Here's yet another early voting post, and the usual caveats apply: when comparing Republicans to Democrats, the best thing to do is to take snapshots in time and put them beside similarly sized snapshots from the same period of time four years ago. This is hard to do, because we often do not have the right data sets, and when we do, they're often incomplete. Note, too, that the early voting procedures in states change every cycle, usually in the direction that makes it easier to cast a ballot.

The biggest question is: What's the denominator? If you use 2008, that's like choosing the 1972 Miami Dolphins as your average NFL team. Democrats won by an average of 7 percent--an epic blowout, where two-thirds of sporadic voters were Democrats. In 2010, neither Republicans nor Democrats are seeing this sort of trend-line. 2006 ought to be the standard for comparison--a midterm election.


But some of you asked me earlier how I could make the tentative claim that the "enthusiasm" or "intensity" gap that's reflected in likely voter surveys is not yet materializing. Basically, if Republicans WERE more excited, then  Republican "sporadic" or "infrequent" voters -- as well as Republican 2008 first time voters -- would be voting at higher early rates than Democrats.

In Colorado, the data is not in agreement with the story. First, note that Denver was late getting out its ballots this year, Democrats are more likely to vote in person, and Democrats turn in ballots later. To date, Democrats have turned in a higher number of ballots as a percentage of Democrats versus Republicans compared to 2006. 48 percent of ballots have been cast by Democrats, and in 2006, 46.7 percent were cast by Democrats. According to GOP internal estimates, the party has a 113-vote lead in the number of identified early voters who've returned ballots, and a 4,000 vote lead overall. But the Democrats were further behind at this point in 2008.

In Colorado, Democrats say they're happy with the voter profile of Democrats who are turning out. If there IS a Republican surge, that means that someone with a different partisan identity but the same voting history should be voting at a different rate. And that's just not true, according to the data to date. (See an internal DNC memo here.)

In Iowa, Republicans have returned a higher percentage of their early requests--roughly 7 out of 11--but Democrats have returned more ballots overall--85,000 out of 135,000. Overall, according to the GOP, the partisan breakdown of what that party calls "convenience" voting shows that Republicans have increased their share relative to Democrats; their overall advantage will probably be half of what it was in 2008.

That reflects Republican enthusiasm matching Democratic enthusiasm, rather than a "lead" per se. Democrats have a nine point margin in ballots cast, but this margin is less than it was in 2006. (Itemized by district, Democrats have a solid margin in the three competitive House districts.)

Now, let's look at some Iowa subroups, courtesy of internal data provided by the Democratic Party. You would expect Republicans designated as "sporadic" voters to be returning their ballots at higher rates than Democrats, regardless of the total number of Republicans registered. But Democrats actually have a slight advantage. Republicans have a small advantage among 2008 first time voters who've registered so far--0.6 percent. These voters comprise about 7.2 percent of all Republicans who've requested ballots so far. In three states measured by the DNC, regular Democratic midterm voters are returning ballots at a higher rate than Republicans, which is important because Republicans usually return their ballots earlier than Democrats.

Take Nevada. First-time voters and sporadic Democratic voters are participating at higher rates than first-time and sporadic Republican voters. 

In turn, the GOP's internal estimates show that voters are returning ballots at a rate that is 7.1 percent more Republican in Clark County and 11 percent more Republican in Washoe County than 2008. For the last four years, Republicans have not been registering voters, but Democrats have. Democrats have higher overall numbers. The Republican argument is that a higher percentage of its much smaller electorate is turning out. But Dems have more returned ballots overall. An internal DNC poll in Nevada shows that 92 percent of Republicans surveyed say they will DEFINITELY vote, compared to 88 percent for Democrats. That's a small gap, but much smaller than it was in June.

In Ohio, Republican internal targeting estimates have the party up statewide by 20,000 returned ballots. Democratic internal modeling shows that they're doing more (and have a cleaner voter file).

In Florida, the Republican swing in early voting--40,000 more than at this point in 2008--is a sign of a good turnout program, because 2008 was a presidential election year in which turnout was significantly higher in general. In 2008, Democrats did poorly in Florida until the very last dates of early voting--the very last week. The jury is still out.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.