HR summary
sbatniji
why_do_managers_use_performance_information.pdf
JPART 20:849–866
The Big Question for Performance Management: Why Do Managers Use Performance Information?
Donald P. Moynihan*, Sanjay K. Pandey†
*University of Wisconsin–Madison; yRutgers University, Newark
ABSTRACT
This article proposes that understanding public employee use of performance information is
perhaps the most pressing challenge for scholarship on performance management.
Governments have devoted extraordinary effort in creating performance data, wagering that
it will be used to improve governance, but there is much we do not know about the factors
associated with the use of that information. This article examines the antecedents of self-
reported performance information use from a survey of local government managers. The
results show that public service motivation, leadership role, information availability,
organizational culture, and administrative flexibility all affect performance information use.
INTRODUCTION
Terms such as ‘‘performance’’ and ‘‘results’’ have become ubiquitous in contemporary
governance. Major administrative reforms are driven by a belief that governments suffer
from a ‘‘performance deficit’’ (Kamensky 1996) that can be best overcome by measuring
the effort and result of government activity. These beliefs are so deeply embedded that they
have been variously described as a ‘‘movement’’ (Radin 2006) and ‘‘doctrine’’ (Moynihan
2008).
The most widespread governmental reform in recent decades has been the requirement
for agencies to track and measure strategic goals, targets, and achievements (Brudney,
Hebert, and Wright 1999; Moynihan 2008). Within our growing state of agents, perfor-
mance goals underpin contractual forms of accountability, the means by which webs of
connected principals and agents allocate responsibility. Citizens, elected officials, and pub-
lic managers have more performance information now than ever. Every year, new rivers
flow into the existing sea of data. These trends are unlikely to be reversed. Performance
management both preceded and outlived the New Public Management and continues to be
viewed as a central plank in the future of governance (Kettl and Kelman 2007).
This article was originally presented at the 2008 meeting of American Political Science Association. We are grateful
for helpful comments from Tony Bertelli, Richard Williams, Pam Herd, Kelly LeRoux, Dale Krane, and Robert
Smith, as well as to three anonymous reviewers from JPART. Address correspondence to the author at
doi:10.1093/jopart/muq004 Advance Access publication on March 1, 2010 ª The Author 2010. Published by Oxford University Press on behalf of the Journal of Public Administration Research and Theory, Inc. All rights reserved. For permissions, please e-mail: [email protected]
Behn has argued that one of three ‘‘big questions’’ for public management research cen-
tersonhowtomeasureperformanceinawaythatfostersachievement,andspecificallyasked,
‘‘how can public managers use measures of the achievements of public agencies to produce
even greater achievements?’’ (Behn 1995, 321). We propose a slightly different question,
which we believe is another big question for public management, and perhaps the biggest
question for performance management: why do managers use performance information?
Whether public officials are actually using performance data to manage is the best indicator
ofwhetherperformancemanagementisworththeeffort(Hatry1999).Withoutknowledgeof
why such use occurs, it becomes difficult to establish the conditions for performance man-
agement success. Van Dooren (2008, 22) argues that, ‘‘if we want to study the successes and
failures of performance movements, we have to study the use of performance information.’’
Determining the actual impact of reforms is exceptionally difficult (Pollitt 2000), but per-
formance information use offers a more tractable measure of success. More broadly, the use
of performance information suggests the type of purposeful and goal-oriented behavior that
elected officials and members of the public say they want from bureaucrats.
Performance information use is important not just to students of administrative reform
but can also inform scholarship on public policy, network theory, principal agent theory,
and other areas of cross-disciplinary interest. Although governments have devoted a great
deal of energy and resources into creating performance information systems, they have
largely neglected the question of how to foster information use. This may be beginning
to change at the federal level, with President Obama’s Chief Performance Officer stating
that, ‘‘the ultimate test of our performance management efforts is whether or not the in-
formation is used’’ (Zients 2009). Although there is some empirical research on this ques-
tion, there is much we do not understand. A group of younger scholars at the most recent
Minnowbrook conference proposed that the performance information use remains one of
the most important yet understudied issues in performance management (Moynihan et al.
forthcoming), whereas Van de Walle and Van Dooren (2008, 2) argued, ‘‘while the pro-
duction of performance information has received considerable attention in the public sector
performance measurement and management literature, actual use of this information has
traditionally not been very high on the research agenda.’’
This article builds upon the existing empirical literature by developing and testing
a model of performance information use on a survey of local government officials. We
treat performance information use as a form of organizational behavior that is influenced
by individual, job, organizational, and environmental factors.
A MODEL OF PERFORMANCE INFORMATION USE
Previous Research
There are a number of empirical pieces on performance information use in the public sector,
and the majority of such research is recent. Heinrich (1999) noted that most empirical ev-
idence came from the private sector. Her study of performance standards in job training
found that data were used but much depended upon the design of the overall performance
system and that there was little to guide designers beyond neoclassical economic arguments
for financial incentives.
Since then, more research has emerged to offer alternatives to neoclassical models,
usually relying on self-reported survey data. It appears fair to assert that this previous
work has not resulted in a common or overarching theory of performance information
850 Journal of Public Administration Research and Theory
use. de Lancer Julnes and Holzer (2001) theorized a basic distinction between rational/tech-
nocraticandpolitical/culturalfactors.Otherresearchhasfocusedmoreonexplicitlypolitical
variables. Bourdeaux and Chikoto (2008) examined the role of the governor, the legislature,
andthestatepoliticalcontext.MelkersandWilloughby(2005)categorizedvariablesinterms
of community characteristics, respondent characteristics, organizational culture, and perfor-
mance measurement characteristics. Moynihan and Ingraham (2004) and Dull (2009) both
placed the role of leadership as central. The study by Askim, Johnsen, and Christophersen
(2008) of municipal benchmarking networks in Norway examined network characteristics,
administrative factors, political variables, task characteristics, and history dependence.
Moynihan and Landuyt (2009) characterized performance information use as an element
of organizational learning and identified structural and cultural variables that predict such
use.
Many of the results of such research are discussed below in the context of specific
hypotheses in order to indicate where this article builds upon and departs from pre-
vious research. Here, we summarize some of the additional findings of this research that
are not directly tied to our model. Some of these findings result from testing similar
variables across models. For example, the provision of adequate resources has been con-
sistently found to be associated with performance information use (Askim, Johnsen, and
Christophersen 2008; de Lancer Julnes and Holzer 2001; Moynihan and Landuyt 2009).
