Top Leadership and Performance of Quasi Government Organizations in Korea

Paper prepared for presentation at the 11th bi-annual Public Management Research Conference, Syracuse, NY, June 2-4, 2011.

Sanghee Park, Research Scholar, School of Policy, Planning, and Development,

University of Southern California selotus@gmail.com

Yoon Jik Cho,

Assistant Professor, Andrew Young School of Policy Studies, Georgia State University

yoonjik@gsu.edu

ABSTRACT

This research explores the determinants of organizational performance of quasi-

autonomous non-governmental organizations, namely quangos, in Korea. Although we

have observed the growth of the quangos in various policy fields, empirical studies

examining the top leadership and performance of quangos remain largely underdeveloped.

Using panel data of 13 quangos in Korea from 1999 to 2007, this research seeks to fill the

gap by examining how several characteristics of top management of quangos, including

politicization of the board, chief executive succession, and insider/outsider origin of chief

executives, affect their performance measured by performance evaluation index (PEI) and

consumer satisfaction index (CSI). Analysis results show that some leadership

characteristics have a significant positive effect on the PEI; while the CSI was more

influenced by contextual factors such as the presidential change and unemployment rate,

while leadership characteristics were not statistically significant.

1

Top Leadership and Performance of Quasi Government Organizations in Korea

INTRODUCTION

Quasi-autonomous non-governmental organizations (hereafter, quangos) have unique

characteristics. In the past, traditional bureaucracies and hierarchical structures took the

lead, but as the idea of New Public Management proliferated since the 1980s, quangos have

been increasing rapidly. Similar terms also include NDPBs (Non Departmental Public

Bodies), NGOs (Non Governmental Organizations), QAOs (Quasi-Autonomous

Organizations), intermediate organizations, and the third sector. While representing no

tangible or clear organizations, quangos provide a continuum on which a diverse range of

bodies can be placed (Flinders and Smith, 1999).

Quangos expend public money and fulfill public functions, but exist with some

degree of independence. Located somewhere between the public and private sector, these

organizations are relatively free from government regulations (Durant 1995), but are

within the reach of governmental influence and political pressures (Koppell 2003, 2007;

Lewis 2003; Bertelli 2005, 2006). For the lack of democratic connection between electors

and government (Skelcher 1998; Cole 2000), quangos have been suffered from the

potential moral hazard and inefficiencies (Lægreid and Roness 2003; Newland 1996; Nolan

2001). Thus, political control through executive appointment of leadership is one of the

essential means to ensure accountability and performance that is directly linked to the

betterment of society and the improvement of public service.

The top management team of quangos, the CEO and the board of directors, are

required to manage complex public service networks, but their leadership is susceptible to

2

both internal and external challenges and pressures (Ingraham and Van Slyke 2006). They

strive to be responsible to the performance of the organizations as well as legitimacy and

democratic accountability to elected officials (Denis et al. 2005). Therefore, it is inevitable

for them to cooperate with and get support from the government department and political

parties in order to elevate performance of quangos.

The major research question of this paper is, “Does top leadership of quangos

matter for performance?” More specifically, “Which characteristics of top management,

including chief executives and board of directors, influence performance of quangos?” Using

panel data of 13 quangos in Korea from a nine-year period, this article attempts to answer

the question. Main characteristics of top management include the degree of politicization of

the boards, chief executive succession, and insider/outsider origin of CEOs.

In the upcoming section, we provide a research framework by reviewing the roles

of top management in quangos and several performance models. Then, several hypotheses

regarding leadership and organizational characteristics will be presented as well as the

description of the data, measurements, and analytic method. Following this, we present the

theoretical model using panel data models and analysis results. It concludes with

implications, limitations, and topics for future research.

RESEARCH FRAMEWORK

Figure 1 illustrates the framework of this research. Three components, including

leadership, organizational, and contextual characteristics are mainly addressed to explain

organizational performance of quangos. As leadership characteristics, we consider three

factors: politicization of boards, chief executive succession, and insider/outsider career of

3

CEOs. Four organizational characteristics including organizational size, organizational age,

government funding proportion, and age of unions are included. Finally, change of

administration and annual average unemployment rate constitute contextual factors.

[Figure 1 about here]

Organizational Performance

Performance is a concept hard to catch, especially in public organizations which contains

multiple goals, principals, and stakeholders (Brewer and Selden 2000; Boyne et al. 2005;

Van Slyke and Alexander 2006). Nevertheless, as Moynihan and Pandey (2004: 421) said,

we are living in “an era of government by performance.” There has been steady concern

about performance and several scholars provided models explaining organizational

performance. For example, Lynn, Heinrich, and Hill (2000) provided a reduced-form model

explaining outcomes of governance with five elements which include environmental factors,

client characteristics, treatments, structures, and managerial roles and actions. Boyne

(2003) indicated five determining factors of performance: resources, regulation, market

structure, organization, and management. Recently, Cho and Ringquist (2011) use four

components in explaining perceived organizational performance: managerial,

organizational, environmental factors, and individual characteristics. Because this research

employs objective measures of organizational performance, not a perceived measure, the

three components, including leadership characteristics (reflecting managerial

characteristics), organizational characteristics, and contextual characteristics, are

considered for the analysis.

The relationship between public management and organizational performance has

4

been studied intensively (e.g. O’Toole and Meier 2003; Boyne 2003), and there is an overall

consensus that management is one of key determinants of performance (Boyne 2003;

Moynihan and Pandey 2004). To build and sustain qualified management of organizations,

the role of leadership cannot be underestimated (Hicklin et al. 2008; Meier et al. 2007;

Forbes and Lynn 2005; Moynihan and Pandey 2004; Yang and Hsieh 2007). In quangos,

both chief executives and board of directors play critical roles of leadership. Working as a

representative of organizations, a chief executive plays critical roles in diverse activities

such as setting up strategic planning, acquiring resources, and making alliances with

internal and external stakeholders (Boyne et al. 2005; Boyne and Dahya 2002). The board

of directors has the power to make important decisions to the organizations, such as hiring

and firing, monitoring and resolving conflicts among the interests (Nicholson and Kiel 2007;

Baysinger and Butler 1985). In summary, top management, including a chief executive and

board of directors, will have substantial influence on organizational performance, and

quangos are not an exception. Now we will review how several characteristics of top

management will make a difference in performance of quangos.

We also include organizational and contextual characteristics that can influence

performance of quangos. Organizational characteristics such as the number of full-time

employees, size of total revenue, organizational age, government fund proportion, and

internal voice makers such as unions will be included to the empirical model. Contextual

factors may have substantial effect on performance, especially in quasi public organizations.

Political pressure is one of the distinctive characteristics making significant changes in

quangos. Additionally, economic factors will determine the general condition of quangos’

performance, as quangos are in between the public and the private sector.

