Does E-Government Promote Accountability?

A Comparative Analysis of Website Openness

and Government Accountability
WILSON WONG* and ERIC WELCH**

Under the global pressure of information technology, the adoption of web-

based technologies in public administration has created a new government-
and-citizen interface. However, whether e-government will unambiguously
lead to a more transparent, interactive, open and hence, accountable, gov-
ernment remains a central question. Applying a framework of global pres-
sure effects on bureaucratic change, this paper conducts an empirical study
on website openness and accountability in fourteen countries. Even when
overall accountability levels rise, the accountability gap between different
national bureaucracies often remains intact as web-based technologies typ-
ically maintain or reinforce the existing practices. The question of whether
e-government promotes accountability depends on what kind of bureau-
cracy one is referring to in the first place. In the current debate about global
convergence and national divergence on the effect of globalization on public
bureaucracies, the spread of e-government provides a case of convergence
in practice rather than in results.

INTRODUCTION

E-government, the adoption of web-based technologies to deliver and

conduct government services, has become a global trend in public admin-
istration.1 In 2000, there were 168 national governments that had their
own websites.2 E-government often comes with a promise to improve
public administration in terms of efficiency, one of the primary values in
public administration (Lee and Perry; Heintze and Bretschneider; Rosen-
bloom). However, e-government has the potential to alter the traditional
relationship between government and citizens by creating a new virtual
government-and-citizen interface. With the popularity of e-government
and the increasing interaction between government and citizens through
the internet, a major question comes: To what extent does e-government
promote public accountability?
Although accountability can be an elusive word in public administra-
tion, it often refers to the answerability of government to the public on

*The Chinese University of Hong Kong

**University of Illinois at Chicago

Governance: An International Journal of Policy, Administration, and Institutions, Vol. 17, No. 2,
April 2004 (pp. 275–297). © 2004 Blackwell Publishing, 350 Main St., Malden, MA 02148,
USA, and 9600 Garsington Road, Oxford, OX4 2DQ, UK. ISSN 0952-1895
276 WILSON WONG AND ERIC WELCH

its performance (Cunningham and Harris; Johnston and Romzek;

Romzek and Dubnick). E-government is often viewed and promoted as
a positive channel for enhancing government accountability and empow-
ering citizens (La Porte, de Jong, and Demchak; Demchak, Friis, and La
Porte 2000). More information delivered in a more timely fashion to citi-
zens is expected to increase transparency of government, empowering cit-
izens to more closely monitor government performance. Enhanced
interactivity of the technology is also expected to improve government
accountability as it makes government more responsive to the needs and
demands of individual citizens.
However, there are arguments against this positive vision of e-
government accountability. Information technology in public organiza-
tions often simply improves their technical efficiency without leading to
significant organizational changes (Heintze and Bretschneider). Instead
of changing the nature of organizations, the role played by information
technology is often no more than reinforcing “existing tendencies” of
organizations (Kraemer and Dedrick).
The empirical research has indicated that computing per se is neither a cen-
tralizing or decentralizing influence. The context in which computing is used
is a much stronger influence on whether organizations centralize or decentral-
ize than is the technology, which can support either type of arrangement. In
general, computing tends to reinforce existing tendencies, and by itself is not
likely to affect organizational structure in significant ways. (Kraemer and
Dedrick, 101)

Applying this logic to the impact of e-government, the new technology

may simply exacerbate the domestic trajectory, enhancing or reducing
accountability levels based on domestic traditions, institutions, and cir-
cumstances. In fact, prior research seems to show that governments with
an authoritarian or paternalistic nature use web-based technologies to
control access to information for the purpose of monitoring citizen behav-
ior to tighten political control of the regime (Welch and Wong 1998). From
this perspective, the relationship between e-government and public
accountability is a conditional one, where change in accountability levels
depends on the context and characteristics of the public organization.
These two very different perspectives on the effect of e-government on
accountability are linked to a fundamental question on the effect of global
pressures on bureaucratic change. In the study of globalization, there is a
heated debate on whether the global pressures, including the global pres-
sure of information technology, will bring to a convergence or a diver-
gence for the organization systems of different nations (Osborne; Pollitt;
Doremus et al.; Hallerberg and Basinger; Kettl). If there is a direct and
positive relationship between e-government and accountability, as sug-
gested by the first paradigm, the convergence theory of globalization will
be supported. With the establishment of more websites by more govern-
ments around the globe, these governments will converge to a global stan-
dard of accountability. In contrast, if the second paradigm is correct that
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 277

e-government accountability is contingent, the spread of e-government

will instead lead to a divergence of accountability among countries and
agencies of different contexts and characteristics.
To address the question of e-government’s effect on accountability and
to shed more light on the impact of globalization, this article conducts an
empirical study of the effect of e-government on accountability in four-
teen countries. It compares the openness of the websites of e-governments
of these countries and tests the impact of the contextual and organiza-
tional factors on their openness. This study attempts to answer the fol-
lowing questions. To what extent does e-government affect accountability
of public bureaucracies? Will e-government lead to a global convergence
in accountability or an accountability divergence among nations and
bureaucracies? If there is any accountability divergence, what national
and organizational characteristics may explain the differences? The paper
is divided into the following sections: a theoretical framework for analy-
sis of the effects of e-government on accountability; hypotheses predict-
ing the effect of contextual and organizational factors on accountability;
countries, data, and method; findings; and discussion.

A THEORETICAL FRAMEWORK OF GLOBAL PRESSURE AND

E-GOVERNMENT ACCOUNTABILITY

This study takes advantage of a theoretical framework of global pressure

and bureaucratic change to analyze the effect of e-government on changes
in accountability (Welch and Wong 1998, 2001b). Globalization creates a
new set of complex and interactive stimuli, demands, and opportunities
in the external environment of national public bureaucracies, whose
origin is not traceable to any particular nation. These stimuli, demands,
and opportunities can be called “global pressures” as they are forces of a
global scope that are putting public bureaucracies worldwide under pres-
sure for change. Examples of global pressures include the global institu-
tion of multinational agreements, the information technology revolution,
security against terrorism, prevention of corruption, empowerment of
nongovernmental organizations, and the public management reform ini-
tiatives. As public organizations are open systems, affecting and effected
by their environments, these set of global pressures may cause changes
in structure, behavior, and other important organizational characteristics
of public bureaucracies, including accountability (Selznick; Thompson;
Aldrich; Rainey; Scott).
With the revolution of web-based technologies, the e-government
movement can be taken as part of the global pressure of information tech-
nology. However, even though there is an identifiable set of global pres-
sures, the response from each public bureaucracy is not necessarily
homogenous. Both comparative public administration and globalization
research suggest that, despite similar demands placed on national public
bureaucracies by global pressures, patterns of adoption of technology and
278 WILSON WONG AND ERIC WELCH

