future prospects for performance management in chinese city governments

http://aas.sagepub.com/Administration & Society

http://aas.sagepub.com/content/42/1_suppl/34SThe online version of this article can be found at:

DOI: 10.1177/0095399710361853

February 2010 2010 42: 34S originally published online 5Administration & Society

Richard M. Walker and Jiannan WuGovernments

Future Prospects for Performance Management in Chinese City

Published by:

http://www.sagepublications.com

can be found at:Administration & SocietyAdditional services and information for

http://aas.sagepub.com/cgi/alertsEmail Alerts:

http://aas.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://aas.sagepub.com/content/42/1_suppl/34S.refs.htmlCitations:

What is This?

- Feb 5, 2010 OnlineFirst Version of Record

- May 13, 2010Version of Record >>

at Bobst Library, New York University on October 18, 2014aas.sagepub.comDownloaded from at Bobst Library, New York University on October 18, 2014aas.sagepub.comDownloaded from

http://aas.sagepub.com/

http://aas.sagepub.com/content/42/1_suppl/34S

http://www.sagepublications.com

http://aas.sagepub.com/cgi/alerts

http://aas.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://aas.sagepub.com/content/42/1_suppl/34S.refs.html

http://aas.sagepub.com/content/42/1_suppl/34S.full.pdf

http://aas.sagepub.com/content/early/2010/02/05/0095399710361853.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml



Administration & Society42(1S) 34S –55S

© 2010 SAGE PublicationsDOI: 10.1177/0095399710361853

http://aas.sagepub.com

Future Prospects for Performance Management in Chinese City Governments

Richard M. Walker1,2 and Jiannan Wu3

Abstract

Contrasting views are presented on the nature and rationale for public management reforms in China. One school argues that they are strongly influenced by international practice, whereas the other holds that they reflect the characteristics of Chinese politics and administration. The authors explore this proposition in the arena of performance management, where they pay particular attention to whether future prospects for performance management are likely to be associated with international practice or Chinese characteristics. The authors’ concern with the future leads them to implement a Delphi study focusing on performance management in city governments, from which they find that performance management has particular Chinese characteristics and that it will continue to be used as a control mechanism. However, expert respondents also indicate the desirability of orienting performance management and evaluation outward to citizens and other key stakeholders, and of enhancing service delivery and performance, in keeping with the international performance management movement. The authors conclude by discussing the research and practice implications of these findings.

Keywordsperformance management, performance evaluation, Chinese city governments, Delphi method

1 University of Hong Kong, China2 Cardiff University, Wales, United Kingdom3 Xi’an Jiaotong University, China

Corresponding Author:Richard M. Walker, Department of Sociology and Kadoorie Institute, 8/F, TT Tsui Building, University of Hong Kong, Pokfulam Rd., Hong Kong, China Email: [email protected]

361853AAS10.1177/0095399710361853Walker and WuAdministration & Society

at Bobst Library, New York University on October 18, 2014aas.sagepub.comDownloaded from


Walker and Wu 35S

In the past 30 years, questions about the performance of public organizations have become prevalent internationally. Performance evaluation, perfor-mance management, or managing for results have become key tools for rais-ing the performance of public agencies, and they remain a central concern in the field of public management research and practice (Behn, 2003; Moynihan, 2006; OECD, 2005; Pollitt & Bouckeart, 2006). Within nations across the globe, the adoption of the New Public Management reforms and management practices from the private sector has accelerated the trend. This change has moved beyond the Anglo-Saxon countries where it was first popularized, and a now-notable example is China, where various forms of performance evaluation have been implemented (Chan & Gao, 2008, 2009; Walker & Moon, 2007; Westcott & Jones, 2007). Although the measurement of performance is complex, and the adoption of performance measurement and evaluation techniques demanding, it has been increasingly accepted as a management tool that permits more informed decisions and has the intention of achieving improvement in public service. It remains an important area therefore for systematic research.

There has been an explosion of performance evaluation in Chinese cities in recent years (Chan & Gao, 2008). The research thus far examines perfor-mance management and evaluation in one locality (Dai & Teng, 2008; Gao, 2007) or for single agencies (Lui, 1995) or focuses on one segment of the performance management system (Chen, Li, & Zhou, 2005). The detailed research questions in this study examine (a) why and how performance man-agement regimes are used, (b) which stakeholders are involved in the pro-cess, and (c) what dimensions of performance are prioritized. In answering these questions, we seek to contribute toward a debate at the heart of public management reform by examining whether current practices and future pri-orities are driven by the international trend toward performance management or by internal concerns within individual countries. This is an important question for reform in China, given China’s political and administrative sys-tem, and a debate has emerged around this issue (Chan & Chow, 2007; Chan & Gao, 2009; Foster, 2006; Ngok & Zhu, 2007). Because we are oriented toward the future and the ways in which performance management priorities may change in the short term, we have adopted the Delphi method. This widely utilized forecasting technique is employed when information about the future is uncertain and requires subjective data. The method tackles prob-lems of uncertainty by drawing on the knowledge of experts and seeking consensus on the topic under investigation.

We have structured the balance of the study as follows. We review the literature on performance management and evaluation, with a focus on the



36S Administration & Society 42(1S)

Chinese context of our study, and tease out evidence on the influences that contributed to the adoption of management reforms. We then discuss the Del-phi method and our samples and measures before we present the findings of our study. These findings indicate the importance of both the local context and international processes; we identified both an ongoing emphasis on man-aging upward to higher levels of political authority and downward within the organization, but we also note a desire to manage outward toward citizens, which reflects the emphasis on accountability found in performance manage-ment elsewhere in the world. Finally, we discuss the implications of these findings, which strengthen the argument for the use of contingency frame-works, in the conclusion.

Management Reform and Managing Performance in ChinaResearch reporting the development of performance management in China presents two somewhat contrasting views. The Chinese perspective is of specific strategies and tools being developed in response to local circumstances (Chan & Chow, 2007; Chan & Gao, 2009; Gao, 2009; Gong, 2009). However, a number of observers see management reform, and performance management in particular, as a response to economic liberaliza-tion and the resulting decentralization of systems, which requires the adop-tion of a range of management practice that are influenced by international practice (Caulfield, 2006; Chou, 2008; Christensen, Dong, & Painter, 2008; Foster, 2006; Painter, 2008). Indeed, Ngok and Zhu (2007) go so far as to argue that “the entire process of administrative reform, ‘with Chinese char-acteristics,’ as many scholars often qualify it, has coincided with the world administrative reform trend” (p. 218).