5

Leadership Characteristics

Board of Directors: Politicization

The board of directors in quangos has the responsibility of enhancing legitimacy and

integrity of public services while improving efficiency of management, where the board

composition and the degree of politicization of the board has been always an issue. Often

referred to the growth of positions occupied by political appointees (Lewis 2008),

politicization of public organizations is commonly justified as a mechanism assuring

democratic accountability (Perry and Rainey 1988). However, there might be a trade-off

between enhanced accountability and potential decrease of flexibility. That requires

empirical evidence, but one cannot find much research on the topic for government

agencies as well as quangos (Lewis 2007, 2009).

Then, what is the expected effect of politicization on performance? There are

competing views. On one hand, politicization may help increase performance by

prioritizing presidential agenda in policy making and implementation (Wood and

Waterman 1991; Nathan 2000; Maranto 2005; Moe 1999). High responsiveness of political

appointees can clarify organizational goals and priority among them. Then, as the goal

setting theory argues, those clear goals may facilitate directed efforts of organizations,

which in turn lead to enhanced performance of quangos (Locke and Latham 1990). In

addition, political appointees are likely to make a difference by inspiring vitality into

routinized organizations (Aberbach and Rockman 2009; Rourke 1984). With political

support from the president and elected officials, organizations led by political appointees

are less likely to face political conflicts and more likely to elicit cooperation between the

6

public and the private sector (Heclo 1987; Bok 2003).

On the other hand, scholars have warned potential negative effects of politicization.

Political appointees are not only less concerned about policies with narrow, short-term

views, but also impair the continuity and stability of the organizations (Pfiffner 1999;

Cohen 1998; Light 1987; Mackenzie 2001; Lewis 2007, 2008). They may suffer from lack of

experience and expertise, which hampers performance (Boylan 2004; Lewis 2007). In

addition, by not understanding bureaucratic culture, they may conflict with career

bureaucrats. Politicization of quangos may be against the original purpose of quangos,

which is to provide public service effectively regardless of regime changes. Since the

politicization of the board of quangos in Korea has been the target of criticism for a long

time (e.g. Lee 2001), we expect a negative relationship between the degree of politicization

and performance of quangos.

Hypothesis 1: The degree of politicization will negatively affect organizational

performance of quangos.

Chief Executive Succession

As top decision makers, the influence of CEOs is substantial in quangos like any other

organization. Since chief executives interpret vague and sometimes conflicting goals and

translate them into policy (Lewis 2009), their influence on performance is to be carefully

assessed. Although the available literature, especially in the business sector, offers rich

insights into the effect on performance in terms of turnover of top management team, the

issue had barely been touched upon in the public sector.

Chief executives are appointed based on various features, including expertise,

7

previous experience, and political affiliation with the current government. Although the

tenure of CEOs is fixed in law, they are vulnerable to political changes and performance

evaluation (McCabe et al. 2008). Out of various characteristics of chief executives, this

research focuses on the chief executive succession and their previous career based on the

insider/outsider origin.

There is much literature examining the links between senior management turnover

and organizational performance, especially in the private sector. However, there are

strikingly only a few studies that deal with succession or turnover of CEOs in the context of

public organizations (Miller 1993; Wilson 1994; Boyne and Dahya 2002; Hill 2005; Boyne

et al. 2009). One can divide the research of chief executive succession into two streams. On

one hand, there was an attempt to identify causes of turnover, viewing it as a reaction to

the organizational problems (Thompson 1967; Salancik et al. 1980), especially in the

private sector (Dalton and Kesner 1983, 1985; Futado and Karan 1990; Cannella 1993;

Boyne et al. 2010). On the other hand, studies investigate chief executive succession as an

explanatory variable (Pfeffer and Davis-Blake 1986; Hambrick and Mason 1984; Hill 2005).

Some of previous studies reported that the leadership influence on organizational

outcomes is somewhat limited (Lieberson and O’Connor 1972; Weiner and Mahoney 1981;

May and Winter 2009; Futado and Karan 1990).

Then, what are the expected effects of chief executive succession on performance of

quangos? Like the politicization of board, one can think of competing arguments for the

relationship. A new leadership can be seen as both a crisis and opportunity (Ingraham and

Van Slyke 2006). The change of CEOs may bring fresh air and energy to quangos suffering

from moral harzards, but also poses threat to the stability of the organizations. Thus,

8

managerial succession or turnover may be harmful to stable and consistent organizations,

but it can be also beneficial to organizations under strategic, fast-changing environments

because a new CEO is more likely to take new initiatives rather than work with old

priorities (Miller 1993). The tenure of CEOs of quangos in Korea is stipulated by the law as

three years, but the contract can be renewed every year based on performance evaluations.

The fixed term of CEOs in the quasi governmental sectors will allow some level of

discretion and autonomy from the political intervention. The importance of managerial and

personnel stability has been emphasized by several management scholars (e.g. O’Toole and

Meier 2003; Terry 1995). Despite the potential inertia, managerial continuity of CEOs

enables quangos to set long-term strategic planning and to follow the plan without losing

managerial efficiency and consistency. Accumulated knowledge will also help improve

performance. Thus, although chief executive succession may bring positive changes by “an

injection of fresh managerial blood” (Boyne and Dahya 2002: 185), we expect that benefits

coming from stability and continuity in organizational management overwhelm the

negativities, such as stagnation, conservatism, and protectionalism that lead to moral

hazard.

Hypothesis 2: The succession of CEO will be negatively associated with performance of

Quangos.

Insider/Outsider Origin of Chief Executives

We also consider the influence of the origins of chief executives, thereby including the

insider/outsider dimension. In a traditional view, outsiders are more likely to pursue and

initiate strategic changes than insiders because outsiders are less bounded to the

9

established administrative structures and values (Futado and Karan 1990; Canella et al.

1993). However, insiders take more flexible and incrementalistic approach to

organizational changes, which might be appropriate to the organizations under the

unpredictable and uncertain situations (Goodstein and Boeker 1991; Dalton and Kesner

1983).

The insider/outsider succession, conceptually linked to the publicness (Boyne et al.

2009; O’Toole and Meier 2006), are commonly discussed with organizational performance

in the private sector (Dalton and Kesner 1983, 1985; Canella et al. 1993). In terms of

expertise, it is important to include previous occupations and experiences of chief

executives (Donahue 2003). It has been generally accepted that outsiders are more able

than insiders, while insiders are well aware of the technical knowledge regarding specific

work and internal operations. In public settings, however, such arguments are liable to be

criticized since types of expertise vary considerably, such as technical, managerial, and

political. Even within the government, in-and-outers who have experience in both the

private and the public sector are increasing demand for the expertise (MacKenzie 1987;

Donahue 2003; Bok 2003).

In Korean quangos, under the authoritarian legacy and top-down decision making

systems, many outsiders with irrelevant backgrounds have been appointed as chief

executives with the political string. It has been regarded as the underlying cause of moral

hazards and corruptions, although quangos have done a critical role in economic and social

development. In a negative sense, nakhasan, a widely used term which means parachute in

Korean, is a type of patronage system designating to the post without considerations of

appropriateness, just like parachuting to the ground. Accordingly, many of previous

10

literature on the political appointment to the quango executives focus on the negative

impacts and reform measures to fix it. Thus, we expect chief executives from inside

organizations are more likely to produce positive performance outcome.

