis, 2014) . Conditions of a focusing event direct official s’ attention to specific evaluative dimensions of a particular community’s problems (Zahariadis, 2014) . This utilization of problems’ indicators, conditions, and focusing events can at least create awareness or trigger widespread attention, a shared concern that some action is required, or a shared perception that the matter falls within the bounds of municipality’s authority (Cobb & Elder, 1972) . For example, the watercourse maintenance project has been postponed for a few years. Yet , heavy rainfalls during monsoon season make a river overflow its banks and caus e prolonged flooding in some areas in the municipality. The municipal clerk and the infrastructure department head survey the affected neighborhoods and estimate the flood damage. While proposing the watercourse maintenance project, they exhibit the detrim ental impact on property (i.e., amount of lost residential buildings, industrial buildings, flooded crops, and increase in animal and livestock health problems) and on people (i.e., the number of missing, 76

PAGE 89

injured, and drowned). They also emphasize that thi s calamity requires a quick response from the municipality , which is directly and wholly responsible for the well being of residents in the jurisdiction. (4) “Mandates”: Budget actors state that the projects they propose are a part of or required by manda tes directly or indirectly enforced by the state statutes and the central government (Mikesel l, 2007) . Responsibilities for making allocation decisions and actually delivering services as a part of implementing the central government’s policies have been delegated to local governments even though those services are only partially or even not at all funded by the central government’s budgets (Radin, 2000) . In order to fulfil these responsibilities, policymakers at the local level usually follow the central government’s policy directions. Budget actors, while discussing their budget proposals, elaborate how the services and policy results of such proposals can serve or promote the policies or the statutes initiated by the central government and i ts agencies. For example, DOLA issues the official letter to request the municipalities to actively take good care of the elderly as the central government prepares the country for an aging society and economy. In order for a budget proposal of setting up a school for the senior citizens to get approved, the social welfare department heads convince their policymakers by showing that the school’s curriculum is in line with the Second National Plan for Older Person 20022021 and that the policy results of sen ior schools match with the central government’s aging society and economy platform —the enhancement of recognition of the elderly, the dignity of old people, and the improvement of their quality of life. (5) “Incrementalism”: Budget actors present and defend their budget proposals by arguing that their proposal is not significantly different from the previous year or insisting that “it is an old stuff” or “it’s so small” (Mikesell, 2007) . As long as the spending ceilings allows, 77

PAGE 90

the department heads show their policymakers that there is no significant change. This can keep the current beneficiar ies satisfied and the political climate unchanged. The current executives— the incumbents in the next local election—still have electoral advantage. However, some small changes are possible. For example, since the Thai new year festivity this year draws mor e local attendants than the previous year, the education department head may ask for additional budget to accommodate more people. Budget Decision . The second outcome of interest for this study is budget decision, which is operationalized as perception and actual change of annual budget. The operationalization of such perception is the budget actors’ view of the impact of information utilization on budget allocation. Based on the studies on performance management, perceptional budget allocation is measured through the questions of how effective has the development and the use of relevant information been in the respondents’ public organization regarding the following issues: affecting cost savings, changing appropriation levels, reducing duplic ative services or programs, eliminating ineffective services or programs, and improving service quality (Bourdeaux & Chikoto, 2008; Melkers & Willoughby, 2005; Willoughby & Melkers, 2000) . For actual budget allocation, the study retrieves the constant baht figures of 2018 and 2019 expenditures in the programmatic functions under the departments of education and public works (Zaltsman, 2009) . The spending levels on education and public infrastructures of the fiscal year 2019 as well as their increases and decreases compared to that of the year 2018 are statistically analyzed as the next section details. Quantitative Data Analysis Th e results of the pilot study allow for mulation of testable regression models of how policymakers use information and how the use of information affect budget allocation. 78

PAGE 91

Regression Models. The first model regresses administrative capacity and information characteristics on information utilization (see Appendix A for the complete list of variables in the model and its measures). This study applies the ordered logistic regression analyses because of the ordinal nature of the degree of information utilization . It also adds into these analytical models organizational factors (i.e., leadership and organizational culture) and political factors (i.e., outsiders’ use of information against the organization) (Landry et al., 2001; Moynihan, Lavertu, & Kamensky, 2012) . The form of model that estimates the l evel of information utilization is: Information Use = 0 + 1Information Characteristics + 2Administrative Capacity + 3Organizational Factors + 4 Political Factors + error For the perceptional allocation, this study regresses administrative capacity, information characteristics, and information utilization on perceptions on municipal budget decisions (see Appendix B for the complete model and measures). This study a pplies the ordered logistic regression analyses because it asks the department heads to rate effectiveness and efficiency (four levels from low to high) of the functional programs for which the department heads are responsible. The models control for commu nity characteristics and political factors (Melkers & Willoughby, 2005) . The form of model that estimates the perceptions among department head on spending efficiency and effectiveness is: Budget Decision = 0 + 1Administrative Capacity + 2Information Utilization + 3Budget Factors + 4Political Factors + errors 79

PAGE 92

In case of actual budget allocation, this study regresses administrative capacity and information utilization on the budget levels (see Appendix C for the complete model and measures). Budgetary factors (population, areas, number of full time employees, budget si ze of the previous year and ownsource revenue) and other control variables are in the model (Lewis, 2013; Zaltsman, 2009) . The functional form that estimates the spending levels is: 20182019 Budget Change = 0 + 1Administrative Capacity + 2Information Utilization + 3Organizational Factors + 4Political factors + 5Budget Factors + error The Qualitative Approach As the second phase of the quantitative mixed methods design, the qualitative component of this study , involves conducting sixteen case study analyses about the department heads’ use of information and the impact of such information use on budget allocation in order to complement and enrich the quantitative findings. Interviews form the basis of data collection. Meanwhile, keywordin context (KWIC), constant comparison analysis, and classical constant analysis constitute the analys es. The following section provides the details regarding the sampling design, cases, instrumentation, and analytical techniques of this qualitative approach. Sampling Design Population and sample. The population of this qualitative approach is the same as that of the quantitative approach. However, this phase relies on the case studies, selecting information rich cases for in depth study in order to document the diverse theoretical concepts as well as to identify important common patterns among such diverse cases (Miles & Huberman, 1994; Onwuegbuzie & Leech, 2007) . 80

PAGE 93

Sampling scheme. Searching for the common patterns of interrelated concepts and maximizing the variation among the cases, this study applies the subgroup sampling design to the qualitative phase (Onwuegbuzie & Leech, 2007) . There are three attributes considered in order to select cases or municipalities for interviews —type of functional program , mu nicipal size, and budget behavior. Taking into account type of functional program and municipal size for sampling design, the quali ta tive component of this study can control for extraneous variables, thereby reducing threats to internal validity (specifically causal validity) (Newman et al., 2003; Singleton & Straits, 2010) . Regarding the budget behavior for each functional program , this study calculat es the inter annual budget changes of years 20182019 of the education and public works departments and then the Z score of those changes in order to statistically divide the municipalities into three different subgroups —increased, unchanged, and decreased budgets (Gerring, 2007) . Cases and observations. Within each subgroups of increase, unchanged, and decreased budgets for education and public works spending, three cases at the mean of Z score are selected in order to include small, medi um, and large municipalities (Onwuegbuzie & Leech, 2007) . However, note that there are limited numbers of the large sized municipalities and most of them are not at the mean of Z score of each subgroup. Consequently, there are only three department heads from the largesized municipalities for the interviews. Tab le 4.4 summarizes these case studies for education and public works departments according to their sizes and the interannual budget changes. 81

PAGE 94

Table 4.4 Interview Sites for the Subgroup Design Municipal Size Inter annual Budget Change Decrease No Change Increase Small E,PW E,PW E,PW Medium E,PW E,PW 2Es,PW Large PW E,PW Note: E and PW refers to the department of education and the department of public works, respectively. Instrumentation. There are two sources of data collection for this study —available documents and materials , and participant responses from field interview s . The numerical data (i.e., annual budgets for programmatic spending) are secondary data deriving from the quantitative approach. Meanwhile, the more subjective and primary data involving motive, interpretation, and explanation of budgetary information and a budget process which include administrative capacity, information characteristics, information utilization, and perceived budget allocat ion (i.e., efficiency and effectiveness) are gathered through field intervie ws of the education and public works department heads (Singleton & Straits, 2010; Teddlie & Tashakkori, 2009) . Interview protocol. In order to allow interviewees to express their own understanding in their own terms and at the same time to test the hypotheses discussed in Chapter 2, this study applies the standardized openended interview in which the exact wording and sequence of an interview guide are mostly predetermined, structured, and based on the theoretical framework and its hypotheses but in a completely openended format (Singleton & Straits, 2010; Teddlie & Tashakkori, 2009) . Ta ble 4.5 lists the interview questions according to the concepts and constructs of interest. 82

PAGE 95

Table 4.5 Interview Protocol Concept Constructs of Interest Questions 1. Structural isomorphism Coercive and normative isomorphism Which budgeting system does your organization adopt? Why is such system adopted? What are managerial structures, standard of procedures, and fiscal information such system requires? 2.Incrementalism Budget incrementalism How has your organization’s budget process changed over the last 4 years or until the recent budget system has been in place? Currently, how does a budget process in your organization usually proceed? What are the sequence and the key activities in such process? On which activity does the budget committee spend quite a long time and why? 3.Budget rationality Budgetary orientations Are you engaged in budgeting? What are your budgetary responsibilities? What is your role during the budget process? How do you interact with other groups of budget actors? Can you prioritize these following competing spending preferences: financial control, efficiency, management techniques, policy issues, and executive’s plan? Why and how so? 4. Capacity to budget Knowledge an d skills What are skills, knowledge, or technical practices necessary for you to initiative, present, and implement budget proposals? Why? 5.Bugetary information Information piece Information characteristics Information provider Which piece of infor mation do you usually think of and bring up during the budget process? Which information do you base your budget decisions on? Which types of information do you find useful (important or adaptable)? Why? Which types of information do you find reliable? Why? What kind of information do you usually receive from the departments? Does your organization have an official unit providing information for policymakers? What kind of information do they provide? How do they get such information? 6. Information utilization Decision making Presentation strategies How do es th at information have a role during your budget drafting, presentation, debate, and discussion? How do you present your projects to the municipal budget committee? How does the committee react to such presentation style? 7. Budget decisions Perceptional budget allocation What factors carry the most weight in the definition of your organization’s budget decisions? How do such factors impact the spending priorities, eliminating the duplicative programs, or cost saving? If those factors involve information and information use, how was information linked to the differences between the 20172018 budgets? In what other ways did each type of information affect allocation at the followin g steps of the budget cycle: formation of the agency’s budget requests, budget debate, and budget approval? 83

PAGE 96

Qualitative Analysis With a priori causal relationship and hypotheses formulated in Chapter 2, this study depends on a confirmatory nature of data analysis. In order to elaborate those relationships among the key concepts, this qualitative analytical phase begins with the keywordin context (KWIC) technique s o as to identify the meanings of such keywords as budget process, efficiency, effectiveness, key information, and budget allocation. Next, it forms those meanings into themes through the constant comparison analysis. Then, th is study groups and counts such themes by the classical content analysis. The results from these three analytical techniques are triangulated and then supplemented and/or complemented with the quantitative findings in the next chapter. KWIC. KWIC can reveal the budgeting norms and pra ctices shared among key budget actors via the meaning of this research’s priori theoretical concepts (Leech & Onwuegbuzie, 2008) —for example , knowledge, supervisors, roles, performance, performance information, information sources, relevance, adaptability, importance, reliability, budget process, decisions, spending levels, and also others as the interview results suggest. Comparing texts that appear before and after such keywords, KWIC helps clarify such technical terms and how they can interrelate under the context of Thailand’s local governments (Leech & Onwuegbuzie, 2007b) . The KWIC helps to provide meaning of those key concepts as well as identify plausible themes which can clarify the interrelations hips among them. Constant comparison analysis. To identify and understand the interrelationships among those meanings of key concepts identified by KWIC, the constant comparison analysis first conducts an open coding where the study chunks the intervie w data into smaller but meaningful segments in order to assign labels to summarize in a word or short phrase of the research’s key concepts (i.e., descriptive coding), to record such concepts in the interviewees’ own language 84

PAGE 97

(i.e., in Vivo coding), and to connote observable and conceptual sequential practices in the data (i.e., process coding) (Leech & Onwuegbuzie, 2007b; Miles et al., 2014; Strauss & Corbin, 1998) . Then, the study applies axial coding, which groups the segments from open coding into similar categories —administrative capacity, the use of information, budget decisions, and others —as antecedent, independent, dependent, and extraneous variables respectively. Finally, the study conducts selective coding, creating themes out of data of each subgroups of programmatic spending by interrelating the segments from the axial coding (i.e., causation coding) and comparing and contrasting with the hypotheses (i.e., hypothesis coding). Classical content ana lysis. The researcher records how the frequent themes are used to determine which key concepts are most discussed throughout the interview data (Leech & Onwuegbuzie, 2007, 2008) . For instance, the researcher finds the number of department heads who identify themselves w ith a particular budget orientation, explain s how they utilize certain information during the local budget process, and assess es whether such information use affects the budget decisions in terms of both actual allocation and their perceptions on cost savi ng, eliminating duplicative programs, and changing spending priorities. Triangulation As a final analytical step, t his research triangulates the results from quantitative and qualitative approaches mainly by allowing qualitative results to confirm, cross validate, or corroborate the quantitative findings so as to build a coherent justification for the themes on how administrative capacity, and performance and nonperformance information and its characteristics, have an impact on information utilizati on during a budget process and eventually on budget allocation (Creswell, 2013; Teddlie & Yu, 2007) . 85

PAGE 98

In conclusion, this study applies the quantitative mixed methods design in order to investigate the relationships between administrative capacity, budget rationality, information utilization, and budget allocations. These concepts were operationalized through the pilot study in order for the study to have the related constructs and the ir measures fit under the context of a developing country’s local budgeting. After that, the study designs the survey instruments so as to collect the data from all 2,441 municipalities in Thailand. Then, the study calculates the inter annual budgets in order for it to serve as a sampling frame from which cases for interviews are chosen. Finally, the study analyzes the quantitative data from the survey instrument and the qualitative data from the inte rviews. The results from these quantitative and qualitative analyses are discussed in detail in C hapter 5 and Chapter 6, respectively. 86

PAGE 99

CHAPTER V QUANTI T ATIVE RESULTS Applying the quantitative mixed methods design, this chapter presents the qua ntitative results especially through how they answer the research questions and (dis)confirm the hypotheses. The quantitative results here are from the ordinal and the multiple regression analyses. The ordinal regression analyses aim at answering the first research question: how do policymakers with different administrative capacities utilize relevant and reliable information during a budget process? The ordinal regression analyses also explore the first component of budget allocation— the perceptional budge t allocation or the efficiency and effectiveness of the municipal spending—which responds to the second research question: how does the information utilization influence budget decisions? Additionally, the multiple regression analysis strives to complement the perceptional budget allocation by investigating the actual budget allocation. For each research question, this study investigates two main local public services for which the education and public works departments are responsible. For the research que stion about the information utilization, this study disaggregates the analyses by groups that use performance information and those that use nonperformance. Prior to proceeding to the regression analyses and answering the research questions, this chapter first provides the overall descriptive statistics of the study variables that operationali ze the theoretical concepts of this study. The Overall Descriptive Statistics of the Survey Participants and the Study Variables Demographic data This study collects the data regarding administrative capacity, budget rationality, the use of information and budget allocation through the survey instrument. This survey is conducted during October 2018, immediately after the municipal budget preparation and approval for the 87

PAGE 100

start of the 2019 fiscal year, running from October 1, 2018 to September 30, 2019. This study maximized the response rate by sending out the questionnaires to all 2,441 municipalities in Thailand. The response rate is 28.23 percent on average, and 27.82 and 28.64 percent for the departments of education and public works, respectively. Additionally, most are by far from the subdistrict municipalities, representing the agricultural and urban/rural fringe neighborhoods with the population density of 920 residents per squared kilometer. Table 5.1 details the demographic data of the municipalities participating in the survey. 88

PAGE 101

Table 5.1 Demographic Data of the Participating Municipalities Panel A: Survey Responses by Types or Sizes of Jurisdictions Education Department Public Works Department Jurisdiction Type (Size) Questionnaires Sent Response Received Respons e Rate Response Received Response Rate Metropolitan Municipality (Large) 30 3 10.00% 1 3.33% City Municipality (Medium) 178 41 23.03% 45 25.28% Sub District Municipality (Small) 2,233 635 28.44% 653 29.24% Total 2,441 679 27.82% 699 28.64% Panel B: Population Density Group (People/Sq.Km) Frequency Percentage Cumulative Percentage Below 100 195 14.09 14.09 100 500 433 31.29 45.38 Min = 0.94, Max = 58,479, >500 1,000 160 11.56 56.94 Mean = 920, Median = 920.38, >1,000 2,000 123 8.89 65.82 Standard Deviation = 3,000.05 2,000 or over 473 34.18 100.00 Panel C: Community Type Type Frequency Percentage Merchandising and trading 119 9.64 Residential 90 7.29 Mixed (urban and suburban) 500 40.49 Agricultural 516 41.78 Others 10 0.81 Total 1,235 100.00 From the perspective of participants, the department heads are dominantly male (62.67 percent) and the half of them are 41 50 years of age. Also, around half of these department heads earned postgraduate degree and most of them (60.68 percent) have just be en in the supervisor positions in the local governments for under five years. The details regarding the demographic data of the survey participants is provided in Table 5.2. 89

PAGE 102

Table 5.2 Demographic Data of Profile of Participants in the Survey the Survey Respondents Panel A: Age Age Group Frequency Percentage Cumulative Percentage Below 30 8 0.61 0.61 31 40 257 19.45 20.06 Min = 26, Max = 64, 41 50 664 50.26 70.33 Mean = 46.48, Median = 46, 51 60 391 29.60 99.92 Standard Deviation = 6.66 60 or over 1 0.08 100.00 Panel B: Gender Frequency Percentage Male 853 62.67 Female 508 37.33 Total 1,361 100.00 Panel C: Educational Level Degree Frequency Percentage Bachelor 625 45.16 Postgraduate 759 54.84 Total 1,384 100.00 Panel D: Length of Service Working as a Supervisor Year Frequency Percentage Cumulative Percentage Below 2.50 275 29.41 29.41 2.51 5.00 294 31.44 60.86 5.01 7.50 117 12.51 73.37 Min = 0, Max = 32 7.51 10.00 164 17.54 90.91 Mean = 70.1, Median = 5, 10.00 20.00 82 8.77 99.68 Standard Deviation = 5.57 20.00 or over 3 0.32 100.00 The Study V ariables This study investigates the relationships among the concepts of the characteristics of information used (reliability and relevance), administrative capacity, budget rationality, information utilization, and budget allocation. It operationalizes these conce pts given the context of Thailand’s local governments and their budget process and related practices. This section provides the insight into each of these concepts based mainly on the results of the pilot study. 90

PAGE 103

Note that this section only presents the descriptive statistics of the survey data without manipulation. For the inferential statistics analyzed by the regression models, the administrative capacity, information reliability, and information relevance are transformed into average values. Meanwhile, t he information utilization and the perceptional budget allocation are grouped into low, moderate, and high levels. Information. As discussed in Chapter 2, the information one deliberates and utilizes should be reliable and relevant to the context in whic h it is being used. During the budget process (especially the budget discussion), almost all of the department heads (96.65%) always utilize the community’s problems and requests. They also heavily ponder their executive policy and priority and values of m oney with 82.71 and 82.65 percent of them utilizing such information items respectively. Meanwhile, policy results and budget ceiling are also utilized but not as frequently as the community’s problems and requests, the executive policy and priority, and values of money. 77.25 and 65.83 percent of the department heads report that they usually choose policy results and budget ceiling respectively for their budget proposals. However, not many of them utilize service quality and professional standards and techniques (only 39.53 and 37.29 percent respectively) for their budget prepa ration and discussion. Table 5.3 summarizes the information pieces the department heads utilize during the budget process. 91

PAGE 104

Table 5.3 Frequency for In formation Pieces Used During the Budget Process Information Piece1 Frequency Percentage N Valid N Missing N Community’s problems and requests 1,270 96.65% 1,314 70 Executives’ policy and priority 1,081 82.71% 1,307 77 Policy results (PI) 1,015 77.25% 1,314 70 Professional standards and techniques 519 39.53% 1,313 71 Service quality (PI) 490 37.29% 1,314 70 Budget ceiling 865 65.83% 1,314 70 Value of money (PI) 1,086 82.65% 1,314 70 1The responders are asked to choose five out of seven information pieces and PI refers to performance information Information characteristics. Based on five types of information they use, the department heads rate the reliability of each information type. This reliability is dependent on the source of information type from which each information head obtains it. The source of information including self observation, per formance contracts and reports, information center, the central government and their agencies, other local governments, contractors or service providers, and the internet are rated in terms of their trustworthiness as a data source and the quality of infor mation they provide. Overall, the reliability of information regarding the community’s problems and requests, the executives’ policy and priority, policy results, budget ceiling, and value of money is relatively low. However, the reliability of information involving professional standards and techniques and service quality are low. Table 5.4 provides the details of these types of information in terms of the scores and frequencies of their sources’ trustworthiness as well as the quality of each information p iece. Note that on the scale of zero to four, zero means none and one, two, three, and four refers to low, relatively low, relatively high, and high, respectively. For relevance of each information type, the department heads consider all information usefu l and applicable to tasks and activities throughout the budget process. Especially the 92

PAGE 105

community’s problems and requests are deemed more useful and applicable (with the average score of 3.65 and 3.43 on usefulness and applicability) than any other type of information. However, other types of information are also highly useful and applicable. Table 5.5 shows the degree of usefulness and applicability of each information piece used during the budget process. 93

PAGE 106

Table 5.4 Descriptive Statistics for Reliability of Each Information Piece Indicator Frequency Descriptive Statistics N 0 1 2 3 4 Mean Median Standard Deviation Valid N Missing N Community’s problems and requests Information Quality 42 (3.59%) 34 (2.90%) 135 (11.53%) 655 (55.94%) 305 (26.05%) 2.98 3 0.90 1,171 213 Trustworthiness 39 (3.49%) 22 (1.97%) 114 (10.22%) 616 (55.20%) 325 (29.12%) 3.04 3 0.89 1,116 268 Executives’ policy and priority Information Quality 198 (17.14%) 27 (2.34%) 125 (10.82%) 540 (46.75%) 265 (22.94%) 2.56 3 1.33 1,155 229 Trustworthiness 188 (17.06%) 25 (2.27%) 101 (9.17%) 514 (46.64 274 (24.86%) 2.60 3 1.34 1,102 282 Policy results Information Quality 268 (22.91%) 18 (1.54%) 97 (8.29%) 559 (47.78%) 228 (19.49%) 2.39 3 1.43 1,170 214 Trustworthiness 252 (22.62%) 12 (1.08%) 89 (7.99%) 514 (46.14%) 247 (22.17%) 2.44 3 1.42 1,114 270 Professional standards and techniques Information Quality 689 (59.91%) 15 (1.30%) 70 (6.09%) 279 (24.26%) 97 (8.43%) 1.20 0 1.53 1,150 234 Trustworthiness 657 (59.89%) 12 (1.09%) 69 (6.29%) 252 (22.97%) 107 (9.75%) 1.21 0 1.55 1,097 287 Service quality Information Quality 730 (62.39%) 14 (1.20%) 50 (4.27%) 255 (21.79%) 121 (10.34%) 1.16 0 1.56 1,170 214 Trustworthiness 691 (61.97%) 11 (0.99%) 41 (3.68%) 245 (21.97%) 127 (11.39%) 1.20 0 1.59 1,115 269 Budget ceiling Information Quality 402 (34.30%) 22 (1.88%) 103 (8.79%) 451 (38.48%) 194 (16.55%) 2.01 3 1.56 1,172 212 Trustworthiness 386 (34.59%) 18 (1.61%) 95 (8.51%) 419 (37.54%) 198 (17.74%) 2.02 3 1.58 1,116 268 Value of money Information Quality 196 (16.70%) 22 (1.87%) 125 (10.65%) 587 (50.00%) 244 (20.78%) 2.56 3 1.30 1,174 210 Trustworthiness 181 (16.19%) 17 (1.52%) 103 (9.21%) 566 (50.63%) 251 (22.45%) 2.62 3 1.30 1,118 266 94

PAGE 107

Table 5.5 Descriptive Statistics for Relevance of Each Information Piece Indicator Frequency Descriptive Statistics N 0 1 2 3 4 Mean Median Standard Deviation Valid N Missing N Community’s problems and requests Usefulness 2 (0.15%) 9 (0.69%) 20 (1.52%) 390 (29.73%) 891 (67.91%) 3.65 4 0.56 1,312 72 Applicability 2 (0.16%) 16 (1.30%) 57 (4.63%) 523 (42.49%) 633 (51.42%) 3.43 4 0.66 1,231 153 Executives’ policy and priority Usefulness 3 (0.23%) 8 (0.61%) 84 (6.42%) 644 (49.24%) 569 (43.50%) 3.35 3 0.65 1,308 76 Applicability 2 (0.16%) 8 (0.65%) 117 (9.47%) 649 (52.55%) 459 (37.17%) 3.26 3 0.66 1,235 149 Policy results Usefulness 3 (0.23%) 1 (0.08%) 60 (4.59%) 560 (42.81%) 684 (52.29%) 3.47 4 0.61 1,308 76 Applicability 1 (0.08%) 7 (0.57%) 59 (4.78%) 580 (46.96%) 588 (47.61%) 3.41 3 0.62 1,235 149 Professional standards and techniques Usefulness 1 (0.08%) 14 (1.07%) 95 (7.29%) 596 (45.74%) 597 (45.82%) 3.36 3 0.67 1,303 81 Applicability 1 (0.08%) 10 (0.80%) 93 (7.48%) 625 (50.28%) 514 (41.35%) 3.32 3 0.65 1,243 141 Service quality Usefulness 2 (0.15%) 6 (0.46%) 70 (5.34%) 627 (47.83%) 606 (46.22%) 3.40 3 0.63 1,311 73 Applicability 1 (0.08%) 4 (0.32%) 79 (6.40%) 647 (52.43%) 503 (40.76%) 3.33 3 0.62 1,234 150 Budget ceiling Usefulness 3 (0.23%) 16 (1.23%) 190 (14.59%) 544 (41.78%) 549 (42.17%) 3.24 3 0.76 1,302 82 Applicability 2 (0.16%) 23 (1.85%) 151 (12.16%) 605 (48.71%) 461 (37.12%) 3.21 3 0.73 1,242 142 Value of money Usefulness 2 (0.15%) 8 (0.61%) 62 (4.75%) 535 (41.03%) 697 (53.45%) 3.47 4 0.63 1,304 80 Applicability 3 (0.24%) 8 (0.65%) 61 (4.95%) 574 (46.55%) 587 (47.61%) 3.41 3 0.64 1,118 151 95

PAGE 108

Administrative capacity. There are six groups of administrative capacity the department heads consider necessary for initiating, drafting, discussing, and defending their budget proposals. This capacity to budget according to their self reporting scores appears to be relatively high especially for the skill and knowledge related to communication skills, intergovernmental relations, strategic planning, and participation budgeting. However, the department heads rate their skill and knowledge regarding project management and accounting and financial mana gement relatively low. Table 5.6 provides the details for each skill or knowledge constituting the capacity to budget. Budget rationality. Budget rationality represents the department heads’ attitudes toward how to prioritize budget values —control (fiscal accountability accuracy), management (efficiency and effectiveness), and policy (strategic pl anning and advocacy for and responsiveness to policy outcome) (David, 1998; Smith, 2003; Willoughby, 2002) . Based on the survey data, 32.94 percent of the department heads report that they identify with conduit —only transmitting information they receive to the budget committee and not presenting any spending preference speci fically. Meanwhile, 25.09 percent of department heads in this study represents the policymaking budget orientation, weighing in both political and analytical information. However, only a few number of department heads (0.89 percent or 12 out of 1,348) identify themselves with techno rational budget orientation who heavily promote financial and budgeting techniques as well as encourage the budget committee of desirable policy priorities and funding strategies. Table 5.7 provides the details of budget orientations of all responding department heads in the survey. 96

PAGE 109

Table 5.6 Frequency and Descriptive Statistics for Administrative Capacity Indicator Frequency 1 Descriptive Statistics N Fac tor Loading s2 0 1 2 3 4 Mea n Media n Standar d Deviatio n Valid N Missin g N Administrative Capacity Participatory budgeting 4 (0.29 %) 15 (1.09 %) 200 (14.58 %) 908 (66.18 %) 245 (17.86 %) 3.00 3 0.63 1,372 12 0.73 Communication skills 0 (0.00 %) 9 (0.66 %) 116 (8.49% ) 970 (71.01 %) 271 (19.84 %) 3.10 3 0.55 1,366 18 0.77 Intergovernmental relations 0 (0.00 %) 9 (0.65 %) 159 (11.56 %) 937 (68.15 %) 270 (19.64 %) 3.07 3 0.58 1,375 9 0.71 Strategic planning and management 1 (0.07 %) 7 (0.52 %) 186 (13.71 %) 902 (66.47 %) 261 (19.23 %) 3.04 3 0.60 1,357 27 0.76 Project and project management 0 (0.00 %) 20 (1.46 %) 267 (19.43 %) 900 (65.50 %) 187 (13.61 %) 2.91 3 0.62 1,374 10 0.77 Accounting and financial management 6 (0.44 %) 46 (3.34 %) 413 (30.01 %) 770 (55.96 %) 141 (10.25 %) 2.72 3 0.71 1,376 8 0.71 1Theoretical Range, 0 = none; 1 = low; 2 = relatively low; 3 = relatively high; 4 = high. 2Principal component analysis variance explained = 55.30%; Cronbach alpha = 83.37 97

