Predicting elections - 13 November 1992

I knew an old man wise in many ways who made a point of inviting you to dinner for the Sunday after each presidential election. And as you sat down to a preliminary drink, he'd twirl his Mandarin moustache and say, "Well, what did I tell you". The first time this happened, I was ready to puff and barbell, "What do you mean, what did you tell me? You were dead wrong." Of course, I might think that, but what down the years I learnt to say was, "You did it again" and was rewarded with a grateful smile.

The thing about this little story is that the man was, I'm convinced, absolutely sincere, he was no better than anybody else at picking the winner but he was a big quiet contributor to both the Republican and the Democrats' candidates, a precaution that is routine with many big corporations and Wall Street houses, and he'd spent so much time before the election being shrewd and knowing about it, that the result didn't shake his conviction that he'd been right all along. He was sensible enough, as I've noticed lots of authorities on various games are, never to bet because the result would be a public demonstration that he was as fallible as the rest of us. But by the following Sunday, I believe he'd convinced himself that the result was what he'd predicted all along.

This form of vanity I'm sure afflicts many more people than my old friend, because so many people are desperate to know the winner before he's won. I don't know how it is in your country, but Americans must be the world's most impatient voters. Almost everybody has caught on to a rule, a guru, a superstition, a method of prediction that we'll ensure their being right on the great day. Some of these indicators are not superstitions, they've been tested time and again, they constitute the fruits of long experience; for instance, it is a maxim of American politics known to schoolchildren that you can't loose the first, the New Hampshire primary and win the election except in 1992 when Bill Clinton came in second to Paul Tsongas an engaging man, he later chuckled that he should have read the oracles correctly, it wasn't in the cards after the 1988 collapse of Mr Dukakis that another Greek from Massachusetts would win the Democrats nomination, let alone the election.

The Democrats never win the White House if they loose Texas, except this year. A little less substantial as a guide, but interesting nevertheless is a county in the western state of Oregon, Crook County, it's a very tiny county, but it's an election district with about 7,000 votes. For 108 unbroken years in 27 successive presidential elections it has called the tune, never failed to give a majority to the man who won. This year, the Crook County winner was George Bush.

Baseball addicts used to wag a warning finger at sceptics who didn't believe that whenever the champion team of the American League won the World Series, a Republican would go into the White House. They have given up on that one since the rule failed three of the last four elections. As you may know the campuses of this country's universities abound in professors of political science, thousands of them. I've often wondered if their so smart at the subject they profess, why don't they go into it? Then I suppose you could say the same about professors of economics who seldom seem to take their own advice and make a fortune.

Well, a professor of political science in a Midwestern university has a rule that he's picked up from, of all people, French wine/grape growers – vintners – an indifferent Beaujolais harvest always foretells a Republican winner. This year, sure enough there was a poor harvest and look what happened.

This professor, who admits defeat this year on the wine/grape test has, however, taken refuge after the event by proclaiming a rule that has never failed in the past hundred or more years, there have been three left-handed presidents James Garfield, Gerald Ford and George Bush, none of them won a second term. If you're making a note of this interesting fact, you may or may not care to add that President Elect Clinton is, you guessed it, a left-hander.

Talking rather meanly as I did about discovering a rule after the event, there's a lady astrologer here who's been going for years, built an impressive reputation for predicting the unpredictable. And when required to call off her more astonishing performances, she will do it promptly. I remember being told by a friend of hers that yes she had predicted the assassination of Bobby Kennedy in that hotel pantry in Los Angeles in 1968. I said, "I don't remember her saying that at the time". Well, I was told she didn't say at the exact time – three days after the murder she declared that on the night before it she'd had a dream, which previewed the whole horror.

I was interested this year to know if she'd made a prediction ahead of time, what we – we rightly call a prediction. She did, so long ago as last summer she assured us that the stars in their courses pointed conclusively to the return of President Bush. If all this teaches us to avoid in the future, all rules of thumb proved superstitions and advice from Frenchmen treading grapes, there is it occurs to me a famous record of the rules of chance or the fallout of coincidences that takes a lot of explaining away. I have taken it up with mathematicians as well as amateurs sleuths says and all of them just goggle and shake their heads, they have no clue to this bizarre series of fated or coincidental facts, it has to do with some of the dates and the events in the life's and deaths of Abraham Lincoln and John F Kennedy.

