As the Democrats wrested control of Congress from the Republicans for the first time since 1994, exit polls revealed some changes in the pattern of faith-based politics, but also much continuity. Democrats won 53 percent of the two-party vote for the House of Representatives, a five percentage-point gain over 2004 (a presidential year) and a six point increase from 2002 (the previous midterm election).
Where did these new Democratic votes come from in religious terms? Focusing on House races (since House races were on the ballot throughout the nation in 2002, 2004 and 2006), we can see that the largest gains for Democrats came from groups with a history of voting Democratic. For example, the Democrats made large gains among Jews (89 percent of the two-party vote in 2006, an increase of 21 percentage points since 2002) and people of other non-Christian faiths (69 percent in 2006, a gain of 13 percentage points over 2002). The victors also improved among voters with no religious affiliation (74 percent in 2006, up 12 points over 2002). All these gains were at least twice the size of the overall shift in the congressional vote. So the Democratic victory was fueled in part by non-Christian voters of one kind or another.
White Catholics have been a key swing constituency recently, and in 2006 they swung Democratic. The Democrats increased their share of the white Catholic vote by six percentage points over 2004, more than the overall shift in the congressional vote. However, this change was largely a recovery from the Republican swing in 2004; the Democrats’ improvement over 2002 was just two percentage points. The white Catholic vote remained closely divided in 2006: 51 percent Democratic and 49 percent Republican.
If nonwhites are included, the Democrats won a solid 56 percent of the entire Catholic vote in 2006, a marked change from 2004. This figure reflects in part the persistently strong Democratic vote of Latino Catholics. In fact, the vote of all minority Christians combined (including black Protestants) displayed strong continuity: three-quarters backed the Democrats in 2006, about the same as in 2004 and 2002.
The initial exit poll results did not define white Protestant communities particularly well, but a careful look at these data and other surveys suggests strong continuity. White evangelical Protestants gave Republican congressional candidates 73 percent of the two-party vote in 2006, down only three percentage points from 2004 and two points from 2002. So this element of the GOP base remained largely intact, contrary to expectations that evangelical discontent would be devastating for Republicans.
Among white mainline Protestants, Democrats gained three and two percentage points respectively over 2004 and 2002. But like white Catholics, mainline Protestants remained closely divided in 2006—52 percent voted Republican and 48 percent Democratic. Thus, the Democrats’ improvement among white Protestants was substantially smaller than the overall shift in the congressional vote. But even such modest changes may have been crucial in close contests.
The relationship between worship attendance and the congressional vote also showed some shifts in 2006. The least observant voters (who never attend worship) voted 69 percent Democratic and 31 percent Republican. In this category, the Democrats picked up seven percentage points over 2004 and 12 points over 2002. In contrast, the most observant voters (who attend worship more than once a week) gave Republican congressional candidates 61 percent and the Democrats 39 percent—virtually the same figures as in 2004 and 2002.
The Democrats improved among those who fall in the intermediate levels of worship attendance, posting gains over 2002 of ten percentage points among yearly attenders, and five percentage points among monthly and weekly attenders. However, partisan differences by level of attendance actually increased in 2006: the difference in the Democratic vote between the least and most observant voters was 30 percentage points (69 minus 39), compared to 25 percentage points in 2004 and 19 percentage points in 2002.
These findings represent the continuation of the gap between the more and less observant voters. Despite the Democratic gains among white Catholics in 2006, the at-least-weekly attenders at worship voted Republican on balance, while less-than-weekly attenders voted more Democratic. (Here the attendance gap narrowed very slightly to seven percentage points in 2006 from eight points in 2004 and 2002.)
Taken together, these patterns suggest that Democrats can gain by mobilizing religious minorities, the less observant and the nonreligious, and by competing for more religiously observant voters at the margins. As for the Republicans, they cannot rely only on the traditionally religious to win national elections. They must compete at the margins for less traditionally religious and secular voters—which is exactly what they did in their 2004 and 2002 congressional victories.