Belcher, who I interviewed for my January Atlantic article on politics and race, views the average voter as a mostly rational figure whipsawed by crosspressures, and his analysis set to figure out the relative force of those pressures. What was it that tilted folks from one side to the other?
To identify issue groupings, Belcher used a statistical technique that reduces a large number of questions or topics to smaller number of basic factors. This technique identified two factors -- 2008 Election Issues -- and the Faith vs. Secular Values axis.

The graph is a little fuzzy, but the points show how voters indentified themselves on issues vis-a-vis where they perceived the candidates to be. The split -- represented by the line -- is roughly horizontal, which means that perceptions of each party on the dominant issues were more important than faith issues in driving the vote year. As Belcher writes, "Voters this cycle were closer to Obama and the Democrats on the issues dimension and the gravitational pull of that dimension this year far outweighed that of the faith vs. secular values dimension."
White voters were clustered close to McCain, and black voters anchored themselves firmly in Obama's camp. Young whites were more secular than older whites, but they were also more clustered toward toward the center, suggesting that Republicans could lure them away with the right amount of attention to the right issues.

More voters identify with the Democratic Party today; Democrats, even after the election, maintain a nine point lead in terms of party identification. For voters under 30, the trend is striking: the party ID lead is 20 points for Democrats. Additionally, a running sample of battleground state voters sureyed from 2004 through 2008 shows a slow but pronounced migration from Republican to Democrat. Of the 11 percent of voters who were new to electorate, 62% of them were under 30, and they identify as Democrats by a margin of two to one -- about more new voters under 30 identify as independents as they do Republicans.
Belcher makes a distinction between "faith" voters and "values" voters, writing that "[t]hose
who place a high importance on faith in their voting choices still swing heavily Republican. But although nearly all faith voters (93 percent) are also values voters, the reverse is not necessarily true. Only half (53 percent) of values voters agree that religious faith was a very important influence in their voting choices. In short, many voters see their values as being about more than solely their religious faith."
