For a simplified example of this problem, consider a hypothetical presidential election with 125 voters, of whom 110 are white and 15 are black, in which Democrats split the white voters 55 each (50 percent) but won 12 of the 15 black voters (80 percent). This is then followed by a midterm election with only 100 voters (smaller turnout), 90 of whom are white, 10 are black, with Democrats again splitting the white voters 45 each (50 percent), but now winning 9 of the 10 non-whites (90 percent). Though Democratic performance among whites is identical in both cycles and slightly better among blacks in the midterm, in which election did whites account for a greater share of all Democratic votes received? You got it: The congressional cycle, where whites account for 83 percent (45 of 54) of all Democratic votes in the midterm, compared to just 82 percent in the presidential (55 of 67). The reason, obviously, is that there are a greater proportion of white voters in the midterm election.

Third Way employs misleading math to produce an artifactual result. In a methodological note they claim they "normalized" the results between the two elections by simply scaling up the total number of votes cast in 2006 to 2004 levels, arguing that this is akin to taking two batters, one who played in 125 games and another in 145 games, and comparing them by scaling up the 125-game performance to a 145-game projection. Doing so would be fine if they were comparing, say, two centerfielders. But using 2004 and 2006 is like comparing a slugging centerfielder with a singles-hitting second baseman and then, just to be fair, running the second baseman through the copier at 127 percent. To their credit, at the end of the report they finally concede that, "In 2008, all things being equal, turnout should benefit Democrats. The lower share of the electorate in 2006 among traditionally Democratic-leaning constituencies was not a failure of the Democratic turnout machine [but] because 2006 was a midterm election ... In 2008, more African-Americans, more young voters, more urbanites, and more low income voters will come out to vote. Theoretically, this should help Democrats retain their majority."

Theoretically? No, practically: If the Democrats, who lost the 2004 presidential race by 2.4 percent nationally, can simply reproduce in 2008 the performance rates among their traditionally favorable, non-privileged groups they enjoyed in 2006, and couple that with a normal turnout for a presidential cycle, it would not matter if Democrats returned to 2004 performance levels among privileged whites, males, and so forth because they could still win the White House. (I will be saying more about this soon, in a longer piece.) If groups like Third Way are intent on proving that "NASCAR men" or "office park dads" are key targets, that's fine. But they are not allowed to use methodological legerdemain.

--Tom Schaller