Hypothesis 3: CEOs from within quangos will be positively associated with

performance of quangos.

Organizational Characteristics

Organizational Size and Age

Organizational size and age is an organizational resource in itself but also facilitates

obtaining and sustaining other resources. According to the contingency theory and

population ecology literature, larger and older organizations are at an advantage over

relations with other organizations and organizational environment (Thompson 1967;

Hannan and Freeman 1989).

Several scholars have examined the size of organizations on performance. Although

size is likely to be treated as an independent variable that shapes organizational structure,

it is sometimes regarded as performance itself as well (Scott 1998). Organizational size is

measured in various indicators, such as the number of full-time employees, the number of

customers, total revenue, and physical capacity and output. But, most studies have used the

number of employees since it reflects “both the capacity of the organization for performing

work as well as the current scale of actual performance” (Scott 1998: 260). Total revenue of

an organization is regarded as direct and visible size of financial resources that is available

to the organizations (Seidel 1991).

Generally, larger organizations have more opportunities to develop and advance

11

experienced staff to the top management team. For example, Boyne (2003) found that the

larger the size, all else being equal, the higher the quality of services. Specifically, the

demographic composition of the top management team is heavily relied upon the

organizational size (Dalton and Kesner 1983; Datta and Guthrie 1994; Furtado and Karan

1990). Dalton and Kesner (1983) investigated the relationship between the

insider/outsider selection and the organizational size of private firms. According to them,

larger firms may have relatively higher rates of internal succession. Moreover, larger

organizations are less vulnerable to political influence (Kaufman 1976; Seidman 1988) by

appointing outsiders with political affiliations. In terms of the political benefits, however,

the patronage motive and benefit of politicization are getting stronger at the same time.

In addition, we have included organizational age as an independent variable that

affects performance. According to Warwick (1975), older organizations are more able to

buffer internal and external pressures and their environments are more favorable, which

make them much easier to produce good performance. Older quangos may have more

experienced staff who have promoted from inside organizations.

Hypothesis 4: Larger and older quangos will show better performance than others.

Government funding proportion

Government funding plays a critical role in providing public service by quasi government

organizations. Since public goods and service apply economies of scale, it is more efficient

and desirable to be provided by one reliable supplier, namely quangos. Quangos hived off

from government are not only short of resources and capacity to provide stable services at

a low price, but also in need of guidance and direction from sponsor department and

12

elected officials.

According to the resource dependency theory, the higher proportion of government

funding means higher dependence on government (Pfeffer and Salancik 1978). Higher

dependency elevates the probability to the managerial intervention from the sponsor

department. That may negatively affect quangos’ autonomy if government imposes its own

political interests over quangos’ management. However, government funding also works as

a stable source for sustainability. Thus, in spite of some potential negative effect, we assume

that the higher proportion of government funding will increase performance of quangos.

Hypothesis 5: Quangos with higher proportion of government funding will show

better performance than others.

Age of Unions

The union of quangos is one of the most important channels of voicing internal demands of

employees. The relationship between existence or militancy of unions and organizational

performance is one of the popular themes in the field of management as well as sociology

and labor economics. Unions of quangos as voice makers wield influence on overall

operation, corporate governance, and appointment of executives. Many studies examining

the effect of union activity on company performance report potential performance gains

from establishing more direct and open channels for collective voice (Mitchell and Stone

1992; Cooke 1994).

Yet, unions of quangos in Korea had been criticized as advocates of their own fringe

benefits such as allowances, welfare, and special leaves in exchange of approving the

decisions from sponsor government department. Unions established during the transition

13

to democracy in 1987 took more active roles in reforming quangos since the end of 1997

financial crisis. Despite the limited access to top decisions, the influence of unions is

substantial to the concerned quangos and appointers, as the resistance or opposition of

unions can be a serious political burden to them. The moral hazard and corruption of

quangos, nevertheless, is partly due to the collusion between union of quangos and political

power. Thus, we expect the unions that have existed longer will be negatively related with

performance of quangos.

Hypothesis 6: Quangos with aged union will show worse performance than others.

Contextual Characteristics

Change of Administration and Unemployment Rate

As contextual characteristics, we include political and economic variables, such as

presidential change and annual average unemployment rate. Since quangos are under

political control, political regime changes are expected to influence quangos’ performance.

According to previous studies, there is a distinctive presidential preference for selecting

executives of quangos (Lewis 2008; Kim and Park 2010). Moreover, performance of

quangos relies on overall economic situations since quangos are in between sectors. We use

annual average unemployment rate in order to capture the vitality of national economy

because it is the most frequently used indicator of economic condition.

METHODS

Data and Measurements

Dependent Variables

14

Table 1 explains measurements of each variable included for the analysis. In this study, two

performance measures are used for dependent variables, including performance evaluation

index (PEI) and consumer satisfaction index (CSI). In Korea, the performance of 13 major

quangos has been evaluated since 1999 as the revision of the law of government

corporations after the financial crisis causing bailout from International Monetary Fund. To

ensure accountability and performance, the performance evaluation team comprised of

outside experts such as professors, certified public accountants, and independent

researchers, was required to assess organizational performance with 100 point ratings

every year. The consumer satisfaction index (CSI) measuring the level of public service

satisfaction is produced by face-to-face interviews with the customers by several private

survey institutions. We use the PEI and CSI from 1999 until 2007 to be consistent with the

previous results as it is assessed by 6 grades to alleviate excessive competition among the

quangos since 2008.

[Table 1 here]

There are potential problems to use the indices as performance variables. The

measurement issues of the indices themselves require specific clarification. Despite the

positive effect brought by the performance evaluation of the quangos, there are continuing

debates over the validity, reliability, and impartiality of the evaluation. The fact that the

indices depend heavily on the judgment of the evaluators implies some degrees of biases by

the halo effect (Lee and Choi 2009; Cooper 1981). Especially for the CSI, several issues are

being raised concerning how to define customers in the public sector, how to measure the

level of satisfaction, how valid the measures and methods, as the concept of satisfaction

refers to the cognitive state as well as the emotional response (Lee and Hong 2005). Since

15

the purpose of performance evaluation is to reward and punish, there is always an

incentive to distort or disguise the results. If there is a room for the intentional distortion

or political influence, performance measurement can obscure as much as it reveals (Pollitt

2006; Talbott 2005). Moreover, the result may be deliberatively kept simple for

communication purposes, or show leniency tendency without any distinct improvement

(Pollitt 2006; Saal et al. 1980). Ingraham et al. (2003) noted that the performance

measurement without consideration of the context can be misleading and destructive in

the public sector.

Despite these caveats, these two indices are regarded as the appropriate second-

best measures to capture the organizational response to the changes of board composition,

leadership succession and tenure. Additionally, the performance evaluation system of

Korea has been recognized as a relatively success to provide a foundation for performance

management which motivates internal staffs. It has been introduced and distributed to the

local government corporations, government RandD research, and other countries (Ministry

of Strategic Planning and Finance 2007).

