CONSUMER SATISFACTION AND COMPLAINING AND
SWITCHING RESPONSES TO DISSATISFACTION WITHIN
THE VOUCHER/CERTIFICATE RENTAL MARKET
Lewis A. Suiter
B.S., Regis University, 1987
M.S., University of Colorado at Denver, 1991
M.S., University of Colorado at Denver, 1993
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
1999 by Lewis A. Suiter
All rights reserved.
This thesis for the Doctor of Philosophy
Lewis A. Suiter
has been approved
Suiter, Lewis A. (Ph.D., Public Administration)
Consumer Satisfaction and Complaining and Switching Responses to Dissatisfaction Within
the Voucher/Certificate Rental Housing Market
Thesis directed by Professor Kathy J. Boyd
This study is concerned with consumer behaviors in the voucher/certificate rental
housing market because vouchers are considered, by many policy makers, a solution to
problems in public housing and privately-owner properties with project-based Section 8
assistance. The popular view is that vouchers give assisted households a choice of housing
in the market, forcing owners to improve performance because renters can switch if
dissatisfied. This view is incomplete.
Because vouchers are a direct subsidy to consumers, organizations that administer
voucher programs should be consumer-centered, helping voucher-assisted consumers
secure desired outcomes and avoid undesirable ones. Increased purchasing power does not
make economically and socially disadvantaged consumers effective participants in the
market automatically. The governments role in assuring efficient operation of markets is
well established; markets that rely on publicly funded vouchers should not be exempted.
Assisted consumers who fail to get full value from the goods and services they
purchase are cheated along with the taxpayers. Effective complaining to sellers, or to third
parties, if sellers are unwilling to fulfill their obligations, is the only means to redress
problems. Switching without complaining does not inform sellers, other consumers, third
parties, or program administrators that product or service quality is not acceptable and does
not maintain the efficacy of the voucher arrangement.
This study represents the first attempt to test the interrelationship between overall
property satisfaction and unit, location, and property management performance among
voucher/certificate-assisted householders and consumer complaining and switching
responses to dissatisfaction within the voucher/certificate rental housing market. This study
tested Hirschmans (1970) exit and voice theory by applying models developed in the market
research literature to consumer satisfaction and responses to dissatisfaction within the
voucher/certificate rental housing market. This study builds on the efforts of Oliver (1997),
Singh and Wilkes (1996), Zeithaml, Parasuraman, and Berry (1990), and Cronin and Taylor
My thanks to the staff of the Graduate School of Public Affairs for their support and to
my advisor for her encouragement
Government Housing Programs.....................................2
Significance of the Research....................................4
Statement of the Problem........................................5
Research Design and Methodology................................13
2A. LITERATURE REVIEW: THE PRIVATE DELIVERY OF PUBLIC GOODS... 16
Characteristics of Goods and Services...............................16
Alternative Delivery Arrangements...................................17
Competitive Market Versus Government Monopoly.......................18
Alternative Rental Housing Delivery Arrangements.......................19
Government Delivery Arrangement.....................................19
Contract Delivery Arrangement.......................................20
Voucher/Certificate Delivery Arrangement............................20
Demographic Characteristics of Voucher/Certificate-Assisted Households ...23
2B. LITERATURE REVIEW: CONSUMER SATISFACTION..................................25
Quality Versus Satisfaction...............................................26
Unit, Location, and Property Management Attribute Performance......36
Properly Management Performance................................38
Demographic Influences on Housing and Neighborhood Satisfaction....40
2C. LITERATURE REVIEW: COMPLAINING AND SWITCHING
RESPONSES TO DISSATISFACTION...........................................41
Consumer Complaint Responses.......................................47
Attitude Toward Switching and Complaining..........................52
Expectancy Value of Complaining....................................52
Prior Complaining and Switching Experience.........................53
Complaining and Switching Responses to Dissatisfaction.........56
Attitude Toward Complaining....................................57
Prior Complaining Experience...................................58
Expectancy Value of Complaining................................58
Attitude Toward Switching......................................59
Prior Switching Experience.................................... 59
3. RESEARCH DESIGN AND METHODOLOGY....................................62
Location of Study and Participants.................................63
Population and Sampling............................................64
Survey Instrument Pre-testing..................................66
Domain Definitions and Measurements................................67
Unit Attribute Performance.....................................67
Location Attribute Performance.................................67
Property Management Attribute Performance......................67
Property Management Service Performance SERV/PERF..............68
Overall Property Satisfaction..................................68
Likelihood of Complaining to Property Management and Third Parties ....69
Attitude Toward Complaining........................................70
Prior Complaining Experience.......................................70
Expectancy-Value of Complaining....................................71
Attitude Toward Switching..........................................71
Prior Switching Experience.........................................72
Analysis of the Data..................................................74
Unit Attribute Performance.........................................75
Location Attribute Performance.....................................75
Property Management Attribute Performance..........................75
Service Performance SERV/PERF......................................76
Overall Property Satisfaction......................................76
Satisfaction Estimated Path Coefficients and Hypothesized Results..76
Complaining and Switching Responses to Dissatisfaction Measures.......78
Likelihood of Complaining to Property Management...................78
Likelihood of Complaining to Third Parties.........................78
Attitude Toward Complaining........................................79
Prior Complaining Experience Property Management...................79
Expectancy Value of Complaining to Property Management.............79
Expectancy Value of Complaining to Housing Authority...............79
Attitude Toward Switching..........................................81
Prior Switching Experience.........................................81
Complaining and Switching Responses to Dissatisfaction Estimated
Path Coefficients and Hypothesized Results.........................81
Demographic Influences on Hypothesized Relationships..................84
5. DISCUSSON AND IMPLICATIONS............................................114
Limitations of the Study.............................................114
Implications far Practice............................................118
Implications for Policy..............................................119
Directions for Further Research......................................121
1.1 Hypothesized Model of Consumer Satisfaction and Complaining
and Switching Responses to Dissatisfaction Within the Voucher/
Certificate Rental Housing Market.........................................12
2.1 Hypothesized Model of Consumer Satisfaction Within the Voucher/
Certificate Rental Housing Market.........................................60
2.2 Hypothesized Model of Consumer Complaining and Switching
Responses to Dissatisfaction Within the Voucher/Certificate
Rental Housing Market.....................................................61
4.1 Hypothesized Model of Consumer Satisfaction Within the Voucher/
Certificate Rental Housing Market.........................................80
4.2 Hypothesized Model of Consumer Complaining and Switching
Responses to Dissatisfaction Within the Voucher/Certificate
Rental Housing Market.....................................................85
2.1. Housing and Neighborhood Problems Reported by Very-Low
Income Households Within Three Rental Housing Delivery
2.2 Moves in the Past Year............................................45
4.1 PHA Representativeness............................................73
4.2 Demographic Representativeness....................................73
4.3 Unit Attribute Performance........................................86
4.4 Location Attribute Performance....................................87
4.5 Property Management Attribute Performance.........................88
4.6 Property Management Service Performance SERV/PERF Tangibles.......89
4.7 Property Management Service Performance SERV/PERF Reliability.....90
4.8 Property Management Service Performance SERV/PERF
4.9 Property Management Service Performance SERV/PERF Assurance.......92
4.10 Property Management Service Performance SERV/PERF Empathy........93
4.11 Overall Property Satisfaction....................................94
4.12 Satisfaction Estimated Path Coefficients and Hypotheses..........96
4.13 Satisfaction Estimated Path Coefficients and Hypotheses With
4.14 Likelihood of Complaining to Property Management.................98
4.15 Likelihood of Complaining to Third Parties.......................99
4.16 Switching Intent................................................100
4.17 Consumer Alienation.................................................101
4.18 Attitude Toward Complaining.........................................102
4.19 Prior Complaining Experience to Property Management.................103
4.20 Prior Complaining Experience to Third Parties.......................104
4.21 Expectancy of Complaining to Property Management....................105
4.22 Value of Complaining to Property Management.........................106
4.23 Expectancy Value of Complaining to Property Management..............106
4.24 Expectancy of Complaining to Housing Authority......................107
4.25 Value of Complaining to Housing Authority...........................108
4.26 Expectancy Value of Complaining to Housing Authority................108
4.27 Attitude Toward Switching...........................................109
4.28 Prior Switching Experience..........................................110
4.29 Complaining and Switching in Response to Dissatisfaction
Estimated Path Coefficients and Hypotheses...........................111
4.30 Complaining and Switching in Response to Dissatisfaction
Estimated Path Coefficients and Hypotheses With Demographic
This research examines the effectiveness of Section 8 rental housing voucher and
certificate1 arrangements from the perspective of consumers. The Section 8
voucher/certificate program provides public rental assistance to eligible households to rent
housing of their choice in the private market. Two primary questions are addressed by this
research: How important are unit, location, and property management service performance
to their overall properly satisfaction? And, if dissatisfied, how likely are they to complain to
property management or third parties, or switch to restore satisfaction?
These questions are important because voucher-assisted consumers, as well as
taxpayers, should get the most for their money in the private rental housing market.
Regulation alone cannot ensure quality housing products and services. Voucher-assisted
consumers must hold rental housing providers (owners and management agents)
accountable for their side of the bargain: delivering housing units, housing services, and
neighborhoods that perform as required by both regulation and the rule of reciprocal
exchange (Taylor 1993, 385). Assisted consumers must be able to command the dignity
and respect due all consumers in the market (Taylor 1993). They should be able to
complain and/or switch when unit, management, or location performance is unsatisfactory.
1 Both programs, Section 8 Vouchers and Section 8 Certificates, are voucher arrangements
that provide a monthly rent subsidy to eligible households enrolled in the programs. No
significant differences exist between the two programs in terms of housing adequacy,
amenities, or assisted householder ratings of their neighborhood (Legerand Kennedy
This research tests the exit ("switching") and voice ("complaining") theory
developed by Albert O. Hirschman, in Exit, Voice, and Loyalty: Responses to Decline in
Firms, Organizations, and States (1970). The consumer satisfaction/dissatisfaction and
switching/complaining models developed and tested in the business marketing research
literature are used to analyze the policy tool of vouchers in the housing sector (Cronin &
Taylor 1992; Parasuraman, Zeithaml, & Berry 1988; Oliver 1997; Singh & Wilkes 1996).
Assisted consumers should first complain to property management; then, if
property management is unresponsive, to third parties prepared to listen and respond. If
complaining is ineffective, they should be able to switch to an alternative property without
being held hostage by lease terms or bureaucratic procedures. They should document the
property management's poor performance for the benefit of public officials and other rental
housing consumers. Economist Arthur Best (1981,12-13) concludes:
"Our economic system is meant to provide values to buyers: it should give
consumers a certain quality of life, and even a certain quantity of life.
Consumer problems actually represent thefts of the values people work to
earn. At present, consumers absorb most problems without making
complaints. Thus, information about problems, which could be a primary
benefit of efficient and fair complaint handling, is denied to government,
business, and the general public. Fair treatment of consumer complaints
should be recognized as a powerful mode of consumer protection, as
valuable as both regulation, which seeks to improve the overall quality of
goods and services, and consumer education programs, which aim to
provide people with better information about products and services and the
ability to use that information."
Government Housing Programs
The Housing Act of 1937 established the federal governments role in delivering
multi-family rental housing to eligible households. It created public housing authorities
(PHAs) to construct and manage safe, decent, and sanitary accommodations as an
alternative to the slum conditions then being provided by the private rental housing market
The Housing and Community Development Act of 1974 ushered in two new multi-
family rental housing delivery programs: project-based Section 8 contract rental assistance
program, and tenant-based Section 8 voucher and certificate rental assistance. Both
increase the role of the private sector in delivering rental housing to eligible households.
Today, publicly financed rental housing in the United States is delivered through five
arrangements of which four are under the jurisdiction of the United States Department of
Housing and Urban Development (HUD)2. The four HUD programs and the cumulative
number of housing units provided within each program are as follows (HUD 1996, 3):
1. Public housing, 1.25 million units
2. Project-based Section 8 rental assistance, 1.8 million units
3. Tenant-based Section 8 programs: Rental Housing Vouchers and Rental
Housing Certificates (voucher delivery), 1.4 million renter households
4. The HOME Investment Partnership (HOME) program (intergovernmental
delivery), 63,000 units. This program is funded through HUD and administered
by State and local governments.
The Section 8 voucher/certificate programs are portable (tenant-based versus
project-based) for use in the private rental market to secure housing. Assisted households
must seek housing that satisfies the quality standards and rent rates established by the
local PHA in accordance with HUD regulations. Vouchers and certificates make it possible
for them to rent single-family houses, condominiums, townhouses, mobile homes, and units
within multifamily properties. The PHA pays the difference between the market rent
2The fifth arrangement is the Low-Income Housing Tax Credit program, regulated by the
Internal Revenue Service. Since its enactment in 1996, this program has produced
approximately 450,000 units (HUD 1996).
established by the landlord and the amount paid toward rent by the household (Goering,
Stebbins, & Siewert 1995; Kennedy & Finkel 1994; Leger & Kennedy 1990).
Because Section 8 voucher/certificate programs are market arrangements, their
effectiveness in delivering satisfactory housing can be measured by examining consumer
satisfaction with products and services and their responses to dissatisfaction.
Significance of the Research
Decreasing governments and increasing the private sector's role in delivering
goods and services through market arrangements is the hallmark of public administration in
the 1990s. Evaluation is necessary to determine how well these programs are working.
This research contributes to the literature on policy evaluation and post-bureaucratic
government (Barzelay 1992; Dilulio 1994; Dilulio, Garvey, & Kettl 1993; Thompson 1993;
This research examines the assumption that voucher arrangements permit the
economically disadvantaged to function as consumers in the market, such that they are
able to secure satisfactory goods and services and, if dissatisfied, effectively seek redress
from sellers, complain to third parties, and/or switch to alternative providers (Best 1981;
Hirschman 1970; Taylor 1993). Assisted consumers, like unassisted counterparts, should
utilize a mix of switching and complaining behaviors to secure and maintain satisfaction
with the products/services they consume.
