Do Politicians' Relatives Get Better Jobs? Evidence from Municipal Elections

We estimate the impacts of being connected to politicians on occupational choice.  We use an administrative dataset collected in 2008-2010 on 20 million individuals and rely on naming conventions to assess family links to candidates in elections held in 2007 and 2010.  We first estimate the value of political connections by applying a regression discontinuity design to close elections in 2007.  Those estimates likely combined the benefits from connections to current office-holders and the cost associated with being related to a losing candidate.  We use individuals connected to successful candidates in the 2010 elections as control group and find that relatives of current office-holders are more likely to be employed in better paying occupations.  Relatives of candidates who narrowly lost in 2007 have lower occupations.  A third-party randomly split our dataset in two and gave us sample 1.  Once the review is completed, we will apply the approved methodoloy to sample 2.