PAGE 110

Table 5.7 Frequency of Budget Rationality and Spending Preference Budget Ori entation Frequency Percentage Conduit (financial rules and regulations) 444 32.94 Accounting (efficiency) 263 19.51 Techno rational (financial and accounting techniques) 12 0.89 Policymaking (specific policy issues) 339 25.09 Advocacy (executive’s policy directions) 322 23.89 Total 1,348 100.00 98

PAGE 111

Information utilization. The survey result s suggest that most of the time the department heads use the information bundle or the information they consider reliable and relevant information mainly for prioritizing programs or spending choices while drafting their budget proposals. Less often, they use such information to support their budget decisions usually at the programmatic level. However, only rarely do they use the information in order to argue against other spending choices. Table 5.8 details the frequency and descriptive statistics for all types of information utilizatio n in this study. Strategy of information utilization. Based on the pilot study, there are five strategies of information use —we’re expert, problem framing, incrementalism, performance management doctrine, and it’s mandate. The survey instrument asks the department heads to specify how frequent they apply each strategy (i.e., always, almost always, sometimes, rarely, and never). This study then changes this scale into a dummy variable —always for one and other for zero. The 18.97 percent of responding depar tment heads states that they always present all the reliable and relevant information to the budget committee by specifying an area of inefficiencies or ineffectiveness, then offering the projects as a solving mechanism, and showing how a variety of target groups (both citizens and local government alike) can benefit from those projects in term of better quality of service, outcomes, efficiency, and effectiveness (Moynihan, 2008b; Mucha, 2009) . However, only 5.27 percent of them report that they would discuss their projects or functional programs to the municipal budget committee by comparing those projects’ or programs’ components with professional or expert standards. Table 5.9 show s the frequency of all strategies of information util ization the department heads apply while presenting their budget proposals. 99

PAGE 112

Table 5.8 Frequency and Descriptive Statistics for Information Utilization for Budget Discussion Indicator Frequency 1 Descriptive Statistics N Factor Loading s2 0 1 2 3 4 Mea n Median Standard Deviation Valid N Missin g N To prioritize programs or spending choices 5 (0.36%) 37 (2.69%) 144 (10.49%) 685 (49.89%) 502 (36.56 %) 3.20 3 0.76 1,373 11 0.79 To compensate with other information 11 (0.80%) 75 (5.48%) 351 (25.66%) 619 (45.25%) 312 22.81% ) 2.84 3 0.87 1,368 16 0.78 To base your questions or investigation 18 (1.32%) 91 (6.66%) 486 (35.55%) 635 (46.45%) 137 (10.02 %) 2.57 3 0.81 1,367 17 0.71 To support your decisions 2 (0.15%) 54 (3.95%) 254 (18.59%) 782 (57.25%) 274 (20.06 %) 2.93 3 0.74 1,366 18 0.70 To argue against other spending choices 35 (2.56%) 136 (9.93%) 479 (34.99%) 568 (41.49%) 151 (11.03 %) 2.48 3 0.91 1,369 15 0.63 Notes: 1Theoretical Range, 0 = never; 1 = rarely; 2 = sometimes; 3 = most of the times; 4 = always. 2Principal component analysis variance explained = 52.35%; Cronbach alpha = 76.88%). The percentage quoted in brackets are those of the valid survey responses. 100

PAGE 113

Table 5.9 Frequency for Strategies of Information Utilization th e Department Heads Always Apply Strategy Frequency Percentage N Valid N Missing N We are experts 67 5.27% 1,272 112 Framing 88 6.93% 1,270 114 Incrementalism 81 6.37% 1,271 113 Performance Management Doctrine 242 18.97% 1,276 108 Mandate 142 11.19% 1,269 115 101

PAGE 114

The perceptional budget allocation. When a sked how their municipality’s budget allocation promotes effectiveness and efficiency, the department heads state t hat information utilization has an impact on their municipality’s spending in terms of service quality improvement and cost saving to the relatively high extent. Meanwhile, it can eliminate ineffective service or pro gram and reduce duplicative services to the relatively low extent. Nonetheless, the information utilization has only limitedly impact on the change in appropriation level. Table 5.10 shows the frequency and descriptive statistics of this perceptional budge t allocation. In conclusion, the descriptive statistics in this section shows the characteristics of survey participants in terms of their communities they represent, their administrative capacity, their budget rationality, their information they usuall y use, and how they use such information. These descriptive statistics are important to understanding information use. Nonetheless, they fail to sugge st a holistic model of information use and budget allocation. T herefore, the causal relationships among those concepts are the subject of the next section, which takes up the regression models. 102

PAGE 115

Table 5.10 Frequency and Descriptive Statistics for Perceptional Budget Allocation Indicator Frequency 1 Descriptive Statistics N Factor Loadings2 0 1 2 3 4 Mean Median Standard Deviation Valid N Missing N Affecting cost savings 7 (0.53%) 14 (1.05%) 86 (6.47%) 904 (67.97%) 319 (23.98%) 3.14 3 0.62 1,330 54 0.80 Changing appropriation levels 57 (4.32%) 88 (6.67%) 403 (30.55%) 664 (50.34%) 107 (8.11%) 2.51 3 0.90 1,319 65 0.75 Reducing duplicative services or programs 42 (3.18%) 44 (3.33%) 245 (18.53%) 822 (62.18%) 169 (12.78%) 2.78 3 0.83 1,322 62 0.69 Eliminating ineffective services or programs 38 (2.87%) 47 (3.55%) 216 (16.33%) 801 (60.54%) 221 (16.70%) 2.85 3 0.84 1,323 61 0.64 Improving service quality 7 (0.53%) 8 (0.60%) 88 (6.62%) 853 (64.14%) 374 (28.12%) 3.18 3 0.62 1,330 54 0.62 Notes: 1Theoretical Range, 0 = none; 1 = low; 2 = relatively low; 3 = relatively high; 4 = high. 2Principal component analysis variance explained = 49.43%; Cronbach alpha = 73.44%). The percentage quoted in brackets are those of the valid survey responses. 103

PAGE 116

As previously mentioned in Chapter 4, the municipal size (i.e., small, medium, and large) is considered an extraneous variable, which hypothetically has an impact on the budget allocation (i.e., spending efficiency and effectiveness and actual budget alloc ation). This section descriptively tests if this impact is statistically significant. The data from the survey are analyzed using Analysis of Variance (ANOVA). ANOVA aims to determine whether budget allocations are different for groups with different munic ipality size at 0.05 percent level of significance as shown in Table 5.11 and 5.12 Table 5.11 ANOVA for a Comparison of Spending Efficiency and Effectiveness according to Municipality Size Spending Efficiency and Effectiveness S um Squares Degree of freedom Mean Squares F Ratio Sig. Both departments Between groups 3.1451 2 1.5725 0.22 0.8038 Within groups 9,343.1977 1,298 7.1981 Total 9,346.3428 1,300 7.1895 Education Department Between groups 6.4147349 2 3.2073 0.54 0.5804 Within groups 3,793.9377 644 5.8912 Total 3,800.3524 646 5.8829 Public Works Department Between groups 26.0366 2 13.0183 1.60 0.2032 Within groups 5,272.9234 647 8.14980 Total 5,298.9600 649 8.16481 For the spending efficiency and effectiveness, the ANOVA tests indicate that there is not a statistically significant difference between the municipality size and spending efficiency and effectiveness reported by the department heads. This holds for the case of the education program (su pervised by an education department), the civil engineering program (supervised by a public work department), and both programs simultaneously. 104

PAGE 117

Table 5.12 ANOVA for a Comparison of Actual Budget Allocation (BTH1,000 per capita) for FY2019 Actual Budget Allocation Sum Squares Degree of freedom Mean Squares F Ratio Sig. Education Department (Total Program Spending) Between groups 100.5374 2 50.2687 20.22 0.0000 Within groups 671.3067 270 2.4863 Total 771.8441 272 2.8377 Education Department (Investment Spending) Between groups 0.0021 2 0.0010 0.00 0.9975 Within groups 89.8147 214 0.4197 Total 89.8168 216 0.4158 Public Works Department (Total Program Spending) Between groups 13.1435 2 6.5718 3.98 0.0205 Within groups 267.3624 162 1.6503 Total 280.5059 164 1.7104 Public Works Department (Investment Spending) Between groups 1.1107 2 0.5554 1.12 0.3285 Within groups 77.2667 156 0.4953 Total 78.3774 158 0.4961 For the actual budget allocation, the ANOVA tests indicate that there is a statistically significant difference actual budget allocation among groups of municipalities with different sizes. Particularly, the differences exist in the total spending of education program and the total spending on civil engineering program. 105

PAGE 118

Information, Administrative Capacity, Budget Rationality, and Information Utilization The first research question asks: how do policymakers with different administrative capacities utilize relevant and reliable information during a budget process? It aims to explain the variation of the degree to which the department heads apply information perceived as reliab le and relevant to influence program prioritization, to complement with oth er information in order to reach budget decisions, to prompt further questions or enquiries, to support a budget decision reached by other means, and to argue that a certain budget choice may not be desirable. This section first elaborates how all respondi ng department heads uti lize information. Then, the subsequent section takes a closer look at each department in order to understand the information utilization under the different policy areas. The Overall Information Utilization among Functional Departme nt Heads Descriptive Statistics. The first group of regression analyses considers the cross section data from the whole sample —all the department heads who participate d in the 2019 annual budget prepar ation and discussion and answered the st udy’s questionnaires. Table 5.1 3 provides the frequency and descriptive statistics for these three regression analyses. Note that the budget orientation and the type of information users (the performance information users) are dichotomous variables and the level of information utilization is ordinal. The table presents them in terms of frequencies and percentage. 106

PAGE 119

Table 5.13 Frequency and Descriptive Statistics for Information Use by Department Heads Variable All Department Heads N Mean SD Control Normative isomorphism 889 3.14 0.79 Coercive isomorphism 889 2.72 0.97 Incrementalism 889 2.62 0.83 Outsiders’ information use 889 2.39 0.99 Leaders’ commitment 889 3.22 0.90 Staff’s commitment 889 3.15 0.81 Rational culture 889 3.56 0.70 Bureaucratic culture 889 3.47 0.81 Group culture 889 3.26 0.74 Development culture 889 3.56 0.83 Independent Information reliability 889 2.99 0.72 Information relevance 889 3.32 0.69 Admin capacity 889 2.98 0.44 Conduit orientation 297 33.41% Accountant orientation 163 18.34% Policymaking orientation 223 25.08% Advocacy orientation 216 24.30% Techno rational orientation 6 0.67% Performance information user 188 21.15 Dependent Low information utilization 342 38.47% Moderate information utilization 275 30.93% High information utilization 272 30.60% Regression result. Table 5.14 provides the results from the regression analyses, indicating the impact of information characteristics (reliability and relevance), administrative capacity, and budget rationality on the level of information utilization for the budget discussion among all department heads. 107

PAGE 120

Table 5.14 Regression Results of Information Utilization for Budget Discussion among all Department Heads Variable (1) (2) (3) (4) Multiple Regression Ordered Logit Ordered Probit Ordered Probit (Robust Standard Errors) Normative isomorphism 0.008 0.023 0.015 0.015 (0.034) (0.085) (0.052) (0.051) Coercive isomorphism 0.008 0.009 0.011 0.011 (0.029) (0.072) (0.043) (0.044) Incrementalism of decision making 0.028 0.082 0.049 0.049 process (0.031) (0.080) (0.048) (0.049) Outsiders’ information use against the 0.005 0.009 0.006 0.006 department (0.027) (0.067) (0.040) (0.040) Leaders’ commitment to info base 0.042 0.097 0.061 0.061 decision making (0.032) (0.082) (0.048) (0.052) Staff’s commitment to info base 0.025 0.071 0.038 0.038 decision making process (0.032) (0.079) (0.048) (0.046) Rational organization culture 0.021 0.054 0.040 0.040 (0.043) (0.108) (0.065) (0.065) Bureaucratic culture 0.003 0.014 0.016 0.016 (0.038) (0.094) (0.057) (0.055) Group Culture 0.040 0.090 0.053 0.053 (0.040) (0.103) (0.062) (0.063) Development Culture 0.038 0.092 0.058 0.058 (0.038) (0.094) (0.057) (0.056) Reliability of information used 0.018 0.031 0.022 0.022 (0.051) (0.128) (0.077) (0.079) Relevance of information used 0.103* 0.315** 0.176** 0.176** (0.053) (0.135) (0.081) (0.083) Administrative capacity 0.570*** 1.483*** 0.896*** 0.896*** (0.062) (0.170) (0.099) (0.100) Accountant (budget rationality) 0.008 0.063 0.033 0.033 (0.075) (0.186) (0.111) (0.116) Policymaking (budget rationality) 0.150** 0.349** 0.214** 0.214** (0.068) (0.166) (0.100) (0.101) Advocacy (budget rationality) 0.207*** 0.562*** 0.326*** 0.326*** (0.070) (0.178) (0.105) (0.112) Techno rational (budget rationality) 0.334 1.008 0.603 0.603 (0.319) (0.919) (0.539) (0.583) Performance Information User 0.060 0.101 0.074 0.074 (0.066) (0.165) (0.099) (0.101) Constant 0.215 5.091*** 3.073*** 3.073*** (0.285) (0.758) (0.449) (0.456) /cut1 6.568*** 3.966*** 3.966*** (0.770) (0.454) (0.462) /cut2 Observations 889 889 889 889 R squared 0.155 Adj. R squared Pseudo R squared 0.080 0.079 0.079 Robust standard errors in parentheses and *** p<0.01, ** p< 0.05, * p<0.10 108

PAGE 121

The results of the ordinal regression analysis reveal that information characteristics, administrative capacity, budget rationality, the type of information users, structural isomorphism, organizational culture, and political factors are to gether related to the use of information for budget discussion. The variables in the model predicted 7.9% of the variability in the information utilization for budget discussion. This study attempts to reduce the heteroscedasticity by introducing the robus t standard errors into the last model. For the multicollinearity, the variance inflation factor (vif) of each independent variable is less than four (ranging from 1.01 to 2.04). Consequently, the multicollinearity may also not be a reason of concern. The examination of each parameter estimate indicates that variables including the administrative capacity and budget rationality of advocacy are statistically significant at the 0.01 level. Meanwhile, the relevance of information and budget rationality of pol icymaking orientation are statistically significant at the 0.05 level. The highest predictor of the information utilization is administrative capacity, with a standardized beta of 0.896. This means that if an average score of administrative capacity of a certain department head is to increase by one point, his or her ordered log odds of using information more comprehensively (i.e., from low to moderate or from moderate to high level category) would increase by 0.896 while the other variables in the model ar e held constant. For the relevance of information, if the average score of relevance is to increase by one point, a department head’s ordered log odds of using information more intensively would increase by 0.176, all else unchanged. However, the ordered l ogit for the department heads with policymaking and advocacy orientations being in a higher level of information utilization category is 0.214 and 0.326 respectively less than those with conduit orientation ceteris paribus. 109

PAGE 122

The results in t his model part ially confirm H1A and fully confirm H1C. In other words, the relevance of information and the administrative capacity are positively associated with the level of information utilization. Additionally, it also partly confirms the H1D that one with policymaking orientation is more lik ely to utilize information than one with advocacy orientation. However, the department heads with policymaking orientation utilize information for budget discussion less than those with conduit orientation. As a result, this fail s to support the H1D hypothesis. Furthermore, the variables of the reliability of information, the technorational orientation, and the performance information users are not statistically significant. Consequently, th e results do not entirely support the H 1A (for the part of the reliability of information) and H1D (for the part of the technorational and policymaking orientations). Also, the comprehensive use of performance information does not contribute to the information uti lization for budget discussion, failing to support H1B. In short, it can be said that among all the department heads in this study, ones with high administrative capacity provided with the relevant information are highly likely to comprehensively utilize the information for the budget discussion. However, if they identify with particular po licy directions either from their own analytical and political consideration or from their compliance to that from their municipal executives, they tend to utilize information less than those who just transmit information between executives and agency and back again without spending preferences. The Overall Information Utilization among Education Department Heads Descriptive statistics. Table 5.15 provides the frequency and statistics for regression analyses related to administrative capacity, budget ratio nality, information utilization, and organization culture reported by the education department heads. 110

PAGE 123

Table 5.15 Frequency and Descriptive Statistics for Information Use by the Education Department Heads Variable All Department Heads N Mean SD Control Normative isomorphism 469 3.17 0.77 Coercive isomorphism 469 2.82 0.94 Incrementalism 469 2.62 0.83 Outsiders’ information use 469 2.40 0.97 Leaders’ commitment 469 3.28 0.86 Staff’s commitment 469 3.16 0.80 Rational culture 469 3.36 0.67 Bureaucratic culture 469 3.48 0.77 Group culture 469 3.26 0.73 Development culture 469 3.40 0.78 Independent Information reliability 469 3.03 0.72 Information relevance 469 3.34 0.72 Admin capacity 469 3.00 0.44 Conduit orientation 197 42.00% Accountant orientation 73 15.57% Policymaking orientation 124 26.44% Advocacy orientation 80 17.06% Techno rational orientation 3 0.64% Performance information user 131 27.93% Dependent Low information utilization 156 33.26% Moderate information utilization 145 30.92% High information utilization 168 35.82% Regression results. Table 5.16 provides the results from regression analyses, indicating the impact of information characteristics (reliability and relevance), administrative capacity, the type of information users, and budget rationality on the level of information utilization for the budget discussion among the education department heads. 111

PAGE 124

Table 5.16 Regression Results of Information Utilization for Budget Discussion among all Education Department Heads Varia ble (1) (2) (3) (4) Multiple Regression Ordered Logit Ordered Probit Ordered Probit (Robust Standard Errors) Normative isomorphism 0.104** 0.263** 0.157** 0.157** (0.049) (0.121) (0.073) (0.074) Coercive isomorphism 0.021 0.045 0.032 0.032 (0.042) (0.103) (0.062) (0.063) Incrementalism of decision making 0.000 0.006 0.000 0.000 process (0.045) (0.113) (0.068) (0.070) Outsiders’ information use against the 0.005 0.007 0.002 0.002 department (0.038) (0.095) (0.057) (0.057) Leaders’ commitment to info base 0.111** 0.269** 0.168** 0.168** decision making (0.048) (0.120) (0.072) (0.073) Staff’s commitment to info base 0.072 0.179 0.112* 0.112* decision making process (0.045) (0.109) (0.067) (0.063) Rational organization culture 0.034 0.086 0.061 0.061 (0.062) (0.153) (0.093) (0.090) Bureaucratic culture 0.041 0.116 0.064 0.064 (0.057) (0.140) (0.085) (0.082) Group Culture 0.132** 0.334** 0.206** 0.206** (0.057) (0.143) (0.087) (0.085) Development Culture 0.028 0.085 0.050 0.050 (0.058) (0.142) (0.086) (0.084) Reliability of information used 0.003 0.020 0.006 0.006 (0.078) (0.196) (0.117) (0.123) Relevance of information used 0.107 0.291 0.169 0.169 (0.077) (0.192) (0.115) (0.120) Administrative capacity 0.508*** 1.302*** 0.784*** 0.784*** (0.089) (0.234) (0.137) (0.139) Accountant (budget rationality) 0.008 0.006 0.004 0.004 (0.109) (0.268) (0.161) (0.163) Policymaking (budget rationality) 0.235** 0.557** 0.345** 0.345*** (0.091) (0.224) (0.136) (0.134) Advocacy (budget rationality) 0.027 0.050 0.023 0.023 (0.104) (0.264) (0.156) (0.168) Techno rational (budget rationality) 0.887* 14.905 5.351 5.351*** (0.460) (757.432) (169.936) (0.252) Performance Information User 0.094 0.206 0.135 0.135 (0.085) (0.212) (0.127) (0.128) Constant 0.147 4.095*** 2.430*** 2.430*** (0.406) (1.039) (0.621) (0.637) /cut1 5.565*** 3.323*** 3.323*** (1.054) (0.627) (0.644) /cut2 Observations 469 469 469 469 R squared 0.160 Adj. R squared Pseudo R squared 0.082 0.082 0.082 Robust standard errors in parentheses and *** p<0.01, ** p<0.05, * p<0.10 112

PAGE 125

The results of the ordinal regression analysis reveals that information characteristics, administrative capacity, budget rationality, the type of information users, structural isomorphism, organizational culture, and political factors are together related to the use of information for budget discussion among the education department heads . The variables in the model predicted 8.2% of the variabil ity in the information utilization for budget discussion. As mentioned earlier, this study attempts to reduce the heteroscedasticity by introducing the robust standard errors into the last model. For the multicollinearity, the variance inflation factor (vif) of each independent variable is less than four (ranging from 1.01 to 2.41). Consequently, the multicollinearity may also not be a reason of concern. The examination of each parameter estimate indicates that administrative capacity, the budget rational ity of policymaking and technorational orientations are statistically significant at the 0.01 level. The highest predictor of the information utilization is administrative capacity, with a standardized beta of 0.784. This means that if an average score of administrative capacity of a certain education department head is to increase by one point, his or her ordered logodds of using information more comprehensively (i.e., from low to moderate or from moderate to high level category) would increase by 0.784 while the other variables in the model are held constant. However, the ordered logit coefficient for the education department head s with policymaking and techno rational orientations being in a higher information utilization category are 0.345 and 5.351 re spectively less than those with conduit orientation ceteris paribus. Additionally, the normative isomo rphism and the group culture among department staff are statistically significant at the 0.05 level and positively associated with the information utilization . Also, the leaders’ and staff’s commitment s to an info rmation based decision making are st atistically and have a negative impact on the information utilization. 113

PAGE 126

Regarding the study’s hypothe ses, the results in this model supports the H1C. In other words, a higher lev el of administrative capacity contributes to a higher level of information utilization. Nonetheless, these results for the education department disconfirm the H1D because education department heads with either policymaking or technorational orientation are likely to utilize inf ormation less than those who have a conduit orientation. Note that t he reliability a nd the relevance of information and the comprehensive use of performance information do not significantly contribute to the information utilization for budget discussion, failing to support H1A and H1B. In other words, the results suggest that the education department heads who possess the relatively high administrative capacity and practice the budgeting principles and elements suggested by the education professionals or experts tend to comprehensively utilize the information during the municipal budget discussion. However, the education department heads w ho are highly technical, either concentrating on their own analytical and political consideration or focusing on the budgeting techniques, are less likely to utilize information during the budget discussion than those who just transfer information back and forth between executives and agency without spending preferences. The organization cultures also affects the level of information utilization among the education department heads. The education department heads who work under the circumstances which promote and focus on their people in the department as well as their employees’ morale and cohesion (Taylor, 2011) tend to highly take into account the relevant and reliable information for their budget discussion. Nonetheless, the organization culture, which thrives on the support from the staff members and the municipal executives or ones outside the informal groups are negatively associated with the use of information among the education department heads. The agency’s top leadership as well as other staff members 114

PAGE 127

demonstrate a strong commitment to achieving results can reduce the extent to which the education department heads utilize information during the budget process. The Overall Information Utilization among the Public Works Department Heads Descriptive statistics. Table 5.17 provides the frequency and statistics that highlight administrative capacity, budget ratio nality, information utilization, and organization culture reported by the public works department heads . Note that the budget orientation is a dichotomous variable and the level of information utilization is ordinal. The table presents them in terms of frequencies and percentage. Table 5.17 Frequency and Descriptive Statistics for Information Use b y Public Works Department Heads Variable All Department Heads N Mean SD Control Normative isomorphism 422 3.11 0.80 Coercive isomorphism 422 2.61 0.98 Incrementalism 422 2.61 0.88 Outsiders’ information use 422 2.36 1.01 Leaders’ commitment 422 3.16 0.94 Staff’s commitment 422 3.12 0.83 Rational culture 422 3.35 0.71 Bureaucratic culture 422 3.50 0.85 Group culture 422 3.26 0.75 Development culture 422 3.30 0.87 Independent Information reliability 422 2.94 0.73 Information relevance 422 3.29 0.66 Admin capacity 422 2.96 0.45 Conduit orientation 100 23.70% Accountant orientation 91 21.56% Policymaking orientation 100 23.70% Advocacy orientation 136 32.23% Techno rational orientation 3 0.71% Performance information user 58 13.74% Dependent Low information utilization 187 44.31% Moderate information utilization 130 30.81% High information utilization 105 24.88% 115

PAGE 128

Regression results. Table 5.18 provides the results from regression analyses, indicating the impact of information characteristics (reliability and relevance), administrative capacity, the type of information users and budget rationality on the level of information utilization for the budget discussion among the public works department heads. 116

PAGE 129

Table 5.18 Regression Results of Information Utilization for Budget Discussion among Public Works Department Heads Varia ble (1) (2) (3) (4) Multiple Regression Ordered Logit Ordered Probit Ordered Probit (Robust Standard Errors) Normative isomorphism 0.087* 0.211* 0.129* 0.129* (0.048) (0.125) (0.077) (0.071) Coercive isomorphism 0.023 0.071 0.040 0.040 (0.039) (0.104) (0.062) (0.062) Incrementalism of decision making 0.035 0.113 0.067 0.067 process (0.043) (0.118) (0.070) (0.072) Outsiders’ information use against the 0.024 0.061 0.033 0.033 department (0.037) (0.098) (0.059) (0.058) Leaders’ commitment to info base 0.005 0.036 0.021 0.021 decision making (0.042) (0.114) (0.067) (0.072) Staff’s commitment to info base 0.037 0.093 0.061 0.061 decision making process (0.044) (0.120) (0.072) (0.072) Rational organization culture 0.013 0.072 0.034 0.034 (0.058) (0.158) (0.095) (0.094) Bureaucratic culture 0.035 0.097 0.065 0.065 (0.051) (0.135) (0.081) (0.080) Group Culture 0.020 0.077 0.051 0.051 (0.056) (0.156) (0.092) (0.097) Development Culture 0.029 0.073 0.045 0.045 (0.049) (0.130) (0.078) (0.077) Reliability of information used 0.029 0.049 0.029 0.029 (0.066) (0.175) (0.104) (0.108) Relevance of information used 0.121 0.450** 0.248** 0.248** (0.074) (0.205) (0.119) (0.126) Administrative capacity 0.601*** 1.717*** 1.051*** 1.051*** (0.088) (0.260) (0.153) (0.147) Accountant (budget rationality) 0.092 0.279 0.178 0.178 (0.105) (0.275) (0.163) (0.176) Policymaking (budget rationality) 0.077 0.204 0.098 0.098 (0.102) (0.263) (0.157) (0.162) Advocacy (budget rationality) 0.254** 0.757*** 0.432*** 0.432*** (0.099) (0.267) (0.156) (0.167) Techno rational (budget rationality) 0.154 0.497 0.212 0.212 (0.441) (1.373) (0.728) (0.955) Performance Information User 0.050 0.189 0.109 0.109 (0.111) (0.293) (0.176) (0.177) Constant 0.429 6.287*** 3.823*** 3.823*** (0.403) (1.184) (0.691) (0.690) /cut1 7.876*** 4.775*** 4.775*** (1.206) (0.700) (0.698) /cut2 Observations 422 422 422 422 R squared 0.197 Adj. R squared Pseudo R squared 0.109 0.109 0.109 Robust standard errors in parentheses and *** p<0.01, ** p<0.05, * p<0.10 117

PAGE 130

The results of the ordinal regression analysis reveal that information characteristics, administrative capacity, budget rationality, the type of information use , structural isomorphism, organizational culture, and political factors are together related to the use of information for budget discussion. The var iables in the model predicted 10.9% of the variability in the information utilization for budget discussion. As mentioned earlier, this study attempts to reduce the heteroscedasticity by introducing the robust standard errors into the last model. For the m ulticollinearity, the variance inflation factor (vif) of each independent variable is less than four (ranging from 1.02 to 1.80). Consequently, the multicollinearity may also not be a reason of concern. The examination of each parameter estimate indicates that administrative capacity, and the budget rationality of advocacy orientation are statistically significant at the 0.01 level. The highest predictor of the information utilization is administrative capacity, with a standardized beta of 1.051. This means that if an average score of administrative capacity of a certain public works department head is to increase by one point, his or her ordered logodds of using information more comprehensively (i.e., from low to moderate or from moderate to high level category) would increase by 1.051 while the other variables in the model are held constant. However, the ordered logit for the public works department heads with advocacy orientations being in a higher information utilization category is 0.432 less than thos e with conduit orientation ceteris paribus. F or the relevance of information, which is statistically significant at the 0.5 level, if the average score of relevance is to increase by one point, a public works department head’s ordered logodds of using information more intensively (i.e., from low to moderate or from moderate to high level category) would increase by 0.248, all else unchanged. Additionally, the normative isomorphism is statistically significant at the 0.10 level. This mean 118

PAGE 131

that if the average score of normative isomorphism is to increase by one point, a public works department head’s ordered logodds of using information more intensively (i.e., from low to moderate or from moderate to high level category) would decrease by 0.129, all else unchanged. The results in this model partially confirms the H1A and fully confirms the H1C. In the other words, the relevance of information and the administrative capacity are positively associated with the lev el of information utilization. Additionally, it also partly confirms the H1D that one with advocacy orientation is less likely to utilize information than one with other budget orientations. Though, that the reliability of information, the technorational orientation, and the performance information users is not statistically significant. Consequently, the results do not entirely support H1A (the reliability of information) and H1D (the technorat ional orientation). Meanwhile, the comprehensive use of perfo rmance information does not contribute to the information utilization for budget discussion, failing to support H1B . In short, it can be said that among all the public works department heads in this study, ones with high administrative capacity provided wi th the relevant information are highly likely to comprehensively utilize the information for the budget discussion. However, if they particularly identify with particular policy directions initiated by their municipal executives, they tend to utilize information less than those who just transmit information between executives and agency and back again without spending preferences. Additionally, the budget norms and practices promoted by the public works professionals also hinders the public works department heads from intensively taking into account the related and reliable information while discussing their budget proposals to the municipal budget committee. 119

