Citation
A study of state health data organizations

Material Information

Title:
A study of state health data organizations an exploration of state-level policy innovation
Creator:
Risker, Donald Christopher
Publication Date:
Language:
English
Physical Description:
ix, 285 leaves : forms ; 29 cm

Subjects

Subjects / Keywords:
Health services administration -- States -- United States ( lcsh )
Health services administration -- U.S. states ( fast )
United States ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Doctor of Philosophy, Public Administration.
General Note:
School of Public Affairs
Statement of Responsibility:
by Donald Christopher Risker.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
28246866 ( OCLC )
ocm28246866
Classification:
LD1190.P86 1992d .R57 ( lcc )

Full Text
A STUDY OF STATE HEALTH DATA ORGANIZATIONS: AN
EXPLORATION OF
STATE-LEVEL POLICY INNOVATION
by
Donald Christopher Risker
B.A., University of Kansas, 1972
M.S., Jersey City State College, 1984
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Administration
1992
t ' 3
? \ 1?


This thesis for the Doctor of Philosophy
degree by
Donald Christopher Risker
has been approved for the
Graduate School of Public Affairs
by
Date
Allan Wallis


Public Administration)
Risker, Donald Christopher (Ph.D.,
A Study of State Health Data Organizations: An
Exploration of State-level Policy Innovation
Thesis directed by Associate Professor E. Samuel Overman
Health data organizations represented a dominant
health care cost containment strategy of the 1980s in the
American states. This investigation explored this policy
change with three research propositions that were culled
from policy formulation, implementation, and innovation
literature: (1) interpretation of the policy "direction"
of health data organizations as top-down or bottom-up
within the federalist or social system, or a integration
of the two perspectives; (2) description of the
relationship between the private sector and public sector
to characterize the interorganizational relationships
where the policy crossed sectors; (3) evaluation of the
role of the cost on the interorganizational sets and the
role of cost in the diffusion of this policy.
The investigation methodology for exploring these
three research propositions was qualitative. Four in-
depth studies of the states of Colorado, Missouri, Oregon
and Pennsylvania were compared to the results of a survey
of the remaining health data organizations. The
investigation demonstrated that the health data
organizations were sensitive to the research propositions
by type of health data organization. Competitively-based
health data organizations showed a great deal of


negotiating and bargaining between interest groups
through their respective organizations. Non-
competitively-based health data organizations tended to
be top-down state-level initiatives. The relationship
between the private sector and the public sector was
varied. Generally, the private sector turned to the
public sector to: establish a safe harbor or neutral
party to contending interest groups, provide state-wide
jurisdictional authority, ensure uniformity of data
collection and systems, and in some states, offer health
data expertise. The cost factor was essentially
negligible. The competitively-based health data
organizations justified the cost on a cost-benefit basis.
The non-competitively-based health data organizations
justified the cost as any other regulatory program. The
comparison of the health data organization phenomenon to
the innovation literature demonstrated different levels
of innovation according to the type of health data
organization.
The form and content of this abstract are approved. I
recommend its publication.
Signed
E. Samuel Overman
iv


CONTENTS
Tables......................................... ix
CHAPTER
1. INTRODUCTION .................................... 1
State Policy-Making Processes .................. 8
Changes at the Federal Level................10
The Health Data Organization Concept .... 12
The "Buy Right" Philosophy..................15
Concepts and Definitions of Innovation ... 17
Innovation and Health Data Organizations . . 23
Contribution .................................. 26
Thesis Arrangement . .........................30
2. LITERATURE REVIEW .............................. 32
State Policy-Making.........................33
The Top-down Literature ....................... 37
The Bottom-up Literature ...................... 40
The Top-down, Bottom-up Synthesis ............. 42
The Private-Public Relationship ............... 47
The Cost Issue..............................55
Innovation Typologies and Research ............ 60
State-Level Policy Innovation Research ... 62
Federal Policy Innovation ..................... 65
Health Policy Innovation: A Mixed View ... 67
State Level Discretion and Innovation . . 68


Federal Control of Health Policy ....... 70
Organizational Innovation ..................... 72
Public Sector Innovation ................... 73
Adaptation, Replication, or Re-invention . 80
Public Sector Innovation ................... 81
Private and Public Sector Innovation ... 90
Research Prescriptions ........................ 92
3. METHODOLOGY......................................99
Design.........................................101
The Sample States..............................105
Data Collection................................109
Data Analysis..................................Ill
Limitations....................................112
Internal Threats .......................... 113
External Threats .......................... 114
4. RESULTS.........................................117
Colorado.......................................117
Missouri.......................................130
Oregon.........................................143
Pennsylvania..................................153
Cross-Case Patterning Analysis ............... 164
The Top-down, Bottom-up Proposition . 168
The Private-Public Relationship ........... 174
Cost........................................181
The National Survey Results .................. 185
Top-down, Bottom-up ....................... 192
vi


Private-Public Relationship ............... 194
Cost.....................................196
Comparison of the In-depth Cases to
the National Survey.........................198
Health Data Organizations Differentiated . . 202
Top-down, Bottom-up ....................... 202
Private-Public Relationship ............... 206
Cost.....................................207
5. CONCLUSIONS..................................209
Competitively-based Health Data
Organizations ................................ 209
Interest Group Bargaining ................. 210
Ideological Basis ......................... 213
Political Middle Ground ................... 217
Non-Competitive Health Data Organizations . . 218
Health Data Organizations and Innovation . . 218
Competitively-based Health Data
Organizations and Innovation .............. 219
Non-Competitively-based Health
Data Organizations and Definitions
of Innovation............................238
Innovation Conclusions ... ................... 241
Practical Implications ....................... 245
Theoretical Implications ..................... 246
Further Research ............................. 247
The Future of Health Data Organizations . . . 250
vii


APPENDIX
A. COVER LETTER AND NATIONAL SURVEY .............. 255
REFERENCES.......................................... 259
INTERVIEWS AND CORRESPONDENCE ...................... 280
viii


TABLES
Table
4.1 The In-depth Case Study Results................165
4.2 National Survey Results ....................... 186
4.3 Competitive Health Data Organizations ......... 204
4.4 Regulatory Health Data Organizations .......... 205
IX


CHAPTER 1
INTRODUCTION
State sponsored health data organizations swept the
American states in the 1980s as a strategy for
controlling health care costs. By 1990 thirty-one states
had passed legislation establishing a state sponsored
health data organization. This represented a spectacular
adoption among the American states of a new, state level
legislative policy. Rosenbaum (1976) observed that new
policies that are established by statute can be expected
to take up to twenty years before they are adopted by
twenty states.
Health data organizations were enacted much more
quickly. Health data organizations were enacted into
statutory law in thirty-one states over a twelve-year
period. In the six-year interval of 1984 to 1990,
fifteen (15) state sponsored health data organizations
were legislated into existence. What was the nature of
this new policy for controlling health care costs? Why
were states eager to adopt this new policy? And central
to this investigation, to what extent was this new policy
innovative?
This investigation was designed to evaluate these
and similar policy questions through an exploratory


research approach to the health data organization
phenomenon. This was a retrospective, exploratory case
study approach (Yin, 1982) which described the
formulation, advancement, and enactment of state
sponsored health data organizations. Three research
propositions were culled from the related policy
formulation, implementation, and innovation literature in
order to map the emergence of the health data
organizations.
The first proposition was whether the health data
organizations could be best described, as a top-down or
bottom-up phenomenon, or as a synthesis of these two
approaches as developed by Elmore (1983) and Sabatier
(1986). A great deal of normative prescription (Elmore,
1978) and research has been developed around both
implementation and innovation in regard to the direction
implementation and innovation take within political and
organizational systems. More recently, mixed models, or
integrative theory, of top-down or bottom-up were
developed in order to construct a more holistic model of
policy formulation and implementation (Barrett & Fudge,
1981; Barrett & Hill, 1984; Elmore, 1983; Hjern, 1982;
Hjern & Hull, 1982; Knoepfel & Weidner, 1982; Sabatier,
1986).
The integrative approach expanded the understanding
of top-down or bottom-up policy formulation and
2


implementation. It has given a broader view of policy
and developed a notion of interorganizational systems.
Much organizational research in recent years has
emphasized the importance of organizational sets for
understanding single organizations, policy formulation,
and policy implementation. Perrow (1986) dramatized the
difference between multiple organizational inquiry and
organizational inquiry that held the single organization
as the central focus of inquiry. Perrow (1986) used the
vehicle of a fictitious network of hospitals, in
"Regional City" (p. 200) to make his point.
The lesson Perrow (1986) drew from his dramatization
was that a complete analysis of an organization requires
a network analysis that reaches across sectors and
governmental levels, including a variety of government
and non-government organizations. Perrow's approach gave
attention to lateral and vertical interaction. The
identification of these institutions and potential policy
actors is what Barrett and Fudge (1981) referred to as
the "implementation structure" (p. 254).
In an approach similar to Perrow's, Osborne (1988)
identified innovations on the part of the states that
demonstrated lateral interaction between government and
business. Thus, the top-down, bottom-up proposition
includes linked organizations as well as the policy
environment. The proposition asks the question of
3


whether the health data organization phenomenon exhibited
interaction or linkages between and among organizations
and policy actors.
The second proposition of this investigation was to
describe the nature of the private-public relationships.
Was there interaction between sectors and what was the
nature of this interaction? Was the interaction typical
or atypical of what is known about innovations? Osborne
(1988) described a high degree of interaction between the
public and private sectors, akin to Perrow's
interorganizational action set in his fictitional
"Regional City." Although unlike Perrow's description,
Osborne's (1988) observations emphasized intense levels
of cooperation that could only be characterized as
private-public partnerships. Did health data
organizations exemplify this type of public-private
partnership? Or was the relationship more adversarial
and competitive?
The third proposition was to what extent did cost
play a role in the enactment or adoption of the state
sponsored health data organization strategy? Was the
health data organization strategy attractive because of a
low entry cost, or was it attractive because of its
benefits or a balance between the two? A significant
amount of research has been developed around scarce
resources and the perceived need for innovation (Wolman,
4


1986). In this literature, cost has also figured
prominently as an attribute of an innovation and wealth
as an attribute of an inventing or adopting organization.
It was natural to ask how these attributes were or were
not related in the case of health data organizations.
Deutsch (1985) in his discussion of broader, sociological
trends of innovation, identified a number of innovation
dynamics around who bears the cost of innovations. He
generally argued that innovation costs provide a map of
the power relationships for a particular innovation.
According to Deutsch (1985), conflict among interest
groups is expressed over who bears the cost of the
innovation. Did the health data organization phenomenon
exemplify the type of interest group activity described
by Deutsch?
Why try to characterize health data organizations as
an innovation at all? What makes the health data
organizations interesting to study from the innovation
perspective? The health data organization strategy
reversed the intent and focus of health care cost
containment regulation at the state level.
The "action" charge of the health data organizations
would have left the question of budgets and rates to be
settled in the market place, not in regulatory
commissions. This placed a different emphasis on how
health data would be used. It would not be used to tell
5


the hospitals what to charge. It would tell purchasers
from whom they should purchase health care services.
Health data organization supporters assumed that
discriminatory purchasing practices would in turn
stimulate competition among providers. Competition then
would determine what providers would, or perhaps more
accurately, could charge. This was a reversal of the
approach of direct regulation.
There have been many sources of health data in past
decades. Hospital discharge data systems, as well as
some other sources of health data, had been developed and
used since the early 1970s. Some health data was
developed out of the Social Security Act amendments of
1972 which established the professional standards review
organizations. The professional standards review
organizations were charged with reviewing the quantity
and quality of care provided for Medicare and Medicaid
hospital patients (Rakich, Longest, & Darr, 1985). The
professional standards peer review organizations
developed data which was primarily related to inpatient
diagnoses and inpatient length of stay.
Data of a different nature was developed under the
certificate of need program (CON) under Section 1122 in
conjunction with P. L. 93-603 of 1974 or the National
Health Planning and Development Act. The CON established
oversight for hospital capital expenditures. States
6


