Does a legislator’s career prior to joining politics impact constituency development? I compile an original dataset of politicians’ biographies and use their entry routes into politics to classify them as ‘parachuters’ — those who are hereditary or part of the local socio-economic or cultural elite — and ‘climbers’, working class politicians who have made their way by rising up the ranks. I use a close election regression discontinuity design to estimate the causal effect of politicians’ identity on local economic growth, measured by night time lights. I document three key results: First, political entry barriers have perverse economic consequences. Lights-GDP elasticity estimates imply that electing parachuters leads to 0.2 percentage point lower growth of GDP per capita per year compared to constituencies where climbers are elected. Second, a candidate’s background is a key determinant of political selection, even after accounting for conventional factors such as politician’s sex, religion and ethnicity. Third, there is suggestive evidence that the impact is driven by misallocation of bureaucratic resources: districts with greater proportion of parachuters have higher turnover of investigating police officers, which in turn is associated with greater economic crime and lower economic growth. These findings indicate at a new mechanism via which elites maintain de facto power and persist over time.