PAGE 132

Administrative Capacity, Information Utilization, and Budget Decisions In order to answer the second research question of “ How does the information utilization influence budget decisions?”, t his study investigates the budget decisions via two perspectives which are the perceptional and the actual budget allocations. The perceptional budget allocations involves the department heads’ perceptions on the efficiency and effectiveness caused by thei r departments’ spending levels. Meanwhile, the actual budget allocations here means the budget allocation levels (in terms of their increase and decrease compared to the previous fiscal year) of the total spending and the investment spending of the educati on and public works departments. The Perceptional Budget Allocation for Both Education and Public Works Departments The efficiency and effectiveness do not go hand in hand with the actual spending levels but are apparently the aims of the performance bas ed budget reform (Andrews & Hill, 2003; Willoughby & Melkers, 2000) . This section therefore investi gates whether and how administrative capacity and information utilization (the level of information utilization during the budget discussion as well as the strategy of information use) are associated with efficiency and effectiveness of the municipal spend ing. S pending efficiency and effectiveness are measured through the department heads’ perceptions on how information use can help promote cost saving, change appropriation levels, reduce duplicative services, eliminate ineffective programs, and improve ser vice quality. Descriptive statistics. There are 744 department heads who often use information for budget discussion to a relatively high extent (2.83 out of 4.00 points) and possess relatively high level of administrative capacity (2.98 out of 4.00). Als o, many of them (20.03 percent) show the ineffective policy issues and propose to the budget committee the solutions based on the performance measures. However, in their opinion, the information utilization can have a 120

PAGE 133

relatively low impact on the spending efficiency and effectiveness (1.88 out of 4.00). Table 5.19 details the descriptive statistics of these factors and other variables for predicting the spending efficiency and effectiveness among municipalities in Thailand. Table 5.19 The Variables Predicting the Perceptional Budget Allocation for the Educati on and Public Works Departments Variable All Department Heads N Mean SD Dependent Perceptional budget allocation (efficiency and effectiveness) 744 1.88 0.85 Independent Administrative capacity 744 2.98 0.43 Information use for budget discussion 744 2.83 0.59 “Problem framing” (Strategy) 49 6.59% “It’s the mandate” (Strategy) 35 4.70% “Performance doctrine” (Strategy) 149 20.03% “We are experts” (Strategy) 90 12.10% “Incrementalism” (Strategy) 49 6.59% Control Commerce (Community Type) 64 8.60% Residential area (Community Type) 51 6.85% Mixed urban and suburban area (Community Type) 292 39.25% Agricultural area (Community Type) 334 44.89% Other (Community Type) 3 0.40% Capital city (Located in the provincial capital) 37 4.97% Population (1,000) 744 11.23 19.91 Staff’s commitment to information based decision making 744 3.16 0.78 Task is motivational 744 2.88 0.80 Normative isomorphism 744 3.16 0.79 Coercive isomorphism 744 2.67 0.97 Incrementalism 744 2.64 0.84 Conflict between the executives and the legislators 744 1.73 0.89 Conflict between the executives and the department heads 744 1.51 0.71 121

PAGE 134

Regression Results. Table 5.20 provides the results from regression analyses, indicating the impact of information utilization for budget discussion, the strategy of information use, and administrative capacity on the level of spending efficiency and effec tiveness reported by the department heads from the educatio n and public works departments. 122

PAGE 135

Table 5.20 Spending Efficiency and Effectiveness Variable (1) (2) (3) (4) Multiple Regression Ordered Logit Ordered Probit Ordered Probit (Robust Standard Errors) Residential (Community type) 0.011 0.002 0.019 0.019 (0.154) (0.406) (0.239) (0.253) Mixed urban and suburban 0.103 0.303 0.165 0.165 (Community type) (0.121) (0.314) (0.186) (0.197) Agricultural (Community type) 0.188 0.547* 0.303 0.303 (0.122) (0.319) (0.188) (0.202) Others (Community type) 0.495 1.072 0.676 0.676* (0.470) (1.004) (0.665) (0.382) Capital city of the province 0.146 0.424 0.244 0.244 (0.144) (0.373) (0.221) (0.236) Population (1,000) 0.002 0.004 0.002 0.002 (0.002) (0.004) (0.003) (0.003) Staff’s commitment to info base 0.009 0.009 0.013 0.013 decision making process (0.037) (0.092) (0.056) (0.055) Motivating tasks 0.012 0.040 0.024 0.024 (0.038) (0.097) (0.058) (0.059) Normative isomorphism 0.020 0.047 0.027 0.027 (0.038) (0.096) (0.058) (0.059) Coercive isomorphism 0.023 0.058 0.040 0.040 (0.031) (0.079) (0.047) (0.048) Incrementalism of the decision 0.136*** 0.360*** 0.215*** 0.215*** making process (0.035) (0.091) (0.054) (0.057) Conflict between the executives 0.017 0.073 0.041 0.041 and the legislators (0.034) (0.086) (0.052) (0.052) Conflict between the executives 0.096** 0.243** 0.147** 0.147** and the department heads (0.044) (0.113) (0.067) (0.069) Administrative capacity 0.436*** 1.169*** 0.700*** 0.700*** (0.074) (0.201) (0.118) (0.121) Information use for budget 0.310*** 0.814*** 0.491*** 0.491*** discussion (0.054) (0.143) (0.085) (0.086) “Problem framing” (Strategy) 0.015 0.018 0.030 0.030 (0.119) (0.305) (0.183) (0.180) “We are experts” (Strategy) 0.076 0.171 0.099 0.099 (0.141) (0.370) (0.219) (0.236) “Performance doctrine” (Strategy) 0.003 0.003 0.000 0.000 (0.074) (0.185) (0.112) (0.113) “It’s the mandate” (Strategy) 0.005 0.068 0.026 0.026 (0.091) (0.228) (0.135) (0.146) /cut1 6.190*** 3.662*** 3.662*** (0.913) (0.532) (0.555) /cut2 7.431*** 4.411*** 4.411*** (0.926) (0.537) (0.561) Constant 0.538 (0.335) Observations 744 744 744 744 R squared 0.183 Adj. R squared 0.161 Pseudo R squared 0.097 0.097 0.097 Robust standard errors in parentheses and *** p<0.01, ** p<0.05, * p<0.10 123

PAGE 136

The results of the ordinal regression analysis reveals that the community characteristics, political factors, structural isomorphism, organizational culture, political factors, information utilization, and administrative capacity are together related to the spending efficiency and effectiveness. The variables in the model predicted 9.7% of the variability in the perceptional budget allocation. As mentioned earlier, this study attempts to reduce the heteroscedasticity by introducing the robust standard err ors into the last model. For the multicollinearity, the variance inflation factor (vif) of each independent variable is less than four (ranging from 1.011.21). Consequently, the multicollinearity may also not be a reason of concern. The examination of e ach parameter estimate indicates that the level of information utilization for budget discussion, the administrative capacity, and the incrementalism of the decision making process are statistically significant at the 0.01 level. The highest predictor of t he spending efficiency and effectiveness is administrative capacity, with a standardized beta of 0.700. This means that if an average score of administrative capacity of a certain department head is to increase by one point, his or her ordered logodds of the level of spending efficiency and effectiveness (i.e., from low to moderate or from moderate to high level category) would increase by 0.784 while the other variables in the model are held constant. Also, for the information utilization, if an average score of information utilization for budget discussion of a certain department head is to increase by one point, his or her ordered log odds of the level of spending efficiency and effectiveness (i.e., from low to moderate or from moderate to high level cat egory) would increase by 0.491, other things being equal. In addition, if an average score of incrementalism of the decision making process reported by a certain department head is to increase by one point, his or her ordered logodds of the level of spending efficiency and effectiveness (i.e., from low to moderate or from moderate to high level category) would 124

PAGE 137

increase by 0.215 ceteris paribus. Furthermore, at the significance level of 0.10, the ordered logit for the department heads working for the commun ities heavily involving with industrial and tourism activities are 0.676 higher than those who working for the commerce community. Finally, the conflict between the executives and the department heads is statistically significant at the 0.05 level. However , this conflict is negatively associated with the spending efficiency and effectiveness. If an average score of the conflict between the executives and the department head is to increase by one point, his or her ordered log odds toward the level of spending efficiency and effectiveness (i.e., from low to moderate or from moderate to high level category) would decrease by 0.215 ceteris paribus. To conclude, for all the department heads in this study, the results confirm that the increase in the information utilization for budget discussion (H2A) and the administrative capacity (H2C) are positively associated with the spending efficiency and efficiency. Though, the strategy for information use (H2B) does not contribute to the perceptional budget allocation. In the other words, it can be summarized that the high level of administrative capacity and information utilization during the budget discussion among the department heads contribute to the high degree of efficiency and effectiveness of budget allocation especially am ong the municipalities involved in such economic activities as tourism and industrial enterprises. The current municipal budget practices also have an important impact on the increase in efficiency and effectiveness of budget allocations for l ocal public goods and services offered by the education and public works departments. Nevertheless, the conflict between the municipal executives and the department heads can significantly lower the spending efficiency and effectiveness of those two types of local goods and services. 125

PAGE 138

Perceptional budget allocations by the Education Departments Descriptive statistics. There are 403 education department heads participating in the survey. They often use information for budget discussion to a relatively high extent (2.93 out of 4.00 points) and possess relatively high level of administrative capacity (3.00 out of 4.00). Some of them (21.34 percent) show the ineffective policy issues and propose to the budget committee the solutions based on the performance measures. In their opinion, the information utilization can have a relatively low impact on the spending efficiency and effectiveness (1.92 out of 4.00). Table 5.21 details the descriptive statistics of these factors and other variables for pr edicting the spending efficiency and effectiveness among the education departments in Thailand’s municipalities. 126

PAGE 139

Table 5.21 The Variables Predicting the Perceptional Budget Allocati on in the Education Department Variable Education Department Heads N Mean SD Dependent Perceptional budget allocation (efficiency and effectiveness) 403 1.92 0.85 Independent Administrative capacity 403 3.00 0.43 Information use for budget discussion 403 2.93 0.55 “Problem framing” (Strategy) 29 7.20% “It’s the mandate” (Strategy) 12 2.98% “Performance doctrine” (Strategy) 86 21.34% “We are experts” (Strategy) 50 12.41% “Incrementalism” (Strategy) 32 7.96% Control Commerce (Community Type) 35 8.68% Residential area (Community Type) 19 4.71% Mixed urban and suburban area (Community Type) 161 39.95% Agricultural area (Community Type) 187 46.40% Other (Community Type) 1 0.25% Capital city (Located in the provincial capital) 24 5.96% Population (1,000) 403 12.15 22.94 Staff’s commitment to information based decision making 403 3.15 0.80 Task is motivational 403 2.86 0.77 Normative isomorphism 403 3.22 0.76 Coercive isomorphism 403 2.76 0.96 Incrementalism 403 2.63 0.84 Conflict between the executives and the legislators 403 1.66 0.86 Conflict between the executives and the department heads 403 1.42 0.67 Regression Results. Table 5.22 provides the results from regression analyses, estimating the impact of information utilization for budget discussion, the strategy of information use, and administrative capacity on the level of spending efficiency and effectiveness reported by the educa tion department heads. 127

PAGE 140

Table 5.22 Spending Efficiency and Effectiveness in the Education Departments V ariable (1) (2) (3) (4) Multiple Regression Ordered Logit Ordered Probit Ordered Probit (Robust Standard Errors) Residential (Community type) 0.045 0.187 0.124 0.124 (0.237) (0.592) (0.367) (0.324) Mixed urban and suburban 0.022 0.071 0.049 0.049 (Community type) (0.175) (0.455) (0.279) (0.267) Agricultural (Community type) 0.016 0.087 0.024 0.024 (0.178) (0.463) (0.282) (0.275) Others (Community type) 0.053 0.192 0.088 0.088 (0.805) (1.604) (1.087) (0.341) Capital city of the province 0.133 0.338 0.216 0.216 (0.184) (0.463) (0.286) (0.259) Population (1,000) 0.002 0.004 0.002 0.002 (0.002) (0.006) (0.003) (0.003) Staff’s commitment to info base 0.054 0.148 0.084 0.084 decision making process (0.049) (0.124) (0.075) (0.074) Motivating tasks 0.091* 0.267* 0.162* 0.162* (0.054) (0.140) (0.084) (0.087) Normative isomorphism 0.050 0.079 0.061 0.061 (0.053) (0.134) (0.081) (0.081) Coercive isomorphism 0.078* 0.213* 0.130* 0.130* (0.044) (0.113) (0.068) (0.068) Incrementalism of the decision 0.139*** 0.348*** 0.215*** 0.215*** making process (0.049) (0.127) (0.076) (0.077) Conflict between the executives 0.024 0.043 0.033 0.033 and the legislators (0.048) (0.122) (0.074) (0.073) Conflict between the executives 0.101 0.256 0.154 0.154 and the department heads (0.065) (0.173) (0.102) (0.107) Administrative capacity 0.438*** 1.122*** 0.692*** 0.692*** (0.102) (0.274) (0.162) (0.163) Information use for budget 0.369*** 0.987*** 0.593*** 0.593*** discussion (0.077) (0.208) (0.123) (0.124) “Problem framing” (Strategy) 0.274* 0.668* 0.401* 0.401* (0.154) (0.395) (0.239) (0.220) “We are experts” (Strategy) 0.430* 1.293** 0.722* 0.722* (0.240) (0.640) (0.380) (0.394) “Performance doctrine” (Strategy) 0.074 0.215 0.128 0.128 (0.097) (0.251) (0.151) (0.152) “It’s the mandate” (Strategy) 0.076 0.220 0.153 0.153 (0.124) (0.311) (0.188) (0.186) /cut1 6.775*** 4.082*** 4.082*** (1.309) (0.773) (0.776) /cut2 8.133*** 4.901*** 4.901*** (1.328) (0.780) (0.785) Constant 0.785 (0.481) Observations 403 403 403 403 R squared 0.215 Adj. R squared 0.177 Pseudo R squared 0.112 0.113 0.113 Robust standard errors in parentheses and *** p<0.01, ** p<0.05, * p<0.10 128

PAGE 141

The results of the ordinal regression analysis reveals that the community characteristics, political factors, structural isomorphism, organizational culture, political factors, information utilization, and administrative capacity are together related to the spening efficiency and effectiveness. The variables in the model predicted 11.3% of the variability in the perceptional budget allocation. As mentioned earlier, this study attempts to reduce the heteroscedasticity by introducing the robust standard errors into the last model. For the multicollinearity, the variance inflation factor (vif) of each independent variable is less than four (ranging from 1.021.36). Consequently, the multicollinearity may also not be a reason of concern. The examination of each parameter estimate suggests that the level of information util ization for budget discussion, the administrative capacity, and the incrementalism of the decision making process are statistically significant at the 0.01 level. The highest predictor of the spending efficiency and effectiveness is administrative capacity , with a standardized beta of 0.692. This means that if an average score of administrative capacity of a certain education department head is to increase by one point, his or her department’s ordered logodds of the level of spending efficiency and effecti veness (i.e., from low to moderate or from moderate to high level category) would increase by 0.692 while the other variables in the model are held constant. Also, for the information utilization, if an average score of information utilization for budget discussion of a certain education department head is to increase by one point, his or her department’s ordered logodds of the level of spending efficiency and effectiveness (i.e., from low to moderate or from moderate to high level category) would increase by 0.593, other things being equal. In addition, if an average score of incrementalism of the decision making process reported by a certain education department head is to increase by one point, the department’s ordered logodds of the level of spending e fficiency and effectiveness (i.e., from low to moderate 129

PAGE 142

or from moderate to high level category) would increase by 0.215 ceteris paribus. Furthermore, at the significance level of 0.10, the ordered logit for the education departments in which the members f ind their education services motivating, the level of spending efficiency and effectiveness would increase by 0.162 ceteris paribus. Finally, the “we are exper t” strategy of information use is statistically significant at the 0.10 level. T he ordered logit for these education departments whose heads apply this strategy are 0.676 higher than those who apply the “incrementalism” strategy of information use. However, the coercive isomorphism of budgeting practice and the application of the “problem framing” str ategy of information by the education department heads are negatively associated with the spending efficiency and effectiveness at the statistical signification level of 0.10. If an average score of the coercive isomorphism of budgeting practice is to incr ease by one point, his or her ordered logodds toward the level of spending efficiency and effectiveness (i.e., from low to moderate or from moderate to high level category) would decrease by 0.130 ceteris paribus. In addition, t he ordered logit for these educaiton departments whose heads apply the “problem framing” strategy are 0.401 lower than those who apply the “incrementalism” strategy of information use. In short, for the education departments, the results confirm that the increase in the information utilization for budget discussion (H2A), the strategy for information use (H2B), and the administrative capacity (H2C) have an impact on the spending efficiency and efficiency. In the other words, these results underscore the fact that the administrative capacity and the information utilization among the education department heads who heavily show their expertise in education services tend to have a positive impact on the increase in the efficiency and effectiveness of the spending for municipal education services. In addition, the organizational factors are also crucial in the levels of spending efficiency and effectiveness among the 130

PAGE 143

education departments. If the tasks related to the provision of educational services appears to be challenging and motiving to the staff members, it can drive the organizations to gain the relatively high levels of spending efficiency and effectiveness. Also, the current municipal budget practices have a positive influence on an increase in efficiency and effectiveness of budge t allocations for educational services. Though, the education department heads who focus on the problem framing—severity of the community’s problems —prevent their departments from gaining the spending efficiency and effectiveness. Additionally, the budget practices highly promoted and enforced by the central government (or the DOLA) can decrease the education services’ efficiency and effectiveness. Perceptional budget allocations by Public Works Departments Descriptive statistics. There are 342 public work s department heads participating in the survey. They often use information for budget discussion to a relatively high extent (2.73 out of 4.00 points) and possess a relatively high level of administrative capacity (2.98 out of 4.00). Also, most of them (18.42 percent) show the ineffective policy issues and propose to the budget committee the solutions based on the performance measures. In their opinion, the information utilization can have a relatively low impact on the spending efficiency and effectiven ess (1.92 out of 4.00). Table 5.23 details the descriptive statistics of these factors and other variables for predicting the spending efficiency and effectiveness among the education departments in Thailand’s municipalities. 131

PAGE 144

Table 5.23 The Variables Forecasting the Perceptional Budget Allocation in the Public Works Departments Variable Public Works Department N Mean SD Dependent Perceptional budget allocation (efficiency and effectiveness) 342 1.75 0.85 Independent Administrative capacity 342 2.98 0.45 Information use for budget discussion 342 2.73 0.62 “Problem framing” (Strategy) 20 5.85% “It’s the mandate” (Strategy) 40 11.70% “Performance doctrine” (Strategy) 63 18.42% “We are experts” (Strategy) 23 6.73% “Incrementalism” (Strategy) 17 4.97% Control Commerce (Community Type) 29 8.48% Residential area (Community Type) 32 9.36% Mixed urban and suburban area (Community Type) 131 38.30% Agricultural area (Community Type) 148 43.27% Other (Community Type) 2 0.58% Capital city (Located in the provincial capital) 13 3.80% Population (1,000) 342 10.15 15.53 Staff’s commitment to information based decision making 342 3.17 0.77 Task is motivational 342 2.91 0.84 Normative isomorphism 342 3.08 0.81 Coercive isomorphism 342 2.56 0.97 Incrementalism 342 2.65 0.85 Conflict between the executives and the legislators 342 1.82 0.91 Conflict between the executives and the department heads 342 1.59 0.75 Regression Results. Table 5.24 provides the results from regression analyses, indicating the impact of information utilization for budget discussion, the strategy of information use, and administrative capacity on the level of spending efficiency and effec tiveness reported by the public works department heads. 132

PAGE 145

Table 5.24 Spending Efficiency and Effectiveness in the Public Works Departments Va riable (1) (2) (3) (4) Multiple Regression Ordered Logit Ordered Probit Ordered Probit (Robust Standard Errors) Residential (Community type) 0.125 0.317 0.160 0.160 (0.209) (0.592) (0.339) (0.376) Mixed urban and suburban 0.231 0.781* 0.404 0.404 (Community type) (0.171) (0.464) (0.270) (0.291) Agricultural (Community type) 0.407** 1.215*** 0.676** 0.676** (0.172) (0.470) (0.274) (0.296) Others (Community type) 0.904 2.317* 1.333 1.333** (0.585) (1.403) (0.889) (0.640) Capital city of the province 0.174 0.824 0.312 0.312 (0.243) (0.734) (0.388) (0.512) Population (1,000) 0.001 0.004 0.002 0.002 (0.003) (0.008) (0.004) (0.006) Staff’s commitment to info base 0.057 0.119 0.085 0.085 decision making process (0.055) (0.141) (0.085) (0.084) Motivating tasks 0.048 0.144 0.086 0.086 (0.055) (0.145) (0.087) (0.086) Normative isomorphism 0.005 0.021 0.008 0.008 (0.056) (0.147) (0.087) (0.090) Coercive isomorphism 0.021 0.046 0.023 0.023 (0.045) (0.119) (0.070) (0.073) Incrementalism of the decision 0.125** 0.354** 0.212*** 0.212** making process (0.052) (0.138) (0.081) (0.085) Conflict between the executives 0.059 0.216* 0.117 0.117 and the legislators (0.049) (0.129) (0.078) (0.075) Conflict between the executives 0.098 0.255 0.154 0.154* and the department heads (0.061) (0.157) (0.095) (0.091) Administrative capacity 0.391*** 1.164*** 0.694*** 0.694*** (0.109) (0.308) (0.181) (0.181) Information use for budget 0.253*** 0.685*** 0.403*** 0.403*** discussion (0.078) (0.215) (0.125) (0.131) “Problem framing” (Strategy) 0.256 0.741 0.381 0.381 (0.189) (0.515) (0.307) (0.310) “We are experts” (Strategy) 0.293* 0.907* 0.503* 0.503 (0.177) (0.510) (0.295) (0.323) “Performance doctrine” (Strategy) 0.105 0.283 0.178 0.178 (0.115) (0.287) (0.173) (0.175) “It’s the mandate” (Strategy) 0.104 0.137 0.111 0.111 (0.137) (0.361) (0.208) (0.233) /cut1 5.658*** 3.292*** 3.292*** (1.316) (0.765) (0.782) /cut2 6.849*** 3.999*** 3.999*** (1.333) (0.771) (0.792) Constant 0.169 (0.470) Observations 342 342 342 342 R squared 0.199 Adj. R squared 0.151 Pseudo R squared 0.114 0.112 0.112 Robust standard errors in parentheses and *** p<0.01, ** p<0.05, * p<0.10 . 133

PAGE 146

The results of the ordinal regression analysis reveals that the community characteristics, political factors, structural isomorphism, organizational culture, political factors, information utilization, and administrative capacity are together related to the spending efficiency and e ffectiveness. The variables in the model predicted 11.2% of the variability in the perceptional budget allocation. As mentioned earlier, this study attempts to reduce the heteroscedasticity by introducing the robust standard errors into the last model. For the multicollinearity, the variance inflation factor (vif) of each independent variable is less than four (ranging from 1.021.35). Consequently, the multicollinearity may also not be a reason of concern. The examination of each parameter estimate suggests that the level of information utilization for budget discussion and the administrative capacity are statistically significant at the 0.01 level. The highest predictor of the spending efficiency and effectiveness is administrative capacity with a standardized beta of 0.694. This means that if an average score of administrative capacity of a public works certain department head is to increase by one point, his or her department’s ordered logodds of the level of spending efficiency and effectiveness would increase by 0.694 while the other variables in the model are held constant. Also, for the information utilization, if an average score of information utilization for budget discussion of a certain public works department head is to increase by one po int, his or her department’s ordered logodds of the level of spending efficiency and effectiveness would increase by 0.403, other things being equal. Furthermore, at the significance level of 0.10, the ordered logit for the public works department heads w orking for the communities heavily involving with agricultural activities as well as with industrial and tourism activities are 0.676 and 1.333 higher than those working for commerce communities . In addition, if an average score of incrementalism of the de cision making process reported by a certain public works department head is to increase by 134

PAGE 147

one point, the department’s ordered logodds of the level of spending efficiency and effectiveness would increase by 0.212 ceteris paribus. However, the conflict bet ween the executives and the public works department heads in a certain municipality is also statistically significant. If an average score of this conflict is to increase by one point, the municipality’s ordered logodds toward the level of spending effici ency and effectiveness would decrease by 0.154 ceteris paribus. In summary, for the public works department, the results confirm that the increase in the information utilization for budget discussion (H2A) and the administrative capacity (H2C) are positiv ely associated with the spending efficiency and efficiency. Though, the strategy for information use (H2B) does not contribute to the perceptional budget allocation for the public works services. In the other words, it can be said that the public works dep artment heads who possess high levels of administrative capacity and comprehensively utilize the relevant and reliable information during the budget discussion can lead their department to achieving high spending efficiency and effectiveness. Note that such association is pronounced among the c ommunities economically dependent on agriculture, industry, and tourism. Additionally, the current municipal budget practices also contributes to the increase in efficiency and effectiveness of budget allocations for local public goods and services offered by the public works departments. Nevertheless, the conflict between the municipal executives and the public works department heads can significantly reduce the spending efficiency and effectiveness of public works se rvices. Actual Budget Allocation The actual budget allocations for the education and public works departments are explained here in terms of budget allocation of FY2019. Note that this study analyzes the 135

PAGE 148

increase and decrease of budget allocation of FY2019 compared that of FY2018 in order not to allow the positive and negative effects to cancel each other out. Additionally, to capture the most current municipal budget process the department heads have experienced. This study considers the total funct ional spending for each department in terms of its natural log . The total expenditure includes such spending items as longterm investments, noncurrent ass ets, premises, maintenance cost, which directly involve developing or improving local public goods a nd services in form of fixed assets as well as operating expenditure which also directly facilitates the local public service provision (i.e., salaries, wages, current assets, and office supplies). The Education Service Programs (Decrease) The education service programs offered by the municipalities in Thailand are under the supervisions of the education departments. They are responsible for the primary schools and nonformal education, the conservation of local cultures and wisdoms, and the events involving sports and recreation. This part considers the 2019 total expenditures on education when it is lower and higher than those of previous year (2018). Descriptive statistics. The variables forecasting the education spending and their descr iptive statistics are shown in th e Table 5.25. 136

PAGE 149

Table 5.25 The Frequency and Descriptive Statistics of Variables Forecasting the Total E ducation Expenditure Variable Public Works Department N Mean SD Dependent Education expenditure 2019 55 14,856,413 35,035,743 Independent Administrative capacity 55 3.05 0.40 Information use for budget discussion 55 3.00 .055 “Problem framing” (Strategy) 3 5.45% “We are experts” (Strategy) 1 1.82% “Incrementalism” (Strategy) 5 9.09% “Performance doctrine” (Strategy) 11 20.00% “It’s the mandate” (Strategy) 5 9.09% Control Population (1,000) 55 10,837.20 11243.25 Area (km 2 ) 55 68.40 168.32 Number of department’s employees 55 13.78 11.70 Incrementalism in decision making process 55 2.54 0.92 Conflict between the executives and the legislators 55 1.56 0.79 Conflict between the executives and the department heads 55 1.32 0.61 Education expenditure 2018 55 21,831,490 70,470,224 Total revenue 2018 55 88,588,437 95,644,843 Regression results. The multiple regression models analyze whether and how the administrative capacity, the information utilization, and strategy of information use play a significant role in the municipalities’ budget allocations for goods and services for which the education departments are responsible. The results from the multiple regression anal yses are presented in Table 5.26. 137

PAGE 150

Table 5.26 The Variables Forecasting the E ducation Expenditure (Decrease) (1) (2) (3) (4) (5) Var iable Baseline Admin Capacity Admin Capacity & Information Use for Budget Discussion Admin Capacity, Information Use for Budget Discussion, & Strategy of Discussion Admin Capacity, Information Use for Budget Discussion, & Strategy of Discussion (Robust Std. Errors) Population (ln) 0.181 0.143 0.161 0.052 0.052 (0.187) (0.189) (0.195) (0.199) (0.135) Area (km 2 ) 0.000 0.000 0.000 0.000 0.000* (0.000) (0.000) (0.001) (0.001) (0.000) Number of department full time 0.016** 0.015* 0.015* 0.013 0.013* employees (0.008) (0.008) (0.008) (0.008) (0.007) Incrementalism of the decision 0.056 0.049 0.048 0.029 0.029 making process (0.088) (0.089) (0.091) (0.090) (0.090) Conflict between the executives 0.048 0.034 0.039 0.011 0.011 and the legislators (0.099) (0.100) (0.102) (0.101) (0.083) Conflict between the executives 0.133 0.159 0.157 0.210 0.210 and the department heads (0.120) (0.122) (0.124) (0.128) (0.150) Spending in education program 0.966*** 0.945*** 0.949*** 0.909*** 0.909*** 2018 (ln) (0.086) (0.088) (0.089) (0.088) (0.053) Total revenue 2018 (ln) 0.180 0.168 0.178 0.208 0.208 (0.247) (0.249) (0.254) (0.252) (0.212) Administrative capacity 0.255 0.261 0.433** 0.433** (0.180) (0.187) (0.204) (0.195) Information use for budget 0.004 0.016 0.016 discussion (0.142) (0.145) (0.068) “Problem framing” (Strategy) 0.044 0.044 (0.081) (0.056) “It’s the mandate” (Strategy) 0.130 0.130 (0.094) (0.102) “Performance doctrine” (Strategy) 0.079 0.079 (0.077) (0.054) “We are experts” (Strategy) 0.147 0.147** (0.088) (0.063) Constant 1.785 2.387 2.470 3.764 3.764 (2.651) (2.688) (2.740) (2.734) (2.866) Observations 57 56 55 55 55 R squared 0.873 0.878 0.879 0.897 0.897 Adj. R squared 0.852 0.855 0.853 0.862 0.862 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.10 138

