This thesis investigates the impacts and determinants of unequal gender relations within households in Uganda. The first topic addressed is the impacts of intrahousehold bargaining power on female agricultural labour supply in Uganda. There have been relatively few studies in Africa and Uganda in particular on this very important issue—mainly due to lack of appropriate data. I have accessed and used a unique merged cross-section dataset that combines individual women’s information and household level indicators of agricultural production. The 2005/06 Uganda National Household Survey (UNHS) and Uganda Demographic and Health Survey (UDHS) merged cross-sectional dataset for the first time captured indicators of women’s status within households and at the same time captured detailed agricultural data—especially relating to: labour allocation, types of agricultural activities, and nature of crop production—whether cash crops or food crops. Our results indicate that women with greater bargaining power contribute more to agricultural production and any continued gender discriminations is bound to affect agricultural production negatively. With regard to coffee production—Uganda’s leading cash crop, we find that increasing household coffee crop activities are also associated with reducing female labour allocation to agricultural production. The second topic adopts an ordinary least squares and marginal probit approaches in examining the impacts of female bargaining power on multi dimensional indicators of household health in the context of a country with very poor indicators of health status like Uganda. We use a sample of 6,600 children aged less than 5 years and 4,700 married women from two cross sectional surveys in 2000/1 and 2006 to examine how female bargaining power impacts on children’s anthropometric outcomes; children’s use of vaccination services; a woman’s nutritional status; and woman’s use of reproductive health services. We find that bargaining power as measured by an index for a couple’s relational indicators has minimal impacts on either children’s or women’s health status. Instead, it is mainly indicators of household welfare status that matter most—despite the presence of free public health care in Uganda. The third topic of the thesis examines female empowerment, as captured by participation in decision-making in the context of a country with strong patriarchal gender norms, such as Uganda. We use a sample of 9,800 married women from two cross-sectional surveys in 2000/1 and 2006, who were asked, “Who has the final say?” with respect to four major household decisions: the woman’s own health; large household purchases; and daily household purchases; and visiting family or relatives. These data allow us to shed light on whether the characteristics for empowerment are similar across different domains of decision-making and the extent to which cultural norms, as captured by ethnicity, constrain female empowerment in Uganda. We find that graduating from secondary school and ethnicity proxies are significantly related to empowerment in different domains. Furthermore, our results are a robust to choice of empowerment indicator used.

Esse estudo tem como objetivo investigar o comportamento da oferta de trabalho dos cônjuges brasileiros e verificar o grau de adequação desse em relação a um modelo específico dentro da abordagem coletiva (\'collective models\'). O modelo coletivo de oferta de trabalho com fatores distributivos oferece uma estrutura teórica para interpretar o processo decisório intra-familiar e seus resultados, as alocações de consumo e oferta de trabalho das famílias, que são Pareto eficientes por hipótese. Fatores distributivos são variáveis que afetam a decisão sobre oferta de trabalho, mas não tem impactos sobre as preferências nem a restrição orçamentária das famílias. As informações relativas à amostra de famílias brasileiras foram obtidas a partir da PNAD e da Estatística de Registros Civis, ambas para o ano de 2004. Os resultados não rejeitam as restrições derivadas do modelo coletivo, tanto em sua forma geral quanto na versão restrita pela imposição de preferências egoístas. Além disso, as variáveis adotadas como fatores distributivos, sex-ratio e \'participação em divórcios\', influenciam, de forma significativa e condizente com a teoria, a oferta de trabalho de maridos e esposas. / This paper has as objective to investigate the Brazilian spouses\' labor supply behavior and to empirically check the adequacy of a specific collective model. The collective labor supply model with distribution factors offers a theoretical structure to interpret the intra-household decision process and its results, the families\' choices of consumption and labor supply, who are Pareto efficient. Distribution factors are variables that affect the labor supply decision, but do not have impacts on the preference relations nor the budget constraint of the families. The sample of Brazilian families had been gotten from the \"PNAD\" and \"Estatística de Registros Civis\", for the year of 2004. The results do not reject the restrictions derived from the collective model, neither in its general form nor in the egoisti preferences form. Moreover, distribution factors, sex-ratio and \'participação em divórcios\', are found to affect labor supply of husbands and wives in the directions predicted by the theory and to be statistically significant.

