ABSTRACT
The persisting unemployment level in almost every OECD country has called for the implementation of a great variety of employment and training measures. Although welfare states are very differently conceived, labour force participation of older workers has been decreasing in all these countries. This suggests that old-age insurance was used as an instrument of employment policies aiming at combating unemployment (Blondal and Scarpetta, 1998). The modes of transition from work to retirement can be considered as a way to reduce labour force indeed (see Casey, p. 1989, 149 and Kohli and Rein, 1991, p. 2). However, in a neo-classical framework the retirement decision is always treated as an individual and rational decision made by older worker who react to incentives provided by a given system (see e.g. Gruber and Wise, 1998 and Riphahn and Schmidt, 1999). Although this dichotomy is helpful to take a stand in the political debate, it is quite irrelevant to the empirical analysis of the context in which older workers' retirement decisions are made. Incentives were actually designed in order to make older workers retire earlier. It should be pointed out, that the retirement instrument has been implemented in old-age insurance systems, which were differently conceived and organised. The questions, which result from these observations, deal first with the working-life patterns and retirement decision of older workers in three countries. Italy,
Great Britain and Denmark have been selected because different traditions underlie their social systems. The Italian system corresponds rather to the Bismarck's tradition whereas the other two stem from the Beveridge's model. Considering this divergence raises then the question of the incentives provided to older workers by each system.
