Use of e-government services in a deeply divided society: A test and an extension of the social inequality hypotheses

Abstract

The issue of e-government use is critical in deeply divided societies where the access to social capital is restricted for minorities. E-government use in these societies may differ by ethnic background, size of locality or both. Israel was chosen as a case since it is an example of a deeply divided society. Using data from the Israel’s Social Survey 2015, it was found that the disadvantaged minority has a lower probability of using e-government as compared to other groups. However, when segmenting this population by size of its localities, the stratification structure differs between large and small localities. The conclusion is that the inequality approaches should consider not only the ethnicity but also the size of locality as a factor differentiating between ethnic groups in terms of the specific Internet use.

Keywords

E-government ethnicity size of locality social diversification social stratification

Introduction

Since the late 1990s, governments all over the world have started to use online space in order to supply information and services (Chen and Gant, 2001). Thus, the term “e-government” was coined. E-government services may include research information, forms, information about public policy, licensing, and so on (Lin et al., 2011). It enables citizens to request information, pay taxes, renew licenses, and so on (Palvia and Sharma, 2007).

There are several positive aspects of e-government. From the public sector perspective, its use contributes to increase in trust in government. The latter encourages obedience to laws and regulations (Tolbert and Mossberger, 2006). Thus, by providing services online, governments “earn” loyal citizens and even encourage them to participate in the work of the government (Lin et al., 2011). From the perspective of citizens, they can contact government institutions online at any time and place without needing to visit any office (Reddick, 2005) or adjust for hours of opening (West, 2004), thus saving personal resources (Wangpipatwong et al., 2005). In sum, e-government is perceived as a way to improve service to citizens, make it more accessible and build trust in the government in the long run (West, 2004).

Nevertheless, the following question should be asked: do all of the sectors of population use these services to the same extent, thereby equally enjoying the benefits of their use? In other words, what is the “structure” of e-government use in terms of ethnic belonging? These questions are of even higher importance in deeply divided societies, where the minority populations, especially those residing in small localities, have a limited access to social capital as compared to the majority population.

To address these questions, two contradicting hypotheses from the sociology of Internet are tested in a current study: the Social Stratification Hypothesis and the Social Diversification Hypothesis. Both of them acknowledge that the access to information and services is unequally distributed between various ethnic groups in a given society. This stems mainly from the fact that ethnicity reflects an ability to access and gain social capital (Mesch et al., 2012). However, from this point, these approaches go in opposite directions. The Social Stratification Hypothesis claims that the existing inequalities in society are amplified on the Internet (Van Deursen et al., 2015), resulting in a better offline outcomes (Van Deursen and Helsper, 2015). Ethnic groups that have greater social capital will both find themselves on the Internet earlier than the disadvantaged ones and have better skills to use it, consequently expanding their resources (Mesch, 2012). This perspective sees the Internet as a direct contributor to the increase in ethnic inequality. In contrast, the Social Diversification Hypothesis claims that disadvantaged ethnic groups will overcome the existing social barriers and gain a better access to social capital using the Internet (Mesch, 2012). The Internet is seen by this approach as an instrument for decreasing the extent of the ethnic inequality.

Both of these hypotheses were tested on the use of Internet for various purposes and found to reflect the ethnic structure, depending on the goal of use. In support of Social Diversification Hypothesis, Mesch et al. (2012) found that Internet users, who belong to the minority group, tend to both search for health information and communicate on health issues more than users belonging to the majority group. In support of the Social Stratification Hypothesis, Mesch (2016) found that the latter have a higher probability of using online health services than the former. However, none of these studies accounted for differences that may exist between the subgroups of disadvantaged minorities, mainly according to the size of locality. The current study will fill this gap.

Israel is an interesting case to study, because it represents an example of a deeply divided society. It is characterized by an extreme residential segregation between the majority (Jews) and the minority (Arabs) groups (Lewin-Epstein and Semyonov, 1992). The majority of the latter resides in small communities on the periphery of Israel (Mesch, 2016). This restricts their opportunities to gain an access to information and services.

The research on ethnic differences in e-government use is important for the theory in the field in light of the fact that there are few studies that examine the effect of the socio-demographic factors in general (Patel and Jacobson, 2008) and ethnic origin in particular (Gauld et al., 2010) on the use of e-government. Most of the studies in this area test theoretical models (Al-Adawi, et al., 2005; Carter and Belanger, 2004; Lin et al., 2011). However, the focus of these models is on studying the intention to use e-government services rather than the actual use. Moreover, they account for the variance in socio-economic background to a very small extent.

The research on ethnic differences in e-government use is also of high importance for actual practice. By studying the phenomenon, it is possible to establish a policy in order to provide all segments of the population with these services such that all citizens, regardless of their ethnic origin, will be able to equally enjoy their benefits.

The current study will employ ethnic origin as a main variable of interest with socio-demographic and situational variables as control variables in order to explain the outcome variability. The goal is to find whether, after controlling for these factors, the ethnic difference in e-government use persists and in which direction. Another goal, which is a novelty of the study for the field, is to find whether the size of locality is a salient factor to consider when discussing the ethnic inequality in e-government use. Thus, the research questions are as follows:

RQ1. How does the ethnicity affect the e-government use?

RQ2. Are there differences between the disadvantaged groups, according to size of locality, in their use of e-government, relatively to both each other and other groups?

This article is organized as follows: First, the conceptual background is presented. Then, the methodology of the study, including definitions of variables, is described. Next, the descriptive, bivariate and multivariate analyses are presented, finishing with the discussion and summary of the results.

E-government—for what purpose?

