Why does raising the early retirement age affect employment?

Why do workers delay retirement when their early retirement age (ERA) rises? Understanding this is essential for predicting the effects of ERA increases in different contexts and assessing distributional impacts of such reforms. We study ten years of reforms in the United Kingdom which increased the ERA for women from 60 to 66. The UK is an attractive setting for this question because there is no financial incentive to retire at the ERA. Credit constraints are found to be important: credit constrained groups experience around a 13-percentage-point larger increase in employment rates as a result of the reform than non-constrained groups. In contrast, groups with greater loss of pension wealth do not see larger increases in employment rates and employers do not delay dismissing employees as a result of the reform. There are smaller, but still large and statistically significant, increases in employment rates of around 11 percentage points for wealthier groups for whom other explanations are shown to be unimportant. This points towards behavioural reasons such as signalling, framing, or social norms also being important.