Final pre-election analysis/predictions

November 2, 2008

Mac Zilber

I have written over 100 articles on this election, and it’s safe to say that I could literally fill a book with what I’ve written on it. It’s rather surreal to see it coming to a close. I have a combination of excitement, as the event that I have essentially written a thesis on is finally approaching, and melancholy, as there will be a certain emptiness in all of us once the most entertaining election in decades is finished. Without further self-indulgence, I will give you what you came for: analysis, tips, and predictions.

The Presidential Race

The main event Tuesday, obviously, is the contest that will decide who becomes the most powerful person in the world. Frankly, the race isn’t very close, with the final Gallup poll showing Barack Obama ahead 53-42. Obama is likely leading by somewhere from 6-9 points, not by 11, given the polling average, and it is worthwhile to go over why I believe there will be little noticeable “Bradley effect.”

The first reason that there is likely to be no Bradley effect is because there is little evidence the Bradley effect exists, if it ever did exist. The Bradley effect is a theoretical phenomenon that predicts that white voters don’t want to tell pollsters they’re voting against a black man, because they don’t want to seem racist. Some of the commonly cited cases of this effect, such as Harold Ford’s senate race in 2006 and Tom Bradley’s first gubernatorial run, had no notable polling inaccuracy, but there have been a few notable cases of the Bradley effect showing up. In 1991, David Dinkins and Douglas Wilder, two black men running for office in New York and Virginia respectively, were leading by double-digits in late polling and ended up winning by only 1-2 points. In 2006, Michael Steele, the former lieutenant governor of Maryland, was essentially tied in late polling in his senate race but ended up losing by a substantial margin. The two former cases happened nearly 20 years ago, while the latter case was not so much a Bradley effect as a Republican effect; Republicans across Maryland underperformed polling in 2006, not just Michael Steele. As you can see, the pieces evidence that a Bradley effect does exist are few and far apart. Even if the Bradley effect did exist, you would be unlikely to see it decide the election. Contrary to what pundits are saying, there are very few “undecided” voters in this election, with about 2-3% of voters identifying as such. If John McCain were to pick up every undecided white voter, he would gain only about 1-2 points, which is not nearly enough to put him over the top. The other factor that must be reckoned with is the fact that there are numerous reasons to believe that many polls are lowballing Obama’s numbers, not overstating them. Many pollsters do not include cellphones in their samples, and therefore tend to under-sample youths and minorities, who use cell phones at a disproportionate rate. Pollsters who do include cell phones in their samples have Obama doing 2-5 points better on average than those who don’t, and, as such, the pollsters who do not have Obama up by 8-13 are likely deflating his numbers. Another reason is that Obama has the best organization of any campaign in the history of the country. With literally about 5 million volunteers and thousands of paid organizers, Obama is likely to have a get-out-the-vote operation the likes of which we have never seen. McCain, on the other hand, has about as many volunteers nationwide as a well-financed senate candidate, and virtually every McCain field office in the country is nearly empty most of the time, according to fivethirtyeight’s Nate Silver. This disparity could also help Obama over-perform the polls. Finally, many surveys have shown that when the third-party candidates’ names are included in polls, the inclusion of their names takes support away from McCain. This could make the difference in states like Georgia and Montana, which have strong support for third-party candidates and are expected to go down to the wire. All of these factors added up indicate that pollsters are, if anything, underestimating Obama’s support, rather than overestimating it, and that any Bradley effect there might be is a rather moot point. Polls are not given much credit by the public, but they have predicted every election correctly within a few points since 1960, and there is no reason to believe that is likely to drastically change.

A quick thing to note, while we’re on the subject of the polls and whether or not they’re accurate, is that the exit polls should be expected to be incredibly inaccurate this year. One reason is because of the fact that so many Americans have voted early this year; almost one third. Since early voters have broken for Obama 2-1, there is an argument to be made that exit polls will undersample them and overrate McCain’s chances. On the other hand, if early voters are properly gauged, then the exit polls will probably overstate McCain’s chances. Why? Because the Democratic candidate has had his numbers overstated by exit polls in every election since 1988, sometimes by as many as 5-6 points. Furthermore, the one place that one would expect a Bradley effect to show up would be in exit polls, as, while most people don’t have any problem telling the truth to a robot on the phone, many people might feel pressured to lie to a pollster in person. This was borne out by the fact that during the primaries, Obama overperformed in the exit polls by seven points on average. Thus, if I would you I would pay zero attention to exit polling information, as it may by off by as many as 5-7 points in favor of either candidate, which is an enormous margin of error.

Now that the issue of whether or not the polls are accurate is out of the way, we can move on to what you should watch for on election night.

