E-Government in China: A Content Analysis of National and Provincial Web Sites

Authors

  • Xiang Zhou

    Corresponding author
    1. Doctoral student in the School of Journalism and Electronic Media, University of Tennessee at Knoxville, with new media as a primary concentration and statistics as a secondary concentration. Her major research interests are international news coverage, the Internet and social change, especially the development of the Internet in China. She was employed with the Xinhua News Agency in Beijing, China, from March 1993 through August 2000.
    Search for more papers by this author

Address: 1611 Laurel Avenue #1117, Knoxville, TN 37916. Tel: (865) 946-6054.xs

Abstract

The Chinese government has actively encouraged and invested in the growth of the Internet to capture the technology's vast commercial potential, while exerting state control. Even before the country was fully connected to the global “network of networks,” China began to implement programs that would facilitate e-government. Despite some valuable examinations of e-government in China, no studies were found that systematically explored the content of both national and provincial Chinese government Web sites. This study fills that gap in the literature by analyzing the content of the opening pages of 177 government Web sites. The study found that broad e-governance, e-knowledge, and e-service functions were implemented at sites of national, costal, and inland government units. Generally, national and costal sites were more sophisticated, but inland sites seemed designed to meet the specific needs of government, citizens, and businesses in inland areas. The coding form and method designed for analyzing these sites holds promise for future researchers and the findings suggest that China may have begun to achieve its goals of facilitating government functions via the Internet.

Introduction and Purpose

In an age of information expansion, the Chinese government has actively encouraged and invested in the growth of the Internet to capture the technology's vast commercial potential while exerting state control over the ways that this “network of networks” is accessed by Chinese citizens. The central government has encouraged the competing state-owned telecommunications providers, such as China Telecom, China Unicom, China Mobile, China Netcom, JiTong Communications Company, and China Railway Telecom, to build their own networks. Faced with the potential political or social challenges that the new medium may bring, the Chinese Communist Party (CCP) has utilized multiple control strategies to maximize official authority (Dai, 1999; Hachigian, 2001; Hartford, 2000; Kalashil & Boas, 2001). These efforts at control began early with the “PRC Interim Regulations Governing the Management of International Computer Networks” issued in February 1996 marking the beginning of systematic regulation of the Internet in China (Tan, 1999).

Efforts of the state to control the Internet did not equate to limiting the technology. To the contrary, the Internet boom in China is centered in state-sponsored strategies for comprehensive “informatization (xinxihua),” which is viewed by the state as critical to China's future growth and international competitiveness. On April 18, 1997, the first national meeting of the National Informatization Work Conference, held in Shenzhen, was opened with an address by former Vice-Premier Zou Jiahua, who set out the “24-character direction for China's informatization that included: overall planning (tongchou guihua), state dominance (guojia zhudao), unified standards (tongyi biaozhun), joint construction (lianhe jianshe), links between government units (hulian hutong), and shared resources (ziyuan gongxiang) (“China Information Chronology”).” During the Ninth Five-Year Plan period, the information industry emerged “prominently as a new growth point of China's national economic development” (China Internet Information Center, 1995). In November 2001, China's State Economic and Trade Commission published the “Tenth Five-Year Plan of Industrial Structure Adjustment.” One of the six major adjustments was “to promote industrialization by informatization” (“China's Information Industry,” 2001).

With the ambition of realizing informatization and the recognition of the potential political challenges brought by the Internet, the Chinese central government has focused on ways, other than controlling the Internet technically and legally, that the Internet can be used to enhance government performance and the public good, to reduce tension between the government and the citizens and to foster economic growth, thus increasing legitimacy of the government, an approach that is labeled “proactive strategies” by some scholars (Kalashil & Boas, 2001). E-government is broadly seen as a way to help the Chinese government, which has begun to recognize the emergence of e-government as having potential to reshape the public sector and build relationships between citizens and the government, to cut the costs of governance and make links to the populace closer by reducing bureaucratic procedures (Zhang, 2001).

As a strategy to drive the information economy, the Chinese government demanded in the early 1990s that all government offices move online, starting with an information-carrying Web site as the first step towards more complex and comprehensive interactive services. Three so-called “Golden” projects launched in 1993, even before the direct full connection to the Internet in China on 20 April 1994, became the roots of e-government in China: the Golden Bridge project (jinqiao), the Golden Card project (jinka), and the Golden Customs project (jinguan).

On January 22, 1999, China Telecom and the State Economic and Trade Commission's Economic Information Center along with the Information Offices of 48 central government departments officially started the Government Online Project (GOP, zhengfu shangwang gongcheng). The primary purposes of the project are to establish the basis for the development of e-government in China, provide more effective coordination between and across governments at different levels, increase public access to government information while reducing government expenses by increasing administrative efficiency, promote procurement of goods and services online, and enhance the informatization of the economy and society in China (Government Online Project [GOP], 2000). This initiative signaled the beginning of a rapid growth phase in the “gov.cn” names that are used for government sites in China. On May 18th, 1998, World Telecom Day, only 145 gov.cn sites existed. Within a few months of the GOP initiative, that number climbed to 1,470 sites representing 720 governmental departments. By 2002, “the Year of E-government” (dianzi zhengwu nian), “more than 220 million government departments” offered electronic services, according to Xie Lijuan, member of the Ninth National Committee of the Chinese People's Political Consultative Conference (CPPCC) (“CPPCC Member Proposes,” 2002). In January 2002, the China Internet Network Information Center (CNNIC) semi-annual report counted 5,864 gov.cn domain names, and the number increased to 7,796 by the following year. National ministries as well as provincial and municipal governments across the country have Web sites. Approximately 4.3% of all Web sites under the “.cn” domain have the “gov” suffix (The China Internet Network Information Center [CNNIC], 2003).

