Determinants of Internet-enabled Transparency at the Local Level

Abstract

The purpose of this study is to examine the premise that increased Internet access leads to higher levels of governmental transparency. In so doing, we assess the impact of geographic, demographic, socioeconomic, and institutional factors on governmental transparency in 816 out of 1055 counties in the twelve Midwestern states. The findings from the most extensive ordered logistic model show that total land area, population density, percentage of minority population, educational attainment, and the council–manager form of government are statistically associated with higher levels of government transparency at the county level via the Internet.

Keywords

transparency local government county Web site Internet access

Introduction

Transparency promotes accountability and provides information for citizens about what their Government is doing. Information maintained by the Federal Government is a national asset. My Administration will take appropriate action, consistent with law and policy, to disclose information rapidly in forms that the public can readily find and use. (President Obama 2009)

Since the 1980s, transparency has been argued by many to be a key element in good governance (Bonsón et al. 2012; Hood 2006; Piotrowski and Ryzin 2007) as it provides citizens a greater opportunity to have control over their government, which may result in a reduction in government corruption, bribery, and other malfeasance (Conradie and Choenni 2014), which consequently contributes to greater knowledge and societal progress (Schauer 2011). Attention to the concept of transparency appears to have been incorporated with the emergence and proliferation of new information technology (IT) and a change in the environment with respect to the fact that the Internet has been adopted as one of several means to realize and achieve greater openness with regard to governmental activities (Bonsón et al. 2012; Margetts 2011).

The Internet has opened a new arena for interaction between governments and citizens, as it not only provides more efficient and cooperative interaction, more efficient service delivery, and more efficient transaction activities (Moon 2002) but also leads to what might be called “Internet-enabled transparency” (Margetts 2011, 518) or “computer-mediated transparency” (Grimmelikhuijsen and Welch 2012, 562). This venue provides citizens with far greater potential to observe and understand what is going on in the government as well as improves the boundaries between citizens and state and opens up governmental processes for greater scrutiny. As access to the Internet increases over time, governments are expected to have more information and readily provide services on their Web sites (Harder and Jordan 2013; Manoharan 2013a; Pina, Torres, and Royo 2010).

Despite globally accepted merits and increased interests, the study of transparency, as a subfield of public affairs, remains relatively new, especially with respect to the Internet or e-government practice (Harder and Jordan 2013; Piotrowski and Ryzin 2007). There is extensive work aiming to materialize the theory of transparency, especially in terms of various dimensions and types of governmental transparency (e.g., Hood and Heald 2006). Still, the literature needs further development with regard to the contexts in which transparency is applied, factors contributing to transparency, and transparency at the local level (Grimmelikhuijsen and Welch 2012). In fact, since the early 2000s, a few public affairs scholars have devoted their efforts to determining the impact of factors influencing the use and development of e-government technologies (England, Pelissero, and Morgan 2011). For example, early studies by Ho (2002) and Moon (2002) provide a good foundation for succeeding studies. In subsequent years, quite a few studies have examined Internet-enabled transparency by focusing on transparency at the international level (e.g., Bonsón et al. 2012; Pina, Torres, and Royo 2007, 2010), state level (e.g., Justice and McNutt 2013; Menifield and Clay 2010), and municipal level (e.g., Borry 2012; del Sol 2013; Welch 2012). The first study of e-government practice at the county level was published in 2007 (Huang 2007). Only recently have studies appeared attempting to account for Web sites as a tool for governmental transparency at the county level (Harder and Jordan 2013; Bernick et al. 2014).

The purpose of this study is to add to the literature by examining factors that affect Web site transparency at the county level in the Midwest region (United States). More specifically, this study tests the premise that increased Internet access leads to higher levels of governmental transparency. We first discuss the concept of governmental transparency and the findings obtained from research studies at the local level. Then, we discuss our analytical methods and data. In the final sections, we articulate our findings and discuss how the findings support and contrast existing studies.

Literature Review

Defining Transparency

For decades, transparency has been argued to be an essential component of good governance and “has attained quasi-religious significance” (Hood 2006, 3). Despite universally accepted merits, there is anything but a commonly accepted definition of transparency, partly because transparency is multidimensional and can take many forms. Heald (2006) argues that transparency can be viewed in different perspectives: upward, downward, outward, and inward. Upward transparency refers to the principal–agent relationship, where the principal can observe the agent’s performance. Downward transparency takes place when the subordinate or population can observe the conduct of the superordinate or government. Outward transparency occurs when those inside the organization can observe what is happening outside the organization. Inward transparency takes place when those outside the organizational can observe the organization’s performance internally (Heald 2006).

