Content area
Purpose
This study uses a bibliometric approach to analyze the overall status of e-government research by revealing patterns and trends that would help gain a broad understanding of global developments in the field and future directions.
Design/methodology/approach
All documents related to e-government published from 2000 to 2019 were extracted from the Scopus and the Digital Government Reference Library databases. Bibexcel, Biblioshiny, and VOSviewer were used to perform the analyses and visualize the science mapping.
Findings
The results showed that 21,320 documents related to e-government research were published and cited 263,179 times. The annual growth rate of e-government research has reached 21.50%. The regression analysis showed that the growth rate is expected to increase gradually over the coming years. Despite the significant role that conference papers play in the e-government literature, the impact of articles far exceeds conference papers. The University of Albany (SUNY) has played an important role in e-government research in terms of production and impact. Furthermore, the study revealed some countries that are expected to play a prominent role in e-government research, as well as several topics that may attract more attention soon.
Originality/value
The results presented in this study and the comprehensive picture obtained about the e-government field make it an effective supplement to the expert evaluation. Thus, researchers, research managers, policymakers, institutions, funding agencies, and governments can rely on it.
1. Introduction
1.1 Background
According to Fountain (2003), e-government research is “research related to the intersection of government practices and information technology”. In the same sense, Dias (2019) assumed that e-government research deals with the study of the use of information and communication technology (ICT) by governments in their different but complementary and interrelated dimensions.
The term e-government appeared in the late 1990s (Grönlund and Horan, 2005), and according to Heeks and Bailur (2007), this term was first used in 1997, while the term electronic government was used before that date in the 1993 U.S. National Performance Review. At that time, the term “electronic” was used to denote modern, new, and future-oriented projects powered by information and communication technologies. Moreover, the term digital government was used interchangeably with electronic government (Scholl, 2017).
Generally, public funding in both the United States and Europe has played a role in the emergence of e-government research through the support of ICT projects directed at improving government operations (Scholl, 2020c). In 1998, the US National Science Foundation released the first appeal for research grant proposals in the digital government research program, and in 1999 began holding annual workshops for holders of this grant (Ciment, 2003; Dawes et al., 2004). Shortly after the United States, the European Union started, under the Fifth and Sixth Framework Programs, to fund research programs related to e-government (Scholl, 2020c). Since then, e-government research began to emerge and grow significantly in size and content (Przeybilovicz et al., 2014). Subsequently, several scientific conferences were held, and specialized journals emerged to discuss and publish e-government research (Grönlund and Horan, 2005).
Between 2001 and 2007, Several journals and conference outlets were dedicated to publishing e-government research, including the International conference on digital government research (DG.O), International Conference on Electronic Government (EGOV), International Conference on Theory and Practice of Electronic Governance (ICEGOV), Hawaii International Conference on System Sciences (HICSS), Electronic Journal of e-Government (ECEG), Journal of Information Technology and Politics (JITP), formerly known as The Journal of E-Government, Electronic Government (EG), International Journal of Electronic Government Research (IJEGR), and Transforming Government: People, Process, and Policy (TGPPP) (Grönlund and Horan, 2005; Harihara and Basden, 2008; Scholl, 2020c). Besides, there were Government Information Quarterly (GIQ) and Information Polity (IP), which have begun to expand their scope and accommodate e-government research (Scholl, 2020c).
In 2005, Electronic Government Reference Library (EGRL) was established to improve the quality of academic research in e-government and to identify peer-reviewed publications in the research area (Scholl, 2020b). The first version of the library was launched in 2006 and renamed in 2008 as the Digital Government Reference Library (DGRL) (Scholl, 2020a).
Thus, the new field of e-government was established and characterized by the richness that originated from its multidisciplinary nature. Several theories and methods from different disciplines are used to explain and understand the issues related to e-government and its various aspects (Dias, 2019; Irani and Dwivedi, 2008). Therefore, we find that the main contributors to e-government research are scholars who have received disciplinary training in public administration, computer science, political science, management information systems, information science, and others (Scholl and Dwivedi, 2014). Thus, no single discipline can claim individual ownership of the e-government field (Scholl, 2017).
Almost twelve years ago, Dwivedi (2009) pointed out that many researchers consider e-government research still in an emerging state, while others believe it is presently moving towards initial maturity. Eight years later, Alcaide-Muñoz et al. (2017a) assert that e-government research is in constant development but has not yet reached a mature stage. To expand our knowledge of the status of e-government research, we present, in this study, a bibliometric analysis of e-government research.
1.2 Purpose of the study and research questions
Studying the behavior of the scientific literature published helps to understand the current status and future directions of a field (Wijewickrema, 2022). Bibliometric studies are used to study performance analysis and scientific mapping of units of a field such as authors, research groups, institutions, and countries (Banshal et al., 2022). Despite the vital role of bibliometric analysis as a statistical tool that provides a quantitative to assess research patterns and trends (de Oliveira et al., 2019; Hossain, 2020), only a few studies have applied bibliometric analysis on e-government research globally. However, the analysis performed in these studies is still partial and limited (see Section 2). Therefore, there is a need to analyze and monitor the characteristics of e-government research globally.
The current study provides a bibliometric analysis of e-government research globally over two decades, including the last five years that were not covered by any studies. It also relies on more complete databases compared to the related studies, especially in light of the recommendation of previous studies that more analyzes of the e-government literature should be conducted (Almeida et al., 2014). Thus, the present study aims to provide an analysis of the overall status of e-government research by revealing patterns and trends that would help gain a broad understanding of global developments in the field and future directions. To this end, the following research questions were formulated:
What are the characteristics of documents related to e-government?
What are the characteristics of authorship in e-government research?
Which sources have the most document production as well as the most influential?
To what extent does the network of cooperation on e-government research spread?
How have subject trends in e-government research evolved over time?
The value of a bibliometric study is that it can clarify ongoing research efforts in a specific topic or field and help achieve a better understanding of the field's evolution over time (Hossain, 2020; Liu and Li, 2016). Besides, bibliometric studies can be used as reliable sources to justify decisions related to research policies, funds, research projects, job offers, etc (Bornmann and Leydesdorff, 2014; de Oliveira et al., 2019; Durieux and Gevenois, 2010). So, the results presented in this study and the comprehensive picture obtained about the e-government field make it an effective supplement to the expert evaluation. Thus, researchers, research managers, policymakers, institutions, funding agencies, and governments can rely on it.
2. Literature review
The evolution of a field can be traced using two approaches, research review techniques and bibliometric methods. The research review helps to develop theories in a field and propose new ones by providing an overview of knowledge development (Webster and Watson, 2002). Moreover, it provides suitable research frameworks to reveal areas that need more attention (Bindu et al., 2019a).
Although there are several reviews in the e-government field (Bannister and Connolly, 2012; Dawes, 2008; Grönlund and Horan, 2005; Irani et al., 2007; Meijer and Bekkers, 2015; Yildiz, 2007), the multidisciplinary nature of e-government has made many researchers focus on reviewing specific topics (Bindu et al., 2019a). For example, there are reviews on e-democracy (Chadwick, 2003), government social media (Medaglia and Zheng, 2017), e-participation (Sæbø et al., 2008), and open government data (Çaldağ and Gökalp, 2022).
In addition to traditional reviews, there is a research review technique, which provides clear and coherent conclusions that are difficult to obtain through other methods such as descriptive analysis or systematic review. It is called a meta-analysis. It is used as a powerful statistical technique to provide general inferences based on the results of previous statistical analyses (Alcaide-Muñoz et al., 2017a). There are some meta-analytic studies that dealt with e-government issues, such as transparency in governments (Alcaide-Muñoz et al., 2017b; Rodríguez-Bolívar et al., 2013), public e-services (Arduini and Zanfei, 2014), and citizen adoption of e-government (Rana et al., 2015). These reviews did not broadly address the e-government field. Therefore, with the increasing volume of documents and the diversity of research trends, can be used bibliometric techniques to provide a more comprehensive picture of e-government research (Bindu et al., 2019a).
