Government Information Quarterly xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Government Information Quarterly

journal homepage: www.elsevier.com/locate/govinf

Engendering inclusive e-government use through citizen IT training

programs
Jaebok Leea,1, Gregory A. Porumbescub,1,
⁎

a
Korea University, South Korea
b
Rutgers University – Newark, United States

ABSTRACT

Around the world, governments turn to information technology (IT) training programs to enhance equitable delivery of online public services to citizens. However,
the effectiveness of these citizen training programs has gone unexplored. This study is motivated by two objectives: 1) To evaluate whether citizen participation in
government training programs is associated with greater e-government use among participants, and 2) To assess whether the strength of this relationship varies
according to whether a citizen is elderly, disabled, or not – those who are elderly or disabled tend to use e-government the least. We use data from South Korean to
examine these objectives. Findings indicate citizen participation in government IT training programs is positively associated with e-government use and that this
relationship is stronger for citizens who are elderly or disabled. These findings highlight the potential of government IT training programs to mitigate the digital
divide. Implications for research and practice are discussed.

1. Introduction e-government going unrealized in segments of the population that stand

to benefit from e-government the most.
Today, optimism surrounding the potential of e-government to To address the issue of user-failure and render online public service
transform relationships between citizens and government has given provision more inclusive, governments in many countries have laun-
way to concerns regarding implementation. Specifically, while the po- ched information technology (IT) training programs. These programs
tential of e-government to improve the responsiveness of public service are open to everyone, but often target vulnerable groups of citizens –
provision is well established, there is growing evidence that it can also those who are older or disabled - and draw on a standardized curri-
engender new forms of inequality and social exclusion (Hsieh et al., culum to teach citizens how to use computers to navigate the Internet
2008; Dunleavy et al., 2006). In turn, governments are now wrestling and access services online (Kim, Cho, & Suh, 2017). Despite the pro-
with the question of how to implement e-government applications in liferation of government IT training programs, there have been no at-
ways that are inclusive. tempts to evaluate their relationship with citizens' use of e-government
When discussing equitable citizen access to e-government and in services. As such, we have no evidence that these policies are working.
particular public services provided online, a major concern pertains to This gap in the literature is problematic in that it points to a very
‘user failure’ – the possibility that clients simply do not use the system limited understanding of steps governments can take to ensure inclusive
(Gauld, Goldfinch & Horsburgh, 2010). These concerns are not without access to public services in this era of digital government (Dunleavy,
merit in that, despite the growing availability of public services online, Margetts, Bastow, and Tinkler 2006). Put differently, understanding the
vulnerable groups of citizens often do not use services provided online extent to which participation in IT training programs is associated with
due to difficulties in navigating the Internet or e-government websites, e-government use among segments of the population who tend to use e-
even when they may know how to use the Internet itself (Gauld et al., government least constitutes a necessary first step in evaluating the
2010; Kvasny & Keil, 2006; Moon 2002). Indeed, a growing body of effectiveness of this prominent policy intending to promote equitable
evidence suggests that vulnerable groups of citizens who stand to access to public services in this age of digital governance.
benefit most from the flexible service delivery arrangements e-gov- To address this important gap in e-government research, we assess
ernment offers – those who are older or lack mobility – tend to use it the the following research question – is citizen participation in government IT
least (Loges & Jung, 2001; Piotrowksi, 2010; Okunola et al., 2017). This training programs associated with greater use of e-government services
lack of inclusiveness results in much of the transformative potential of among groups of citizens who typically use e-government the least? We

⁎
Corresponding author.
E-mail address: [email protected] (G.A. Porumbescu).
1
Author order is alphabetical. The authors contributed equally to this article.

https://doi.org/10.1016/j.giq.2018.11.007
Received 10 August 2018; Received in revised form 23 October 2018; Accepted 22 November 2018
0740-624X/ © 2018 Elsevier Inc. All rights reserved.

Please cite this article as: Lee, J., Government Information Quarterly, https://doi.org/10.1016/j.giq.2018.11.007
J. Lee, G.A. Porumbescu Government Information Quarterly xxx (xxxx) xxx–xxx

