Who uses e-government?

Who Uses E-government? Examining the Digital Divide in E-government Use Taewoo Nam and Djoko Sigit Sayogo University at Albany, State University of New York

{tnam, dsayogo}@ctg.albany.edu ABSTRACT This empirical study examines the digital divide in e-government adoption and profiles e-government users, by analyzing the data from the national random-sampled survey that the Pew Internet and American Life Project conducted via telephone interviews on American citizens in 2009. The path analysis suggests four main findings. First, socio-demographic conditions strongly matter for e-government use. Younger generations and socioeconomically advantaged people use e-government more than their counterparts. Second, perceived usefulness of e-government contributes to actual use of e-government. Third, the effect of trust in government on e-government adoption is indirect through perceived usefulness rather than directly causal. Those with higher levels of trust in overall government would likely perceive value of e-government, and then those who perceive potential benefits from using e-government adopt e-government. Last, Internet use intensity is highly associated with e-government use intensity.

Categories and Subject Descriptors H.4.2 [Information Systems Applications]: Type of systems – e-government applications

General Terms Management, Performance, Human Factors, Theory

Keywords E-government, Digital divide, Usage divide, Trust in government, Technology acceptance, Perceived usefulness

1. RESEARCH MOTIVATION The common definition of e-government refers to the use of information and communication technologies (ICTs)–– predominantly the Internet––to deliver information and services to citizens, businesses, government agencies, and other entities [13,20,48,58,64]. From the conventional conceptualization, the functionality and utility of e-government are broadly divided into two categories [37]: internal (potential as effective and efficient managerial tools within the public sector) and external (facilitating government’s linkage with various external stakeholders such as citizens, businesses, and other governmental units). Two primary obstacles to further e-government development have been identified: institutional and organizational Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. ICEGOV 2011, September 26-28, 2011, Tallinn, Estonia. Copyright 2011 ACM 978-1-4503-0746-8…$10.00

barriers in the internal and supply side [37], and social exclusiveness in the external and demand side [43]. Both internal and external directions of e-government maturity are equally important, but this study focuses on digital exclusion in egovernment adoption, which has become the area where potentials of ICTs fundamentally are undermined. E-government has been touted as an effective tool to improve the way by which citizens interact with governments for contact, transaction, information, and governance. Such an enthusiastic expectation, however, has confronted a formidable barrier that impedes the nationwide spread of technology-enabled benefits to diverse segments of the population. That impediment is a digital divide. E-government brings risks as well as benefits by creating the digital divide that deepens the disadvantage of already socially disadvantaged citizens [19]. By providing electronic services to a select group of people, government agencies miss the opportunity to interact with and solicit feedback from a larger portion of the population [4]. A volume of studies concerned and examined the impact of the digital divide on the extent to which e-government potentials are achieved [1-4,6,12-3,20,24,48-9,52]. Another stream of empirical research has untangled why some people adopted e-government services but others don’t––socio-demographic and psychological antecedents, determinants, and/or predictors of e-government usage in an individual level [18,21,34,39,40,43-4,46,51,54]. So far academics and practitioners have already profiled egovernment adopters with a variety of lens (i.e., socioeconomic stratification, generational gaps, political and civic participation, trust based on social capital, technology acceptance, and diffusion of innovations). Notwithstanding, there is something to augment and contribute to the ongoing discussions of e-government adoption. This study will re-establish the causal relationship among the usual factors in a comprehensive but parsimonious way, and conduct the path analysis that previous research has not done much. By analyzing the data from the Pew Internet and American Life Project, this study aims to empirically examine the causal impact of key determinants (i.e., a set of personal attributes, perceived use value, trust in government, and Internet use intensity) on the variation in e-government use, and also takes a further step toward testing the influence of socio- and techno-psychological characteristics on e-government use. The paper will offer academic and practical implications, beyond confirming the already-demonstrated profiles of e-citizens. The paper is structured into six sections, including the foregoing introduction. Section 2 reviews the literature, and then establishes the research framework and hypotheses. Section 3 details measurements and method. Section 4 presents results of the path analysis. Section 5 discusses policy implications and research limitations. The final section addresses concluding remarks.