The international literature contends that performance management has been adopted as a mechanism to attain accountability, whether this is politi-cians seeking to hold bureaucrats accountable or for purposes of internal management, and to enhance the efficiency and effectiveness of government and increase its responsiveness (OECD, 2005). This line of reasoning draws on public choice theory, which argues that information is vested in bureau-crats and the provision of performance data to politicians and citizens can reduce the information asymmetry (Boyne, Farrell, Law, Powell, & Walker, 2003). To use Moore’s (1995) terminology of managing upward, downward, and outward, performance management is used across all these directions—upward from the bureaucracy to provide information to politicians, who in turn use it downward to hold officials to account; it is also used outward to



Walker and Wu 37S

permit citizens to question politicians and their servants. Offering empirical support of the internationalization hypothesis, Foster (2006) presents a detailed case study of the implementation of, and subsequent nationwide adoption of, the Service Promise System in the city of Yantai. Although Fos-ter (2006) notes the “CCP’s [Chinese Communist Party’s] continuing aver-sion to being bound by clear and enforceable rules and by public opinion” (p. 241), he ultimately concludes that the Service Promise System is an interna-tional tool from the New Public Management rubric and represents “a serious attempt to make the bureaucracy more customer-orientated and professional” (p. 221). Foster’s study, then, sits within broader arguments and justifications for the adoption of performance management (also see Caulfield, 2006; Christensen et al., 2008).

Three arguments associated with Chinese characteristics have been pre-sented for the adoption of performance management: to manage central–local relations, to enhance state capacity, and to manage and control local officials (Gao, 2009). Performance management is the modern tool the Chi-nese central government has adopted to exert influence over the many admin-istrative levels of government and, in particular, over the great distance between the central and local governments in the policy-implementation chain. The structure of Chinese government means that performance man-agement is concerned not only with service performance but also with main-taining a stable, harmonious society and promoting economic and sustainable development. The hierarchical nature of government means that performance management is “used chiefly to ensure that local officials comply with higher-level policy priorities” and that “by binding local officials’ target accomplishments to their career future the target-based responsibility system guarantees that local officials will follow the directives that come down from above” (Chan & Gao, 2008, p. 8).

The primary aim within China is to manage upward to the higher levels of government and downward to ensure that the local level implements nation-ally identified performance targets. This suggests that there is little drive attached to managing outward for “accountability,” with the ultimate aim of “enhanced outcomes for society” (OECD, 2005, p. 56). This assertion can be illustrated with evidence from Gao’s (2009) case study in Zhouzhi County in Xi’an City. Gao’s study concludes that performance management has been introduced to assist with state capacity building within government and to enhance policy implementation. The performance targets used are able to achieve the desired outcomes because they specify the policy objectives of a higher level of government, which holds lower level cadres accountable for ensuring that CCP policies are implemented (Caulfield, 2006; Gao, 2007).




These outcomes are achieved through the Objective Responsibility System (ORS), which was initially promulgated in the mid-1980s to meet the needs of Chinese government (Chan & Gao, 2008).

The conflation of politics and bureaucracy in the Chinese system of gov-ernment has important implications for the role of stakeholders and their con-sequent ability to influence performance management. We draw on the work of Mitchell, Agle, and Wood (1997, p. 869), who argue that stakeholders express three characteristics. Power implies the ability of one actor to influ-ence another actor and to get that individual to do something that the indi-vidual would otherwise not have done, by means that may be coercive or normative or via incentives. Legitimacy derives from norms, values, and beliefs in a social system and promotes the authority of some stakeholders over others. Finally, different stakeholders are able to confer urgency on a public agency; urgency is likely to be measured by the time and extent to which it is acceptable to delay responding to the stakeholder.

The “Chinese characteristics” hypothesis suggests that internal stakehold-ers are those most likely to be powerful, legitimate, and have urgency. If this is the case, the main pattern of management would be accountability upward to higher levels of government and management downward to direct improved performance because civil servants and politicians are conflated into one body, as the CCP shadows bureaucratic positions. In this unitary and highly concentrated system, conflicts of interest can, nonetheless, arise. For example, audit bodies are not structurally independent; they can be commis-sioned by and report to the same agency or be required to audit departments that are responsible for personnel matters within the audit agency (Li, 2009). Alternatively, decentralization has increased the freedom of local cadres, which has resulted in them promoting their own interests above those of the party and national policies (Gao, 2009). The role of external stakeholders in public management processes has been shown to be limited, but useful to assist with capacity building (Li, 2009). On the other hand, there is some evidence that a wider set of factors may drive current practice and future priorities—regimes also now focus on citizens and have developed charac-teristics of outward management, in keeping with international practices (Christensen et al., 2008).

Method“The Delphi Method is an exercise in group communication among a panel of geographically dispersed experts . . . [it] allows experts to deal sys-tematically with a complex problem or task” (Ziglio, 1996, p. 9). The key



Walker and Wu 39S

characteristics of the Delphi method include repetitive processes, panelist anonymity, controlled feedback, and a group statistical response (Alder & Ziglio, 1996; Linstone & Turoff, 2002). The Delphi method structures and refines large amounts of information to achieve and improve judgments and decision making. The notion of informed judgment is central to policy mak-ing and to the Delphi methodology, which seeks to acquire systematic con-text-dependent information from experts by using rounds of questionnaires (Engles & Powell Kennedy, 2007). Considerable expertise now exists on the Delphi method, and it is being widely used in the social sciences (Landeta, 2006) and among public administration scholars (see Critcher & Gladstone, 1998; Proeller, 2006; Torres, 2005).

We adopted the Delphi method to examine our research questions because it allows us both to draw a baseline of current practice from expert actors and to consider the direction of change in the future—the alternative is to look backward, which is replete with bias and recall errors. We focus on future scenarios to ascertain the extent to which respondents prioritize the more international aspects of performance management as against Chinese prac-tices, without differentiating between these aims, and to draw on their informed judgment about their locality to achieve a wider view for China more generally. Surveying experts in this way also involves balancing researcher and subject bias, that is, ensuring that respondents are impartial while having an interest in the research topic. Subject bias may be exacer-bated where individuals responding to the survey are likely to be affected by decisions made within the policy area under investigation. However, one important aspect of the Delphi approach is that it seeks to bring about con-sensus between participants by sharing information. Recognizing this key difference between Delphi and other research methods assists in addressing some of these concerns.