Public administrators must continuously evaluate the effectiveness of voucher
arrangements from the perspective of the consumers by:
1. Assessing the consumers overall satisfaction with the goods and services
acquired in the private market with public assistance.
2. Ensuring that complaining and switching behaviors are both active and effective
means of maintaining satisfaction.
3. Facilitating switching.
4. Increasing the effectiveness of consumer complaining through individual skill
development and the creation of mechanisms that increase the sensitivity of private
housing providers to consumer complaints (Hirschman 1970).
This research provides an instrument to gather baseline data which future research
can utilize to examine the efficacy of voucher systems as market delivery arrangements.
To date, PHAs have had no access to theory-based instruments to evaluate the
effectiveness of the market-like rental housing voucher/certificate programs from the
consumers perspective. In the future, this instrument can be adapted to evaluate other
market-like arrangements within the public sector in general as the trend toward
Statement of the Problem
Since the 1970s, the primary aim of the privatization movement has been to reduce
governments role of the private sector in delivering of public goods and services.
Proponents of privatization consider competitive markets a superior means of efficiently
delivering all but the most "public" goods and services compared to government
monopolies. They advocate market arrangements that reduce governments role in
financing, arranging, producing, and delivering goods or services. The dominant keywords
for this movement are competitive, enterprising, market-oriented, and customer-driven
government (Bish & Ostrom 1973; Donahue 1989; Gore 1993; Osborne & Gaebler 1992;
Kirlin, Ries, & Sonenblum 1977; Ostrom & Ostrom 1977; Savas 1977,1987).
The voucher system is unique because government directly subsidizes consumers
(Savas 1977,1987). Voucher arrangements presume that the only difference between
assisted and unassisted consumers is purchasing power. Voucher-assisted consumers,
not government bureaucrats, evaluate price and quality and choose firms that provide
satisfactory goods and services. If dissatisfied, assisted consumers, like unassisted, can
switch to a competitor (Bridge 1977; Ostrom & Ostrom 1977; Savas 1977,1987). Section 8
rental housing voucher/certificate holders find and lease housing of their choice in the
private market, provided the housing meets minimum quality standards and rent level
This is the first application of Hirschmans (1970) theory to vouchers and the
housing sector. Hirschman's (1970) model bridges the public and private sectors. This
research applies consumer behavior theories and models developed in the marketing
research literature to voucher/certificate rental housing market. The exit and voice
premises of Hirschmans (1970) theory are as follows:
1. Consumer switching ("exit") is the termination of one consumer-supplier
relationship and the initiation of a relationship with an alternative supplier.
Switching in response to dissatisfaction with product or service quality is
neither the only nor the most effective means of informing firms that quality has
2. Consumer complaining ("voice") is an alternative and complementary means of
informing firms that quality has declined. Complaining is a direct attempt by the
consumer to communicate the existence of a dissatisfying condition, such as
poor product quality or service performance, and secures a specific remedy. 3
3. When dissatisfied with products and services, consumers tend to rely on
switching as a means of restoring satisfaction because switching is a private,
'secref vote in the anonymity of a supermarket (16). Consumer switching
does not provide firms with the information needed to restore performance and
4. Effective complaining is difficult because it is messy, direct and
straightforward rather than roundabout Voice is political action par excellence"
(16). Consumers base their decision to complain on past experience with the
costand effectiveness of complaining. The effectiveness of complaining
depends on the sensitivity of firms to the value of this behavior.
5. Organizations should be redesigned to make them more responsive to
Hirschman uses "exit" to describe the behavior of customers, members, and
citizens who terminate their relationship with firms, organizations, political parties, and
states. Since this research applies only to consumers and firms, the term "switching" is
more appropriate to consumer-supplier relationships where consumers terminate one
relationship and initiate another with an alternative supplier (Keaveney 1995).
The term "voice" defines direct individual and collective attempts to communicate
dissatisfaction to firms, organizations, political parties, and states. The term "complaining"
is used to denote the individual consumer's attempt to communicate a dissatisfying
condition such as poor product quality or service performance, and secure a specific
remedy. This research examines the likelihood that voucher/certificate-assisted rental
housing consumers will complain to property management and/or third parties.
Competitive markets are dominated by exit responses to dissatisfaction, while
families, churches, and states are dominated by voice responses to dissatisfaction, where
exit through expulsion is an organizational notan individual decision (Hirschman 1970,76).
Monopolistic markets are dominated by voice responses to dissatisfaction because exit is
not an option. Collective voice is an effective means of restoring satisfactory performance.
In less competitive markets (e.g., fewer suppliers and substitute products or services)
switching is more difficult, which makes complaining more necessary and cost-effective.
Yet, within such markets firms may ignore complaints with little concern about the adverse
consequences of loosing a few dissatisfied consumers. According to Hirschman (1970,
In the economic sphere such "lazy" monopolies which "welcome
competition" as a release from effort and criticism are frequently
encountered when monopoly power rests on location and when mobility
differs strongly from one group of local customers to another. If, as is
likely, the mobile customers are those who are most sensitive to quality,
their exit, caused by the poor performance of the local monopolist, permits
him to persist in his comfortable mediocrity. This applies, for example, to
small city or "ghetto" stores, which lose their quality-conscious patrons to
better stores elsewhere.
When quality-conscious consumers with knowledge of and access to alternative providers
switch in lieu of complaining, they leave behind those consumers who are less able to
switch and less likely to complain effectively (Hirschman 1970).
Within less competitive market structures, consumers should be able to seek
redress from sellers indirectly through third parties such as the Better Business Bureau or
government agencies. If consumers are unable to resolve problems by complaining to
sellers, corrective public policies should either facilitate consumer switching (e.g., referral
services) or amplify consumer complaining (e.g., legal action) (Best 1981).
Hirschman's (1970) loyalty construct is not included in this research because the
construct has not been defined and operationalized in a manner that permits its application
to the context of this research. Assisted households may continue to live in unsatisfactory
housing for a variety of reasons other than loyalty. Research has identified other
behaviors, such as neglect, disloyalty, passivity, and fear, that hold exit at bay without
activating voice (Cannings 1992; Farrell & Rusult 1992; Rusbult, Farrell, Rogers & Mainous
1988; Rusbult, Zemrodt, & Gunn 1982; Saunders 1992; Withey & Cooper 1989). Therefore
most research focuses on exit and voice as key dependent constructs (Singh 1990b).
According to Hirschman (1970), dissatisfied consumers rely on switching and do
not complaining, which makes economic sense for consumers in competitive markets
offering relatively simple goods and services (e.g., groceries). There is no motivation to
expend the effort to inform the seller that performance is unsatisfactory.
Consumer switching does not inform management that performance has
deteriorated. Management may attribute declining sales to general fluctuations in
consumer demand, prompting decisions to cut costs, increase sales efforts, or use
accumulated resources to continue operations until the market improves (Hirschman 1970).
When sellers restore quality in response to consumer complaints, there are four
1. The complaining consumer.
2. The firm.
3. Non-complaining consumers doing business with the firm.
4. All consumers in general.
The complaining consumer demands full value for goods and services and seeks
to maintain his/her dignity as a participant in a market transaction (Taylor 1993).
Complaining restores satisfaction saves the consumer the costs of switching (e.g., search
costs) (Best 1981; Hirschman 1970).
Businesses that encourage and reward complaints gain valuable information
concerning the quality of their products and services and the opportunity to correct lapses
in product and service performance. Firms that respond affirmatively to valid complaints
demonstrate their respect for the dignity of the consumer (Taylor 1993). In addition, firms
avoid the mistake of attributing declining sales to market conditions when product/service
performance failures are to blame (Best 1981; Hirschman 1970, Keaveney 1995).
Consumers unable to detect a decline in quality benefit from the complaints of
those who make management aware of it. Consumers who complain benefit others and
improve the market system (Best 1981; Day 1984; Hirschman 1970; Keaveney 1995).
This research evaluates the effectiveness of the voucher/certificate program from
the perspective of the assisted householder by examining three primary questions. First,
how do assisted consumers perceive the performance of their unit, location, and property
management and do these perceptions influence their overall property satisfaction?
Second, how likely are they to complain to property management or third parties, or switch,
if dissatisfied with their housing? Third, how is the likelihood of complaining and switching
influenced by consumer alienation, prior complaining and switching experience, attitude
toward complaining and switching, and expectancy-value judgments of complaining? A
total of twenty domains are identified for study. The hypothesized inter-relationships
among these domains are presented in Figure 1-1, the Hypothesized Model of Consumer
Satisfaction and Complaining and Switching Responses to Dissatisfaction Within the
Voucher/Certificate Rental Housing Market.
Overall property satisfaction is measured using a scale previously tested, validated,
and reported in the marketing literature (Oliver 1997). The hypothesized antecedents of
overall property satisfaction are unit attribute performance (Casey 1992; Galster 1987;
Newman & Duncan 1979), location attribute performance (Casey 1992; Newman & Duncan
1979; Varday 1983), property management attribute performance (Basile 1994; Isler, Drury,
& Wellborn 1971; Isler, Sadacca, & Dairy 1974; Stegman 1972), are measured using
exploratory scales based on attributes presented in the literature. The five dimensions of
property management service performance (SERV/PERF): tangibles, reliability,
responsiveness, and assurance are measured using scales developed by Parasuraman,
Zeithaml, & Beny 1988 and Zeithaml, Parasuraman, & Berry 1990 and tested by Cronin &
Three responses to dissatisfaction are measured: the likelihood consumers will
complain to the property management Likelihood of Complaining to Property Management
(Singh 1989,1990a, 1990b, 1991; Singh & Wilkes 1996); Likelihood of Complaining to
Third Parties (Singh 1989,1990a, 1990b, 1991; Singh & Wilkes 1996); and the likelihood
that assisted consumers will switch housing; Switching Intent (Hirchman 1970; Keaveney
The hypothesized antecedents of complaining to property management and third
parties or switching are presented in Figure 1.1. Consumer alienation is global attitude or
feelings toward business, business owners, and the marketplace (Allison 1978, Singh &
Wilkes 1995). Prior complaining experience is the frequency of complaining and switching
(Singh & Wilkes 1995). Attitude toward complaining and switching refers specifically to a
consumer-supplier context (Singh & Wilkes 1995). Expectancy-value (EV) judgement is
the consumers cost and benefit analysis for complaining (Singh & Wilkes 1995).
Figure 1.1 Hypothesized Model of Consumer Satisfaction and
Complaining and Switching Responses to Dissatisfaction Within the Voucher/Certlflcate Rental Housing Market
and empathy. It Is presented a one domain to simplify the model.
Likelihood of Complaining to Property Management, Likelihood of Complaining to Third
Parties, and the hypothesized antecedents are measured by scales previously tested,
validated, and reported in the marketing research literature (Singh 1989,1990a, 1990b,
1991; Singh & Wilkes 1996). Three direct questions regarding the likelihood of switching in
the next 90 days or next year, and whether the householder has already located another
unit measure switching intent Prior switching experience is based directly on the number
of moves made in the past five years (Casey 1992).
Research Design and Methodology
The research was conducted within three PHAs: Richmond and Newport News,
Virginia, and Greensboro, North Carolina. They were selected because their housing
markets had a good supply of affordable rental units (HUD 1996; Greensboro 1995;
Newport News 1995, Richmond 1995). A supply of affordable units is necessary for
switching. In addition, they manage their programs substantially the same without special
programs to encourage switching or complaining. Also, Directors and senior staff at each
PHA were interested in supporting research aimed at improving customer service.
Two focus groups with voucher/certificate-assisted householders in Richmond,
Virginia, helped develop the survey instrument They focused on consumers' description
and definition of their housing unit, its location, the property management, and their overall
housing consumption experience. The survey was pre-tested. Participants in the focus
groups and survey pre-test were selected at random from assisted householders who had
previously completed recertification and, therefore, would not be recertified during the
survey period. Participants in the focus groups and pretest were paid a $20 stipend.
The survey was conducted at the administrative offices of each PHA where
householders are initially certified and annually recertified, and administered by a
designated PHA staff person whose duties did not include certifications/recertifications. The
PHA representative asked householders who had completed recertification to participate in
a survey on housing satisfaction being conducted by a graduate student from the University
of Colorado at Denver. They were verbally informed that participation in the survey was
completely voluntary and the survey had nothing to do with their recertification.
The survey was administered at the time of recertification because (1) it was
convenient for the householder, PHA, and the researcher; (2) it was ideal time because
householders were reflecting on and evaluating their housing consumption experience; and
(3) householders who intend to switch must do so at the end of their lease which often
coincides with recertification. Householders who intend to switch units had made their
decision and were in the process of securing another unit Those who did not intend to
switch had made a conscious decision to continue renting their present unit.
Each participant, including focus groups and pre-tests, was provided an Informed
Consent Form as required by the Graduate School, University of Colorado at Denver, to
ensure they knew what the study was about, who was conducting it, that participation was
voluntary, and that they could discontinue at any time. The Informed Consent Form
explicitly stated there was no relationship between participation and the participants
recertification in the Section 8 program. Participates signed the Informed Consent Form
and were given a copy, a pen, if required, and a quiet place to complete the survey. The
survey was not completed in the office or presence of the PHA staff member conducting
the recertification. After completing the survey, householders placed them in sealable
envelopes addressed to the principal investigator, in care of the Graduate School of Public
Affairs, University of Colorado at Denver. Envelopes were collected by the designated
PHA staff person and mailed directly to the Norwest Public Policy Research Program,
Graduate School of Public Affairs, University of Colorado at Denver, for data entry.
Secondary sources provided data on voucher/certificate-assisted householders
nationwide and within the study population (Burke 1996; Casey 1992, HUD 1996), vacancy
rates and the market structure within each PHA jurisdiction (U.S. Census Bureau 1997;
Greensboro 1995; Newport News 1995; Richmond 1995).
The unit of analysis was the householder, the person or one of the persons in
whose name the house or apartment is owned or rented. Prior to 1980, the concept head
of household was used (Casey 1992, 4). The number of voucher/certificate-assisted
householders in the three jurisdictions was 3,989. The number of assisted householders
who completed their annual recertification for continued eligibility in the Section 8
voucher/certificate program over the six and one-half month survey period was 1,556. A
total of 433 surveys were received.
