It is increasingly apparent that use of the world wide web is growing rapidly, but we may not have realized just how far it has come in such a short period of time. In 1996, 40 million people had access to the Internet (Selnow, 1998, p. xiv). By 1998, this figure had grown almost 50%, as 57 million people now have access to the Internet (Scenic Digital, 1999). A recent Pew Research Center Poll found that 41% of Americans use the Internet in 1999, up from 23% in just two years ("Internet Users," 1999). Buie predicts that by 2000, "more than half of Americans will be connected to the Internet" (1999). While many voters still do not have Internet access, tens of millions do, and there can be no doubt that the number of people with Internet access will continue to increase substantially as the price of equipment decreases and software browsers become more user friendly.

Of course, some Internet users are not politically active -- but a surprising number use the web to obtain political information. Mallery reported that "modem owners have a much higher probability of voting (Skiba, 1995, p. 20A). Bucy, D’Angelo, and Newhagen (1999) found that "respondents who said they voted also reported heavier media use across all channels, including. . . the Internet" than non-voters (p. 342). An exit poll from the 1996 election found that of those who voted, "twenty-six percent are regular Internet users" (Buie, 1999, emphasis original). Furthermore, many people use the web expressly to obtain political information. An AT&T public opinion poll in 1996 revealed that 65% of respondents wanted to research candidate issue positions on the web (Gach, 1996). Buie reported that a Washington Post poll found that "one-third of voters said the Internet was a source of information about the election" (1999). Seventy percent of people aged 18 to 25 believe the Internet is a useful source of information on "issues important to the country. They were more likely to look to the Internet than to more traditional sources of information" ("Young Voters," 1999, p. 8A). Thus, the world wide web is increasingly viewed as a source of information, especially for younger adults. Some voters visited non-partisan sites like PoliticsNow, Politics1, or Project Vote Smart, but IntelliQuest reported that 24% of Internet users had accessed a candidate site in 1996 (Stone, 1996, p. 50).

There is also an indication of voter influence from information obtained via the Internet. First, information is obtained from on-line sources. More Internet users could identify Bob Dole as the candidate who proposed a 15% tax cut than non-users, and more users associated Bill Clinton with the phrase "bridge to the future" than non-users (Pew, "One in Ten," 1996). This information may influence their voting decisions. Chandrasekaran (1996) reported that after the 1996 elections, "9% of voters surveyed said information they found on the Internet influenced their vote. That figure translates into about 8.5 million people nationwide" (p. A4). Buie believes that "millions of [Americans] will use their computers to participate in the political process" in the 2000 election (1999). Hansen (1999) found that viewing presidential primary web sites in early 1999 changed perceptions about candidates, but not levels of trust or support. The Internet, as a completely new medium for political campaigning, is already beginning to exert its potential for disseminating information and thereby affecting the voting behavior of millions of voters.

With a growing number of people connected to the web, many of whom actively seek political information, we should not be surprised that politicians are beginning to tap the vast potential of the Internet for their political campaigns. In 1996 "campaigns often provided their website addresses (URLs) in brochures, on the air, and in speeches. Bob Dole gave his URL on national television in the second presidential debate" (Selnow, 1998, p. 88). Dole’s world wide web site generated considerable traffic: it "lured a grand total of three million hits, 10,000 people have joined the e-mailing list, and 1,700 have signed on to volunteer." Similarly Buchanan’s site averaged about 10,000 visits per day (Dongen, 1996, p. 21). The Clinton/Gore campaign, the Republican and Democratic Parties, and other candidates and parties also had Internet sites in the 1996 election (Stone, 1996). Thus, political candidates and parties are increasingly utilizing the Internet as a medium for communicating with voters.

Campaigns & Elections surveyed 270 candidates, staff members, and consultants who were working on 1998 campaigns in local, state, and federal elections. Sixty-three percent of campaigns reported having a web site and 20% planned to create one. Almost ninety percent thought that the Internet is important and is already changing, or will change, political campaigning ("Campaigns & Elections," 1999). The survey identified (from most to least important) ten purposes of the Internet in politics: giving voters information about candidacy, giving information about policy, giving political information and campaign news, communicating with supporters and endorsers, providing information to voters about how and where to vote, recruiting volunteers, providing information and news about the community, seeking voter opinions on issues, raising money, and attacking the opposition. Only 4.7% of respondents thought their web sites were unsuccessful, although many said it was too early to tell if they had accomplished their goals ("Campaigns & Elections," 1999). Nearly every site (97.1%) included biographical information about the candidate. Most contained discussion of policy (89.9%), links to other web sites (78.2%), and some feedback, communication, or survey option (52.4%). Many (38.2%) solicited volunteers or financial contributions. Only a few sites provided audio or video (6.5%) or chat rooms (4.1%; "Campaigns & Elections," 1999).

