Election Forecast: Three Weeks Out

In the past two weeks, since the emergence of the Mark Foley affair, every political analyst has raised his forecast of likely Democratic gains in the House, and most are forecasting a Democratic takeover, requiring at least a 15 seat pickup. Yesterday, Jay Cost wrote an article for Real Clear Politics that attempted to assign some probability assessments alongside each major forecaster's range of outcomes. Cost did not attempt to analyze individual races.

In a September 5th article, six weeks back, I tried to compare how 5 major political analysts viewed the most competitive House races. At that time, the analysts were suggesting that the Democrats would pick up between 12 and 20 seats. Today, the range is 18 to 27. Chris Bowers provides a range of 19—27 , after allowing for 1 or 2 GOP pick—ups. Stuart Rothenberg has a range of 18—25 seats. Larry Sabato has  a range of 18—22 seats . Charles Cook says a pickup of 30 is more likely than a pickup of 15 . Robert Novak has not updated his forecast from the end of August, though presumably he would be more pessimistic about GOP chances today than before, if he did so. In my earlier article, the consensus of the five forecasts was that Democrats were ahead in races for 9 seats, even in two others, and only slightly behind in 15 other races.

Evaluating the Bowers, Rothenberg, Sabato and Cook reports available today, those numbers have changed quite a bit. There are 17 seats currently held by the GOP in which the consensus view is a Democratic lead. But there are varying degrees of agreement on the status of the 17 races. In 4 races, there is consensus the Democrat is ahead— Arizona 8  (open), Florida 16 (open— formerly Foley), Texas 22 (open— formerly DeLay), and Colorado 7 (open). It is worth noting that in Florida 16 and Colorado 7, there are some polls suggesting the races are still competitive. Arizona 8 is for all practical purposes  a write—off by the GOP at this point, and in Texas 22,  GOP hopes ride with  a write in vote for a candidate with a hypenated name.

Three of four analysts believe the Democrats are ahead in Pennsylvania 10 (Sherwood), Indiana 2 (Chocola), Indiana 8 (Hostettler), Iowa 1 (open) and Ohio 18 (open).  Two of four believe Democratic candidates are ahead in
New York 26 (Reynolds), Indiana 9 (Sodrel), and Pennsylvania 6 (Gerlach).

One of the four believes Democrats are ahead in New York 24 (open), North Carolina 11 (Taylor), Ohio 15 (Pryce), New Mexico 1 (Wilson), and Pennsylvania 7 (Weldon).  In the 17 races identified to this point, any analyst who did not believe the Democrat was ahead, was forecasting that the race was a tossup.

What is of interest is that while the analysts are predicting Democratic gains of 18 seats to maybe 30, they can only find 17 seats in which at least one of them believes the Democrat is now ahead, and only 12 seats on which at least two agree that the Democrat is ahead. In the other five, only one analyst of four believes the Democrat is ahead, and this would qualify as a tossup race.  No other seat is identified by any of the analysts as leaning to the Democrats.

This would suggest that while the GOP's chance of retaining control has clearly slipped, it may not be impossible for them to eke out a small majority by winning almost all of the close or leaning races. 

The four analysts are not equally circumspect about assigning a lead to the Democrats in individual races. Bowers has the Democrats ahead in 17 races, and even in 8 others. In all he looks at 61 GOP—held seats, and considers 22 of them likely GOP holds, or 39 in play.  Rothenberg has the Democrats ahead in 10 races, and even in 14 others. In all, he looks at 52 GOP held seats, and considers 13 of them likely to be held by the GOP, or 39 in play, the same numbers as Bowers. Charles Cook calls the Democrats ahead in only 4 races, and calls 25 others tossups (even). He looks at 60 GOP held seats and believes 43 of them are in play.   Finally, Sabato has the Democrats ahead in 11 races, and rates 16 others as tossups. He considers 62 GOP held seats, and regards 20 of them as likely GOP holds, or 42 in play.  As the election nears, the four analysts are far more consistent in their analyses than earlier in the cycle.

As an avid poll—watcher myself, I do not agree with the consensus view in a few races. I think the Democrats are in a very strong position in North Carolina 11, Ohio 15 and Pennsylvania 7, (especially with the latest news on the Weldon family), three races in which only one of the four analysts now give the Democrats the lead.  The Reynolds race in New York 26 also appears to be in great jeopardy.

Several of these races have been seriously affected by the Foley scandal — Florida 16, Ohio 15, New York 26 most prominently. Pennsylvania 10, Pennsylvania 7 , Texas 22, and Ohio 18 are all races with the whiff of local scandals. Florida 13 (open seat now held by Katherine Harris) and Minnesota 6 (open seat held by Mark Kennedy) also appear to have come  much more into play since the Foley scandal broke (more on these later). 

It is quite possible that it will not be the Iraq war that determines control of Congress, but the Abramoff, and Foley scandals, combined with the sleaze in Ohio's state government. Without Abramoff and Foley, the GOP would be coasting in Texas 22, Fla. 16, Fla. 13, Ohio 18, and New York 26, and ahead in Ohio 15. You can't give away as many as half dozen seats in a year when the political environment is bad for your party to begin with and expect to hold on. 

