This column has repeatedly -- some would say relentlessly -- argued since late last summer that Democrats are in real danger of losing their House majority in November. So it's only fair to analyze why top Democratic strategists, in an admittedly bad year for their party, think they still have a good chance to retain their hold on the chamber, albeit with greatly reduced numbers.
First, they figure they will pick up four seats held by Republicans. These include Michael Castle's at-large seat in Delaware and Mark Kirk's 10th District seat in Illinois (both men are running for the Senate). The strategists also believe they can defeat two incumbents in overwhelmingly Democratic districts: Joseph Cao in Louisiana's 2nd District and Charles Djou in Hawaii's 1st. They think a couple of other seats are vulnerable, including Dan Lungren's in California's 3rd and Charlie Dent's in Pennsylvania's 15th, but the first four are clearly their best shots. If Democrats can win that quartet, Republicans would have to grab 43 seats held by Democrats to score a net gain of 39, the number they need to claim the barest majority of 218-217.
Next, Democrats look at their 16 most endangered open seats and believe, based on the party's strong performances in special elections over the past 18 months, that their worst-case scenario would be to lose eight. Those 16 are AR-01, AR-02, IN-08, KS-03, LA-03, MA-10, MI-01, NH-02, NY-29, PA-07, RI-01, TN-06, TN-08, WA-03, WV-01, and WI-07.
In the Democratic calculus, if the party picks up the four Republican seats and loses just eight of its 16 most vulnerable open seats, it would be down only four at that point. That means 35 Democratic incumbents would have to lose for the House to flip. Even with a large number of theoretically vulnerable Democratic members, say 70, at least half of them would have to lose for the GOP to regain the majority.
http://www.cookpolitical.com/node/7659