Essays on political economy and migration
- Autores: Gustavo José Fajardo Luis
- Directores de la Tesis: Mónica Martínez Bravo (dir. tes.), Manuel Arellano (codir. tes.)
- Lectura: En la Universidad Internacional Menéndez Pelayo (UIMP) ( España ) en 2016
- Idioma: español
- Tribunal Calificador de la Tesis: Guillermo Caruana (presid.), Joan Monràs (secret.), Michael Callen (voc.), Claudio Ferraz (voc.), Noam Yuchtman (voc.)
- Materias:
- Texto completo no disponible (Saber más ...)
- Resumen
As economists seek to improve our understanding of the determinants of institutional dynamics, a new body of work explores the behavior of agents that had typically been marginalized.
As a key actor in many weakly institutionalized countries, the military has started to receive increased attention. In Chapter 1 of this thesis, I study what leads individual members of the military to support ocers that intervene in politics through coup d'etats. More speci cally, I show that coups polarize the military along hierarchical lines.
The formation of political preferences has been subject of some previous research in the economic literature. These studies have looked at the population at large instead of focusing on speci c groups of individuals. Thus, the political preferences of Army ocers is a topic that remains unexplored. The question is intuitively appealing, though, as Army members are unique in their ability to impose their preferences on the whole of society through the use of force. Additionally, learning about the preference formation of ocers can have important lessons regarding the consolidation of elites, the micro-foundations of con ict, an the like.
Another dimension in which the military is distinct is its organizational culture, which stresses discipline and respect for the chain-of-command are de ning values. On the other hand, results in social psychology indicate that hierarchy and power relations shape the way in which people process information and perceive the actions of each other. Therefore, coup d'etats could trigger very di erent responses within the military according to hierarchical relations.
To test this hypothesis, I take advantage of an instance in Venezuelan history where the political preferences of individual members of the Army towards Hugo Chavez became observable. Chavez was a Lieutenant Colonel of the Venezuelan Army when he led a failed coup in 1992, several years before he became President through elections. My measure of political preferences comes from a petition drive launched (in 2003) by the political opposition to the government of Chavez. The purpose of the petition drive was to collect signatures in order to activate a recall vote against him. The list of petition signers was eventually coded into a software called Maisanta, which was made public. I match that information with the roster of Venezuelan Army ocers.
With this data, I estimate the e ect of outranking Chavez in 1992 (the year of the coup) on the probability of signing the petition drive against him. To obtain a causal estimate, I employ a fuzzy regression discontinuity design, where the assignment variable is the graduation year of ocers. The existence of minimum time requirements for promotion in the Army create a binding constraint in the eligibility to outrank Chavez, and this eligibility serves as an instrument for the outrank variable.
In the reduced form results, I observe a clear discontinuity in the probability of signing the petition across cohorts. Ocers graduating up to 1973 display higher signing rates than those graduating in 1974 or more recently. The 1973-1974 threshold coincides precisely with the eligibility constrain I mentioned. Individuals who graduated up to 1973 were eligible to outrank Chavez in 1992, while younger ones were not. Indeed, in my full speci cation, I observe that outranking Chavez increased the probability of signing the petition by approximately 30 percentage points.
I argue that the mechanisms driving my results are of a behavioral nature. Hierarchical relations a ect the attitudinal responses to the coup, which translate into preferences towards the political agenda advanced by the insubordinates. An alternative explanation based on economic incentives would argue that, at the time of the petition drive, post-1973 ocers expected higher returns from Chavez's continuation in power. I test this by gathering additional data on appointments of Army ocers during the rst period of Chavez's administration (between the moment when he assumed oce and the petition drive). I compare the ocers who outranked Chavez (during the coup year) and those who did not and nd that the latter group received no preferential treatment. If anything, those who outranked him had slightly higher appointment rates. This is inconsistent with an explanation of the main results based on di erential economic incentives, and thus favors the baseline behavioral mechanism.
In Chapter 2 of this thesis I turn my attention from non-electoral to electoral politics. I study the e ects of term limits, i.e. rules that establish a cap in the number of times a given individual is allowed to serve in a given oce, on politicians.