organizational change are discernable and the domestic context sur-

rounding public bureaucracy often acts as significant intervening factor.
(Farazmand; Heady 1996b; Aberbach, Putnam, and Rockman; Riggs;
Welch and Wong 2001a, 2001b). In elaborating their framework of global
pressure and bureaucratic change, Welch and Wong (2001a) wrote:
The framework shows that global pressures affect public organizations directly
and indirectly through domestic contexts. Despite significant variation in the
structures and processes of public organizations worldwide, global pressures
create common exigencies upon bureaucracies that result in predictable reac-
tions or changes by public organizations. Moreover, the model suggests that
elements of the domestic context filter the effects of global pressures in pre-
dictable ways. As a result, the domestic context either offsets or reinforces the
change induced by the global pressure on public bureaucracy. (Welch and Wong
2001a, 511)

How will e-government affect accountability? In the framework, there

are two major sources of change for accountability. First, the global pres-
sure of information technology has a direct impact on accountability.
Second, there is an indirect impact of change brought by the domestic
context. In filtering the global pressure of e-government, the domestic
context can either reinforce or offset the influence of e-government on
accountability. It is possible that even if e-government leads to an increase
in the mean accountability level of all nations, the accountability level of
a particular government may still recede because its domestic context
offsets the direct and positive impact of e-government on accountability.
Studies on the effects of global pressures on nations also have a poten-
tial to contribute to the debate on convergence theory: institutions in dif-
ferent countries, including public bureaucracy, tend to converge to a
common pattern (Osborne; Pollitt and Bouckaert; Doremus et al.;
Hallerberg and Basinger; Kettl; OECD 1993, 1995). The interesting and
important questions are not only about convergence or divergence but
also about the contextual factors that influence national level administra-
tive change toward or away from the global trends (Pollitt; Scharpf; Welch
and Wong 2001a). In this study, to specify the domestic context, we focus
on two set of variables: the context of the national civil service system and
the characteristics of the public agencies. The former has a more national
and institutional focus and the latter has a more agency-specific and orga-
nizational focus.
Website and web-based technologies often form a core and indis-
pensable part of any e-government. It is also the focus and contact point
of the new electronic government-and-citizen interface created under
e-government. We would therefore focus on the attributes of websites
in operationalizing e-government accountability in the study. Different
degrees of openness in websites can also expose public organization
tendencies toward accountability (Welch and Wong 1998, 2001a). Change
in the level of website openness represents the revealed level of change
in accountability of the public agency.
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 279

The challenge for accountability in public administration is that many

of the expectations for performance from multiple legitimate sources are
often changing and contradictory (Cunningham and Harris; Johnston and
Romzek; Romzek and Dubnick). However, citizens would not be able to
hold their government accountable if they do not know what it is doing
and have no channel for interacting with it. As long as public organiza-
tions are ultimately accountable to the citizenry, transparency and inter-
activity would be two critical elements for the accountability function of
government.
The Cyberspace Policy Research Group (CyPRG) develops the con-
nection between website openness and e-government accountability in its
comparative research of websites with a focus on these two key elements
of accountability.3 CyPRG defines government websites openness to be a
function of transparency and interactivity. In the CyPRG study, trans-
parency refers to the extent to which an organization reveals work and
decision processes and procedures. Website transparency is equivalent to
“a layman’s basic map of the organization as depicted in the information
on the site [and] reveals the depth of access it allows, the depths of knowl-
edge about processes it is willing to reveal, and the level of attention to
citizen response it provides” (La Porte, de Jong, and Demchak). Interac-
tivity refers to the quality of communication between agency and citizen.
“[It] is a measure of the level of convenience or degree of immediate feed-
back [provided]” (La Porte, de Jong, and Demchak).
CyPRG hypothesizes that greater openness is associated with greater
accountability (Demchak, Friis, and La Porte 1998, 2000; La Porte, de Jong,
and Demchak). A more transparent government allows citizens to
monitor the performance of public organizations more easily through the
increase in the availability of information (Reichard). A more interactive
public organization enhances accountability by being more responsive to
the preferences of the citizenry. As our theoretical understanding of
accountability is similar to the CyPRG constructs, this paper adopts
CyPRG definitions and measures of website openness in measuring e-
government accountability and uses publicly available data from CyPRG.

HYPOTHESES OF E-GOVERNMENT ACCOUNTABILITY

In general, there are two different approaches of measuring the impact of

domestic context on e-government accountability. They are the direct
measure approach and the indirect measure or interaction approach (La
Porte, de Jong, and Demchak; Welch and Wong 2001a). The direct
measure approach refers to the use of direct measurement of the domes-
tic context, such as using the type of political regime and GDP per capita
to capture the influence of political and economic dimensions of the
domestic context, respectively. However, the use of the direct measure
approach often leads to some major statistical problems in empirical
studies.4 Prior research using many of the direct measures often resulted
280 WILSON WONG AND ERIC WELCH

in weak findings that explained little of the variations in website open-

ness (La Porte, de Jong, and Demchak).5
The interaction approach refers to using measurements that capture the
interaction of the different segments of the domestic context (political,
economic, social) for empirical analysis. There has been some success in
experimenting the interaction approach to build empirical models to
explain the impact of global pressure on bureaucratic change (Welch and
Wong 2001a). The interaction approach is also attractive because it focuses
on the interaction among the political, economic, and social dimensions
of the domestic context. This approach makes it able to capture richer
information and allow the analysis to get closer to the in-depth and pro-
found meanings of the theoretical concepts applied in the study. It also
allows researchers to benefit from the major concepts from well-grounded
theories developed on the study of the public bureaucracy, such as the
national civil service system. Because of the above reasons, this study
adopts the interaction approach as its primary approach.6
For hypotheses building, Heady (1996a) provides useful classifications
of national civil service systems. Four major dimensions of Heady’s
framework are adopted in the paper: relation to political regime, qualifi-
cation requirements, role of state, and sense of mission.7 Overall, these
dimensions serve as an index of the power or role of the civil service
system in relation to other elite and power groups in either the major
functions of the civil service system or the core domains of society.
The “relation to the political regime” construct concerns the power of
the civil service. It ranges from minimal independent power, “ruler
responsive,” to maximal power under a “military responsive” regime.
“Qualification requirements” captures bureaucrat involvement in civil
service qualification decisions. At one end, “patrimonial,” political rulers
determine civil service qualifications, at the other, “bureaucratic determi-
nation,” civil servants are in charge. Because these two dimensions are
thought to be measuring the same underlying construct, they are com-
bined into a new variable “political autonomy” in our empirical analysis.
“Role of state” represents the degree of state intervention and pene-
tration in decision making in the polity. This dimension can also be
viewed as a measure of the degree of involvement or intervention by
the civil service in society. It ranges from “traditional,” in which civil ser-
vants play a limited role, to “centrally planned,” where their role is the
greatest.
The fourth dimension, “sense of mission,” captures civil service values.
At one end of this scale, “compliance” requires strict conformity by
bureaucrats to political directives, while at the other, “guidance” systems
portray systems in which civil servants consider themselves to be most
able to intervene, lead and direct. In systems in which guidance is the
sense of mission, civil servants would express a tendency to dominate in
public governance as they would view themselves as the “most legitimate
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 281