This paper focuses on studies that dealt with the field of e-government from the bibliometric perspective. The bibliometric studies are “very important because they do not only serve as a synopsis of existing research, but also as an identifier of emerging trends, gaps, and areas for future studies, offering a descriptive state of the art in the field of e-Government” (Alcaide-Muñoz et al., 2017a).
Several studies applied bibliometric analysis to e-government literature. Some of these studies analyzed e-government research globally (Almeida et al., 2014; Cheng and Ding, 2012; Ismayilova, 2014), while others focused on a geographical area (Dias, 2016, 2019; Przeybilovicz et al., 2014), subject field (Alcaide-Muñoz and Rodríguez-Bolívar, 2015; Rodríguez-Bolívar et al., 2016), or specific sources (Dwivedi, 2009; Dwivedi and Weerakkody, 2010; Erman and Todorovski, 2010; Grönlund, 2004; Joseph, 2013).
Cheng and Ding (2012) conducted a bibliometric analysis on 2,232 documents using the CiteSpace II software to analyze data from the Web of Science. They concluded that the hotspots in e-government research were “cross-sectoral collaboration”, “construction of e-government” and “security infrastructure design”, while “performance evaluation” was the research front of e-government. Almeida et al. (2014) analyzed 4,225 documents from Web of Science using HistCite software. They found that the e-government research testified a reasonable decrease, and the top 10 countries, led by the United States and the United Kingdom, accounted for 79.50% of all citations in this field. On the other hand, Ismayilova (2014) used Google Scholar to analyze 381 e-government materials published during 2000–2014 and found that the most researched subjects were “Technological and Developmental Issues” and “E-Gov applications” and that the most productive authors were from the United States, Singapore, and the United Kingdom.
Moreover, Przeybilovicz et al. (2014) analyzed 124 e-government research published in Brazilian journals and conferences (2007–2012). They indicated that the production of the publications was prolific. However, they could not determine whether it was increasing or decreasing, they found that there was little use of a conceptual domain, and a collaboration between the authors was limited to a few relationship networks unlike a collaboration between institutions. In another study, Dias (2016) analyzed the Portuguese e-government research indexed in Scopus from 2005 to 2014. He concluded that the number of documents published has consistently grown, as well as the number of citations received by those documents. The most frequent research topics of e-government research were “interoperability and service integration”, “strategies and methodologies” and “quality, accessibility, and usability”. More recently, Dias (2019) conducted a bibliometric analysis on 1,129 e-government research published in the Ibero-American (IA) Community in Scopus (2003–2017) and found that the production in IA is still rising when compared to the world. Besides, e-government research is very heterogeneous in the IA Community.
Furthermore, Alcaide-Muñoz and Rodríguez-Bolívar (2015) applied bibliometric analysis to examine e-government research in ISI in the field of “information science and library science” (2000–2014). They indicated that e-government research in the field is still scarce. While the most frequent research topics were “the implementation of e-government”, “program evaluation and planning”, “e-participation and digital democracy”, “social behavior and user-centered studies” and “information dissemination”. Recently, Rodríguez-Bolívar et al. (2016) analyzed documents (2000–2012) indexed in ISI in the fields of “Public Administration” and “Information Science and Library Science”. They found that “the research approach to e-government remains immature” with various research gaps, especially in developing countries. The most frequent research topics were “e-participation and the use of Web 2.0”, “new technologies and the modernization of management procedures”, and “online public services”.
Finally, several studies focused on specific sources. As for conferences, Grönlund (2004) examined 170 documents published in three major e-government conferences (DEXA, HICSS, and ECEG) to assess the maturity of the e-government field. He found that the e-government field is indeed immature “because theory generation and theory testing are not frequent, case stories (no theory, no data) and product descriptions (no analysis or test) are”. Besides, Erman and Todorovski (2010) used the bibliographic data on documents published in EGOV (2005–2009) to study collaborations by social network analysis. They found that the most influential authors are from outside the EGOV community.
On the other hand for journals, Dwivedi (2009) analyzed 41 e-government articles published in TGPPP (2007–2008). He concluded that “analytical, descriptive, theoretical and conceptual methods were the most dominant research approaches utilized” and that most contributions come from authors “with an information systems background, followed by business and computer science and IT”. Similarly, Dwivedi and Weerakkody (2010) analyzed 90 papers published in IJEGR during 2005–2009. They found that there were “imbalances in terms of author's discipline, gender, and background” and that “very few authors contributed to more than one article in IJEGR”. Finally, Joseph (2013) applied bibliometric analysis on data retrieved from GIQ. He concluded that there was no specific topic dominating e-government research and that the primary focus regions for e-government research were Europe, North America, and Asia.
In addition to traditional bibliometric methods, novel methods have been used to trace the evolution of research. Science mapping is a bibliometric method used to display the structural and dynamic aspects of a scientific field by recognizing the relationships between fields, topics, documents, authors, or countries (Cobo et al., 2011a, b). This method has been used in a limited number of bibliometric studies in the field of e-government (Alcaide-Muñoz et al., 2017a; Bindu et al., 2019a; Corrales-Garay et al., 2019).
For Example, Bindu et al. (2019b) analyzed the research collaboration in e-governance using the co-author network to assess its influence on the advancement of the field. Their results showed significant growth in research collaborations since 2000. Recently, Dias (2020) reviewed 59 articles published between 2002 and 2018 to provide a conceptual map related to e-government to propose an empirical model for determinants of e-government implementation through local governments.
It is noted from the previous review that the most recent studies that discuss e-government research globally did not cover the five past years. Besides, these studies focused on the Web of Science database or Google Scholar. Therefore, we think that these studies may have used small samples, thus they are difficult to rely on them to produce comprehensive indicators of the status of e-government research globally. In addition to restricting some of these studies to specific countries, these studies have relied on an insufficient sample period (Przeybilovicz et al., 2014). Nonetheless, focusing some studies on a specific field is inconsistent with the fact that e-government is a multidisciplinary domain. Despite the importance of the journals and conference proceedings that some studies focused on, consideration of a specific information source “severely limits the unreached theories and practical implications” (Alcaide-Muñoz and Rodríguez-Bolívar, 2015). Moreover, “it may be difficult to extrapolate these findings and make general comments about the entire field of e-government research” (Joseph, 2013).
However, the absence of comprehensive bibliometric studies could mean that “an interesting aspect of e-government research has remained unexamined” (Rodríguez-Bolívar et al., 2016). Thus, some studies recommended conducting a study on e-government research on the international level, so that this study covers a wider period and depends on a more comprehensive database (Almeida et al., 2014; Joseph, 2013; Przeybilovicz et al., 2014). That is why we present this study.
3. Methodology
3.1 Data source
The Scopus database and the DGRL have been utilized to extract the analyzed data in this study. Scopus database was chosen for its wide coverage of intellectual production compared to other databases, especially Web of Science (Aghaei Chadegani et al., 2013; Aksnes and Sivertsen, 2019; Vieira and Gomes, 2009), thus implying a larger sample. A similar approach was used by Dias (2019) in his study of e-government research in the Ibero-America Community.
Furthermore, the DGRL is a specialist database in the e-government field. Version 16.6 of this database contains 14,940 peer-reviewed works on e-government (Scholl, 2021a). However, we found that many records in DGRL lacked sufficient information to analyze and answer our study questions. Most of this unavailable information was the authors' addresses and citation data. To overcome this problem, the following steps were followed in the data collection stage.