From the macro perspective, there are two main motivations for implementing an e-government initiative. First, there is an acknowledged decline in trust in government (Tolbert and Mossberger, 2006), political participation and civic engagement (Di Gennaro and Dutton, 2006). Thus, the online presence of governments may restore public trust (Al-Shafi and Weerakkody, 2010). Second, e-government projects are one requirement of a modern society, not just because everything goes online. Due to social changes, such as the emergence of information society (Castells, 2010), social institutions are seeking ways to keep on functioning efficiently (Almarabeh, 2011). While the citizens experience a constant improvement in online services provided by the private sector, they expect the same from the public sector (Al-Shafi and Weerakkody, 2010). This is important in light of the findings suggesting that the perception of the efficiency of e-government (in terms of perceived usefulness, perceived ease of use, etc.) positively affects both the intention to use e-government and its actual use (Carter and Belanger, 2005; Mahadeo, 2009). The latter positively affects trust in government (Tolbert and Mossberger, 2006).

From government to citizen (user)

Unfortunately for the governments, not everybody adopts their online projects in general and as a primary/important channel of interaction with them in particular. Therefore, the focus should be primarily on a micro perspective in order to discover the factors of e-government use.

As previously mentioned, various theoretical concepts try to explain the intention to adopt technology or its actual use on a micro level. The most frequently used concept for this purpose is TAM (Al-Adawi et al., 2005; Carter and Belanger, 2005; Lin et al., 2011). Its core assumption is that the intention to adopt any technology depends on the attitudes toward it, which in turn are a function of perceived usefulness and perceived ease of use (Carter and Belanger, 2005). However, this model in all of its further variations only accounts to a very small extent for the social background of the users and to no extent for their ethnic background. Therefore, the most suitable theoretical framework for a current study is one that accounts for gaps in Internet use according to socio-demographic background, that is, a digital divide approach (Dodel, 2016; Mesch, 2012, 2016; Van Deursen et al., 2015; Van Deursen and Helsper, 2015).

It is commonly known that existing social inequalities tend to “migrate” into the online sphere (Dodel, 2016; Van Deursen et al., 2015). E-government use is known as unequally distributed over population (Gauld et al., 2010), with this inequality related to the “digital divide” (Mesch et al., 2012). In the beginning of the digital era, this term referred to having access to computers and Internet, what is called “first-level digital divide” (Van Deursen and Helsper, 2015). The issue of access plays nowadays a much smaller role than in the past due to high rates of the Internet penetration around the globe. Therefore, the main topic of discussion in sociology of Internet is now the differences among groups in terms of its use (“second-level digital divide”) (Mesch, 2012, 2016) and the offline benefits of such use (“third-level digital divide”) (Van Deursen and Helsper, 2015).

Studies have already “painted” a “portrait” of Internet and e-government users. They are mostly young, highly educated and high income (Gauld et al., 2010; Woolley and Peterson, 2012). Gender and ethnicity effects vary between the goals of use. Men use e-government more than women (Gauld et al., 2010), but the latter dominate in the health-related use (Mesch et al., 2012). Ethnic minorities are found to use health services less than the members of the majority group (Mesch, 2016), but search for health information on the Internet more than the latter (Mesch et al., 2012).

Ethnicity, residential patterns and e-government use

When discussing the role of ethnicity, it can be assumed that the e-government use varies among ethnic groups. In deeply divided societies, ethnicity reflects residential segregation (Mesch et al., 2012). This is an important factor to consider since it may limit the opportunities of minority groups to access services and information (Mesch, 2016). Given the inconsistency in results on the effect of ethnicity on various Internet uses (see Introduction), but knowing that the former may have some effect on the latter, it is hypothesized that

H1. Ethnicity affects e-government use.

Despite the predominant role of the ethnicity in accessing social capital, size of localities must not be ignored, especially for disadvantaged minorities. Residing in larger localities means having a better Internet infrastructure (Dodel, 2016). Thus, it may contribute to the increase in social stratification and lead to the extended use of e-government by the majority population. This may also benefit with the members of disadvantaged minorities and lead them to use e-government more than their counterparts residing in small localities. However, residing in small localities may serve a motivator for extensive use of the Internet due to the limited opportunities and resources provided by these localities. This because public services are mostly situated at larger localities (Edmiston, 2003). Thus, it may contribute to the decrease of social inequality (Mesch, 2016), so that minority population will use e-government more than the majority population. However, since it is still unknown whether residing in small localities in an advantage or a disadvantage for the disadvantaged minorities in terms of e-government use, it is suggested that

H2. Disadvantaged minorities use e-government to different extent, according to size of their localities (small as opposed to large)—both one compared to another and compared to other ethnic groups.

Additional factors explaining the e-government use

The theoretical framework employed in this study accounts for the fact that there are factors beyond ethnic background that may explain the e-government use. Therefore, it is important to control them. Below, a detailed explanation on how each variable may affect the studied phenomenon is provided.

Effects of socio-economic variables

Education

Individuals with higher education express a greater readiness to use e-government (Patel and Jacobson, 2008) and actually use it more than lower educated ones (Carter and Weerakkody, 2008; Reddick, 2005). This is because higher educated people have all the resources required to use the Internet (Basu, 2004). In addition, they tend to adopt innovations more than lower educated ones (Al-Shafi and Weerakkody, 2010).

Age

According to Reddick (2005), users aged between 55 and 64 tend to search for information on e–government sites less than the younger users. This is because the older population can be less open toward adoption of new technologies (Phang et al., 2005). In contrast, younger users are indifferent to adoption of technologies (Al-Shafi and Weerakkody, 2010).

Gender

Al-Shafi and Weerakkody (2010) found that the gender composition of the adopters is different from that of the non-adopters. It can be explained by differences in online communication patterns of both men and women (Patel and Jacobson, 2008). Men tend to be more task-oriented as compared to women. The latter have lower self-efficacy and a larger fear of computers (AlAwadhi and Morris, 2009).