2. THEORETICAL AND EMPIRICAL BACKDROPS 2.1 Multidimensional Issues of the Digital Divide The digital divide is a complex, dynamic, and multifaceted concept [10]. It captures the gap, separation, distinction, disparity, or gulf between haves and have-nots in terms of various resources and competences related to ICTs, but its multidimensional, multifaceted nature denies a simple dichotomy between haves and have-nots, connoting a more complicated, complex social (nontechnical) phenomenon [24]. Access is fundamental and basic to the digital divide, and little else is possible without access. The nature of that access is not without ambiguity. The concept of access evolves into different and successive kinds of access to digital technologies: motivational access, physical access, skills access, and usage access [60]. Given such multiple dimensions of the digital divide, there are three crucial links between e-government and the digital divide. First, the usage access can be specified in the context of egovernment adoption. Inclusiveness in actual usage of egovernment provokes a higher concern in the multi-layered concept of the digital divide. Sipior and Ward’s [48] multidimensional perspective highlighted three facets of the digital divide: Internet access, computer skills, and e-government inclusion. There are more elaborated models of the e-government adoption divide. According to Gurstein [22], the term “access” extends from access to ICT infrastructure and resources to access to e-government (electronically enhanced service delivery and information dissemination) and e-governance (electronically enhanced decision-making process including varied stakeholders). Becker et al. [6] presented the e-inclusion gap model. E-inclusion denotes participation for all in the digital, knowledge-based information society. The hierarchical model of e-inclusion gap stratifies different levels: the gap between the total population and Internet users, the gap between Internet users and e-commerce users, the gap between e-commerce users and e-government information users, and the gap between e-government information users and e-government transactional service users. A broader context of the digital divide embraces the effective use of electronic means [22], which refers to “the capacity and opportunity to successfully integrate ICTs into the accomplishment of self or collaboratively identified goals.” The effective use contextualized in the citizen-government relationship means adoption of a variety of e-government tools and services. The second connection captures the impact of the digital divide on e-government. The digital divide determines the extent to which e-government benefits accrue and whether e-government initiatives result in success or failure [24]. E-government has not lived up to its possibilities and potentials [34]. Government agencies have a charge to make their information and services available to everyone [4]. However, the uneven distribution of computer access and skills biases the governments’ ability to make their online services equally accessible and beneficial. Adoption of e-government is limited to those who have access to the technology and possess the skills necessary to utilize eservices. Furthermore, there is an apparent distinction between potential use (by physical access haves) and actual/effective use of e-government services [22,59]. Last but not least, not merely does the digital divide hinder egovernment, but e-government also leads to a new level or dimension of the digital divide [4]. E-government itself represents technical innovation which certain members of society can be inevitably excluded from. Benefits from e-government may mobilize only the technically savvy while disenfranchising those

who have less experiences and technical know-how. As with all novel technological trials, some individuals will inevitably be left behind by the technology. McNeal et al. [34] found that egovernment appears to a double-edged sword, motivating egovernment use for some (the young and women) while magnifying existing gaps based on socioeconomic status (SES). Therefore, the new technological tools of e-government hold benefits only for some segments of the population.

2.2 Determinates of E-government Adoption 2.2.1 Personal Attributes A rich body of empirical studies strongly have confirmed the determining effects of demographic, socioeconomic, generational, and geographical––hereafter, socio-demographic––differences on the digital divide in access, skills, and e-government use [34,40,41,65]. The multiple layers of the digital divide are interrelated with each other. Non-users of e-government are hampered by the lack of internet access and technological efficacy [31]. Technologically less skillful individuals who are older, Latino, African American, less educated, or less affluent are less likely to take part in online government interactions [40]. The rapid uptake of new technologies is occurring among most groups of the American population in large, regardless of those personal attributes [1], but noticeable gaps still exist between different levels of those attributes. The skeptic finding of little change in the socio-demographic context of e-government adopters is prevalent in empirical investigation profiling their sociodemographic composition [34]. The explanatory power of socio-demographic conditions does not only come from a cumulative set of evidence but also from a solid theoretical background. Innovation diffusion theory strongly bolsters the decisive impact of socio-demographic attributes on innovation diffusion. As innovation refers to “an idea, practice, or object that is perceived as new by an individual or other unit of adoption” [47], using e-government services is taken as “a new practice and can be seen as an innovation for each individual Internet user” [18]. Rogers’ [47] model specifies the hierarchy of innovation adopters: innovators, early adopters, early majority, late majority, and laggards. Dimitrova and Chen [18] viewed citizens who already use e-government services as early adopters. Early adopters of any innovation share common characteristics: young, well educated, and higher incomes [47]. A wide array of empirical findings reveal that those more likely to use e-government functions (e.g., transaction, information, contact channel, and e-participation) include younger citizens, citizens with higher levels of income and education, and citizens who frequently use the Internet for other tasks [1,2,43-4,46,49,53,63]. In Becker et al.’s [6] study, even technology-advanced countries in Western Europe show e-government exclusiveness of seniors citizens, less educated people, and residents in thinly populated regions. Not all but some of empirical studies found the egovernment usage stratified by gender and ethnicity [4,7,20,21,32]. A piece of evidence has shown that Republicans are more likely to be interested in e-government than Democrats [53]. Thomas and Streib [52] claimed that the digital divide is even more pronounced among government Web site visitors than among Internet users in general, so the e-government use divide is wider than the access divide. Although the continuing diffusion of computers and Internet access across the population may be eroding the basic digital divide, the digital divide for government Web site visitors persists.

There is a paradox surrounding the e-government use divide. The neediest, citizens who are the primary users of government services, are least likely to access the Internet, thereby getting poorly connected to electronic transformation of those services. The socioeconomically disadvantaged status increases the risk that households will be disconnected from the Internet [19], and the benefits that could accrue from e-government are at risk. Since the neediest often depend upon public services for their daily needs, their least use of e-government services delivered in a more efficient and convenient way is an unpleasant fact to government and society [40].

2.2.2 Trust in Government Socio-demographic effects on the digital divide do not explain all. A set of psychological factors such as trust and perception strongly matter for the gap in e-government use. The egovernment use is associated with increased trust [34]. Trust in government is a nebulous and contested notion [8,28]. Trust generally exists when one party has confidence in another’s reliability and integrity [38]. Miller and Listhaug [35] views trust in government as the “judgment of the citizenry that the system and the political incumbents are responsive, and will do what is right even in the absence of constant scrutiny.” Trust in government is also “a central indicator of the public’s underlying feeling about its polity” [42]. A group of studies considered the relationship between trust and e-government. Since two conflicting conclusions exist, the causal relationship between e-government use and trust in government is controversial [34]. Some studies found that higher levels of trust in government are correlated with more intensive e-service use [3,13-4,46,55,63,66]. Tolbert and Mossberger [54] found that trust was positively related to visiting Web sites at the local government level. Others showed the absence of significant relationship between citizens’ usage of e-government and trust in government [51,56,64]. Most empirical studies considered trust in government as an outcome variable [15,29,54-5,63,64]. By contrast, Goldfinch et al. [21] revealed that trust in government causes the greater use of e-government measures.