established monetary thresholds for review. Hospitals
and nursing homes were required to submit detailed
proposals for approval prior to building expansions or
acquiring equipment. Some states developed rate and
budget review processes to complement the CON.
These cost containment efforts on the part of the
federal government and the states had limited success in
most states. Focused on hospital expansions, equipment
acquisitions, rate setting and budget review, these
federally stimulated initiatives had threatened provider
autonomy. In some states, the threat to provider
autonomy had considerable impact on implementation of the
review process (Singleton, 1980). For some states, this
meant that these state rate setting and budget review
initiatives were short lived.
Why were states willing to take a different course
of action? Certainly a part of the answer was in the
rapidly escalating health care costs in the states.
Health care costs rose one-fifth faster than inflation in
the 1970s and 1980s. During this time period, the cost
of hospitalization more than doubled the general rate of
inflation (Dresang & Gosling, 1989). Perhaps more
importantly at the state level, there has been a close
linkage between welfare costs and health care costs.
When state welfare recipients increased so did state
7


health care costs (Dresang & Gosling, 1989; Rivlin,
1991) .
These alarming costs sent public policy experts and
business leaders looking for solutions. The social and
political noise caught the attention of the provider
community and health care interest groups. The provider
community and the health care interest groups were intent
on being a part of any solution that evolved out of this
activity. This set the political table for interaction
among these various sectors. It also set the table for
the ideological shift of the 1980s. What was eventually
served up was the state sponsored health data
organization strategy. Two important factors that impact
the acceptance and advancement of state policies,
particularly innovative policies, are the state policy-
making processes, and political changes that cause policy
shifts at the federal level.
State Policy-Making Processes
The chosen pathways and outcomes of state policy
efforts are difficult to study. One of the major
difficulties in researching state policies is that the
states vary widely in regard to structure and function.
Some states have "strong" governors, often designated by
the number of high-ranking officials that they can
appoint, some have "weak" governors. States also vary
8


according to their legislatures. States differ on the
support, funding, and salaries that they offer their
legislatures (Hrebnar & Thomas, 1987). The relative
strength of these sources of power in the state
governments may determine the relative strength of
political action committees (PACs) and the relative
strength of interest group politics. For example, it has
been argued that states with weak governors and/or
legislatures are particularly vulnerable to interest
group politics and strong lobbies (Hedrick & Zeigler,
1987).
Elazar (1972) identified specific state political
cultures that may determine the relative strength of
state institutions and the solutions these institutions
seek to governmental problems. Elazar (1972) developed a
typology of state cultures. Elazar's typology suggested
that specific political cultures have preferences for how
the legitimate role of government is defined and how the
process of governance should be carried out. Whether
structure and function precede culture, or the other way
around, both influence the policy outputs of state
government.
Yet these variances make comparisons problematic,
especially in the area of policy studies where the
phenomenon is not easily distinguished from its context
(Yin, 1982). Polsby as recently as 1984 observed that
9


state and local innovations have been little studied.
However, since Reagan's new federalism, states have been
the subject of intensified comment and research, and
researchers have been much more sanguine in their
assessment of state level innovation and governmental
effectiveness (Osborne, 1988; Reeves, 1990; Schneider,
1989).
Changes at the Federal Level
Reagan's new federalism engendered both positive and
negative consequences that forced the states to act as
policy initiators. The severe economic downturn at the
beginning of his administration created severe budgetary
shortfalls for a number of the American states. The
shortfalls sent state governments looking for innovative
methods for improving economic conditions and funding
needed programs. As Rivlin (1991) asserted "With less
federal help available, states had to meet rising needs
on their own" (p. 11). Many of the innovative efforts
were developed to close budgetary shortfalls. These
innovative efforts often included public-private
partnerships, as well as substantial interaction between
local and state government (Osborne, 1988).
The Reagan administration altered the flow of
revenue sharing with the states by reducing the number of
categorical grants and changing some of the categorical
10


grants to block grants.1 On the positive side, this
change did away with a great deal of red tape for state
and local governments and gave discretion to the state
and local governments to spend federal transfer dollars
as local needs dictated. Critics charged that this left
open the possibility that monies would still not be spent
efficiently. As an example of one type of innovation
created by these changes and in an effort to meet the
efficiency challenge, Pennsylvania developed a program to
avoid the problems of inefficiency. In the approach they
developed, state and local governments would have to
compete for grants based on the quality of their programs
(Osborne & Gaebler, 1992). States during this time took
on more of the character of Brandeis' phrase
"laboratories of democracies" (Osborne, 1988). The
innovative efforts by Pennsylvania and many other states
on a broad spectrum of topics, from education to economic
JDirect federal payments to state and local
government have consisted of grants-in-aid in categorical
and block grants, during the time between 1972-1986,
categorical grants-in-aid represented the bulk of general
revenue sharing. Categorical grants are for specific
programs that require state and local matching funds.
State and local governments are held accountable through
detailed reports. Federal agencies also frequently hold
the power of approval over plans. The federal agencies
conduct financial audits and may develop additional
regulations. See D. Berman (1991) State and local
politics (pp. 31-36). Dubuque, IA: Wm. C. Brown
Publishers. For a discussion of the implementation
realities of federal grants, see Ingram, H. (1977).
Policy implementation through bargaining: the case of
grants-in-aid. Public Policy. 25(4), 500-526.
11


development, increased the interaction and partnership
between state and local governments.
The health data organization strategy was ushered in
on the cusp of this new era in government. Naturally,
these conditions raised the question of whether and to
what extent health data organizations reflected these new
trends in government and the private sector. Were
partnerships between the public and private sectors
formed? Was there similar interaction between local and
state governments on the health data organization
strategy?
The Health Data Organization Concept
State sponsored health data organizations were
enacted by a majority of the states to establish or
augment existing state administrative structures to
collect, publish, and disseminate health data.2 The
"action" charge of most of the health data organizations
was to produce data that would be consumed in the health
2 Data was the most commonly used term in the health
data organization enabling legislation and by the policy
players. Often writers on the subject of data and
information choose to discriminate between different
levels such as data, information, knowledge, etc. See
Harland Cleveland's (1985) the Knowledge Executive (pp.
21-22), New York: E.P. Dutton. The intent of health data
organizations was to produce aggregate data to support
the decisions of health care purchasers. Data and
information were understood to be synonymous for the
purposes of this thesis.
12


care market place,3 but where was this data to come from?
And who was to receive it?
The principal providers of the health data in the
states were the acute care hospitals. Other health care
providers, such as physicians, were considered, but
initially the practical difficulties of collecting data
from physicians overwhelmed the most sanguine regulators.
This data collection problem was compounded by the need
to produce the comparative data that large purchasers
would need in order to distinguish between efficient and
inefficient providers of health care. Hence, in most
states, the provider came to be known as the state's
acute care hospitals. An additional part of the logic of
requiring only hospitals to report was that about 44% of
3The literature on consumer choice theory crosses at
least three different disciplines and has been developing
for approximately twenty years. First, there is an ample
body of literature that has been developed in the
behavior sciences dealing with individual choice and
decision-making under differing assumptions and
circumstances. See Lanzetta, J. T., St Kanareff, V. T.
(1975). Information cost, amount of payoff, and level of
aspiration as determinants of information seeking in
decision making. Behavioral Science. 7(4), 459-473.
Second, economists have similarly developed complex
models of choice in environments of variable amounts of
information see Grether, D. M., & Plott, C. P. (1982).
Economic theory of choice and the preference reversal
phenomenon. The American Economic Review. 72, 569-584.
And third, the discipline of marketing has developed a
complex literature on information and consumer choice,
see Murray, K. B. (1991). A test of services marketing
theory: Consumer information acquisition activities.
Journal of Marketing, 55. 10-25.
13


the national health care dollar went to hospitals (Coile,
1990).
The intended consumer of the information varied from
state to state. In some states, the intended consumer of
the information was the public. In other states, the
intended consumer was the state government. The public
for most states implied the large, corporate purchasers
of health care services, not the individual consumer, or
person on the street. In fact, many supporters and
advocates of the health data organization strategy held
that the individual had too little or no control of their
choice of hospital for the data to be relevant to them.
Supporters and detractors of the health data organization
strategy alike argued that individuals would have
difficulty in accurately interpreting health data.
The collection, publication, and dissemination of
health data by the state sponsored health data
organization was intended to allow major purchasers of
health care to practice rational, price-sensitive
purchasing of health care services. Or in the case of
those health data organizations that were appended to
existing rate setting efforts, to artificially stimulate
competition through dictating rates.
14


The "Buy Right" Philosophy
The state sponsored health data organization
strategy emphasized that the dissemination of health data
was intended to allow major purchasers of health care to
explicitly or implicitly practice the "buy right"
philosophy of Walter McClure, Ph.D.4 It was the
responsibility of the health data organization to produce
information that could be used in the market place. It
was the responsibility of the market place to use that
information to distinguish among the providers of health
care. The purchasers were to distinguish among providers
based on price, and when possible, based on the quality
of services.
McClure maintained that the systematic practice of
the "buy right" strategy on the part of purchasers would,
over time, restructure the health care market place.
McClure argued that the current health care market place
was unsound. The "buy right" strategy, according to
McClure, would instill a sound market place in health
care services (W. McClure, personal communication, June
4McClure worked at InterStudy, a Minneapolis-based
think tank founded by Paul M. Ellwood, Jr., another
leading advocate of the market model in health care and
developer of the health maintenance organization model.
In 1981, McClure departed InterStudy and founded the
Center for Policy Studies, which also is headquartered in
the Twin Cities. McClure developed what has become known
as the "buy right" strategy.
15


14, 1991). The "buy right" strategy was summarized by
McClure (1985):
The buy right strategy can be understood as follows.
When a purchaser asks a provider to render high
quality care but do it efficiently, the purchaser
really asks the provider to get the best health
results for patients but do it with fewer and less
costly services. In other words, do top-notch care,
but earn less on each patient.
When quality providers are asked to earn less
on each patient, there is only one way to reward
them: Send them more patients. This is the central
insight of the buy right principle, (p. 41)
McClure reported that he was directly influential in the
states of Colorado, Iowa, Pennsylvania, and locally in
Cleveland, Ohio (W. McClure, personal communication, June
14, 1991). Nevertheless, to the extent that McClure's
philosophy reflected market place mechanisms, the "buy
right" philosophy was the guiding philosophy for other
health data organizations as well.
Interestingly, McClure felt that he failed to
correctly guide the data efforts in the states of
Colorado, Iowa, and Pennsylvania. McClure stated he had
not intended for these states to adopt state sponsored
health data organizations. He would have preferred for
the business community to have worked with the provider
16


community, as currently is the practice in Cleveland,
Ohio (W. McClure, personal communication, June 14, 1991).
The ultimate goal of the health data organization
strategy was to restructure the health care services
market place through the dissemination of information.
This restructuring would then invoke the "Invisible Hand"
of Adam Smith (Heilbroner & Malone 1987, p. 2) to govern
all the selling and purchasing of health care services.
McClure's national strategy was to see the "buy right"
strategy and its market restructuring take hold in all
metropolitan markets. The restructuring would result in
lower health care costs for all. The savings could then
be used to address the underserved rural and the
uninsured poor (W. McClure, personal communication, June
14, 1991). It is this specific emphasis that made the
health data organization strategy different from cost
containment efforts in the past, and underscores the
value of assessing the health data organization strategy
as a state level innovation.
Concepts and Definitions of Innovation
The concepts and definitions of innovation tend to
vary according to discipline and the sociological level
at which innovation is studied. The multiple levels are
partly due to the pervasive interest in innovation.
Innovation research has become a major area of research
17


and speculation, primarily because of its social and
economic implications (Kanter, 1984; Merritt & Merritt,
1985; Zaltman, Kotler, & Kaufman, 1972). Indeed,
successful innovation has been placed at the very heart
of the American economy (Kanter, 1983). Kanter asserted
that "...where America leads, it leads because of
innovation" (p. 20). Downs and Mohr (1976) observed that
another reason for the popularity of innovation research
is that its research provides .generalizability that
has become rare as social science becomes increasingly
specialized" (p. 700).
Innovation can be studied at the macro-sociological
level as well as the micro-sociological level. At the
macro-sociological level, innovation theory is concerned
with changes in systems of thought, such as Darwinian
theory (Merritt & Merritt, 1985). At the micro-
sociological level, innovation theory is concerned with
organizational processes that lead to managerial and
product innovation. Policy innovation is a hybrid
between these two points on the innovation continuum,
containing some elements of both macro-phenomena and
micro-phenomena.
Innovations spread through macro-social and micro-
social systems by a process called diffusion. Rogers
(1977) set forth four elements he considered essential to
the analysis of diffusion of innovations:
18


