The structural approach allows to disentangle the two counteracting mechanisms that drive the employment effects of partial retirement: 1) the crowd-out from full-time employment, and 2) the movement from early retirement or unemployment to partial retirement. It also allows to investigate the isolated role of financial compensations in a partial retirement program. The analysis is based on a unique administrative dataset that collects biographical information on full employment histories and combines information on partial retirement take-up with information on individual pension levels. I perform counterfactual policy simulations that analyse the role of partial retirement combined with financial subsidies and an increased normal retirement age. The results show negative employment effects but potentially positive fiscal consequences as well as a reduction in pension income inequality when partial retirement is introduced. Wage and pension compensations in partial retirement do not substantially affect employment behavior but play a substantial role in the fiscal consequences and distributional effects of partial retirement. An increased normal retirement age decreases negative employment effects and improves the effect on fiscal balances from partial retirement.