Ce travail de thèse a pour objectif d’étudier l’impact de l’introduction de la taxation des héritages sur la croissance et l’offre de travail, en considérant la diversité des transferts familiaux intergénérationnels. En effet, la transmission familiale peut être éducative, culturelle, patrimoniale, ou encore perçue comme un transfert en temps. Toutes ces formes de solidarités familiales génèrent des externalités, qui impactent différemment la croissance et l’offre de travail, ce qui peut affecter l’efficacité des politiques fiscales. Ainsi, l’impôt successoral réduit l’incitation à épargner mais peut accroître l’investissement éducatif ou les transferts en temps, ce qui peut affecter positivement la productivité des ménages et l’offre de travail. Nous développons ici des modèles théoriques à générations imbriquées avec altruisme envers les descendants. La thèse est composée de trois chapitres. Le premier chapitre permet d'étudier l’impact de la non-disponibilité de la dette publique sur la politique de redistribution intergénérationnelle mise en place par le gouvernement, en utilisant uniquement l'impôt sur les revenus du travail et l'impôt successoral. Il permet aussi d’analyser son effet sur la croissance économique et les transferts familiaux intergénérationnels, consistant en des legs et des dépenses d’éducation, en mettant en évidence le rôle central de la taxation de l’héritage. Le second chapitre propose un modèle avec legs et transferts de temps descendants, dont l’objectif est de montrer les différences entre la taxation de l’héritage et la taxation du capital de cycle de vie, sur le comportement des ménages. Nous montrons que l’utilisation de la taxation de l’héritage à la place de celle du capital peut être une reforme Pareto-améliorante, en fonction de l’effet de la réforme sur l’offre de travail. Enfin, le troisième chapitre s’intéresse aussi à la comparaison entre taxation du capital et taxation de l’héritage, dans un modèle où les dynasties sont différentes en termes de productivité et de niveau d'altruisme. Ce chapitre démontre qu’appliquer l'impôt successoral à la place de celui du capital, peut améliorer à long terme, le bien-être des moins altruistes et, dans certains cas, peut être Pareto-améliorante, si les ressources disponibles pour les plus altruistes augmentent avec la réforme. / This thesis analyzes the impact of inheritance taxation on growth and labor supply, considering the diversity of intergenerational family transfers, such that bequests, parent's education spendings or time transfers. These forms of family solidarity generate externalities, which impact growth and labor supply, and affect the effectiveness of tax policies. Concerning inheritance tax which reduces the incentive to save, it can also increase educational investment or time transfers, which can positively affect household productivity and labor supply. For this purpose, we use overlapping generations models with altruism towards offspring. The thesis is divided into three chapters. The first chapter studies the impact of public debt on intergenerational transfers and on human capital growth, using a simple tax structure with labor and bequest taxes. In this model, parents augment their children's income through education and bequest. When public debt is not available, we show that the long run growth is higher thanks to an increase of the gap between the two taxes, which underlines the role of inheritance taxation. The second chapter proposes a model with rational altruism textit{`a la} Barro, where time transfers and bequests are available to parents. We analyze a shift from capital income tax towards inheritance tax, leaving constant the capital labor ratio. We show that this reform may increase welfare of all generations. Welfare improvement mainly depends on the effect of the reform on the labor supply. This tax reform is also implemented in the third chapter where we consider that dynasties differ in productivity and altruism. We show that the tax reform increases the welfare of less altruistic dynasties but decreases welfare of the most altruistic one. Extending the model with time transfers and elastic labor supply, we identify situations where the tax reform is Pareto improving.

