For Whom the 'Retirement' Bell Tolls: Inter-temporal Comparisons Using the 1994 and 2002 Canadian General Social Survey

Data from the 2002 and 1994 General Social Survey are used to analyze the determinants of retiring due to mandatory retirement and the expected age of retirement in Canada. Changes between 1994 and 2002 are decomposed into a component attributable to shifts in the composition of respondents and the other component to changes in the preferences and constraints of respondents, the latter playing a very important role for both outcomes. Specifically, between 1994 and 2002 we find: (1) a 1.3 percentage point drop in the probability of retiring due to mandatory retirement, with that drop due to preference/constraint changes substantially reducing the probability of retiring due to mandatory retirement but being almost offset by a change in the composition of the workforce that increased the probability of retiring due to mandatory retirement; and (2) a 3.7 year increase in the expected age of retirement, with that increase being almost exclusively attributable to preference and constraint shifts. The implications of these findings for employers, employees and policy makers are