The Content and Effect of Political Advertising in U.S. Campaigns
Abstract and Keywords
Political advertising, especially negative advertising, is a prominent feature of contemporary political campaigns in the United States. Campaigns use advertising strategically to persuade citizens their candidate is preferable to the alternatives; to mobilize like-minded supporters to get out to the polls to cast a ballot for their candidate; and to acquire citizen-personal information, so they can more effectively target individuals with appropriate persuasive or mobilizing messages. Online advertising is growing, but television advertising volume has largely been on the rise, too, with 2014 being a plateau. Evidence about trends in advertising content and effects of advertising on citizens come from television advertising in particular.
Over the past decade, candidates have consistently sponsored a majority of advertising on the airwaves although their share does appear to be declining in legislative races. Interest group sponsorship of political advertising has grown, especially in Senate and presidential races, taking advantage of recent legal changes in the campaign finance landscape. Negativity is the dominant form of television advertising, constituting more than 65% and as much as 75% of all congressional general election ads (and as much as 87% of presidential ads) on air since 2006. Parties and interest group sponsors are more likely to air negative advertising by candidates, but candidates do not refrain from going negative. In fact, candidate negativity comprises roughly half of all negative ads on air. Negative ads are more likely to cite specific sources and therefore are generally considered more substantive. TV advertising is unlikely to contain partisan or ideological cues, in part, because it is targeted at swing voters.
Early studies of advertising cast doubt on their effectiveness, but more recent work suggests that advertising effects are small (mattering at the margin in the most competitive contests) and often conditional. That is, advertising effects often vary in relation to characteristics of the messages being aired, the individuals who view them, and contextual factors relating to the campaign more generally. Scholarship suggests that advertising has persuasive but short-lived influence on citizens and that advertising volume and negativity may aid mobilization efforts (although the influence of negativity may be conditioned upon ad characteristics and timing).
Technological advances in the way TV advertising is deployed is increasing campaigns ability to target citizens in a fashion similar to online advertising, which has implications for how well researchers can continue to study it. Scholars have made considerable progress in studying 21st-century advertising effects, but a number of logistical hurdles and unanswered research questions remain.
Americans dislike political advertising—especially negative advertising. Most citizens would prefer to encounter fewer advertisements on the airwaves in the course of federal campaigns; few see negative advertisements as useful in aiding voters in learning about candidates, and more than half believe that negative ads cause voters to “stay home” on Election Day. Further, an overwhelming majority (more than 80%) of Americans believe that running no campaign advertisements at all is preferable to running mostly negative advertisements (Brooks, 2000). The public’s disdain for negative advertising could hardly be clearer.1
Nevertheless, contemporary congressional and presidential campaigns are highly negative, and have become more negative in the last decade and a half. For example, purely positive advertisements2 made up only slightly more than 14% of advertisements aired in the most-recent 2012 presidential election (Fowler & Ridout, 2013), the lowest since at least the 2000 contest between George W. Bush and Al Gore. How can researchers make sense of such a stark asymmetry between campaign advertisers’ behavior and voters’ advertising preferences?
No doubt part of the answer lies in the diverse set of empirical evidence documenting that campaign negativity has the ability to increase citizen knowledge about politics and current affairs without necessarily depressing turnout at the polls. Moreover, political advertisements (especially those that are negative) are designed and carefully crafted to be persuasive. They may not be popular, but political ads (and negative advertisements in particular) play an important role in shaping citizens’ electoral decisions. The extent to which advertising persuades, however, is conditional on the characteristics of particular messages, the characteristics of individual voters, and (at times) the combination of the two.
This overview is organized as follows. First, we describe how campaign advertising in the United States3 aids political elites in accomplishing three sets of strategic goals: vote choice effects (altering citizens’ electoral preferences at the ballot box), mobilization (influencing whether or not citizens turn out to vote), and the acquisition of voters’ personal information. Second, we make use of presidential and congressional advertising data from the Wesleyan Media Project (WMP) and the Wisconsin Advertising Project based on Kantar Media/CMAG tracking to describe a variety of advertising trends taking place over the past decade. Third, we turn to a scholarly literature describing political advertising effects—the extent to which advertisements influence citizens’ vote choice, turnout propensities, knowledge about politics, and views about the American political system writ large. We conclude by highlighting avenues for future research in the study of campaign advertising.
Why Advertise? The Strategic Motives of Campaign Advertising
The central goal of campaigning is to persuade the electorate via political messaging to adopt attitudes, and to mobilize them to take actions consistent with political elites’ goals (Hovland, 1959; Lazarsfeld, Berelson, & Guadet, 1944). While political advertising may also have important implications related to what citizens know about public affairs, the impressions citizens have about politicians, and the extent to which they place trust in the American political processes and its institutions, political campaigns first and foremost hope to convince voters to support their candidate at the polls. Consequently, social scientists have—for several decades—made use of experimental methods (Gosnell, 1927; Hovland, Janis, & Kelley, 1954) and panel survey research (Berelson, Lazarsfeld, & McPhee, 1954; Lazarsfeld, Berelson, & Guadet, 1944) to study persuasion in response to campaigning and political advertising.
McGuire (1985) conceptualized persuasion as the extent to which political messages are able to alter citizens’ attitudes, and thereby shape their behavior. For example, a candidate advertisement might be considered effective if it (1) leads individuals to increase their favorability toward the favored candidate and (2) thereby encourages them to show up in support of that candidate at the ballot box on Election Day. One might say that behavioral change, in this theoretical framework, ought to be mediated by attitude change.
However, it is worthwhile to question whether or not attitudinal change must lead to behavioral change in order to be considered to be persuasive. As Valentino and Nardis (2013) point out, real instances of attitude change—whether in terms of directionality (changing viewpoints from one “side” to another) or intensity (embracing a particular point of view with more or less strength)—may not always lend themselves to easily identifiable changes in behavior, and therefore fail to satisfy McGuire’s definition of persuasion. Consider, for example, a particularly uncivil attack ad that makes an individual less enthusiastic about the candidate they initially favored. Exposure to this ad might create more ambivalence toward the candidate, decreasing the likelihood that the individual turns out to vote. If that individual was already planning on staying home on Election Day, despite having a somewhat more negative view of the candidate, it would be difficult to observe behavioral change associated with attitude change (a problem of observational equivalence).
Further, Ridout, and Franz (2011) draw a useful conceptual distinction between direct and indirect forms of persuasion. Direct persuasion, which can be thought about as “vote choice effects,” occurs when citizens update their vote choice in response to exposure to persuasive political communication. Indirect persuasion, which can be thought about as “mobilization effects,” occurs when advertisements alter citizens’ likelihood of heading to the polls on Election Day. Indirect persuasion is distinct from explicit mobilization efforts even if they can have a similar effect. For example, negative political advertisements focused on altering citizens’ vote choice late in campaign cycles may not directly persuade them to switch their electoral allegiances from one candidate to another, but may instead lead them to experience psychological conflict—making them less enthusiastic about their electoral choice, and encouraging them to stay home on Election Day (Krupnikov, 2011).
It is useful to distinguish between campaigns’ persuasive and mobilization advertising efforts because they involve different audiences and therefore different tactics. Traditionally, campaigns tend to target undecided voters primarily through persuasive appeals in television advertising while their mobilization efforts to convince and remind like-minded partisans to show up at the polls tend to be concentrated online, through paid online advertising and uploaded videos intended to be shared and or incorporated into other shareable user-generated content (Fowler, Franz, & Ridout, 2016). (Of course, there are plenty of other persuasion and mobilization efforts that occur outside of the scope of this review of advertising—on the ground canvassing and direct mail to name a few—as well.) The division of primarily persuasive efforts on TV versus primarily mobilization focused efforts online (get-out-the vote campaigns that do not need to persuade strong partisans for whom to vote but rather convince them that showing up is important) makes some sense. Partisans are likely to share mobilizing messages with other like-minded partisans, whereas swing voters can be harder to reach through social media and networks. However, this traditional division between online and TV advertising is disappearing as addressable television (targeting of ads by household rather than by geographic media market) becomes more of a reality, and campaigns work to capitalize on their rich individual-level data in online persuasive efforts as well. Besides mobilization efforts, there is one additional goal of paid online advertising worth noting: Campaigns frequently use it to acquire citizens’ personal information—specifically their email addresses—in an effort to stay in communication with them throughout the campaign, to test messaging and to solicit donations (see Fowler, Franz, & Ridout, 2016).
Although scholars are starting to study online advertising (Broockman & Green, 2014; Ridout et al., 2015), the bulk of what we know about advertising in campaigns comes from studies of television advertising in particular. In what follows, therefore, we focus on what is known about advertising efforts on television and what we know about their effect.
Advertising Trends in the 21st Century
Before reviewing a literature that details a wide range of advertising effects, we first seek to situate our review in what we know about contemporary trends in advertising over the last decade. The data featured here represent all televised political advertising from the 2006 midterms up through the 2014 midterm election cycle aired in markets across the country (2006 is the only year for which we do not have data on all 210 media markets in the country). The analyses presented here are limited to advertisements aired in U.S. House, U.S. Senate, and presidential contests, and consist only of ads aired after September 1 in each year (the traditional start date of the general election). All advertising content was coded by the Wisconsin Advertising Project for 2008 (Goldstein et al., 2011) and the Wesleyan Media Project (2006 and 2010–2014; see Fowler, Franz, & Ridout, 2014, 2015), based on data collected by Kantar Media/CMAG.
Figure 1 plots the sheer volume of advertisements aired across each election cycle between 2006–2014.4 The left-hand panel of Figure 1 displays the total number of ads aired over time in House and Senate races, while the right panel displays presidential advertising from 2008 and 2012.