Previous work on this area has already established empirically a causal link between term limits and the behavior of elected ocials. However, the existing research has focused on the electoral accountability e ect of term limits. That is, by forbidding reelection, the incentives to act in accordance to the interests of the electorate disappear for term-limited ocials. The electoral accountability e ect predicts that the same individual will behave di erently when the term-limit binds vis-a-vis when it does not bind. Evidence in favor of this prediction has been found in several instances. A noteworthy example of this is Ferraz and Finan (2011), who show that Brazilian mayors misallocate more resources when they are term limited.
Another e ect of term limits is that they reduce the value of holding a given oce, since they restrict the total rents that individuals can extract. Although the existence of this value-of-oce e ect is generally accepted, empirical assessments of its consequences are much more scarce. Diermeier et al. (2005) calibrate a model of congressional careers in the US and conduct a policy experiment of introducing term limits. They nd that this would have a sizable negative e ect on the value of congressional seats, leading to a signi cant increase of voluntary exit from oce by congressmembers. Smart and Sturm (2013) provide a theoretical framework to reconcile the negative consequences of the electoral accountability e ect with the popularity enjoyed by term limits. In their model, the value-of-oce e ect plays a central part. They argue that lower value of holding oce leads to \truthful" behavior by incumbents which, in turn, facilitates their screening.
The reduction in the value of oce could a ect individuals who aspire to hold the termlimited oce. Chapter 2 uses observational data to test that hypothesis. Namely, I assess whether the bill sponsorship activity of members of the US House responds to changes in the term limits of the Gubernatorial oce of the state they represent. A novelty of this study is exploring whether the e ects of term limits go beyond oceholders themselves and reach other political actors. As motivating evidence, I show i) that the US House works as a stepping stone towards the Gubernatorial oce, and ii) that being successful at passing legislation is electorally relevant. Regarding the main hypothesis, the results indicate that imposing tighter term limits for the Gubernatorial seat decreases the bill sponsorship activity of representatives. These ndings suggest that politicians do internalize changes in the value of a given oce even when they don't currently hold it, but aspire to hold.
Finally, Chapter 3 is devoted to the topic of international migration. In joint work with Emilio Gutierrez and Horacio Larreguy, we investigate how household with migrants abroad respond to shocks in destination labor markets.
There is growing interest in the role that remittances can play in order to leverage families out of poverty. Consequently, economists have devoted e orts to study the ways in which remittance ows a ect the economic decisions of origin households. A lot of attention has been put in trying to decipher what are the main uses families give to these international income ows: consumption, small business investments, agricultural investment, schooling, etc. However, the literature so far has not been conclusive. Some authors have also explored the way in which labor market conditions at destination correlate with the labor force participation of the family members who remain in the origin households.
In Chapter 3 we turn to analyzing the e ect of labor market shocks at destination on subsequent migration decisions, which we assume to be made at the household level. We start by providing a theoretical framework where families have to choose how to distribute their labor force between the foreign and domestics labor markets, while facing a minimum consumption requirement. We show how a negative show in foreign labor income can trigger di erential responses across income levels. For poorer families, the income e ect generated by the shock is large and they need to compensate the income loss by sending additional migrants to the foreign market (which continues to be more attractive than the domestic one). High-income families, in turn, respond by reducing their migration rate.
Our empirical analysis is done using Census data for Mexico, where a non-negligible fraction of families have members working in the US. We divide Mexican households in income quintiles and estimate the e ect of unemployment changes in the US on subsequent migration decisions. The pattern of the estimates is largely consistent with the intuition of the theoretical framework. For low-income families, the negative shocks in the US increase the probability of sending additional members abroad. The opposite occurs for higher-income households.
Our ndings are also informative to the branch of the literature that studies the selection of international migrants. Namely, we argue that the labor market conditions in receiving communities have a direct impact on familias in sending communities. But the impact is proportionally higher for lower-income households, which shapes the composition of subsequent migrant ows. Considering this interaction beteen destination labor market shocks and pre-existing stock of migrants expands the focus of previous work which sees selection patterns as a function of di erences in expected wages and migration costs.
