This article examines the prominence of Web sites of major and minor parties in the United States and the United Kingdom, comparing features such as search capabilities, membership forms, information on party organization and issues, characteristics of graphics, and currency of updates as well as their relative quality and sophistication. We also look at the prominence of major and minor parties in newspapers and magazines and in various search engines and sites for political junkies. We find that minor parties have a greater presence on the Web in the United Kingdom than in the United States, but even so, the sites of major parties in both countries are more prominent and sophisticated than those of minor parties, and major parties generally receive more media coverage than minor parties, both on-line and off-line. The data suggest that the established interests dominating most of the communications, transactions, elections, and political processes of advanced industrialized countries are extending their influence to these processes in cyberspace.
Pioneers of cyber-communication display an unwavering optimism that has carried this medium during the early days .... The establishment, however, quickly caught on .... Afraid to be a breakthrough behind, the big players in the Fortune 500--such as GE-NBC, Disney, ABC, Westinghouse-CBS, and the national news organs--sent their claims takers into Cyberspace. Their money has had quite an effect on the evolution of the medium. Gary Selnow (1998:xxii)
The Normalization of Cyberspace
Before the World Wide Web (WWW) came to prominence, the most active, influential, and numerous inhabitants of cyberspace were members of research and educational communities who shared values conducive to the advancement of science. For many of them, the Internet was like a Lockean state of nature: All individuals were free and equal. Each was a producer and a consumer, though some were more likely to consume than to produce. The dominant ethos was altruism, and there was a spirit of mutual aid. Even though some were more skilled, there was no rigid division of labor. The sparsely settled land stretched from virtual horizon to virtual horizon. There were hardly any real politics--certainly no concerted efforts to manipulate the opinions and actions of a mass public. Communities formed around common interests, and differences of opinion spurred thought and discussion. The inhabitants were more innocent than their alter egos in the real world. They led simple lives on-line, and life was fun, if not particularly profitable.
Although the governments of the United States and other technologically advanced nations had underwritten most of the development of the Internet, the behavioral norms drew more from the libertarian tradition of minimal governance than from the bureaucratic rules of the modern regulatory state. By and large, inhabitants' behavior conformed to a loose set of informal rules, or "netiquette," which allowed separate independent self-governing networks to develop. That government-owned "backbone" lines should not be used for private business for profit was one of the few rules that extended across cyberspace. As recently as 1994, the law firm of Canter and Siegel caused a scandal by circulating an unsolicited advertisement to Usenet newsgroups.(n1) The ubiquitous banners and other types of advertisements only became prevalent in 1996.
Today, a mass audience of consumers, entertainment and information seekers, and curious tourists far outnumber those with significant technical expertise, and even among the technologically competent, the Internet is not the same. Amateurs and hobbyists are being pushed aside by professionals. When these professionals began to see time spent on the Internet as work rather than recreation, things changed forever. Now people are paid to construct Web pages, to provide content and maintenance for Web sites, or to log on to the Web regularly as part of their job. Political, economic, social, and recreational life on the Net for the mass public is increasingly designed and guided by Web professionals. Some of the early stalwart settlers on-line have dropped off in disgust. Those who remain do not need their own Frederick Jackson Turner to tell them that the frontier is fading fast.
Nowadays, the corporate interests that dominate most of the communications, transactions, elections, and political processes of capitalistic industrialized countries have extended their influence to the communications, transactions, elections, and political processes of cyberspace. If current trends continue, it is unlikely that new values and behaviors originating in cyberspace will lead to substantial restructuring of politics, commerce, or any other set of socioeconomic arrangements in the real world (Resnick 1998). Minor parties in the United States once had the jump on major parties on-line, but by 1996, the major parties had recovered and had largely taken over. Just as major American corporations and established media have come to dominate mom-and-pop businesses and media, so they have reshaped the Web (Margolis et al. 1997; Selnow 1998).
We originally derived this interpretation of the evolution of political party activity on the Web largely from observations made and data collected before the American presidential election of 1996. We call it the "normalization hypothesis." It runs counter to the popular notion that the Internet provides the means for minor political parties and marginal political interest groups to compete with major parties and interests on a more or less equal basis in cyberspace, but it is hardly conclusive. The Internet is a complex and changing medium, about which it is notoriously difficult to generalize. It behooves any researcher, therefore, to check the extent to which his or her initial observations and interpretations hold both over time and outside the United States.
Indeed, in late 1996, Rachel Gibson and Stephen Ward collected systematic data concerning the features of British political parties on the Web. Their data, which combined survey questionnaires sent to party officials with observations of Web sites, suggested "that, contrary to Margolis's assessment of U. S. parties, ... there is significant equalization of the communications playing field. The internet is offering the minor parties a more equal footing to compete with their major counterparts" (Gibson and Ward 1997:14). Minor parties in Britain seemed more competitive with major parties on the Web than did their American counterparts, at least as measured by the presence of graphics, split screens, flashing icons, links to other sites, and the use of professional Web site designers.
Stimulated by Gibson and Ward's findings, this article updates our work on U.S. political parties and extends our study to make some relevant comparisons with parties in the United Kingdom. We examine the prominence of Web sites of minor parties in the United States and the United Kingdom relative to those of major parties. We also compare features of these Web sites, such as search capabilities, membership forms, information on party organization and issues, characteristics of graphics, and currency of updates, to estimate their relative quality and sophistication. In addition to these comparisons, we look at the relative prominence of major and minor parties in newspapers and magazines and in various search engines and sites for political junkies on the Net. We supplement our Web site data with information gained from interviews with Webmasters or others knowledgeable about the party Web sites.
Hypotheses and Methods
The normalization hypothesis asserts that as the Internet develops, patterns of socioeconomic and political relationships on-line come to resemble those of the real world. Applied to political parties, this hypothesis implies that just as the major parties dominate the sphere of everyday domestic politics, so they come to dominate cyberspace. This proposition runs counter to a popular alternative that might be called the "inherent-equalization hypothesis," which derives its plausibility from the extraordinarily powerful capabilities of the Internet as a medium for communication. Unlike the standard broadcasting and print media, where information flows from one to many, the Internet permits many-to-many reciprocal flows. Interconnected and interactive, the Internet is a network that purportedly has no privileged center. Every "netizen" has the means to create and distribute information, not just to consume it (Hauben and Hauben 1997).