As displayed in the figure, the total number of post-September airings have generally been on the rise, eclipsing 1.1 million airings in both 2010 and 2012. While 2012 pulverized advertising volume records over the full course of the campaign (see Fowler & Ridout, 2013), the 2010 cycle had slightly higher total airings from September 1 to Election Day.5 By 2014, however, advertisement airings plateaued somewhat. The figure also shows that airings in Senate races have kept pace with airings in House races—even surpassing them in several years (e.g., 2012 and 2014). Scholars noting the decreased competitiveness of House elections may not be particularly surprised by this finding (Abramowitz, Alexander, & Gunning 2006; Cox & Katz, 1996). Even though the number of Senate seats up for election in any given cycle is far smaller than the 435 House seats up for reelection, airings in senatorial contests have kept pace with (or surpassed) House airings in every year in the series.
The right-hand panel performs the same analysis on advertisement airings in the 2008 and 2012 presidential contests. Overall, the number of advertisements aired in presidential contests increased between the two cycles nearly doubled (from nearly 310,000 in 2008 to over 635,000 in 2012).
Next, we display the share of ad airings by each sponsor: candidates; political parties (for ease of presentation, we combine coordinated efforts between parties and candidates with the party-sponsored airings); and interest groups. The upper left-hand panel of Figure 2 displays this proportion for House and Senate races combined; the upper right-hand panel does the same for presidential contests; and the bottom-left and right-hand panels do the same for House and Senate races (respectively).
Two important trends emerge from this figure. First, the share of advertisements aired by candidates is larger than the proportion aired by either parties or interest groups for every year in the series, and for every race in the dataset. Moreover, candidate-sponsored spots constitute a majority of all ads aired in every race in each election year although the proportion appears to be declining in congressional races. Of course, these aggregate trends do not necessarily indicate that interest groups and other sponsors air fewer spots than candidates in all races. For particularly competitive contests, which often garner attention from private donors and other groups across the country (Gimpel, Lee, & Pearson-Merkowitz, 2008), interest group airings have exceeded candidate activity, which occurred in some 2014 Senate contests (see Fowler & Ridout, 2014).
Second, the data point to an interesting divergence between interest group and political party airings. Prior to 2010, party-sponsored spots constituted a sizable proportion of ad airings in both the House and the Senate combined (25% in 2006 and 34% in 2008), but since then that percentage has declined. Interest group sponsorship of ads, by contrast, has increased as a percentage of total ad buys. In 2010, interest group sponsorship rose sharply in House and Senate races to 14% from only 2% of all House and Senate (combined) advertisements aired in 2008, and continued to rise in 2012 and 2014. Further, while interest group sponsorship (27% of ads aired) was already more frequent than party sponsorship (13% of ads aired) in the 2008 presidential contest, the proportion of spots sponsored by interest groups increased considerably (to 31%) in 2012.
Although interest group airings rose in both sets of races, the effect was more pronounced in Senate contests. The sharp rise in interest group airings between 2008 and 2012 was met with a corresponding decrease in party activity in the same time span. In contrast, House contests showed a more modest increase in interest group airings. Overall, then, it should be clear that interest group activity is on the rise, especially in competitive races and when control of the chamber is at stake.
The sponsorship trends we present here are consistent with claims that legal changes in the campaign finance landscape have affected sponsorship patterns in advertising. Specifically, interest group sponsorship increased following the 2010 Supreme Court decision in Citizens United v. The Federal Election Commission, even if the increase may not appear as dramatic in the general election period as some have claimed. The figure shows that interest group airings were comparatively low in relation to party-sponsored airings in 2006 and 2008, and additional research suggests that the same was true in for congressional spots aired in 2002 and 2004 (Fowler, Franz, & Ridout, 2016). The 2000 election featured considerably higher levels of interest group activity in House races (about 18% of ads aired in House races, a percentage comparable to 2014 involvement),6 which is also consistent with claims that BCRA depressed interest group sponsorship following its passage (in 2002) before Citizens reopened the floodgates. Further, it is important to note again that our figures only display the share of sponsorship during the general election, and other work has shown that interest group activity in recent post-Citizens cycles is even more prominent during primaries (see Fowler & Ridout, 2013, 2014).
Trends in Tone
It is nearly impossible to talk about political advertising without also discussing advertising tone and negativity in particular. Negative spots have dominated the airwaves in recent elections comprising 65% of all congressional advertisements aired from September 1 through Election Day in 2006 and 66% of all those aired in 2008. Negativity peaked in 2010 congressional races at 75% of general election airings and receded somewhat in subsequent years (to 68% in 2012 and 70% in 2014), but nevertheless remain very high. Presidential campaigns are even more overwhelmingly negative. Nearly three quarters (73%) of general election airings in the 2008 contest between Barack Obama and John McCain were negative, and that figure grew to nearly 87% in the 2012 contest between Obama and Mitt Romney.
As these data suggest, contemporary congressional and presidential campaigns are predominantly negative, especially in competitive contests (Fowler & Ridout, 2013, 2014). In addition, we also know that the use of negativity differs across sponsors. Figure 3 documents different sponsorship trends in the use of negative and positive advertising, in congressional campaigns. The top panel displays the proportion of all negative and contrast (upper left-hand panel) and positive (upper right-hand panel) airings in each general election cycle accounted for by each type of sponsor. The bottom panel of Figure 3 shows the proportion of each sponsor’s airings that are negative.
Focusing first on the top panels, it is clear that candidates, parties, and interest groups play different roles in contributing to overall levels of campaign negativity and positivity. Whereas candidate-sponsored negativity has consistently comprised about half of all negativity on the airwaves between 2006–2014, party-sponsored negativity was more prominent than interest group-sponsored negativity prior to the 2010 midterms. In 2008, for example, party-sponsored negativity comprised 49% of all advertisements aired in congressional races, while interest group negativity accounted for only about 3%. In sharp contrast, only 26% of negative ads aired in the 2010 midterms were put on the airwaves by parties, whereas 17% were aired by interest groups. For the remainder of elections in the series, parties and interest groups have played roughly equal roles in contributing to overall levels of negativity in congressional campaigns.
Conversely, nearly all positive advertisements aired in congressional general election cycles between 2006–2014 have been candidate-sponsored advertisements. Candidates consistently sponsor more than 85% of all positive ads on the airwaves. However, it is not the case that candidates avoid going on the attack. As the bottom panel of Figure 3 demonstrates, candidate airings in congressional elections between 2006–2014 tend to be more or less balanced in terms of how many negative messages they air (hovering around the 50% mark). Negative messages comprised more than 60% of all candidate advertisements aired in only one election cycle in the series: the highly negative 2010 midterms, which may in part have been driven by slow economic recovery. In contrast, the overwhelming majority of interest group and party advertisements aired are negative. More than 90% of party-sponsored ads were negative in every election cycle displayed here, and interest group airings ranged from a low of 75% negative (in 2006) to a high of 93% negative in 2010.
What these findings suggest is that, while parties and interest groups play prominent roles in contributing to overall levels of advertising in congressional campaigns, the lion’s share of advertisements that they air are negative. Consequently, parties, and interest groups account for roughly half of all negativity in congressional races on the airwaves in most election cycles, but never account for more than a sixth of positive advertisements aired (at best).
Why the discrepancy in the deployment of negativity? As discussed earlier, voters dislike negative advertising and may punish sponsors of negativity. However, a growing line of research suggests that sponsors do not experience this “backlash” equally. In particular, interest groups are not on the ballot and the paid for byline frequently says explicitly that they are not connected to a candidate, which decreases the likelihood that their favored candidate will be punished for the negativity they air (Brooks & Murov, 2012; Dowling & Wichowsky, 2013, 2015). Candidates may therefore have an incentive to allow interest groups to engage in negative campaigning on their behalf—they gain the benefit of the attack without the penalty associated with the backlash. Consistent with this view, Figure 3 demonstrates that interest group-sponsored negativity has become increasingly common over time in congressional contests (the presidential airings data reveal a similar trend).7 Of course, these data cannot discern whether or not this trend is demonstrative of candidates’ conscious efforts to let others “do the attacking,” and the figure also makes clear that candidates themselves do not shy away from negativity—even if they employ it less often than their party and interest group counterparts.
The Substance of Advertising
Next, we attempt to extend a few important insights from Geer’s 2006 book In Defense of Negativity. Geer argues that negative advertisements are “asymmetrically constrained” to focus on issues relevant to voters compared to positive advertisements, in part because they make claims requiring external information sources to back up them up. As Mattes and Redlawsk (2014) argue, despite disliking negative campaigning in the abstract, voters are unlikely to punish sponsors of negativity when specific negative ads focus primarily on policy issues and may even report viewing these types of advertisements as “helpful” in decision-making.
Figure 4 plots the proportion of negative (left-hand panel) and positive (right-hand panel) advertisements that “cite” at least one outside source (e.g., a newspaper, website, piece of legislation), as well as the proportion that focus exclusively on policy issues (and not candidates’ personal characteristics or some combination of the two). Because data limitations prevent us from analyzing citation trends in 2008, we carried this analysis out for congressional elections only: extending the trend line across the missing data point in the process for the “source documentation” line.
As shown in the figure, the proportion of advertisements focusing exclusively on policy issues is similar across positive and negative advertisements. However, as expected, that proportion is higher for negative advertisements, for every year in the series. On the other hand, there are stark differences in the extent to which positive and negative advertisements “cite” external sources. More than half of negative advertisements aired in each election in the series (again, excluding 2008 due to data constraints) document at least one source, while less than a quarter of positive advertisements do the same (with 2014 posing the only exception: where 30% of positive advertisements cited an external source), supporting Geer’s “asymmetric constraint” hypothesis.