Esse estudo tem como objetivo investigar o comportamento da oferta de trabalho dos cônjuges brasileiros e verificar o grau de adequação desse em relação a um modelo específico dentro da abordagem coletiva (\'collective models\'). O modelo coletivo de oferta de trabalho com fatores distributivos oferece uma estrutura teórica para interpretar o processo decisório intra-familiar e seus resultados, as alocações de consumo e oferta de trabalho das famílias, que são Pareto eficientes por hipótese. Fatores distributivos são variáveis que afetam a decisão sobre oferta de trabalho, mas não tem impactos sobre as preferências nem a restrição orçamentária das famílias. As informações relativas à amostra de famílias brasileiras foram obtidas a partir da PNAD e da Estatística de Registros Civis, ambas para o ano de 2004. Os resultados não rejeitam as restrições derivadas do modelo coletivo, tanto em sua forma geral quanto na versão restrita pela imposição de preferências egoístas. Além disso, as variáveis adotadas como fatores distributivos, sex-ratio e \'participação em divórcios\', influenciam, de forma significativa e condizente com a teoria, a oferta de trabalho de maridos e esposas. / This paper has as objective to investigate the Brazilian spouses\' labor supply behavior and to empirically check the adequacy of a specific collective model. The collective labor supply model with distribution factors offers a theoretical structure to interpret the intra-household decision process and its results, the families\' choices of consumption and labor supply, who are Pareto efficient. Distribution factors are variables that affect the labor supply decision, but do not have impacts on the preference relations nor the budget constraint of the families. The sample of Brazilian families had been gotten from the \"PNAD\" and \"Estatística de Registros Civis\", for the year of 2004. The results do not reject the restrictions derived from the collective model, neither in its general form nor in the egoisti preferences form. Moreover, distribution factors, sex-ratio and \'participação em divórcios\', are found to affect labor supply of husbands and wives in the directions predicted by the theory and to be statistically significant.

Debates on whether microfinance remains an effective measure to eradicate poverty and empower women have continued with a bigger question of if an alternative model is available to outweigh the problems of group-based solidarity-based lending scheme. This dissertation aims to study if and how a Lembaga Perkreditan Desa (LPD) – a community owned microfinance – affects women's agency in household and society, and most importantly to observe if it has long term effects on the reconstruction of gendered roles and relations. This study employed participation observation and semi-structured interviews for data collection because each method allowed the exploration of multi layered information and tacit values that other data collection methods do not provide. I spent four months conducting participant observation with female LPD clients from four villages and eight semi-structured interviews around Bali. This study concludes that social capital affects LPD's performance. Impacts of social capital on LPD are posited to occur through the immersion of LPDs into the village governance system that renders members' loyalty, trust, and respect, and the adoption of shared customary laws to name a few. LPD is also proven to strengthen social capital by increasing interdependence among community members; boosting the members' sense of belonging, trust, and responsibility for community development. However, LPD does not necessarily foster women's social capital. This study also found that access to LPD corresponds to women's agency in the household decision-making process. Three features of cooperative decision-making in household are 1) the ability to switch roles in the loan application process, 2) any LPD related financial decisions will involve women's opinion or approval in it, and 3) women have the control over the allocation and repayment of the loans from LPD. Furthermore, I argue that LPD facilitates women's social mobility by ensuring that their access to LPD remains intact. Moreover, LPD causes intergenerational impacts when women are involved in the lending-saving mechanism. Lastly, this study argues that LPD has long-term effects on the reconstruction of gendered roles and relations in Balinese society. / Doctor of Philosophy / This dissertation investigates the impacts of Lembaga Perkreditan Desa (LPD) – a community-owned village bank – to Balinese women from the perspective of four female LPD clients residing in various villages in Bali. In general, the study aims to gain understanding about rural Indonesian women in navigating their opportunities in male-oriented microfinance system. The study found that members of society – female and male – develop a shared understanding, norms, trust, and resources to strengthen their ties among each other called as community capital. My dissertation found that there exists a two-way relationship, instead of one way, between social capital and LPD. Derived from those reciprocal relationship, I argue that LPD may empower women in household and society, though it also may disempower them due to the LPD’s rules. Furthermore, LPD affects women’s capacity to make decisions in their households. These decisions including the ability to choose the right and most practical role when applying for loans and most importantly the capacity to manage the use of loans. LPD corresponds to women’s economic attainment and mobility across villages, yet, it does not correspond to the development of their professional networks. Moreover, LPD creates impacts across generation due to its lending mechanism, payment procedures, and financial benefits from allocating the loans for income generating activities. Lastly, if the LPD transforms the current and future relationships among men and women in household and society? my study found that LPD, through its immersion into the village system and customary values that guide the life of all community members, paves the way for women’s empowerment today and in the future.