and best-equipped group for setting and achieving goals” (Heady 1996a,
220).
Because the adoption of information technology often occurs within
a domestic context, hypotheses relate the dimensions of national civil
service systems and organizational characteristics to concepts of e-
government accountability. Two competing views provide two series of
alternate hypotheses.9 On the one hand, a positive view can be taken on
the accountability orientation of the civil service. Bureaucrats will have the
natural tendency to respect accountability and the professional responsi-
bility to attain it. Therefore, high political control and constraints imposed
on the public bureaucracy can cause e-government accountability to be
diminished. Following this thinking, we can assume that there should be
a linear and positive relationship between the civil service dimensions and
e-government accountability. E-government accountability should rise
with the independence and power of the civil service in society.
While public bureaucrats are taken as professional and responsible
managers in the positive approach, under the public-choice approach,
there are alternative and negative views on the nature of the civil service.
Like the power-seeking politicians, bureaucrats are self-interest-
maximizing individuals (Niskanen; Downs). Once they are in control,
they will do exactly what the politicians do to protect their power base,
even at the expense of the interest of other political participants and the
general public. Therefore, they must be monitored and constrained to a
certain extent before accountable behavior can be expected from
them.
In a global context, national polities may seek to stem the flow of power
from the nation state to global institutions through national policies that
protect their power and authority (Farazmand; Cleveland). As informa-
tion is an important source of power, nations would tend to limit infor-
mation disclosure and openness as one means of maintaining national
political control under globalization (Kraemer and Dedrick; Cleveland).
Similarly, bureaucrats who view themselves as legitimate leaders and
enjoy a high level of independence may limit the ability of external enti-
ties to review decisions or contact responsible parties (Reichard). High
concentration of power in the hands of the bureaucracy causes central-
ization and control of information by the bureaucracy in order to secure
its own power.
With the two different and opposing views, two sets of competing
hypotheses are set. Under the positive view, a linear and positive rela-
tionship is expected. Under the negative view, a nonlinear U-shaped rela-
tionship is expected. This means that only when the civil service is being
situated in a competitive environment, with proper checks and balances
from other political and social actors, will it take the virtue of account-
ability in e-government seriously.
These two sets of competing hypotheses are stated as below:
282 WILSON WONG AND ERIC WELCH

Linear relationship

H1: The greater the level of political autonomy of the civil service, the
higher the e-government accountability of the public bureaucracy.
H2: The greater the role of state (civil service) in society, the higher the
e-government accountability of the public bureaucracy.
H3: The greater the sense of mission of the civil service, the higher the
e-government accountability of the public bureaucracy.

Nonlinear relationship

H4: High and low levels of political autonomy of the civil service lead
to reductions in e-government accountability while a moderate
level of political autonomy leads to increases in e-government
accountability (U-shaped curve).
H5: High and low levels of the role of state (civil service) in society lead
to reductions in e-government accountability while moderate levels
of state role lead to increases in e-government accountability (U-
shaped curve).
H6: High and low levels of sense of mission of the civil service leads to
lower levels of e-government accountability while moderate levels
of sense of mission of the civil service leads to higher levels of
e-government accountability (U-shaped curve).

In setting up the hypotheses for organizational characteristics, we dis-

tinguish between internally focused (justice, education, and labor sectors)
and externally focused agencies (defense, finance, and immigration).10
Internally focused agencies have missions primarily associated with
national issues, while externally focused agencies have missions associ-
ated with a substantial international component. Externally focused
agencies usually have a stronger need to use the internet to interact with
parties outside the national border. Disclosure or availability of informa-
tion is also a symbol of trust, modernity, and global citizenship that may
be necessary for competitive vitality and political legitimacy of a nation
(Strang and Meyer; DiMaggio and Powell). Therefore, it is hypothesized
that:

H7: Bureaucracies of externally focused sectors have higher e-

government accountability than those of internally focused
sectors.

Finally, as the open economy is often believed to be a major driving

force for e-government (Welch and Wong 2001b), it is hypothesized
that:
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 283

H8: The more open the economy public bureaucracies face, the higher
their e-accountability.

THE COUNTRIES, DATA, AND METHODS

Fourteen countries covering five continents are included in the study:

Australia, Canada, China, Egypt, France, Germany, Indonesia, Japan,
Korea, the Netherlands, New Zealand, Singapore, the United Kingdom,
and the United States. These countries are chosen primarily because they
represent a wide variation in the policy variables we are interested in for
the study. Thus both developed and developing countries are included.
Although they are not a random sample of all the e-government coun-
tries, they should still serve as useful reference points for similar coun-
tries in their respective regions.
The study uses two primary data sources: the CyPRG database on the
transparency and interactivity of national agency websites and Ferrel
Heady’s framework for distinguishing among civil service systems.
Coding for the Heady constructs was based primarily on Heady’s own
work.11 According to Heady’s work, each dimension is measured on a
continuum from low to high. For example, the “relations to political
regime” measures independent power of the civic service in a regime
where nations classified as “ruler responsive” are coded between 1 and
2.5, nations coded as “single party responsive” are coded between 2.6 and
5.0, and so on. The maximum score for each dimension is 10.
CyPRG has collected data on website transparency and interactivity
since the inception of the project in 1995. Transparency and interactivity
are two elements of openness, which is considered to be a reasonable
proxy measure of accountability. As mentioned above, transparency
measures the amount of data available on agency websites and interac-
tivity measures the ease with which users are able to access data or people
(Demchak, Friis, and La Porte 2000; CyPRG 1; CyPRG 2). Both measures
represent tallies of predetermined qualities of the agency website.
The transparency measure represents tallies of website qualities in five
areas (ownership, contacts, issue or organizational information, citizen
consequences, and timeliness of data) and interactivity measures the
presence or absence of qualities in four areas (ownership, reachability,
issue or organizational information, and citizen consequences). Each of
the substantive areas contains within it a set of measurable criteria which
are scored 0 or 1 according to their presence or absence. Scores are
summed across areas to provide an overall measure of transparency
or interactivity. Openness is a linear sum of transparency and
interactivity.12
As we are primarily interested in bureaucratic change, we calculate the
dependent variable as the score of 2000 minus the 1997 score.13 Change
score has been found to be a better measure for accountability as it cap-
tures the dynamics of change and policy choice in e-government better
284 WILSON WONG AND ERIC WELCH

