1 Digital innovations in the public sector have attracted increased attention from researchers since the take-off and rapid development of e-government in the past decade, especially in developed countries. E-government is defined as the process of implementation, diffusion and use in public administration of Information and Communication Technologies (ICT) to achieve innovative forms of information and public services provision (OECD, 2003). Its diffusion has mainly been studied by researchers and private companies as a gradual linear process which starts with an initial web presence and ends with an integration-transaction and/or participative phase of e-service development (Layne, Lee, 2001; OECD, 2003; Cap Gemini, 2005; Torres et al., 2005; UN, 2008; Coursey, Norris, 2008; Zhang et al., 2014). Based on this gradual model, a national index of e-government diffusion has been built and annually actualized by national reports (OECD, 2003; UN, 2008). In turn, such an index has usually been used by empirical studies that take the national level as the unit of analysis to explain e-government diffusion. The main conclusion was that progress from information to the integration and eventually the transformation phase has not occurred either within European or US local governments (Torres et al., 2005; Ruano de la Fuente, 2013; Attour, Longhi, 2014; Norris, Reddick, 2013).
2 As they thus focus mainly on the planned diffusion process, those studies, however, overlook any spontaneous diffusion processes (Zhang et al., 2014). Furthermore, they do not offer a general view of the type of e-government services developed. To fill this gap, one other group of studies focused on factors explaining website implementation by municipalities and highlighted the role of several groups of factors: technological, organizational, environmental and cultural factors (Zhao et al., 2014). Other studies identified the e-services that local governments offer on their websites. Two categories of e-services have then been highlighted: information and communication services, and transaction-based services. While not all the services provided on local government websites are included in the authors’ surveys, several factors have been identified as impacting local e-government diffusion. The role of governmental forms, political institutions, organizational forms and public policies have, for example, been tested; but mostly in the specific case of American states (Karch, 2007). There are few, if any, studies that have applied this framework to European local governments.
3 More generally, drawing mainly on the theory of innovation diffusion as a general framework, e-government has been analyzed either as a technological innovation or as an administrative policy innovation. Technological factors and administrative policy factors have either been examined separately or, when they have both been considered, it was either in studies examining website implementation or e-government services development. Consequently, we do not clearly know if each identified factor similarly impacts website implementation and e-government services diffusion by local governments.
4 The aim of our research is to examine the role of such factors in the two steps of e-government diffusion: website adoption and the number of online services offered. By doing so, the objective is to answer the following research question: do technological and institutional factors similarly impact website implementation by local governments and the level of their e-government services development? To answer this question, this paper examines whether digital innovations in public administrations, such as e-government, follow a technological innovation process or an administrative policy innovation process of diffusion. Accordingly, original data were collected for the period January 2013 to December 2015 [2] (and regularly updated during the data collection period) to measure the level of development and diffusion of e-government in 5,108 French municipalities located in the French Grand Est region.
5 This paper is organized as follows. Section 2 presents the twofold framework mobilized by our research. Section 3 summarizes the theoretical model and hypotheses tested. Section 4 presents the data and the methodology used to explain website implementation and e-government services development. Section 5 describes the empirical findings and discusses the variation in e-government implementation and levels of development.
Theoretical Basis of E-Government Diffusion Analysis: A Twofold Framework
6 E-government diffusion has usually been studied either as a technological innovation or as an administrative policy innovation. The technological innovation literature defines e-government services as a specific case of services innovation, conceptualized as “a process of transmission of technology from capital goods sectors to a specific user sector that employs it to supply existing as well as new services” (Arduini et al., 2010, p. 258). This approach uses the territorial context-specific characteristics and demand conditions as the determinants of innovation diffusion. Context-specific characteristics include population size, internal competency [3], economic and tourist factors (Walker, 1969; West, 2000). Demand conditions are usually described as ‘citizen pressure’ [4] (Ho, Ni, 2004). All these determinants have been largely tested in empirical works showing their positive and significant effect on national and local government attitudes to innovation. However, in these previous works, the dependent variable is either a national index of e-government (Arduini et al., 2010; Lee et al., 2011; Zhang et al., 2014), or the probability that an online website has been implemented by a municipality (Finney, Yoon, 2011). Other studies, focusing on the local level of e-government diffusion, have tested the role of context-specific characteristics to explain why a municipality invests more in transactional e-services than in non-transactional ones (Coursey, Norris, 2008; Norris, Reddick, 2013). In sum, how a municipality’s context-specific characteristics impact the two steps of e-government diffusion – namely website implementation and online services development – has been examined separately. It is therefore difficult to compare their role in the two steps of e-government implementation and diffusion, as the authors’ surveys are quite different and did not adopt the same level of analysis (local versus national surveys).