CHAPTER 2, PART A
LITERATURE REVIEW: THE PRIVATE DELIVERY
OF PUBLIC GOODS
Arguments for privatization rest upon three foundations. First, public and private
goods and services can be differentiated based upon consumption and exclusion
characteristics. Second, goods and services can be financed, produced, and delivered
through a variety of arrangements that specify distinct roles for the public and private
sectors. Third, the proponents of privatization consider competitive markets superior to
government monopoly as an effective means of delivering all but the most "public" goods
Characteristics of Goods and Services
Economists classify goods and services are classified as public goods, common-
pool goods, toll-goods, and private goods based on two characteristics: consumption and
exclusion. Public goods are subject to joint consumption because it is difficult (too costly)
to exclude anyone from benefiting from the good. A public park is one example. Private
goods are consumed individually and exclusion is easily achieved at a low cost Food,
clothing, medical care, housing, and most consumer goods and services are private goods.
Each individual consumer determines the amount of private goods and services that
he/she is willing or able consume (Bish & Ostrom 1973; Pappas & Hirschey 1987; Ostrom
& Ostrom 1977; Savas 1977,1987). Private goods such as food, shelter, clothing,
emergency medical services, and education are considered vital to human survival and
development. If the private market is unable or unwilling to provide these goods and
services at a price affordable to all citizens, especially the poor, the government and/or
private charitable organizations arrange to provide them at little or no direct cost to those in
need (Savas 1987).
Alternative Delivery Arrangements
Privatization relies on market arrangements to reduce the role of government in
financing, arranging, producing, and delivering goods and services. According to E. S.
Savas (1977,1987), the four primary arrangements to deliver public goods and services
are as follows.
1. contracting with private firms to produce and deliver goods and services,
2. franchising the operation of government-owned facilities to private firms that charge
3. funding grants to private institutions that provide services, and
4. funding and distributing vouchers that permit recipients to purchase services
directly from private producers of their choice in the competitive market
Arguments for the use of vouchers rather than direct government delivery are
based on fundamental belief in the efficiency and effectiveness of the marketplace.
Because firms must compete for customers, only the most efficient survive. Unlike
government bureaucracies, inefficient and ineffective private firms go out of business.
Second, existing private sector distribution channels eliminate the need for duplicate
channels in the public sector. Finally, voucher-assisted consumers, like unassisted
consumers, are free to purchase goods and services from firms of their choice based on
individual preferences and assessments of price and quality. Consequently, consumers,
not government bureaucrats, decide what goods and services will be produced (Bridge
1977; Pappas & Hirschey 1987).
Competitive Market Versus Government Monopoly
Proponents of privatization believe competitive markets are superior to government
monopolies as a means of delivering all but the most "public" goods and services.
According to Savas, Competition is the key to achieving better and more cost-effective
public services; a monopolist arrangement, whether government or private, is an invitation
to poor performance (1987, 272).
Competitive markets are characterized by a large number of buyers and sellers of
an essentially identical product, where the participants' transactions are so small in relation
to the total industry output that they cannot affect the price of the product (Pappas &
Hirschey 1987, 366). In addition, buyers and sellers have equal information on the cost,
price, and quality of the products (Pappas & Hirschey 1987).
Since pure competition and pure monopoly market structures seldom exist, market
structures fall within a range of less competitive to more monopolistic (i.e., "loose
monopolies) (Andreasen 1984,1985; Hirschman 1970; Singh 1990a). The key
characteristics of loose monopolies are as follows;
1. The number of firms in the market is limited by restrictive licensing.
2. Consumer information about alternative products/services is limited.
3. Consumer information about alternative suppliers is limited.
4. Products/services are complex and/or infrequently purchased, making it more
difficult to evaluate quality.
5. Consumer switching and/or complaining is inhibited by custom or norms (e.g.,
getting a second opinion is considered disloyal) (Andreasen 1984,1985;
Hirschman 1970; Singh 1990).
Researchers have concerns about loose monopolies within the Section 8
voucher/certificate arrangement (Goering, Stebbins, & Siewert 1995; Kennedy & Finkel
1994). Local rental markets are affected by the supply of available housing, landlord
preferences, racial discrimination, Section 8 discrimination, PHA operations and local
program characteristics, and the Fair Market Rents established by HUD. John Goering,
Helene Stebbins, & Michael Siewert (1995, 39-40) conclude that:
Although the research evidence on this issue remains largely inferential, its
seems to suggest that, by absorbing a significant share of the Section 8
population, this submarket may exacerbate both racial isolation and
concentrations of poverty; if applicants in tenant-based rental assistance
programs are referred only to neighborhoods where other assisted families
liveor limit their search to these areas-they will resegregate themselves
and perpetuate economic and racial isolation and its consequences.
In addition, an economic barrier to switching may exist because the costs of
switching are a relatively high percentage of household income (Goering, Stebbins, &
Siewert 1995; Kennedy & Finkel 1994).
Alternative Rental Housing
Government financed rental housing for VLI households is delivered through three
arrangements: government, contract, and voucher.
Government Delivery Arrangement
The Housing Act of 1937 established the federal government's role in the delivery
of multifamily rental housing to VLI households. In response to the slum conditions
provided by the private sector, the Act created PHAs to construct and manage safe,
decent, sanitary accommodations for the working poor (Scobie, 1975). Today, local PHAs
own and manage multifamily housing properties serving VLI households that pay 30
percent of their income to the PHA to lease a public housing unit HUD establishes
minimum standards for housing quality and eligibility criteria. The PHA is required to
maintain the unit in a decent, safe, and sanitary condition (HUD 1996).
Contract Delivery Arrangement
The Housing and Community Development Act of 1974 created Section 8 project-
based New Construction and Substantial Rehabilitation contract rental assistance program.
This program increased the role of the private sector in delivering rental housing to VLI
households by providing rental subsidy payments directly to landlords who contract to
provide specific rental units to eligible households. Government pays the difference
between the HUD approved contract rent and the amount the household is required to pay
toward rent (HUD 1996).
Eligible households apply directly at the property's leasing/management office. If
no unit is available, the household can request to be placed on a waiting list. Lease terms
generally follow a form approved by HUD and/or the PHA. The rental rate must comply
with the regulatory requirements of the Section 8 voucher/certificate program. Eligible
households must occupy the contract unit in order to receive the rent subsidy (HUD 1981).
The Housing and Community Development Act of 1974 created the Section 8
certificate program to provide eligible low-income households a portable rent subsidy
(tenant-based versus project-based). The Section 8 voucher program was established in
1987. Both programs operate in the same manner, except voucher-assisted households
are permitted to pay more than 30 percent of their income toward rent if they choose a unit
in excess of the PHA's payment standard. Certificate-assisted households must pay 30
percent of their income toward rent and the rate cannot exceed the PHA's payment
standard (Goering, Stebbins, & Siewert 1995; Kennedy & Finkel 1994; Leger& Kennedy
Low-income households must apply for vouchers/certificates at their local PHA.
Once issued a voucher/certificate, the householder must be recertified annually to continue
participation in the Section 8 voucher/certificate program.
Vouchers/certificates can be used in the private rental market to secure any type of
housing provided it meets occupancy requirements, quality standards and rent levels
established by the local PHA in accordance with HUD regulations. Once the unit and lease
are approved, the household occupying the unit pays its portion of rent directly to the
landlord. The PHA pays the difference between that and the market rent. (Goering,
Stebbins, & Siewert 1995; Kennedy & Finkel 1994).
Vouchers/certificates can be applied to single-family houses, condominiums,
townhomes, mobile homes, and units in multifamily properties. Regulations require the
number of bedrooms in the unit to be appropriate for the number of persons in the
household (Code of Federal Regulations (CFR), Part 24, Sections 882 & 887). Except for
the monthly rate and amount of the security deposit, provisions of private leases are not
subject to HUD's control. PHAs may require landlords to use a model lease (Goering,
Stebbins, & Siewert 1995; Kennedy & Finkel 1994; Leger & Kennedy 1990).
Once the householder finds a unit that meets the size and rent requirements, PHA
inspectors examine it. The Section 8 regulations explicitly recognize that numbers of
assisted households may not be able to detect certain defects or hazards. All units leased
by assisted households are inspected prior to occupancy and at least once each year at or
near the time of recertification to ensure that minimum housing quality standards are
maintained (CFR, Part 24, Sections 882 & 887).
The most common measure of rental unit availability is the vacancy rate compiled
and reported by the U.S. Census Bureau. A vacancy rate of five percent is associated with
market equilibrium. Lower rates indicate a limited supply and higher rates indicate many
units are available. The number and types of rental units available to voucher/certificate
assisted households is limited by the PHA's rental payment standard, which establishes the
maximum that PHA will pay based on the number of bedrooms and bathrooms in the unit
The payment standard cannot exceed, but may be less than, the Fair Market Rents (FMRs)
established by HUD and published annually for every city and/or county in the nation.
FMRs are set at or below the 40th percentile of the housing markets rent distribution
(Goering, Stebbins, & Siewert 1995, 37). In 1993, the percentage of vacant units below
the FMRs was 7.4 percent nationally, 5.9 percent in the northeast, 7.6 percent in the
Midwest, 8.5 percent in the south, and 7.2 percent in the west (HUD 1996).
Units available to voucher/certificate-assisted householders are limited further by
the number of owners or property managers who are familiar with the Section 8
voucher/certificate program and willing to participate. Nationally, 24 percent of property
owners and managers are very familiar and 34 percent are somewhat familiar with the
Section 8 program (U.S. Census Bureau 1997).
A study of 33 PHAs revealed that 24 percent of households rely on lists of owners
or management agents provided by the PHA; approximately 22 percent of assisted
households rely on friends. Daily newspapers are an important source of rental information
for 18 percent (Kennedy & Finkel 1994).
Aside from visual clues, such as cleanliness, general condition, age of appliances,
appearance of the leasing office and property staff, little outside information is available
about the property or its management For example, information such as criminal activity in
a particular location is not readily available from either the property management or the
PHA. Word-of-mouth is the primary source of information on which assisted households
base their pre-rental decisions.
Demographic Characteristics of
Nationally, the average voucher/certificate assisted household is comprised of 2.8
persons and earns $8,500 per year. More than 20 percent of these households earn less
than $5,000 per year, and approximately four percent earn more than $20,000. The
majority (60 percent) are single parents of which 83 percent are females. Nearly 10
percent of the householders are younger than age 25, 57 percent are between ages 25 and
44 years, 18 percent are between ages 45 and 61 years, and 16 percent are age 62 or
older. The racial and ethnic distribution is: white, 46%; African-Americans, 37%;
Hispanics, 14%; Asian/Pacific, 2%; and Native Americans, 1% (Burke 1996).
Within each PHA, the choice of rental housing is limited by federal program
regulations, local program management, the market supply of units in the market, the
number of property managers aware of and willing to participate in the Section 8 program,
and by information on the location and quality of available units. Despite these limitations,
voucher/certificate-assisted households generally have a choice of housing, as evidenced
by high utilization rates (Goering, Stebbins, & Siewert 1995; HUD 1996; Kennedy & Finkel
1994; Leger & Kennedy 1990). This research extends beyond whether voucher/certificate-
assisted households are able to switch from one property to another because there is an
adequate supply of units.
CHAPTER 2, PART B
LITERATURE REVIEW: CONSUMER SATISFACTION
According to Oliver (1997,13):
Satisfaction is the consumers fulfillment response. It is a judgement that a
product or service feature, or the product or service itself, provided (or is
providing) a pleasurable level of consumption-related fulfillment including
levels of under- or overfulfiilment
Satisfaction is distinct from product/service performance or quality. Satisfaction is
related to an "outcome and a comparison referent" (Oliver 1997, 14). Satisfaction is not the
absence of complaints, loyalty, or repeat purchasing (Oliver 1997).
Satisfaction/dissatisfaction may be experienced during the consumption
experience and expressed as a final judgement of the consumption outcome.
Dissatisfaction is simply the "displeasure of underfulfillment" and/or the "displeasure of
overfulfillment if it is unpleasant (Oliver 1997,14).
Consumers compare initial observations of performance to their expectations
(confirmation/disconfirmation), need fulfillment, and perception of "what might have been" if
alternative products or services had been selected (regret) (Oliver 1997,216). If
sufficiently "aroused" by their appraisal of product or service performance, consumers may
begin to worry whether the product/service will perform as expected (dissonance) and/or
why the product/service failed to meet expectations (attributions)(Oliver 1997). Once
dissonance and/or attributions are evoked, the consumers product response is more
complex, consisting of the primary effect plus specific types of distinct emotions resulting
from the consumer's analysis of the product outcome (Oliver 1993,1994,1997).
These emotions include negative affects, such as anger, disgust sadness, fear,
shame, guilt, and surprise; positive affects, such as joy, pleasure, surprise, delight or
interest If the product/service performance does not "evoke the attribution process," the
consumer's overall response will be determined by the primary affective tone of the
consumption experience (Oliver 1997, 342).
Quality Versus Satisfaction
Quality judgments are based on beliefs or expectations regarding product or
service performance. These beliefs are formed prior to consumption by word-of-mouth,
consumer reports, advertising, and cues from the product/service that indirectly relate to
performance or quality, such as brand name, price, business or employee appearance, or
expensive surroundings (Bitner 1990, Oliver 1997). Consumers evaluate quality based on
measurements, comparisons, and performance evaluations. This is a cognitive process.
Satisfaction, in contrast, is"both a cognitive and an affective response" (emphasis in
original, Oliver 1997,178).