3.2 Data collection
The data set related to e-government research in the last two decades from 2000 to 2019 was retrieved in two phases:
In the first phase, the DGRL has been relied upon as a starting point for collecting data related to the e-government field. The records available on the DGRL published during 2000–2019 amounted to 13,319 records. These records were searched in the Scopus database to obtain complete data for each record, including citation data.
In Scopus, 5,722 records available in the DGRL were searched with a DOI field. As for the remaining 7,597 records in the DGRL, a search strategy was used to search for them in Scopus. This search strategy was based on the first author's last name, date of publication, and the main title of the work.
Thus, 10,399 records were retrieved from the Scopus out of 13,319 records available on the DGRL. This number represents about 78% of the DGRL records in the period of the analysis.
As for the second phase of data collection, several previous studies were relied on (Almeida et al., 2014; Dias, 2019; Lv and Ma, 2019) to formulate a search strategy with various terms related to e-government in the title, abstract, and keyword fields. 19,944 records were retrieved from Scopus using the formulated strategy (Appendix).
In general, a total of 30,313 records were retrieved in the two phases. Overlapped records were deleted depending on the EID field, which is a unique code available on each record in Scopus.
Non-research works such as retracted, editorial, note, short survey, erratum, business article, letter, data paper, undefined, and conference review were excluded.
3.3 Data pre-processing
The 10,399 records overlapped between Scopus and the DGRL have been reviewed by the DGRL team, and thus it was previously confirmed that they are within the coverage of the e-government field.
Regarding other records retrieved from Scopus only and not available in the DGRL database, the inclusion of strategy terms in all three content fields (title, abstract, and keywords) simultaneously was considered acceptable evidence that the work was within the field scope.
As for the works in which strategy terms were mentioned in only one field (4,809), we reviewed them manually to ensure they belonged to the scope of the study based on bibliographic information.
The data was filtered, and lists of authors, countries, institutions, sources, and keywords were created to detect duplicate and misspelled items.
Finally, the record duplicates were detected based on author names, titles, and sources. Thus, the final data set consisted of 21,320 records.
3.4 Data analysis and visualization
Bibliographic data such as author names, affiliations, document title, year of publication, source title, document type, research area, document language, and the number of citations were collected to determine quantitative and qualitative trends and identify the most productive and influential items. The authors' keywords were collected to discover current and future subject trends. Finally, data related to institutions and countries were extracted from the author's address to identify the collaboration networks.
We used some bibliometric tools to achieve the objective of the study. Bibexcel (Persson et al., 2009) and Biblioshiny (Aria and Cuccurullo, 2017) were used to analyze bibliographic data such as date of publications, types of publications, and language of publications to determine the characteristics of e-government research. Besides, bibliometric indicators such as the number of publications, the number of citations, and the h-index were used. Finally, the regression analysis was used to generate a time curve for the number of documents.
Moreover, VOSviewer (van Eck and Waltman, 2021) was used to identify collaboration networks and subject trends for e-government research. We used the statistical cluster and keyword co-occurrence techniques as an effective tool to understand the research topics, provide a clear picture of the knowledge structure (Wang et al., 2020), and identify the future research directions and common topics in e-government research (Zhu et al., 2020).
4. Results and discussion
4.1 Data set distributions
The total number of documents retrieved from 2000 to 2019 was 21,320 and received 263,179 citations. Of these, 4,065 (19.07%) were open access documents, which received 65,099 (24.74%) citations. Open access documents usually have higher citations than others (Amjad et al., 2022). Therefore, the average citation for open access documents was as high as 16 citations per document, compared to the overall average citation for study documents of 12.34 citations.
4.1.1 Number of documents and evolutionary trend of their growth and impact
Figure 1 shows that the number of documents increased gradually from 2000 to 2008 and reached a peak at the beginning of the second decade in 2010–2011. However, the number of documents began to decline after that in the period from 2012 to 2015. Then it began to rise gradually again from the middle of the second decade to its end. We could notice that the total production doubled in the second decade, although a gradual decline in 2012 and 2015.
This increase is due to the continued interest in several topics, such as “e-services” and “local e-government”. Moreover, there has been a noticeable interest in topics related to “e-democracy”, “e-participation”, and “transparency”, as well as a rise in interest in new topics such as “smart cities”, “open government”, “open government data, OGD”, and “social media” (see Section 4.4.2). Many governments have invested heavily in providing widespread and proactive online services (Alcaide-Muñoz et al., 2017a).
Furthermore, there are currently more outlets that help publish more research contributions in the field (Scholl, 2020c). Especially the annual growth rate of e-government research has reached 21.50%. This impressive development shows that e-government research is rapidly growing. Hence, it reflects the increasing global interest in the field.
To predict future growth trends, the linear regression model was used to create a time curve for the number of documents (Figure 1). The results of the regression model demonstrated that there was a significant relationship between the number of documents and the explanatory variable (the time). It is shown that the strength of the relationship between the dependent variable (the number of documents) and the explanatory variable (the time) is high (%93.19). The curve indicates that although the number of documents decreased in the middle of the second half of the last decade (2015), the increase that began in 2016 is expected to continue gradually.
On the other hand, the situation is not the same for the impact. The number of citations increased precariously from 2000 to reach a peak in 2012. Figure 1 illustrates that although the number of citations witnessed a boom at the beginning of the second decade, it decreased from the middle to the end of the second decade.
However, the impressive increase in the number of citations at the beginning of the second decade undoubtedly indicates the growing importance of e-government research. It is despite the marked decline in the number of citations at the end of the decade. This decrease is an expected phenomenon because documents usually start to receive citations within a reasonable time after publishing, and it needs some time to get highly cited (Almeida et al., 2014).
4.1.2 Types of documents
It is clear from Figure 2a that conference papers were at the top of the list with 49.74% of the total documents, followed by articles (41.03%), book chapters (8.33%), and books (0.89%).
Looking at the situation over time, it is noticed that the order of document types did not differ much. However, the number of journal articles published gradually increased from the beginning of the first decade until the end of the second decade, which may indicate an increase in the role of articles in the future. As for conference papers, it has also witnessed a gradual increase from the beginning of the first decade until the beginning of the second decade. Then it started declining before rosing again at the end of the second decade.
Most documents related to e-government appear in conference papers. This can be explained by the fact that e-government started as a practitioner field (Grönlund, 2004). Besides, there are many annual conferences dedicated to e-government, such as “DG.O”, EGOV, and “ICEGOV” as well as the smaller tracks in the PACIS, AMCIS, and ECIS conferences (Grönlund and Horan, 2005; Heeks and Bailur, 2007; Scholl, 2020c).
Figure 2b shows that despite the significant role that conference papers play in e-government literature, the impact of journal articles far exceeds conference papers over time. This phenomenon may be because the number of open journal articles related to e-government is more than (2,265 articles by 25.89% of the total articles) the number of open conference papers (1,591 by only 15% of the total conference papers). It is well known that documents published according to open access get more citations than others (Gargouri et al., 2010; Lawrence, 2001; Piwowar et al., 2018; Tang et al., 2017). Accordingly, open articles received 54,477 citations with 20.70% of the total citations, while open conference papers received only 9,753 citations, representing 3.71% of the total.
4.1.3 Languages of documents
Admittedly, English is the universal language of scientific research. Figure 3 shows that English has dominated most e-government research in both production and impact. Documents written in English reached 20,663 documents with 261,729 citations, accounting for 96.92 and 99.45% respectively. Spanish came in second place, with 176 documents (0.83%), followed by Chinese (0.72%), Portuguese (0.46%), German (0.27%), Russian (0.17%) and French (0.17%). The other languages reached 20 are distributed over 99 documents, representing 0.46% of the total number of documents (Figure 3a).
As for the impact, the number of citations for documents in Spanish and Portuguese exceeded the number in Chinese documents. Moreover, despite the paucity of documents, the impact of Lithuanian documents has gone beyond German, Russian, and French (Figure 3b).