In the 2000 presidential election cycle, Steve Forbes became the first candidate to announce his campaign via the Internet (and other candidates made available text and video from their announcement speeches on their web sites). Forbes declared that the Internet is "a way to bring politics and power back to the people in a way that was not possible in the era of mass media" like radio and television ("Forbes Begins," 1999, p. 8A). Another indication of the growing importance of the Internet in political campaigns is the fact that in March of 1999, almost one year before the first primary of the 2000 campaign, the leading presidential candidates had Internet sites: Lamar Alexander (Lamar Alexander for President), Bill Bradley (Bill Bradley for President), Pat Buchanan (Buchanan 2000: Patrick J. Buchanan for President), George W. Bush (George W. Bush Presidential Exploratory Committee), Elizabeth Dole (Elizabeth Dole for President Exploratory Committee), Steve Forbes (Steve Forbes National Online Headquarters), John Kasich (Kasich 2000), Alan Keyes (Keyes 2000 - Alan Keyes for President), John McCain (John McCain for President), Dan Quayle (Quayle 2000), and Bob Smith ("Bob Smith for President). Al Gore opened his site on April 6 (Gore 2000). There are also some web pages devoted to drafting potential candidates (e.g., Tom Harkin, Fred Thompson). The list of campaign (or exploratory committee) web pages is remarkable given the early date. The Internet as a medium for political campaigning is clearly here to stay. As Birdsell, Muzzio, Taylor, and Krane (1996) observed, web "is well on its way to becoming an important medium for campaign communication" (p. 34).

However, this potentially important medium -- new but with a staggering and rapidly increasing audience reach -- is not well understood. There is relatively little scholarly literature on the Internet as a medium of political campaigning. Nor have candidates figured out how best to use this new tool. Faucheux, editor-in-chief of Campaigns & Elections, declared that "The Internet is not only the newest [campaign] medium, it is the fastest-growing and least understood" ("63%," 1999). Witt, executive editor of the PoliticsNow web site, explained that campaign sites in 1996 were not interactive because "the campaigns don’t know who their target audiences are on the net" (quoted in Stone, 1996, p. 44). Hurwitz observed that 1996 political campaign "site developers were so mired in the broadcast model that they overlooked the extraordinary benefits of getting information from voters, thereby discarding the medium’s greatest gift" (Selnow, 1998, p. xxiii). More work is needed for scholars, candidates and their advisors, the media, and voters to understand the potential, and the pitfalls, of the world wide web for political campaigning.

This essay will explore the emerging campaign medium of the world wide web with an analysis of presidential web sites in the pre-primary season of the 2000 campaign. First, we will elucidate the advantages and disadvantages of campaigning on the world wide web. Then we review the literature on the Internet as a medium for political campaigns. Third, we develop a set of design criteria for evaluating political web sites. Next, we analyze presidential candidate web sites in the pre-primary period of the 2000 campaign, evaluating them with the criteria we developed. Finally, we discuss the implications of our investigation for political campaigning.

As we discuss the advantages and disadvantages of Internet campaigning, we will focus on the pros and cons for candidates. However, we will also consider the voters’ point of view.

Advantages of the Internet for Campaigning

The Internet has nine advantages for political candidates. First, of course, the world wide web is an additional medium that can supplement traditional campaign media. Political campaigns have a variety of media available for conveying their messages, including stump speeches, newspaper advertisements, direct mail, radio spots, television spots, talk shows, and debates. Each medium has its own advantages and disadvantages, and while there is overlap in the audiences they possess, each medium reaches a different configuration of voters. Candidates employ the mix of media that best fits their budget, message, and target audience. The Internet adds another possibility, increasing the candidate’s options. However, it has several other advantages that are unique to this medium.