3:00 - The first state we are going to get results from, assuming they have a better votecounting process now than they did in the primaries, is Indiana. The early results from Indiana will be a bit misleading, because they will be leaving out Lake County and LaPorte county, which are both Democratic strongholds. I would avoid reading too much into the early Indiana results, but if they’re close, it could be indicative of a good night for Obama.

4:00 - Virginia, Florida, and the western portion of Indiana have their polls close at 4:00 (several other states do too, but they don’t really matter that much). We don’t know exactly when we will get the results from these states, but the odds are that Virginia will be the first swing state that is called. If Obama wins Virginia, which he probably will, he will be the president. The networks may not call the election for him then, but that will be the point when those who are in the know will know that he has prevailed. If Obama wins Indiana or Florida, it also means he has won the election, but, while he is likely to win at least one of these two states, neither of them is going to tell us who won or who lost. Obama can lose Indiana and Florida and still have an 85% chance of winning the election, while McCain cannot win if he loses either state. Also worth watching is the Georgia race, as a victory in Georgia for Obama is both indicative of a landslide, of a groundswell of support in the south, and of the Democrats winning 60 senate seats.

4:30 - North Carolina, Ohio, and West Virginia’s polls close at this time, and by the time these states are called, we will probably know who has won the election. If Obama wins Ohio or North Carolina, he is the victor. If Obama loses North Carolina, it doesn’t mean much (though if he loses it by more than 5 points, there might be some concern), but if by this point he has lost Virginia, Indiana, Ohio, North Carolina, and Florida, then this election is going to be a nail-biter. If Obama wins West Virginia, which I highly doubt is plausible, then it means that he is drastically over-performing his polling numbers and may reach as many as 401 electoral votes.

5:00 - Pennsylvania is probably the only state that matters that has its polls closing at this time. Obama is almost certain to win Pennsylvania, and Pennsylvania is perhaps the single most important state for McCain to win if he wants any chance of coming out of this election victorious. If McCain has won Ohio, Florida, North Carolina, Indiana, Missouri, Virginia, and Pennsylvania by this point, then he’ll probably win. If he has won Ohio, Florida, North Carolina, Indiana, Missouri and Pennsylvania, but hasn’t won Virginia, then the election is going to come down to Nevada. Even though Missouri is perhaps the closest state in the polls right now, and closes at this time, I don’t really think there’s much use in saying anything about it because if McCain loses Missouri, he loses the election, and there isn’t really much more to say on the subject.

6:00 - The polls in Colorado, New Mexico, North Dakota, and Arizona close. Putting New Mexico on the list of states that matter is more of a habit than anything; it was near the top of the list for so long. New Mexico is almost certainly going towards Obama, as is Colorado, due to the combination of early voting and a heavy advantage in the polls towards Obama. Colorado is probably one of the four most important states, along with Pennsylvania, Virginia, and Ohio, but it is unlikely to be the difference-making state, for the sole reason that Obama leads by 10 in the most recent poll of the state. North Dakota is an interesting state, as there has been a very small amount of polling done and it has all showed a dead-even race. For polling junkies like myself, North Dakota is the single hardest state to predict in this election. The main reason North Dakota is of interest, however, is because that would be the state that would be indicative the impressiveness of his victory, as Bush won the state by 28 points and it hasn’t gone to any Democrat since Lyndon Johnson. Arizona, similarly, is a state that would only be window-dressing on an Obama victory, as it is McCain’s home state, and is one of the 6 states that LBJ couldn’t capture. Obama likely won’t win Arizona, but he trailed by only one percent in a recent poll there, and McCain is visiting his home state today, so evidently it is in play.

7:00 - The last states that are in play close at this time: Nevada and Montana. Nevada is one of the five states that could actually make a difference in this election, and, if the election is incredibly close, it could come down to this state. The likelihood is that we’ll know who won the election by now, but, if we don’t, we will by the time Nevada is called. Montana is a state that I doubt will go for Obama, but it has polled within the margin of error and is notoriously difficult to poll.

9:00 (10:00 for a few counties) - The Alaskan polls close at this time. Originally thought to be a race that McCain had in the bag, due to Palin, there was a poll that came out today showing McCain up by only 3.5 points. What should be intriguing about Alaska’s election is the fact that, since the race may be called by as early as 3:30 in Alaskan time, we may see a scenario in which few Alaskans bother to vote. On the one hand, that could favor McCain, as Alaska may be one of the few states in which McCain has more enthusiastic supporters than Obama does, but it could also favor Obama, as people will be going to the polls to vote against Ted Stevens and Don Young, rather than to vote for president. I have an unscientific hypothesis that a not-insignificant portion of the voters who were planning to vote against both Republican congressional candidates (60%+) will end up siding with the Democrat for president as well, which could make the difference if the presidential race became a minor race because it was already decided.

