We estimate the effect of getting elected on future income development of political candidates. We present a bootstrap approach for measuring electoral closeness, which can be used to implement a regression discontinuity design in any electoral system. We apply the method to the Finnish proportional open list system. Being elected to parliament increases annual earnings initially by €20,000, and getting elected to a municipal council by €1,000. The returns to office for parliamentarians accrue mainly through higher wages during the time in office. Our results imply that the returns to office may be large even in the absence of corruption.