Language proficiency

It is argued that cultural factors such as language proficiency should be taken into account when discussing e-government adoption (Mahadeo, 2009). Proficiency in local language increases human capital, which, in turn, brings larger benefits in various domains of life (Mesch, 2003). This may be reflected in the Internet use either. Therefore, individuals with better local language proficiency are expected to have better access to and better skills in using e-government services.

Effects of the situational variables

Frequency of Internet use

The literature claims that experienced users have a higher probability of using e-government services (Welch and Hinnant, 2003). This is because the frequent Internet use increases the familiarity with its different applications, including the governmental one (Welch and Hinnant, 2003). In addition, the more people use the Internet, the more they trust it (AlAwadhi and Morris, 2009). Since trust reduces the perceived risk (Al-Adawi et al., 2005), it is not surprising that trust of the Internet is positively associated with the intentions to use e-government services (Belanger and Carter, 2008; Sang et al., 2009).

Downloading files

Individuals who perform various actions online are expected to perform the same actions when using the e-government services. Positive attitudes toward technology increase the probability of performing these actions on all possible domains, including the e-government (Gauld et al., 2010).

E-government in Israel

The history of e-government in Israel traces its roots back to the end of the 1980s when the government was striving to improve governmental services using computerized media. In 1997, the government decided to prepare the state for the information age. The decision included making changes in legislation and work patterns in order to make the public sector more efficient (Rotem, 2007). In May 2002, the Israeli government started to implement the e-government project (B. Chen and Rashty, 2002). The vision of e-government project in Israel included improving the service for citizen and the image of government and decreasing the scope of bureaucracy (Goldshmidt, 2012).

Nowadays, the Israeli government portal offers services for citizens, the private sector and tourists wishing to visit the country as well as students and members of the Jewish Diaspora. The portal also offers online forms, with the possibility of carrying out various online transactions through the portal (United Nations, 2014).

Ethnicity in Israel

Israel is a multinational state. The ethnic layer differentiates between Jewish and Arab population (Bar-Haim and Semyonov, 2015; Mesch, 2012). Jewish population is divided according to country of origin (Raijman et al., 2015), further distinguishing between Israeli-born Jews, immigrants from the former Soviet Union (hereinafter FSU) and non-FSU immigrants. Below, each segment of the population is presented.

Israeli Jews are the majority population. They constitute about 70% of Jewish population in Israel (Bar-Haim and Semyonov, 2015). The main distinction is between Jews of European-American and of Asian-African origin (Lewin-Epstein and Semyonov, 1992).

Immigrants from the FSU are the largest immigrant group in Israel (Raijman et al., 2015), constituting about 15% of total population and about 20% of the Jewish population in Israel (Mesch, 2012). Since 1989, approximately a million FSU immigrants have come to Israel (Mesch, 2012). To some extent, they tend to preserve a cultural distinction from the majority group, operating mass-media channels in Russian, having their own political representatives, and so on (Al-Haj, 2004).

A large part of the non-FSU immigrants has come in recent years from Ethiopia, Western and Central Europe, North and South America with much less from Asian and North African countries. Except for immigrants from Ethiopia, immigrants from these countries have a lower tendency than FSU immigrants to reside in concentrated neighborhoods (Raijman et al., 2015).

Israeli Arabs constitute about 19% of total population of Israel (Abbas and Mesch, 2016). This population is heterogeneous in terms of both religious composition and socioeconomic status (Bar-Haim and Semyonov, 2015). However, a common feature to all of the groups in this sector is a residential segregation. Most of the Arab-speaking population resides in small peripheral localities, with only 10% living in mixed cities (Saabneh, 2015), while only seven urban localities in Israel (out of 101) defined as “mixed” (Lewin-Epstein and Semyonov, 1992).

The residential segregation of Arab population is expressed in a number of difficulties. First, the major urban centers are relatively distant from the segregated communities (Lewin-Epstein and Semyonov, 1992). Second, public services are allocated by place of residence. Therefore, there are ethnic differences (in favor of the majority population) in terms of access to those services, which are even replicated in use of similar online services (Mesch, 2016). Third, the state invests less resources in the infrastructure in the periphery (Mesch and Talmud, 2011), resulting in a larger percent of users in Jewish sector rather than in Arab sector (Mesch, 2012).

Considering all these, the effect of ethnic origin on the use of e-government will be tested controlling for other variables. In order to better understand this use, first, the general Internet use will be regressed on socio-demographic variables. Then, the e-government use will be predicted applying the same socio-demographic and situational variables. As a first step, the methodology of the study is presented in the following chapter.

Method

General description

The data for the current study were attained from the Israeli 2015 Social Survey (Israel Central Bureau of Statistics [CBS], 2015). This dataset is a large representative sample of the Israeli population (Lewin and Stier, 2018). Social Surveys are carried out annually by Israeli CBS among individuals aged 20 and older permanently living in Israel. Each year, the survey consists of two parts: a core questionnaire and a variable module devoted to one or two topics requesting more detailed data than in the core questionnaire. The data are collected via face-to-face computer-assisted personal interviews (CAPI), each of which lasts for about an hour (CBS, n.d.).

The current dataset is particularly suited for the current research because it includes several items about computer and Internet use for various purposes, including to obtain services from governmental institutions.

Since I first predict the Internet use, I include the entire sample (N = 7078) in my analysis. It consists of 73.7% of total contacted persons in the survey framework (my own calculation, based on CBS, 2017).

The multivariate test used in this study is the Heckman’s probit regression (Lancee and Bol, 2017; Yamada, 2017). This type of the regression integrates two models in which one is a “consequence” of another. The assumption of the current dataset is that only those who use the Internet may (or may not) use e-government services. Since such assumption can lead to a sample selectivity bias when treating by separate models (Yamada, 2017), it should be controlled using an appropriate method.

