Measuring Change in Personal Economic Well-Being

Abstract

Individual-level investigations of the link between personal economic well-being and political preference commonly turn up findings at odds with aggregate, overtime analyses. Although various theoretical explanations for this anomaly have been advanced, we focus on a different culprit: the possibly frail measure employed in survey based studies. We develop improved measures of change in personal economic well-being and use a national survey of the American electorate to test them. Our findings are striking: although the traditional question used by survey researchers is quite valid, its unreliability has caused the effect of personal economic well-being on political evaluations to be substantially underestimated.