(1) the tracing of an innovation, (2) over time, (3)
through specific channels of communication, and (4)
within a social structure, (p. 12)
Walker (1969) in his seminal work on state innovation did
not choose to distinguish invention and adoption or
diffusion. Walker defined an innovation as "a program or
policy which is new to the states adopting it, no matter
how old the program may be or how many other states may
have adopted it" (p. 881). Starling (1976/1988)
identified three characteristics of policy innovation.
First, he argued that innovation should be defined as "an
inventive alternative [that] builds on a new combination
of available knowledge concerning properties of the
social system" (p. 302). Second, he stated that "an
inventive alternative must be the product of a mental
effort above the average" (p. 303). Third, he stated
that "an inventive alternative must be productive" (p.
304). Later, Polsby (1984) offered a similar definition
of innovation, but also did not differentiate between
adoption and invention.
Similarly, Rosenbaum (1976) in his analysis of
innovative land use legislation, defined an innovative
legislative act as:
a statute which represents a distinct departure from
past state practicethat is, it establishes a new
19


state policy and program or extends an existing
state policy and program into a novel area. (p. 4)
However, Rosenbaum (1976) did delineate between a state
of origin for the innovative policy and a diffusion
process to other states.
Non-policy innovation theorists have also exhibited
divergent definitions of innovation and varying concerns
with invention and diffusion. Rogers and Kim (1985)
argued that an innovation is "an idea, practice or object
perceived as new by an individual or other relevant unit
of adoption" (p. 87). Similarly, Kanter (1983) argued
that:
Innovation is the generation, acceptance, and
implementation of new ideas, processes, products, or
services. It can thus occur in any part of a
corporation, and it can involve creative use as well
as original invention, (p. 21)
Wolman (1986) called this type of innovation definition
"organisational innovation" (p. 160). By "organisational
innovation" he meant that if the idea was perceived as
new by the relevant reference group, then it constituted
an innovation.
On the other hand, Becker and Whisler (cited in
Zaltman, Duncan, & Holbeck, 1973) defined innovation "as
the first or early use of an idea by one of a set of
organisations with similar goals" (p. 11). In a similar
20


vein, Pettigrew (1973) offered that innovation could be
understood to be "the adoption of a change which is new
to an organisation and to the relevant environment" (p.
11).
Downs and Mohr (1976) broadly defined innovation as
"...the adoption of means and ends that are new to the
adopting unit" (p. 701). Wolman (1986) described this
set of definitions as "systemic innovation" (p. 160).
In "systemic innovations" argued Wolman (1986), as in the
case of Rosenbaum's (1976) description of innovation,
early adopters are inventors, later adopters are
concerned with changing the organization, not with
invention.
The definition or understanding of innovation has
spread to the concepts of adoption as well. There has
been general agreement that just 'newness' does not
constitute an innovation (Wolman, 1986). Adoption is
requisite to invention becoming innovation. Moreover,
adoption patterns have been a matter of concern to policy
innovation theorists. Walker (1969) was concerned with
receptivity of innovative policies among the American
states. Rosenbaum (1976) made some observations about
diffusion patterns. Savage (1978) amended Walker's
interpretation by better defining the possible adoption
date for the states. Rogers (1983) described five types
of innovation adopters: (1) innovators; (2) early
21


adopters; (3) early majority; (4) late majority; and (5)
laggards (p. 23).
Going farther into the question of adoption of an
innovation, it has been argued that innovation may not
have occurred unless it has been adopted and implemented
(Downs & Mohr, 1976). Rogers and Shoemaker (1971) made
the distinction between "symbolic versus action
innovation-decisions" (p. 21). Rogers and Shoemaker
(1971) posited that innovations have two components: "(1)
an idea component, and (2) an object component. Idea
adoption is symbolic; object adoption requires action"
(p. 21). It has been argued that the idea component and
the object component are two separate stages that require
separate study (Downs & Mohr, 1976). Kanter (1983)
similarly emphasized this distinction: "Application and
implementation are central to this definition; it
involves the capacity to change or adapt" (p. 21). Yin
(1979) after having acknowledged many of the
aforementioned definitions, emphasized that an
"innovation is an artifact of some sort that is newly
introduced into an organization" (p. 45). Yin (1979) was
also concerned with implementation. He viewed
innovations as having three broad stages: (1) "initiation
and adoption," (2) "implementation," and (3)
"routinization" (pp. 4-5).
22


Barrett and Hill (1984) refined the notion of
implementation even further. They argued that
implementation is an effort to produce "change" (p. 222)
in organizations. Hence, by defining innovation as
simply "change" (p. 222) they link innovation and
implementation in a stroke. They acknowledged Yin's
notion of routinization, but they argued that even though
innovations may eventually disappear in organizations,
rule making continues the process of mutual adaptation.
In their words, "The fact that bargaining is less
visible does not mean that it does not go on" (Barrett &
Hill, 1984, p. 228).
An 'innovation' tends to be the product of the
definition rather than the result of inherent qualities.
Some definitions of innovation appear a better
description of the subject at hand, health data
organizations, than others, but the matter is worth a
closer look.
Innovation and Health Data Organizations
The health data organization phenomenon fitted
neatly with the policy innovation definitions of Walker
(1969), Starling (1976/1988), and Polsby (1984). Walker
emphasized 'newness' as the essential condition for
innovation. Certainly, health data organizations were a
new, state level cost containment approach. This new
23


cost containment strategy also stressed a novel use of
health data. Health data was no longer the prerogative
of regulators, but an essential element of the health
care market place. The use of health care information in
this manner would lead to competition and ultimately the
restructuring of the health care market place.
However, the policy literature of Walker (1969),
Starling (1976/1988), and Polsby (1984) on innovation
focused principally on two dynamics of innovation:
enabling legislation, the top-down analytical
perspective, and newness. As is apparent from the brief
discussion on the definition of innovation, there are
many other dynamics to implementation and innovation,
such as the bottom-up perspective. The definitions of
Walker (1969), Starling (1976/1988), and Polsby (1984)
assumed a top-down character to policy innovation.
Therefore, their definitions are not likely to provide a
full understanding of the health data organization
phenomenon. Sabatier (1986) commented that top-down
research is made more facile by the saliency of
legislation, as might be seen in the work of Walker,
Starling, and Polsby, but it may not provide a complete
picture of the policy formulation and implementation
process.
24


For example, by 1981 Walker had abandoned the top-
down research approach and had begun to write of policy
communities. O'Toole (1986) commented that:
Some researchers (those using a 'top-down'
perspective) have focused primarily on those aspects
of the implementation process putatively accessible
to or controllable by those who formally adopt
official policies, or mandates, (p. 182)
Bottom-up research, contrarily, deals with "a network of
actors involved in the planning, financing, and execution
of the relevant governmental and non-governmental
programs" (Sabatier, 1986, p. 33).
Plainly, health data organizations officially began
with state legislation and a new policy perspective, or
the top-down perspective. Health data organizations met
a few of the definitions of innovation, but the
literature is sufficiently ambiguous and diverse to
require that an analysis of health data organizations
include multiple perspectives, including a bottom-up
analysis and organizational set theory. Perrow (1986)
argued that only through the full establishment of the
"background" (p. 206) of a single organization does the
"figure" (p. 206) of the organization become clear.
Without the background, the figure disappears Perrow
maintained.
25


Berman (1992) was much more explicit in his
understanding of the notion of 'background.1 Berman
(1992) defined 'background' as the "socio-political
circumstances" (pp. 7-8) in which a new (new to that
system) model is introduced and implemented. In
addition, Berman (1992) chose to distinguish this notion
of 'background' from 'foreground.' The 'foreground'
Berman (1992) asserted includes organizational
characteristics such as leadership, staff capabilities
and training, and resource allocation decisions. It
becomes, then, difficult to describe health data
organizations as a de facto innovation, or to describe
them as top-down or bottom-up without considering an
interorganizational network.
Contribution
This exploratory investigation was designed to
assess how health data organizations can best be
described when compared to the relevant policy outcome,
implementation, and innovation literature. The
contribution of this investigation extended from the
three research propositions culled from the literature:
top-down or bottom-up, private-public relationship, and
cost. These three broad propositions interconnect across
a large portion of the policy, implementation, and
innovation literature, so that although the approach of
26


this investigation was exploratory, the research
propositions provided a template against which the health
data organization phenomenon could be measured.
Policy research has tended to focus on either a top-
down or bottom-up perspective (Sabatier, 1986). The
health data organization phenomenon offered a unique
opportunity to interpret a new state-level policy from
both perspectives and toward an integrative theory of
top-down and bottom-up.
Alexander (1982) observed to date that previous
implementation literature fell into three categories.
The first category according to Alexander (1982) included
"definitions of implementation with valuable
conceptualizations of the implementation process and
schemes for its analysis. Most of these acknowledged
their predecessors but paid little heed to them and
rarely offered any empirical evidence in the form of case
analyses" (p. 133). The second category presented case
studies with little if any theoretical base. The third
category included "a few works" (Alexander, 1982, p. 133)
that combined a conceptual framework with case studies.
This analysis of the health data organization
phenomenon attempted to meet the conditions of the third
category of literature, offering a conceptual framework
and case studies. For example, the top-down or bottom-up
proposition can be understood in at least three
27


implementation and innovation contexts: federalist,
sociopolitical, and organizational. This investigation
matched these conceptualizations of top-down and bottom-
up against specific cases, an under-represented area
according to Alexander (1982).
The investigation explored the relationship of the
private sector to the public sector. This exploration
was intended to better explain the formulation,
enactment, and implementation of a new policy which was
understood in some cases to have originated in the
private sector and crossed to the public sector. From
this exploration it was expected that the reasons for the
private sector turning to the public sector will be made
more explicit. This could provide a concrete policy map
for private-public initiatives, if not private-public
partnerships. What was the nature of the private-public
relationship in the health data organization initiative?
Was it like the partnerships described by Osborne? Were
multiple levels of government involved in the health data
organization strategy? Such an exploration may also
highlight the policy pitfalls and implementation
difficulties, as well as successes in this important
relationship.
In addition the private-public relationship should
ferret out questions of partnership and conflict within
interorganizational networks. The relationship of
28


interorganizational networks to innovation diffusion has
been noted and discussed, but has lacked research
(Kimberly, 1981). The case studies were developed to
sound out these relationships and add to this literature.
The private-public relationship should, in turn,
assist in the explanation of the role that cost played in
the adoption of the health data organization strategy.
Cost is an explicit concern in innovation research.
Adoption of a new policy may be reflective of wealth
(Downs & Mohr, 1979) or tight budgets (Wolman, 1986). Or
cost may be reflective of the research process or the
implementation process (Downs Sc Mohr, 1979). The
question of cost also may explain various interest group
behavior and the states' willingness to innovate. As
noted, Deutsch (1985) described interest group activity
based on winners and losers in the face of innovation.
Was there evidence of this type of activity when the
health data organizations were being enacted? Was the
health data organization strategy adopted because of
state fiscal austerity (Wolman, 1986)? And, if so, was
it adopted because of the sincere belief that the
strategy would be effective or for symbolic purposes to
signal to the public that 'something' was being done?
(Edelman, 1971; Rogers & Shoemaker, 1971). Or was the
strategy adopted because it was perceived as both
effective and low cost?
29


And as a final contribution, what can the
exploration of the health data organization phenomenon
through these three propositions reveal about the health
data organization phenomenon and state policy innovation
Thesis Arrangement
Chapters 2 through 5 are arranged in the following
manner: Chapter 2 explains and describes the relevant
literature on state policy-making, implementation, and
innovation theory as they relate to the propositions of
this research. The literature review of innovation
theory includes public and private theory and how these
theories stack up against the various types of
innovation.
Chapter 3 describes the methodology of the
investigation. The design of the investigation is
presented along with the data collection techniques and
analytical techniques. The chapter concludes with the
limitations of the investigation.
Chapter 4 presents the results of the investigation
Chapter 4 contains the in-depth case studies of Colorado
Missouri, Oregon, and Pennsylvania. The results of the
patterning analysis among the in-depth case studies are
explained. The results of the national survey of health
data organizations are displayed and the patterning
analysis between the in-depth case studies and the
30


national survey results are discussed. Chapter 4
concludes by differentiating the health data
organizations based on the three research propositions.
Chapter 5 opens with a summary and explanation of
the findings relative to the various types of health data
organizations. Chapter 5 discusses the relationship
between health data organizations and the innovation
theory reviewed in Chapter 2. Chapter 5 concludes with
some suggestions for future research and speculation on
the future of the health data organizations.
31


CHAPTER 2
LITERATURE REVIEW
The literature review has been organized around the
research propositions of this investigation: top-down or
bottom-up, private sector-public sector relationship, and
cost. The literature that supported and illuminated
these propositions is complex, elaborate, and sometimes
contradictory. An effort was made to focus on
representative scholars who are exemplars of a given
position, such as Sabatier (1986) representing the top-
downers and Hjern (1982) representing the bottom-uppers.
The top-down or bottom-up literature was introduced
with a discussion of state-level, statutory policy-
making. It included several definitions of how top-down
or bottom-up might be understood from various
perspectives. The bottom-up literature emphasized the
bottom-up implementation research. The private sector-
public sector relationship literature examined the
private-public partnership theme. The cost literature
examined issues related to the costs of innovation, the
attractiveness of the policy, interest party activity,
and the relative willingness of the states to invent or
adopt innovations. The innovation literature was
reviewed relative to foregoing literature.