Lincoln was elected in 1860, Kennedy in 1960, both were shot in the head from behind with their wives at their side. Lincoln's assassin John Wilkes Booth was born in 1839, Kennedy's assassin Lee Harvey Oswald was born in 1939, both murderers were Southerners and both were shot before they could be brought to trial. Lincoln was immediately succeeded by Vice President Johnson born in 1808; Kennedy was immediately succeeded by Vice President Johnson who was born in 1908. Lincoln's secretary was named Kennedy who advised the president not to go to the theatre; Kennedy's secretary was named Lincoln who advised the president not to go to Dallas. Booth shot Lincoln and ran from the theatre into a warehouse; Oswald shot Kennedy and ran from a warehouse into a theatre.

Looking back over the prophecies and the analysis and the educated guesses written before the election before tossing them into the garbage, we can see now two miscalculations, one before, one after the election. The first was the general sense that Mr Bush had surged dangerously close to Governor Clinton in the last week and might very well overtake him and win the election. I don't believe this was ever more than a remote possibility because though the figures marking Bush's advance were correct, they were still recording the popular vote, the numbers of people who might be changing their minds scattered or distributed across the country. Across the country is the key to this mistake, no pollster so far as I know has ever gone out to test the balance of the popular vote in the big key states and come up with an electoral poll with that is a prediction of the likely number of electoral votes to be won by each candidate.

Now there's nothing mysterious about the electoral system in spite of its 18th- century origin. Each state is allotted a collective vote, which is made up of the number of its congressmen plus its two senators for example, California has 52 representatives in the House and like every other state, however big, however small, two senators – total 54. And the election goes to the candidate who collects the most electoral votes, if you win the total popular vote in the state by only one vote; you win the whole electoral caboodle.

Now what we should have remembered when the Bush surge was reported was that he might go ahead of Clinton in the polls and still loose. In fact, in a way that's what happened between them. Bush and Perot got 57% of the popular vote, Clinton only 43%, but what matters is where you got your majorities. Anyone expecting a neck-and-neck finish should have looked first at the insuperable early advantage that Clinton had in the polls from the evidence of polling in California, New York and Pennsylvania, he was way ahead in the first two almost certain to carry Pennsylvania – three of the biggest the most populous five states in the Union, which meant that Clinton started with 106 electoral votes in three states. In the early hours of that Tuesday evening, the only not so sure sates from the polls for Bush were Nevada, Utah, Virginia, Indiana and South Carolina, together totalling only 42.

In the end as we all know, Clinton wound up with a handsome 370 electoral votes against Bush's 168, because Perot was very impressive popular vote, 19% half of all Bush's, because it was too scattered to give him a majority over the others in any state he got no electoral votes. Still, Governor Clinton, like Woodrow Wilson, like Harry Truman is what we call a minority president, his total vote is way below the combined votes of his rivals, so it was no landslide.

The minority leader, the Republican leader in the Senate, the very powerful Senator Bob Dole keep saying that the Republicans can mount a very formidable opposition because quote 57% of the American voters voted against Clinton. On paper that's true, but it's not true in life or politics, surveys after the election showed that Mr Perot took almost exactly the same percentage of votes from each of the other two, 40%. In other words, if there had been no Perot, 40% of the people for him would have voted for Clinton and in that result add those votes about 8 million to Clinton and he has 51 million for him against Bush's 38 million. It's not a mandate but it's a rather heavy vote of confidence by most Americans in the prospect of Mr Clinton's presidency.

THIS TRANSCRIPT WAS TYPED FROM A RECORDING OF THE ORIGINAL BBC BROADCAST (© BBC) AND NOT COPIED FROM AN ORIGINAL SCRIPT. BECAUSE OF THE RISK OF MISHEARING, THE BBC CANNOT VOUCH FOR ITS COMPLETE ACCURACY.

Letter from America audio recordings of broadcasts ©BBC. Letter from America scripts © Cooke Americas, RLLP. All rights reserved.