Despite this rapid growth of e-government in China, no studies were found that explored the content of both national and provincial Chinese government Web sites systematically. Although Zhang (2001, 2002) provided a macroscopic overview of e-government in China and documented a brief history of the GOP initiative, he quantitatively examined only the Web site of the GOP Navigation and Service Center (http://www.gov.cn), a portal site through which China's e-government Web sites at different levels can be reached. This study fills a gap in the literature by quantitatively analyzing the “front office” of Chinese e-government - that is, the content of the opening pages of government Web sites at both national and local levels. It explores the extent to which the Web sites have met the official expectations four years after the start of the national project. A key secondary purpose of this study is to develop an instrument for identifying and measuring the features of e-government Web sites.

Literature Review and Research Questions

Governments around the world are embracing electronic government. From industrialized countries to developing ones, national and local governments are moving online (Improvement & Development Agency [IdeA], 2002; Taylor Nelson Sofres, 2002). China is one of the countries following this trend even though its goals might not be exactly the same as Western countries who are promoting e-government as a way of realizing e-democracy (Phil Noble & Associates, Inc. [PN&A], 2001; Bonham, Seifert, & Thorson, 2001). The “Tenth Five-Year Plan” makes it clear that China is focusing on IT as it is promoting construction of network infrastructure and preparing for the realization of e-government in 2005 (Planning Commission Government of India). The boom comes in part from a belief that technology can transform government's often-negative image (Bonham et al, 2001; National Audit Office [NAO], 2002). Therefore, e-government is broadly seen as the use of information and communication technology to promote more efficient and effective government.

The World Bank Web site on e-government offers the following definition of the term:

E-Government refers to the use by government agencies of information technologies (such as Wide Area Networks, the Internet, and mobile computing) that have the ability to transform relations with citizens, businesses, and other arms of government. These technologies can serve a variety of different ends: better delivery of government services to citizens, improved interactions with business and industry, citizen empowerment through access to information, or more efficient government management. The resulting benefits can be less corruption, increased transparency, greater convenience, revenue growth, and/or cost reductions. (The World Bank Group, 2003)

A study of e-government projects found three primary categories: e-governance, e-service, and e-knowledge. E-governance refers to linking citizens, stakeholders, and elected representatives to participate in governance of communities. E-service involves securing and providing government services by electronic means. E-knowledge means using communication technologies to gain knowledge (IDeA, 2002).

Tang (2000) suggested a primary function of e-governance in China should be to carry out the functions of the Chinese government on the Internet. This includes government-to-citizen connections, through features such as online forums and government-to-government connections through features such as intranets. The GOP includes an intranet system designed in part to make existing information available to all relevant institutions (Zhang, 2001).

E-service refers to specific government services that can be offered online. This includes business-related services such as license applications and interactive consulting. It also includes citizen-oriented services such as application for residency permits, weather and traffic information, and free e-mail services (Moon, 2002). Many contacts between the local government and citizens happen when specific services are wanted or needed. Examples include passport services, birth certificates, and entry into official databases (Fursich & Robins, 2002).

The concept of e-knowledge refers to governments' use of the Internet to disseminate information resources. The Chinese governments at all levels are by far the largest owner of information resources in the country. Estimates are that government agencies own and operate more than 3,000 databases (Lu, Du, Zhang, Ma, & Le, 2002). Qi (2000) suggested that at least five types of information should be available on Chinese government Web sites: 1) information about governmental duty and roles; 2) documents, files, databases, and other information owned by the government; 3) information about networking clerical work such as electronic file centers; 4) information about governmental routine affairs; 5) information about trading and transaction markets. News reports in China and in the U.S. press indicate that dissemination and control of information are central to the Chinese Internet (McMillan & Hwang, 2002). A report from the U.S. Embassy reports that strides have been made in advancing e-knowledge:

Chinese regulations have gone from being largely unpublished, confidential, in-house edicts…to published texts that are more detailed with each iteration. Some Chinese government agencies have even published proposed regulations online and asked for comments, although this is still very unusual (“Kids, Cadres and ‘Cultists’ All Love It: Growing Influence of the Internet in China,” 2001).

An additional form of e-knowledge is dissemination of image-building information about the government (Zhang, 2001). Examination of e-government projects in other countries has shown that promotional uses of the Internet may sometimes overshadow governance, service, and knowledge functions (Fursich & Robins, 2002).

Scholars have suggested there are three sectors of e-government. Government-to-citizen (G2C) e-government should make it easier for citizens to interact with governmental agencies for everything from obtaining marriage licenses to paying taxes. Government-to-business (G2B) initiatives help facilitate activities such as procurement, licensing, and other activities that help facilitate business-based economic growth. Finally, government-to-government (G2G) functions support intergovernmental operations and involve sharing data and conducting electronic exchanges between governmental actors (Bonham, et al., 2001).

As illustrated in Table 1, each of the sectors of e-government could exist within each of the three categories (e-governance, e-service, and e-knowledge). Table 1 provides examples of G2C, G2B, and G2G sector activities in each category of e-government.

Table 1.  Summary of e-government categories and sectors.
 Categories of E-Government
SectorsE-GovernanceE-ServiceE-Knowledge
Government-to-CitizenEncouraging citizens to become involved in participatory access to the political process through online discussion forums, chat rooms, etc.Provide electronic opportunities to conduct activities such as making payments or obtaining consulting services for individuals.Provide information about governmental activities that are important to citizens.
Government-to-BusinessProviding businesses with opportunities to give input on business regulationsMake payments, file for licenses, download files needed for business, etc.Deliver information relevant to specific business needs.
Government-to-GovernmentFacilitate communication among government agencies to enhance interrelationships among agenciesExchange files between agencies; provide internal government services.Intergovernmental access to government information systems.

