This thesis studies the effect of information and cultural barriers on the demand for health care in both developed and developing countries. Chapter 1 exploits the randomized research design of PROGRESA, a conditional cash transfer programme implemented in rural Mexico, to study whether health interventions can have indirect effects on the propensity to screen for gender specific diseases. We show that among women not entitled to a cash transfer there is a higher likelihood of being screened for cervical cancer as result of greater social acceptance of the test and increased awareness of potential risk factors. Chapter 2 discusses whether attendance at health and nutrition sessions as one of the requirements for receiving a transfer, affects the health behaviour of eligible adults. Using data from the randomized design of the PAL nutritional programme, implemented in rural Mexico, we show that there is a lower propensity among women for a large waist circumference, which is driven by reduced calorie intake based on better nutrition knowledge. Chapter 3 examines whether the quality of primary care affects the uptake of mammography and colonoscopy among individuals aged 50 plus, in eight European countries. We find that better quality general practitioner are significantly increases the propensity to undergo screening. Finally, Chapter 4 looks at whether the costs involved in acquiring health information are an important determinant of the decision to buy supplementary private health insurance and whether this explains in part the large cross country variation in supplementary private health insurance coverage across European countries. We find evidence that both education and proxies for cognitive ability act as substitutes for quality of health promotion in the propensity to sign a supplementary private insurance.

This thesis develops computational and applied tools to study differentiated products. The core of the thesis focuses on Berry, Levinsohn, and Pakes’s [1995] (BLP hereafter) model of differentiated products. First, I examine how polynomial-based methods for multi-dimensional numerical integration improve the performance of the model. Unlike Monte Carlo integration, these rules produce reliable point estimates and standard errors as well as increasing the accuracy and execution speed of the estimation software. Next, I conduct a large scale simulation study to investigate both the asymptotic and finite sample behavior of the BLP model using the traditional instruments formed from characteristics of rival goods and also supply-side cost shifters, which are necessary for asymptotic identification. The final part of the thesis evaluates the 2003 merger of Morrisons and Safeway by combining a discrete/continuous choice model of demand with census data to construct a geographic distribution of demand. I use this distribution to model the interaction between the location of consumers and stores, focusing on the welfare implications of the merger.

This thesis employs microeconometric methods to understand determinants and effects of individual behaviour relating to educational and migration choices. First, I make use of a rich and little used dataset, from the Learning and Education Achievement in Punjab Schools (LEAPS) project, to study the market for schools in rural Pakistan, drawing lessons for school markets across developing countries whenever possible. The LEAPS includes a unique combination of student (test score), school (quality location, teacher characteristics), and household data, allowing me to estimate rich models of the demand for schools (using standard methods for estimating demands for differentiated products with multiple attributes, widely used in IO). In parallel, since many schools in rural Pakistan are private, it is also possible to study the supply of schools. Second, using the same dataset (LEAPS) I examine the determinants of children's schooling in Pakistan. I estimate a dynamic structural model of school attendance in which the benefits of schooling are uncertain, and the costs of schooling change (probably increase) as the child grows. The model allows for simple dynamic interactions in parental-decision making. One of the main advantages of having a structural model is the possibility of performing policy experiments. In this context I quantify the effect of demand and supply side policies on school attendance. Finally, using US Census data from 1900 to 1930, I investigate how many immigrants in the U.S. adopted the American first names and why they did so, i.e., I study what factors determine name change. A change in the first name from the ethnic name from origin countries to the American name can be viewed as a change in social identity. Also a change in the name can be associated with an intention to assimilate. The early waves of the US Census are especially interesting for this purpose because they include names of respondents. As far as I know, this data has not been used before.

In chapter one I study the welfare optimal allocation of a number of identical indivisible objects to a set of heterogeneous risk-neutral agents under the hypothesis that money is not available. Agents have independent private values, which represent the maximum time that they are willing to queue to obtain a good. I show that a priority list is optimal when hazard rates of the distributions of values are increasing. Instead, queues are optimal in a symmetric setting with decreasing hazard rates. In the second chapter, I study a model in which the use of both market (e.g. auctions) and non-market mechanisms (e.g. lotteries and priority lists) for the allocation of scarce public resources can be rationalized. Agents are risk-neutral and heterogeneous in terms of their monetary value for a good and their opportunity cost of money. The designer wants to allocate a set of goods to the agents with the highest values. The designer screens agents on the basis of their observable characteristics, and extracts information on their willingness to pay using market mechanisms. I show that both market and non-market mechanism can be optimal depending on the prior information. In the last chapter I study a dynamic market model where trade for a single object is bilateral, constrained by an exogenous network structure and conducted under asymmetric information. The model provides a insights into how the position of a player in a network affects his bargaining outcomes. First, traders who provide monopolistic access to valuable portions of the trading network become intermediaries, and obtain a payoff advantage. Second, the earlier an intermediary obtains the object in the trading chain, the higher is his expected payoff. Finally, inefficient outcomes are possible, and are jointly caused by the network structure and by asymmetries of information.