State Policy-Making
Elazar (1972) in his American federalism: A view
from the States developed a typology for the political
cultures in the American states. Elazar held that the
national political culture is a combination of three
fundamental political cultures. He contended that these
three political cultures are mixed and occasionally
overlap in the same state; nevertheless, he argued that
dominant state cultures could be identified. He labeled
the three cultures individualistic, moralistic, and
traditionalistic.
The individualistic culture, according to Elazar,
interprets the appropriate mechanism of democratic
governance as that of the market place. It does not
accept regulation or intervention as appropriate
mechanisms of governance. In this definition of
governance, government is less concerned with doing good.
In a individualistic culture, governmental action is
limited to establishing and maintaining a level playing
field in the market place. Elazar (1972) emphasized that
in this individualistic culture with its de-emphasized
role for intervention, government is viewed as just
another kind of business to go into, simply a way of
making a living. Consequently, political life within the
33


individualistic culture is determined more by personal
relationships than by political parties.
In contrast to the individualistic culture, the
moralistic culture has at the center of its concern, the
commonwealth. In the moralistic culture, the tools and
power of government are intended to search for the means
to a better society and to establish that better society.
Individualism in the moralistic culture is muted by its
overriding concern for the public good. Governmental
intervention for the public good is encouraged. Public
service in this culture, then, is highly valued. Unlike
political life in the individualistic culture Elazar
argued, public life can be practiced even at the expense
of personal relationships. Political parties in
moralistic cultures primarily serve the purpose of
identifying and offering candidates.
The traditionalistic culture maintains a mixed view
of the market place. The traditionalistic culture looks
more to established hierarchy and prominent families for
governance. As a consequence, political parties hold
little importance for individuals in the traditionalistic
culture. Traditionalistic political cultures, according
to Elazar are suspicious of the rationalizing efforts of
government because they threaten to interfere with the
established order; consequently, Elazar deemed
34


traditionalistic political cultures to be inherently
ant i-bureaucrati c.
To Elazar, these different political cultures were
likely to affect state-level innovation and innovation
diffusion. Elazar maintained, drawing on Walker's (1969)
article "The diffusion of innovation among the American
states," that the moralistic cultures were those that
would face new problems first. The individualistic and
traditionalistic political cultures would lag behind the
moralistic culture on innovation. Elazar did not
explicitly state how far behind. However, the subject of
policy and implementation research has largely been
legislation not political cultures.
The top-down implementation research literature has
been especially concerned with the legislative event
(Sabatier, 1986). Dresang and Gosling (1989) identified
seven state-level policy making processes:
(1) the traditional single-bill legislative process;
(2) the budgetary process, as a distinct variant of
the former; (3) the formal administrative rule-
making process; (4) the informal process of
administrative determination, outside of formal rule
making; (5) the process of regulatory commission
review and decision; (6) the process of judicial
decision making; and (7) the processes of direct
35


citizen democracy, including the initiative
referendum, and recall, (pp. 147-148)
Dresang and Gosling (1989) noted that there are
similarities among these legislative processes, but they
also emphasized that each type of legislative procedure
signified a substantive political difference. Dresang
and Gosling highlighted three legislative processes as
being the most important. Two of these processes were
characterized by Alan Rosenthal (cited in Dresang &
Gosling, 1989) as the legislative assembly line and the
executive-legislative policy-making process. The third
was identified as the budgetary process.
The difference in these three processes is how they
produce legislation and their political saliency. The
legislative assembly line in Rosenthal's view pumps out
bills rapidly. The expectation of passage varies
considerably for bills produced in this manner. The
executive-legislative policy-making process is much more
deliberate according to Dresang and Gosling (1989). As a
result, the political ante is generally much higher. The
ante is increased as government participants and
important interest groups get drawn into the enactment
process.
Dresang and Gosling (1989) observed that the
budgetary process is often dominated by gubernatorial
influence. Consequently, the budgetary process can
36


sometimes overshadow the executive-legislative policy-
making process. They argued that this gubernatorial
domination can reduce the legislative influence by
limiting the legislature to its appropriation committees.
Dresang and Gosling (1989) concluded that the executive-
legislative policy-making process and the budgetary
process are the most substantive policy-making processes
at the state level.
The Top-down Literature
The top-down, proposition holds relevance from many
different perspectives. Socio-politically, the top-down,
question holds relevance for question posed by Dahl
(1961) in his classic work Who Governs? Dahl (1961)
raised the issue of whether New Haven was predominantly
governed by elites or elite groups, whether New Haven was
governed by heterogeneous groups, or pluralism. Dahl
sided with the pluralistic interpretation, but other
researchers researching other sites who admittedly used
different methodologies, did not (Polsby, 1980). While
this controversy remains unresolved, and possibly
unresolvable (Polsby, 1980), both results offer viable
ways to analyze the governance process.
The top-down perspective can also be interpreted
within the American federalist system. In this instance,
top-down refers to action taken by the federal government
37


that requires state or local governments to enact or
implement federal mandates at the state or local level
(Stone, 1991). Top-down can also imply state and local
compliance with administrative rules and regulations for
existing programs (Berman, 1991).
However, most implementation research until the
early 1980s, excepting Lipsky in 1971 and Berman in 1978
(cited in Sabatier, 1986), dealt with statutes. This
research has been referred to as top-down. This top-down
perspective emphasizes a "policy decision and focuses on
the extent to which its objectives are attained over time
and why" (Sabatier, 1986, p. 32). As a consequence, a
equally valid argument can be made that statutory law,
according to implementation research, represents a top-
down policy initiation at any level of government.
Although, it is worth reiterating that even top-down
state and local policy has been understudied (Polsby,
1984). The policy innovation literature in particular
reflected this top-down bias. This literature was more
fully discussed under the policy innovation literature.
Top-down can also be interpreted at the
organizational level.5 At the organizational level the
5This thesis presents implementation literature that
assumes similarities between policy implementation and
organizational implementation. A close examination of
this parallel is outside the scope of this investigation.
Crozier (1964) in his classic work The bureaucratic
phenomenon was one of the first to explore the
38


top-down perspective has taken on a strongly normative
tone (Elmore, 1978). Elmore (1978) identified this top-
down implementation approach as "systems management" (p.
191). The systems management approach implies that top
management sets the organizational agenda, communicates
it to the lower ranks, and closely monitors the
implementation of the agenda. The organization is
impacted by the environment, but "is always goal-directed
and value-maximizing" (p. 191). The top-down approach in
organizations has both an implementation component, such
as implementing strategic plans, strategic management,
and/or implementation of research and development
policies (Hart, 1992). These organizational processes
are closely linked to organizational efforts to innovate.
Nevertheless, there is ample controversy among
organizational theorists about how organizations can best
go about planning and/or encouraging innovation (Drucker,
1985; Kanter, 1984; Loveridge & Pitt, 1990; Pearson,
similarities between organizations and their larger
social context, describing organizations as being
isomorphic of the society. Elmore (1978), commenting on
social program models, wrote that implementation could be
viewed as process of conflict and bargaining. Bacharach
and Lawler (1980) in their work Power and politics in
organizations dealt explicitly with this dynamic. They
argued that the sociological literature has failed to
develop a theory of organizational action. Their theory
of action meant that organizations were to be viewed as
micro-political systems. Bacharach and Lawler's view of
organizational action centered on work groups, interest
groups, and coalitions.
39


1988; Quinn, 1985; Schrage, 1989). Efforts have also
been made to distinguish between private sector
implementation and innovation and public sector
implementation and innovation (Altshuler & Zegans, 1990;
Behn, 1988, 1991; Elmore, 1991; Golden, 1990).
The Bottom-up Literature
The bottom-up literature oppositely mirrors the
top-down interpretations. Socio-politically, bottom-up
implies grass-root, pluralistic efforts to affect the
process of governance. Bottom-up within the federalist
system implies state and local governments initiating
ideas, programs, or statutes that are adopted by the
federal government. For example, Rovner (1988) reported
the recent welfare reform effort as "The issue that
boiled up from the states to capitol hill" (p. 17).
The bottom-up vantage in implementation research
emphasizes "strategic initiates from the sector, from
street level bureaucrats or local implementing officials,
and from other policy subsystems" (Sabatier, 1986, p.
32). The bottom-up approach according to Sabatier also
identified:
contacts as a vehicle for developing a network
technique to identify the local, regional, and
national actors involved in the planning, financing,
and execution of the relevant governmental and non-
40


governmental programs. This provides a mechanism
for moving from street level bureaucrats (the
'bottom') up to the 'top' policy-makers in both the
public and private sectors Hjern et al., 1978; Hjern
and Porter, 1981; Hjern and Hull, 1985.
The bottom-up interpretation for organizations has
the same components as bottom-up for policy. In fact,
Hjern and Hull (1982) cited Simon's organizational
analysis of the impact of three interest groups,
customers, employees entrepreneurs on organizational
efficiency as providing a policy research model. They
commented:
The importance of Simon's theory is that it opens up
the way to adopting an organizational perspective in
policy research which fits the implementation
research agenda. This perspective allows the
researcher to incorporate into his research design
the institutions of government and parliament as two
actors among otherswhose goals, strategies and
resources have then to be researched co-equally with
those of other actors in the context of the policy
being analysed. (p. 108)
Elmore (1978) offered two organizational models in this
category. The first model is "implementation as a
bureaucratic process" (p. 1991). This process traces the
effects of first-level bureaucrats on policy
41


implementation. The second model is the organizational
development model. This model calls for the organization
to include individuals in the decision making process: a
process of mutual accommodation between the policy makers
and the implementers. Other organizational theorists
have referred to this as a 'generative' perspective
(Hart, 1992; Kanter, 1984) as it pertains to
organizations. In this instance, strategic plans,
strategic management, and innovation are generated from
the lower ranks, receive support (implementation), and
then flow back down the organization and/or across the
organization to gather resources for implementation or
realization. These theories were reviewed in-depth in
the innovation section in this chapter.
The Top-down. Bottom-up Synthesis
In the policy implementation literature, there have
been at least three principal efforts to combine the
elements of the top-down, bottom-up approach. Sabatier
(1986) cited Knoepfel and Weidner who in 1982 produced a
complicated, "massive" (p. 37) research project. The
full results of their research are in German only. Two
works by Knoepfel (1981, a & b ) and one by Knoepfel and
Weidner (1982) available in English have been used here
to present their point of view. Sabatier also cited
42


