Air war? Campaign advertising in the 2016 Presidential Election

In many respects, the 2016 Presidential Election was unlike any other. One particularly unique feature of the campaign was a sizable asymmetry in the number of advertisements aired on behalf of each of the major party candidates. 

Despite being vastly outspent on the airwaves, President-Elect Donald Trump won more than 300 votes in the Electoral College. However, his victory should not imply that political advertisements are ineffective at winning votes. Instead, the final election tally begs scholars and observers of American politics to rethink conventional wisdom about campaigning on television. 

In what follows, I raise (and attempt to answer) several questions about the state of advertising in 2016 and its implications for what scholars know about their effectiveness.

A War on the Airwaves?

If the 2016 campaign was a battle for control of the airwaves, the fight was one sided (at best). While both sides saw fewer advertisements aired on their behalf than did each respective party nominee in 2012, data from the Wesleyan Media Project (WMP) suggests that pro-Clinton airings (489,142 from June 8 – October 30, 2016) were about three times greater than pro-Trump airings (99,441). Clinton’s dominance on the airwaves held fairly steady throughout the campaign. In contrast to the view that Trump might make a late push to flood the airwaves with advertisements before the campaign concluded, WMP data show that pro-Clinton advertisements outnumbered pro-Trump ads by nearly 2:1 in the final two weeks of the campaign

There were also several important asymmetries in the sponsorship of advertisements on both sides. While candidates sponsored the majority of all ads aired in their favor, Clinton received substantial help from interest groups (more than ninety thousand airings in her favor), whereas Donald Trump received absolutely none (although several interest groups were actively involved in airing anti-Clinton advertisements). Interestingly, Donald Trump aired fewer advertisements overall than did Bernie Sanders in the Democratic Primary, and the overall tone of ads aired were somewhat more positive than those aired in 2012.

Does Campaign Advertising Change Minds?

In the past, political scientists have found that asymmetries in advertising totals have important consequences for candidates’ electoral fates. Several scholars have demonstrated that advertising advantages can increase support for a candidate, even independently of mobilization efforts “on the ground.” Political scientists John Sides and Lynn Vavreck find that support tends to respond to short-term airing advantages. But, because candidates typically keep pace with each others’ advertising spending, these effects usually cancel out. 

The 2016 election offers a unique opportunity for scholars to study a campaign in which advertising was more one-sided, and may prove to turn conventional wisdom on its head. Consistent with conventional wisdom, Donald Trump picked up narrow, and unanticipated, victories in Wisconsin and Michigan; states where he held moderate to high advertising advantages in the final two weeks of the campaign, in some media markets. Further, the candidate with the most advertisements aired on her behalf also won the popular vote. 

Yet, there is also reason to rethink the conventional wisdom. In some states where Clinton held heavy advertising advantages in the final weeks of the campaign (e.g., Ohio, Pennsylvania, and Florida), she ultimately lost. Further, Trump’s advertising edges in Virginia and Colorado during the same timeframe ultimately did not win him either state. 

How will Scholars Make Sense of 2016?

We can never truly know what the popular and state vote totals would have looked like had political advertising not been present. One way to ascertain the effectiveness will be to turn to public opinion surveys collected after the final vote was tallied.

Political advertisements create “naturally occurring” experiments on a daily basis, because media markets often cross state boundaries. If voters live in markets that receive substantial advertising because they overlap with a battleground state (where candidates will also have strong ground games), but do not actually live in a battleground state themselves, it is possible to isolate the effect of advertising independently of other factors that might also shape vote choice. 

For example, the Erie market in Pennsylvania (a battleground state) overlaps with New York (a strongly Democratic state). In the final two weeks of the campaign, Trump held a significant advertising advantage in Erie. If voters in that part of New York became more likely to vote for Trump at the election’s conclusion, advertisements may have indeed shaped their vote choice. 

Advertisements also have the potential to do more than alter citizens’ vote intentions. Exposure to campaign advertising has been shown to boost citizens’ knowledge about ,and interest in, the presidential campaign, for example. 

The 2016 Election will almost certainly challenge conventional wisdom about presidential campaign advertising. Scholars now have an opportunity to empirically which aspects of conventional wisdom were upheld, and which need further attention.