Applying this second hypothesis to democratic politics and political parties on the Internet yields two important empirical claims that run counter to the normalization hypothesis: (1) Cyberspace should show a tendency toward the reduction of the domination of major political parties. We should observe the leveling effects of the Internet's connectivity in a relatively higher prominence and visibility of the minor parties on-line compared to off-line. (2) On-line, the parties themselves should reflect the leveling effects of the interactive communication in cyberspace. The old media model of one-to-many flows of information simply reinforces the iron law of oligarchy in political parties. If the Internet as a medium exerts a powerful counterpressure toward equality and political participation, we should be able to observe it in the way political parties operate on-line in contrast to off-line. This is a claim that Gibson and Ward tried to substantiate.
The normalization hypothesis suggests a different picture, however, evidenced in three main predictions: (1) Major parties will have a greater presence on the Web and will deploy more sophisticated Web techniques than will minor parties. (2) Minor parties may have a greater relative presence in cyberspace than they have in other mass media, but the major parties' presence on the Web still will dwarf that of the minor parties. Moreover, there will be little or no evidence that minor parties' "greater relative presence in Cyberspace" has any substantive importance for electoral success in the real world.(n2) (3) Party representatives and Webmasters will not emphasize using their Web sites to increase members' participation in party policy making or otherwise to enhance internal party democracy; rather, they will emphasize organizational advantages, such as using them for recruiting new party members or for communicating with party activists or promoting party candidates and platforms.
To examine these predictions, we used several measures to estimate the leveling of the playing field for party competition in cyberspace. For our purposes, political parties can be likened to brand names for commodities. Consumers may know a great deal about a product and then look for a brand that best suits their preferences; or they may know very little about the desirable features of a product, or even whether they wish to purchase it at all, and turn to trusted brand name manufacturers for advice and information. Analogously, citizens may be interested and informed about political issues and then look for the party or parties that support their policy preferences; they may feel or know little about the issues and then look toward trusted political parties for information and guidance; or like consumers in general, citizens may become unhappy or bored with the political products that are readily available and may wish to survey the market to find out what else is being offered. The Internet certainly can facilitate any of these actions.
The inherent-equalization hypothesis implies that a leveling of the playing field among major and minor parties should be reflected in the relative ease with which citizens can find and interact with various political parties on-line. One natural way for this process to occur is to come across a news story on- or off-line that mentions one or more political parties. The story suggests to interested readers that knowing more about the policies or candidates of one or more political parties would be relevant to a better understanding of the issues involved. Although ordinary citizens who desire such an understanding might turn to traditional sources of information--newspapers, political journals, party manifestos, or radio or television programs on politics, netizens might search for relevant party Web sites to find more about the issues from the perspectives of different parties.
One of the ways in which the major parties manifest their domination over minor parties is in news coverage. A simple--albeit crude--measure of this is the number of specific references to particular political parties in the stories. Minor parties continually complain that the mass media restrict their ability to present their views to the public. They claim that if their positions were better known, their support among the population would grow. If the Internet is a powerful new medium for equality, we would expect to see some evidence of a loosening of the stranglehold of major parties on news stories. We investigated whether anything like this phenomenon is occurring.
We employed the following measures to determine the relative prominence or "Web presence" of political parties in various media, both on- and offline. When possible, we looked at these measures not only in the recent past, but also thirty days before major elections when more coverage of political events and issues should appear.
Web sites listed in Yahoo
Backpointers to Web sites found using AlexaPPC(n3)
Hits on Web sites found using AlexaPPC
References in major newspapers and magazines found using Lexis-Nexis
References in on-line news sites, such as Allpolitics or Yahoo News
References in Usenet newsgroups found using DejaNews
In addition, we employed the following indicators to measure the relative sophistication of party Web sites:
Currency: recent updates of site found using AlexaPPC
Pages that comprise Web site found using AlexaPPC
Number of graphics per page on top-level pages
Presence of active graphics
Presence of tables of contents
Search capacity for site
Databases on organization, rules, principles, issues, officers, and so on
Forms for membership, volunteering, receiving newsletters, and so on
Sales or means of fund-raising other than membership
Entertainment
Chat rooms or discussion groups
Personalized information
Findings
Web Presence
Our first measure of party prominence, the number of first-level listings to American party Web sites found in Yahoo, indicated that the situation is much the same as it was in 1996. Web sites of the two major parties predominate, along with a handful of minor parties' sites. The 158 Democratic and 252 Republican party sites comprise 59.5 percent of the 632 listed in late July 1998. Together with the Libertarian, Green, Reform, and Communist party sites, they account for more than 93 percent of all American party sites. These percentages are nearly unchanged from 1996 (Figure 1).(n4)
The same measure applied to major political parties in the United Kingdom yields surprisingly similar results. The 75 Conservative, Labour, and Liberal Democrat party sites comprise 63.0 percent of the 119 Web sites listed. As the Green Party is the only minor party with 10 or more sites listed, however, overall concentration of the Web sites is somewhat less dense than in the United States. Excluding the Greens, minor party sites still account for 28.6 percent of the list (Figure 2).(n5)
Arguably, however, Yahoo may provide relatively more exposure to minor parties than they receive in the real world. After all, each party gets one line, regardless of the number of Web sites it has listed. In truth, we simply don't know how many people actually go to Yahoo's party pages to find a particular party's Web site rather than come across that site by some other means, such as finding a reference to it while browsing for information on a news story or a public policy issue. Here is where "Web presence," as measured by links or "backpointers" from other Web sites to a party's site, must be considered (Hill and Hughes 1998:143).(n6) Generally speaking, we presume that the greater the number of external links that point to a particular party's Web site, the greater the chance of attracting visitors to that site.
The normalization hypothesis predicts that major parties will have a significantly greater Web presence than minor parties. The inherent-equalization hypothesis can tolerate the major parties' having a temporary advantage, but it predicts that ultimately, Web presence of most parties will become nearly equal. Moreover, if the equalization hypothesis holds, we would expect to find that the advantages of major parties over minor parties on the Web would be significantly smaller than their advantages over minor parties in the established news media.
