Mark Blumenthal moderates a debate over the GOP's prospects for retaining control of the House. Basically, the historical evidence shows that it would be hard. Wave elections happen infrequently, and they're often a rejection a the incumbent party. It seems especially true in the last few decades that voters are motivated by revulsion against, rather than enthusiasm for, the incumbent party. All of that suggests Democrats won't have a big wave until Republicans hold the White House.
But, there's an important caveat here: age. Midterm elections have long seen disproportionate numbers of older voters. Older people vote at higher rates, and mid-term elections are low-turnout affairs. In the past, this fact didn't matter much, because age was not a highly salient characteristic of voting. Over the last few elections, though, a huge generation gap has opened up, with the young tilting strongly Democratic and the old strongly Republican.
The 2010 election was largely a result of the generation gap. To oversimplify, the old (conservative) portion of the 2008 electorate showed up, and the young (liberal) portion stayed away. Democrats borrowed a lot of seats in 2008 with a swollen electorate filled with young voters who weren't likely to stay engaged in 2010. But the corollary of that is that Republicans borrowed those seats right back in 2010 with a disproportionately old electorate that doesn't reflect what 2012 will look like. The mere fact of having a presidential race will make the House electorate substantially more Democratic.
Indeed, it's entirely possible that, if the age gap continues, the Congressional vote will continue to swing back and forth like this, with Democrats picking up seats in presidential election years, and losing them in off-years. Obviously, those swings could be obscured by swings going the other way -- Democrats could gain seats in a midterm election if the president is a really unpopular Republican, etc. Anyway, my main point is that the salience of age is a new and very important factor in American politics, and it provides some reason for caution in using history to forecast future results.