(La Porte, de Jong, and Demchak; Welch and Wong 2001a).14 Correlations
between transparency and interactivity are high for 1997 (r = 0.61) and
2000 (0.62) data. Combining transparency and interactivity into one vari-
able to represent openness has some merit as the Chronbach Alpha
correlation coefficients is 0.75 and 0.77 for 1997 and 2000, respectively.
National agencies from the fourteen countries that have a website data in
1997 and 2000 form the data for our study, a total of 267 agencies.
To measure national economic openness, we use a combined measure
of 1997 exports as a percentage of GDP and 1997 imports as a percentage
of GDP.15 Figures were taken from World Bank statistics (World Bank).
For the external/internal focus variable, externally focused agencies were
coded 1 and all others were coded 0. We also divided agencies into three
groups: external political, economic and industrial, and domestic public
service. The external political group included defense, executive, finance,
and foreign agencies. The economic and industrial category included
science and technology, communications, industry and trade, and trans-
portation and infrastructure categories. The domestic public service
group included culture, education, health, social services, and library-
related agencies. As a final alternative, we coded a number of sectors as
dummy variables and included them in the regression analysis. The three
different coding techniques for agency-specific characteristics required
three separate models.
Descriptive statistics for all variables are shown in Table 1. We conduct
Ordinary Least Squares (OLS) regression analysis in which dependent
variables of openness, transparency, and interactivity are regressed on
measures of the civil service systems variables, the organizational char-
acteristic variables, and the economic openness variable. For each regres-
sion, we recoded “political autonomy”, “role of state”, and “mission” into
three-level dummy variables to test for nonlinear effects. The “mission”
variable—the shared and self-perceived values of the bureaucrats on its
role in society—is a critical variable that tests directly our two competing
perspectives of the accountability orientation of bureaucrats. Therefore,
to better test the two competing hypotheses, two sets of regression are
run for the variable: one for the linear relationship and one for the non-
linear relationship.

FINDINGS

Figure 1 indicates that all the attributes of accountability (transparency,

interactivity, and openness) generally increased across all countries
between 1997 and 2000, with a substantial jump in both measures in 1999
and 2000. Aggregate findings indicate some support for the convergence
theory that the general level of accountability of the countries has
increased across time. However, this ignores that there are considerable
variations among countries not only in terms of the level of accountabil-
ity but also in terms of the slope and direction of change (see Appendices
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 285

TABLE 1
Descriptive Statistics of Independent and Dependent Variables in the Study
(n = 267)

Variable Mean Standard Minimum Maximum

Deviation

Mission 6.72 1.53 4 10

Political regime 7.68 0.92 5 10
Qualification requirements 8.11 0.79 4 10
Role of state 5.45 1.28 4 8
Personnel management 7.96 0.75 6.75 10
Political autonomy 7.89 0.78 4.5 10
External sector 0.47 0.50 0 1
External political sector 0.22 0.42 0 1
Industry & trade sector 0.09 0.28 0 1
Domestic service provision 0.17 0.37 0 1
sector
Exports as a percent of GDP 35.34 34.17 10 187
Imports as a percent of GDP 33.27 30.40 9 170
Transparency change 6.21 4.34 -15.0 17.3
1997–2000
Interactivity change 1997–2000 5.61 4.34 -17.0 27.8
Openness change 1997–2000 11.83 7.62 -32.0 31.8

FIGURE 1
Website Data Trends (All Countries)

25

20

15

10

0
1997 1998 1999 2000

Transparency Interactivity Openness

1 and 2 for individual trends of all fourteen countries).16 These differences

are probably due to the elements of national and organization differences,
the primary subjects of the study.
Findings from regression analysis are arranged into three tables (Tables
2, 3, and 4).17 Results from the regression analysis indicate relatively
strong and consistent support for the set of nonlinear hypotheses.18 First,
286 WILSON WONG AND ERIC WELCH

TABLE 2
Regression Results (Standardized Coefficients) with External and Internal
Sector Variable

Transparency Interactivity Openness

Change Change Change

Intercept (medium reference) 7.11 (1.12)*** 7.35 (1.20)*** 14.45 (2.04)***

Mission 0.32 (0.16)** -0.31 (0.17)* 0.01
Socioeconomic context
Low -5.36 (0.66)*** -0.05 (0.70) -5.42 (1.20)***
High -1.34 (1.01) -2.18 (1.09)** -3.52 (1.84)**
Political autonomy
Low -1.83 (0.96)** -3.73 (1.04)*** -5.55 (1.76)***
High -1.56 (0.97)* -2.51 (1.04)** -4.07 (1.76)**
Economic openness -0.08 (0.04)** 0.07 (0.04) -0.01 (0.07)
External sector 0.37 (0.46) 1.03 (0.49) 1.40 (0.83)*
n 267 267 267
Adjusted R-square 0.30 0.18 0.23
Model significance *** *** ***

***p < 0.01, **p < 0.05, *p < 0.10.

TABLE 3
Regression Results (Standardized Coefficients) with Sector Variables of
External Political, Economic & Industrial, and Internal Public Service

Transparency Interactivity Openness

Change Change Change

Intercept (medium reference) 7.69 (1.09)*** 7.90 (1.19)*** 15.59 (2.00)***

Mission 0.31 (0.15)** -0.32 (0.16)** -0.01 (0.27)
Socioeconomic context
Low -5.10 (0.64)*** 0.18 (0.70) -4.92 (1.17)***
High -1.11 (0.98) -1.90 (1.07)* -3.00 (1.79)*
Political autonomy
Low -1.69 (0.94)* -3.80 (1.03)*** -5.49 (1.72)***
High -1.47 (0.94) -2.58 (1.02)** -4.04 (1.71)**
Economic openness -0.07 (0.04) * 0.08 (0.04)* 0.01 (0.07)
Sectors
External political -1.16 (0.60)** 0.10 (0.66) -1.06 (1.10)
Economic & industrial 0.16 0.38 (0.62) 0.53 (1.04)
Internal public service -2.49 (0.65)*** -2.39 (0.70)*** -4.88 (1.17)***
n 267 267 267
Adjusted R-square 0.34 0.21 0.23
Model significance *** *** ***