Table 1, which summarizes observations on the Web presence of American political parties, defies a simple acceptance or rejection of either hypothesis. The table includes only those parties with more than 349 backpointers.(n7) Clearly, the Democratic and Republican National Committee (DNC and RNC) sites, with more than 3,000 backpointers apiece, have many more links leading back to them than do their rivals, but by no means do their sheer numbers add up to the majority, as they did for Yahoo site listings. Indeed, the Libertarian Party's site by itself has nearly 80 percent as many links as the DNC and RNC. Moreover, all the minor parties included have complex sites made up of many Web pages, and they also manage to keep their Web sites as fresh as the majors. Finally, the traffic to their sites falls into the same general category as that of the DNC.(n8)
The measures displayed, however, probably underestimate the Web presence of the Democratic and Republican parties. Because the U.S. party structure is federal, many of the major party sites that attract attention belong to the fifty state party organizations or to the myriad county, city, and other subnational political units. The Libertarians' organization on the Net improved considerably compared to 1996. By July 1998, nearly all of the state affiliates had Web sites of their own, and the College Libertarians of the University of Illinois at Urbana-Champaign no longer turned up as one of the party's most elaborate sites. In contrast, however, the Communist Party site listed only twenty regional affiliates. Of these, only twelve had e-mail addresses, and only five--Connecticut, Illinois, Minnesota, Missouri-Kansas, and New York--had Web sites. The Greens listed thirty-five regional offices in the United States (and six in Canada), each with Web pages and e-mail, but this still paled in comparison to the Web structures of the Democratic and Republican subnational organizations.
Notwithstanding its prowess in Web site construction, the Libertarian Party has fewer resources than the major parties to pay to have its sites advertised broadly or placed prominently on party or candidate lists by those search engines that charge for such prominent placements. It also gets less attention from general "nonpartisan" political sources.(n9) Nor do its candidates or leaders have the personal notoriety or celebrity to match that of many of their Democratic and Republican counterparts who maintain their own personal Web sites. Beyond this, the major parties' Web presence tends to be strengthened through having elected officials listed--often with their party affiliations--on governmental Web sites at federal, state, and local levels. The data in Figure 1 and Table 1 suggest that the Democratic and Republican parties still maintain a dominant Web presence that reflects their advantages in the real world. Although the Libertarian Party improved its Web presence relative to theirs between August 1996 and July 1998, other minor parties showed no consistent gains. Still, it is possible that some, such as the Reform Party, could also improve their Web presence in the foreseeable future. The Web sites of most minor parties, however, appear to reflect their marginal existence in the real world. In sum, the data tend to support the normalization hypothesis more than they support the equalization hypothesis regarding major and minor parties in the United States, but the support is hardly conclusive.
When we turn to the United Kingdom (Table 2), the data again show that the major national parties--Conservative, Liberal-Democrat, and Labour-have a greater number of backpointers than do their minor party rivals.(n10) In terms of backpointers and Web pages, the Web presence of the British parties is somewhat less than that of the American parties, but their actual traffic appears to be just as substantial. In contrast to the United States, where minor national (and international) parties like the Libertarians, Reform, Communists, Greens, or Natural Law have the most prominent sites, we find several regional nationalist parties' Web sites as the most prominent.(n11) Once again, the measures probably underestimate the prominence of the major parties' Web presence because the major parties have many more constituency and affiliated organizations with their own Web sites than do the minor parties. The Liberal Democrats have a reasonably attractive home page with (as we shall see later) a number of sophisticated features, but their site has relatively fewer pages than the Conservative and (we believe) Labour parties, and they are surprisingly slow in updating most of them. As found for the minor parties in the United States included in Table 1, the Web sites of the minor parties in Britain included in Table 2 generally have more pages and more recent updates than do those of the minor parties that were not included. Once again, however, even though the pattern tends to confirm the normalization hypothesis regarding the dominant presence of the major parties, neither the normalization nor the equalization hypothesis receives unqualified support.
Presence in the News Media
When we examine the presence of the parties in major newspapers, television news, and Web guides to the news, the dominance of the Democratic and Republican parties is immediately apparent (Table 3). Whereas the backpointer ratio for the Libertarians' Web site in comparison to both the DNC and RNC Web sites was greater than 2:3, the ratio of coverage in the major newspapers (Lexis-Nexis) was less than 1:7 in 1998 and was even smaller during the last weeks of the 1996 election campaign. Worse yet, the Libertarian party received no mention at all on the evening news toward the end of the 1996 campaign, nor did it receive any mention through July 1998. In contrast, the ratio of the 1998 newspaper coverage of the Communist and Green parties to that of the Democratic and Republican parties exceeds that of their backpointers. Many of these stories, however, focus on either Communist and Green parties abroad or on investigation of the Chinese Communist Party's alleged campaign contributions to the Democratic Party in 1996. Television news coverage of both minor parties through July 1998 is virtually nil: The sole mention of either is a comment by a party official, Mary Lou Greenberg, on the scholarly publication of a new edition of the Communist Manifesto. Ross Perot's candidacy--and the millions of dollars he spent on it--resulted in substantial media coverage of the Reform Party during the last weeks of the 1996 presidential campaign. In 1998, however, the ratio of the Reform Party's newspaper coverage to that of the major parties dropped substantially below that of its Web backpointers, and the Reform Party received no television news coverage at all through July 1998.(n12) Similarly, the other minor parties have received smaller proportions of coverage in newspapers and on television than the ratios of their Web backpointers to those of the major parties.
