This thesis considers the role of parents, teachers and peers in the school decision making processes of children at the end of compulsory education in England. This stage represents, in fact, the first and most important school transition when pupils will have to choose whether to enter post-secondary education or not, and in cases where they do, whether to choose an academic course or a vocational one, knowing that this will affect their next transition at the age of 18. This thesis is amongst the ones to most fully analyse the role of significant others in children’s education. All the quantitative analyses in this thesis are done using the Longitudinal Study of Young People in England (LSYPE). Most of the statistical modelling of this data is done using multivariate regression analysis. Some of the results are also based on evidence from qualitative interviews with children in their last year of GCSEs in two comprehensive schools in England and children attending an apprenticeship scheme in the London area. With regards to educational aspirations, minority students are those who show the highest and most stable aspirations during years 9 to 11, while White English working class students, especially boys, have lower and unstable aspirations. Among the explanatory factors for these results, along with social class and ethnicity, parental aspirations, friends’ plans and individual attitude to education have the strongest correlation with the intentions to stay on in school after year 11. Moreover, parental aspirations did not appear to differ with regards to social class, suggesting a different mechanism than the one indicated by Breen and Goldthorpe (B&G) (2000). Also, the fact that minority students have very high aspirations (and are high achievers), do not confirm the principles of the relative risk aversion theory by B&G. Among the types of parental involvement in children’s education, participation in school-related activities and feelings towards school and supervision of children’s school work seemed to have a positive impact on children’s entering A-levels in year 12, although the results did not highlight differences with regards to social class and ethnicity. Evidence from the qualitative interviews showed different results with regards to helping with homework - only educated parents do that – and with regards to supporting and encouraging their children’s aspirations, which is more effective with minority and middle class parents. Considering peer relationships, the evidence from qualitative interviews suggested a very small influence of peers, especially schoolmates, in children’s school decision processes; peers are, in fact, perceived as someone to share plans and common interests with, but not as well-informed and trustworthy sources such as family. Moreover, interviews suggest that school choices are not the results of long-term plans, and children treat school transitions as separate stages. This does not support Morgan’s model of prefigurative and preparatory commitment.

Cette thèse cherche à approfondir la nature et la forme des relations entre les inégalités éducatives et le développement. Elle s’inscrit dans le prolongement des analyses engagées sur les liens éducation/croissance et inégalités/croissance, en essayant d’apporter un éclairage complémentaire sur ces deux relations. Elle vise à retracer de manière stylisée l’évolution des inégalités éducatives particulièrement dans les pays en développement et à caractériser la non-linéarité de la relation à partir de l’estimation de modèles non-paramétriques et semi-paramétriques. Cette thèse est constituée de trois chapitres auxquels correspondent des objectifs, des bases de données et des méthodologies spécifiques. Dans un premier chapitre, nous proposons une nouvelle base mondiale sur les inégalités d’éducation. La majorité des travaux sur la relation entre capital humain et développement économique ont principalement appréhendé la mesure du capital humain à travers des mesures de l’éducation en utilisant notamment la moyenne d’années de scolarisation (stock du capital humain). Notre base de données, qui présente une mesure alternative du capital humain, tend à améliorer sensiblement le mode de calcul des inégalités de l’éducation. Elle exploite toute la richesse des données désagrégées, corrige les pondérations inappropriées et affine certaines hypothèses réductrices sur les durées des cycles d’enseignement et les niveaux d’éducation retenus. Nous avons aussi généralisé la formule proposée par Berthélemy (2006) sur l’indice de Gini de l’éducation. Le domaine de variation possible de cet indice est identifié graphiquement selon la moyenne d’années de scolarisation et les durées cumulées des cycles d’enseignement. Nous mettons en évidence, dans le cadre du chapitre II, l’existence d’une relation non linéaire entre les inégalités dans l’éducation et le développement économique en utilisant des modèles non-paramétriques et semi-paramétriques qui n’exigent pas de formes fonctionnelles prédéfinies à l’avance. Plusieurs phases sont ainsi mises en évidence : les trois premières sont repérées seulement par rapport aux niveaux de développement ; deux autres sont identifiées à la fois par des seuils de développement et d’inégalité d’éducation ; une sixième et dernière phase est définie par rapport au seul niveau d’inégalité d’éducation. Nous montrons que c’est dans la troisième et cinquième phases que la réduction de l’inégalité d’éducation présente l’impact le plus bénéfique sur le développement économique.Au-delà du schéma général mis en évidence sur le plan transnational dans les chapitres I et II, nous explorons dans le chapitre III la nature de cette relation au plan régional dans le cas du Maroc, pour lequel nous disposons de données aux niveaux communal et provincial. La non-linéarité de la relation est aussi confirmée. La troisième phase repérée au chapitre II est subdivisée, dans le cas des provinces marocaines, en deux sous phases qui présentent un impact différencié selon un seuil de développement et d’inégalité d’éducation. / This thesis seeks to deepen the nature and the shape of the relationships between educational inequalities and development. It goes along with the prolongation of the analyses undertaken about the relationships between both education & growth; and inequality & growth, by trying to shed additional light on these two. It aims to retrace, in a schematic way, the evolution of educational inequalities particularly in the developing countries; and also to characterize the nonlinearity of this link using nonparametric and semiparametric estimation models.This thesis consists of three chapters that correspond to specific objectives, databases and methodologies. In the first chapter, we propose a new dataset on the inequalities of education. Most of the studies on the relationship between human capital and economic development have mainly apprehended the measurement of human capital through quantitative education indicators, using namely the average of years of schooling (human capital stock). Our database, which presents an alternative measure of human capital, tends to improve significantly the way in which inequalities in education are calculated. It employs all the abundance of disaggregated data, corrects inappropriate weightings and refines some reductive assumptions about the durations of schooling cycles and the levels of education. We have also generalized the formula proposed by Berthélemy (2006) on the Gini index of education. The possible variation range of this index is graphically identified according to the average years of schooling and the cumulative duration of the schooling cycles. In Chapter II, we reveal the existence of a nonlinear relationship between inequalities in education and economic development using nonparametric and semiparametric models that do not require predefined functional forms. Several phases are therefore highlighted: the first three are identified only according to the level of development; then two other phases are recognized by combining thresholds of both development and education inequality; the sixth and final phase is defined by the educational inequality level alone. We show that the phases during which the reduction of educational inequality presents the most beneficial impact on economic development are the third and the fifth.Beyond the general outline highlighted at the transnational level in Chapters I and II, we explore in Chapter III the nature of this relationship at the regional level in the case of Morocco, for which we have data at both the municipal and provincial levels. We also confirm the nonlinearity of the relationship. The third phase, identified in Chapter II, is divided to two sub-phases in the case of the Moroccan provinces which have a differentiated impact according to a threshold of development and inequality of education.

The thesis deals with the construction of inequalities in the access to the tertiary education with the special emphasis on entrance exams to the universities. The aim is to explore how the entrance exams as the gate to the universities are perceived by students - the actors who are the most affected by these inequalities in education, according to relevant research of these inequalities. The thesis studies how these actors perceive their position and situation within the entrance process, what situations (if any) are constructed as unequal by these actors, who is affected by such inequalities and how does he/she copes with them. The research question is seen from the perspective of interactionist constructivism and answered by applying methods of grounded theory and semi-structured in-depth interviews.