Some students at Hamilton College published a study saying they are, at least for the sample size of pundits they tracked. Now, much as I'd like to believe this, the data was taken from 2007-8, and a lot of the predictions are of the "you'll see, my party's presidential candidate is going to win this race!" variety. They all say that, and 2008 was a Democratic year. If you studied politicians' predictions in 2009-10, the results would probably be a lot different.
That analysis doesn't apply to everyone, though. The most accurate prognosticator of the whole bunch was liberal Nobel Prize-winning economist Paul Krugman, who got almost all his predictions right. These were not generally political forecasts, but economic ones, and he saw the much of what happened with the collapse of the economy earlier, and more clearly, than almost anyone else. He deserves a lot of credit.
On the other end of the spectrum, conservative columnist Cal Thomas is rarely right about much, and that's true almost every year. Anyone in either party who makes such partisan projections is likely to be wrong a lot, and he was.
The full study is here.