Initial performance evaluation may confirm or disconfirm expectations. Negative
disconfirmation occurs "when performance falls short of expectations in an absolute sense"
and negative disconfirmation includes "both the nonoccurrence of highly probable,
desirable events and the occurrence of remote-probability negative events" (Oliver 1997,
Positive disconfirmation occurs when performance exceeds expectations and when
"highly probable undesirable events do not occur or when very improbable desirable events
do" (Oliver 1997,105). Expectation levels "that are too low will always produce a positive
disconfirmati'on" (Oliver 1997,109). Zero disconfirmation or "confirmation" obtains when
high-probability events occur and low-probability events do not in about the frequency one
would predict, regardless of the valence of the event (Oliver 1997,105). Expectations are
confirmed and, because disconfirmation is absent, there is little arousal or attribution
processing. Expectations become "passive and not actively processed, typically because
they have become permanently coded into the consumers schema (e.g., the expectation
that a reliable clock will keep correct time)" (Oliver 1997, 339). Consequently,
disconfirmation perceptions do not appear as long as performance remains within an
acceptable or tolerable level.
According to Alden Speare, Jr., (1974), physical and/or social environmental
factors create stress and reduce residential satisfaction. "[I]f dissatisfaction passes some
unspecified threshold, then the household or some component of it will develop a desire to
move" (Speare 1974,186). Only those households that develop a desire to move that will
begin weighing the costs and benefits of doing so (Speare 1974). If the consumer's
expectations are met or the extent of disconfirmation remains within the "zone of
indifference" or "latitude of acceptance," the consumer may simply accept suboptimal
performance without becoming dissatisfied (Oliver 1997,112).
Products/services attributes are designed to meet consumer needs. Marketing
efforts aim to make consumers aware of the need-fulfilling benefits of products/services
with particular attributes. According to Oliver (1997):
Generally, what the product has will operate at lower-level need fulfillment,
and what the consumer gets will operate at higher-level need fulfillment
The reason is that the translation of product attribute performance is made
by the consumer. Lower-level needs tend to be functional in nature, and it
is the product features that give the product its ability to perform its function
Oliver (1997) identifies three categories of product/service attributes that mediate
satisfaction: monovalent dissatisfier, monovalent satisfier, and bivalent satisfier. What the
product has or does not have will operate at a lower-level need fulfillment What the
consumer gets (e.g., benefits) will operate at a higher-level need fulfillment (Oliver 1997).
A monovalent dissatisfier is a functional lower-level attribute that is "essential but
unprocessed, only capable of causing dissatisfaction when flawed" (Oliver 1997,152). The
presence of a bathroom, an essential attribute, is not processed as a source of satisfaction,
it only causes dissatisfaction when plumbing or fixtures fail to operate properly. Research
on housing satisfaction has primarily examined the influence of monovalent dissatisfiers
(Casey 1992, Cook & Bruin 1993; Galster 1987; Newman & Duncan 1979; Varady 1983).
Monovalent satisfiers operate when consumers translate product attribute
performance to higher-order need fulfillment (Oliver 1997,151). In the neighborhood
context, satisfaction is primarily related to benefits of a location that has less crime,
congestion, and blight, and more grocery and other stores. The appearance of the property
and proximity to friends and family are also important location attributes. Research on
neighborhood satisfaction has primarily examined the influence of monovalent satisfiers
(Anthony, Weiderman, & Chin 1990; Casey 1992, Newman & Duncan 1979; Speare 1974;
Bivalent satisfiers "can cause both satisfaction and dissatisfaction" (Oliver 1997,
151). Public transportation may be a positive attribute if the householder needs it; on the
other hand, noise and congestion generated by public transportation may cause
Consumers often experience regret after they have been forced to decide between
alternative product/service offerings. If the selection is not satisfactory, alternative offerings
will be viewed as better choices. This judgement may be based on additional information
about the alternative offerings or the consumer's imagination (Oliver 1997). According to
Oliver (1997, 233);
Consumers would generally wish to avoid regret and to make efforts to
minimize its impact through low-risk decisions or through bolstering
immunity to regret.
Consumers seldom rejoice over product/service outcomes because "positive
outcomes give the consumer what was expected" (Oliver 1997,222-223). As indicated
above, the consumer sufficiently disturbed by product/service performance may begin to
worry about whether the product/service purchased will perform as expected (dissonance)
and/or why the product/service failed to meet expectations (attributions) (Oliver 1997).
According to Oliver (1997), dissonance "is an apprehension over events yet to
come," as opposed to "disconfirmation and other comparative processes where the
purchase outcomes are known* (emphasis in original, 247). Dissonance results from a
personal decision to purchase one product/service over another. The purchase decision
excludes alternative products/services that may perform better than the product/service
selected. Prior to actually consuming the product/service, the consumer is faced with the
anxiety that he/she has made a mistake. "This state of anticipated regret (to distinguish it
from the realization of regret) leads to a general feeling of apprehension on the part of the
consumer" (emphasis in original, Oliver 1997, 243).
To reduce dissonance, consumers "will evaluate the desirability of the chosen
alternative and reduce the desirability of the unchosen before the outcomes of the choice
are known" (Oliver 1997,249). Other dissonance reducing strategies include denying that
a attribute is important; revaluing an attribute, making it less important; looking for attributes
that were not initially apparent ("every cloud has a silver lining"); selective forgetting;
denying responsibility for the decision, assigning blame (attribution) to someone else.
Reaction to a consumption outcome begins with evaluation of performance;
disconfirmation begins the attribution process (Oliver 1997,280). According to Oliver
The primary evaluation of success or failure results in a primary positive or
negative affect, setting the general affective tone of the consumption
Depending upon the outcome and the extent of disconfirmation, the consumer
attributes success or failure to him/herself or others, such as the salesperson,
manufacturer, or the property management, attempting to determine "why did this happen?"
(Oliver 1997, 266). The consumer decides whether he/she is responsible (internal
attribution) or if responsibility rests with another party (external attribution). Successful
outcomes are less likely than failures to result in attributions to others; we take credit for
success and blame others for failure (Oliver 1997).
Response to and expectations of redress for a product or service failure is affected
by perceived likelihood the failure will recur (Bitner 1990, Oliver 1997). If attributed to a
stable cause, we expect the failure to occur again and are more likely to demand refunds.
If attributed to an unstable cause, the consumer may be willing to try the product or service
again (Oliver 1997).
Attributions are outcome-dependent and result in distinct emotions that are
attribution-dependent not outcome-dependent (Oliver 1997,281). According to Oliver
Successes attributed to oneself should result in pride and a sense of
personal competence. When success is attributed to others (as in
assistance) or to unexpected outside interventions (such as when a
stranger gives a tip on a horse race), both gratitude and pleasant surprise
should result. Failure attributions operate in the same way. A loss or
failure caused by others is shown to result in anger and, in this case,
unpleasant surprise. Finally, internally attributed failure results in guilt and
regret (over forgone opportunities).
Initial evaluation of a consumption experience is followed by confirmation, resulting
in primary positive or negative affect, known as"unappraised affect, the simple
pleasure/displeasure or'feel good' (or bad) response" (emphasis in original, Oliver 1997,
317). Confirmation may be followed by a process of cognitive appraisal; a comparison of
the outcome to expectations, need fulfillment, quality, and what might have been"
(dissonance). This sets the stage for outcome-dependent reactions (attributions) discussed
above. Two emotional stages are initial or primary affect (unappraised affect) and
appraised affect (cognitive appraisal).
The type and intensity of emotion resulting from the initial outcome evaluation
depends upon expectations and confirmation. When expectations are confirmed (zero
disconfirmation) there is little arousal (e.g. contentment, unemotional, indifference) or
attribution processing. According to Oliver (1997, 312):
Pride, competence, confidence, and relief dominate success attributions;
when others are involved, gratitude is found. Surprise and relief occur with
luck. Anger and frustration, with evidence of guilt and fear, dominate
According to Oliver (1997, 327):
Because many consumer surveys typically use only a single-item scale to
measure satisfaction, one consumer's satisfaction 'apple' becomes
indistinguishable from another's satisfaction 'orange.'
Two satisfied refrigerator owners may differ in their "location in the satisfaction-
emotion space" (Oliver, 1997, 327). One is ecstatic over the novelty of owning a new
refrigerator that dispenses ice. The other, who has owned an identical refrigerator four
years is satisfied that it continues to operate as it should (Oliver 1997). A scale that
measures the arousal or engagement dimension of the consumer's response is needed .
Oliver (1997) explores consumer's arousal or engagement responses within four general
states of satisfaction: pleasure, delight, relief, and contentment
Pleasure may be positive or negative (unpleasant). Within a pleasure-as-
sati'sfaction state, the primary affect may be happiness (positive) or unhappiness
(negative). Depending upon the level of arousal or interest, expectations are processed
and appraisal takes place. According to Oliver (1997,341):
"Although not all aspects of the consumption processing model may be
fully accessed, or accessed at moderate levels of arousal, satisfaction-as-
pleasure is modeled by both the nonprocessing (primary affect) and the
When consumers experience both pleasure and surprise from product or service
performance, they experience delight. To "judge the surprisingness of the outcome,"
consumers must process expectations (Oliver 1997, 341). In addition, "discontinuation
processing will be present since both expectations and performance assessments will be
active" (Oliver 1997,432).
Relief occurs when a product or event eliminates a negative situation, in the way
that an aspirin eliminates a headache. "By definition, satisfaction-as-relief requires that
performance be processed" (Oliver 1997, 342). If full relief is not attained, the consumer
may experience resignation.
"Contentment is a passive response and results when pleasant homeostatic states
are maintained or prolonged" (Oliver 1997, 339). Continuous consumption experiences,
like rental housing, are processed differently than encounter-specific experiences.
Consumers in continuous consumption situations may no longer actively process their
experience. Expectations may be lowered to a level consistent with experience.
Product/service attribute failures may be attributed to internal, stable causes (Oliver 1997;
Olson & Schober 1993). Within this state, consumer satisfaction is expressed as
"acceptance or tolerance" as long as "performance remains within acceptable levels"
(Oliver 1997, 340).
According to Oliver (1997, 340):
This one satisfaction state alone may explain why reports of satisfaction
levels in surveys are inordinately high, approaching 90 percent If a survey
focuses on an ongoing situation, most subjects will be responding from a
satisfaction-as-contentment perspective and the remaining satisfaction
states will become "mixed in" at varying levels of satisfaction/
dissatisfaction. Thus, one should not expect to find high levels of reported
Olson and Schober (1993) define the state of being satisfied with objectively
unsatisfactory living conditions" as the satisfaction paradox (173), the result of a
contradictory motivational status of the individual which represents a cognitive dissonance
(Olson & Schober 1993,188).
This cognitive dissonance consists in the feet that simultaneously the
motivation to change the situation is high, because of realistic appraisal of
the same and the resulting low satisfaction with life and the motivation to
change the situation is low, because the individual has, through the
experience with ineffective poverty programs, learned that he is not in
control of the determinants of his poverty situation and it therefore makes
no sense at all to act (Olson & Schober 1993,188).
Olson and Schober (1993,189) identify four symptoms of learned helplessness in
the case of instrumental coping" (based on Seligman 1978):
1. an internal attribution of failure.
2. a negative expectation of the controllability of the determinants of poverty in the
3. a motivational deficit concerning the willingness to participate in instrumental
coping strategies, and
4. instrumental coping strategies have been perceived to be ineffective for them
and the coping strategies that involve participation in emergency programs are
associated with high effectiveness.
This may account for the high satisfaction levels reported by VLI unassisted and
assisted renter households in the literature. Connie H. Casey (1992) reports that 42
percent of voucher/certificate-assisted householders, 37 percent of public housing
residents, and 26 percent of VLI unassisted householders rate their housing as the best
they have occupied. Similarly, nearly 30 percent of householders in all three groups rated
their neighborhood the best in which they have lived.
Performance problems of unit and location attributes are summarized in Table 2-1,
(Casey 1992). There is little difference in percentage of residents reporting problems
related to unit attributes across the three arrangements. Neighborhood problems are less
for voucher/certificate assisted householders than public housing and unassisted. This is
due to the increased purchasing power and location flexibility provided by
vouchers/certificates (Goering, Stebbins, & Siewert 1995).
Table 2.1 Housing and Neighborhood Problems Reported by Very-Low Income
Householders Within Three Rental Housing Delivery Arrangements (Percentage of
Problem Certificates Housing Unassisted
Signs of rats in last three months 7% 12% 8%
No working toilet at some time in last three months Blown electrical fuses or circuit breakers in last 9% 12% 7%
three months 11% 15% 13%
Holes in floors and walls, cracked walls and peeling paint 16% 20% 24%
Water leakage from inside the structure 22% 18% 17%
Water leakage from outside the structure Crime, noise, traffic, litter, poor local services, undesirable industrial uses, and other people as 12% 8% 14%
neighborhood problems 42% 46% 57%
Unit. Location, and Property
There are three dimensions to a rental housing property: the unit, its location, and
the property management. The unit is a tangible space composed of walls, floors, ceilings,
windows, doors, rooms, closets, cabinets, appliances, plumbing fixtures, and plumbing,
electrical, and heating systems. Location is where the product is delivered, and includes
attributes both present and absent. For example, grocery and stores that provide goods
and services should be present On the other hand, crime and vandalism, trash and litter,
noise and congestion from traffic, and loud neighbors should be absent Property
management performs the services needed to operate and maintain the unit and its
The literature has focused primarily on the relationship between core attributes of
the unit and its location, demographic characteristics of the household, and satisfaction.
Much research on housing satisfaction has confused consumer rating of attribute
performance with satisfaction. According to Oliver (1997), this approach is better than
conducting no research at all" (38). Examining problems with product and/or service
attributes provides useful insights "as to the levels and potential causes of satisfaction, but
does "not explain why the particular feature is a problem (or a benefit) to the consumer"
(Oliver 1997,38). The primary problem with this approach is that it "ignores the
psychological processing of performance," providing little insight in the workings of the
consumer's mind" (emphasis in original, Oliver 1997,39). Consequently, the researcher
cannot know the extent to which performance ratings are related to consumer expectations
(negative discontinuation), regret, or dissonance (Oliver 1997).
Like most functional products, rental housing fulfills lower-level consumer needs.
Attributes such as appliances, plumbing, electrical, and heating systems are "essential but
unprocessed, only capable of causing dissatisfaction when flawed," and when fulfilled,
never contribute to satisfaction (Oliver 1997,152). They are monovalent dissatisfiers
Location is where the unit is delivered to the consumer. The term neighborhood is
not used because, according to Lee, Oropesa, Kanan (1994).
