(Page 9 of 24) - CAMPAIGNING FOR CONGRESS ON THE WORLD WIDE WEB AND THE IMPLICATIONS FOR STRONG DEMOCRACY authored by Gulati, Girish.

There were 690 home pages that were available for the content analysis. This number

excluded non-working sites and sites where there was any uncertainly that about whether or not
it had been updated since the last week in October. Also excluded were sites that were simply
links to national party sites. Each of the working sites were analyzed for the presence/absence of
various political content that have been analyzed before and for other content that have
implications for the controversy outlined above (Gulati 2004; Klotz 1997; Williams et. al. 2002).
The specific items that were coded are presented in the Appendix.

The presentation of the results begins by describing the percentage of candidates who

campaigned on the Web in 2002 by party, incumbency, gender, and campaign receipts. These
indicators were chosen because they allow us to assess whether the Internet truly has the
potential to be a medium for leveling the playing field and providing new voices with a forum to
communicate their ideas. In addition, the results are presented by a number of constituency
characteristics that measure the level of demand for Internet content: age, income, education,
ethnicity, and competitiveness. This is followed by a description of the findings from the content
analysis.

Web Campaigning in 2002

Web Presence

In 2002, third-party candidates and Independents continued to lag significantly behind the

candidates of the two major parties with regard to taking their campaigns on to the Web (see
table 1). Whereas as 76% of the Republicans and 67% of the Democrats running for the House
had a campaign website in 2002, only 40% of the minor-party candidates had a website in that
same year. For all parties, however, these percentages represented an increase from the previous
election cycle. Among Senate candidates, moreover, the gap between major- and minor-party
candidates actually decreased, although it still remained significantly large. Whereas 85% of the
Republican Senate candidates and 88% of the Democratic Senate candidates had a campaign
website in 2002, 59% of the minor-party candidates campaigned with a website in 2002. These
differences remained statistically significant, moreover, even after controlling for a number of
constituency and electoral factors.

[Table 1 about here]

It might also be important to note that the difference between the proportion of

Democratic and Republican candidates to launch a web campaign was statistically significant. It
is difficult to tell if this is a new phenomenon or represents the continuation of an existing trend
since none of the previous studies of web campaigns separated the major-party candidates. A
study of incumbent House members’ websites before every member of Congress had an official
site indicates, however, also found that Republicans were more likely to have websites than
Democrats. Much of this difference is attributed to Speaker Gingrich’s apparent fascination with

To rule out the possibility of observing a spurious relationship in the bivariate analyses, I estimated a

multivariate logit model of web presence that included chamber, party, incumbency status, gender, constituency age,
constituency education, constituency income, constituency ethnicity, competition, campaign finance receipts, and
constituency ideology as independent variables. References in the text to having controlled for other variables are
based on the estimates from this model.