7 On the administrative policy innovation literature side, the theoretical framework usually used suggests that innovation adoption decisions and diffusion are a complex part of policymaking. From there, to analyze e-government diffusion, authors added two other factors to the one related to context-specific characteristics (Berry, Berry, 1999): competition and learning (strictly related to geographical learning and diffusion), and mandates (normative or coercive). Competition and learning factors derive from social learning process theory (Rogers, 1995), stating that the social learning process has a positive regional effect on policy diffusion (McNeal et al., 2003). Thus, according to the social learning model, a national or local government will be more likely to adopt a policy if a neighboring government has already done so (Berry, Berry, 1999). Imitating decisions made by national or local authorities that are geographically or socio-economically proximate helps to minimize the risks and is likely to contribute to the achievement of results like those achieved by their counterparts (Berry, Berry, 1999, 2007).
8 Inspired by the institutional isomorphism concept of DiMaggio and Powell (1983) [5], mandate factors are relayed through the vertical diffusion of accepted management and policy standards. In cases of national interaction, innovation diffuses horizontally from one government to another. This diffusion process could stem from a pressure on all local governments (despite the autonomy they have) to conform to nationally or regionally accepted standards. As written by Berry and Berry (1999, p. 226): “Sometimes the pressure is what DiMaggio and Powell (1983) label ‘coercive’, when federal mandates give state governments little choice. In other cases, there is “normative” pressure on state officials to adopt the best practices in other states. State officials tend to be socialized into shared norms by common professional training (such as the Masters in Public Administration degree) and by interaction in professional associations (e.g. the National Emergency Management Association)”.
9 Mandates can thus take place through formal and informal interaction between elected officials and decision-makers who are early adopters and non-adopters. The sharing of experience enabled by communication networks can hasten the adoption of innovation by local governments that may initially be hesitant (Berry, Berry, 2007, p. 226). Nevertheless, innovation policies and directives issued by national governments can be difficult to implement at the local level. Consequently, innovation implementation may require adaptation or discretionary decisions calling for consultation with higher levels of government, based on national directives (Walker, 1969). Indeed, as suggested by the vertical innovation diffusion theory, a local government will be more likely to embrace an innovation if its implementation is part of a broader plan decided at a higher level of government (Berry, Berry, 1999).
10 While competition and learning and citizen pressure factors have already been considered in empirical works studying e-government diffusion at both the local and the national level (Dang Nguyen et al., 2013; Norris, Reddick, 2013), arguments derived from mandate dimensions has been investigated only when the unit of analysis is the national level (Lee et al., 2011). Examining all three factors of Berry and Berry’s (1999) framework in a same model and considering a same unit of analysis would help to better understand e-government diffusion. By doing so, the aim is also to try to unify the e-government diffusion model. Indeed, the administrative policy innovation diffusion framework has mostly been addressed by a range of US studies at different levels of government (local, state and national). But as the role of all three determinants of the administrative policy innovation diffusion framework has not been examined for the same unit of analysis and the same outcome variables, empirical findings suggest that there is no single e-government diffusion model (Norris, Reddick, 2013). Furthermore, few, if any, studies mobilizing policy diffusion models focus on European countries or municipalities. And, in both US and European literature on local e-government diffusion, there is little or no work on the mandatory dimension of administrative policy innovation diffusion theory.
11 Based on this literature review, to answer our research question, a theoretical model including all factors derived from technological innovation and administrative policy innovation frameworks is built and detailed in the next section.
The Research Framework and Hypothesis Development
12 The following figure depicts the theoretical model and hypothesis development derived from the literature discussed in the last section.