Dependent variables

Use of the Internet

The respondents were asked to mention if they had used the Internet via PC or mobile telephone during the 3 months prior to the survey. Those who had not used the Internet represent an omitted category.

Use of e-government

The respondents were asked to mention if they had used the Internet to obtain services from government agencies, such as downloading or filling the forms, approvals or certificates, during the 3 months prior to the survey. Those who had not used the Internet for these purposes represent an omitted category.

Independent variables

Ethnicity

Similar to Mesch (2016), this variable is computed using several existing variables in a database. Four main populations were of interest: Israeli Jews, Arabs, FSU immigrants and non-FSU immigrants. However, since not only ethnicity or immigration status but also the spatial residence is of interest, the Arab population was subdivided by size of its localities: small (rural or up to 50000 residents) and large (more than 50000 inhabitants). The distinction between small and large localities is acknowledged in sociological studies (Haberfeld and Cohen, 2007; Lewin-Epstein and Semyonov, 1992).

Age

Following the existing knowledge about a non-linear effect of age on the use of online services (Dodel, 2016), age was recoded into four categories: 20–29 (omitted), 30–44, 45–59 and 60 + .

Education

The original “highest diploma received” variable was recoded into a dichotomous one, with individuals obtained non-academic education as a reference category.

Gender

The original variable was recoded into a dichotomous one, with female users omitted.

Hebrew proficiency

Studies in sociology of migration in Israel (Amit, 2018; Raijman et al., 2015) assess this variable using three abilities: reading, writing and speaking Hebrew. However, since the main type of proficiency required in use of e-government is understanding the written text, only reading ability is used in order to measure this variable in a current study. The scale of the original item was inversed, so that “1” means no proficiency at all while “5” means very good proficiency.

Frequency of Internet use

Originally, this variable had three categories: “Every day or almost every day,” “Once or twice a week” and “Less than once a week.” It was recoded into a dichotomous variable, with those who use Internet on an everyday basis (“heavy users”) omitted.

Downloading files—a dichotomous variable. Those that did not perform this action during 3 months prior to survey were omitted.

Results

Descriptive results

Table 1 presents the general description of the sample.

Table 1.

Sample statistics (in percents) of Internet and e-government users and non-users by independent variables.

Variables	Internet		E-government
Variables	Users	Non-users	Users	Non-users
Internet users	–	–	43.1	56.9
Gender
Male	49.8	46.6	51.8	48.3
Female	50.2	53.4	48.2	51.7
Education
Academic	41.1	17.5	59.9	40.1
Non-academic	58.9	82.5	35.4	64.6
Age
20–29	25.2	10.8	21.7	27.9
30–44	34.3	16.9	40.0	30.0
45–59	22.7	23.5	24.5	21.2
60+	17.8	48.8	13.7	13.7
Reading in Hebrew
Not at all	1.7	18.6	0.3	2.8
Small extent	4.2	10.6	0.9	6.7
Medium extent	5.9	8.8	3.4	7.8
High extent	11.4	16.7	7.7	14.1
Very high extent	76.9	45.4	87.8	68.6
Frequency of use
“Light”	9.0	–	3.2	13.5
“Heavy”	91.0	–	96.8	86.5
Downloading files
Yes	67.3	–	78.4	58.8
No	32.7	–	21.6	41.2

Source: Israel’s Social Survey 2015.

As seen, the differences between groups can already be seen on the descriptive level. A higher relative percent of individuals with academic education use both the Internet (93.2%) and e-government (59.9%) as compared to individuals with non-academic education (80.6% and 35.4%, respectively). Older individuals use the online space less than younger ones (Internet: 55.3% among 60 +-aged as opposed to 88.8% among 20–29-aged; e-government: 33.1 among the former as opposed to 37.1% among the latter). Users with a higher Hebrew reading proficiency use online space more than those with lower ability (Internet: 85.2% among those with a very high level of reading as opposed to 23.5% among non-readers; e-government: 49.2% among the former as opposed to 6.5% among the latter). In addition, higher relative percent of “heavy” users (about 46%) as compared to that of “light” users (15%) uses e-government. Finally, there is a higher relative percent of e-government users among individuals that download files (50.2%) as opposed to those that do not download them (28.2%).

Bivariate analyses results

Before discussing the results of the multivariate analyses—it is important to understand how the variables are correlated without controlling for other variables. Therefore, several Chi-square and Pearson correlation tests were performed. The results are presented in Table 2.

Table 2.

Results of the cross tabulation and Chi-square test on differences between categories of ethnic origin in regards to the Internet and e-government use.

Category	Internet			E-government
Category	Out of users	Users	Non-users	Out of users	Users	Non-users
Not differentiating between Arabs
Jewish	64.0	85.6	14.4	72.3	49.7	50.3
Arab	14.4	62.4	37.6	8.8	26.9	73.1
FSU	17.0	80.6	19.4	14.7	38.0	62.0
Non-FSU	4.7	82.0	18.0	4.2	39.0	61.0
Chi-square	287.80**			135.19**
Cramer’s V	0.22			0.17
Differentiating between Arabs
Jewish	64.0	85.6	14.4	72.3	49.7	50.3
Arab—large localities	2.8	65.2	34.8	1.6	25.2	74.8
Arab—small localities	11.6	61.7	38.3	7.2	27.3	72.7
FSU	17.0	80.6	19.4	14.7	38.0	62.0
Non-FSU	4.7	82.0	18.0	4.2	39.0	61.0
Chi-square	289.08**			135.38**
Cramer’s V	0.22			0.17

FSU: former Soviet Union.

p < .01.

