Senate Elections Swing II

As I discussed on March 30 Senate elections have a lot to do with the previous class of Senators and nothing to do with what happened in the last election and, don’t always follow a President’s mid-term ratings. That’s not to say environment has nothing to do with the elections. The last 3 times 12 or more House seats shifted at least 5 Senate seats also did. Of the 13 times before that, however, only 4 had a corresponding Senate gain.

In 1982 the Democrats picked up 27 House seats but only 2 Senate seats. In 1984 Republicans were +16 in the House (due to surge and decline) but actually lost 2 Senate seats. This despite Ronal Reagan’s landslide victory.

The reverse is also true. In 1986 the Democrats picked up 8 Senate seats and only 5 House seats. They were +11 in the House net from 1982 and 1984, so there was no decline. The election wasn’t strong enough for a surge. The Republicans had picked up 12 Senate seats in 1980. Losing a significant number of seats was inevitable.

The Senate isn’t immune to a party’s popularity. If the House gain is a surge there’s usually a corresponding Senate gain. Thus the Democrats winning big in the House and Senate in the last 2 elections. A party gaining 5 or more Senate seats in consecutive elections for one Senatorial class has happened 3 times before. Democrats gained 8 seats in 1930, but this followed losing 8 in 1918 and 1924. So when they gained 7 seats in 1936 it was a new surge. It was one followed by a 9 seat decline in 1942. Likewise the Republican gains in 1942 and 1948 were a Democratic decline with the former and a Republican surge with the latter.

The Republicans gained 9 seats in 1994. Democratic gains in 2000 were almost inevitable. Republicans won the White House, lost only 1 House seat but lost 5 Senate seats. The Democrats won another 5 in 2006. Because the net for 1994-2000 was Republicans +4 this could either be a Republican decline left over from 1994-2000 or a new Democratic surge. Regardless I wouldn’t expect Republican gains in 2012 if it isn’t a heavy Republican year.

In 2004 the Republicans picked up 4 seats after neither party picking up any in 1994. The last time there was a +4 gain without a surge the previous class election was 1966 for the Republicans. Democrats gained 2 Senate seats despite the Nixon landslide.

When a party gains a significant number of seats following two elections where the other party didn’t gain they lose seats. This didn’t happen once. Republicans gained 7 seats in 1938, no net in 1944, and a Republican gain of 5 in 1950. These did follow a 12 seat Democratic gain in 1932. The Republicans shouldn’t make Senate gains this year. Yet it appears they will break the statistical certainty that has ruled Senate elections since there was a direct election of Senators in 1914.