The sprint to the finish
The Baehr Essentials
A week from today, I will provide my state—by—state forecast for the Presidential election and the Senate. So that gives me one more week to hedge. Both President Bush and Senator Kerry still have a realistic shot at being elected on November 2nd (or sometime after that in the courts). President Bush, however has a little bit of a wind at his back, in what still shapes up as a very close race. In the Senate, the GOP is doing much better, and is positioned to pick up several seats.
Starting with the Senate races, both Georgia (a pickup for the GOP with Johnny Isakson) and Illinois (a pickup for the Democrats with Barack Obama) have been conceded by both parties for months. But the Republican candidates have now moved ahead in several open seat races. Former Rep. Tom Coburn is a few points ahead of Democratic Congressman Brad Carson in Oklahoma in most surveys, though one newspaper poll out today shows Carson with a 7 point lead. With President Bush expected to win the state by over 20%, Coburn looks like he is now favored to hold Don Nickles's seat for the GOP.
In South Carolina, Republican Congressman Jim Demint has made a few missteps (as has Coburn in Oklahoma), but appears to have maintained a 4 or 5 point lead over Democrat Inez Tenenbaum, to replace Democrat Fritz Hollings. This would be a pickup for the GOP. In North Carolina, Republican Congressman Richard Burr has moved ahead of Democrat Erskine Bowles for the seat being vacated by John Edwards. This one is not 'done' yet, but Burr has all the momentum for another GOP pickup. In Florida, there are mixed polling results but most surveys show Republican Mel Martinez ahead of Democrat Betty Castor, though by small margins in virtually all cases, in the race to succeed Democrat Bob Graham.
In Louisiana, the election on November 2nd will only be decisive if one of the candidates running reaches 50% of the popular vote. If no candidate hits 50%, then a runoff will occur several weeks later, with the top two facing off. At the moment, Republican Congressman David Vitter is well ahead of three Democrats, but a few points short of 50%. If a runoff occurs, this race would be a tossup, but Vitter has a chance to hit 50% with Bush's strong effort in the state. Republicans could win all five open seats in the South, all now held by Democrats. If the Presidential race were not so close, I would expect a Bush visit to Louisiana to help Vitter hit 50%. If there is a runoff, this will surely occur.
Three races remain too close to call. In Colorado, the Republican nominee Peter Coors has battled from behind to pull even with Democrat Ken Salazar. The polls are in rude disagreement on this one, ranging from a 7 point Salazar lead (Zogby) to leads of 1 to 5 points for Coors. In Alaska, former Democratic governor Tony Knowles has maintained a small lead throughout the race against GOP incumbent Lisa Murkowski. President Bush will win the state by more than 20%, and may pull Murkowski through, but Knowles is probably a slight favorite.
And finally, there is the mother of all races this cycle— the South Dakota Senate seat. Over thirty million dollars have been poured into a state with only 700,000 people, and after all the ads and door—to—door campaigning, the race appears tied. In 2002, Republican candidate John Thune lost by just over 500 votes, as Democrats mined votes on Indian reservations for incumbent Tim Johnson. In 2002, Democrats used the argument that Johnson's re—election was needed to preserve Tom Daschle's power as Majority Leader. Johnson won, but the GOP retook the Senate anyway, so this argument will not work as well this time around, since Democrats' chances of regaining control are slim to none. Polls show the race tied, but I think Daschle is a slight favorite.
I feel pretty safe saying that Democrats will not recapture the Senate and the GOP is poised to pick up a few seats. There is one state, however, where Republicans could in for a nasty surprise. Kentucky Senator Jim Bunning was coasting to an easy re—election victory until a few weeks back. Then a series of odd statements and a weak debate performance have tightened the race considerably. Two recent polls show Bunning up over his under—funded Democratic opponent Dan Mongiardo by 6% to 14%. If the 6% number is correct, then Bunning may be in free fall, and this seat could be up for grabs. Bush will win the state by 15—20%, and maybe Bunning's best hope to pull through.
