With the advent of graphic browsers for the World Wide Web, users have surged onto the Internet. Jupitermedia's network user address estimates show that the number of U.S. residents online jumped from fewer than 18 million in 1994 to more than 164 million in December 2000 (www.nua.com/surveys/how_many_online/n_america.html). Such rapid change makes the assessment of the Internet's impact on American elections a precarious undertaking.

Unable to afford significant advertising and mostly ignored by the dominant mass media, minor political parties and interest groups became early adopters of the Web in hopes of exposing their candidates and policies to a broader public. The Republicans, Democrats, and mainline interests soon followed, and political analysts and commentators as diverse as Dick Morris (Vote.Com, 1999) and Andrew L. Shapiro (The Control Revolution, 1999) predicted mostly positive democratic changes for election campaigns and for policymaking. Notwithstanding their plausibility, these predictions stood upon weak empirical foundations. Projected largely from anecdotal accounts, they illustrated how selected parties and interest groups had successfully utilized the Internet's communication capabilities to advance their causes. Less optimistic projections, such as the report of the National Telecommunications and Information Administration (Falling Through the Net II, 1998), also remained plausible.

Throughout this period of rapid growth, Bruce Bimber and Richard Davis—separately and now together—have stepped back from particulars and sought instead to assemble and generate representative and reliable data relevant to the assessment of the influence of the Internet on various areas of democratic politics (e.g., Bruce Bimber, Information and American Democracy, 2003; Richard Davis, The Web of Politics, 1999). In Campaigning Online, they continue this effort with a multimethod study that employs historical as well as survey research, content analysis of Web sites, interviews with webmasters and campaign officials, and laboratory experiments in four major cities. They present not only the most comprehensive data currently available regarding how presidential candidates (and statewide candidates in Missouri) used the Web during the 2000 campaign, but also the most thorough analysis of the impact of this usage and its implications for campaign strategy and for democratic theory.

The survey research consisted of five parts: two national preelection telephone surveys of approximately 1,000 respondents; two national postelection telephone surveys of 300, each drawn from previously interviewed preelection Web users; and one statewide survey of 500 Missourians. The laboratory experiments involved the first 210 people who accepted invitations issued randomly by telephone, approximating a goal of 50 for each city. The content analysis focused on the Web sites of George Bush, Ralph Nader, Al Gore, Pat Buchanan, and John Hagelin, as well as those of Missouri's candidates for U.S. senator, governor, and secretary of state. The elite interviews were conducted by telephone or in person with knowledgeable managers or webmasters from all these campaigns (excepting Mel Carnahan's) and with similarly placed individuals who had worked on presidential nomination campaigns in spring 2000 and in statewide races in 1998. A methodological appendix provides details.

The analysis focuses on 1) comparing how candidates present themselves online with how they present themselves through more traditional offline media, and 2) how online campaigning affects “voters' knowledge level, attitudes and behavior” (p. 8). The authors find that online campaigning is producing incremental changes: somewhat disappointing news for those who expected a radically democratic restructuring of electoral politics, but something of a vindication for those who have studied mass electoral behavior over the past 60 years.

At this juncture, the contents of websites suggest—and most webmasters and campaign managers confirm—that the principal aims of the candidates' presence online are to reinforce and mobilize their political bases. These loyal supporters are most likely to turn out and most likely to convey information to friends, relatives, and co-workers about their candidates' personalities, policies, and positions on public issues. Unlike advertising in the mass media, placing information on the Web does not greatly increase the likelihood that messages will reach those who do not actively seek information about the election. Only 9% of adults viewed one or more candidate websites. This compared to 40% who paid at least “a little attention” to the presidential campaign in newspapers and 77% who remembered seeing at least one campaign advertisement on television (pp. 102–3).

Despite their limitations, websites do open new avenues of communication. Journalists and other influential individuals may visit, especially to gain “rapid responses” that answer opponents' charges or react to breaking news that could affect the election result. Donations may be solicited on site, and the accompanying information may be used (with the donor's permission) to expand Internet mailing lists of supporters. Moreover, the laboratory experiments did reveal an issue-based “learning-liking” syndrome. If websites can attract voters who seek information mostly about a candidate's positions on issues, rather than about a candidate's personal traits, the voters are likely to come away feeling more informed and more positive toward the candidate.

Nevertheless, the large amount of political information on the Web is offset by users' increased control over the information they choose to access. If you build a website, who will come? Mostly partisan males with above-average incomes and education: much the same people whose political participation is above average in the real world. The authors conclude that paradoxically, as the traditional mass media cut down on coverage of political news, the users' control of exposure to information on the Web may exacerbate the divide between the politically active and the politically indifferent.

Some of the authors' conclusions hark back to the findings of the pioneering work of Paul F. Lazarsfeld, Bernard Berelson, and Hazel Gaudet (The People's Choice, 1944). Their study, which included exposure to radio during the 1940 presidential campaign, concluded that the campaign's effects were primarily the reinforcement of political predispositions, secondarily the reactivation of previous voting habits, and only thirdly the conversion of partisans. Even though radio and later television gradually changed American political campaigns, they did not remake the playing fields. Major parties and interest groups incorporated these new media into their political repertoire and used them for their own purposes.

Cautioning readers that one cross-sectional study cannot capture the swiftly changing nature of Internet usage, the authors are nonetheless brave enough to theorize the future of campaigning online. In contrast to those of most political commentators, however, their theories have solid empirical foundations, and they can be tested over time.

Aimed at general readers as well as students of politics, Campaigning Online is clearly written and its argument is straightforwardly presented. To my mind it deserves a wide readership. Let me add one quibble, however. The authors—or the publisher—used endnotes instead of an organized list of references, thus inconveniencing anyone who wants to look further into this important area.