answer this question using citizen level data collected from South Korea international recognition, winning the best practice in electronic tax
(n = 8950). Drawing upon this data and insights offered by work on the administration award by the OECD in 2006 (MOI, 2016).
technology acceptance model (Davis 1989), we offer nuanced insight All told, the South Korean government has signaled a sustained
into the extent government IT training programs are effective in miti- commitment to using e-governments as a means of addressing the
gating the digital divide and fostering more inclusive access to public growing complexity of citizens' needs. For this reason, citizens' ability
services online. This is done by examining the strength of the re- to access and utilize technologies plays an ever critical role in de-
lationship between participation in government IT training programs termining the way citizens experience the State. In an effort to ensure
and e-government use across two groups of citizens that previous re- that no one is left behind in this era of digital governance, the South
search suggests use e-government the least – the elderly and the dis- Korean Government passed the ‘National Information Law,’ which was
abled. As is explained in greater detail in the following section, South enacted in 1996, and requires local governments to provide IT training
Korea is a relevant context to address this research question for the programs to the general population as well as to vulnerable segments of
reason that it is a world leader in e-government provision – it ranked the general population (Kim et al., 2017). Accordingly, the law defines
first among all UN member states for three consecutive UN Global E- several vulnerable segments of the general population, including the
Government Surveys in 2010, 2012 and 2014, and places in the top 5 in elderly and the disabled and establishes a regulatory body to ensure
the 2016 online service index (UN E-Government Survey, 2016). Fur- that local governments are providing IT programs that are accessible by
thermore, the South Korean government, in collaboration with local the general population as well as by vulnerable segments of the popu-
governments has been active in administering e-government training lation (#35 of the law). Therefore, to comply with the law local gov-
programs in hopes of boosting e-government use among groups of ci- ernments are obliged to promote these training programs among
tizens that use e-government the least (Kim, Cho, & Suh, 2017). These members of the community they serve. IT training programs are orga-
factors make the South Korean context well suited for evaluating the nized in local government buildings, welfare centers, community cen-
effectiveness of e-government training programs. ters, post office, school, or even for-profit institutions local govern-
Our findings reveal that citizen participation in government IT ments contract out with. The law also regulates curriculum that is
training programs is significantly associated with e-government use and offered to ensure consistency. The topics that training programs must
that the magnitude of this relationship is stronger for citizens who are cover include: 1) basics of computer or Internet use, 2) accessing gov-
elderly and disabled than it is for members of the general population. In ernment and private websites for purpose of obtaining information and
turn, a key implication of our findings is that government IT training services (#34–4 of the Act). All told, the curriculum includes modules
programs can play a vital role in combatting the digital divide and that walk participants through the different steps of how to use a
promote more equitable access to public services online in this era of computer (turn it on, connect to the Internet), navigate the Internet,
digital governance. and obtain services online. In general, courses last 20 h a month.
The remainder of this article is organized as follows. The next two
sections provide an overview of e-government in Korea and reviews 3. Literature review
research examining citizens' use of e-government and e-government
training programs. Following this, we describe the data, measures, and 3.1. Why do citizens use e-government?
methods used, and review the results of our analysis. We conclude with
a summary of the key findings and provide policy recommendations for The digital divide, in part, results from a heterogeneous distribution
future initiatives. of user failure, such that certain segments of the population system-
atically under consume digital goods and services. Historically, efforts
2. E-government in South Korea to explain determinants of the digital divide in general and citizens'
access to e-government more specifically focused on citizens' access to
South Korea is viewed as a world leader in e-government. Internet enabled devices and a stable Internet connection. In other
Furthermore, South Korea's role as a world leader in e-government is words, citizens used e-government because they had the tools to do so.
not limited to the Central government alone. Rather, studies also find To this end, efforts to mitigate the digital divide and promote more
that Seoul is a world leader in e-government when compared to major equitable access to e-government services online emphasized expanding
municipalities worldwide according to a number of international sur- citizens' access to the tools they need to access e-government services
veys, such as the Digital Governance in Municipalities Worldwide online, namely the Internet and Internet enabled devices. However,
Survey (Rutgers' E-Governance Institute, 2014). with declining costs of Internet subscriptions and Internet enabled
There are several examples of the South Korean Government's mobile devices, the issue of access is becoming less poignant. Indeed,
commitment to e-government. For example, Minwon24 (www.egov.go. today in most developed nations nearly all citizens now have access to
kr) is a one stop service portal used by the Ministry of Interior (MOI) to the Internet and internet-enabled devices that allow them to use e-
support civil petitions and provide public services. Minwon24 has government (Gonzales, 2016; Porumbescu & Im, 2015). All told, steady
grown in popularity, seeing the number of petitions filed increase from advances in technology have helped ensure more balanced access to
62.3 million to 64.4 million and the number of citizens using the portal Internet enabled devices and, as such e-government. At the same time,
to obtain public services increasing from 21.1 million to 50.1 million these developments highlight the importance of a second dimension of
between 2010 and 2013 (MGAHA, 2016: 35). While there is a general the digital divide, which is citizens' confidence that they are actually
trend of greater e-government use in South Korea as well as in many able to use technology to access e-government services (Porumbescu,
other areas in the world, the achievements of Minwon24 in bolstering e- 2016). Here, research on the technology acceptance model (TAM) is
government use have been acknowledged by the United Nations, which particularly useful in understanding how citizens' confidence in their
awarded the platform the UN Public service award in 2011 (MOI, ability to use technology is responsible for shaping their use of e-gov-
2016). Another example of the South Korean Government's commit- ernment.
ment to e-government excellence is the HomeTax portal (www. The TAM, which extends Fishbein and Ajzen's (1975) theory of
hometax.go.kr) launched by the National Tax Service Agency, which reasoned action, is an early and profound attempt to explain why in-
allows citizens to pay taxes, file tax returns, and obtain tax related dividuals might use technology to do things they could otherwise do
documents online, without visiting tax offices. Between 2010 and 2013, offline (Davis 1989). As such, the TAM is frequently used to explain
the number of portal users grew from 12.2 million to 22.3 million, with citizens' use of e-government (Shyu & Huang, 2011). In the TAM, per-
the number of electronic tax filings growing from 20.5 million to 22.3 ceived usefulness (PU), perceived ease of use (i.e., confidence in ability
million (MGAHA, 2016). Like Minwon24, this service has also received to use a technology) (PEU) and an implicit cost benefit analysis are

