Here are two things we can safely say about this election:

1. It will be a high-turnout election.

2. High turnout will benefit the Democrats.

Here's how we know it will be a high-turnout election:

First and foremost, this is a very high interest election. Data across a wide range of polls have persistently shown that voters are expressing more interest in this campaign and are following it more closely than they were at comparable points in the 2000 and 1996 campaigns. These indicators suggest that, on the basis of interest alone, voter turnout could be comparable to that in 1992.

And besides high interest, this is an election where there have been high levels of new registrations and voter contact by the "ground games" of both parties. Therefore, not only are voters more interested, they are also more likely to have been provided with the opportunity to mobilize that interest and convert it into voting on election day (or before). This suggests that turnout could potentially surpass that in 1992.

Here's how we know that high turnout is likely to benefit the Democrats:

The basic reason is simple. Democrats enjoy support from a number of "peripheral" constituencies this year whose participation levels are typically low and can be difficult to get to the polls. But in a high turnout election, electoral intensity draws these constituencies into the process and tends to produce not only an increase in their turnout--after all, most groups will experience at least some increase in turnout--but an increase in turnout that is higher than that of more mainstream constituencies. Therefore, the higher the turnout, the higher the payoff for the Democrats, because their peripheral constituencies are disproportionately mobilized into the process.

One such constituency is young voters (18-29). While there have been exceptions, and Kerry's lead has varied over time, most polls most of the time have shown Kerry with a healthy lead over Bush among young voters in general, and college students in particular.

Another constituency is new voters. Again, while not all polls agree and Kerry's lead has varied over time, most polls most of the time have shown Kerry with a strong lead over Bush among new voters.

Underscoring the new voters pattern is Democratic success in generating new registrations this year, particularly, of course, in the battleground states. While there is some dispute over who won these "registration wars"--and we may not have a final answer until after the election when more and better data will be available--I am persuaded that the Democrats have had greater success registering voters where they matter the most. And a high turnout election is just what is needed to bring these new registrants, whose participation rates are typically less than that of already-registered voters, to the polls.

Another key--perhaps the key--constituency is minority voters, whose support for the Democrats is exceptionally high. Bush's black support generally registers in the 7-10 percent range (though there are exceptions; see my recent discussion of one of those exceptions) and Bush's Hispanic support has been running around 30 percent in nationwide or multistate polls of Hispanics. (Such polls--as opposed to regular national polls--provide for the Spanish-language interviewing and other arrangements needed to get a proper sample of Hispanic voters and therefore provide better measures of Hispanic sentiment.)

Finally, recent work by Victoria Lynch of the DLC, based on National Election Studies data, shows that peripheral voters in general--those who are not highly committed to voting and tend to surge in and out of the electorate depending on their interest in the election--tend to lean naturally toward the Democrats, not the Republicans. As the DLC's memo on the report summarizes these tendencies:

Peripheral voters are much more like Democrats than Republicans in supporting an activist government; in their commitment to equal opportunity; and in their rejection of cultural conservative "wedge issues." Demographically, peripheral voters are more like Democrats than Republicans in that they are relatively younger, less educated, more likely to consider themselves "working class," less likely to attend worship services regularly, and much more likely to self-identify as ideological "moderates" rather than conservatives. Indeed, this analysis casts a lot of doubt on Republican claims that non-voting Christian conservatives are a big part of the pool of "mobilizable" peripheral voters -- in part because these voters are disproportionately disengaged from civic as well as political involvement, and do not readily follow opinion-leaders, much less the "voter guides" distributed in churches that they do not regularly attend.

The last point is important because it helps debunk Karl Rove's infamous--and specious--claim that there were 4 million "missing" conservative white evangelical voters in the 2000 election who could potentially be turned out in this election. (If further debunking of the missing 4 million is needed, let me recommend Marisa Katz' demolition job on The New Republic's website.)

So, in sum, a high turnout election seems very likely and a consequent advantage for the Democrats very likely as well. And the Democrats' ground game seems to be running in high gear and fully capable of maximizing that advantage (see Harold Meyerson's excellent new piece on Democratic mobilization efforts). In a close election--and that seems a very distinct possibility--this turnout advantage for the Democrats could not only be important, but decisive.