Partisan and Ideological Cues
We conclude our content analysis by drawing attention to a rapidly growing literature in political science. Recently, scholars have drawn attention to the premise that the mass public’s evaluations of Democrats and Republicans have become polarized on the basis of affect. One mechanism by which affective polarization is thought to occur is through campaign messaging (of which political advertising is one form). Candidates for public office from one political party denigrate and attack members of the other in widely aired and expressly partisan messages, encouraging partisans to dislike members of the other party on a highly personal level (Iyengar, Sood, & Lelkes, 2012; Iyengar & Westwood, 2014). While there is strong reason to believe that partisans’ affect toward opposing partisans has certainly become polarized along affective lines in recent years, our content analyses cast some doubt on the possibility that negative campaigning is the mechanism by which these effects occur.
Figure 5 plots the proportion of all general election congressional advertisements that make mention of candidates’ political parties and ideological affiliations. The left-hand side of the figure displays the proportion of advertisements aired that make mention of the favored candidate’s party (either the sponsor, or the promoted candidate in a party or interest group spot), the opposing candidate’s party (the “target” of negativity in attack or contrast ads), and the proportion of ads making no mention of party at all.8 The right-hand side captures whether or not the phrases “liberal” and “conservative” were mentioned at all in the content of all advertisements aired (WMP did not code for whether this word was mentioned specifically to describe the favored or targeted candidate). As was the case in Figure 4, data limitations prevent us from conducting these analyses on liberal and conservative mentions in 2008—so we limit our analyses to congressional races only, and extend the time series line through the missing 2008 data point.
If out-party denigration in political campaigning leads to affective polarization, we might expect to see that a large number of political ad airings make mention of opposing candidates’ political parties. In contrast, for every congressional election cycle in the series, the proportion of advertisements mentioning neither the favored nor opposing candidate’s party vastly outnumber those that do. The proportion of advertisements making no mention of party whatsoever ranges from a low of 83% in 2006 to a high of 95% in 2008. Further, for almost every year in the series (with the exception of 2012), the proportion of favored candidate party mentions was at least as high or higher than the proportion opposing candidate party mentions.9 Mentions of ideology tell a similar story. Since the Wesleyan Media Project does not code for whether or not individual candidates are labeled as liberals or conservatives, our analyses are likely to overreport the amount of ideological candidate labeling in ads. Still, the proportion of ads mentioning neither liberals nor conservatives never dips below 90% of all ads aired in the (four) election cycles for which data was available. Mentions of conservatism were highest in 2012 (6% of all ads aired), whereas mentions of liberalism were highest in 2006 (again, 6% of all ads aired). Whatever concerns may be raised about advertising, it is clear that they do not emphasize partisan or ideological loyalties.
Research on Political Advertising Effects
Thus far, we have described a number of trends relating to the volume and content of advertisements aired by different groups of political players. In what follows, we summarize scholarly knowledge about advertising’s influence. Before describing contemporary evidence, however, we first briefly review how media effects have been conceptualized and studied in the past. Early theorizing following concern over WWII events posited a “hypodermic needle” model of media influence—that citizens would uncritically accept the positions put forward by political and media elites in persuasive messages (see Shaw, 1979, for a review). Yet empirical results from early panel studies revealed that citizens’ electoral preferences generally changed only slightly over the course election cycles, leading scholars to conclude that advertising had only minimal effects (Lazarsfeld, Berelson, & Guadet, 1944). Concurrently, scholars studying media effects in experimental settings also noted that few citizens seemed to update their attitudes in response to persuasive messaging (Hovland, 1959; Hovland et al., 1949), although they sometimes found nontrivial media effects (Hovland, Janis, & Kelly, 1954).
Early behaviorist research and the advent of the Michigan Model offered a potential explanation for the minimal effects thesis. The mass public is generally uninterested and unknowledgeable about most political matters (Campbell et al., 1960; Converse, 1964), and is therefore unlikely to pay attention to (or make sense of) persuasive political communication. To the extent that anything was thought to influence citizens’ vote choice preferences, the Michigan School argued it was citizens’ unwavering psychological attachments to partisan labels that best explained electoral preferences. More recently, Finkel (1993) offered contemporary support for the so-called minimal effects thesis—showing that citizens’ partisan identities play a key role in shaping vote choice. To the extent that citizens deviate from simply voting for candidates who share their partisan sympathies, most of the variation in citizens’ electoral preferences occurs early in election cycles: before intensive campaign efforts following partisan conventions in presidential election cycles (see also Gelman & King, 1993). Similarly, Holbrook (1994) noted that campaign events (e.g., debates, nominating conventions) in presidential contests have some ability to explain individual voters’ electoral preferences, but found that changes in evaluations of the macroeconomy played a much stronger predictive role.
Despite some level of contemporary support for the minimal effects thesis, it has nevertheless been successfully challenged on many fronts. For one, studies demonstrated that even if media did not always change minds, it could set the agenda by altering the mix of issue considerations that citizens brought with them to the ballot box (Iyengar, Peters, & Kinder, 1982; Krosnick & Kinder, 1990) Further, after correcting for random measurement error in variables attempting to measure citizens’ exposure to political messaging, media exposure has been shown to play a significantly larger role in shaping voters’ issue positions and attitudes toward presidential candidates (Bartels, 1993). As will become evident shortly, many studies of political advertising have also offered reason to doubt the minimal effects hypothesis.
Importantly, all of the studies mentioned thus far were primarily concerned with searching for (or debunking) what we might call main media effects. Advertisements and the mass media have the ability to persuade “in the main” when the mass public adopts new attitudes or behavioral tendencies in response to exposure to messaging (as opposed to not being exposed). These effects are difficult to uncover in part because in competitive contexts, campaign messages tend to be evenly balanced, canceling out the potential persuasive influence of each side (Zaller, 1996). Moderated, or conditional media effects, are sometimes much larger than main media effects, and can offer an empirical rejection of the minimal effects thesis for (1) messages with a particular set of characteristics not shared by other messages (e.g., Druckman, Peterson, & Slothuus, 2013; Mutz, 2007) and (2) certain subgroups in the electorate (e.g., Huber & Arceneaux, 2007; Krosnick & Kinder, 1990; Zaller, 1992). For example, with regard to message characteristics, when messaging suggests intense elite conflict on an issue, partisans are more likely to rely on party cues relative to messages absent the conflict cues (Druckman, Peterson, & Slothuus, 2013).
With respect to the conditional effects on subgroups of the public, moderates on the sophistication spectrum may be most influenced by media messages, as they pay enough attention to be exposed to these messages, but do not have predispositions that are as strong as those as political sophisticates (Zaller, 1992). Moreover, we also know that message characteristics may interact and be moderated by individual-level characteristics: such as individuals’ personality traits and dispositions (e.g., Gerber et al., 2013; Kam & Simas, 2010), cognitive styles (e.g., Kruglanski, Webster, & Klem, 1993; Rahn, 1993;), partisanship and ambivalence toward party labels (e.g., Lavine, Johnston, & Steenbergen, 2012), and levels of political sophistication (e.g., Zaller, 1992).
Overall, this brief historical overview suggests that persuasive communication has the ability to alter the saliency of some considerations at the expense of others in general, and persuasive effects might be more pronounced among certain subsets of the mass public, or for certain types of messages. Keeping these two points in mind, we now turn to a discussion of the extent to which political advertisements are thought to influence citizens’ vote choice.
Vote Choice Effects
Exposure to political advertising, in general, has been shown to play a role in shaping citizens’ electoral choices, as well as their attitudes toward candidates running for office. At the individual level, scholars have noted that individuals living in media markets where one candidate airs more advertisements than another are ceteris paribus more likely to report voting for (or planning to vote for) the candidate with an advertising advantage (Franz & Ridout 2007; Fridkin & Kenney, 2004; Ridout & Franz, 2011). Ridout and Franz (2011) find, for example, that in the 2000 election a half standard deviation increase in the share of pro-Gore advertisements, when compared to a half standard deviation increase pro-Bush advertisements, resulted in more than a 30 percentage point increase in the likelihood of voting for Gore (from a 30% to over a 60% chance of voting for Gore). These effects were almost perfectly reversed given an identical increase in the share of pro-Bush advertisements, relative to Gore ad exposure. Furthermore, in the aggregate, others have noted that—in the most recent presidential election—tracking poll advantages tended to closely mirror short-term advertisement share advantages (Sides & Vavreck, 2014).
These studies all offer support for the notion that political advertisements can persuade individuals to update their electoral selections. Nevertheless, it could be the case that increases in political advertising are accompanied by other alterations in campaign strategy, such as more intensive efforts designed to recruit voters “on the ground.” How can researchers determine that advertisements alone cause citizens to update their electoral preferences?
Huber and Arceneaux (2007) offer one such approach, by taking advantage of the fact that media markets occasionally cross state boundaries. Their quasi-experimental approach shows that intense advertising campaigns in battleground states (where “ground game” efforts are likely to be quite strong) produced spillover effects in non-battleground states (where “ground game” tactics are presumably weaker) that share a media market. Controlling for citizens’ partisan and previously formed electoral commitments in both panel and cross-sectional surveys, a one standard deviation increase in pro-Bush advertising exposure increased the likelihood that citizens cast their ballot for him by more than 9% (and decreased the likelihood that they would do the same for Gore by about 8%). Gerber and colleagues (2011) offer another solution to the aforementioned causality issue. These researchers teamed up with Rick Perry’s gubernatorial campaign to create and air an advertisement on his behalf—varying the intensity of advertisement exposure across a number of different media markets, and tracking public opinion simultaneously. Consistent with the idea that advertising advantages produce short-term persuasive effects (Hill et al., 2013; Sides & Vavreck, 2014), they found that increased exposure meaningfully improved Perry’s (predicted) vote share, but for no longer than seven days. After one week, the effects largely dissipated.
