Polls taken immediately after the Democratic Conventio showed President Bush trailing Bill Clinton, the Democratic challenger, by truly enormous margins: 22 points according to a Gallup survey for Cable News Network and USA Today; more than 30 points in others taken slightly later.
Everyone expected this margin to shrink after the Republican Convention, and it did: Down to ten points according to Gallup's first post-GOP Convention poll for CNN/USA Today; to anywhere from 8 to 15 points in other polls taken after opinion had had a few days to settle down.
Not surprisingly, a riot of speculation is breaking out about the chances for a GOP comeback. The discussion has so far been largely data-free and highly impressionistic. This does not have to be, however; twice in modern history (that is, since TV, which definitely excludes the 1948 "miracle") Republican candidates have rallied from far behind to win (or, in the 1976 case, to lose in a photo finish after forgetting where Poland is).
This history is very instructive, and it contains both good news and bad news for the Bush campaign.
First the good news: Both of the comebacks -- 1976 and 1988 -- were managed by James Baker.
But now the bad news: In both cases, the long climb back up was not, in fact, very long. In both cases, it began immediately after the Democratic Convention.
Consider the polling results from 1976 and 1988. The main inference is obvious: We have all heard about "convention bounce," but in fact, when the GOP is recovering from the depths, the improvement begins well before the GOP's conclave a month or so later. (After the 1988 Democratic Convention, readers will recall, Michael Dukakis famously went on vacation to reign as "Governor of the Berkshires." This left the field to George Bush, who, the Gallup Poll suggests, used the time well). both 1976 and 1988, this initial rise, added to the "convention bounce," brought the GOP much of the gains it was ever to make.
By contrast, after the Republican Convention, the rest of the race was trench warfare -- until the very end, when, perforce, remaining undecideds finally had to make up their minds. (Note that in 1988, most appear to have broken for Mr. Dukakis as his campaign moved rhetorically to the left in the campaign's waning days.)
It is difficult to see any signs of such momentum in the current GOP campaign.
Though one or another poll occasionally startles by suggesting narrower margins, there are excellent reasons for crediting Gallup as especially reliable (not least because the organization thoroughly absorbed the lessons of the 1948 debacle, and now strives mightily to filter out non-voters). As a consequence, one is squeezed to the conclusion that the president's progress has been feeble indeed.
At this late date -- now well into the "incremental" stage where additional voters have historically been won almost one at a time, the president needs to break all records for making up ground after the conventions. Of course, he may do it, but one must suspect that the constant invocation of the diety at the GOP convention was less an electoral ploy than a hard-headed assessment of where his best hope now resides.
As an exercise, however, it is worth asking how an "October surprise" might affect the race. At the moment, the most likely "surprise" is probably the unveiling of some sort of Mideast peace settlement along the lines of Jimmy Carter's Camp David Accord, though it is interesting to note that some timetables mentioned by the press in connection with the South African talks also work out rather well for a timely White House announcement.
Though conventional wisdom holds that most Americans care little about foreign affairs, Camp David gave the slumping Mr. Carter a sizable lift: 11 points in a Gallup Poll taken shortly after. This poll, however, was of the normal "presidential approval" type -- it did not pit the president against a challenger. Making the very generous -- indeed, absurd -- assumption that three quarters of the 11% might have been induced to vote for Carter against a challenger, we arrive at an "upper bound" estimate of the electoral effect of an October surprise: perhaps 8 points at the very best (and most improbable: a better projection is that the president would be lucky to see even half of this).
One can also roughly estimate how a mudslinging campaign, along the lines of 1988, will fare. Making the assumption that all opinion change in 1988 after the GOP convention was the result of skillful mud throwing (and that the GOP cannot realistically expect to do better than it did then in this department), we arrive at the figure of about half a point a week as the "best estimate" of what the 1992 Bush effort can hope for.
In recent days, the Wall Street Journal and some Republicans have begun comparing Mr. Bush's chances to that of British Prime Minister John Major, who triumphed earlier this year. But British campaigns are very short by American standards, whereas the message of the 1976 and 1988 races seems to be that by the time the conventions are over, many Americans feel they have seen enough to decide.
In politics it is foolish ever to say "never." Every time the race closes up, Republican hearts -- along with those of the media -- will flutter. Still, the historical record on presidential comebacks indicates that the president is probably too far behind at too late a stage in the election to win -- unless, of course, God really is a registered Republican.
Thomas Ferguson is professor of political science and senior fellow at the John W. McCormack Institute of Public Policy at the University of Massachusetts, Boston.