4.2 Characteristics of the authorship on e-government research
4.2.1 Author level
In general, 30,539 authors participated in publishing the retrieved documents, with an average of 1.43 authors per document. The total number of single-authored documents was 4,882 (22.9%) documents produced by 3,445 (11.28%) authors, while the co-authors' documents reached 16,438 (71.1%) produced by 29,031 (95.06%) authors.
Table 1 provides the authors' productivity data using Lotka's Law (Lotka, 1926). According to Lotka's Law, the theoretical result is that about 60% of authors make only one contribution in a field. Regarding the e-government field, the percentage of these authors is 71.64%. We conclude that there is a clear dispersion in the e-government research, as the contribution of many authors is limited to only one document. This can be explained by the multidisciplinary nature of the e-government field, which requires theories and methods from different disciplines to explain and understand the issues related to e-government and its various aspects (Dias, 2019; Irani and Dwivedi, 2008; Liu and Avello, 2021).
Moreover, Figure 4 shows the top 20 authors, according to the number of documents, citations, and h-index. The production of the top 20 authors did not exceed 10% of the total documents, while the influence of the top 20 authors exceeded 25% of the total citations. Janssen, M. (from the Delft University of Technology in the Netherlands) topped the list regarding documents, citations, and h-index followed by Gil-Garcia, J.R. (from State University of New York Albany in the United States) in second place in terms of the number of documents, citations, and fifth in terms of h-index.
It is noted that the only author that was mentioned in the h-Index ranking and did not appear among the top 20 authors in terms of the number of documents or citations is Grönlund, Å. (from Örebro University in Sweden), who came in seventeenth place.
It is also noticed that the authors' rank has changed according to the type of arrangement (the number of documents, citations, and h-index), except Janssen, M., who topped the ranking in the three lists together.
According to Figure 5, only three of the twenty authors had documents since 2000 (Pardo, T.A., Wimmer, M.A., and Tambouris, E.). The authors' production witnessed booms, most of which were during the second decade, except for Weerakkody, V. (in 2009) and Peristeras, V. (in 2007). However, the production of the 20 authors did not follow a clear pattern of growth. While some authors' production has risen in recent years, such as Janssen, M., Gil-Garcia, J.R., Charalabidis, Y., and Rodríguez Bolívar, M.P., the production of some other authors such as Tarabanis, K., Irani, Z., Loukis, E., Jaeger, P.T., and Bertot, J.C. has decreased.
As for the influence of these authors, the number of citations has decreased in recent years in general, except for citing Dwivedi, Y.K. In addition, the number of citations reached its peak at the beginning of the second decade in most cases.
4.2.2 Institutional level
The results showed that there are 11,789 institutions to which researchers publishing e-government research are affiliated. Figure 6 shows the top 20 institutions in e-government literature. It is noticeable that even though the production of the top 20 institutions did not exceed 15% of the total documents, the impact of the top 20 institutions has exceeded 30% of the total citations.
The University at Albany (SUNY) ranked first, both in terms of the number of documents and citations. The Delft University of Technology came second in terms of the number of documents, followed by Brunel University London. However, Delft University of Technology ranked fourth in terms of the number of citations, after Brunel University London and the University of Maryland.
Moreover, it shows that some institutions were among the top 20 influential institutions, but they were not among the top 20 most productive institutions. These institutions included Virginia Polytechnic Institute and State University, Utrecht University, Texas A and M University, University of California, University of Illinois, and the University of Oxford.
It is found that United States institutions are the most productive and influential. As four American institutions appear in the top, most productive institutions in e-government research, followed by Greece (3 institutions), the United Kingdom (2 institutions), and the Netherlands (2 institutions). Additionally, there are eight American institutions in the most influential institutions, followed by the United Kingdom (3 institutions), the Netherlands (3 institutions), and Greece (2 institutions).
4.2.3 International level
Figure 7 shows geographic distributions of e-government research as it was published by researchers in 158 countries. While Figure 8 presents the top 20 countries on e-government research sorted by the number of documents and citations. The results illustrate that more than a third of the documents (34.43%) were published by only three countries: The United States (17.06%), the United Kingdom (9.24%), and China (8.13%).
The United States and the United Kingdom led e-government research in terms of impact. They accounted for nearly half of the total number of citations with 32.75 and 14.91%, respectively. The Netherlands ranked third with 7.2% despite producing only 4.55% of documents, followed by Australia (5.18%) and Spain (5.07%).
Although China came third in terms of the number of documents, it came in tenth place in terms of the impact, with 3.6% of citations. This may be because most Chinese researchers published their research on e-government in conference papers, which accounted for 67.82% of the total documents and which turned out to be less impact compared to journal articles (see Section 2.1.2).
Although some countries were among the top 20 influential countries, they were not among the top 20 most productive countries, such as Mexico, Denmark, Taiwan, Singapore, Ireland, Norway, and France. We conclude from this result that the quality of the research plays a vital role in ranking the most influential countries in e-government research.
By comparing the above results at the institution level with the international level (Figures 6 and 8), there are five institutions among the most productive institutions in e-government research that do not belong to any of the twenty most productive countries. This indicates that the field of e-government research is attracting the interest of various institutions around the world, which may belong to countries other than the major players in the publication in the field. These institutions are the National University Singapore (Singapore), Centro De Investigación Y Docencia Económicas (Mexico), National University of Ireland (Ireland), University of Agder (Norway), and Copenhagen Business School (Denmark). Besides, it is observed that the countries to which the five institutions belong are among the most influential countries (Figure 8), which confirms the previous conclusion.
4.3 Sources of e-government research
Figure 9 shows the top 20 sources of e-government research, ranked by the number of documents. The top 20 sources produced 8,565 documents with 40.17% of all the documents. ACM International Conference Proceeding Series (ACM) was at the top of the most productive sources of e-government research with 1,706 documents (8% of all the documents), followed by Lecture Notes in Computer Science (LNCS) with 1,582 documents (7.42%), and Government Information Quarterly (GIQ) with 745 (3.49%).
Figure 9 also shows the core sources of e-government research according to Bradford's Law (Bradford, 1934), which numbered 10 sources.
The ranking of the most influential sources is shown in Figure 10. Overall, the top 20 influential sources accounted for 50.55% of the citations received by e-government documents. The results confirm the highly influential role that GIQ plays, as it alone received 17.79% (46,814) of the total citations in the e-government research, even though its production does not exceed 3.5% of the documents. It is due to the high quality of the papers published in GIQ because of a rigorous peer-review process and the stability of the journal publication sequence. Thus, it predicts the continuing influential role of GIQ in the future.
LNCS came second, with 9,790 citations (3.72%), followed by the Hawaii International Conference on System Sciences (HICSS) which received 9,635 (3.66%), and the Public Administration Review (PAR) which received 8,966 (3.41%) though it had only 67 documents (0.31%).
It is noticed that the impact of journals is greater than the impact of conferences, although conferences produce a higher number of documents in the field. This coincides with the results for document types (see Section 2.1.2). For example, The ACM that published the most documents (8%) came in fifth place in terms of the number of citations (4.28%).
In general, journals received 204,947 citations (77.87%) of the total citations, while conference proceedings received only 48,764 citations (18.53%) of the total citations. Interestingly, this result copes with a previous study that retrieved results from the web of science (Almeida et al., 2014). Nevertheless, conferences remain good channels for scientific and academic publishing, as they are very used to present new ideas and documents in the form of work in progress (Almeida et al., 2014).
Moreover, Figure 11 shows that some sources reached their peak production during the first decade, such as LNCS, ELECTRONIC GOVERNMENT (EG), and IFIP ADVANCES IN INFORMATION AND COMMUNICATION TECHNOLOGY (IFIP AICT), while most of the sources reached their peak production in the second decade, such as ACM, GIQ, and Transforming Government: People, Process and Policy (TGPPP).