Margetts (2011, 518) also argues that the term transparency can be perceived through various points of view, including “openness, surveillance, accountability, simplicity and notions of rule-governed, predictable governance processes … fairness, and efficiency.” For instance, whereas Piotrowski and Ryzin (2007, 308) define transparency as “the ability to find out what is going on inside a public sector organization through a wide range of avenues,” Grimmelikhuijsen and Welch (2012, 563) assert that transparency is “the disclosure of information by an organization that enables external actors to monitor and assess its internal workings and performance.” These definitions imply that transparency can be understood from two interdependent perspectives: an outside-in perspective, focusing on the seeking role of external entities, and an inside-out perspective, highlighting the responsibility of the organization to inform and open up. This study focuses on the inform and open up perspective of transparency, as it sees the Internet or a government Web site as a tool that enhances the government’s provision of information and services, at the same time, allows the population to observe inward.

The Internet, Transparency, and Local Government

Over the past decade, the percentage of U.S. households with Internet access has increased dramatically from 54.7 percent in 2003 (Day, Janus, and Davis 2005) to 74.4 percent in 2013 (File and Ryan 2014). As the population’s access to the Internet has increased, governments are expected to have information and services readily provided on their Web sites (Harder and Jordan 2013; Pina, Torres, and Royo 2010). Much of the current literature examining government Web sites focused on trust in government (Grimmelikhuijsen 2012; Grimmelikhuijsen and Meijer 2014; Grimmelikhuijsen et al. 2013; S. Kim and Lee 2012; Welch, Hinnant, and Moon 2005). This study, however, focused on research that examined variables that influence levels of government Web site transparency. This is important as the use of government Web sites has dramatically increased over time. Despite their efforts to do so, many governments have not kept pace with demand as information and services offered on their Web sites are fairly stagnant, partially contributing to less governmental transparency (England, Pelissero, and Morgan 2011).

Some authors have argued that there had been few attempts to identify the determinants of individuals’ demand for transparency due to two factors: “the subfield of transparency has just recently gained momentum, and the concept of demand for transparency is difficult to measure” (Piotrowski and Ryzin 2007, 307). In fact, there have been a number of studies assessing the influence of various factors on the availability and accessibility of services and information on local government Web sites—from structural, demographic, socioeconomic, to organizational and individual perspectives (England, Pelissero, and Morgan 2011, 23). However, the majority of these investigations were examined at the municipal level. This section reviews early studies of Internet-enabled transparency at the local level and recent studies that specifically focused on the county level.

Ho (2002) examined the relationship between socioeconomic and organizational factors and cities’ progressiveness in Web development. He hypothesized that “cities with larger minority populations and a lower per capita income are less likely to adopt progressive web design because there may be insignificant citizen demand for web-based services” (Ho 2002, 439). The author found that the minority population and average years of operating the Web sites had an impact on Web site development. By contrast, the total population and per capita income did not have a significant relationship with the progressiveness of city government’s efforts in Web site development.

Moon (2002) examined the implementation and effectiveness of e-government at the municipal level and assessed the impact of size and type of municipal governments on the adoption of e-government. Based on a survey conducted in 2000 by the International City/County Management Association (ICMA) and Public Technologies Inc., the findings indicated that the size of a municipality and the adoption of e-government were positively related: municipalities with larger populations tended to have their own Web sites and were more likely to be early adopters of e-government practices than municipalities with smaller populations. The findings also suggested that the type of municipal government had an impact on the adoption of e-government practices: the council–manager governments were likely to adopt Web technologies faster and were more proactive in employing e-government than their mayor–council counterparts (Moon 2002).

Piotrowski and Ryzin (2007) examined the public’s demand for transparency at the local level by first identifying various characteristics of citizens related to the demand for governmental transparency. The characteristics that they identified include (1) demographic attributes—that is, sex, race, education attainment, income, and age; (2) a number of general political attitudes and orientations; and (3) some other contextual correlates. Using the data derived from a 2005 online survey, the authors conducted an exploratory factor analysis which reduced the transparency items into five scales, that is, fiscal transparency, safety transparency, principled transparency, good government transparency, and the amount of documents obtained in the last few years. The authors found that people who perceived the government as sufficiently open demanded less transparency, whereas those who saw the government as closed and those who regularly contacted the government tended to demand more transparency. Citizens’ confidence in government officials had an inverse relationship with their demand for transparency. While conservatives were more concerned with safety transparency, their liberal counterparts were more concerned with principled and good governance transparency. Age was found to have an inverse relationship with obtaining governmental documents, implying that older residents were less likely to use the Internet on a regular basis. Data also showed that race and the type of community (i.e., urban, suburban, or small) were not related to attitudes toward transparency (Piotrowski and Ryzin 2007).

Huang (2007) conducted a comprehensive study of county Web site adoption, examining various socioeconomic factors related to the adoption. The author found that 56.3 percent (1744 out of 3099) of the counties had Web sites. The findings indicated that the total population, the percentage of population change, ethnicity, education, home ownership, median value of housing units, income, private nonfarm employment, and retail sales per capita were statistically associated with the provision of information and service via government Web sites.

