Take Ford's famed Poland gaffe: Few people recall that Jimmy Carter's lead over the incumbent shrank over the course of the debate period. And while the single Reagan-Carter debate is recalled as the Gipper's watershed moment where voters finally broke for him, the fact is that Reagan had led in the average of tracking polls since the late spring.
Why don't debates have a greater impact? First, most voters tend to have made up their minds by the time debates occur, a phenomena likely to be exaggerated this year when there seem to be fewer undecided voters than usual.
Second, while the Romney campaign seems to bank on distinct moments in time to turn the race—first the vice presidential rollout was going to restart the campaign, then the Republican National Convention was going to spark a Romney surge—the fact is that elections are akin to aircraft carriers. They generally don't turn on a dime.
And this makes intuitive sense. After nearly four years, people have a pretty good sense of Barack Obama, and by now they've developed a notion of Mitt Romney. A debate moment might crystallize voters's views, but the events are not going to change the fundamentals of the race. People like Obama and think that he understands their concerns, and neither statement is true of Romney. And the debates are unlikely to change the fact that Romney is a tone-deaf candidate who has surrounded himself with an inept campaign team; or that he has been crippled by the need to straddle the sizable gap between his base's policy preferences and those of the mainstream voters on issues like Medicare and taxes.
None of this is to suggest that the debates will be devoid of drama. In fact, as I suggested a few weeks ago, there's good reason to think that the debates might provide memorable drama. Suppose the first debate or two pass without the momentum change for which the Romney campaign is hoping. What sort of Hail Mary nonsense might he trot out in his last electoral gasps?