View article

Do media affect the behavior of people who are not their intended audience? Do people ignore media outlets which are hostile to them? Our paper addresses these questions by looking at the influence of Serbian public radio on anti-Serbian nationalism in a post-conflict border region of Croatia. The effect is identified from the exogenous variation in the availability of the signal, mostly generated by village elevation and amounts of forests nearby. We find that in places with a good quality of signal of Serbian public radio people were more likely to vote for extreme nationalist parties. These results are consistent with IV estimation in which predicted signal availability is used to instrument for actual radio availability and with larger sample results. We also find some weak evidence that people in places with Serbian public radio are less likely to vote for moderate nationalist and post-communist parties and also less likely …