Figure 1 – E-government theoretical model of diffusion
Figure 1 – E-government theoretical model of diffusion
13 Context-specific characteristics rely on technological factors. Different types of those factors have been tested by e-government literature: the role of economic resources (Baldersheim, Øgärd, 2008), technical resources (Moon, Norris, 2005; Arduini et al., 2010), tourism activity (Dang Nguyen et al., 2013) and the role of citizen pressure that is a common factor in administrative policy innovation theory (Weare et al., 1999; Ho, Ni, 2004; Finney, Yoon, 2011; Lee et al., 2011; Dang Nguyen et al., 2013; Norris, Reddick, 2013).
14 For example, based on 75 municipalities and 54 regional administrations of four Nordic countries (Denmark, Finland, Norway and Sweden), Baldersheim and Øgärd (2008) showed that two economic resources – unemployment rate and demographic change – have an impact on e-government development (which was measured through the analysis of information and communication features of Nordic municipalities’ web pages). From there, we formulate the following first hypothesis H1: The unemployment rate has a significant and positive effect on municipalities’ choice to implement and develop a website (H1a) and on the level of development of the municipality website (H1b).
15 In terms of technical resources, Moon and Norris (2005) found the lack of technology as a key barrier to the advancement of e-government. In their model, Arduini et al. (2010) introduced the wide availability of broadband connections as an important technical resource for the provision of advanced e-government services. Their results suggested that advanced communication infrastructures do not influence the start-up of e-government but do affect e-government intensity. From there, we formulate the following second hypothesis H2: the level of access to high-speed broadband has a significant and positive effect on the probability that a municipality implements its own website (H2a) and on the level of municipality website development (H2b).
16 Dang Nguyen et al. (2013) introduced tourism activity in their model as key context specific characteristics. They assumed and validated that the more tourist activities the municipality has, the greater the probability that the local government of this municipality has a website. They also validated that the more the municipality is specialized in tourist activities, the greater the number of features linked to e-government on the website. We then carry out the following assumption H3: the probability of having a website (H3a) and the level of development of the municipality website (H3b) are high for tourist municipalities.
17 Citizen pressure has been measured in empirical works by the city’s population size or by several variables including median income, population age, level of education, and socio-professional categories. It is generally shown that citizen pressure has a significant effect on local authorities’ attitudes to e-government. However, the results are not consistent with both website implementation and website development. For example, Moon and Norris (2005) showed that while there were no statistically significant relationships between government size and plans to adopt a website, size was positively associated with adoption for all online transaction e-services. On the contrary, Finney and Yoon (2011) point out that, in the United States, municipalities that are identical in terms of population size often take similar approaches to websites and online resources. Considering a sample of municipalities localized in the north-west of France, Dang Nguyen et al. (2013) found a positive and significant relationship between population size and both website implementation and development. Based on this, we formulate the following hypothesis H4: there is a significant and positive correlation between municipality population density and the probability that the former implements a website (H4a) and high development of its level (H4b).
18 Arduini et al. (2013) analyzing the e-government development [6] of 4,471 Italian municipalities with more than 20,000 inhabitants, confirmed the important role of age as a barrier to e-government demand and supply. The authors tested the role of demographic change, calculated by evolution of the share of the older population, and found a negative coefficient. From there, the authors concluded that there is a negative relationship between e-government development by a municipality and the rate of its elderly population. Thus, we formulate the following hypothesis H5: the age of the population has a significant impact on the probability that the municipality has a website (H5a) and on the level of development of the municipality website (H5b).
19 In relation to the level of education and socio-professional categories of a city’s population, Lee et al. (2011) used a human capital index that is composed of the adult literacy rate and educational levels. They found that the index is strongly associated with e-government development and concluded that countries with a high rate of educated human capital tend to develop advanced e-government applications. Similarly, considering the local level of e-government development, Norris and Reddick (2013) found a positive and significant relationship between e-government development and a high average of the population’s educational attainment. We then assume the following hypothesis H6: the greater the well- qualified population in the municipality, the higher the demand for a municipal website (H6a) and the higher the level of development of the municipality website (H6b).