Borry (2012) conducted an exploratory study of governmental transparency determinants by analyzing the contents of 566 municipal Web sites in New Jersey. The author specifically focused on council meetings documents available on the Web sites and created five categories of transparency based on the degree to which documents were available online. Multinomial logistic model was used to estimate the effect of a set of independent variables on transparency levels. The findings showed that the total population of an area and the percentage of the rural population were significantly associated with municipalities with no Web sites. When the total population increased, the municipality was more likely to maintain a Web site, and when the percentage of the rural population increased, the municipality was less likely to maintain a Web site. Findings also showed that per capita income was positively associated with higher Web site transparency, whereas the percentage nonwhite population and median age had an inverse relationship with higher Web site transparency.

Manoharan (2013a) conducted a nationwide study of Web site adoption at the county level by examining institutional, contextual, and socioeconomic factors that influenced Web site adoption. Data for the independent variables were collected via an online survey of county Web site administrators. Data for the dependent variables were derived from an evaluation of the county Web site contents based on twenty-four measures, which were each coded on a scale of 0–3 according to the availability of information for certain measures. Multivariate regressions were used to predict county Web site adoption. The findings showed that Web site longevity, number of IT employees, organizational size, collaboration with nonprofit organizations, the total population, and the percentage of the population with a bachelor’s degree were positively associated with Web site adoption. By contrast, form of government (i.e., the commission form), geographic area and location, population density, and per capita income were not significant predictors.

In another study, Manoharan (2013b) examined the data discussed earlier by exclusively focusing on three dimensions of e-government practice: e-information, e-transaction, and e-participation. For the e-information dimension, which closely resembles the Internet-enabled transparency concept, findings for the multivariate regression showed that IT contracting, degree of elected officials’ support, external collaboration with other counties, and the percentage of high school graduates were positively associated with the provision of reliable and sufficient information on the Web site.

Harder and Jordan (2013) conducted a content analysis of county Web sites in Arkansas using transparency indicators constructed from the literature, laws, and organizations that focused on the subject. The authors found that population change and the percentage of high school graduates were a significant predictor of transparency levels in Arkansas. In addition, they found that the percentage of the population 65 years and older, the percentage of the white population, median household income, and the percentage of the population below the poverty level were not significant in explaining levels of county-level Web site transparency.

More recently, Bernick et al. (2014) studied fiscal transparency in 400 counties that were randomly selected from counties with a population of 100,000 or more residents. The authors first evaluated the contents of the Web sites and then categorized fiscal transparency into four categories—from no information to sophisticated information—based on fiscal information provided on the Web sites. Findings from generalized logistic regression showed counties with an appointed manager, more minority residents, and a larger non-English-speaking population were more likely to have higher level of transparency. By contrast, counties with higher economic stress (as measured by poverty level, among others) and an older population were more likely to exhibit lower levels of transparency.

The aforementioned studies examined a wide range of factors influencing the degree of governmental transparency with regard to Web-based information and service provisions. In sum, the factors included demographic characteristics, such as population size and ethnicity; socioeconomic characteristics, such as per capita personal income and educational attainment; and institutional, organizational, and municipal characteristics. As the review of the literature shows, there are both similarities and disagreements among the studies. In order to expand this research area in the United States, this study investigated the importance of Internet access and geographic characteristics as well as other characteristics examined by scholars.

Transparency Hypotheses

With respect to the variables that were found salient in the literature, this study tests the following hypotheses. Table 1 provides a summary of the independent variables and the expected direction of the relationship.

Table 1.

Independent Variables and Expected Relationships.

Variable	Expected relationship	Operationalization of independent variables (source)
Internet access	+ Coefficient	Internet access connections over 200 kilobits per second in at least one direction per 1000 households (FCC 2014)
Land area	− Coefficient	Total land area in square miles (Census Bureau http://www.census.gov/prod/www/decennial.html)
Population density	+ Coefficient	Population density in persons per square mile (Census Bureau http://www.census.gov/prod/www/decennial.html)
Population change	+ Coefficient	Percentage of population change between 2010 and 2000 (Census Bureau http://www.census.gov/prod/www/decennial.html)
Minority	+ Coefficient	Percentage of nonwhite population (Census Bureau http://www.census.gov/data.html)
Age	− Coefficient	Median age of population (Census Bureau http://www.census.gov/data.html)
Education	+ Coefficient	Percentage of population with a bachelor’s degree and beyond (Census Bureau http://www.census.gov/data.html)
Income	+ Coefficient	Per capita personal income (Census Bureau http://www.census.gov/data.html)
Poverty	− Coefficient	Percentage of population below the poverty level (Census Bureau http://www.census.gov/data.html)
Form of government	+ Coefficient	Forms of government as reported in the Municipal Year Book: commission, council–manager, and council-elected executive (ICMA 2010)

Note: FCC = Federal Communications Commission; ICMA = International City/County Management Association.

