1994, like 1982, was a second-year election (of a presidential term). 2006, like 1998, is a sixth-year election. Much has been written about the “six-year itch,” which has historically spelled large House losses for the party which occupies the White House. But the key word is historically. Consider how many seats the party in control of the White House has lost or gained in the last six sixth-year elections:

• 1950     -28
• 1958     -49
• 1966     -48
• 1974     -49
• 1986       -5
• 1998      +5

It is easy to see from the above how the concept of the six-year itch gained currency – but it’s also easy to see that the last time it made its presence felt was a third of a century ago. Things have changed since 1974. Most significantly, redistricting/gerrymandering has drastically reduced the number of competitive seats. The 1950 – 1974 elections are no more relevant to predicting 2006 results than the election of 1802 is.

There is another factor in play here: the coattail effect. Actually, in this case, it’s the absence of a Bush coattail. Most presidents win election by wider electoral margins than George W. Bush won by in 2000, and they often sweep a number of their fellow party members into Congress with them. But then, two years later, many of these coattail congressmen are swept back out office when they cannot ride the president’s back. The Democrats lost 35 House seats when Ronald Reagan decisively won the presidency in 1980 – but they gained 27 of them back two years later.

Because Bush never had a coattail to begin with – he lost the popular vote in 2000 and the Democrats gained one House seat that year – he didn’t have a reverse coattail in 2002 (his party actually gained seven seats). In fact, the relatively narrow margins in the last five biennial elections simply haven’t produced the kinds of large House swings that would make the GOP vulnerable to a large corrective counter-swing. Bush also had no particular 2004 coattail whose beneficiaries would now be vulnerable – his party gained two House seats two years ago.

There is one important caveat regarding the above: 9/11. Without 9/11, there probably wouldn’t have been a seven-seat GOP House gain in 2002. Seven seats isn’t a huge swing, but it’s a larger gain than the Republicans would likely have realized in the absence of 9/11 – they might even have lost a few seats instead. So it’s possible that 2006 will see a swing that would otherwise have occurred in 2002 but which has been artificially delayed by security concerns.