Routine use of "neighborhood" in survey research and in everyday
discourse obscures the "elastic" character of the term (Converse & Presser
1986:18). Unlike cities or counties, neighborhoods typically are neither
formal governmental jurisdictions nor clearly demarcated territorial units.
Rather, they are social constructs named and bounded differently y
different individual (Guest & Lee 1984; Haney & Knowles 1978; Hunter
1974; Lee 1968). This variation is especially marked with respect to
perceived neighborhood scale.
Units with the same attributes may exist in several locations within the same
neighborhood. Differences in location attributes vary from block to block. The term location
is used to "encourage respondents to assign their own scale to the concept" (Lee,
Oropesa, & Kanan 1994, 252).
Location attributes such as the absence of crime, litter and trash, traffic congestion
and noise, and loud neighbors and the presence of grocery and other stores are benefits.
They are monovalent satisfiers that, when frustrated, contribute modest amounts to
dissatisfaction (Oliver 1997,152).
Property Management Performance
Property management is the service dimension of housing consumption. Service
performance affects perception of quality and "service quality is an antecedent of consumer
satisfaction" (Cronin & Taylor 1992,65). Responsibility for property management falls on
the property owner, an employee of the owner, ora professional properly management
agent. In large multi-family apartment properties, a property manager may live and work at
the apartment complex. Depending upon size (total number of units), the property
management may employ several people to perform leasing, rent collection, repair,
maintenance, and custodial services. Property management may also contract for services
such as trash collection, utilities, elevator maintenance, and security. Ultimately,
management is responsible for marketing, administration, maintenance and operations,
and financial management of units and common areas, and for the performance of location
attributes. Property management should be vitally interested in correcting location
problems that reduce marketability and increase operating costs (e.g., security expenses)
(Anthony, Weidemann, & Chin 1990; Basile 1994; Isler, Drury, & Wellom 1971); Isler,
Sadacca, & Drury 1974; Stegman 1972). There is a paucity of research on management
performance. One study of low-income single parents (Anthony, Weidemann, & Chin
1990) identified a statistically significant bivariant correlation between management and
According to J. Joseph, Jr., and Steven A. Taylor (1992), the consumer's
perception of service quality is based on performance measures. How service is delivered
is as important as the quality of the service (Parasuraman, Zeithaml, & Berry 1988;
Zeithaml, Parasuraman, & Berry 1990). For example, stopping a water leak quickly, even if
the repair caused other damage to walls or floors and the patch was temporary, is more
likely to result in high consumer satisfaction than a slower high quality response (Oliver
According to Zeithaml, Parasuraman, & Berry (1990), service quality is composed
of five dimensions: tangibles, reliability, responsiveness, assurance, and empathy.
Tangibles related to the appearance of physical facilities, equipment, personnel, and
literature about the firm and its services. Reliability is related to the performance of
services in a dependable manner and the maintenance of accurate records.
Responsiveness is related to managements willingness to help consumers and provide
prompt service. Assurance is related to employee courtesy and their trustworthiness.
Empathy is related to the staffs ability to give personal attention and have the consumers
best interests at heart Reliability is the most critical dimension, followed by
responsiveness, assurance, empathy, and tangibles (Ziethaml, Parasuraman, & Berry
Demographic Influences on Housing
and Neighborhood Satisfaction
Much of the research on housing and neighborhood satisfaction has correlated
household characteristics with housing and neighborhood satisfaction and switching
behaviors (Cook & Bruin 1993; Galster 1987; Lee, Oropesa, & Kanan 1994; Newman &
Duncan 1979; Speare 1974; Varady 1983). In general, older (generally mid-40s and
above) householders and those who have lived in the same unit for many years reported
higher satisfaction and lower desire to switch frequently regardless of satisfaction level. The
presence of friends and family in the neighborhood increases overall satisfaction with the
unit and the neighborhood. These findings are consistent with Oliver's (1997) description of
consumer responses within a "satisfaction-as-contentment" state.
Higher levels of housing dissatisfaction and increased switching is found among
younger householders (generally age 30 and younger) who are experiencing changes due
to life-cycle stages. Households with a large number of persons generally live in
overcrowded conditions and are among those most dissatisfied with their units. (Cook &
Bruin 1993; Galster 1987; Lee, Oropesa, & Kanan 1994; Newman & Duncan 1979; Speare
1974; Varady 1983).
CHAPTER 2, PART C
LITERATURE REVIEW: COMPLAINING AND SWITCHING
RESPONSES TO DISSATISFACTION
Hirschman (1970) identifies exit and voice as alternative and complementary
consumer responses to dissatisfaction with quality of goods and services. He bridges the
private and public sectors by linking both economic and political theories to the
marketplace. Economic theory relies on exit as the primary means by which consumers
inform management that its performance has slipped and corrective actions are needed to
restore quality in order to retain customers; political theory relies on voice. Consumers
should regard switching and complaining as complementary responses to dissatisfaction;
both may be used to restore satisfaction with the quality of the goods/services they
consume (Hirschman 1970).
Exit and voice dominate separate domains: exit dominates competitive markets
and voice dominates "such primordial human groupings as family, tribe, church, and state,"
where exit through expulsion is an organizational not an individual decision (Hirschman
1970,76). Together, they dominate a third domain, political parties and voluntary
organizations, wherein, participants have a special attachment to an organization known
as loyalty" (Hirschman 1970,77). According to Hirschman (1970), "loyalty holds exit at bay
and activates voice" (78). As previously indicated, Hirschmans (1970) loyalty construct is
not included in this research because the construct has not be defined and operationalized
in a manner that permits its application to the context of this research.
Switching when dissatisfied is the consumers most common experience in the
marketplace, a "secret vote in the anonymity of a supermarket" (Hirschman 1970,16).
This is the sort of mechanism economics thrives on. It is neat-one either
exits or one does not; it is impersonalany face-to-face confrontation
letween customer and firm with its imponderable and unpredictable
elements is avoided and success and failure of the organization are
communicated to it by a set of statistics.
Keaveney (1995) identified "more than 800 critical behaviors of service firms that
caused consumers to switch service firms from which eight general categories were
developed: core service failures, failed service encounters, price, inconvenience,
competition, ethical problems, involuntary switching, and other. The most significant cause
of consumer switching is core service failures such as mistakes, billing errors, and service
catastrophes. Second most significant is service encounter failures, personal interactions
between customers and employees of service firms" (Keaveney 1995,76). More than 17
percent of the subjects in Keaveneys (1995) study switched because of unsatisfactory
responses by employees to services failures: (1) reluctant responses, (2) failure to
respond, and (3) patently negative responses (Keaveney 1995,77). Firms that do not
encourage and facilitate consumer complaining fail to capture important feedback and
satisfaction recovery is unlikely (Best 1981; Hart, Heskett, & Sasser 1990; Heskett, Sasser,
& Hart 1990).
Switching is an active behavior among renter households. More than one-thind
(37%) of all renter households moved in 1989. Table 2.2 compares the percentage of
households reporting a move in the past year by type of housing assistance (Casey 1992).
Table 2.2 Moves in the Past Year (percent of total households)
All Public Voucher- Private VLI
Renter Housing Assisted Project-based Unassisted
Households Households Households Households Households
37% 16% 28% 22% 33%
A lower rate of switching by public housing householders is expected because fewer
alternatives are available. VLI householders in project-based Section 8 units also have
fewer housing options. Interestingly, the percentages of assisted and unassisted VLI
householders switching housing are comparable.
The extent to which switching results in a better home or neighborhood is
summarized in Table 2.3.
Table 2.3 Comparison of Switching Outcomes
Percentage of households that moved within the last 12 months
Current home compared to previous home is: All Renter Households Public Housing Households Certificate/ Voucher- Assisted Households Private Project- based Households VLI Unassisted
Better 44% 73% 52% 49% 44%
Worse Current neighborhood compared to previous neighborhood is: Better 23% 5% 16% 26% 24%
35% 44% 45% 39% 34%
Worse 20% 20% 16% 31% 20%
Proportionally high improvement is expected among householders moving to public
housing because many are moving from very adverse conditions. The proportion of
voucher/certificate assisted householders who find better housing and neighborhoods is
greater than both private project-based and VLI unassisted households. This likely is due
to two factors: greater flexibility of vouchers/certificates compared to project-based
assistance and increased purchase power of voucher/certificate householders compared to
unassisted householders. Nevertheless, the percentage of voucher/certificate
householders securing better housing and neighborhoods by switching seems low.
Hirschman's (1970) second premise is that consumer complaining is an alternative
and complementary means of informing firms that quality has declined. In contrast to
switching, complaining is neither quiet nor impersonal. He describes complaining as
"messy" because, whether quiet or loud, it requires the consumer to express his/her
opinion rather than quietly switch to another provider; it is "political action par excellence"
Complaining is defined as:
Any attempt to change, rather than escape from an objectionable state of
affairs, whether through individual or collective petition to the management
directly in charge, through appeal to a higher authority, or through various
types of actions or protests, including those meant to mobilize public
opinion (Hirschman 1970, 30).
According to Best (1981), "complaining can serve two functions: compensating
individual buyers, and collecting information valuable to all buyers."
Properly monitored consumer complaints, collectively, can serve as a fund
of information to help businesses improve service to customers, to help
government improve law enforcement and regulatory actions, and to help
consumers plan purchases to avoid sellers or particular products and
services with poor records of problem occurrence or complaint resolution
Consumer complaints can also serve a third function, service recovery.
Complaints are opportunities to turn consumer dissatisfaction into satisfaction through
effective service recovery policies and procedures (Hart, Heskett, & Sasser 1990; Heskett,
Sasser, & Hart 1990).
Day (1980) classifies complaints into three broad purposive categories: seeking
redress from the seller, complaining to others about the dissatisfying experience (negative
word-of-mouth), and boycotting the offending product, brand, store, or service. Singh
(1988) examined the validity of these classifications by analyzing behavioral and non-
behavioral complaint behaviors resulting from dissatisfying experiences with grocery stores,
automobile repair shops, medical care providers, and banks and financial services. Singh
(1988,104) proposes the following classification schema for consumer complaint behaviors
(CCB) based on identifying the object toward which the responses are directed:
Voice CCB is directed to objects that are external to the consumer's social
circle (i.e., informal relationships) and are directly involved in the
dissatisfying exchange (e.g., retailer, manufacturer).
Third party CCB includes objects that are external to the consumer, as in
the voice CCB, but they are not directly involved in the dissatisfying
transaction (e.g., Better Business Bureau, legal agencies, newspapers,
etc.)(emphasis in original).
Private CCB objects are not external to the consumer's social net and are
also not directly involved in the dissatisfying experience (e.g., self, friends,
relatives, etc.) (emphasis in original).
Dissatisfied consumers can complain to sellers (voice CCB), third-parties (third
party CCB), and/or privately to family and friends (private CCB). This research examines
the complaint behaviors of assisted renter householders in relation to property
management and third parties such as the PHA, local government agencies, and the news
media. Negative word-of-mouth (private CCB) is not included in this study because
complaints to friends and family members do not directly inform property management or
third parties that product or service performance has declined (Richins 1983; Singh 1988;
1989; 1991; Singh & Wilkes 1995).
According to Singh and Wilkes (1996), consumer complaint responses are
"intermediate behaviors" that increase the probability of achieving goals but do not
guarantee success in that goal attainment (351). Consumer complaining is problematic
due to factors such as lack of control (e.g., a retailer is involved), resource scarcity (e.g.,
lack of time to complete complaint procedures), and environmental contingencies (Singh &
Wilkes 1996, 351). The complexity and costs of complaining make it more likely that higher
income, better educated, and higher occupational prestige consumers will complain and do
so more effectively, especially concerning judgmental deficiencies such as workmanship.
In addition, whites complain more than blacks within each socio-economic category (Best
Orbell and Unos (1972), research on complaining within the context of
neighborhood problem solving examined the "proneness" of black and white residents to
complain in response to dissatisfaction with neighborhood problems (477). They compared
complaint responses of to a hypothetical question related to actions by the city government
that respondents would consider unjust or harmful. Consistent with Hirschman's (1970)
thesis that complaining should increase as market conditions become more monopolistic,
blacks were more likely than whites to complain because switching was not considered a
viable option. In addition, younger blacks were more likely than older blacks to complain
(Orbell & Uno 1972).
Switching and complaining behaviors are affected by consumer alienation; attitude
toward switching/complaining; prior switching/complaining experience; and expectancy-
value judgments, and the assessment of the costs and benefits of switching/complaining
(Allison 1978; Andreasen 1984,1985; Day 1984; Hirschman 1970; Singh 1989,1990a,
1990b, 1991; Singh & Wilkes 1996). These antecedents were selected based on Singh
and Wilkes (1996) criteria for including them in their model of complaint response
First, we preferred antecedents that were rooted in relatively strong
theoretical frameworks. For instance, the inclusion of expectancy value
(EV) constructs was desirable due to their roots in political and economic
frameworks (e.g., Hirschman 1970). Second, we favored antecedents that
had been used individually in past research but whose relative or
differential impact was largely unknown. For instance, attitude toward
complaining and attribution of blame have been tested individually, but few
studies have used both constructs within a single study to unravel their
The concept of human alienation has been variously defined as powerlessness,
meaninglessness, normlessness, isolation, and self-estrangement (Seeman 1959). Allison
(1978) determined that within the context of the marketplace, "the concept of alienation
was more homogenous when treated a one overall syndrome than when broken down into
variants" (569). Consumer alienation from the marketplace is a unidimensional construct
defined as a general feeling of separation from the marketplace. According to Allison
Such a state includes a lack of acceptance of or identification with market
institutions, practices, and outputs as well as feelings of separation from
the self when one is involved in the consumption role. The marketplace
encompasses all institutions engaged in the distribution of goods and
services to the consumer as well as all facilitating agencies.