Except for ACM and GIQ, the top 20 sources of e-government research production witnessed instability in the number of documents produced, punctuated by mutations in production at intermittent intervals. GIQ and ACM have seen a gradual increase in document production over the past two decades, except for 2015, which saw a marked decline in ACM production.
On the other hand, most of the sources were not stable in terms of the number of documents. For example, although the production of LNCS increased significantly during the period 2002–2004, it subsequently declined and then increased again in the period 2009–2011. Also, the European Conference on E-Government (ECEG) witnessed an unstable increase until the first half of the second decade (2000–2013) and then began to decline during the second half (2014–2019).
It is noted that all sources published documents during the first decade except for PUBLIC ADMINISTRATION AND INFORMATION TECHNOLOGY due to their appearance at the beginning of the second decade. Although documents of the International Conference on Electronic Government (EGOV) have been published in just four years (2006–2009) at the end of the first decade, it was ranked 20 on the list of sources.
Regarding the influence of the more prolific sources, most of the sources experienced an unstable situation regarding the number of citations they obtained. However, there were mutations in some years, such as a rise in citations for documents published in GIQ, HICSS, and IJEGR in 2012. It is noted that most of these mutations occurred in the first decade and the beginning of the second decade, except for IFIP ADVANCES IN INFORMATION AND COMMUNICATION TECHNOLOGY (IFIP AICT) (in 2001) and SOCIAL SCIENCE COMPUTER REVIEW (SSCR) (in 2003–2004).
4.4 Mapping analysis
In this section, we performed a science mapping analysis of documents related to e-government by VOSviewer.
4.4.1 Co-authorship analysis
Katz and Martin (1997) defined a research collaboration as “the working together of researchers to achieve the common goal of producing new scientific knowledge”. Direct collaboration between two or more researchers is the basic unit of cooperation. However, there are other levels of collaboration such as collaboration between research groups within the department, between institutions, between countries, etc (Katz and Martin, 1997).
To describe the collaboration relationship in documents related to e-government from 2000 to 2019, we make a co-authorship analysis to measure publication links between authors, institutions, and countries.
4.4.1.1 Authors' collaboration
The analysis identified 30,539 authors, of which 463 authors published at least 10 documents. There were 334 of 463 authors connected. In Figure 12, the size and color of the nodes (items) describe the degree of collaboration and clusters of authors. If there is a line between two authors, then this means that there is a cooperation relationship between them. Besides, the larger nodes indicate a more collaborative relationship.
It shows that the ten authors with the highest total link strength were Janssen M. (total Link Strength = 392 times with Links = 80), Gil-Garcia J.R. (288 with 43), Charalabidis Y. (280 with 28), Weerakkody V. (240 with 32), Tarabanis K. (213 with 31), Luna-Reyes L.F. (203 with 21), Dwivedi Y.K. (202 with 26), Pardo T.A. (190 with 39), Tambouris E. (165 with 43), and Irani Z. (150 with 23).
The authors published at least 10 documents that were classified into 35 clusters. More than half of the production of these authors was concentrated in only 9 clusters:
The green cluster (951 documents with 24,629 citations) included 27 authors, headed by Gil-Garcia J.R. (178 with 5,362), Pardo T.A. (105 with 5,294), and Luna-Reyes L.F. (102 with 1,267).
The red cluster (721 with 13,780) included 30 authors, headed by Weerakkody V. (132 with 3,597), Irani Z. (72 with 2,404), and Kamal M.M. (38 with 832).
The blue cluster (398 with 17,371) included 20 authors, headed by Meijer A.J. (41 with 2080), Bannister F. (40 with 936), and Shareef M.A. (30 with 1,176).
The yellow cluster (398 with 12,042) included 14 authors, headed by Dwivedi Y.K. (119 with 4,673), Rana N.P. (56 with 2072), and Gupta M.P. (40 with 908).
Mint cluster (393 with 7,227) included 9 authors, headed by Charalabidis Y. (145 with 2,450), Loukis E. (70 with 994), and Zuiderwijk A. (54 with 2,308).
The purple cluster (393 with 8,596) included 8 authors, headed by Janssen M. (298 with 7,145), Saxena S. (24 with 244), and Wagenaar R. (15 with 205).
The pink cluster (372 with 5,514) included 11 authors, headed by Tarabanis K. (108 with 1,568), Tambouris E. (83 with 1,006), and Macintosh A. (39 with 1,014).
The light blue cluster (336 with 6,503) included 12 authors, headed by Wimmer M.A. (93 with 871), Scholl H.J. (79 with 2,970), and Klischewski R. (32 with 669).
The light orange cluster (289 with 3,393) included 9 authors, headed by Ojo A. (93 with 1,046), Janowski T. (60 with 1,018), and Estevez E. (57 with 786).
4.4.1.2 Institutional collaboration (institutions)
The analysis of the institutions to which the authors of e-government documents were affiliated revealed that of 11,789 institutions, 580 institutions published at least 10 documents. There were 553 of 580 institutions connected. Figure 13 shows that the ten institutions with the highest total link strength were the University At Albany, SUNY (Total Link Strength = 299 Times with Links = 80), Delft University of Technology (250 with 43), Brunel University London (193 with 28), Centro De Investigación Y Docencia Económicas (166 with 32), Universidad De Las Americas Puebla (134 with 31), University of The Aegean (120 with 21), Swansea University (119 with 26), University of Macedonia (111 with 39), National and Kapodistrian University of Athens (92 with 43), and Danube University Krems (88 with 23).
Institutions can be categorized into 23 clusters; each cluster is shown with a distinct color in Figure 13. More than half (55.6%) of the production of these institutions was concentrated in only 8 clusters:
The red cluster (1,412 documents with 41,858 citations) included 52 institutions, headed by the University of Maryland (144 with 7,682), Seoul National University (69 with 1,046), and Syracuse University (63 with 2,135).
The cyan cluster (1,332 with 29,048) included 37 institutions, headed by Brunel University London (326 with 7,729), University of Twente (123 with 2,962), and Erasmus University Rotterdam (96 with 1,796).
The green cluster (1,182 with 26,282) included 47 institutions, headed by Copenhagen Business School (102 with 2,202), University of California (93 with 2,618), and Virginia Polytechnic Institute and State University (65 with 4,183).
The yellow cluster (1,095 with 12,956) included 41 institutions, headed by the University of the Aegean (186 with 2,786), the University of Piraeus (72 with 732), and the University of South Africa (54 with 292).
The light orange cluster (966 with 20,375) included 19 institutions, headed by Delft University of Technology (367 with 7,598), Swansea University (128 with 4,742), and University of Oxford (72 with 2,309).
The pink cluster (949 with 12,019) included 28 institutions, headed by the National and Kapodistrian University of Athens (213 with 2,305), University of Koblenz-Landau (94 with 576), and University of Ljubljana (69 with 729).
The purple cluster (940 with 20,319) included 38 institutions, headed by the National University of Singapore (114 with 3,424), Nanyang Technological University (49 with 1,140), and Griffith University (43 with 761).
The blue cluster (918 with 11,559) included 46 institutions, headed by Wuhan University (72 with 173), California State University (56 with 1,200), and Tsinghua University (52 with 491).
4.4.1.3 International collaboration (countries)
The co-authorship network of countries is shown in Figure 14. Of the 158 countries, 146 are connected. Figure 14a shows that the ten countries with the highest total link strength were the United States (total Link Strength = 1,420 times with Links = 89), the United Kingdom (1,060 with 81), Netherlands (546 with 56), Germany (492 with 56), Greece (466 with 43), Australia (454 with 66), China (430 with 52), Spain (421 with 52), Italy (395 with 43), and Canada (292 with 56).
