I was subjected to an Ann Coulter op-ed piece this morning where she repeated the Republican talking point that Democratic gains this year weren't even in step with historic opposition performance 6 years into a President's term. Republicans who actually bother to give a figure say the average is about 30 House seats, and they say it's the average for the past "several decades."

Really?

By my count,

Democrats actually picked up 5 House seats in 1998 while the Senate stayed even.

Democrats picked up 5 House seats and 8 Senate seats in 1986.

Democrats picked up 49 House seats and 4 Senate seats in 1974.

That pretty well covers the past several decades. The mean gain for the party opposing the President is under 17 House seats.

It's possible, however, that Rs really meant to include earlier elections; I went back and tried to figure it out (not easy with some of these elections; I gave up on 1822 and 1794) and excluding those two years, there was a historical average of a bit under 30 losses for the president. So I guess if we want to compare the Democrats' gains to those amassed against Ulysses S. Grant (loss of something like 97 seats) or FDR (loss of 72), then yes, they are below average. Barely.

(A few notes on methodology: I got my data from the House of Representatives Clerk. When there were multiple parties, I treated the small parties like the opposition. When the House increased in size, which happened a lot back in the day, I would add the President's losses to the opposition's gains and divide by two. Finally, there were big realignments in 1822 and 1794, so I didn't count those elections, but I don't think they represented large net movements against the respective administrations. Historians among the readers, feel free to correct me. Had I included them, the average might have dropped to as few as 25 seats).