[If you missed it, please see the Program Note for an explanation of the repeat columns this week.]

[Just for fun, right after the Democrats swept the 2006 midterm elections and took both houses of Congress, I thought I would take a crack at making some predictions for 2008. I've never been afraid of jumping in with my analysis long before others get around to it, even if it proves eventually to be laughably wrong.

So here is how the 2008 election looked, to me, a year and a half ago. I concentrated mostly on the Senate, and if I were dividing up the races today, obviously I would switch some of these states from one category to another. But I think my final prediction of 57 Democrats (55 Democrats plus two independents who vote with them) to 43 Republicans is still among the most likely outcomes this year. I could be off by one or two, but I bet not much more than that.

So here are my ridiculously-early 2008 predictions, to provide an interesting look back.]

[This column originally ran November 29, 2006, just after the 2006 midterm elections.]

My 2008 Election Picks

At the risk of appearing slightly delusional, allow me to predict the 2008 election. No, not the main Hillary / McCain / Obama / Giuliani fracas, as that field is pretty crowded with wild predictions already. Pundits everywhere are focused on the presidential horserace, but there's going to be a bigger battle fought in 2008 for the Hill. So I'd much rather attempt to predict the outcome of the Senate.

The ideology and spin are already lining up on both sides of this contest. On the right, the talking points are all about how Republicans lost the past election by not being conservative enough. True Conservatives (the story goes) are going to rescue the party in 2008 by returning to their small-government roots, which will win them both houses, beginning anew the reconstruction of the Permanent Republican Majority.

On the left, the new conventional wisdom is that "the 2006 election was the end of the Goldwater/Reagan conservative era, forever." The public has woken back up, and will continue to reliably send Democrats to Congress from now on.

It should be noted that at least one of these viewpoints is going to be proven wrong in 2008. So what are the Democrats' chances of further gains in Congress in the 2008 elections?

An early look at who will control the 2008 party campaign committees comes from Chris Cillizza at washingtonpost.com, which is well worth a read. For the Senate (he reports), Chuck Schumer's amazing success in 2006 was rewarded by giving him a second term as Democratic Senatorial Campaign Committee (DSCC) Chairman. This is good news, as he ran against overwhelming odds and still pulled the "majority" rabbit out of his hat. Even Robert Novak is singing Schumer's praises these days.

The jockeying for the Chair of the Democratic Congressional Campaign Committee (DCCC) gets a little more interesting. There are three candidates up for Speaker-Elect Pelosi's consideration. Two of them are protégés of Rahm Emanuel, and the third is a protégé of Pelosi herself. Considering how successful each of them would be has to be colored by your thoughts on Rahm Emanuel's success at gaining a majority in the House this year. There are two camps: the Deaniacs who insist that Howard Dean's Democratic National Committee (DNC) and their "50-state strategy" won the victory; and then the supporters of Emanuel who will tell you the gains are all due to the DCCC. Look for whoever Pelosi ultimately selects to have to prove themselves to the other faction.

The Numbers
Calling 435 races for the House two years in advance is beyond my crystal ball's abilities; but in the Senate there will only be 33 races, making it much easier to prognosticate.

First, some raw numbers. Republicans hold 21 of the seats up for election in 2008, the Democrats only 12. This is good news or bad news, depending on whether you believe it's easier to play offense or defense. Almost all agree, however, that it's easier to take a vacant seat than trying to unseat an incumbent. So the next set of numbers to examine is age.

Democrats seem to have an advantage here. Of the fifteen Senators who will be 65 or older on election day, nine of them are Republicans. If you look at the 70-or-older group, it gets even better: of the ten total, seven are Republicans, and only three Democrats. As the example of Strom Thurmond proves, however, age alone doesn't mean a whole lot. Frank Lautenberg [D-NJ] (who will be 84 on election day); Ted Stevens [R-AK] (who will also be 84); and John Warner [R-VA] (who will be 81) may all decide to run -- and all may even win. But as insurance actuaries will tell you, a certain percentage of the oldest members may not be running due to voluntary retirement, health reasons, or even death. Actuarial tables can give you accurate percentages, but they cannot predict which individual Senators will be affected. The raw numbers show that there are more older Republican Senators than Democrats, so there will probably be more Republican-held seats open in 2008 than Democratic-held seats.