Internet Access

The previous research argues that Internet access impacts government transparency levels. More specifically, as the percentage of households with access to the Internet increased over time, governments are expected to be more transparent by providing more information and services via their Web sites. Therefore:

Hypothesis 1: Counties with more Internet access are more likely to exhibit a higher degree of governmental Web site transparency.

Since the availability of broadband Internet may vary by geographical area, counties with larger geographical areas may have limited access to the broadband Internet, which limits the capability to provide online information and services. Therefore:

Hypothesis 2: Counties with larger land areas are less likely to exhibit a higher degree of governmental Web site transparency.

The literature examined indicates that the availability of broadband Internet varies from urban to rural communities. More specifically, scholars found that Internet access and availability increases as the total population increases (Garcia-Murillo 2005; J. Kim, Bauer, and Wildman 2003). Therefore:

Hypothesis 3: Counties with higher population density are more likely to exhibit a higher degree of governmental Web site transparency.

Demographic

Population growth was found to be a salient predictor of governmental transparency (Harder and Jordan 2013; Huang 2007; Manoharan 2013a; Moon 2002). Therefore:

Hypothesis 4: Counties that experience population growth are more likely to exhibit a higher degree of governmental Web site transparency.

The presence of a nonwhite population was found to be a statistically significant predictor of governmental transparency (Bernick et al. 2014; Borry 2012; Ho 2002; Huang 2007). Therefore:

Hypothesis 5: Counties with a larger minority population are more likely to exhibit a higher degree of governmental Web site transparency.

Older residents were found to be less likely to use the Internet on a regular basis (Piotrowski and Ryzin 2007), and age was found to have an inverse relationship with higher levels of transparency (Bernick et al. 2014; Borry 2012). Therefore:

Hypothesis 6: Counties with older residents are less likely to exhibit a higher degree of governmental Web site transparency.

Socioeconomic

Educational attainment of the population was found to have a positive relationship with governmental transparency at the local level (Harder and Jordan 2013; Huang 2007; Manoharan 2013a, 2013b). Therefore:

Hypothesis 7: Counties with higher levels of educational attainment are more likely to exhibit a higher degree of governmental Web site transparency.

Income was found to be positively associated with governmental transparency (Borry 2012; Huang 2007). Therefore:

Hypothesis 8: Counties with higher levels of per capita personal income are more likely to exhibit a higher degree of governmental Web site transparency.

Existing studies found that poverty had an inverse relationship on government transparency (Bernick et al. 2014). Therefore:

Hypothesis 9: Counties with higher levels of poverty are more likely to exhibit a lower degree of governmental Web site transparency.

Institutional

Form of local government in general was found to be a statistically significant predictor of technology and innovation adoption (Nelson and Svara 2012). More specifically, municipalities with a council–manager form of government were found to be more proactive in providing information and services on the government Web site (Moon 2002). Therefore:

Hypothesis 10: Counties with a council–manager form of government are more likely to exhibit a higher degree of governmental Web site transparency.

Data and Methods

This study examined the effect of various county government characteristics on Internet-enabled transparency. The aforementioned hypotheses were tested using five sets of county government characteristics: Internet access, geographical, demographic, socioeconomic, and institutional characteristics. For the purpose of this study, the analysis included 816 Midwest counties in twelve U.S. states—Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin (U.S. Census Bureau 2000). The dependent variable is categorical in nature with five categories. Ordered logistic regression models were employed to estimate the influence of these factors on Midwest county Web site transparency.

The measure of county Web site transparency was retrieved from county Web site evaluations conducted by the Sunshine Review. The Sunshine Review was a nonprofit organization established in 2008, with the primary focus of building awareness for and promoting transparency, openness, and accountability among state and local governments (Sunshine Review 2011a). As of January 2012, the Sunshine Review evaluated over 6000 government Web sites, including every county, state, and capital city in the United States (Sunshine Review 2011b). In so doing, they first conducted a survey asking over 100 organizations that focused on transparency to provide “a list of website transparency features that citizens in any part of the United States should be able to find when they visit the websites of counties, cities, school districts, and state agencies” (Sunshine Review 2011c). Based on the survey findings, they developed a set of criteria evaluating county Web sites. The criteria, called “the 10-point transparency checklist,” included ten items: budget, meetings, elected officials, administrative officials, permits and zoning, audits, contracts, lobbying, public records, and local taxes (Sunshine Review 2011c; see http://ballotpedia.org/Transparency_checklist for a list of full definitions and premises). In completing their analysis, a team of Sunshine Review associates identified the appropriate county Web site to review and analyzed the contents of the Web site according to the availability and accessibility of information relative to the checklist. A point was given to each of the ten items if it completely met the criterion and no point was awarded if it partially met the criterion. The points were then aggregated and a letter grade was assigned. Web sites awarded the full ten points were given an A+, 9 = A−, 8 = B, 7 = B−, 6 = C, 5 = C−, 4 = D, 3 = D−, and 0–2 = F (McMurray 2013).