***p < 0.01, **p < 0.05, *p < 0.10.

findings in all three tables show generally consistent nonlinear effects of

“role of state” on transparency, interactivity, and openness. Low and high
“role of state” are very often found to be significantly negatively associ-
ated with transparency, interactivity, and openness. These findings
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 287

TABLE 4
Regression Results (Estimates and Standard Errors) Testing the Nonlinear
Relationship of the Mission Variable

Transparency Change Interactivity Change

Intercept (reference) 8.23 (0.95)*** 7.83 (1.05)***

Mission
Low 1.62 (1.72) -3.97 (1.91)**
High 2.25 (0.86)*** -2.60 (0.95)***
Socioeconomic context
Low -5.05 (0.68)*** -1.54 (0.75)**
High -3.27 (1.58)** 1.52 (1.75)
Political autonomy
Low -2.30 (0.90)*** -3.03 (1.00)***
High -1.68 (0.92)* -3.94 (1.02)***
Economic openness -0.03 (0.05) -0.03 (0.05)
Sector
Finance 0.29 (0.73) 2.33 (0.81)***
Industry & trade 1.56 (0.75)** 0.60 (0.83)
Executive -3.47 (1.23)*** -2.48 (1.37)*
Foreign 0.44 (0.92) 0.04 (1.02)
Culture -5.39 (0.95)*** -5.67 (1.05)***
Government operations 2.79 (1.38)** 2.27 (1.53)
Justice 2.64 (0.73)*** 0.92 (0.81)
Health 0.78 (1.15) -0.73 (1.27)
n 267 267
Adjusted R-square 0.44 0.30
Model significance *** ***

***p < 0.01, **p < 0.05, *p < 0.10.

support hypothesis H5. Websites in nations in which the role of the civil
service is high and low tend to be less open than in nations with mixed
competitive systems.
We also find consistent nonlinear relationship between “political
autonomy” and transparency, interactivity, and openness for most regres-
sion runs. Excepting the transparency model in Table 3, all other results
show that low and high political autonomy are negatively related while
medium level of political autonomy is positively associated with trans-
parency, interactivity, and openness. This suggests that in nations where
either politicians or bureaucrats hold a high degree of independent
power, website openness and hence accountability are lower than in
nations where competitive mechanisms are more prevalent.
Tables 2 and 3 show the testing of the linear relationship of the mission
variable. It is found to be positively associated with transparency but
negatively associated with interactivity. The effects cancel each other out
when mission is regressed on openness. This contradicts our expectation
288 WILSON WONG AND ERIC WELCH

that the direction of association between independent variables and trans-

parency and interactivity would be the same.
Partly due to the problematic findings and partly due to our special
interest in the mission variable, it is tested for nonlinear relationships in
Table 4. Results show a positive linear relationship between mission and
quantity of data (transparency), significant for high levels of mission ori-
entation (H3 is partially supported). However, findings also show a non-
linear relationship with interactivity in which low and high mission levels
are negatively associated but moderate levels are positively associated
(H6 is also partially supported). All other results are similar with the
models in Tables 2 and 3. The new findings appear to suggest that inter-
activity and transparency are measures of different phenomena. There-
fore, we did not combine the two variables into the openness variable for
this final set of regression. We interpret that these findings on the mission
variable indicate two mechanisms of technology use by the bureaucracy
or rulers in general. Different technology uses can have different impli-
cations on governance. The new findings imply that in countries where
bureaucrats hold a stronger sense of mission, websites increasingly
provide more data and information, but provide increasingly restricted
access to the agency.
Findings also show consistent evidence that agency-specific organiza-
tional characteristics matter for accountability. In Table 2, external sector
is significantly positively associated with openness. H6 is supported.
In Table 3, recoding of the external/internal sector variable into three
categories finds a negative relationship between public service sector
agencies (health, education, etc.) and all three dependent variables. Exter-
nal political sector agencies (finance, trade, foreign) are also negatively
associated with transparency. Many significant findings are also found for
the sector variables in Table 4. The adjusted R-square is the largest for the
regression model with dummy variables for the sectors. Therefore, in
terms of explanatory power, using sector-based dummy variables to
model interactivity, transparency, and openness is more appropriate and
provides richer information. While the independent variables that repre-
sent Heady’s civil service dimensions show how the bureaucratic envi-
ronment effects openness, agency-specific indicators show that
organizational characteristics also determine how technology shapes
accountability.
Finally, it is surprising to find that economic openness is not signifi-
cant for website openness in all regression models.19 H7 is not supported.
It is possible that when information technology becomes a common global
pressure for nations and its technology is increasingly available and
affordable, the normative and policy pressure to have an enhanced
and sophisticated website may be more relevant than economic concerns
and necessity. Accountability may have become more policy driven than
economically and technically driven.
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 289

DISCUSSION

Consistent with existing research on information technology and organi-

zation change, this study finds that e-government often only exacerbates
the existing nature and attributes of public bureaucracies (Bovens and
Zouridis; Welch and Wong 2001b; Kraemer and Dedrick; Kraemer,
Dedrick, and King). Information technology does not act out of context
in affecting public organizations. E-government accountability is more
about nations and bureaucracies than simply about technology per se.
The effect of e-government on accountability of public organization is
affected both by the civil service system within which it is embedded and
by its agency-specific organizational characteristics. The nonlinear and U-
shaped bureaucratic response to e-government accountability suggests
that bureaucrats think and behave as rulers to control information to con-
solidate power in their adoption and management of e-government. A
pure administrative state governed by bureaucrats, without the checking
of elected politicians and the civil society, may lead to a less accountable
and open government (Aberbach, Putnam, and Rockman). A strong polit-
ical regime with a weak civil service will also lead to similar drop in
accountability. This implies that competition or proper share of power
between the political elite and the civil service will be an important
factor in determining whether e-government enhances or reduces
accountability.
Adding on previous studies but following the same line of logic, our
study shows that not only the general characteristics of the civil service
affect accountability, the specific characteristics of agency also matter.
Although there is no clear-cut generalization about the impact of the
agency variables yet, it can be deducted from the findings that agencies
often determine their website openness with a strategic mind-set on
whom they see as their target user groups.
The set of interest groups relevant to government agencies can be clas-
sified into clients, those who receive the services, and constituents, who
demand the service (Viteritti). If the website of the agencies is used to
serve the clients but the clients do not happen to be the constituents
who provide political support for the agencies, there may be a tendency
for the agencies to provide less interactive and transparent services
through their websites. This may explain why the executive political
agencies have a negative relationship with transparency. Executive
political agencies usually depend less on the website to communicate and
cultivate their constituencies. Furthermore, a higher level of secrecy and
isolation from the public can sometimes serve them well to prevent
unnecessary exposure of information which can be politically dangerous
to the agencies.
Departing from previous studies, however, our study poses the
question that interactivity and transparency can each represent a differ-
290 WILSON WONG AND ERIC WELCH