The existence of dedicated learning forums is also associated with use (Askim, Johnsen,
and Christophersen 2008; Moynihan and Landuyt 2009). Evidence from US state govern-
ments and Norwegian municipal governments associates performance information use with
more liberal political settings (Askim, Johnsen, and Christophersen 2008; Bourdeaux and
Chikoto 2008; Moynihan and Ingraham 2004). There is mixed evidence on size. Larger US
state governments have been associated with performance information use (Bourdeaux and
Chikoto 2008; Moynihan and Ingraham 2004), but smaller governments were more likely
to use performance data among Norwegian municipalities. Political competition (Askim,
Johnsen, and Christophersen 2008) has been found to be positively associated with use,
although other models have not found significant results (Moynihan and Ingraham
2004), and in some instances, political conflict has been shown to have a negative or non-
significant effect (Dull 2009). The presence of basic bureaucratic competence and expertise
in performance management is associated with use (Bourdeaux and Chikoto 2008; Dull
2009). Measures of legislative involvement vary in their influence on use among executive
branch officials, ranging from positive (Bourdeaux and Chikoto 2008), negative (Moynihan
and Ingraham 2004), to nonsignificant (Dull 2009).
Other findings are specific to particular studies. The following variables have been
found to be positively associated with performance information use: administrative stabil-
ity (Askim, Johnsen, and Christophersen 2008); internal requirements and lower levels of
government (de Lancer Julnes and Holzer 2001); inclusion of organizational members in
performance management processes (Melkers and Willoughby 2005); and chief executive
power (Bourdeaux and Chikoto 2008). Factors that have been negatively associated with
performance information use include efforts by the central agency to control the policy
agenda (Moynihan and Ingraham 2004) and measurement challenges (Dull 2009).
Qualitative work has examined fewer variables and has generally not attempted to con-
struct formal models. But it has identified some common findings that overlap with much of
the quantitative work cited above. Leadership and organizational culture are recurring
Moynihan and Pandey The Big Question for Performance Management 851
themes (Broadnax and Conway 2001; Franklin 2000; Moynihan 2005). Radin (2006) points
outthatsometasksaremore compatiblewithperformancemanagementthanothers,whereas
Ammons and Rivenbark (2008) find that the quality of performance data matters.
Conceptualizing Performance Information Use as Organizational Behavior
We conceptualize performance information use as a form of organizational behavior. Like
other forms of organizational behavior, employees have discretion about whether and the
degree to which they engage in it but are influenced by the social context and formal sys-
tems in which they work. We test categories of variables consistent with this conceptual-
ization, incorporating individual beliefs, job attributes, organizational factors, and
environmental influences.
Ourmodeldrawsuponsomeofthevariablesusedinpreviousresearch,althoughweseek
tousealternativemeasuresoftheunderlyingconcept.Forexample,ratherthanmeasurelead-
ershipsupportforperformancemanagement,wemeasurewhethertheleaderisinageneralist
of specialist position. Rather than measure public participation in performance management
routines, we measure if more general forms of participation matter. The model also includes
variables whose relationship with the dependent variable has been previously untested,
including public service motivation (PSM), reward expectations, role clarity, task-specific
experience, the role of budget officials, and the influence of professional organizations.
Individual Beliefs
There is significant evidence that altruistic beliefs affect public employee behavior in
ways that benefit the organization. Much of this evidence falls under the rubric of the
PSM concept (see Perry and Hondeghem 2008). High PSM employees exhibit higher levels
of organizational commitment, enjoy higher job satisfaction, experience greater job
involvement, and require less extrinsic rewards (Pandey and Stazyk 2008). There is also
evidence that PSM fosters positive citizenship behavior both internal (Pandey, Wright, and
Moynihan 2008) and external to the organization (Brewer 2003; Houston 2006).
There has, thus far, been surprisingly little investigation into how PSM might affect
organizational decision making. There are two reasons to assume that PSM might foster
performance information use. First, performance information use involves costs for the
employees. Performance information use is a behavior that imposes costs on the employee.
It displaces traditional modes of decision making and heuristics, while adding another de-
cision criterion, making decision processes more rather than less complex. Although using
performance data might generate organizational benefits, individual benefits are unlikely or
uncertain. It thus resembles a form of extra-role behavior where employees make gifts of
their time and effort to the organization without the expectation of individual reward. 1 Such
behavior is likely to be exhibited by employees driven by prosocial or altruistic motives.
1 It is worth noting that the accuracy of this assumption may vary with the context public officials find themselves in.
We believe this characterization is accurate for our sample of local government officials, and most government officials
generally. For such officials, performance information use is difficult to observe and therefore an impossible-to-enforce
behavior that is not obviously tied to self-interest. But in contexts with high-powered financial or tenure incentives tied
to performance data—such as those in performance contracts—performance information use may be related to their
self-interest. There are also some instances where organizations formally organize processes of mandatory
performance information use (e.g., Behn 2007), thereby removing the voluntary component of performance
information use. In such settings, performance information use is less likely to be a form of extra-role behavior.
852 Journal of Public Administration Research and Theory
Second, it is plausible that high PSM employees identify with and care about the achieve-
ment of organizational goals. It has been proposed that PSM fosters higher individual and
organizational performance, although without definitive evidence (Brewer 2008). If high
PSM employees see performance information use as a means of achieving organizational
goals, they may also see it as a means to fulfill their desire to serve.
H1 Managers with higher levels of PSM are more likely to use performance information.
Job Attributes
Job attributes represent the interplay between the individual and organizational level
(Wright 2001). In contrast to the intrinsic approach represented by PSM, other approaches
assume self-interest. From this perspective, extrinsic motivators and incentives may foster
performance information use. Organizations do not reward performance information use
but do reward individualized performance to varying degrees. If individuals perceive that
their organization links pay and promotion with goal achievement, employees have an ex-
trinsic incentive to use performance information (Jennings and Haist 2006).
H2 Managers who perceive a link between extrinsic rewards and performance are more
likely to use performance information.
Another factor that affects the types of decisions individuals make, and by extension
their tendency to use performance data, is their organizational role. Previous research tends
to treat role differences in terms of membership of different organizations and of seniority
or hierarchical level. Both qualitative (Radin 2006) and quantitative (Askim, Johnsen, and
Christophersen 2008; Dull 2009) research report that variation in the use of performance
data depends upon the type of program an employee worked in. 2 Other research tests the
influence of seniority, finding no relationship (de Lancer Julnes and Holzer 2001; Dull
2009) or a positive relationship (Moynihan and Landuyt 2009).
