Nonprofit, nonpartisan journalism. Supported by readers.


Raw data on midterm history

Raw data on midterm history
By Eric Black

Pretty soon, we’ll know some numbers, including at least an early partial count of how many Dem-held seats in the U.S. House will turn red. A lot of forecasts suggest that the final number will be between let’s say 35 and 65, and the word “historic” may be thrown around a lot.

Below you’ll find a table of the House seats lost (or gained) by the president’s party in the last 19 midterms going back to 1934, FDR’s first midterm.

The overall record is 16 cases in which the president’s party lost ground in the House, versus three instances of gains. The gains are all extremely small, and some of the losses are very big. The biggest on this chart is a loss of 71 seats by FDR’s Democrats in 1938, his second midterm. But that is not the biggest midterm loss ever. The pattern of the presidents’ parties losing seats at the midterm goes way back into the 19th century. Abraham Lincoln’s Republicans lost 22 seats in the midterm election of 1862 – that’s the middle of the Civil War and bear in mind, the most anti-Lincoln states were out of the union at the time. (Lincoln’s home district in Springfield, Ill. even flipped from Repub to Dem.)

The biggest midterm clobberization ever appears to be 1894, with Dem Pres. Grover Cleveland struggling through a bad economy in his second term. The Repubs picked up – hold your breath – 130 seats in that one, and that was when the full House numbered only 357. (It’s 435 now.) Yikes.

Article continues after advertisement

But sticking to relatively modern times on the table below, you’ll see that although the president’s party almost always loses ground, there’s a wide variety of margins, which I suppose is the main point. There seems to be no correlation at all between how the president’s party did at the midterm and how the president did in the next election, if he sought reelection.

So, if the Repub gains break 50, it will be among the biggest midterm pickups in modern times, but they would need 71 to be the biggest. When we see some real numbers tonight or in the morning, the chart below will give you some idea where it fits into the historical pattern.

Year       President            House seats lost/gained by Prez’s party

1934       FDR                       +9

1938       FDR                       -71

1942       FDR                       -55

1946       Truman                  -45

1950       Truman                  -29

1954       Eisenhower             -18

Article continues after advertisement

1958       Eisenhower             -48

1962       JFK                          -4

1966       LBJ                         -47

1970       Nixon                      -12

1974       Ford                        -48

1978       Carter                      -15

1982       Reagan                    -26

1986       Reagan                     -5

1990       Bush I                      -8

Article continues after advertisement

1994       Clinton                     -52

1998       Clinton                      +5

2002       Bush II                     +8

2006       Bush II                    -30

2010       Obama                     ???

My source for all these numbers is The American Presidency Project. Their chart also shows the approval ratings of the various presidents in the months leading up to the midterms.

What about the Senate?

Their chart also has the Senate results, which seems to be even less predictable. There are three instances in which the president’s party gained seats in the Senate while losing ground in the House. If, as expected, the Repubs pick up 7-10 Senate seats this cycle, it would be among the biggest such pickups post-Hoover history. But the biggest were losses of 13 Senate seats by Eisenhower’s Repubs in 1956 and 12 by Truman’s Dems in 1946.