Religion's Role in the 2006 Election

The "God Gap" Persists, but Other Gaps Are Much Larger

December 5, 2006

Leading journalists and distinguished scholars gathered in Key West, Fla., in December 2006 for the Pew Forum's biannual Faith Angle Conference on religion, politics and public life. To help journalists better understand the role religion played in the 2006 midterm election, Pew Forum Senior Fellow John Green and American Enterprise Institute Resident Fellow Karlyn Bowman analyzed polling data to address such issues as the extent to which Democrats closed the "God gap," which religious groups were "in play" this election, and whether or not religion polarizes voters. The discussion was moderated by Michael Cromartie, Vice President, Ethics & Public Policy Center.

The survey data referred to by John Green and some of the data referred to by Karlyn Bowman, may be downloaded from PDF of Tables 1-17 Other Forum resources on religion's role in the 2006 election include an analysis of post-election poll data and a transcript of a discussion featuring political activists Eric Sapp and Charmaine Yoest. In the following edited excerpts, ellipses have been omitted to facilitate reading.

JOHN GREEN: Generally speaking, white evangelical Protestants tended to stay in the Republican column, with the Democrats making some small gains from 2004 to 2006, and then from 2002 to 2006. In 2006, the Democrats made a 3-percentage-point gain over their vote in 2004 and a 2-percentage point gain over 2002.

These changes were small, but in a group as large and Republican as white evangelicals, even small shifts can be important. Of course, these are national figures. This change was concentrated in certain key races. A good example would be Ohio. The data we have from Ohio is not as good as what we have from national exit polls, but there is reason to believe the governor-elect, Ted Strickland, did very well among evangelicals in that state, while basically breaking even among white Protestants as a whole. It was quite an important gain.

We didn't see gains like that everywhere in the country, and that is why we see only a small change in the white evangelical category at the national level. This point is especially important given the large number of close elections in 2006.

White Mainline Protestants showed a similar pattern to white evangelicals, about a 3-percentage-point shift. However, the shift was entirely off of 2004 and not off of 2002.

The big change occurred among white Catholics, and there we have two stories. One is a significant shift in the congressional vote from 2004 to 2006, a 5-percentage-point change in the Democratic column. But notice that is much less of a change from 2002 when the Republicans won this group by a small margin. We might conclude there are a lot of swing voters among white Catholics, and in 2006, they swung Democratic. But in 2004, they had swung Republican. They may swing differently in the future.

The other thing to note is that even in 2006, despite a 5-percentage-point gain among white Catholics for the Democrats, the white Catholic community was still evenly divided, with 50 percent for the Democrats and 49 percent for the Republicans.

If one looks at these three large white Christian communities – evangelical Protestants, mainline Protestants and Roman Catholics – one could conclude the partisan gap in those communities narrowed because the Democrats picked up votes in each case, but especially among Catholics. But among unaffiliated voters, non-white voters, Jews and other faiths the gap actually widened. Here religious groups that were already Democratic in 2004 and 2002 became much more Democratic in 2006.

Thus, the Democrats made gains everywhere in 2006, but they made the biggest gains in groups already most in their favor. Part of what happened in 2002 was the Democrats ate into Republican religious constituencies, but they were also able to mobilize more votes from their own religious constituencies. In these data, the group that had the biggest percentage gain was Jews. It's a small community, but, again, in certain races it could have been critical, such as in the Northeast where Democrats did very well.

Another one of the largest changes was among unaffiliated voters. The unaffiliated voters have tended to vote Democratic for a long time, but notice the consistent change over time from 2002 to 2004 to 2006. That is a very important change a lot of people haven't noticed.

Let's turn to Table 2 and Table 3. These tables look at worship attendance, a powerful predictor of the vote. These tables measure religious behavior, which is connected to religious beliefs. The exit polls don't have any beliefs measures. Thus, we can't look at a "God gap" per se; all we can look at is a worship-attendance gap.

Table 2 looks at attendance in the way that it is typically portrayed, dividing respondents into those who attend weekly or more and those who attend less than weekly. You can see quite a difference. The weekly attenders tend to be Republican, even in 2006, and the less than weekly attenders tend to be much more Democratic. But notice the shift in the Democratic advantage across the tables. At the bottom, we calculate the gap, and even in this fairly simple measure, it was bigger in 2006 than in 2004. So while in some measures of religious affiliation, we saw the partisan gap narrow, it widened by other measures.

KARLYN BOWMAN: First, as many of you already know, the God gap, while real, is only one of the many gaps in our politics and it is not the largest. Second, we see powerful continuity and stability in many questions about religion. Some of this is surely what Peter Berger called yesterday "vicarious religion," which I agree is important. Third, although we are polarized and divided on paper, we do not appear to be polarized in practice, and by that I don't mean religious practice; I mean in ordinary life.