Countries in the same color in Figure 14a show that they are in the same cluster and may have more cooperation with each other. There are a total of 18 clusters of countries. More than 74.41% of the e-government literature was published by only 6 clusters as follows:
The brown cluster (4,531 documents with 102,057 citations) included 7 countries, headed by the United States (3,637 with 86,197), Canada (514 with 9,941), and Mexico (293 with 5,625).
The green cluster (3,614 with 28,767) included 21 countries, headed by China (1,734 with 9,472), Malaysia (466 with 3,184), and Japan (213 with 1,187).
The cyan cluster (2,356 with 25,353) included 12 countries, headed by Germany (999 with 9,837), Spain (925 with 13,346), and Argentina (103 with 604).
The purple cluster (2,172 with 17,192) included 12 countries, headed by India (816 with 6,735), Indonesia (529 with 2,333), and Sweden (509 with 5,525).
The pink cluster (1,684 with 16,880) included 7 countries, headed by Italy (901 with 10,482), Portugal (346 with 2,624), and Turkey (187 with 2,426).
The red cluster (1,565 with 11,039) included 22 countries, headed by Russian Federation (349 with 1,896), Belgium (213 with 2,921), and Estonia (135 with 872).
We conclude from Figure 14a that most countries displayed a clear geographic pattern of cooperation. Where the brown cluster mainly includes North American countries, the cyan cluster consists of South American countries, as well as two European countries, Germany and Spain. The cooperation of Northern and Central European countries is also shown in the red cluster, such as Russian Federation, Estonia, the Czech Republic, Romania, Serbia, and Slovakia, as well as Belgium, a Western European country.
The green cluster mainly included Asian countries such as China, Malaysia, Japan, Singapore, Saudi Arabia, Iran, the United Arab Emirates, and Pakistan. Likewise, the purple cluster included South Asian countries such as India, Indonesia, Thailand, Bangladesh, and Sri Lanka. Finally, the blue cluster included mainly African countries such as South Africa, Nigeria, Kenya, Ghana, Namibia, and Mauritius.
It is noticed that the developed countries have broad cooperation relations, unlike the developing countries. Although there is cooperation among many developing countries as shown in Figure 14a, the cooperation among them as well as the cooperation with leading countries is still limited. It is evident in countries such as Tunisia, Algeria, Namibia, Ghana, Libya, Yemen, Morocco, Cambodia, Kenya, Cameroon, Kazakhstan, Mauritius, Ukraine, and Costa Rica.
It indicates that one of the main reasons for the gap in e-government research in developing countries is the weakness of international cooperation, especially with the leading countries in the field. Therefore, the developing countries should enhance cooperation with the leading countries to transfer successful practices and experiences, which will enhance their chances of success.
Moreover, in Figure 14b, countries were color-coded based on the average publication year. Colors toward the blue end of the spectrum indicate an earlier date for the average publishing year of countries, while countries that have a later average publishing year are indicated by colors toward the red end. 2012 is the most year in terms of average publication with 12,355 documents, followed by 2013 (4,525) and 2014 (2014).
Most items in Figure 14b appear between green and yellow. 2014 is the most average publishing year for 28 countries, followed by 2015 (25 countries), 2013 (22 countries), and 2012 (20 countries).
The past five years (2014–2019) represented the average publishing year for several countries (colored toward the red end of the spectrum in Figure 14b), which predicts the important role that these countries can play soon. India comes on top of these countries (Published 816 documents with total Link Strength = 232 times and 55), followed by Indonesia (529, 165, and 29), Malaysia (466, 156, and 31), Russian Federation (349, 73, and 31), Portugal (346, 157, and 38), South Africa (332, 122, and 33), Saudi Arabia (171, 123, and 32), Estonia (135, 64, and 25), Ecuador (82, 63, and 13), and Colombia (80, 37, and 12).
4.4.2 Co-occurrence analysis
Keywords provide information about the main content of the document and can also be used to determine search trends in a specific area (Lee et al., 2020). Therefore, regarding the concepts and terms used in research on e-government, we analyzed networks of co-occurrences of keywords. To this end, we scanned all documents in the dataset (Scopus output files) using VOSviewer to extract authors' keywords. Scopus-generated keywords were excluded in order not to rely on ex ante preprocessed and pre-manipulated data. The author keywords were relied upon because these authors are best suited to describe the content of their documents. The documents that contained author keywords amounted to 16,477.
For analytical purposes, keyword term-consolidation tactics were employed, for example, consolidating the keyword terms “egov,” “e-government,” “eGovernment,” “electronic government,” “digital government,” etc. into one keyword.
We chose a value of 10 occurrences for the minimum number of occurrences that keywords must achieve to be included in the co-occurrence network. The analysis identified 26,792 keywords, of which 860 met the threshold of at least 10 occurrences, such that the final network comprised 860 keywords.
Figure 15 shows the keywords co-occurrence network. A node represents a keyword, and the bigger the node is, the more documents the keyword has. The final network keywords appeared a total of 38,015 times across all documents. The keywords are categorized into 11 clusters, represented by colors in Figure 15a. These clusters reflected the most prominent topics of research on e-government to date. These eleven main themes can be labeled as:
“E-government services, acceptance and adoption” contained 143 keywords that occurred 6,545 times, headed by “e-services”, “trust”, and “adaptability”.
“Local e-government information systems and interoperability” contained 95 keywords that occurred 4,163 times, the most important of which were “local e-government”, “interoperability”, and “information systems”.
“E-government implementation and case studies” contained 88 keywords that occurred 2,800 times, including “case study”, “business process modeling”, and “information security”.
“Transparency in e-government and public administration” contained 85 keywords that occurred 3,011 times, headed by “transparency”, “public administration”, and “accountability”.
“E-democracy (e-participation) and social media” contained 85 keywords that occurred 5,264 times, including “e-democracy”, “social media”, and “e-participation”.
“Public services and digital divide” contained 73 keywords that occurred 2,131 times, headed by “digital divide”, “public services”, and “knowledge management”.
“Open data and open government” contained 71 keywords that occurred 4,183 times, including “open government”, “open data”, and “big data”.
“Smart cities and Internet of things” contained 68 keywords that occurred 3,020 times, including “smart cities”, “Internet of things”, and “citizen participation”.
“ICT and cloud computing in the public sector” contained 64 keywords that occurred 3,576 times, headed by “ICT”, “public sector”, and “cloud computing”.
“Security and privacy in e-voting and e-identity” contained 48 keywords that occurred 1,832 times, including “security”, “Privacy”, and “e-voting”.
“GIS and e-health” contained 38 keywords that occurred 1,043 times, headed by “geographic information systems, gis”, “e-health”, and “open source”.
The identified keywords were then color-coded by VOSviewer based on the average publication year. In Figure 15b, colors toward the blue end of the spectrum indicate an earlier date for the appearance of the keyword, while keywords that appeared later are indicated by colors toward the red end.
To trace the eleven main topics and identify the evolution of the researchers' use of the keywords, we divided the two decades into four time periods (Figure 16). The literature in the period 2000–2004 contained 1,525 keywords, of which 29 met the threshold of at least 10 occurrences. This is due to the lack of documents produced during this period. In addition, many authors did not consider including keywords in their research.
This early period of e-government research (2000–2004) witnessed an interest in implementing ICT in public organizations to modernize public administration and provide public services via the internet. Therefore, As shown in Figure 16a, the term “Internet” was the most frequently used by 56 times, followed by “ICT” (50 times). It is evident through the inclusion of the implementation of ICT within political agendas and strategic programs of governments (Gil-García and Pardo, 2005).
Interestingly, a topic such as “e-democracy” (35 times) emerged at this early stage of the e-government field, and interest in this topic increased during the following period (2005–2009), as shown in Figure 16b.