In this study we sampled all members of the Performance Government Management Research Association, a group established by Chinese Public Administration Society under the auspices of the Office of the State Council and Ministry of Civil Affairs. Its aims are to promote research on and the practice of performance management in the public sector. Membership is based on expertise in performance management through academic reputation or civil service position. There are a total of 151 members—80 practitioners, 66 academics, and 5 from enterprises.

We invited all association members to join the Delphi process and received positive responses from 34 members, split equally between academics and practitioners. Reasons for declining the invitation ranged from being too busy to concerns about the sensitivity of the topic. Attrition in Round 2




reduced our respondent pool to 26 experts, and analysis is based on these matched respondents. Panel members were drawn from different economic regions of China: 14 from the east (including Beijing and Shanghai), and 6 each from mid-China (e.g., Hubei and Jiangxi provinces), and 7 from west China (including Chongqing and Shaanxi provinces). Of the academic par-ticipants, 11 were full professors and 3 were associates, while the practitio-ners split half and half between departmental and divisional levels. All panelists had published on the topic, with a mean of 5.4 articles for the aca-demics and 3 for the practitioners. The assessments by academics and practi-tioners were similar, with only a handful of statistically significant differences.

Delphi statements were written in English and translated to Chinese prior to the survey being administered. To ensure the translation was accurate, the Chinese-language version of the survey was translated back into English to ensure consistency in interpretation. This exercise led to a number of small changes made in the characters used in the final Chinese-language version of the survey.

We draw our measures from our reading in performance measurement literature and current practice within China. (The statements in the Delphi appear in Table 2 through 5.) The first part of the survey we report examined “Evaluation reasons: Why undertake performance evaluation in your city?” Detailed measures were taken from Behn’s (2003) study on the purposes of performance measurement and were supplemented with questions of innova-tion (service, process, and partnership) and cost reduction. We then moved on to examine “Results application: Use of the performance evaluation ranking and rating results in your city,” for which items were drawn from knowledge of current practice among Chinese city authorities. “Stakeholders: Different groups of key actors consulted during your city’s performance evaluation,” were examined in the third part of the study, where we quizzed respondents about a range of key stakeholders inside and outside the organization (Andrews, Boyne, & Walker, 2006). In the final part of the survey we exam-ined the dimensions of performance used in each city government (see Boyne, 2002). In each section, respondents were given the option of supple-menting the closed questions we asked.

Three questions, on a 5-point Likert-type scale, were posed against each statement (see Table 1). In Round 1, respondents were asked about current priorities to establish a baseline against which the future could be judged. Questions 2 and 3 examined future prospects—desirability and feasibility. These two items were also included in Round 2, where we also probed respondents on the extent to which they were confident in their assessments



Walker and Wu 41S

Table 1. Instructions to Respondents

Item Questions Measurement scale

Current priority

To what extent are the below uses of the performance evaluation results a key priority now?

1 = not a priority and 5 = main priority

Future desirability

How desirable is it that the below uses of perfor-mance evalu-ation results be extensively included in the next 3 years?

1 = not desirable and 5 = desirable

Future feasibility

What is the feasibility of the below uses of performance evaluation results being extensively included in the next three years?

1 = low and 5 = high

Confidence How confident are you in your predication of future feasibility?

1 = unreliable (great risk of being wrong, of no use as a decision basis)

2 = risky (substantial risk of being wrong, not willing to make a decision based on this alone)

3 = reliable (some risk of being wrong, will-ing to make a decision based on this but recognizing some chance of error)

4 = certain (low risk of being wrong, deci-sions based on this will not be wrong because this is a “fact”)

of feasibility. Delphi panel members were instructed to “think about perfor-mance evaluation as it is currently developed in your local city” and consider this “over the next three years.”




Two rounds were conducted: Round 1 in February 2008 and Round 2 in June the same year. We sought consensus on the questions concerning the future and posed these questions in Round 1 and 2. In Round 2 we fed back to each participant the group mean, together with his or her own score, and asked if they wished to change their rating.

FindingsFindings are presented in four parts: reasons for the use of performance man-agement, results application, stakeholders, and dimensions of performance. Each table presents mean scores for current priorities, future desirability and future feasibility, and the percentage agreement reached after two rounds. We do this by indicating the percentage of respondents who gave a response of 4 or 5 on the 5-point Likert-type scale and signal agreement by a simple major-ity. Given that our purpose was to look at current priorities and future pros-pects, we also report differences of means tests to identify if the gap between current priorities and future desirability and future feasibility is statistically significant. Finally, we commented on the confidence of our respondents in relation to future feasibility for each question they answered.

The means, percentage in agreement, and t-test results for reasons for undertaking performance evaluation are reported in Table 2. Top current pri-orities for undertaking performance management were focused on managing upward and downward; however, there was some evidence of managing out-ward (Moore, 1995). Managing upward was identified as the first- and sec-ond-highest priorities (“How can a city government convince political superiors and legislators that it is doing a good job?” and “How can city gov-ernment improve sustainable development?”). Tied third was a measure of internal control (“How can leaders ensure that subordinates are doing the right thing?”), reflecting the importance of managing downward to ensure that city governments got the job done, and process innovation (“How can governments achieve new ways to deliver services, e.g., IT, one-stop ser-vices?”), which attaches some importance to managing outward to citizens through innovation. Managing outward also figured as an important reason for undertaking evaluation, and Delphi participants ranked service innova-tion fifth (“How can the city government put in place new services to citi-zens?”). Although these were the highest-ranked items, none of the scores exceed 4 on the Likert-type scale.

The lowest priorities were found for measures that operationlized “learn-ing” (“Why is what working or not working?”) and “budgeting” (On what fields, groups, or projects should the government spend the public’s



43S

Tabl

e 2.

