Retirement behavior in the state civil service.

Summary Modeling retirement behavior is an essential issue for both practitioners and researchers. For pension plan managers, it is indeed crucial to be able to make long-term projections of the financial equilibrium - or imbalance - of the plans for which they are responsible, and the importance of changes in retirement behavior in response to programmed changes in pension scales is considerable. For academics, particularly in the fields of public economics and labor economics, understanding the way in which individuals choose to stop working is an ambition that remains highly topical. Work has made it possible to understand the intertemporal dimension of retirement choices, but the explanatory capacity of the models remains weak, both in terms of explaining temporal changes and international differences in retirement behavior. In this context, studying the retirement behavior of civil servants is a way of making progress in these two dimensions. Indeed, little work has been done on civil servants using French data. The fact that civil servants in France can remain in employment until the desired retirement age allows the researcher to rule out labor demand effects, which are often difficult to model convincingly, and which are likely to play an important role in the choice of retirement benefits for private sector employees. This advantage is counterbalanced by a characteristic of the pension scale in the civil service, which is extremely marked by age limits, and therefore less likely to involve a financial trade-off between continuing to work and retirement. The 2003 reform and its aftermath have changed the incentives to retire in the public sector, which suggests an increased capacity to estimate models of retirement behavior.