During the 1970s, research on market practices in poor neighborhoods determined
that low-income black consumers in Philadelphia were shopping outside their market area
because local stores were charging the more and offering limited selections (Goodman
1973). A study of business and consumer practices among low-income Mexican-
Americans in East Los Angies revealed that they paid more for groceries, furniture,
appliances, clothing, and automobiles than consumers in other areas of the city (Sturdivant
1973). A similar study in Denver revealed that most residents of low-income
neighborhoods shopped for groceries outside their community to avoid the higher prices
charged by smaller neighborhood stores (Berry & Solomon 1973).
Based on his study of the low-income marketing system in Los Angeles following
the 1965 Watts riots, Sturdivant (1973,264) concludes:
The most direct contact between the poor and the business community is
at the retail level. The greatest opportunity to assist and revolutionize the
daily lives of the poor rests in the retailing communities serving poverty
areas. Very little may be gained if they are confronted with a shopping
situation that generally offers them higher prices, inferior merchandise,
high-pressure selling, hidden and inflated interest charges, and a
degrading shopping environment Such conditions are closely related to
the frustrations that have produced the spectacle of looted and burned
stores throughout the nation (emphasis in original).
Research by Pruden and Longman (1972) on alienation within a market context
examined and compared the extent of anomie, attitudes toward marketing and belief in
government control among Mexican-Americans and Anglo-Americans. They determined,
NOTE TO USERS
Page(s) not included in the original manuscript
are unavailable from the author or university. The
manuscript was microfilmed as received.
This reproduction is the best copy available.
complain to the service provider, third parties, or friends and family (negative word-of-
mouth). They conclude that:
Alienated consumers apparently view sellers as relatively less responsive,
tend to have more positive attitudes toward complaining, and perceive
greater benefits in exit and negative word-of-mouth responses. Thus, for
managers, alienated consumers pose problems as they appear to defy
recovery and propagate their dissatisfaction (through negative word-of-
mouth). By definition, alienation is less amenable to direct influence by
individual managers. Collective action at the industry level may be
necessary. Industries such as automotive repair, medical care, and
banking services may especially wish to consider collective actions to
manage consumer alienation (Singh & Wilkes 1996, 364).
Attitude Toward Complaining and Switching
Attitude toward complaining is "conceptualized as the overall affect of 'goodness or
'badness' of complaining to sellers and is not specific to a given episode of dissatisfaction"
(Singh & Wilkes 1996,353). Attitudes can develop based on information provided by
others without any prior experience with a product or service, affecting the consumer's
expectations. Negative word-of-mouth is an attempt by dissatisfied consumers to influence
the attitudes of friends and family (Richins 1987, Singh 1988). It is reasonable to expect
the consumer's attitude toward complaining to affect his/her intention to complain in the
same way that attitudes affect the consumer's intention to repurchase a product or re-
patronize a service (Oliver 1997). The consumer's attitude toward switching has not been
specifically measured in the marketing research literature.
Expectancy Value of Complaining
Dissatisfied consumers can switch, complain, do both or neither. Their choice is
based upon their judgment of the potential costs and benefits of a particular choice, as well
as the likelihood that desired results will be achieved (Day 1984, Singh & Wilkes 1996).
When dissatisfied with products or services, consumers prefer switching over complaining
because latter is more time-consuming than switching and requires development to be
effective (Hirschman 1970).
Racial discrimination's adverse influence on switching is well documented in the
housing mobility literature (Orbell & Uno 1972; South & Deane 1993; St John, Edwards, &
Wenk 1995; Varady 1983). Residential racial discrimination "presents another cost to
blacks in the form of a diminished ability to seek new-and, for many, improvedliving
quarters" (Deane & South 1993). Residential racial discrimination creates conditions of
loose monopoly within the rental housing market structure.
Prior Complaining and Switching Experience
According to Singh and Wilkes (1996):
Prior experience is posited to directly influence EV [expectancy value]
judgments and attitudes. The extent (frequent/infrequent) of past
complaining experiences becomes assimilated into an individuals attitude
toward complaining. Moreover, prior experiences affect an individual's
cognition about, for instance, how a retailer or manufacturer probably
would respond to voiced complaints and the associated costs and/or
This is consistent with Hirschman's (1970) premise that switching and complaining
behaviors are affected by the sensitivity of firms to consumer switching and/or complaining
behaviors. Although Hirschman (1970) theorized consumer switching would decline when
market structures became more monopolistic, Andreasen (1984,1985) and Singh (1950)
determined that a higher percentage of consumers switched in response to dissatisfaction
with medical care (monopoly with competition market structure) than groceries (competitive
market structure). Switching is easier than complaining, especially when social norms and
customs that influence patient-physician and client-attorney relationships discourage
complaining (Andreasen 1984,1985).
The literature reveals that satisfaction in a continuous consumption experience,
such as housing rental, is likely to be high despite specific unsatisfactory episodes,
because "expectations have become passive and are not actively processed" (Oliver
1997). Consequently, consumer satisfaction measures must tap the dimensions of
satisfaction: overall evaluation of performance and quality, need fulfillment, expectations,
cognitive dissonance, regret, attribution, and emotions, supporting the need for a multi-
dimensional measure in context of Section 8 voucher/certificate rental housing and an
examination of unit, location, and property management performance as antecedents of
overall property satisfaction.
Switching in response to dissatisfaction, the dominant behavior, does not inform
sellers that product or service performance has deteriorated. Switching in response to
dissatisfaction with unit performance is more likely than switching in response to
unsatisfactory location performance. The literature indicates that household life-cycle
changes are the primary cause of unit and neighborhood dissatisfaction. The relationship
between property management performance and overall property satisfaction and switching
is not substantively addressed by the research literature.
Effective consumer complaining benefits the seller, other consumers, and the
dissatisfied consumer. Consumer complaining, when supported by third-parties, can be
more effective than regulations in maintaining product and service quality. No research
literature on consumer complaining in response to dissatisfaction within the
voucher/certificate arrangement was identified.
Consumer satisfaction and consumer responses to dissatisfaction have been
extensively researched in the business marketing literature. This review has examined
numerous studies that explicitly define and measure the satisfaction, complaining, and
switching constructs of Hirschman's (1970) theory. These measures are applicable to the
voucher/certificate rental housing context The domains of study for this research and the
hypothesized relationships between the domains follow.
The following relationships are hypothesized between overall property satisfaction
and unit location, and management attribute performance and property management
service performance SERV/PERF tangibles, reliability, responsiveness, assurance, and
Hypothesis 1: The more positive unit attribute performance, the greater overall
Hypothesis 2: The more positive location attribute performance, the greater overall
Hypothesis 3: The more positive property management attribute performance, the
greater overall satisfaction with the property.
Hypothesis 4a: The more positive property management service performance
SERV/PERF tangibles, the greater overall property satisfaction.
Hypothesis 4b: The more positive property management service performance
SERV/PERF reliability, the greater overall property satisfaction.
Hypothesis 4c: The more positive property management service performance
SERV/PERF responsiveness, the greater overall property satisfaction.
Hypothesis 4d: The more positive property management service performance
SERV/PERF assurance, the greater overall property satisfaction.
Hypothesis 4e: The more positive property management service performance
SERV/PERF empathy, the greater overall property satisfaction.
Figure 2.1, Hypothesized Model of Consumer Satisfaction Within the Voucher/Certificate
Rental Housing Market, presents hypotheses 1 through 4e.
Complaining and Switching Responses to Dissatisfaction
The following relationships are hypothesized between overall property satisfaction
and the likelihood of complaining to property management likelihood of complaining to third
parties, and switching intent
Hypothesis 5a: The more positive overall property satisfaction, the lower the
likelihood of complaining to property management.
Hypothesis 5b: The more positive overall property satisfaction, the lower the
likelihood of complaining to third parties.
Hypothesis 5c: The more positive overall property satisfaction, the lower switching
The following relationship is hypothesized between the likelihood of complaining to
third parties and the likelihood of complaining to property management
Hypothesis 6: The greater the likelihood complaining to property management, the
greater the likelihood of complaining to third parties.
The following relationships are hypothesized between consumer alienation toward
the marketplace and attitude toward complaining, expectancy value of complaining to
property management and third parties, likelihood of complaining to property management
and third parties, switching intent, and attitude toward switching.
Hypothesis 7a: The greater consumer alienation, the more positive attitude toward
Hypothesis 7b: The greater consumer alienation, the less positive expectancy-
value of complaining to property management
Hypothesis 7c: The greater consumer alienation, the more positive expectancy-
value of complaining to the housing authority.
Hypothesis 7d: The greater consumer alienation, the lower the likelihood of
complaining to property management
Hypothesis 7e: The greater consumer alienation, the greater the likelihood of
complaining to third parties.
Hypothesis 7f: The greater consumer alienation, the greater the likelihood of
Hypothesis 7g: The greater consumer alienation, the more positive attitude toward
Attitude Toward Complaining
The following relationships are hypothesized between attitude toward complaining
and the likelihood of complaining to property management and third parties.
Hypothesis 8a: The more positive attitude toward complaining, the greater the
likelihood of complaining to property management
Hypothesis 8b: The more positive attitude toward complaining, the greater the
likelihood of complaining to third parties.
Prior Complaining Experience
The following relationships are hypothesized between prior complaining experience
and the expectancy value of complaining to property management and third parties,
attitude toward complaining, and the likelihood of complaining to property management and
Hypothesis 9a: The greater prior complaining experience to property management,
the more positive the expectancy-value of complaining to property management
Hypothesis 9b: The greater prior complaining experience to property management,
the greater the likelihood of complaining to property management.
Hypothesis 9c: The greater prior complaining experience to property management,
the more positive attitude toward complaining.
Hypothesis 9d: The greater prior complaining experience to third parties, the more
positive attitude toward complaining.
Hypothesis 9e: The greater prior complaining experience to third parties, the more
positive the expectancy-value of complaining to the housing authority.
Hypothesis 9f: The greater prior complaining experience to third parties, the
greater the likelihood of complaining to third parties.
Expectancy Value of Complaining
The following relationship is hypothesized between the expectancy values of
complaining to property management and the likelihood of complaining to property
Hypothesis 10a: The greater the expectancy-value of complaining to property
management, the greater the likelihood of complaining to property management
The foliowing relationship is hypothesized between the expectancy value of
complaining to the housing authority and the likelihood of complaining to third parties.
Hypothesis 10b: The greater the expectancy-value of complaining to the housing
authority, the greater the likelihood of complaining to third parties.
Attitude Toward Switching
The following relationship is hypothesized between attitude toward switching and
Hypothesis 11: The more positive attitude toward switching, the greater the intent
Prior Switching Experience
The following relationships are hypothesized between prior switching experience
and switching intent and attitude toward switching.
Hypothesis 12b: The greater prior switching experience, the more positive attitude
Figure 2.2, Hypothesized Model of Consumer Complaining and Switching Responses to
Dissatisfaction Within the Voucher/Certificate Rental Housing Market, presents hypotheses
5a through 12b.
The definition and measurement of each domain is described in the following
chapter. Also described are the study site, study population, and the secondary and
primary research methods employed to collect data.
i: The greater prior switching experience, the greater the intent to
Figure 2.1 Hypothesized Model of Consumer Satisfaction
Within the Voucher/Certificate Rental Housing Market
Property Management Service Performance SERV/PERF
Figure 2.2 Hypothesized Model of Consumer Complaining and Switching Responses to
Dissatisfaction Within the Voucher/Certlflcate Rental Housing Market
RESEARCH DESIGN AND METHODOLOGY
Location of Study and Participants
Three PHAs participated in the study: Richmond (Virginia) Redevelopment and
Housing Authority, Newport News (Virginia) Housing Authority, and the Greensboro (North
Carolina) Housing Authority. The rental markets within all three jurisdictions were favorable
for consumer switching, an assessment was based on market vacancy rates for units
affordable to voucher/certificate-assisted households. Section 8 voucher/certificate
programs at all three were managed and administered in a relatively uniform manner, an
assessment based on review of program policies, procedures, and interviews with program
staff. None of the PHAs operated special mobility programs to encourage and facilitate
moving to areas of low minority or poverty concentration. All three PHAs maintained similar
programs to inform owners about the Section 8 voucher/certificate program and none
conducted special programs to facilitate switching (Goering, Stebbins, and Siewert 1995;
DeLancer interview 1997; Madison interview 1997; Mom's interview 1997).
Executive Directors and senior staff were interested in improving the quality of their
programs and services and were willing to support this research by allocating staff
resources to distribute and collect survey information and materials. In addition, the
Richmond Redevelopment and Housing Authority assisted in the identification and
selection of three groups of voucher/certificate-assisted households to participate in focus
groups and pre-testing the survey instrument
Greensboro (1995) reported that nine percent of units affordable to households
with incomes below 80 percent median income (low-income) were vacant Richmond
(1995) reported a rental vacancy rate of 8.7 percent but did not differentiate based on unit
affordability by low-income households. Richmond (1995) reported 20,741 affordable units
for its 7,889 low-income households. Newport News (1995) reported that 11 percent of
vacant units were affordable by moderate income households (incomes below 80 percent
of area median family income). According to PHA staff, voucher/certificate-assisted
households had not reported any difficulty locating or leasing units (DeLancer interview
1997; Madison interview 1997; Morris interview 1997).
The number of owners renting units to voucher/certificate-assisted households
indicates the concentration of units by owner in the market, a measure of loose monopoly
conditions. The greater the number of assisted households in smaller properties, the
greater the potential number of owners serving the market Nationally, 31 percent of
Section 8 voucher/certificate-assisted households occupied units in properties comprising
1-49 units. In Richmond, half (52%) were situated in properties composed of one to 49
units. In Greensboro, North Carolina, nearly 80 percent of the units were situated in
properties composed of one to 49 units. In Newport News, Virginia, approximately 70
percent of the units were situated in properties composed of one to 49 units. These
percentages indicate that there is potentially less ownership concentration in these PHA
jurisdictions compared to the nation.
Population and Sampling
The population is voucher/certificate-assisted householders. A householder is the
person in whose name the house or apartment is rented. Data reported for householders
refers to individuals. The term household refers to all of the persons occupying a unit,
therefore, household characteristics describe the occupants as a group.