Mea

ns, S

tand

ard

Dev

iatio

ns, a

nd P

erce

ntag

e in

Agr

eem

ent

for

Rea

sons

to

Eval

uate

Eval

uatio

n re

ason

sC

P, M

(SD

)FD

, M

(SD

)FD

>3%

FF,

M (

SD)

FF >

3%C

, M

Prom

ote:

How

can

a c

ity g

over

nmen

t co

nvin

ce p

oliti

cal

supe

rior

s an

d le

gisl

ator

s th

at it

is d

oing

a g

ood

job?

3.71

(1.

04)

3.96

(0.

62)

793.

72 (

0.94

)56

3.00

Prom

ote:

How

can

a c

ity g

over

nmen

t im

prov

e su

s-ta

inab

le d

evel

opm

ent?

3.59

(1.

01)

4.08

(0.

80)

733.

54 (

0.95

)46

3.00

Prom

ote:

How

can

lead

ers

ensu

re t

hat

subo

rdin

ates

ar

e do

ing

the

righ

t th

ing?

3.48

(1.

16)

3.77

(0.

99)

653.

77 (

0.76

)65

3.08

Proc

ess

inno

vatio

ns: H

ow c

an a

city

gov

ernm

ent

achi

eve

new

way

s to

del

iver

ser

vice

s, e.

g., I

T, o

ne-

stop

ser

vice

?

3.48

(1.

16)

4.00

* (0.

89)

774.

00 (

0.75

)81

3.16

Prom

ote:

How

can

a c

ity g

over

nmen

t en

hanc

e th

e sa

tisfa

ctio

n of

citi

zens

?3.

38 (

0.92

)3.

81* (

0.80

)65

3.54

(0.

81)

503.

12

Serv

ice

inno

vatio

n: H

ow c

an a

city

gov

ernm

ent

put

in

plac

e ne

w s

ervi

ces

for

citiz

ens?

3.36

(1.

15)

4.23

* (0.

82)

853.

73* (

0.72

)65

3.04

Mot

ivat

e: H

ow c

an le

ader

s m

otiv

ate

empl

oyee

s to

do

the

thin

gs n

eces

sary

to

impr

ove

perf

orm

ance

?3.

32 (

1.18

)3.

81* (

0.80

)65

3.50

(0.

76)

502.

96

Acc

ount

abili

ty: T

o ta

ke r

espo

nsib

ility

for

actio

ns, i

n pa

rtic

ular

, mis

take

s3.

32 (

1.35

)3.

96* (

0.92

)77

3.68

(0.

80)

483.

04

Red

uce

loca

l cos

ts: E

nsur

ing

that

adm

inis

trat

ive

cost

s ar

e no

t to

o hi

gh3.

24 (

1.16

)4.

00* (

0.94

)77

3.65

(0.

75)

502.

96

Part

ners

hip

inno

vatio

ns: H

ow c

an a

city

gov

ernm

ent

wor

k w

ith o

ther

age

ncie

s an

d bo

dies

to

deliv

er

inno

vatio

ns?

2.76

(1.

05)

3.50

* (0.

76)

503.

04 (

0.82

)19

2.69

Budg

et: O

n w

hat

field

s, gr

oups

, or

proj

ects

sho

uld

a ci

ty g

over

nmen

t sp

end

the

publ

ic’s

mon

ey?

2.72

(1.

17)

3.69

* (0.

79)

583.

50* (

0.86

)50

2.92

Lear

ning

: Why

is w

hat

wor

king

or

not

wor

king

?2.

63 (

1.01

)3.

58* (

0.86

)62

3.42

* (0.

81)

382.

84

Not

e: C

P =

cur

rent

pri

ority

; FD

= fu

ture

des

irab

ility

; FF

= fu

ture

feas

ibili

ty; C

= c

onfid

ence

.*p

< .0

5 (t

-tes

t re

sults

for

diffe

renc

e C

P an

d FD

and

CP

and

FF).




money?”). The lack of attention on “learning” is perhaps an indication of the way in which higher levels of government drive the performance evaluation regime and the priorities of the cities being examined. The low priority accorded to “budgeting” is likely a reflection of the gap between resource allocation and performance—the unitary system leads to top–down perfor-mance target setting and bottom–up resource allocation.

Assessments of future desirability saw four of these five items retain their high rank, the exception being on downward management (“How can leaders ensure that subordinates are doing the right thing?”) The reason “Ensuring that administrative costs are not too high” was seen as desirable in the future, but not feasible, indicated that respondents did not see cost pressures as a reason for undertaking performance evaluation. Reasons for undertaking per-formance management identified as feasible in the future placed emphasis across the process of managing upward, downward, and outward. We would have expected that managing upward and downward would be the most important reasons for undertaking performance management, given our prior discussion. The two items measuring this (“How can a city government con-vince political superiors and legislators that it is doing a good job” and “How can leaders ensure that subordinates are doing the right thing”) generated high future feasibility ranks, but not future desirability. These findings sup-port existing evidence on the role of performance management in controlling lower levels of government, but the lack of statistical significance between current practice and future priorities implies resigned acceptance.

Service and process innovations, ranked as the top five current priorities, were seen as desirable in the future and feasible to implement, suggesting that finding new ways to deliver services is likely to become an important reason for undertaking performance management in the short term. Innova-tion at the local government level in China has become important, particu-larly with the sponsorship of provincial- and national-level innovation awards. The respondents saw service innovations as more feasible to imple-ment than process innovation (see t-test results), and they were confident in these responses. These findings are somewhat at odds with research suggest-ing that performance management is used only as a system of control and capacity building. A further factor at odds with prior evidence on the ratio-nale for performance management is the lack of future feasibility accorded to sustainable development. Sustainable development ranked high for future desirability, but not for future feasibility. Sustainable development is a major CCP policy priority, yet it is not seen as a reason for undertaking performance management. The fifth-highest-ranked reason for the feasibility for undertak-ing performance management was “to take responsibility for actions, in



Walker and Wu 45S

particular mistakes,” implying a growing role for accountability within per-formance management.

Our survey participants were less likely to agree on future feasibility, with eight items recording an agreement score of 50% or below. However, agree-ment was highest for the top four items, which implies that the role of perfor-mance management goes beyond managing upward and downward alone. Respondents offered scores greater than 3 for their assessments of feasibility, indicating that they felt their projections to be reliable.