Even when using Web guides to discover references to political parties, the chances of finding Democrats or Republicans remain far greater than those for other parties. Aside from three articles that mentioned the Green or Reform parties, the Yahoo news guide, which searches stories from the past week, contained references only to the Democratic and Republican parties. We have also found that Democratic and Republican party organizations, officeholders, and candidates get more prominent display in politically neutral information sites, such as Project Vote Smart (http://www.vote-smart.org/), Campaign Central (http:// www.clark.net/central/parties.htm), and Agora (http://www.agora.stm.it/ politic/home.htm).(n13)
Only in Usenet newsgroups does the ratio of minor party references to major party references consistently exceed that of Web backpointers. Here the ninety-seven thousand mentions of the Communist Party nearly match the ninety-nine thousand mentions of the Republicans, and the smallest ratio--Libertarians to Democrats--still exceeds 1: 3.(n14) Usenet groups have a long history on the Net, and they include many "old hands" with good technical skills. Usenet participants are also more active politically than most ordinary citizens. As the popularity of the Web has expanded the population of cyberspace, Usenet participants have also become a minority who tend to be more active politically than most Web users (Fisher et al. 1996 ;Taylor 1999). Their postings that mention minor parties appear to be atypical, not unlike the activists who write letters to the editor, make calls to radio talk shows, or respond to television evangelists (Clausen et al. 1965; Hughes and Hill 1998; Johnston 1989).
In Britain, the dominance of major parties over minor parties seems less absolute in both mass and Web media than in the United States (Table 4). True, in the weeks leading up to the 1997 General Election, the Labour and Conservative parties received seven and five times the number of references, respectively, as the Liberal Democrats or Sinn Fein in major newspapers. However, the pattern has changed considerably in 1998: Nearly as many newspaper stories mention Northern Irish parties, especially Sinn Fein and Ulster Unionists, as mention the Conservative or Labour parties. Where evening television news in the United States covered Labour, Conservative, Sinn Fein, and Ulster Unionists during the last weeks of the 1997 campaign, the Irish parties have received the lion's share of mentions through July 1998.(n15) Sinn Fein and the Ulster Unionists receive similarly prominent mentions in Allpolitics and Yahoo News summaries for the United Kingdom and Ireland. Surprisingly, in contrast to the pattern for major and minor American political parties, the Conservative and Labour parties receive substantially more mentions than do the Irish parties in Usenet newsgroups. However, even here references to "Conservative" and "Labour" also include references to Conservative and Labour parties of countries other than Britain.
The mass media and the Web media may pay scant attention to most minor British political parties, but they clearly focused on the Northern Irish parties during the first seven months of 1998. Whether this represents a new prominence for regional political parties or just a short-term anomaly that stems from 1998's hopeful progress in settling the long-standing "troubles" in Northern Ireland remains to be seen. Our data do not cover a sufficiently long period to make any confident judgment. That British minor parties get proportionately more coverage than do their American counterparts, however, does tell us that we cannot generalize about the normalization hypothesis on the basis of data gathered mostly from observing American politics.
Web Site Features
Internet lore suggests that attractive, sophisticated Web sites are a necessary (though not sufficient) condition for using the World Wide Web to increase an organization's viability (Miley 1996; Ubois 1996). It follows that if minor parties' Web sites truly serve to increase their viability, their features must be comparable to those of the major parties' Web sites. The normalization hypothesis predicts that the features of major parties' sites will outshine those of the minor parties; the equalization hypothesis predicts that the minor parties' sites will have comparable features to those of the major parties. Although measuring any single feature may not distinguish the quality of a Web site, measures of a variety of features can uncover patterns and highlight differences (see Hill and Hughes 1998:141-54). Table 5 summarizes the similarities and differences among the national party sites in both the United States and the United Kingdom with regard to ten selected features.
For the United States, the evidence is once again supportive of the normalization hypothesis. The first three columns of Table 5 indicate a near perfect pattern of Democratic and Republican party sites having more features than those of their rivals. We also find, however, that Web sites of the six better-linked minor parties (listed in Table 1) have somewhat more of these features than the Web sites of the other minor parties. Moreover, when we consider subnational party sites, the differences are likely to be even larger. We recall that nearly 60 percent of all American party Web sites belong to the Democratic or Republican parties and another 33 percent belong to the Libertarian, Green, Communist, or Reform parties, whose features are included in the second column. The relatively large numbers of these subnational sites are likely to reinforce the prominence of these parties over others in cyberspace.(n16) The American pattern suggests that the major parties maintain their dominance in cyberspace as they do in the real world. That several minor parties show a strong Web presence and also maintain sophisticated Web sites, however, indicates that the equalization hypothesis cannot be dismissed.
For the United Kingdom, the picture is less clean The minor parties' Web sites tend to show as many as or more features than those of the major parties. The Labour and Conservative party sites stand out only in terms of search capability and collection of personal information. Even though overuse of graphics can make a Web site cluttered and confusing, sophisticated Web sites usually make generous use of graphic links. Thus the Labour and Conservative parties' relatively sparse use of graphics is somewhat surprising. Moreover, the lesser concentration of British party Web sites among the major and prominent minor parties in comparison to American party sites also serves to mitigate the major parties' dominance. It is possible that the Labour and Conservative parties, like the Democrats and Republicans in 1994, are only beginning to put significant amounts of their prodigious resources into the Internet. However, at this time, the data regarding Web site sophistication tend to refute the predictions of the normalization hypothesis.
Organizational Use
To accomplish their primary purpose of electing candidates to office, viable party organizations are normally hierarchical. Even within the decentralized American party system, the separate state, county, and city organizations adopt this organizational structure. Indeed, America's legendary party bosses derived their power from subnational party units, usually big city or county organizations. The equalization hypothesis suggests that as party organizations move onto the Web, they will become more democratic internally. Among other things, they will use the Internet's communication capabilities not merely to cut costs and increase organizational efficiency, but also to increase rank-and-file participation in strategic planning and decision making. In contrast, the normalization hypothesis expects that party officials will use party Web sites primarily to increase the efficiency or to enhance the effectiveness of traditional forms of communication and organization.
We contacted Conservative and Labour party officials in Britain in August 1997 and Democratic and Republican officials in the United States during the winter and spring of 1996 concerning these matters. Our samples were limited and our interviews were informal, but the observations of our informants were nearly uniform. In both Britain and the United States, the officials reported that the Web sites' primary values were in facilitating members' and activists' ability to get information useful for organizing, defending, or otherwise promoting the party and its candidates. Indeed, during the 1996 presidential primary campaign, several officials described use of the Internet as a symbolic commitment to technological advancement or else a preemptive move to ensure their party's or candidate's presence on any new media. The officials expected the electoral impact of the Internet to grow, but at present it was relatively unimportant in comparison to that of television, radio, and (in the United Kingdom) mass-circulation tabloids.
