It is a lot harder to reliably forecast the likelihood of a change in control for the House than for the Senate.  There are 435 House races, and roughly 65 of them are drawing some degree of competitive attention this year, far more than in prior cycles.

There are few independent polls that have been released on the competitive House races.  Democratic challengers have raised substantial sums in many races, and the Democratic Congressional Campaign Committee has as much cash on hand as the Republican Congressional Campaign Committee.  The RNC has much more cash on hand than the DNC (4 times as much), but the Democratic Senate Committee has almost twice as much cash on hand as the Republican Senate Committee. Among all 3 national committees in total, the GOP holds only a modest cash on hand advantage.

With a much greater number of Republican than Democratic seats in both Houses that are at risk of changing hands, the GOP's advantage may not be enough for all the candidates who need it.

The Senate

The Senate has only 33 races this cycle, and the Democrats need to pick up  6 seats  to gain control. Given that 18 of the 33 seats this year are Democrat—held seats, at one time, this was considered likely to be a good year for GOP pickups. No more.

The Democrats did a better job recruiting challengers, and the political environment is, quite obviously, not favorable to Republicans this year, with the President's approval rating still south of 40%.  How the Democrats can gain control of the Senate is pretty straightforward: they need to hold their most vulnerable seats (ranked by vulnerability) in New Jersey, Maryland, Washington, Minnesota and Michigan, and then win the needed six from among Pennsylvania, Ohio, Montana, Rhode Island, Missouri, Virginia and Tennessee, and possibly Arizona or Nevada (also ranked from most vulnerable to least). 

At the moment, three seats are indicating a possible change in control: Pennsylvania, Ohio and Montana are leaning towards the Democrats, for a net gain of 3 seats. Rhode Island (if Chafee is the Republican nominee), Maryland (if Mfume is the Democratic nominee) and New Jersey, are tossups. The Republican candidates hold narrow leads in Missouri, Virginia and Tennessee.

There is also the wild card faint possibility of Joe Lieberman getting re—elected as an independent in Connecticut and choosing not to caucus with his former Party, which has effectively abandoned him. The Democrat leads in Ohio, Pennsylvania and Montana are all in the 3 to 6 point range,  and are hardly conclusive at this stage. As I have been saying all year, it is unlikely the Democrats will gain control of the Senate, but likely they will pick up several seats.

The House

In the House, Democrats need to pick up a net of 15 seats to gain control. For several months, many of the leading political analysts expressed deep skepticism about the Democrats' ability to do that, and could not foresee where the likely pickups might occur. That is no longer the case. I have reviewed the latest analyses by Larry Sabato, Stuart Rothenberg, Charles Cook, Robert Novak, and blogger Chris Bowers of MyDD.com.  All were completed in the last week, and there is remarkable consistency both in identifying the races where a seat could turn over and in noting how few opportunities exist for the Republicans to pick up a Democrat—held seat.  In fact, in only one race, do any of the analysts rate any Republican challenger as either ahead or even. That is the 8th district in Illinois, where Dave McSweeney is challenging Democrat Melissa Bean.  And even here, the consensus view is still a narrow Bean lead.

At the moment, there are 232 Republican held seats in the Congress. In the 2004 Presidential election, George Bush beat John Kerry in 255 of the 435 Congressional districts. So there are many more Democrats in Congress in districts where Bush won, than Republicans holding seats in districts where Kerry won (more than twice as many in fact). This would suggest many more opportunities exist for Republicans to pick up seats than Democrats. But again, this year, the political environment is so nasty for the GOP that the opportunities for seat changes seem to be virtually all on one side. In fact, it is possible that the GOP could win more Democratic held Senate seats than Democrat—held House seats this year.

The analysts have been raising their estimates of the likely number of seats the Democrats will pick up in recent reports.  Sabato is now forecasting a Democrat gain of 12 to 15 seats, Cook thinks Democrats will pick up seats in the mid—teens. Rothenberg sees a 15—20 seat Democrat gain. Novak is forecasting a 14 seat or larger gain for the Democrats, and Bowers a 12—19 seat gain. So all the analysts are in the same general prediction range, right around the point where the House turns over majority control to the Democrats. Nobody is saying that the 15 seat pickup is locked in, but rather that it is there for the taking.

Summing the forecasts into a consensus view

To make some sense of the five reports, I attempted to lay them out side by side to see where they agree and disagree on individual races.  I used a simple point system to come up with a consensus view for the five forecasts.

If an analyst thinks the Republicans are likely to win a seat, that is a +2. If the analyst thinks the seat is leaning to the Republicans, that is a +1. If a seat is a tossup, that is a zero. If a seat is leaning to the Democrats, that is a —1. If a seat is likely to go to the Democrats, that is a —2. In general, a seat that is leaning might be associated with up to a 5 or 6 point lead in a race, slightly larger than a poll's stated margin or error.

On the whole, Sabato and Cook are more reluctant to call a race as leaning, and more likely to call a race a tossup than Rothenberg, Novak or Bowers. Sabato has the Democrats ahead in two GOP held seats, and shows 14 other GOP—held seats as tossups.  He calls Illinois 8 a Democrat—held tossup seat. Cook has the Democrats ahead in one GOP—held race, and calls 18 other GOP—held seats tossups. Rothenberg has the Democrats ahead in 7 GOP—held seats, and calls 10 other GOP—held seats tossups. Novak has McSweeney ahead in Illinois 8, and has Democrats ahead in 15 GOP—held seats (11 of which he calls toss—ups leaning Democratic). Bowers has the Democrats ahead in 9 GOP—held seats, 5 of which he calls leaning tossups, and has 7 other GOP held seats as pure tossups.