2.2.3 Perceived Usefulness The consideration of perceived usefulness draws upon Davis’s [17] technology acceptance model. The model goes beyond the demographic classification of adopters, elucidating psychological predispositions (perceived usefulness and perceived ease of use). Perceived usefulness denotes “the extent to which a person believes that using the system will enhance his or her job performance” [61]. In the e-government context, many studies have corroborated the influence of perceived usefulness on the intention and intensity of e-government usage. The term “perceived usefulness” is interchangeably used with perceived benefits and perceived use value in the prior literature. Perceived usefulness, benefit, or use value is psychologically based on utilitarianism of technology use. E-government users define use value in terms of information and service functionality [29]. Perceived value in e-government service significantly raises usage intentions [12,36]. In general the perceived benefit of technological innovation positively influences e-government adoption decisions [9]. Perceived usefulness determines the degree of actual e-government adoption [14,25-6,30,45,49,57,62]. Experienced users of e-government expectedly perceive greater value than less frequent users, suggesting that user efficacy influences value perceptions, but

non-users of e-government may also perceive potential value in egovernment [29].

2.3 Research Model and Hypotheses Drawing on theoretical and empirical discussions up to this point, this study constructs the research model sketched in Figure 1. According to the previous literature, different levels of the digital divide are mutually related––for example, the access divide influences the e-government use divide to a substantial extent. In this sense, e-government use intensity moves with Internet use intensity to some degree. Personal attributes including sociodemographic profiles and political affiliation influence egovernment use. The following path diagram shows the influence is exerted in a complicated way. Socio-demographic conditions determine Internet use intensity, and then general Internet usage affects perceived usefulness of e-government. The degree of perceived usefulness ultimately influences e-government use. Another causal direction flows from trust in government to perceived use value then to actual e-government usage. The path model hypothesizes four causal relationships. Figure 1. The Path Model H1

Personal Attributes

Internet Use Intensity

H4 E-Government Use

Trust in Government

Perceived Value of E-Government Use

H3

H2

Causal Correlated

Hypothesis 1. The set of personal attributes has a significant impact on the degree of e-government use. Hypothesis 2. Trust in government increases the degree of e-government use. Hypothesis 3. Perceived value of e-government use increases the degree of e-government use. Hypothesis 4. Internet use intensity increases the degree of e-government use.

3. MEASUREMENTS AND EMPIRICAL STRATEGY This study analyzes the publicly-available data (entitled 2009 Government Online) from the national random-sampled survey that the Pew Internet and American Life Project conducted via telephone interviews on American citizens during December in 2009. The data for the analysis (N = 1,199) is extracted from the original cross-sectional dataset (N = 2,258) by keeping only relevant variables and excluding missing observations in those variables. Table 1 exhibits the demographic distribution of the sample. This study considers personal attributes and political partisanship as a set of explanatory variables. Those personal backgrounds include age, sex, race, education (school attainment in years), annual household income, and residential area. Table 2 describes measures in detail.

Table 1. The Demographic Composition of the Sample N = 1,199

Categories

Generation Y (born after 1976) 24% Age M = 51, S.D = 18 Generation X (born between 1965 and 1976) 18% Min=18, Max=95 Baby Boomers (born between 1946 and 1964) 39% Dutifuls (born before 1946) 19% Male 45% Sex Female 55% Caucasian 80% Race Non-Caucasian 20% High school incomplete 5% Education High school graduate 23% Some college level 29% Four-year college graduate 25% Post-graduate education 18% $30,000 or less 24% Annual $30,001 to $50,000 22% household $50,001 to $75,000 18% income $75,001 to $100,000 15% $100,001 or more 21% Rural 21% Residential Suburban 52% place Urban 27% Democrat 37% Self-reported Republican 26% Partisanship Independent or others 37% Source: www.PewInternet.org/Shared-Content/Data-Sets/2009/December2009--Government-Online.aspx

To test the hypotheses, this study relies on the path analysis. Ordinary least squares (OLS) regressions for constructing the hypothesized path model are run, using the indexed variables of egovernment use and e-government value. Standardized coefficients estimated in the OLS regressions are presented as path coefficients in terms of direct, indirect, and total causal effects of hypothesized relationships (Figure 2). Since this study focuses on causal impacts, the estimation of non-causal correlated effects is omitted in the presentation of the path model. The validity of the research model needs to be tested. The likelihood ratio test does not affirm the structural difference between the just-identified model (including all possible causal paths) and the over-identified model (excluding some possible causal paths). The path model in Figure 1 does not include causal flows from personal attributes to perceived use value and from trust in government to Web use intensity because those relationships are not grounded on the literature. The Chi-square statistics (χ2 = 3.15, p = 0.52, d.f = 2), given the degree of freedom as the number of restrictions (two excluded possible causal paths in the path diagram), supports that the exclusion of possible causal relationships does not make any significant distinction from the full model.

