Within Reach

A year ago, with Republican victories in the 2004 elections still fresh and with President Bush's job-approval ratings still above 50 percent, Democrats' chances of capturing the House looked fairly slim. Today, however, with Bush's approval ratings hovering around 42 percent, the possibility of a Democratic takeover -- although less than 50-50 -- is very real.

Even though House Democrats need a net gain of only 15 seats this November, that task is fairly daunting because the playing field is so small: Few House Republicans are retiring, and few of the Republican incumbents who ought to be vulnerable are in districts that actually are competitive.

Nevertheless, the Democrats do have a real shot at ending the GOP's control of the House. A simple statistical model that Cook Political Report Senior Editor Amy Walter and I developed several years ago suggests that the 2006 House election will be very close.

On an Excel spreadsheet, we gave each of the 435 House districts a numerical score based on our assessment of the chances of its going Republican this November. The score is computed by taking into account a number of factors, including the historical voting patterns of that district, the caliber of this year's candidates, the strength of the campaigns, and the political atmosphere in the district.

If a district looks solidly Republican, we give it a score of 1.0. If a district is rock-solid Democratic, we assign it a score of 0.0. If the parties have an equal chance of winning in a given district, we give the district a .5 score. Very few districts are in the toss-up territory; the vast majority fall close to the partisan extremes -- 1.0 or 0.0.

Adding up the probabilities of those districts ending up in Republican hands yields an overall estimate of the number of seats that the GOP will hold after the election. We then factor in a margin of error of a few seats to come up with the range of seats likely to be gained or lost by the two parties.

This method is apt to underestimate the turnover in seats if a national "wave" election materializes. But that is a danger that we can keep in mind when analyzing the data.

The current model predicts a net Democratic gain of 10 seats. When we factor in a four-seat margin of error, the model projects that as of today, under current conditions, and without speculating about what the national political environment may do between now and November, Democrats will make a net gain of six to 14 seats. With a bit of luck -- or a wave of any size -- Democrats might even grasp their gold ring: 15 seats.

To be prepared to make the most of a wave, if one occurs, Democrats need to recruit strong challengers in states where filing deadlines have not yet passed. They also need, of course, to raise plenty of money.

Republicans must continue to hold down their retirements in competitive districts, raise money, and pound Democratic challengers in those GOP-held districts where Democrats have a chance of taking over.

The not-so-secret weapon of the National Republican Congressional Committee in recent elections has been its use of money to rip into Democratic challengers and Democratic open-seat candidates early and often. Early attacks have thus enabled the NRCC to create negative impressions of those contenders before they could define themselves positively.

In the past few campaign cycles, the NRCC had enough money to go on the attack very early and very effectively. The danger for Republicans this time is that while the committee has much more money than the Democratic Congressional Campaign Committee -- $19 million at the end of 2005, compared with $15.7 million for the Democrats -- the GOP's resources might not be sufficient to douse all the anti-Republican fires that could erupt around the nation.

Will the GOP have enough money to prevail? That's the question keeping Republican strategists awake at night.