Selnow (1998) declared that the Internet should not be considered merely another medium because it is the "master medium." The Internet can include text and photos like newspapers, audio like radio, and audio, text, and video like television. The quality of video on the Internet doesn’t yet match video on television, but the technology is persistently improving. Although he doesn’t discuss direct mail, candidates can use the Internet to distribute the same text that appears in these messages without any postage costs.

As indicated above, the potential audience for a world wide web site is huge and growing. Although many people lack Internet access, millions are connected, and that number rises daily. Furthermore, as indicated above, about a quarter of voters visited candidate web sites in 1996 and many reported that the information obtained on the Internet had influenced their vote. A third advantage is the huge and insistently increasing size of the potential Internet audience.

Fourth, Tedesco, Miller, and Spiker (1999) explained that "web pages are much less expensive than televised political ads" (p. 53). When you consider how much contemporary candidates spend on television advertising, this advantage comes into focus quickly. For example, in 1996, Dole, Clinton, and Perot jointly spent about $200 million on television advertising (Devlin, 1997). Selnow (1998) also argues that using the Internet is far more convenient, or costs campaigns less time, than producing television spots:

You need writers, producers, camera operators. . . For television spots, you need someone to place the ads and buy the time. . . And if the candidates appear in the spots, they must prepare, report to a studio, and powder up before they sit in front of the camera. . . . The Internet posed no such barriers. In fact, campaigns were confronted by an army of young workers and volunteers [to develop web pages]. (p. 83)

So, using the Internet to disseminate campaign messages involves less cost than other media in terms of time, personnel, and money.

A fifth advantage of the Internet is that it provides an incredible opportunity for audience adaptation and personalization of campaign messages. Selnow (1998) explains that "Through e-mail, it allows the delivery of individual messages and personal appeals to voters. Through chat rooms, it offers candidates direct access to virtual communities of voters. Through the web it allows voters to seek out items of personal interest about the candidate and the campaign" (pp. xxi-xxii). The ability to personalize messages through e-mail, to obtain direct (virtual) access to voters, and to permit voters to seek out information that interests them personally is very important.

We particularly want to elaborate on the third aspect of personalization or adaptation. Different voters have different interests. In the 1996 campaign, for example, an NBC News/Wall Street Journal poll asked which factor was most important in determining their presidential vote: 65% of respondents said that the candidates’ positions on the issues was most important, while 27% said the candidates’ personal character was most important. A web site can provide information on both the candidate’s character and on his or her positions on the issues. Visitors have the power to decide which information to access when they visit a web site. Voters who consider character paramount can spend most, or all, of their time on a web site learning about the candidate’s character. On the other hand, those who care more about issues or policy can devote their time to acquiring information about the candidate’s issue positions. That same poll asked respondents which issue was most important, finding that 33% selected jobs and the economy, 18% education, 18% the budget deficit, 14% Medicare and health, 9% taxes, and 9% crime and drugs. Candidate web pages can include position statements and discussions of accomplishments in each issue area, and visitors who care about issues have the power to choose which topics to explore, and how much time to spend on each. Thus, if a web page is well-designed, individual voters have the power to learn about the topics that matter most to them -- and to ignore or skim the topics of little or no interest. No other medium has this incredible potential to permit a multitude of diverse auditors to each tailor a rhetor’s message to suit their own individual interests and concerns.

Sixth, the Internet is, at least potentially, interactive: "unlike all the other media, which send but do not receive, the Internet can interact with its users as individuals" (Selnow, 1998, p. 42). Candidates can (and do) include informal polls on their web sites to assess the opinions of voters who visit their sites. Candidates can solicit feedback about their messages or policy positions through a web site. Candidates can, themselves or through a staff person, interact directly with voters. Candidates can include a chat room so that interested voters can interact in "virtual space" -- and the campaign can monitor and learn from those discussions. The Internet is more interactive than any other medium besides face-to-face interaction (and perhaps the telephone). And the world wide web is clearly more efficient than dyadic interaction, which, by definition, is one-on-one contact. Depending only on the increasingly sophisticated hardware and software, thousands (or more) of voters can interact with a candidate’s web site simultaneously.