Another test of how organizational roles matter is to distinguish between specialist
(i.e., function specific) and generalist roles. The model controls for whether the respondent
is a generalist leader (a City Manager or Assistant/Deputy City Manager) or the head of
a task-specific agency. There are good reasons to believe that leaders with more general re-
sponsibilities are less likely to use performance information than those with task-specific
responsibilities. First, a leadership role requires dealing with external stakeholders and
political negotiation. This leaves less time for management activities. Second, the more
seniortheleader,thelessspecializedtheknowledge.Unlikeagency-levelofficials,generalist
leaders lack a sense of context to interpret performance information—they struggle to know
why performance is good or bad or what to do about it. There are some exceptions, such as
athe‘‘stat’’modelthatseesseniorofficialsquestionagencyofficialsonperformancedata,but
theseexceptionsareregardedasinnovationspreciselybecausetheyareunusual(Behn2007).
H3 Task-specific leaders are more likely to use performance information than generalist
leaders.
2 In an alternative version of the model, we controlled for the specific functions our respondents manage, using
dummy variables for Finance/Budgeting, Public Works, Personnel/HR, Economic Development, Parks and
Recreation, Planning, and Community Development. Consistent with the research cited above, we found that some
functions were more likely to use performance data than others. Specifically, a function characterized by a clear task
and direct service to the public (Parks and Recreation) was significantly related to performance information use.
Moynihan and Pandey The Big Question for Performance Management 853
A final role factor is the amount of experience that an individual has with a specific
position. Organizational learning theory points to the importance of task knowledge to
learning (Moynihan 2005). Managers rarely learn directly from quantitative numbers,
but from interpreting these numbers, making sense of what they mean given their knowl-
edge of the context in which they work. Individuals with a deep knowledge of task are
therefore advantaged in the ability to apply performance data. Lewis (2008) finds that bu-
reau-specific experience was a significant predictor of why some managers’ received higher
program evaluation scores than others. One measure of task-specific knowledge is time in
a particular position. 3
H4 Managers with greater task-specific experience are more likely to use performance
information.
Organizational Factors
All levels of government have devoted considerable effort into creating and disseminating
performance information. This effort reflects a supply-side approach to performance infor-
mation, assuming that provision of performance data is the key to its use (Jennings and
Haist 2006). There is some research support for this view. Various measures that track
the availability of performance information have been found to be positively associated
with use (Bourdeaux and Chikoto 2008; de Lancer Julnes and Holzer 2001; Moynihan
and Ingraham 2004; Moynihan and Landuyt 2009). By contrast, Melkers and Willoughby
(2005) find that the availability of measures in budget documents actually reduces the in-
fluence of these measures in the budget process and suggest that this may be the result of
information overload.
Ammons and Rivenbark (2008) point out that it is not only the existence of measures
but also the ability to tie these measures to management systems that fosters use. We test
whether the perceived availability of measures and the linkage of these measures to man-
agement processes (via benchmarking, strategic planning, customer service measurement,
and linking to budgets) are associated with performance information use.
H5 Managers who perceive performance information is available and tied to
management systems are more likely to use it.
A contrast to the supply-side view is the demand-side approach. Although the supply-
side approach assumes that information availability drives use, a demand-side approach sug-
gests that simple access to data is not enough. Managers must want to use performance data.
This demand is shaped by the organizational environment and cultural norms. There is ev-
idence to support the idea that organizational culture matters. Case research illustrates the
importance of building an organizational culture that is supportive of performance informa-
tion use (e.g., Broadnax and Conway 2001; Franklin 2000). Yang and Hsieh (2006) find that
3 It is possible that experience in the same position may indicate an inability to be promoted and, therefore, might be
negatively associated with management practices perceived as positive. But this is not a significant concern since our
respondents are agency heads, Assistant City Managers, or City Managers. It is also possible that time in organization
might have a negative effect on performance information use since employees may become jaded. Melkers and
Willoughby (2005) find that organizational tenure is negatively related to one of their three measures of performance
information use. We did test the effects of organizational tenure in an alternative specification of the model and did not
find it to have a significant linear or curvilinear relationship. Because of its strong correlation with our measure of
experience, we drop it from the final model.
854 Journal of Public Administration Research and Theory
organizational support for performance measurement is correlated with perceptions that it is
effective. There is also evidence that goal or mission orientation is associated with perfor-
mance information use (de Lancer Julnes and Holzer 2001; Moynihan and Landuyt 2009).
One limitation of this previous research is that it defines culture narrowly, in terms of its
orientation toward performance information systems. The findings tell us that cultures that
are supportive of mission or performance systems are more likely to see those systems used,
but tell us little about how more basic aspects of the organizational culture relate to perfor-
mance information use. Some broad cultural concepts are likely to be relevant. For example,
if managers are in an environment that rewards innovation and allows them to question exist-
ing routines, they are more likely to use performance data. But if they are in an environment
thatemphasizesproceduralcontinuityandwarnsagainstrisktaking,theyarenotlikelytouse
performance data. We therefore hypothesize that where managers perceive a developmental
culture,theywillbeencouragedtouseperformanceinformation.Developmentalculturesare
associated with a focus on the organization, flexibility, adaptability and readiness, growth,
andresourceacquisition(QuinnandRohrbaugh1981;ZammutoandKrakower1991;Pandey,
Coursey, and Moynihan 2007). Developmental cultures correlate with self-reported meas-
uresoforganizationaleffectiveness(MoynihanandPandey2005),althoughdeLancerJulnes
and Holzer (2001) found that a similar measure of organizational culture was not associated
with performance information use among their respondents.
H6 A developmental organizational culture fosters performance information use.
Flexibility is another organizational factor likely to shape performance information
use. If managers have the freedom to experiment with processes, they have an incentive
to examine performance data to find rationales for innovation. If managers are restricted in
their ability to pursue process change, insights derived from examining performance data
are less likely to be useful, and therefore, the incentive to use data are reduced.
Performancemanagementsystems,reflectingadoctrinalconnectionwiththeNewPub-
lic Management, were often presented as part of a package of reforms that also included the
provisionofgreaterautonomytomanagers,inparticulargreaterdiscretionwithfinancialand
human resources (Moynihan 2008). For example, Schick (2001) argues that performance
measurement should be presaged by wider organizational changes, including greater man-
agerial flexibility, if it is to succeed. Studies have found that flexibility and its flipside, cen-
tralization are antecedents to organizational learning (Moynihan and Landuyt 2009; Schulz
2001) and performance information use (Willis, Mastrofski, and Weisburd 2007).
H7 Managers who perceive decision flexibility are more likely to use performance
information.