4. WHO USES E-GOVERNMENT? This section examines who is more likely to be a heavy or light user of e-government. The correlation as the pre-regression evidence merits attention. Different types of e-government use and perceived value variables are associated with each other in a moderate or high level, but trust in government is not correlated with adoption and use value of e-government.

Table 2. Variable Descriptions and Univariate Statistics E-government use (M=5.09, S.D=4.09, Min=0, Max=20) Summating the four categories (Cronbach α = 0.75) • Transactional service (M=1.46, S.D=1.22, Min=0, Max=5) Summating the following binary responses (Cronbach α = 0.67) 1) Renewing a driver’s license or auto registration (32%), 2) Applying for a fishing, hunting or recreational license (11%), 3) Paying a fine such as a parking ticket (12%), 4) Downloading government forms (44%), and 5) Looking up what services a government agency provides (47%) • Information (M=2.60, S.D=2.35, Min=0, Max=10) Summating the following binary responses (Cronbach α = 0.75) 1) Information about a public policy or issue (49%), 2) Advice or information about a health or safety issue (26%), 3) Recreational or tourist information (33%), 4) Official government documents or statistics (37%), 5) Information about benefits (21%), 6) Information about how to apply for a government job (16%), 7) Government data on data.gov or usaspending.gov (17%), 8) Information about the campaigns of elected officials (14%), and 9) Text of any legislation (24%) • Participation (M=0.33, S.D=0.70, Min=0, Max=4) Summating the following binary responses (Cronbach α = 0.67) 1) Participating in an online town hall meeting (3%), 2) Posting comments, queries or information in a blog, online discussion, or online forum about public issue (10%), 3) Uploading photos or videos online about public issue (7%), and 4) Joining a group online that tries to influence policies (12%) • Government 2.0 tools (M=0.71, S.D=0.08, Min=0, Max=5) Summating the following binary responses (Cronbach α = 0.66) 1) Becoming a fan of a government agency on its social networking site (9%), 2) Following a government agency or official on Twitter (7%), 3) Reading the blog of a government agency or official (15%), 4) Receiving email alerts from a government agency (15%), and 5) Receiving text messages from a government agency/official (4%) Trust in government (M=2.26, S.D=0.66, Min=1, Max=4) Averaging the following three items (Cronbach α = 0.77) • Trust in federal government (M=2.15, S.D=0.73) 1) Never (17%), 2) Some of the time (55%), 3) Most of the time (25%), and 4) Just about always (3%) • Trust in state government (M=2.27, S.D=0.73) 1) Never (13%), 2) Some of the time (51%), 3) Most of the time (32%), and 4) Just about always (4%) • Trust in local government (M=2.40, S.D=0.76) 1) Never (11%), 2) Some of the time (45%), 3) Most of the time (38%), and 4) Just about always (6%) Perceived use value (M=10.49, S.D=1.98, Min=3, Max=12) Summing the responses to the following statements (Cronbach α = 0.71) • E-gov provides information to the public (M=3.66, S.D=0.66) 1) Very important (2%), 2) Somewhat important (4%), 3) Not too important (20%), and 4) Not important at all (74%) • E-gov allows people to complete tasks (M=3.57, S.D=0.75) 1) Very important (4%), 2) Somewhat important (4%), 3) Not too important (22%), and 4) Not important at all (70%) • E-gov allows people to contact officials (M=3.61, S.D=0.70) 1) Very important (3%), 2) Somewhat important (4%), 3) Not too important (23%), and 4) Not important at all (70%) Web use intensity (M=5.43, S.D=1.72, Min=1, Max=7) 1) Never (5%), 2) Less often (4%), 3) Every few weeks (4%), 4) 1-2 days a week (13%), 5) 3-5 days a week (15%), 6) About once a day (22%), and 7) Several times a day (37%) Note 1. The original questions about e-government use asked about experiences during the recent year (Have you ever done ---- during the recent year?). The question about trust in government was “How much of the time can you trust ----?”. Note 2. A reliability test was carried out using Cronbach’s alpha (α), which measures the internal consistency of research constructs. The recommended minimum acceptable limit of reliability alpha for exploratory study is 0.60 [16,23].

Table 5. Estimation of Causal Effects

Table 3. Correlation Matrix Perceived value Trust in government Web use intensity Age Male Caucasian Education Income Urban Suburban Democrat Republican * p < 0.05

E-gov use 0.35* 0.05* 0.36* -0.13* 0.06* 0.01* 0.31* 0.25* 0.06* 0.03* 0.04 0.01

Perceived value

Trust

Web use

0.11* 0.17* -0.18* -0.07* -0.02 0.15* 0.06* 0.04* -0.04* 0.09* -0.03

0.05* -0.07* 0.01* 0.01 0.10* 0.02 0.01 -0.04 0.13* -0.04*

-0.14* 0.04* 0.07* 0.23* 0.23* 0.02* 0.09* 0.00 -0.01

Table 4. OLS Regressions Constructing the Path Model N = 1,199 Perceived value Trust in Government Web use intensity Personal attributes Age Male Caucasian Education in years Income in $1,000 Urban Suburban Rural (reference) Political affiliation Democrat Republican Independent (reference) Constant

E-gov use b Beta 0.674*** 0.264 (0.077) -0.116 -0.018 (0.185) 0.553*** 0.226 (0.075)

Web use b Beta

0.313*** 0.130 (0.063) 0.178 0.191 (0.027)