This study used the 2010 county Web site data evaluations conducted by the Sunshine Review. For ease of understanding, this study combined the grades A+ and A−, B+ and B, and so on, thus reducing the number of categories from nine to five. Each of the combined grades were assigned a numerical value: A = 4, B = 3, C = 2, D = 1, and F = 0.

Findings

For meaningful interpretation, summary statistics were presented in two tables. Table 2 included summary statistics for the discrete variables. The dependent variable and the Web site transparency grade were ordinal variables and form of government was a nominal variable. There are 1055 counties in the Midwest region of the United States and 239 of those counties (22.65 percent) had no organizational Web site. This number was similar to that reported in Manoharan (2013b; 23.5 percent), but lower than that reported by Harder and Jordan (2013; 53.33 percent), and higher than that reported by Bernick et al. (2014; 9 percent). To control for the a priori effect of having no Web site, the analysis included only the 816 counties that had a Web site.

Table 2.

Descriptive Statistics of Discrete Variables.

Variable	Frequency	Percentage	Cumulative
Ordinal variable
Web site transparency grade (code)
A (4)	24	2.94	2.94
B (3)	41	5.02	7.96
C (2)	154	18.87	26.83
D (1)	286	35.05	61.88
F (0)	311	38.11	100
Binary variable
Form of government
Council–manager (1)	210	25.74	25.74
Other forms (0)	606	74.26	100

Note: N = 816.

Table 2 shows that a majority of the counties (38.11 percent) provided relatively limited information on the Web site. Only twenty-four (2.94 percent) of all county Web sites provided the most extensive information. Council–manager counties represented about a fourth of the counties and 74.26 percent of the counties had either a commission or a council-elected executive.

Table 3 included summary statistics for the continuous variables in the study. The data in this table showed wide variability in terms of geographic, demographic, and socioeconomic characteristics. For example, the population density data ranged from less than 1 resident per square mile in Sioux County, Nebraska, to more than 5496 residents per square mile in Cook County, Illinois. Similarly, the minority population data ranged from <1 percent in Garfield County, Nebraska, to 89.34 percent in Menominee County, Wisconsin.

Table 3.

Descriptive Statistics of Continuous Variables.

Variable	Mean	Standard deviation	Minimum	Maximum
Internet access and geographic variables
Internet access (×1000)	3.32	0.69	1	5
Total geographical area (square miles)	711.55	471.48	61.89	6245.75
Population density (per square mile)	153.83	434.06	0.63	5496.57
Demographic variables
Population change (%)	1.42	10.74	−22.97	108.29
Minority population (%)	7.94	8.33	0.78	89.34
Median age (years)	41.02	4.70	24.3	55.2
Socioeconomic variables
Bachelor’s degree attainment (%)	19.80	7.31	7.5	53.8
Per capita income (log)	10.02	0.14	9.56	10.58
Poverty (%)	12.41	4.21	3.9	31.8

Note: N = 816.

Table 4 contains the summary data for our regression model. In conducting our analysis, we also determined if we had multicollinearity problems using the variance inflation factor (VIF). The mean VIF score was 1.53 (minimum 1.08 and maximum 1.94). In general, a VIF larger than 10 provides evidence for concern (O’brien 2007). A test for proportional odds assumption (Brant test) showed that the assumption was not violated, and therefore the ordered logistic model was applicable.

Table 4.

Ordered Logistic Regression Analysis Results.

