When Do Campaign Effects Persist for Years? Evidence from a Natural Experiment

The forthcoming article “When Do Campaign Effects Persist for Years? Evidence from a Natural Experiment” (https://doi.org/10.1111/ajps.12488) by Kai Jäger is summarized by the author below.

Election campaigns are a cornerstone of democracies. The ability of parties to sway the public via election campaigns draws the attention from academics, journalists, and historians alike. The vast academic literature on election campaigns generally suggests, however, that campaign effects are, at best, minimal and short-lived. Given this consensus and the methodological difficulty to conduct long-term studies due to potential confounders, we do not know much about the longevity of campaign effects and what political behavior can be durably shaped by election campaigns. 

A unique political event in Germany constitutes a natural experiment to evaluate whether political campaigns can last for nearly a decade: In the 2007 state elections of Bremen, the local right-wing conservative party Citizens in Rage (Bürger in Wut, BIW) missed the 5-percent threshold for parliamentary representation by a mere vote. Due to vote counting irregularities in a precinct in the city of Bremerhaven, the Constitutional Court of Bremen ordered a re-vote in this single precinct over a year later. BIW increased its vote share in the re-vote from 4.35 to 27.57 percent, successfully passing the threshold for parliamentary representation. What preceded the re-vote was a one-sided election campaign focusing on law and order by BIW frontrunner Jan Timke in the precinct. 

I use three different research strategies to evaluate whether BIW’s re-vote campaign could have a long-lasting impact. First, I compared the election results of BIW in the re-vote precinct to the neighboring precincts in subsequent elections, finding that BIW’s vote share has on average increased by about 4.2 percentage points in the re-vote precinct. 

Second, I conducted an observational shoe-leather study, in which I evaluate whether the precinct area has installed more warning signs on their property than the structurally identical neighborhood. Warning signs are an indicator of security-sensitive behavior and suggest that residents have been influenced by BIW’s law-and-order campaign. I find that residents of the re-vote precinct were 13 percentage points more likely than their neighbors to have warning signs.  

Third, I invited the residents of the re-vote area and the structurally similar surroundings to a mail survey to test for attitudinal differences. The survey shows that both groups do not differ in their support for right-wing programmatic positions, but that residents of the re-vote area were 15 percentage points more likely to consider BIW as the most competent party on security. In addition, I also find that the re-vote has negatively affected some levels of trust in the democratic system, as residents of the re-vote area were more likely to believe that election fraud reoccurred in the 2015 Bremen election, in which the right-wing Alternative for Germany (Alternative für Deutschland, AfD) barely missed the five-percent threshold by a few votes. 

There are broader lessons for election campaigns from this unique case, as my study shows that long-term effects can exist if parties are able to run a dominant campaign, which could even shape non-political behavior. Thus, campaigning in uncompetitive first-past-the-post constituencies or in non-election periods might turn out to have long-term benefits.

About the Author: Kai Jäger is a Lecturer in Political Economy, Department of Political Economy, King’s College London. The research “When Do Campaign Effects Persist for Years? Evidence from a Natural Experiment” (https://doi.org/10.1111/ajps.12488) is now available in Early View and will appear in a forthcoming issue of the American Journal of Political Science.