-0.024** (0.008) 0.577* (0.236) -0.088 (0.311) 0.512*** (0.082) 0.181** (0.059) 0.597† (0.323) 0.361 (0.290) Omitted

-0.091 -0.020*** (0.003) 0.071 0.137 (0.102) -0.008 0.381* (0.150) 0.194 0.197*** (0.037) 0.098 0.114*** (0.027) 0.066 0.344* (0.149) 0.045 0.399** (0.134) Omitted

0.449 (0.279) 0.309 (0.293) Omitted

0.054

-7.343*** (1.073)

0.034

Perceived value b Beta

-0.190 0.041

0.181 0.150 0.092 0.120

0.061 0.018 (0.115) -0.153 -0.040 (0.129) Omitted

F (17,1181) 33.73*** 14.21*** Adjusted R2 0.2809 0.1523 *** p < 0.001, ** p < 0.01, * p < 0.05, † p < 0.10 Robust standard errors in parentheses Note. Beta is full-standardized coefficient.

Hypothesis 1. The set of personal attributes has a significant impact on the degree of e-government use. The result of OLS regressions demonstrates the determining effects of some personal attributes on e-government use. The path model posits that the causal impact of personal backgrounds is both direct and indirect through Web use intensity, as illustrated in Figure 2. Frequent Internet users would likely belong to the demographic categories of the younger, Caucasians, the bettereducated, the more affluent, and urban or suburban dwellers. The generational effect is expectedly obvious in both direct and indirect effects. The older are less likely to use e-government services. E-government use also shows a significant gender gap. Men use more e-government services than women. The racial gap between Caucasians and non-Caucasians is minute, compared to the gender gap, in the magnitude in the standardized coefficient, and it’s not significantly predicted. The effect of education and income as measures of SES is significantly strong. There is a large gap between the least affluent/educated and the most affluent/educated in the degree of using e-government services. The potential benefits of e-government indeed bypass those in lower SES [40]. Since the less educated and affluent often depend upon public services for their daily needs, their lower use of egovernment is unavoidably disconcerting. E-government services are least likely to be delivered to individuals falling into a lower ladder in “ascriptive hierarchy,” [50] which represents historically institutionalized and structuralized participatory inequality.

0.086

4.404*** (0.289)

Direct effect Indirect effect Total effect Perceived value 0.264* – 0.264* Trust in government -0.018 0.034* 0.016 Web use intensity 0.226* – 0.226* Age -0.091* -0.044* -0.135* Male 0.071* 0.011* 0.082* Caucasian -0.008 -0.022* -0.030 Education 0.194* 0.048* 0.242* Income 0.098* 0.040* 0.138* Urban 0.066 0.025* 0.091 Suburban 0.045 0.032* 0.077 Democrat 0.054 0.005 0.059 Republican 0.034 -0.011 0.023 * p < 0.05 Note. As sketched in Figure 2, the indirect effects of personal attributes on the outcome variable are exerted through Internet use and through both Internet use and perceived value of e-government use. The effects are calculated by [(βWebUse * βH4) + (βWebUse * 0.151 * βH3)]. Total effects are the sum of direct and indirect effects.

9.158*** (0.220) 35.01*** 0.0872

On the other hand, urban residents use more e-government services, compared to rural dwellers. There are two possible reasons for the fact. First, the service-needy can be located in urban areas more than in rural areas. Second, the urban governments provide citizens with more diverse e-government services. In addition, the effect of political partisanship on e-government usage is examined. The regression result does not validate Tolbert and McNeal’s [53] previous finding that individuals who selfidentify as Republicans involve more in public life as e-citizens (e.g., political/civic activities and interactions with e-government for service, information, and transaction).

Hypothesis 2. Trust in government increases the degree of e-government use. The effect of trust in government on e-government use is hypothesized to be direct and/or indirect through perceived use value. Only the indirect effect is significantly predicted, and thus

the intermediated effect of perceived use value on the causality of trust in government to e-government use is significant. Those with higher levels of trust in government are likely to perceive benefits from using e-government, and then those who perceive use value of e-government actually adopt e-government. Hence

decomposing the effect of trust into the direct and indirect ones challenges existing findings of the considerable association between trust and e-government use. The nature of the association could be indirect through perceived value.

strengthen social inclusion and counter the emergence and deepening of socioeconomic divides [19]. E-government cannot become universal yet, and thus other channels must not be closed off or contracted and have to be kept open [33,52].

Hypothesis 3. Perceived value of e-government use increases the degree of e-government use.

5.2 Research Limitations and Directions

Those who perceive e-government value are likely to be heavy users of e-government. Perceived use value significantly predicts the difference between heavy and light users of e-government services. Charted in Figure 2, perceived use value does not significantly intermediate the effect of Internet use intensity on egovernment use because Internet use intensity does not determine the level of perceived usefulness of e-government. Hypothesis 4. Internet use intensity increases the degree of e-government use. The result of the linear estimation demonstrates that the degree of e-government adoption proportionately rises with the frequency of using the Internet in general. However, Internet use intensity does not lead to higher levels of perception on usefulness of egovernment. Thus, the impact of Internet use on e-government adoption is more direct rather than intermediated through perceived value of e-government use.