Before I get to my actual predictions, a few caveats. I'm no statistician, so I'm probably going about the process all wrong (although I humbly point out I accurately picked every Senate race this year, so I'm batting a thousand so far...). The main data I considered were: the vote split between the last two Senate races in each state, the presidential vote in that state in 2004, and (to a lesser extent), the current governor's party. I also took into account the age of the incumbent and their seniority (terms served as a senator). What I did not look at (due to lack of time) was the incumbents themselves, their voting records, or their probable opponents.

I have divided the races into six categories, "safe," "at risk," and "easy pickup" for each party. I have to say, the numbers in general look good for the Democrats' chances.

Safe Republican
Total: 10

Alabama, Georgia, Idaho, Kansas, Kentucky, Mississippi, Nebraska, South Carolina, and Texas are on this list. Maine is also most likely to remain Republican, unless Susan Collins decides that the winds are changing in the Northeast, and decides to run as a Democrat (which is not likely). Of the other nine, a case could be made that Kentucky and Nebraska are soft for the GOP, but I don't currently buy it.

At-Risk Republican
Total: 8

Alaska's Ted Stevens will be 84 years old, and hence might not run. If the seat is open, it may be more competitive than you'd think. Republicans won the 2004 Senate race here, but only by a margin of 49/45.

Minnesota's Norm Coleman only won in 2002 by 50/47 (against Walter Mondale and the memory of the recently dead Paul Wellstone), and since then the state has gone for Gore in 2004 (51/48) and in 2006 voted for a Democratic Senator by a 58/38 margin. This will also be the most fun race to watch in 2008 (assuming Al Franken runs).

New Mexico's Pete Domenici has been around since 1973 (six terms), and will be 76 years old. If he steps down, an open race will be very obtainable for the Democrats. The 2006 Senate race went 70/30 for the Democrats, and they've got a Democratic governor. If Domenici runs, however, he'll most likely win a seventh term.

North Carolina's Elizabeth Dole will be 72 years old. Although this state has gone Republican in both 2004 and 2002 Senate races and voted for Bush in 2004, the margins are small enough that with an open seat (if Dole retires), Democrats might have a chance.

Oklahoma's James Inhofe is going to be 73, so the seat may be open. Even so, it's a relative longshot for the Democrats.

Tennessee's Lamar Alexander is on the list since the 2006 Tennessee Senate race was so close. The Democrats would have to field a pretty good candidate to have a chance, however.

Virginia's John Warner will be 81, and has been in the Senate since 1979. If he steps down, and if the Democrats field a candidate as acceptable to the military as Jim Webb was, they may have a chance. Virginia's demographics are undergoing a slow change that benefits Democrats in general. And not only did Democrats win the Senate seat this year, they also held the governor's mansion.

Finally, Wyoming is a real longshot, but worth mentioning. Although the state votes overwhelmingly Republican as a rule, the race for their single House seat (a statewide race, in other words) went down to the wire in 2006. The Republican won, but just barely. And they've got a Democratic governor, so it's within the outside bounds of possibility.

Easy Republican Pickup
Total: 0

I just don't see any seats held by Democrats that would be easy pickups for Republicans next time around. Possibly South Dakota, but I prefer to list it as a risky Democratic seat.

Easy Democratic Pickup
Total: 3

Colorado could be an easy pickup for Democrats. The state's demographics are swinging slowly Democratic, and they just picked up a House seat (one of only seven) in 2006, leaving them with a 4-3 advantage. They also won the governorship in 2006, and the 2004 Senate race went 51/47 for the Democrats. Wayne Allard's last win was with a thin 51/46 margin, back in 2002. If the Democrats come up with a decent candidate, they should have an easy shot at this state.

