This paper investigates the value of firm political connections to US congressional candidates using a regression discontinuity design. In a sample of close special elections occurring at times unrelated to firm-specific economic events or broader political events, I compare the abnormal returns...