Elmore's (1985)6 effort to combine his notions of
'backward mapping' and 'forward mapping' toward an
integrated theory. The third effort was made by Sabatier
(1986).
Knoepfel and Weidner (1982) noted that there are
structural concerns that affect the policy implementation
process as well as the processes that surround the
implementation process in their study of air quality
control programs in Europe. First, they described the
program "core" (p. 104) as it relates to a regulatory
program. The program core contains "some set of
performance standards which translate the general goals
contained in laws or governmental programs into more
specific guidelines for action" (p. 104). Secondly, they
identified the "inner program shell" (p. 104). The
"inner program shell" consists of the decisions made
relative to how individual cases would be processed.
Thirdly, since regulatory policies require "an
organizational and instrumental foundation" (p. 104),
this leads to the development of an "outer program shell"
(p. 104), according to Knoepfel and Weidner (1982). The
"outer program shell" was crucial to the interpretation
6The dating of this work depends on whether the
conference paper is used (1983) or the date of its serial
publication (1985). Sabatier used the serial publication
date. The bibliography for this thesis reflects the 1983
conference.
43


of variation in implementation to Knoepfel and Weidner.
They argued that it is not so much the control effort
that varied the program, but the delegation of
responsibility to different levels of administration.
They identified three levels of controls:
...a full national program, where the decisions
regarding practically all control dimensions are
made at the national level; a partial national
program, where the national level regulates only the
inner and outer program shells; and a so-called rump
program, where national decisions deal with the
outer program shells alone, (p. 105)
How then is implementation affected? The threshold
guidelines for administrative intervention or action can
be altered, depending on where they are set. Knoepfel
and Weidner observed that in the case of air pollution
standards or the "inner program shell," the number of
high pollution days can depend on the perodicity of
readings. The implementation structure or "outer program
shell" is affected when interested parties determine that
they are more effective at a local, regional, or national
level and lobby to place the enforcement at the level
that most benefits them. Since, for Knoepfel and
Weidner, none of these elements are independent of each
other, and all stand to strengthen or weaken a program,
they concluded that "...implementation research will have
44


to be extended to an integrated analysis of policy that
joins both the implementation and program formulation
process" (p. 107).
Elmore (1983) argued that 'forward mapping' or the
top-down approach gives only a partial view of the policy
analysis process. Elmore (1983) asserted that
"specifying alternatives and assessing their expected
effects is only part of the analytic problem" (p. 7).
According to Elmore (1983), the nature of policy analysis
requires that the logic be reversed:
Reversing the logic has two effects on our analysis.
First, it provides insurance against unanticipated
effects, so that if things start to go wrong in the
implementation process we have an intelligent
response. But second, and more importantly, it
changes the content of the policy options we
recommend.
Policy analysts use terms like 'iterative' to
describe this process of reasoning back and forth
between first choices and expected effects.
Regardless, of what you call it, reversible logic
carries an important message for both analysts and
policymakers: Specifying the expected relationship
between implements and their effects is only half
the analytic processthe forward mapping half, if
you will. The other half consists starting with the
45


choices confronting people at the 'outcome' end and
playing the consequences of those choices back
through the sequence of decisions to first choices
the backward mapping half, if you will. (p. 8)
Elmore concluded that effective policy-making requires
both 'forward mapping' and 'backward mapping.' The
exclusion of 'backward mapping' risks overestimating the
control of policy-makers and sacrifices effective policy
alternatives that are understood to provide the desired
outcome.
Sabatier (1986) in his effort to integrate the top-
down literature with the bottom-up literature conceded
that the correct starting place is:
a policy problem or subsystemrather than a law or
other policy decisionand then examine the
strategies employed by relevant actors in both the
public and private sectors at various levels of
government as they attempt to deal with the issue
consistent with their objectives, (pp. 38-39)
Sabatier then concluded that the concerns of the 'top-
downers' should be integrated. This requires that the
concerns of the 'bottom-uppers' receive attention, such
as:
(1) the effects of socio-economic (and other)
changes external to the policy network/subsystem on
actor's resources and strategies; (2) the attempts
46


by various actors to manipulate the legal attributes
of governmental programs in order to achieve their
objectives over time; and (3) actors' efforts to
improve their understanding of the magnitude and
factors affecting the problem-as well as the impacts
of various policy instruments-as they learn from
experience (Sabatier, 1986, p. 39).
This synthesis offered by Elmore (1983) and Sabatier
(1986) provided a context not only for analyzing policy,
but also provided a context for analyzing organizational
networks and action groups. If the policy analyst traces
a policy formulation from a group of interested policy
actors to the stage of enactment, and through the
implementation, which, according to Browne and Wildavsky
(1983/1984), also changes the enacted policy, the policy
analyst should be rewarded with knowledge of how the
organizational networks interacted. This dissertation
focused particularly on the private-public relationships
expressed in this complex web of interaction.
The Private-Public Relationship
Increased interaction between the private and public
sectors has been dated from the postwar era (Hjern,
1982). Hjern (1982) noted that this increased
interaction has "had an impact upon the ways in which
politics and administration are linked" (p. 302). Hjern
47


(1982) commented that this increased interaction includes
the "formulation, aggregation, and execution of policy
interests" (p. 302).
There are at least two ways to understand the
private-public relationship. One way, and perhaps the
most obvious, is to personify all private organizations
as the private sector and to personify all public
organizations as the public sector. A second way, and
subtler approach, is to distinguish among levels of
government and to recognize a "private-public"
relationship between different levels of government and
all privately structured organizations including private
non-profit organizations. As Hjern (1982) commented:
To understand how politics and administration are
linked now requires an understanding of more than
just how clause is added to clause. It also takes
an understanding of how organization is linked to
organization, (p. 307)
These "linked" organizations have been designated as
clusters. Clusters can be issue or policy specific and
loosely or tightly coupled (Meyer & Scott, 1983; Perrow,
1986). Loosely coupled systems or clusters are marked by
sporadic interaction and alternation of organizations
and/or individuals. Tightly coupled organizations, it is
postulated (Perrow, 1986), share many points of
48


interaction. Perrow (1986) characterized the health care
system as an example of a tightly coupled system.
Meyer and Scott (1983) also described
interorganizational linkages as "centralized" and/or
"formalized" (p. 165). They observed that
interorganizational sets can display "fragmentation" or
"federalization" (p. 165). Fragmentation in the
interorganizational set refers to "multiple uncoordinated
linkages between units at differing levels in the field"
(p. 165). Federalization refers to "the degree to which
there exist multiple uncoordinated linkages from two or
more 'higher' levels impinging on a third level, for
example, national and state links of the same type to a
county unit" (p. 165).
Perrow (1986) observed that "Each system has its own
network of public and private organizations, but the two
systems are only loosely joined to each other; there are
different agencies, different politicians, and different
commercial groups in each cluster" (p. 202). Meyer and
Scott (1983) noted similarly that each organizational
field may vary "in terms of whether they contain a
bureaucratic or a professional sovereign" (p. 167).
Aldrich and Whetten (1981) emphasized that these
systems should have some conceptual limitations. Aldrich
and Whetten (1981) identified three types of such
systems: (1) organization sets, (2) organization action-
49


sets, and (3) organization networks.7 Organization set
theory is similar to Merton's concept of the individual
role set according to Evan (cited in Aldrich & Whetten,
1981): "that complement of role relationships which
persons have by virtue of occupying a particular social
status" (p. 386). Aldrich and Whetten (1981) defined an
action set as "a group of organizations that have formed
a temporary alliance for a limited purpose" (p. 387).
Refining their model further, Aldrich and Whetten defined
a network as the "totality of all the units connected by
a certain type of relationship" (p. 387).
Meyer and Scott (1983) identified many different
characteristics of organizational networks. They argued
that if single organizations can be distinguished in the
field, then such notions as "density," variety of "types"
of organizations, functional "specificity" (p. 165) can
be developed. They further differentiated the
7The notion of organizational networks is one of two
dominant theories of interorganizational relations.
Aldrich's (1979) Organization and Environment has been
described as an effort at "a macroinstitutional theory
for analyzing organizational transactions in the
political economy" (Bacharach & Lawler, 1980, p. 4).
Perrow (1986) called this the "population-ecology" (p.
208) model. The model ostensibly would have more to
offer the private sector than the public sector as it
emphasizes evolutionary competition and niche analysis,
but Scott and Meyer (1983) found the model gives little
attention to connections in organizational populations.
Similarly, Freeman and Barley (1990) found this model
inadequate for describing the complexity of
interorganizational relationships.
50


organizational network structure from its "elements."
The elements include "cultural elements," "historical
aspects," societal context," "network elements," and
"world system context" (pp. 164-172). Scott and Meyer
maintain, however, that it is not just the organizational
network form that is important. They emphasized that the
content of organizational networks is important. Content
includes such notions as "resource" exchange,
"interpenetrations," types of "decisions," and what
types of "interlevel controls" are used.
Aldrich and Whetten (1981) argued that networks
evolve out of an adaptive response to the environment.
They conceded that their model of interorganizational
networks was hypothetical and speculative, not empirical,
but they also argued that it is "a very powerful device
for synthesizing the research and theorizing that does
exist." They perceived governmental intervention as one
of the more disruptive forces on networks, local networks
in particular. They argued that governmental
intervention often results in hierarchical controls that
confine local action. And while Aldrich and Whetten
(1981) apply their model to many different "research
foci" (p. 394), the most important of these is innovation
for this investigation.
51


Innovation diffuses much more readily along
interorganizational networks (Aldrich & Whetten, 1981;
Roessner, 1979). Elkin (1983) commented that:
organizational members can also consider innovations
through an external political process, i.e., through
interaction with environmental elements,
particularly other organizations, (p.372)
Glaser (1981) found that innovations have a greater
"durability" (p. 181) when the organizational staff has
the opportunity to participate "in a network of adopters
and to observe other applications of the innovation" (p.
181). Aldrich and Whetten (1981) posited that "if
information flows relatively freely throughout a
population or organizations, selective diffusion of
innovations among organizations is an important selection
mechanism for changes that do not involve eliminating
entire organizations" (p. 394). Roessner (1979)
commented that:
Professional associations appear to be a key
component of the communication networks that operate
horizontally among similarly-placed state and local
officials in different cities and states, and
vertically among functional area counterparts at the
local, state, regional, and federal levels, (p. 196)
Roessner also emphasized that professional organizations
provide networks for individuals or what Yin (cited in
52


Roessner, 1979) called "natural points of entry into the
incentive system."
Warren, Rose, and Bergunder (1974) reported their
experience with innovation and the interorganizational
field as it related to the model cities program. They
focused less on the notion of innovation than on the
interorganizational state of affairs that tends to be
associated with various innovations. They developed
three categories of innovation: "primary, secondary, and
gross" (p. 86). A primary innovation is a radically
different "analytical paradigm" (p. 86) that threatened
the basis of an organization. Secondary and gross
innovations were more modest program adjustments that
"were hardly distinguishable from ongoing, routinized
operations of the organizations involved" (p. 103).
Out of a total of 606 innovations, they found but 3
primary innovations. This led them to conclude that
"...it is difficult to see how serious innovation can
arise from strategies that accept the existing
institutionalized thought structure" (p. 103). Perhaps
more critically to the understanding of organizational
networks is their observation that when such radically
different ideas come forth, they are met with what the
researchers described as a "reinventing-blunting-
repelling sequence" (p. 103). The "reinventing-blunting-
repelling sequence" is akin to the negotiating and
53


bargaining process that attends implementation. They
concluded that as long as community decisions are made
within the institutionalized thought structure,
organizations would have to be content with minor changes
in programs and relationships. It is, perhaps, with
these limitations among public organizational sets that
calls for private-public partnerships have taken on a
sense of urgency as a means to increasing the likelihood
of "primary" innovation.
There have been many recent calls for a greater
private-public partnership between government and
business from the federal level to the local level
(Osborne, 1988; Osborne & Gaebler, 1992; Reich, 1984).
Osborne (1988) and Osborne and Gaebler (1992) presented
convincing evidence that the private-public partnership
increased in the 1980s. The private-public partnership
increased through newly formed organizational networks
comprised of private and public organizations. This
increase was in response to different conditions in the
American states. It was Osborne's (1988) argument that
these new relationships led to innovative programs and
services. He argued that this could only be accomplished
by eschewing common assumptions about what the business
of government is and what its relationship to the private
sector should be.
54


As noted, the precipitating conditions for these
innovations8 varied among the states, although all were
described as being led by a governor or by gubernatorial
staff. These innovations ranged from education in
Arkansas to housing the poor in New York. The
innovations occurred in different arenas, but a common
thread for all of them was a lack of resources to provide
services or a need to stimulate economic growth.
Governments are keen on the issue of cost, and so is the
innovation literature, but in a perplexing way.
The Cost Issue
The issue of cost has been discussed and researched
from at least two major perspectives: the amount of
resources of the inventing or adopting organization and
the cost as an attribute of the innovation itself (Cyert
& March, 1963). Downs and Mohr (1976) observed that cost
is a primary attribute ("primary attribute" is discussed
more fully in the innovation typology section of
innovations):
...the determinants of high-cost innovations would
seem to be markedly different from those of low-cost
Osborne did not choose to define innovation. He
tended to speak in terms of a "new paradigm." His
underlying assumption seemed to be that a new way of
doing things, especially by organizations that crossed
sectors, constituted an innovation.
55


innovations. Wealth or resources would clearly
predict the former differently from the latter.
Resource levels might explain much of the variation
in the adoption of elaborate EDP systems by
corporations but be relatively unrelated to the same
corporations' adoption of new supervisory techniques
or job enrichment programs, (p. 702)
It has been asserted that the cost of a particular
innovation has many different dynamics, including
feasibility, adoptability, and implementation, but this
has been more a matter of observation than empirical
findings. Downs and Mohr (1979) split costs into two
categories: (1) decision costs, or the costs associated
with "arriving at a decision" (p. 397) and (2)
implementation costs, or the costs of implementing the
innovation to what Downs and Mohr called the "fair-trial
point" (p. 397). The "fair-trial point" is
differentiated from the "token-adoption point" (p. 387).
The "fair trial point is the extent of use at which the
adopter has enough experience with the innovation to
assess its costs and benefits accurately" (p. 387). The
"token-adoption point" is met when an adoption occurs on
such a small scale as to encounter no risk at all. This
reasoning led them to conclude that "reversibility" (p.
390) is an important determinant of adoptability:
56