ICT applications benefit societies in improving democratic models. For example, the adoption of electronic voting systems indicates the ability to positively influence democratic deliberation and citizen participation in politics (Bannister and Connolly, 2007). Therefore, during the period 2005–2009, researchers focused on ICT applications used to enable and enhance the formal voting procedure and e-participation in general. It is evident by the significant increase in the use of some terms compared to the previous period, such as “ICT” (233 times), “e-democracy” (200 times), and “Internet” (132 times).
However, the 2005–2009 period (Figure 16b) witnessed the expansion of research topics in the e-government field. The keywords increased to 6,122, of which 165 met the threshold of at least 10 occurrences. Furthermore, many new terms were used, especially those related to e-participation and sharing and integrating information, such as: “interoperability” (93 times), “e-participation” (80 times), “ontology” (71 times), “digital divide” (70 times), “accessibility” (48 times), “adaptability” (34 times), “semantic web” (33 times), “transparency” (31 times), “GIS” (23 times), “e-procurement” (23 times), “web 2.0” (21 times), “identity management” (20 times), “semantics e-government” (19 times), “information sharing” (18 times), and “open source” (18 times).
Moreover, the term “e-health” (18 times) began to be used in the e-government during this period to refer to the “use of ICT in support of health and health-related fields, including health-care services, health surveillance, health literature, and health education, knowledge and research” (World Health Assembly, 2005). Then, it received increasing attention during the following periods, as shown in Figure 16c,d.
By the beginning of the second decade (2010–2014), the e-government literature increase reflected in the use of the terms, as shown in Figure 16c. The analysis identified 10,379 keywords, of which 303 met the threshold of at least 10 occurrences. The interest in many terms continued, such as: “ICT” (353 times), “e-participation” (261), “e-democracy” (235 times), “interoperability” (143 times), “transparency” (142 times), “Internet” (134 times), “local e-government” (127 times), “public administration” (123 times), and “e-services” (112 times). Besides, many new terms have also emerged. The most used terms were “open government” (252 times), “social media” (225 times), “open government data, OGD” (132 times), “cloud computing” (99 times), “smart cities” (92 times), “government 2.0” (45 times), “linked data” (43 times), “service quality” (38 times), and “big data” (37 times).
The above terms reflect a growing interest in e-participation during 2010–2014, especially at the social networks level and web 2.0 tools as interactive communication channels with citizens. In this direction, researchers have attempted to analyze the use of social media for various purposes, such as political campaigns (Williams and Gulati, 2012) and anti-corruption (Bertot et al., 2010). Moreover, researchers have discussed the relationship between e-participation and trust in local government (Kim and Lee, 2012).
This period also witnessed an increased interest in the topic of transparency, which appeared for the first time during the previous period (2005–2009). Many governments are working to raise the openness and transparency of information disclosure, which leads to less corruption (Alcaide-Muñoz et al., 2017a).
With the increased interest in transparency issues, the use of many topics that support the efforts of openness, anti-corruption, participation, and collaboration have emerged and increased. The “open government” and related terms such as “open government data” were the most prominent. Although open government data has become an important movement between government administrations around the world, research is still limited in this research area.
At the end of the second decade (2015–2019), despite a marked increase in interest in open government. The majority of research has been conducted by Western, democratic, and developed societies, with very little information on its implementation in less developed countries (Safarov, 2020). Thus, future research can be directed to fill this research gap.
In this period (2015–2019), the analysis identified 15,300 keywords, of which 412 met the threshold of at least 10 occurrences. The interest in several terms that appeared in the previous period also continued, especially: “smart cities” (939 times), “open government” (581 times), “open government data, OGD” (468 times), and “social media” (438 times). On the other hand, there is less interest in using some general terms such as “interoperability” (103 times), “digital divide” (89 times), “Internet” (83 times), “ontology” (58 times), and “web 2.0” (38 times).
Thus, the focus of researchers in this period (2015–2019) is on case studies and tools of public services delivery in the context of what is currently called “smart cities”, such as “intelligent transportation systems”, “healthcare”, “education”, “public safety”, and “energy and water management applications”. Besides, the researchers focus on the “Internet of Things” (IoT) which is one of the innovative technologies that are being applied in smart cities to transform many services into digital form, improve the lives of citizens, and address various services, security, and administrative challenges (Evertzen et al., 2019).
Despite the clear interest in smart cities, little research has concentrated on actual practices of citizen participation in smart cities to date (Meijer and Rodríguez-Bolívar, 2015). Moreover, there is a need to assess the contribution of smart city governance to economic growth (Meijer and Rodríguez-Bolívar, 2015). It would therefore be beneficial to direct future research to discuss these issues.
Recently there has been a trend to use more specialized terms. It indicates that e-government studies are paying more attention to how to benefit from modern web applications and artificial intelligence in the e-government field. It is shown through the use of specialized terms such as: “Internet of things” (197 times), “blockchain” (95 times), “digitalization” (72 times), “artificial intelligence (ai)” (67 times), “cyber security” (63 times), “digital transformation” (52 times), “mobile applications” (46 times), “co-creation” (46 times), “machine learning” (45 times), “sentiment analysis” (43 times), “systematic review” (32 times), and “natural language processing” (31 times).
Analysis shows that blockchain technology is among the topics that have attracted interest recently. There is a noticeable focus on the impact of blockchain technology on various sectors such as government and healthcare (Alam et al., 2022). It can provide quantitative and qualitative improvements in government services, more transparency, access to government information, and information participation across different organizations (Hou, 2017). Thus, the issue of how blockchain technology can contribute to the development of e-government can be a fertile subject of research during the coming period. Also, the application of blockchain technology is related to “information security”, “cost”, and “reliability” issues (Hou, 2017). These topics can be discussed and focused on during future research.
Recently, artificial intelligence techniques have been used to advance “the current state of e-government services” to reduce processing times, “reduce costs”, and “improve citizens' satisfaction” (Al-Mushayt, 2019). These issues could constitute research topics soon.
All the above confirms the results shown in Figure 16, which indicate that topics such as “smart cities”, “open government”, “open government data, OGD”, “social media”, “Internet of things”, and “artificial intelligence and data analysis” have attracted more attention in recent years and may attract more attention soon.
5. Conclusions and implications
5.1 Research conclusions
This paper presents a bibliometric analysis of e-government research during the last twenty years (2000–2019) based on the Scopus and the DGRL databases to monitor the status of e-government research globally and track research trends in the field over the past two decades. It is found that the number of documents related to e-government from 2000 to 2019 reached 21,320 documents and received 263,179 citations.
The results showed a highly significant relationship between the number of documents and the time variable, which predicts the gradual growth of e-government research. This is supported by the existence of many research topics that have attracted the interest of a large segment of researchers around the world recently.
Furthermore, there are now an adequate number of publication outlets (Figure 9) that were not available at the beginning of e-government research (Scholl, 2020c). However, we recommend stakeholders increase the number of publication outlets to accommodate the expected increase in e-government research.
It is interesting that despite the significant role played by conference papers in e-government literature, the impact of journal articles far exceeds conference papers. It is concluded that open access is the most prominent reason for this phenomenon. Our analysis shows that the number of open access articles published in journals outnumbers the conference papers. It is also reflected in the citations.
According to Bradford's Law (Bradford, 1934), the number of core sources for e-government research is 10 sources. ACM was at the top of the most productive sources of e-government research, followed by LNCS and GIQ. The results confirm the highly influential role that GIQ plays, as it alone received 17.79% of the total citations in the e-government research, followed by LNCS (3.72%).
The analysis shows that the proportion of researchers who contributed to only one work is higher than the theoretical proportion of Lotka's Law (Lotka, 1926). Hence it can be argued that there is a clear dispersion in e-government research. This dispersion can be explained by the multidisciplinary nature of the e-government field, which requires theories and methods from different disciplines to explain and understand the issues and various aspects related to e-government (Dias, 2019; Irani and Dwivedi, 2008; Liu and Avello, 2021).
