There have been two major polls (I don't consider Fox a "major") since the middle of June--The Washington Post and Gallup--and in both of them Kerry leads Bush among registered voters (in the Post poll by 8 and in Gallup by 4). Not bad for John Kerry, not bad at all.

Some are dissatisfied, however. Why isn't he farther ahead? What about some of the minor polls where he isn't doing do well? Why does Kerry's lead sometimes vanish, even in the major polls (as indeed might happen tomorrow for all I know)?

Calm down. Say your mantra. And meditate on these recent words from Frank Newport, editor-in-chief of The Gallup poll:

Based on historical patterns, Bush's job approval rating is thus underperforming the pattern of presidents who have won re-election. In the broadest sense, Bush's job approval rating has generally been remarkably stable this year, averaging about 50% (which is a symbolic dividing line for an incumbent seeking re-election) since mid-January. The current downtick in his ratings puts him below the pattern of successful presidents. Having a rating below 50% (as is the case with his last four ratings) is not a good sign for an incumbent. If Bush wins this November, he would be the first president since Harry Truman to come from a below 50% rating to win re-election.

The trial heat patterns of the three presidents who eventually lost were erratic enough, however, to suggest that fluidity is the norm rather than the exception in trial heat ballots at this point in the campaign.

Feel better (at least if you're a Kerry partisan)? Eventually, of course, Kerry does need to take and maintain a solid lead, but it is unrealistic to expect that to happen this early in the campaign. Perhaps after the Democratic convention such a pattern might begin to emerge, but that would be the earliest.

So expect fluidity to continue for awhile. As many have pointed out, deciding you want to fire the incumbent (where we are now) is a different decision than deciding you definitely want to hire John Kerry (the stage of the campaign we are moving into). Once we reach that stage, many swing voters will come off the fence and the political picture should be clarified.

And, make no mistake about it, despite apparent shrinkage in the swing voter pool, there are still plenty of these voters around. According to a just-released report by the Pew Research Center, about a fifth of voters can be classified as swing at this stage of the campaign--sizable, though still down 6-11 points from the proportions of swing voters at analogous points in the 1992, 1996 and 2000 campaigns.

In the Pew data, swing voters who currently express a preference are about equally divided between Bush and Kerry. However, their approval ratings for Bush on the economy (36 percent) and Iraq (34 percent) are notably low. In addition, swing voters in their June data are heavily moderate (49 percent) and independent/no party preference (45 percent) or Democratic (36 percent). Thus, John Kerry appears to be well-positioned to make headway among these voters over time.

Two other interesting notes from the Pew report. They pooled their data from April to June to get enough respondents to look at Bush approval ratings in some key battleground states. Generally, the rank order of approval rating follows from the 2000 results: Bush approval is higher in states he carried in 2000; lower in states he did not. But there is one important exception. Among the nine states they provide data for, Ohio is ranked dead-last in Bush approval, with just 41 percent approval and 46 percent disapproval..

Another intriguing finding is an apparent narrowing of the "religiosity gap"--that is, the tendency for those who attend church more often to vote Republican with far greater frequency than those who attend less often. According to the Pew data, the gap in Bush support between those who say they attend church every week and those who attend seldom or never is now 14 points, compared to 27 points in the 2000 VNS exit poll.

Who knows if this will hold up in this year's election, but it's food for thought. After all, we have not always had the relationship observed in 2000 between church attendance and support for Republican candidates. For example, in the 1980's, there appears to have been only a weak relationship between church attendance and Republican support. But that relationship became quite noticeable in 1992, strengthened in 1996 and strengthened some more in 2000.

Who's to say that relationship might not start heading in the opposite direction? Despite Bush's best efforts, he has had little success inflaming the culture wars and we are now 4 years past the sex scandals that dogged the last years of the Clinton-Gore administration. Keep an eye on this one as we head toward November.