Householders who expressed an interest in participating were provided an
Informed Consent Form that explained the purpose of the research and briefly described
the survey instrument. The form emphasized the voluntary and confidential nature of the
survey that participants were not required to answer every question or complete the survey,
and were free to discontinue at any time. The form stated there was no relationship
between the decision to participate and recertification. It included the name, address, and
telephone number of the Office of Academic Affairs at the University of Colorado at Denver,
and participants were encouraged to contact the University if they had any questions
concerning their rights. Participants received a copy of the signed consent form from the
PHA representative. Consent forms were collected separately from surveys and were
transmitted directly to the Norwest Public Policy Research Center, Graduate School of
Public Affairs, University of Colorado at Denver.
Participants were given a survey with an envelope addressed to the researcher at
the University of Colorado at Denver, marked confidential, with instructions to place the
completed survey in the envelope, seal it, and returning it to the PHA representative. They
were provided a pen or pencil, if needed, to complete the survey consent form and
instrument and a quiet place to work, outside the presence of the person conducting the
Two focus group interviews were conducted to refine the survey instrument.
Twelve householders from Richmond were randomly selected to participate in focus group
interviews. A letter was mailed to householders who had recently completed their annual
recertification and would not be participating in the survey. The letter invited respondents
to participate in one of two focus group interviews. The first fourteen respondents that
called the PHA were selected for the interviews. Each participant was provided an
Informed Consent Form. At the conclusion of the interview, each participant was provided
a copy of the Consent Form and paid a $20.00 stipend.
The interview concerned the definition of unit, location, management company, and
property. The terms unit versus apartment were discussed out of concern that
apartment was too limiting and did not include houses or condominiums. There was no
confusion over unit in the context of rental housing within the voucher/certificate
The researcher was concerned that neighborhood was too broad, encompassing
areas beyond the more immediate location (Guest & Lee, 1984), which were of primary
interest in this research.
This distinction was examined in the focus group interviews. Although the meaning of
location was clear to the participants, there was no consensus that location was distinct
Management company" was changed to property management" as a result of the
focus groups. Participants expressed confusion between management company, used
throughout the survey to define the party or entity responsible for the operation of the
property, and the PHA management
Finally, the concept property was discussed as an overarching term that
characterized the overall housing consumption experience: the unit (product), location
(product delivery), and management (service). There was no strong consensus on any
single term that encompassed all three rental housing dimensions. Content analysis of the
tape recorded sessions revealed the consistent use of property when discussing the unit,
its location, and management.
Survey Instrument Pre-testing
The survey instrument was pre-tested by 10 randomly selected householders from
the Richmond PHA. A letter was mailed to approximately 30 householders who had
recently completed their annual recertification, and not be participating in the survey,
inviting the first twelve respondents to pre-test a survey instrument on consumer
satisfaction and consumer responses to dissatisfaction. Each participant was provided an
Informed Consent Form and copy of the signed form. The pre-test resulted in a slight
reorganization of the survey questions. No other issues affecting the design or wording of
the instrument were identified. Each participant was paid a $20.00 stipend.
Domain Definitions and Measurements
Unit Attribute Performance
The unit is a private dwelling space leased and occupied by a household. The
space consists of a living room, kitchen, bathroom, and bedroom(s). The unit is the "good"
or product purchased. Unit performance is based on the consumer's comparison of
expected performance to actual performance of the core attributes of the unit appliances
that work; plumbing, electrical, and heating systems that operate properly; walls, floors, and
ceilings that are in sound condition; and the absence of roaches, mice and other vermin in
the unit (based on Casey 1992, Galster 1987, Newman and Duncan 1979).
Location Attribute Performance
Location is the place where the unit (the "good") is "delivered." The location is
comprised of adjacent and nearby residential, commercial, and industrial activities. In
general, the core attributes of location for an urban dwelling unit are the absence of crime,
vandalism, noisy neighbors, litter and trash, traffic congestion and street noise. The
presence of nearby grocery and other stores is a positive attribute of location (based on
Casey 1992, Lee 1981, Newman and Duncan 1979, Varady 1983).
Property Management Attribute Performance
Property management is the "service" component The tasks and responsibilities
of property management are generally those of the owner or a professional agent who
operate the rental business and maintain the unit and its immediate location. Property
management hires office, maintenance, and janitorial staff; contracts for repair, security,
trash removal and other services; and leases the unit, collects rent enforces leases and
house rules, makes routine repairs to the unit, and maintains the common areas around the
Property Management Service Performance SERV/PERF
Service performance was measured using Cronin and Taylors (1992) service
quality performance (SERV/PERF) scale, based on Parasuraman, Zeithaml, and Berry
(1988) and Zeithaml, Parasuraman, and Berry (1990). Five dimensions of service quality
are measured by this scale: Tangibility, Reliability, Responsiveness, Assurance, and
Perception of service quality is affected by interactions with service provider
employees, the appearance of the service provider's office, equipment, and employees,
and the consumers participation in the service experience (Bitner 1990; Zeithaml,
Parasuraman, and Berry 1990). Service quality is more difficult to measure than product
quality because outcomes are not the only measure of performance. Consumers evaluate
service quality based on the manner in which services are delivered, including the behavior
and attitudes of employees that perform services (Oliver 1997; Parasuraman, Zeithaml, and
Berry 1988; Zeithaml, Parasuraman, and Berry 1990).
Overall Property Satisfaction
Overall housing satisfaction is the global experience with the unit, its location, and
the property management Overall housing satisfaction develops over time and is not
significantly influenced by an individual event such as a broken water pipe, provided that
the broken pipe is promptly and properly repaired. Overall housing satisfaction is
measured using the "Universal Product Or Service Consumption Satisfaction Scale"
developed and tested by Oliver (1997, 343).
Likelihood of Complaining to Property Management
and Third Parties
A complaint is the verbal expression of dissatisfaction with the unit, its location,
and/or the property management Complaints are goal oriented; the consumer seeks to
achieve some desirable outcome. Complaints may be directed toward management and/or
third parties (e.g., the public housing authority). The likelihood of complaining to property
management and the likelihood of complaining to third parties are both measured.
Critical incident technique method was used to construct these measures.
Householders were asked to recall a dissatisfying experience with their unit, its location, or
the property management before responding to the seven scale items. Critical incidents
are defined as any event, combination of events, or series of events between the
householder and the property management that cause the householder to choose a
particular complaint response (Singh and Wilkes 1996).
Switching intent is the degree to which a householder has made a decision to
move or is contemplating a move at his/her next opportunity (generally one-year).
Voucher/certificate-assisted households must give notice approximately 90 days prior to
the end of their lease. In most instances, they either renew their lease or give notice of
their intent to move at the time of their annual recertification for program eligibility. Since
this survey was administered at the time of annual recertification, the majority of
respondents indicating their intent to move in the next 90 days had acted on their decision
by notifying their current property manager and, perhaps, searching for another unit
Otherwise, the householder's decision to move will be delayed from one month to one year,
depending upon the terms of the lease.
Consumer alienation is a "generalized affect" representing "an individuals global
feelings about the marketplace, behavior of firms/business, and consumption of products
and services" (Singh and Wilkes 1996,354).
Attitude Toward Complaining
Attitude toward complaining was defined as the overall affect of "goodness" or
"badness" of complaining to sellers (Singh and Wilkes 1996, 353).
Prior Complaining Experience
Prior complaining experience was defined as the frequency of complaints about the
unit, location, and/or management to the property management and/or the housing
authority. The time frame for recalling prior complaints was the past year, roughly the
period since the householder initially certified or recertified for participation in the
voucher/certificate program, to ensure the incident related to the present unit, location, or
Expectancy Value of Complaining
The expectancy value of complaining was the individual's cost-benefit analysis of
complaining. Cost was the perceived difficulty of complaining, and benefit was the
perceived chances that complaining would bring the desired result (Singh and Wilkes
1996). Complaints could be directed to the property management and/or the housing
authority (third party), thus they were measured separately.
The critical incident technique method was used to construct this measure.
Householders were asked to recall a dissatisfying critical incident (e.g., an individual,
combination, or series of events between the householder and the property management
that cause the householder to expect and value certain responses by the property
management and the housing authority third party (Keaveney 1995) before answering.
Expectancy Value of Complaining to Property Management. The expectancy-
value of complaining to property management was measured by 6 items: three items for
expectancy and three items for value (Singh 1990, based on Day 1984 and Bagozzi 1982).
Expectancy Value of Complaining to Housing Authority. The expectancy-value of
complaining to the housing authority was measured by six items: three items for
expectancy and three items for value(Singh 1990, based on Day 1984 and Bagozzi 1982).
Attitude Toward Switching
Attitude toward switching is defined as the overall affect of "goodness" or
"badness" of switching to another product provider (based on Singh and Wilkes 1996,353).
Prior Switching Experience
Prior switching experience was the extent to which a householder had moved in
the past five years, a time frame based on Casey's (1992) finding that 67 percent of
voucher/certificate assisted households had been in their unit five years and less. Prior
switching experience was measured with one frequency based item: In the past five years,
how many times have you moved from one property to another?
A total of 433 surveys were obtained from the 1,556 householders recertified over
a six and one-half month period, a 28 percent response rate. The representativeness of
the sample is summarized in Tables 4.1 and 4.2. The population is the total number of
voucher/certificate-assisted households within each PHA. The percentage is the
proportion of the total number of assisted households in all three PHAs.
Table 4.1 PHA Representativeness
Greensboro PHA Population SamDle
Number 1,088 108
Percent 25% 25%
Number 1,357 177
Percent 34% 42%
Newport News PHA
Number 1,859 148
Percent 47% 34%
Total 3,989 433
Table 4-2 Demographic Representativeness
Female 92% 95%
Age 24 and under 5% 7%
Ages 25 to 44 64% 67%
Ages 45 to 61 17% 10%
Age 62 and over 11% 5%
Table 4.2 Continued
Black 93% 93%
Hispanic 1% 2%
White 6% 3%
Asian and Other 1% 1%
Household Size 2.8 3.0
Analysis of the Data
All survey questions used 7-point Likert scales. For attitude questions (unit,
location, and management attribute performance, management service performance
SERV/PERF, and overall property satisfaction), 1=Strongly Agree and 7=Strongly
Disagree. For behavioral questions (likelihood of complaining to property management
and third parties, and switching intent), 1=Very Likely and 7=Very Unlikely. For questions
concerning frequency of prior complaining, 1=Very Often and 7=Never. When survey were
questions worded so that a high response had a negative connotation, rather than the
typically positive, and loaded onto a factor, that response was reverse coded when the
scale was computed (indicated by [R] in the tables below).
For each measure, a factor analysis was conducted using all items with the
potential to be related (see Tables 4.3 through 4.11 and 4.14 through 4.27). Items for
which a factor analysis yielded loadings of .600 and higher are listed first, with factor
loadings on the left and means for the overall sample as well as of the three PHAs on the
right. The scale score, computed by adding all items that loaded significantly and dividing
the sum by the number of items follows these, along with the scale alpha reliability. Based
on wording, some items on the scale had to be reverse scored. Any items deleted from the
scale are reported at the bottom. Analyses of variance were performed on all items as well
as scale scores to identify significant (p<.05) differences among PHAs. Footnotes identify
and explain any significant differences.
Unit Attribute Performance
The four items that related to unit attribute performance concerned problems with
appliances, roaches and mice, the electrical system, and leaks from outside. Cracks and
holes in the walls and floors, frequent plumbing problems did not load on the scale. Unit
comfort loaded high on the scale but its inclusion reduced alpha reliability (Table 4.3).
Location Attribute Performance
Three negatively constructed statements related to location attribute performance:
crime and vandalism, litter and trash, and traffic noise and congestion around unit. Lower
overall scale scores indicate respondents were more positive about location attributes
(e.g., agreement with positive statements) (Table 4.4).
Property Management Attribute Performance
The four items that related to tangible services were availability of emergency
repairs, repairs being done right the first time, feeling safe inside and outside the unit, and
adequate lighting. As with location attribute performance, this scale concerned physical
consequences; people management was not related (e.g., lease terms, house rules)
Service Performance SERV/PERF
Five dimensions of service performance are measured: tangibles, reliability,
responsiveness, assurance, and empathy. All four items concerning appearance of the
property, equipment, and employees related to perception of service performance loaded
on the tangibles scale (e.g., the appearance of the property is in keeping with the type of
service provided by the property management) (Table 4.6). Rve items relate to reliability
loaded on the scale, which primarily concerned property management's keeping its
promises, keeping accurate records, and dependability (Table 4.7). Five items related to
responsiveness, especially management's accessibility, promptness, and willingness to
help residents (Table 4.8). The four items on the assurance scale concerned employee
politeness, management and employee trustworthiness, and sympathetic or reassuring
responses to resident problems (Table 4.9). Five items related to empathy, especially
managements and employees attentiveness to problems and interest in the needs of
residents (Table 4.10).
Overall Property Satisfaction
Eight items, seven related to overall performance evaluation and quality, purchase
evaluation, positive affect, success attribution, need fulfillment, and satisfaction. One item,
related to failed expectations. (Table 4.11)
Satisfaction Estimated Path Coefficients and Hypothesized Results
The relationships between Overall Property Satisfaction and each of the attribute
and SERV/PERF domains were tested through multiple regression analyses. The
coefficients are presented in Figure 4.1 and Table 4.12, Hypothesized Model of Consumer
Satisfaction Within the Voucher/Certificate Rental Housing Market
Unit Attribute Performance. There was no significant relationship between unit
attribute performance and overall property satisfaction. Hypothesis 1 is not supported.
Location Attribute Performance. There is a significant positive relationship
between location attribute performance and overall property satisfaction, r=.18, p.=.004,
supporting Hypothesis 2 that the more positive location attribute performance, the greater
overall property satisfaction.
Property Management Attribute Performance. There is a significant positive
relationship between management attribute performance and overall property satisfaction,
r=.26, p=.001, supporting Hypothesis 3 that the more positive property management
attribute performance, the greater overall satisfaction with the property.