The most important finding here is a highly significant but relatively weak correlation between the ethnicity and both Internet and e-government use. As seen, most of the Internet (64%) and e-government (72.3%) users are Israeli-born Jews with FSU immigrants ranked second in terms of the e-government use. The Arab population has a low percent of the e-government users (only about 9% of users are Arabs), but not the lowest. Thus, when the ethnic origin variable is treated without decomposing the Arab population by size of locality, there is a clear evidence of ethnic stratification, whereas the majority group is on the top of the stratification system.

As seen, the decoupling of Arabs living in large localities from those who live in small ones complicates the picture. The percentage of Arab Internet and e-government users living in small localities is much higher than that of their counterparts residing in large localities. Thus, in this case, we find evidence for both the stratification (relatively to the Jewish and FSU populations) and diversification (smaller as opposed to larger localities) hypotheses.

In the next stage, Pearson correlations were performed between the rest of the variables and e-government use. The results are provided in Table 3.

Table 3.

Results of Pearson correlation test between the situational variables, control variables and the dependent variables.

	M	Ac	Age	RH	LU	D	IU	EGU
Male	1.0	−.05**	−.04**	.08**	−.01	.04**	.03*	.03*
Academic		1.0	.13**	.02	−.11**	.06**	.17**	.24**
Age			1.0	−.26**	.09**	−.19**	−.30**	−.02
Reading Hebrew				1.0	−.09**	.04**	.35**	.23**
Light users					1.0	−.19**	–	−.18**
Download						1.0	–	.21**
Internet use							1.0	–
E-government use								1.0

p < .05, **p < .01.

As seen, most of the results support the findings from the previous studies. Male users use both Internet (r = .03, p < .01) and e-government (r = .03, p < .01) more than female as is the case of people with academic education (as opposed to those with a non-academic one) (r = .17, p < .01 and r = .24, p < .01, respectively). Age is negatively correlated with Internet use (r = –.3, p < .01) but is not correlated with the e-government use (r = –.02, n.s.). This will not prevent entering this variable into a multivariate analysis since the literature indicates a non-linear correlation between these variables (Dodel, 2016). In addition, higher levels of reading in Hebrew are positively correlated with both Internet (r = .35, p < .01) and e-government use (r = .23, p < .01). Furthermore, “light” users have a lower probability to use e-government as opposed to “heavy” ones (r =–.18, p < .01). Finally, those that download content from the Internet also use the e-government more often as opposed to those that do not download it (r = .21, p < .01).

Multivariate analysis results

Although bivariate results are valuable and interesting, they are limited since they do not control for other variables that may challenge the findings (Mesch and Talmud, 2006). Therefore, in order to estimate the probabilities of using the e-government, the Heckman’s probit regression analysis was performed. In Table 4, the first model assessed the Internet use while the second model assessed the e-government use. In all of the tables, the result of the Spearman test on sample selectivity is non-significant, meaning there was no selectivity bias found.

Table 4.

Coefficients (standard errors) and odds ratios of the Heckman probit regression predicting the probability of using the Internet and e-government services.

	Model 1 (Internet)	Model 2 (E-government)
Not distinguishing between Arabs
Ethnicity (Israeli Jews = 1)	0.24** (0.07)	0.35** (0.08)
Ethnicity (FSU = 1)	0.70** (0.10)	0.22 (0.13)
Ethnicity (non-FSU = 1)	0.41** (0.13)	0.24 (0.13)
Gender (Male = 1)	0.13** (0.05)	0.10* (0.05)
Education (Academic degree = 1)	0.70** (0.06)	0.52** (0.11)
Age (30–44)	−0.20** (0.07)	0.23** (0.06)
Age (45–59)	−0.55** (0.07)	0.29** (0.09)
Age (60 +)	−0.99** (0.08)	0.02 (0.17)
Reading in Hebrew	0.32** (0.03)	0.28** (0.07)
Frequency of Internet use (Light users = 1)	–	−0.55** (0.10)
Viewing or downloading content online (yes = 1)	–	0.49* (0.05)
Constant	−0.40 (0.16)	−2.38** (0.51)
Distinguishing between Arabs, comparing to Arabs from large localities
Ethnicity (Israeli Jews = 1)	0.12 (0.14)	.49** (0.14)
Ethnicity (FSU = 1)	0.58** (0.15)	.36* (0.17)
Ethnicity (non-FSU = 1)	0.30 (0.17)	.38* (0.17)
Ethnicity (Arabs from small localities = 1)	−0.15 (0.15)	.18 (0.14)
Gender (Male = 1)	0.13** (0.05)	.10* (0.05)
Education (Academic degree = 1)	0.70** (0.06)	.51** (0.10)
Age (30–44)	−0.21** (0.07)	.23** (0.06)
Age (45–59)	−0.55** (0.07)	.30** (0.08)
Age (60 +)	−0.99** (0.08)	.03 (0.07)
Reading in Hebrew	0.32** (0.03)	.28** (0.07)
Frequency of Internet use (Light users = 1)	–	−.55** (0.10)
Viewing or downloading content online (yes = 1)	–	.49** (0.05)
Constant	−0.28 (0.19)	−2.50** (0.49)
Distinguishing between Arabs, comparing to Arabs from small localities
Ethnicity (Israeli Jews = 1)	0.27** (0.08)	.31** (0.09)
Ethnicity (FSU = 1)	0.73** (0.10)	.18 (0.13)
Ethnicity (non-FSU = 1)	0.45** (0.13)	.20 (0.13)
Ethnicity (Arabs from large localities = 1)	0.15 (0.15)	−.18 (0.14)
Gender (Male = 1)	0.13** (0.05)	.10* (0.05)
Education (Academic degree = 1)	0.70** (0.06)	.51** (0.10)
Age (30–44)	−0.21** (0.07)	.23** (0.06)
Age (45–59)	−0.55** (0.07)	.30** (0.08)
Age (60 +)	−0.99** (0.08)	.03 (0.16)
Reading in Hebrew	0.32** (0.03)	.28** (0.07)
Frequency of Internet use (Light users = 1)	–	−.55** (0.10)
Viewing or downloading content online (yes = 1)	–	.49** (0.05)
Constant	−0.44 (0.16)	−2.32** (0.48)

p < .05 **p < .01.