In the summer of 1998, there were as yet only about eight million adults on the Net in Britain, but around sixty million or so in the United States (http:/ /www.nua.ie/surveys/how_many_online/index.html). When these numbers reach a majority of electors, as expected sometime early in the next millennium, officials predicted that their parties would make significantly more use of the sites for persuading and mobilizing the electorate directly, as opposed to focusing on organizing and communicating with party activists. In no case, however, did our interviewees emphasize either expectations or concerns for using the Internet to increase the participation of members or supporters in the internal affairs of their parties. Indeed, Labour and Conservative party officials reported that during the 1997 campaign, they each operated a Web site for party officials and activists that was separate from their public site. The sites were used for intraorganizational e-mail, distribution of information such as policy positions and candidates' campaign schedules, and coordination of campaign efforts (see also Gibson and Ward 1998; Roper 1998; Selnow 1998).(n17)
Conclusions
Although the results are by no means unequivocal, the data generally support the normalization hypothesis as applied to political parties in the United States. Not only do a majority of party Web sites belong to the Democrats or Republicans, but these sites have more backpointers, greater numbers of pages, more sophisticated features, and somewhat greater likelihood of being visited than do those of the minor parties. Nevertheless, minor parties appear relatively more frequently on the Web than they do in various media in the real world. The Libertarian, Green, Communist, and Reform parties stand out from other minor parties in this regard.
Web sites of the Conservative and Labour parties in Britain do not make up as large a proportion of party Web sites as do those of the Democrats and Republicans in the United States. Although these two parties also receive more coverage than do minor parties both on and off the Web, the pattern indicates that minor parties get relatively more attention than was the case for American parties. Indeed, for the first seven months of 1998, coverage of Sinn Fein and the Ulster Unionists equaled or exceeded that of the major parties on several measures. By most measures, the Liberal-Democrats, Scottish Nationalists, and the Social Democratic and Labour Party (SDLP) also received significantly higher coverage than did the remaining minor parties. Whether this coverage of regional parties represents a trend toward equalization or just a short-term surge in attention stemming from recent progress toward peace and reconciliation in Northern Ireland remains to be seen. In our view, the preponderance of data still supports the normalization hypothesis, but further study over time is needed to discern any trend.
Regardless of which hypothesis best fits the data, we should keep in mind that the most popular political party Web sites--indeed political Web sites in general--usually attract less traffic than do the most popular commercial sites. This pattern of commercial dominance, however, is not surprising. In cyberspace, as in the real world, most people's interest in government and politics pales in comparison to their interest in business, commerce, sports, entertainment, and matters of family or personal concern. Notwithstanding the acclaimed importance of public policy decisions, communication about government and politics comprises only a small proportion of the traffic on the Internet.(n18) By and large, the business of the Internet--especially the business of the Web--is business.
Cyberspace is now maintained by private corporations and is populated mostly with consumers of information, products, and services.(n19) Concerns about democratic politics play only a small part in the hype about the Internet in the popular press. A Lexis-Nexis UNIVerse (http://web.lexis-nexis.com/ universe/) search of articles in major newspapers from May 1994 through April 1998 averaged well over one thousand articles every month that mentioned "Internet" or "World Wide Web" or "Information Superhighway (or highway)." However, only 398 of these--fewer than 1 percent--also contained words with the roots "politic" or "democra" within one hundred words of the aforementioned terms. Even the party officials and Webmasters with whom we talked generally emphasized using the Internet for communicating information to the faithful and organizing them more effectively, not for developing more democratic procedures for conducting party business. In the tradition of western capitalism, the information superhighway is here mainly to advertise and sell products and services, not to improve the democratic quality of British and American politics and civic life (Caruso 1996).
In our view, the data on party presence on the Web and in the mass media tend to support the predictions of the normalization hypothesis. We admit that the situation is more ambiguous in Britain, where minor parties show a relatively greater presence than in the United States. Nonetheless, the inherente-qualization hypothesis seems to be fighting against the tide of marketing and commercialization that has swept the Internet. As the World Wide Web drew more people into cyberspace, the established groups have taken notice of the new medium, if only, as Gary W. Selnow argues in our epigraph, to protect their privileged political position in the real world. Moreover, in ordinary times most people neither know nor care very much about politics. Citizens tend to favor established parties and interests, therefore, simply because they are more familiar with them than with their alternatives.
We find few signs that this trend is likely to change, but in the great tradition of our discipline, we remind our readers of the incompleteness of the findings, we note the tentativeness of our conclusions, and we enjoin ourselves and others to engage in further study of electoral politics on the Internet over time.
Notes
We would like to thank Melicent Homan for research assistance for this article.
(n1.) Peter Lewis, "An Ad (Gasp) in Cyberspace," New York Times, Apr. 19, 1994:D1.
(n2.) The critical question remains: What impact, if any, will a minor party's relatively greater Web presence have on that party's political fortune over time? Until we collect more data, we can only offer speculative answers to this question.
(n3.) We use statistics given by AlexaPPC (http://www.Alexa.com). As of February 1999, Alexa uses the term links in instead of backpointers.
(n4.) The percentages for the two major parties varied between 56.5 percent and 60.2 percent between January and August 1996, and the top six parties accounted for 90 percent of the sites. The total number of sites grew from 191 in January 1996 to 477 in August (Margolis et al. 1997:61, 75). As might be expected, after the 1998 election the total number of parties listed by Yahoo declined from the 632 found in late July. Yahoo listed only 556 party organizations on January 31, 1999. The declines, however, were solely among minor parties. The Democratic Party sites increased to 169 (30.4 percent), the Republican to 220 (39.6 percent), and the Libertarian to 68 (12.3 percent). The Reform Party stayed at 35 sites, but the Green Party sites declined to 24, the U.S. Taxpayers to 9, and the Socialists and Communists to 5 sites each.
(n5.) The number of U.K. party Web sites listed by Yahoo, in contrast to those in the United States, increased to 123 between July 1998 and February 1999. The Labour and Liberal Democrat parties each lost one site. The number of Conservative and Green party sites remained the same, but the number of Web sites of various Socialist parties increased to 9.