Independent Variables: Three Characteristics of Top Management

The degree of politicization of board is challenging to define and measure because the

intention of the appointers and political affiliation of every board member of quangos

cannot be easily identified (Skelcher 1998; Donahue 2003). As a result, the primary

occupation, the largest part of the prior career is the most commonly used as a proxy

variable to the ideological orientation and the extent of the political influence on the

appointees. The degree of politicization of the board of directors, in this paper, is measured

16

by the ratio of the number of politicians, military personnel, and private consultants or

experts that have political or partisan affiliations, to total number of board members of an

organization.

Strictly speaking, the board members are all political appointees in a sense that

they are appointed under the political consideration of the president and the minister of

government departments. Although the political mechanism for choosing board-level

decision-makers can be varied, politicians, military personnel, and consultants from private

sectors are likely to have the political supports from presidents especially in Korea.

Although simplification through operational definition was inevitable, there might be some

concerns about using the prior jobs as a proxy variable for politicization that contains

intentions of the appointer and political affiliations of the appointees. Undoubtedly, there

are some cases requiring a degree of judgment in classifying several categories; however, it

is rather clear to judge someone who was promoted from the organizations and

government departments.

The managerial change is coded as a dummy variable, where the years of chief

executive succession events are coded as 1. The leadership quality of chief executives is

assessed by insider or outsider origin within the quangos. Chief executives promoted from

the inside individual organizations are coded as 1. Major careers of the chief executives are

obtained mainly from Who’s Who in Korea (on-line version) and public organization

information system (www.alio.go.kr), which contains variety of information on corporate

governance as well as business management.

[Table 1 here]

17

Control Variables: Organizational and Contextual Characteristics

As organizational characteristics, we consider organizational size and age, government

funding proportion, and age of unions. Organizational size of quangos is measured in two

ways: the number of full-time employees and the size of total revenue. Organizational age is

defined as [(the current year minus the year of establishment) plus one]. If a quango was

established in 1983, the age of the organization is 27 [(2009 minus 1983) plus 1].

Government funding proportion is measured by the ratio of government funding to total

revenue of a quango. The age of unions is measured in the same way as calculating

organizational age.

As contextual characteristics, we include change of administration and

unemployment rate; change of administration (0, 1) is coded 0 for President Kim Dae-jung

(1999~2002), 1 for President Roh (2003~2007). The annual average unemployment rate is

provided by the Korean Statistical Information Service (KOSIS).

Analytic Method

The scope of the analysis covers 13 quangos from 1999 to 2007 that are classified as

government corporations.1 Recent data was excluded from the analysis because of the

scheme of performance evaluation was recently changed. As we use panel data, we are able

to distinguish group effects from individual effects, reduce the problem of collinearity

1 The 13 quangos include Korea National Housing Corporation, Korea Coal Corporation, Korea

National Oil Corporation, Korea Expressway Corporation, Korea Water Resource Corporation,

Korea Electric Power Corporation (KEPCO), Korea Agro-Fisheries Trade Corporation (AT), Korea

Rural Community Corporation, Korea Trade-Investment Promotion Corporation (KOTRA), Korea

National Tourism Corporation, Korea Minting and Security Printing Corporation. Korea Resource

Corporation, Korea Land Corporation.

18

among explanatory variables, and control the effects of missing or unobserved variables

(Hsiao 2007). Below is our model equation.

it + nXnit +

nX nit + nX nit + μi +ηt +εit

As the data set is an unbalanced panel, Ti represents a different time scope for each

panel group.2 is a scalar, is 3 1 and Xit is the it-th observation on three explanatory

variables reflecting leadership characteristics. Likewise, γ is 5 1 and X it for organizational

characteristics; δ is 2 1 and X it for contextual factors. μi stands for unobserved

characteristics of i-th quango or cross-sectional specific components, and ηt stands for

unobserved characteristics of t year or time-series specific components, and εit is the

idiosyncratic errors of the model. Panel data can be estimated using several different

models according to the assumption of the characteristics of residuals. In this paper, the

results of four models are presented.

The panel generalized least square (GLS) assuming heteroscedasticity and

autocorrelations among the residuals estimates the ordinary least square by pooling the

time series and cross sectional observations. This is the weighted average estimate of

between effects and fixed effects models. The between regression model on group means

assumes that the residuals are heterogeneous among panel groups (μi ) but not correlated

in the same group. The fixed effects model assumes μi and ηt as the fixed parameters to be

estimated in the model. If the μi are thought of as fixed parameters to be estimated, most

variations of dependent variables can be explained by the variations of residual μi

2 For example, Korea Rural Community Corporation was established in 2000.

19

representing the characteristics of individual panel groups. The within group estimators of

least square dummy variable will be presented in the third column. The random effects

model assuming the residuals as random variables with a normal distribution will estimate

generalized least square. It is more appropriate to the estimation with randomly drawn

individual groups from a large population (Hsiao 2007). We conducted the Hausman test to

determine whether some elements of Xit are correlated with the cross section specific

components (μi) and which models are more appropriate to the estimation of our models.

RESULTS AND DISCUSSIONS

Descriptive statistics and correlations among the study variables are presented in Table 2

and Table 3. From Table 2, the two performance measures have similar mean score, but

customer satisfaction has larger standard deviation.

[Table 2 and 3 here]

Results for the panel data analysis of organizational performance are presented in

Table 4 and Table 5. It displays all the result of the pooled GLS models, Between effects,

Fixed effects (within), and Random effects models. A quick glance at signs and significance

suggests that our hypotheses are partly supported and requires specific explanations. The

panel data models for each dependent variables, performance evaluation index (PEI) and

consumer satisfaction index (CSI), show estimation results for each of the pooled GLS, the

between regression, fixed effects, and random effects models. Based on the F-test and

Hausman test, the random effects model seems more appropriate to estimate for PEI.

[Table 4 here]

According to hypothesis 1, we expected the coefficient estimate for PEI relative to

20

the degree of politicization to be negative. The results of model 1, however, do not support

this expectation as the politicization of the boards is found to increase performance

evaluation score. Although filling the boards with the political allies has been regarded as

the cause of corruption, it seems that the politicization of the board members brings

positive effects on performance evaluation by engaging diversity and freshness to the

organizations trapped in moral hazard. The chief executive succession also increase the PEI

score by adopting new managerial strategies. The leadership quality of chief executives

with inside origins showed expected positive sign in Hypothesis 3, but failed to reach the

significance level. Although not entirely persuasive, these results are at least suggestive

that top leadership of quangos might have lasting effects on performance evaluation.

Among the organizational factors, the ratio of government funding and the size of

total revenue were statistically significant and positively related to the PEI. Estimates

suggest that one percent increase in total revenue leads to an approximately one point

increase in the performance evaluation index. The positive significant sign of government

funding implies that the more financially supported from the government, the higher the

score. It supports the assumption underlying many studies of accountability and

cooperation between sponsor department and quangos as extended arms of government.

