This article seeks to understand the effect unified government has on seat loss in midterm elections.
Using data from the 1950 to 2018 midterm elections, I use Ordinary Least Squares regression models to determine the effect of unified government on seat loss.
I find that unified government increases the number of seats a president’s party loses during a midterm election. In addition, unified government reduces the number of seats saved by presidential approval and increases surge and decline effects.
Unified government explains why Democrats have traditionally performed worse during midterm elections. The results also indicate that rather than a surge and decline, midterm elections are a counter-surge to the surge in support the president’s party receives in the previous presidential election.