Most of the studies of e-government to date have focused on democracies - particularly those in Europe and the Americas (see for example: Bonham et al., 2001; Ho, 2002; IDeA, 2002; Van Wert, 2002). A few have studied Asian situations in more detail (see for example: Holliday, 2002; Thompson, 2002) but tended to focus on democracies. Zhang (2002) noted that it is important to consider e-government in China separately from that of other nations because in centrally managed states, such as China, core assumptions about e-government may be different from those in other parts of the world. Specifically, democratic participation of citizens might be conceived differently in China than in other countries with e-government initiatives. Zhang found that the most common functions at the portal Web site of China's GOP Navigation and Service Center (http://www.gov.cn) were providing services and agenda-setting information. However, it may not be appropriate to generalize the findings to China's e-government, because his study examined only one government Web site, a portal site oriented to establish standards for China's e-government, and to help users understand the initiative, retrieve updated information about China's e-government and navigate e-government Web sites at all levels.

A case study of the development of the Zhongguancun Digital-Park program (Zhongguancun Shuziyuanqu, http://www.zhongguancun.com.cn) in Beijing described the program as one of the earliest and most advanced e-government pilot projects in China. Begun in 2000, the initiative was designed to support a technology business district in northwest Beijing. The program includes five functions coordinated through a central Web site: e-application, e-registration, e-reporting, e-administration, and e-consulting. Despite challenges ranging from security to compatibility, the system has been found to improve efficiency and transparency of government services to this economic development program (Lin, Zhu, & Hachigian, 2001).

Information inequalities exist between China's rich coastal areas in the east and its inland regions, due to historical and economic disparities, reflected by the idea of “one China, four worlds” (Hu, 2001). Internet infrastructure and citizens' economic resources, which affect ability to access the Internet, are stronger in the southern and eastern parts of China than in the northern and western areas (Tang, 2000). This uneven economic and Internet development may lead to differing experiences and expectations in different parts of the country.

Three research questions arise from this brief review of the literature:

RQ1: What type of e-governance is available and how does content vary among Web sites for the national government, and Web sites for inland and coastal provinces within China?

RQ2: What type of e-service is available and how does content vary among Web sites for the national government, and Web sites for inland and coastal provinces within China?

RQ3: What type of e-knowledge is available and how does content vary among Web sites for the national government, and Web sites for inland and coastal provinces within China?

Method

Sampling & Analysis Unit

To address the questions detailed above, a content analysis was conducted. Unlike Zhang's study (2002), the study reported here did not include an examination of the Web site of GOP Navigation and Service Center itself. Instead, the list of Web sites under the Navigation Center (Daohang Zhongxin), including 77 national and 821 provincial Web sites, was used as the sampling frame, from which a stratified random sample of 25 national and 160 provincial e-government sites were produced.2 Among the provincial sample, 80 sites were randomly drawn to represent respectively the two regions of the coastal and inland provinces or municipalities.3 In most cases, the Web sites selected for the study were directly linked from the GOP site. In the cases where the link was not functioning, Yahoo China! (Yahu Zhongguo, http://cn.yahoo.com) was used to search for the URL address by typing the Chinese name of the site on the GOP. If the Web site still could not be located, it was dropped from the sample. In total, 177 Web sites were actually examined. Considering the size of the sample frame (898), the sample of 19.71 percent of Web sites is large enough to allow for generalization about e-government Web sites in China.

Analyzing Web sites introduces many challenges to the content analysis process (McMillan, 2000). However those challenges can be overcome with careful planning for issues such as sampling, intercoder reliability, and unit of analysis. The primary unit of analysis for this study was the “home page” or opening screen of the Web site. By using the home page as the unit of analysis, Web sites of varying sizes can be more effectively compared. Ha and James (1998) reported that Web Techniques estimates that Web sites range from one page to 50,000 pages. They argued that coding an entire site could be extremely time-consuming and introduce biases based on the size of Web sites. Furthermore the home page is central to Web-based communication because it provides a kind of “front door” to all the messages contained in the site. Most visitors to a Web site decide whether they will continue to browse a site based on their impressions of the home page (Ha & James 1998).

Coding Scheme

A coding form was developed that quantified types of e-governance, e-service, and e-knowledge. The typology developed in the literature and summarized in Table 1 provided the initial framework for developing this coding form. Earlier studies that examined e-government sites in other countries were also used as a baseline for developing the coding form (Fursich & Robins, 2002; Ho, 2002; Moon, 2002; PN&A, 2001). Additionally, a pre-test of ten leading Chinese sites, such as Capital Online (shoudu zaixian), China Shanghai (zhongguo Shanghai), and Shenzhen Government Online (Shenzhen zhengfu zaixian), identified China-specific coding categories and specific types of e-government functions as described below. Pre-testing of these full-featured sites made it possible to obtain a list of functions that was as complete and sensible as possible. Two trained coders reviewed the coding form prior to the start of coding and agreement was reached on the overall structure and content of the coding form, thus assuring face validity.4

In addition to content that was provided in detail on the home page, hyperlinks, including text links, image links, and navigation bars, which appeared on the home page, were also coded. The primary coding categories were for the three primary types of e-government: e-governance, e-service, and e-knowledge. Some hyperlinks served as a category or a navigation gateway from the home page to a subsequent page on which two or more overlapping functions of e-government (e.g., both e-governance and e-knowledge) could be found. These “gateway” links were coded as assorted e-government functions. Hyperlinks that were for mere Web page functions (e.g. language options, bookmarks, and Web ads) were not coded as having e-government functions because the issue of Web page design is not the major concern in this study. Because of the exploratory nature of this study, the coding scheme included an “other” category that allowed for coding of unexpected functions (see Table 2).