2
J. Lee, G.A. Porumbescu Government Information Quarterly xxx (xxxx) xxx–xxx

central to explaining why an individual will decide to accept and use 3.2. Hypotheses
certain technology (Davis 1989). PU posits that individual intentions to
use a particular technology grow in relation to perceived usefulness of As noted earlier, despite the proliferation of such programs in
the technology – the more useful it is perceived, the greater an in- Korea, as well as elsewhere in the world, there is little empirical work
dividual's intentions to use it. PEU qualifies the effects of PU, suggesting assessing whether citizen participation in government IT training pro-
that in addition to perceiving a new technology as useful, individuals grams are associated with e-government use. Therefore, while the
must also perceive it as easy to use – if not, then the perceived costs benefits of these policies are intuitive at a conceptual level, there is
associated with learning to use a new technology will outweigh the little evidence that speaks to their effectiveness.
perceived benefits and individuals will be less likely to use the tech- Drawing on insights provide by the TAM outlined above, we address
nology (Davis 1989: 321; Zheng and Schachter, 2016). Drawing on the this important gap in the e-government literature and predict a positive
TAM, research indicates that low levels of PEU corresponds to a lack of relationship between participation in government IT training programs
digital literacy, and therefore constitutes a primary determinant of the and e-government use – participation in IT training programs will
digital divide today (Molnár, 2003; Kim, Cho, & Suh, 2017). On the imbue participants with higher levels of digital literacy and greater
other hand, it is also important to note that studies show most segments confidence, which in turn will enhance perceived ease of use (PEU) of
of the population, irrespective of PEU levels and subsequent levels of online public services, thus resulting in greater uptake (Lee and Kim
digital literacy tend to positively evaluate the utility of IT and e-gov- 2014). In other words, IT training programs will impart on participants
ernment (Steele et al. 2009; Nam 2012). In other words, just because greater confidence in their ability to use e-government websites,
citizens think e-government is difficult to use does not mean they do not thereby leading to greater e-government use. Second, we predict that
see the value in it. the strength of the relationship between citizen participation in gov-
All told, evidence indicates that the crucial differentiating factor ernment IT training programs and levels of e-government use will be
explaining variation in levels of e-government use is citizens' levels of strongest for groups of citizens who tend to use e-government the least,
PEU, or their self assessed ability to use technology to access govern- namely the elderly and the disabled. This is because individuals be-
ment online. This is problematic in that in many instances those with longing to groups that tend to use e-government the least also, on
lower levels of PEU belong to segments of the population that rely most average, tend to have lower levels of digital literacy when compared to
on government services, for example the elderly and the disabled the general population. In this sense, there is a diminishing returns type
(Colesca & Dobrica, 2008). As a result, a key risk governments run in relationship in that gains from participation in government IT training
this era of digital government is that the growing reliance on e-gov- programs will be largest for those who have the lowest levels of digital
ernment applications to engage, inform, and serve the public will re- literacy, in this case those who are elderly or disabled, because they are
inforce or even enhance rather than overcome existing social chasms. learning how to do more than their peers who have higher levels of
To address this dimension of the digital divide, interventions at- digital literacy. These points lay the foundation for the following hy-
tempting to enhance citizens' (self assessed and objective) ability to use potheses:
technology and e-government are essential. Against this backdrop,
H1. Citizen participation in government IT training programs is
government IT training programs play an important role for two key
positively associated with e-government use.
reasons. First, they are essential to imparting on citizens least likely to
use e-government the technical skills and knowledge needed to use e- H2. The positive relationship between citizen participation in
government for specific purposes. As Eshet-Alkalai and Chajut (2010) government IT training programs and e-government use will be
show that general technology training programs targeting the elderly, stronger for citizens who are elderly or disabled when compared to
which focus on enhancing their technical understanding of how to use the general population.
different devices for specific purposes are highly effective in improving
participants ability to use these devices to achieve a predetermined set
of task. These results are echoed elsewhere as well (Piotrowski & Borry, 4. Method
2009; Kim, Cho, & Suh, 2017). To this end, government IT training
programs can play an important role in ensuring citizens possess a basic 4.1. Data
technical skillset that is needed to access specific e-government func-
tions. Second, government IT training programs operate as a forum for The data we use comes from a 2011 survey conducted by the
experiential learning for participants that enhance PEU. This is because National Information Society Agency (NIA). This data incorporates re-
training programs allow participants to actively develop knowledge and presentative samples of citizens from three target populations in South
skills pertaining to technology by using technology in a supervised Korea: the disabled, the elderly (excluding disabled), and members of
environment (Kolb and Kolb 2005). By acting as a forum for experi- the general population (excluding those who are elderly or disabled).
ential learning, government IT training programs do not just impart Including samples that address these groups of citizens in particular is
technical knowledge that allows participants to understand how to valuable for the purposes of this research as those who are disabled or
access a digital services discussed in class. Rather, they also allow elderly tend to use e-government least (Gauld& Horsburgh, 2010).
participants supervised space to figure out how to use forms of digital Data were collected using face-to-face interviews with structured
services, such as e-government that are not addressed in class through questionnaires. Both multi-stage stratified and proportional stratified
worked examples. In other words, the experiential component to IT sampling procedures were used to ensure representativeness of the
training programs fosters a more flexible understanding of how to use three target populations (NIA, 2012).
technology to access government online, thus enhancing citizens' po- More specifically, the target populations are defined as follows
tential to independently access a broader set of public services online. (NIA, 2012): (1) The general target population: citizens over 20 to
In this respect, these programs are not only effective because they imbue 50 years old and do not possess a disability, but are otherwise re-
participants with a flexible understanding of how to use e-government, but presentative of the entire national population. The general population
they also enhance participants' confidence to use e-government as well sample generated 1929 usable responses that are representative of the
(Clark and White 2010). Taking a broader perspective, by providing general population on dimensions of age, gender and province. (2)
participants with technical and flexible forms knowledge and enhan- Disabled target population: individuals aged 20 to 69 who suffer from
cing their confidence, participation in government IT training programs mental retardation, brain lesions, hearing or language disorders, and
can play a critical role in fostering more positive perceptions regarding visual handicaps and who have enrolled themselves on the national
e-government's ease of use. disabled citizens list. The disabled population sample generated 3497

3
J. Lee, G.A. Porumbescu Government Information Quarterly xxx (xxxx) xxx–xxx

Table 1
The number of respondents by population types (persons, %).
Age > 17 Disabled Elderly Ordinary pop.

