I've been arguing for quite a while that the key numbers to look at are not the horse race numbers between Kerry and Bush but rather all the indicators that show voters losing confidence in the incumbent. We've had those in abundance for a couple of months, including indicators that show declining voter confidence in Bush's handling of the war on terrorism, once his seemingly impregnable electoral advantage.

But now that Kerry seems to have taken a small lead in the horse race and now that the media have finally absorbed the fundamental fact that Bush is doing poorly, not well, in comparison to previous incumbent presidents, perhaps it's time to strike a few cautionary notes. After all, the election is still over five months away, the lead may change hands again several times, and Kerry's position, while strong, is hardly unassailable.

A first cautionary note worth paying attention to is provided by Matthew Yglesias in an article on the The American Prospect website. As Yglesias rightly points out, the elections about which we have relevant polling data only go back to 1948, which is a mighty small dataset. Hence, just as caution was well-advised when it appeared by historical standards that Bush would win, so is caution well-advised now when it appears that, by those standards, he is likely to lose. Moreover, as Yglesias points out, if you expand the number of cases under consideration by including elections where a sitting vice president runs to succeed an incumbent president (1960, 1988, 2000) as sort of quasi-incumbent elections, the historical picture looks a little cloudier.

Personally, I still think Bush is in a great deal of trouble. But we should be cautious about relying too much on the historical record in assessing his likelihood of losing.

Another cautionary note is struck by Terence Samuel, also on the Prospect website. His article, "Chicken Littles Recant", points out how quickly Democrats tend to go from being more depressed than they should be by political trends (oh no, Kerry's only running even with Bush; he should be ahead by 10 points; disaster looms!!) to being excessively optimistic (Kerry's ahead, Bush is sinking fast--Kerry's going to win by a landslide!!)

As he points out:

This is an up moment, but it was only a few weeks ago that some influential but unnamed Democrats were wringing their hands on the front page of The New York Times about how Kerry was blowing their big opportunity to win back the White House.

Let's try to remember this and not get similarly silly the next time the polls and news cycle go south for Kerry.

Finally, Josh Marshall cautions us not to conclude from the current good news that now is the time for Kerry to ratchet up his aggressiveness and take center stage away from the president. As he puts it:

....partisan polarization will intensify in the coming months. And that will help the president in many ways, getting some of the attention off him and on to Kerry. But a judgment about the president like the [negative] one I've described above, once made, can be hard to unmake. And for the moment, with so many of the president's actions delivering abysmal dividends to the nation he's led, that judgment is being made against the president. So, for the moment, I'm not sure having Kerry give Bush center stage is such a bad thing.

Of course, that doesn't really tell us when Kerry should turn up the heat. But it's a reasonable point that we should not necessarily assume that a bit of good news means that time is now.