Do Online Advertisements Increase Name Recognition or Favorability of Political Candidates in the United States?

David E. Broockman
Fieldwork by:
California, United States of America
2,984 and 3,577 voters in each district
2012 - 2012
Target group:
  • Voters
Intervention type:
  • Digital and mobile
  • Information

The massive audience offered by the Internet has made online political advertising increasingly popular in the United States. Researchers randomly selected segments of a political candidate’s constituency to evaluate the impact of online Facebook advertisements on the name recognition and favorability of political candidates. Voters randomly exposed to the ads were in some cases more likely to recall the ads, but no more likely to recognize or positively evaluate the candidates they depicted.

Policy issue

In the United States in 2012, about 85 percent of individuals used the Internet and one-third of the adult population logged into Facebook alone at least once per day. This enormous audience has made online political advertising increasingly popular. For example, in 2012, presidential candidates Barack Obama and Mitt Romney spent approximately 25 percent of their advertising dollars on Internet ads (equivalent to about 100 to 150 million US dollars). Despite the amount of spending dedicate to online advertisements, their impact remains largely unknown. In order to assess the effects of online advertising, researchers evaluated the impact of a high volume of Facebook advertising on name recognition and favorability of political candidates.

Context of the evaluation

Researchers examined the effects of week-long Facebook advertising campaigns leading up to the November 2012 US elections for two different contexts: 1) a little-known Republican candidate running for state legislative office; and 2) a viable Democratic candidate running for Congress. The first candidate ran in a predominantly white, rural district, while the second candidate’s Congressional district was more densely populated. The timing of the advertising campaigns also differed slightly. The first candidate’s advertisements were displayed one month before the election, while the second candidate’s advertisements were displayed one week beforehand.

Details of the intervention

To understand the effectiveness of Facebook advertising, researchers grouped registered voters by age, gender, and location, and then randomly assigned them to receive or not receive the Facebook political ads. Both candidates purchased the maximum volume of ads that Facebook allows, so voters in the treatment group were typically exposed to many dozen ads over the course of the week-long campaign. Following the campaigns, independent firms conducted polls of registered voters in the candidates’ constituencies. The polls included questions about whether respondents knew the candidates’ names, had favorable impressions of them, whether they recalled seeing material on the Internet about the candidates, and Facebook usage. In the first experiment, of the 32,029 registered voters randomly assigned to treatment and control groups, live interviews were completed with 2,984 of them. In the second, of the 261,150 registered voters randomly assigned to treatment and control groups, automated interviews were completed with 3,557 of them.

Results and policy lessons

In both experiments, exposure to Facebook ads did not increase recognition of a candidate’s name or increase positive assessments of the candidate. The studies’ findings differ only regarding recall of the advertisements. In the first study, individuals assigned to the treatment group were no more likely to recall seeing the ads, while individuals in the second study who were assigned to the treatment group were 5.3 percentage points more likely to recall seeing the ads. Taken together, the results cast doubt on the ability of online advertisements to have meaningful impacts on political attitudes or election outcomes. However, many questions remain open: For instance, further evaluation may determine whether more dynamic or prominent ads are more effective, or whether more personalized outreach from known sources (e.g. a Facebook friend) are more effective than impersonal advertisements.

Broockman, David E., and Donald Green. 2014. "Do Online Advertisements Increase Political Candidates’ Name Recognition or Favorability? Evidence from Randomized Field Experiments." Political Behavior 36(2): 263-289.