Over a decade ago, Schick (1997) conjectured that public sector reform could see an
evolution in the role of the budget officials, moving away from a traditional emphasis on
micro-financial controls, to a sustained attention to the performance of agencies, both in
terms of technical efficiency but also how they allocate their resources. Some reforms, such
as the ‘‘stat’’ model (Behn 2007), or the Program Assessment Rating Tool (Moynihan
2008), are based on the logic that the performance that agencies made must be challenged
if performance data are to be taken seriously. From this perspective, an adversarial dialogue
fosters use. A contrary perspective is that an adversarial exchange can devolve into
a ‘‘gotcha’’ approach, fostering defensiveness and even gaming (de Haven-Smith and Jenne
2006). Consistent with this criticism, some have proposed that homogeneity of beliefs
Moynihan and Pandey The Big Question for Performance Management 855
within organizational settings will encourage use since organizational actors have a com-
mon base of trust, understanding, and cooperation (Jennings and Haist 2006; Moynihan
2008). To test whether a more adversarial discourse between agency budget and other staff
fosters or discourages performance information use, the model includes a measure of bud-
get official willingness to challenge the plans and actions of department heads.
H9 The willingness of budget staff to adopt an adversarial stance affects performance
information use.
EXTERNAL FACTORS
To examine the role of external factors, we test how the influence of citizens and external
professional organizations affects performance information use. It is possible that the use of
performance information is a form of bureaucratic behavior that is disconnected from the
public. Managers may believe that the public may evince strong support for performance
but actually care little about the details. On the other hand, results-based reforms draw from
popular sentiment that governments were not as effective as they could be. There is sig-
nificant evidence that perceived citizen support for or involvement in performance man-
agement processes facilitates use. Research suggests that perceived citizen demand for
performance-based accountability encourages performance information use (Moynihan
and Ingraham 2004; Poister and Streib 1999). de Lancer Julnes and Holzer (2001) find
that support among external interest groups (in the form of elected official/citizens) for
performance management also fostered use. Yang and Hsieh (2006) find that stakeholder
participation is a positive predictor of the perceived effectiveness of performance measures,
whereas Ho (2006) finds that citizen involvement in performance measurement practices
increases the perceived usefulness of data in the eyes of elected officials.
But we know little about whether more common forms of citizen participation—such
as budget hearings, customer surveys, citizen phone calls, or emails—foster performance
information use. One might assume that more participatory governments would be under
greater pressure to demonstrate performance and, therefore, more likely to use performance
data. Consistent outreach to and feedback from the public may create a pressure on man-
agers to justify decisions, legitimate programs, and seek additional support from stakehold-
ers (Van de Walle and Boivard 2007). But there is some reason to suggest that the
relationship might be negative. In the battle for administrative attention, there may be
a trade-off between information derived from citizen input and performance data. Partic-
ipation has been portrayed as at odds with the technocratic decision approaches (Moynihan
2003). Public managers have sometimes proven reluctant to meaningfully incorporate cit-
izen input into performance management systems (Heikkila and Isett 2007; Poister and
Streib 1999) or view such participation as a way of legitimating rather than informing de-
cisions (Moynihan 2003). Bureaucrats that use performance data may see less need to listen
to citizen input and vice versa.
H10 Perceptions of citizen participation affects performance information use.
In the public sector, new innovations are often fostered by professional norms (Roy
and Seguin 2000). Professional organizations such as the American Society for Public Ad-
ministration, Governmental Accounting Standards Board, National Academy of Public Ad-
ministration, the National Council of State Legislatures, and the International City/County
856 Journal of Public Administration Research and Theory
Managers Association have issued recommendations and guidance that encourage perfor-
mance management systems. Case research suggests that such professional guidance in-
fluences central agency officials and the adoption of results-based reforms (Moynihan
2008), but there is no clear evidence on whether it encourages agency managers to use
performance data.
H11 Managers influenced by professional organizations are more likely to use
performance information.
Because we rely on a national survey of municipal managers, the model controls for
a number of additional environmental factors, including the size of government (in terms of
per capita expenditures), income per capita, the homogeneity of population, population
size, and region.
DATA COLLECTION
The data for this study were collected in phase four of the National Administrative Studies
Project (NASP-IV). NASP-IV is a multimethod study, a key part of which is a survey ad-
ministered to a nationwide sample. The theoretical population of interest for NASP-IV was
comprised of senior managers in US local government jurisdictions with populations over
50,000. The general managers included the city manager and assistant/deputy city man-
agers (29% of respondents), whereas the rest of the respondents headed key departments.
The sample design and construction for the NASP-IV study was aided by the International
City/County Management Association, who provided an initial contact list of respondents,
which were verified, updated, and expanded, by the research team. When the study con-
cluded, 1,538 respondents completed the survey, a response rate of 46.4%. The respondents
come from 545 different jurisdictions. The mean age of respondents was 51.4 years. As
expected, a sizable majority were male (71%), white (86%), highly educated (more than
60% with graduate degrees), and well compensated (68% with salaries over $100,000).
Most measures used in the study have been tested and validated in earlier studies;
some measures were developed in earlier administrations of NASP and others were written
and/or refined for NASP-IV. Appendix provides additional detail on the survey items, in-
cluding the source of previously used questions, and Cronbach alphas for indexes. Descrip-
tive statistics and a correlation matrix are provided in table 1. The correlation matrix does
not show correlations between independent variables indicative of collinearity.
