Indicators of social segregation between colleges in departments

The entropy index is used by the Ministry of National Education to measure social segregation between colleges, i.e. disparities according to the social composition of the students they host. This file presents the entropy indices and their breakdown according to the education sector (public/private under contract) at different geographical levels: national, academic, departmental. Data are calculated at different geographical levels: at national level (academy label=“whole France”), this field covers metropolitan France and DROMs; \- at the level of each academy (department label = “Together of the academy”); \- at the level of each department; Data from academies that include only one department appear only at the level of the department concerned (example: Paris). Students are categorised according to their manager’s PCS. For couples, priority is given to the father’s PCS, even if it is the mother who has been identified as the primary manager. PCS are grouped into 4 categories: − highly favoured: entrepreneurs with ten or more employees, senior managers and intellectual professions, teachers, school teachers; − favoured: intermediate professions (except school teachers), retired executives and intermediate professions; medium: farmers farmers, craftsmen and traders (and corresponding pensioners), employees; disadvantaged: workers, retired workers and employees, unemployed who have never worked, inactive other than retirees. Uninformed PCS are grouped together with disadvantaged PCS. The calculations in the fields “Public College Entropy Index”, “Private College Entropy Index”, “Public-Private College Entropy Index”, “Public Colleges’ Contributions to the Entropy Index”, “Contributions of Private Colleges Contracted to the Entropy Index” and “Contributions of Public-Private Gaps to the Entropy Index” exploit the fact that the Entropy Index is decomposable. Example of reading: in the Alpes-de-Haute-Provence, in 2014, social segregation between colleges, as measured by the entropy index, was 0.039 (0.036 if we consider only public colleges). 83 % of social segregation could be attributed to segregation within public colleges, 3 % to that existing in private colleges, and finally 14 % came from differences in the diversity of social categories represented in the public and private sectors, taken as a whole. **Methodology on the construction of the college field for study ** The field covers public and private colleges under contract. Some colleges are excluded from the entropy calculation for a given school year. This is the case for some of the colleges that have an unencumbered CSP rate greater than 25 % in a given year N, 2 cases: 1- The previous year (N-1), the rate was less than 25 %: in this case, the data for year N are counted. The college is retained in the calculation of the entropy index with its imputed data. Method of imputation: the distribution according to the CSP of year N-1 is applied to the number of pupils in year N. Example: number of students for the college in year N = 600 with the uninformed PCS for more than 25 % of them. It is observed that in the year N-1, the number of pupils was divided as follows: Highly favoured = 22 %, Favored = 13 %, Average = 27 %, Disadvantaged = 33 %, Uninformed = 5 %. The staff for year N shall be charged as follows: Highly favoured = 22 % * 600 = 132 students, Favored = 13 % * 600 = 78, Average = 27 % * 600 = 162, Disadvantaged = 33 % * 600 = 198, Not informed = 5 % * 600 = 30. 2- The previous year (N-1), the rate was also higher than 25 %: in this case, the college is excluded from the calculation of entropy for year N. **References ** For details on the measurement of segregation and in particular the properties of the entropy index, see: Givord P., Guillerm M., Monso O. and Murat F. (2016), “How to measure segregation in the education system? A study of the social composition of French colleges”, Education & Training, n°91, pp. 21-51, DEPP. https://halshs.archives-ouvertes.fr/halshs-01447178 For the most recent publication using this data, see: Guillerm M., Monso O. (2022), “Evolution of the social mix of colleges”, Information Note, n°22.26, DEPP. https://halshs.archives-ouvertes.fr/halshs-03727756