This paper studies the impact of mergers and acquisitions on prices and markups in the Mexico City retail gasoline market following price deregulation. Using a difference-in-differences design, I find that merging stations significantly increased prices and markups relative to non-merging stations. To understand what drives these changes, I estimate a structural model of spatial competition in which stations are differentiated by location along consumers' commuting routes and firms compete on prices. I use the model to decompose price changes into three channels: increased market concentration, changes in marginal costs, and changes in product quality. My counterfactuals reveal that increased concentration accounts for only a small fraction of the observed increases. Instead, stations that switched upstream suppliers after a merger lowered their wholesale costs, accounting for part of the higher markups. The remaining price increases are consistent with improvements in unobserved station quality. These findings highlight that post-merger price increases need not reflect market power, particularly when ownership changes coincide with other operational changes.

Criminally accused candidates in India win at three times the rate of non-criminals upon nomination, despite evidence that voters penalize criminal records and that criminal politicians worsen governance outcomes. We argue that this pattern is driven by strategic nomination by political parties. We develop a model in which parties trade off a candidate's electoral viability against a reputational cost for fielding a criminal. Because this cost is borne regardless of the election outcome, parties nominate criminals only in competitive districts where candidate quality can be decisive, not in safe seats or hopeless ones. Criminals thus win more upon nomination, not because voters prefer them, but because parties nominate them only when they are expected to tip the race. Using data from four Indian parliamentary elections, we document an inverse-U relationship between district competitiveness and criminal nomination, consistent with the model's predictions.