Our dependent variable is the response to the question: ‘‘I regularly use performance
information to make decisions,’’ with the response ranging from 1 (strongly disagree) to 6
(strongly agree). The dependent variable is a relatively broad indicator of managerial use of
performance data but is appropriate given the development of empirical research. Although
the use of single-item measures is sometimes criticized, research in diverse areas such of
job satisfaction, performance appraisal, job targets, and marketing find that single items are
often not less reliable than multiple response items (Bergkvist and Rossiter 2007; Gardner
et al. 1998; Wanous and Hundy 2001). Indeed, Gardner et al. (1998) point out that single-
item measures avoid the risk of aggregating multiple measures whose inter-item correlation
is due common method variance. Previous work on performance information use by man-
agers tends to reflect a unidimensional understanding of information use, which is that it is
purposeful, resulting in improved outcomes. Indeed, previous research that reported dif-
ferent measures of use found that these measures were so highly correlated that they were
Moynihan and Pandey The Big Question for Performance Management 857
Table 1 Descriptive Statistics and Correlation Matrix
Variable Range Mean SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Performance
information use
1–6 4.067 1.343
2. Public service
motivation
8–30 25.464 3.607 0.261
3. Reward
expectation
4–20 11.927 3.764 0.193 0.124
4. Generalist
leader
0–1 .286 .452 20.017 0.117 0.188
5. Task-specific
experience
0–36 7.366 6.800 0.100 0.045 20.006 20.053
6. Information
availability
4–24 15.479 4.403 0.520 0.246 0.344 0.147 0.106
7. Developmental
culture
3–15 11.528 2.456 0.297 0.284 0.206 0.051 0.113 0.336
8. Flexibility 2–12 9.36 1.973 0.214 0.203 0.197 0.126 0.092 0.262 0.248
9. Budget staff
takes
adversarial role
1–6 3.73 1.451 0.115 0.093 0.076 0.099 20.023 0.15 0.106 0.073
10. Citizen
participation
6–42 29.84 6.368 0.276 0.2957 0.168 0.129 0.006 0.404 0.220 0.174 0.111
11. Professional
influence
1–7 6.149 1.041 0.173 0.305 0.078 0.007 0.009 0.146 0.157 0.107 0.067 0.171
12. Population
size (log)
10.825–15.122 11.599 .691 0.065 0.031 0.027 0.089 20.098 0.16 0.019 0.000 0.109 0.088 0.018
13. Income per
capita
9340–63015 22252.25 6545.395 0.026 0.008 0.132 0.047 0.068 0.198 0.048 0.037 0.009 0.061 20.027 20.126
14. Population
homogeneity
0.175–0.953 0.588 0.162 20.030 20.037 0.015 20.039 0.063 20.029 0.001 0.014 20.031 20.031 20.009 20.351 0.244
15. Government
size
165–9452 1746.96 1074.863 20.010 0.032 20.064 0.001 0.003 0.055 20.014 20.040 0.027 0.037 0.012 0.221 0.059 20.081
16. Western
region
0–1 0.398 0.49 20.036 20.036 0.090 0.063 20.054 0.031 0.018 0.027 20.008 0.004 20.057 0.05 0.017 20.229 20.193
17. Northeastern
region
0–1 0.044 0.205 20.011 0.018 20.176 20.043 0.05 20.067 20.055 20.161 0.058 20.036 20.035 20.058 0.043 20.039 0.2693 20.175
18. Midwestern
region
0–1 0.232 0.422 20.049 0.015 20.099 20.047 0.027 20.100 20.045 20.041 20.019 20.043 20.0349 20.083 0.091 0.390 20.0763 20.444 20.123
8 5 8
aggregated into a single scale (Bourdeaux and Chikoto 2008; de Lancer Julnes and Holzer
2001; Dull 2009). As with previous research, we rely on self-reported indicators. This ap-
proach brings limitations, in particular the potential for an upward response bias. Because
we are taking a behavioral approach, we focus on individual estimates of their own use,
rather than individual perceptions of wider group use, as most previous research has done.
Since the dependent variable is ordinal, we employ an ordered probit analysis. 4 In order
to ease interpretation of the results, we employed MEOPROBIT method (Cornelissen 2006),
whichisameansofre-estimatingorderedprobitresultsinawaythatprovidesmarginalprob-
ability estimates that are similar to ordinary least-squares coefficients. The estimate repre-
sents how a one-unit increase in the independent variable results in a mean change on the
dependent variable. Examples of such interpretations are provided below (table 2).
RESULTS AND DISCUSSION
In terms of our individual-level variable, we find that PSM is positively correlated with
reported performance information use. A one-unit increases on the 30-point PSM scale
results in a .04 increase in the 6-point performance information use scale. The finding also
contributes additional circumstantial evidence to the question of whether PSM fosters
higher performance. Brewer (2008, 146) notes ‘‘there is little empirical evidence on the
PSM—performance relationship.’’ The results here find that PSM is positively associated
with a form of behavior that is a logical contributor to both higher individual and orga-
nizational performance.
The finding on PSM is also important when considered in relation to the nonresult for
the measure of reward expectations. The findings suggest that performance information use
in our sample of city managers is driven by altruism rather than self-interest. This is an
important result because contemporary reforms often attempt to use performance indicators
to create contract-like arrangements. These reforms rest on what may be incorrect assump-
tions about what drives performance information use in the public sector. Our results sug-
gest that performance systems should be designed to appeal to a sense of public service
rather than to reward expectations, particularly in tasks where contractual arrangements
will be incomplete.
Of course, this finding may not apply universally to all public officials. Much
depends on the nature of the performance data and how closely it is linked with extrinsic
reward. One possible criticism of the result is that the reward expectations experienced by
our sample, like most governmental officials (Perry 1986), are simply inadequate to test
whether high-powered incentives would foster performance information use. 5
The
4 One concern with ordinal regressions is that they violate the parallel regression assumption that the relationship
between each pair of outcome groups is the same, that is, that the coefficients are related in the same way to any
category of the outcome variable relative to the other categories of the outcome variable. To deal with this problem we
also ran the model using the partial proportional odds model, a form of the generalized ordered logit model
(see Williams 2006 for more detail). The results were equivalent to the ordered probit results we report here. Because
the ordered probit approach we employ here lends itself to greater ease of interpretation, we dropped the partial
proportional odds model at the suggestion of a reviewer.
5 Although we cannot definitively exclude the possibility that significantly higher rewards than those experienced by
respondents would change performance information use, we note that the mean score for reward expectation is above
the midpoint for the scale and that there is substantial variation in the measure (see table 1). Therefore, even among
those who perceive high reward expectations among a sample that appears to experience reasonable variation in
rewards, we do not see significantly greater use of performance information.
Moynihan and Pandey The Big Question for Performance Management 859
findings on PSM, and extrinsic reward, and performance information use may be quite
different in contexts where extrinsic motivators are high powered and/or where managers
feel that performance systems are being used to undermine the altruistic goals that mo-
tivate their behavior (Dias and Maynard-Moody 2007; Weibel, Rost, and Osterloh 2009).
It may also be the case that for our sample, there is a relatively weak link between the type
of effort managers believe will be rewarded and what is measured by performance sys-
tems. Incentives are likely to foster to performance information use where there is a strong
direct link between individual reward and performance data. But this is difficult if for no
other reason then that for most public services, outputs are the result of the interplay of
many individuals, and the precise contribution of any individual is hard to determine
(Perry 1986). As a result, most individual-level reward systems rely on a mix of subjective
and informal assessments of performance. Doing well on such systems depends on a wide
variety of factors, of which performance information use is unlikely to be prominent
(Feldman 1981).
The role an actor plays matters. Being a generalist leader (city managers and deputy/
assistant city managers) resulted in a .35 mean reduction on the 6-point performance in-
formation use scale. It is important to note that there is evidence that leaders, including
elected officials, can inspire performance information use among managers. But much
of this research suggests that this influence comes not from the direct use of information,
Table 2 Results of Performance Information Use Model
Ordered Probit
Marginal Effects p Score
Individual beliefs
Public service motivation 0.038 0.001
Job attributes
Reward expectation 0.005 0.595
Generalist leader 20.346 0.000
Task-specific experience 0.006 0.290
Organizational factors
Information availability 0.146 0.000
Developmental culture 0.051 0.001
Flexibility 0.042 0.024
Budget staff take adversarial role 0.023 0.338
External factors
Citizen participation 0.011 0.062
Professional influence 0.066 0.063
Controls
Population size (log) 20.022 0.703
Income per capita 20.057 � 1024 0.317 Population homogeneity 20.221 0.371
Government size 0.005 � 1022 0.127 Western region 20.048 0.571
Northeastern region 0.262 0.141
Midwestern region 20.006 0.960
Note: Ordinal logistic model: N 5 1,132; Wald x 2 5 432; Psuedo R
2 5 .116 (cut points: 1 5 1.828; 2 5 2.627; 3 5 3.287; 4 5 4.209;
5 5 5.260). Marginal effects and p values calculated with Stata’s ‘‘MEOPROBIT’’ by Cornelissen (2006).
