We study causal effects of two early retirement reforms. Reform 1 increased normal retirement age stepwise from 60 to 63. Simultaneously, it became possible to use early retirement with benefit discounts. Reform 2 increased the age of early retirement stepwise from 60 to 63. We investigate behavioral responses to the reforms using administrative data and difference-in-differences strategies. We find strong and significant causal effects of both reforms. Individuals postponed retirement, stayed employed longer, postponed unemployment, and shifted to alternative pathways into retirement. The overall use of the retirement system declined by about 1.5 and 2 months per person after each of the two reforms. Individuals with low pension wealth and those who were affected immediately by the reform responded more strongly.