Master of Science/Doctor of Clinical Psychology / Figures from the Australian Bureau of Statistics show that over 40% of all Australian children moved at least one time in the census period from 1996 to 2001 (ABS, 2001). The literature varies in the impact that this has on children. The purpose of this study was to examine the associations between residential relocation, resilience and the emotional, behavioural and academic adjustment of children 8-12 years of age who had moved. Risk factors as identified in the literature as well as the relative impact of resilience were examined. By studying how adjustment occurs in the context of resilience, possible areas for prevention and intervention may be developed for the large numbers of children who move. Results showed that the sample population was in the normal range in academic and behavioural terms. The sample was found to have repeated more grades than average; however the children did not exhibit significant behavioural or emotional consequences. A number of demographic factors have been indicated in the literature as affecting adjustment after residential relocations, yet these were generally not found to be significantly associated with adjustment for this study population. Socioeconomic status was the only factor other than resilience to have been significantly associated with adjustment. Possibly due to the developmental stage of the participants, only the resilience subscales of interpersonal strength and school functioning were found to be significant in their positive association with adjustment, leading to fewer behavioural and academic problems. While the children in this study have all had the potential stress of moving house, the demographic characteristics of this sample would suggest that they might not have had to encounter multiple life challenges or adversities. This conclusion may help explain the lack of significant effects of demographic factors on the adjustment of the children in this sample. Results highlight the importance of good schooling and that the core business of schools in building and enhancing the intellectual functioning of children, is a vital component in the development of resilience. These findings suggest that different aspects of resilience may be important for different developmental stages and different life stressors. The distinction between cause and effect when examining resilience factors is discussed and it is suggested that outcomes in one context may be treated as influences upon outcomes in another context.

Master of Science/Doctor of Clinical Psychology / Figures from the Australian Bureau of Statistics show that over 40% of all Australian children moved at least one time in the census period from 1996 to 2001 (ABS, 2001). The literature varies in the impact that this has on children. The purpose of this study was to examine the associations between residential relocation, resilience and the emotional, behavioural and academic adjustment of children 8-12 years of age who had moved. Risk factors as identified in the literature as well as the relative impact of resilience were examined. By studying how adjustment occurs in the context of resilience, possible areas for prevention and intervention may be developed for the large numbers of children who move. Results showed that the sample population was in the normal range in academic and behavioural terms. The sample was found to have repeated more grades than average; however the children did not exhibit significant behavioural or emotional consequences. A number of demographic factors have been indicated in the literature as affecting adjustment after residential relocations, yet these were generally not found to be significantly associated with adjustment for this study population. Socioeconomic status was the only factor other than resilience to have been significantly associated with adjustment. Possibly due to the developmental stage of the participants, only the resilience subscales of interpersonal strength and school functioning were found to be significant in their positive association with adjustment, leading to fewer behavioural and academic problems. While the children in this study have all had the potential stress of moving house, the demographic characteristics of this sample would suggest that they might not have had to encounter multiple life challenges or adversities. This conclusion may help explain the lack of significant effects of demographic factors on the adjustment of the children in this sample. Results highlight the importance of good schooling and that the core business of schools in building and enhancing the intellectual functioning of children, is a vital component in the development of resilience. These findings suggest that different aspects of resilience may be important for different developmental stages and different life stressors. The distinction between cause and effect when examining resilience factors is discussed and it is suggested that outcomes in one context may be treated as influences upon outcomes in another context.

[pt] Em 2003, o Congresso Nacional aprovou uma lei regulamentando o crédito consignado, o que causou um grande impacto no crédito a pessoas físicas. No capítulo 1, forneço evidências que o crédito consignado teve um efeito pequeno, mas não desprezível, nas escolhas ocupacionais. Em particular, eu encontro evidências que o crédito consignado está associado positivamente com escolhas ocupacionais com características de empreendedorismo. Esse efeito é mais forte em indivíduos que estão numa faixa etária onde dificilmente será um aposentado e residem em famílias com aposentados e pensionistas, o que sugere a existência de mecanismos de transferência intra-familiares. No Capítulo 2, eu apresento evidências de que o crédito consignado tem um efeito pequeno, mas não desprezível, nos indicadores de saúde dos idosos. Em particular, o crédito consignado está: 1) negativamente associado com as taxas de mortalidade nas faixas etárias de 60 a 69 anos e 70 a 79 anos. 2) negativamente associado com indices de dificuldades físicas. Esse efeito é maior em famílias com menor renda. 3) positivamente associado com um número de consultas médicas por ano. Usando uma abordagem diff-diff, encontro evidência de que um possível mecanismo é um aumento de gastos com saúde nas famílias com maior proporção de ganhos oriundos de aposentadorias nas suas rendas. / [en] In 2003, the Brazilian Congress passed a law regulating payroll lending, which had a large impact on consumer lending (Coelho et al [2011]). In Chapter 1, I present evidence that that payroll lending had a small but non-negligible impact on occupational choice. In particular I find that payroll lending is positively associated with more entrepreneurial-like occupations. The effect is stronger for individuals in age ranges unlikely to have retirees in families with retirees or pensioners, suggesting that intra-family transfer mechanisms are operative. In Chapter 2, I present evidence that that payroll lending had a small but non-negligible impact on health outcome for elderly citizens. In particular, we find that Payroll Lending: 1) is negatively associated with death rates on age ranges of 60-69 and 70-79 years old; 2) is associated with less physical disabilities. The effect is stronger for individuals in families with lower incomes; 3) is associated with more medical consultations per year. Using a diff-diff approach, I also find evidence that one possible mechanism is more health care spending in families with more proportion of retirement benefits in their income.