ent dimension of the accountability relationship between citizenry

and government under e-government. Findings show that the two attrib-
utes sometimes elicit different responses to the same domestic context.
This means that transparency and interactivity can serve different
and separated political and strategic functions for the bureaucracy.
Bureaucrats can therefore use the web as a tool for information dissemi-
nation on the one hand while trying to use it to limit interaction on the
other hand.
For example, in a civil service system of a high mission level, bureau-
cracies show greater transparency but place greater interactivity restric-
tions in their websites. More transparency may help the mission-oriented
agency to “socialize” the public with the ideologies and visions of the
agency. Governments can place what they want people to know, or what
they believe they have a duty or desire to share in the public domain, yet,
prefer to limit direct interaction. There is a danger that the new govern-
ment-and-citizen interface created by e-government may simply be used
as additional channel for more political propaganda and political control
rather than real accountability enhancement. Considering the possibility
of this, website establishment should not be taken automatically as an
accountability enhancement without a careful evaluation of the website
attributes.
To sum up, all the findings of the civil service system variables and
agency-specific variables point to the same direction. Technology is often
not adopted for the sake of acquisition alone, it is adopted and then
adapted to meet institutional and organizational needs as defined by key
decision makers. When the agency-specific organizational variables give
the empirical model larger explanatory power, more contextual informa-
tion of the public organizations are usually needed for more in-depth
interpretation. Therefore, linking each major organizational variable with
accountability of e-government more directly, pinpointing the relevancy
of these variables, and further elaborating on the different dimensions of
public accountability under e-government should be some of the major
future directions of research in e-government.
Equally important, the study sheds some new light on the globaliza-
tion debate of global convergence and national divergence on e-
government. In some ways, convergence perspectives are upheld by our
findings. With e-government, public accountability in general is increas-
ing over time under the global information technology pressure. Never-
theless, important domestic factors result in divergence in e-government
accountability at both the national and organizational levels. Although
the direct effect of the global pressure of information technology, that is
the adoption of e-government, will lead to a general and overall rise in
accountability, the indirect effect of the domestic context will lead to a
divergence in accountability among the countries and agencies.
As a result, the general level of accountability rises in a global sense
but the accountability gap, the actual and absolute difference between
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 291

countries and agencies in accountability, will probably be maintained or

even widened. In other words, the difference in accountability among
nations and agencies cannot be narrowed simply by the introduction and
spread of web-based e-government technology. This is similar to the effect
of the internet in the business world. As internet is available for all firms,
it ceases to be an advantage for enhancing a firm’s competitiveness over
other firms (Porter).20 Similarly, as e-government technology is available
for and adopted by most governments, it ceases to be the technology that
makes a decisive difference in narrowing the accountability gap among
governments of different nations.
As web-based technologies become widely available and affordable,
e-government will become more policy driven than technology and
economic driven. It will be the normative pressure of the global com-
munity and the domestic context that drive the growth and change of e-
government (DiMaggio and Powell; La Porte, de Jong, and Demchak).
Public accountability expressed by e-government will therefore become
more and more a conscious policy choice that reflects both national and
organizational characteristics.
Pollitt provides a very useful classification in summing up the con-
vergence issues in globalization and public management. He carefully
elaborates the concept of convergence and classifies four different stages
of convergence: discursive convergence, decisional convergence, practice
convergence, and results convergence. A significant contribution of his
work is pointing out that convergence at one stage does not necessarily
imply convergence at the next stage and the convergence process can be
discontinued at any stage.
The significance of this for our analysis is that it supports our earlier sugges-
tion that convergence could be taking place at one or more stages without
necessarily doing so at all four. In other words there may be a considerable
convergence of discourse and/or of decisions, without anything like the
same degree of convergence of practice (and still less of results). (Pollitt,
487)
In e-government, it only converges up to the stage of practice in the
sense that e-government has been widely adopted by nations around the
globe. The adoption of the technology itself is a global convergence.
However, the convergence of results in e-government in terms of account-
ability has not happened so far. Differences in national and organizational
factors have led more toward national divergence, rather than global con-
vergence, in public accountability under e-government.
Introducing e-government without the corresponding institutional
reform of the civil service system and organizational reform of the agen-
cies may only lead to limited success in enhancing accountability. The
effect of technology on organizational change should therefore never be
overstated. It is simply a myth that e-government will automatically and
dramatically change the accountability nature of public organizations.
The question of “whether e-government promotes accountability” cannot
292 WILSON WONG AND ERIC WELCH

be answered completely without knowing what kind of bureaucracy one

is referring to in the first place.

ACKNOWLEDGMENT

This research project is jointly funded by the visiting fellowship awarded

to the first author by the Center for Northeast Asian Policy Studies, Brook-
ings Institution, and a research grant from the Research Grant Council of
Hong Kong (RGC ref. no.: CUHK 4223/02H). The authors would also like
to thank the anonymous reviewers for their valuable comments.