Table 3 presents the assessments of the Delphi participants for the pros-pects of result application. Current priorities typically related to downward management: “implementation, reward and punish,” “achieve performance improvement,” “communicate and coordinate objectives to subordinates,” and to “motivate staff”— emphasizing the “command and control” processes argued to typify Chinese performance management. These items remained highly rated by the Delphi panelists for future desirability and feasibility, achieving high levels of agreement, with confidence ratings greater than 3. A wider role for these items was also envisioned, including human resources (“motivation for staff”) and “achieve performance improvement.” The per-spective of managing outward (“open evaluation results to the public”) received a high rating, moving from near the bottom of current priorities to being the highest rated for future desirability and feasibility.

The experts were all of the opinion that in comparison with the current priorities means, all uses of performance management results were desirable in the future; the mean scores ranged from 3.50 to 4.46, and all were statisti-cally significant. However, the participants were slightly more circumspect about future feasibility, for 7 of 11 means were significantly different from current priorities. The four items given the lowest rank for current priorities received similarly low ranks for future feasibility. Confidence scores were high for top-rated feasibility measures—all were greater than 3, meaning they are seen to be reliable.

Delphi panel participants were asked to consider priorities and desirability and feasibility for 6 external and 13 internal stakeholders (Table 4). Current priorities were clear: the three CCP organs and the People’s Congress scored highest, with mean scores greater than 4. This is not unexpected given the key roles played by the CCP committees in policy and personnel appoint-ment and the important role played by the People’s Congress in approving the annual performance report. The mayor, two bureaus, and the next highest level of government followed closely behind. The external stakeholders trailed the majority of the internal actors by some distance, with mean scores between 2 and 3. These results reaffirm the importance of the political and



46S

Tabl

e 3.

Mea

ns, S

tand

ard

Dev

iatio

ns, a

nd P

erce

ntag

e in

Agr

eem

ent

for

Use

of R

esul

ts

Res

ult

appl

icat

ion

CP,

M (

SD)

FD, M

(SD

)FD

>3%

FF, M

(SD

)FF

>3%

C, M

Impl

emen

t, re

war

d, a

nd p

unis

h3.

80 (

1.12

)4.

35* (

0.80

)81

4.19

(0.

90)

773.

15To

ach

ieve

per

form

ance

impr

ovem

ent

3.56

(1.

08)

4.46

* (0.

71)

884.

19* (

1.13

)73

3.23

Dia

gnos

e an

d co

ntro

l pro

blem

s3.

48 (

1.16

)4.

15* (

0.83

)81

3.65

(1.

09)

652.

92C

omm

unic

ate

and

coor

dina

te o

bjec

tives

to

subo

rdin

ates

3.32

(0.

85)

4.15

* (0.

78)

853.

92* (

1.02

)73

3.08

Mot

ivat

e st

aff

3.32

(1.

07)

3.96

* (0.

60)

813.

77* (

0.82

)62

3.12

Adj

ust

wor

k ar

rang

emen

t3.

20 (

1.15

)3.

69* (

0.79

)50

3.69

* (0.

79)

583.

12A

lloca

te o

rgan

izat

iona

l res

ourc

es3.

12 (

1.05

)3.

81* (

0.85

)54

3.62

* (0.

98)

462.

92O

rgan

izat

iona

l lea

rnin

g an

d de

velo

pmen

t3.

04 (

1.06

)3.

81* (

0.90

)58

3.54

(0.

81)

503.

00O

pen

eval

uatio

n re

sults

to

the

publ

ic3.

00 (

1.19

)4.

54* (

0.65

)92

4.19

* (0.

98)

773.

19Su

pply

tra

inin

g op

port

uniti

es2.

96 (

1.31

)3.

50* (

0.99

)58

3.27

(0.

87)

352.

96Pr

epar

e fo

r bu

dget

ing

2.88

(1.

09)

3.77

* (0.

95)

583.

38* (

0.90

)35

2.77

Not

e: C

P =

cur

rent

pri

ority

; FD

= fu

ture

des

irab

ility

; FF

= fu

ture

feas

ibili

ty; C

= c

onfid

ence

.*p

< .0

5 (t

-tes

t re

sults

for

diffe

renc

e C

P an

d FD

and

CP

and

FF).



47S

Tabl

e 4.

Mea

ns, S

tand

ard

Dev

iatio

ns, a

nd P

erce

ntag

e in

Agr

eem

ent

for

Stak

ehol

ders

Con

sulte

d

Stak

ehol

ders

CP,

M (

SD)

FD, M

(SD

)FD

>3%

FF, M

(SD

)FF

>3%

C, M

Exte

rnal

sta

keho

lder

s

Aca

dem

ic e

xper

ts3.

33 (

1.13

)4.

46* (

0.65

)92

4.19

* (0.

90)

773.

35

Busi

ness

: sta

te e

nter

pris

es3.

29 (

0.86

)3.

65* (

0.63

)65

3.35

(0.

75)

503.

00

Busi

ness

: pri

vate

firm

s2.

72 (

1.02

)3.

62* (

0.90

)50

3.35

* (1.

02)

422.

92

Indi

vidu

al u

sers

2.52

(1.

08)

3.92

* (0.

91)

723.

50* (

0.93

)46

3.04

C

itize

ns: t

he g

ener

al p

ublic

2.48

(1.

08)

3.92

* (0.

93)

773.

35* (

1.02

)46

2.96

W

orke

rs u

nion

2.28

(1.

02)

3.31

* (1.

01)

423.

00* (

0.57

)28

3.04

Inte

rnal

sta

keho

lder

s

CC

P co

mm

ittee

4.32

(0.

90)

4.27

(0.

96)

734.

31 (

0.84

)85

3.23

C

CP

disc

iplin

e in

spec

tion

4.20

(0.

91)

4.27

(0.

87)

814.

46 (

0.71

)88

3.44

C

CP

orga

niza

tion

depa

rtm

ent

4.16

(1.

03)

4.36

(1.

04)

844.

40 (

0.83

)88

3.28

Pe

ople

’s co

ngre

ss4.

04 (

0.81

)4.

23 (

0.76

)88

4.23

(0.

86)

813.

27M

ayor

4.00

(1.

14)

4.23

(0.

76)

884.

23 (

0.99

)85

3.27

A

udit

bure

au3.

92 (

1.15

)4.

12 (

0.78

)84

4.08

(0.

98)

733.