Did not participate in government IT training program 2670(76) 3171(90) 1553(81)

Participated in government IT training program 827(24) 353(10) 376(19)
Sub-group total 3497(100) 3524(100) 1929(100)
Total 8950

usable responses that are reflective of the population of disabled citi- services they obtain using e-government. Participation in a government
zens on dimensions of age, gender, province and type of disability. (3) IT training program is a binary variable, with the reference case being
Elderly target population: individuals over the age of 50 at the time they not participating in a government IT training program. We also include
were interviewed. Those who were above the age of 50 but disabled two interaction terms to examine how the relationship varies according
were counted as disabled and not elderly. The number of usable re- to whether one is disabled/elderly or a member of the general popu-
sponses for the elderly sample was 3524 and responses were reflective lation (reference case). If the strength of the relationship does not differ
of the general population of elderly on dimensions of age, gender, and (e.g., the interaction term is insignificant) than this suggests these
province. programs are unsuccessful in closing the digital divide. We include
Table 1 shows the number of respondents, according to their en- three models to enhance the transparency of our analysis. In the first
rollment in government training programs. model, we provide estimates for key variables (elderly, disabled, and
participation in a government IT training program). In the second
4.2. Measure model, we include control variables. In the third model, we include
interaction terms.
4.2.1. Independent variables Finally, we are also interested in understanding the correlates of
Participation in an IT training program is measured using a two-part participation in government IT training programs for each of the three
question. The first part of the question asks participants: “have you ever samples included in our analysis – the elderly, disabled, and the general
participated in a Computer, Internet, and/or information & commu- public. This information adds depth to our analyses because it offers
nication technology training program?” Respondents were given the insight into which segments of the samples benefit most from these
option to answer either yes, or no. For those who answered ‘yes,’ a the programs. This information also offers insight into whether certain
second part of the question was provided: “was the program you par- segments of the three populations we assess are systematically opting
ticipated in funded by the government and held in a welfare center, out of participating, thus helping to identify opportunities for enhan-
community center, local government building, post office, or school cing the effectiveness of this policy – a point that will be discussed at
(including at home program for the disabled).” Here, participants also greater length in the discussion and implications. To identify correlates
had the option to respond yes, or no. Participants who responded yes to of citizen participation in government IT training programs for each of
both questions were coded as having used a government IT training the samples, we assess three sets of bivariate correlations – for the
program. sample of elderly citizens, disabled citizens and citizens belonging to
the general population. We focus on socio-demographic correlates of
4.2.2. Dependent variables: internet utilization for e-government for public participation in government IT training programs. Given the cross-
information sectional structure of the data, bivariate correlations are used instead of
In this article, e-government use is evaluated using the following more complex forms of analyses such a multivariate regression for the
question: “Of all the public services you use, what percentage do you reason that this form of analysis is more straightforward yet affords
use the internet for?” Values can range from zero to 100%. This mea- readers substantively similar insights.
sure of e-government use is novel in that it looks at it assesses e-gov-
ernment use as a proportion of all public services a citizen uses. This 5. Findings
provides more nuanced insight when compared to measures that simply
gauge e-government use, without taking into consideration the total The results of our negative binomial regression analyses can be
amount of public services a citizen uses. found in Table 4.
Hypothesis 1 predicts that citizen participation in government IT
4.2.3. Control variables training programs will be positively associated with e-government use,
In our model, we control for socio-economic variables (age, sex, Models (1) and (2) are used to examine this hypothesis and offer evi-
income, education level, occupation), as well as whether or not re- dence of a positive relationship between participation in a government
spondents have a computer at home and how much time they spend on IT training program and levels of e-government use. Specifically, those
the Internet a day. We control for these variables because they may who participate in government IT training programs make greater use
influence individual's adoption or utilization of e-government. Table 3 of e-government than those who do not participate in government IT
provides the descriptive statistics of total samples. training programs. Moreover, as model (2) demonstrates this relation-
ship holds when controlling for ostensibly relevant factors, such as
4.3. Model whether respondents have access to Internet enabled devices and the
amount of time they spend on the Internet. All told, models (1) and (2)
Our dependent variable varies in relation to the number of public turn back robust evidence in support of hypothesis 1.
services a citizen accesses online. This means that our dependent Hypothesis 2 predicted that the positive relationship between ci-
variable, e-government use, should be treated as a count variable. tizen participation in government IT training programs and e-govern-
Moreover, our dependent variable is overdispersed. Therefore, to esti- ment use will be stronger for citizens who are elderly or disabled when
mate relationships using an over dispersed count variable we use a compared to the general population. To explore this hypothesis, first we
negative binomial regression model (NBR). examine patterns in e-government among those who are disabled and
In our model, we estimate the relationship between a citizen's par- elderly when compared to the general population. Model (1) shows how
ticipation in a public IT training program and the proportion of public levels of e-government use vary according to whether a citizen is

4
J. Lee, G.A. Porumbescu Government Information Quarterly xxx (xxxx) xxx–xxx

Table 2
Variables.
Source: NIA (National Information Society Agency).
Variables Questionnaires