5. FURTHER DISCUSSIONS 5.1 Practical Implications Governments have cared the digital divide, and social and national initiatives have contributed to closing the digital divide. This paper addresses a deeper issue beyond the gradually closing divide in access: Who lags behind benefiting from e-government use? Based on the findings from the path analysis, this study suggests two major orientations for governments. First, governments need to educate citizens (especially, the service-needy, the technology-illiterate, and the socioeconomically disadvantaged) about the value of egovernment so that citizens keep aware of usefulness of egovernment [27]. Those who are more aware of and comfortable with e-government will be more likely to use e-government. Besides, the path model analysis affirmed that perceived use value intermediates the causal effect of trust in government on egovernment usage intensity. Since citizens clearly distinguish and dichotomize their trust in e-government between institutional trust in government and process trust in the Internet channel, governments should simultaneously consider the two components underlying trust in e-government [5,55]. Trust in overall government affects perceived value of e-government use, which ultimately determines the degree of e-government usage. Second, governments should keep service channels for segments of the population underserved by Internet connectivity and least likely to be online. The e-government use divide is taken as being a particular problem for governments who cannot choose their customers [1]. Many public services are provided specifically for vulnerable or low-income groups who are least likely to have access to the technology and skills for using the technology. To extent to which they are the biggest users of government services, governments should continue providing services through multiple channels at least in the short term to prevent excluding those who do not have access to the Internet, but ultimately in the long term making efforts to connect the unconnected and wire the unwired. Governments should integrate ICTs into communities in ways that

Empirical research on the divide in e-government adoption and usage needs to construct a more comprehensive albeit parsimonious model capturing the relationships among key factors. Future studies will be able to further develop this research with three directions of improvement. First, the model could be elaborated in a more complicated manner. This study does not include all possible factors that might be relevant for the path model; for example, other theoretical constructs from the technology acceptance model (perceived needs), trust in technology, and civic mindedness and attitude. The model will be advanced by adding other attitudinal and psychological measures. Second, in addition to elaborating the model, sophisticating measurements will clarify the relationships among variables. Current surveys offer multiple items to be collapsed into a single indicator. Further research can establish a solid model by deriving common theoretical factors from a variety of psychological, attitudinal, and behavioral measures. Third, the models and hypotheses in prior empirical research face a high level of complexity, especially when endogenous and exogenous relationships are identified. The causal arrows in the path model of this study tap on previous findings. However some discussions may support different directions of the arrows or simultaneous causalities. If a majority of findings gained by new empirical investigation overturn the hypothesized model, the model should be modified reflecting changes in theory-based arguments.

6. CONCLUDING REMARKS The presence of the e-government use divide may constrain the workings of today’s government to only some segments of the whole population. Government’s imperative to mitigate the skeptic reality in this digital age seems quite a tough job, given the fact that e-government further cements the marginalization of already marginalized groups in society. An irony is that egovernment to envision e-inclusion might empower those already powerful [21]. However, leverages of personal attributes and background (in particular, socioeconomic profiles) on egovernment use are not impossible to be weakened, when governments consider the lessons from this study. Psychological factors affect the degree of e-government use. Perceived usefulness or perceived value of e-government is key to boosting e-government adoption across various segments of the population. Trust in government also influences e-government adoption through its effect on perceived use value. Government needs to keep doing efforts to close the divide in general use of the Internet because of the great association between Internet use intensity and e-government use intensity.

7. REFERENCES [1] Akman, İ., Yazici, A., Mishra, A., & Arifoglu, A. (2005). EGovernment: A global view and an empirical evaluation of some attributes of citizens. Government Information Quarterly, 22(2), 239-57. [2] Bélanger, F., & Carter, L. (2006). The effects of the digital divide on e-government: An empirical evaluation. In

Proceedings of the 39th Hawaii International Conference on System Sciences (HICSS-39, Kauai, Hawaii, Jan 4-7). [3] Bélanger, F., & Carter, L. (2008). Trust and risk in egovernment adoption. Journal of Strategic Information Systems, 17(2), 165-76. [4] Bélanger, F., & Carter, L. (2009). The impact of the digital divide on e-government use. Communications of the ACM, 52(4), 132-5. [5] Bart, Y., Shankar, V., Sultan, F., & Urban, G. L. (2005). Are the drivers and role of online trust the same for all web sites and consumers? A large-scale exploratory empirical study. Journal of Marketing, 69(4), 133-52. [6] Becker, J., Niehaves, B., Bergener, P., & Räckers, M. (2008). Digital divide in eGovernment: The eInclusion gap model. In M. A. Wimmer, H. J. Scholl & E. Ferro (Eds.), Electronic Government: Proceedings of the 7th International Conference, EGOV 2008 (Turin, Italy, Aug 31 - Sep 5) (Vol. 5184, pp. 231-42). Berlin/Heidelberg, Germany: Springer. [7] Bimber, B. (1999). The Internet and citizen communication with government: Does the medium matter? Political Communication, 16(4), 409-28. [8] Blind, P. K. (2007). Building trust in government in the twenty-first century: Review of literature and emerging issues. In Proceedings of the 7th Global Forum on Reinventing Government: Building Trust in Government (Vienna, Austria, Jun 26-29). Available at http://unpan1.un.org/intradoc/groups/public/documents/un/un pan025062.pdf. [9] Bretschneider, S., Gant, J., & Ahn, M. (2003, Oct 9-11). A general model of e-government adoption and diffusion. In Proceedings of the 7th Public Management Research Conference, Georgetown University, Washington, D.C.