40

Pers

onne

l bur

eau

3.92

(0.

95)

3.65

(1.

06)

653.

84 (

0.90

)68

3.24

Pr

ovin

cial

gov

ernm

ent

3.92

(1.

04)

4.08

(1.

09)

814.

15 (

0.88

)77

3.19

St

atis

tics

bure

au3.

76 (

1.05

)4.

35* (

0.85

)85

4.35

* (0.

80)

813.

42

Supe

rvis

ion

bure

au3.

72 (

0.84

)4.

35* (

0.80

)81

4.15

* (0.

83)

813.

19

Peop

le’s

cons

ultin

g co

nfer

ence

s3.

52 (

1.19

)4.

04* (

0.92

)77

3.88

(1.

11)

653.

15

Cou

nty

or d

istr

ict

gove

rnm

ents

3.44

(0.

96)

3.77

(0.

95)

653.

81* (

0.90

)65

3.31

Fi

nanc

ial b

urea

u3.

17 (

0.89

)4.

12* (

0.65

)92

3.92

* (0.

76)

923.

19

Not

e: C

P =

cur

rent

pri

ority

; FD

= fu

ture

des

irab

ility

; FF

= fu

ture

feas

ibili

ty; C

= c

onfid

ence

; CC

P =

Chi

nese

Com

mun

ist

Part

y.*p

< .0

5 (t

-tes

t re

sults

for

diffe

renc

e C

P an

d FD

and

CP

and

FF).




administrative arms of government, who possess power, urgency, and legiti-macy. CCP and government power is derived from its higher position in the hierarchy of government. The actions of the CCP and government are deemed to promote legitimacy through the norms, values, and beliefs within the sys-tem of government. The hierarchical nature of Chinese government confers urgency—for example, claims from the political arm of government call for immediate attention by part of the bureaucracy of government.

Assessments of future desirability promoted the external stakeholders, “academic experts,” to the most important group to be consulted. This how-ever does not represent a sea change in future desirability, because all other external stakeholders are rated below internal stakeholders, and for future feasibility, panelists rate academics as the seventh most feasible group to consult. There are, nonetheless, increased mean scores for the future desir-ability for external stakeholders in comparison to current priorities (all means are greater than 3, and the increases are statistically significant). The percent-age in agreement on future desirability are beyond a majority in all cases, except the Workers’ Union. The increases in the means for users, citizens, business, and academics may be interpreted to suggest that they have legiti-mate claims to be consulted in the operation of performance management regimes, but that they do not possess power or urgency. The mean scores for future feasibility, on first glance, would appear to suggest that external stake-holders will play an important role in the future—mean scores remain greater than 3 and are significantly different from current priorities. However, more detailed examination indicates that although it may be more feasible to involve them in the future, they are not ranked as high as internal stakehold-ers. For example, if external stakeholders are contrasted with internal ones they are ranked lowest except for academics.

Views of the expert Delphi panel were similar for future feasibility. Politi-cal organs remain the most highly rated, together with the statistics bureau. Statistically significant increases in mean scores were seen for county or dis-trict governments and three bureaus: financial, statistics, and supervision. It is important to note that our expert participants were not able to agree on the future feasibility of consulting external stakeholders (except academics) and were not confident in their assessments. Nonetheless, reasons for and uses of performance management pointed toward a greater outward orientation. If this continues, it would not be unreasonable to expect a growing role accorded to external stakeholders. However, in the short term, the evidence points toward the ongoing importance of managing upward to the CCP and bureaus.

Formal effectiveness, probity, citizen satisfaction, and accountability and quantity were the dimensions of performance most likely included in perfor-mance evaluations (Table 5). Experts saw citizen satisfaction, accountability, and responsiveness as the most desirable dimensions of performance to be



49S

Tabl

e 5.

Mea

ns, S

tand

ard

Dev

iatio

ns, a

nd P

erce

ntag

e in

Agr

eem

ent

for

Dim

ensi

ons

of P

erfo

rman

ce

Dim

ensi

ons

of p

erfo

rman

ceC

P, M

(SD

)FD

, M (S

D)

FD >

3%FF

, M

(SD

)FF

>3%

C, M

Form

al e

ffect

iven

ess;

e.g.

, GD

P gr

owth

ach

ieve

d;

perc

enta

ge o

f chi

ldre

n pa

ssin

g fo

rmal

exa

ms

3.96

(1.

00)

4.00

(0.

75)

733.

92 (

0.74

)69

2.92

Prob

ity; n

umbe

r of

bri

bed

polit

icia

ns o

r ci

vil

serv

ants

3.70

(1.

15)

4.00

(1.

02)

733.

65 (

1.02

)58

2.69

Citi

zen

satis

fact

ion;

e.g

., ov

eral

l sat

isfa

ctio

n w

ith

the

loca

l gov

ernm

ent

by c

itize

ns3.

58 (

1.32

)4.

38* (

0.64

)92

3.92

(0.

63)

773.

04

Acc

ount

abili

ty; e

.g.,

com

plai

nts

to lo

cal a

utho

rity

om

buds

man

; pub

lic p

erce

ptio

n of

the

ext

ent

to w

hich

the

may

or is

acc

ount

able

for

his

or

her

actio

ns

3.38

(0.

97)

4.15

* (0.

83)

813.

62 (

0.94

)62

2.58

Qua

ntity

; e.g

., nu

mbe

r of

ser

vice

rec

ipie

nts

3.38

(0.

88)

3.88

* (0.

71)

773.

77 (

0.76

)65

2.81

Qua

lity;

e.g.

, per

cent

age

of p

upils

tau

ght

in o

ver-

crow

ded

scho

ols

3.25

(0.

79)

3.69

* (0.

74)

543.

73* (

0.72

)58

2.73

Use

r sa

tisfa

ctio

n; e

.g.,

user

s’ s

atis

fact

ion

with

w

aste

col

lect

ion,

per

cent

age

of li

brar

y us

ers

satis

fied

with

sta

ff m

embe

rs a

nd h

ours

ope

n

3.21

(1.

06)

3.88

* (0.

65)

813.

77* (

0.71

)69

2.88

Res

pons

iven

ess;

e.g.

, the

res

pons

e tim

e of

em

er-

genc

y se

rvic

es3.

17 (

0.87

)4.

