A One-shot Deal on the Spot: How Vote Buying Affects Electoral Behaviour. Experimental Evidence from Mexico

For the 2012 Mexican Presidential Elections, about 50 million voters went to polls to elect more than two thousand posts. The runner-up attributed the defeat to a massive vote-buying mobilisation in favour of the front-runner. Reports from electoral observers supported that version. Did vote buying modify voters choices? Although the literature has approached vote buying from several angles, there remain disputes and gaps in our understanding of the mechanisms involved and their direct implications for electoral outcomes. In this dissertation, I assess both, asking i) how are Mexican voters confronted by vote-buying strategies, ii) what mechanisms for targeting and buying votes do parties deploy, and iii) how strong are the effects on voting choices. First, I propose an extended two-stage model of vote-buying mobilisation to frame the analysis and to resolve conflations and confusions in previous research. Second, I employ a mixed-methods research design, analysing thousands of phone calls reporting vote-buying to a national hotline service, a series of semi-structured interviews with brokers, and a list experiment embedded in a nationally representative survey in Mexico. Qualitative evidence from calls and interviews confirm the two-stage model: that activists begin to target voters long time before polling days by knocking on doors, proffering rewards as an exchange for votes and compiling lists of electors. Near and during polling days, activists conduct the second mobilisation strategy to monitor voters and ensure compliance by distributing benefits broadly across the country. Survey evidence shows that 15% of those electors switching voting choices near polling days were contacted by activists during the Election Day, which suggests that further research on vote buying should be more attentive to the timing of the exchange. This research contributes to the literature on vote buying in three ways. First, it extends theoretical approaches of models of vote-buying mobilisation. Second, it provides qualitative evidence from both citizens and brokers to understand mechanisms of targeting and buying votes. Third, it highlights some indirect questioning strategies -including but not confined to the list experiment- that are helpful for estimating vote-buying.