...because a non-reversible innovation would have
involved an organization in excessive costs if the
organization were to change its mind shortly after
adopting, (p. 390)
Downs and Mohr (1979) argued further that costs and
benefits can be viewed as "properties of the choice
situation" (p. 397) so that perceptions of costs and
benefits can and do vary widely from one organization to
another.
Perry and Danziger (1980) found in their study of
computer innovations that decreased uncertainty of the
costs of an innovation increased the probability of
adoption. Nelson and Sieber in 1976 (cited in Kimberly,
1981) commented that "there is some evidence that the
publicity value of an innovation may greatly enhance the
probability of its being adopted irrespective of
assessments of quality or of cost" (p. 96). Wolman
(1986) reviewed a number of recent efforts to assess the
role of cost in innovation. He noted that Cyert and
March (1963) argued that organizations that are
financially strapped are more likely to innovate.
However, Cyert and March (1963) also argued that
slack organizational resources act as an enhancement to
innovation. Cyert and March (1963) speculated that this
apparent contradiction was actually a reflection of
different types of innovation. Cyert and March suggested
57


that slack induced innovations occur when the innovation
has the support of a subunit and does not face scarcity.
Gap based innovations occur when the innovation can be
directly linked to problem solving. Zammuto, according
to Wolman (1986), came up with a similar finding.
Organizations that innovate because of a performance gap
are more likely to choose technical innovations and
production-process innovations.
Nevertheless, there has not been a great deal of
research to support either position, according to Wolman
(1986). G. Downs found initially a relationship between
organizational liquidity and innovation, but this finding
was arrived at by only considering the total budget of
the agency, rather than the demand for use. In another
related study, G. Downs did not have the same finding,
but Mohr using the size of the organization's budget
duplicated G. Downs' initial findings. Wolman (1986)
concluded his article with a call first of all for a
clear definition of innovation. Wolman (1986) also
called for clearer categorization of innovations and
organizations. He argued that "The determinants of
technological innovations are not likely to be the same
as those of policy or service innovations" (p. 179).
Deutsch (1985) took another tack with the cost
issue. He argued broadly that innovations also involve
costs and benefits, but he asserted that neither costs
58


nor benefits are evenly distributed. Deutsch (1985)
argued a priori that this uneven distribution leads to
expressions of power relationships:
Typically, particularly in private enterprise
economics, innovations occur also, as in investment,
by shifting costs to groups weaker in the market or
weaker in the political arena, (p. 25)
Deutsch's emphasis on the self-interested aspect of
innovation led him to conclude that:
Groups that mainly receive benefits will learn to
accept and foster an innovation which other groups
will learn to fear and oppose, (p. 25)
The opposition would then consist of "coalition formation
or direct action to prevent it" (p. 22). He saw in this
opposition the potential to block an innovation. He
determined that the terms of cost benefit would require
that costs for not innovating need to be included as much
as the cost of innovating.
Deutsch (1985) saw a particular advantage to the
cost benefit approach to innovation with regard to the
private-public relationship. He asserted that
interesting research could be conducted showing the
innovation curves for the private sector and public
sector. He hoped that this would answer such questions
as:
59


What are the patterns of innovation in the public
sector? Are they rising more steeply for an
innovation spreading within one and the same
bureaucracy, but slowly and less smoothly from one
bureaucracy to another? (p. 23)
And finally, he argued that the interplay between the
private and public sector should be analyzed. He gave
the example of the automobile, innovated by the private
sector, but which gave rise to the large system of public
highways. It is curious that Deutsch did not invert the
relationship to ask what public efforts have stimulated
private productivity.
Innovation Typologies and Research
The many different definitions of innovation were
presented in Chapter 1. It has been noted that these
varying definitions have contributed to the varying
results of innovation research (Downs Sc Mohr, 1976;
Wolman, 1986). It has also been argued that the very
attribute of innovation or the notion of innovativeness
has contributed to the instability of research results
(Downs & Mohr, 1976; Elkin, 1983). Innovativeness and
adoptability were more the concern of the political
science or policy outcomes literature on innovation.
Wolman (1986) referred to this literature as an
assessment of the "socio-economic" (p. 172) factors. For
60


purposes of this review, this literature has been
referred to as the socio-economic approach.
Zaltman and Holbeck (1973) identified five types of
innovation: (1) product or service innovation, as an
organizational product; (2) production process
innovations, relating to the production of goods and
services; (3) organizational structure innovations; (4)
people innovations, changes in the way people interact;
and, (5) policy innovations, or changes in organizational
strategies. Warren, Rose, and Bergunder (1974) divided
innovations into three categories: primary, secondary,
and gross as specified on page 21.
Downs and Mohr (1979) chose to differentiate
innovations by primary and secondary attributes. The
identification of primary and secondary attributes
according to Downs and Mohr permits rejection of a
t
unitary theory of innovation and offers the opportunity
to develop theories that explain why distinct types of
innovation are adopted. To illustrate the use of primary
and secondary attributes, Downs and Mohr (1979) chose the
attribute of cost. They argued that if cost, high or
low, is a primary attribute, then any organization, rich
or poor, will "place the innovation in the same cell of
the typology" (p. 702). Contrarily, when a typology is
based on secondary attributes, the "classification of the
innovation depends on the organization that is
61


contemplating its adoption" (p. 702). The research
implications of this observation by Downs and Mohr (1979)
are discussed at the end of this chapter.
State-Level Policy Innovation Research
Historically, a number of researchers have taken a
socio-economic approach that has largely been concerned
with policy outputs, or the legislative top-down
approach. Their research emphasized such factors as
interparty conflict, and degree of district
malapportionment; other researchers added percentage of
urbanization of population, degree of industrialization,
education of population, and per capital income (Fenton &
Chamberlayne, 1969). The results varied. Fenton &
Chamberlayne (1969) concluded that the findings among
these research efforts demonstrated that political
competition and socio-economic conditions affect policy
outputs.
Walker's "The Diffusion of Innovations among the
American states" (1969) continued the policy outcome
analysis by examining similar conditions, but he sought
to weigh how these conditions relate to innovations. In
other words, he posed the question of which states are
more likely to innovate and can regional patterns be
detected? Walker emphasis on innovation and diffusion
set his work apart from the strictly policy output
62


research. Walker did not reach any definitive
conclusions. Walker (1969) saw relative 'innovativeness'
and adoptability as a quality of the states, not of the
innovation itself. However, he did not find regional
leaders among the states. He concluded that states that
adopted innovations more readily tended to have a more
urban population and a higher per capita income. Most
importantly, Walker commented that the many
intergovernmental contacts between entities led to
complex patterns that were difficult to uncover. Walker
(1969) did not develop this interorganizational theme
until later in 1981. By 1981, Walker was writing about
"policy communities" (p. 80) in which policy
professionals would be motivated to pursue innovation in
order to enhance their careers, abandoning the policy
output approach. Hjern and Hull (1982), citing Scharpf
1977, commented "What ultimately discredited policy
output studies was the lack of any explicit
organizational perspective" (p. 107). Nevertheless, this
stream of literature continued to develop along the same
lines.
Savage (1978) reported that Gray (cited in Savage,
1978) had questioned not only the "quantitative measures
devised by Walker (1969), but also the very validity of
the construct itself" (pp. 212-213). Using a time series
approach, Savage demonstrated that innovativeness as a
63


state trait varies over time among most American states.
Savage concluded that nearly half the states exhibit a
constant adoption across different types of innovative
policies. Hence, Savage rejected Gray's argument. He
maintained that 'innovativeness' is a trait of the
states, but that it may be issue and time sensitive. For
Savage, this evidence underscored the role of the states
as experimental laboratories.
Rosenbaum (1976) conducted a study of the diffusion
of innovative land use policies among American states.
Rosenbaum was the first of the policy outcome theorists
to insist on a distinction between diffusion and
adoption. Citing Walker (1969), Rosenbaum (1976)
observed two patterns in adoption: A slower "linear" (p.
6) diffusion rate, that increases evenly over time, and a
"non-linear" (p. 6) adoption pattern that showed clusters
of increasing adoption among the states. Rosenbaum
(1976) related the former pattern to regional adoption
and the latter pattern to more of a national adoption
pattern. States within the same region tended to adopt
innovations in a cluster pattern over a relatively short
time according to Rosenbaum.
Rosenbaum (1976) found in his study of the diffusion
of innovative land use that the findings of Walker did
not hold true. He observed that only in California,
which led the West, and Virginia and Florida, which led
64


the South, did more "populous, affluent, urbanized, and
industrialized states" (p. 79) set policies for the
region. He attributed this departure from Walker's
findings to the relative urgency of land use legislation
for particular states. He did note that he found the
usual pattern of regional adoption of innovation.
Rosenbaum's work built on Walker's (1969) analysis.
It went further to differentiate between the innovation
stages of adoption and invention. It expanded the
understanding of the diffusion of legislative acts, but
the work shed little light on why and how innovations
come about.
Federal Policy Innovation
Nelson Polsby (1984) in his Political Innovation in
America tried in part to answer these and related
questions. Polsby used case studies of federal
legislation from which to develop his theories of
innovation. Polsby's analysis was not as slanted toward
a top-down analysis as the state-level innovation
research. Polsby emphasized policy innovation as a
response to a need, in describing this need, Polsby
provided a rich description of the context of enactment.
Polsby (1984) identified two types of innovative
policy enactment: acute and incubated. The acute
enactment was typified by a short period of gestation
65


from the idea stage to enactment. This implied that
little time or energy is expended to develop or
investigate alternatives. According to Polsby, the
accepted alternative is invented during the search
process along the lines of Simon's notion of
"satisficing" (p. 150n). Polsby further observed that
the political conflict in this case is truncated and
constrained. The justification for the new policy lacks
ideological appeal. Consequently, there is little, if
any, partisanship.
The incubated enactment was the opposite of the
acute enactment. The incubated policy faces substantial
conflict and opposition as it moves more slowly through
the legislative process. As this conflict increases, the
legislation may become more ideologically based and more
partisan.
Polsby (1984) suggested seven "dimensions" (p. 14)
of policy innovation from the cases he analyzed: (1)
Timing: the period of time between concept and statute;
(2) Specialization: degree of involvement of experts as
opposed to political actors; (3) Existence of agreement
in the decision making subculture, or degree of consensus
on the nature of the problem and its solution; (4)
Political Saliency: public support and attention; (5)
Political Conflict: referred to the difference between
acute and incubated policies; (6) Research: the question
66


of spontaneity versus technical design; (7) Staging:
whether the solution or need came first. These are the
basic dimensions that Polsby found "useful" (p. 14) to
describe policy innovations.
Polsby (1984) ended his analysis with two
observations. First, that policy innovation can be
separated from other political events. Second, he
commented that there are two distinct phases in the
innovation process. They are "invention" and "search"
(p. 173). He used the word invention to include policy
entrepreneurs, think tanks, and those who generally make
a living through developing policy alternatives. The
search phase was designated as the process through which
policy options are identified. Polsby concluded that the
study of innovation is the study of the interaction
between invention and search.
Health Policy Innovation: A Mixed View
Stone (1991) argued that health policy innovation
within the American federalist system is a top-down
process, alternatively constrained and limited in action
by the federal government. She acknowledged that the
orthodox view is that innovation "bubbles up." Stone
identified the many influences and pressures that should
have magnified the potential for innovation at the state
67