Overall, this research suggests that exposure to political advertising in general can yield persuasive effects. But what about conditional effects? Might some characteristics of advertisements allow them to be better (or worse) at persuading voters than others?
One element of ad content that can enhance persuasive effectiveness is advertising tone. Negative advertisements are particularly well-suited to encourage individuals to alter their electoral preferences. In both presidential and senatorial contests, increased exposure to campaign negativity from one side usually rewards the side doing the attacking (Ridout & Franz, 2011), although—at least in Senate races—it may be the case that challengers benefit more than incumbents (Lau & Pomper, 2004). Negative ads have also been shown to be effective at altering citizens’ affective impressions about candidates, such that attack and contrast ads encourage citizens—especially those who pay little attention to politics—to view targeted candidates more negatively (Fridkin & Kenney, 2004). Though recent meta-analyses have not found consistent evidence that increased negativity results in behavioral electoral outcomes that are beneficial to attackers (Lau et al., 1999; Lau & Pomper, 2004; Lau, Sigelman, & Rovner, 2007), they do suggest that more negativity is associated with more negative affective impressions of targeted candidates, which is a first step in behavioral change.
Under certain conditions, however, negative advertisements may not always produce the effects that campaign strategists intend. The “backlash effect” occurs when voters punish candidates at the polls (or in the ballot box) for airing negative advertisements (Brooks & Geer 2007; Kaid, 1997; Mattes & Redlawsk, 2014; Pinkleton, 1997). Overall, Ridout and Franz (2011) find that the impact of negativity on vote choice tends to more closely resemble strategists’ “intended effects.” Consistent with this view, Mattes and Redlawsk (2014) note that voters generally view most negative information (the notable exception being lies and defamation, see Fridkin & Kenney, 2004) as quite helpful in making political decisions, particularly when negative campaigning is referred to without expressly using the term “negative.” In contrast to other work, they find little evidence of backlash toward the sponsors airing negative messages in the laboratory environment.
Other work, however, suggests that some sponsors of negativity may be more likely to experience backlash than others. Though negativity aired by both incumbents and challengers in Senate contests has been shown to decrease affective evaluations of targeted candidates, incumbents—but not challengers—may receive an electoral punishment for doing so (Fridkin & Kenney, 2004; Lau & Pomper, 2004). In the laboratory environment, negative advertisements sponsored by candidates tend to result in more backlash (negative affect toward sponsors) than unfamiliar interest groups (Brooks & Murov, 2012; Dowling & Wichowsky, 2013) and—to some extent—political parties (Dowling & Wichowsky, 2015).
Therefore, candidates may benefit from letting unfamiliar interest groups “do their dirty work” (Dowling & Wichowsky, 2015; Fowler & Ridout, 2010) as the extent to which negative advertisements persuade has been shown to be mediated by perceptions of credibility in the sources of sponsorship (Weber, Dunaway, & Johnson, 2012). That is to say that citizens use the sponsors of advertisements as “source cues” (see also Eagly & Chaiken, 1993; Lupia, 1994)—cognitive shortcuts that aid them in making inferences concerning whether or not they view the claims made in the advertisements as credible. Because citizens are more trusting of groups whom they know little about than they are of candidates or parties, candidates may have an incentive to avoid “going negative” themselves, and let increasingly active, yet unknown to the public, interest groups take that action on their behalf (Ridout, Franz, & Fowler, 2015).
In addition to advertising tone, the emotional appeals featured in political advertisements have been shown to have strong effects with respect to vote choice. Consistent with Affective Intelligence Theory (Markus, Neuman, & MacKuen, 2000), advertisements that appeal to fear (which are almost always negatively toned) can lead individuals to experience anxiety—which has been shown to encourage them to reconsider their prior beliefs and political affiliations (Brader, 2005, 2006). Those experiencing anxiety in the context of negative campaigning are more likely to consider supporting candidates who offer a remedy to a potentially dangerous status quo than those who are exposed to negativity without such emotional appeals.
Political advertising expenditures are often the single-largest expense that contemporary federal political campaigns must account for in their budgets (Fowler, Franz, & Ridout, 2016; West, 2009), and while the academic literature has not always agreed with political strategists’ emphasis on messaging, there is plenty of evidence to suggest that advertising can matter, even if its effects are not always large. Advertising exposure in general has been shown to play a key role in shaping individual vote choice intentions. The tone of advertisements, as well as the emotional appeals they make, can also shape vote choice. However, some of the effects that negativity and emotional appeals exert on citizens’ political behavior are more closely tied to whether or not citizens show up to vote in the first place. We take up those issues in turn.
In addition to persuading citizens to change their vote choices, political advertisements also influence whether or not citizens turn out to the polls. At the level of individual voters, frequent television viewers residing in media markets with high levels of advertising across states are more likely to turn out to vote on Election Day (Freedman, Franz, & Goldstein, 2004) and report that they intend to turn out (Hillygus, 2005). However, in the aggregate, validated turnout levels do not seem to vary across levels of advertisement exposure after accounting for differences in campaigning tactics between and within state boundaries (Krasno & Green, 2008). While these differences pose an interesting empirical puzzle, it is important to point out that neither account suggests that overall levels of advertising exposure decrease turnout.
The same cannot be said, however, about negative advertising. The premise that exposure to campaign negativity may decrease citizens’ willingness to turn out to the polls on Election Day has been a rich source of scholarly research over the past two decades. Early studies of the link between negativity and turnout found that—in both the experimental environment and in the aggregate—exposure to attack advertising was associated with lower levels of turnout as well as the propensity to cast ballots on Election Day (Ansolabehere et al., 1994; Ansolabehere & Iyengar, 1995). As Geer (2006) notes, analyses of presidential campaign advertisements dating back to the early 1960s offer limited bivariate support for this hypothesis: more negative campaign cycles did tend to correlate with lower levels of turnout. Many studies have challenged the demobilizing effects of negativity on the grounds that they either increase (e.g., Finkel & Geer, 1998; Franz & Ridout, 2007; Goldstein & Freedman, 2002) or fail to decrease (e.g., Garramone et al., 1990; Green & Krasno, 2008; Wattenberg & Brians, 1999) voter turnout.
Finkel and Geer (1998), for instance, found that while aggregate turnout levels and campaign positivity have decreased over time, they have not always done so simultaneously. For example, the highly negative 1980 presidential campaign actually boasted somewhat higher levels of turnout overall than the more positive 1976 contest. The duo further used individual-level survey data to model voter turnout intentions in relation to negative advertising exposure. They found that increased exposure to campaign negativity between 1960 and 1992 did not depress turnout intentions for most voters, though it slightly decreased it for political independents (in certain models). Franz and Ridout (2007) expanded on this individual-level approach by matching respondents’ self-reported television viewership habits (i.e., whether or not they tune into particular programs at a certain time) to advertisement frequency and content data (for methodological background and validation, see Freedman & Goldstein, 1999; Ridout et al., 2004) in order to estimate citizens’ exposure to negative advertising. Overall, they find that exposure to negativity not only fails to depress turnout, but may actually slightly increase it—a finding consistent with broader meta-analytical assessments of the field (Lau, Sigelman, & Rovner, 2007), and earlier investigations using similar methodological approaches (Goldstein & Freedman, 2002). In the aggregate, Krasno and Green (2008) too do not find evidence that exposure to negativity depresses turnout, although they also do not find evidence of a mobilization effect in response to negative campaigning in the aggregate.10
It may also be the case that the link between negativity and turnout is moderated by a host of factors pertaining to advertisement content. For example, Kahn and Kenney (1999) find that campaign civility in senatorial campaigns (whether or not negativity focuses on “legitimate” political issues or “inappropriate” and derisive personal characteristics of candidates) can have both mobilizing and demobilizing effects on turnout. Survey analyses of voters’ turnout intentions show that civil negativity can boost electoral participation, whereas uncivil attacks depress it. Brooks and Geer (2007) offer a potential individual-level mechanism by which such effects may occur: as voters tend to dislike uncivil ads, and view their content as less credible than more civil forms of negativity. However, that same experimental trial also showed that incivility in general did not depress voters’ intentions to turn out. The important takeaway from this research is that different forms of advertisement negativity can play an important role in influencing whether or not citizens turn out to vote.
Another element of advertising content that may influence turnout is the extent to which advertisements appeal to some emotions and not others. Exposure to negative advertisements that also attempt to instill anxiety in those watching them, as well as positive advertisements that attempt to do the same with enthusiastic emotional content, is associated with higher levels of reported turnout in experimental studies (Brader, 2005, 2006).
Further, the most recent work on the subject suggests that the timing of negative campaigning may influence whether or not citizens turn out to vote. When negative advertisements aired late in campaigns conflict with citizens’ prior electoral selections, citizens are thought to experience vote choice ambivalence (conflicting negative and positive evaluations about the candidate for whom they wish to vote), and are subsequently less likely to turn out to vote (Krupnikov, 2011). In the 2004 presidential election, for instance, Krupnikov (2011) found that a 60% increase in exposure to negative advertising targeted at voters’ preferred candidates was associated with nearly a 15% reduction in the likelihood that they would turn out to vote. Overall negativity throughout the campaign, however, had no relationship to voter turnout. Conceivably, this could be the case because negativity aired earlier on in campaigns—when voters are still making up their minds about for whom they should cast their vote—does not conflict with their electoral choices.
Advertisements also influence voter turnout when news media outlets pick up and amplify certain advertisements in their coverage of the campaign. Studies suggest that news coverage of advertising is common but also that it much more likely to focus on negative ads or new tactics (Fowler & Ridout, 2009; Ridout & Smith, 2008), which could increase and reinforce the persuasive benefits of the original ad even when the ad is mentioned in a critical way (Pfau & Louden, 1994), but may also increase the risk of backlash for candidates (though see Mattes & Redlawsk, 2014, for a counter perspective on backlash).
