Campaigns Don't Count

Throughout most of 2004, the presidential election was too close to call in the polls. Journalists and politicians alike insisted that its outcome would depend on developments in the Iraq war, on the state of the economy, and on the strength of each candidate's campaign.

But one veteran columnist in Ohio — author Martin Gottlieb, of the Dayton Daily News editorial page — wrote in a January 2004 column that President George W. Bush "has won" the election. He repeated this point several times throughout the year. How could he be so confident?

After decades of observation, Gottlieb has concluded that the best system for predicting the outcome of an American election‹and, more important, for explaining American politics‹is one devised by Professor Allan J. Lichtman of American University in Washington. It's a system that emphasizes the importance of substance over imagery.

Campaigns Don't Count takes the reader briskly through recent American political history, taking note of predictions of all sorts. What emerges is a new way to think about American democracy. Gottlieb's case will intrigue journalists and anyone else interested in American politics.

Fred Barnes wrote the May 4, 1992 cover story in The New Republic, a magazine for political sophisticates: "Loser: Why Clinton Can't Win." It needs to be paused over here, not because it is one more mistake, or because there's any point in picking on Barnes. The special value in his piece is that he presented it not merely as his view, but as the view of the insiders, the people in the know. It is a remarkably well researched piece, if you consider talking to political insiders to be research, which, of course, by the standard rules, it is. The list of people he quoted and paraphrased as agreeing with him is impressive.

The piece begins, "Vin Weber, the Republican representative from Minnesota, remembers the exact moment he concluded Bill Clinton will never become president." (It was Feb. 6, the day the Wall Street Journal printed new evidence that Clinton had dodged the draft, this coming a couple of weeks after a flare-up in the Gennifer Flowers adultery scandal.)

Another passage: "Tubby Harrison, the pollster for Paul Tsongas this year and Michael Dukakis in 1988, has called Clinton 'a dead stone loser' in the fall." In this passage, Barnes notes that loser Dukakis jumped ahead of Vice President George Bush in the polls when he reached the point in the primaries where he clinched the nomination. Clinton didn't.

Another passage on Harrison: "Every party figure he's talked to, the consultant says, 'thinks it's going to be a disaster.' Why? Because Clinton has problems in four separate areas: the character issue, voter turnout, the South, and foreign policy. These problems are all the more troublesome because there's no clear way to overcome them." (On character, Barnes cited four polls making his point that the public had doubts that were not going away. He called the poll numbers "staggering.")

Barnes quotes respected Republican pollster Richard Wirthlin saying, "A candidate can't get on the playing field unless he pays the ticket of honesty and integrity."

And an unidentified "Democratic strategist:" "Once Clinton's negatives are established and allowed to grow deeper and deeper, they're almost impossible to blow away."

Barnes unearths a Clinton problem beyond the usual list: low turnout in the Democratic primaries. He quotes former Sen. Paul Tsongas (a candidate in the early primaries) and then Georgia Secretary of State Max Cleland (later a Democratic U.S. senator), not to mention Bush's Southern coordinator, plus academic turnout expert Curtis Gans on the difficulty that low-turnout poses for Clinton. Barnes notes that blacks turned out in smaller numbers in the 1992 primaries than in 1988 (when the Rev. Jesse Jackson was in the race), and says, "I can't imagine how Clinton will lure them in the fall when he couldn't in the spring."

Barnes concludes, "Absent a scandal or economic collapse, Clinton's a goner."

The economy improved.

In 1973, White House lawyer John Dean told the Senate Watergate Committee that earlier that year he had told President Richard Nixon that a "cancer" was growing on his presidency. The next day, Professor Milton Rakove, a specialist in urban affairs, said this to his graduate class at the University of Illinois in Chicago:

"He's got to resign."

At first I didn't know whom the professor was talking about. But he obligingly explained to the blank faces in front of him:

"Nixon. He's got to resign. That's how this ends: Nixon resigns."

I was dumbstruck. Presidents don't resign. And Nixon had just been re-elected the previous fall in a 49-state landslide. And, though Watergate was, indeed, giving him all manner of trouble, few observers had raised the question of resignation publicly. I knew this. I was obsessed with Watergate.

But Dr. Rakove didn't simply raise the question. And he didn't simply make a prediction. He was making an announcement.

I searched my understanding of things for anything that would lead me toward the conclusion that Nixon must ultimately resign. I found nothing.

For months after that, I searched the expressed thoughts of people in journalism, academe, politics and elsewhere for any thoughts of a resignation, and I found almost none. As for certitude that Nixon would resign, absolutely none.

When Nixon resigned, a year and half after Rakove's announcement, I thought of Rakove.

Specifically, I thought that's the kind of understanding of politics I want. That's the kind I respect. Not the kind that explains events after the fact. Not the kind that analyzes things. Not the kind that tells you what "factors" are at play. (I was getting to the point where even I could do that.) I wanted the kind that predicts things. Flatly, boldly. Correctly.

Prediction is what cuts through the crap.

Journalists and most other political people — including academics — generally treat prediction in politics as a game, a sideshow. Occasionally journalists venture into boldness when contemplating the future, but most often with tongue in cheek. Later everybody has a good laugh about who was wrong about what. If you make a prediction — whether it's right or wrong — you get credit for being a good sport, and that seems to be the only characteristic that gets measured.