Table 2.  Coding scheme.
  National*  Coastal*  Inland* 
 NMean RankX2NMean RankX2NMean RankX2
  1. *p < 0.001

E-governance243.9271.530804.06200.082734.2135.367
E-knowledge 4.65  3.86  3.95 
E-service 2.94  3.72  2.76 
Assorted e-government functions 1.21  1.32  1.50 
Other Web functions 2.29  2.04  2.59 

Types of E-Governance Functions5

As detailed in Table 1, three types of e-governance functions were identified in the literature: government-to-citizen, government-to-business, and government-to-government. However, in developing the coding scheme, it became evident that the manifest content of publicly available Web sites allowed for examination of two basic e-governance functions: engaging or representing citizens (government-to-citizen functions) and facilitating communication among government agencies (government-to-government functions). This is not to suggest that government-to-business functions are not available in China. It merely reflects the fact that those functions do not seem to appear at the Web sites that are made available to all citizens through the Government Online Project from which this sample was drawn.

Engaging or Representing Citizens

Civic engagement in governance has been seen as an important way to promote citizen empowerment, provide “voice” for those outside government, and connect citizens to government by listening to citizens and supporting accountability, thus strengthening good governance and increasing people's trust in their governments (Pacific Council on International Policy, 2002). The pre-test observation found various Web tools, such as online discussion forums, online surveys or polls, chat rooms, online forms for public concerns or complaints, that were used to engage citizen participation. Additionally, links to personal Web sites of officials and email or online mailbox to officials or policymaking bodies were identified to help tighten the connections between government and citizens.

Facilitating Communication Among Government Agencies

Internal processes of e-governance between governmental units occur primarily via the government secure intranet. This internal system enables the integration between systems, networks and data within public administration, and the “back-office” capability to deliver coordinated government services. Examining this type of government-to-government use of ICT is beyond the scope of this study. However, some e-governance functions were still found in this study in the “front office” of e-government, such as facilitating the horizontal and vertical interconnection of government bodies and making the interaction between government and its employees and inter-agency relationships more convenient and inexpensive by providing internal email boxes, and access to multiple related government sites, such as Web sites of the branches of government, as well as to provinces or municipalities. Two major types of hyperlinks on the home page under the subcategory of portal function were identified. Vertical linkage referred to the existence of links to lower/higher levels of government, whereas horizontal linkage represented the existence of links to same-level agencies or entities. Those links leading to a Web page other than e-government Web sites, e.g., the Web sites of search engines like Baidu and Yahoo China! were grouped into “other linkage.”

Types of E-Service Functions

Online delivery of services benefits both government and its constituents by lowering costs and making services more accessible. Electronic services have been introduced and are being used and developed in China's e-government Web sites. The service delivery has evolved through a variety of functions. Rather than trying to somewhat artificially divide these services into the category of individuals being served (citizens, business, and government), two major types of services that apply to all categories of individuals were identified in this study:

Non-Interactive Services

This refers to information or advice that is published online and that is designed to help citizens or businesses efficiently carry out their daily activities. These e-service hyperlinks allow only a passive relationship between the Web site and the user. The user only has to click on a link to receive the information or advice. The content was coded as “timely information for daily life” if it was the kind of service that would be accessed frequently. It was coded as “consultative information” if it was information that would only be consulted on rare occasions. For instance, the links leading to a page covering weather, air quality or traffic information, up-to-date food pricing, movie timetables, local maps, job information, and so on, were grouped into timely information for daily life. By contrast, consultative information included help-yourself data on various guidelines and processes, such as specialist advice and step-by-step instructions on a wide array of procedures (e.g. passport application, marriage and motor vehicle registration, business operational guide, license application, etc.).

Interactive Services

This coding category refers to content that facilitates interaction between the Web site and the user. These services require more than a simple “click” to obtain information. Typically, the user has more control over content and the Web site is more responsive to user input. Six major interactive functions were identified in this study: 1) E-filing. For instance, individuals are able to check their tax status online on some Web sites. Other examples include license application or renewal, registration of corporations and partnerships, online release of information about commodities, civil service status tracking, application for residency permit or renewal, record updates, and so on. 2) E-payment. In filing their taxes online, for example, taxpayers key-in and validate data themselves. Other examples include online payment systems for fines, utility bills, or permit fees. 3) Downloading files, such as legal forms for consumers, motor vehicle certificates, export documents, and a variety of forms for application procedures. 4) Searchable consultative information for individuals or businesses. In some Web sites, for instance, users can search information or documents by common subjects. 5) Subscription/registration service. Some Web sites require membership registration to receive a certain service. For instance, companies that wish to do business with the public sector do not need to search through newspapers or the Web for information about bidding opportunities. Instead, they need only to register a single time in the areas in which they do business. Whenever a request is issued from a public agency, the system will automatically send e-mail to all the private companies registered in that selected area, minimizing response time and providing an equal opportunity for all firms. 6) E-mail service. Some government Web sites provide e-mail service for their constituents.

Types of E-Knowledge Functions

The Web, with its virtually unlimited capacity for information and ease of frequent updating, is uniquely capable of helping to disseminate e-knowledge. Through its ability to spread accurate, updated and comprehensive information, the provision of e-knowledge through the Web can be a powerful tool for social control. For instance, nothing is more powerful in combating corruption than conducting transactions openly and with public knowledge of the rules and criteria to be applied (Bhatnagar, 2003). Again, the pre-test suggested that the three categories of e-knowledge identified in Table 1 might be better examined using functional divisions rather than division by who is receiving the knowledge. Thus, as illustrated in Table 2, three primary categories of e-knowledge were identified.