The Presidential race is a lot trickier. Rather than states falling out of play, amazingly enough, more states seem in play this week than last week. Examples include: Hawaii (4) where two weekend polls put Bush up by a point. Hawaiians have not voted for Republicans for President except a few times in their re—election bids. Hmm. Gore won Hawaii by 18% in 2000. On the other hand, Arkansas (6), a state assumed to be safe for Bush given the GOP tide in the Deep South, is tied according to one newspaper poll which appeared over the weekend. I tend not to believe an outlier poll, if there is no confirmation from other sources. Two weeks ago one poll showed Bush up 5 in Oregon (7). All other surveys have Kerry ahead, and I think he is likely to win the state. I suspect Arkansas is still a Bush state, but by less than anyplace else in the South except for Florida (which is a Southern state only in its North), and perhaps Virginia (13). Polipundit reports that a new Connecticut poll will be out today showing that state tied as well. President Bush has been doing better in many states in the Northeast than he did in 2000, though he still trails in all of them — New Jersey (15), Maryland (10), Delaware (3), Connecticut (7), among them. The proximity of the 9/11 attacks seems to have moved some former suburban Republican voters, who had strayed to the Democrats in the Clinton era, back to the GOP. But I suspect the trend is not strong enough in any of these states for a Bush victory. Assuming all these states are just teases, the real race will be fought in eleven states, five of them won by Bush last time, and six by Gore.
Bush is clearly ahead in two of the contested red states from 2000 — Nevada (5), and Colorado (9). No poll has put Kerry ahead in Nevada for months. The same was the case for Colorado until a new Zogby tracking survey showed Kerry up 4 last night. I do not believe it. Colorado has been polled frequently the last two weeks and Bush has maintained 5 point or greater leads in every survey. The Zogby survey also showed Democrat Ken Salazar up 7 for the Senate when most other polls show him trailing by a small amount. I suspect Zogby has a small unreliable sample in Colorado. But Bush partisans may not want to dismiss Zogby as a lousy pollster when they see his other state results — which put Bush up 5 in Ohio, and 3 in Florida, and also have the President ahead in a bunch of blue states — New Mexico, Iowa and Wisconsin. Zogby says Kerry is up 1 in Minnesota and 2 in Pennsylvania, hardly comfortable leads in either state. Only in Michigan, does Zogby have Kerry well ahead — by 10%. Oddly enough, a Detroit News daily tracking poll shows Bush pulling away to a 5% lead in Michigan (17), a stat where most pundits thought Kerry was safely ahead two weeks ago.
Is Michigan really in play? I suspect so — given that Kerry is going there today, and Bush will spend two days there midweek. Eastern Michigan has a large number of African Americans, and Arab and Muslim voters. The Arabs and Muslims will back Kerry strongly this year. But a recent survey suggests Bush could double his vote percentage among African Americans. That plus inattention to the state (no Kerry visits since Sept 15th) may have allowed the state to slip into play. The Bush team has campaigned hard in Western Michigan, and may pull in a big margin in this more conservative part of the state.
The website realclearpolitics.com provides all the daily national and state polls, and the average of the polls in the battleground states . If you believe the averages, then Bush would want Election Day to be tomorrow. For only in New Hampshire (4) is Kerry ahead in a red state, and there by less than 1%. Bush, on the other hand, has the lead in Iowa (7), Wisconsin (10), New Mexico (5), and even Minnesota (10), though by small amounts in each. Bush also leads in Florida (27) and Ohio (20), but in both cases by less than 1%. Rasmussen which has been conducting daily tracking polls in both states for a month, has Bush up 4 in Ohio, and even in Florida.
If Kerry can not win either state, he will be back in the Senate in January. The conventional wisdom has been that he needs to pick off one of the two, and hold Pennsylvania, where he has been ahead consistently in every poll by a small amount for a month. It may be that one will not be enough and that Kerry has to win both to get elected. If Bush in fact sweeps the Upper Midwest — Iowa, Wisconsin and Minnesota total 27 Electoral College votes, the same as Florida. If Bush loses New Hampshire (4), but picks up New Mexico, he could lose Florida and still win by taking the three Midwest states. He could even lose one of the three Upper Midwest states, and still win, if he wins New Mexico, while losing Florida.