(n6.) Alexa automatically counts the number of links to the designated Web page. Hill and Hughes enter URLs in Infoseek's "advanced search" function to count the number of external links to any given site (http://www.infoseek.com/find?pg=advanced_ www.html&ud9=advanced_ ;www). (Hill and Hughes then multiply the links by the number of Web pages that comprise the Web site.) Although both Alexa and Infoseek generally produce the same order regarding numbers of external links, Alexa provides a number of additional measures that are useful for our analysis.
(n7.) Most of the remaining minor parties had many fewer backpointers to their sites as well as less complex sites that averaged fewer pages. At least three had not updated their Web sites for more than ninety days. The Patriot Party had ceased to maintain one of its two sites (and disappeared altogether by February 1999); the White Nationalist Party dropped from the listing sometime between April and mid-July 1998, and the Workers Party site existed primarily to sell The Worker, a weekly Marxist periodical.
(n8.) In April through June 1998, the categories that AlexaPPC characterized as having "moderate" or more traffic were limited as follows: top 10, 100, 1,000, 10,000, 25,000, 100,000, and below 100,000 (but apparently above one million).
(n9.) Compare the prominence of Libertarian links (see Table 3), for instance, with those associated with the Democratic and Republican parties on sources like Yahoo news or Allpolitics. (Allpolitics is a partnership of CNN and Time magazine.) In the summer of 1996, Open Text, then a search engine, introduced a "preferred listing" procedure that allowed advertisers to pay for having their advertisement--labeled as such--appear among the first ten results of a relevant search (Sandy McMurray, "Surf the Net--But First a Word from Our Sponsor," The Toronto Sun, June 26, 1996: 48). Goto.com, a spinoff of Idealab of Los Angeles, will advance this concept one step farther: They will sell prominent listings to the highest bidders. As long as a site meets general criteria of relevance, it will be listed near the top, regardless of its centrality to the client's inquiry (Laurie J. Flynn, "With Goto.com's Search Engine, the Highest Bidder Shall Be Ranked First," New York Times, Mar. 18, 1998: DS).
(n10.) Excepting Sinn Fein, Table 2 includes all parties that Alexa lists as having more than 130 backpointers. That Alexa reports only one backpointer for Sinn Fein seems implausible. Infoseek's advanced search facility uncovered only twenty-four external hyperlinks, however, affirming Sinn Fein's surprisingly few direct connections with other Web sites. We suspect that the site's high number of hits stems from the IRA's long and often violent struggle for Northern Ireland's independence from Britain, the election of a new Northern Ireland Assembly, the negotiations surrounding the April 1998 agreement on the future of Northern Ireland, and the violence associated with the 1998 Orange marches.
(n11.) Plaid Cymru maintains a sophisticated, up-to-date Web site that even sends visitors a cookie. It has 144 pages but only sixty-seven backpointers. It also falls into the same category of traffic (below one hundred) as the others. The Social Democratic and Labour Party (SDLP) site's URL is listed under the rubric of its Internet service provider (ISP) (www.indigo.ie), so we cannot get its statistics from Alexa. The party's self-count was 26,358 visitors between August 13, 1996, and August 3, 1998. For comparison, the (American) Pan Sexual Peace Party, whose URL is also listed under its ISP (www.neosoft.com), counted 46,479 visitors between January 27, 1996, and August 3, 1998.
(n12.) The Vanderbilt TVnews archive lists fourteen stories on the Reform Party of candidate and later Minnesota governor Jesse Ventura from August 1998 through January 1999. Only one abstract of these stories--CBS Evening News, Nov. 4, 1998--mentions the Reform Party itself, however.
(n13.) Agora and Campaign Central (Losser and Associates) also offer commercial services, such as Web site design and Web marketing.
(n14.) The figures are inexact. Due to limitations of the DejaNews internal search engine, mentions of the parties include many that refer to foreign "Democratic," "Communist," or "Green" parties, not just parties in the United States. Similar problems occur in using the internal search engines of the Vanderbilt TVnews Archive and of Allpolitics. Even the more powerful search engines, such as those of Yahoo and Lexis-Nexis, differ in their capabilities.
(n15.) The evening news contained thirty-three stories that mentioned Gerry Adams from January 1, 1998, through August 5, 1998, and all of them also mentioned Sinn Fein. In contrast, of seventy-four stories that mentioned Prime Minister Blair, only one also mentioned the Labour party. As far as U.S. television news is concerned, Tony Blair is a man without a party.
(n16.) We do not have a systematic sample of subnational Web sites, but our visits indicate that many also have sophisticated features, particularly those of the Democratic and Republican parties.
(n17.) Between August 1, 1998, and February 15, 1999, the Conservative Party incorporated its Web pages into its new "Listening to Britain" campaign. At http://www.conservativeparty.org.uk/ltb/index.html, the campaign is described as an exercise in listening to people in all walks of life. "We want to identify their aspirations and the challenges, problems and opportunities that they believe Britain will face in the future. As we have identified the problems which will face Britain in the early years of the next century, so we can develop policies, based on Conservative principles, to meet those challenges." The site provides a means for rank-and-file members to use the Web mail to voice their opinions to the party leadership and to facilitate inviting party leaders to local meetings at which issues of public policy would be discussed. Interestingly, however, as of February 15, 1999, the campaign also included organizing special mailing lists for business and church groups to receive information from the leadership on a centralized top-down basis. See http://www.conservative-party.org.uk/cblu/form.html and http:// www.conservative-party.org.uk/ltb/6.htm.
(n18.) For links to relevant sites, see Georgia Tech's Graphics, Visualization and Usability Center's site (http://www.gvu.gatech.edu/user_surveys/others/).
(n19.) GVU Surveys (http://www.cc.gatech.edu/gvu/user_surveys); for links to additional sources on Internet surveys, see http://www.gvu.gatech.edu/user_surveys/others/); Rohit Khare, "Demographic Status Report," 1995 (http://www.w3.org/pub/WWW/ Demographics / StatusReport.htmld/index.html); and Jared Sandberg, "Rush to Claim Turf on Internet Ends Tradition of Free 'Domain Names,'" Wall Street Journal, Sept. 14, 1995: B2.