Alors que la question du vieillissement de la population dans les pays d’Afrique subsaharienne ne semble pas être particulièrement préoccupante pour l’instant, les conditions de vie des personnes âgées s’imposent graduellement comme un enjeu majeur de prise en charge sociale au regard de la croissance rapide de leur effectif. Les systèmes formels de retraite, d'assurance maladie et de protection sociale destinés aux personnes âgées vivant dans la plupart de ces pays demeurent quasiment inexistants, malgré la ratification et l’adoption de plusieurs conventions et recommandations internationales visant l’amélioration de leurs conditions de vie. Compte tenu de ces faiblesses, la famille est très souvent présentée comme la principale source de soutien des personnes âgées dans le besoin à travers leurs arrangements résidentiels, notamment la cohabitation avec d’autres personnes (conjoint, enfants adultes ou autres parentés), mais aussi par le biais des transferts financiers ou matériels qu’elles reçoivent en provenance de la famille qui vit en dehors de leur ménage. L’objectif principal de la présente thèse est de contribuer aux connaissances actuelles sur les conditions de vie et de santé des personnes âgées en examinant leurs modes de cohabitation, les transferts de ressources qu’elles pourraient recevoir du réseau familial externe au ménage, les liens entre ces deux formes de soutien et leurs corrélations avec l’état de santé des personnes âgées vivant en contexte d’Afrique subsaharienne (le Sénégal). Afin d’atteindre cet objectif, trois articles scientifiques empiriques ont été rédigés en utilisant d’une part les données des trois derniers recensements de la population du Sénégal (1988, 2002 et 2013) et d’autre part, la deuxième Enquête Pauvreté et Structure Familiale au Sénégal réalisée en 2011. Le premier article propose un portrait des arrangements résidentiels des personnes âgées notamment les situations où elles vivent avec/sans leur conjoint et/ou leurs enfants adultes selon le sexe et examine leur évolution à travers le temps. Les résultats suggèrent que la réalité du vieillissement est très différente pour les hommes et pour les femmes. Les hommes sont plus susceptibles de vivre accompagnés par leur conjointe et au moins un enfant adulte tandis que les femmes âgées sont beaucoup plus susceptibles de vivre sans conjoint mais avec un enfant adulte pour des raisons de veuvage, de séparation ou de divorce. En revanche, vivre seul tout comme vivre à la fois sans conjoint et sans autre adulte demeure un mode de cohabitation très marginal chez les personnes âgées au Sénégal, quel que soit leur sexe. Par ailleurs, les résultats ont aussi permis de constater qu’au fil des récentes décennies, il n’y a pas eu de changements majeurs dans les modes de cohabitation des personnes âgées, bien qu’on note des prémices d’une augmentation de la proportion des hommes qui vivent seuls ou sans conjoint ni adulte notamment chez les générations plus récentes. Le second article vise à améliorer notre compréhension des mécanismes de soutien familial dans le contexte des pays en développement en examinant dans quelle mesure les modes de cohabitation des personnes âgées sont associées à la réception de transferts financiers ou matériels provenant de la famille qui vit en dehors de leur ménage. Les résultats mettent en évidence un avantage net des femmes à recevoir des transferts familiaux nets positifs par rapport aux hommes pour certains arrangements résidentiels. Il ressort également que le fait de vivre sans mari ni enfant adulte augmente significativement la probabilité pour les femmes âgées de recevoir un soutien de la part de membres de la famille non-cohabitants, comparativement à celles qui vivent à la fois avec un conjoint et un enfant adulte plus jeune. Toutefois, ces différences ne se sont pas révélées significatives chez les hommes âgés. Dans le cadre du troisième article de la thèse, nous avons examiné dans quelle mesure la composition du ménage des personnes âgées et les transferts qu’elles reçoivent de la famille hors ménage sont associés à leur état de santé. L’utilisation d’une stratégie analytique prenant en compte des potentiels biais d’endogénéité a permis de constater que les femmes âgées qui vivent sans conjoint et sans adulte ont une probabilité plus élevée de déclarer une moins bonne santé que celles qui vivent avec l’un ou l’autre, ou les deux. En revanche, les différences en matière de santé perçue selon les types d’arrangements résidentiels ne se sont pas révélées significatives chez les hommes. Par ailleurs, il existe un désavantage