An Inside View of Congressional Campaigning on the Web

Abstract

This paper offers an insider perspective of United States Congressional campaigning by exploring political marketing on the web. We offer theoretical frameworks that predict how campaigns view their websites (e.g., perceptions of likely audiences), how campaigns use their websites (e.g., content posted), and how these views and usages have evolved (or not) over time. We test our predictions with a unique data set from surveys of political marketers involved with the creation and maintenance of congressional campaign websites between 2008 and 2014. Consistent with our expectations, we find uniform views across campaigns about perceived website users (e.g., engaged voters). However, we also find support for our expectation of fundamental differences—between incumbent and non-incumbent campaigns—in what is posted on campaign websites. We also find some, but not many, changes in website usage over time. We conclude that differential marketing motivations result in campaigns that depart from the normative ideal of engaged dialogues that facilitate representation.

KEYWORDS:

• congress
• political campaigns
• staff
• web

Notes

Our work follows others who have conducted similar surveys (e.g., Stromer-Galley et al. 2003; Foot and Schneider 2006) or interviews (e.g., Kreiss 2012); however, we present a more updated and larger data set with a distinct focus.

Our focus is variation across candidates, and thus it may be that in more competitive races, there are generally more frequent visits by all potential audiences but the relative proportion of visits by audience will remain the same.

As mentioned above, this does not mean that the inherent capacities of websites free campaigns from rhetorical and political constraints, just that they are more likely to see their websites, as compared to other media, as a better venue for promoting their overall campaign message, especially to a general audience of voters.

Our predictions in this section echo extant work by identifying incumbency-challenger status as a critical determinant of campaign behavior over a range of strategies (Trent, Friedenberg, and Denton 2011, 82–88; Jacobson 2013, 105–113).

In cases in which we could not locate contact information, we would—if available—submit a message directly to the campaign on its website (i.e., an online inquiry).

As noted, we implemented the survey during four distinct election campaigns. Given the anonymity of the respondents, we have no way to know whether any (or how many) respondents completed the survey in multiple years. What we do know, however, is that respondents reported being highly informed about website constructions, as just noted. In addition, while Nyhan and Montgomery (2015) show that consultants tend to work only for one party or another, we are unaware of evidence concerning using the same consultants over multiple years. Also, note that many campaigns relied on volunteers (as reported below) who may be even more likely to rotate. Most important, we phrased our questions in terms of what “the campaign” is doing so our presumption is that respondents, regardless of whether they had previously participated in a survey, answered in ways that reflect the current campaign on which they work. In the end, though, we do recognize this limitation—an unavoidable reality as anonymity was likely critical to ensure responses.

Other potential sources of influence include state parties, national parties, or other politicians, but we did not measure these.

We also asked respondents how often other campaign material included the site’s URL. We find that respondents estimated that an average of 91.26% of other campaign material (e.g., television ads, direct mailers) included the campaign website address. This further suggests that campaigns see their websites as an informational hub and ideal place to present their overall message to voters at large. Campaigns continue to drive traffic to these general sources of campaign information.

This is not to say, however, that campaigns see their websites as static or unoriginal “brochures” (see Foot et al. 2003; Druckman, Kifer, and Parkin 2007). To assess this, we also included measures that gauged how often the websites were updated and assessed the originality of the websites (the latter was asked only on our 2014 survey). We find that the majority of campaigns view their websites as fairly dynamic, updating information every few days (35.73%) or even daily (27.84%). The vast majority (72.62%) of our 2014 respondents also described their campaign websites as more than moderately original or unique (when asked to rank them on a seven-point scale from not at all original to very original the average is 4.34, SD = 1.52, N = 84). All of this suggests that campaigns have a specific way of looking at their websites when it comes to presenting information and communicating with voters. They typically see their websites as a platform for presenting an original and dynamic overview of their campaign message while relying more heavily on social media for direct communication to those who have established a connection with the campaign.

Figure 4 does not include the goal of “collecting data for analysis of campaign goals and strategies” since we only collected data on that goal in 2014. In 2014, the mean value for that goal was 4.49 (SD = 1.92; n = 183).

To compute these values, we set all other independent variables to their mean and reran our models using Clarify.

Logically, we also find that respondents from incumbent campaigns are more likely (82.88%) than those from non-incumbent campaigns (53.38%) to complain about negativity on their opponent’s website.

We also find that race competitiveness is associated with a greater probability of going negative. Here we see that negativity has a 97.7% probability of showing up on candidate websites in toss-up races, compared to a 79.7% probability in leaning races and a 36.0% probability in races solidly favoring one party or the other. Both of our negativity findings follow past research on the content of congressional campaign websites and reconfirm that, unlike candidates who are relatively safe (often including incumbents), those who are coming from behind or find themselves in tight races are more likely to draw contrasts with their opponents (see Druckman, Kifer, and Parkin 2010).