Projections of elderly people with a need for autonomy support - EPP4 model (Insee-Drees)

The model "EPP4 - Projections of elderly people in need of support for self-reliance" is the result of joint work between Drees and INSEE. It was mobilised for the projections made public in November 2025 ([Dufeutrelle et al., 2025](https://drees.solidarites-sante.gouv.fr/publications/700-000-seniors-en-lose-dautonomy-additional-dici-2050)). The data available here are projections of the number of people aged 60 or over between 2021 and 2070, with distributions by sex, age group and level of loss of autonomy, all within the departmental framework. The definition of dependency used is based on the GIR estimated from the 2022 Self-Employment Survey (Drees). This leads to higher numbers of staff losing autonomy than current APA beneficiaries. Description of the columns in the datasets: Column name | Description ---|--- DEP | Department. SEX | Sex. 1 : men, 2: women TRAGE | Age range vol_seniors| Total number of individuals vol_seniors_perte_autonomy| Number of persons with loss of autonomy (estimated IRG 1-4) vol_seniors_perte_autonomy_severe| Number of persons with severe loss of autonomy (estimated IRG 1 to 2) vol_seniors_menage| Number of people living at home vol_seniors_perte_autonomy_menage| Number of people with loss of autonomy (estimated IRM 1-4) living at home vol_seniors_perte_autonomy_severe_menage| Number of people with severe loss of autonomy (estimated IRM 1 to 2) living at home vol_seniors_institution| Number of people living in Ehpad or similar (Ehpa non Ehpad, USLD) vol_seniors_perte_autonomy_institution| Number of persons with loss of autonomy (estimated IRG 1-4) living in Ehpad or similar vol_seniors_perte_autonomy_severe_institution| situation of severe loss of autonomy (in GIR estimated 1 to 2) living in Ehpad or similar esp_vie_60_ans| Life expectancy at age 60 esp_vie_spa_60_ans| Life expectancy without loss of autonomy at age 60 The projections are available according to several scenarios relating to the evolution of life expectancy, and several hypotheses on the evolution of the loss of autonomy. There is one file per cross-scenario x hypothesis. The life expectancy scenarios correspond to those of INSEE's Omphale projections. "edv_basse" corresponds to the low life expectancy scenario, "edv_haute" to the high life expectancy scenario and "central" to a central scenario (https://www.insee.fr/information/3683517). Assumptions of changes in dependency are defined as follows: Assumption | Definition ---|--- Optimistic | gains in life expectancy at 60 years in autonomy are higher (by 10%) than gains in life expectancy at 60 years; time spent in moderate dependence recedes. ‘intermediate’ | Life expectancy gains at age 60 correspond entirely to independent life-year gains. ‘pessimistic’ | The share of years of moderate dependence in life expectancy at 60 remains constant; life expectancy gains thus correspond for one part to fully autonomous life and for another to years of moderate dependence, in proportion to the share observed today. These data are used as input to the Livia model to calculate the distribution of seniors in different types of living places (ordinary housing, independent residences and Ehpad).