What follows is based on an analysis of national Presidential and House popular vote totals since 1948. I did not look at House seats, but the national vote. I did it while watching the All-Star Game (which was awesome, if you didn't see it).
There is a steady pattern to our congressional elections. A larger than usual caucus of a president's party is swept into office with that president. Two years later, the president's party loses ground, unless the year is 1986, 1990, or 2002 (or even 1978 or 1998, when the president's party suffered only minor setbacks). Pass another two years, and one of two things happens: 1) the president (or a successor of his own party) is reelected, while his party draws down a vote share that exceeds that of the previous midterm, but almost never reaches the heights of the president's first election ; 2) a new president (of a different party) is elected, and their party is riding high relative to their historical performance and the cycle starts again.
So, what does the cycle portend for the likely upcoming Obama administration? Well, a party taking over the White House after large gains in the previous midterm in a familiar occurrence in post-WWII politics. That's what happened in 1960, 1968, and 1976. In those elections, the incoming president's party either did no better than it did in the previous midterm (1968) or actually did considerably worse (1960, 1976). Now that's fine, since our majority is decently large enough to survive a hit in the 2010 midterms and maybe even through the 2014 midterms as well.
At the same time, big change comes from big majorities. Our majority is not big by historical standards. Perhaps there is a way to break out of the cycle that so many post-WWII administrations have followed. The Reagan administration did just that. Their electoral high-point was actually the 1986 midterm, when House Republicans only lost by 2.5%. No other post-WWII administration has done so well so late in their cycle. Assuming Obama wins this election and the next one, it will be incumbent on Democrats to figure out what Reagan figured out in the 1980s. If we don't, we're likely looking at a short-lived majority. If we do, we're looking at a real seismic shift in politics long past the end of his administration. In Reagan's case, that impact was that he got his party within striking distance for Newt Gingrich to eventually lead them to the promised land for the first time in forever, an eventuality that would have seemed impossible in the 1970s (the years leading up to Reagan).
Here's another way to look at Reagan's impact on his own party.
Democrats fell from a 5.8% win in Truman's first election to a virtual tie in his last midterm
Republicans fell from a .1% win in Eisenhower's first election to an 11.9% loss in his last midterm
Democrats fell from +9.5% to +2.5% under Kennedy/Johnson
Republicans fell from -1.8% to -16.6% under Nixon/Ford
Democrats fell from +10.8% to +8.7% under Carter
Republicans rose from -2.7% to -2.5% under Reagan
Republicans rose from -7.9% to -7.2% under Bush
Democrats fell from +5.1% to -0.9% under Clinton
Republicans fell from +0.3% to -7.9% under Bush