Legend for Chart: A - Web Site Criteria B - Party Democratic C - Party Republican D - Party Libertarian E - Party Communist F - Party Green G - Party Reform H - Party Natural Law I - Party Democratic Socialist A B C D E F G H I Backpointers 3,179 3,089 2,444 1,425 1,117 826 663 649 Web pages 317 > 1,000 > 1,000 > 1,000 83 133 429 > 1,000 Updates 1-7 days 1-7 days 1-7 days 1-7 days 1-7 days 1-7 days 1-7 days 1-7 days No. of hits: ranked in top (in thousands) Below 100 25 Below 100 Below 100 Below 100 Below 100 Below 100 Below 100
Legend for Chart: A - Web Site Criteria B - Party Conservative C - Party Liberal Democrat D - Party Labour E - Party Alliance Party of Northern Ireland F - Party Scottish National G - Party Sinn Fein H - Party Ulster Unionist A B C D E F G H Backpointers 1,014 662 660 447 421 1 224 Web pages 293 60 NA 68 NA NA 133 Updates 1-7 days 6 months-2 years 1-7 days 1-7 days 1-7 days NA 1-7 days No. of hits: ranked in top (in thousands) Below 100 Below 100 25 Below 100 Below 100 25 Below 100 NA = # not available.
Legend for Chart: A - Databases B - Party Democratic C - Party Republican D - Party Libertarian E - Party Communist F - Party Green G - Party Reform H - Party Natural Law I - Party Democratic Socialist A B C D E F G H I Lexis-Nexis[a] 725 751 103 371 195 119 30 75 Lexis- Nexis[b] 1,254 1,151 118 77 60 375 153 15 Evening TV news[c] 27 9 0 1 (39) (10) (0) (6) 0 0 0 0 (1) (1) (0) (1) Evening TV News[d] 56 101 0 0 (80) (101) (0) (22) 8 79 0 0 (10) (86) (0) (2) AllPolitics CNN/Time[e] 831 841 46 17 65 346 55 3 Yahoo Weekly News USA[f] 24 13 0 0 2 1 0 0 DejaNews 1,000s[g] 130 99 46 97 57 50 61 68 [a] Last six months -- Search paradigms: Democrat!, Communist, Green, Reform, and Socialist within same sentence as Party AND story also contains: (America! OR USA OR United States); Republican and Natural Law within same sentence as Party; August 6-7, 1998 (http://web.lexis-nexis.com/universe). This is a worldwide database that includes major European newspapers. [b] Oct. 2-Nov. 5, 1996 -- Same search paradigms as[sup a]. [c] Jan. 1-Aug. 5, 1998 -- Search paradigms Name AND Party. Summaries then checked for mention of USA party. Resultant count is listed. Total stories found is in parentheses (http://tvnews.vanderbilt.edu/). [d] 1996 -- Same search paradigms as[sup c]. [e] News Archive as of Aug. 6, 1998 -- Search paradigms: "Democratic Party" AND (America OR American OR USA OR United States OR Clinton OR Gore); "Republican Party" AND (America OR American OR USA OR United States OR Dole OR Gingrich); "Libertarian Party"; "Communist Party" AND (America OR American OR USA OR United States); "Green Party" AND (America, OR USA OR United States OR Nader); "Reform Party" AND (America OR American OR USA OR United States OR Perot); "Natural Law Party"; "Democratic Socialist" OR "Socialist Party" (http://allpolitics.com/ 1998/index.html). [f] July 31-Aug. 6, 1998 -- Search Paradigms: "Democratic Party" excluding foreign countries; "Republican Party"; "Libertarian Party"; "Socialist Party"; for others: "Communist Party," "Green Party," and "Reform Party," each with "AND USA" (http://dailynews.yahoo.com/headlines/). [g] Jan. 1-Aug. 7, 1998 -- Search paradigms: "Name AND Party" in English language (http://www.dejanews.com/).
Legend for Chart: A - Databases B - Party Conservative C - Party Liberal Democrat D - Party Labour E - Party Alliance Party of Northern Ireland F - Party Scottish National G - Party Sinn Fein H - Party Social Democratic and Labour Party (SDLP) I - Party Ulster Unionist A B C D E F G H I Lexis-Nexis[a] 1,685 436 2,787 105 489 2,060 413 1,228 Lexis- Nexis[b] 1,219 238 1,604 11 121 229 71 58 Evening TV News[c] 0 0 2 (3) (0) (2) 2 0 38 (2) (0) (38) 2 23 (2) (23) Evening TV News[d] 23 1 28 (23) (1) (30) 0 2 40 (0) (5) (40) 2 11 (2) (11) AllPolitics CNN/Time[e] 13 2 26 2 0 23 0 3 YahooWeekly News UK/Ireland[f] 8 4 12 2 2 12 0 6 DejaNews 1,000s[g] 62 8.3 54 1.3 2.1 21 6.7 3.8 [a] Feb. 1-Aug. 7, 1998 -- Search paradigms: "Conservative" in same sentence as "party"AND (UK OR United Kingdom OR Britain OR British OR Tory OR Tories); "Liberal" immediately preceding "Democrat! "AND both terms in same sentence as "party" AND (UK OR United Kingdom OR Britain OR British); "Labor" or "Labour" in same sentence as "party" AND (UK OR United Kingdom OR Britain OR British); "Alliance" immediately preceding "party" AND (Ireland OR Britain OR British); "Scottish" immediately preceding "National" and both terms in same sentence as "party"; "Sinn Fein"; SDLP; "Ulster Unioni!". [b] Apr. 1-May 1, 1997 -- Same search paradigms as[sup a]. [c] Jan. 1-Aug. 5, 1998 -- Search paradigms: "Conservative AND Party" OR "Tory OR Tories'; "Liberal AND Democrat"; "Labour Party" or "Labor Party"; "Alliance AND party AND Ireland"; "Scottish AND National"; "Sinn AND Fein'; "Social AND Democrat* AND Labour AND Party" OR "SDLP'; Ulster AND Unionist. [d] 1997 -- Same search paradigms as[sup c]. [e] News Archive as of Aug. 8, 1998 -- Search paradigms: "Conservative Party" and ("UK" or "United Kingdom" or "Britain" or "British" or "Major"); "Liberal Democrat Party"; ("Labor Party" or "Labour Party") AND ("Blair" or "UK" or "United Kingdom" or "Britain" or "British"); "Alliance" and "Party" and "Ireland"; "Scottish National Party"; "Sinn Fein"; "SDLP" OR "Social Democratic"; "Ulster Unionist." [f] Aug. 2-9, 1998 -- Search Paradigms: "Conservative Party" OR "Tory" OR "Tories"; "Liberal Democrat"; "Labour Party"; "Alliance AND Party AND Ireland"; "SNP" OR "Scottish National Party"; "Sinn Fein"; "SDLP" OR "Social Democrat*"; "Ulster Unionist" (http://www.yahoo.co.uk/headlines/). [g] Jan. 1-Aug. 7, 1998 -- Search Paradigms: "SDLP"; "Alliance AND party AND Ireland"; "Ulster AND Unionist"; for all others: "Name AND Party"; all searches in English language.