Since performance evaluation includes overall achievement of the planned objectives, the

relationship with government is essential to improve effectiveness of primary activities of

function, efficiency of management, and outcome of the activities.

The strength of unions measured by the age of union also significantly affect

performance evaluation score and the relationship was in the expected direction. As

quangos have been regarded as agents without a principal, they are attractive tools to

21

government officials who are interested in the top positions of quangos as well as to the

staff members of quangos who seek more fringe benefits from the organization. For that

reason, unions of quangos are likely to constitute a vicious circle of moral hazard and

inefficiencies. In line with the expectations, the coefficient estimates of the variable are

negative.

The signs for the contextual variables concerning partisan interests and economic

conditions generally fit what might be expected and they came close to significant, but they

failed to reach the threshold. Despite the potential vulnerability, the change of

administration and unemployment rate did not significantly affect performance evaluation

score.

[Table 5 here]

In Table 5, we displayed analysis result for Model 2 with the dependent variable of

consumer satisfaction index (CSI). Since the Wu-Hausman test of the contrast between the

within and random effects estimates cannot reject the null hypothesis that (Xit, μi) is

uncorrelated, we prefer the within estimates on the efficient grounds. Although the sign of

the explaining variables concerning top leadership characteristics showed the expected

direction, they all failed to reach the significance level. The overall fit is less satisfying. On

the contrary, the organizational and contextual variables turned out to be strong

determinants of consumer satisfaction scores.

The control variables for organizational characteristics revealed to influence CSI

statistically significantly are organizational age and size of total revenue, government

funding proportion. The positive coefficient estimates suggest that our theoretical

expectations are supported: when other things being equal, larger and older quangos with

22

higher proportion of government funding will be likely to get higher scores in consumer

satisfaction. Since consumer satisfaction to the public service is partly based on the trust

level among general public, old quangos with accumulated expertise and experience might

have better relationships with public. The higher proportion of government funding to total

revenue leads to positive results in the CSI, which is also true for the PEI. Financial

characteristics of quangos have a statistically significant positive effect on both the

performance evaluation and consumer satisfaction index.

Additionally, Table 5 clearly shows the importance of the contextual variables for

performance measured by consumer satisfaction. In the case of the CSI, the importance of

contextual factors seems consistent and reliable. The political change of administration to

the recent President is to be significant and positively related to the higher satisfaction

score. The result can be interpreted as the evidence for the improvement of performance

during the President Roh Administration, however, at least primarily, for the influence of

the political changes on performance of quangos. Meanwhile, the annual average

unemployment rate was negatively related to the index, reflecting the overall decrease of

satisfaction during the economic downturn. Therefore, it seems organizational and

contextual characteristics that matter for CSI rather than top leadership characteristics.

CONCLUSIONS

This article set out to develop a preliminary model of organizational performance of

quangos and examine the effect of top leadership on the two performance indices,

Performance Evaluation Index (PEI) and Consumer Satisfaction Index (CSI). This model

was applied to 13 quangos in Korea.

23

Analysis results of Model 1 show that leadership characteristics concerning

politicization of the boards and chief executive succession have a significantly positive

effect on the PEI. Organizational characteristics such as the size of total revenue,

government funding proportion, and the age of union have a distinctive effect as well, while

contextual factors such as political change and unemployment rate were not significant. In

Model 2, on the contrary, consumer satisfaction index was more influenced by contextual

factors such as the presidential change and unemployment rate, while leadership

characteristics were not statistically significant. As many contextualists argue, the reason

might be because leadership affects performance in complicated and indirect ways

(Goldston 2001; Van Slyke and Alexander 2006), though more research is desired.

This study has some implications for public administration and research and

theory. Most importantly, it attempted to contribute to leadership research in quasi

government sectors with panel studies that control organizational and contextual factors.

Since leadership literature has shown conflicting results (Van Slyke and Alexander 2006),

an attempt to assess the relative influence to performance indices has implications to

leadership studies as well as performance management. Second, this study presented

evidence to explain why politicization is still very powerful around the quasi government

sector, not just in the executive branch. The positive effect on performance of quangos in

Korean context provides comparative bases on these fields. The result of this research

bears out the difficulties of studying performance in public organizations, which is mostly

confusing and conflicting to each other across the indices. Despite the importance and the

duration of this topic, defining and measuring performance has always been one of the big

challenges to scholars in the field of public management.

24

REFERENCES

Aberbach, D. J. and B. A. Rockman. (2009). “The Appointments Process and the

Administrative Presidency”, Presidential Studies Quarterly 39(1): 38~59.

Baysinger, Barry D. Henry N. Butler. (1985). “Corporate Governance and the Board of

Directors: Performance Effects of Changes in Board Composition”, Journal of Law,

Economics, & Organization 1(1): 101~124.

Bertelli, A. M. (2005). “Governing the Quango: An Auditing and Cheating Model of Quasi-

Governmental Authorities,” JPART 16: 239~251.

Bertelli, A. M. (2006). “Delegating to the Quango: Ex Ante and Ex Post Ministerial

Constraints”, Governance 19(2): 229~249.

Bok, Derek. (2003). “Government Personnel Policy in Comparative Perspective,” in J. D.

Donahue and J. S. Nye (Ed) For the People: Can We Fix Public Service? Washington:

Brookings: 255~272.

Boylan, R. T. (2004). “Salaries, Turnover, and Performance in the Federal Criminal Justice

System.” Journal of Law and Economics 47(1): 75~92.

Boyne, G. A., O. James, N. Petrovsky. (2009). “What Do New Agency Chief Executives Make a

Difference to Policies and Performance? A Theoretical Model of Insider/Outsider

Distinctiveness in the Public Sector.” Presented Paper in 10th Annual Public

Management Research Conference.

Boyne, G. A. and J. Dahya. (2002). “Executive Succession and the Performance of Public

Organizations.” Public Administration 80(1): 179~200.

Boyne, G. A. (2003). “Sources of Public Service Improvement: A Critical Review and

Research Agenda.” JPART 13(3): 367~394.

Boyne, G. A., K. J. Meier, L. J. O'Toole, R. M. Walker. (2005). ”Where Next? Research

Directions on Performance in Public Organizations.” JPART 15: 633~639.

Boyne, G. A., O. James, P. John, N. Petrovsky. (2010). “Does Public Service Performance

Affect Top Management Turnover?” JPART, 20 (Supplement 2): i261~i279.

Brewer, G. A., and Selden, S. C. (2000). “Why Elephants Gallop: Assessing and Predicting

Organizational Performance in Federal Agencies.” JPART 10(4): 685~712.

Cannella Jr., A. Albert., and M. Lubatkin. (1993). “Succession as a Sociopolitical Process:

Internal Impediments to Outsider Selection.” Academy of Management Journal 36(4):

763~793.

Cho, Y. J. and E. J. Ringquist. (2010). “Managerial Trustworthiness and Organizational

Outcomes,” JPART 21(1): 53~86.