This thesis investigates the effect that market institutions have on economic outcomes such as employment and innovation. The market institutions under study are those that determine the conditions in product, labour and capital markets. Of particular interest is how the effect of institutional changes in one market depends on the conditions in another, or depends on the nature of innovation by the firm. The first chapter describes the matching of patents at the European Patent Office to firm accounts data for all registered firms across fifteen European countries. This constitutes a valuable new dataset for research in innovation that is used for much of the empirical work in this thesis. The second chapter investigates the impact of product market competition on unemployment, and how this depends on labour market institutions. It uses differential changes in regulations across OECD countries to find that increased competition reduces unemployment, more so in countries with strong unions. The third chapter investigates how the effect of product market competition on innovation depends on financial institutions. Using exogenous variation in competition in manufacturing industries this chapter finds that the positive effect of competition on innovation is larger in countries with good financial institutions. The fourth chapter investigates the effect of employment protection legislation on innovation. The theoretical effect of employment protection legislation on innovation is ambiguous, and empirical evidence is thus far inconclusive. This chapter finds that within multinational enterprises overall innovation occurs more in subsidiaries located in countries with high employment protection, however radical innovation occurs more in subsidiaries located in countries with low employment protection.

This thesis employs microeconometric methods to understand determinants and eects of individual behavior relating to educational choice and consumer demand. Chapter 2 studies the intergenerational eects of maternal education on a range of children's outcomes, including cognitive achievement and behavioral problems. Endogeneity of maternal schooling is addressed by instrumenting with schooling costs during the mother's adolescence. The results show substantial intergenerational returns to education. The chapter studies an array of potential channels which may transmit the eect to the child, including family environment and parental investments. The following chapter 3 investigates the eect of studying abroad on international labor market mobility later in life for university graduates. As source of identifying variation, this work exploits the introduction and expansion of the European ERASMUS student exchange program. Studying abroad signicantly increases the probability of working abroad, and the chapter provides evidence on the underlying mechanisms. Chapter 4 compares labor market outcomes between rm-based apprenticeships and full-time vocational schooling alternatives, exploiting the idea that variation in apprenticeship availability aects the opportunities individuals have when they grow up. The chapter documents how variation in vacancies for apprenticeships aects educational choice. The results show that apprenticeship training leads to lower unemployment rates at ages 23 to 26, but there are no signicant dierences in wages. Chapter 5 develops a new approach to the measurement of price responsiveness of gasoline demand and deadweight loss estimation. It uses shape restrictions derived from economic theory to match a desire for exibility with the need for structure in the welfare analysis of consumer behavior. Using travel survey data, the chapter shows that these restrictions remove the erratic behavior of standard nonparametric approaches. Investigating price responsiveness across the income distribution, the middle income group is found to be the most responsive.

In the first chapter, co-authored with Emanuela Galasso, we evaluate an original large-scale intervention in Madagascar (SEECALINE) that focuses on promoting correct breast-feeding, complementary feeding and hygiene practices. We find that the program helped 0-5 year old children, in the participating communities to bridge their gap in weight-for-age z-score and the incidence of underweight. The program also had significant effects in protecting height-for-age and reducing the incidence of stunting. We also show that SEECALINE can have very different effects on the anthropometric status of children, depending on the educational level of the mother. We find that the program improved height-for-age only for children whose mothers had at least secondary level education. We propose an explanation based on interaction effects between proxies of birth conditions and maternal education. More educated mothers meet the necessary conditions that reinforce the behavioral change enabling program effects. We provide evidence that access to public health facilities during birth and early childhood is necessary for translating behavioral change into improvements in children‟s health status. Chapter 1 leaves the question of differential take-up by maternal education unanswered. The heterogeneous effects could be due to lack of adoption of practices by the least educated mothers. In the second chapter I apply difference-in-difference and propensity score weighting techniques to identify causal impact of the program availability on behavioral change and show that least educated mothers adopted the recommended practices. This complements the evidence presented in chapter 1 that although improved knowledge of child-care may be necessary it is not sufficient to translate into improvements in nutritional outcomes. In the third chapter co-authored with Emanuela Galasso and Jeffrey Yau, we estimate the returns to differential lengths of exposure to SEECALINE. We address this question using information available only on program participants. To that end, we develop a methodology that circumvents this data hurdle and estimate returns to differential lengths of exposure using administrative data. We find that the differential returns are decreasing over time, though they do not dissipate to zero. These results provide suggestive evidence that the returns to the program reflect learning effects from the intervention.

