Dangerous flexibility -- retirement reforms reconsidered.

Flexible retirement is supposed to increase labour supply of older workers without touching the third rail of pension politics, the highly unpopular increase of the retirement age. While this may have intuitive appeal, this paper shows that it might be wishful thinking. Economic theory tells us that flexible retirement policies can have a zero or positive effect on labour force participation (LFP) while the effect on hours worked can be positive or negative depending on the distribution of leisure preferences. Thus, the overall effect is ex ante unclear. Empirical results from nine OECD countries show that the effect on LFP is ex post small and positive while the effect on hours worked is negative. Overall, there is no evidence of the desired positive effect on total labour supply (TLS). We conclude that the flexibility reforms enacted so far are dangerous instruments if one wants to increase TLS because they postpone or even replace the installment of more effective policies and may, even worse, reduce total labour volume. [ABSTRACT FROM AUTHOR]