20 In addition to citizen pressure, administrative policy innovation theory considers the role of two other factors: competition and learning, and mandates. Competition and learning are generally measured through the behavior of neighboring municipalities. For example, Finney and Yoon (2011) found that the implementation of a website is dependent on the decisions of nearby neighbors of similar size. However, they found weaker evidence of interdependence in the decision to utilize the web for financial transactions. Similarly, at country level, Lee et al. (2011) showed that a nation that is adjacent to one or more neighboring countries with high e-government performance is more likely to have a higher e-government ranking. Dang Nguyen et al. (2013) validated the same assumption, but at the local level of e-government. From there, we formulate the following hypothesis H7: if there are more municipalities with websites in the same inter-municipality, the probability that the municipality has implemented a website is greater (H7a) and the level of development of the municipality’s website is higher (H7b).
21 Finally, as Berry and Berry’s (1999) model considers mandates to be a key factor of administrative policy innovation adoption, we added the following hypothesis to test H8: departmental affiliation has a significant effect on the municipality’s website implementation (H8a) and on the level of the municipality’s e-government services (H8b). Indeed, this variable has not yet been tested in empirical findings. Lee et al. (2011) have considered the normative side of the mandates factor by testing how the rank of a country in its level of transparency could impact its rank in e-government development.
22 Section 4 below presents the data we built and used to test these eight hypotheses.
Data collection and Methodology
23 France’s territorial and institutional organization comprises four layers: national government, 13 metropolitan regions, [7] 101 departments and 35,461 municipalities. Municipalities can link up and form inter-municipalities [8] so as to facilitate common economic development projects. Because the regions where municipalities are located are largely representative of the relevant technological, institutional and economic innovation context (Arduini et al., 2010, p. 258), we focus on municipalities [9] and observe how each department or former regional affiliation affects their e-government strategies. The present analysis focuses specifically on French municipalities located in the Grand Est region, which comprises ten departments: Bas-Rhin and Haut-Rhin (Alsace), Meurthe-et-Moselle, Meuse, Moselle and Vosges (Lorraine) and Ardennes, Aube, Haute-Marne and Marne (Champagne-Ardenne, see Maps 1a and 1b). Currently, there are 5,195 municipalities in this region [10].
Map 1a – French Grand Est region
Map 1a – French Grand Est region
Map 1b – Departments of Grand Est
Map 1b – Departments of Grand Est
Data Collection and Presentation
24 We collected data for the period January 2013 to December 2015 on the Grand Est municipalities (referred to as e-government in the Grand Est municipalities (2013-15)). Of the 5,108 municipalities in Grand Est, 1,509 have a website. However, as Map 2 shows, the proportion of municipalities having websites varies according to the former region, with two thirds in Alsace but only a quarter in Lorraine and one in six in Champagne-Ardenne. This geographical similarity suggests a “neighborhood” effect, as in Berry and Berry (1999, 2007).
Map 2 – Geographical distribution of municipalities with a website
Map 2 – Geographical distribution of municipalities with a website
25 The number of services and functions offered by each municipality through their official websites was recorded and assessed using the stepwise linear model of e-government diffusion. The resulting database includes details on the type of websites provided by each municipality in Grand Est (individual sites, general sites created by agglomerations of several municipalities, sites established by associations, etc.), the number of websites, the type and level of information provided, and the presence of local or central interactive/transactional e-government services. We obtained a list of 67 services. Graph 1 shows that the most frequent services offered by municipality websites in Grand Est include: an organization chart of the city hall (70% of all cases), associations (63%), schools (55%), cultural events (52%), newsletter (49%), information about waste recycling (49%), and the Municipal Council agenda (46%). We classify the services into six categories: local e-administration, central e-administration, e-information, local life, e-democracy and social networks (cf. Table 1). Of these categories, the most prevalent is e-democracy, with services offered on average by more than 41% of municipal websites. Information on local life is provided by 31% of the websites, with central and local e-administration services provided by between 18% and 26% of municipalities. Services related to social networks were offered by less than 10% of websites on average.
26 The level of development of e-government is in line with the findings in the literature. The services offered by municipality websites are primarily the provision of information, with a small number of additional services available on the websites of the region’s largest cities.