Beyond the 17 races in which at least one analyst believes the Democrat is ahead there are many races that are considered pure tossups, or with a very slight lean to the GOP. In a Democratic wave election, most of these races would go the Democrats, which is why the forecasts of 25 to 30 seats as the high end gains for the Democrats are out there.

Three races are regarded as even by all four analysts — Connecticut 2 (Simmons), Illinois 6 (open), and Virginia 2 (Drake). In 8 other races, three of the four analysts believe the race is even, and one forecasts that the GOP is ahead. These 8 are: Connecticut 4 (Shays), Minnesota 6 (open), Ohio 1 (Chabot), Washington 8 (Reichert), Wisconsin 8 (open), Florida 22 (Shaw), Kentucky 4 (Davis), and Florida 13 (open). The 17 races with the Democrats ahead, the 3 that are even, and the 8 that are very slightly leaning to the GOP, are the real battleground for the Republicans. If the 12 races with the Democrats favored by 2 or more of the 4 analysts, are seen as clearly leaning to the Democrats, then they would need to win only 3 of the remaining 16 tossup races to gain control, which seems not that difficult a task, especially since 5 of the 16 are open seat races, and many Democratic challengers are well funded.

In one other race, Connecticut 5 (Johnson), one of the four analysts believes the race is a tossup, and the other three believe Johnson is ahead. There is no other House seat in the country now held by the GOP, where any of the four believe the race is even. All see the GOP ahead in all other races. This would seem to suggest that the more extravagant forecasts by some, like James Carville, of potential 50 seat pickups for the Democrats is not likely.

Nonetheless, in some of the other races, the GOP candidate is viewed as only slightly ahead, and the race is not over. A few Republican—held seats beyond the 29 mentioned so far could turn over.  Seats in this category mentioned by all 4 analysts include: Arizona 5 (Hayworth), California 11 (Pombo), California 4 (Doolittle), Kentucky 3 (Northrup),  Nevada 3 (Porter), Pennsylvania 8 (Fitzpatrick), New York 20 (Sweeney), and New York 29 (Kuhl). Three of four analysts include a few other seats as potential takeovers for the Democrats: Nevada 2 (open), New Jersey 7 (Ferguson), and Texas 23 (Bonilla). In all, these 11 races, added to the other 29 put 40 GOP held seats on the board as potentially competitive.

A major problem for the GOP this cycle has been the paucity of seats now held by Democrat that are targets for GOP pickups. In fact, there is not a single Democratic held seat in the country in which any analyst believes the race is even or the GOP challenger ahead. The closest races are believed to be the following five: Illinois 8 (Bean), Georgia 8 (Marshall), Georgia 12 (Barrow), Iowa 3 (Boswell), and the open seat in Vermont. Three other Democratic held seats are mentioned as long—shot possibilities: Texas 17 (Edwards), West Virginia 1 (Mollohan), and Ohio 6 (open). The national GOP seems to be focusing on Illinois 8 and Georgia 8 in the stretch drive.

I think a Democratic takeover of the House is likely at this point, if not certain. The most likely scenario is a small Democratic majority, not a blowout win taking the Democrats to 240 seats. 

Finally, a few words about the Senate.  As with the House, Democratic prospects are much brighter than earlier in the year. There is a chance of a Democratic takeover, though it is less likely than in the House. In the Senate races, there has been much less movement the last month than in House races.

The Democrats remain solidly ahead (5—10 point leads) in 4 Republican held seats — Pennsylvania (Santorum), Montana (Burns), Rhode Island (Chafee), and Ohio (DeWine). Two GOP held seats are neck and neck races — Tennessee (open), and Missouri (Talent). George Allen (R) holds a slight lead in Virginia, and Robert Menendez (D) a small lead in New Jersey. The best GOP scenario would be to lose only 3 of its seats , and win New Jersey, Ohio, Missouri, Tennessee and Virginia. 

This would leave the GOP with 53 Senate seats, and is an unlikely scenario.  Winning Ohio is not impossible, but would be difficult at this point. More likely is for the Republicans to lose the 4 of their seats in which they are now trailing, and split the remaining 4. This would leave the Senate at 50—50. If the three Southern/border state races are all held, the GOP will do no worse than 51 seats. 

In a worst case scenario for Republicans, the GOP would lose all the close races. In this case the Democrats would wind up with 52 seats. This also is unlikely. Republicans still hold out hopes of a victory in Washington (Cantwell) and Maryland (open seat), and Democrats have not given up on Arizona (Kyl). But these three races seem to be firmly in the hands of the party now holding them at the moment. At one time, Democrats thought Jack Carter could run a strong race in Nevada, but he appears to be as ineffectual  a candidate as his father Jimmy was a President. The likely Senate outcome is for each party to be in the 48—52 range; hence a 50—50 outcome can be thought of as a central tendency at this point.

Richard Baehr is the chief political correspondent of American Thinker.