Rick Santorum, despite being a Catholic, does significantly less well among Catholic Republicans nationally than does Mitt Romney, who is a Mormon. Santorum, on the other hand, does better than Romney among Protestants.
The data on which I base these conclusions - an aggregate of our Gallup daily tracking of Republicans nationwide in February -- are presented in the table below.
Catholic Republicans supported Romney over Santorum by a 12-percentage-point margin in February. Protestant Republicans supported Santorum over Romney by a four-point margin.
It comes as no great shock to note that 88% of Mormon Republicans supported Romney.
Those Republicans who have no religious identity split their support about evenly between Romney and Ron Paul, with Newt Gingrich and Santorum behind. (I didn't include in this table Republicans who identify with other, non-Christian religions because of low sample sizes).
Now, as you have no doubt noted by this time, the data show that the religiosity of Republicans also makes a big difference in candidate preference -- within both Protestant and Catholic segments of Republicans.
Santorum leads Romney by 14 points among Protestant Republicans who attend church weekly. Romney ties Santorum among those who attend nearly every week or monthly, and Romney moves to a nine-point lead among Protestant Republicans who seldom or never attend church.
The same thing occurs within Catholic Republicans. Santorum wins by five points among weekly church-going Catholics, but loses badly to Romney among Catholics who are less frequent church attenders.
So, when we put it all together, we find only two segments included in this analysis in which Santorum beats Romney: 1) weekly-church-attending Protestants and 2) weekly-church-attending Catholics.
Keep in mind that these data are an aggregate for all of February, a month that saw Santorum surge and then fall back in support. But the central conclusion is clear: Santorum's base of strength lies among highly religious Republicans, regardless of their religion.
Now, let's look ahead to Super Tuesday, when Republicans in 10 states will be going to the polls. Our basic assumption is that -- everything else being equal -- Santorum should do best in the states that have the highest percentage of weekly church going Republicans.
Here are the data:
Everything else being equal, Santorum should do best in the states at the top of the list -- the most religious states, and Romney should do best in the states at the bottom of the list -- the least religious states.
Tennessee and Oklahoma -- both highly religious states with half or more of Republican residents saying they attend church weekly -- should tilt toward Santorum.
Vermont, Massachusetts, and Alaska -- all three with a third or less of Republican residents saying they attend church weekly -- should tilt toward Romney.
Of the remaining states, our "church attendance predictor" suggests that Idaho, North Dakota, and Georgia could tilt toward Santorum, and Virginia should tilt toward Romney.
The critical state of Ohio sits right in the middle.
Of course, there are other factors involved in the voting behavior of Republicans in these states. Romney is from Massachusetts, and Vermont is contiguous to Massachusetts, so his almost-certain victory in those states is going to be due as much to his history there as to the relative low rate of religiosity among Republicans in those states.
Since Newt Gingrich and Santorum are not on the ballot in Virginia, Romney will do well in that state no matter what the level of religiosity of its Republican residents. Gingrich is from Georgia, so he would be predicted to do at least relatively well there, even if the state were full of atheist Republicans.
Now, what about Ohio? That's the most crucial state of all next Tuesday. As I've been reviewing, Romney victories in the New England states and perhaps in Alaska and Virginia are pretty much expected, and Romney losses in Georgia, Tennessee, and Oklahoma are pretty much expected as well. But Ohio is much more of a bellwether state. A Romney victory there could go a long way toward cementing his front-runner, inevitable status, and a Romney loss there could do real damage to his claim to be the inevitable nominee.
What does the relative religiosity of Republicans in Ohio tell us? Well, one clue comes from a comparison of the religiosity of Republicans in Ohio and in Michigan. We find that Republicans' religious behavior is almost identical in these two states. To be specific, 42% of Republicans in Michigan attend church weekly, compared to the 44% in Ohio. And 36% of Michigan Republicans seldom or never attend church, compared to 32% of Republicans in Ohio.
Romney won Michigan by 3% points on Feb. 27. The relative similarities of the religiousness of Republican residents in Ohio compared to Republicans in Michigan predict a similarly close race in Ohio next Tuesday.