Independent Variables	Model 1		Model 2		Model 3		Model 4
Independent Variables	Coefficients	Odds ratio	Coefficients	Odds ratio	Coefficients	Odds ratio	Coefficients	Odds ratio
Internet access	0.560*** (0.103)	1.751			−0.029 (0.118)	0.972	0.031 (0.126)	1.032
Land area	0.000 (0.000)	1.000			0.000 (0.000)	1.000	0.000** (0.000)	1.000
Population density	0.002*** (0.000)	1.002			0.001*** (0.000)	1.001	0.001*** (0.000)	1.001
Population change			0.019** (0.008)	1.019	0.017** (0.008)	1.017	0.010 (0.009)	1.010
Minority			0.047*** (0.011)	1.048	0.019* (0.011)	1.019	0.028** (0.012)	1.028
Median age			−0.008 (0.016)	0.992	−0.017 (0.016)	0.983	−0.028 (0.017)	0.973
Education			0.092*** (0.012)	1.096	0.080*** (0.013)	1.083	0.098*** (0.014)	1.103
Per capita income			1.473*** (0.550)	4.362	1.250** (0.552)	3.489	0.578 (0.588)	1.782
Poverty			−0.023 (0.550)	0.977	−0.218 (0.020)	0.978	−0.0.7 (0.023)	0.964
Council–manager					0.946*** (0.155)	2.575	0.668*** (0.177)	1.951
State dummies	No		No		No		Yes
Cut1 constant	1.649*** (0.345)		15.686*** (5.609)		12.963** (5.628)		5.222 (6.020)
Cut2 constant	3.288*** (0.360)		17.480*** (5.617)		14.847*** (5.635)		7.291 (6.024)
Cut3 constant	4.966*** (0.388)		19.340*** (5.627)		16.810*** (5.645)		9.364 (6.033)
Cut4 constant	6.191*** (0.432)		20.588*** (5.629)		18.118**** (5.645)		10.680* (6.034)
Summary statistics
LR χ² (21)	127.70		224.52		277.68		367.73
Probability > χ²	.0000		.0000		.0000		.0000
Pseudo-R ²	.0600		.1055		.1305		.1728

Note: Standard errors within parentheses. N = 816. LR = logistic regression.

*p < .10.

**p < .05.

***p < .01.

In order to test the stated hypotheses, four ordered logistic regression models were run separately, each with a different set of independent variables. Model 1 included the three independent variables that were the primary focus of this study: Internet access, land area, and population density. Model 2 included the variables commonly reported in the literature. Model 3 combined the variables used in models 1 and 2 and also controlled for the form of government. Model 4 was the most extensive and takes into account the notion that the roles and responsibilities of county governments may vary from one state to another, and thus controls for the state effects are employed. Estimates obtained from the three models are presented in Table 4, which reports the coefficients as well as the odds ratios.

As previously stated, the main focus of this article was to determine the impact of three independent variables on government transparency. Model 1 shows that as Internet access and population density increased, counties were more likely to have greater Web site transparency. The proportional odds ratio indicated that for an one-unit increase in Internet access, the odds of having a grade A transparency score versus the combined grades of B, C, D, and F was 1.75 times greater, given the other variables were held constant.

Model 2 included all of the independent variables in the model minus the three in model 1. The data show that population change, minority population, education, and per capita income had a positive impact on transparency, while median age had a negative impact. The variable that is particularly interesting in this model is per capita income. We found that not only was it statistically significant at the 1 percent level, but the variable had a relative high proportional odds ratio of 4.36.

Models 3 and 4 contain all of the independent variables. However, model 4 did not include a dummy variable for states. The data in model 3 show that population density, population change, minority population, education, per capita change, and form of government had a positive impact on transparency levels. The Internet access variable was not significant. The most interesting finding in this model are the data for form of government. The coefficient indicated the log odds of council–manager counties having a higher level of transparency was 0.95 higher than counties with another form of government, when all the other variables were held constant. The proportional odds ratio indicated that for council–manager counties, the odds of having a grade A transparency score versus the combined grades of B, C, D, and F was 2.58 times greater than for counties with another form of government.

The findings in Model 4, which controlled for state effects, were similar to model 3, but with two caveats. First, population change was no longer significant and land area was significant. Second, the effect of a council–manager form of government was lower than that in model 3, with a proportional odds ratio of 1.95.

Discussion and Conclusions

The major contribution of this study to the literature is that it includes the extent of Internet access in explaining Internet-enabled transparency. To the best of our knowledge, this is the first study in the United States to include such a variable in the transparency literature (Alcaraz-Quiles, Navarro-Galera, and Ortiz-Rodriguez 2015). While we found Internet access to be a significant predictor of Internet-enabled transparency in the simplest model, the finding did not hold true in the most extensive model. This does not negate the fact that the variable is an important factor in assessing transparency levels and Internet access. A cursory analysis often gives the impression that the greater the extent of Internet access, the higher the transparency levels. However, when taking into account some other factors, the extent of Internet access does not necessarily account for transparency (reject hypothesis 1). Suffice it to say that Internet access is dependent of education and perhaps population density.

With respect to the role of government and the notion that transparency is a key to good governance, we perceive that it is important for government officials and scholars to note what sort of factors influence transparency in government. Our findings show that a large majority of the counties in the Midwest region in the United States provide relatively limited information on their Web sites, and only a small portion provide the most extensive information. Thus, levels of transparency vary widely across the region.

