Partisanship, media and the objective economy: Sources of individual-level economic assessments

Abstract

Economic voting studies have repeatedly shown that voter's assessment of incumbent economic performance is important for the vote decision. However, there is little work explaining how individuals form their economic assessments. Utilizing individual-level data from Turkey, we find that variation in retrospective assessments can actually be predicted by individual income growth rates over the previous year, and the association is stronger for pocketbook assessments. Nonetheless, partisanship and media are important sources of bias, especially for sociotropic assessments. Controlled for partisanship, viewers of pro-government media are more likely to think that the national economy has done better than their own household over the last year, and also more likely to believe that the economy would fare worse if the incumbent is replaced. The findings testify both to the capacity of the individuals to anchor their assessments to personal experience, and to the media's ability to weaken this anchor.