12* (

0.77

)85

3.85

* (0.

67)

772.

88

Equi

ty; e

.g.,

perc

enta

ge o

f res

iden

ts in

diff

er-

ent

grou

ps s

atis

fied

with

cul

tura

l and

leis

ure

faci

litie

s

3.13

(0.

99)

4.00

* (0.

80)

773.

54 (

0.76

)54

2.62

Cos

t pe

r un

it of

out

put;

e.g.

, cos

t of

hou

seho

ld

was

te c

olle

ctio

n pe

r ho

useh

old

2.46

(1.

22)

3.77

* (0.

86)

653.

31* (

0.79

)38

2.65

Not

e: C

P =

cur

rent

pri

ority

; FD

= fu

ture

des

irab

ility

; FF

= fu

ture

feas

ibili

ty; C

= c

onfid

ence

.*p

< .0

5 ( t

-tes

t re

sults

for

diffe

renc

e C

P an

d FD

and

CP

and

FF).




used in the future, with more than 80% of the experts agreeing on these assessments. Three of the dimensions of performance highly rated as having feasible use in the future focused on outward-looking dimensions of perfor-mance: citizen satisfaction, user satisfaction, and responsiveness. The remaining two high-rated dimensions of performance were formal effective-ness and quantity, suggesting an orientation toward service delivery. Expert respondents were not very confident in their assessments, except for citizen satisfaction, where the mean score was greater than 3. Interestingly, cost issues were the least important item, followed by equity—cost issues are prevalent internationally, but equity is typically absent from most perfor-mance management regimes.

ConclusionsIn this article, we have explored the short-term future for government perfor-mance management and evaluation in Chinese cities and posited two com-peting hypotheses for the factors influencing the use of performance management: Chinese characteristics, which would see an emphasis on man-aging upward and downward, or international practice, which would in addi-tion be associated with managing outward and other practices. Because our orientation was toward the future, we implemented a Delphi study. Findings support the Chinese characteristics thesis, because many facets of perfor-mance management regimes now and in the future are geared toward control and capacity building, as expressed through managing upward and down-ward and consulting internal stakeholders. However, the Delphi panel respondents did also see a wider role for performance management, one that embraced managing outward to citizens, innovating, and using human resource management practices to achieve performance improvements.

These findings reflect the nature and structure of Chinese local govern-ment, which seeks to control the behavior of lower levels of government, but they also suggest that traces of some rationales for performance management in Western countries may be penetrating the Chinese model. They also rein-force evidence presented elsewhere that China has not blindly adopted man-agement reforms from other places, and when they have been adopted, they have been blended to the local context (Chan & Chow, 2007; Ngok & Zhu, 2007). This research suggests the importance of contingency theory and indicates that one-size-fits-all approaches to public management reform are replete with problems (see Boyne et al., 2003). The adoption of perfor-mance management regimes in China is likely to be influenced by interna-tional processes and then made suitable for the local context. By achieving



Walker and Wu 51S

implementation suited to the local context, performance management regimes are more likely to deliver locally successful outcomes—the challenge for China is achieving this across a large, disparate country. Additional research on the mechanisms and adoption strategies of management reforms in Chi-nese agency would provide more systematic justification of such claims, while providing invaluable insights.

One of the central tensions for managers in Chinese local governments will be to balance desires for more outward and user/citizen-oriented approaches to and uses of performance management with expectations about who should be consulted during the processes. Wider evidence from else-where indicates that there is value in bringing these stakeholders into the process, that they can produce more deliberative results. Change over the coming years may see more information provided to the public; it will be interesting to see if the public in turn come to demand more from their city governments. Future research could examine this and build on this study’s results by investigating multiple cases in different provinces across China to ascertain whether the adoption and use of performance management varies by locality.

The Delphi technique could be of value to policy makers when looking to develop policy and consider future options. It is argued to be of particular use when data are soft and uncertain. Uncertainty resulting from the global eco-nomic crisis has become ever more apparent during the period of this research (2008-2009) and this insecurity is unlikely to change, which emphasizes the ongoing relevance of tools such as the Delphi method. Although the tech-nique requires certain skills and adroitness in its application, it is within the grasp of most public agencies. It can be used to assist with planning new poli-cies and strategies at any level of government. It is therefore a potentially viable new tool for the armory of public management in the search for more informed evidence-based policy and practice.

Having promoted the Delphi technique, we need to draw attention to a number of its limitations: credibility, applicability, and auditability (Engles & Powell Kennedy, 2007). Internal validity is based on the experience of the panelists: Did they find the results credible? Credibility is tested on an ongo-ing basis: Delphi panels are predicated on communication and feedback, for results are fed back to the panel at each stage. Although divergence of views is to be expected across the large geographical area examined in this project, panelists had the option to query the findings as they emerged. Auditability was achieved through the use of objective and transparent methods and con-sistent and systematic data collection, coding, and analysis. Readers looking for wider lessons from this study need also be aware of its external validity.




Chinese government and politics is of a particular type, and thus the detailed results are likely to be primarily applicable to this context. Having said this, our findings point toward the importance of the use of performance manage-ment to enhance the performance of public agencies and to facilitate manag-ing upward, downward, and outward. Applicability will be uncovered over time, as the pertinence of these findings become clear in the future practice of performance management in China. Clearly, this study may be limited by the relatively small number of experts on our panel.

The findings here provide support for the processes of managing upward and downward but point toward a subtle shift toward managing outward in performance management and evaluation. Although additional research is required to answer questions about the performance consequences of perfor-mance management in China, the views collected from the expert members of our performance management Delphi imply that these approaches and techniques have common effects irrespective of location.

Authors’ Note

We acknowledge the research assistance of graduate students in the School of Public Policy and Administration, and the following in particular: Zhongkai Huang, Chunping Hu, Lei Zhang, Yanru Huang, Bo Bai.

Declaration of Conflicting Interests

The authors declared no conflicts of interests with respect to the authorship and/or publication of this article.

Funding

The authors wish to acknowledge the China NFSC General Program (70573083; 70873092); New Century Excellent Talents Program, China Ministry of Education (NCET-06-0851) and University of Hong Kong Basic Research Grant (200711159049).

References

Alder, M., & Ziglio, E. (Eds.). (1996). Gazing into the oracle: The Delphi method and its application to social policy and public health. London, UK:Jessica Kingsley.