Cohen, D. (1998). “Amateur Government”, JPART 8(4): 450~497.

Cole, M. 2000. "Quangos: UK Ministrial Responsibility in Theory and Practice", Public Policy

and Administration 15: 32~45.

Cooper, W. H. (1981). “Ubiquitous Halo”, Psychological Bulletin 90(2): 219~244.

25

Cooke, W. N. (1994). “Employee Participation Program, Group-based Incentives, and

Company Performance: A Union-Nonunion Comparison,” Industrial and Labor

Relations Review 47(4): 594~609.

Dalton, D. R. and I. F. Kesner. (1983). “Inside/Outside Succession and Organizational Size:

The Pragmatics of Executive Replacement,” Academy of Management Journal 26(4):

736~742.

Dalton, D. R. and I. F. Kesner. (1985). “Organizational Performance as an Antecedent of

Inside/Outside Chief Executive Succession: An Empirical Assessment,” Academy of

Management Journal 28(4): 749~762.

Datta, D. K., and James P. Guthrie. (1994). “Executive Succession: Organizational

Antecedents of CEO Characteristics,” Strategic Management Journal 15(7): 569~577.

Denis, J-L., A. Langley, and L. Rouleau (2005). “Rethinking Leadership in Public

Organizations.” in L. Ferlie, and Pollitt, Oxford Handbook of Public Management.

Oxford University Press.

Donahue, D. J. (2003). “In-and-Outers: Up and Down?” in J. D. Donahue and J. S. Nye (Ed.),

For the People: Can We Fix Public Service? Washington, D. C: Brookings Institution

Press.

Durant, F. R. (1995). “Public Policy, Overhead Democracy, and the Professional State

Revisited,” Administration and Society 27(2): 165~202.

Forbes, M. and L. E. Lynn. Jr. (2005). “How does Public Management Affect Government

Performance? Findings from International Research.” JPART 15: 559~584.

Flinders, M. V. and S. J. Martin. (1999). Quangos, Accountability and Reform: The Politics of

Quasi-Government. Political Economy Research Centre, University of Shefield.

Futado, E. P. H., and Vijay Karan (1990). “Causes, Consequences, and Shareholder Wealth

Effects of Management Turnover: A Review of the Empirical Evidence,” Financial

Management 19.

Goodstein, J. and W. Boeker (1991). "Turbulence at the Top: A New Perspective on

Governance Structure Changes and Strategic Change," Academy of Management

Journal 34(2): 306~330.

Hambrick, D. C. and P. A. Mason. (1984). “Upper Echelons: The Organization as a Reflection

of its Top Managers.” Academy of Management Journal 9(2): 193~206.

Hannan, M. T. and J. Freeman. (1987). Organizational Ecology. Harvard University Press.

Heclo, Hugh. (1987). “The In-and-Outer System: A Critical Assessment,” in Mackenzie, G.

Calvin. The In-and-Outers: Presidential Appointees and Transient Government in

Washington, The Johns Hopkins University Press. 196~216.

Hicklin, A., L. J. O'Toole, Jr., and K. J. Meier. (2008). “Serpents in the Sand: Managerial

Networking and Nonlinear Influences on Organizational Performance.” JPART 18(2):

253~273.

Hill, G. C. (2005). “The Effects of Managerial Succession on Organizational Performance.”

JPART 15: 585~597.

26

Hsiao, C. (2007). Analysis of Panel Data. Cambridge University Press.

Ingraham, P. W., P. G. Joyce, A. K. Donahue. (2003). Government Performance: Why

Management Matters, The Johns Hopkins University Press.

Ingraham, P. W. and D. M. Van Slyke. (2006). “The Path Ahead for Public Service Leadership”

American Review of Public Administration 36(4): 392~394.

Kaufman, H. (1976). Are Government Organizations Immortal? Washington D.C: Brookings

Institution.

Kim, B. S. and S. Park. (2010). “The Political Appointment of Executive of Public

Organizations–The Influence of Legal Institutions and Presidencies” Korea Public

Administration Review 44(2): 85~109. (in Korean)

Koppell, G. S. J. (2003). The Politics of Quasi-Government: Hybrid Organizations and the

Control of Public Policy. Cambridge and New York: Cambridge University Press.

Koppell, G. S. J. (2007). “Political Control for China’s State-Owned Enterprises: Lessons from

America’s Experience with Hybrid Organizations”, Governance 20(2): 255~278.

Lee, C. G., and S. R. Choi. (2009). “Halo Effect in Performance Evaluation on Public

Enterprises,” Korean Public Administration Review 43(3): 151~172. (in Korean)

Lee, K. H., and W. K. Hong. (2005). “The Measurement and Use of Customer Satisfaction of

Public Administration: Focused on the Discontinuous Approach of Satisfaction,”

Korean Public Administration Review 39(4): 105~124. (in Korean)

Lee, M. S. (2001). “Political Appointment and Management Performance of Executive

Officers of Government-Invested Enterprises”, Korean Public Administration Review

35(4) (in Korean)

Lewis, D. E. (2003). Presidents and the Politics of Agency Design. Stanford University Press:

Stanford, California.

Lewis, D. E. (2007). “Testing Pendleton's Promise: Do Political Appointees Make Worse

Bureaucrats?” Journal of Politics, 69(4), 1073~88.

Lewis, D. E. (2008). The Politics of Presidential Appointments: Political Control and

Bureaucratic Performance, Princeton University Press.

Lewis, D. E. (2009). “Revisiting the Administrative Presidency: Policy, Patronage, and

Agency Competence. Presidential Studies Quarterly 39(1): 60~73.

Lægreid, P. and P. G. Roness. (2003). “Administrative Reform Programs and Institutional

Response in Norwegian Central Government, in J.J. Hesse, C. Hood and B.G. Peters

(Eds.) Paradox of Civil Service Reform. Berlin: Duncker & Humblot.

Lieberson, S. and J. F. O’Connor. (1972). “Leadership and organizational performance: A

study of large corporations.” American Sociological Review 37(1): 117~30.

Light, C. Paul. (1987). “When Worlds Collide: the Political and Career Nexus,” in Mackenzie.

1987. The In-and-Outers: Presidential Appointees and Transient Government in

awWashington. Baltimore: Johns Hopkins University Press. 156~173.

Locke, E. A., and Latham, G. P. (1990). A Theory of Goal Setting and Task Performance. Upper

Saddle River, NJ: Prentice Hall.

27

Lynn, L. E., C. J. Heinrich, and C. J. Hill. (2000). “Studying Governance and Public

Management: Challenges and Prospects,” Journal of Public Administration Research

and Theory 10(2): 233-262.

Mackenzie, G. C. (1987) The In-and-Outers: Presidential Appointees and Transient

Government in Washington, The Johns Hopkins University Press.

Mackenzie, G. C. (2001). “The State of the Presidential Appointment Process,” in G. C.

