Honesty is the best policy. When and why are governments (not) punished for breaking electoral pledges?

The electoral consequences of (unpopular) reforms are often theorized, but seldom measured and tested. Based on a new dataset that combines election pledges, policy outputs, and economic performance data with the electoral gains and losses of 220 governments in 18 countries between 1971 and 2012, we add to the ongoing debate on policy feedback by investigating how welfare reforms affect electoral gains and losses. In line with recent advances in the literature, we do not find that electoral punishment for cutbacks is a universal phenomenon. Rather, our results attest to the importance of four theoretical key distinctions: Whether the reforms concern life-course or labour market risks; whether the government is characterized by relatively high or relatively low electoral vulnerability; whether the cabinet is in an economic gains or losses domain; and whether policy changes were ‘signalled’ to the electorate in election pledges or not (as measured by the ‘pledge gap’, the gap between implemented policies and electoral promises). The results help us to qualify recent arguments concerning the link between parties, policies, institutions, and voters.