en matière de perception de la santé chez les femmes qui n’ont pas reçu de transferts de la famille hors ménage, mais chez les hommes le rôle de ces transferts reste mitigé. Globalement, les résultats de cette thèse suggèrent que les gouvernements devraient tenir compte des arrangements résidentiels des personnes âgées et du soutien du réseau familial externe à leur ménage afin d'optimiser l'impact des politiques et des interventions sur leur santé. / While the issue of population aging in sub-Saharan African countries does not seem to be of particular concern at present, the living arrangements of the older people are gradually becoming a major social issue given the rapid growth of their numbers. Generalized formal pension, health insurance and social protection systems for the older people living in most of these countries remain almost non-existent, despite the ratification and adoption of several international conventions and recommendations aiming to improve their living conditions. Given these limitations, the family is most often presented as the main source of support for older people in need through their living arrangements, including cohabitation with others (spouse, adult children, or other relatives), but also through the financial or material transfers received from family living outside their household. The main objective of this thesis is to contribute to the current knowledge on the living and health conditions of the older people, by examining their living arrangements, the resource transfers they might receive from the family network outside their household, the links between these two forms of support and their correlations with the health status of older people living in a sub-Saharan African context (Senegal). To achieve this objective, three empirical scientific articles were written using data from the last three population censuses in Senegal (1988, 2002, and 2013) and from the second Poverty and Family Structure Survey in Senegal conducted in 2010-2011. The first article presents a description of the residential arrangements of older people, especially situations where they live with their spouse and/or adult children, according to gender, and examines their evolution over time. The results suggest that the reality of aging is very different for men and women. Men are more likely to live accompanied by their spouse and at least one adult child, while older women are much more likely to live without a spouse but with an adult child due to widowhood, separation, or divorce. On the other hand, living alone as well as living without a spouse and without another adult remains a very unusual for both older men and women in Senegal. The results also show that over recent decades, there have been no major changes in other cohabitation patterns among the older adults, although there are signs of an increase in the proportion of men living alone or without a spouse or adult, especially among the more recent generations. The second paper aims to improve our understanding of family support mechanisms in the context of developing countries by examining the extent to which older people's living arrangements are associated with the receipt of financial or material transfers from family living outside their household. The results show a net advantage of women over men in terms of receiving positive net family transfers for some residential arrangements. It also shows that living without a husband or adult child significantly increases the likelihood that older women will receive support from non-cohabiting family members, compared to those living with both a spouse and a younger adult child. This pattern, however, was not significant for older men. In the third article of the thesis, which examines the extent to which older adults' household composition and the transfers they receive from non-household family are associated with their health status, it was found that older women who live without a spouse or other adult have a higher likelihood of reporting poorer health than those who live with one or both. In contrast, differences in self-rated health by type of living arrangement were not found to be significant for men. In addition, there was a disadvantage in self-rated health for women who did not receive non-household family transfers, while for men the role of such support was less clear. Overall, the results of this thesis suggest that governments should take into account older adults' living arrangements and non-household family support in order to optimize the impact of policies and interventions on their health status.