A seventh advantage is that candidates can provide longer, more complete messages to voters. Knisley, director of electronic communication for the Republican National Committee, explained that "I see the Internet. . . as an extension of the 30-second sound bite, which doesn’t handle issues very well. The Internet allows us to offer more depth for the topics presented in other places" (Selnow, 1998, p. 82; see Tedesco, Miller, & Spiker, 1999). We know that television spots are predominantly 30 seconds long (Benoit, in press), a length of time that does not permit candidates to discuss issues in depth. Web pages provide candidates with the opportunity to develop their policy positions in depth (of course, some politicians may prefer to remain vague on some issues). As noted above, a voter can read, listen to, and watch as much information as the web site provides on topics that matter to each individual voter. Web pages have the potential to provide a great deal of information for voters who are interested in a topic. As we will see next, the news media can only be counted on to provide voters with ever-decreasing sound bites from candidates’ messages.

Of course, the candidates’ messages are not the only places voters can obtain information about the candidates: there is also the nightly news. However, we may not realize that the network news is simply not the optimum venue for voters to obtain information about candidates and their issue positions. First, both the number (Steele & Barnhurst, 1996) and length of campaign stories (Hallin, 1992) has decreased 20% from 1968-1988. Second, when they do cover the campaign, the news media devote more time to the competitive nature of the campaign than to the issues of the election. Lichter, Noyes, and Kaid (1999) lamented "the media’s tendency to cover the ‘horse race’ aspect of the campaign while offering relatively little information about substantive policy issues" (p. 4). Third, voters hear less and less from the candidates themselves when the media does cover the candidates’ messages. Hallin (1992) reported that when candidates are allowed to speak on the news (e.g., in video-clips and interviews), the average quotation from a political candidate in the news shrunk from 43 seconds in 1968 to only 9 seconds in 1988. Lichter, Noyes, and Kaid (1999) found that this figure dropped to a mere 8.2 seconds in 1996 (p. 6). So, the news is far from an ideal source of information about the issues in a presidential campaign.

Furthermore, the fact that the news rarely lets candidates speak for themselves means that the media functions as a filter, determining what parts of a candidate message reaches viewers and offering interpretations of those messages:

Press reports insert the journalist between the candidate and the voter with three effects. First, they filter information from interviews or political events. What gets through the filter may not be what the candidate intended. . . Second, reporters interpret. . . . The context, explanations, even a description of the candidate’s appearance can affect audience perceptions of the candidates and their messages. . . . Third, news stories look for angles that may not square with the candidates’ agendas. (Selnow, 1998, pp. 90-91; see also Tedesco, Miller, & Spiker, 1999)

A site on the World Wide Web, in contrast, gives candidates the opportunity to communicate directly to voters, without permitting the mass media to select what parts of a candidate’s message voters will see, interpret the meaning of that message, or comment on the significance or implications of that message (cf. Lichter, Noyes, & Kaid, 1999). Selnow (1998) explained that "You set the agenda and discuss the things you find important. If you don’t want to bring up the fact that you misled kids on the spelling of potato, you don’t have to. Instead, you can brag about your proper spelling of most vegetables, or about your constituent services or your war record" (p. 85). A representative from Dole’s 1996 campaign explained that the Internet "offers a direct avenue of communication between the Dole campaign and the American people" (Skiba, 1995, p. 20A). Thus, a candidate’s Internet site is completely under the candidate’s control, unmediated and unfiltered by the news media.

Another advantage of the Internet was outlined by Trevor Kaufman, a member of the company that designed the Clinton/Gore site in 1996. He explained that "One of our missions is making sure the site is very modular so the campaign can do some rapid-response updating as we get closer to the election" (Stone, 1996, p. 44). Because it is much quicker to edit and reload a web page than, say, to film and distribute a television spot, the Internet is a very quick response medium.

There are five advantages of the Internet for voters as well as these pluses for candidates. Tedesco, Miller, and Spiker (1999) argued that "The biggest advantage of the Internet for campaigning is that it feeds the public’s desire for information" (p. 63). Second, Internet access is increasingly available as the cost of computers and connections fall. Some people access the Internet via public (e.g., library) terminals. Third, with the explosion of information available on the world wide web, a voter can obtain information when the voter wants to receive it (of course, web surfers should be careful to evaluate quality of the information they obtain via the Internet, because there are fewer checks than with traditional media). Just as VCR’s permit television viewers to watch programs when it is convenient, the Internet allows voters to seek information about political candidates whenever they want. A fourth advantage is that the web empowers users in a second way: Voters can control the type and amount of information they access. They can read about the candidates’ character, or accomplishments, or issue positions, while ignoring other information. Internet users can skim the information available on the web or they can go into as much depth on a topic as the site provides. Fifth, the Internet provides access to multiple sources (opposing candidates, parties, supporters, media, or public interest sites like Project Vote Smart), so web users can seek partisan, opposing views or search for sites that are neutral. In short, the Internet provides a wealth of information that voters can use when and how they choose.