Property Management Service Performance SERV/PERF. There was a marginally
significant positive relationship between SERV/PERF tangibles and overall property
satisfaction, r=.13, p.=.058. Hypothesis 4a is marginally supported; the more positive
property management service performance SERV/PERF tangibles, the greater overall
There was a significant positive relationship between SERV/PERF reliability and
overall property satisfaction, r=.19, p.=.028, supporting Hypothesis 4b that the more
positive property management service performance SERV/PERF reliability, the greater
overall property satisfaction.
There was no relationship between SERV/PERF responsiveness, assurance, or
empathy and overall property satisfaction. Hypotheses 4c, 4d, and 4e were not supported.
Complaining and Switching Responses to
Likelihood of Complaining to Property Management
The two items that affected the likelihood of complaining both asked the likelihood
they would complain. The question that asked the likelihood they would not complain did
not load on the factor (Table 4.13).
Likelihood of Complaining to Third Parties
All four items related to the likelihood of complaining to the housing authority, local
news media, government agencies, or taking legal action against the property
management loaded on this factor (Table 4.14).
Only item, likelihood of moving in the next 90 days, emerged as a good indicator of
switching intent (Table 4.15).
Six items loaded on the alienation factor. The concerned the extent to which
businesses were honest and concerned about consumers, and the consumer's influence
over products sold (Table 4.17). Items were stated negatively but reverse scored for the
overall scale, such that higher scale scores reflect higher alienation.
Attitude Toward Complaining
The three items that expressed attitude toward complaining concerned
participants own behavior, not that of others. What others do was not systematically
related (Table 4.18).
Prior Complaining Experience
All three items asking the frequency of complaining during the past year, to
property management about problems the unit, its location, or employees. (Table 4-19)
Three items related to the frequency of complaining during the past year, to the
housing authority, another government agency, or the new media loaded on the factor
Expectancy Value of Complaining to Property Management
Each of three expectancy items was multiplied by a parallel value item related to
property management taking action to solve a problem and give better service, benefiting
the respondent and other residents. The three products were summed and divided by
three to compute the expectancy value. (Tables 4.21,4.22, and 4.23).
Expectancy Value of Complaining to Housing Authority
Each of two expectancy items was multiplied by a parallel value item related to
housing authority taking action to solve problems. The two products were summed and
divided by two to compute the expectancy value (Tables 4.24, 4.25, and 4.26).
Figure 4.1 Hypothesized Model of Consumer Satisfaction
Within the Voucher/Certificate Rental Housing Market
Dashed lines between domains Indicate hypothesized relationships that wore not significant.
Attitude Toward Switching
Interesting, attitude toward complaining did not include the behavior of others,
while items that determined attitude toward switching concerned specifically attitude toward
other people who switched. After reverse coding, higher scores on attitude toward
switching reflected disapproval of people who do (Table 4.27).
Prior Switching Experience
Prior switching experience was measured by responses to one item, How many
times have you moved in the past five years? (Table 4.28)
Complaining and Switching Responses to Dissatisfaction
Estimated Path Coefficients and Hypothesized Results
The strength and direction of the relationship between CRE Property Management,
CRE Third Parties, and Switching Intent and Overall Property Satisfaction, prior
experience, attitudes, consumer alienation, and the expectancy value of complaining was
tested through multiple regression analyses. The coefficients are presented in Figure 4.2
and Table 4.29, as the Hypothesized Model of Consumer Complaining and Switching
Responses to Dissatisfaction Within the Voucher/Certrficate Rental Housing Market.
There was no significant relationship between the likelihood of complaining to
property management or third parties and overall property satisfaction. Hypotheses 5a and
5b were not supported.
Hypothesis 5c, a negative relationship between switching intent and overall
property satisfaction was supported, r=-.48, p.=.000. The more positive overall property
satisfaction, the lower switching intent.
Complaining Hierarchy. There was a significant positive relationship between the
likelihood complaining to property management, the likelihood of complaining to third
parties, r=.25, p.=.000. Hypothesis 6 is supported.
Consumer Alienation. A significant positive relationship between consumer
alienation and attitude toward complaining, consistent with Hypothesis 7a, r=-.25, p.=.002.
There is also a positive relationship between alienation and the likelihood of complaining to
third parties, consistent with Hypothesis 7e, r=-.20, p.=.002. As alienation increases,
likelihood of complaining to third parties increases. Contrary to Hypothesis 7g, as
alienation increases, so does attitude toward switching, r=.19, p.=.001.
Among hypotheses not supported were: 7b, the greater consumer alienation, the
less positive expectancy-value of complaining to property management 7c, the greater
consumer alienation, the more positive expectancy-value of complaining to the housing
authority; 7d, the greater consumer alienation, the lower the likelihood of complaining to
property management; and 7f, the greater consumer alienation, the greater the likelihood
Attitude Toward Complaining. Attitude toward complaining was not correlated with
the likelihood of complaining to property management (Hypothesis 8a). There was a
positive relationship between attitude toward complaining and the likelihood of complaining
to third parties, supporting Hypothesis 8b that the more positive attitude toward
complaining, the greater the likelihood of complaining to third parties, r=.17, p.=.006.
Prior Complaining Experience. There was a negative relationship between prior
complaining experience to property management and the expectancy value of complaining
to property management, such that past complaining experience was correlated with lower
expectancy value, r=-21, p.=.001. This contradicts Hypothesis 9a that complaining
experience would increase the prospects for successful outcomes, and, therefore,
complaint-experienced householders would be more likely to complain.
There is a positive relationship between prior complaining experience to property
management and the likelihood of complaining to property management, supporting
Hypothesis 9b that the greater prior complaining experience to property management, the
greater the likelihood of complaining to property management, r=.15, p.=.044.
There were no significant relationships between prior experience complaining
experience to property management or third parties and attitude toward complaining.
Hypotheses 9c and 9d were not supported.
The relationship between prior experience complaining to third parties and the
expectancy value of complaining to the housing authority was negative, contradicting
Hypothesis 9e that the greater prior complaining experience to third parties, the more
positive the expectancy-value of complaining to the housing authority, r=.-17, p.=.006.
The relationship between prior complaining experience and the likelihood of
complaining to third parties was significant and positive, supporting Hypothesis 9f that the
greater prior complaining experience to third parties, the greater the likelihood of
complaining to third parties, r=.24, p.=.000.
Complaining Expectancy Value. There was a positive relationship between the
expectancy value of complaining and the likelihood of complaining to property
management, consistent with Hypothesis 10a that the greater the expectancy-value of
complaining to property management, the greater the likelihood of complaining to property
management, r=.28, p.=.000. There was a positive relationship between the expectancy
value and the likelihood of complaining to the housing authority or to third parties,
consistent with Hypothesis 10b that the greater the expectancy-value of complaining to the
housing authority, the greater the likelihood of complaining to third parties, r=.14, p.=.030.
Attitude Toward Switching. There was negative relationship between attitude
toward switching and switching intent, counter to Hypothesis 11 that a positive attitude
toward switching would increase switching intent; instead the data show that positive
attitude is associated with a decreased intent, r=-.12, p.=.039.
Prior Switching Experience. There was no relationship between prior switching
experience and switching intent or attitude toward switching. Neither Hypothesis 12a or
12b is supported: the greater prior switching experience, the greater the intent to switch
and the more positive attitude toward switching.
Demographic Influences on
Three demographic variables were added to path analyses: age, education, and
household size. In relation to overall property satisfaction, no significant association was
identified. In relation to consumer complaining and switching, two demographic variables
were significant age related to Attitude Toward Complaining, r=.15, p.=.015 and
household size related Likelihood of Complaining to Third Parties, r=-.15, p.=.015.
Coefficients for Satisfaction are presented in Table 4.13; coefficients for Complaining and
Switching Responses to Dissatisfaction are presented in Table 4.30.
Figure 4.2 Hypothesized Model of Consumer Complaining and Switching Responses to Dissatisfaction
Within the Voucher/Certlflcate Rental Housing Market
Dashed lines between domalnes Indicate hypothesized relationships that were not significant.
Table 4.3 Unit Attribute Performance
Factor Loading Items Included In Scale All Greensboro Richmond Newport News
.721 I seldom have problems with my appliances. 3.17 3.09 3.20 3.18
.660 Roaches and/or mice are seldom a problem in my unit 3.58 3.50 3.61 3.61
.618 I seldom have problems with the electrical system in my unit. 3.69 4.07 3.70 3.41
.518 Water leaks from outside my unit seldom occur. 4.18 4.26 4.06 4.27
Scale ri Alpha reliability .62 Items Not Included In Scale 3.64 357 3.65 87 3.66 146 3.68 124
-.402 The walls and floors of my unit have holes and cracks. 5.00 5.01 5.10 4.87
-.273 I often have problems with my plumbing. 4.88 4.51 4.99 5.03
.535 The heating system keeps my unit comfortable. 3.24 3.19 3.42 3.04
1=Strongly Agree, 7=Strongly Disagree
Table 4.4 Location Attribute Performance
Factor Loadinq Items Included In Scale AH Greensboro Richmond Newport News
.787 There is a lot of litter and trash around my unit. [R] 5.39 5.45 5.23 5.55
.772 Traffic congestion and noise is a problem around my unit [R] 5.06 31 4.70 4.93 5.48
.502 Crime and vandalism is a problem around my unit. [R] 4.19 4.32 3.90 4.44
Scale 2.77 u 2.90 2.96 2.44
Alpha reliability .51
Items Not Included In Scale
-.226 A variety of grocery and other stores are located near my unit. 3.4^ 2.75 3.34 2.59
-.269 Loud neighbors are not a problem around my unit 4.41 4.36 4.46 4.38
1=Strongly Agree, 7=Strongly Disagree
31 Newport News residents disagreed more strongly with the statement that
congestion and noise were problems, compared to Greensboro and Richmond
residents, F(2,387)=3.69, p=.026.
b/ Newport News residents were more positive about location attribute
performance than were others, F(2,383)=3.21, p=.041.
Richmond residents were less positive about the statement that a variety of
grocery and other stores are available near their unit, compared to Greensboro
and Richmond residents, F(2,376)=4.74, p=.009.
[R] Survey responses were reverse coded when the scale was computed.
Table 4.5 Management Attribute Performance
Factor Loadina Items Included In Scale A|] Greensboro Richmond Newport News
.797 Repairs to my unit are done right the first time. 3.65 3.64 3.92 3.33
.750 The hallways, stairs, walkways, and parking areas serving my unit are well lighted. 3.12 3.29 3.29 2.77
.695 The security services provided by the property management make me feel safe inside and outside my unit 4.04 37 3.89 4.39 3.70
.537 Emergency repair services are not always available when you need them. [R] 4.38 4.13 4.44 4.48
Scale 3.63 u 3.64 3.86 3.33
q 348 87 144 117
-.430 Alpha reliability Items Not Included In Scale The property management does not enforce the lease terms or house rules fairly. 4.98 4.99 5.02 4.93
1=Strongly Agree, 7=Stongly Disagree
87 Richmond residents did not agree as strongly that security services make
them feel safe, compared to Greensboro and Newport News residents,
F(2, 372)=3.21, p.=.041.
b/ Newport News residents were more positive about management attribute
performance than were others, F(2,345)=3.33, p.=.037.
[R] Survey responses were reverse coded when scale was computed.
Table 4.6 Property Management Service Performance SERV/PERF Tangibles
Loadina Items Included In Scale AH Greensboro Richmond News
.730 The property managements office and equipment is visually appealing. 3.36 37 3.35 3.77 2.89
.719 The property managements employees are well dressed and
appear neat. 2.73 2.61 2.98 2.54
.656 The property management has up to date equipment and technology. 3.67 b/ 3.45 4.03 3.42
.576 The appearance of the property is in keeping with the type of service provided by the property management 3.07 3.10 3.08 3.04
Scale 3.25 67 3.17 3.49* 3.03
ri 317 79 127 111
Alpha reliability .59
1=Strongly Agree, 7=Strongly Disagree
37 Newport News residents agreed more strongly that managements office and
equipment was visually appealing, compared Greensboro and Richmond
residents, F(2, 372)=5.86, p.=.004
b/ Richmond residents did not agree as strongly that property management has
up to date equipment and technology, compared to Greensboro and Newport
News residents, F(2, 399)=3.45, p.=.033.
07 Newport News residents were more positive about property management
service performance SERV/PERF tangibles than were others, F(2, 314)=3.40,
Table 4.7 Property Management Service Performance SERV/PERF Reliability
Factor Loadinq Items Included In Scale AH Greensboro Richmond Newport News
.847 The property management is dependable. 3.12 37 2.72 3.45 3.01
.821 When the property management promises to do something by a certain time, it does so. 3.35 3.33 3.59 3.36
.789 The property management keeps its records accurately. 2.88 b/ 2.57 3.24 2.69
.775 The property management provides its services at the time it promises to do so. 3.34 3.10 3.59 3.21
.719 Employees get adequate support from the property management to do their jobs well. 3.29 3.10 3.52 3.17
Scale 3.22 2.95 3.47 3.14
T| 336 86 136 114
Alpha Reliability .85
1=Strongly Agree, 7=Strongly Disagree
* Richmond residents did not agree as strongly that management is
dependable, compared to Greensboro and Newport News residents,
F(2,397)=4.16, p.= 016.
b/ Richmond residents did not agree as strongly that management
keeps accurate records, F(2,409)=4.40, p.=.013.
Table 4.8 Property Management Service Performance SERV/PERF Responsiveness
Loadina Items Included In Scale AN Greensboro Richmond News
.752 You do not receive prompt service from the property managements employees. [R] 4.48 4.75 4.21 4.60
.732 Employees of the property management are too busy to respond to resident requests promptly. [R] 4.73 5.02 4.68 4.52
.677 The property management does not tell its residents exactly when services will be performed. [R] 4.83 4.88 4.93 4.67
.650 Employees of the property management are not always willing to help the residents. [R] 4.90 4.89 4.80 5.03
Scale n 3.24 338 3.11 3.31 3.24
Alpha reliability .66
1=Strongly Agree, 7=Strongly Disagree
[R] Survey responses were reverse coded when the scale was computed.