It is hard to see Bush winning if he loses both states, unless he wins all the Upper Midwest states, New Mexico, and also steals Michigan. At the start of the year, it would have seemed impossible for Bush to win Michigan, but not Ohio. But the Democrats' registration drive has been far more successful in Ohio than Michigan. Too successful, one might argue given that there are now 848,000 registered voters in Franklin County (home of Columbus) and only 813,000 adults over the age of 18 according to a 2003 US census estimate .
Pennsylvania is another state with a registration anomaly. Philadelphia has lost 13% of its population since 2000, but has 24% more registered voters than in 2000. Bill Clinton will be back on the campaign trail today in Philadelphia, then Miami later today, and Broward County, Florida tomorrow. Clinton's efforts should bolster the Kerry ticket with black voters, single women, and independents, all categories in which he has been lagging.
The national polls have shown a consistent Bush lead — sometimes small — a point or less, and in about half the cases, 5 points or more. Both campaigns have off the record stated they think Bush is ahead nationally by 3 to 4 points. But both sides consider the battleground states to be much closer. Bush's national lead is coming in part from reducing the large Gore margins in the Northeast to more respectable percentage losses this year, and from big leads in many of the Southern states and Texas. Losing a state by less than in 2000 or winning a state by more than in 2000, will not add Electoral College votes in either case, though it might result in a popular vote win, which could add legitimacy to an Electoral College victory.
But a national poll lead is important for providing a sense of momentum to the President's effort. Mickey Kaus has argued that there could be a lawsuit avoidance strategy for some voters, who are loosely connected to either candidate. If Bush is up 3% on Election Day in the polls, maybe some swing voters will select Bush to move his margin outside the margin of lawsuit error. Nobody, other than the DNC, the Kerry campaign, the New York Times, and the 20,000 lawyers waiting in the wings, seem anxious to create a repeat of Florida this time around in a collection of states.
Several small states could also play a role in determining a winner. The Democrats have paid a lot of attention to Colorado (9), and were probably cheered by the Zogby poll last night. While Colorado may be closer than last time, Bush is still ahead. Nevada (5) and New Hampshire were both very close wins for Bush last time. Kerry's campaign admits trailing in Nevada, and his people seem more confident about New Hampshire (4). New Hampshire has surprised Democrats before— John Sununu came from behind to win the Senate seat in 2002. Having a summer home in Maine, I know that not all Northern New Englanders look fondly on Massachusetts politicians ('Massholes' are what they call them), much as they love the Red Sox.
New Mexico is the best small state prospect as a pickup for Bush. After trailing most of the year, he has opened up a lead in every recent poll, sometimes by as much as 5%. The President and Senator Kerry were in the state this weekend. New Mexico is another state where there have been many new registrations, an advantage for the Democrats, if they can turn the new voters out to the polls.
In the most recent two weeks, both campaigns have focused much more on Florida than Ohio. This makes sense for Kerry, given that Florida has 27 Electoral College votes and Ohio has 20. Winning Florida is like taking Ohio, and also Iowa. Florida is also less predictable than Ohio. Florida has Nader on the ballot, probably less than a 1 per cent showing this time around. In Ohio, Nader is off. In both states, federal courts have ruled that provisional ballots can not be counted if the voter shows up in the wrong precinct. Republicans have mounted a vote challenge operation in Ohio, suspecting that many new registrants are duplicate voters, or otherwise not entitled to vote. Ohio has been a more Republican state than Florida in recent elections. But Republican strength in Florida may be underestimated. If panhandle voters support Bush overwhelmingly, much as their neighbors do in Georgia and Alabama, that could offset to some extent a strong Kerry lead in Southeast Florida. In 2000, the panhandle vote dropped off after major news networks incorrectly stated that polls had closed across the state, when they were still open for another hour in the central time zone portion of the state .
Both campaigns are making a huge effort on turning out their voters. It is hard to imagine that there are many undecided voters left unless they stick to ESPN for their campaign coverage.
I would make one pick today — Red Sox in 6, but I won't, for fear of jinxing the Beantowners.