“Congressional elections, held in the middle of a president’s term, are usually referenda on a president and his policies. Only twice has a president’s party gained seats in his first midterm election.”
“The Republicans [walked] away from the 1946 midterms with gains of 56 seats in the House and 13 in the Senate—and majorities in both houses…”
- Top 10 Historic Midterm Elections by T.A. Frail/Smithsonian October 13, 2010
As of tonight - that is, the evening of November 8, 2018 - it looks like the Democrats are expected to gain 36 seats in the House of Representatives. How does that stack up to previous midterms? Glad you asked:
* Estimates based on confirmed gains/losses (https://www.vox.com/a/midterms) and anticipated further gains/losses (https://fivethirtyeight.com/features/the-16-races-still-too-close-to-call/), as of November 7, 2018.
What conclusions to draw? I don’t think there are any strong conclusions to draw from midterm elections, except that the president’s party usually loses seats. At the very least, midterm outcomes don’t seem to predict which party will win the White House in the next electoral cycle. Not that the media has shrunk from framing these outcomes as saying something about historical trends. Take midterm coverage for 1994, 2010, and 2018, all very bad years for the then-current White House. Here are some day-after headlines from the New York Times:
G.O.P. Captures House, but Not Senate (2010)
Women Lead Parade of Victories to Help Democrats Win House (2018)
The above headlines reveal three different spins on midterm outcomes: historical blip (1994), partial setback (2010), and triumphant (2018). All seem to imply that history is moving in one direction.