PAGE 151

The results of the multiple regression analyses in Table 5.24 reveal that the demographics, political factors, information utilization, and administrative capacity are together related to the actual budget allocations in the education departments. The variables in the models predicted approximately 86.2% of the variability in the actual budget allocation. The W hite test confirms the homoscedasticity so that there should not be concern over the heteroscedasticity of the samples in these models. However, the interpretation of these model s should be done with caution. These models are possibly affected by the multicollinearity because their values of VIF for the 2018 total education expenditure and the population are higher than 4 (6.93 and 6.43 respectively). The examination of each par ameter estimating the 2019 education expenditure suggests that the 2018 spending level with a standardized beta of 0.909 is statistically significant at the 0.01 level. This confirms the prevalence of budgetary incrementalism. Meanwhile, the information ut ilization for budget discussion and the administrative capacity are statistically significant at the 0.05 level. This means that if an average score of administrative capacity of a certain education department head is to increase by one point (out of four point), his or her organization’s education spending may increase by 43.3% while the other variables in the model are held constant. Also, if the education department heads apply the “we are experts ” strategy while discussing their budget proposals, their departments’ budget allocation would increase by 5.4% higher those apply the strategy of incrementalism. The basic budget factors including the number of department staff and the municipality’s areas are also statistically significant at the 0.10 level. As suggested earlier, ones should interpret these models with caution. It can be said that the information utilization during the budget discussion, the strategy of information use, and the 139

PAGE 152

administ rative capacity are important factors determining the level of education expenditure, supporting H2A, H2B, and H2C. The Educa tion Service Programs (Increase) Descriptive statistics. The frequency and the descriptive statistics of variables such as the admi nistrative capacity, the information utilization, the strategy of information, and budget factors estimating the education spending a re shown in the Table 5.27. Table 5.27 The Frequency and Descriptive Statistics of Variables Predicting the E ducation Expenditure Variable Public Works Department N Mean SD Dependent Education expenditure 2019 119 17,608,826 43,492,762 Independent Administrative capacity 119 3.00 0.41 Information use for budget discussion 119 2.91 0.54 “Problem framing” (Strategy) 2 1.68 “We are experts” (Strategy) 3 2.52 “Incrementalism” (Strategy) 5 4.24 “Performance doctrine” (Strategy) 28 23.53 “It’s the mandate” (Strategy) 17 14.29 Control Population (1,000) 119 11394.79 12678.18 Area (km 2 ) 119 47.83 70.38 Number of department’s employees 119 14.50 11.45 Incrementalism in decision making process 119 2.57 0.87 Conflict between the executives and the legislators 119 1.56 0.80 Conflict between the executives and the department heads 119 1.35 0.58 Education expenditure 2018 119 15,202,407 95,106,377 Total revenue 2018 119 38,776,425 128,094,353 Regression results. The multiple regression models analyze whether and how the administrative capacity, the information utilization, and strategy of information use play a significant role in the municipalities’ budget allocations for goods and services for which the 140

PAGE 153

educatio n departments are responsible. The results from the multiple regression anal yses are presented in Table 5.28. 141

PAGE 154

Table 5.28 The Regression Results for the E ducation Expenditure (Increase) (1) (2) (3) (4) (5) Variable Baseline Admin Capacity Admin Capacity & Information Use for Budget Discussion Admin Capacity, Information Use for Budget Discussion, & Strategy of Discussion Admin Capacity, Information Use for Budget Discussion, & Strategy of Discussion (Robust Std. Errors) Population (ln) 0.175** 0.180** 0.180** 0.098* 0.098 (0.072) (0.073) (0.073) (0.053) (0.062) Area (km 2 ) 0.001 0.001 0.001 0.000 0.000 (0.001) (0.001) (0.001) (0.000) (0.000) Number of department full time 0.005 0.005 0.005 0.001 0.001 employees (0.004) (0.004) (0.004) (0.003) (0.003) Incrementalism of the decision 0.036 0.035 0.036 0.058** 0.058** making process (0.040) (0.040) (0.041) (0.029) (0.026) Conflict between the executives 0.032 0.035 0.034 0.033 0.033 and the legislators (0.047) (0.048) (0.048) (0.036) (0.022) Conflict between the executives 0.070 0.074 0.074 0.014 0.014 and the department heads (0.059) (0.061) (0.061) (0.049) (0.047) Spending in education program 0.702*** 0.701*** 0.700*** 0.821*** 0.821*** 2016 (ln) (0.041) (0.041) (0.041) (0.033) (0.094) Total revenue 2016 (ln) 0.081* 0.082* 0.084* 0.063* 0.063 (0.046) (0.047) (0.047) (0.034) (0.042) Administrative capacity 0.034 0.027 0.025 0.025 (0.085) (0.089) (0.068) (0.048) Information use for budget 0.018 0.071 0.071 discussion (0.068) (0.051) (0.076) “Problem framing” (Strategy) 0.037 0.037 (0.025) (0.029) “It’s the mandate” (Strategy) 0.021 0.021 (0.029) (0.022) “Performance doctrine” (Strategy) 0.015 0.015 (0.026) (0.029) “We are experts” (Strategy) 0.013 0.013 (0.030) (0.027) Constant 1.797*** 1.876*** 1.888*** 0.900 0.900 (0.683) (0.716) (0.720) (0.545) (0.743) Observations 130 129 129 119 119 R squared 0.896 0.896 0.896 0.951 0.951 Adj. R squared 0.889 0.888 0.887 0.945 0.945 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.10 The results of the multiple reg ression analyses in Table 5.28 reveal that the demographics, political factors, information utilization, and administrative capacity are together 142

PAGE 155

related to the actual budget allocations in the education departments. The variables in the models predicted approximately 94.5% of the vari ability in the ed ucation budget allocation. The W hite test fail to reject the heteroscedasticity assumption so that there should be concern over the issues of heteroscedasticity of the samples in these models. These models may not be affected by the multicollinearity because their values of VIF for each variable are lower than 4 (ranging from 1.07 to 2.76). However, the interpretation of these models should proceed with caution. The examination of each parameter estimating the 2019 education expenditure suggests that the 2018 spending level with a standardized beta of 0.821 is statistically significant at the 0.01 level. This confirms the prevalence of budgetary incrementalism. Additionally, the incrementalism in the decision making process is statisti cally significant at the 0.05 level. This means that if an average score of incrementalism in the decision making process of a certain department is to increase by one point (out of four point), his or her organization’s education spending may increase by 5.8% while the other variables in the model are held constant. In conclusion, it can be said that the information utilization during the budget discussion, the strategy of information use, and the administrative capacity are not important factors determi ning the level of education expenditure , failing to support H2A, H2B, and H2C, if the education departments have the higher level of expenditure than the previous current year. The Public Works Expenditure (Decrease) Descriptive statistics. The frequency and the descriptive statistics of variables such as the administrative capacity, the information utilization, the strategy of information, and budget factors forecasting the municipalities’ spending on public w orks are shown in the Table 5.29 . 143

PAGE 156

Table 5.29 The Frequency and Descriptive Statistics of Variables Predicting the Publ ic Works Expenditure Variable Public Works Departments N Mean SD Dependent Total spending on public works 61 7,244,874 9,045,962 Independent Administrative capacity 61 2.96 0.48 Information use for budget discussion 61 2.73 0.60 “Problem framing” (Strategy) 2 3.28% “It’s the mandate” (Strategy) 6 9.84% “Performance doctrine” (Strategy) 11 18.03% “We are experts” (Strategy) 4 6.56% “Incrementalism” (Strategy) 2 3.28% Control Population density (persons/km 2 ) 61 769.68 1,461.72 Number of department full time employee 61 8.62 5.74 Incrementalism of the decision making process 61 2.79 0.86 Conflict between the executives and the legislators 61 1.70 0.86 Conflict between the executives and the department heads 61 1.57 0.69 Spending in public works programs 2018 61 9,485,349 10,426,723 Total revenue 2018 61 65,682,121 64,966,291 Commerce (Community type) 3 4.92% Residential (Community type) 4 6.56% Mixed Urban & Suburban (Community type) 29 47.54% Agricultural (Community type) 25 40.98% Other (Community type) 0 0% Regression results. This study applies the multiple regression models to analyze whether and how the administrative capacity, the information utilization, and strategy of information use play a significant role in the municipalities’ budget allocations for goods and services for which the public works depar tments are responsible. The results from the multiple regression anal yses are presented in Table 5.30. 144

PAGE 157

Table 5.30 The Regression Results for the Publ ic Works Expenditure (Decrease) (1) (2) (3) (4) (5) V a riable Baseline Admin Capacity Admin Capacity & Information Use for Budget Discussion Admin Capacity, Information Use for Budget Discussion, & Strategy of Discussion Admin Capacity, Information Use for Budget Discussion, & Strategy of Discussion (Robust Std. Errors) Population density 0.000 0.000 0.000 0.000 0.000 (0.000) (0.000) (0.000) (0.000) (0.000) Number of department full time 0.008 0.010 0.010 0.009 0.009 employees (0.007) (0.009) (0.009) (0.013) (0.011) Incrementalism of the decision 0.094* 0.083 0.090 0.078 0.078 making process (0.055) (0.057) (0.060) (0.065) (0.057) Conflict between the executives 0.006 0.023 0.028 0.004 0.004 and the legislators (0.055) (0.062) (0.064) (0.072) (0.076) Conflict between the executives 0.075 0.068 0.053 0.060 0.060 and the department heads (0.072) (0.077) (0.082) (0.083) (0.098) Commerce (Community type) 0.072 0.095 0.085 0.094 0.094 (0.236) (0.245) (0.249) (0.246) (0.174) Residential (Community type) 0.002 0.058 0.095 0.039 0.039 (0.193) (0.212) (0.246) (0.257) (0.240) Mixed Urban & Suburban 0.124 0.126 0.124 0.099 0.099 (Community type) (0.103) (0.107) (0.109) (0.116) (0.105) Spending in education program 0.876*** 0.854*** 0.849*** 0.853*** 0.853*** 2018(ln) (0.068) (0.071) (0.073) (0.081) (0.163) Total revenue 2018 (ln) 0.370*** 0.389*** 0.404*** 0.399*** 0.399* (0.114) (0.122) (0.125) (0.134) (0.223) Administrative capacity 0.019 0.039 0.086 0.086 (0.114) (0.118) (0.122) (0.118) Information use for budget 0.068 0.003 0.003 discussion (0.090) (0.105) (0.107) “Problem framing” (Strategy) 0.006 0.006 (0.060) (0.073) “It’s the mandate” (Strategy) 0.077 0.077 (0.067) (0.074) “Performance doctrine” (Strategy) 0.085 0.085 (0.058) (0.063) “We are experts” (Strategy) 0.032 0.032 (0.057) (0.049) Constant 4.459** 4.377** 4.683** 4.748** 4.748** (1.871) (2.117) (2.191) (2.299) (2.217) Observations 67 64 63 61 61 R squared 0.876 0.858 0.859 0.882 0.882 Adj. R squared 0.854 0.828 0.826 0.837 0.837 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.10 145

PAGE 158

The results of the multiple regression analyses in Table 5.30 reveal that the demographics, political factors, information utilization, and administrative capacity are together related to the actual budget allocations in the public works departments. The variables in the models predicted approximately 83.7% of the variability in the actual budget al location for public works. The W hite test confirms the homoscedasticity assumption so that there should not be concern over the heteroscedasticity of the samples in these models. These models are possibly unaffected by the multicollinearity because their values of VIF for all variables are less than 4 (ranging from 1.18 to 2.22). However, the interpretation of these models should be done with caution due to the limited numbers of the sample (61). The examination of each parameter estimating the 2019 public works expenditure suggests that the 2018 spending level and the 2018 total revenue with a standardized beta of 0.853 and 0.399 are statistically significant at the 0.01 and 0.05 level respectively. This also highlights t he prevalence of budgetary incrementalism. In sum, it can be said that the information utilization during the budget discussion, the strategy of information use, and the administrative capacity are not the important factor determining the level of public works expenditure in this case , failing to support H2A, H2B, and H2C. The Public Works Expenditure (Increase) Descriptive statistics. The frequency and the descriptive statistics of variables such as the administrative capacity, the information utilization, the strategy of information, and budget factors forecasting the municipalities’ spending on public w orks are shown in the Table 5.31. 146

PAGE 159

Table 5.31 The Frequency and Descriptive Statistics of Variables Predicting the Publ ic Works Expenditure Variable Public Works Departments N Mean SD Dependent Total spending on public works 53 9,965,725 16,056,019 Independent Administrative capacity 53 2.95 0 .4 9 Information use for budget discussion 53 2.77 0.53 “Problem framing” (Strategy) 1 1.89% “It’s the mandate” (Strategy) 8 15.09% “Performance doctrine” (Strategy) 15 28.30% “We are experts” (Strategy) 5 9.43% “Incrementalism” (Strategy) 3 5.66% Control Population density (persons/km 2 ) 53 690.566 959.5428 Number of department full time employee 53 10.57 10.98 Incrementalism of the decision making process 53 2.66 0.85 Conflict between the executives and the legislators 53 1.91 0.90 Conflict between the executives and the department heads 53 1.62 0.79 Spending in public works programs 2018 53 5,952,704 5,374,344 Total revenue 2018 53 105,439,647 256,284,707 Commerce (Community type) 4 7.55% Residential (Community type) 4 7.55% Mixed Urban & Suburban (Community type) 22 41.51% Agricultural (Community type) 23 43.40% Other (Community type) 0 0% Regression results. This study applies the multiple regression models to analyze whether and how the administrative capacity, the information utilization, and strategy of information use play a significant role in the municipalities’ budget allocations for goods and services for which the public works departments are responsible. The results from the multiple regression anal yses are presented in Table 5.32. 147

PAGE 160

Table 5.32 The Regression Results for the Publ ic Works Expenditure (Increase) (1) (2) (3) (4) (5) V ariable Baseline Admin Capacity Admin Capacity & Information Use for Budget Discussion Admin Capacity, Information Use for Budget Discussion, & Strategy of Discussion Admin Capacity, Information Use for Budget Discussion, & Strategy of Discussion (R obust Std. Errors) Population density 0.000 0.000 0.000 0.000 0.000 (0.000) (0.000) (0.000) (0.000) (0.000) Number of department full time 0.024** 0.024** 0.018* 0.014 0.014 employees (0.011) (0.011) (0.010) (0.011) (0.009) Incrementalism of the decision 0.282** 0.298** 0.197* 0.277** 0.277** making process (0.107) (0.115) (0.115) (0.121) (0.115) Conflict between the executives 0.024 0.043 0.050 0.061 0.061 and the legislators (0.103) (0.116) (0.109) (0.119) (0.082) Conflict between the executives 0.156 0.179 0.138 0.104 0.104 and the department heads (0.124) (0.133) (0.126) (0.125) (0.089) Commerce (Community type) 0.180 0.199 0.047 0.084 0.084 (0.368) (0.379) (0.368) (0.362) (0.247) Residential (Community type) 0.972** 0.992** 0.960** 1.355*** 1.355*** (0.448) (0.471) (0.443) (0.496) (0.404) Mixed Urban & Suburban 0.257 0.244 0.199 0.319 0.319 (Community type) (0.216) (0.221) (0.208) (0.214) (0.255) Spending in education program 0.659*** 0.634*** 0.618*** 0.651*** 0.651*** 2018(ln) (0.085) (0.096) (0.090) (0.092) (0.143) Total revenue 2018 (ln) 0.025 0.015 0.022 0.019 0.019 (0.090) (0.093) (0.088) (0.085) (0.112) Administrative capacity 0.077 0.083 0.060 0.060 (0.215) (0.211) (0.221) (0.191) Information use for budget 0.500** 0.459** 0.459* discussion (0.189) (0.203) (0.247) “Problem framing” (Strategy) 0.088 0.088 (0.108) (0.098) “It’s the mandate” (Strategy) 0.200* 0.200** (0.103) (0.089) “Performance doctrine” (Strategy) 0.100 0.100 (0.078) (0.083) “We are experts” (Strategy) 0.077 0.077 (0.099) (0.095) Constant 6.691*** 7.271*** 8.252*** 7.821*** 7.821*** (1.823) (2.044) (1.955) (2.013) (1.996) Observations 60 57 57 53 53 R squared 0.662 0.612 0.666 0.746 0.746 Adj. R squared 0.593 0.518 0.575 0.633 0.633 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.10 148

PAGE 161

The results of the multiple regression analyses in Table 5.31 reveal that the demographics, political factors, information utilization, and administrative capacity are together related to the actual budget allocations in the education departments. The variables in the models predicted approximately 63.3% of the variability in the actual budget al location for public works. The W hite test confirms the homoscedasticity so that there should not be concern over the heteroscedasticity of the samples in these models. However, the interpretation of these models should be done with caution. These models are possibly affected by the multi collinearity because their values of VIF for the variable of communities located in the residential areas and the population are higher than 4 (4.9). The examination of each parameter estimating the 2019 public works expenditure suggests that the 2018 spending level and the municipalities located in residential areas with a standardized beta of 0.651 and 1.355 respectively are statistically significant at the 0.01 level. This also shows that budgetary incrementalism prevails. Also, the municipality locate d in residential areas may spend 13.55% higher than those located in the tourist spots or industrial areas. Furthermore, the “it’s the mandate” strategy of information use is statistically significant at the 0.05 level. This implies that if the public works department heads relate their budget proposal to the mandates from the central government, their departments would receive 20.0% higher in budget allocation than those applying the strategy of incrementalism. However, the incrementalism of decision makin g process are negatively associated with the public work spending at the statistical significance level of 0.05 and 0.10 respectively. In sum, it can be said that the strategy of information use is an important factor determining the level of public work s expenditure , supporting the H2B . Nonetheless, t he administrative capacity is not significantly associated with the public work spending level , 149

PAGE 162

failing to support H2C . Additionally, the information utilization during the budget discussion is indeed negati vely associated with the public works expenditure , failing to support H2A . In conclusion, this chapter quantitatively analyzes the survey data in order to identify the determinants of information utilization and budget allocation according to the first an d the second subresearch questions respectively. Table 5.33 summarizes the hypothesized determinants, which is statistically significant, based on the results from the regression analyses. Table 5.33 Summary of Regression Results for the Information Utilization Department Heads H1A (Relevant or reliable information) H1B (Performance information users) H1C (Administrative capacity) H1D (Techno rational and policymaking) Education and public works Education , , and mean entirely support, partially support, and do not support the corresponding hypotheses respectively It can be said that administrative capacity, relevant information, and budget orientations (i.e., policymaking, technorational, and advocate) are the determinant of information utilization among department heads during the budget process. However, the per formance information is not a significant determinant. For the spending efficiency and effectiveness, Table 5.34 summarizes the results from the regression analyses. Based on this table, administrative capacity, information utilization, and strategy of information use ( “we’re expert”) are the determinant s of spending efficiency and effectiveness in the programs under the supervision of an education department and a public works department. However, the strategy of information use is statistically significant only for the case of educational program. 150

PAGE 163

Table 5.34 Summary of Regression Results for the Spending Efficiency and Effectiveness Spending Efficiency and Effectiveness H2A (Information Utilization) H2B (Strategy of information use) H2C (Administrative capacity) Education and public works departments Education department and mean entirely support and do not support the corresponding hypotheses respectively For the actual budget allocation in Table 5.35, strategy of information use and administrative capacity are the determinants of programmatic spending under education and public works departments. However, the regression analyses estimating the actual budget allocations apparently risk internal validity or the casual linkage between strategy of information use and administrative capacity and budget allocations due to the fact that there exist hetero scedasticity and multicollinearity . Accordingly, this study cautiously limits its interpretation and generalization based the results related to these actual budget allocations. Table 5.35 Summary of Regression Results for the S pending Efficiency and Effectiveness Budget Allocation H2A (Information Utilization) H2B (Strategy of information use) H2C (Administrative capacity) Education department (Increase) Educational department (Decrease) Public works department (Increase) Public works department (Decrease) * Note: and mean entirely support and do not support the corresponding hypotheses respectively * means that the hypothesis is statistically significant but it is of the opposite direction 151

PAGE 164

CHAPTER VI THE QUALITATIVE RESULTS The mixed methods design of this study applies the quali tative approach to complement the quantitative resul ts explained in the last chapter. This chapter qualitatively demonstrate s the concepts of interest especially by means of the causal mechanism as well as by the example budget practices that the department heads mention during the interview. The first group of these concepts address answering the first research question —how do department heads with different administrative capacities utilize reliable and relevant information during a budget process? Then the second group provides the answers for the second re search question—how does the information utilization influence budget decisions? The Determinants of Information Utilization Asked about the factors determining how they utilize the information during a budget process, the department heads first explain how they utilize the information. Then they mention roughly six groups of factors which play a role on their information utiliz ation. Accordingly, this section begins with the description of information utilization among the department heads. Later it elaborates the related six concepts influencing the levels of such information utilization. Information Utilization There are two department head s that report barely us ing the information while initiating local development projects. They strictly follow the mayor’s policy directions and also believe that the mayor actually knows more about the community than they do. Though, convers ely, seven department heads give examples of how they actually use the information during the budget process. 152

PAGE 165

To draft a budget proposal. First, five department heads discuss how relevant information is used in drafting and supplying all the details for budget proposals. Typically, the budget proposal format demands such details as the community’s problem, rationale for proposed solutions, budget and costing data , target groups, and policy outcomes. Besides, the department heads depend on this information in order to roughly conduct a possibility study for the projects aiming at solving a specific problem. They also prioritize those projects according to the possibility of study. To support their decisions. Second, two department heads elaborate that they us e relevant information to comple ment other factors driving a decision making. For example, even though there is a specific request from the community, the public works department heads often thoroughly weigh in such request with the construction standar ds. Factor s Affecting Information Utilization Administrative Capacity The administrative capacity here refers to skills, knowledge, or technical capability the department heads consider necessary for initiating, presenting, discussing, and defending their budget proposals. D epartment heads identify nine capacities as indispensable for playing their role during the budget process. Law and regulations. First, the knowledge regarding rules, laws, and regulations . Four department heads state that they must be able to understand budgetary cycle and required activities, f ulfil department’s responsibilities and key functions, and also comply with new or updated mandates from DOLA. Professional expertise. Second, two department heads insist that they must possess and master their expertise. The case in point in this study’s context includes the specializations in 153

PAGE 166

education services and construction management. Not only mastering it, they must be able to train or transfer such knowledge to their superiors and other municipality staff. Promotion of civic engagement. Third, t wo department heads perceive that the skills promoting civic engagement is also crucial. For instance, they have to engage in the town meeting such as moderating the meeting, attentively listening to problems and demands raised by the local residents, carefully gathering such information and accurately reflecting it into the community’s development plan. Furthermore, they must painstakingly seek opinions and suggestions from every group of stakeholders. Financial management. Fourth, four department heads t hink that they are required to be adept at prioritizing the department’s projects according to both budget constraints and demands from the residents, applying cost accounting in order to estimate the cost of each departmental project or program, mastering Microsoft Office programs especially Excel, and estimating the department’s revenues and expected annual budgets allocated by the municipality, the central government, and other agencies. Project management. Fifth, one department head specifies that worki ng at the local government means that you must be skillful in project management. One should be able to draft a complete project and provide such details as rationales, process, target groups, estimated budget , and expected results; monitor the project as it proce eds; and eventually evaluate the projects especially via a participant satisfaction survey. Public relations and communication. Sixth, five department heads identify that the skills involving public relations are absolutely important. For instan ce, one must be able to coordinate and congenially work with the stakeholders and make sure the stakeholders understand and willingly participate in the municipality’s activities. 154

PAGE 167

Negotiation. Seventh, one department head reports that he must be good at negotiating. In doing so, he has to not only inform the residents of the projects they may not fully support but also talk back and forth between the policymakers and the residents in order to reach an agreed upon policy direction. Strategic planning. Ei ghth, three department heads consider the strategic planning skills as part and parcel of their everyday job. This is because they have to apply strength, weakness, opportunity, and threat ( SWOT ) analysis to the community development plans and finally draf t the strategic plans for the department. Motivation and empowerment techniques. Ninth, one department head says that it is also important to be able to motivate their subordinates. He states that from time to time he empowers his staff but also focus on the performance, not the detailed work processes. Budget Rationality Spending preferences among the department heads vary and most of the time are latent. However, a few department heads during the interview s reveal their attitudes toward the municipal spending for their functional programs. These prioritized preferences are expressed as categorized two budget orientations. Accounting orientation. The first group of department heads identifies (two from publi c works department heads and two from education department heads) themselves with an accounting orientation who highly values accounting principles, efficiency, and effectiveness. The fiscal roles they often play include to make sure that the standard price and cost accounting are correct and transparent; that the public services or policy results from a certain project are worth its investment or payment; that the department’s functional programs and projects are 155

PAGE 168

cost saving (i.e., using the materials from local sources or creating jobs for the local workforce); and that such investment or payment effectively solves the community’s problems. Techno rational orientation. The second group, two education department heads, represents the techno rational budge t orientation, balancing the policy outcomes and fiscal resources while initiating and preparing their budget proposal s . They first contemplate the local problems and the service demands from the residents. Next, they carefully take into consideration the municipality’s responsibilities, target groups, and available funds. Then, they prioritize those problems or demands. They eventually initiate the projects or programs responding to those prioritized problem or demands in form of local development plans. Structural isomorphism The department heads explain how they participate in their organization’s budget processes especially through the managerial patterns from two main sources—called coercive and normative isomorphism. Coercive isomorphism. Regarding the coercive isomorphism of local budgeting process, three education department heads underscore the role of DOLA upon the local governments’ budgeting practices. DOLA creates the performance budgeting’s work process and standard operating procedures. It provides the releva nt trainings and workshops and require s the attendance and participation of the municipal clerks, accountants, and financial managers. These train ings and workshops mostly cover the planning process, the local development directions, the budget proposal formats, and the practice of strategic planning. The local staff must rigorously follow these practices because they are one of the key issues of financial and accounting appraisa l conducted by the office of auditor general. 156

PAGE 169

Normative isomorphism. For the normative isomorphism of local budgeting, one of the public works department heads explains that his department’s budget proposals must follow the standards of construction mana gement. The construction plans proceed under the monitoring and with the approval of the construction engineer. Information Used during the Budget Process There are approximately eleven types of information the department heads utilize during the budget process. Different information types re present different perspectives of urgency and importance of policy issues as the following examples show. Community problems. First, three department heads identify that they often start their budget proposals with t he alarming problems happening in the communities. This includes, for instance, infectious diseases among schoolchildren, heavy rains during the monsoon season, fire, traffic congestion, and car accidents. Public service clients’ satisfaction. Second, thr ee department heads often consider performance information while initiating their budget proposals. They especially take into account the results of satisfaction survey (i.e., service quality, outcome, and policy impact), the praise from the target audience, and the comments from the event participants. Budget ceilings or constraints. Third, five department heads contemplate their budget constraints while prioritizing their projects in the local development plans. The previous year operating budgets and capital budgets, the revenue forecasts, the own source revenue statistics, and the intergovernmental fiscal transfers from the central government. Policy platforms. Fo urth, four department heads state that the policy platforms (i.e., organization’s vision st atements, elements of mission, and policy goals and purposes set by the executives) are also crucial for the departments’ plans. The residents vote for certain mayors 157

PAGE 170

mainly because of their policy packages or highlighted policy domains. Some departments s tate that they must strictly follow these policy directions. Community demands and requests. Fifth, service demands and requests from the communities are also paramount. Five department heads emphasize the indispensability of this type of information. Hav ing to gather this information is the reason why they are required to attend the town meetings, to prepare the stakeholder analysis, and to seek th e community leaders’ opinions. The required service quality and standards. Sixth, the standard of local publ ic services i s necessary, especially since the central government enforces the use of monitoring and evaluation system called the Local Performance Assessment (LPA) on all local governments. One department head states that the LPA aims to improve the local service delivery in terms of its quality and its delivery process that is transparent, fair, and accountable. Policy results. Seventh, five department heads discuss how the information regarding policy results ha s a significant impact on their budget proposals. National accolades for local students, literacy rate, higher household income, health, quality of life are examples of this type of information. This type of information has an advantage over others because it is easy for the budget actors to understand and use it to compare their performance with other local governments. Programmatic outputs. Eighth, one department head mentions the output of service is also important. Road construction is the case in point. Its routes, length, width, and height are the details the department head must include in their budget proposal and elaborate it to the municipal budget committee. 158

PAGE 171

The central government’s plans and policies. Ninth, the policy direction of the central governments is another type of information to which the department heads pay close attention. From time to time, the central government issues warnings, the request for cooperation, and the mandate from the DOLA. Three department heads note the inclusion of these agendas into their budget proposals. Crisis and emergencies. Tenth, emergencies are absolutely important when the municipalities plan their local development plans. Despite the significant impro vement in infrastructure, the localities are still disaster prone. The municipalities still have to tackle with drought, flashflood, and fire. One department head puts these emergencies into his plans in order to lessen their impacts or to prevent the loss in the future. Professional knowledge. Eleventh, seven department heads point out the importance of professional knowledge. This type of information includes road construction, soil density, city planning, and geography. It is indispensable for the publi c works department heads to draw their construction plans . The information sources All of the information types mentioned in the last part are derived from several sources. However, the department heads state that they receive the relevant information pr imarily from these following information sources. Self observation. The first information source from which the department heads gather related data is to observe such data by themselves. Ten department heads explain that self observation in this context includes such inquiries as observing the classroom activities, asking the school teachers about their suggestions on education service delivery, and walking through the potential construction sites with the community leaders. 159

PAGE 172

The central government’s agencies. The second information source the department heads rely on is the central government’s agencies. Two department heads state that they must know the responsible agency for each issue of interest. For instance, the education department staff search fo r the statistics of contagious diseases among student children from the department of disease control. Service providers or contractors. The third information source is the municipalities’ subsidiaries or the municipal entities which provide local public goods or services to the residents within a certain jurisdiction. The elementary schools, for example, keep and analyze their operating and capital budgets. The education department heads can retrieve such information while drafting their budget proposals . The Summary of the Determinants of Information Utilization Each qualitative explanation of a certain factor , above , which is called a theme supplies details for answering the first research question. Table 6.1 provide the summary of these details aiming to explain the factors accounted for the variation in information utilization during the municipal budget process. 160