Graph 1 – Proportion of municipal websites offering various items
Graph 1 – Proportion of municipal websites offering various items
Table 1 – Categories of e-government services
Table 1 – Categories of e-government services
27 To estimate more precisely the level of development of municipalities’ e-government offers, we constructed two variables. The first summarizes the number of services available on the website. On average, municipal websites in Grand Est offer around 18 e-services (cf. graph2). Only one website offered more than 50 services (56 services). The second variable is a weighted sum of the level of development [11] of each service offered on the website. We obtained a mean of 26.11 with a maximum of 98 (see Graph 3). The database was complemented by the addition of three further data fields: departmental affiliation of each municipality; INSEE (the French National Institute for Statistics and Economic Studies) data on the municipality’s socio-economic characteristics; and data on the extent of high-speed broadband coverage in each municipality. [12] Thus the full database allows comparison of the impact of internal competencies and administrative policy innovation diffusion in e-government diffusion in Grand Est.
Graph 2 – Sum of e-services offered on level a website
Graph 2 – Sum of e-services offered on level a website
Graph 3 – Weighted sum of the development of e-government services
Graph 3 – Weighted sum of the development of e-government services
Econometric Regressions: Methodology
28 To answer our research question, we constructed three explained variables. The first, ‘website implementation’, is a binary variable that takes the value 1 if the municipality has created a website and 0 otherwise. The model is a binary logit which explains the probability of a municipality in Grand Est having a website. To analyze the degree of development of municipalities’ websites, we use two different variables: the sum and the weighted sum of the services offered on the website (cf. Graphs 2 and 3). These are continuous variables that allow estimation of OLS models [13]. To test our hypotheses (cf. Section 3), we introduced several explanatory variables into the regressions, after checking for any multicollinearity problems. These exogenous variables are classified in two groups. The first group is related to the municipality’s specific context characteristics that we adapted from the literature on technological innovation diffusion:
- Demand size (represented by two variables: population density, and number of jobs),
- Economic resources (following Baldersheim and Øgärd (2008), we use the unemployment rate),
- Technical resources (determined by the proportion of housing in the city with an internet connection higher than 8 Mbits) and
- Tourism activity (a municipality is considered as a tourist municipality if there is a camping facility, hotel or tourist office [14]).
30 The second group of explained variables was chosen based on the dimensions of the administrative policy innovation diffusion model:
- Mandates (measured by the municipality’s affiliation with a department [15]),
- Competition (proportion of municipalities with a website in the same inter-municipality),
- Citizen pressure (taking account of age, education level (occupational category) and size of the population). Note that population size as a proxy for citizen pressure is used in technological innovation diffusion and administration policy innovation diffusion models.
Results and Discussion
32 This section presents and discusses the results of the three econometric models that are presented in Table 4. Column 1 of this table presents the results of the logit ‘website implementation’ model. Column 2 presents the results of two OLS ‘website development’ models [16] (the number of e-services on the municipal website in the first sub-column and the weighted sum of e-services in the second sub-column). In all three models, not all but the majority of the hypotheses concerning the role of specific context characteristics, citizen pressure, mandates, competition and learning factors, hold. Generally, the three econometric models’ results suggest that the provision of a website by a municipality can be understood as a technological innovation whose diffusion is explained mainly by economic factors and territorial competition, while the level of development and type of e-government offer correspond to administrative policy innovation, which is diffused vertically. We discuss these results in the following sub-sections.
Table 2 – Results of Logit model ‘website implementation’ and OLS ‘website development’ models
Table 2 – Results of Logit model ‘website implementation’ and OLS ‘website development’ models
The Role of Context-Specific Characteristics on E-Government Diffusion in the French Grand Est Municipalities
33 Among contextual factors, the results in Table 4 suggest that the most influential ones are size of demand and the economic environment. These results suggest that, in contrast to the findings in Baldersheim and Øgärd (2008), e-government diffusion in French municipalities does not follow an economic catch-up logic but rather confirms the municipalities’ economic advantages. Indeed, surprisingly, the local unemployment rate does not have a significant effect on the probability that a French Grand Est municipality implements a website (not supporting hypothesis H1a). Conversely, this variable has a significant and high effect in ‘website development’ models (supporting hypothesis H1b). Such a result suggests that economic resources are less powerful factors influencing website implementation by French Grand Est municipalities than the demand size factor (number of jobs). On the contrary, it is a more powerful factor than demand size when considering the development of existing municipalities’ websites.