Government Information Delivery

If a Web page covered government-centered information, it was included in this subcategory, which was divided into three groups: 1) administrative information about the Web host and its subordinate agencies that focuses on informing the public about the objectives, roles and duties of the concerned governmental unit, 2) information about governmental routine affairs and political processes such as routine meetings, the leaders' affairs and activities, and governmental projects, and 3) information about rules, regulations, and policies, including a variety of government bulletins, news briefs, announcements, notices and full-text documents about the rules, regulations and policies in a specific field.

Some information was provided by the governmental unit but was not government-oriented. This was grouped into the category of general information delivery, including multiple sub-categories as detailed in Table 2. The final category, searchable government information database(s), was differentiated from “government information delivery” by its interactive nature with users, although the content of both could be similar.

Coding & Intercoder Reliability

Unlike numerous previous studies in analyzing Web content (see for example: Aikat, 2000; Ha & James, 1998; Ho, 1997; Schultz, 1999), in this study, the categories were not coded by simply noting whether or not a coding item existed. Such coding schemes are limited because the data generated are usually statistically analyzed by using Chi-square tests that are limited in their ability to analyze differences among groups. Simply noting the presence or absence of a function did not enable the researcher to differentiate between the home pages of two Web sites if one had 30 links to a specific type of function while the other only had three links. To understand thoroughly how the governments at different levels in China used available resources on the Internet, the researcher decided to adopt the method of counting the frequency for each item developed in the coding scheme. Another reason for coding this way was that e-knowledge, one of the three major e-government categories in this study, mainly dealt with information delivery. Instead of simply noting whether a certain type of information was delivered, the researcher was more interested in how much the type of information was delivered and how important the information was.

Because home pages were examined in detail (often requiring selecting hyperlinks to determine the type of content that they linked to), it was important to ensure inter-coder reliability. This in-depth coding procedure was particularly sensitive to changes in content that might happen frequently at the Web sites that were information-intensive. Therefore, to address the instability of Web sites, both coders coded sites on the same day . Also, a pre-test found that a few of the examined Web sites were in the process of being redesigned, which would have seriously undermined the intercoder reliability if they had not been coded on the same day.

Each link was coded as a single frequency for the appropriate coding item under a certain category or subcategory. However, if more than one link connected to the same subsequent page, the second or later link was not coded. Two coders examined 30 Web sites (5.9 percent of n=177) to test the validity of the coding instrument and obtain a measure of intercoder reliability. The coders first coded ten randomly selected Web sites, which produced a Holsti coefficient of 0.87. After some variables were collapsed and definitions for controversial variables were clarified, coders analyzed another twenty randomly selected Web sites and the Holsti coefficient was improved to 0.91. All Web sites were coded between January 20 and March 2, 2003.

Data Analysis and Results

Home page size may vary and content may be quite different based on the nature of institutions and the targets of their Web sites. The distribution tests showed that the frequency distributions of primary variables in this study were all quite un-normal (p < 0.001). Therefore, it was appropriate to examine research questions using non-parametric tests rather than parametric tests, such as t-test and ANOVA, which assume normal distribution.

Research questions not only examined specific types of e-governance, e-service, and e-knowledge content available at the Web sites but also compared content between the national Web sites and the sites for coastal and inland provinces. To examine the research questions, the raw numbers of links from the home pages were coded for the key content types (for example, within e-governance a key content type was facilitation of citizen involvement; the total number of links to online discussion forums, chat rooms, etc. were coded within this content type). To explore differences among the key functions (the first half of each research question), the number of links for each key content type was compared. To explore the second half of each research question, which examined differences in implementation at the different levels of government, the proportion of each key content type to the total number of links was examined.

As far as the issue of data analysis is concerned, this study differs from previous studies. In many of previous studies (see for example Aikat, 2000; Dillon & Gushrowski, 2000; Ha & James, 1998; Ho, 1997; Parker, 2002; Schultz, 1999), each examined dimension or function was checked at each sample Web site and the proportion of the Web sites with the specific dimension or function was compared with the other dimensions or functions. This method is limited in its ability to see in detail how different functions were distributed and what functions were projected at each Web site, and what differences existed between functions within the Web sites at the same level and across the Web sites at different levels. To overcome such limitations, this study performed two major types of statistical tests, Friedman tests and Mann-Whitney tests, based on the frequencies of each function and category. In the former tests, the frequencies of functions and categories were ordered at each Web site, and then the mean rank of the frequency of each function and category within the sites at the same level was calculated and compared. In this way, a horizontal comparison of functions at the same e-government level was performed. In the Mann-Whitney tests, the proportion of the frequency of each function and category to the total links on the home page was ranked across the sites at two different levels. Then the mean rank of the proportion at each level was calculated and compared. In this way, we could see how each function differed across the different government levels.

Before examining the specific research questions, it should be noted that the national Web sites had the largest mean total number of links (78.13), followed by the coastal Web sites (70.70) and the inland Web sites (65.91). The Mann-Whitney tests show that the mean rank of the total numbers of links of the national Web sites significantly differed from that of the inland Web sites (z= -1.781, p= 0.075), while the coastal Web sites showed no significant differences from either the national (z= -1.20, p= 0.23) or inland sites (z= -1.414, p= 0.157). The highest variance in the total numbers of links was found at the inland sites (4085.048) with some having very few links and some being fairly sophisticated. The national sites were most consistent with a variance of 1795.505 and the coastal sites had a variance of 2997.149. Also, the inland Web sites were more likely to have dead links on their home pages than the sites at the other two levels. The mean rank of the proportions of dead links to the total links on the home pages of the inland Web sites were significantly higher than those on the home pages of the national (z= - 4.537, p < 0.001) and coastal sites (z= - 6.875, p < 0.001), while no significant difference in the number of dead links was found between the national and coastal Web sites (z= - 0.231, p= 0.817).

Table 3 provides an overview of the mean rank for e-government categories for the three government levels. E-governance was the dominant category for the coastal and inland Web sites while e-knowledge dominated the national Web sites. E-knowledge was in a strong second place for both the coastal and inland sites while e-governance took the second place for the national sites. For all three government levels, e-service was the third-most important category.

