Looking ahead to election night, I thought I'd offer some optimistic assessments of how it might go, hour by hour. Obviously, it's a lot easier to do an optimistic Obama assessment than an optimistic McCain assessment, given that realclearpolitics.com gives Barack Obama a significant lead in states with 274 electoral votes, four more than are needed to win, and John McCain a significant lead in states with only 132 votes. But if this election at this point seems less close than 2000 or 2004 did at this stage, it also seems closer than 1988 or 1992 or 1996 did at this stage.
So let's start with the optimistic Obama scenario, with the states arranged according to the last poll closings in those states (all eastern time). The networks, including Fox News, where I will be working, do not report results until those times, which are the earliest that any state can be "called" for either candidate.
Comments: GA: Current polling shows Obama narrowly behind; even taking the Obama-optimistic assumption that he's doing a little better, an accurate exit poll would make this state too close to call. IN: Ditto. VA: Here the assumption is that the October 1-27 polling (which with just a few exceptions showed Obama 6 to 12 points ahead) is right while the post-October 27 polls (showing Obama 3 or 4 points ahead) is wrong.
Comments: The most recent NC polls are so close that even a somewhat better result for Obama would not make this state callable when polls close. In Ohio, the October 29-30 Mason-Dixon poll is the first nonpartisan poll to show McCain ahead (by 2 percentage points) since an October 19 Fox News poll and an October 16-17 Mason-Dixon poll. Until Mason-Dixon got into the mix, the realclearpolitics.com average edge for Obama in OH was 5 percentage points or more.
Comments: Obama has been leading FL polls pretty consistently since September 28 but seldom by more than 5 percentage points. So I think that it will be hard to call it off the exit poll or until a considerable number of returns have come in. Everyone remembers how FL got miscalled and miscalled again in 2000, and it's a diverse enough state that you can't assume that an unusually good showing for Obama in election returns from one part of the state is going to be replicated in other parts. But super-Obama optimists can put it in the clear Obama category if they like. As for MO, even an Obama performance on the high side of recent polling leaves it in the TCTC category. In PA, the recent close Mason-Dixon and Rasmussen polls still leave the realclearpolitics.com Obama margin at 7 percentage points.
Comments: Obviously, at this point, the results look pretty woeful for McCain. The 86 Obama electoral votes not yet recorded from NY (31) and CA (55) would put him up to 221, and that's without the 79 votes we've put in the TCTC category. But as reports of actual votes in exit poll precincts come in, enabling us to calculate WPE (within-precinct error, which in the primaries tended to overstate the actual Obama vote), we can have a better idea of how much to trust the exit poll, and as actual election returns are transmitted from the states over the Associated Press wire, states that couldn't initially be called off the exit poll can be called. This is an Obama-optimistic scenario, so let's assume that enough GA returns have come in to show us some time before 9 p.m. that Obama has reduced the usual Republican lead in the Atlanta suburbs and that Obama, with unusually high Democratic percentages in the Indianapolis area, is roughly equal with McCain in most of Indiana, with Lake County (Gary, 26 percent black) holding out its results as it did in the primary (then, to prevent Hillary Clinton from giving a victory speech on primary night). So in this scenario, GA and IN go for Obama, and the running totals look like this before 9 p.m.
Comments: The most recent polls have shown McCain ahead by only a hair in AZ. We're assuming Obama does a little better than his current realclearpolitics.com margin of 5 percent in CO. Some Obama enthusiasts may argue that this scenario should give him the 1 electoral vote allotted to the Second Congressional District of NE; I don't think the few statewide numbers support that. ND may be close, but it's unlikely to be callable for Obama, particularly since few if any of the people on decision desks have any experience with any close elections in the state.
Comment: No state's polls close at this hour. But by this time, it may be possible to call NC for Obama (remember, this is an Obama-optimistic scenario); it depends on how fast the tabulated vote comes in. As for the other TCTC states, the results in FL and MO depend too much on jurisdictions that to my recollection tend to report late—Broward, Palm Beach, and Miami-Dade counties in FL, Jackson and St. Louis counties and St. Louis City in MO. All but Miami-Dade are heavily Democratic; Democrats in St. Louis City have a habit of getting a local judge to hold open the polls there for an hour or so after they close in the rest of the state. So let's say NC is called for Obama, while FL, MO, AZ, and ND are not. The totals going into 10 p.m.:
Comments: I think the best an Obama optimist can hope for from MT is TCTC. But I think that the current 5.8 percentage-point margin for Obama in the realclearpolitics.com average for NV justifies it as clear Obama in an optimistic scenario.
This puts Obama over the top just after 10 p.m. But without calls in GA, IN, and NC, he would be at 233, well under 270. And note that in this optimistic-for-Obama scenario, we are assuming he wins some close states even where he is behind in current polls, as well as those where his current poll leads are extremely small, in addition to those where the most recent polls show McCain as trailing. Even so, help is coming at 11 p.m.