PAGE 173

Table 6.1 Summary of Determinants of Information Utilization Concept Theme Frequency Percentage 1 1. Administrative capacity 1.1 Rules, laws, and regulation 4 25.00 1.2 Expertise 2 12.50 1.3 Civic engagement 2 12.50 1.4 Financial management and accounting 4 25.00 1.5 Project management 1 6.25 1.6 Public relations 5 31.25 1.7 Negotiation 1 6.25 1.8 Strategic planning 3 18.75 1.9 Motivation skill 1 6.25 2. Budget rationality 2.1 Accounting orientation 4 25.00 2.2 Techno ration al orientation 2 12.50 3. Structural isomorphism 3.1 Coercive isomorphism 3 18.75 3.2 Normative isomorphism 1 6.25 5. Information pieces 5.1 Community problems 3 18.75 5.2 Performance indicators 3 18.75 5.3 Budget ceilings 5 31.25 5.4 Policy platforms 4 25.00 5.5 Community demands and requests 5 31.25 5.6 Standard of public service delivery 1 6.25 5.7 Policy results 5 31.25 5.8 Output 1 6.25 5.9 Mandates 3 18.75 5.10 Emergency 1 6.25 6. Information source 6.1 Self observation 10 62.50 6.2 Central government agencies 2 12.50 6.3 Municipality’s entities 1 6.25 1 From the total of 1 6 department heads The Determinants of Budget Allocation When interviewed about budget allocation, the department heads discuss ed the strategies of information use and other determinants. The following parts describe these determinants in detail. 161

PAGE 174

Strategies of Information Use Just answer the questions. First, six department heads solely focus ed on answering the questions from the budget committee during the budget discussion session. It is fundamentally their main responsibility in attending such session. However, they must be well prepared for potential questions especially by making sure that they truly understand all the details of the projects they are proposing. The questions from the committee are mostly about costing, the purchase of some items, and the modes of service delivery. Focusing on the target groups. Second, six department heads usually explain to the municipal budget committee how and what their target groups receive from the proposed programs or projects. They highlight the benefits or positive impacts a certain budget proposal gives to the residents, the community, and even to the organizations. The y also introdu ce the key mechanisms directly delivering those benefits or impacts so that the committee can be sure of such success. Highlighting urgency of the problem. Third, six department heads argue the urgency of their projects around the relevant problems the r esidents are tolerating. They mention, for instance, the costs the affected people have to bear and the worst case scenario or the opportunity loss if the projects are unfunded. Presenting the project’s efficiency. Fourth, there are two department heads heavily discuss their proposed projects’ effectiveness or efficiency to the budget committee. For example, they present how those projects can save t he municipality’s budgets by using the local materials and asking for volunteers from the community. Align ing the projects with policy platforms. Fifth, two department heads concentrate on following the policy platforms set by the municipality’s executives. These directions always 162

PAGE 175

guide the budget allocations in their municipalities. These department heads are not willing to propose projects that seem to go against those policy directions. Lessening the intensity of disasters or emergencies. Sixth, three department heads usually explain their projects by showing how they can respond to the emergency or how they can lessen the problem’s intensity or severity. Additionally, sometimes they show the opportunity loss or how the problematic situation can get worse if the municipality does not choose to fund those projects. Supporting the mandates. Seventh, one department head reports that the mandate from the central government directs how he plans his department’s local development plans. He contemplates the national agendas as well as the local development policy from DOLA before outlining the department’s budget proposals. Picturing the output. Eighth, two department heads elaborate the output their projects provide during the budget discussion. For example, they detail the building construction and sometimes provide the pictures to the budget committee. Prov iding technical knowledge. Ninth, two department heads thoroughly present their projects through their technical knowledge. Technical knowledge, for instance, provide a reason why they choose a certain design or a particular choice of construction over the others. They strongly make sure that the committee understand s it so that the committee can technically explain the technical components to the residents later on when the municipality carries out such construction project. Other determinants Related in formation pieces. First, the specific types of information are actually the key for budget allocation. Ten de partment heads report that information as the community problems 163

PAGE 176

and demands, policy results (especially with identified target groups), the budget ceilings (i.e., projected ownsource revenues and estimated grants and transfers from the central government), and the executive’s policy directions. Project values. Second, one department head mentions that the budget allocation depends on the cost of each project. A project which requires high amount of financial resource is unlikely to be approved since a municipality rarely has enough budget to do so. Such a huge project is therefore divided into much smaller projects , which are mostly implemented over the next few fiscal years. The central government’s policy platforms. Third, four department heads admit that budget committee s are often likely to align their allocation decisions with the central government’s policy directions. These policy directi ons come in forms of cooperation as well as mandates. Political factors. Fo urth, two department heads discuss ed the importance of political factors in the local governments’ budget allocation. For example, if a municipal election approaches, the incumbent s are likely to spend significantly more on public works projects. In addition, the projects aiming at the target group at the neighborhood level usually get financially supported. Budget distribution. Fifth, the distribution of financial resources to eac h function is also a critical issue for the budget allocation decisions. One department head point s out that the committee usually makes sure that every community or neighborhood has a balanced share of the municipality’s financial resources. Other departments’ projects. Sixth, the projects of other departments also play a crucial role in the municipality’s budget allocation. Two department heads mention that they carefully 164

PAGE 177

draft their budget proposals to avoid redundancy with other departments budge t committee s are known to dismiss a budget request that finds the other department’s plans more important or more urgent. The auditor general office’s comments. Seventh, one department head states that the budget committee are against the projects the au ditor general usually disapproves. To get the projects revoked by the auditor general can create an administrative scandal. It is not worth risking their effort, which can eventually become at least futile or at worst scandalous. Table 6.2 below summariz es the determinants of budget allocation all the department heads mention during the interview. Table 6.2 De terminants of Budget Allocation Concept Theme Frequency Percentage 1 1. Strategy of information use 1.1 Just answering the questions 6 37.50 1.2 Focusing on target groups 6 37.50 1.3 Highlighting urgency of the problem 6 37.50 1.4 Presenting the projects’ efficiency 2 12.50 1.5 Aligning the projects with policy platforms 2 12.50 1.6 Lessening the intensity of disasters or emergencies 2 12.50 1.7 Supporting the mandates 1 6.25 1.8 Picturing the output 2 12.50 1.9 Providing technical knowledge 2 12.50 2. Information related factors 2.1 Information pieces 10 62.50 2.2 Project value 1 6.25 2.3 Central government’s policy platforms 4 25.00 2.4 Political factors 2 12.50 2.5 Budget distribution 1 6.25 2.6 Other departments’ projects 2 12.50 2.7 The auditor general office’s comments 1 6.25 1 From the total of 16 department heads The results from this qualitative component and those of quantitative component comprehensively discussed in the previous section are triangulated in the next two chapters in 165

PAGE 178

order to answer the study’s research questions as well as discussion the findings based on all the results reported in this chapter. 166

PAGE 179

CHAPTER VII DISCUSSION ON INFORMATION UTILIZATION Based on the results of Chapter 5 and Chapter 6 , this chapter discusses those results mainly in response to the first research question of how do the policymakers with different administrative capacities utilize relevant and reliable information during a budget process? Particularly, this chapter first provides the answers to this research questions by describing the patterns determining the information utilization among the department heads from both functions (i.e., education and public works). After that, it discusses and investigates the findings ba sed on the crucial concepts and their interrelation in those patterns especially from the perspective of the oretical and practical relevance considering the extant literature and the context of Thailand’s decentralization previously delineated in Chapter 2 and Chapter 4, respectively. Overall and FunctionBased Information Utilization Different organization units delivering different local public services tend to have different patterns determining the level of information utilization. This study focuses on two functions which are key to Thailand’s local development —education and public works. In order to directly answer the first research question, the following parts in this section offer the overall pattern influencing the information utilization and the specific patterns in each function. Overall information utilization RQ1 — A: Overall, how do the policymakers with different administrative capacities utilize relevant and reliable information during a budget process? In case of the department heads from both functions of education and public works, it can be said that department heads who possess the relatively high level of administrative capacity (i.e., being adept in soliciting civic engagement, conducting financial and accounting analyses, 167

PAGE 180

applying pr oject management, making use of public relations and communication, and administering strategic planning) and the highly relevant information (i.e., community problems and demands, public service satisfaction report, service quality, policy results, and p rofessional knowledge) are likely to intensively utilize such information for initiating and discussing their budget proposals to their municipal budget committee. However, if the department heads have been committed to preset policy directions guided by their own political beliefs or the mayor’s policy platforms, this commitment can keep their levels of in formation utilization low . Particularly, if the department heads identify themselves with policy making and try to pursue t he policy directions which they believe desirable according to their own policy and financial analyses or if they identify themselves with an advocacy budget orientation and restrictively follow the policy directions initiated by the municipal executives, the se department heads tend to utilize all the relevant information throughout the budget process less than ones identifying themselves wit h conduit budget orientation who only transmits all the information from their staff to the budget committee. This in formational behavior highlights the symbolic use of information which involves using information to legitimate and sustain predetermined positions or confirm programs the information users wish to promote (Feldman & March, 1981; Knorr, 1977; Weiss, 1977) . Information Utilization by the Education Department Heads RQ1 — B: For the educational services, how do the policymakers with different administrative capacities utilize relevant and reliable information during a budget process? Like the previous cases of overall department heads, the education department heads’ administrative capacity and budget rationality is strongly associated with their level of information utilization. Especially, the education department heads with budget rationality of 168

PAGE 181

technorational orientation who simultaneo usly pay attention to several budgetary factors such as policy outcomes, fiscal resources, service demands, organization’s missions, target groups, and budget constraints tend to modestly utilize the related information for initiating and discussing budget ing. Somewhat different from the case of all department heads, the organization culture appears to have a significant impact on the information utilization among the education department heads. On one hand, the self driven organization culture (i.e., the c lose knit and family like work environment and the budgeting practice shared among the education professionals themselves) appears to have a positive impact on the information utiliz ation. On the other, the support from the outsiders (i.e., the executives and other staff members ) are negatively associated with the information utilization. This finding underscores the fact that a group’s sentiments, which involve social values, noneconomic motive s , and interests, are fundamental in informational behaviors of budget actors (Dingwall & Strangleman, 2007) . Information Utilization by the Public Works Department Heads RQ1 — B: For the public works services, how do the policymakers with different administrative capacities utilize relevan t and reliable information during a budget process? Like the case of all department heads, the public works department heads’ administrative capacity and the relevant information are positively associated with their level of information utilization while budget rationality of advocacy orientation is negatively associated with it. Additionally, the organization cultures in terms of coercive structural isomorphism (the budget practice enforced by the DOLA) tends to decrease the information utilization by the public works department heads. 169

PAGE 182

Discussion and Findings Given the results shown in Chapter 5 and Chapter 6, this section triangulates those results in order to discuss the interrelations of such results with the related theories and the current studies identified in Chapter 2. This section extricates the interrelations explaining the information utilization and budget allocation as follow s . Administrative Capacity, Relevant Information, and Information Utilization In general the administrative capacity and relevant information together drive the information ut ilization. Administrative capacity strongly facilitates the department heads’ involvement in gathering and processing the relevant information, thereby increasing the level of its utilization during a budget process (Moynihan, Lavertu, et al., 2012; Sun & Gargan, 1993) . T he study finds that the type of administrative capacity matters while the attitude towards spending does not. This section details the process of how administrative capacity together with relevant information leads to the increase or the decrease in information utilizat ion during a budget process. From Capacity to Budget and Information Production to Information Utilization during the Local Budget Process Generally speaking, corresponding to the personal judgement based on the principal agent theory (Ellsberg, 1961) and the previ ous studies concentrating on administrator tenure (Goodman & Clynch, 2004) as well as the academic discipline ( Lee, 1991) , the results of this study reassert s that administrative capacity is strongly associated with the information utilization through a budget process. Specifically, the administrative capacity under the context of Thailand’s local budgeting involves the capacity to budget among the frontline public administrators who ac tually participate in gathering, processing, and disseminating relevant 170

PAGE 183

information (Arguello, Callan, & Shulman, 2008; Blazenko, 1987; Spence, 1973; Witting, 2015) , or, in other words, who actively engage in the information production in order to serve the market of local public goods and services (Blazenko, 1987; Spence, 1973) . The following components of capacity to budget in this study underscore this finding. Participatory Budgeting. The information discussed during the town meetings are in dispensable for the local development plans. In order to have access to and al so gath er such information, budgeteers have to actively take part in the town meetings and understand the residents’ opinions. An education department head in the east highlights this capacity by stating that “civic engagement is a must for improving our schooling and educational services. The municipality alone cannot go straight to our education goals. Opinions and suggestions from all groups of stakeholders are necessary for our policy priority and budget proposals.” Also, a public works department head from the south describes the importance of civic engagement by explaining that “I always participate in the community meetings, painstakingly listening to the residents and bringing their requests into our plans. Later, based on such plans, the executives and the communities together choose the projects to be funded.” Communication skills. Being a good listener or a good negotiator is indeed crucial for gathering the important information for the local development plans. An education department head from the east points out that “in order to properly respond to all stakeholders and literally embrace them into our department plans, we experience the educational issues preferably first hand by directly discussing the educationrelated problems and demands with the s tudents, the parents, the teachers, and the municipal executives before drafting our budget proposals.” An education department head from the north also adds that “Not everyone agrees with a certain educational policy. Before including a policy agenda into the budget proposal, sometimes we 171

PAGE 184

have to repeatedly explain it to the whole community. We go out, negotiating back and forth between the executives and the community members until that agenda sounds acceptable for all stakeholders.” Financial management and accounting. Despite the support from the finance department staff, the func tional department heads find that possessing financial management and accounting skill is indispensable for them to calculate and allocate funds for each project. A public works department from the east states that “we must be able to estimate the budget size our department is likely to be allocated so that we can properly prioritize and choose the potential projects worth our limited budgets. ” In addition, a public works departm ent head from the south explains that while drafting and defending his budget proposals, “the department staff must be able to estimate both the total cost and the unit cost of the project. We can show, for example, for this type of building, how much it w ould cost or for this local road, how much the municipality has to pay per square meter. So that we can compare and contrast the costs of possible projects and defend why we choose one instead of another.” A public works department head from the north also describes that “the budget allocation and the total revenue from the previous years are the basis for estimating our next year ’ s budget size. We have to distribute it across every community so as to solve all of their pressing problems.” Strategic planning. The department heads must be able to think ahead strategically and anticipate the changes potentially affecting their local development plans. One public works department from the east posits that “we have to predict how the city is sprawling as well as how other types of urbanization can affect our initial spatial plans. These conditions finally result in the increase in demand of all types of infrastructures —tap water, sewage, electricity. Our budget proposal should be forwardlooking, embracing these infrastructure demands and avoiding the 172

PAGE 185

potential problems. Remember that if not provided today, the residents are going to come back at us and repeatedly make requests for such infrastructures anyway.” Thus, skills in participatory budgeting, comm unication, financial management and accounting, and strategic planning are critical for information production and information use. These four components of capacity to budget depict how the elements of administrative capacity can promote the information production. With these capacities, department heads bring into the municipal budget planning, formation, and approval such relevant information as community problems, values of money for each project, the mayor’s policy and priority, and budget constraints or ceilings especially throug h two causal mechanisms—professional judgment and involvement in the informational processing. Professional judgment as one of the solving mechanisms of informati on a symmetry represents technical competence, knowledge, and specialization and is significantly accumulated through training and experiences (Karl, 1976) . It technically helps the public administrators value the relevance of the information when the perfect or complete information regarding local public services is not guaranteed (Ellsberg, 1961) . Specifically, regarding the information asymmetry, these frontline public a dministrators are highly involve d in information production by choosing what information to collect (Rich, 1975) as in transmitting that information, cognizing that information, referencing that information in budget proposal, attempting to adopt such information, influencing choices and decisions based on that information, and giving rise to application and extension of that information to policymakers concerned ( Knott & Wildavsky, 1980; Landry et al., 2001) Their roles i n information production result in the more effective utiliz ation of relevant information since this information assists the department heads in planning their budget requests, comparing 173

PAGE 186

one choice of spending to another, and prioritizing the community problems and their possible solutions . To sum up, in the logic al consequence, it can be said that the higher level of capacity to budget particularly in areas of participatory budgeting, communication, financial management and accounting, and strategic planning among the frontline public administrators leads to the higher level of their involvement in information production. This results in the higher level of availability of relevant information which brings about the higher level of information utilization. Type of Budget Rationality, Relevant Information, and Information Utilization A different story about the use of information emerges where department heads have an ideological perspective of budgeting. Budget rationality reflects another component of administrat ive capacity —the policymakers’ attitudes about economic ideology particularly towards government spending or spending preferences (Reck, 2000; Willoughby, 2002) —and significantly influences how they selectively and purposefully attend to, encode, and process information (McCue, 1999) . The budget rationality of advocacy, policymaking, and technorational orientat ions are all associated with low level of information utilization when comparing to the conduit orientation, which represents a relatively neutral preference on public spending or on unspecific fiscal policy directions. “Policy” Budget Norm and Symbolic Use of Information. This finding of the study suggests the highly influential impact of the policy budget norm on information ut ilization during the local budget process in Thailand. The policy budget norm , which mirrors the strategic planning and advocacy for and responsiveness to a particular policy direction, appears to have a stronger impact on the information utilization than do other norms which focus either on control 174

PAGE 187

(fiscal accountability and accuracy) or management (efficiency and effectiveness) (David, 1998; Smith, 2003; Willoughby, 2002) . No matter from which source this policy direction originally comes (from the mayors’ priority, the comprehensive policy analyses, or the sophisticated budgetary techniques from DOLA for the case of the advocating, policymaking, and technorational orientations, respectively), it predominates other norms and leads the department heads to legitimate and sustain such predetermined policy positions (Amara, Ouimet, & Landry, 2004; Weiss, 1977) . As a result, policymakers use information as political ammunition or only to confirm the programs they wish to promote and reflects the advocating budget orientation (Weiss, 1977) . This highlights the symbolic use of information. A few department heads mentioned their emphasis on the use of policy platform initiated by their mayors and/or their municipal executives. A public works department head from the s outh asserts that “the mayor is someone for whom the majority of the resident voted. More importantly, eventually the budget distribution all comes to the hand of the mayor. So I would rather follow his policy and priority from the start.” The public works department head from the north also strongly follows his mayor’s policy, mentioning that while outlining his budget proposal, “I first put the mayor’s policy into our local development plan then estimate the costs of the projects supporting such policy wi thin the estimated budget assigned to m y department.” In addition, a public works department head from the east states that “the mayor’s vision on local development is something I have to follow. What is the point of proposing the project conflicting with the vision? To have our department’s budget cut? Or doing so to indirectly encourage the mayors to allocate the funds to other departments instead?” The Co Existence of Instrumental, Conceptual, and Symbolic Use of Information. Nonetheless, it is worth not ing that this symbolic use of information coexists with instrumental 175

PAGE 188

and conceptual uses which feature applying information to solve clearly predefined problems as a vital part of a rational decision making process and/or to indirectly and unspecifically i nfluence actions as a general enlightenment (Amara et al., 2004) . To some extent similar to these two latter uses, the policy norm possibly coexists with control and management norms in the case of the department heads with the technorationa l and policymaking orientations, which directs the information utilization toward the rational decision making process. Relevant Information and Information Utilization The relevant information in this study includes both performance information (value of money, policy results, and service quality) and other information (community problems, professional techniques, budget ceiling, and executive’s policy and priority). The results suggest that the type of information is not associated with the level of information utilization. The department heads who usually deliberate performance and o ther information do not utilize the information at different levels. Even though the DOLA promotes the use of performance measures among the municipalities, the results highlight the equal importance of non performance information. Local Knowledge. The descriptive statistics in Table 5.4 suggest that the most used information pieces among the department he ads are the community’s problems and requests (96.65%) and executive’s policy and priority (82.71%). Both of them are information which are related to local contexts or settings and often contain specific characteristics, circumstances, events, and relati onship as well as important understanding of their meaning, degree of verification, and status to common sense, causal empiricism, or thoughtful speculation and analysis (Corburn, 2003; Lin dblom & Cohen, 1979) . 176

PAGE 189

An education department head from the central area highlights the vitality of these two types of local knowledge, stating that “problems and requests from the local people and policy directions from the municipal executive are communitybased. Our works are c ommunityfocused and we go handin hand with the community. So that we always ask the loca l people of what they want, what they can contribute. We compromise this information with the executive’s policy platform. Our department is also responsive to their policy and priority. Anyways, these two sources of information are often insync because t he may or and the executives represent local people and obviously depend on their votes.” A public works department head from the south highly values the executive’s policy and priority, asserting that “truthf ully, the mayor and other executives know every well what and where the problem is. Admittedly, they know and understand the community more than we do because they always mingle with the local residents. They attend every single ritual or ceremony celebrating or happening in the community. Ordination, w edding, funeral, you name it!” In conclusion, the administrative capacity especially possessing skills in participatory budgeting, communication, financial management and accounting, and strategic planning can promote information utilization during the l ocal budget process among department heads. However, budget rationality of advocacy, policymaking, and technorational orientations is all associated with low er level of information utilization. Organization Cultures and Information Utilization Beside t he concepts of interest —administrative capacity and budget rationality, there is other concept which determines the level of information utilization by department heads. Based on the results, organization cultures also matter. 177

PAGE 190

The organization culture is an important determinant in organizational development as well as effectiveness (Schraeder, Tears, & Jordan, 2005; Sminia & Van Nistelrooij, 2006) . Thereby, as a means of organizational development, the budget reform must take into account the organization culture. This study’s results show how structural isomor phism, group culture, and organizational support for change influence the information utilization in the context of Thailand’s local governments. C oercive Structural Isomorphism The coercive isomorphism or t he enforcement of the strategic performancebase d budget system by DOLA is negatively associated with the information utilization through a budget process determining the spending level of public works services. A few public works department heads explain how they can comply with this new budget practic e especially for the procurement process. One public works department head from the east mentions that “there is this new procurement done electronically. It seems too complicated for us. Our staff still lacks of any electronical skills. Usually we have a hard copy to back it up.” Another one from the east adds that “DOLA invited us to attend the local budgeting training. We learn and understand how to prepare the local development plans as well as its budgeting system. However, we cannot really follow it b ecause we do not have enough resource and staff to do so.” One from the north states that “we completely follow the budget practices suggested by the DOLA. Comply with it or mess with the Office of Auditor General. No worries! We absolutely don’t want to r isk it.” This apparently points out that this new budgetary structure becomes a pure cost from the viewpoint of efficiency because it is incompatible with current organizational conditions especially in terms of administrative capacity (Meyer & Rowan, 1977) . Additionally, the concentration on accounting and budgetary principles that DOLA imposes on the public works 178

PAGE 191

departments’ procurement procedures suggest s the hierarchical structure of control and authority from the central government to the local governments requiring the latter’s conformity (White & Smoke, 2005) . Additionally, the department heads are obligated to pursuing the technical improvement rather than the policy results. These control and authority structure as well as technical obligation accentuate a mechanical form of management system, which unintentionally increases a tendency of vertical (top down) interaction between superior and subordinate in the organization (Burns & Stalker, 1961) . Under these conditions , t he department heads’ information utilization during the budget process (i.e., the discussion and exchange of the information with the budget committee) can be relatively low. Nor mative Structural Isomorphism. However, the DOLA’s new budget practice is sometimes implemented through the combination of normative isomorphism and group culture shared by informal group members. The normative isomorphism is positively associated with the information utilization throughout a budget process especially for determining the spending level of educational services. The normative isomorphism under this study’s context means the budget practice or other related standard of procedure widely promote d by the educations professionals and widely accepted by the educational services staff. Many department heads during the interview imply the application of performance budgeting. They create the budget proposals and defend them based on the students’ perf ormance and achievement, comprehensively utilizing the standard scores of Ordinary National Educational Test (O NET) or Advance d National Educational Test (A Net) and comparing these scores to those of other local governments. A few education heads explain how they prepare the budget proposal based on such test scores. An education department head from the north points out that, to effectively def end his budget proposal to the municipal budget committee, “ONET is apparently the most valid and 179

PAGE 192

effective measure for students’ performance. I always show them whether our students’ average ONET score is higher (or lower) than the national, regional, and provincial ones. I also provide the details of how each of our four primary school s performs in terms of ONET. Which school should be monitored more closely? Why is it so and what to do in order to bring up their ONET?” An other education department head from the north says that “I cite the ONET score a lot because it is the standardized score. Ministry of Education and its regional office (the Primary Education Service Area Office) provide the recommendations on how to effectively use and interpret it.” Additionally, one from the central region gives the example of how she prepares the budget proposal for teacher training projects, expressing that “I analyze both ONET and ANET scores of our schools. Which academic field (mathematics, language, sciences, or social sciences) our students perform poorly? I propose that the teachers in such f ield should go attend the training.” An education department head from the east provides the additional information regarding her education service based training. “The Primary Educational Service Area Office holds and invites us to join the educational service training. They also have the teachers from the excellent schools winning a lot of education al accolades show us how they manage their teaching programs and provide extra curriculum activities. Their examples inspire us and, more importantly, show us the concrete path to be an excellent educational service provider.” Group Culture. Additionally, among the education department heads, group culture related to the c ohesion and morale tends of the people in the department, tends to be positively associ ated with the level of information utilization. This suggests that there exists an informal force which draws the departmental members together and creates a sense of connectedness and 180

PAGE 193

harmony, which provides guidance allowing the staff to be more supporti ve of the department’s mission (Schraeder et al., 2005) . Note that both normative structural isomorphism (the budget practice shared among the educational departments) and the group culture underscore an organic form of management systems in terms of the characteristics of the contributive nature of special knowledge and experience on educational service provision and the lateral direction of communication resembling co nsultation rather than command, respectively (Burns & Stalker, 1961) . Consequently, it can said that the institution affiliation which highlights organic form of management systems supports the information utilization by the education department heads (Burns & Stalker, 1961; Moynihan, 2008) . Organizational Support for Change in Budget Practice. In order to implement and promote the ne w budget practice called the strategic performance based budget system , the municipalities’ chief executives and managerial leaders understand that such change typically invol ves a political process so they build internal supports for this procedural chang e and reduce resistance to it through their commitment to such new budget practice (Fernandez & Rainey, 2006) . However, the demonstration of a strong commitment to achieving results by municipal top leadership and the overall staff is indeed negatively associated with the information utilization by the education department heads. One possible explanation to this inverse relationship is that the education department heads may perceive this commitment to the new budget practice as a pseudo c ontrol (Hofstede, 1978) which unexpectedly drives defensive routines (Argyris, 1990) . The education department heads can possibly perceive their leaders’ and other staff members’ st rong commitment to the strategic performancebased budget system as a hierarchical control measure which outlines the 181

PAGE 194

mechanistic form of management system (Burns & Stalker, 1961) . They potentially react to this commitment by suspicion and resistance which unintentionally encourage s more hierarchical communi cation and thus highly restricts upward flow of information from the department heads to the municipal budget committee (Henri, 2006) . This process of pseudo control and defensive routines eventually leads to the low information utilization. In conclusion, it can be said that the organization cul ture which helps department heads naturally adopt a new budget practice (i.e., group culture and normative isomorphism) can promote their information utilization. Meanwhile, the organization cultures , which involve conformity and control, such as a leader’ and other staff members’ strong commitment to the strategic performancebased budget system and coercive isomorphism can limit the information utilization. Theoretical Contributions As articulated in Chapter 2, the main theories this study appl y are theories of information economics and the budget rationality model. Meanwhile, their relationships with information utilization are possibly mediated by other theoretical concepts such as isomorphism. This part discusses how the results confirms and adds the fine theoretical nuances to these theories. Theories of Information Economics The results regarding the administrative capacity, relevant information, and information utilization reassert the crucial role of information production that frontline public administrators play in the local public services market. Relying on their professional judgment or their capacity to budget, the department heads (agents) provide the relevant information to their municipal decisionmaking process or for the municipal executives and legislators (principals) to make budget decisions appropriately (Blazenko, 1987; Heinkel , 1982; Ross, 1973) . Owing to the 182

PAGE 195

proximity in every locality, the chain of principal agent relationship in the local governments (in this case from citizen to politician to bureaucrat ic superior) does not imply a long hierarchy so that the principals and agents in the local context here can easily observe, gather, validate, disseminate, and even crosscheck the relevant information without substantial cost (Moe, 1984) . Budget Rationality Model The results pertaining to the budget orientations accentuate how the budget norms or the spending preferences dictate the types of information gathered and deliberated as well as to which extent a certain public administrator utilize s them throughout a local budget process (McCue, 1999) . Specifically, advocacy, policymaking, and technorational orientations are associated with low level of information utilization. This show s the strong influence the policy budget norm inversely has on the information utilization mostly due to the predetermined policy directions and the symbolic use of information (Knorr, 1977; Weiss, 1977) . Structural Isomorphism This study confirms the influence of sociological institutionalism (the culturally specific practice assimilated into organization not necessarily to enhance their formal meansends efficiency but as a result of the transmission of cultural practice more generally) on information utilization (Hall & Taylor, 1996) . Specifically, this proposition is pronounced when this institutionalism is intertwined with the organization culture. The normative isomorphism and organization culture of group culture can bring up the conformity of new budgetary structure and its desirable informational behaviors in form of increased information utilization during the budget process. On the opposite end, the finding regarding coercive isomorphism and t he formal organizational support that budget change s are negatively associated with the information utilization. 183