Thus, the Internet as a medium for political campaign messages has numerous advantages for candidates and voters alike. We believe that the most important advantages are that it is interactive, unfiltered, and that it empowers voters to learn what they want when they want to learn it. As the Internet continues to grow, these advantages will become even more important.

However, we must recognize that the Internet has drawbacks as well as benefits. We identify four primary disadvantages of Internet campaigning for candidates and reject one potential limitation. First, despite the explosion in Internet access described earlier, many voters still do not use the Internet. Some are not connected to the world wide web and will never access the Internet. Most research has found that certain demographic groups (younger, better educated, more affluent) are disproportionately represented on the web. However, one interesting recent development emerged in Pew Research Center poll which found that those who use the Internet increasingly resemble the general population ("Internet Users," 1999). Nevertheless, the fact remains that even if predictions for growth hold true, half of the population will not have Internet access in 2000.

Second, those voters who use the Internet may not visit the candidates’ sites. An absolutely crucial truth is the fact that web sites must attract viewers. As Adelaide Elm, a founding board member of project Vote Smart, explained, the Internet "doesn’t come into their home just because they’ve got the TV on; they have to make the effort to go get it" (Ubois, 1996, p. 58). Television spots, for example, appeal to the audience who is attracted to the programming in which the spot is broadcast. A candidate advertising on "The CBS Evening News," for example, pays for the opportunity to communicate with people who tune in to see the news. But with a web site, voters must be attracted to visit the candidates’ web sites and then be enticed to read some of the material available on it. Similarly, television spots are seen multiple times when time is purchased on several different television shows watched by a viewer or when time is purchased on different episodes of the same show. However, candidates must give people a reason to return to their sites if they want voters to have repeated exposure to their web messages or to view updated or added information. Thus, candidates must work to realize the huge potential audience of Internet users. We would note that many people throw away direct mail without reading those campaign messages, and even the much vaunted television spots can only guarantee a potential audience (i.e., viewers can ignore a political spot or change channels).

Third, one of the advantages of web sites discussed above is that they are relatively inexpensive, compared with other media. However, there are costs (both monetary and staff time) involved in using the Internet, and these costs are directly tied in to some of the potential advantages of this medium. A web site which is regularly up-dated is more costly than one that is created and then ignored. Selnow (1998) noted that one of the complaints against many 1996 sites is that "They assembled a page early in the race, then abandoned it" (p. 80). Digitizing video and audio to realize the multi-media potential of the Internet requires time and money. Diversifying a web site so that it includes materials appealing to a wide variety of voters (allowing audience personalization and adaptation) is more costly than constructing a simpler web site. Interacting with voters takes time. Obtaining an easy Internet address, its URL (universal resource locater, or web address), can be costly. In 1999, for example, someone bought the rights to the URL "www.forbes2000.com" and later sold it at considerable profit to Forbes’ campaign ("Candidates Find," 1999). Thus, while web sites may be a bargain, the better the site, the more it will cost the campaign in both time and money.

There are also technological liabilities to this new technology, a fourth limitation of the Internet. Many web users access the Internet via modems of varying speed. Graphics and video take time to load, and those accessing the web are prone to move on to another Internet site if a page takes too long to access. Given the present state of installed browser software, options like video and audio require additional "plug-in" software which must be downloaded and installed by the user. If the user doesn’t have the required software and is unwilling to take the time to obtain it, that voter will move on to another part of the site (skipping that particular message) or leave the candidate’s site entirely. These potential glitches could even mean more than a lost opportunity to communicate the candidate’s message: Frustration could alienate voters.