RQ1. What type of e-governance is available and how does content vary among Web sites for the national government, and Web sites for inland and coastal provinces within China?

As can be seen in Table 4, at each E-government level, the dominant content type was the portal function, providing links to other Web sites, including government sites, media sites, etc. The other major content types of e-governance were online forums and e-mail links provided for citizens to express their concern or complaints about current situations or policies, and online surveys or polls concerning a certain issue. There were significant differences among the e-governance functions at each level of government (p < 0.001).

Table 4.  Friedman tests for differences in e-governance functions at each e-government level.
  1. p< 0.001, df = 10

  2. The first three highest ranks in e-governance are highlighted, indicating the frequency numbers of the links to the referred functions tended to be larger than the others at each e-government level. Similar in the following tables of Friedman tests.

  National*  Coastal*  Inland* 
NMean Rank X2NMean RankX2NMean RankX2
 24 134.00380 490.78073 354.147
Engaging/Representing Citizens         
Online discussion forum(s) 5.71  5.20  5.58 
Online survey(s)/poll(s) 5.96**  6.04  5.63 
Chat room(s) 4.90  4.76  5.42 
Online forms for public concern/complaints 6.63**  7.48**  5.89** 
Link(s) to personal Web site(s) of official(s) 4.90  4.77  5.08 
E-mail/online mailbox to official(s) or policymaker bodies 5.54  6.81**  6.29** 
Assorted or others 5.10  5.15  5.70 
Facilitating Communication         
Portal function 10.90**  10.89**  10.20** 
Internal email-box or other intrafunction(s) 5.96**  5.34  5.77 
Assorted or others 4.90  4.85  5.30 
Assorted Functions or Others 5.52  4.71  5.14 

Table 5 highlights e-governance functions for which differences were found between government levels. The coastal sites more often provided access to engage citizen involvement than did the inland sites, which significantly more often provided links to other Web sites. Detailed analyses within each subcategory found that the difference between the national and coastal sites in the subcategory of engaging and representing citizens was that the coastal sites were more likely to provide e-mail links or online mailbox that enable citizens to government communication than the inland sites (z= -2.596, p= 0.09). More differences between the coastal and inland sites were found, in that the coastal sites significantly more likely provided online forms (z= -4.376, p< 0.001), online surveys or polls (z= -2.498, p= 0.012) and e-mail links or online mailbox to officials or policymaker bodies (z= -2.291, p= 0.022) than the inland sites, which didn't significantly differ from the national sites in each content type of this subcategory.

In examination of the functions of facilitating communication among government agencies, differing patterns were found in how hyperlinks were used by different levels of government. While no differences existed between the national and coastal Web sites, the inland sites were more likely to take advantage of links as a portal function, with the proportion of these kinds of links significantly larger than those of the national sites (z= -2.514, p= 0.012) and the coastal sites (z= -2.561, p= 0.010), thus causing the significant overall differences of the inland sites than the sites at the other two government levels in the implementation of facilitating government communication. Detailed analyses found that the national sites were more likely to provide links to subordinate agencies than were either the coastal (z= -0.6014, p < 0.001) or inland sites (z = .0.000, p < 0.001), while there was no significant difference between the coastal and inland sites. By contrast, both the coastal (z= -2.441, p= 0.015) and inland sites (z= -2.179, p= 0.029) were more likely than the national sites to provide links across the national and local levels. Again, no difference was found in how the coastal and inland sites used such links.

Table 6 examines differences in the levels of government for e-governance as well as e-service and e-knowledge and will be referenced for each of the research questions. With regard to e-governance, significant differences were found between the different levels of government. The inland Web sites devoted significantly more of their total links to the e-governance functions than did the national (z= -2.313, p= 0.021) and coastal Web sites (z= -2.238, p= 0.025), while the national and coastal Web sites did not significantly differ from each other (z= -1512, p= 0.130).

RQ2. What type of e-service is available and how does content vary among Web sites for the national government, and Web sites for inland and coastal provinces within China?

Two key functions, non-interactive and interactive, were found with regard to e-service. Table 7 shows that the content type of consultative information dominated at each e-government level, where timely information for daily life was also found important. At the coastal and inland levels, the function of downloading various files was another primary function, while links to searchable consultative information was popular at the national Web sites. The other relatively frequently found functions included e-filing and subscription or registration services. The two least-often occurring types were e-payment and e-mail services. The other interactive functions, such as online shopping and e-procurement, were seldom found. There were significant differences among the e-service functions at each level of government (p < 0.001).

Table 7.  Friedman tests for differences in e-service functions at each e-government level
  1. p< 0.001, df = 10

  National*  Coastal*  Inland* 
NMean Rank X2NMean RankX2NMean RankX2
 24 58.95080 337.86273 140.204
Non-interactive Services         
Timely information for daily life 6.71  6.54  6.34 
Consultative information 8.71  10.13  7.98 
Assorted or others 4.88  3.95  5.09 
Interactive Services         
E-filing 5.92  5.96  5.62 
E-payment 4.88  4.13  5.09 
Downloading files 6.27  6.38  6.29 
Searchable consultative information 6.38  5.79  5.82 
Subscription/registration service 5.79  5.16  5.32 
E-mail service 5.10  4.22  5.38 
Assorted or others 5.79  6.45  6.26 
Assorted E-Service Functions or Others 5.58  7.29  6.82 

As illustrated in Table 6, significant differences were found in the overall implementation of e-service among the government levels. The coastal Web sites devoted significantly more of their total links to e-service than did either the national (z= -4.106, p < 0.001) or inland Web sites (z= -6.679, p< 0.001). However, no significant difference was found between national and inland Web sites (z= -1.197, p= 0.231).