New Hampshire went through the biggest swing from "red" to "blue" in 2006. In 2004, the state voted 66/34 Republican for the Senate race, but since then Democrats have taken over the state government in a big way. Democrats also won both of the state's House seats in 2006 from Republican incumbents. John Sununu is only a one-term Senator, and won his first race by only 5 points in 2002. Democrats should take this state easily.

Lastly, Oregon seems like an easy pickup for the Democrats. Gordon Smith is a two-term Senator who won his last race in 2002 by a 56/40 margin, but since then the state voted Democratic for President (52/48) and for Senator (63/32) in the 2004 election. If the Democrats field a decent candidate, they should take advantage of this state's blue trend.

At-Risk Democratic
Total: 6

Arkansas started off on my safe list, but in fairness I moved it here. Mark Pryor's only a one-term Senator, and the state has flip-flopped on which party voters prefer. A longshot for Republicans, but a possibility.

Iowa voted Republican in 2004's Senate race 70/28. But Harkin's probably got the seat sewn up. He's a four-term Senator, and won his seat 54/44 back in 2002.

Louisiana politics requires a voodoo doctor to predict in normal times, and these are anything but. Mary Landrieu will probably win, but if the Republicans run only one candidate and there are several Democratic challengers, she could wind up losing. There are many who place a large part of the Katrina blame on her (rightly or wrongly), and that will work against her. But, again, Louisiana's bizarre election system is impossible on the best days to understand. A Republican longshot.

New Jersey's Frank Lautenberg will be 84 years old. If he retires, the seat will be open, and New Jersey politics is a full-contact sport. Anything could happen, even though the state leans Democratic in general.

South Dakota's last two Senate races were razor-thin affairs, with one win to each party. In 2002 Tim Johnson won 50/49 for the Democrats, but in 2004 it went 51/49 against Tom Daschle. So this is a real tossup state, the closest the Republicans have to an "easy pickup" state.

West Virginia's pretty solidly Democratic when it comes to the Senate, but Jay Rockefeller will be 71, so there's an outside possibility it will be an open seat. I almost put this one in the "Safe Democratic" category, though, since I don't think it'll ultimately matter.

Safe Democratic
Total: 6

Delaware, Illinois, Massachusetts, Michigan, Montana, and Rhode Island all seem pretty safe to me at this point. A case could be made that Montana is at risk, but I doubt it at this point.

Adding It Up
So we've got 10 safe GOP seats, 6 safe Democratic seats, and 17 in the middle. If you split the risky states evenly, you come up with a total of 16 Democratic, 17 Republican, with a total pickup of 4 seats for the Democrats (the Senate would then be 55/45).

The best case scenario (realistically) for the Republicans would be holding all their risky seats and picking up half the Democrats' risky seats, for a total of 21 Republican, 12 Democratic -- exactly the split we have now.

The best case scenario (realistically) for the Democrats, using the same formula, would be 14 Republican, 19 Democratic, for a pickup of 7 seats (Senate would be 58/42).

My Picks
I'm going out on an optimistic limb here, and predicting Democrats will only lose one risky state (SD), pick up all three easy pickups, and win four of the Republican risky states (MN, NM, TN, VA).

Here's where it will stand, the day after the election in 2008:

Democratic: DE, IL, MA, MI, MT, RI, AR, IA, LA, NJ, WV, CO, NH, OR, MN, NM, TN, VA

Republican: AL, GA, ID, KS, KY, ME, MS, NE, SC, TX, AK, NC, OK, WY, SD

Total Republican loss / Democratic gain: 6 seats.

Senate convenes January 2009 with: 55 Democrats, 2 Independents (Democratic-leaning), and 43 Republicans.

These predictions will probably turn out to be wildly wrong, of course. But since I'm the only one rash enough to call the races so early, perhaps others can use these numbers as a benchmark for their own picks.

-- Chris Weigant