Tom Holbrook has a nice discussion of how much conventions help the nominated candidates. Here are his estimates of the convention bumps for the 1964-2004 elections:
Typically, conventions do help the nominated candidate (more so than any other campaign event or activity, e.g., debates, advertisements, etc.). But there is also variation in the size of the convention “bump.” Holbrook writes:
Two things in particular seem to drive the size of the bumps. First, candidates who are running ahead of where they “should” be (based on the expected election outcome) tend to get smaller bumps, and those running behind their expected level of support get larger bumps…
…The other key factor is the timing of the convention. The earliest convention tends to get a bigger bump, and there is some evidence that going appreciably earlier exacerbates this effect. One reason for this is that the first convention is held by the out-party, whose candidate generally less familiar to the general public.
Currently, Obama and McCain are running about where they are expected to (see here). And, unlike in most elections, the Democratic and GOP conventions are very close together. This may make it difficult to get any noticeable bump if the two conventions in essence cancel each other out before polls can register any effects. (The Labor Day weekend will complicate polling.) All of this suggests small convention bumps. Holbook promises a prediction for how big they’ll be. Stay tuned to his blog.
(And see also Charles Franklin’s data and Mark Blumenthal’s commentary at Pollster.com.)