Skip links

Senate 2006: Scratching Beneath the Surface of the Sixth-Year Itch

Every election year is different, and the results of midterm elections have varied dramatically. Still, can recent history suggest anything about the 2006 Senate results? At the very least, our modern electoral experience can set goals for Democrats and Republicans alike. Let’s look at the last half-century of midterm elections for the U.S. Senate from 1950 to 2002. Note that we are including all midterm elections, both the first midterm election of a presidency and the second one in the sixth year of the two-term presidency. On average, the president’s party has lost three Senate seats in each of those 14 elections.

The Democrats should aim to equal or beat the historical spread in 2006, bringing Republicans to 52 seats or fewer. The GOP and the White House have the opposite goal: to reduce the size of the average loss, or even post a gain above the Republicans’ current 55 seats.

Both sides will play games with the numbers. (We know you are shocked speechless.) The games are easy to play, and to justify. Senate elections are idiosyncratic, not least because only a third of the seats are up in any election year. Moreover, some years see many open seats without incumbents, maximizing the chances for turnover, while other years have almost no open seats, minimizing the level of change. The mix of Republican Red and Democratic Blue states hosting contests also matters. If Democrats are playing on a primarily Red field, they are at a disadvantage, and the same is true in reverse for the Republicans. Finally, the events of six years prior, when the same set of Senate seats was last up for election, are critical. A landslide for one party six years ago almost certainly will mean that the lucky party of yore is very exposed, and possibly due for an “evening out” of their good luck, or perhaps reckoning in the near-zero-sum game of politics. Some weaker candidates likely won six years earlier in states that actually favor the opposition party. Odds are, the same tidal wave at the same height won’t be available twice in a row.

Apply these guidelines to 2006. At most five or six Senate seats out of 33 will be open, without an incumbent, making party turnover harder, though not impossible. There are slightly more Democratic Senate seats up in 2006 (18) than Republican seats (15), giving the GOP potentially more targets and making the six-seat gain needed for Democratic control very difficult to achieve. Additionally, six years ago, in the 2000 elections, Democrats picked up four net seats from the GOP, with several of those new senators regarded as among the weaker incumbents up for reelection in 2006. Prognosis: A tough, uphill struggle for the Democrats in 2006.

Now, let’s leaven that conclusion with some equally impressive realities. The Senate numbers for midterm elections are actually all over the map, and the average of three lost seats for the White House is deceptive. Eisenhower and the GOP lost 13 Senate seats in 1958, and Clinton’s Democrats scratched 9 seats in 1994. Contrast this with Clinton’s no-change Senate midterm in 1998, and a loss of just one seat for Eisenhower in 1954 and Bush Sr. in 1990. The Democrats must hope to revisit 1958–which was an omen for their Presidential takeover in 1960–or perhaps the Democrats would prefer to have their version of 1994. Anything is possible in American politics, and history proves it.

You can be sure the White House will be touting yet another number in 2006. Here at the Crystal Ball, we’ve discussed at length the “sixth year itch” election in two-term, eight-year presidencies (see There have been six of these elections in the post-World War II era (1950, 1958, 1966, 1974, 1986, and 1998). The average loss for the White House in these sixth year elections has been six Senate seats–double the overall midterm average loss of three seats. Thus, if Republicans just barely hold onto the Senate in 2006–a sixth-year election–the Bush White House will try to declare a major historical victory! Of course, just try selling a 50-50 Senate with Vice President Cheney breaking the tie as a glorious outcome.

Instead, the White House will be hoping to match the remarkable achievements in four specific midterm elections in the last five decades. In 1962, riding a perceived triumph in the Cuban Missile Crisis only a couple of weeks prior to election day, JFK’s party added three Senate seats to its total. In 1970, hard campaigning by President Nixon and Vice President Agnew produced a two-seat Senate gain for the GOP. In 1982, despite a severe recession and the loss of 26 U.S. House seats, President Reagan held onto the Senate and even added a Republican seat. And in the example known best to the Bush White House, the GOP gained two seats and captured Senate control from the Democrats in 2002.

But there’s one fly in the ointment: All four of these triumphant White House midterm elections were in a President’s first term. Never in modern times has a president been able to add Senate seats in the dreaded sixth-year election. Of course, look at George W. Bush’s remarkable electoral record so far. Sooner or later, there’s always a first time for every mark in the record book.