Moreover, there were 11,789 institutions and 158 countries to which the researchers publishing e-government research are affiliated. The University at Albany (SUNY) ranked first among these institutions, both in terms of the number of documents and citations. As for the countries, the United States and the United Kingdom led e-government research in terms of the number of documents and citations.
Cluster analysis showed that the developed countries have broad cooperation relations with others, unlike the developing countries. Therefore, one of the main reasons for weak e-government research in developing countries is poor cooperation, especially with the leading countries in the field. Consequently, developing countries should enhance cooperation opportunities with leading countries in e-government research, which will enhance their chances of success.
The analysis also showed some countries that could play a vital role in e-government research in the future. At the forefront of these countries were India, Indonesia, Malaysia, Russian Federation, Portugal, and South Africa. Moreover, based on the analysis of the authors' keyword co-occurrence, several topics such as “smart cities”, “open government”, “open government data, OGD”, “social media”, “Internet of things”, and “artificial intelligence and data analysis” have attracted more attention in recent years and may attract more attention soon.
5.2 Research implications
It is common practice to use the results of bibliometric studies as reliable sources to justify decisions related to research projects, research policies, funding, employment, etc (Bornmann and Leydesdorff, 2014; de Oliveira et al., 2019; Durieux and Gevenois, 2010). Thus, our results can help guide stakeholders to better practices that will improve the performance of their tasks.
For example, according to Scholl (2021b), the DGRL “has potentially begun to play a formative role in the evolution of” the e-government research study field. However, the current study revealed some weaknesses in this database. The DGRL lacks comprehensiveness in its coverage of e-government research. Furthermore, it lacks some essential data in many records, such as citation data and author affiliations. Therefore, the DGRL team can benefit from the results of the study in addressing the mentioned weaknesses and improving the future role that the database plays in e-government research. Also, it is valuable to direct future research to analyze e-government research using other databases such as Web of Science and Google Scholar. Besides, the coverage and overlap between databases related to e-government can be compared.
There is no doubt that the development of e-government in emerging countries contributes to promoting e-government research in those countries and vice versa (Dias, 2019). Analysis showed that several countries have recently started to pay attention to e-government research. Many developing countries had limited cooperative relations. Therefore, these countries should expand the cooperation network in e-government research, especially with the leading countries in the field. Algeria, Costa Rica, Morocco, Yemen, Colombia, Bahrain, Ghana, Kazakhstan, and Ukraine are among these countries. It will help these countries too to benefit from the pioneering experiences and expertise in e-government.
The results help researchers and institutions involved in e-government research to identify gaps and fertile issues in the field. For example, the results showed a growing interest in e-health in e-government research. Recently, COVID-19 has spread in the world, causing the death and injury of millions of people. In light of this crisis, there has been a noticeable increase in the application of e-health all over the world (Webster, 2020). Thus, e-government initiatives in this direction can be an effective area that must be explored in the future. Moreover, researchers could study the impact of COVID-19 on e-government research.
Although there was a clear interest in smart cities, little research focused on the actual practices of citizens' participation in smart cities and on evaluating the contribution of governance to economic growth (Meijer and Rodríguez-Bolívar, 2015). Therefore, it would be beneficial to direct future research to discuss these issues.
There is a need to conduct more research on how blockchain technology can contribute to the development of e-government, as well as to discuss related issues such as “information security”, “reduce cost” and “reliability”. Similarly, there is a need to address the use of artificial intelligence techniques to advance the current state of e-government services.
There was a clear impact of open access on the volume of citations received. It would be that useful to study this phenomenon in more detail in future research. It is similar to some studies that discussed this phenomenon in other fields (Gargouri et al., 2010; Tang et al., 2017).
It is well known that researchers' productivity and citations are used as indicators in promotion, employment, and participation in research projects (Almeida et al., 2014). Journal articles are usually highly regarded by officials. With this aspect, the results help researchers to identify the high-quality sources in e-government research to publish their works. On the other hand, the results make it easier for funding agencies to make decisions concerning funding and employment.
6. Limitations
Even though this study presented many arguments based on its findings, it also carries some limitations. This research used documents published between 2000 and 2019. Therefore, the analysis did not include papers published from 2020 to 2022. This period witnessed the spread of the COVID-19 pandemic globally, which affected all aspects of life. Thus, it would be valuable to study the impact of COVID-19 on e-government research and discuss the e-government role that contributed to the control of this epidemic.
Moreover, there is a limitation regarding the coverage of Scopus and DGRL databases used because they do not include all published e-government research such as master and doctoral theses. So, further studies are recommended to analyze e-government research in other additional types of documents.
Figure 1
The evolutionary trend and the linear regression for the number of documents related to e-government and the annual citation structure during 2000–2019
[Figure omitted. See PDF]
Figure 2
Distribution of documents (a) and citations (b) according to type of documentes (2000–2019)
[Figure omitted. See PDF]
Figure 3
Distribution of documents (a) and citations (b) according to language
[Figure omitted. See PDF]
Figure 4
Top 20 authors in terms of the number of documents, citations, and h-index in the e-government literature (2000–2019)
[Figure omitted. See PDF]
Figure 5
The top 20 authors' production and their influence over time
[Figure omitted. See PDF]
Figure 6
Top 20 institutions in terms of number of documents, citations in the e-government literature (2000–2019)
[Figure omitted. See PDF]
Figure 7
Global geographic distributions on e-government research (2000–2019)
[Figure omitted. See PDF]
Figure 8
Top 20 countries in terms of number of documents and citations in the e-government literature (2000–2019)
[Figure omitted. See PDF]
Figure 9
Top 20 sources on e-government research during 2000–2019, ranked by the number of documents
[Figure omitted. See PDF]
Figure 10
Top 20 sources on e-government research during 2000–2019, ranked by the number of citations
[Figure omitted. See PDF]
Figure 11
The top 20 sources' production and their influence over time
[Figure omitted. See PDF]
Figure 12
Network map of authors published e-government research
[Figure omitted. See PDF]
Figure 13
Network map of institutions of e-government research
[Figure omitted. See PDF]
Figure 14
Network map of countries of e-government research; (a) 146 countries had at least 10 publications and connected to each other; (b) Distribution of countries according to the average appearance time
[Figure omitted. See PDF]
Figure 15
Co-occurrence analysis of e-government research during 2000–2019; (a) Mapping of keywords in the research area, which are divided into 11 clusters; (b) Distribution of keywords according to the average appearance time
[Figure omitted. See PDF]
Figure 16
Co-occurrence analysis of e-government research over four time periods; (a) Mapping of keywords in the research area during 2000–2004, which contained 29 keywords met the threshold of at least 10 occurrences; (b) Mapping of keywords in the research area during 2005–2009, which increased to 165 keywords met the threshold of at least 10 occurrences; (c) Mapping of keywords in the research area during 2010–2014, which identified 303 keywords met the threshold of at least 10 occurrences; (d) Mapping of keywords in the research area during 2015–2019, which identified 412 keywords met the threshold of at least 10 occurrences
[Figure omitted. See PDF]
Productivity of authors according to Lotka's Law
| N. of documents published | N. of authors | % OF total authors |
|---|---|---|
| 1 | 21,877 | 71.64 |
| 2 | 3,964 | 12.98 |
| 3 | 1,534 | 5.02 |
| 4 | 744 | 2.44 |
| 5 | 440 | 1.44 |
| 6 | 307 | 1.01 |
| 7 | 214 | 0.70 |
| 8 | 142 | 0.46 |
| 9 | 105 | 0.34 |
| 10 and more than | 1,212 | 3.97 |
© Emerald Publishing Limited.