From the table, it can be seen that all of the associations that were significant in the bivariate analysis are also significant controlling for other variables in the multiple equation. As to the first model, all of the coefficients are significant. This is not the case in the second model. Here, the main finding is that Israeli Jews have a higher probability to use the e-government as compared to Arabs (b = 0.35, p < .01). The rest of the ethnic origin coefficients are non-significant, thus partially supporting the Hypothesis 1. The rest of the associations are as same as on the bivariate level. The finding for age, as expected, supports the non-linearity assumption.

When decomposing the Arab population and comparing the ethnic groups to the Arabs living in large localities, a clear stratification system can be observed. Both Israeli Jews (b = .49, p < .01), FSU (b = .36, p < .05) and non-FSU immigrants (b = .38, p < .05) have a higher probability to use e-government as compared to the reference group. There is no difference between the two Arab subgroups.

When treating the Arabs from small localities as a reference group, the picture becomes different. Similar to the second model in the Table 4, only Israeli Jews have a higher probability to use the e-government services more than the reference group (b = 0.31, p < .01). Therefore, the results shown in Table 4 provide partial support for the Hypothesis 2.

Figure 1 presents the predicted probabilities for ethnic groups in regards to e-government use: It can be seen that the predicted probability of the disadvantaged minority varies according to the size of locality, thus providing additional support for the Hypothesis 2.

Figure 1.

Predicted probabilities of the e-government use by origin.

Discussion

This study was designed to examine the effects of the ethnic origin and size of locality on use of the Internet to obtain services from government agencies, thereby contributing to the research in two ways. First, it increases understanding of the ethnic structure of e-government use in the deeply divided society. Second, it enables the assessing of the effect of size of locality of the disadvantaged minorities on e-government use. The results partially support the study hypotheses.

First, regarding Hypothesis 1, it can be seen that the groups do differ by ethnic layer in terms of the e-government use. When treating Arabs as a single unit, the only significant difference found is between them and the Israeli Jews. Thus, the Hypothesis 1 is partially confirmed, providing a support to the stratification approach. Similar to health services (Mesch, 2016), the use of e-government services are stratified according to the access to social capital. Thus, the “offline” stratification system in the Israeli society is transferred into the online space, at least in terms of health and e-government services.

Second, Hypothesis 2 is also partially supported. Although both Table 2 and Figure 1 show that Arabs from small localities have a higher probability of using e-government as compared to those living in large ones, the regression analysis did not yield a significant difference. However, the multivariate results also show that the structure of differences between each of the Arab population categories and the rest of ethnic groups varies. These results provide a partial justification for the extension of the social inequality approaches by including a size of locality as discriminant factor alongside with the ethnicity. This means that in order to better understand the structure of use of the specific Internet domain, the disadvantaged population should be decomposed and treated by size of localities in which it lives.

Regarding the effects of other socio-economic variables, some aspects should be mentioned. Education remains one of the key factors affecting e-government use. This supports similar findings in the e-government literature (Carter and Weerakkody, 2008). Regarding age, it can be seen that those aged 20–29 have a smaller probability of using the e-government services as compared to all other age groups (except 60 +), corresponding to the findings of the Uruguayan study (Dodel, 2016). Interestingly, no gender effect was found, contradicting the literature to a certain extent (Al-Shafi and Weerakkody, 2010; Patel and Jacobson, 2008). This may signal closing the “second-level” gender divide in terms of e-government use in Israel.

The effects of the situational variables support the existing theoretical claims. Less frequent Internet users have a lower probability of using e-government, thus reflecting the lower level of online experience and familiarity with the Internet applications (Welch and Hinnant, 2003). In addition, downloading files increases the probability to use e-government.

In summary, this study adds to the thin volume of literature dealing with the differences in e-government use according to socio-economic background. It validates the results of the studies of Gauld et al. (2010) and Dodel (2016). The findings of the current study emphasize the role of the socio-economic background in explaining the use of e-government services by showing the presence of advantaged and disadvantaged ethnic groups in terms of access to social capital. Similar to Gauld et al. (2010), it also emphasizes the type of locality as an important factor differentiating between e-government users and non-users. This shows that the digital divide in e-government use is still here and should be studied and understood in greater depth.

Beyond the theoretical contribution, the current study has also an implication for policy. Public sector officials should pay attention to the findings of the study and strive to make the e-government more accessible to minorities, especially those residing in remote areas, in order to apply the principles of transparency and democratic participation (Bertot et al., 2010; Norris and Reddick, 2003). In addition, governmental bodies responsible for the ICT infrastructure should concentrate their attempts to improve this infrastructure in the localities in which disadvantaged populations reside.

Limitations and future directions

This study is not without limitations. First, the e-government use was represented by only one variable, which only assessed if individuals used or did not use the Internet to obtain government services. The use of e-government may include different actions, including searching for information or performing transactions, whereas each action may be presented by a separate variable (for example, see Al-Adawi et al., 2005). Thus, future studies should assess e-government use with more than one variable. Second, the item that had been employed to assess the e-government use had no explanation on which governmental agencies may provide these services—whether these are ministries or other state authorities (such as Bituach Leumi [Israeli national institution of social insurance]).