But physicists can make predictions within their realm. They can say flatly what will happen when two objects collide, or something. They really know their subject.

Of course, we can simply take the attitude that politics isn't physics — that nobody ever said politics is an exact science — and leave it at that. But we aren't required to leave it at that.

Before deciding whether to leave it at that, we should face this fact: when the alleged political experts treat prediction as a laughing matter, they serve their own interests. They free themselves of a form of oversight, of one way their listeners might judge whether the "experts" really understand things, or whether they just sound good.

The "experts" in politics are generally accorded the "expert" label as the result of some professional achievement. They've been award-winning reporters for some time. Or they've written important books or won advanced degrees from prestigious schools. They've been consultants to successful candidates. Or they've been successful candidates themselves. They are respected in their professions. And if you — the reader — are satisfied listening to the people the professionals listen to, fine. But what if the professionals are confused as a class?

Many people have attempted to study political phenomena somewhat the way physicists study physics. We have a name for these people: political scientists. They — or some of them — look for patterns in political phenomena precisely so that (among other reasons) political events might become more predictable. This book is not a brief for or against these people. My own experience with them almost 30 years ago did not leave me believing they are the hope for mankind. I thought they tended to get bogged down in unfascinating issues, and that they ended up making predictions that were too hedged or too small to be interesting. Or just too obvious.

As a matter of fact, a fair number of political scientists have turned their attention to the primary subject of this book: How does one predict the outcomes of elections, and what do successful predictions tell us about how the political system works? But most of these predictive schemes do not go very far and, for other reasons, are not all that exciting. (Some are considered in Chapter 3.)

But suppose somebody does come along who can make really useful predictions about a broad range of races, and can make them early in the year. Shouldn't that accomplishment be taken seriously?

In 1984 I became an editorial page writer for the Dayton Daily News in Ohio. In that capacity, when I wasn't writing the unsigned editorials that are presented as the opinion of the newspaper on the controversies of the day, I was writing what I thought were David Broderish signed columns. They were about local, state and national politics. In the editorials, I was a moderately liberal Reagan basher. In the columns — my love — I was often an ideologically detached journalist (as in this book). Some people think detachment is a stretch. They are wrong. Piece of cake. Since when is abandoning principles difficult?

I found, in fact, that writing a political column was the easiest work I had ever done. Because I had been following political events way too closely for 20 years and had been aware of the biggest events for closer to 30, I had all manner of historical analogies to make to current events, all manner of precedents to cite for current behavior. Besides drawing analogies, I highlighted ironies. Ironies are the most common commodity in politics. You trip over them on the way to work. Then you get paid for pointing at them. Nobody knows why.

Sometimes I stated opinions about the prospective course of political events, but never the kind of opinions that could be disproved. I didn't opine that one candidate would prove stronger than another, just that one would have a lot of support, or would get a lot of people mad. I wrote the kind of stuff that looks like more than it is, or that, at least, the writer hopes looks like more than it is.

The cowardice was particularly shameful because, if I had dared to make a real prediction and been proven wrong about something, I would have paid no price. Pundits don't get fired or docked or even bawled out for being wrong, any more than they benefit from being right. It's not about that. It's about sounding like you know what you're talking about.

That is too easy to be considered respectable work.

A fundamental rule of punditry is that nothing ever happens that can't be easily explained away. The rule is especially true when the subject is election outcomes. In long campaigns, many things happen that seem likely to help one candidate. Many other things happen that seem likely to help the other. Picking what you want to use after the fact as the explanation for an outcome is trivially easy — always, no matter what point you're trying to make.

In 1984, say, if Ronald Reagan demolishes Walter Mondale, despite your pre-election insistence that the American people are really more in tune with Mondale, you have a whole menu of explanations to choose from: People don't really know what Reagan stands for; half the people didn't vote; Mondale didn't articulate a clear liberal alternative. Piece of cake.

If, however, Mondale defeats Reagan, and you are writing from, say, a conservative perspective, the explanation for the election outcome is just as obvious: The liberal media never gave Reagan a chance.

As for the non-ideological, non-partisan pundits, they, too, were ready for anything. Mondale defeated Ronald Reagan so badly in one debate that if it had lasted 10 more minutes, it would have been stopped, because nobody wants to see a president get hurt. Reagan told a story that made no point whatsoever, except that he was seriously confused.

If Reagan had lost the election, many pundits would have been perfectly content to say he lost it in the debates. It would have been obvious. The pundits had been treating debates as momentous, turning-point events for years. This would have just shown that they were right.

Somehow, though, the fact that the debates turned out not to matter — that voters turned out not to care much whether an incumbent president can debate — didn't result in a great number of newspaper columns expressing confusion about the outcome.

Not many writers said, "Mondale's debate skills, compared to those of his elderly opponent — put on top of his refreshing honesty about the deficit, his daring and galvanizing choice of a woman for vice president, and his shameless but clever willingness to give the traditional Democratic interest groups whatever they wanted — should have been enough to bring victory. I can't figure out what happened."

Yet I knew in my heart that if Mondale had won, all those factors would have been cited as the obvious explanations. I knew I would have been doing it.

Eventually, you have to start to wonder about this profession. I had the sense that I was playing a game with no rules, which hardly mattered, because there was also no score. You just do it. Then you came back and do it some more.

After a while, though, you start to think maybe there should be some rules.