PAGE 196

Empirical Contributions to the Current Literature and Empirical Studies The influence that administrative capacity and information characteristics have on the information utilization is the center of this study’s first research question. Its answers and how the study answers such question add new empirical perspectives to the current literature as follows. Capacity to Budget: Alternative Constructs of Administrative Capacity Under the umbrella of the determinants of information utilization, most of the current studies usua lly operationalize administrative capacity of policymakers in terms of th e educational background such as their length of education in college level (Nielsen & Baekgaard, 2015) , their fields of education and school (Kroll, 2015b) , their educational levels (Moynihan & Ingraham, 2004; Moynihan, Pandey, & Wright, 2012; Zhang, Van de Walle, & Zhuo, 2016) , and in terms of the work experiences such as years in their supervisor positions (Dull, 2009; Melkers & Willoughby, 2005; Moynihan & Pandey, 2010; Taylor, 2011) . The studies exploring the administrative capacity in terms of the publ ic administrators’ skills and knowledge to directly carry out the budgetary tasks is still limited. This study, thus, attempt s to develop and tailor new measures of the administrative capacity so as to meet local budgeting challenges in the developing countries (Farazmand, 2009; Tankha, 2009) . It offers the alternative measures of administrative capacity —the capacity to budget which reflects the skills essential for drafting, analyzing, presenting, and defending the budget proposals —and empirically proves that this persp ective of administrative capacity is indeed a strong determinant of information utilization among all the frontline public administrators delivering both educational and public works services in Thailand’s local governments . 184

PAGE 197

Frontline Public Administrators Generally speaking, the game of public policy formulation and implementation is no longer solely dominated by politicians ; management matters as much as policy design does (Denhar dt, 2002; Ingraham et al., 2003) . Often many current studies investigate the use of information among the chief executives and the legislators (Bourdeaux & Chikoto, 2008; Goodman & Clynch, 2004; Hou et al., 2011b; Raudla, 2012) while earlier a few paid attention to that of agency budget analysts or department heads (McCue, 1999; Thurmaier, 1992) . The information utilization here involves the latter group of policymakers (the department heads who directly provides local pub l ic services to the communities), because they have usually been overlooked in the current public management literature (Dimitrijevska Markoski & French, 2019) , in order to assess the new empirical elements (i.e., capacity to budget and budget norms) into the pattern of information utilization. It is found that frontline public administrators with high level of knowledge and skills in in participatory budgeting, communication, fina ncial management and accounting, and strategic planning are likely to comprehensively utilize information during a budgeting process. Local Knowledge The extant studies overwhelmingly attend to how the performance measures, a type of information, are used and impac t the budgetary decision making (Laegreid et al., 2008; Melkers & Willoughby, 2005; Moynihan, 2008a, 2015; Pollitt, 2006) . Meanwhile, some studies regarding the information utilization in the field of public policy center solely around local knowledge as a type of information (Corburn, 2003; Maiello, Viegas, Frey, & Jos, 2013; Smit, 2005) . T his study emphasize s the simultaneous us e of both types of information, performance i nformation and local knowledge . I t highlights the fact that in the context of the developing country the 185

PAGE 198

performance information is equally important as local knowledge. The department heads during the interviews confirm t he necessity of local knowledge. T his is consistent with wide recognition among development experts that local knowledge is an important type of information, for example, many African academics and researchers have perceived local knowledge infrastructure as important component of human de velopment (Kolawole, 2015) . The Do minance of Policy Budget Norm across the different Budget Orientations Similar to other studies based on the U.S. experiences (Thurmaier, 2001) , th e results in this study confirm the relationships between the budget orientations and information utilization. However, this study empirically investigates this relationship not in terms of the types of information the budget actors use (Goodman & Clynch, 2004; Willoughby, 2002) but in terms of the level of their information utilization. The results suggest that the policy budget norms (strategic planning and advocacy for and responsiveness to policy outcome) is more dominant than control (fiscal accountability accuracy) and management (efficiency and effectiveness). Based on this dominance, the results add the new findi ng to the current literature —budget actors without policy budget norm, those identifying with conduit and accounting budget orientations , can actually utilize information more comprehensively than those identifying with technorational, policymaking, and advocating orientations. Structural Isomorphism and Organization Culture Generally, inst itutional forces cannot simply be neglected in research on performance measurement and management in the public sector (Ashworth, Boyne, & Delbridge, 2009) . The current st udy provides the sequential process of how the performance budgeting reform at the national level evolves over time in Thailand (Sutheewasinnon et al., 2016) . This study investigates how the isomorphic pressures influence the budgeting process in terms of 186

PAGE 199

information utilization at the local level. T he results of this study argue that the interaction between isomorphic pressure and organization culture can promote the adoption of performance budgeting among the local governments. Nevertheless, while the current studies assert that the chief executives’ and other staff members’ commitment to the performance budgeting is positively associated with the information utilization (Dull, 2009; Moynihan et al., 2012) , in contrast to the literature , this study finds that the higher level of such commitment together with the coercive isomorphism tends to lower the i nformation utilization during the municipal budget process. Practical Contributions The findings point out to two possible mechanisms that can promote the higher level of information utilization among the local frontline public admini strators in developi ng counties — the customized capacity building program and the institutional design encouraging the information utilization. Even though this study monitors the performance budget reform, the mechanism proposed here involves the human resource development . E ve ry kind of public sector reform eventually and always ends up with human resource development because reforms are based on the notion that sufficient capable administrators cannot always be found (Ingraham, 2009; Tankha, 2009) . Meanwhil e, institutional support is also necessary for smoothing and organically supporting the change in informational behaviors. The Capacity Building Program The capacity building program here involves and aims at the capacity to attend to the available information ( Simon, 1973) . The finding in the previous section reports the relationships between the administrative capacity (both capacity to budget and the budget orientations) and the information utilization. In order to increase the level of information utilization among the local 187

PAGE 200

frontline public administrators, this stud y suggests that the department heads with the policy norms are the highest priority to attend the capacity building programs because they are unlikely to comprehensively utilize information during a local budget process. Hence, in order to increase the information utilization, the capacity building program should assign a priority to them before others. Institutional Design There are two institutional support systems for smoothing and facilitating the behavior al change that hopefully can bring the higher level of information utilization among the frontline public administrators at the local level in developing countries. First, the related capacity building programs should be led by the professionals in the fields so as to naturally promote the behavioral change regarding how the department heads utilize the information. Second, the commitment to the information utilization should not be from the chief executives or other staff mem bers but from the workgroup to which each department head belong. Possibly, the loc al governments should allow department heads to choose the staff members they prefer to join the capacity building program with them. This can indirectly promote the strong group culture whose members sha re the same training experiences, thereby together supporting the department heads to more comprehensively use information during the municipal budget process. 188

PAGE 201

CHAPTER VIII DISCUSSION ON BUDGET ALLOCATION This chapter discusses how the concepts of interest make a logical and causal connection to the municipality’s budget decisions in terms of both actual and perceptional budget allocation. It begins with answers to the second research question of this study: how does the information utilization influence budget decisions ? After that, it discusses the main concepts from the theoretical framework such as incrementalism, administrative capacity, information utilization, and the context of Thailand’s decentraliz ation . This chapter discusses how budget ary incrementalism, the management components, and the intergovernmental relation have impact on budget allocation. Finally, this chapter considers how the findings from the discussion contribute to the theoretical d evelopment, the current literature, and practitioners. Actual and Perceptional Budget Allocation Actual Budget Allocation How does the information utilization influence budget decisions? In general, the data support that department heads use information but how and when they use it depends on their administrative capacity and budget rationality. Then this information use influences the budget decisions in the municipalities. For the educational services, if the education department heads possess a relative ly high level of administrative capacity and apply the “we’re experts” strategy of information use, then it is likely that the municipal budget committee allocates more budget to their departm ents than those departments where the department heads possesses relatively low administrative capacity and apply other strategy of information use. However, the effect of incrementalism (i.e., the previous year spending ) exists in the municipal budget allocation. 189

PAGE 202

Moreover, other basic budget factors —the numbers of full time employees and the jurisdiction’s areas— are also associated with the level of budget allocation. Regarding public works servi ces, the previous year’s spending and the budget factors — the residential areas and each municipality’s total revenue —drive the increase in the department’s budget allocation, similar to the experience in educational services. The differences is that if the department heads apply the “it’s mandate” strategy of information use, then it is likely that the municipal budget committee allocates more budget to their departm ents than to those departments where the department heads possess the relatively low administrative capacity and apply other strategies of information use. Additionally, t he information utilization is indeed negatively associated with the municipality’s budget allocation for public works services. Perceptional Budget Allocation Since the increase or decrease in actual budget allocation do es not necessarily portray the efficiency and effectiveness of such allocation, this study investigates the perceptions of the department heads regarding their mun icipal budget allocation in key service areas. Overall, the higher administrative capacity and information utilization, the higher level of efficiency and effectiveness of municipal budget allocation. In addition, the difference in policy directions between the chief executive and the department heads can lower efficiency and effectiveness in the local servic e programs. Furthermor e, incrementalism in the budgetary decision making process inhibits the efficiency and effectiveness of budget allocation. In other words, if there exists the change in budget practice, it can increase the efficiency and efficiency of budget allocation. 190

PAGE 203

Also, other budget factors have impact on perceptional budget allocation. In this case, the community that economically depends on nonagricultural activities —tourism and industry —can drive the increase in efficient and effective budget allocation. Th ese causal links also hold for budget allocation in public works services. However , for the educational services, if the education department heads apply the “we’re experts” strategy of information use and find their programs motivating or challenging, the n it can lead to the increase in efficiency and effectiveness of budget allocation. Conversely, if they apply the “problem framing” strategy, then there exists the decrease in efficiency and effectiveness of budget allocation. Given the factors influencing the municipal budget allocation, this study arrange s them into four discussing points —incrementalism, the contextual characteristics, administrative capacity, and the use of information. The following section elaborates them in detail. Discussion and Findings Budgetary Incrementalism and Municipal Finance T he re are two source s of budgetary incrementalism in the municip al budget process, which are the gradual change via considering and adhering to the previous year spending levels and the change in budgetary decision making process . Previous Year Budget Allocation The spending levels of 2019 are firmly determined by 2018 levels for both educational and public works service programs. An education department head from the north explains his budget preparation, stating that “I first look back at the budget the department received in the last three or five years. Then I adjust it a little bit to make it sound appropriate considering some what’s going on the community.” Another education department head from the north adds that “a 191

PAGE 204

sharp rise is pretty rare. I prepare my budget drafts by picturing our last year spending and how the com munity reacted to it.” Also, a public works department head from the south also underscores the prevalence of the previous year budget allocation, expressing that “I review the municipality’s both revenues and spending ceiling from the last year. Then I try to choose the projects and dra ft my budget accordingly.” Incrementalism in the Budget Process The budgetary decision making process of the municipalities in Thailand demonstrates some changes. These changes are seen though such budget reform measures as the application of information technology (to install and to apply the e procurement system), the promotion of participatory budgeting, and the coverage of local development plan (from three to four years). Since the DOLA aims at modifying local budgeting processes through such measures, the department heads report the influence of such endeavors. An education department head from the Northern municipality gives an example of change in the local budget practice. He point s out that “the new budget practice requires t he higher level of citizen engagement. The change of any local sub programs must be reviewed by both the municipal budget committee and the civic forums. Previously, the civic forum was unnecessary. ” A public works depart ment head from the east adds a few examples highlighting the changes in the local budget process, stating that “we shift our local plan coverage from three to four years. Additionally, this four year plan is somehow rigid because a municipality cannot move an operational plan in a certain y ear to another. The new budget practice also demands the electronic standard price system to be installed and implemented. ” 192

PAGE 205

Spending Levels VS Spending Efficiency and Effectiveness Undeniably, there is an underlying tendency toward budgetary incrementalism in municipal finance so that budgetary incrementalism is still the dominant way of understanding the budgetary decision making in the government ( True, 2000; Wildavsky, 1964) . The budgetary incrementalism in this study is positively associated with both of the municipalities’ spending programs (education and civil engineering) . H owever, due to the low level of incrementalism in the budget process or the changes in budget practices, t he greatest gain from an adoption of new budget practice is realized in the management of frontline departments and not in regards to their spending a mounts (Willoughby & Melkers, 2000) . How Management Matters in Local Budgeting Despite the prominence of budget incrementalism, it does not prevail at all time s and a management system can disrupt the municipal finance’s underlying tendency towards budgetary incrementalism. What public administrators and management systems do inside a certain public orga nization and how they do it has an impact on how well such organization perform (Ingraham et al., 2003) . Even though this study does not investigate the character istic s of the municipality’s management in Thailand (i.e., financial management, human resource management, information technology management, and capital mana gement), to some extent, it still shows how, when, and under conditions management matters through its results that point out to the key levers that drive the municipality’s perf ormance: leadership influence and emphasis, integration and alignment, and managing for results. Leadership Influence and Emphasis This study concentrates on the local frontline public administrators (the education and public works department heads), not the formal political leaders (mayors, chief executives, or 193

PAGE 206

leadership team members) and administrative leaders (city managers or municipal clerks). Even though the latter two groups constitute the official budget committee members who cast their vote s for or against the budget proposals, the frontline public administrators’ technical know how as well as specialized knowledge about how to solve difficult problems, representing the monopolistic power of bureaucracy, can enhance their power potential, pla y off opposing economic interests, and aggrandize such power over other institutions and stakeholders groups (including citizens and politicians) (Hummel & Stivers, 2010; Riggs, 1998; Weber, 1968) . Empirically, the results of this study also confirm that technical skills, professional know ledge, and specialization among the department heads has an impact on the municipal budget allocation. The political/career relationships as well as tension between the political executives and the career management cadre are also a significant determinan t of municipal budget allocation (Ingraham, Thompson, & Eisenberg, 1995) . Specifically, in t his study, the shared d irection or emphasis on the same local development policy and priority by these two groups is favorable for an organization’s performance. Meanwhile, the tension or conflict is adversely associated with the efficiency and effectiveness of budget allocation . Some department heads discuss this policy congruence by highlighting how their chief executives’ policy platforms plays a role in budget allocation. They assert that they must consider and translate their chief executives’ visions and goals into substant ive actions (Ingraham et al., 2003) . A department head from the east states that “the mayor’s vision specifies how the community should grow and develop. It is no use proposing the opposite. Otherwise, the budget we should get allocated will be redirected to other departments instead.” A public works department head from the south points out that “The mayor is elected by the community. As career officials, we should not violate or overrule the 194

PAGE 207

local development policy and priority initiated by the mayor who acts on the community’s behalf.” Another public works department head from the south gives some details regarding the budget allocation determined by the mayor’s policy platform, stating that “The mayor’s policy platform is a key determinants of budget allocation. His policy platform, the municipal development directions and other urgent policy agendas initiated by his team receive about 40 percent of the total spending. The remaining 40 percent goes to the current budget.” Integration and Alignment E xcellent man agerial performance is fundamentally dependent on the extent to which its management systems are integrated and aligned. In the other words, such management systems must be orchestrated as a unified, cohesive whole with shared values, common goals, and ali gned objectives primarily accomplished through three key activities —the exercise of leadership, the resource allocation, and the use of information (Ingraham et al., 2003) . This study provides the evidence highlighting the last activity which has an impact on the municipal finance. On one hand, this study focuse s on the level of attention that local frontline public administrators give information and their willingness to share such information during the budgeting process in order to facilitate (or thwart) the municipal committee’s decisions on budget allocation (Ingraham e t al., 2003; Melkers & Willoughby, 2005) . On the other hand, though it is almost impossible to predict when and where an input of certain information is likely to have an impact on a particular budgetary decision, the strategy of information use dictate s whether such information confirm s a policy position already held or the intuition of budget actors (Rich, 1991; Weiss, 1977) . Accordingly, the use of information in this study is considered 195

PAGE 208

through a couple of perspectives —the level of information utilization and the strategy of information use. The Level of information Utilization. As the quantitative results suggest, if the department heads comprehe nsively utilize the information for drafting and defending their budget proposals, it can increase the efficiency and effectiveness of the budget allocation in both educational and public works service programs. This empirically confirms the causal linkage between the informationbased decision making process and the budget allocation. In the other words, the budget practice promoting the information utilization can instrumentally drive budget process toward better decisions, eventually resulting in manager ial improvement and program effectiveness (Caiden, 2010; Schick, 1966; Weiss, 1977) . Strategy of Information Use and Performance Information. H ow department heads present and discuss their budget proposals to the municipal budget committee matters . The results suggest that the strategy of “we’re expert” also leads the local budgeting to the efficient and effective budget allocation, especially for educational services. However, elaborating certain problems the community experiences through the strategy of “problem framing” tends to decrease the efficiency and effectiveness of budget allocation. “We’re Expert”. Focusing on the elaboration of problem condition or severity can harm the efficiency and effectiveness. The education department heads, thus, concentrate on the results or educational accomplishment of education services instead. Particularly, they present and discuss their budget proposal by comparing their proposed programs’ results in terms of technical feasibility, affordability, or value acceptability with that of professional or expert standards. For example, an education department head from the east explains that “we present to the executives how each program solves the community’s problems, how well it can respond and 196

PAGE 209

help the affected residents, and what it can offers to its target groups. Value of money is also the important issue. The program whi ch can benefit the target group the most will get approved.” An education department head from the north also states that during the budget discussion, “if asked, I answer with professional explanation. I provide the empirical data of how my department’s programs can be a policy success. Policy results such as academic accomplishment among school students are discussed and presented because they are concrete and understandable.” An other education department head in the north also applies this approach, stat ing that “Benefits to the students according to educational principles are the best way to present my budget proposal. For example, I highlight how the students can communicate with the foreigners, how they have greater mathematic and scientific skills, or how the toddlers in the municipal nursery show their normal physical growth and mental development.” “Problem Framing”. One education department head from the central region recounts how she present s her budget proposal to the budget committee, stating t hat “we must gather the information directly from our communities and our target groups. All information used has to be clear and from all perspectives. Especially, we elaborate to the budget committee how the target groups struggle to fight those problems by themselves. So that it is urgent for the municipality to do something about it.” Performance Budgeting for Educational Services. Based on the positive impact of the strategy of information use on the budget allocation, the education department heads highlight information such as policy results, service quality, value of money, and professional standards and techniques during the budget process. This adoption of performance budgeting for education services shows the promising progress toward the resource allocation that expands, modifies, or terminates public programs according to their programs’ performances. 197

PAGE 210

Intergovernmental Relations: The Indomitable Centralization Even though the municipalities mostly plan, prepare, approve, and execute their own budgets, the influence of the central government and its centralization power —the major characteristic of l ocal government in Thailand—prevails in municipal finance (Nagai et al., 2008) . Particularly for the public works services, in order to have a positive influence the budget allocation, the department heads present their budget proposals by relating their programs to the mandates directly or indirectly enforced by the state statutes and the central government. In the other words, the central government’s policy and DOLA’s policy directions are key determinants of municipal budget allocation. Some public works departments cite those policy directions while presenting and defending the ir budget proposals to the municipal budget committee. This is because the municipalities should comply with the higher level governments’ policy (both the central government and the provincial administrative organization). A public works department head f rom the north points out that “If there exists a policy enforced by the central government, we have to design the programs in order to follow that policy. The budget committee will have to allocate the budget to fund such programs. A public works department head from the east expresses that “There is request from the Provincial Administrative Organization that we construct the local roads to connect with their main roads. Also, we consider the central government’s policy and priority. This year they focus on quality of life.” In the same vein, another public works department heads from the east also mentions that “the municipality’s spending level on public works does not decrease. Look at the local road construction. We usually propose the plan to construct local roads into the community areas which can serve the economic development plan initiated by either the central government or the provincial administrative organization.” 198

PAGE 211

Theoretical Contributions The theoretical framework in Chapter 2 concentrates on a few theories, increment alism and information economics, which explain municipal budget allocation. This section describes how this study’s findings affirm or give additional views on such theories as follows. Budgetary Incrementalism. Budgetary increme ntalism still prevails in municipal finance and its budgeting. The budget allocations in education and civil engineering programs are based mainly on their previous year spending, often making the change piecemeal because it is heavily historical , looking backwards more than forward and indifferent about objectives (Wildavsky, 1978) . Disjointed Incrementalism. The local budgeting process and its allocation in Thailand is theoretically located somewhere between the rational co mprehensive and the incremental decision making process ( Lindblom, 1959) . Due to the change of the local budget practice toward the performance budgeting, this study’s findings confirm that there exists the comprehensive information utilizati on during the budgeting process. In particular, department heads use empirical aspects of problems, possible consequences, and analysis of policy results , especially for the educational services ( Lindblom, 1979) . Contributions to the Current Literature This study aims to provide the details on the causal linkage between the use of information and the budget allocation. The findings give additional empirical perspectives to the current studies as follows. Local frontline administrators as key policy/budget actors. Most of the current literature pays att ention to the use of information among key budget actors such as the chief executives or the elected officials (Buylen & Christiaens, 2016; Ho, 2006; Nielsen & Baekgaard, 199

PAGE 212

2015; Ter Bogt, 2004) , the financial and budget analysts (Melkers & Willoughby, 2005) , and the public managers (Moynihan & Pandey, 2010; Taylor, 2011; Zhang, Van de Walle, & Zhuo, 2016) . However, the studies focusing on department heads are still limited. This study affirms that the department heads’ information use has a positive impact on the performance of the local budgeting (i.e., efficient and e ffective budget allocation). Consideration of the public organization’s performance should include the factors reflecting the characteristics of frontline public administrators into its equation of management performance linkage (O’Toole & Meier, 2014) . This is because department heads are not passive, neutral actors in the public management’s black box but policy actors who play a substantial and legitimate role in shaping public policy (Ingraham et al., 2003) . Additionally, this study show s that these frontline public administrators at the local level possess leadership rooted in their individual managerial and technical skills and local knowledge and pl ay such important roles as decision makers, strategists, and analysts during the budget process (Lynn, 1996) . Operationalization of Information Use. Beside the level of information utilization more commonly investigated in the public policy literature (Knott & Wildavsky, 1980; Taylor, 2011) , this study also adds the strategies of information use ( Mikesell, 2007; Zahariadis, 2014) based on the findings from the pilot study to explore whether and how they impact the budget allocation. The results show that the “we’re experts” and “it’s mandate” strategies have impact on spending efficiency and eff ectiveness. This finding confirm s the significant roles of information presentation in local budgeting and allocation highlights the influence of bounded rationality and behavioral economics such as framing effects (Kahneman & Tversky, 1979; Tversky & Kahneman, 1986) and selection biases (Zahariadis, 2014) on the local decisionmaking process. 200

PAGE 213

Operationalization of Budget Allocation. The current studies on the relationship between the use of information and budget decision consider the latter from one dimension— either the actual spending levels (Ho, 2011; Zaltsman, 2009) or the perceptions on efficiency and effectiveness of budget allocation (Goodman & Clynch, 2004; Melkers & Willoughby, 2001; Melkers & Willoughby, 2005) . However, this study investigates and affirms the impact of information use on both actual and perceptional budget allocation among municipalities in Thailand. In particular, the actual budget allocation involves spending levels and the perceptional budget allocation are spending efficiency and effectiveness perceived by department heads. Additionally, this study investigates two main functional programs of municipalities — educational and public works services —and compares and contrasts the results from these two spending programs. Management Matters for Budget Decisions in Developing Countries. Management certainly does matters but the questions which should be asked are “ how and under conditions management matter?” (Ingraham et al., 2003) . This study offers an empirical answer to such question, arguing that under the cont ext of a developing country, with at tempt s to adopt the performance budgeting in the face of strong decentralization, the frontline public administrators’ comprehensive use of information especially through comparing their proposed programs’ results with technical feasibility, affordability, or value acceptability actually has a positive impact on the municipalities’ budget allocation for educational public services. Practical Contributions The findings of this study provide the implications on how to improve the local budgeting process which can bring about more efficiency and effectiveness of the budget allocation among the local governments. 201

PAGE 214

First, the frontline public administrators (or the department heads in this study) are possibly the group of budget actors who can indirectly improve the budget decisions. With their comprehensive use of information and their strategies to present and discuss the budget proposals, they are able to f acilitate the information based or rational decisionmaking process. For example, with the analysis of policy results and value of money as well as the standard of classroom activities, the education department heads can make the budget committee understan d and make their decision based on such performance information. Second, the different functional programs may require different approaches to enhance their spending efficiency and effectiveness. For example, the performance management and budgeting is m ore conducive to cost saving, eliminating duplicative or ineffective programs, and improving service quality for the educational services. Meanwhile, the public works service programs may not have to focus on performance information as much as the educatio nal service programs. Third , the collaboration between the local government’s chief executives and the frontline public administrators is the key to improving budgetary decisions. There should be more preliminary discussions or budgeting workshops in ord er for them to have shared understanding on the policy and priority of their local development plans. This policy harmonization can enhance the spending for the local government’s educational or public works services. 202

PAGE 215

CHAPTER IX CONCLUSION To summarize the study, this chapter revisits the research questions, provide s the theoretical relevancy, and present the policy implications in order to portray an overview picture of the study’s results and findings. Additionally, this chapter offers a future research direction based on the study’s limitations or the threats to legitimation of the mixed methods research design. Research Question s Overall, this study asks: How do policymakers with different administrative capacities utilize information during a budget process in order to have an impact on budget decisions? In particular, it first asks: How do policymakers with different administrative capacities utilize relevant and reliable information during a budget process? The results affirm that administrative capacity —skills in participatory budgeting, communication, strategic planning, and financial management and accounting—can promote the information utilization during local budgeting process. However, budget rationality in terms of policymaki ng, advocacy, and techno rational orientations, which feature a predetermined po licy direction, and the organization culture especially coercive isomorphism limit information utilization among de partment heads. Then this study asks: How does the informati on utilization influence budget decisions? The results suggest that despites the existence of budgetary incrementalism, management matters in municipal finance. In other words, the change in budget practice, the level of information utilization, the strate gy of information use, and professional judgement have impact on municipal budget allocation. 203

PAGE 216

Theoretical Relevancy Budgetary incrementalism exist s in Thailand’s municipal finance. Nonetheless, at times the local budgeting process reflects a rational comprehensive or informationbased decision making process. According to the theories of information economics and budget rationality model, the releva nt information (both performance and other information) as well as individual characteristics (i.e., frontline public administrators’ capacity to budget and budget norms toward public spending) are key to moving the local budgeting toward a rational decisi onmaking process. In order words, the deliberation of performance measures such as policy results and value of money together with local knowledge such as community programs and requests among department heads drives the local budgeting process to be an i nformationbased decision making. Policy Implications It is possible to make a local budgeting process less incremental and more rational especially toward the new budget practice promoted by the performance reform. Together with the influence of the new ly adopted budget practice and the central government on municipal budget allocations, the local policymakers can enhance their budgeting process es via promoting the information use especially by the department heads. First, they should encourage the depa rtment heads to gather and consider relevant information including not only performance measures (i.e., value of money, service quality, and policy results) but also other necessary information such as local knowledge (i.e., community problems and requests ), professional standards and techniques, and budget constraints. Second, local policymakers can design competency building activities which aim at enhancing such knowledge and skills as participatory budgeting, project management, and accounting and fina ncial management. 204

PAGE 217

Third, together with the capacity building programs, local policymakers should motivate the department heads to consider and apply more of the local fiscal discipline and financial and account techniques to their budget preparation and discussion in order to bala nce their propensity to draft the budget proposal based mainly on the local development plans. Fourth, in order to adopt the new budget practice (i.e., the performance based management and budgeting) according to the Financial and Fiscal Discipline Act of 2018 , the central government and local policymakers should cons ider cooperat ing with or invit ing the professional networks to participate in guiding the new budget practice into the local budgeting process. Professional guidance will increase the legitimacy of DOLA’s reforms and may encourage adoption by local staff members. Additionally, the local executives should promote the activities that can strengthen understanding among informal groups of department heads . This is because this group culture can indirectly promote and institutionalize the implementation of the performancebased management and budgeting. As a result, these organization cultures of normative isomorphism and group culture can lead to the higher level of information utilization in a local budgeting process. Then , this comprehensive information utiliz ation can eventually bring about the more efficient and effective use of local resource—which is one of the main objectives of the decentralization policy. Threats to Legitimation and Future Research Even though this research attempts to carefully select cases according to the guiding theories and to triangulate the results from the survey data and interviewing data, there are some threats to legitimation for this research design. First, both quantitative and qualitative approaches in this study collect c ross section data so that the study does not account for the change of policymakers’ informational behaviors. There is a tendency t hat the research findings are likely 205

PAGE 218

to vary across time. In other words, the temporal validity is possibly compromised because such findings cannot be generalized to other time periods ( Onwuegbuzie, 2003) . The future research should take into account informational behaviors and budget decis ions of the budget actors over time or collect panel data in order to avoid the risk of temporal invalidity. Second, even though this study attempts to investigate the strategy of information use, it still does not analyze in detail the information characteristics which includes, for instance, measurement process and disclosure of information. Particularly, some information is soft (qualitative, context dependent, and able to be directed to one way or another) while some is hard (quantitative, easy to stor e and transmit in impersonal ways) , but both have implications on decision makings (Petersen, 2004) . Thi s is because, in cognitive and social psychology, decisions makings and attitude formation are affected by numbers (Olsen, 2015) as well as by narratives (McBeth, Jones, & Shanahan, 2014) . These informational characteristics, which this study does not take into considerat ion , are omitted variables that may affect the relationship between the information types, its utilization, and its influence on budget decisions, thereby causing spurious causal association in this study (Gerring, 2007) . On this ground, it is possible that the findings concerning causal relationships suffer from internal validity. Thus, future research should take the nature of information (soft or hard) into its consideration. Third, the concepts of information characteristics as well as information utilizations are measured by the constructs based on personal perception, thereby not necessarily reflecting the actuality. This also causes a threat on construct validity, a type of internal validity (Van de Ven, 2007) . This study proposes future research to measure the information utilization from, for example, scrutinizing policymakers’ speech for the presence of fiscal information during budget debates (Buylen & Christiaens, 2016) . 206