I have brought two handouts. The first handout (PDF) (Table 7 is a simplified version) looks at how different groups voted in House elections. I too include the caution John did: The 2006 data are still being cleaned, so some of those numbers may change a little.

The relevance of the first document for this session is as a reminder of those other gaps in our politics. They are larger than the God gap. In the 1960s and 1970s, we talked about the generation gap. Many people argued that young and old would be deeply divided on issues such as Social Security and that we would have generational warfare. More than 40 years later, we are still waiting for that to happen.

The attention religion has received in recent years is healthy, but it is nonetheless interesting why the God gap appears to get more attention than the gap between married voters and those who are not married, between blacks and whites, between voters with a high school education and those with a post-graduate one. The religious attendance gap, which came to public attention about the same time as the gender gap in the 1970s, is fairly new, and that is surely one of the reasons for the attention.

Frequent church-goers, as John Green has pointed out, were more likely to vote for Nixon and for Reagan, but the attendance gap doesn't clearly emerge in presidential voting behavior until 1992 when we saw a double-digit attendance gap.

I remember a conversation I had with the late Richard Beal, who worked with Richard Wirthlin, who polled for Governor Reagan and President Reagan. Rich told me not only were they looking at 22 different groups of women in their gender-gap analysis in the 1980s, but that they were also using a religiosity index as a voter screen, in addition to a patriotism index, which told them a lot more about how people were going to vote than did the familiar broad polling question, "if the election were held today, for whom would you vote."

If you look at John's Table 2 and calculate the net for the gap in attendance overall, you see in 2006 the gap between those who attended church weekly and those who did so less than weekly was 36 percentage points. That is up from 33 points in 2004 and 27 points in 2002. It is about the same size as the gap you'll see in my handout between married and non-married voters – 32 points this year – and smaller than the gap between voters with the lowest and highest reported incomes – 37 points this year – and is obviously dwarfed by the race, partisan, and ideological gaps overall.

Just as a point of comparison, in Gallup's recall data from the 1960 election, there was a gap of 80 percentage points between Protestants and Catholics in the presidential vote.

The second handout (PDF) is a compilation of poll data from different sources, which tells a story of remarkable stability and continuity. Because it's probably so familiar to you, I didn't include Gallup's question first asked in 1948 about whether or not you believed in God, when 94 percent said they did. When Pew repeated the question in August 2006, the response was 91 percent. Nor did I include a question first asked in 1952 when 87 percent told the pollster, Ben Gaffin, they were absolutely certain that God exists. Forty years later, absolute certainty has declined somewhat, but it is still a fairly healthy 72 percent.

As Table 8 shows, in 1937, 73 percent of those surveyed by Gallup said they were members of a church or a synagogue. In 2006, nearly 70 years later, that response was 64 percent.

Although there are reasons to be skeptical of reported church attendance figures in polls, Gallup's question [about church or synagogue attendance] shows remarkable stability over nearly seven decades. (Table 9) The question Gallup has been asking for the past 15 years about church attendance again shows great stability, so too the question about having a Bible in your home.

If the vast majority of our voters claim to believe in God, can belief in God be so clearly linked to political differences? [Let's look] at views about how George Bush talks about his religion and uses it in governing. I wish these were fuller trends but they are not. But if the country were deeply polarized by God, these numbers would look much different. In the Pew trend, (Table 11) around a quarter think Bush talks about his religion too much with around about 10 percent saying too little. A solid majority say it's about right.

There has been a decline in the proportion in the CBS question (Table 12) saying they like the way Bush talks about his religion, though it is still a bare majority. It's unclear to me whether that decline is about Bush's discussion of his religion or about a more generalized dissatisfaction with him.

The CBS question [shown in Table 13] is particularly interesting. Far more people say big business has too much influence on the Bush administration than feel that way about the religious right.

The data on religion in public life again suggest to me we are not deeply divided on religious matters or on the role of religion in public life. The Fox News question (Table 14) shows a significant number of people think religion is under attack. A CBS question shows more people in 2006 than 2004 think people with strong religious beliefs face discrimination.

When you care deeply about something, as Americans do about their religion, you tend to worry it's under attack or losing its influence. Those aren't surprising responses. People who think religion is losing its influence in our lives and on our government leaders overwhelmingly see that as a bad thing. There is strong national agreement that religion should not be excluded from public life and that the courts have gone too far in taking it out of public life. The data in Table 16 generally show people feel the Republican Party is friendlier to religion than the Democratic Party, but only small numbers regard either party as unfriendly.

The larger picture shown by all of these data suggests to me an enormous amount of stability and movement towards more tolerance and acceptance, not the divisions emphasized in the "gapology" literature. We see worries about religion losing its influence in public life, but overall, a public that does not appear to be deeply dissatisfied with the role that religion is playing in public life.

Read the full transcript

Read the full report at pewforum.org