We study how party discipline varies with electoral mandates and career concerns. In our model, partisan benefits induce discipline insofar as leaders rally enough support among their legislators to make their promises credible. Absent reelection, constituency concerns offset the leader’s promises, leaving district legislators less disciplined than their list copartisans. Reelection makes these benefits electorally valuable, directly improving district legislators’ local valence and list legislators’ renomination prospects, and indirectly strengthening the leader’s credibility in both tracks. We test the empirical implications of the model in Mexico’s mixed-member Congress around the reform introducing consecutive reelection. We measure disagreement with party leaders in votes and speech across seven Congresses (2006–2024): disagreement falls in both tracks after the reform, most among district incumbents. Leaders, in turn, redirect partisan benefits toward district incumbents ex ante and reward the disciplined while punishing dissenters ex post. Accountability reforms that leave ballot access, list placement, and campaign finance in party hands can strengthen the organization they were meant to check.

Governments, whether autocracies or democracies, seek to influence the media. Autocracies use strategies such as propaganda and censorship, but these tools are limited in democracies. In this paper I show how governments can still influence the media through political favoritism. By reallocating advertising toward aligned outlets, a government enables them to invest in their newsrooms and raise their news production. Combining the universe of federal advertising contracts in Mexico with the full text of national print media, I use the 2018 transition from a right-wing to a left-wing administration and estimate a difference-in-differences specification. Outlets that share the incumbent’s ideology become more likely to receive a contract. These allocations reflect favoritism and not the audiences a government would need to reach, whether the demographic a campaign names or the incumbent’s own voters. Aligned outlets then hire more journalists, publish more articles, cover more news events, and report more facts in each. Their circulation also rises relative to misaligned outlets. Favoritism toward competing firms therefore reallocates productive capacity across a market instead of transferring rents within it, a mechanism of political control that operates through market competition.

Does broader political representation improve local governance? I use population thresholds in Mexican state law that fix the size of a municipal council, comparing municipalities just above a threshold with those just below, and link novel data on the characteristics of every council member to the output that council produces in office. Crossing a threshold seats 12 percent more representatives, and they are negatively selected: councils gain more members with a basic level of education and with no prior political experience, while parity rules hold the share of women fixed. Governance does not improve. Larger councils approve about half as many initiatives, a third fewer per member, across a narrower agenda, and their alignment with the preferences of citizens decreases; spending is unchanged in level and composition. I interpret this as negative selection at the margin of the party list: the supply of candidates is smooth at the threshold, but lists are closed and ranked, and parties concentrate their strongest candidates where the return to quality is highest.