Table 6.  Mann-Whitney tests for differences in link proportions of three major functions over total links between e-government levels
FunctionLevelMean RankSum of RanksZ*PValue (two-tailed)
  1. Grouping variable: E-government level

 national44.33764.00-1.5120.130
 coastal54.951064.00  
E-governancenational37.48599.50-2.3130.021
 inland52.79899.50  
 coastal69.342533.00-2.2380.025
 inland85.395234.00  
 national30.33728.00-4.1060.000
 coastal59.154732.00  
E-servicenational54.921318.00-1.1970.231
 inland47.053435.00  
 coastal99.817985.00-6.6790.000
 inland52.003796.00  
 national74.101778.50-4.0000.000
 coastal46.023681.50  
E-knowledgenational64.211541.00-3.0520.002
 inland44.003212.00  
 coastal73.695895.00-0.9680.333
 inland80.635886.00  

Table 8 examines differences in e-service more specifically by comparing how the three government levels implemented both non-interactive and interactive e-service. The general pattern found in Table 6 is supported by this data. No significant difference was found between the national and inland sites. The costal sites were significantly more likely than the national sites to provide consultative information for individuals and businesses (z= -3.139, p= 0.002). The differences between the coastal and inland sites were much more comprehensive and significant, largely existing in each function, except for e-mail service.

RQ3. What type of e-knowledge is available and how does content vary among Web sites for the national government, and Web sites for inland and coastal provinces within China?

Table 9 summarizes how the content types of e-knowledge were delivered at each government level. It shows that news concerning the industry or field under the administrative domain of the host government agency of the Web site, followed by information about government rules or regulations or policies and information about government routine affairs or political processes, dominated on the home pages of the national Web sites. At the coastal and inland levels, the three dominant content types were all related to government information delivery. Overall, information about government rules or regulations or policies and information about government routine affairs or political processes seemed to be important at each government level. Significant differences existed among the e-knowledge functions at each level of government (p < 0.001).

As for the differences between the e-government levels, Table 10 shows that the coastal Web sites were less likely to deliver government information than either the national or inland Web sites, which didn't significantly differ from each other. But the coastal Web sites were more likely to provide searchable government databases than the inland Web sites. Compared to these two levels, the national Web sites tended to devote the larger proportion of the total links, whose mean rank was significantly higher than the other two, to general information delivery. Based on the detailed analyses of each general information type, this was the case for news of a specific industry, international news, and provision of academic news or information from “experts.” The national sites also provided more general knowledge education than did the inland sites. However, there were two instances where the coastal and inland sites outperformed the national sites: introductory material about the area and local news.

Table 10.  Mann-Whitney tests for differences in e-knowledge between e-government levels
FunctionLevelMean RankZaP Value (two-tailed)
  1. Grouping variable: E-government level.

 national65.60-2.4270.015
 coastal48.57  
Government information deliverynational50.02-0.3090.757
 inland47.99  
 coastal66.79-2.8650.004
 inland87.28  
 national72.60-3.7230.000
 coastal46.47  
General information deliverynational63.85-3.1180.002
 inland43.38  
 coastal74.66-0.5430.587
 inland78.54  
 national44.88-1.4610.144
 coastal54.79  
Searchable government databasesnational55.48-1.6300.103
 inland46.17  
 coastal87.81-3.6130.000
 inland63.93  
 national58.90-1.2180.223
 coastal50.58  
Assorted or others in E-knowledgenational61.23-2.6070.009
 inland44.98  
 coastal83.31-1.9450.052
 inland70.09  

The detailed analyses also found that as a general rule, the coastal and inland sites were most alike in their delivery of general information. However, the inland sites outperformed the coastal sites for introductory material about the area under the administrative domain of the Web host (z= -2.747, p= 0.006). The coastal sites outperformed the inland sites for news concerning the industry or field under the administrative domain of the Web host (z= -2.819, p= 0.005).

In other significant differences, the inland sites were more likely than the coastal sites to provide administrative information about the Web host and its subordinate agencies (z= -2.487, p= 0.013). By contrast, the coastal sites were more likely than the inland sites to post government rules, regulations, and policies (z= -3.322, p= 0.001).

As illustrated in Table 6, significant differences were found in the implementation of e-knowledge between the government levels. The national Web sites devoted significantly more links to e-knowledge than did either the coastal (z= -4.000, p < 0.001) or inland sites (z= -3.052, p= 0.002). No significant difference was found between the coastal and inland Web sites (z= -0.968, p= 0.333).

Discussion and Suggestions for Future Research

In summary, this study found that the national sites were the “largest” with more links on their home pages and with least variance in home page size than either the inland or coastal Web sites. This suggests that these sites run by national government ministries seem to be the most information rich. The Web sites for coastal provinces had the second highest number of links and both the national and coastal Web sites seem to be better maintained than the inland sites as demonstrated by the higher number of “dead links” at the inland sites. However, an important distinction needs to be made between the national and coastal sites. While the coastal Web sites had fewer links on home page than the national sites, the sites tended to be more organized in the way that they presented information and services on home page. The coastal Web sites were more likely to use “functional” links that lead to more depth of information while the national Web sites have more “text” links such as news titles. In actuality, the depth of information may be greater at the coastal Web sites.

The national sites were dominated by e-knowledge with particularly strong focus on the posting of rules, regulations, and policies. E-governance was the strongest category of e-government for both the coastal and inland sites. Portal functions were the most popular form of e-governance at all government levels, but features that engage citizens were also found. The coastal Web sites consistently outranked the inland sites in terms of the amount of citizen involvement features at Web sites. E-services were also a key component of many Web sites. The costal sites tended to provide more services than did the sites at the other two levels, especially the inland sites.

