This paper examines the influence of private equity firm (GP) political contributions on public pension funds' investment decisions using micro-data on investments in private equity (PE) funds. Employing a regression discontinuity design comparing GPs donating to winning versus losing candidates in close U.S. state elections, I find that post-election pensions' tendency to invest are 10 times higher in GPs donating to winner assigned as or appoint their board member. Effects are pronounced for candidates seeking elections afterwards and weakest in states with high public corruption oversight. Connection-based PE funds underperform non-connected ones, partly attributed to abnormal management fees and lower subscription rates.