level, citing Beyle (1988), Reeves (1990), and Schneider
(1989) who saw the states as revitalized.
State Level Discretion and Innovation
Beyle (1988) argued that changes in gubernatorial
power and legislative authority in the last few decades
made the states more progressive. He emphasized that the
governors of this new found power felt that innovation
and success were synonymous. As a result, the governors,
according to Beyle, saw innovation as their major goal.
Reeves (1990) characterized the states as
"reinvigorated" (p. 83). Reeves' observation was based
on the increased emphasis on the state as a partner in
the federalist system. Reeves' treatment of the state
initiated innovations was limited however. She pointed
to specific state innovations but did not define the
concept of innovation. She felt that the states provided
"laboratories of democracy" (p. 85) for Congress. She
claimed that Congress has benefitted because of the
states' willingness to experiment to see what works and
what does not.
Schneider (1989) argued that not only are the states
re-invigorated, but they are effective initiators of
public policy. She thought this to be particularly true
of the states relative to the medicaid program.
Schneider felt that much of the states' control over
68


public policy extended from the powers of the governor.
She chose to differentiate "the formal and informal power
of the governor" (p. 909). She asserted that the
informal power of the governor was the stronger of the
two. Schneider interpreted the states as far more
influential in the federalist system than mere "conduits"
of federal health policy goals, but her examples were
narrow in their impact.
First, Schneider asserted that the states determine
"who qualifies for care as well as the type of care they
can receive" (p. 910). Second, Schneider imputed a great
deal of impact to the states' regulatory control over
health care providers. Third, she drew attention to the
control states have over funding of various health care
programs. She noted that in concert with these control
mechanisms, governors can in addition influence health
programs through appointments to health care regulatory
bodies, the issuance of public statements to draw
attention to health care issues and problems. She noted
that one or all of these efforts can be used to effect
budgetary adjustments for health care programs. Are
these instruments, or what Elmore (1983) called
implements, sufficient to counter federal power in health
policy? Stone answered no.
69


I
Federal Control of Health Policy
Stone (1991) argued in her essay that the federal
government bullied the states into innovative health care
programs through a carrot and stick approach, leaving
little initiative to the states. Though Stone's
interpretation went against the orthodoxy of the day, it
was not without anecdotal evidence. Rovner (1988)
reported thlat Republican Governor Michael Castle of
Delaware when commenting on state welfare reform stated:
It's a1 policy that actually bubbled up to the
federal level, opposed to almost any health and
social service policy in the last 50 years, which
started at the federal level and went back down. (p.
17)
Stone put into the category of federally mandated or
federally induced health policy "innovations" such
initiatives as neighborhood health programs, health
maintenance organizations and peer standards review
organizations. These would have to be uncontested
examples in that they all received their enabling
legislation at the federal level. Her argument for the
Omnibus Budget Reconciliation Act of 1981 as a federally
I
sponsored and controlled inducement for innovation seemed
i
less convincing.
Stone interpreted language in the health policy
literature that described the changes at the state level
70


in response to the Omnibus Budget Reconciliation Act of
1981 as written in a telltale "passive voice" (p. 7).
Stone argued that this amounted to the states being
granted the permission to act or implement, not the
states acting independently. She cited the "innovation"
of states picking up options that expanded eligibility to
pregnant women into this category. Among such changes as
this. Stone (1991) commented, "There was scarcely room
for anything worthy of the name innovation" (p. 7).
Two problems become evident here. One is the matter
defining innovation: a problem not unique to Stone's
argument. Second, the matter is whether program
adjustments fall into the category of innovation. For
example, Bergthold (1988) wrote an analysis of
Massachusetts' passage of Chapter 372, a prospective
payment system for all payors in Massachusetts. Chapter
372 was very much a response to the Omnibus
Reconciliation Act of 1981. Chapter 372 evolved from a
state and community political process that developed the
uniform, prospective method of reimbursement for all
categories of payors in the State of Massachusetts.
Bergthold held that this was an innovation, albeit
without defining innovation, underscoring that innovation
is in the eye of the beholder.
Stone focused much of her attention on the national
effort to create local coalitions of health care
71


insurers, providers, and health care professionals. She
determined that they were not local coalitions based on
their representation and activities. She argued that
they had patterned themselves on the Dunlop Group of Six,
had failed to create broad community representation, and
some had accepted Robert Woods Foundation money to
develop. These facts not withstanding, although
certainly there is room for argument with each of these
tenets, Stone's argument lies fallow for lack of an
understanding of the notion of innovation.
Organizational Innovation
Organizational theorists have claimed that the
innovation process in organizations is or should be one
of main functions in organizations (Ranter, 1984; Rogers
& Kim, 1985). Organizational innovation has been
researched with many of the same concerns as that of the
larger socio-political system: economic benefit and
renewal. Generally, researchers have tried to discover
why some organizations innovate while others do not, and
what makes an innovation particularly adoptable
(Kimberly, 1981; Perry & Danziger, 1980). However, a
substantial portion of the literature has been normative,
descriptive and/or prescriptive (Behn, 1991; Ranter,
1984; Osborne, 1988; Osborne & Gaebler, 1992). In
essence, the argument is that American socioeconomic
72


survival is dependent, especially at current living
standards, on innovation. If we must innovate, then, how
can we best organize to do so?
These calls for innovation include all sectors of
the economy, and they vary in the degree to which they
specify what is meant by innovation. Some differentiate
between innovation in different sectors of the economy,
others do not. Some differentiate between types of
innovation, and others do not, but they all agree on the
importance of innovation to such an extent that many
writers have observed an "innovation bias" in both the
private and public sectors of the economy (Altshuler &
Zegans, 1990; Behn, 1991; Ven De Van, 1986). Innovation
bias implies that innovation is always good for the
social system or organization. Elmore (1991) cautioned
that the "innovation bias" can obstruct efforts at hard-
nosed evaluation of the effectiveness of a given
innovation.
Public Sector Innovation
Public sector innovation research has focused on
technological innovation and managerial innovation. Two
representative studies of public sector technological
innovations were presented first. The largely
descriptive and normative managerial literature was
presented second.
73


Technological Innovation. Roessner (1979) sought to
explain the apparent lack of innovation in the public
sector. He asked if it were a problem of "demand,
supply, or infrastructure" (p. 189). He decided all
three factors had some influence. Along the way, he
acknowledged that there existed evidence that "public
organizations operate under a set of incentives different
from that of their private counterparts, and that these
differences may cause misleading conclusions concerning
the innovativeness of public organizations" (p. 191).
Roessner (1979) recommended that the public technology
policy could be improved by the federal government acting
as a capacity builder toward state and local governments
and to utilize professional networks or Yin's "natural"
points of entry.
Perry and Danziger (1980) studied the adoptability
of innovations. They stated that:
the adoptability question asks, what characteristics
seem to affect the likelihood that one innovation or
another will be adopted by an organization? (p. 462)
Their study was designed to be sensitive to the problem
of primary and secondary attributes of innovations noted
by Downs and Mohr (1976). The study tested for four
independent variables: "(1) relation of the innovation to
organizational domain, (2) integration, (3) risk, (4)
need" (p. 464) for ten computer innovations. They
74


concluded that the innovations were more adoptable when
there was:
(1) greater visibility of the innovation; (2) less
uncertainty about the cost and evaluation of the
innovation; (3) greater EDP staff competence to
implement the innovation; and (4) a higher level of
objective need for the innovation, (p. 482)
Interestingly, they noted that adoptability was not
associated with greater or lesser developed professional
infrastructures, leading them to conclude that
professional networks did not assist in communicating the
innovations.
The technological innovation literature showed a
greater sensitivity to the larger sociological framework
of the innovation literature. Some of the recent
managerial literature has looked more to the private
sector, the normative literature of Ranter (1984) for
prescriptions, and applied them to public organizations
to determine if public organizations can optimize
innovation opportunities.
Managerial Innovation. Two types of organizational
processes have been identified by Golden (1990) as common
methods for organizational innovation in the public
sector. She developed her model from a review of public
sector organizations cited by the 1986 Ford Foundation
Awards Program for Innovations in State and Local
75


Government. She identified one approach as a somewhat
rigid, single event occurrence which she called the
"policy planning" (p. 220) model. The second approach
she called the "groping along" approach after Behn's 1988
article "Groping Along."
The "policy planning" model, according to Golden,
developed from an innovative idea. The innovative idea
was then carefully translated .into statute and policy.
This "policy planning" model maintained a close adherence
to pre-established implementation parameters through
managerial control techniques. This is, of course, an
exact description of Sabatier's (1986) description of the
top-down implementation literature and Elmore (1978)
description of "systems management." The same
weaknesses and strengths of Sabatier's model can be
attributed to Golden's (1990) "policy planning" model.
Golden's (1990) "groping along" model took a
different tact. The initial policy idea was less
important. The emphasis is on a public agency's ability
to solve a difficult problem in an innovative way through
continuous adjustment to both the implementation process
and the environment. Behn (1988), originator of the
"groping along" model, explained the "groping along"
process as:
...even the best manager must grope along. He tests
different ideas and gauges their results. Then he
76


tries different combinations and permutations of the
more productive ideas. Rather than develop a
detailed strategy to be followed unswervingly, a
good manager establishes a specific directiona
very clear objectiveand then gropes his way
towards it. He knows where he is trying to go but
is not sure now to get there. So he tries numerous
things. Some things work. Some do not. Some are
partially productive and are modified to see if they
can be improved. (p. 645)
Behn's view invokes the perspective of the bottom-uppers
with the exception that it tends to concentrate on
administrative personnel instead of the full complement
of policy actors or the notion of an implementation
structure. This is reinforced by Golden's citation of
the work of Browne and Wildavsky (1983/1984). Prior to
Browne and Wildavsky, Majone and Wildavsky (1979/1984)
had reasoned that administrative discretion was not only
unavoidable but requisite to effective policy
implementation. They argued that to the extent
administrative discretion is exercised, the original
intent is changed. Consequently, the policy outcomes
could be expected to change. Browne and Wildavsky (1983/
1984) argued that implementation is both adaptive and
exploratory. The arguments of Majone and Wildavsky
(1979/1984) and Browne and Wildavsky (1983/1984) added to
77


Behn (1988) and Golden's (1990) argument that innovation
can result from the sum effect of continuous and adaptive
managerial changes. And this is in fact what Golden
concluded from her review of public sector innovations;
they mostly come about through an evolutionary process.
Similarly, Rabinovitz (1991) perceived public sector
innovation as evolving modestly. She argued that
innovation in the public sector is likely to approximate
"...a change in the balance of existing resources and
processes" (p. 4). She observed that "...the truth is
that many public leaders have very little interest in
such innovation as a disembodied phenomenon" (p. 5).
Public leaders according to Rabinovitz want concrete
answers to specific problems, not innovation nurseries.
Elmore (1991) viewed innovation in the public sector as
strengthening ties with constituencies. However, Elmore
was less impressed with the value or utility of
innovation or the incremental, evolutionary theory of
innovation. He found this approach to innovation to lack
staying power. Elmore remarked that educational
innovation tends to ebb and flow according to policy
cycles determined by the prevailing theory of the day.
Elmore argued that even though public education is "awash
in innovation" (p. 5), public education has failed in
Yin's (1979) term to "routinize" the innovations into the
larger system.
78


Elmore (1991) summarized his observations with the
three following observations: (1) Innovation as the
hallmark of a good organization has yet to be
demonstrated; (2) Organizations do not necessarily
continue to improve because of their innovations; and,
(3) Public organizations, because of their "complex core
technologies," (p. 50) do not easily change, particularly
by applying the customary techniques of public policy and
management. He concluded by reiterating that when
theories of innovation are applied to smaller segments of
an industry, it does not insure that the theory will
extend to the industry as a whole.
Elmore's analysis of the relationship of segmented
innovations to the larger industry underscores another
type of innovation bias. As much as there is a positive
bias toward the notion of innovation, there is an equal
bias that innovation especially in human services, should
have applicability in similar systems. Rabinovitz (1991)
stated that Peter Goldmark, former New York State Budget
Director, said, "Replication...is just as important a
step as innovation, but it has been little studied and is
inadequately understood" (p. 8). She added that Anthony
Shorris, former Finance Commissioner of the City of New
York, observed that there is a "vacuum" (Rabinovitz,
1991, p. 9) in knowing what other governments have tried
or are trying. There is an increasing interest in the
79


notion of replication, but not necessarily the same sense
of the word Goldmark intended.
Adaptation, Replication, or Re-invention
The notions of adaptation, replication, or re-
invention extend from the recognition that innovations
must be implemented not just decided on. The terms
adaptation and replication come from implementation
literature, the term re-invention was first hinted at by
Charters and Pellegrin in 1972 (cited in Rogers & Kim,
1985). Rogers defined re-invention as the degree to
which an innovation is changed during the implementation
process (Rogers & Kim, 1985).
More recently, Berman (1992) and De Lone (1990)
writing on the subject of public education and Glaser
(1981) writing on mental health took related but
different stances on the matter of replication. Berman
took essentially the bottom-up perspective. Berman
(1992) argued that the insistence, such as Elmore (1991)
made on the replication of programs in the larger system
or to other sites, has been misguided. Recognizing from
the bottom-up perspective that adaptation is more or less
inevitable, Berman argued it is not so much replication
that is needed, but outcome information, so that
effectiveness can be measured.
80