860 Journal of Public Administration Research and Theory
but by communicating credible commitment to performance systems, through symbols, the
allocation of resources, and leadership attention (Askim, Johnsen, and Christophersen
2008; Bourdeaux and Chikoto 2008; Dull 2009; Ho 2006; Melkers and Willoughby
2005; Moynihan and Ingraham 2004). It may be possible for generalist leaders to encour-
age performance information use even as they do not follow their own advice. It is also
important to note that the findings are relative—they do not suggest that generalist leaders
do not use performance data, but that specialist leaders report that they use data more,
controlling for the other factors in the model. Although task-specific experience does mat-
ter, time in current position is not significantly related to reported performance informa-
tion use.
Three of the four organizational factors are positively related to reported performance
information use. Consistent with previous research and a supply-side perspective on data
use, we find that access to and integration of information into performance management
systems predicts greater use. A one-unit increase on the 24-point information availability
scale results in a .146 mean increase on the performance information use scale.
Aspects of the organizational culture also matter. The model shows that having a more
open, innovative, and risk-taking culture is associated with higher reported performance
information use for our sample. A one-unit increases on the 15-point developmental culture
scale resulted in a .05 mean increase in the performance information use scale. Having
greater flexibility increases reported performance information use. A one-unit increase
in the 12-point flexibility scale is correlated with a .024 mean increase in the performance
information use scale.
Taken together, the findings on information systems and organizational culture un-
derline the need for a balanced approach to considering organizational factors. The results
provide additional evidence that the potential for learning from performance information
necessarily requires not just a supply-side approach that ensures that useful information is
easily available but also a demand-side approach that fosters norms consistent with infor-
mation use (see also Moynihan and Landuyt 2009). The other organizational factor, an
adversarial stance by budget officials, was not significantly associated with reported per-
formance information use.
The findings on external factors are not definitive. Both professional influence and
participation are positively related to performance information use but are significant only
at the 10% level. The findings therefore do not imply a clear rejection of these variables but
instead suggest the need for additional testing and imply the need for additional specifi-
cation of environmental variables that might affect performance information use. Although
none of the controls proved significant, there are additional environmental factors—such as
social capital—that have plausible influences on performance information use (Tavits
2006) but where suitable data at the municipal level is lacking. 6
6 In some specifications of the model, an alternative measure of population homogeneity—percent of whites—was
negatively related to performance information use at marginal levels of significant. This is consistent with a finding by
Bourdeaux and Chikoto (2008). However, this result did not hold once we used the more nuanced indicator of
homogeneity in the form of the Herfindahl index.
Moynihan and Pandey The Big Question for Performance Management 861
CONCLUSION
This article argues for examining performance information use as a key variable if we are to
develop systematic knowledge about contemporary governance. The usual caveats and
flaws of cross-sectional survey data apply. In particular, the potential that common source
bias inflates the relationship between independent variables and the dependent variable
cannot be dismissed. But in testing such relationships, capturing a large number of com-
parable individual responses on items that are difficult to externally observe—such as PSM
or culture—is a key requirement, necessitating survey-based approaches. In addition, the
use of previously tested measures of the majority of variables, as well as the relatively low
correlation between these measures, suggests discriminant validity among the items tested
and reduces concerns about common source bias.
The article offers a number of contributions. First, it suggests that performance in-
formation use is more likely to be driven by altruism rather than self-interest among gov-
ernment officials. A logical next step would be to examine how this finding varies in
different contexts—does it, for example, hold in high-powered contract settings where in-
trinsic motivation may be crowded-out? Is the result even more pronounced in settings
where employees feel that their intrinsic motivations are being satisfied?
Another contribution is to test the effect of other well-established predictors in orga-
nizational theory on performance information use for the first time, including organiza-
tional culture, flexibility, and professionalism. The findings not only provide new
information on the antecedents of performance information use but also inform other re-
search literatures, by demonstrating an additional effect of variables such as organizational
culture. For variables that have been previously tested in some forms, the model offers new
insights by using broad-based independent variables. There is a tendency in previous em-
pirical work to test the aspect of a concept that is likely to have the strongest connection
with performance management. As a result, we know that leadership/political support for
performance management matters, goal-oriented cultures matter, and that citizen support
for and involvement in performance management processes matters. Such narrow construc-
tion of independent variables helps to answer the question: ‘‘how can organizations foster
performance information use?’’ But we know less about how more commonly occurring
and broadly constructed organizational concepts matter to performance information use.
Such variables help answer the question ‘‘which organizations are more likely to succeed
with performance management?’’ This article tests many such variables, offering evidence
on how more broadly constructed measures of leadership, citizen participation, and orga-
nizational culture matter.
The findings cannot be considered definitive but take one additional step toward a bet-
ter understanding of why public employees use performance information. In particular, one
should be cautious about extrapolating the factors that shape performance information use
at the local level to other levels of government since the closeness of local government
officials to the public and the actual services delivered may foster an attention to perfor-
mance management not seen at other levels of government (de Lancer Julnes and Holzer
2001; Jennings and Haist 2006). Further research can profitably examine performance in-
formation use in different settings, testing a variety of variables using both quantitative and
qualitative techniques.
862 Journal of Public Administration Research and Theory
APPENDIX Variable Measurement
Variable (Source)
Performance Information use I regularly use performance information to make decisions (1 5 strongly disagree, 6 strongly agree).
Public service motivation
(Perry 1996)
Meaningful public service is very important to me. I am often reminded by daily events about how dependent
we are on one another. Making a difference in society means more to me than personal achievements. I am not
afraid to go to bat for the rights of others even if it means I will be ridiculed. I am prepared to make sacrifices
for the good of society. (1 5 strongly disagree, 6 strongly agree; Cronbach alpha 5 .849)
Reward expectation (Rainey
1983; Spector 1997)
If I accomplish my work objectives, it increases my chances for a pay raise. Fulfilling all my job
responsibilities does little to improve my chances for a promotion. Raises are too few and far between.Pay
structures and personnel rules make it hard to reward a good employee with higher pay here. (1 5 strongly
disagree; 5 5 strongly agree; Cronbach alpha 5 .704.)