All of these trends taken together tend to support the notion that the governments of more economically developed coastal areas are often leading Web site development in China (Tang, 2000). While the national sites might provide more information, it is the coastal Web sites that allow individuals more opportunity to participate in the governance process.

Although the inland sites provided the fewest number of links, they were best at providing certain types of information. Most notably, the inland sites seem to be doing a good job of building awareness of their provinces or municipalities by providing content such as an introduction to the province or municipality and local news.

China's leadership has held out the vision that information technology will allow the country to decentralize decision-making while allowing the central government to monitor and control the economy. This vision seems to be reflected in Web site content. The central government develops and disseminates information while the provincial governments engage citizens in decision-making processes. Posting government regulations and policies online at various levels of e-government is no longer “unusual.” Citizens are able to obtain various documents of regulations and policies, such as regulations on higher education admission and health care, policies of personnel promotion and “human talent,” and local privilege policies for foreign investors, from the Web sites of the concerned government agencies, which is very helpful for them to be engaged in decision-making.

Examination of the Web sites suggests the central government initiatives aimed at driving the information economy through the development of Web sites at every stage of government may be working. While the sites are not identical or even equal, there is some evidence that a kind of unified model is evolving that incorporates e-governance, e-service, and e-knowledge into Web sites at both the national and provincial level.

A secondary goal of this study was to develop a coding instrument for analysis of Web sites. The literature suggested a typology that seemed promising as the framework for a coding document (see Table 1). However, during pre-testing it became clear that it was difficult to differentiate between target audiences at many e-government sites. Thus, the three basic categories of e-government (governance, service, and knowledge) were retained. But rather than trying to force the coding scheme to reflect the sectors (government-to-citizen, government-to-business, and government-to-government), coding categories were allowed to emerge from the data as illustrated in Table 2. This coding scheme, along with the method of coding number of features rather than simple presence or absence of features, should prove useful to scholars examining e-government in many different counties and contexts.

Despite the care taken in designing, implementing, and analyzing this study, a few limitations must be noted. Future studies might consider sampling not only national and provincial sites but also sites of municipalities across the country. This could provide more insight into the depth of success that the central government is (or is not) having in implementation of its policy of universal e-government availability.

Future studies might also consider examining e-government Web sites from the citizens' perspective. It does the government little good to build Web sites if they are not used. Future studies could use surveys, interviews, focus groups and other techniques to learn about how citizens use e-government Web sites. What factors encourage them to go to the sites? What relative value do they place on e-governance, e-service, and e-knowledge?

Another key question that should be addressed in future research is how citizens access e-government Web sites. Do they have access at home, through their work, at a public library, or cyber-café? What are the implications of access on use of e-government Web sites? If people have to pay a per-use fee for Internet access, will that limit their use of e-government Web sites?

Future studies should examine the effects of e-government Web sites. Are citizens more informed? Are citizens able to receive feedback in a timely manner? Do they feel more enabled to participate in governance? Are they receiving more services? Are they more knowledgeable?

This study suggests that the Web has at least begun to help the Chinese government address its primary needs in the age of information expansion. Government Web sites do integrate information networks. In particular, they provide e-governance, e-service, and e-knowledge. Both time and additional research are needed to determine if these Web sites will also enhance government performance, reduce tension between the government and citizens, and foster economic growth.

Acknowledgments

The author very much thanks Dr. Sally J. McMillan for her help with conceptualizing and proofreading the paper. (An earlier draft of the paper was presented at the conference “China and the Internet: Technology, Economy, and Society in Transition” held May 30-31, 2003, at the University of Southern California.) The author would also like to thank the anonymous reviewers and the editor for their constructive suggestions.

Footnotes

  • 1

    The five autonomous regions, such as Inner Mongolia and Tibet, and the four municipalities under the jurisdiction of the Central Government, such as Beijing and Shanghai, fall into the provincial category based on the articles of the Chinese Constitution concerning the present-day administrative divisions.

  • 2

    In order to overcome the limitation of the uneven sizes of the sample at the two levels, the researcher adjusted the proportion of the sample for each level. More Web sites had been added into the list of the Navigation Center by the time when this paper was accepted by the journal.

  • 3

    The appendices provide the detail of the sample list.

  • 4

    The author would like to thank Ms. Linda M. Lee for her help with coding. All Websites were in Chinese requiring coders who were fluent in the language. She and the author were graduate students studying in the United States when they were involved in coding.

  • 5

    The detailed description of types of e-government functions helps readers to understand how the coding scheme has been developed and what content the categories in the coding scheme cover.

  • 6

    For historical and economic reasons, the two provinces, Yunnan and Hainan, were categorized in this group, even though they have their own coastal lines.

Appendices

Figure Appendix 1..

Examined Web sites at the national level.

Table Appendix 2..  Examined Web sites at the provincial level
 Province/MunicipalTotal SitesSampleWrong URLUnable to LocateURL by YahooExamined Site(s)
CoastalLiaoning4040//4
 Hebei2410//1
 Tianjing2010//1
 Beijing139280//28
 Shandong30///0
 Jiangsu5690//9
 Shanghai83233/323
 Zhejiang3750//5
 Fujian710//1
 Guangdong5481/18
TOTAL 4638040480
Inland6Heilongjiang2261/16
 Jiling1930102
 Inner Mogolia710//1
 Shaanxi1330//3
 Henan3270106
 Anhui42100//10
 Jiangxi1220//2
 Hunan521201011
 Hubei2980107
 Guangxi1130//3
 Guizhou1330102
 Sichuan2640//4
 Chongqing1960//6
 Shanxi30///0
 Ningxia20///0
 Gansu310//1
 Qinghai410//1
 Xinjiang1040//4
 Tibet00///0
 Yunnan2760204
 Hainan60///0
TOTAL 3528017173