PAGE 219

Fo urth, some r egression analyses in Chapter 5 has risk to internal validity. The model fit value is relatively low in for the regression analyses estimating the info rmation utilization. Meanwhile, the regression models estimating the actual budget allocations are affected by the problems of heteroscedasticity and multicollinearity. Future research should take this problems into account and, for example, includes other variables into the analyses that fit the developing context. Fifth , the study’s unit of observation both for the information utilization and the perceptional budget allocation is a department head (or a frontline public administrator). However, there are still other main budget actors such as mayors, chief executives, legislators, and local clerks who also play a role in the budget process. The future study should include the information use and perceptions on budget allocation from these budget actors in order to triangulate data from different sources or from multiple groups of local budget actors, thereby eventually providing the study with the more comprehensive and accurate results (Creswell, 2013; Leech & Onwuegbuzie, 2007) . Sixth, the current studies on information utilization identify measures of administrative capacity , such as administrative tenure (Goodman & Clynch, 2004) and academic discipline (Lee, 1991) , as an important determinants . This study measures administrative capacity in terms of ability to prepare a budget proposal (i.e., skills in financial manage ment and accounting, strategic planning, participatory budgeting, communication, networking, and project management) and finds that it significantly encourages the information utilization in Thailand. In order to promote the external validity of this addit ional operationalization ( Onwuegbuzie et al, 2007) , the future study in other settings should use the ability to prepare a budget proposal as a construct of administrative capacity and see whether and how it affects information utilization . 207

PAGE 220

Seven, this study focuses on the instrumental use of information—to use information in order to reach the best decision and to promote the rational decision making process. However, it finds that the symbolic use also matters. In order to promote the internal validity of this component of information utilization, the future studies on how information utilization has an impact on budget decisions should systematically take into consideration the concept of symbolic use of information so as to see whether tha t this type of information utilization is also a determinant of budget decisions. In conclusion, despite the existence of budgetary incrementalism in municipal finance (Davis et al., 1966; Wildavsky, 1964) , administrative capacity and information use among frontline public administrators can make budgeting process es more rational and eventually enhance budget allocation to be more efficient and effective. This affirms that management matters in the developing countries’ budgetary decision making (Ingraham et al., 2003) . To strengthen this argument, future research should investigate other modes of information use (i.e., speeches or policy statements during a budget debate) and take account the roles of all groups of local budget ac tors in order to explain of budget allocations over years. 208

PAGE 221

REFERENCES Akerlof, G. A. (1970). The market for “lemons”: Quality uncertainty and the market mechanism. Quarterly Journal of Economics , 84(3), 488–500. https://doi.org/10.2307/1879431 Alchian, A. A., & Demsetz, H. (1972). Production costs, information costs, and economic organization. American Economic Review , 62(5), 777–795. https://doi.org/10.1109/EMR.1975.4306431 Alijarde, M. I. B. (1997). The usefulness of financial reporting in Spanish local governments. Financial Accountability and Management , 13(1), 17–34. Alt, J. E., Lassen, D. D., & Skilling, D. (2002). Fiscal transparency, gubernatorial approval, and the Scale of Government: Evidence from the States. State Politics & Policy Quarterly, 2(3), 230–250. Amara, N., Ouimet, M., & Landry, R. (2004). New evidence on instrumental, conceptual, and symbolic utilization of university research in government agencies. Sci ence Communication, 26(1), 75–106. Ammons, D. N., & Roenigk, D. J. (2015). Performance Management in Local Government: Is Practice Influenced by Doctrine? Public Performance & Management Review , 38(3), 514– 541. Andrews, M., & Hill, H. (2003). The impact of traditional budgeting systems on the effectiveness of performancebased budgeting: A different viewpoint on recent findings. International Journal of Public Administration, 26 (2), 135–155. Appleby, P. H. (1949). Big democracy . New York, NY: Alfred A. Knop f. Arguello, J., Callan, J., & Shulman, S. (2008). Recognizing citations in public comments. Journal of Information Technology and Politics , 5(1), 49–71. Argyris, C. (1990). The dillemma of implementing contols: The case of managerial accounting. Accounting, Organizations and Society , 15(6). Arrow, K. J. (1962). Economic welfare and the allocation of resources for invention. In R. Nelson (Ed.), The rate and direction of inventive activity: Economic and social factors (pp. 609–626). Boston, MA: Princeton University Press. Ashworth, R., Boyne, G., & Delbridge, R. (2009). Escape from the iron cage? Organizational change and isomorphic pressures in the public sector. Journal of Public Administration Research and Theory, 19(1), 165–187. https://doi.org/10.1093/jopart/mum038 Asian Development Bank. (2018). Thailand: Economy. Retrieved February 24, 2017, from https://www.adb.org/countries/thailand/economy Askim, J., Johnsen, A., & Christophersen, K. A. (2008). Factors behind organizational learning from benchmarking: Experiences from Norwegian municipal benchmarking networks. Journal of Public Administration Research and Theory , 18(2), 297–320. 209

PAGE 222

Barriball, K. L., & While, A. (1994). Collecting data using a semi structured inte rview: A discussion paper. Journal of Advanced Nursing, 19(2), 328–335. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/8188965 Bertelli, A. M. (2007). Law and Public Administration. In E. Ferlie, L. E. Lynn, & C. Pollitt (Eds.), The Oxford Handbook of P ublic Management (pp. 134–155). New York, NY: Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199226443.003.0007 Blazenko, G. W. (1987). Managerial preference, asymmetric information, and financial structure. The Journal of Finance , 42(4), 839– 862. https://doi.org/10.1111/j.15406261.1987.tb03915.x Boggs, J. P. (1992). Implicit models of social knowledge use. Science Communication, 14(1), 29–62. Bourdeaux, C., & Chikoto, G. (2008). Legislative influences on performance managment reform. American Journal of Evaluation, 68 (2), 242–243. Bourdeaux, C., Chikoto, G., Bourdeaux, C., & Chi, G. (2008). Legislative influences on performance management reform. Public Administration Review , 68(2), 253–265. Brace, P. (1991). The changing context of state political economy. The Journal of Politics , 53 (2), 297–317. Breton, A. (2009). Information shrounding and the governmental supply of goods and services: An economic perspective. In L. M. Imbeau (Ed.), Do they walk like they talk?: Speech and action in policy processes (pp. 31–51). Berlin, Germany: Springer Science & Business Media. https://doi.org/10.1007/9780 387896724 Buchanan, J. M., & Tullock, G. (1962). The calculus of consent: Logical foundation and constitutional democracy . Indianapolis, IN: Liberty Fund, Inc. https://doi.org/10.1038/024026a0 Burns, T., & Stalker, G. M. (1961). The management of innovation. Oxford, UK: Oxford University Press. Buylen, B., & Christiaens, J. (2016). Talking numbers? Analyzing the presence of financial information in councilors’ speech during the budget debate in Flemish municipal councils. International Public Management Journal , 19(4), 453–475. https://doi.org/10.1080/10967494.2015.1064502 Caiden, N. (2010). Challenges confronting contemporary public budgeting: Retrospectives/prospectives from allen schick. Public Administration Review , 70(2), 203– 210. https://doi.org/10.1111/j.15406210.2010.02148.x Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitraitmultimethod matrix. Psychological Bulletin , 56(2), 81–105. https://doi.org/10.1037/h0046016 Campbell, T. S., & Kracaw, W. A. (1980). Information production, market signalling, a nd the theory of financial intermediation. Journal of Finance , 35(4), 863–882. https://doi.org/10.2307/2329297 210

PAGE 223

Caplan, N. (1979). The TwoCommunities Theory and Knowledge Utilization. American Behavioral Scientist , 22 (3), 459–470. https://doi.org/10.1177/000276427902200308 Cash, D. W., Borck, J. C., & Patt, A. G. (2006). Countering the loading dock approach to comparative analysis of El Nio/Southern Oscillation (ENSO) forecasting systems. Science, Technology, & Human Values , 31(4), 465 –494. https://doi.org /10.1177/0162243906287547 Christensen, R. K., & Gazley, B. (2008). Capacity for Public Administration: An analysis of meaning and measurement. Public Administration and Development , 28(2), 265–279. Cobb, R. W., & Elder, C. D. (1972). The dynamics of agenda setting. In J. Shafrizt, K. Layne, & C. Borick (Eds.), Classics of Public Policy (pp. 128–136). New York, NY: Pearson Longman. Coe, C. K. (2003). Usually engaging local budget analysts during budget execution. State & Local Government Review , 35(1), 48–56. Collins, K. M. T., Onwuegbuzie, A. J., & Sutton, I. L. (2006). A model incorporating the rationale and purpose for conducting mixed methods research in special education and beyond. Learning Disabilities: A Contemporary Journal , 4(1), 67–100. Corburn, J. (2003). Bringing local knowledge into environmental decision making: Improving rrban planning for communities at risk. Journal of Planning Education and Research , 22, 420–433. https://doi.org/10.1177/0739456X03253694 Corburn, J. (2007). Community knowledge in environmental health science: Co producing policy expertise. Environmental Science and Policy , 10(2), 150–161. Creswell, J. W. (2013). Qualitative, quantitative, and mixed methods approaches. Research Design , 1–26. https://doi.org/10.2307/3152153 Dahler Larsen, P. (2007). Evaluation and Public Management. In Evan Ferlie, L. E. Lynn, & C. Pollitt (Eds.), The Oxford Handbook of Public Management (pp. 615–639). New York, NY: Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199226443.003.0027 David, I. T. (1998, July 1). The true financial manager. Government Executive. Davis, O. A., Dempster, M. A. H., & Wildavsky, A. (1966). A theory of the budgetary process. The American Political Science Review , 60(3), 529–547. De Renzio, P., & Masud, H . (2011). Measuring and Promoting Budget Transparency: The Open Budget Index as a Research and Advocacy Tool. Governance , 24(3), 607 –616. https://doi.org/10.1111/j.14680491.2011.01539.x Denhardt, R. B. (2002). Symposium on Citizenship, Governance, and Adm inistrative Capacity: An introduction. Public Organization Review , 2(1), 3–4. Dimaggio, P. J., & Powell, W. W. (1983). The iron cage revisited: Isomophism and collective rationality in organization fields. American Sociological Review , 48(2), 147–160. Dimi trijevska Markoski, T., & French, P. E. (2019). Determinants of public administrators’ use of performance information: Evidence from local governments in Florida. Public Administration Review , 00, 1–11. https://doi.org/10.1111/puar.13036 211

PAGE 224

Dingwall, R., & St rangleman, T. (2007). Organizational cultures in the public services. In Evan Ferlie, L. E. Lynn Jr., & C. Pollitt (Eds.), The Oxford Handbook of Public Management (pp. 468–490). New York, NY: Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199226443.003.0021 Downs, A. (1967). Inside Bureaucracy. Long Grove, IL: Waveland Press Inc. Downs, Anthony. (1957). An economic theory of democracy . New York, NY: Harper & Row Publishers. Dull, M. (2009). Results model reform leadership: Questions of credi ble commitment. Journal of Public Administration Research and Theory , 19(2), 255–284. https://doi.org/10.1093/jopart/mum043 Dunn, W. N. (1980). The twocommunities metaphor and models of knowledge use: An exploratory case survey. Science Communication, 1(4 ), 515 –536. Eisenhardt, M. (1989). Agency theory: An assessment and review. The Academy of Management Review , 14(1), 57–74. https://doi.org/10.5465/AMR.1989.4279003 Ellsberg, D. (1961). Risk, ambiguity, and the Savage axiom. Quarterly Journal of Economics , 75(4), 643– 669. Faibisoff, S., & Ely, D. P. (1974). Information and information needs . Washington, D.C.: Office of Education (DHEW). Failing, L., Gregory, R., & Harstone, M. (2007). Integrating science and local knowledge in environmental risk management : A decision focused approach. Econogical Economics , (64), 47–60. https://doi.org/10.1016/j.ecolecon.2007.03.010 Farazmand, A. (2009). Building administrative capacity for the age of rapic globalization: A modest prescription for the twenty first century. Public Administration Review , 69(6), 1007–1020. Feldman, M. S. (1989). Order without design: Information production and policy making. Standford: CA: Stanford University Press. Feldman, M. S., & March, J. G. (1981). Information in organizations as signal a nd symbol. Administrative Science Quarterly, 26 (2), 171–186. Fernandez, S., & Rainey, H. G. (2006). Managing Successful Organizational Change in the Public Sector. Public Adminisration Review , 66(2), 168–176. https://doi.org/10.1111/j.15406210.2006.00570.x Fischer, F. (2000). Citizens, experts, and the environment . London, England: Duke University Press. Frederickson, H. G., Smith, K. B., Larimer, C. W., & Licari, M. J. (2012). The Public Administration theory primer (2nd ed.). Boulder, CO: Westview Press. Friedman, M., & Savage, L. J. (1948). The utility analysis of choice involving risk. Journal of Political Economy , 56(4), 279–304. 212

PAGE 225

George, A. L., & Bennet, A. (2004). Case studies and theory development in the social sciences. Cambridge, MA: The MIT Press. https://doi.org/10.1017/S1537592707070491 George, A. L., & Bennett, A. (2005). Case studies and theory development in social sciences . Cambridge, MA: MIT Press. Georgescu Rogen, N. (1954). Choice, expectations, and measurability. The Quarterly Journal of Economics , 68(4), 503– 534. Gerring, J. (2007). Case study research: Principles and practices. Cambridge, MA: Cambridge University Press. Gerring, J. (2012). Social Science Methodology: A Unified Framework . New York, NY: Cambridge University Press. Gerring, J. (2017). Case Study Research: Principles and practices (2nd ed.). Boston, MA: Cambridge University Press. Gilmour, J. B., & Lewis, D. E. (2006a). Assessing performance budgeting at OMB: The influence of politics, performance, and program size. J ournal of Public Administration Research and Theory, 16(2), 169–186. https://doi.org/10.1093/jopart/muj002 Gilmour, J. B., & Lewis, D. E. (2006b). Does performance budgeting work?: An examination of the Office of Management and Budget’s PART scores. Public Administration Review , 66(5), 742– 752. Goodman, D., & Clynch, E. J. (2004). Budgetary decision making by executive and legislative budget analysts: The impact of political cues and analytical information. Public Budgeting & Finance , 24(3), 20–37. Goodnow, F. J. (1900). Politics and Administration: A Study in Government . New York, NY: Rusell & Rusell. Hall, P. A., & Taylor, R. C. R. (1996). Political Science and the three new institutionalisms. Political Studies , 44, 936 –957. https://doi.org/Article Hannan, M. T., & Freeman, J. (1977). The population ecology of organizations. American Journal of Sociology , 82 (5), 929. Heinkel, R. (1982). A theory of capital structure relevance under imperfect information. The Journal of Finance , 37(5), 1141–1150. Henri, J. F . (2006). Organizational culture and performance measurement systems. Accounting, Organizations and Society , 31(1), 77 –103. https://doi.org/10.1016/j.aos.2004.10.003 Hildreth, W. B., Yeager, S. J., & Miller, G. J. (2011). Finance managers’ propensity to sa ve. Journal of Public Budgeting, Accounting & Financial Management , 24(2), 1–35. Hindmoor, A. (2006). Rational choice . New York, NY: Palgrave Macmillan. Hirshleifer, J. (1973). Where are we in the theory of information? American Economic Association, 63(2) , 31–39. 213

PAGE 226

Ho, A. T. K. (2006). Accounting for the value of performance measurement from the perspective of midwestern mayors. Journal of Public Administration Research and Theory , 16(2), 217– 237. https://doi.org/10.1093/jopart/mui046 Ho, A. T. K. (2011). PBB in American Local Governments: It’s More than a Management Tool. Public Administration Review , 71(3), 391–401. Hofstede, G. (1978). The poverty of management control philosophy. Academy of Management Review , 3(3), 450–461. https://doi.org/10.5465/amr.1978.4305727 Hlmstrom, B. (1979). Moral hazard and observability. Bell Journal of Economics , 10 (1), 74–91. https://doi.org/10.2307/3003320 Hood, C. (1991). A public management for all seasons? Public Administration, 69(Spring), 3–19. Hood, C. (2007). Public Management: The word, the movement, the science. In Ewan Ferlie, L. E. Lynn, & C. Pollitt (Eds.), The Oxford Handbook of Public Management (pp. 7–26). New York, NY: Oxford University Press. Hou, Y. (2003). What stabilizes sta te general fund experience in downturn years --Budget stabilization fund or genderal fund unreserved undesignated balance? Public Budgeting and Finance , 23(3), 64–91. https://doi.org/10.1111/15405850.2303004 Hou, Y., Lunsford, R. S., Sides, K. C., & Jones, K. A. (2011). State performance based budgeting in boom and bust years: An analytical framework and survey of the states. Public Administration Review , 71(3), 370– 388. https://doi.org/10.1111/j.15406210.2011.02357.x Hummel, R. P., & Stivers, C. (2010). P ostmodernism, bureaucracy, and democracy. In R. F. Durant (Ed.), The Oxford Handbook of American Bureacracy (pp. 1–16). New York, NY: Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199238958.003.0014 Ingraham, P. W. (2009). Striving for bala nce. In C. Ferlie, Ewan: Lynn Jr., Laurence E.: Pollitt (Ed.), The Oxford Handbook of Public Management (pp. 1–11). https://doi.org/10.1093/oxfordhb/9780199226443.003.0023 Ingraham, P. W., Joyce, P. G., & Donahue, A. K. (2003). Government Performance: Why Managment Matters . Baltimore, MD: John Hopkins University Press. Ingraham, P. W., Thompson, J. R., & Eisenberg, E. F. (1995). Political managment strategies and polical/career relationships: Where are we now in the federal government? Public Administration Review , 55(3), 263– 272. Jakobsen, M. L. F., & Mortensen, P. B. (2016). Rules and the Doctrine of Performance Management. Public Administration Review , 76(2), 302–312. Jeong, M.G. (2007). Local political structure, administrative capacity, and revenue pol icy choice. State & Local Government Review , 39(2), 84–95. Johnson, R. B., & Onweugbuzie, A. J. (2004). Mixed methods research: A research paradigm whose time has come. Educational Researcher , Vol. 33 No, 14–26. 214

PAGE 227

Jones, M. D., Peterson, H. L., Pierce, J. J ., Herweg, N., Bernal, A., Raney, H. L., & Zahariadis, N. (2015). A river runs through it: A multiple streams meta review. Policy Studies Journal , 44(1), 1–24. Joyce, P. G. (2008). Does more (or even better) information lead to better budgeting?: A new per spective. Journal of Policy Analysis & Management , 27(4), 945–975. Joyce, P. G. (2011). The Obama Administration and PBB: Building on the legacy of federal performance informed budgeting. Public Administration Review , 71(3), 356–367. Julnes, P., & Holzer, M. (2001). Promoting the utilization of performance measures in public organizations: An empirical study of factors affecting adoption and implementation. Public Administration Review , 61(6), 693– 708. Economics. The American Economic Review , 93 (5), 1449–1475. Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263 –292. https://doi.org/10.1017/CBO9781107415324.004 Karl, B. D. (1976). Public Administration and American History: A Century of Professionalism. Public Administration Review , 36(5), 489. https://doi.org/10.2307/974229 Katz, D., & Kahn, R. L. (1966). Th e social psychology of organization. New York, NY: Wiley. Kettl, D. F. (2005). The global public management revolution. Washington, D.C.: Brookings Institute Press. Key, V. O. (1940). The lack of a budget theory. The American Political Science Review , 34(6 ), 1137–1144. Kim, Y. (2011). The pilot study in qualitative inquiry: Identifying issues and learning lessons for culturally competent research. Qualitative Social Work , 10(2), 190 –206. https://doi.org/10.1177/1473325010362001 Kingdon, J. W. (2003). Agenda, alternatives, and public policies (2rd ed.). New York, NY: Longman. Kittipanyasiri, A. (2015). Planning and budgeting for capital improvements in Thai municipalities: Problems in the development of Thai municipal self governing capacity. Journal of African & Asian Local Government Studies , 4(1), 33–52. Klenk, N. L., & Hickey, G. M. (2011). Government science in forestry: Characteristics and policy utilization. Forest Policy and Economics , 13(1), 37–45. Knorr, K. (1977). Policy makers’ use of social science knowledge: Symbolic or instrumental? In C. H. Weiss (Ed.), Using Social Science Research in Public Policy making (pp. 165–182). Toronto, Canada: Lexington Books. Knott, J., & Wildavsky, A. (1980). If dissemination is the solution, what is the problem? Knowledge: Creation, Diffusion, Utilization, 1(4), 537–578. 215

PAGE 228

Kolawole, O. D. (2015). Twenty reasons why local knowledge will remain relevant to development. Development in Practi ce, 25(8), 1189–1195. Kong, D. (2005). Performance based budgeting: The U.S. experience. Public Organization Review , 5(2), 91–107. Kopits, G., & Craig, J. (1998). Transparency in government operations (Occastional Paper No. 158) (Vol. 158). Washington, D.C . Kroll, A. (2015a). Drivers of performance information use: Systematic literature review and directions for future research. Public Performance & Management Review , 38(3), 459 –486. Kroll, A. (2015b). Exploring the link between performance information use and organizational performance: A contingency approach. Public Performance {&} Management Review , 39(1), 7–32. https://doi.org/10.1080/15309576.2016.1071159 Krueathep, W. (2014). Bad luck or bad budgeting: An analysis of budgetary roles underpinning poor m unicipal fiscal conditions in Thailand. Public Budgeting and Finance , 34(3), 51 –72. Krueathep, W., Nudam, J., & Termtawewut, O. (2015). The practices of local government borrowing in Thailand. Journal of Political Science , 36(3), 1 –26. Laegreid, P., Roness , P. G., & Rubecksen, K. (2008). Performance informaiton and performance: Integrated system or loose coupling. In W. Van Dooren & S. Van de Walle (Eds.), Perfomance information in the public sector: How it is used (pp. 42–57). New York, NY: Palgrave Macmil lan. Landry, RJEAN, Amara, N., & Lamari, M. (2001). Climbing the ladder of reserach utilizaiton. Science Communication, 22(4), 396–422. https://doi.org/10.1177/1075547001022004003 Landry, Rejean, Lamari, M., & Amara, N. (2003). The extent and determinants of the utilization of university research in government agencies. Public Administration Review , 63(2), 192– 205. Lazarsfeld, P. F. (1935). The art of asking why in marketing research: Three principles underlying the formulation of questionnaires. Natioinal Market Review , 1(1), 26–38. Retrieved from http://www.jstor.org.login.bibproxy.whu.edu/stable/pdf/4291274.pdf?_=1464466834642 Lee, J. (2012). The administrative broker: Bureaucratic politics in the era of prevalent information. The American Review of Publ ic Administration , 43(6), 690 –708. Lee, R. D. (1991). Educational characteristics of budget office personnel and state budgetary processes. Public Budgeting & Finance , 11(3), 69 –79. Lee, R. D. (1992). The use of executive guidance in state budget preparati on. Public Budgeting & Finance , 12(3), 19–31. Leech, N. L., Barrett, K. C., & Morgan, G. A. (2011). IBM SPSS for intermediate Statistics: Use and interpretation (4th ed.). New York, NY: Routledge. Leech, N. L., & Onwuegbuzie, A. J. (2007a). A typology of m ixed methods research designs. Quality and Quantity , 43 (2), 265–275. https://doi.org/10.1007/s11135007 91053 216

PAGE 229

Leech, N. L., & Onwuegbuzie, A. J. (2007b). An array of qualitative data analysis tools: A call for data analysis triangulation. School Psycholog y Quarterly, 22(4), 557– 584. https://doi.org/10.1037/10453830.22.4.557 Leech, N. L., & Onwuegbuzie, A. J. (2008). Qualitative data analysis: A compendium of techniques and a framework for selection for school psychology research and beyond. School Psychol ogy Quarterly , 23(4), 587–604. https://doi.org/10.1037/10453830.23.4.587 Leland, H. E., & Pyle, D. H. (1977). Informational asymmetries, financial structure, and financial intermediation. The Journal of Finance , XXXII (2), 371–387. Levine, C. H. (1978). Or ganization decline and cutback management. Public Administration Review , 38(4), 316–325. Lewis, B. (2013). Local government capital spending in Indonesia: Impact of intergovernmental fiscal transfers. Public Budgeting and Finance , 33(1), 76–94. https://doi.org/10.1111/j.15405850.2013.12002.x Lindblom, C. (1979). Still muddling, but not through. Public Administration Review , 39(6), 517– 526. Lindblom, C. E. (1959). The Science of “muddling through.” Public Administration Review , 19(2), 79–88. Lindblom, C. E., & Cohen, M. D. (1979). Usable knowledge: Social science and social problem solving. New Haven, CT: Yale University Press. Liu, X., Lindquist, E., Vedlitz, A., & Vincent, K. (2010). Understanding local policymaking: Policy elites’ perceptio ns of local agenda setting and alternative policy selection. Policy Studies Journal , 38(1), 69–91. https://doi.org/10.1111/j.1541 0072.2009.00345.x Lorsuwannarat, T. (2017). Public participation in budgeting: The new path of budget reform in Thailand. Inte rnational Journal of Public Administration, 40 (5), 385–400. https://doi.org/10.1080/01900692.2015.1126730 Lundin, M., & berg, P. (2014). Expert knowledge use and deliberation in local policy making. Policy Sciences, 47(1), 25–49. Lynn, L. E. (1996). Publi c management as art, science, or profession. Ann Arbor, MI: University of Michigan Press. Lynn, L. E. (2006). . New York, NY: Routledge. Maiello, A., Viegas, C. V., Frey, M., & Jos, J. L. (2013). Public managers as catalysts of knowledge coproduction? Investigating knowledge dynamics in local environmental policy. Environmental Science and Policy , 27, 141–150. McBeth, M. K., Jones, M. D., & Shanahan, E. A. (2014). The narrative policy framework. In P. A. Sabatier & C. M. Wei ble (Eds.), Theories of the Policy Process (3rd ed., pp. 225–266). Boulder, CO: Westview Press. 217

PAGE 230

McCue, C. P. (1999). The impact of objective and empathic dispositions on local government budget analysts’ spending preferences. Public Budgeting and Finance , 19 (1), 89–114. https://doi.org/10.1046/j.02751100.1999.01158.x Melitski, J., & Manoharan, A. (2014). Performance measurement, accountability, and transparency of budgets and financial reports. Public Administration Quarterly , 38(1), 38– 70. Melkers, J. E. , & Willoughby, K. G. (2001). Budgeters’ views of state performance budgeting systems: Distinctions across branches. Public Administration Review , 61(1), 54–64. Melkers, J., & Willoughby, K. (1998). The state of the states: Performance based budgeting requirements in 47 out of 50. Public Administration Review , 58(1), 66–73. Melkers, J., & Willoughby, K. (2005). Models of performance measurementt use in local governments: Understanding budgeting, communication, and lasting effects. Public Administration Review , 65(2), 180– 190. Meyer, J. W., & Rowan, B. (1977). Institutionalized organizations: Formal structure as myth and ceremony. American Journal of Sociology , 83(2), 340. https://doi.org/10.1086/226550 Mikesell, J. (2007). Fiscal administration: Analysis and application in the public sector (7th ed.). Belmont, CA: Thomson, Wadsworth. Miles, M. B., & Huberman, M. A. (1994). Qualitative Data Analysis (2nd ed.). Thousand Oaks, CA: Sage Publications Ltd. Miles, M. B., Huberman, M. a, & Saldana, J. (2014). Qua litative data analysis: A methods sourcebook . Los Angeles, CA: Sage Publications Ltd. Miller, G. J. (2005). The political evoluation of principal agent models. Annual Review of Political Science , 8(1), 203 –225. https://doi.org/10.1146/annurev.polisci.8.082103.104840 Moe, T. M. (1984). The new economics of organization. American Journal of Political Science , 28(4), 739– 777. Morse, J. M. (1997). Approaches to qualitative quantitative methodological triangulation. Nursing Research , 40(1), 120–123. Moynihan, D. P., & Ingraham, P. W. (2004). Integrative leadership in the public sector: A model of performance information use. Administration & Society , 36(4), 427– 453. https://doi.org/10.1177/0095399704266748 Moynihan, Donald P. (2006). What do we talk about when we talk about performance? Dialogue theory and performance budgeting. Journal of Public Administration Research and Theory , 16(2), 151– 168. Moynihan, Donald P. (2008a). Advocacy and learning: An interactive dialogue approach to performance information use. I n W. Van Dooren & S. Van de Walle (Eds.), Perfomance information in the public sector: How it is used (pp. 24–41). New York, NY: Palgrave Macmillan. 218

PAGE 231

Moynihan, Donald P. (2008b). The dynamics of performance management . Washington, D.C.: Georgetown University Press. Moynihan, Donald P. (2015). Uncovering the circumstances of performance information use findings from an experiment. Public Performance & Management Review , 39(1), 33– 57. Moynihan, Donald P., Lavertu, S., & Kamensky, J. (2012). Does involvement in performance management routines encourage performance information use?: Evaluting GPRA and PART. Public Administration Review , 72(4), 592–604. https://doi.org/10.1177/0734371X9101100307 Moynihan, Donald P., Pandey, S. K., & Wright, B. E. (2012). Setting the table: How transformational leadership fosters performance information use. Journal of Public Administration Research and Theory , 22(1), 143– 164. https://doi.org/10.1093/jopart/mur024 Moynihan, Donald P, & Pandey, S. K. (2010). The big question for performance management: Why do managers use performance information? Journal of Public Administration Research & Theory, 20(4), 849–866. Mucha, B. M. J. (2009). Explaining the doctrine of performance management. Governm ent Finance Review , (June), 88–90. Mullins, D. R. (2007). Local budget process. In A. Shah (Ed.), Local Budgeting (pp. 213–264). Washington, D.C.: The World Bank. Nagai, F., Funatsu, T., & Kagoya, K. (2008). Central local government relationship in Thailand. In F. Nagai, N. Mektrairat, & T. Funatsu (Eds.), Local Government in Thailand: Analysis of the Local Administrative Organization Survey . Tokyo, Japan: Japan External Trade Organization. Retrieved from http://www.ide.go.jp/English/Publish/Download/Jrp/147.html NESCB, & Thamasat University. (2009). Decentralization and the budget for social services at Tambon administrative level: Thailand. Bangkok, Thailand: UNICEF. Newman, I., Ridenour, C. S., Newman, C., & DeMarco, G. M. (2003). A typology of research purposes and its relationship to mixed methods . (A. Tashakkori & C. Teddlie, Eds.), Handbook of mixed methods in social and behavioral research. Thousand Oaks, CA: Sage Publications Ltd. Nielsen, P. A., & Baekgaard, M. (2015). Performance information, blame avoidance, and politicians’ attitudes to spending and reform: Evidence from an experiment. Journal of Public Administration Research and Theory , 25(2), 545–569. Niskanen, W. A. (1971). Bureaucracy and Representative Government . Chicago, IL: Aldine Atherto n. O’Toole, L. J., & Meier, K. J. (2014). Public management, context, and performance: In quest of a more general theory. Journal of Public Administration Research and Theory , (August 2013), 237–256. 219