Summary

This study was held in order to test the social inequality hypotheses and extend them by decomposing the disadvantaged population. The results provide a partial support for the Stratification Hypothesis (Mesch, 2016). In addition, they provide partial justification for extension of the social inequality hypotheses by adding size of locality as a differentiating factor together with ethnicity. Future studies should continue this trend, testing the extended social inequality hypotheses in other countries or other domains of the Internet use.

Footnotes

Acknowledgements

A previous version of the article was presented on the Stratification Seminar, Department of Sociology, 2017. The author would like to thank Prof. Meir Yaish, Prof. Rebeca Raijman, Dr. Asaf Levanon, Prof. Vered Kraus, Prof. Ilan Talmud, Dr. Alisa Lewin (Department of Sociology, University of Haifa), and Prof. Rita Mano (Department of Human Services, University of Haifa) for their useful comments and suggestions.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author biography

Dennis Rosenberg is a PhD candidate at the department of Sociology and Antropology at University of Haifa.

References

Abbas

Mesch

(2016) Do rich teens get richer? Facebook use and the link between offline and online social capital among Palestinian youth in Israel. Information, Communication & Society 21(1): 63–79.

Al-Adawi

Yousafzai

Pallister

(2005) Conceptual model of citizen adoption of e-government. In: Simone

Chandru

Nathan

(ed.) The second international conference on innovations in informational technology (IIT’05), Dubai, UAE, 26–28 September, pp. 1–10.

AlAwadhi

Morris

(2009) Factors influencing the adoption of e-government services. Journal of Software 4(6): 584–590.

Al-Haj

(2004) Immigration and Ethnic Formation in a Deeply Divided Society: The Case of the 1990S Immigrants from the Former Soviet Union in Israel. Leiden: Brill, pp. 1–246.

Al-Shafi

Weerakkody

(2010) Factors affecting e-government adoption in the State of Qatar. In: European, Mediterranean and Middle Eastern conference on information systems (EMCIS 2010) (eds Irani

Ghoneim

Ali

et al .), Abu-Dubai, UAE, 12–13 April, pp. 1–23. London: Brunel University.

Almarabeh

(2011) Knowledge management of e-government in Jordan. International Journal of Advanced Corporative Learning 4(4): 4–9.

Amit

(2018) Identity, belonging and intention to leave of first and 1.5 generation FSU immigrants in Israel. Social Indicators Research 139(3): 1219–1235.

Bar-Haim

Semyonov

(2015) Ethnic stratification in Israel. In: Sáenz

Embrick

Rodríguez

(eds) The International Handbook of Demography of Race and Ethnicity. Dordrecht: Springer, pp. 323–337.

Basu

(2004) E-government and developing countries: an overview. International Review of Law Computers & Technology 18(1): 109–132.

10.

Belanger

Carter

(2008) Trust and risk in e-government adoption. The Journal of Strategic Information Systems 17(2): 165–176.

11.

Bertot

Jaeger

Grimes

(2010) Using ICT to create a culture of transparency: e-government and social media as openness and anti-corruption tools for societies. Government Information Quarterly 27(3): 264–271.

12.

Carter

Belanger

(2004) Citizen adoption of electronic government initiatives. In: Proceedings of the 37th annual Hawaii international conference on system sciences (HICSS’04), Big Island, HI, 5–8 January, pp. 1–10. Washington, DC: IEEE.

13.

Carter

Belanger

(2005) The utilization of e-government services: citizen trust, innovation and acceptance factors. Information Systems Journal 15(1): 5–25.

14.

Carter

Weerakkody

(2008) E-government adoption: a cultural comparison. Information Systems Frontiers 10(4): 473–482.

15.

Castells

(2010) The Rise of the Network Society: The Information Age: Economy, Society, and Culture. 2nd ed. Chichester: John Wiley & Sons.

16.

Chen

Rashty

(2002) E-Government in Israel: The Five Layers Model of E-Government. Jerusalem, Israel: Ministry of Finance, General Accountant Office.

17.

Chen

Y-C

Gant

(2001) Transforming local e-government services: the use of application service providers. Government Information Quarterly 18: 343–355.

18.

Di Gennaro

Dutton

(2006) The internet and the public: online and offline political participation in the United Kingdom. Parliamentary Affairs 59(2): 299–313.

19.

Dodel

(2016) E-government’s hidden inequality: why spending on online services can be regressive and how to avoid it. In: Proceedings of the 9th international conference on theory and practice of electronic governance (ICEGOV ’15–16), Hammamet, Tunisia, 1–3 March, pp. 68–74. New York: ACM Press.

20.

Edmiston

(2003) State and local e-government: prospects and challenges. The American Review of Public Administration 33(1): 20–45.

21.

Gauld

Goldfinch

Horsburgh

(2010) Do they want it? The “demand side” of e-government in Australia and New Zealand. Government Information Quarterly 27(2): 177–186.

22.

Goldshmidt

(2012) E-Government in Local Authority. Jerusalem, Israel: Knesset Research and Information Center (in Hebrew).

23.

Haberfeld

Cohen

(2007) Gender, ethnic, and national earnings gaps in Israel: the role of rising inequality. Social Science Research 36(2): 654–672.

24.

Israel Central Bureau of Statistics (CBS) (2015) The 2015 social survey. Received by request from Israel Social Science Data Center (ISDC) at the Hebrew University. Available at: http://isdc.huji.ac.il

25.

Israel Central Bureau of Statistics (CBS) (2017) The 2015 social survey: response and non-response rates, by reason (in Hebrew). Available at: http://www.cbs.gov.il/publications17/seker_hevrati15_1670/pdf/table_h.pdf

26.