NOTES
1. According to the American Society for Public Administration (ASPA), e-
government is defined as: “the pragmatic use of the most innovative infor-
mation and communication technologies, like the internet, to deliver
efficient and cost effective services, information and knowledge. It is an
unequivocal commitment by decision-makers to strengthening the partner-
ship between the private citizen and the public sector.” (definition taken
from http://www.apsanet.org) Broadly speaking, e-government may also
include the use of communication technologies other than the internet for
service production and delivery. Because of the research purpose of the
study, we will mainly focus on the use of internet and other web-based tech-
nologies in our study of e-government.
2. This statistics is taken from the government website data of the Cyberspace
Policy Research Group (CyPRG), available online at http://www.cyprg.
arizona.edu.
3. CyPRG is a research group funded by the National Science Foundation and
is based at the University of Arizona, Tucson and George Mason Univer-
sity. It has collected data on the openness of websites of national govern-
ments in the world since 1996.
4. Because of the high correlation among the direct measures, multicollinear-
ity is one of the major statistical problems encountered by the approach.
5. For example, in an empirical study conducted by La Porte and his col-
leagues (La Porte , de Jong, and Demchak), among all the direct measures
of the domestic context they use, website openness is found to be related
with national income and similar wealth measures. All the direct social and
political measures are found to be insignificant.
6. Partly due to the data nature and limitation, the direct approach will also
be applied in operating some of the variables in the study, such as economic
openness and some of the agency-specific organizational characteristics.
7. One of the dimensions of Heady, the focus for personnel management, is
not adopted in the study. To a certain extent, this dimension is slightly dif-
ferent in nature from other dimensions. Instead of capturing the relative
power of the civil service in different domains of society, it tends to capture
more about the locus and operational arrangement of the personnel func-
tion. Second, prior work has found high correlation between the two vari-
ables: “role of state in society” and “focus for personnel management”
(Welch and Wong 2001a). A similar result was found in this data set (r =
0.47), however, the Chronbach Alpha was not high enough (r = 0.58) to merit
a combination of these variables. Problems of multicollinearity preclude the
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 293

inclusion of both variables in the regression model. As a result, “focus

for personnel management” was dropped from the model. This variable
also has the lowest intercoder correlation and the least variation among
nations.
8. “Relation to political regime” and “qualification requirements” capture the
political autonomy of the civil service in relation to the political regime in
internal and external affairs respectively. There are also statistical reasons
to back up such combination. Preliminary analysis found high correlation
(r = 0.71) between two coded categories from the Heady framework. The
Chronbach Alpha coefficient was 0.82, validating a linear combination
of the standardized variables into a new variable that we call “political
autonomy.”
9. The competing views relate to traditional views on the accountability ori-
entation of bureaucracies. This debate can actually be dated back to the
debate between Finer and Friedrich.
10. These sector distinctions were developed and coded by CyPRG.
11. The coding of seven countries (Egypt, France, Indonesia, Japan, Korea, the
United Kingdom, and the United States) are done by Heady in his own
work (1996a). The other eight countries are coded by the researchers
independently according to the Heady characteristics. Readers can refer to
Welch and Wong (2001a) for more explanation of the coding method and
samples of the coding results. The coding results are then averaged to obtain
one measure of each dimension for each country. Intercoder reliability, mea-
sured as the correlation between the two sets of researcher-coded data,
ranged between 0.89 and 0.94.
12. CyPRG data and the detailed definition of each criterion are available online
at http://www.cyprg.arizona.edu. Although the titles of some interactivity
and transparency areas are identical, the measurement qualities are not. We
do not reprint the criteria here due to space considerations.
13. Because relatively few websites were assessed in the first two years of the
CyPRG study, we rely on the most recent four years of CyPRG data from
1997 through to 2000.
14. A weakness of average score of accountability is that it is only a measure of
a single point. It often reflects “momentary lags in technology application,
familiarity with new processes or other transitory phenomena,” instead of
policy choices of government. The change score measures the change in the
level of e-government accountability of a government across time in com-
parison with its own accountability in previous years. As a comparison and
change measure, it can measure the degree of willingness of public officials
and managers to make their organizations more accountable through e-
government when the technology, knowledge and experience of allowing
them to do so are already available.
15. The alpha correlation coefficient between these two variables was high
enough (r = 0.98) to validate a linear combination of the standardized
variables.
16. For example, the interactivity gap between the US and France has been
significantly narrowed in 2000. At the same period, the interactivity
gap between France and the Netherlands has been significantly widened.
17. All models are statistically significant and statistical tests show that nor-
mality assumptions were not violated and that there were no problems of
multicollinearity.
18. Subsequent runs of this model, alternatively replacing low and high
dummy variables with the medium dummy variable, showed similar evi-
dence of a strongly significant nonlinear relationship.
294 WILSON WONG AND ERIC WELCH

19. In Table 2, economic openness is not significant for interactivity. Although

it is significant for transparency in both Table 2 and Table 3, it has the wrong
sign and very small coefficients.
20. According to Porter, the adoption of internet in the business world will not
give the firm a sustainable competitive advantage. As the internet technol-
ogy becomes widely available to all firms, it will mean that no firm will
have a competitive edge by the use of the internet. Therefore, what inter-
net does is raising the competitiveness of all firms, instead of the relative
competitiveness of any specific firm. It changes the competition landscape
at the market level, but will not give any individual firm a competitive edge
in competing in the new landscape.

REFERENCES

Aberbach, Joel, Robert Putnam, and Bert Rockman. 1981. Bureaucrats and Politi-
cians in Western Democracies. Cambridge, MA: Harvard University Press.
Aldrich, Howard. 1979. Organizations and Environments. Englewood Cliffs, NJ:
Prentice-Hall.
Bovens, Mark, and Stavros Zouridis. 2002. From Street-Level to System-Level
Bureaucracies: How Information and Communication Technology is Trans-
forming Administrative Discretion and Constitutional Control. Public Admin-
istration Review 62:174–184.
Cleveland, Harlan. 1993. Birth of a New World: An Open Moment for International
Leadership. San Francisco: Jossey-Bass.
Cunningham, Gary, and Jean Harris. 2001. A Heuristic Framework for Account-
ability of Government Subunits. Public Management Review 3:145–165.
Cyberspace Policy Research Group (CyPRG), 1. Available online at http://
www.cyprg.arizona.edu/hypo_content.htm
Cyberspace Policy Research Group (CyPRG), 2. Available online at
http://cyprg.arizona.edu/cqms_content.htm
Demchak, Chris C., Christian Friis, and Todd M. La Porte. 1998. Configuring
Public Agencies in Cyberspace: A Conceptual Investigation. In I. T. M. Snellen
and W. B. H. J. van de Donk, eds., Public Administration in an Information Age:
A Handbook, pp. 179–196. Amsterdam: IOS Press.
———. 2000. Webbing Governance: National Differences in Constructing the Face
of Public Organizations. In G. David Garson, ed., Handbook of Public Informa-
tion Systems. New York: Marcel Dekker.
DiMaggio, Paul J., and Walter Powell. 1983. The Iron Cage Revisited: Institutional
Isomorphism and Collective Rationality in Organizational Fields. American
Sociological Review 48:147–160.
Doremus, Paul N., William W. Keller, Louis W. Pauly, and Simon Reich. 1998.
The Myth of the Global Corporation. Princeton, NJ: Princeton University
Press.
Downs, Anthony. 1967. Inside Bureaucracy. Boston: Little Brown.
Farazmand, Ali. 1999. Globalization and Public Administration. Public Adminis-
tration Review 59:509–522.
Finer, Herman. 1941. Administrative Responsibility and Democratic Government.
Public Administration Review 1:335–350.
Friedrich, Carl. 1940. Public Policy and the Nature of Administrative Responsi-
bility. In C. Friedrich and E. Mason, eds., Public Policy. Cambridge, MA:
Harvard University Press.
Hallerberg, Mark, and Scott Basinger. 1998. Internationalization and Changes
in Tax Policy in OECD Countries: The Importance of Domestic Veto Powers.
Comparative Political Studies 31:321–352.
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 295