Generalist leader 1 5 City Manager/Assistant/Deputy City Manager; 0 5 leader of task-specific agency (Finance/Budgeting, Public
Works, Personnel/HR, Economic Development, Parks and Recreation, Planning, and Community Development).
Task-specific experience Years in present position
Information availability
(adapted from Brudney,
Hebert, and Wright 1999)
Please indicate the extent to which your organization has implemented each of the following: Benchmarks for
measuring program outcomes or results. Strategic planning that produces clear organization mission statements.
Systems for measuring customer satisfaction. (1 5 not at all; 6 5 fully) Performance information is integrated in
my department’s budget preparation process. (1 5 strongly disagree; 6 5 strongly agree). Cronbach alpha 5 .791.
Developmental culture
(adapted from Zammuto
and Krakower 1991)
My department is a very dynamic and entrepreneurial place. People are willing to stick their necks out and take risks.
The glue that holds my department together is a commitment to innovation and development. There is an emphasis
on being best.My department emphasizes growth and acquiring new resources. Readiness to meet new
challenges is important. (1 5 strongly disagree; 5 5 strongly agree; Cronbach alpha 5 .791).
Flexibility My department is able to shift financial resources within its budget to accomplish its mission. My department
is able to shift non-financial resources within its budget to accomplish its mission. (1 5 strongly disagree;
6 5 strongly agree; Cronbach alpha 5 .734).
Budget staff takes
adversarial role
Please indicate the extent to which the budget staff performs the following functions: Challenge plans and actions
of department heads (15 not at all; 6 5 a great deal).
Citizen participation How important are the following methods of gaining citizens feedback for your department? Town hall meetings;
Budget hearings; Citizen/customer surveys; Citizen feedback via the web; Direct contact via phone, mail, e-mail office
visit; Indirect contact via elected officials (1 5 not important at all 7 5 very important; Cronbach alpha 5 .739).
Professional influence I use my profession to help set standards for what I consider good performance for myself
(1 5 strongly disagree, 7 5 strongly agree).
Population size Jurisdiction population in 2000, Census data
Income per capita 2000 census data
Homogeneity Herfindahl index (sum of squares of different racial groups), 2000 Census data
Government size Per capita total expenditures in 2000, Census data
Region Regions according to census definitions, Southern region is reference variable
8 6 3
REFERENCES
Ammons, David N., and William C. Rivenbark. 2008. Factors influencing the use of performance data to
improve municipal services: Evidence from the North Carolina benchmarking project. Public
Administration Review 68:304–18.
Askim, Jostein, Åge Johnsen, and Knut-Andreas Christophersen. 2008. Factors behind organizational
learning from benchmarking: Experiences from Norwegian municipal benchmarking networks.
Journal of Public Administration Research and Theory 18:297–320.
Behn, Robert D. 1995. The big questions of public management. Public Administration Review
55:313–24.
———. 2007. What all mayors would like to know about Baltimore’s CitiStat performance strategy.
Washington, DC: IBM Center for the Business of Government.
Bergkvist, Lars, and John R. Rossiter. 2007. The predictive validity of multiple-item versus single-item
measures of the same constructs. Journal of Marketing Research 44:175–84.
Bourdeaux, Carolyn, and Grace Chikoto. 2008. Legislative influences on performance management re-
form. Public Administration Review 68:253–65.
Brewer, Gene A. 2003. Building social capital: Civic attitudes and behavior of public servants. Journal of
Public Administration Research and Theory 13:5–26.
———. 2008. Employee and organizational performance. In Motivation in public management: The call
of public service, ed. James L. Perry and Annie Hondeghem, 101–7. Oxford: Oxford Univ. Press.
Broadnax, Walter D., and Kevin J. Conway. 2001. The social security administration and performance
management. In Quicker, better, cheaper: Managing performance in American government, ed. D.
Forsythe, 143–75. Albany, NY: Rockefeller Institute Press.
Brudney, Jeffery L., F. Ted Hebert, and Deil S. Wright. 1999. Reinventing government in the American
States: Measuring and explaining administrative reform. Public Administration Review 59:19–30.
Cornelissen, Thomas. 2006. MEOPROBIT: Stata module to computemarginal effects after estimation of
ordered probit. http://ideas.repec.org/c/boc/bocode/s456788.html (accessed August 12, 2009).
de Haven-Smith, Lance, and Kenneth C. Jenne. 2006. Management by inquiry: A discursive accountability
system for large organizations. Public Administration Review 66:64–76.
de Lancer Julnes, Patria, and Marc Holzer. 2001. Promoting the utilization of performance measures in
public organizations: An empirical study of factors affecting adoption and implementation. Public
Administration Review 61:693–708.
Dias, Janice J., and Steven Maynard-Moody. 2007. For-profit welfare: Contracts, conflicts, and the
performance paradox. Journal of Public Administration Research and Theory 17:189–211.
Dull, Matthew. 2009. Results-model reform leadership: Questions of credible commitment. Journal of
Public Administration Research and Theory 19:255–84.
Feldman, Jack M. 1981. Beyond attribution theory: Cognitive biases in performance appraisal. Journal of
Applied Psychology 66:127–48.
Franklin, Aimee L. 2000. An examination of bureaucratic reactions to institutional controls. Public
Performance and Management Review 24:8–21.
Gardner, Donald D., L. L. Cummings, Randall B. Dunham, and Jon L. Pierce. 1998. Single-item versus
multiple-item measurement scales: An empirical comparison. Educational and Psychological
Measurement 58:898–915.
Hatry, Harry. 1999. Performance measurement: Getting results. Washington, DC: Urban Institute.
Heikkila, Tayna, and Kimberley Roussin Isett. 2007. Citizen involvement and performance management
in special-purpose governments. Public Administration Review 67:238–48.
Heinrich, Carolyn J. 1999. Do government bureaucrats make effective use of performance information?
Journal of Public Administration Research and Theory 9:363–93.
Ho, Alfred Tat-Kei. 2006. Accounting for the value of performance measurement from the perspective of
city mayors. Journal of Public Administration Research and Theory 16:217–37.
Houston, David J. 2006.‘‘Walkingthe walk’’of publicservice motivation: Public employees and charitable
gifts of time, blood, and money. Journal of Public Administration Research and Theory 16:67–86.
864 Journal of Public Administration Research and Theory
Jennings, Edward T., and Meg Patrick Haist. 2006. Putting performance measurement in context. In The
art of governance, eds. Patricia W. Ingraham and Laurence E. Lynn, 173–94. Washington, DC:
Georgetown Univ. Press.
