It is clear from the analysis I encountered at the 2010 American Political Science Association meetings in Washington that Democrats are facing a beating in the 2010 House elections.
Prominent midterm election forecasters predict a GOP House in 2011; GOP gains must be at least 39 seats for them to take control. Forecasters build election projections using variables such as presidential job approval and economic performance. Two forecasters with excellent track records are James Campbell of SUNY-Buffalo and Alan Abramowitz of Emory University. Campbell predicts a 51-seat GOP house gain in 2010; Abramowitz foresees a 43-seat gain. Election savant Larry Sabato of the University of Virginia predicts a 47-seat GOP pickup.
The only pro-Obama prediction I encountered came from Robbin Mellen of Mississippi State University. Drawing on historical evidence of the impact of presidential visits in tossup House races, Mellen predicts that a tight Obama focus on these contests might yield 221-224 Democratic House seats in 2010 (a majority comprises 218 seats). That, of course, assumes that other factors, such as those in forecasters' models, don't overshadow presidential efforts.
So we'll have a test this fall of the power of presidential campaign effects. The lay of the land is starkly threatening for Obama, but there may be some chance that careful campaigning on his part might stem the threatening tide in House races. Given his faltering popularity, that's probably a long shot. If Obama prevails, it will be an unprecedented triumph of a president over the broader campaign environment--so Obama supporters shouldn't become too optimistic about that transpiring.