PAGE 232

Olsen, A. L. (2015). The numerical psychology of perfor mance information: Implications for citizens, managers, and policymakers. Public Performance & Management Review , 39(1), 100–115. Onwuegbuzie, a. J., Witcher, a. E., Collins, K. M. T., Filer, J. D., Wiedmaier, C. D., & Moore, C. W. (2007). Students’ perc eptions of characteristics of effective college teachers: A validity study of a teaching evaluation form ssing a mixedmethods analysis. American Educational Research Journal , 44(1), 113–160. https://doi.org/10.3102/0002831206298169 Onwuegbuzie, A. J. (2003). Expanding the framework of internal and external validity in quantitative research. Research in the Schools , 10 (I), 71 –89. Retrieved from http://psycnet.apa.org/psycinfo/2004 11894008 Onwuegbuzie, A. J., & Leech, N. L. (2006). L inking research questions to mixed methods data analysis. The Quantitative Report , 11 (3), 474–498. the Sampling Process More Public. The Qualitative Report , 12(2), 19–20. https://doi.org/10.1007/s1113500690003 Onwuegbuzie, A. J., Onwuegbuzie, A. J., Collins, K. M. T., & Collins, K. M. T. (2007). A Typology of Mixed Methods Sampling Designs in Social Science Research. The Qualitative Report , 12(2), 281–316. Ottoso n, J. M. (2009). Knowledge for action theories in evaluation: Knowledge utilization, diffusion, implementation, transfer, and translation. In J. M. Ottoson & P. Hawe (Eds.), KNowledge utilization, diffusion, implementation, transfer, and translation: Impli cations for eveluation--New directions for evalutaion. New York, NY: Wiley InterScience. Pandey, S. K. (2015). Performance information Use: Making progress but a long way to go. Public Performance & Management Review , 39(1), 1–6. https://doi.org/10.1080/15309576.2016.1071158 Petersen, M. A. (2004). Information: Hard and soft . Pollitt, C. (2006). Performance management in practice: A comparative study of executive agencies. Journal of Public Administration Research and Theory , 16(1), 25–44. https://doi.org/1 0.1093/jopart/mui045 Potter, C., & Brough, R. (2004). Systemic capacity building: A hierarchy of needs. Health Policy & Planning, 19(5), 336– 345. Radin, B. a. (2000). The Government performance and resultsaAct and the tradition of federal management reform : Square pegs in round holes? Journal of Public Administration Research and Theory, 10(1), 111–135. https://doi.org/10.1093/oxfordjournals.jpart.a024260 Raudla, R. (2012). The use of performance information in budgetary decisionmaking by legislators: Is estonia any different? Public Administration , 90 (4), 1000– 1015. https://doi.org/10.1111/j.14679299.2012.02041.x Reck, J. L. (2000). Ethics and budget all ocation decisions of municipal budget officers. Journal of Business Ethics, 27 (4), 335–350. 220

PAGE 233

Rich, R. F. (1975). Selective utilization of social science related information by federal policymakers. Inquiry , 12(3), 239–245. Rich, R. F. (1991). Knowledge creation, diffusion, and utilization: Perspectives of the founding editor of knowledge. Science Communication, 12 (3), 319–337. Riggs, F. W. (1998). Public administtation in America: Why our uniqueness is exceptional and import ant. Public Administration Review , 58(1), 22–32. Riggs, F. W. (2006). The prismatic model: Conceptualizing transitional societies. In E. E. Otenyo & N. S. Lind (Eds.), Comparative public adnistration (Research in public policy analysis and managment) (Vol. 15, pp. 17–60). Oxford, UK: Emeral Group Publishing Limited. https://doi.org/10.1016/S07321317(06)150022 Riley, J. G. (1975). Competitive signalling. Journal of Economic Theory , 10, 174–186. Ross, S. A. (1973). The economic theory of agency: The principal’s problem. The American Economic Review , 63(2), 134–139. Rothschild, M., & Stiglitz, J. (1976). Equilibrium in competitive insurance markets: An essay on the economics of imperfect information. In G. Dionne & S. E. Harrington (Eds.), Foundation of insur ance economics: Readings in economics and finance (pp. 629–649). New York, NY: Springer Science+Business Media, LLC. Rubin, I. (2015). Past and future budget classics: A research agenda. Public Administration Review , 75(1), 25–35. https://doi.org/10.1111/puar.12289 Salamon, L. M. (1987). Of market failure, voluntary failure, and third party government: Toward a theory of government nonprofit relations in the modern welfare state. Nonprofit and Voluntary Sector Quarterly, 16(1 –2), 29–49. Schick, A. (1966). T he road to PPB: The stages of budget reform. Public Administration Review , 26(4), 243– 258. Schick, A. (1990). The capacity to budget. Washington, D.C.: Urban Institute Press. Studi es in Comparative International Development , 38(3), 32–56. Schraeder, M., Tears, R. S., & Jordan, M. H. (2005). Organizational culture in public sector organizations: Promoting change through training and leading by example. Leadership & Organization Devel opment Journal , 26(6), 492– 502. https://doi.org/10.1108/01437730510617681 Scott, W. R. (1987). The adolescence of institutional theory. Administrative Science Quarterly, 32(4), 493. Seifert, J., Carlitz, R., & Mondo, E. (2013). The open budget index (OBI) as a comparative statistical tool. Journal of Comparative Policy Analysis: Research and Practice , 15(1), 87– 101. 221

PAGE 234

Seldon, A. (2002). Public choice or political sovereignty. In G. Tullock, A. Seldon, & G. L. Brady (Eds.), Government failure: A primer in public choice (pp. 129–136). Washington, D.C.: CATO Institute. Selznick, P. (1948). Foundations of the theory of organization. American Sociological Review , 13(1), 25–35. https://doi.org/10.1177/0038040712456559 Shaw, J., Danese, C., & Stocker, L. (2013). Spanning the boundary between climate science and coastal communities: Opportunities and challenges. Ocean and Coastal Management , 86, 80–87. https://doi.org/10.1016/j.ocecoaman.2012.11.008 Simon, H. (1957). Administrative behavior: A study of decisionmak ing process in administrative organization . New York, NY: The Free Press. Simon, H. A. (1946). The Proverbs of Administration. Public Administration Review , 6(1), 53. https://doi.org/10.2307/973030 Simon, H. A. (1973). Applying information technology to or ganization design. Public Administration Review , 33(3), 268. https://doi.org/10.2307/974804 Simon, H. A. (1985). Human Nature in Politics: The Dialogue of Psychology with Political Science. The American Political Science Review , 79(2), 293– 304. Singleton, R. A. J., & Straits, B. C. (2010). Approaches to Social Research (3rd ed.). Thousand Oaks, CA: Sage Publications Ltd. Sminia, H., & Van Nistelrooij, A. (2006). Strategic management and organization development: Planned change in a public sector organization. Journal of Change Management , 6(1), 99– 113. https://doi.org/10.1080/14697010500523392 Smit, B. (2005). Teachers, local knowledge, and policy implementation: A qualitative policy practice inquiry. Education and Urban Society , 37(3), 292– 306. https://doi.org/10.1177/0013124505275426 Smith, R. W. (2003). Ethical norms in public budgeting: Evolution for devolution? Journal of Public Budgeting, Accounting & Financial Management , 15(2), 205–227. Spence, M. (1973). Competitive and optimal responses to signals: An analysis of efficiency and distribution. Journal of Economic Theory , 7(3), 296–332. Stigler, G. J. (1961). The economics of information. Journal of Political Economy , 69(3), 213– 225. Stone , D. (1997). Policy Paradox: The Art of Political Decision Making . New York, NY: W.W. Norton & Company. Strauss, A., & Corbin, J. M. (1998). Basics of qualitvative research: Grounded theory procedures and techniques . Los Angeles, CA: Sage Publications Ltd. Sun, T.W., & Gargan, J. J. (1993). Determinants of administrative capacity: The case of Taiwan. International Review of Administrative Sciences , 59, 277–290. 222

PAGE 235

Sutheewasinnon, P., Hoque, Z., & Nyamori, R. O. (2016). Development of a performance management system in the Thailand public sector: Isomorphism and the role and strategies of institutional entrepreneurs. Critical Perspectives on Accounting , 40, 26 –44. Suwanmala, C. (2007). Thailand: Civic participation in subnational budgeting. In Anwar Shah (Ed.) , Paticipatory Budgeting (pp. 127–154). Washington, D.C.: The World Bank. Talbot, C. (2007). Performance Management. In Ewan Ferlie, L. E. Lynn Jr, & C. Pollitt (Eds.), The Oxford Handbook of Political Institutions (pp. 491–518). New York, NY. Tamkin, P., Hillage, J., & Willison, R. (2002). Indicators of management capacity: Developing a framework. Management . London, England: Council for Excellent in Management & Leadership. Tankha, S. (2009). Building administrative capacity in developing countries: SWAT teams or Beat Cops? Public Administration Review , 61(6), 1028–1030. Taylor, J. (2007). The usefulness of key performance indicators to public accountability authorities in East Asia. Public Administration and Development , 13(2), 341–352. https://doi.org/10 .1002/pad Taylor, J. (2011). Factors influencing the use of performance information for decision making in Australian state agencies. Public Administration, 89(4), 1316 –1334. https://doi.org/10.1111/j.14679299.2011.02008.x Teddlie, C., & Tashakkori, A. (2009). Foundations of mixed methods research: Integrating qualitative and quantitative approaches in the social and behavioral sciences . Washington, D.C.: Sage Publications Ltd. Retrieved from http://www.phweingarten.de/zesa/Prof._Dr._Bur ke_Johnson_Mixed_Methods_PRIMER.pdf Teddlie, C., & Yu, F. (2007). Mixed Methods Sampling: A Typology With Examples. Journal of Mixed Methods Research , 1(1), 77– 100. https://doi.org/10.1177/2345678906292430 Ter Bogt, H. J. (2004). Politicians in search of performance information?: Survey research on dutch aldermen’s use of performance information. Public Management , 20(August), 221– 252. Terman, J. N., & Feiock, R. C. (2015). Improving outcomes in fiscal federalism: Local political leadership and administrati ve capacity. Journal of Public Administration Research and Theory , 25(4), 1059–1080. Thurmaier, K. (1992). Budgetary decisionmaking in central budget bureaus: An experiment. Journal of Public Administration Research and Theory , 2(4), 463–487. Thurmaier, K. (2001). Budget rationality in a policy oriented state budget office. International Journal of Organization Theory & Behavior , 4(1/2), 133. Thurmaier, K. M., & Willoughby, K. G. (2001). Policy and politics in state budgeting . Armork, NY: M.E. Sharpe. Tomes , B. W. E., Berger, A. B., & Bassett, M. E. (2011). Using unassigned funds to balance the budget. Government Finance Review , (54), 45–48. 223

PAGE 236

True, J. L. (2000). Avalanches and incrementalism: Making policy and budgets in the United States. The American Review of Public Administration , 30 (1), 3 –18. True, James L., Jones, B. D., & Baumgartner, F. R. (2007). Punctuated Equilibrium Theory: Explaining stability and change in public policymaking. In P. A. Sabatier (Ed.), Theories of the Policy Process (2nd ed., pp. 155–187). Boulder, CO: Westview Press. Tullock, G. (1965). The politics of bureaucracy. Washington, D.C.: Public Affairs Press. Tullock, G. (2002). People are people: The elements of public choice. In G. Tullock, A. Seldon, & G. L. Brady (Eds.), Government failure: A primer in public choice (pp. 3–16). Washington, D.C.: CATO Institute. Tversky, A., & Kahneman, D. (1986). Rational choice and the framing of decisions. Journal of Business, 59 (4), S251–S278. United City and Local Government. (2016). Subnational governments around the world: Structure and finance . Paris, France: Organisation for Economic Cooperation and Development. Van de Ven, A. (2007). Engaged Scholarship: A Guide for Organizational and Social Research. New York, NY: Oxford University Press. Van Dooren, W. (2008). Nothing new under the sun?: Change and continuity in the Twentieth Century performance movements. In W. Van Dooren & S. Van De Walle (Eds.), Perfomance information in the public sector: How it is used (pp. 11–23). New York, NY: Palgrave Macmillan. van Teijlingen, E., & Hundley, V. (2002). The importance of pilot studies. Nursing Standard, 16(40), 33– 36. https://doi.org/10.7748/ns2002.06.16.40.33.c3214 Weber, M. (1968). Basic sociological terms. In G. G oth & C. Wittich (Eds.), Economy and society. Berkeley, CA: University of California Press. https://doi.org/10.4324/9781315775357 Weible, C. M. (2008). Expert based information and policy subsystems: A review and synthesis. Policy Studies Journal , 36(4), 615– 635. Weiss, C. H. (1977). Research for policy’s sake: The enlightenment function of social research. Policy Analysis, 3(4), 531 –545. Weiss, C. H. (1989). Congrssional committees as users of analysis. Journal of Policy Analysis & Management , 8(3), 411. https://doi.org/10.2307/3324932 White, R., & Smoke, P. (2005). East Asian decentralizes. In World Bank (Ed.), East Asia Decentralizations: Making local government work (pp. 1–24). Washington, D.C.: World Bank. Wildavsky, A. (1964). The politics of the budge t process. Boston, MA: Little Brown and Company. https://doi.org/10.1111/15405850.00996 Wildavsky, A. (1978). A budget for all seasons?: Why the traditional budget lasts. Public Administration , 38(6), 501–509. 224

PAGE 237

Williamson, O. E. (1971). The vertical integr ation of production: Market failure considerations. The American Economic Review , 61 (2), 112–123. Willoughby, K. G. (2002). A multiple rationality model of budgeting: Budget office orientations and analysts’ roles. In A. Khan & B. W. Hildreth (Eds.), Budge t Theory in the Public Sector (pp. 104–123). Westport, CT: Greenwood Publishing Group. Willoughby, K. G., & Finn, M. A. (1994). Organizational professionalism and technological sophistication: Budget offices in the South. Public Productivity & Management R eview , 18(1), 19–35. Willoughby, K. G., & Melkers, J. E. (2000). Implementing PBB: Conflicting views of success. Public Budgeting and Finance , 20(1), 85–120. Wilson, W. (1887). The study of Administration. Political Science Quarterly, 2(2), 197 –222. https: //doi.org/10.1038/scientificamerican1209 38 Witting, A. (2015). Measuring the use of knowledge in policy development. Central European Journal of Public Policy , 9(2), 54–71. Wolman, H., & Page, E. (2002). Policy transfer among local governments: An information theory approach. Governance, 15(4), 477–501. Wongpredee, A., & Sudhipongpracha, T. (2014). The politics of intergovernmental transfers in Northeast Thailand. Journal of Developing Societies , 30 (3), 343–363. https://doi.org/10.1017/S0007123400007481 Wood, B. D. (2010). Agency theory and the bureaucracy. In R. F. Durant (Ed.), The Oxford Handbook of American Bureaucracy (pp. 181–206). New York, NY: Oxford University Press. https://doi.org/10.1093/oxfordhb/9780199238958.003.0008 Wooldridge, J. (2010). Econometric Analysis of Cross Section and Panel Data. London, England: The MIT Press. https://doi.org/10.1515/humr.2003.021 World Bank. (2012). Central Local Government Relations in Thailand: Improving service delivery. Bangkok: World Bank.Yetano, A. (2013). What drives the institutionalization of performance measurement and management in local government? Public Performance & Management Review , 37 (1), 59–86. https://doi.org/10.2753/PMR15309576370103 Zahariadis, N. (2014). Ambiguity and multiple streams. In P. A. Sabatier & C. Weible (Eds.), Theories of the Policy Process (pp. 25–58). Boulder, CO: Westview Press. Zaltsman, A. (2009). The Effects of Performance Information on Public Resource A llocations: A Study of Chile’s Performance Based Budgeting System. International Public Management Journal , 12(4), 450–483. https://doi.org/10.1080/10967490903328931 Zhang, H., Van de Walle, S., & Zhuo, Y. (2016). Does trust in the performance measurement organization influence how public managers use performance information? Public Performance & Management Review , 40(2), 409 –430. 225

PAGE 238

Appendix A The Ordinal Regression Model for Information Utilization The indicator measuring the information use for decision making involves the five items assessing the extent to which information pieces are used for budgetary decision making (Taylor, 2007, 2011) . Controlling for structural isomorphism, incrementalism, context factors, and organizational culture from the department heads’ opinions, this study regresses this information use for budgetary decisi on making on information characteristics, administrative capacity, and budget rationality. 0 1Information Characteristics 234Structural Isomorphism 567Organizational Culture The indicators for each variable in the multiple regression model above are tabulated in the Table 1 below. Table A1 Variables and indicators for the multiple regression explaining the infor mation use for budget decision making Variable Variable Name Indicators (Question Items and Rating) Dependent Information use for budget decision making How often do the following activities happen? (0 = never, 1 = rarely, 2 = sometimes, 3=most of the time, and 4 = always) (1) The information pieces you use influence your budget decisions especially in terms of program prioritization (2) You use those information pieces in conjunction with other data to reach budget decisions (3) You use those information pieces t o prompt further budgetary questions or enquires during budget discussion or debate (4) You use those information pieces to support a budget decision reached by other means (e.g., professional experience) (5) You use those information pieces to argue that a certai n budget choice may not be desirable Independent Administrative capacity How would you rate the following budgeting skills of yours? (0 = none, 1 = low; 2 = relatively low, 3 = relatively high, and 4 = high) 226

PAGE 239

Variable Variable Name Indicators (Question Items and Rating) (1) Participatory budgeting (i.e., soliciting citi zen opinions, promoting local participation, preference modeling) (2) Communication skills (i.e., data presentation, expressing and discussing opinions) (3) Intergovernmental relations (i.e., networking and collaborating with other local governments, the central government, and its provincial agencies) (4) Strategic planning and management (i.e., problem structuring, analysis of strengths, weaknesses, opportunities, and threats, and detaining strategy plans) (5) Project management (i.e., stakeholder analysis, project feasi bility, monitoring, and evaluation,) (6) Accounting and financial management (i.e., budgetary techniques, cost benefit analysis, financial ratios, risk management) Information characteristics (Information piece) What are information pieces you frequently use while preparing budget proposals and discussing or debating projects and programs in budget proposals of yours or other departments? (Dichotomous variables) (1) Value of money (i.e., costing, pricing, and standard costs) (2) Service quality (3) Policy results (i.e., political, social, and/or economic implications) (4) Spending ceiling (based on past years’ revenues and the next year revenue forecast) (5) Mayoral policy directions and priorities (6) Communal problems and demands (7) Professional techniques and standards Information characteristics (Reliability) Where do you receive or can assess to or which organization does provide you each of those information pieces you frequently use? (Dichotomous variables) (1) Self observation (i.e., field work and casual talk with local citizens o r service recipients) (2) Performance contracts and reports (past, current, and next year) (3) Information center (i.e., the department of plans and policy analysis or the clerk office) (4) The central government and its agencies (5) Other local governments (6) Contractors or service providers (7) The internet 227

PAGE 240

Variable Variable Name Indicators (Question Items and Rating) In your opinion, to which extent is each of information providers above capable of data gathering, processing, and analysis? (0 = none, 1 = low; 2 = relatively low, 3 = relatively high, and 4 = high) Conside ring their integrity and benevolence, to which extent do you think each of those information providers as a trusted and legitimate information producer? (0 = none, 1 = low; 2 = relatively low, 3 = relatively high, and 4 = high) Information characteristics (Relevance) Based on the information pieces you use, to which extent do you think each of those information pieces you frequently use is useful? (0 = none, 1 = low; 2 = relatively low, 3 = relatively high, and 4 = high) Based on the information pieces you use, to which extent do you think each of those information pieces you frequently use is applicable? (0 = none, 1 = low; 2 = relatively low, 3 = relatively high, and 4 = high) Budget rationality (Budget orientations) Which of the following spending condition do you give the most weight when drafting your budget proposal? (Dichotomous variables) (1) Financial rules and regulations (2) Efficiency (i.e., costing, standard costs, and value of money) (3) Specific policy issues that needs attention (4) Executive’s policy directions (5) Financial and accounting techniques (i.e., performance management) Control Structural isomorphism Do you agree that the central government force your municipality to adopt the informationbased decision makings? (1 = disagree, 2 = quite disagr ee, 3 = quite agree, and 4 = agree) Do you agree that your municipality adopt the informationbased decision makings because it is what other local government do so? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Incrementalism Do yo u agree that your municipality’s budget process has drastically changed over the last four years? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Political factors (External pressure) Outsider’s information use Do you agree that other budget actors, other organizations, the local people, or the public can use information pieces (both performance information and others) against your department? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) 228

PAGE 241

Variable Variable Name Indicators (Question Items and Rating) against the department) P olitical factors (Stakeholder support) Leader commitment to information based decision making Staff’s commitment to information based decision making Do you agree that the city mayor support the activities related to the information gathering, processing, and analysis in the municipality? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Do you agree that the municipality’ staff members support the activities related to the information gathering, processing, and analysis in the municipa lity? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Organizational culture (Rational culture) Do you agree that most of municipal staff are concerned with getting the job done? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Organizational culture (Bureaucratic culture) Do you agree that bureaucratic procedures generally govern how the staff do their work? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Organizational culture (Group culture) Do you agree that most of the staff members in your municipality seem to share a lot of themselves? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Organizational culture (Development culture) Do you agree that your municipality is rea dy to change the work process in response to an emergency or if necessary? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) 229

PAGE 242

Appendix B The Regression Model for Perceptional Budget Allocation The equation that estimates the impact of administrative capacity, fiscal information’s characteristics and the ladder of information utilization on the perceptional budget allocation is (Ho, 2006; Melkers & Willoughby, 2005) : 0 1Administrative Capacity 23Strategy of Information Use 4Budget Factors + errors Table B1 below operationalizes all related variables in the equation above and captures the extent to which the department heads report having perceptions toward a par ticular sets of activities or factors. The measures of these perceptions are 0 = to no extent; 1 = to a small extent; 2 = to a moderate extent; 3 = to a great extent; and 4 = to a very great extent. Table B1 The Variables in the Model for Perceptional Bud get Allocation Variable Variable Name Indicators (Question Items and Rating) Dependent Perceptional budget allocation) An average of the extent to which respondent think the development and use of fiscal information been effective in his/her organization in the following five kinds of budget decisions (0 = to no extent; 1 = to a small extent; 2 = to a moderate extent; 3 = to a great extent; and 4 = to a very great extent): (1) affecting cost savings (2) changing appropriation levels (3) reducing duplicat ive services or programs (4) eliminating ineffective services or programs (5) improving service quality Independent Administrative capacity An average of the extent to which respondent routinely utilize reliable and relevant information for budget decision makin g based on the question: How would you rate the following budgeting skills of yours? (0 = none, 1 = low; 2 = relatively low, 3 = relatively high, and 4 = high) 230

PAGE 243

Variable Variable Name Indicators (Question Items and Rating) (1) Participatory budgeting (i.e., soliciting citizen opinions, promoting local participation, pre ference modeling) (2) Communication skills (i.e., data presentation, expressing and discussing opinions) (3) Intergovernmental relations (i.e., networking and collaborating with other local governments, the central government, and its provincial agencies) (4) Strategi c planning and management (i.e., problem structuring, analysis of strengths, weaknesses, opportunities, and threats, and detaining strategy plans) (5) Project management (i.e., stakeholder analysis, project feasibility, monitoring, and evaluation,) (6) Accounting and financial management (i.e., budgetary techniques, cost benefit analysis, financial ratios, risk management) Information utilization An average of the extent to which respondent routinely utilize reliable and relevant information for budget decision making based on the question: How often do the following activities happen? (0 = never, 1 = rarely, 2 = sometimes, 3=most of the time, and 4 = always) (1) The information pieces you use influence your budget decisions especially in terms of program prioritization (2) You use those information pieces in conjunction with other data to reach budget decisions (3) You use those information pieces to prompt further budgetary questions or enquires during budget discussion or debate (4) You use those information pieces to support a budget decision reached by other means (e.g., professional experience) (5) You use those information pieces to argue that a certain budget choice may not be desirable Strategy of Information Use Which approach do you apply for presenting the information in your budget proposal to your budget committee? (1) Doctrine of performance management (2) We are the experts (3) Framing (4) Mandates (5) Incrementalism Control (Budget factors) Community type Whether (1) or not (0) the respondent considers his/her jurisdiction (1) a trade/merchandise community or 231

PAGE 244

Variable Variable Name Indicators (Question Items and Rating) (2) a residential community (3) a mixed community, representing both urban and suburban characteristics (4) an agricultural community (5) others Population Number of population in respondent’s jurisdiction Location in a city area Whether (1) or not (0) the respondent considers his/her jurisdiction is located in the capital city of the province Control (Organizational Culture) Staff’s commitment to information based decision making Do you agree that the municipality’ staff members support the activities related to the information gathering, processing, and analysis in the municipality? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Task is motivational Extent to which respondent find it easy to motivate employees to be more results oriented in their programs, operations, and projects (0 = to no extent; 1 = to a small extent; 2 = to a moderate extent; 3 = to a great extent; and 4 = to a very great extent). Control (Structural Isomorphism) Coercive Do you agree that the central government force your municipality to adopt the informationbased decision makings? (1 = disagree, 2 = quite disagree, 3 = quit e agree, and 4 = agree) Normative Do you agree that your municipality adopt the information based decision makings because it is what other local government do so? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Control (Incrementali sm) Incrementalism Do you agree that your municipality’s budget process has drastically changed over the last four years? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Control (Political Factors) Conflict between the executives and th e legislators How do you rate the level of the differences of policy directions among the executives and the legislators in your municipality (0 = to no extent; 1 = to a small extent; 2 = to a moderate extent; 3 = to a great extent; and 4 = to a very great extent)? Conflict between the executives and the department heads How do you rate the level of the differences of policy directions among the executives and the department heads in your municipality (0 = to no extent; 1 = to a small extent; 2 = to a moderate extent; 3 = to a great extent; and 4 = to a very great extent)? 232

PAGE 245

Appendix C The Regression Model for Impact of Administrative Capacity and Information Utilization on Actual Budget Allocation Somewhat like the equation estimating the percepti onal budget allocation, the equation that estimates the impact of administrative capacity, information’s characteristics and the level of information utilization on the perceptional budget allocation is of organization level. Nonetheless, their control var iables are budget characteristics of each spending programs such as political contents, revenue generation, and budget size of the previous years (Gilmour & Lewis, 2006b; Ho, 2011; Zaltsman, 2009) . The spending programs of interest under the context of Thailand’s local budgeting include education and public works. These programs the literature suggests that the linkage between information utilization and budget allocation exists. The equation that estimates the impact of administrative capacity, information characteristics, and information utilization on actual budget allocation is: 0 1234BUDRAT 56789PROFF 10BGTSZ + 11REVGE 0 1Administrative Capacity 23Strategy of Information Use 444Political Factors + errors i ( i = 0 11) are the coefficients and the independent and control variables (i.e., organizational and political factors). The equation above is a crosssection model focusing on the budget allocation of FY2019. However, some variables are of FY2018, especiall y ones 233

PAGE 246

representing the budgetary incrementalism. Table C1 details the operationalization of related variables. Table C1. The Variables in the Model Estimating Impact of Administrative Capacity, Fiscal Information’s Characteristics, and Information Utiliza tion on Actual Budget Allocation (FY 20142015) Variables Derived from Models Measures of Variables Dependent Variable Increase Natural log of budget allocation of FY 2019 for the observations whose budget allocation of FY 2019 is more than that of FY 2018 Decrease Natural log of budget allocation of FY 2019 for the observations whose budget allocation of FY 2019 is more than tha t of FY 2018 Independent Variable Information utilization An average of the extent to which respondent routinely utilize the information during budget discussion Administrative capacity An average of the extent to which respondent routinely utilize re liable and relevant information for budget decision making Strategy Strategy of information use based on the answer to the question of whether (1) or not (0) respondent reports that they always presents their budget proposal with one of the following str ategies: “problem framing”, “it’s the mandate”, “performance doctrine”, “we are experts”, and “incrementalism” Political Factors Incrementalism Do you agree that your municipality’s budget process has drastically changed over the last four years? (1 = disagree, 2 = quite disagree, 3 = quite agree, and 4 = agree) Conflict between the executives and the legislators How do you rate the level of the differences of policy directions among the executives and the legislators in your municipality (0 = to no extent; 1 = to a small extent; 2 = to a moderate extent; 3 = to a great extent; and 4 = to a very great extent)? Conflict between the executives and the department heads How do you rate the level of the differences of policy directions among the executives and the department heads in your municipality (0 = to no extent; 1 = to a small extent; 2 = to a moderate extent; 3 = to a great extent; and 4 = to a very great extent)? Organizational Factors Population density Number of population per squared kilometer Employee Number of department full time employees 234

PAGE 247

Variables Derived from Models Measures of Variables Budgetary Factors Community Type Whether (1) or not (0) respondent represents one of the following type of community: commerce, residential, agricultural, mixed, or others. Budget Allocation Natural log of the budget allocation of FY 2018 Revenue Natural log of the total revenue of FY 2018 235