and information channels. Social Science Computer Review, 24(2), 171-88. [19] Dugdale, A., Daly, A., Papandrea, F., & Maley, M. (2005). Accessing e-government: Challenges for citizens and organizations. International Review of Administrative Sciences, 71(1), 109-18. [20] Edmiston, K. D. (2003). State and local e-government: Prospects and challenges. The American Review of Public Administration, 33(1), 20-45. [21] Goldfinch, S., Gauld, R., & Herbison, P. (2009). The participation divide? Political participation, trust in government, and e-government in Australia and New Zealand. Australian Journal of Public Administration, 68(3), 333-50. [22] Gurstein, M. (2003). Effective use: A community informatics strategy beyond the digital divide. First Monday, 8(12). Available at http://131.93.53.231/www/issues/issue8_12/gurstein/index.ht ml. [23] Hair, J., Anderson, R., Tatham, R., & Black, W. (1998). Multivariate Data Analysis (5th ed.). Upper Saddle River, NJ: Prentice-Hall. [24] Helbig, N., Gil-García, J. R., & Ferro, E. (2009). Understanding the complexity of electronic government: Implications from the digital divide literature. Government Information Quarterly, 26(1), 89-97. [25] Horst, M., Kuttschreuter, M., & Gutteling, J. M. (2007). Perceived usefulness, personal experiences, risk perception and trust as determinants of adoption of e-government services in The Netherlands. Computers in Human Behavior, 23(4), 1838-52.

[10] Bruno, G., Esposito, E., Genovese, A., & Gwebu, K. L. (2011). A critical analysis of current indexes for digital divide measurement. The Information Society, 27(1), 16-28.

[26] Huang, W., D’Ambra, J., & Bhalla, V. (2002). An empirical investigation of the adoption of e-government in Australian citizens: Some unexpected research findings. Journal of Computer Information Systems, 43(1), 15-22.

[11] Carter, L. (2008). E-government diffusion: A comparison of adoption constructs. Transforming Government: People, Process and Policy, 2(3), 147-61.

[27] Jaeger, P. T., & Thompson, K. M. (2003). E-government around the world: Lessons, challenges, and future directions. Government Information Quarterly, 20(4), 389-94.

[12] Carter, L., & Bélanger, F. (2005a). The influence of perceived characteristics of innovating on e-government adoption. The Electronic Journal of e-Government, 2(1), 1120.

[28] Job, J. (2005). How is trust in government created? It begins at home, but ends in the parliament. Australian Review of Public Affairs, 6(1), 1-23.

[13] Carter, L., & Bélanger, F. (2005b). The utilisation of egovernment services: Citizen trust, innovation and acceptance factors. Information Systems Journal, 15(1), 5-25. [14] Carter, L., & Weerakkody, V. (2008). E-government adoption: A cultural comparison. Information Systems Frontier, 10(4), 473-82. [15] Cohen, J. E. (2006). Citizen satisfaction with contacting government on the Internet. Information Polity, 11(1), 51-65.

[29] Kolsaker, A., & Lee-Kelley, L. (2008). Citizens’ attitudes towards e-government and e-governance: A UK study. International Journal of Public Sector Management, 21(7), 723-38. [30] Kumar, V., Mukerji, B., Butt, I., & Persaud, A. (2007). Factors for successful e-government adoption: A conceptual framework. The Electronic Journal of e-Government, 5(1), 63-76.

[16] Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16, 297-334.

[31] Lee-Kelley, L., & James, T. (2003). E-government and social exclusion: An empirical study. Journal of Electronic Commerce in Organizations, 1(4), 1-16.

[17] Davis, F. D. (1989). Perceived usefulness, perceived ease of use and user acceptance of information technology. MIS Quarterly, 13(3), 319-40.

[32] Losh, S. C. (2003). Gender and educational digital chasms in computer and Internet access and use over time: 1982-2000. IT and Society, 1(4), 73-86.

[18] Dimitrova, D. V., & Chen, Y.-C. (2006). Profiling the adopters of e-government information and services: The influence of psychological characteristics, civic mindedness,

[33] Margetts, H. (2006). E-government in Britain: A decade on. Parliamentary Affairs, 59(2), 250-65.

[34] McNeal, R. S., Hale, K., & Dotterweich, L. (2008). Citizengovernment interaction and the Internet: Expectations and accomplishments in contact, quality, and trust. Journal of Information Technology & Politics, 5(2), 213-29.

[48] Sipior, J. C., & Ward, B. T. (2005). Bridging the digital divide for e-government inclusion: A United States case study. The Electronic Journal of e-Government, 3(3), 13746.

[35] Miller, A. H., & Listhaug, O. (1990). Political parties and confidence in government: A comparison of Norway, Sweden and the United States. British Journal of Political Science, 20(3), 357-86.

[49] Sipior, J. C., Ward, B. T., & Connolly, R. (Forthcoming). The digital divide and t-government in the United States: Using the technology acceptance model to understand usage. European Journal of Information Systems. DOI:10.1057/ejis.2010.64.

[36] Mills, A., Carter, L., & Belanger, F. (2010). Conceptualizing public service value in e-government services. In Proceedings of the Americas Conference on Information Systems (Lima, Peru, Aug 12-5).

[50] Smith, R. M. (1993). Beyond Tocqueville, Myrdal, and Hartz: The multiple traditions of America. American Political Science Review, 87(3), 549-66.

[37] Moon, M. J. (2002). The evolution of e-government among municipalities: Rhetoric or reality? Public Administration Review, 62(4), 424-33.

