Fifty-two weeks from tommorow, the voters of the United States will elect a president? Who will it be?
The polls suggest that Edward M. Kennedy has substantial lead over his rivals for the Democratic presidention nomination -- including Jimmy Carter -- and that Ronald Regan is miles ahead among Republicans. And if Reagan and Kennedy square off next November, the polls suggest, Kennedy will win in a romp.
Wrong. Althought political types and reporters sometimes treat them as holy writ, public opinion polls, particularly those taken last year or more before the vote, are anything but safe predications of the outcome on election day.
There is living proof of this simple truth in the White House right now. A year before Carter won the presidency, his "standing" in the polls was more like "lying down."
In a Gallup Poll among Democratic voters in October 1975, the name "Carter" showed up only in the small type listing of "all others" -- indicating that he had the support of less than 3 percent of the voters in his own party.
Similarly, Carter's predecessor as the Democratic nominee -- George McGovern in 1972 -- looked like an also-ran at best polls taken a year before that election. The October 1971 Gallup Poll of Democratic voters showed McGovern tied for fourth place among potential Democratic candidates; he was the choice of 6 percent of he voters in his party.
And the Democratic nominee before McGovern -- Hubert H. Humphrey in 1968 -- stood fourth in the polls a year before the election, with the backing of 6 percent of his party's voters.
Among Republicans, however, the polls do better. A review of Gallup Polls for the last five elections shows that the candidate favored by Republican voters a year before the election has become the Republican nominee every time.
At this point the careful reader may see a rough rule of thumb emerging: polls of Republicans mean something and polls of Democrats often don't. But anyone who draws that conclusion has not been careful enough.
In both parties, the year-early polls have been skewed somewhat by circumstance.
Among the Republicans, the skewing influenc has been incumbency. In three of the last five presidential elections, the GOP candidate either was the incumbent vice president (Richard M. Nixon in 1960) or the president (Nixon in 1972, and Gerald R. Ford in 1976), and thus it is hardly surprising that the year-early favorite hung on to become the nominee.
The Democratic experience, meanwhile, has been skewed by a factor called Kennedy. In four of the last five elections, as in the current campaign, the year-early leader among Democrats has been a fellow named Kennedy -- John in 1959, Robert in 1967, and Edward in 1971 and 1975.
Among the Democrats, in fact, being a Kennedy has generally been a more potent political attribute than an incumbent.
Although then-President Lyndon B Johnson was his party's favorite a year before the 1964 election, by the fall of 1967 Johnson ranked well behind Robert F. Kennedy in preference polls among Denocratic voters.
President Carter has ranked well behind the current Kennedy for the past two years.
There is no better example, perhaps, of how mysterious the polling business really is than a record of the relative popularity of Edward Kennedy and Jimmy Carter among Democrats. The numbers that emerge when this survey is taken seem to have no relationship to developments in the real world.
When a sample of Democrats was asked in June 1978 whom they would like the party to nominate for president in 1980, 55 percent picked Kennedy and 31 percent were for Carter -- a 24-point margin for Kennedy.
Then, in October, Carter worked his diplomatic miracle at Camp David and his approval rating soared. But in the next Kennedy-Carter preference surveys a few weeks later, Kennedy actually increased his margin over the president to 26 points.
Since last August, things have gone badly for Carter; inflation worsened, the Cuban troops situation festered, and his approval rating plummeted. During this time of trial, though, Carter has somehow managed to close the gap somewhat between himself and Kennedy.
If the results of these "party preference" polls seem confusing, another type of survey -- known in polling circles as "trial heats" -- can be equally unclear.
In a "trial heat" poll voters are asked to choose between two candidates, one from each party. Over the years, year-early trial heats have been about as reliable -- which is to say, not terribly reliable -- as the preference polls in predicting the election results.
In October 1959, a Gallup "trial heat" showed that Richard Nixon had a comfortable lead over John F. Kennedy among a nationwide sample of voters. A year later, Kennedy defeated Nixon by a narrow margin.And four years ago "trial heats" showed Ford beating all Democratic challengers except one Edward Kennedy, who was not a candidate then.
In 1963 and 1971, however, trial heats showed, accurately, as it turned out, that the incumbent presidents were safe bets for reelection.
In 1967, there was no exact trial heat because no pollster thought to gauge sentiment in a Nixon-Humphrey contest.
The polls a year before the 1968 election did show, though, that Nixon was likely to beat any Democrat.
All of this is not to suggest that polls are useless or inevitably inaccurate. They do reflect, and provide useful information on, one key point: the election of a president is a highly volatile process, particularly a year before the votes are cast.
As a rule, the polls predict results more accurately as election day draws near. But not always.
The final Gallup Poll of 1976 showed Ford with a small lead over Carter among voters nationwide.
Had that survey been accurate, Carter would probably be home in Plains right now. Which is where he will be in a year or so if you could believe the polls.