Despite nice charts about policy and message and all that, elections turn on the economy. If the economy is doing well, the party in charge will stay in charge. If people are scrambling to pay their mortgage and have homes that are decreasing in value, they'll vote for the other guy. Incumbents still have a huge advantage, but in tight races, the outsider party will win. All this happens, despite the fact that political officials have largely no control over major swings in the economy.
That's American politics in a nutshell. Class dismissed.
So, what's going to happen in November? What's going to happen in two years? The economy is still in the dumps, and all indications are that it will stay sucky for a long time. Not a good scenario for Democrats.
Matt Yglesias and Kevin Drum look at a recent survey by the Washington Post that shows that even though a slight majority of people believe that Democrats better represent people like them, more people say they will vote for Republicans for Congress.
Apt. 11d 
Isn’t there another consideration, which is that, in all cases, the party that holds the presidency tends to lose seats at every other level? There are occasional elections that buck the trend slightly, but the overall pattern is clear, at least since 1980. Evidently, the president serves as a lightning rod. So one would expect the Democrats to lose seats in Congress, governorships, state legislative positions etc. even in the best of cases.
LikeLike
It’s quick n dirty, but this chart shows that since WWII, the party of a first-term president loses seats in Congress in the mid-term about 80% of the time. Exceptions: JFK +4 Senate in 1962, Nixon +2 Senate in 1970, and GWB +2 Senate and +6 House in 2002. Reagan also eked out no net losses in the Senate in 1982.
LikeLike