34 Apart from this contextual factor, in line with previous work (Ho, Ni, 2004; Finney, Yoon, 2011; Lee et al., 2011; Norris, Reddick, 2013), the three others appear as main determinants of French municipalities’ decisions to both create and develop a website. Indeed, as expected, population density, a high level of broadband internet access, and tourism have a positive effect on both website implementation by a municipality and the development of a quantitative (the number of e-services) and qualitative (the weighted sum of e-services) offer of e-government services. These results support hypothesis H2, H3 and H4. For example, a one per cent rise in the local unemployment rate increases the sum of e-services by 0.22 and the weighted sum of e-services by 0.34, other things being equal. Also, being a tourist municipality increases the sum of e-services by 2.5 and the weighted sum of e-services by 4.2, other things being equal. Tourism has a significant effect, with the probability of the presence of a website being 12.2 points higher for tourist municipalities, all other things being equal (supporting H3a). This finding suggests that a website is a good way of promoting the municipality (Dang Nguyen et al., 2013).
35 In contrast to Arduini et al. (2010), who found that broadband access does not influence the startup of e-government but does affect e-government intensity, our results show that broadband access has a significant and positive influence both on the municipality’s decision to establish a website and on the development and types of e-services offered. This unexpected disagreement with the findings of Arduini et al. (2010) can be explained by the presence of areas with low levels of internet access (e.g. Meuse). On the other hand, the results for website level of development are consistent with Arduini et al. (2010) and with Grubesic and Murray’s (2004) findings: the technological level of the network infrastructure is associated with higher rates of generation and diffusion of advanced communication services.
36 And in relation to population density, we observe that the variable ‘population density’ has a positive and significant effect in the three tested models, while the variable ‘number of jobs’ only has a positive and significant impact on the ‘website implementation’ model. This could be explained by the fact that population density not only represents demand size but also the number of electoral votes at the local level (Dang Nguyen et al., 2013). Indeed, this variable is also one indicator usually used to examine the role of citizen pressure. It also appears as the most powerful one among all those used to measure the role of citizen pressure.
Citizen Pressure and E-Government Diffusion in the French Grand Est Municipalities
37 The results of the three econometric models do indeed show that citizen pressure has an important role in e-government diffusion in the French Grand Est municipalities. However, this role seems different in the ‘website implementation’ model compared to the ‘website development’ model. Indeed while, as expected, the larger the population, the greater the probability of website provision and the greater the level of development of this website (supporting H4), the effect of the distribution by occupational category is different in the ‘website implementation’ model and in the ‘website development’ model. We found that the proportion of farmers in the population has a negative effect on the probability of having a website (confirming H6a). If the proportion of agricultural workers increases by 1%, the probability of having a website falls by 0.7 per cent.
38 Furthermore, we find that municipalities that tend to emphasize the number of their services (sum of services model) rather than their quality (weighted sum of services model) are more sensitive to citizen pressure (measured by ‘population density’). This is in line with Lee et al. (2011) who introduced population size as a control variable in their analysis of national e-government rankings. Indeed, if we examine the role of citizen pressure measured by the level of education of the population, it seems that municipalities are influenced by the presence of higher educated professionals and especially in relation to the quality of their e-government services offer. Lee et al. (2011) suggest that the presence of well-educated citizens increases the potential for government to use e-government services. In addition, our results show that the influence of a more highly educated population affects the level of development of the e-government services offered: a rise of 1 per cent in this proportion increases the sum of e-services by 0.19, and the weighted sum of e-services by 0.28, other things being equal (confirming H6b). This effect can be interpreted as demand on the part of well-educated individuals. Thus, to explain e-government website provision, it appears that citizen pressure is more relevant when it is measured by the population’s educational level than by demand size (population density) or other variables. This is also supported by the not significant role of the distribution of age factor, neither for website implementation nor for website development (not supporting H5).
39 In the same vein, the role of competition and learning seems not to impact in a similar manner the ‘website implementation’ model and the ‘website development’ model.