It could be argued that another disadvantage is that Internet campaigning is a waste of time because web sites are unlikely to attract, or persuade, voters from the opposing party. Jamieson and Birdsell, for example, write that presidential debates rarely change the voting intention of committed partisans (1988, p. 161). It is easy to imagine this argument applied to web sites. However, web sites are potentially important for two important reasons besides converting committed partisans. First, during the primaries, party affiliation cannot be used to select a party’s nominee because all candidates belong to the same party. Thus, campaign messages, like web pages, have the potential to influence even committed partisans in the primary phase of the campaign. Given the fact that some incumbents are weakened by factors such as the economy (e.g., Gerald Ford in 1976, Jimmy Carter in 1980, George Bush in 1992), it is possible that several candidates from the opposing party could defeat the incumbent. If so, this means that the primary campaigns in such years may actually determine the outcome of the general election (i.e., who gets to face the weak incumbent). Web sites are a potential source of information during primaries.

Second, neither party controls enough of the electorate to assure a victory in the general election. No third party candidate has come close to winning the presidency in recent years, so it is essential to have the support of the Republican or the Democratic Party to win the election. However, neither party has enough voters to determine the outcome of the election. According to the Center for Political Studies National Election Study, more people identified themselves as independents (32%) than Republicans (29%); Democrats also fell short of a majory of voters (38%; Flanigan & Zingale, 1998, p. 61). Candidates must attract independent and uncommitted voters (and the few party members who do defect) in order to win the general election, and it is possible for those voters to be persuaded by messages like candidate web sites. Thus, we do not consider it to be a significant disadvantage that web sites may not be able to convert committed partisans.

There are also disadvantages to the Internet for web users. First, there is the question of access. While hardware is getting cheaper and software more user-friendly, many people are not connected and may never adopt this medium. As Selnow explained, "Millions of people browsed the web during the 1996 elections, but many more millions stayed away, their absence a matter of cost, inertia, and fear" (p. xxiii, emphasis original).

A second limitation of the World Wide Web is that users must find a site before they can use it. Internet search engines often generate a plethora of "hits" that hide the sites a user seeks. For example, searching on "Steve Forbes" brings up the site "Steve Forbes for President" (http://www.forbes-2000.com) which opens with two pictures of Forbes, one between two flags, a logo ("Steve Forbes 2000: National Online Headquarters"), and the statement that Forbes is the "first person in U.S. history to announce their [sic] presidential campaign on the Internet." However, this site wasn’t created by Forbes’ campaign: It sells a computer program offered by the "sponsor" of this site. Searching for "John McCain brought up a non-official web site for John McCain which may confuse visiters into thinking that they have visited his official web page (http://www.mccainforpresident.com). Searching for "George W. Bush" brought up site for his 1998 gubernatorial campaign, which could create some confusion. Internet users who encounter difficulty finding a candidate web site may give up searching or decide to visit instead one of the non-candidate sites that was pulled up in their search. Nor does guessing at the URL always work. For example, "Gore2000" brings up an entrepreneur who sells Gore campaign buttons and bumper stickers. Forbes’ official site (despite the logo on the other site must mentioned) is http://www.forbes2000.com -- without the dash. Thus, people must be able to find the candidates’ web sites for this medium to have the opportunity to inform or persuade voters.

Third, each candidate site is, by definition, intensely partisan rather than objective. Stone (1996) was particularly pessimistic, declaring that campaign sites in 1996 were "filled with meaningless rhetoric, irritating propaganda." He wrote that the Clinton/Gore site "is nearly devoid of objective information for voters looking to educate themselves" and considered the Dole/Kemp site to be similar (p. 44; see also Tedesco, Miller, & Spiker, 1999). It is unclear to us that we ought to expect a candidate’s site to be objective. Accordingly, people should evaluate the source when seeking information via the Internet just as they should evaluate the quality of other sources of information. The information available on the Internet tends not to be subjected to the same scrutiny and quality control accorded to other media. For these reasons, voters would be well-advised to visit both (or several) candidates’ sites, and neutral sites as well, to try to get a complete picture. Finally, voters must be careful not to be confused into thinking that a parody site is sponsored by a candidate (Warnick, 1998).

Thus, the world wide web has a variety of advantages and disadvantages for both candidates and voters. In our estimate, the advantages far outweigh the disadvantages for each group. This new medium is already large, it is growing, and political candidates and voters are actively using it. It is imperative that we develop an understanding of this new campaign medium.

