La suppression de la sectorisation est-elle une bonne chose ?

School enrolment zoning is the object of a deep debate in France with in the background the virtues of school competition. Towards its objective of social co-education, its assessment can seem quite disappointing. The map of school catchment areas is indeed a relatively ineffective instrument in the urban spaces marked by a strong spatial segregation, which it even tends to increase. Besides, its functioning is strongly unegalitarian (with by-pass strategies implemented by wealthier and/or insiders and screening devices of schools). In this context, the system ends exactly in what against what it tries to fight : a strong polarization of schools, strong disparities and poor performances in internationals comparisons. Its abolition announced for 2010 and the related stake in competition can appear as a means to unbolt the system and to increase average quality. Applied econometrics realized from foreign experiments show however that efficiency gains to be waited are at best modest. On the other hand, competition rather tends to increase inequalities. We develop a model where pupils differ with respect to their parental income and their ability, and where achievements depend on own ability and on peers?abilities. Each school chooses an admission policy to maximize a quality index. We show that abolition of school enrolment zoning will imply a stronger school segregation even more based on academic results. A computational example also demonstrates that this stronger scholastic segregation will probably come with a more pronounced social segregation. All in all, the debate can then seem relatively orthogonal to the challenges which the school has to raise. The crucial questions for the future of the school indeed concern more social mixity, redistribution of the means in the direction of the pupils in academic failure, or still the improvement of the educational spending efficiency.