The Final Prediction

E.J. Dionne, Jr.
EJ Dionne
E.J. Dionne, Jr. W. Averell Harriman Chair and Senior Fellow - Governance Studies

November 5, 2000

Political punditry, like sportswriting, is rooted in traditions and expectations. People who love a particular sport and people who love elections just “know” that certain things are true and that other things are highly likely. They offer predictions based on their belief in the reliability of what has come before.

The 2000 campaign has been confounding to pundits, and not just because there are, oh, 1.2 million or so tracking polls. Election Night will be fun for many reasons, not the least of which is that many venerable assumptions are cracking apart this year. States that aren’t supposed to be competitive are, voting blocs that are supposed to move a certain way aren’t, and rules about who benefits from what kind of turnout are being rewritten.

What follows is a guide to the list of assumptions that have been proven wrong so far and others that might go by the board after the returns are in.

The battle is always fought in the same battleground states. We act as if things that were true in the last election will be true in the next one. This is by no means a foolish assumption: If you looked at maps of the border states at the time of the Civil War—Tennessee is a good example—and found out which counties had opposed seceding from the Union, you’d have a pretty good idea of which counties would vote Republican in 1944 or even 1984. On the other hand, by 1984, many of the old Confederate/Democrat counties also had become Republican, at least in presidential elections.

This year, it was assumed that if a state voted for Democrat Michael Dukakis in 1988, it had to be safe for Al Gore. Not so. Gore has had to work his heart out in Wisconsin, Washington, Minnesota and West Virginia, among other places. Similarly, some Republicans were shocked that Florida—long considered a Republican bastion—has bedeviled George W. Bush.

A Washington state with Microsoft and the burgeoning high-tech industry is very different from one dominated by Boeing and longshoremen. Microsoft’s employees and suppliers are more interested in the politics of antitrust than in the politics of unions.

Colorado has become more Republican for the same reason that California has become more Democratic. Many conservative former Californians moved to Colorado, reducing the Republican vote in one state and expanding it in the other.

In Florida, Cuban Americans were long the mainstay of Republican strength. But second- and third-generation Cubans are less Republican than the first generation. And the arrival of immigrants from Central and South America has made Cuban Americans a minority among Florida Latinos.

A lot of things are going on here: the issues change, the demographics change (people move in and out of states) and voters change their minds about where they take their cues. Jim Humphreys, a Democrat in a surprisingly close race for the 2nd Congressional District seat in Democratic West Virginia, says “almost everywhere is competitive” now because you can no longer assume that a state’s “political culture or its economy is static.”

Humphreys also argues, rightly, that local political cultures are less powerful than they once were as voters no longer rely so much on “local TV stations, local radio stations and local newspapers.” Cable, satellite and the Web let people diversify (and also narrow) their sources of information.

Low turnout always means the Democrats are staying home. This is an old assumption, based on the broadly accurate premise that upscale people are the most likely to vote, and that upscale people tend to be more Republican than their poorer neighbors. In the past, Republicans prayed for rain on Election Day; a couple of years back, when several key congressional races in Ohio and Kentucky were in the balance, a Republican operative called me around noon to say with glee, “It’s good news for us. It’s raining all across the Ohio Valley.”

It’s perfectly possible these assumptions will hold up this year, because Republicans seem so eager to recapture the White House from Bill Clinton. But don’t count on it. If you get word of a low turnout at 3 p.m. on Election Day, you can’t be as sure as you once might have been that it means a Bush victory. It might mean a replay of the 1998 elections, when Republicans had a hard time getting their base out and Democrats didn’t.

Organized labor was successful in turning out its members that year, and is putting even more people and money to work this time. So are environmental groups and the NAACP.

Republicans, having learned from 1998, know their faithful need more prodding than they once did. The Republican National Committee is spending an estimated $80 million to get the party’s supporters to come out on Tuesday.

Another wild card is the news of Bush’s conviction 24 years ago for driving under the influence of alcohol. Republicans could be angered by the last-minute disclosure and vote in force, or be dispirited and stay home.

An issue that works in one campaign will work the same way the next time around. The Republicans won elections for two decades after the Civil War by urging Union Army veterans to “vote as they shot.” The Democrats won seven of nine presidential elections after the Great Depression by running against Herbert Hoover.

But not every issue is like the Civil War or the Depression. In 1996, Bill Clinton did very well among suburban voters because of a panoply of issues, including his support for gun control. This year may be different because the newly energized NRA has gone all out for Bush. As Richard Moore, president of the city commission in Gallipolis, Ohio, says, some voters “fear that if Gore is elected, they’ll have to register all their guns.”

Moore, a Bush supporter, adds: “Maybe the NRA exaggerates a little bit.” It does, but it may be working. Gore is worried enough that he has soft-pedaled his position on guns. The results on Tuesday may hang, in part, on whether pro-gun voters help Bush more than pro-gun-control voters help Gore.

The losing party always suffers from the gender gap. When Democrats win, it’s said that Republicans have a terrible problem winning the votes of women. When Republicans win, it’s said that Democrats have a terrible problem with white men.

But the gender gap can stay exactly the same from one election to another and either statement can be justified. If Bush wins, he’ll do better among men than women and if Gore wins, he’ll do better among women than men. That’s because, all things being equal, women are more Democratic than men (or, if you prefer, men are more Republican than women). The interesting questions are whether the gender gap grows or narrows, and why.

You can always count on close elections to be close on Election Night. It’s been said for weeks that this is one of the closest elections in recent history. But it’s quite possible that the final result will not be close at all, either because one side will turn out in larger numbers, or because of a large shift by undecided voters in one direction. (Such a shift in the final days turned the close 1980 election into a big win for Ronald Reagan over Jimmy Carter.)

So have mercy on all those offering predictions on these pages. Their calls might have been right on the day they made them, and wrong on Tuesday.

E.J. Dionne is a Washington Post columnist and a senior fellow at the Brookings Institution.