Does higher office always lead to more favoritism? We argue that firms may lose their benefit from a connected politicians ascent to higher office, if it entails stricter scrutiny that may reduce favoritism. Around close Congress elections, we find RDD-based evidence of this adverse effect that a politicians win reduces his former classmates rms stock value by 3.2% after a week. Exploiting the entry of Craigslist across the U.S., we find that state-level scrutiny drives this effect. It further varies with politicians power, firm size and governance, and connection strength, and diminishes as a politicians career concern fades over time.