Let's take a look at where we actually gained seats last night: FL-16, LA-6, KS-2, TX-22.
And look at places where we well could have, but fell short: PA-11, WI-8, NH-1.
Notice a pattern here? These all fit the conventional target of weak freshman Democrat (PA-11 excepted) elected in a conservative area. The gains all fit this criteria and they were all south of the Mason-Dixon.
Much is made of the fact that we do not have a single representative in New England, but I am equally if not more concerned about the fact that we seem to be losing our grip on the conservative areas in blue states, as evidenced by losses or too-close-to-call races in MD-1, MI-7, and MI-9. The fact that we backslid 9 points from 2004 in Pennsylvania (tied to an underperformance from final House race polls), and 12 points in Michigan despite the fact that these were "bitter" Hillary states suggests something is seriously amiss. Zell Miller's "a national party no more" could easily apply to the GOP if this continues.
I am beginning to think that looking at House and Senate races and doing descending sorts on Bush 2004 percentages has become a seriously crippling way of targeting House seats. The President won 250 or so House districts in 2004. But building a majority solely by picking seats off this target list becomes seriously problematic. To win 218 races from this grouping, you'd need to win 87% of races. To win 200, leaving a handful to come from blue districts, you'd need to win 80%.
These need-to-win percentages are simply too high. First, you have personally popular Democrat incumbents in many of these seats. And second, any district that was within 10 points of the national median at the Presidential level is (at best) only a lean to one party or another for Congress.
We need to expand the map.
We need to get to a place where we only need to win 60% or two thirds of our "winnable" races. And that requires expanding the definition of a winnable race.
This is not a fifty state strategy. It is a 435 district strategy. Parties and Presidential candidates trade within narrow bands within states. Local candidates trade in very wide bands, enabling them to win "safe" seats from the other party. All the field organizers in the world did not flip Alaska, North Dakota, and Montana at the Presidential level, and states like Indiana, Florida, and North Carolina were either attainable because of the national swing, or next door to Illinois.
We need to contest all 435 seats -- there are some, like the inner cities, that we will never get. But for everything else, we should act like these are winnable seats for Republicans, if not this time, then next time, or the time the seat opens up.
We need to encourage good repeat candidates in Democratic seats. The average second-time candidate who won in 2006 won just 42% in their previous bid. A lot of people figure they won't run if they're not absolutely sure they can win. We need to create a culture, like the one that exists in Britain, where it's expected that you'll have one or two elections to hone your skills as a candidate before winning. Barack Obama was a failed House candidate in 2000.
In many ways, our failure in the House was a failure of recruiting. A Diana Irey could have knocked off Murtha this year. In other ways, we will need to strengthen our focus in rebuilding in the blue states so we can credibly raid seats on the other side.