Labour markets of developing countries are typically characterised by low unemployment but high informality. In Latin America, about half of the workforce is informal. This includes wage workers without registration and those self employed who do not make social security contributions. Informality is an issue because it has been associated with low productivity jobs and low general human capital investment, despite being a source of employment. The first two chapters focus on extending and estimating labour market models to evaluate the impact of labour market policies on welfare, employment, informality, and wages in developing countries. The first chapter presents a model with search frictions where workers and firms decide whether to be formal or informal. The second chapter assumes firms’ sector is exogenous however allows for workers to become self employed. In common, simulations using these two different frameworks show that increasing the cost of informality has a small impact on unemployment and informality levels. Such policies reduce informal sector wages which are on average the lowest in the economy. Consequently, wage inequality increases. Nonetheless, results show that welfare may improve significantly. The first reason is the increased competition in the formal sector which may occur if firms can choose sector. The second reason is a large increase in formal labour force size because of improved rent-sharing between firms and workers since the latter can enter self employment. In either case, formal wages rise, along with all workers’ welfare. The third chapter provides an empirical test for the impact of enforcement of labour legislation on measures of workers’ welfare. Stricter enforcement increases compliance with mandated benefits (registration, social security and minimum wage). However, there are two tradeoffs, one between the provision of mandated benefits and wages, and another between mandated and important optional job benefits, such as private health.

In this thesis we present a new take on two classic problems of game theory: the "multiarmed bandit" problem of dynamic learning, and the "Colonel Blotto" game, a multidi- mensional contest. In Chapters 2-4 we treat the questions of experimentation with congestion: how do players search and learn about options when they are competing for access with other players? We consider a bandit model in which two players choose between learning about the quality of a risky option (modelled as a Poisson process with unknown arrival rate), and competing for the use of a single shared safe option that can only be used by one agent at the time. We present the equilibria of the game when switching to the safe option is irrevocable, and when it is not. We show that the equilibrium is always inefficient: it involves too little experimentation when compared to the planner solution. The striking equilibrium dynamics of the game with revocable exit are driven by a strategic option-value arising purely from competition between the players. This constitutes a new result in the bandit literature. Finally we present extensions to the model. In particular we assume that players do not observe the result of their opponent's experimentation. In Chapter 5 we turn to the n-dimensional Blotto game and allow battlefields to have different values. We describe a geometrical method for constructing equilibrium distribution in the Colonel Blotto game with asymmetric battlfield values. It generalises the 3-dimensional construction method first described by Gross and Wagner (1950). The proposed method does particularly well in instances of the Colonel Blotto game in which the battlefield weights satisfy some clearly defined regularity conditions. The chapter also explores the parallel between these conditions and the integer partitioning problem in combinatorial optimisation.

My thesis dissertation focuses on the temporariness of migration, its diverse effects as well as on migration selection. The first paper, A Dynamic Model of Return Migration analyzes the decision process underlying return migration using a dynamic model. We explain how migrants decide whether to stay or to go back to their home country together with their savings and consumption decisions. We simulate our model with return intentions and perform policy simulations. The second paper, Remittances and Temporary Migration, studies the remittance behaviour of immigrants and how it relates to temporary versus permanent migration plans. We use a unique data source that provides unusual detail on the purpose of remittances, savings, and return plans, and follows the same household over time. Our results suggest that changes in return plans lead to large changes in remittance flows. The third paper, Savings, Asset Holdings, and Temporary, analyzes how return plans affect not only remittances but also savings and the accumulation of assets. We show that immigrants with temporary return plans place a higher proportion of savings in the home country and have accumulated a higher amount and share of assets and housing value in the home country (compared to the host country). Finally, the fourth paper, Migrant Selection to the U.S.: Evidence from the Mexican Family Life Survey (MxFLS), studies the selection in terms of skills of recent migrants to the United States using the MxFLS. We highlight the important age gradient of migration, the different education attainment between age cohorts in Mexico and show the implications when analyzing migrant selection. Our claim is that in order to properly study the self-selection of migrants, it is necessary to compare migrants to non-migrants of the same age cohort.