More On Bounce
The extent of the Kerry-Edwards predicament following the no-or-negative-bounce Boston Democratic Convention compared with the certain-but-not-how-much-bounce of the Republican New York Convention can be sensed by considering how many historical precedents were shattered this year, as noted by the Gallup Organization:
The lowest bounce recorded is that seen for Kerry following this year's Democratic convention. Support for Kerry actually declined by one point among registered voters (and two points among likely voters) over the course of the convention -- a "negative bounce." Kerry is not the first candidate to experience no boon from his convention. George McGovern saw no change in support for his candidacy spanning the Democratic convention in 1972. ....
1. The order in which the conventions occur. Looking at all bounces between 1964 and 2000, there appears... to be a real advantage in terms of a ticket's bounce when that party's convention comes first. ....
2. Whether the party holding the convention is currently in the White House. Including 1992 in the calculations, challengers look like they have a significant edge over incumbents (7.1 points for challengers vs. 5.5 points for incumbents).
3. The party effect. ... When including 1992, the Democrats surge ahead on this comparison, leading Republicans by more than a full point: 6.9 points vs. 5.6 points. ....
Out of the 17 elections since 1936, 12 included an incumbent seeking re-election -- as Bush is now. Of the nine successful incumbents, all but one (Truman) were ahead as of Labor Day. Of the three unsuccessful incumbents, only one (Jimmy Carter in 1980) was ahead on Labor Day, while the other two (Ford in 1976 and Bush in 1992) were behind. While both Ford and Bush managed to shrink their Labor Day deficits by Election Day, neither managed to win.
Clearly, the odds favor Bush's re-election if he can establish a lead over Kerry on Labor Day, and the odds obviously become greater the larger Bush's lead over Kerry is at that point. But it is important to note that in more competitive elections such as the one this year, the candidates' standing on Labor Day, alone, is not necessarily strongly correlated with the outcome. In elections in which the Labor Day gap was five points or fewer, the leading candidate prevailed only twice (in 1936 and 1940) and lost three times (in 1960, 1980, and 2000). It is important to note that in 1960 and 2000, the leads were not outside the polls' margins of error, so the candidates could be considered to have been tied on Labor Day.
Note: The 1992 election was arguably unusual, and Gallup includes a separate analysis excluding that election - an analysis which is deleted above and which in some respects tends to make Kerry-Edwards look less like an alarming failure. However, every election has its own unusual features. Gerald Ford, for example, was an unusual "incumbent." I can see no special reason to specifically exclude 1992 instead of, say, 1976 - so I don't.
UPDATE: Gallup reports
that Bush-Cheney leads by 7% immediately following the Convention.