Heady, Ferrel. 1996a. Configurations of Civil Service Systems. In Hans Bekke,

James Perry, and Theo Toonen, eds., Civil Service Systems in Comparative Per-
spective. Bloomington: Indiana University Press.
———. 1996b. Public Administration in Comparative Perspective, 5th ed. New York:
Marcel Dekker.
Heintze, Theresa, and Stuart Bretschneider. 2000. Information Technology
and Restructuring in Public Organizations: Does Adoption of Information
Technology Affect Organizational Structures, Communications, and
Decision-Making. Journal of Public Administration Theory and Research 10:801–
830.
Johnston, Jocelyn, and Barbara Romzek. 1999. Contacting and Accountability in
State Medicaid Reform: Rhetoric, Theories, and Reality. Public Administration
Review 59:383–399.
Kettl, Donald. 1997. The Global Revolution in Public Management:
Driving Themes, Missing Links. Journal of Policy Analysis and Management
16:446–462.
Kraemer, Kenneth, and Jason Dedrick. 1997. Computing and Public Organiza-
tions. Journal of Public Administration Research and Theory 7:89–112.
Kraemer, Kenneth L., Jason Dedrick, and John Leslie King. 1995. The Impact of
Information Technology on City Government in the United States. CRITO Working
Paper. Irvine: Center for Research on Information Technology and Organiza-
tions, University of California.
La Porte, Todd M., Martin de Jong, and Chris Demchak. 2002. Democracy and
Bureaucracy in the Age of the Web. Administration and Society 34:411–426.
Lee, Geungjoo, and James Perry. 2002. Are Computers Boosting Productivity: A
Test of the Paradox in State Governments. Journal of Public Administration
Theory and Research 12:77–102.
Niskanen, William. 1971. Bureaucracy and Representative Government. Chicago:
Aldine, Atherton.
Organisation for Economic Co-operation and Development (OECD). 1993. Public
Management: OECD Country Profiles. Paris: Public Management Service
(PUMA), OECD.
———. 1995. Public Management Developments Update 1995. Paris: Public Manage-
ment Service (PUMA), OECD.
Osborne, Stephen. 2001. Public Management Reform: A Case for National
Divergence or Global Convergence. Public Management Review 3:451.
Pollitt, Christopher. 2001. Clarifying Convergence: Striking Similarities and
Durable Differences in Public Management Reform. Public Management Review
3:471–492.
Pollitt, Christopher, and Geert Bouckaert. 2000. Public Management Reform: A
Comparative Analysis. Oxford: Oxford University Press.
Porter, E. Michael. 2001. Strategy and the Internet. Harvard Business Review
March:63–78.
Rainey, Hal G. 1998. Understanding and Managing Public Organizations. San
Francisco: Jossey-Bass.
Reichard, Christoph. 1998. The Impact of Performance Management on Trans-
parency and Accountability in the Public Sector. In Annie Hondeghem, ed.,
Ethics and Accountability in a Context of Governance and New Public Management.
Amsterdam: IOS Press.
Riggs, Fred W. 1994. Global Forces and the Discipline of Public Administration.
In Jean-Claude Garcia-Zamor and Renu Khator, eds., Public Administration in
the Global Village. Westport, CT: Praeger.
Romzek, Barbara, and Melvin Dubnick. 1987. Accountability in the Public
Sector: Lessons from the Challenger Tragedy. Public Administration Review
47:227–238.
296 WILSON WONG AND ERIC WELCH

Rosenbloom, David. 1992. Public Administrative Theory and the Separation of

Powers. In Jay Shafritz and Albert Hyde, eds., Classics of Public Administration,
3rd ed. Pacific Grove, CA: Books/Cole.
Scharpf, Fritz W. 1997. Introduction: The Problem-Solving Capacity of Multi-Level
Governance. Journal of European Public Policy 4:520–538.
Scott, W. Richard. 1998. Organizations: Rational, Natural and Open Systems, 4th ed.
Englewood Cliffs, NJ: Prentice Hall.
Selznick, Philip. 1966. TVA and the Grass Roots. New York: HarperCollins.
Strang, David, and John W. Meyer. 1993. Institutional Conditions for Diffusion.
Theory and Society 22:487–511.
Thompson, James. 1967. Organization in Action. New York: McGraw-Hill.
Viteritti, Joseph. 1990. Public Organization Environments: Constituents, Clients,
and Urban Governance. Administration and Society 21:425–457.
Welch, Eric, and Wilson Wong. 1998. Public Administration in a Global Context:
Bridging the Gaps of Theory and Practice between Western and Non-Western
Nations. Public Administration Review 58:40–49.
———. 2001a. Global Information Technology Pressure and Government
Accountability: The Mediating Effect of the Domestic Context on Website
Openness. Journal of Public Administration Theory and Research 11:509–538.
———. 2001b. Effects of Global Pressures on Public Bureaucracy: Modeling a New
Theoretical Framework. Administration & Society 32:371–402.
World Bank. 1998. World Development Indicators. Washington, DC: The World Bank.

APPENDIX 1
Website Transparency Trends
18
16
14
12
10
8
6
4
2
0
1997 1998 1999 2000
All Countries Australia Canada China
Egypt France Germany Indonesia
Japan Korea Netherlands New Zealand
Singapore United Kingdom United States
 DOES E-GOVERNMENT PROMOTE ACCOUNTABILITY? 297

APPENDIX 2
Website Interactivity Trends
14

12

10

0
1997 1998 1999 2000

All Countries Australia Canada China

Egypt France Germany Indonesia
Japan Korea Netherlands New Zealand
Singapore United Kingdom United States