Overall, advertising exposure in general may boost turnout levels (or, at the very least, not decrease them), and exposure to campaign negativity in particular has the potential to mobilize as well although these effects may be conditioned upon ad characteristics and the timing of citizen exposure to criticisms. News media amplification of advertising may further boost persuasion and may also drive up citizen perceptions of campaigns being very negative, especially when the coverage is focused on negativity as a strategic tactic (Ridout & Fowler, 2012).
The Downstream Consequences of Political Advertising
In addition to altering vote choice and turnout, exposure to political advertising (both positive and negative) has been shown to have a wide range of consequences pertaining to citizens’ knowledge about electoral politics and views of the political system. Dealing first with knowledge, increased exposure to advertising in general has been shown to boost citizens’ abilities to recall candidates’ names, to correctly guess where candidates stand on highly salient issues, and to increase interest in politics (Freedman, Franz, & Goldstein, 2004; Ridout et al., 2004). Further, as campaigning efforts become more intense, citizens appear to make use of information about political issues, the performance of the incumbent president, and impressions about the economy when formulating attitudes about presidential (Bartels, 1993) and Senate candidates (Kahn & Kenney, 1997).11 In general, then, many scholars have found that intense campaign advertising efforts, and exposure to advertising more generally, can be particularly useful in boosting knowledge about politics.
Consistent with the idea that political advertisers expertly craft spots in order to be informative to even the most unengaged voters—perhaps even serving as information supplements for those who need it most (Freedman, Franz, & Goldstein, 2004)—early studies of learning effects in response to political advertising exposure found that politically inattentive individuals learned more from advertisement exposure than those who are more politically aware (Patterson & McClure, 1976). More recent analyses, though, have casted doubt on this optimistic view of campaign learning effects. In both panel survey (Freedman, Franz, & Goldstein, 2004) and experimental (Valentino, Hutchings, & Williams, 2004) studies, scholars have not consistently found that low-information voters learn more from increased exposure to campaign advertisements than those who already hold high levels of knowledge about politics and current affairs. This is not to say that low-information voters fail to learn from campaign advertisements altogether, as both of the aforementioned studies do find sizable learning effects for even the least attentive citizens. However, despite already possessing a wealth of information about politics and current events, high-information voters tend to learn from campaign advertisements in such a way that is not statistically discernible from that of less attentive voters.
The extent to which citizens learn from advertising is shaped by certain ad content features, such as tone and the emotional appeals they make. Meta-analyses of the negative advertising literature (Lau, Sigelman, & Rovner, 2007) have shown that increased exposure to negative advertisements increases the extent to which citizens remember the information to which they were exposed (e.g., Brader, 2005; Geer & Geer, 2003; Lang, 1991) and boosts knowledge about political campaigns (e.g., Craig, Kane, & Gainous, 2005; Stevens, 2005). This may be due to the idea that humans are more attentive to negatively valenced information than positively valenced information (Lang, Dhillon, & Dong, 1995; Soroka & McAdams, 2015). Consequently, negative advertisements are more likely to feature appeals to politicians’ stances on issues that are highly salient in public discourse, and are more likely to provide documentation of the sources pertaining to the claims that they make (Geer, 2006; see also Figure 4 in this chapter). When negative advertisements do this, citizens tend to view negativity as being helpful in making political decisions. Perceived helpfulness increases the extent to which citizens view negative ads as believable and appropriate (and decreases the extent to which they are angered by negativity),12 which then play a role in shaping vote choice (Mattes & Redlawsk, 2014). However, this is not to say that positive ads do not have the potential to play a role in shaping citizens’ abilities to learn about politics. Because negative ads are more likely than positive ads to include discussions of salient political issues, it could be the case that positive advertisements that feature as much information as negative spots are similarly able to foster learning effects (see the argument and experimental evidence in Stevens, 2012).
Psychologically, Affective Intelligence Theory offers additional insight into why exposure to negatively toned advertising may be more memorable and informative than other forms of campaigning. Negative messages, especially those that also make emotional appeals to fear, can be thought about as stimulating the extent to which citizens search for additional information about a threatening stimulus (e.g., a politician’s actions in office, or issue position), pay attention to information that runs contrary to their prior beliefs, and render evaluations of candidates that eschew prior partisan commitments and focus instead on issue stances (Brader, 2005, 2006; Markus, Neuman, & MacKuen, 2000). When evaluating hypothetical candidates in the laboratory environment, Brader (2005, 2006) shows that exposure to negative advertisements that also featured appeals to fear (relative to those that do not make appeals to fear) may stimulate the willingness to search for more information, and (slightly) improve the recall of relevant information in the ad.
However, the emotional content of political advertising is not without potential harmful effects on public opinion and policy outcomes. For example, citizens’ negative racial attitudes are thought to be recalled into memory more readily when individuals experience anger (Banks & Valentino, 2012; Banks, 2015). Consequently, the presence of angry emotional cues in advertisements that make subtle (implicit) mentions of race can encourage citizens to rely on their negative attitudes toward blacks when evaluating political issues and candidates for office (Banks, 2015; Banks & Bell, 2013). In general, angry emotional cues are quite common in political advertising—present in nearly 75% of advertisements aired in the 2012 general election (Fowler & Ridout, 2013). Interestingly, it is important to point out that exposure to negative advertisements in general (i.e., irrespective of the emotional appeals that they might make) does not necessarily make respondents experience anger (Mattes & Redlawsk, 2014). Though individuals are slightly more likely to feel angry after viewing negative ads than positive ads, only those who are exposed to particularly defamatory attacks experience high levels of anger.
In addition to concerns that negative advertising exposure may depress citizens’ willingness to turn out to vote, scholars have also considered whether or not campaign negativity affects political efficacy, trust in government, and views of the system more broadly with mixed results. In an experiment, Brooks and Geer (2007) note that while citizens view trait-based negative attacks as “unfair,” incivility does not decrease perceptions of political efficacy and may even increase the extent to which citizens are interested in politics and turn out to vote. Other research supports the view that negativity does not harmfully affect efficacy and trust (Geer, 2006; Jackson, Mondak, & Huckfeldt, 2008). In their meta-analysis of the relationship between negative advertising and feelings of political efficacy, Lau and colleagues (2007) noted small, yet consistently negative, relationships between exposure to negative advertising and feelings of trust in government, affect toward the government, and perceptions of efficacy (e.g., Stevens, 2002; Thorson et al., 2000). Notably, although the literature is divided, the evidence supporting a more harmful view of negativity in U.S. elections does not suggest that the effects are large.
“Microtargeting” and the Future of Political Advertising
Campaigns have long sought to tailor their messaging to appeal to particular audiences, a tactic called microtargeting. Traditionally, this strategy was most easily deployed through campaign mailers and direct contact with voters, but there is some evidence to suggest that campaigns employed very crude advertising targeting strategies early on. In the 1980s, for example, candidates could geotarget their ads based on the electoral histories of individuals living in particular districts (for whom past electoral outcomes could be easily ascertained), but—due to uncertainty about individual voters’ preferences—were generally limited to airing “get out the vote” messages to individual voters (Hillygus & Shields, 2008). However, technological developments—both in video production and in campaign data—have made the targeting of advertising easier. The proliferation of microtargeting strategies and its application to television targeting represents an important development in campaigning.
Franz (2013) notes that while overall levels of television viewership are on the rise in the 21st century, the amount of televised political information to which citizens are exposed has declined (see also Prior, 2007). Airing a single set of advertisements on a handful of networks is therefore likely to miss out on targeting potentially persuadable voters. Thus, campaigns are adopting microtargeting strategies that allow them to identify the television viewing habits of potentially persuadable voters, and efficiently tailor advertisement content and airing strategies to their unique habits and preferences (Ridout et al., 2012). As Franz (2013) details, campaigns can do this by conducting large surveys, asking voters to describe their media usage habits (as well as their political preferences and other demographic information), and match responses to voter registration lists. They can then make use of publicly available information (or purchased data pertaining to those individuals’ consumer histories) in order to develop media viewership and product consumption profiles of undecided and otherwise-persuadable voters (e.g., political independents). This information might be useful to campaigns in selecting the television channels and other media outlets on which they opt to air their messages, and gives campaigns the opportunity to market different messages to different subsets of the voting public. Campaigns can also use this information to inform programmatic advertising strategies—the automated process by which they bid on opportunities to advertise to select audiences on certain Internet platforms (Fowler, Franz, & Ridout, 2016). For example, campaigns can bid on the opportunity to display content to viewers in a particular location, using the Internet for a particular purpose (e.g., a search engine query for a candidate’s name), and/or meeting particular demographic traits (e.g., all web users under a certain age threshold).
However, the ability to engage in the type of data gathering and management described by Franz (2013) above, and assess the likelihood that any individual is persuadable, changed this process dramatically. As George W. Bush strategist Sara Taylor put it, “We could identify who exactly should be mailed, on what issues, and who should be ignored completely” (quote from: Hillygus & Shields, 2008). Consistent with this, Hillygus and Shields (2008) focusing on the 2000 and 2004 presidential campaigns found that a quarter of party and candidate-funded mailers focused on “wedge issues,” with 8% focusing on highly divisive moral wedge issues—whereas only 1% of television advertisements did the same (on both fronts). Campaigns thus seem to use highly contentious issues to narrowly target voters directly, with a special emphasis on reaching out to weak partisans and independents. However, in their more general forms of communication with the electorate, campaigns prefer focus on more consensual policy issues. This strategy avoid the potential pitfalls (e.g., a decreased likelihood for supporting the sponsor of an advertisement) associated with “mistargeting” voters—airing messages directed toward a specific audience toward a broader one, resulting in backlash (Hersh & Shaffner, 2013). Ridout et al. (2012) combine Wisconsin Advertising Project tracking data with Scarborough survey data to show that presidential candidates in recent cycles are increasingly distributing their ads across different programs on the basis of partisanship, suggesting that the tactics of microtargeting have moved to advertising.