When we consider our hypotheses in totality, we find mixed results. The data show that total land area, population density, percentage of minority, education attainment, and the council–manager form of government are statistically significant predictors of Internet-enabled transparency. These findings both confirm and negate the findings of previous researchers. For example, while the effect of education on transparency appears to be the most consistent finding in the literature (Huang 2007; Manoharan 2013a; Harder and Jordan 2013), we also noted that the rural/urban (population density) dichotomy and the education variable are important factors in assessing transparency levels. Hence, as governments create strategic plans that include growth models, they should consider not only the budgetary ramifications of growth but also the fact that educated residents want more Web-based interaction with government. This finding was reinforced by a recent Census Bureau report indicating that some of the cities and counties in Florida and California had population increases greater than 10,000 persons per month during the period 2013–2014 (Census Bureau Population Estimates Reveal Metro Areas and Counties that Propelled Growth in Florida and the Nation 2015).

This area of research holds much promise for future research as technology continues to improve all over the United States. For example, studies examining the impact of Internet services on the provisions of governmental services and educational outcomes should lead these research efforts. In fact, we perceive that advancement and the availability of Internet services will play a key role in reducing educational disparities around the country. Last, we suggest that future researchers utilize other statistical tools to analyze this data. For example, models that examine unidirectional data may provide more robust findings.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Alcaraz-Quiles

Francisco J.

Navarro-Galera

Andres

Ortiz-Rodriguez

David

. 2015. Factors determining online sustainability reporting by local governments. International Review of Administrative Sciences 81:79–109. doi:10.1177/0020852314541564.

Bernick

E. L.

Birds

J. M.

Brekken

Gourrier

A. G.

Kellogg

L. D.

. 2014. Explaining county government fiscal transparency in an age of e-government. State and Local Government Review 46:173–83. doi:10.1177/0160323X14556819.

Bonsón

Torres

Royo

Flores

. 2012. Local e-government 2.0: Social media and corporate transparency in municipalities. Government Information Quarterly 29:123–32.

Borry

E. L.

2012. Exploring determinants of governmental transparency: The case of municipal websites as a tool for proactive dissemination. In E-governance and civic engagement: Factors and determinants of e-democracy, ed. Manoharan

Holzer

, 25–47. Hershey, PA: Information Science Reference.

Census Bureau Population Estimates Reveal Metro Areas and Counties that Propelled Growth in Florida and the Nation. 2015. Release number: CB15-56. Accessed June 6, 2015. https://www.census.gov/newsroom/press-releases/2015/cb15-56.html

Conradie

Peter

Choenni

Sunil

. 2014. On the barriers for local government releasing open data. Government Information Quarterly 31:S10–17. doi:10.1016/j.giq.2014.01.003.

Day

Jennifer C.

Janus

Alex

Davis

Jessica

. 2005. Computer and internet use in the United States: 2003 Current population reports. Washington, DC: U.S. Census Bureau.

del Sol

Daniel A.

2013. The institutional, economic and social determinants of local government transparency. Journal of Economic Policy Reform 16:90–107. doi:10.1080/17487870.2012.759422.

England

Robert E.

Pelissero

John P.

Morgan

David R.

. 2011. Managing urban America. 7th ed. Washington, DC: CQ Press.

10.

FCC (Federal Communications Commission). 2014. Internet access services reports. http://www.fcc.gov/reports/internet-access-services-reports (accessed January 25, 2015).

11.

File

Thom

Ryan

Camille

. 2014. Computer and internet use in the United States: 2013 American community survey reports. Washington, DC: U.S. Census Bureau.

12.

Garcia-Murillo

Martha

. 2005. International broadband deployment: The impact of unbundling. Communication & Strategies 57:83–105.

13.

Grimmelikhuijsen

Stephan G.

2012. Linking transparency, knowledge and citizen trust in government: An experiment. International Review of Administrative Sciences 78:50–73. doi:10.1177/0020852311429667.

14.

Grimmelikhuijsen

Stephan G.

Meijer

Albert J.

. 2014. The effects of transparency on the perceived trustworthiness of a government organization: Evidence from an online experiment. Journal of Public Administration Research and Theory 24:137–57. doi:10.1093/jopart/mus048.

15.

Grimmelikhuijsen

Stephan G.

Porumbescu

Gregory

Hong

Boram

Tobin

. 2013. The effect of transparency on trust in government: A cross-national comparative experiment. Public Administration Review 73:575–86. doi:10.1111/puar.12047.

16.

Grimmelikhuijsen

S. G.

Welch

E. W.

. 2012. Developing and testing a theoretical framework for computer-mediated transparency of local governments. Public Administration Review 72:562–71. doi:10.111/j.1540-6210.2012.02532.x.

17.

Harder

Carolyn T.

Jordan

Meagan M.

. 2013. The transparency of county websites: A content analysis. Public Administration Quarterly 37:103–28.

18.

Heald

David

. 2006. Varieties of transparency. In Transparency: The key to better governance? ed. Hood

Heald

, 25–43. New York: Oxford University Press.

19.

Alfred Tat-Kei

. 2002. Reinventing local governments and the e-government initiative. Public Administration Review 62:434–44.

20.

