Radio Advertisements to Increase Electoral Competition in the United States

Costas Panagopoulos
United States
49 cities
2005 - 2006
Target group:
Outcome of interest:
Electoral participation Voter behavior
Intervention type:
Nudges and reminders
AEA RCT registration number:


Incumbents in elections may enjoy an advantage over their challengers if voters are more likely to vote for candidates they are more familiar with. Mass media campaigns that provide information about both candidates may increase the competitiveness of elections by familiarizing voters with the lesser-known challengers. Researchers examined the effect of radio advertisements on electoral competition through a randomized evaluation and found that radio advertisements that stated the names of both incumbents and challengers, reminded listeners about the date of the upcoming election, and encouraged them to vote increased competitiveness in municipal elections in the United States.

Policy issue

Incumbents often easily defeat their challengers in political elections. In the United States, for example, reelection rates for members of the House of Representatives have exceeded 90 percent since the 1940s. One potential explanation for this incumbent advantage is that voters are more willing to vote for incumbents simply because they are more familiar with them than their opponents. Such voting tendencies can increase the risk of keeping an incompetent official in office while a better candidate exists. Providing information on candidates through mass media, such as radio broadcasting, could diminish such incumbent advantage. However, there is little empirical evidence on the real-world impact of radio information campaigns on voting.

Context of the evaluation

This study focused on municipal elections in the United States in November 2005 and 2006. Municipal elections are usually low-profile events that attract little media attention. Campaigning via television or newsprint in these elections is rare because of the associated high cost and difficulty in using these media to target a geographically compact area. This lack of media coverage means challengers in these elections often have difficulty gaining notoriety, so information might have a large effect on incumbent advantage. Radio advertisement are particularly relevant in this context because they can be targeted to a local area while TV and newspaper advertisements typically can only be targeted by metropolitan area.

Details of the intervention

Researchers randomly assigned cities to receive nonpartisan radio advertisements to study the effect of such advertisements on electoral competitiveness. Each participating city was paired with another that was comparable in terms of voter turnout and incumbents’ vote share in previous mayoral elections, whether mayoral elections were partisan or not, and whether the upcoming mayoral election was contested. One city in each pair was randomly selected to receive the advertisements while the other was not.

Throughout the week prior to the municipal elections in 2005 and 2006, a scripted 60-second message stating the names and party affiliations of each of the candidates running for mayor, reminding listeners about the date of the upcoming election, and encouraging them to vote was aired on various radio stations in the treatment cities. All of the messages were created by a professional political consulting and media firm, were non-partisan, and did not evaluate the candidates.

Taking into account the different local costs of airing radio advertisments, researchers varied the size of the audience the advertisements reached as low (50 GRP1 ), medium (70 GRP), or high exposure (90 GRP). They took care to ensure that the advertisements were broadcasted at peak listening times in all locations to capture as wide a local audience as possible.

Results and policy lessons

Researchers found that the radio advertisements reduced the average vote share received by the incumbent. Each additional GRP reduced (by 0.078 percentage points) the incumbent’s average vote share. The incumbent experienced a 3.9 percentage points drop in gain in vote shares between consecutive elections in cities with low exposure, a 5.5 percentage points drop in cities with medium exposure, and 7.0 percentage points drop in cities with high exposure. The estimated effects of radio ads fall short of conventional levels of statistical significance, however, suggesting the need for further studies of this type.

These results tentatively suggest that promoting awareness of upcoming elections and publicizing even minimal candidate information can contribute to competitiveness of elections by reducing the incumbent advantage.

Panagopoulos, Costas, and Donald P. Green. 2008. "Field Experiments Testing the Impact of Radio Advertisements on Electoral Competition." American Journal of Political Science 52(1): 156–168.

GRPs are used to measure the size of an advertisement’s audience and factor in both the reach (the percentage of the target population that has heard the ad at least once) and frequency (the number of times the audience hears the ad) of an advertisement. GRP is calculated by multiplying the reach by the frequency. For example, 100 GRP can mean 50 percent of the audience heard the advertisements an average of two times, or 25 percent of the audience heard them an average of four times, or any other combination that equals 100.