What explains the wide variation across countries in the use of vote buying and policy promises during election campaigns? We address buy ip votes this question, and account for a number of stylized facts and apparent anomalies regarding vote buying, using a model in which parties cannot fully commit to campaign promises. We find that high vote buying is associated with frequent reneging on campaign promises, strong electoral competition, and high policy rents. Frequent reneging and low party competence reduce campaign promises. If vote buying can be financed out of public resources, incumbents buy more votes and enjoy an electoral advantage, but they also promise more public goods. Vote buying has distributional consequences: voters targeted with vote buying pre-election may receive no government benefits post-election. The results point to obstacles to the democratic transition from clientelist to programmatic forms of electoral competition: parties may not benefit electorally from institutions that increase commitment. Vote buying is a common practice in democracies, but its use varies widely across countries. The fraction of respondents to a global survey who said that vote buying occurred often in their country ranged from 12 percent in the OECD to 56 percent in South Asia; and from 4.3 percent of Dutch respondents to 75.8 percent of Brazilian.1 What explains the large variation across countries in the use of vote buying? The literature on vote buying has mostly focused on the effectiveness of this strategy from the politicians’ perspective, given the costs and reciprocity issues it raises. To account for the emergence of vote buying, the literature emphasizes politician incentives to target benefits to individual voters. From the voter’s perspective, however, a key feature of vote buying is that its benefits are obtained pre-election, circumventing the commitment problems inherent in campaign promises. We argue that the timing of vote buying is critical to understanding its use by politicians.