De Lone (1990) took more of a top-down and
prescriptive stance. He argued there are we11-documented
successes in education and that these should be very
nearly replicated throughout the country. He allowed for
some adaptation, but only those structural elements not
essential to outcomes could be allowed to adapt to local
needs. Essential elements, he argued must be replicated.
Glaser (1981) contended that on the balance of most
innovations, Berman (1992) was correct at least in regard
to "durable" (p. 183) innovations. Glaser stated
"...innovations that last over time typically change in
format and operation to some degree" (p. 183). Larsen
and Agarwala-Rogers (cited in Glaser, 1981) argued in
their 1978 research that:
Re-invention occurs with greater frequency than
complete adoption. This high degree of re-invention
most likely occurs in response to the realities of
the local situation and local community. Each
organization has its own structurevalues,
obligations, commitmentsand to maintain itself,
must filter out those ideas or aspects of an idea
that are not congruent with its overall position
thus reinvention.
The assumption that adopters are passive
audiences is not accurate. Equally, invalid is the
assumption that innovations have a universal and
81


constant applicability. Differences in community
concerns and state and local policy have a dramatic
effect on the applicability of any innovation.
If data from this study, indicating that re-
invention is more common than complete adoption, are
supported by other research, re-invention cannot be
treated as an interesting variation on a basic
model; rather, it becomes the basic model.
Innovations must be introduced as general concept
and include recommendations to facilitate
appropriate adaptations in local settings to fit
local conditions, (p. 183)
The notion of innovation re-invention, then, has a strong
implementation theme and has its own top-down, bottom-up
theorists.
Private Sector Innovation
Private sector innovation has tended to be more
closely associated with technical or product innovation
(Ranter, 1984), but like the public sector, there has
been an increased interest in managerial innovations
(Taylor, 1990; Ranter, 1984). The technical or product
orientation of the private sector is probably a
reflection of profit motive in the private sector. In
fact, both the private and public sector literatures
evinced a bias toward innovation in the sense that both
82


the private and public sectors discuss innovation in
positive terms. However, much of the private sector
literature failed to distinguish among different types of
innovation. The private sector literature tended to
assume a general understanding of the notion of
innovation and its desirability regardless of the type of
innovation, managerial or product.
The private sector literature can be divided into
three groupings: (1) those that would restrict innovation
to Research and Development (R & D), or what Christensen
and Clark (1992) called "deliberate" innovation (p. 352);
(2) those that would argue that organizations must
continually innovate throughout organizations at all
levels, or what Christensen and Clark (1992) called
"emergent" (p. 352) or what Hart (1992) called generative
and, (3) those who would emphasize the "firm-in-sector"
perspective (Loveridge & Pitt, 1990, p. 5).
Both the deliberate and the emergent theories of
private sector innovation can be interpreted within a
dynamic and complex environment. This is a complex
environment that may impinge or enhance innovation by way
of such dynamics as market structure and intensity of
competition (Nelson & Winter, 1982). Especially, crucial
is the return on investment in R & D, and a related
concern, the efficiency of R & D investment. In
addition, anti-trust implications can inhibit a company
83


with a large market share from seizing an innovative
opportunity. Nelson and Winter (1982) held that
innovation can negatively affect short run "achievement
of Pareto optimality" (p. 116). Modeling these various
concerns is in the purview of the economist and
econometrics, but there is also a literature that
discusses managerial philosophies of innovation in the
private sector.
Drucker (1985) argued from the R & D or deliberate
innovation stance when he insisted that innovation
requires a systematized approach. In his view, this
systematized approach is more consistent with innovation
theory. Drucker maintained that innovation was not the
result of a entrepreneurial personality, "despite much
discussion these days" (Drucker, 1985, p. 67). In fact
Drucker commented that "innovation is work rather than
genius" (p. 72). Drucker argued that "when all is said
and done, what innovation requires is hard, focused,
purposeful work. If diligence, persistence, and
commitment are lacking, talent, ingenuity, and knowledge
are of no avail" (p. 72).
Schrage (1989) took the "emergent" approach to
innovation. Schrage, reviewing mostly technological
innovations, emphasized the importance of major
breakthroughs. He stated that the "most successful
corporate innovation systems aren't 'systems' at all.
84


They are environments..." (p. 44). Schrage emphasized
this is an environment that has "tolerance" for
intelligent failure. However, Schrage placed "corporate
self-deception" (p. 48) at the heart of the obstacles to
innovation, arguing that companies do not always want
what they say they want.
Christensen and Clark (1992) referred to the
"emergent" approach as existing on a continuum, many
different views are represented along this continuum.
Pearson (1988) also emphasized "creating and sustaining"
(p. 99) organizational environments that support and
enhance innovation. Pearson insisted these efforts be
focused and directed toward the realistic strategic goals
of the organization, otherwise, the costs of maintaining
slack resources will be too great. He offered a five
point formula for innovation success:
(1) creating and sustaining a corporate environment
that values better performance above everything
else; (2) structuring the organization to permit
innovative ideas to rise above the demands of
running the business; (3) clearly defining a
strategic focus that lets the company channel its
innovative efforts realistically, in ways that will
pay off in the market; (4) knowing where to look for
good ideas and how to leverage them once they're
found; (5) going after good ideas at full speed,
85


with all the company's resources brought to bear.
(P- 99)
If there is a point of divergence in the debate on
private sector innovation literature, it is on the size
of the organization. There is a theoretical controversy
about the innovation efforts and successes of small
organizations versus large organizations (Peters, 1991;
Nelson & Winter, 1982; Quinn, 1985). Peters (1991),
emphasizing the breakthrough definition of innovation,
cited efforts to create spin-offs from larger
organizations by starting a couple of entrepreneurs in an
area of needed technology. If the venture is successful,
it then becomes an affiliate of the larger corporation,
if success continues, it becomes a total spin-out from
the affiliate. Peters observed that it was difficult to
imagine a large corporation adapting to this approach.
Quinn (1985) maintained that small inventors have
less at risk. They invent at the lowest cost with little
overhead. Large bureaucracies with their accounting
principles aimed at explaining overhead constrain the
resources that a company can sink into new projects.
Although Quinn (1985) described managing innovation as a
venture in managing "controlled chaos" (p. 73), he
opined:
Many features of small company innovators are also
applicable in big companies. With top-level
86


understanding, vision, a commitment to customers and
solutions, a genuine portfolio strategy, a flexible
entrepreneurial atmosphere, and proper incentives
for innovative champions, many more large companies
can innovate to meet the severe demands of global
competition, (p. 84)
Ranter (1984), who has urged innovation on American
enterprise, has been one of the more frequently quoted
proponents of the "emergent" theory of innovation. She
held similarly to Quinn (1985) and Pearson (1988) that
the key to innovation is establishing organizational
environments that are conducive to innovation. Ranter's
work was based on the analysis of existing firms held by
reputation to be innovative firms. Ranter concluded that
organizations need to: (1) develop a participative style
of management; (2) integrate organizational structures
and functions, not segmentalize; (3) encourage a prideful
organizational culture; (4) make organizational power
sources available to innovators; (5) improve lateral
communication; (6) reduce hierarchy; and (7) have a
generative approach to the company's strategic plan,
allow the process to flow up before it flows down the
organization.
The company's strategic plan is an integral part of
innovation efforts, especially its technological
innovations. Freeman and Barley (1990) offered the
87


organizational network perspective for understanding a
firm's strategy over time. Freeman and Barley integrated
two theories: (1) organizational network, and (2) the
organizational population ecology model. They
demonstrated the explanatory strength of this model by
describing Genentech, a biotechnology firm in its
environment. Their approach disclosed complex
relationships among suppliers, joint venture
organizations, licensing efforts, and cooperative
research. In emphasizing these relationships, Freeman
and Barley bring the discussion full circle. A company's
strategy represents its "deliberate" efforts at
structural, in this case, if not product innovation. In
fact, the overriding context for private sector
"deliberate" innovation is the organization's strategic
plan.
Taylor (1990) in his interview with Paul Cook,
former CEO of Raychem, elicited comments regarding most
of the foregoing concerns from Cook. Cook argued for the
need for talented people in an environment where
innovation is expected. Cook emphasized that companies
must bring products to the market swiftly, paying close
attention to the details it takes to get them there, and
holding down the costs on the way to market. Cook stated
that to get and maintain the employee motivation that
innovation requires, the company must offer individual
88


recognition, "patient capital," and a willingness to
obsolete your own products. He viewed innovation as a
way to compete without competing; it is a way to develop
products that competitors cannot match.9 He recognized
that this approach can dampen profits, temporarily, but
he argued his approach ensures long term returns. Cook
agreed with Peters that large companies have to find
pathways to innovation. Size gets in the way of
innovation, according to Cook.
Private sector innovation literature is also marked
by a conglomeration of theory, practice, and empirical
inquiry. In this brief discussion of private sector
innovation10, there was much that related to the central
concern of this investigation. The tension between
theories of "deliberate" innovation versus "emergent"
innovation echo the concerns of the top-downers and the
bottom-uppers, the need for collaborative environmental
relationships, and an acknowledgment of the potential
costs and potential benefits of innovation.
This notion is more fully developed by Prahalad, C.
K., & Hamel, G. (1991). The core competence of the
corporation. In Cynthia A. Montgomery and Michael E.
Porter (Eds.), Strategy (pp.277-299). Boston: Harvard
Business Review Press.
l0Like so many areas of innovation theory, the
private sector literature is broad. The effort here is a
representative effort, not an exhaustive one.
89


Private and Public Sector Innovation
There have been several efforts to distinguish
between the characteristics of private sector innovation
from public sector innovation or what Nelson and Winter
(1977) referred to as innovation "selection" (p. 71)
environments (Altshuler & Zegans, 1990; Behn, 1991;
Roessner, 1979). With the exception of Roessner's (1979)
examination of private and public selection environments,
most of the 11 selection" environment observations have
been largely a priori.
Altshuler and Zegans (1990) enumerated the
prescriptions of private sector innovation and tried to
demonstrate a priori that the public sector is limited in
its ability to create an environment conducive to
innovation. For Altshuler and Zegans, the public sector
lacks the decentralization, individual rewards,
managerial discretion, and freedom of risk taking
requisite to innovation. Altshuler and Zegans offered a
definition of innovation, but they made no effort to
distinguish among policy innovation, managerial or
administrative innovation, and product innovation. They
concluded that it is more difficult to innovate in the
public sector because of the structural and functional
constraints on the public sector. Behn (1991) made a
similar argument, though he added a prescriptive element,
that of trust. He argued that if the public sector can
90


gain the trust of its constituency, it will then be
afforded the requisite conditions for innovation.
Organizational size would seem to be playing a
factor in this private-public comparison. Perhaps, large
public organizations have more in common with large
private organizations in the matter of innovation. It
would seem reasonable that any discussion of private
sector innovation and public sector innovation would
ideally differentiate among the many types of innovation
before coming to conclusions about the two sectors.
In addition, most of the analyses of private
innovation versus public innovation tend to focus on
managerial differences. This emphasis seems to have led
to underestimating the constraints on private sector
innovation, particularly with regard to the market place
as noted by Nelson and Winter (1982). For example,
Walker (1981) comments:
In the private sector, when the market is working
without externally imposed constraints, firms are
able to evaluate the possible benefits of an
innovation according to its potential profitability;
and especially if competitors have already begun to
employ or sell the innovation, it is not difficult
for a firm to make judgments about the innovation's
impact on their position in the market. If business
firms are unwilling or unable to adopt profitable
91