Israel Central Bureau of Statistics (CBS) (n.d.) The essence of the survey its objectives Available at: http://www.cbs.gov.il/www/skarim/social_surv/what_survey_e.pdf

27.

Lancee

Bol

(2017) The transferability of skills and degrees: why the place of education affects immigrant earnings. Social Forces 96(2): 691–716.

28.

Lewin

Stier

(2018) Material status, gender, and material hardship: evidence from Israel. Advances in Life Course Research 35: 46–56.

29.

Lewin-Epstein

Semyonov

(1992) Local labor markets, ethnic segregation, and income inequality. Social Forces 70(4): 1101–1119.

30.

Lin

Fofanah

Liang

(2011) Assessing citizen adoption of e-government initiatives in Gambia: a validation of the technology acceptance model in information systems success. Government Information Quarterly 28(2): 271–279.

31.

Mahadeo

(2009) Towards an understanding of the factors influencing the acceptance and diffusion of e-government services. Electronic Journal of E-Government 7(4): 391–402.

32.

Mesch

(2003) Language proficiency among new immigrants: the role of human capital and societal conditions: the case of immigrants from the FSU in Israel. Sociological Perspectives 46(1): 41–58.

33.

Mesch

(2012) Minority status and the use of computer-mediated communication: a test of the social diversification hypothesis. Communication Research 39(3): 317–337.

34.

Mesch

(2016) Ethnic origin and access to electronic health services. Health Informatics Journal 22(4): 791–803.

35.

Mesch

Talmud

(2006) The quality of online and offline relationships: the role of multiplexity and duration of social relationships. The Information Society 22(3): 137–148.

36.

Mesch

Talmud

(2011) Ethnic differences in internet access. Information, Communication & Society 14(4): 445–471.

37.

Mesch

Mano

Tsamir

(2012) Minority status and health information search: a test of the diversification hypothesis. Social Science & Medicine 75(5): 854–858.

38.

Norris

Reddick

(2003) Local e-government in the United States: transformation or incremental change? Public Administration Review 73(1): 165–175.

39.

Palvia

SCJ

Sharma

(2007) E-government and e-governance: definitions/domain framework and status around the world. In: Foundations of e-government (eds Agarwal

Ramana

), Hyderabad, India, 28–30 December, pp. 1–12. Hyderabad, India: IECG.

40.

Patel

Jacobson

(2008) Factors influencing citizen adoption of e-government: a review and critical assessment. In: Proceedings of 16th European conference on information systems (ECIS 2008) (eds Golden

Acton

Conboy

et al .), Galway, 9–11 June, pp. 1058–1069. AIS.

41.

Phang

Sutanto

et al . (2005) Senior citizens’ adoption of e-government: in quest of the antecedents of perceived usefulness. In: Proceedings of the 38th annual Hawaii international conference on system sciences (HICSS’05), Big Island, HI, 3–6 January, pp. 130a. Washington, DC: IEEE Computer Society.

42.

Raijman

Semyonov

Geffen

(2015) Language proficiency among post-1990 immigrants in Israel. Journal of Ethnic and Migration Studies 41(8): 1347–1371.

43.

Reddick

(2005) Citizen Interaction with e-government: from the streets to servers? Government Information Quarterly 22(1): 38–57.

44.

Rotem

(2007) E-government in Israeli local authorities: comparative analysis (in Hebrew). Available at: https://en-social-sciences.m.tau.ac.il/sites/socsci-english.tau.ac.il/files/media_server/government/pdf2/2007/6-%20EGov.pdf

45.

Saabneh

(2015) Arab-Jewish gap in life expectancy in Israel. The European Journal of Public Health 26(3): 433–438.

46.

Sang

Lee

J-D

Lee

(2009) E-government adoption in ASEAN: the case of Cambodia. Internet Research 19(5): 517–534.

47.

Tolbert

Mossberger

(2006) The effects of e-government on trust and confidence in government. Public Administration Review 66(3): 354–369.

48.

United Nations (2014) United Nations E-government Survey 2014: e-government for the future we want. Available at: https://publicadministration.un.org/egovkb/portals/egovkb/documents/un/2014-survey/e-gov_complete_survey-2014.pdf

49.

Van Deursen

AJAM

Helsper

(2015) The third-level digital divide: who benefits most from being online. In: Robinson

Cotten

Schulz

et al . (eds) Communication and Information Technologies Annual. Bingley: Emerald Group Publishing Ltd., pp. 29–52.

50.

Van Deursen

AJAM

Vanan Dijk

JAGM

Ten Klooster

(2015) Increased inequalities in what we do online: a longitudinal cross sectional analysis of internet activities among the Dutch population (2010 to 2013) over gender, age, education, and income. Telematics & Informatics 32(2): 259–272.

51.

Wangpipatwong

Chutimaskul

Papasratorn

(2005) Factors influencing the adoption of Thai e-government websites: Information quality and system quality approach. In: Proceedings of the fourth international conference on e-business 2005, Bangkok, Thailand, 19–20 November, pp. 19–20. Bangkok, Thailand: Assumption University Press.

52.

Welch

Hinnant

(2003) Internet use, transparency and interactivity effect on trust in government. In: Proceedings of the 36th Hawaii international conference on system sciences (HICSS’03), Big Island, HI, 6–9 January, p. 7. Washington, DC: IEEE Computer Society.

53.

West

(2004) E-government and the transformation of service delivery and citizen attitudes. Public Administration Review 64(1): 15–27.

54.

Woolley

Peterson

(2012) Efficacy of a health-related Facebook social networking site on health-seeking behaviors. Social Marketing Quarterly 18(1): 29–39.

55.

Yamada

(2017) Migration, education investment and gender in transition: evidence from Tajikistan in the 2000s. Available at: https://ssrn.com/abstract=2912084