[51] Sweeney, A. D. P. (2007). Electronic government-citizen relationships exploring citizen perspectives. Journal of Information Technology & Politics, 4(2), 101-16.

[38] Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing. Journal of Marketing, 58(3), 20-38.

[52] Thomas, J. C., & Streib, G. (2003). The new face of government: Citizen-initiated contacts in the era of egovernment. Journal of Public Administration Research and Theory, 13(1), 83-102.

[39] Mossberger, K., Tolbert, C. J., & McNeal, R. S. (2008). Digital Citizenship: The Internet, Society, and Participation. Cambridge, MA: The MIT Press. [40] Mossberger, K., Tolbert, C. J., & Stansbury, M. (2003). Virtual Inequality: Beyond the Digital Divide. Washington D.C.: Georgetown University Press. [41] Neu, C. R., Anderson, R. H., & Bikson, T. K. (1999). Sending Your Government a Message: E-mail Communication Between Citizens and Government. Santa Monica: Rand Corp. [42] Newton, K., & Norris, P. (2000). Confidence in public institutions: Faith, culture, or performance? In S. J. Pharr & R. D. Putnam (Eds.), Disaffected Democracies: What’s Troubling the Trilateral Democracies? (pp. 52-73). Princeton, NJ: Princeton University Press. [43] Niehaves, B., & Becker, J. (2008). The age-divide in egovernment – data, interpretations, theory fragments. In M. Oya, R. Uda & C. Yasunobu (Eds.), Towards Sustainable Society on Ubiquitous Networks: Proceedings of the 8th IFIP Conference on e-Business, e-Services, and e-Society (I3E 2008, Sep 24-6, 2008, Tokyo, Japan) (Vol. 286, pp. 279-87). Boston: Springer. [44] Niehaves, B., & Plattfaut, R. (2010). What is the issue with Internet acceptance among elderly citizens? Theory development and policy recommendations for inclusive egovernment. In M. A. Wimmer, J.-L. Chappelet, M. Janssen & H. J. Scholl (Eds.), EGOV 2010: Proceedings of the 9th IFIP WG 8.5 international conference on Electronic government (pp. 275-88). Berlin/Heidelberg: SpringerVerlag. [45] Phang, C. W., Sutanto, J., Kankanhalli, A., Li, Y., Tan, B. C. Y., & Teo, H.-H. (2006). Senior citizens' acceptance of information systems: A study in the context of e-government services. IEEE Transactions on Engineering Management, 53(4), 555-69. [46] Reddick, C. G. (2004). Citizen interaction with egovernment: From the streets to servers? Government Information Quarterly, 22(1), 38-57. [47] Rogers, E. M. (2003). Diffusion of Innovations (5th ed.). New York: Free Press.

[53] Tolbert, C. J., & McNeal, R. S. (2003). Unraveling the effects of the Internet on political participation? Political Research Quarterly, 56(2), 175-85. [54] Tolbert, C. J., & Mossberger, K. (2003). The effects of egovernment on trust and confidence in government. In Proceedings of the Annual National Conference on Digital Government Research (dg.o 2003, Boston, May 18-21). [55] Tolbert, C. J., & Mossberger, K. (2006). The effects of egovernment on trust and confidence in government. Public Administration Review, 66(3), 354-69. [56] Torres, L., Pina, V., & Acerete, B. (2005). E-government developments on delivering public services among EU cities. Government Information Quarterly, 22(2), 217-38. [57] Tung, L. L., & Rieck, O. (2005). Adoption of electronic government services among business in Singapore. Journal of Strategic Information Systems, 14(4), 417-40. [58] United Nations. (2002). Benchmarking E-government: A Global Perspective. New York: American Society for Public Administration. [59] van Deursen, A., van Dijk, J., & Ebbers, W. (2006). Why egovernment usage lags behind: Explaining the gap between potential and actual usage of electronic public services in the Netherlands. In M. A. Wimmer (Ed.), Electronic Government: Proceedings of the 5th International Conference, EGOV 2006 (Krakow, Poland, Sep 4-8, 2006) (Vol. 4084, pp. 269-80). Berlin/Heidelberg, Germany: Springer. [60] van Dijk, J. A. G. M. (2005). The Deepening Divide Inequality in the Information Society. Thousand Oaks, CA: Sage. [61] Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology acceptance model: Four longitudinal field studies. Management Science, 46(2), 186204. [62] Wangpipatwong, S., Chutimaskul, W., & Papasratorn, B. (2008). Understanding citizen’s continuance intention to use e-government website: A composite view of technology acceptance model and computer self-efficacy. The Electronic Journal of e-Government, 6(1), 55-64. Available at

http://www.ejeg.com/volume-6/vol6iss1/Wangpipatwong.pdf. [63] Welch, E. W., Hinnant, C. C., & Moon, M. J. (2005). Linking citizen satisfaction with e-government and trust in government. Journal of Public Administration Research and Theory, 15(3), 371-91. [64] West, D. M. (2004). E-government and the transformation of service delivery and citizen attitudes. Public Administration Review, 64(1), 15-27.

[65] Wilhelm, A. (2000). Democracy in the Digital Age: Challenges to Political Life in Cyberspace. New York: Routledge. [66] Yazici, S., & Cengiz, E. (2010). Towards an e-government performance evaluation model from the citizen’s perspective. In Proceedings of the International Conference eGovernment and eGovernance (ICEGEG 2010, Antalya, Turkey, Mar 112).