The Role of Competition and Learning Factors in E-Government Diffusion in the French Grand Est Municipalities
40 According to factors related to the competition and learning model, our results suggest that the proportion of municipalities with one website in the same inter-municipality has a positive and significant influence on the probability that a municipality implements a website, on the one hand, on the probability this municipality offers several e-services on this website, on the other; but it has no effect on the level of online development of e-government services. Provision of a website is more likely if neighboring municipalities with websites are more numerous (confirming H7a). As the result of Table 4 Column 1 suggests, a one per cent increase in the number of municipalities with a website in the inter-municipality increases the probability of the focal municipality providing a website by 0.5 of a point, other things being equal. This finding reflects an imitation effect. However, although the proportion of municipalities with one website in the inter-municipality has a significant effect in the “sum of e-services” model, this does not apply to the “weighted sum of e-services” model. The role of the competition dimension seems to be less important when qualitative aspects are considered (H7b only partially confirmed), suggesting that municipalities motivated by competition issues are adopting a quantitative strategy (“sum of e-services”) of e-government diffusion more than a qualitative strategy (“weighted sum of e-services”). The role of mandates factors seems to confirm the existence of two types of e-government development strategies in the municipalities of Grand Est.
The Role of Mandates Factors on E-Government Diffusion in the French Grand Est Municipalities
41 In relation to the mandate dimension of administration innovation policy diffusion, we found that department affiliation plays a significant role both in website implementation and development. Our results also suggest that the provision of e-government services differs according to the department affiliation of the municipality. Indeed, we found differences among departments (confirming H8b): e-services are more numerous in Moselle and Marne and to a lesser extent in Haut-Rhin and are infrequent in Vosges. When we take account of the nature of the e-services provided, this hierarchy changes slightly. We find that the best developed websites are in Meurthe-et-Moselle, followed by Marne and Moselle, with Haut-Rhin ranked last among these, while the lowest levels of development are in Vosges and Meuse, which are more rural departments. This confirms the existence of two types of e-government development strategies in the Grand Est municipalities.
42 Some municipalities prioritize quantity while others prioritize quality. Those adopting a quantity strategy seem to be motivated more by enhancing their interaction and proximity to citizens, reflecting a normative logic. For those favoring the quality of the services offered, it seems that e-government services are developed to enhance the performance and quality of public administration services. E-government is seen as a way to improve, but not to transform, the traditional relationship between public administrations and citizens by delivering one-directional services and information (Norris, Reddick, 2013). Norris (2005) showed, and Norris and Reddick (2013) confirmed, that e-government is adopted mostly by municipalities to provide information and services online, and to provide citizen access to government officials.
43 The mandate dimension shows that there is pressure on local governments to conform to accepted management and policy standards decided at a higher level (Berry, Berry, 1999). These standards may be coercive or normative (DiMaggio, Powell, 1983). In our case, we would suggest that in departments where municipalities have prioritized the quantity of services, e-government policy standards are more coercive than normative. Among those that prioritize quality, e-government policy standards are more normative than coercive.
Conclusion
44 Applying the administrative policy innovation diffusion theory model (Berry, Berry, 1999, 2007) to a European country, this paper examined the role of technology innovation factors and administrative policy innovation factors on e-government diffusion and development in the French Grand Est municipalities. The empirical results show that website implementation is explained mostly by competition, citizen pressure, and a favorable economic context, while the level of development and type of e-government services offered is affected by the vertical diffusion of accepted management and policy standards. Access to high-speed broadband and tourism activity are two internal competencies influencing both the first step (website implementation) in e-government diffusion and its level of development. This finding implies that website implementation is governed by a technological innovation diffusion logic, while the level of development responds to a vertical diffusion logic. Furthermore, the econometric results confirm findings in the US literature on e-government diffusion, showing that there is no single e-government diffusion model. Our results suggest that in the French case, there are two types of e-government diffusion and development model. Some municipalities favor a quantitative strategy responding to an economic and competition logic, while others favor a qualitative strategy to enhance their relationships with citizens.
45 From a policy standpoint, our results show that departmental affiliation has a significant and relatively strong influence on the type and level of e-government services offered by municipalities. Consequently, since local e-government in Grand Est has yet to reach the transformative phase (as in Coursey, Norris, 2008), policy aimed at improving the level of development of local e-government services at departmental level could be positive.