Hill and Hughes (1997) analyzed Internet usenet group discussions, concluding that "a large quantity of political discussion is taking place across the Internet" (p. 24). Warnick (1998) examined parody on the web, but did not analyze serious candidate web pages. Klotz (1997) reported that in 34 Senate races in 1996, 50 of 68 candidates had a web page. These pages tended to include a photo of the candidate and "links to a biography, issues section, and contact information" (p. 483). Klotz (1998) found that 94% of congressional web sites in 1996 were positive, 4% comparative, and 2% oppositional. Some web pages had no attacks: 47% of pages by challengers, 69% by incumbents and 77% by open-seat candidates were positive.

Margolis, Resnick and Tu (1997) analyzed presidential web sites in the 1996 primary campaign. They found that "Republican and Democratic party sites outnumbered all other American party sites" (p. 65). Clinton did not have an official candidate site until July 10. They described standard features of candidate web sites: signing a guest book, volunteering to help the campaign, fact-sheets and speeches, and links to political party organizations, groups, and people (p. 71). The Dole/Kemp web page included an interactive map that would reveal the names of Dole’s endorsers by state, a game involving trivia about Dole, screen-savers and images that could be down-loaded by visitors, audio clips. Not surprisingly, they found the sites from the Republican and Democratic Parties, and their candidates, to be more sophisticated than many of the minor party and minor candidate sites.

Selnow (1998) characterized presidential candidate sites in the 1996 general election campaign:

Sites posted biographical sketches and the obligatory photos, sometimes with family and flag, always with the candidate grinning or staring longingly into the future. They carried quotable quotes and provided links to speeches, policy positions, and flattering news stories. At the end it was their custom to offer e-mail links to the candidate. (pp. 81-82)

More detail on the 1996 campaign sites can be found in the study by Tedesco, Miller, and Spiker (1999), who reported that virtually every serious presidential contender’s campaign included a www home page in 1996" (p. 52). Every Republican web site, they noted, "provided information concerning how to make financial contributions, how to get on low-technology mailing lists, and some information regarding policy positions" (p. 52).

The Dole/Kemp site had six major links: "About the Team" consisted of biographical information on the candidates and their families, photos, and a link to Dole’s hometown, Russell Kansas; "The Dole Agenda" discussed 25 issues, made speeches news releases, and issue briefs available, and had information tailored to every state; "Get Involved" permitted visitors to volunteer to join the campaign, to subscribe to a mailing list, or to contribute money to the campaign; "Dole Interactive" consisted of trivia games, "On the Campaign Trail" reported on the candidate’s travels; and "News Room" dispensed speeches, press releases, policy memos, (Tedesco, Miller, & Spiker, 1999, pp. 57-58). This web page also included a recipe for pecan roll cookies and allowed visitors to calculate the tax cut they would receive if the Republican ticket wins the election.

The Clinton/Gore web page opened with eighteen links "ranging from a welcome message from Vice President Gore to a Volunteer of the Week link" and "The Presidential Debates, The Briefing Room, The Electoral College Computer, Bill Clinton and Al Gore, What You Can Do, 21st Century Express, the CG’96 Channel, and Search this Web site" (Tedesco, Miller, & Spiker, 1999, p. 59, emphasis original). The site emphasized information about issues, including links to the text of the presidential debates (with rebuttals to attacks made by Dole), briefing memos, and news releases. Text and audio versions of speeches by Bill Clinton, Hillary Rodham Clinton, Al Gore, and Tipper Gore were available. The web site included statements from Clinton supporters as well. The web page followed Clinton’s train journey to the Democratic National Convention. It allowed visitors to download a "grassroots action kit," to volunteer, to subscribe to a mailing list. A "Volunteer of the Week" was highlighted. The CG’96 Channel "delivered the latest news from the campaign in audio, video, and graphic files directly to desktop users" (Tedesco, Miller, & Spiker, 1999, p. 61).

Thus, campaign web sites, which are increasingly becoming de rigeur, have several elements in common. Most have biographical information; photographs of the candidate, often with family and/or supporters; issue information; speeches and press releases; and opportunities to volunteer, contribute money, and sign up for e-mail newsletters. Some had audio or video files and links to other web sites. Scholars have not examined candidate web pages, though, for web design features. Nor do we know whether presidential pre-primary sites in 1999 are following this lead.