After running the numbers for all five analysts, the consensus view is that nine seats held by Republicans are viewed as leaning to the Democrats, and one, Tom DeLay's old seat (Texas 22), is seen as likely to switch to the Democrats. This last forecast seems pretty reliable since Republicans will have to write in their nominee in this district.

Contested GOP seats

The 9 seats that are leaning to the Democrats (with the Bush 2004 vote percentage in parenthesis) are as follows:

Arizona 8 (open seat), 53%

Colorado 7 (open seat), 48%

Connecticut 4 (Chris Shays' seat), 46%

Iowa 1 (open seat), 46%

Indiana 8 (John Hostettler's seat), 53%

Indiana 9 (Mike Sodrel's seat), 59%

Kentucky 4 (Geoff Davis' seat), 63%

Pennsylvania 6 (Jim Gerlach's seat), 48%

North Carolina 11 (Charles Taylor's seat), 57%

Even among the 9 seats, several are very close to breakeven using my point system (and some analysts have the Republicans ahead in North Carolina 11 and Connecticut 4). Only Indiana 9,  Pennsylvania 6, Arizona 8, and Iowa 1 have some degree of consensus that they are leaning Democratic, and even in these races, one or more of the analysts thinks the race is a tossup.

Two Republican held seats come out exactly even in the consensus view:

Connecticut 2 (Rob Simmons), 44%,

Indiana 2 (Chris Chocola), 56%.

Fifteen Republican—held seats are seen as leaning to the Republicans, some very narrowly (especially the first 7 seats below).  These include:

Florida 22 (Clay Shaw's seat), 48%

Wisconsin 8 (open seat), 55%

New Mexico 1 (Heather Wilson's seat), 48%

Ohio 18 (open seat, formerly Bob Ney's), 57%

Illinois 6 (open seat), 53%

Virginia 2 (Thelma Drake's seat), 58%

Ohio 1 (Steve Chabot's seat), 51%

Minnesota 6 (open seat), 57%

Connecticut 5 (Nancy Johnson's seat), 49%

New York 24 (open seat), 48%

Pennsylvania 7 (Curt Weldon's seat), 47%

Pennsylvania 8 (Mike Fitzpatrick's seat), 48%

Pennsylvania 10 (Don Sherwood's seat), 60%

Arizona 5 (J.D. Hayworth's seat), 54%

Ohio 15 (Deborah Pryce's seat), 50%

Finally, 12 Republican—held seats are viewed as likely to stay Republican, though they remain competitive. If there is a big Democratic wave, some of these seats could fall to the Democrats as well.

Washington 8 (Dave Reichert's seat), 48%

Colorado 4 (Marilyn Musgrave's seat), 58%

Florida 13 (open seat), 56%

Nevada 3 (John Porter's seat), 50%

New York 20 (John Sweeney's seat), 54%

New Hampshire 2 (Charles Bass' seat), 47%

California 11 (Richard Pombo's seat), 53%

New York 29 (Randy Kuhl's seat), 56%

Kentucky 3 (Anne Northrup's seat), 51%

Texas 23 (Henry Bonilla's seat), (new district)

Florida 9 (open seat), 57%

New Jersey 7 (Mike Ferguson's seat), 53%

In total, the Republican—held seats that are either leaning to the Democrats or tossups number 12. Add to this, the 15 Republican seats that are considered to be leaning to the Republicans, and you get to 27 very vulnerable GOP—held seats. Including only the 7 leaning Republican seats that are narrowly leaning to the GOP, one comes up with 19 top targets for the Democrats where they are ahead or about even. 

Can the Democrats snatch 15 of the 19 or 15 of the 27? It is certainly possible. Nine of the 27 best targets for the Democrats are open seats, which are much more prone historically to turn over to the other party, and in another 10 districts the Bush percentage in 2004 was 51% or lower. Among the next 12 vulnerable GOP held seats, one is an open seat, and in 4 others the Bush percentage was 51% or less in 2004.

Other than Texas 22, there is no seat among any of the other top 26 Republican—held seats that are being targeted by the Democrats that could not potentially be held by the GOP. It is, however, highly unlikely that such a scenario will play out. If the President's approval rating remains where it is today and the Democrats win just half of the 26 very vulnerable GOP held seats, plus Texas 22, the Democrats will wind up right around where they need to be, just one seat short of a  majority.

Contested Democrat seats

If the Republicans want to make a Democratic takeover of the House more difficult, they need to put some Democratic held seats on their side of the board. The consensus of the five analysts, is that they are unlikely to accomplish this. Other than Illinois 8, (56% Bush), where the race is viewed as very tight by all but Chris Bowers, the other two Democratic held seats that are viewed as  leaning Democratic are:

West Virginia 1 (Alan Mollohan's seat, 58%

Georgia 12 (John Barrow' seat) (new district). 

Seven other Democrat—held seats are viewed as competitive, though likely to remain Democratic. These include:

Georgia 8 (Jim Marshall's seat), new district

Iowa 3 (Leonard Boswell's seat), 50%

Louisiana 3 (Charles Melancon's seat), 58%

Ohio 6 (open seat), 51%

South Carolina 5 (John Spratt's seat), 57%

Texas 17 (Chet Edwards' seat), 70%

Vermont (at large open seat), 39%

Could the GOP steal one of the Southern seats or Vermont? Maybe, but I would not bet on their picking off a handful. In the two Georgia seats, the incumbents are running in new districts,which adds a bit of uncertainty to their races. That is also the case with Henry Bonilla, Republican in Texas 23, who will run in a less Republican district than in 2004).

The Republican problem in defending the House in 2006 is evident in the lists above. 39 Republican—held seats are considered as in play, and at best 10 Democratic held seats.  From now until Election Day, the GOP will be playing defense.  As the Chicago Bears have shown all too often, that is usually not enough to win.

Richard Baehr is the chief political correspondent of The American Thinker.