46 Our findings suggest avenues for future research. First, the proportion of municipalities with one website in the same inter-municipality could be refined to explain the level of e-government development by constructing additional variables for the number of proximate municipalities with the same levels of e-government. Doing so would allow a better comparison of the role of the competition model with other dimensions of administrative innovation policy diffusion. Second, we measured the mandate dimension of policy diffusion using municipality departmental affiliation. Coerciveness or normativity of policy standards decided at a higher level (Berry, Berry, 1999) in the specific context of French municipalities located in Grand Est was deduced using the two e-government development strategies (quantity or quality) identified. These standards could be better characterized by examining what coercive and normative means – as defined by DiMaggio and Powell (1983) – and how these factors could be measured more precisely.
Notes
-
[1]
Acknowledgements: This research was supported by the French Grand Est Region and the University of Lorraine.
-
[2]
Two motivations explain the choice of this period. First, several governmental plans have been adopted by the French agenda of e-government. Most of them were adopted at the beginning of the 2000s. These planned strong objectives in terms of e-government diffusion with the deadline of 2007. Studies conducted after this period showed that not all the objectives of these plans have been successfully achieved (Dang Nguyen et al., 2013; Attour, Longhi, 2014). Second, the last national program adopted was in 2008 (Digital France 2012, adopted in October 2008). Through 154 proposals, this plan aims to increase the development of e-government at the local and national level by the end of 2012. We thus assumed that, at the end of 2012, all local governments that were willing to transpose the national plan at the local level had made the principal efforts they were able to make in terms of e-government diffusion.
-
[3]
Technical and managerial capacities are specific technological assets adopted by the local authority in the past and capable of facilitating the implementation of innovation.
-
[4]
Citizen pressure, measured by several variables including median income, population age, level of education and socio-professional categories, is shown to have a significant effect on local authorities’ attitudes to e-government.
-
[5]
DiMaggio and Powell (1983) find institutional isomorphism to be a powerful concept to understand the political environments that may influence organization. They identified three mechanisms through which institutional isomorphic change occurs: coercive process, mimetic process, and normative-leading process (Huault, 2009).
-
[6]
Arduini et al. (2013) measured the development of e-government by a composite indicator (a Front Office Index) describing the availability (number of thematic areas covered) and the quality (level of interactivity) of e-government services offered by municipalities.
-
[7]
On 1st January 2016, there were 22 French metropolitan regions. Grand Est includes three former regions: Alsace, Champagne-Ardenne and Lorraine.
-
[8]
An inter-municipality is a local government structure specific to France, involving cooperation between two or more municipal governments to accomplish common goals.
-
[9]
The choice of municipal data is coherent with the theoretical model, which includes hypotheses like competition, learning and mandates.
-
[10]
This region was chosen for two reasons. The first is that it is a Region where the development of e-government varies widely from territory to territory. The second is that the Conseil Régional of Grand Est subsidizes this study.
-
[11]
The level of development of the services was assessed using the stepwise linear model of e-government implementation and development. The scale of the weightings contains five values. A simple information provision is coded 1, the possibility of download is coded 2, the opportunity to submit a document is coded 3, the possibility to enter information is coded 4, and if it is possible to obtain all the services online, this service is coded 5 (transactional level). Of course, if the item is absent, the code is 0.
-
[12]
These data were collected from the Observatoire France Très Haut Débit website.
-
[13]
We choose linear models because the distribution of the sum of items and the distribution of the weighted-sum of items are linear. So, the ordered logit models are not appropriate.
-
[14]
With this definition, 785 towns are considered as tourist towns. If we restrict this definition, so that all three facilities are present, only 83 towns are tourist towns.
-
[15]
In our research, we assumed that municipalities belonging to the same department may adopt a similar level of e-government services due to departmental mandates. Despite the fact that in France, as in the US context, e-government adoption at the local level is mostly voluntary, several policy initiatives have been taken and transposed from the regional to the departmental level in terms of the digital development of territories. Among these policy initiatives, in France, each department could adopt their policy plan for enhancing ICT implementation and diffusion. An analysis of the French department policy plan indicated that each one has included action plans in terms of local e-government development.
-
[16]
Here the population of interest is the 1,169 municipal Grand Est websites.