Comments: OR votes entirely by mail; Edison-Mitofsky conducts a telephone poll of those who have voted, but decision desks will need a fairly big Obama margin to justify calling the state for him, given the limitations of phone polling. There's a similar, though more limited, problem in WA, where more than half the votes will probably be cast early or absentee. But the optimistic Obama scenario is not troubled by these things. And, to leave the optimistic Obama scenario behind, and to project from current polls, there is a good chance he will be declared the winner of the election at 11 p.m. But there's a serious chance he might not, depending on how the actual tabulated vote comes back in states impossible to call from the exit polls alone. Finally:
Now let's turn to the optimistic McCain scenario. I am going to indulge optimism a bit more on this side, given that McCain is where he is in current polls.
Comments: Give McCain a little bit better than his current 3 percentage-point lead in GA, and it should be callable for him. But I don't think McCain optimism gets you to a call in IN. The four most recent polls in VA show McCain behind by only 3 percent or 4 percent; this justifies a TCTC rating.
Comments: We're putting FL into TCTC, on the same line of reasoning that we put OH there; since October 22, only one poll has shown Obama leading by more than 5 percentage points, while most have been in the 2 percentage-point range. Polling in MO since October 22 has given McCain a slight edge but not enough to take it out of TCTC given its history of late-reporting, heavily black jurisdictions. McCain has made a last-minute appearance in NH, but polling there even discounted for McCain optimism doesn't justify taking it out of the clear Obama category. Our most controversial position here is putting PA in the TCTC category. One possible justification is that polls have shown the votes in the various regions to run counter to longtime patterns, as I have written on here and here.
Comments: The attentive reader will have noted that Obama passed McCain with the 8 p.m. states but that a lot of states are stacked up in the TCTC category. Let's assume, in line with McCain optimism, that as the election returns pour in from IN and NC, it becomes clear that Obama has fallen short and that these states can be called for McCain. This leaves VA, OH, FL, MO, and PA as TCTC.
Comments: Optimism gives McCain his home state. And recent polling tends to justify a McCain optimist in calling CO TCTC. Several MI polls have shown high percentages of undecideds (a Bradley effect in a state where Detroit's Mayor Kwame Kilpatrick was forced out of office and is headed to jail?), but Obama has been pretty consistently running over 50 percent, and the McCain campaign notoriously pulled out of the state a month ago. Have the polls tightened up enough in NM to justify a McCain optimist rating it TCTC? I don't think so. But perhaps McCain optimism is justified in ND. Note that under these assumptions, McCain passes Obama in electoral votes assigned.
Comments: No poll closings at this hour, but perhaps we will have enough election returns, in this optimistic McCain scenario, to say that FL, OH, and MO are going to McCain. Are these optimistic projections? You bet they are. They suggest that current polling has been farther off than polling usually is—or that there has been some late swing in McCain's direction from when the last polling was conducted (October 31 in FL and OH, October 30 in MO). But this is an optimistic McCain scenario. If you want to read a straight-line projection from current polling, look elsewhere.
Comments: McCain's lead in current polling in MT is small, but optimism raises it to clear McCain. The most recent three polls in NV show Obama with 4 percentage-point and 5 percentage-point leads, and McCain optimism converts that to TCTC.
Our optimistic McCain scenario has left McCain with a narrow lead over Obama in electoral votes, and with four states still TCTC: VA (13), PA (21), CO (9), NM (5). McCain in this setting can win without PA, but in that case must win in all three of VA, CO, and NM. Or if he wins PA, he can get to 270 by winning any one of those three while losing the other two.
All of this, of course, assumes some shift of opinion from where polling indicates it is—you have to do that if you are optimistic for McCain, because he is plainly behind. Such a shift is possible, but in the view of most of the people I talk to, of both parties, unlikely.
Still, I think this vindicates McCain's strategy of seriously contesting PA. Without PA in the TCTC category, even this optimistic strategy leaves McCain short of 270 unless he carries VA, CO, and NM, the latter two of which have looked very unlikely all summer and fall. And I think the evidence is pretty strong that McCain has shifted the race some distance in his way in PA over the past month. Whether it's far enough is far from clear.
I find it fascinating that, by indulging plausible optimistic views in favor of each of two candidates (all right, maybe stretching it a bit for the one who is behind), you can come up with such different electoral vote totals—352 to 132 for Obama in one case, 247 to 243 for McCain in the other. Back in April, I speculated on what would have happened if Mitt Romney had gotten 3 percent more of the vote and John McCain 3 percent less in the Florida primary January 26 and the Super Tuesday primaries February 5. With the Republicans' winner-take-all delegate allocation rules, the difference would have been enormous. After Super Tuesday, McCain actually led Romney by about 516 to 207 delegates. If just 3 percent of voters had switched from McCain to Romney, I calculated that Romney would have been ahead by 382 to 365 delegates. The Republican race would have looked entirely different.
So it is with these optimistic scenarios. The Electoral College is a winner-take-all system, and small changes in public opinion can make an enormous difference in the result. To be sure, opinion in the primaries tends to be more fluid than in general elections, but we have seen shifts of 3 percent in the last couple of days in presidential general elections before. In the realclearpolitics.com map of how all the states would go if they voted for the candidate now leading (by however small a margin) in the polls, Obama gets 353 votes and McCain 185. But if you subtract 3 percent from Obama's poll numbers and add 3 percent to McCain's, six states with 89 electoral votes shift from the Obama column to the McCain column and McCain leads in electoral votes by 274 to 264. Which looks a lot like our optimistic McCain scenario. But the betting today is still a lot closer to the optimistic Obama scenario.