Mackenzie (Ed.), Innocent Until Nominated. Washington: Brookings

Maranto, Robert. 2005. Beyond a Government of Strangers: How Career Executives and

Political Appointees Can Turn Conflict to Cooperation, Lexington Books. USA.

May, P. J. and S. C. Winter (2009). “Politicians, Managers, and Street-Level Bureaucrats:

Influences on Policy Implementation,” JPART 19(3): 453-476.

McCabe, B. C., R. C. Feiock, J. C. Clingermayer, C. Stream. (2008). “Turnover among City

Managers: The Role of Political and Economic Change,” PAR 68(2): 380~386.

Meier, K. J., O'Toole, L. J., Jr., Boyne, G. A., and Walker, R. M. (2007). “Strategic Management

and the Performance of Public Organizations: Testing Venerable Ideas against Recent

Theories,” (Author abstract) JPART 17(3): 357(21).

Miller, D. (1993). “Some Organizational Consequences of CEO succession,” Academy of

Management Journal 36(3): 644~659.

Mitchell, M. W. and Joe A. Stone. (1992). “Union Effects on Productivity: Evidence From

Western U.S. Sawmills,” Industrial and Labor Relations Review 46(1): 135~145.

Moe, T. M. (1999). “The Politicized Presidency,” in P. J. Pfiffner (Ed.), The Managerial

Presidency, Texas A&M University Press.

Moynihan, P. D. and S. K. Pandey. (2004). “Testing How Management Matters in an Era of

Government by Performance Management.” JPART 15: 421~439.

Nathan, P. R. (2000). So You Want to Be in Government? Albany: Rockefeller Institute Press.

Newland, C. A. (1996). “Transformational Challenges in Central and Eastern Europe and

Schools of Public Administration” Public Administration Review 56(4): 382~389.

Nicholson, J. G. and G. C. Kiel. (2007). "Can Directors Impact Performance? A Case-based

Test of Three Theories of Corporate Governance." Corporate Governance 15(4):

585~608.

Nolan, B. C. (2001). Public Sector Reform: An International Perspective, Basingstoke:

Palgrave.

O'Toole Jr., L. J., and Meier, K. J. (2003). “Plus ça Change: Public Management, Personnel

Stability, and Organizational Performance.” JPART 13(1): 43~64.

Perry, J. L. and H. G. Rainey. (1988). ‘The Public-Private Distinction in Organization Theory:

A Critique and Reseach Strategy”, The Academy of Management Review 13(2):

182~201.

Pfeffer, J. and G. Salancik. (1978). The External Control of Organizations: A Resource

Dependence Perspective, New York: Harper & Row.

28

Pfeffer, J. and Alison Davis-Blake. (1986). “Administrative Succession and Organizational

Performance: How Administrator Experience Mediates the Succession Effect,”

Academy of Management Journal 29(1): 72~83.

Pfiffner , P. J. (1999). “Can the President Manage the Government?” in P. J. Pfiffner (Ed.), The

Managerial Presidency (2th ed.): Texas A&M University Press.

Pollitt, C. (2006). “Performance Management in Practice: A Comparative Study of Executive

Agencies.” JPART 16(1): 25~44.

Rourke, E. F. (1984). Bureaucracy, Politics, Public Policy. Boston: Little Brown.

Saal, Frank E., Downey, Ronald G., Lahey Mary A. 1980. “Rating the ratings: Assessing the

Psychometric Quality of Rating Data,” Psychological Bulletin 88(2): 413~428.

Salancik, G., Berry. M. Staw, Louis R. Pondy. (1980). “Administrative Turnover as a

Response to Unmanaged Organizational Interdependence,” Academy of Management

Journal 23(3): 422~437.

Scott, R. W. (1998). Organizations: Rational, Natural, and Open Systems, 4th Ed. Prentice Hall,

Upper Saddle River, New Jersey.

Seidel, R. J. (1991). “Resource Interdependence: The Relationship between State Agencies

and Nonprofit Organizations,” PAR 51(6): 543~553.

Seidman, H. (1988). “The Quasi World of the Federal Government,” The Brookings Review 6:

23~27.

Skelcher, C. (1998). The Appointed State: Quasi-governmental Organizations and Democracy,

Buckingham: Open University Press.

Talbott, C. (2005). “Performance Management,” in L. Ferlie, and Pollitt, Oxford Handbook of

Public Management. Oxford University Press.

Terry, Larry D. (1995). Leadership of public bureaucracies: The administrator as conservator.

Thousand Oaks, CA: Sage Publication.

Thompson, D. J. (1967). Organizations in Action. NY: McGrew-Hill.

Van Slyke, D. M. and R. W. Alexander. (2006). “Public Service Leadership: Opportunities for

Clarity and Coherence.” American Review of Public Administration 36(4): 362~374.

Warwick, D. P. (1975). A Theory of Public Bureaucracy. Cambridge, MA: Harvard University

Press.

Weiner, N. and T. A. Mahoney. (1981). “A Model of Corporate Performance as a Function of

Environmental, Organizational, and Leadership Influences,” Academy of Management

Journal 24(3): 453~470.

Wilson, P. A. (1994). “Power, Politics, and Other Reasons Why Senior Executives Leave the

Federal Government,” PAR 54(1): 12~19.

Wood, D. B. and R. W. Waterman. (1991). “The Dynamics of Political Control of the

Bureaucracy,” The American Political Science Association 85(3): 801~828.

Yang, K., and Hsieh, J. Y. (2007). “Managerial Effectiveness of Government Performance

Measurement: Testing a Middle-Range Model,” PAR 67(5): 861~879.

29

Figure 1 Research Framework

Organizational Performance Performance Evaluation Index (PEI) Consumer Satisfaction index (CSI)

Organizational Characteristics Organizational Size (Number of employees, Total Revenue) Organizational Age Government Funding Proportion Age of Union

Contextual Factors Change of Administration Unemployment rate

Leadership Characteristics Degree of Politicization Chief Executive Succession Insider/Outsider Origin

30

Table 1 Data and Measures

Variables Descriptions Measures Data Sources

Dependent Variables

Organizational performance

Performance Evaluation Index (PEI) Performance Evaluation Report issued by the Ministry of Planning and Finance,

Public Organization Information System (www.alio.go.kr),

Korean Statistical Information Service (KOSIS) (http://kosis.kr),

Who’s Who in Korea,

Each agency’s websites and white papers.

Customer Satisfaction Index (CSI)

Leadership Variables

Politicization of the board of directors

(the number of former politicians, military personnel, private consultants/ total number of board members)×100

Chief Executive Succession

(0,1) coded 1 if there is management succession or turnover of CEOs

Previous career of CEOs

(0, 1) coded 1 if the chief executives have insider origin.