Cette thèse étudie la formation des familles, leur organisation et les systèmes d’éducation. Les deux premiers chapitres traitent du choix du conjoint et du partage des ressources au sein du couple. Je modélise tout d’abord comment les individus se rencontrent et décident de se mettre en couple ou non en fonction de leur niveau d’éducation, de leur revenu et leur apparence physique. J’utilise des données américaines où j’observe qui est marié avec qui et pour combien de temps, ce qui me permet d’identifier les préférences des individus. Le deuxième chapitre analyse les effets du mariage sur certaines tendances économiques comme les inégalités de revenu ou les offres de travail. Dans ce chapitre, les personnes en couple se partagent leur revenu et choisissent ensemble la meilleure organisation pour travailler et accomplir les taches domestiques et élever les enfants. Avec des données britanniques, j’identifie les transferts de revenu qui existent entre les hommes et les femmes et je montre que ces transferts augmentent le travail des hommes mariés et diminuent celui des femmes mariées. Les deux derniers chapitres étudient l’efficacité du collège français et notamment la pratique du redoublement. Le troisième chapitre mobilise des méthodes de décomposition pour évaluer dans quelle mesure la baisse des scores des élèves français aux tests PISA peut être attribuée à des changements dans les caractéristiques des élèves ou à des changements dans les rendements de l’éducation. Enfin, le dernier chapitre se concentre particulièrement sur le redoublement et utilise des données de panel sur les collégiens français pour évaluer l’impact du redoublement sur leurs résultats. / This dissertation deals with family formation, family organization and education systems. The first two chapters study how people choose their partners and how they share their income. First, I focus on couple formation and I model how people meet and decide to match or not. People can choose their partner according to their education level, their wage and their physical attractiveness. Using American data, I observe who matches with whom and who stays single and for how long to recover the preferences of individuals in terms of mating. The second chapter attempts to understand how the efficiency and the sorting of the marriage market could impact economic outcomes such as income inequalities or labor supplies. In this chapter, when people marry, they share their income and decide how much each of them will work on the market and at home to raise children or do the housework. Using British data, I recover the amount of monetary transfers which exist between household members and show that these transfers make married women work less on the market and married men work more. The last two chapters of this dissertation focus on the French education system and on the impact of grade retention policies. In the third chapter I use decomposition methods to assess to which extent the decrease in French student’s score at PISA tests can be attributed to the changes in student’s characteristics or to the changes in school returns. Finally, in the last chapter, I use an estimation strategy to get rid of this selection effect and we use a panel data on French High School students to evaluate the impact of grade retention on their scores.

This dissertation deals with consumption in Swedish households between 1913 and 2001. More specifically, it asks whose resources matter most in determining consumption patterns. As a second question, the dissertation also attempts to establish whether the fact that simple covariance between a spouse’s background variables implies that the spouse has any influence at all over the household’s consumption decisions. The theoretical background is mostly drawn from literature regarding intra-household allocations: on the one hand cooperative game theory and on the other hand sociological theory. Cooperative game theory establishes influence, say or power within the household as a function of the marriage’s or cohabitation’s alternative cost, i.e., the difference between the utility level for a married or cohabiting person as opposed to a single person. Sociological theory considers the contribution one makes to the total level of utility in the household, whether in the form of monetary income, household work or as something else. This is in part conceptualized as a difference between exit and voice. The dissertation’s statistical analysis uses three surveys of household expenditure conducted in 1913, 1952 and 1999-2001. They give us an excellent picture of what they actually purchased during that year. The sample sizes are 552, 596 and 3,501, respectively. The dissertation’s main result is that human capital is a previously underestimated determinant of influence in consumption decisions. As the female stock of human capital increases, so does her influence over the household’s consumption decisions. In an attempt to determine the level of democracy within households, the dissertation uses a complementary data source: a questionnaire called “The Swedish People 1955”. Here, one of the questions directed to females was whether they checked with their husbands before deciding on a purchase, as a measure of intra-household democracy. This was then regressed upon the female share of total income, ideological position and two forms of human capital, one general and one for household work. Both forms of human capital lead to democratic households, which is taken to mean that human capital is important not only because it increases labor opportunities in the event of divorce (exit) but also because it increases female voice.