E-government use Of all the public services you use, what percentage do you use the internet for (eg., issuing and reading petition documents, paying taxes and
public utilities (0–100%, an interval of 10%)
ICT training program Have you ever participated in an information & communication technology training program?
If you replied ‘yes,’ were the programs you participated in fully or partially funded by government and conducted at welfare centers, communities
centers, local governments, post offices, or schools (including at home program for the disabled)
1 = Participated in a government IT training program, 0 = Has not participated in a government IT training program
Age Age
Gender 0 = Female
1 = Male
Possession of Internet device 1 = has internet devices at home or office
0 = does not
Internet use The number of hours spent using the Internet per day
Education 1 = Primary or lower secondary school, or no formal education
2 = Middle school
3 = High school
4 = Higher education (e.g., university, college, polytechnic)
Income 1 = if < 500 dollars per month
2 = if > 500 and < 1000 per month
3 = if > 1000 and < 1500 per month
4 = if > 1500 and < 2000 per month
5 = if > 2000 and < 2500 per month
6 = if > 2500 and < 3000 per month
7 = if > 3000 and < 4000 per month
8 = if > 4000 dollars per month

disabled or elderly. The results indicate that, when compared to participation in government IT training programs helping to expand e-
members of the general population, respondents who are disabled or government use (Tables 2, 5–7).
elderly are significantly less likely to use e-government to access public Given the relationships uncovered, we are also interested in
services. Model (2) indicates that these results hold, even when ac- knowing more about who is participating. More specifically, are certain
counting for socio demographic variables, such as one's age or educa- demographic groups among the elderly, disabled and general public
tion, or how frequently one uses the Internet. Further, model (2) in- systematically opting out of participating?
dicates that, when controlling for relevant background variables, the As can be seen across the three tables, certain similarities and dif-
elderly are the least active in using e-government, followed by the ferences are observed. Irrespective of whether one is elderly, disabled
disabled. Together, model (1) and (2) indicate that citizens belonging to or a member of the general population, we see that factors such as age,
these groups of the population lag behind in their use of e-government possession of an Internet enabled device and levels of Internet usage are
when compared to members of the general population. positively associated with levels of participation in government IT
Next, we assess how the strength of the relationship between par- training programs. Interestingly gender is not significantly related to
ticipation in a government IT training program and e-government use participation in any of the groups whereas higher levels of income are
varies according to whether a participant is elderly, disabled, or a for the disabled group, but not for the elderly or general populations.
member of the general population. At an initial level, our interaction This may be because those who are disabled will be less mobile, thereby
terms are both positive and significant thereby offering evidence that making costs of transportation to training programs higher. One critical
the strength of the relationship between participation in a government difference between those who belong to more vulnerable segments of
IT training program and e-government use does indeed differ according the population and those belonging to the general population however,
to membership in one of these three segments of the population. More is education. Namely, for the sample of respondents recruited from the
specifically, we find that participation in government IT training pro- general population, we observe no significant relationship between
grams is more strongly associated with greater e-government use for education and participation indicating that levels of participation did
those who are elderly or disabled than for those who are not elderly or not vary significantly according to how well educated a respondent is.
disabled. Once again, our findings hold in the presence of relevant However, for the samples recruited from the population of elderly ci-
control variables. These findings offer support for hypothesis 2. They tizens and the sample of disabled citizens, we see that participation in
also speak to the potential of government IT training programs to close government IT programs varies significantly across different levels of
the digital divide in that they offer preliminary evidence of respondent education. This points to an issue of reach for these pro-
grams – namely, while they may generally be associated with greater
use of e-government, and more effective for those who are elderly or
Table 3
Sample characteristics. disabled, a more nuanced perspective indicates that the standardized
curriculum used in these programs may preclude progress for those
Mean Std. dev. Min Max with less education, which perhaps signals special learning needs.
E-government use 6.54 18.14 0 100
Gov IT training program 0.17 0.38 0 1
Age 54.17 13.77 18 92 6. Discussion and conclusion
Gender(1 = male) 0.54 0.50 0 1
Possession of internet device 0.75 0.43 0 1 As governments continue to expand their online presence they
Internet use(hour/day) 1.13 1.64 0 12 confront a range challenges and opportunities (Porumbescu, Vrabie,
Education 2.42 0.97 1 4
Ahn & Im, 2012). One key challenge is the notion of digital divides and
Household income 4.37 2.23 1 8
Total 8950 growing inequality in this era of digital governance. Scholars have
outlined different ways in which the growth of digital services and e-

5
J. Lee, G.A. Porumbescu Government Information Quarterly xxx (xxxx) xxx–xxx

Table 4
Negative binomial regression results.
Model 1 Model 2 Model 3

Coef. S.E Coef. S.E Coef. S.E

Intercept 2.71 ⁎⁎⁎

0.11 −2.16 ⁎⁎⁎
0.40 −2.08 ⁎⁎⁎
0.40
Gov IT training program (ref. Non-public program) 1.04⁎⁎⁎ 0.13 0.91⁎⁎⁎ 0.12 0.46⁎⁎ 0.21
Population (ref. Gen. pop.)
Disabled −1.63⁎⁎⁎ 0.14 −0.33⁎⁎⁎ 0.13 −0.45⁎⁎⁎ 0.14
Elderly −1.98⁎⁎⁎ 0.14 −0.49⁎⁎⁎ 0.13 −0.59⁎⁎⁎ 0.14
Disabled⁎participation 0.58⁎⁎ 0.27
Elderly⁎participation 0.61⁎⁎ 0.30
Internet device 2.44⁎⁎⁎ 0.16 2.41⁎⁎⁎ 0.16
Internet use (hour/day) 0.57⁎⁎⁎ 0.05 0.57⁎⁎⁎ 0.05
Age −0.05⁎⁎⁎ 0.00 −0.05⁎⁎⁎ 0.00
Gender (Male = 1) 0.38⁎⁎⁎ 0.09 0.38⁎⁎ 0.09
Education (ref. elementary or below) middle 0.41⁎⁎ 0.16 0.42⁎⁎ 0.16
High school 1.56⁎⁎⁎ 0.16 1.54⁎⁎⁎ 0.16
College above 2.15⁎⁎⁎ 0.20 2.13⁎⁎⁎ 0.20
Income (ref. level < $5000/month) 0.44⁎ 0.23 0.43⁎ 0.23
Level (< $10,000/mon) 0.81⁎⁎⁎ 0.22 0.81⁎⁎⁎ 0.22
Level (< $15,000/mon) 0.82⁎⁎⁎ 0.22 0.85⁎⁎⁎ 0.22
Level (< $20,000/mon) 1.27⁎⁎⁎ 0.22 1.39⁎⁎⁎ 0.22
Level (< $25,000/mon) 1.43⁎⁎⁎ 0.23 1.46⁎⁎⁎ 0.23
Level (< $35,000/mon) 1.46⁎⁎⁎ 0.22 1.47⁎⁎⁎ 0.22
Level (> $40,000/mon) 2.03⁎⁎⁎ 0.24 2.08⁎⁎⁎ 0.24

