UC Berkeley

The chapters in this dissertation study political economy and development economics topics related to the decline of the Mexican Institutional Revolutionary Party (PRI), one of the longest-lasting authoritarian governments of the twentieth century. Chapter 1 provides an introduction linking the main topics, hypotheses, and results. Chapters 2 and 3 focus on the role of mass media diversity and unsustainable media capture, respectively, in the Mexican democratic transition. Chapter 4 examines how fractured political power across levels of government as a result of the collapse of the PRI centralized state, led to higher violence levels from the war against organized crime launched by the National Action Party (PAN) in 2007. Chapter 5 concludes with a summary of the most important findings and contributions.

Using a unique panel dataset that provides local broadcast media coverage and ownership data for each of the 1,556 radio and broadcast television outlets in the country from 1990 to 2012, Chapter 2 studies the effect of media diversity on the PRI and opposition parties' electoral performance as well as on turnout, and shows how local media diversity, particularly in the radio market, contributed to the Mexican PRI authoritarian regime's radical municipal electoral decline. Conditional on time-varying observables and controlling for municipal and year fixed effects, the chapter develops three main sets of results. First, I show that increases in local media diversity, particularly from the local radio market, had a large significant negative effect on mayor municipal voting outcomes for the PRI and a significant positive effect on the electoral performance of the left of center opposition Party of the Democratic Revolution (PRD). Second, I show that both local radio and broadcast television plurality had a positive effect on turnout, that local broadcast television ownership diversity had a negative effect on turnout, and that media exposure matters more for electoral participation than ideological diversity. Third, my analysis shows that the most popular measure of media diversity used in the literature, media plurality, is an incomplete measure of diversity and/or competition. Ownership is of central importance when studying media's effects on voting behavior. Results hold after controlling for overall media exposure, democratization trends, and turnout and are robust to different measures of media diversity, PRI electoral outcomes, as well as an alternative ownership concentration measure.

Chapter 3 develops an extension of Besley and Prat's (2006) canonical media capture framework with a three-period political agency retrospective voting model to understand how new media licenses are granted, when media capture occurs, and the effects it has on political outcomes and voters' welfare. The model shows that it is more costly to capture media when media have a higher commercial motive, when there are regulatory structures that make bribing harder, when the initial number of free media is higher, when there is lower expected media loyalty, and when the cost of rebribing media is higher. It also shows that transparency and efficient news production influence the cost of media capture indirectly, and that media independence, initial media plurality, and media concentration have an ambiguous effect on the cost of capture. Moreover, the optimal number of new licenses and of outlets bribed in period 1 are both functions of the cost of bribing in period 1 relative to the cost of bribing in period 2. The optimal number of outlets captured in the second period is ambiguous, and suggests that the extent of capture when license-granting is possible may be context-specific and needs to be evaluated empirically. In addition, the theoretical results show that the equilibrium with unsuccessful media capture in period 2 yields higher audience-related revenues, turnover, and voter welfare than successful media capture in both periods and lower audience-related revenues, turnover, and voter welfare than without media capture in both periods. The model provides an adequate framework to study license-granting as a additional means of media capture and suggests that media freedom regulatory frameworks, market incentives, and limited direct government ownership, may not be enough to contain capture.

Chapter 4 investigates one of the many consequences of the Mexican democratic transition studied in Chapters 2 and 3: institutional coordination failures. The collapse of a centralized state meant that the federal government was no longer able to ensure cooperation from local governments, a key factor to ensure the correct implementation and effectiveness of the war against organized crime launched by the PAN administration at the start of 2007. The chapter studies the role of coordination between federal and state governments in containing violence from the war against organized crime. Within a municipal and year fixed effects framework and controlling for various socioeconomic and demographic characteristics, the empirical analysis exploits an exhaustive dataset from 2006 to 2011, and finds that the lack of coordination among levels of government, measured with party alignment, had a significant positive effect on violence. The chapter also surveys various theories that help explain the increase in violence related to organized crime in Mexico, it studies the reasons why political coordination is decisive in the effectiveness of the fight against organized crime, and discusses public policy implications.