Kamensky, John. 1996. Role of the ‘‘reinventing government’’ movement in federal management reform.
Public Administration Review 56:247–55.
Kettl, Donald F., and Steven Kelman. 2007. Reflections on 21st century government management.
Washington, DC: IBM Center for the Business of Government.
Lewis, David E. 2008. The politics of presidential appointments: Political control and bureaucratic
performance. Princeton, NJ: Princeton Univ. Press.
Melkers, Julia, and Katherine Willoughby. 2005. Models of performance-measurement use in local
governments: Understanding budgeting, communication, and lasting effects. Public Administration
Review 65:180–90.
Moynihan, Donald P. 2003. Normative and instrumental perspectives on public participation: Citizen
summits in Washington, D.C. American Review of Public Administration 33:164–88.
———. 2005. Goal-based learning and the future of performance management. Public Administration
Review 65:203–16.
———. 2008. The dynamics of performance management: Constructing information and reform.
Washington, DC: Georgetown Univ. Press.
Moynihan, Donald P., and Noel Landuyt. 2009. How do public organizations learn? Bridging structural
and cultural divides. Public Administration Review 69:1097–105.
Moynihan, Donald P., Sergio Fernandez, Soonhee Kim, Kelly LeRoux, Suzanne J. Piotrowski, and
Bradley Wright. Forthcoming. Performance measurement amidst governance complexity. Journal of
Public Administration Research and Theory.
Moynihan, Donald P., and Patricia W. Ingraham. 2004. Integrative leadership in the public sector: A model
of performance information use. Administration & Society 36:427–53.
Moynihan, Donald P., and Sanjay K. Pandey. 2005. Testing how management matters in an era of
government by performance management. Journal of Public Administration Research and Theory
15:421–39.
Pandey, Sanjay K., and Edmund C. Stazyk. 2008. Antecedents and correlates of public service motivation.
In Motivation in public management: The call of public service, eds. James L. Perry and Annie
Hondeghem, 101–17. Oxford: Oxford Univ. Press.
Pandey, Sanjay K., David H. Coursey, and Donald P. Moynihan. 2007. Organizational effectiveness and
bureaucratic red tape: A multi-method study. Public Performance and Management Review
30:398–425.
Pandey, Sanjay K., Bradley E. Wright, and Donald P. Moynihan. 2008. Public service motivation and
interpersonal citizenship behavior: Testing a preliminary model. International Public Management
Journal 11:89–108.
Perry, James L. 1986. Merit pay in the public sector. Review of Public Personnel Administration 7:57–69.
———. 1996. Measuring public service motivation: An assessment of construct reliability and validity.
Journal of Public Administration Research and Theory 6:5–22.
Perry, James L., and Anne Hondeghem. 2008. Motivation in public management: The call of public
service. Oxford: Oxford Univ. Press.
Poister, Theodore H., and Gregory Streib. 1999. Performance measurement in municipal government:
Assessing the state of the practice. Public Administration Review 59:325–35.
Pollitt, Christopher. 2000. Is the emperor in his underwear? An analysis of the impacts of public man-
agement reform. Public Management Review 2:181–99.
Quinn, Robert E., and John Rohrbaugh. 1981. A competing values approach to organizational effec-
tiveness. Public Productivity Review 5:122–40.
Radin, Beryl. 2006. Challenging the performance movement: Accountability, complexity, and democratic
values. Washington, DC: Georgetown Univ. Press.
Rainey, Hal G. 1983. Public agencies and private firms: Incentive structures, goals, and individual roles.
Administration & Society 15:207–42.
Moynihan and Pandey The Big Question for Performance Management 865
Roy, Claude, and Francine Seguin. 2000. The institutionalization of efficiency-oriented approaches for
public service improvement. Public Productivity & Management Review 23:449–68.
Schick, Allen. 1997. The changing role of the central budget office. Paris, France: OECD.
———. 2001. Getting performance measures to measure up. In Quicker, better, cheaper: Managing per-
formance in American government, ed. Dall Forsythe, 39–60. Albany, NY: Rockefeller Institute.
Schulz, Martin. 2001. The uncertain relevance of newness: Organizational learning and knowledge flows.
Academy of Management Journal 44:661–81.
Spector, Paul E. 1997. Job satisfaction: Application, assessment, causes, and consequences. Thousand
Oaks, CA: Sage.
Tavits, Margit. 2006. Making democracy work more? Exploring the linkage between social capital and
government performance. Political Research Quarterly 59:211–25.
Van Dooren, Wouter. 2008. Nothing new under the sun? Change and continuity in twentieth century
performance movements. In Performance information in the public sector: How it is used, eds.
Steven Van de Walle and Wouter Van Dooren, 11–23. Houndmills, UK: Palgrave.
Van de Walle, Steven, and Tony Bovaird. 2007. Making better use of information to drive improvement in
local public services: A report for the Audit Commission. Birmingham, AL: School of Public Policy,
Univ. of Birmingham.
Van de Walle, Steven, and Wouter Van Dooren. 2008. Introduction: Using public sector information. In
Performance information in the public sector: How it is used, eds. Steven Van de Walle and Wouter
Van Dooren, 1–8. Houndmills, UK: Palgrave.
Wanous, John P., and Michael J. Hundy. 2001. A single-item reliability: A replication and extension.
Organizational Research Methods 4:361–75.
Weibel, Antoinette, Katja Rost, and Margit Osterloh. Pay for performance in the public sector-benefits and
(hidden) costs. Journal of Public Administration and Theory. Advance Access published May 27,
2009, doi:10.1093/jopart/mup009
Willis, James J., Stephen D. Mastrofski, and David Weisburd. 2007. Making sense of COMPSTAT.
Law & Society Review 41:147–88.
Williams, Richard. 2006. Generalized ordered logit/partial proportional odds model for ordinal dependent
variables. The Stata Journal 6:58–82.
Wright, Bradley E. 2001. Public sector work motivation: Review of current literature and a revised
conceptual model. Journal of Public Administration and Theory 11:559–86.
Yang, Kaifeng, and Jun Yi Hsieh. 2006. Managerial effectiveness of government performance mea-
surement: Testing a middle range model. Public Administration Review 67:861–79.
Zammuto, Raymond F., and Jack Y. Krakower. 1991. Quantitative and qualitative studies of organiza-
tional culture. Research in Organizational Change and Development 5:83–114.
Zients, Jeffery. 2009. Testimony before the Committee on Homeland Security and Governmental Affairs,
United States Senate. September 29, 2009
866 Journal of Public Administration Research and Theory
Copyright of Journal of Public Administration Research & Theory is the property of Oxford University Press /
UK and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright
holder's express written permission. However, users may print, download, or email articles for individual use.