Organizational Variables

Organizational size The number of full-time employees

The size of total revenue (million)

Organizational age (The current year - year of establishment) + 1

Government Funding Ratio

(government funding/ total revenue)×100

Age of Union (The current year - year of establishment) +1 (0 coded if there is no union)

Contextual Variables

Presidential change

(0, 1) coded 1 for President Roh

Unemployment rate

The yearly average rate of national unemployment

31

Table 2 Descriptive Statistics

Variables Obs Mean Std. Dev. Min Max

Performance Evaluation 116 75.792 5.2575 60.24 87.61

Customer Satisfaction 115 74.710 11.345 44 96.3

Degree of politicization 116 53.681 41.795 0 100

Succession of CEOs 116 .3275 .4713 0 1

Insider/Outsider 116 .8965 .3058 0 1

Number of Fulltime Employees 113 3764.7 6149.8 296 34608

Organizational Age 116 37.241 12.164 1 58

Government Funding Proportion 110 21.496 25.577 0 104.92

ln(Total Revenue) 105 14.085 1.7172 11.619 17.376

Age of Union 116 15.646 12.590 0 53

Presidential Change 116 .5603 .4984 0 1

Unemployment Rate 116 3.756 .9245 3 6.3

Table 3 Correlations Among Variables

Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

(1)Performance Evaluation 1.000

(2) Customer Satisfaction 0.103 1.000

(3)Degree of Politicization 0.304* 0.154 1.000

(4) Succession of CEOs 0.086 0.183* 0.105 1.000

(5) Insider/Outsider 0.003 -0.054 -0.174 -0.426* 1.000

(6) Number of Employees 0.245* -0.020 0.170 -0.190 0.087 1.000

(7) Organizational Age -0.19* 0.208* -0.152 -0.116 -0.04* 0.020 1.000

(8) Government Funding 0.135 -0.062 -0.112 0.035 -0.115 -0.289 0.078 1.000

(9) ln(Total Revenue) 0.337* -0.010 0.104 -0.009 0.010 0.619* -0.277 -0.459* 1.000

(10) Age of Union -0.359* 0.131 -0.069 0.324* -0.228 -0.39* 0.388* 0.279 -0.371* 1.000

(11) Presidential Change 0.142 0.818* 0.199* 0.285* -0.186 -0.031 0.156 -0.008 0.051 0.137 1.000

(12) Unemployment Rate -0.186 -0.696* -0.218* -0.052 -0.025 0.056 -0.086 0.091 -0.046 -0.094 -0.51* 1.000

Note) Coefficients with mark asterisk are statistically significant at Bonferroni-adjusted .05 levels.

32

Table 4 Performance Evaluation Index (PEI)

Pooled

Panel GLS

Between

Regression

Fixed Effect

(within)

Random

Effect

Leadership Characteristics

Politicization of the Boards .0176**

(.0075)

.0568

(.0302)

.0211**

(.0104)

.0246**

(.0101)

Chief Executive Succession 2.8089***

(.7749)

3.211

(1.5210)

1.293

(1.203)

1.857*

(.9803)

Insider/Outsider 1.7425

(1.221)

.2418

(2.584)

.6840

(1.702)

1.180

(1.418)

Organizational Characteristics

Number of Employees -.0000

(.0001)

.0000

(.0001)

.0004

(.0002)

.0000

(.0001)

ln(Total Revenue) .9540***

(.3024)

.7986

(.2910)

1.084*

(2.290)

1.085***

(.3492)

Organizational Age .0458

(.0336)

.0334

(.0293)

.7351

(.6720)

.0180

(.0425)

Government Funding

.0861***

(.0153)

.0869**

(.0158)

.0749**

(.0406)

.0878***

(.0174)

The Age of Union -.2135***

(.0453)

-.1270

(.0615)

-.7215

(.6152)

-.2013***

(.0404)

Contextual Characteristics

Presidential Change -.1076

(.7968)

-126.73

(60.329)

.1802

(1.600)

.2054

(.9876)

Unemployment Rate -.7783*

(.4569)

-72.596

(36.398)

-1.060

(.7265)

-.7471

(.5498)

Constants 61.965***

(5.229)

404.48

(172.64)

73.342**

(30.117)

63.092***

(6.699)

Model Specifications

Number of Obs 104 104 104 104

Number of Groups 13 13 13 13

R-Squared (overall) − .0068 .0824 .4492

Wald Chi2/ F Wald chi2(10)

= 80.19

F(10,2)

= 15.82

F(10,81)

= 1.18

Wald chi2(10)

= 75.85

Prob> Chi2 Prob > chi2

= .0000

Prob >F

= .0609

Prob > F

=.3108

Prob > chi2

=.0000

Note 1) Panel: heteroskedastic, Correlations: common AR(1) coefficient for all panels (.1274)

Note 2) F test that all u_i=0, F(12, 81) = 1.18 Prob > F = 0.310

Note 3) Hausman test: under H0: difference in coefficients in the fixed and random effects model are

not systematic, Prob > Chi2 = 0.0862

33

Table 5 Consumer Satisfaction Index (CSI)

Note 1) Panel: heteroskedastic, Correlations: common AR(1) coefficient for all panels (.0848)

Note 2) F test that all u_i=0, F(12, 81) =10.45 Prob > F = .0000

Note 3) Hausman test: under H0: difference in coefficients in the fixed and random effects model are

not systematic, Prob > Chi2 = .0000

Pooled

Panel GLS

Between

Regression

Fixed Effect

(within)

Random

Effect

Leadership Characteristics

Politicization of the Boards -.0015

(.0108)

-.0129

(.0358)

.0001

(.0094)

-.0029

(.0135)

Chief Executive Succession 1.559

(1.160)

.0263

(1.925)

-.5986

(1.079)

1.293

(1.299)

Insider/Outsider 2.459

(1.813)

9.143*

(2.450)

.0988

(1.553)

1.415

(1.927)

Organizational Variables

Number of Employees .0001

(.0001)

.0002

(.0001)

.0000

(.0002)

.0002**

(.0001)

ln(Total Revenue) -.5908

(.4609)

-1.165*

(.3531)

4.647**

(2.053)

-.8081*

(.4632)

Organizational Age .0203

(.0499)

.0644

(.0344)

2.686***

(.6027)

.0173

(.0564)

Government Funding .0114

(.0218)

.0448

(.0258)

.1222***

(.0364)

.0262

(.0236)

The Age of Union .0185

(.0435)

-.0758

(.0727)

.2871

(.5516)

.0139

(.0540)

Contextual Variables

Presidential Change 13.625***

(1.141)

121.95

(44.493)

3.748**

(1.435)

14.405***

(1.308)

Unemployment Rate -5.579***

(.6316)

59.861

(27.185)

-1.956***

(.6516)

-5.816***

(.7288)

Constants 91.682***

(8.064)

-208.61

(126.48)

-94.094***

(27.020)

95.475***

(8.669)

Model Specifications

Number of Obs 103 103 103 103

Number of Groups 13 13 13 13

R-Squared (overall) − .0349 .0584 .8067

Wald Chi2/ F Wald chi2(10)

= 419.49

F(10,2)

= 16.63

F(10,80)

= 86.26

Wald Chi2(10)

= 384.00

Prob> Chi2 Prob > chi2

=.0000

Prob > F

=.0580

Prob > F

=.0000

Prob > chi2

=.0000