Andrews, R., Boyne, G. A., & Walker, R. M. (2006). Subjective and objective measures of organizational performance: An empirical exploration. In G. A. Boyne, K. J. Meier, L. J. O’Toole, Jr., & R. M. Walker (Eds.), Public services performance: Perspectives on measurement and management (pp. 14-34). Cam-bridge, UK: Cambridge University Press.



Walker and Wu 53S

Behn, R. (2003). Why measure performance? Different purposes require different measures. Public Administration Review, 63, 586-606.

Boyne, G. A. (2002). Concepts and indicators of local authority performance: An evaluation of the statutory frameworks in England and Wales. Public Money and Management, 22(2), 17-24.

Boyne, G. A., Farrell, C., Law, J., Powell, M., & Walker, R. M. (2003). Evaluating public management reform. Buckingham, UK: Open University Press.

Caulfield, J. L. (2006). Local government reform in China: A rational actor perspec-tive. International Review of Administrative Sciences, 72, 253-267.

Chan, H. S., & Chow, K. W. (2007). Public management policy and practice in west-ern China: Metapolicy, tacit knowledge, and implications for management inno-vation transfer. The American Review of Public Administration, 37, 479-498.

Chan, H. S., & Gao, J. (2008). Performance measurement in Chinese local govern-ments. Chinese Law & Government, 41(2-3), 4-9.

Chan, H. S., & Gao, J. (2009). Putting the cart before the horse: Accountability or per-formance? Australian Journal of Public Administration, 68(Suppl. 1), S51-S61.

Chen, Y., Li, H., & Zhou, L.-A. (2005). Relative performance evaluation and the turn-over of provincial leaders in China. Economic Letters, 88, 421-425.

Chou, B. K. P. (2008). Does “good governance” matter? Civil service reform in China. International Journal of Public Administration, 31, 54-75.

Christensen, T., Dong, L., & Painter, M. (2008). Administrative reform in China’s central government: How much ‘learning’ from the West’? International Review of Administrative Science, 74, 351-371.

Crticher, C., & Gladstone, B. (1998). Utilizing the Delphi technique in policy discussion: A case study of a privatized utility in Britain. Public Administration, 76, 431-449.

Dai, G., & Teng, W. (2008). Performance evaluation at local government depart-ment level in China: A case study of People’s Municipal Government of Hangzhou (PMGH) [Mimeo]. Nottingham Business School, UK: Nottingham Trent University.

Engles, T. C. E., & Powell Kennedy, H. (2007). Enhancing a Delphi study on family-focused prevention. Technological Forecasting and Social Change, 74, 433-451.

Foster, K. W. (2006). Improving municipal governance in China: Yantai’s pathbreak-ing experiment in administrative reform. Modern China, 37, 221-250.

Gao, J. (2007). Performance management in the People’s Republic of China during the market reform era: A case study of two counties in Shaanxi Province. Doctoral disserta-tion, Department of Public and Social Administration, City University of Hong Kong.

Gao, J. (2009). Governing by goals and numbers: A case study in the use of perfor-mance measurement to build state capacity in China. Public Administration and Development, 29, 21-31.

Gong, T. (2009). Institutional learning and adaptation: Developing state audit capac-ity in China. Public Administration and Development, 29, 33-41.




Landeta, J. (2006). Current validity of the Delphi method in social sciences. Techno-logical Forecasting and Social Change, 73, 467-482.

Li, L. C. (2009). Decision-making in Chinese local administrative reform: Path dependence, agency and implementation. Public Administration and Develop-ment, 29, 79-87.

Linstone, H. A., & Turoff, T. (Eds.). (2002). The Delphi method: Techniques and applications. Retrieved from http://is.njit.edu/pubs/delphibook/.

Lui, K. C. (1995). State control and performance evaluation: The case of a state-owned enterprise in China. International Journal of Public Sector Management, 8(6), 29-50.

Mitchell, R. K., Agle, B. R., & Wood, D. J. (1997). Towards a theory of stakeholder identification and salience: Defining the principle of who and what really counts. Academy of Management Review, 22, 853-886.

Moore, M. (1995). Creating public value. Strategic management in government. Cambridge, MA: Harvard University Press.

Moynihan, D. P. (2006). Managing for results in state government: Evaluating a decade of reform. Public Administration Review, 66, 77-89.

Ngok, K., & Zhu, G. (2007). Marketization, globalization and administrative reform in China: A zigzag road to a promising future. International Review of Adminis-trative Sciences, 73, 217-233.

OECD. (2005). Modernisation government: The way forward. Paris, France: Author.Painter, M. (2008). From command economy to hollow state? Decentralisation in

Vietnam and China. Australian Journal of Public Administration, 67, 79-88.Pollitt, C., & Bouckeart, G. (2006). Public management reform: A comparative anal-

ysis. Oxford, UK: Oxford University Press.Proeller, I. (2006). Trends in local government in Europe. Public Management

Review, 8, 7-29.Torres, L. (2005). Service charters: Reshaping trust in government—The case of

Spain. Public Administration Review, 65, 687-699.Walker, R. M., & Moon, M. J. (2007). Performance measurement in Asia: Measure-

ment, tools and practices. In OCED Asian Centre for Public Governance, Com-parative Studies for Better Governance in Asian Countries (pp. 45-92). Seoul, South Korea: OCED Asian Centre for Public Governance.

Westcott, C. G., & Jones, L. R. (2007). Managing for results and performance in Asia: Assessing reform initiative in the public sector. International Public Management Review, 8, 56-102.

Ziglio, E. (1996). The Delphi method and its contribution to decision-making. In M. Alder & E. Ziglio (Eds.), Gazing into the oracle: The Delphi method and its application to social policy and public health (pp. 3-34). London, UK: Jessica Kingsley.



Walker and Wu 55S

Bios

Richard M. Walker is a Professor of Public Management and Policy at the Universities of Hong Kong and Cardiff. His research focuses on public management (the determinants of performance, strategic management, innovation, and red tape), urban management, and environmental policy and sustainable development.

Jiannan Wu is Tengfei Professor of Xi’an Jiaotong University. His research interests include government performance and organization theory. He serves as the Associate Dean of the School of Public Policy and Administration.



future prospects for performance management in chinese city governments

Documents