The Mystery Pollster deals with the concept that challengers pick up undecided voters in the last days of Presidential elections and states that there’s some evidence for it.
The problem is that correlation and causation are two entirely different things. Consider the last few times that a challenger has run against an incumbant President. Each time in recent political history, there has been other factors that are at play here. 1964 had Barry Goldwater’s self-destruction against Johnson. 1972 had McGovern also self-destruct by running far against the mainstream of American politics. In 1976 you had the shadow of Watergate hanging over the head of Gerald Ford. 1980 saw the Reagan Revolution sweep up Carter’s economic and political malaise. In 1992 Ross Perot cost Bush the election.
The problem with extracting a general trend for this is that while undecideds may have swung towards the challenger, that correlation may not have a general causation that holds true from election to election.
In this election, I don’t think it will hold true. First of all, the number of undecideds is very low. A lot of those people just won’t vote. Indeed, both candidates are playing to their bases right now and not really trying to get swing voters at this point. There’s a reason for this. If you’ve not decided now, you’re not all that likely to ever decide. Even if we get a decisive move by fence-sitters, it won’t necessarily shift the race enough to make a huge difference.
Secondly, the GOP has a vastly better GOTV apparatus than they did in 2000. The 2002 election saw Republican turnout that was dramatically higher than predicted. Efforts like the 72-Hour Task Force have put the GOP’s GOTV efforts at a rough parity with the Democrats. In key states like Colorado and Iowa, the Republicans have registered just as many voters as the Democrats have.
And the difference being that Republican voters are far more likely to go to the polls. Your average Republican (statitistically speaking) tends to be older, well-off, married, and a regular churchgoer. Your average Democrat skews younger, less well-off (or ultra-rich), unmarried, and non-religious. Typically the former votes in larger numbers than the latter. Especially once you peel away the fraudulent registrations, I don’t think that the much-vaunted Democratic machine is as effective as people think. In fact, just look at Howard Dean. Dean’s organization was supposed to be a juggernaut – and yet even before “I Have A Scream” Dean’s campaign crashed and burned. The reason is that the Howard Dean/Michael Moore/MoveOn.org activist is a very vocal minority, but a minority in the end. In fact, such people actively turn off moderate voters with the intensity of their anti-Bush rhetoric.
In the end, I think this race will be a squeaker unless there’s a real break in the race. If there is a break in this race, I very much doubt it will be for Kerry. This election isn’t like other elections. It’s a wartime election with the Republicans achieving parity with the Democrats in terms of party ID and GOTV operations in which the President leads on the most important issue by double-digit margins in nearly every poll.
That doesn’t mean it’s a shoe-in for the President, however, it also means that those who are using old models to predict what I still firmly believe will be a realigning election might get quite the surprise when some of the old theories don’t predict what happens in under two weeks from now.
I’m not clinging to the “challenger takes all the undecideds” rule the way John Zogby is. I could easily see a nightmare scenario playing out where the undecideds hold their nose and vote for Bush as a means of avoiding post-election conflict. However, the internals of the polls (at least Zogby’s polls) show that the undecideds are highly critical of Bush, and he doesn’t expect them to vote. With turnout expected to extremely high, I wouldn’t count on them not showing up at the polls. Many will turn out for local or state races even if they still have misgivings about the Presidential race.
Early indications do not suggest that the GOP base is as energized as the Democratic base. If the GOP base scores at a 9 on an enthusiasm rating of 1-10, the Dems are at a 10.5. Early voting by so many Democratic activists will free up more foot soldiers for the GOTV ground operation. Unless there’s a radical swing towards the incumbent, a scenario which is certainly possible, Kerry’s odds are good, particularly with the two major swing states moving his direction.
As an aside, it’s time to give up the “Perot cost Bush the election” fantasy. It’s not reasonable to suggest that Perot voters stole so disproportionately from Bush that Bush could have went from 168 electoral votes to 270 without Ross. And your list of scenarios why embattled incumbents lost should include one more. In 2004, Bush-43 had led us into a ruinous military quagmire against a nation that posed no threat to us.
Yet the GOP has registered just as many if not more voters than the Democrats in Colorado and Iowa. Most polls show more Democrats who support Bush than Republicans for Kerry.
In 1992 Perot got 19% of the vote, the vast majority of which came from Bush voters. Yes, it was enough to have cost Bush several key states. Perhaps Bush would still have lost without Perot, but there’s no doubt that Perot’s candidacy sunk any chance of Bush 41 winning.