Log likelihood −11,967 −11,202 −11,199

Observations 8950

⁎
p < .1.
⁎⁎
p < .05.
⁎⁎⁎
p < .001.

government will influence extant forms of inequality, as well as create differently, these findings confirm the presence of a digital divide in
new forms of inequality. Yet, while the causes and effects of digital that some citizens are benefitting more from investments in digital
divides have attracted a great deal of attention, there have been few provision of public services than others. These findings contribute to
attempts to identify and evaluate strategies for closing such disparities. literature on the digital divide by extending our understanding of de-
Namely, what effective paths are there toward ensuring more equitable terminants of the digital divide by demonstrating how belonging to
access to public services online? certain social groups – the elderly and the disabled - shapes e-govern-
One often-used policy to address this issue is government provision ment uptake. These findings also indicate efforts to transition to e-
of IT training programs. These programs intend to afford participants government as a primary means of delivering public services and gov-
the skills and confidence they need to access e-government portals and ernment information creates the potential to reinforce existing social
public goods and services online. Moreover, because they are often free, chasms, and further marginalize vulnerable segments of the population.
government IT training programs constitute a critical means of ensuring In this sense, these findings highlight the need for future research to
that everyone, and in particular those who belong to segments of so- identify and evaluate interventions capable of promoting more equi-
ciety that may have the greatest difficulty in accessing public services table access to digital services.
online, is able to effectively interact with the state using the Internet. The second key finding of this study comes from demonstrating that
However, despite their popularity, there have been no empirical eva- government IT training programs represent one type of intervention
luations of these programs. As a result, the expansion of these policies that can play a crucial role in increasing e-government uptake among
has essentially not been based on evidence, but rather hunches. This all segments of the population. We also find that the relationship be-
study, to the best of our knowledge, represents the first empirical effort tween participation in government IT training programs and e-gov-
to assess the effectiveness of these policies. Our findings provide some ernment use is strongest for those who are older or those who are
insight into this important gap in the literature. We overview key disabled. Indeed, participation in IT training programs is more strongly
findings and discuss contributions below. associated with e-government uptake among those who are disabled
An initial key finding of this study comes from demonstrating that and elderly than for members of the general population. This finding
those who are elderly or have a disability have significantly lower levels contributes to e-government research by illustrating the vital role IT
of e-government use than members of the general population. Put training programs play in promoting e-government use, especially

Table 5
Correlations for sample of elderly respondents.
Training program Internet device Internet use Age Gender Education

Training program 1
Internet device 0.2174⁎ 1
Internet use 0.3615⁎ 0.4519⁎ 1
Age −0.1955⁎ −0.3515⁎ −0.4088⁎ 1
Gender 0.0598 0.0952⁎ 0.1063⁎ −0.0908⁎ 1
Education 0.1894⁎ 0.3490⁎ 0.4098⁎ −0.4763⁎ 0.2182⁎ 1
Income −0.056 0.3167⁎ 0.2112⁎ −0.2566⁎ 0.0359 0.4381⁎

⁎
p < .01.

6
J. Lee, G.A. Porumbescu Government Information Quarterly xxx (xxxx) xxx–xxx

Table 6
Correlations for sample of disabled respondents.
Training program Internet device Internet use Age Gender Education

Training program
Internet device 0.154⁎
Internet use 0.289⁎ 0.318⁎
Age −0.113⁎ −0.364⁎ −0.366⁎
Gender 0.017 0.081⁎ 0.165⁎ −0.090⁎
Education 0.215⁎ 0.360⁎ 0.460⁎ −0.595⁎ 0.259⁎
Income 0.156⁎ 0.418⁎ 0.326⁎ −0.497⁎ 0.112⁎ 0.554⁎

⁎
p < .01.