We examine the effect of joint custody on marriage, divorce, fertility and female employment in Austria using individual-level administrative data, covering the entire population. We also use unique data obtained from court records to analyze the effect on post-divorce outcomes. Our estimates show that joint custody significantly reduces divorce and female employment rates, significantly increases marriage and marital birth rates, and leads to a substantial increase in the total money transfer received by mothers after divorce. We interpret these results as evidence against Becker-Coase bargains and in support of a mechanism driven by a resource redistribution that favors men giving them greater incentives to invest in marriage specific capital. (authors' abstract) / Series: Department of Economics Working Paper Series

Une des marques du développement est la résorption du décalage entre la distribution spatiale de l’activité économique et celle de la population. Cette thèse étudie les conséquences d’une telle résorption sous l’effet de la migration ou d’une redistribution géographique de l’activité économique. Le premier chapitre s’intéresse à la relation entre migrants internationaux et foyers d’origine. Il met en lumière l’importance, pour les envois de fonds et les investissements, des croyances des migrants et de l’asymétrie d’information, aggravée par la distance, entre migrants et destinataires de ces envois. Le deuxième chapitre explore la transformation de l’économie d’accueil sous l’effet d’un afflux de migrants venus des campagnes en ville, en Chine. Il mesure d’abord leur impact sur le marché du travail à destination, puis examine comment ils affectent l’allocation des facteurs de production et les contraintes rencontrées par les entreprises sur les marchés du travail et du capital. Le troisième chapitre étudie l’impact sur le long terme d’un vaste programme d’industrialisation réalisé en Chine, et montre un retour de fortune. Ce renversement est imputé aux distorsions introduites sur le marché du travail local par la présence de grands complexes manufacturiers. Ce chapitre met en évidence le rôle de la migration pour surmonter ces imperfections et amener l’étape ultérieure de la transformation structurelle : le passage de l’industrie lourde à la production de biens de consommation et de services. / A key characteristic of the process of economic development is the shrinking spatial mismatch between economic activity and population. This thesis analyzes what happens when this spatial mismatch is reduced, as people’s places of residence and work or the geographical distribution of economic activity is altered. Chapter I deals with the relationship between international migrants and their households of origin. It sheds light on the importance for remittances and investments of migrants' beliefs and the information asymmetry between remittance senders and recipients that distance aggravates. Chapter II explores the transformation of the receiving economy due to an influx of rural-to-urban migrants in China. It first quantifies the effect of immigrants on the labor market at destination, and then investigates their impact on the reallocation of production factors and factor-market constraints faced by urban firms. Chapter III focuses on the long-term impact of a large industrialization plan in China. It reveals a reversal pattern due to the distortions in local labor markets induced by the presence of big plants. It highlights the role of migration in overcoming such imperfections and in bringing about the later stage of structural transformation—from heavy industry to consumption goods and services.

Les conséquences des chocs économiques négatifs sur les ménages ont été documentés extensivement, mais on en sait beaucoup moins sur la manière dont ces chocs sont transmis aux individus à travers la médiation du ménage. Le ménage contribue-il à modérer l'effet des chocs négatifs ? Dans quelle mesure le choc économique pèse-t-il dans la négociation familiale ? À partir de données sur la crise économique argentine de 2001, je montre d'abord que les femmes en couple ont une plus grande probabilité de devenir actives si leur mari a fait l'expérience d'un choc de revenu. Ensuite, je montre que le cycle économique importe dans les décisions d'investissement en capital humain. Sur le long terme, les profils de salaire et d'employabilité des hommes argentins sont affectés de manière persistante par les conditions économiques initiales au moment de l'obtention du diplôme. Enfin, je considère la dimension “man-cession” de la crise économique de 2009 en Espagne et montre que la part des ressources du ménage reçues par les femmes pour leur consommation privée augmente avec la diminution de l'écart des taux de chômage hommes-femmes, confortant l'hypothèse que les chocs négatifs modifient le pouvoir de négociation des individus au sein du ménage. / The consequences of adverse aggregate shocks on households have been repeatedly documented, but far less has been said on the way they are passed over to individuals through the mediation of the household. Does the household contribute in mitigating the effects? Or does the economic shock rather invite itself at the family negociating table? Using the Argentine 2001 economic crisis as a natural experiment, I first show that married women are more likely to enter the labor market if their husband experienced a loss in income, giving credit to the insurance mechanism. Then, I show that the business cycle matters for investments in education, and that long run labor outcomes of Argentine men are persistently affected by the initial conditions upon graduation. Finally, I consider the “Mancession” dimension of the Great Recession in Spain and demonstrate that the resource share accruing to wives for own consumption increases together with the decreasing unemployment gap, which comes in support to the bargaining hypothesis.