Hood

Christopher

. 2006. Transparency in historical perspective. In Transparency: The key to better governance? ed. Hood

Heald

, 3–23. New York: Oxford University Press.

21.

Christopher

Hood

Heald

David

, eds. 2006. Transparency: The key to better governance? Vol. 135. New York: Oxford University Press.

22.

Huang

Zhenyu

. 2007. A comprehensive analysis of U.S. counties’ e-government portals: Development status and functionalities. European Journal of Information Systems 16:149–64. doi:10.1057/palgrave.ejis.3000675.

23.

ICMA (International City/County Management Association). 2010. The municipal year book 2010. Washington, DC: ICMA Press.

24.

Justice

Jonathan B.

McNutt

John G.

. 2013. Social capital, e-government, and fiscal transparency in the states. Public Integrity 16:5–24. doi:10.2753/PIN1099-9922160101.

25.

Kim

Bauer

J. M.

Wildman

S. S.

. 2003. Broadband uptake in OECD countries: Policy lessons from comparative statistical analysis. Paper presented at the 31st Research Conference on Communication, Information and Internet Policy, Arlington, VA.

26.

Kim

Soonhee

Lee

Jooho

. 2012. E-participation, transparency, and trust in local government. Public Administration Review 72:819–28. doi:10.1111/j.1540-6210.2012.02593.x.

27.

Manoharan

Aroon

. 2013a. A study of the determinants of county e-government in the United States. The American Review of Public Administration 43:159–78. doi:10.1177/0275074012437876.

28.

Manoharan

Aroon

. 2013b. A three dimensional assessment of U.S. county e-government. State and Local Government Review 45:153–62. doi:10.1177/0160323X13494858.

29.

Margetts

Helen

. 2011. The internet and transparency. The Political Quarterly 82:518–21. doi:10.1111/j.1467-923X.2011.02253.x.

30.

McMurray

Kristin

. 2013. 2013 Transparency report card: Bringing state & local government to light. Chesterfield, VA: Sunshine Review.

31.

Menifield

Charles E.

Clay

Joy A.

. 2010. Achieving transparency in state lottery websites. In E-government website development: Future trends and strategic models, ed. Downey

Ekstrom

C. D.

Jones

M. A.

, 320–41. Hershey, PA: Information Science Reference.

32.

Moon

M. Jae

. 2002. The evolution of e-government among municipalities: Rhetoric or reality? Public Administration Review 62:424–33.

33.

Nelson

Kimberly L.

Svara

James H.

. 2012. Form of government still matters: Fostering innovation in U.S. municipal governments. The American Review of Public Administration 42:257–81. doi:10.1177/0275074011399898.

34.

O’brien

Robert M.

2007. A caution regarding rules of thumb for variance inflation factors. Quality & Quantity 41:673–90. doi:10.1007/s11135-006-9018-6.

35.

Obama

2009. Transparency and Open Government [Memorandum for the Heads of Executive Departments and Agencies]. Washington, DC: The White House.

36.

Pina

Vicente

Torres

Lourdes

Royo

Sonia

. 2007. Are ICTs improving transparency and accountability in the EU regional and local governments? An empirical study. Public Administration 85:449–72.

37.

Pina

Vicente

Torres

Lourdes

Royo

Sonia

. 2010. Is e-government leading to more accountable and transparent local governments? An overall view. Financial Accountability & Management 26:3–20. doi:10.1111/j.1468-0408.2009.00488.x.

38.

Piotrowski

Susan J.

Ryzin

Gregg G. Van

. 2007. Citizen attitudes toward transparency in local government. The American Review of Public Administration 37:306–23. doi:10.1177/0275074006296777.

39.

Schauer

Frederick

. 2011. Transparency in three dimensions. University of Illinois Law Review 2011:1339–57.

40.

Sunshine Review. 2011a. Government transparency. http://sunshinereview.org/index.php/Government_Transparency (accessed January 25, 2012).

41.

Sunshine Review. 2011b. Sunshine review: About. http://sunshinereview.org/index.php?title=Sunshine_Review:About&oldid=70075 (accessed January 25, 2012).

42.

Sunshine Review. 2011c. Transparency checklist. http://sunshinereview.org/index.php/Transparency_Checklist (accessed January 25, 2012).

43.

U.S. Census Bureau (Producer). 2000. Census regions and divisions of the United States. Accessed September 6, 2014. http://www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf

44.

Welch

Eric W.

2012. The relationship between transparent and participative government: A study of local governments in the United States. International Review of Administrative Sciences 78:93–115. doi:10.1177/0020852312437982.

45.

Welch

Eric W.

Hinnant

Charles C.

Jae Moon

. 2005. Linking citizen satisfaction with e-government and trust in government. Journal of Public Administration Research and Theory 15:371–91. doi:10.1093/jopart/mui021.