among vulnerable segments of the population. When evaluated through programs are fairly popular – 24% of the disabled, 10% of the elderly
a lens of the TAM, we also suggest these effects are likely because of the and 19% of the general populations samples participated in a govern-
role participation in IT training programs play in improving citizens' ment IT training program. That said, this study is a preliminary eva-
objective and self-assessed ability to use e-government. Yet at the same luation of the effectiveness of these programs. Further research that
time, we also find that these programs are not equally effective for all of explores the impact of different styles of curriculum, training formats,
those who fall into the category of elderly or disabled. Indeed, we and instruction is needed to tease out greater nuance in terms of how
identify a correlate that may be uniquely significant to understanding these programs are performing. Indeed, because these programs use a
participation among those who belong to vulnerable pockets of the general curriculum that is not tailored to the educational needs of any
population – education. All told, IT training programs are not a silver specific segment of the population, it is important for future research to
bullet when it comes to promoting equitable access to public services provide better insight into methods of enhancing the effectiveness of
provided online, though they do appear to be a good start. these programs. To this end, as governments continue to transition to
Before discussing implications of our findings it is helpful to over- new modalities of public service provision, it is essential training pro-
view some key limitations. First of all, data for this study are cross- grams be altered to keep pace so as to not only ensure that technolo-
sectional, meaning that we are unable to assert a clear temporal or- gical progress enhances efficiency and effectiveness of public service
dering of the variables of interest to this study. For this reason, studies provision, but also equality and equity.
that draw on waves of data collected over multiple years, or an ex- A second key implication of this study pertains to the importance of
perimental design, is needed to tease out causality in terms of the re- ensuring access and participation in such programs. Given the provi-
lationship between participation in IT training programs and e-gov- sional evidence that participation in government IT training programs
ernment use. That said, our findings do provide a solid platform to is positively associated with levels of e-government use, a key challenge
inform future, more elaborate research designs that are capable of confronting these programs is how to bolster citizen participation in
teasing out causality. Second, this study focuses on the relationship them, especially among more vulnerable segments of the population,
between participation in IT training programs and e-government use in who may lack the resources needed to participate. Indeed, as our bi-
one specific context – South Korea. As was explained earlier, local variate correlations indicate, some correlates of participation are un-
governments in South Korea are obligated to provide IT training courses ique to certain segments of the population. This means that in order to
and furthermore. Moreover, when compared European or North ensure that these programs are broadly effective in reaching citizens
American contexts, the population of South Korea is relatively less di- who would benefit from such programs, and not just helping those who
verse. To gain a more robust understanding of this relationship it is are easiest to help (e.g., cream-skimming) distinctive targeting strate-
important to explore the extent to which the patterns of relationships gies are necessary. Put differently, generic recruitment efforts will likely
we uncover in this study hold in other national contexts. Third, while fall short of mobilizing more vulnerable segments of the population.
this study makes an effort to account for confounding factors, such as With respect to the third implication, our findings offer evidence to
socio-economic variables and one's use of the Internet, future research suggest that the growing migration of government functions to online
is needed in order to explore the extent to which these findings will formats can contribute to reinforcing extant inequity in terms of the
hold in the presence of alternative sets of control variables. provision of public services – those who belong to vulnerable segments
The limitations of this study notwithstanding, our findings offer of the population tend to use e-government less, even in highly con-
valuable implications for research and practice. First key implication of nected societies, such as South Korea. To this end, while research on e-
this study is that government IT training programs can play a critical government suggests that the proliferation of technology can render
role in efforts to mitigate the digital divide. Indeed, we see that these public services easier to access, a more nuanced perspective suggests
participation in these programs is positively a greater usage of e-gov- these effects will vary significantly in practice. Therefore, as govern-
ernment. Specifically, those who participate in these programs obtain a ments continue to draw upon new forms of technology to inform and
greater proportion of public services online than those who do not. Our engage citizens, a key implication of this study pertains to the im-
descriptive statistics can also be interpreted to suggest that these portance of identifying strategies to develop a more comprehensive

Table 7
Correlations for sample of general population respondents.
ICT training program Internet device Internet use Age Gender Education

Training program 1
Internet device 0.109⁎ 1
Internet use 0.110⁎ 0.244⁎ 1
Age −0.136⁎ −0.394⁎ −0.259⁎ 1
Gender −0.046 0.058 0.098⁎ −0.024 1
Education 0.065 0.390⁎ 0.289⁎ −0.462⁎ 0.146⁎ 1
Income 0.023 0.389⁎ 0.157⁎ −0.359⁎ 0.033 0.468⁎

⁎
p < .01.