There is evidence to suggest that message tailoring in advertising can be effective (Fowler, Franz, & Ridout, 2016; Strach et al., 2015), but there is also work on campaign mailers suggesting that targeting (or rather mistargeting in particular) can backfire. Hersh and Schaffner (2013) show that voters receiving microtargeted campaign mailers are no more likely to support sponsoring candidates than those who are exposed to more general appeals, but are considerably less likely to support the sponsors of identical yet mistargeted campaign mailers. Do the same findings apply to political advertising? Moreover, as Hillygus and Shields (2008) contend, politicians likely broadcast different messages to more general audiences than those they microtarget. With increases in targeting and continued fragmentation of advertising platforms, what are the consequences for voters, vote choice, and learning? Will the public perceive a unified campaign and will they be able to hold candidates accountable for campaign promises when different promises can be made to different voters?
These questions will only further grow with the latest technological advances in television advertising deployment. The new generation of ad targeting, which is likely to take off during the 2016 election, is called addressable advertising (Fowler, Franz, & Ridout, 2016). Addressable advertising will enable campaigns to target individual household targeting rather than relying solely on geographic and program-based filters. This means that the advertising any one individual sees may not be the same as his or her neighbor. As will be discussed below, these microtargeting developments represent a significant challenge for researchers wanting to study campaigns and campaign tactics systematically.
Conclusions and Challenges of Studying Advertising in the 21st Century
In recent election cycles, millions of political ads have flooded television airwaves in the United States and only recently have shown signs of slowing. These airings have grown more negative over time and the most recent cycles have seen a surge of interest group activity, particularly in competitive contests. Though Americans dislike negative campaigning, and political advertising more broadly, our review of the scholarly literature suggests that political elites have much to gain by “going negative.” There is plenty of evidence to suggest that advertising (and especially negative advertising) is memorable and can be persuasive and little evidence to suggest that negativity demobilizes the public although the timing of exposure to negativity may be key. Advertising and negative advertising has also been linked to increased learning and information seeking on the part of voters and while there may be some downside to negativity linked to citizen attitudes about the system, there is quite a bit of work challenging the notion that there are large, harmful long-term consequences of advertising. Figure 5 suggests that advertising in particular may not provide much partisan fodder to contribute to polarization and recent research could not find any influence of advertising on traditional measures of polarization (Ridout, Franz, Fowler, & Goldstein, 2015) in contrast to work examining media coverage of the topic (Levendusky & Malhotra, 2015).
As with the broader literature on media effects, evidence is robust that the influence of advertising is conditioned by various elements of their content—the tone, the emotional appeals they make, and how late/early in a campaign they are aired—along with characteristics of the audience receiving the messages. Although we know a lot about how individual features of advertising and how individual characteristics of viewers condition the reception of ad messages, there is little work comparing ads in terms of persuasiveness—either in comparing the relative contribution of multiple content features or in comparing different ads to each other.
Technological developments—both in video production and in campaign data—make the continued development of microtargeting strategies all but inevitable. The continued development of multifaceted advertising strategies on multiple online and social media platforms will further complicate the study of political advertising. In particular, campaigns have also expanded their advertising efforts to encompass a variety of platforms to include Facebook, YouTube, Twitter, Instagram, and Snapchat among others, and unlike the commercial tracking data provided by Kantar Media/CMAG for television advertising, there is no comparable source tracking online advertising across different platforms. This lack of data combined with the growing use of online ads presents a growing challenge for researchers especially in outlining how the content of advertising across platforms may differ, but increasing academic and practitioner collaborations hold much promise for assessing advertising influence. For example, Broockman and Green (2014) collaborated with local and national political campaigns in order to vary whether or not members of geographic constituencies were exposed to Facebook ad banners. After airing these ads for a week, follow-up phone interviews with individuals who were (and were not) treated with the banners showed that Facebook advertising exposure had no discernable impact on whether or not they viewed the sponsoring candidate more positively, or recalled that candidate’s name.
Future work should explore differences across platforms and how variations in content may influence online effectiveness on other outcomes of interest. Might advertising on other online and mobile platforms be more successful at facilitating repeated contact conducive to learning? Online advertisements can directly link citizens to additional online campaign content; for example, banner ads encouraging citizens to click on their content can redirect individuals to campaign websites or donation pages. Even if online advertisements fall short of persuading voters, can they be useful in redirecting web traffic to campaign pages or increasing donation cash flows? In other words, scholars could do more to study the effectiveness of advertising in meeting the acquisition (in addition to the persuasive and mobilization) goal of campaigns.
Similar to the challenges of tracking advertising across online platforms, the rise of addressable advertising on television, enabling household rather than geographic targeting of advertising, will further complicate studies of ad content and real-world exposure. There is little doubt that experiments and field experimentation will continue, but systematic studies of campaign strategy in advertising will become harder as traditional TV ad tracking data becomes a smaller portion of the total ad strategy. On the other hand, if academic and practitioner collaborations continue, academics may have more firsthand access to strategic decision-making.
Despite important methodological and logistical challenges, scholars have made considerable progress in understanding the content and influence of political advertising in the 21st century. The literature suggests that political advertising in general (and negativity in particular) plays an important role in U.S. democracy with largely positive or at least non-negative effects. As campaigns become more adept and precise at targeting the citizens they intend to reach, however, scholars have expressed some concerns that the democratic benefits of “spillover” advertising—those ads reaching a wider swath of the U.S. public who pays little attention to politics—may decrease (see Franz, 2013). Further, the mechanisms for holding politicians accountable for the promises and claims they make in a more fragmented advertising environment present a particular challenge not only for scholarship but also for citizens and good government organizations. Although digital advertising represents the largest growth sector for campaign ads, due to its reach, television advertising will an remain important part of campaign strategies for the foreseeable future. As such, the search for both main and moderated campaign effects will continue even as the landscape and scope of political advertising shifts and challenges in measurement and tracking grow. Moreover, as the landscape evolves, understanding how the content of advertising and viewer characteristics individually moderate and collectively interact to influence attitudes and behaviors among the public may be even more important.
Abramowitz, A. I., Alexander, B., & Gunning, M. (2006). Incumbency, redistricting, and the decline of competition in U.S. House elections. Journal of Politics, 68(1), 75–88.Find this resource:
Ansolabehere, S., & Iyengar, S. (1995). Going negative: How attack ads shrink and polarize the electorate. New York: Free Press.Find this resource:
Ansolabehere, S., Iyengar, S., Simon, A., & Valentino, N. (1994). Does attack advertising demobilize the electorate? American Political Science Review, 88(4), 829–838.Find this resource:
Banks, A. J. (2015). Anger and racial politics. Cambridge, U.K.: Cambridge University Press.Find this resource:
Banks, A. J., & Bell, M. A. (2013). Racialized campaign ads the emotional content in implicit racial appeals primes white racial attitudes. Public Opinion Quarterly, 77(2), 549–560.Find this resource:
Banks, A. J., & Valentino, N. A. (2012). Emotional substrates of white racial attitudes. American Journal of Political Science, 56(2), 286–297.Find this resource:
Bartels, L. M. (1993). Messages received: The political impact of media exposure. American Political Science Review, 87(2), 267–285.Find this resource:
Berelson, B. R., Lazarsfeld, P. F., & McPhee, W. N. (1954). Voting: A study of opinion formation in a presidential election. Chicago: University of Chicago Press.Find this resource:
Brader, T. (2005). Striking a responsive chord: How political ads motivate and persuade voters by appealing to emotions. American Journal of Political Science, 49(2), 388–405.Find this resource:
Brader, T. (2006). Campaigning for hearts and minds: How emotional appeals in political ads work. Chicago: University of Chicago Press.Find this resource:
Broockman, D. E., & Green, D. P. (2014). Do online advertisements increase political candidates’ name recognition or favorability? Evidence from randomized field experiments. Political Behavior, 36(2), 263–289.Find this resource:
Brooks, D. J. (2000). Negative Campaigning Disliked By Most Americans. Gallup News Service. Retrieved from http://www.gallup.com/poll/Find this resource:
Brooks, D. J., & Geer, J. G. (2007). Beyond negativity: The effects of incivility on the electorate. American Journal of Political Science, 51(1), 1–16.Find this resource:
Brooks, D. J., & Murov, M. (2012). Assessing accountability in a post-Citizens United era: The effects of attack ad sponsorship by unknown independent groups. American Politics Research, 40(3), 383–418.Find this resource:
Campbell, A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). The American Voter. Ann Arbor: University of Michigan Press.Find this resource:
Converse, Philip E. (1964). The Nature of Belief Systems in Mass Publics. In David Apter (Ed.), Ideology and Discontent (pp. 207–261). New York: Free Press.Find this resource:
Cox, G. W., & Katz, J. N. (1996). Why did the incumbency advantage in U.S. House elections grow? American Journal of Political Science, 40(2), 478–497.Find this resource:
Craig, S. C., Kane, J. G., & Gainous, J. (2005). Issue related learning in a gubernatorial campaign: A panel study. Political Communication, 22(4), 483–503.Find this resource:
Dowling, C. M., & Wichowsky, A. (2013). Does it matter who’s behind the curtain? Anonymity in political advertising and the effects of campaign finance disclosure. American Politics Research, 41(6), 965–996.Find this resource:
Dowling, C. M., & Wichowsky, A. (2015). Attacks without consequence? Candidates, parties, groups, and the changing face of negative advertising. American Journal of Political Science, 59(1), 19–36.Find this resource:
Druckman, J. N., Peterson, E., & Slothuus, R. (2013). How elite partisan polarization affects public opinion formation. American Political Science Review, 107(1), 57–79.Find this resource:
Eagly, A. H., & Chaiken, S. (1993). The psychology of attitudes. Harcourt Brace Jovanovich College Publishers.Find this resource:
Finkel, S. E. (1993). Reexamining the “minimal effects” model in recent presidential campaigns. Journal of Politics, 55(1), 1–21.Find this resource:
Finkel, S. E., & Geer, J. G. (1998). A spot check: Casting doubt on the demobilizing effect of attack advertising. American Journal of Political Science, 42(2), 573–595.Find this resource:
Fowler, E. F., Franz, M. M, & Ridout, T. N. (2014). Political advertising in 2010. Version 1.0 [dataset]. Middletown, CT: The Wesleyan Media Project, Department of Government at Wesleyan University.Find this resource:
Fowler, E. F., Franz, M. M, & Ridout, T. N. (2015). Political advertising in 2012. Version 1.0 [dataset]. Middletown, CT: The Wesleyan Media Project, Department of Government at Wesleyan University.Find this resource:
Fowler, E. F., Franz, M. M, & Ridout, T. N. (2016) Political advertising in the United States. Boulder, CO: Westview.Find this resource:
Fowler, E. F., & Ridout, T. N. (2009). Local television and newspaper coverage of political advertising. Political Communication, 26(2), 119–136.Find this resource:
Fowler, E. F., & Ridout, T. N. (2010). Advertising Trends in 2010. Forum, 8(4), Article 4.Find this resource:
Fowler, E. F., & Ridout, T. N. (2013). Negative, angry, and ubiquitous: Political advertising in 2012. Forum, 10(4), 51–61.Find this resource:
Fowler, E. F., & Ridout, T. N. (2014). Political advertising in 2014: The year of the outside group. Forum, 12(4), 663–684.Find this resource:
Franz, M. M. (2013). Targeting campaign messages. In T. N. Ridout (Ed.), New directions in media and politics (p. 113). Routledge.Find this resource:
Franz, M. M., Freedman, P., Goldstein, K., & Ridout, T. N. (2008). Understanding the effect of political advertising on voter turnout: A response to Krasno and Green. Journal of Politics, 70(1), 262–268.Find this resource:
Franz, M. M., & Ridout, T. N. (2007). Does political advertising persuade? Political Behavior, 29(4), 465–491.Find this resource:
Freedman, P., Franz, M., & Goldstein, K. (2004). Campaign advertising and democratic citizenship. American Journal of Political Science, 48(4), 723–741.Find this resource:
Freedman, P., & Goldstein, K. (1999). Measuring media exposure and the effects of negative campaign ads. American Journal of Political Science, 43(4), 1189–1208.Find this resource:
Fridkin, K. L., & Kenney, P. J. (2004). Do negative messages work? The impact of negativity on citizens’ evaluations of candidates. American Politics Research, 32(5), 570–605.Find this resource:
Garramone, G. M., Atkin, C. K., Pinkleton, B. E., & Cole, R. T. (1990). Effects of negative political advertising on the political process. Journal of Broadcasting & Electronic Media, 34(3), 299–311.Find this resource:
Geer, J. G., & Geer, J. H. (2003). Remembering attack ads: An experimental investigation of radio. Political Behavior, 25(1), 69–95.Find this resource:
Geer, J. G. (2006). In defense of negativity: Attack ads in presidential campaigns. Chicago: University of Chicago Press.Find this resource:
Gelman, A., & King, G. (1993). Why are American presidential election campaign polls so variable when votes are so predictable? British Journal of Political Science, 23(4), 409–451.Find this resource:
Gerber, A. S., Gimpel, J. G., Green, D. P., & Shaw, D. R. (2011). How large and long-lasting are the persuasive effects of televised campaign ads? Results from a randomized field experiment. American Political Science Review, 105(1), 135–150.Find this resource:
Gerber, A. S., Huber, G. A., Doherty, D., Dowling, C. M., & Panagopoulos, C. (2013). Big five personality traits and responses to persuasive appeals: Results from voter turnout experiments. Political Behavior, 35(4), 687–728.Find this resource:
Gimpel, J. G., Lee, F. E., & Pearson-Merkowitz, S. (2008). The check is in the mail: Interdistrict funding flows in congressional elections. American Journal of Political Science, 52(2), 373–394.Find this resource:
Goldstein, K., & Freedman, P. (2002). Campaign advertising and voter turnout: New evidence for a stimulation effect. Journal of Politics, 64(3), 721–740.Find this resource:
Goldstein, K., Niebler, S., Neiheisel, J., & Holleque, M. (2011). Presidential, congressional, and gubernatorial advertising, 2008, combined file [dataset]. Initial release. Madison, WI: University of Wisconsin Advertising Project, Department of Political Science at the University of Wisconsin-Madison.Find this resource:
Gosnell, H. F. (1927). Getting out the vote. An experiment in the stimulation of voting. Chicago: University of Chicago Press.Find this resource:
Hersh, E. D., & Schaffner, B. F. (2013). Targeted campaign appeals and the value of ambiguity. The Journal of Politics, 75(02), 520–534.Find this resource:
Hill, S. J., Lo, J., Vavreck, L., & Zaller, J. (2013). How quickly we forget: The duration of persuasion effects from mass communication. Political Communication, 30(4), 521–547.Find this resource:
Hillygus, D. S. (2005). The missing link: Exploring the relationship between higher education and political engagement. Political Behavior, 27(1), 25–47.Find this resource:
Hillygus, D. S., & Shields, T. G. (2008). The persuadable voter: Wedge issues in presidential campaigns. Princeton, NJ: Princeton University Press.Find this resource:
Holbrook, T. M. (1994). Campaigns, national conditions, and U.S. presidential elections. American Journal of Political Science, 38(4), 973–998.Find this resource:
Holtz-Bacha, C, & Kaid L. L. (2006). Political advertising in international comparison. In L. L. Kaid & C. Holtz-Bacha (Eds.), The SAGE handbook of political advertising. Thousand Oaks, CA: SAGE.Find this resource:
Hovland, C. I. (1959). Reconciling conflicting results derived from experimental and survey studies of attitude change. American Psychologist, 14(1), 8.Find this resource:
Hovland, C. I., Janis, I. L., & Kelley, H. H. (1954). Communication & Persuasion. New Haven, CT: Yale University Press.Find this resource:
Hovland, C. I., Lumsdaine, A. A., Sheffield, F. D., Hovland, C. I., Lumsdaine, A. A., & Fred, D. (1949). Experiments on Mass Communication. Princeton, NJ: Princeton University Press.Find this resource:
Huber, G. A., & Arceneaux, K. (2007). Identifying the persuasive effects of presidential advertising. American Journal of Political Science, 51(4), 957–977.Find this resource:
Iyengar, S., Peters, M. D., & Kinder, D. R. (1982). Experimental demonstrations of the “not-so-minimal” consequences of television news programs. American Political Science Review, 76(4), 848–858.Find this resource:
Iyengar, S., Sood, G., & Lelkes, Y. (2012). Affect, not ideology a social identity perspective on polarization. Public Opinion Quarterly, 76(3), 405–431.Find this resource:
Iyengar, S., & Westwood, S. J. (2014). Fear and loathing across party lines: New evidence on group polarization. American Journal of Political Science, 59(3), 690–707.Find this resource:
Jackson, R. A., Mondak, J. J., & Huckfeldt, R. (2008). Examining the possible corrosive impact of negative advertising on citizens’ attitudes toward politics. Political Research Quarterly, 62(1), 55–69.Find this resource:
Kahn, K. F., & Kenney, P. J. (1997). A model of candidate evaluations in Senate elections: The impact of campaign intensity. Journal of Politics, 59(4), 1173–1205.Find this resource:
Kahn, K. F., & Kenney, P. J. (1999). Do negative campaigns mobilize or suppress turnout? Clarifying the relationship between negativity and participation. American Political Science Review, 93(4), 877–889.Find this resource:
Kaid, L. L. (1997). Effects of the television spots on images of Dole and Clinton. American Behavioral Scientist, 40(8), 1085.Find this resource:
Kam, C. D., & Simas, E. N. (2010). Risk orientations and policy frames. The Journal of Politics, 72(02), 381–396.Find this resource:
Klotz, R. (2010). The sidetracked 2008 YouTube campaign. Journal of Information Technology & Politics, 7(2/3), 110–123.Find this resource:
Krasno, J. S., & Green, D. P. (2008). Do televised presidential ads increase voter turnout? Evidence from a natural experiment. Journal of Politics, 70(1), 245–261.Find this resource:
Krosnick, J. A., & Kinder, D. R. (1990). Altering the foundations of support for the president through priming. American Political Science Review, 84(2), 497–512.Find this resource:
Kruglanski, A. W., Webster, D. M., & Klem, A. (1993). Motivated resistance and openness to persuasion in the presence or absence of prior information. Journal of Personality and Social Psychology, 65(5), 861.Find this resource:
Krupnikov, Y. (2011). When does negativity demobilize? Tracing the conditional effect of negative campaigning on voter turnout. American Journal of Political Science, 55(4), 797–813.Find this resource:
Lang, A. (1991). Emotion, formal features, and memory for televised political advertisements. Television and political advertising, 1(1), 221–243.Find this resource:
Lang, A., Dhillon, K., & Dong, Q. (1995). The effects of emotional arousal and valence on television viewers’ cognitive capacity and memory. Journal of Broadcasting & Electronic Media, 39, 313–327.Find this resource:
Lau, R. R., & Pomper, G. M. (2004). Negative campaigning: An analysis of U.S. Senate elections. Rowman & Littlefield.Find this resource:
Lau, R. R., Sigelman, L., Heldman, C., & Babbitt, P. (1999). The effects of negative political advertisements: A meta-analytic assessment. American Political Science Review, 93(4), 851–875.Find this resource:
Lau, R. R., Sigelman, L., & Rovner, I. B. (2007). The effects of negative political campaigns: A meta‐analytic reassessment. Journal of Politics, 69(4), 1176–1209.Find this resource:
Lavine, H. G., Johnston, C. D., & Steenbergen, M. R. (2012). The ambivalent partisan: How critical loyalty promotes democracy. Oxford: Oxford University Press.Find this resource:
Lazarsfeld, P. F., Berelson, B., & Gaudet, H. (1944). The people’s choice: How the voter makes up his mind in a presidential election. New York: Duell, Sloan and Pearce.Find this resource:
Levendusky, M., & Malhotra, N. (2015). Does media coverage of partisan polarization affect political attitudes? Political Communication, 33(2), 283–301.Find this resource:
Lupia, A. (1994). Shortcuts versus encyclopedias: Information and voting behavior in California insurance reform elections. American Political Science Review, 88(01), 63–76.Find this resource:
Markus, G. E., Neuman, W. R., & MacKuen, M. (2000). Affective intelligence and political judgment. Chicago: University of Chicago Press.Find this resource:
Mattes, K., & Redlawsk, D. P. (2014). The positive case for negative campaigning. Chicago: University of Chicago Press.Find this resource:
McGuire, W. J. (1985) Attitudes and attitude change. In G. Lindzey & E. Aronson (Eds.), Handbook of social psychology (pp. 1130–1187). Chicago: Rand McNally.Find this resource:
Mutz, D. C. (2007). Effects of “in-your-face” television discourse on perceptions of a legitimate opposition. American Political Science Review, 101(4), 621–635.Find this resource:
Mutz, D. C. (2011). Population-based survey experiments. Princeton, NJ: Princeton University Press.Find this resource:
Patterson, T. E., & McClure, R. (1976). The unseeing eye: The myth of television power in national elections. New York: Putnam.Find this resource:
Pfau, M., & Louden, A. (1994). Effectiveness of adwatch formats in deflecting political attack ads. Communication Research, 21(3), 325–341.Find this resource:
Pinkleton, B. (1997). The effects of negative comparative political advertising on candidate evaluations and advertising evaluations: An exploration. Journal of Advertising, 26(1), 19–29.Find this resource:
Prior, M. (2007). Post-broadcast democracy: How media choice increases inequality in political involvement and polarizes elections. Cambridge, U.K.: Cambridge University Press.Find this resource:
Rahn, W. M. (1993). The role of partisan stereotypes in information processing about political candidates. American Journal of Political Science, 437(2), 72–496.Find this resource:
Ridout, T. N., & Fowler, E. F. (2012). Explaining Perceptions of Advertising Tone. Political Research Quarterly, 65(1), 62–75.Find this resource:
Ridout, T. N., Fowler, E. F., Branstetter, J., & Borah, P. (2015). Politics as usual? When and why traditional actors often dominate YouTube campaigning. Journal of Information Technology & Politics, 12(3), 237–251.Find this resource:
Ridout, T. N., Franz, M., Fowler, E. F. (2015). Sponsorship, disclosure, and donors: Limiting the impact of outside group ads. Political Research Quarterly, 68(1), 154–166.Find this resource:
Ridout, T. N., Franz, M., Fowler, E. F., & Goldstein, K. M. (2015). The long-term and geographically-constrained effects of political advertising on political polarization. Paper presented at the 2015 APSA Political Communication Pre-Conference. San Francisco, CA.Find this resource:
Ridout, T. N., Franz, M., Goldstein, K. M., & Feltus, W. J. (2012). Separation by television program: Understanding the targeting of political advertising in presidential elections. Political Communication, 29(1), 1–23.Find this resource:
Ridout, T. N., & Franz, M. M. (2011). The persuasive power of campaign advertising. Philadelphia: Temple University Press.Find this resource:
Ridout, T. N., Shah, D. V., Goldstein, K. M., & Franz, M. M. (2004). Evaluating measures of campaign advertising exposure on political learning. Political Behavior, 26(3), 201–225.Find this resource:
Ridout, T. N., & Smith, G. R. (2008). Free advertising: how the media amplify campaign messages. Political Research Quarterly, 61(4), 598–608.Find this resource:
Shaw, E. F. (1979). Agenda setting and mass communication theory. International Communication Gzette, 25(2), 96–105.Find this resource:
Sides, J., & Vavreck, L. (2014). The gamble: Choice and chance in the 2012 presidential election. Princeton, NJ: Princeton University Press.Find this resource:
Soroka, S., & McAdams, S. (2015). News, politics, and negativity. Political Communication, 32(1), 1–22.Find this resource:
Stevens, D. (2002). Explaining heterogeneity in the effects of negative advertising. Presented at the annual meeting of the Midwest Political Science Association, Chicago.Find this resource:
Stevens, D. (2005). The two routes for effects of negative advertising. Presented at the annual meeting of the Midwest Political Science Association, Chicago.Find this resource:
Stevens, D. (2012). Tone versus information: Explaining the impact of negative political advertising. Journal of Political Marketing, 11(4), 322–352.Find this resource:
Strach, P., Zuber, K., Fowler, E. F., Ridout, T. N., & Searles, K. (2015). In a Different Voice? Explaining the Use of Men and Women as Voice-Over Announcers in Political Advertising. Political Communication, 32(2), 183–205.Find this resource:
Thorson, E., Ognianova, E., Coyle, J., & Denton, F. (2000). Negative political ads and negative citizen orientations toward politics. Journal of Current Issues & Research in Advertising, 22 (Spring), 13–41.Find this resource:
Valentino, N. A., Hutchings, V. L., & Williams, D. (2004). The impact of political advertising on knowledge, Internet information seeking, and candidate preference. Journal of Communication, 54(2), 337–354.Find this resource:
Valentino, N. A., & Nardis Y. (2013). Political communication: Form and consequence of the information environment. In L. Huddy, D. O. Sears, & J. S. Levy (Eds.), The Oxford handbook of political psychology, 559–590. Oxford: Oxford University Press.Find this resource:
Wattenberg, M. P., & Brians, C. L. (1999). Negative campaign advertising: Demobilizer or mobilizer? American Political Science Review, 93(4), 891–899.Find this resource:
Weber, C., Dunaway, J., & Johnson, T. (2012). It’s all in the name: Source cue ambiguity and the persuasive appeal of campaign ads. Political Behavior, 34(3), 561–584.Find this resource:
West, D. M. (2009). Air wars: Television advertising in election campaigns, 1952–2008. Washington, D.C.: CQ Press.Find this resource:
Zaller, J. (1992). The nature and origins of mass opinion. Cambridge, U.K.: Cambridge University Press.Find this resource:
Zaller, J. (1996). The myth of massive media impact revived: New support for a discredited idea. In D. C. Mutz, P. M. Sniderman, & R. A. Brody (Eds.), Political Persuasion and Attitude Change (pp. 17–78). University of Michigan Press.Find this resource:
(1.) While Americans clearly dislike negative advertising in general, Mattes and Redlawsk (2014) point out that they are more favorable toward specific elements of content typically featured in negative ads (e.g., criticizing an opponent’s voting record or political views), provided that this is done without actually making mention of “negativity” (as merely mentioning the term might make respondents feel social desirability pressures to express dislike).
(2.) That is, those which do not attack another candidate (“attack ads”), or promote one while attacking another (“contrast ads”).
(4.) Data for 2006 include only the top 100 markets, not all 210. Other work has shown the increase in volume persists if we limit to the top 75 markets, which were consistently tracked over time (see Fowler, Franz, & Ridout, 2016).
(6.) These figures include all advertisements aired during the calendar year between January 1st and Election Day.
(7.) Rising from 27%t of negative ads aired in the 2008 general election contest to 33% in 2012, while party-sponsored negativity fell from 21% to about 4% of airings in the same time span.
(8.) The Wesleyan Media Project also codes for whether or not both parties were mentioned in the content of advertisements, and if the favored candidate’s party is mentioned in passing, in the “Paid for By” (PFB) line in each spot (i.e., the text required by Bipartisan Campaign Reform Act or BCRA that discloses who sponsored the advertisement). For the ease of visual interpretation, given the uniformly low percentages of party mentions in campaign advertisements, we omit these trend lines from our analyses. Ads mentioning both parties never comprise more than 1% of advertisements aired in the series, and ads mentioning the favored candidates’ parties in the PFB line range from a low of a third of 1% in 2014, to a high of 7% in 2010.
(9.) This result becomes even more apparent when factoring in additional favored candidate party mentions that are only mentioned in passing, in advertisements’ PFB lines.
(10.) Whether or not negativity increases turnout has been an important source of methodological debate. For more information, see Krasno & Green (2008) and Franz, Freedman, Goldstein, & Ridout (2008).
(11.) While Kahn and Kenney (1997) do not expressly measure political advertising exposure in making these claims, they do estimate citizens’ exposure to intense campaign efforts in widely-circulated newspapers: a measure which they expect to be highly correlated with advertising exposure.
(12.) This is not to say that negative advertisements that focus on issues fail to make citizens feel angry—like the emotional cues present in ads might. Rather, this finding suggests that anger toward negative campaigning in general is suppressed when advertisements focus on political issues.