My prediction is that Obama will probably win the election with somewhere from 353 to 364 electoral votes, and there probably won’t be much drama in the outcome. Many of the other elections that are happening tomorrow will have far more dramatic finishes.

California’s Ballot Propositions

Proposition 4 and 8 are likely to be the closest races of the night, as the polls have been consistently within the margin of error. Proposition 4, which would force teens to notify their parents before having an abortion, is more likely than not to pass, while proposition 8, which would ban gay marriage, is more likely to fail than to pass, based on the most recent polling. The bullet-train proposition, proposition 1A, is also likely to come right down to the wire. It is currently leading 47-42 in the polls, but usually the majority of undecided voters vote “no” on propositions. Propositions 3, 10, and 12 look very likely to pass. Prop 3 is a bond measure for building children’s hospitals, proposition 10 would enact some of T. Boone Pickens’ environmental reforms (which are opposed by nearly every major environmental group), and proposition 12 is a measure to give loans to veterans.

Senate Races

The most interesting senate races tomorrow will be in Minnesota and Georgia, as these are the two true tossup seats. In Minnesota, it should be interesting to see whether independent Dean Barkley, who currently is polling at about 15%, will have his support hold up. If he can maintain that support, Republican Norm Coleman might have a shot at victory. If Barkley loses a third or more of his support, then Al Franken will likely carry the day. In Georgia, Republican Saxby Chambliss will likely emerge with the most votes if there is no wave of black turnout. What could make this the single most interesting race to watch tomorrow, however, is the fact that it takes 50% to win Georgia’s senate seat. The libertarian candidate is expected to get 2-5% of the vote, and if the race is close enough, this could mean that nobody will reach the 50% barrier required to win the seat outright, and the race will go to a runoff on December second. There is a good chance that such a runoff would determine whether or not the Democrats would win a filibuster-proof majority in the senate, and, as such, the runoff race may end up being one of the most widely followed senate races in history, as all the other races will be over and there will be a huge amount riding on the results. My prediction is that Chambliss will win tomorrow, but I think he’ll struggle to get a runoff-proof victory.

Much of the rest of what I have to say about the senate races already has been discussed in my previous articles on the topic.

Local Elections

Each city council candidate has their own reason to feel confident and their own reason to feel nervous. Mayor Lieber can feel confident knowing that he has the most name-recognition and has spent the most money of any candidate in this election by far, but can feel nervous in knowing that certain elements of the city have built up an antipathy towards him and that he may receive a backlash from the “sour grapes” leaflet. Councilor Javandel can feel confident in being an incumbent who doesn’t really have many detractors, but can feel nervous in the fact that “Javandel” yard signs are by far the least common yard signs in Albany, due to Javandel having less than one tenth the budget of any other candidate. Ellen Toomey can feel confident in the fact that “Toomey” yard signs are incredibly common, and in the fact that she has all the prominent endorsements, but may feel a bit uneasy in the fact that her alliance with Lieber, Panian, Atkinson, and Wile, while it hasn’t entirely disappeared, has become less strong over the last month of the campaign. Leo Panian can feel confident in the fact that he has one of the best-financed campaigns, and in the fact that his predecessors, Marge Atkinson and Joanne Wile, prevailed in 2006, but can feel nervous in the fact that he has been less successful in branding himself and differentiating himself from the other candidates than his opponents have. Peggy Thomsen can feel confident in the fact that she has raised more money from citizens than anybody except Lieber, but can feel nervous because of the fact that she was the individual whose positions the “Sour Grapes” flyer most egregiously misrepresented. Nick Pilch can feel confident due to the fact that he has spent more money than any candidate except Robert Lieber, but may be uneasy in the fact that, despite these expenditures, he has struggled to gain name recognition. Overall, we have no polling data to predict what will happen in the municipal races tomorrow, and we only can guess based on unscientific and qualitative factors.

The Future of the Muse?

Just because the election will be over in a couple of days won’t stop this blog from being updated frequently with relevant news and analysis on local, national, and international government. The Muse isn’t going away any time soon.

And that’s the muse

Exercise your freedom of speech: Send an email to themuse [at] albanyhighcougar [dot] net and I will answer any question or comment you may have.

2 Responses

1. 5:56 PM on November 3rd, 2008maxine conyers-lee:

   I need to print this out and refer to it tomorrow. You’re the most comprehensive one-stop news source I know of.

2. 8:22 AM on November 4th, 2008Griff:

   Are you related to Jon Sitter?!?

Leave a Comment

All comments are moderated. If your comment contains profanity or libelous content, please don't waste our time.

Name: Some basic HTML allowed

Your comment will be posted when it is approved.