7
J. Lee, G.A. Porumbescu Government Information Quarterly xxx (xxxx) xxx–xxx

understanding of how this digital transformation is shaping relation- disadvantaged. MIS Quarterly, 97–126.
ships between different segments of the population and the state. As our Kim, Y.-D., Cho, Y., & Suh, Y. (2017). A study of the effectiveness of information literacy
education among the elderly: A focus on digital literacy and quality of life. Korean
findings indicate, this gap in our understanding of the effects of e- Journal of Public Administration, 55(1), 229–259.
government is problematic in that this neglect can result in further Kvasny, L., & Keil, M. (2006). The challenges of redressing the digital divide: A tale of two
marginalization of vulnerable segments of the population. In addressing US cities. Information Systems Journal, 16(1), 23–53.
Lee, J., & Kim, S. (2014, January). Active citizen e-participation in local governance: Do
this gap, we drew upon the TAM to suggest that perceived ease of use as individual social capital and e-participation management matter? System Sciences
a critical determinant of e-government use and, in turn identified IT (HICSS), 2014 47th Hawaii International Conference on (pp. 2044–2053). IEEE.
training programs as one potential intervention. Going forward, future Loges, W. E., & Jung, J.-Y. (2001). Exploring the digital divide: Internet connectedness
and age. Communication Research, 28(4), 536–562.
research would do well in identifying alternative theoretical frame- Ministry of Government Administration and Home Affairs (MGAHA) (2016). Best prac-
works that help to distill other determinants of citizens' use of e-gov- tices in e-government of Korea. [online], available: http://unpan1.un.org/intradoc/
ernment services and in turn search for interventions that address these groups/public/documents/UNGC/UNPAN043625.pdf, Accessed date: 25 November
2017.
determinants. Doing so will improve our understanding of interventions
Ministry of the Interior (MOI) (2016). Building better governance: The Korean case. Vol. 1.
that are capable of comprehensively addressing inequity that results Ministry of the Interior.
from the digital divide. Molnár, S. (2003). The explanation frame of the digital divide. Proceedings of the summer
As technology stands to transform features of public services (e.g., school, risks and challenges of the network society (pp. 4–8). .
Moon, M. J. (2002). The evolution of e-governmnet among municipalities: rhetoric or
more responsive or easier to access), core challenges, such as those reality? Public Administration Review, 62(4), 424–433.
pertaining to the inclusive delivery are as relevant today as they have Nam, T. (2012). Citizens' attitudes toward open government and government 2.0.
ever been. Indeed, as innovation and the new forms of technology International Review of Administrative Sciences, 78(2), 346–368.
National Information Society Agency (NIA) (2012). NIA.2011 analysis on digital divide and
adopted by governments help to reshape the structure of relationships policy implication report. NIA.
between citizens and the state, inequity remains as a stable feature of Okunola, O. M., Rowley, J., & Johnson, F. (2017). The multi-dimensional digital divide:
the public administration landscape. To this end, as the functions of Perspectives from an e-government portal in Nigeria. Government Information
Quarterly, 34, 329–339.
government progressively shift to online formats, our findings speak Porumbescu, G. A. (2016). Placing the effect? Gleaning insights into the relationship
strongly to the importance of governments not just working to identify between citizens’ use of e-government and trust in government. Public Management
more flexible methods of service delivery, but to also identify methods Review, 18(10), 1504–1535 (Chicago).
Piotrowksi, S. J. (2010). The operationalization of municipal transparency: Primary ad-
of ensuring that progress is inclusive and felt by all segments of society. ministrative functions and intervening factors. Transparencia Y Privacidad, 1(1), 1–31
We offer evidence to suggest that training programs can play an im- (Chicago).
portant role in this capacity, while also highlighting avenues for further Piotrowski, S. J., & Borry, E. L. (2009). Transparency and local government websites.
Handbook of research on strategies for local e-government adoption and implementation:
research.
Comparative studies (pp. 390–407). IGI Global.
Porumbescu, G., & Im, T. (2015). Using transparency to reinforce responsibility and re-
Acknowledgement sponsiveness. The handbook of public administration (pp. 120–136). (3rd edition). San
Francisco, CA: John Wiley.
Porumbescu, G., Vrabie, C., Ahn, J. H., & Im, T. B. (2012). Factors influencing the success
Porumbescu's work was supported by a National Research of participatory e government applications in Romania and South Korea. Korean
Foundation of Korea Grant from the Korean Government [NRF- Journal of Policy Studies, 27(1), 1–21.
2017S1A3A2065838]. Rutgers' E-Governance Institute (2014). Digital governance in municipalities
worldwide—Sixth global e-governance survey: A longitudinal assessment of municipal
websites throughout the world. Newark: Rutgers, The State University of New Jersey.
References Shyu, S. H.-P., & Huang, J.-H. (2011). Elucidating usage of e-government learning: A
perspective of the extended technology acceptance model. Government Information
Quarterly, 28(4), 491–502.
Clark, J., & White, G. (2010). Experiential learning: A definitive edge in the job market.
Steele, R., Lo, A., Secombe, C., & Wong, Y. K. (2009). Elderly persons' perception and
American Journal of Business Education, 3(2), 115–118.
acceptance of using wireless sensor networks to assist healthcare. International
Colesca, S., & Dobrica, L. (2008). Adoption and use of e-government services: The case of
Journal of Medical Informatics, 78(12), 788–801.
Romania. Journal of Applied Research and Technology, 6(3), 204–217.
United Nations (UN), E-Government Survey (2016). [online], available: http://
Davis, F. (1989). Perceived usefulness, perceived ease of use, and user acceptance of
workspace.unpan.org/sites/Internet/Documents/UNPAN96407.pdf, Accessed date:
information technology. MIS Quarterly, 13, 318–341.
25 November 2017.
Dunleavy, P., Margetts, H., Bastow, S., & Tinkler, J. (2006). New public management is
Zheng, Y., & Schachter, H. L. (2016). Explaining citizens' e-participation use: The role of
dead—Long live digital-era governance. Journal of Public Administration Research and
perceived advantages. Public Organization Review, 1–20.
Theory, 16(3), 467–494.
Eshet-Alkalai, Y., & Chajut, E. (2010). You can teach old dogs new tricks: The factors that
affect changes over time in digital literacy. Journal of Information Technology Jaebok Lee is a visiting research fellow in the Department of Public Safety and Social
Education: Research, 9, 173–181. Cohesian at the Korean Institute for Public Administration. She was a visiting scholar at
Fishbein, M., & Ajzen, I. (1975). Belief, Attitude, Intention, and Behavior. Reading: Addison- the School of Public Affairs and Administration, Rutgers University–Newark from 2015 to
Wesley Publishing Company, Inc. 2016. Her research interests include, nonprofit organizations, social service delivery
Gauld, R., Goldfinch, S., & Horsburgh, S. (2010). Do they want it? Do they use it? The systems, e-government, and individual behavior and preference.
‘Demand-Side’ of e-Government in Australia and New Zealand. Government
Information Quarterly, 27(2), 177–186. Gregory A. Porumbescu is an assistant professor in the School of Public Affairs and
Gonzales, A. (2016). The contemporary US digital divide: From initial access to tech- Administration. He is also the associate director of the Transparency and Governance
nology maintenance. Information, Communication & Society, 19(2), 234–248. Center.
Hsieh, J. P.-A., Rai, A., & Keil, M. (2008). Understanding digital inequality: Comparing
continued use behavioral models of the socio-economically advantaged and