Legend for Chart: A - Web Site Criteria B - United States Democrats and Republicans C - United States Other Parties in Table 1 D - United States Other Minor Parties E - United Kingdom Labour and Conservative F - United Kingdom Other Parties in Table 2 G - United Kingdom Other Minor Parties A B C D E F G Median no. of graphics: top-level pages 9 (N=2) 5.5 (N=6) 3 (N=17) 2 (N=2) 3 (N=5) 4 (N=27) Sites with active graphics 100% 33% 53%[a] 0% 80% 30% Sites with tables of contents 100% 67% 53% 100% 80% 74% Sites with search capacity 100% 17% 18% 100% 20% 11% Sites with party databases 100% 100% 100% 100% 100% 100% Sites with forms 100% 100% 100% 100% 100% 93% Sites with sales/fundmising 50% 33% 29% 100% 0% 15% Sites with entertainment 50% 0% 0% 0% 0% 4% Sites with chat rooms/ Discussion groups 50% 50% 35% 0% 20% 11% Sites with personalized information 100% 67% 65% 100% 40% 41% [a] Includes one with audio though no active graphics.
References
Caruso, Denise. 1996. "The Net Nobody Knows." Utne Reader 75 (May-June):41-49.
Clausen, Aage, Philip E. Converse, and Warren. E. Miller. 1965. "Electoral Myth and Reality." American Political Science Review 59:321-36.
Fisher, Bonnie, Michael Margolis, and David Resnick. 1996. "Surveying the Internet: Democratic Theory and Civic Life in Cyberspace." Southeastern Political Review 24(Sept.):399-429.
Gibson, Rachel K., and Stephen J. Ward. 1997. "U.K. Political Parties and the Internet: Prospects for Democracy." Working Papers in Contemporary History and Politics 13, European Studies Research Institute, University of Salford.
Gibson, Rachel K., and Stephen J. Ward. 1998. "The Internet and Intra-party Democracy: Discourse and Dissent." Presented at the annual meeting of the American Political Science Association, Boston, Sept. 3-6.
Hauben, Michael, and Rhonda Hauben. 1997. Netizens: On the History and Impact of Usenet and the Internet. Los Alamitos, CA: IEEE Computer Science Press.
Hill, Kevin A., and John E. Hughes. 1998. Cyberpolitics: Citizen Activism in the Age of the Internet. Lanham, MD: Rowman and Littlefield.
Hughes, John E., and Kevin A. Hill. 1998. "The Political Culture of New Media Users: A Snapshot of Talk Radio and Internet Users." Presented at the annual meeting of the Midwest Political Science Association, April 23-25, 1998, Palmer House Hilton, Chicago.
Johnston, Michael. 1989. "The Christian Right and the Powers of Television." In Manipulating Public Opinion, ed. Michael Margolis and Gary A. Mauser. Pacific Grove, CA: Brooks/Cole.
Margolis, Michael, David Resnick, and Chin-chang Tu. 1997. "Campaigning on the Internet: Parties and Candidates on the World Wide Web in the 1996 Primary Season." Press/Politics 2(1):59-78.
Miley, Michael. 1996. "Good Interactive Sites Take Time, Hard Work." Mac Week (Apr. 1): 32-34.
Resnick, David. 1998. "Politics on the Internet: The Normalization of Cyberspace." In The Politics of Cyberspace, ed. Chris Toulouse and Timothy W. Luke.
Roper, Juliet. 1998. "New Zealand Political Parties Online: The World Wide Web as a Tool for Democratization or for Political Marketing?" In The Politics of Cyberspace, ed. Chris Toulouse and Timothy W. Luke.
Selnow, Gary W. 1998. Electronic Whistle-Stops: The Impact of the Internet on American Politics. Westport, CT: Praeger.
Taylor, Roger. 1999. "Net Spreads to the Mainstream." Financial Times (Jan. 30-31).
Ubois, Jeff. 1996. "Ten Essential Steps for Maintaining a Web Site." MacWeek (Apr. 1):28-30.
Biographical Notes
Michael Margolis is Professor of Political Science at the University of Cincinnati. Margolis and Resnick are the authors of Politics as Usual: The Cyberspace "Revolution" (Thousand Oaks, CA: Sage Publications, forthcoming).
David Resnick is Associate Professor of Political Science and Director of the Center for the Study of Democratic Citizenship at the University of Cincinnati.
Joel D. Wolfe is Professor of Political Science at the University of Cincinnati. His most recent book is Power and Privatization: Choice and Competition in the Remaking of British Democracy (New York: St. Martins Press, 1996).
Address: University of Cincinnati, Department of Political Science, PO Box 210375, Cincinnati, OH 45221-0375; phone: 513-556-3300; fax: 513-556-2314; e-mail: Michael. Margolis@uc.edu; David. Resnick@uc.edu; Joel. Wolfe@uc.edu.
Paper submitted October 2, 1998; accepted for publication January 20, 1999.
~~~~~~~~
By Michael Margolis; David Resnick and Joel D. Wolfe