Justin Sullivan, Getty Images
BOSTON, MA - MARCH 06: Republican presidential candidate, former Massachusetts Gov. Mitt Romney arrives at a Super Tuesday night gathering on March 6, 2012 in Boston, Massachusetts. Mitt Romney celebrated Super Tuesday wins in Massachusetts, Vermont and Virginia. Super Tuesday could play a vital role in the nomination of the Republican candidate as ten states hold their primaries or caucuses on the same day. (Photo by Justin Sullivan/Getty Images)

Mitt Romney held serve on Super Tuesday, to borrow a tennis metaphor, but he couldn’t break serve. The game he had to win was Ohio, and he did so — barely. The game he needed to put the match away was Tennessee, and there he failed.

Romney dominated the delegate fight Tuesday night, winning 199, by one estimate, compared to 80 for Santorum and 68 for Gingrich. The math for Santorum now looks daunting. He would need to win 65 percent of the remaining delegates to secure the nomination. Still, he remains in the game, draining energy and cash from Romney, who still struggles to close the deal.

Once again, evangelicals were Romney’s nemesis. The national media is making much of Romney’s failure to win over the GOP base, but this masks the real demographic story: 73 percent of the voters in Tennessee were evangelicals, and they chose Santorum over Romney by 19 points.

Stepping behind the veil of religious identity, one of the most effective survey questions this year asks how important it is that a candidate “shares your religious beliefs.” On its face, it is a strange question in a polity that eschews religious tests. But for that very reason it has been valuable throughout this primary season. Note that it specifically targets religious doctrine, not shared values. A “great deal” response thus appears to be a clear assertion of voting based mainly on religious identity.

Forty-three percent of Tuesday’s Tennessee voters said that it mattered a “great deal” to them that the candidate “shared” their religious beliefs; 53 percent of these went to Santorum, 25 percent to Gingrich and just 16 percent to Romney.

This compares with the 39 percent of Michigan’s primary that was evangelical, and 24 percent who selected “a great deal” on religion identity. Sixty-three percent of these went to Santorum.

As in Michigan, there is almost no daylight between the “very conservative” vote and the “great deal” religious alignment vote in Tennessee; 41 percent called themselves “very conservative,” and they tilted to Santorum 48 percent to 18 percent. The two groups overlap to such a degree that any analysis that focuses on ideology is likely missing the real story.

The combined Tennessee anti-Romney vote on the religious ID question was 76 percent — a stunning 60 percent gap. Precisely one-third of Tennessee voters may have been unavailable to Romney based on religion alone.

In other Bible Belt news, 72 percent of Oklahoma voters were evangelical, but only 30 percent said the candidate’s religion mattered “a great deal” to them — 13 percent fewer than in Tennessee. Of these, Santorum again won 53 percent and Gingrich 23 percent with Romney stuck at 18 percent. Oklahoma evangelicals thus appear to be more cosmopolitan than Tennessee’s, but those voters who did prioritize religious alignment behaved identically to their Tennessee fellows.

In Georgia, 64 percent of voters were evangelicals and 37 percent said religion mattered “a great deal.” Romney got just 9 percent of these, with Santorum and Gingrich splitting a whopping 85 percent. This is likely somewhat skewed because of Gingrich’s home turf advantage here, but the 9 percent floor for Romney is still remarkable.

In Massachusetts, just 15 percent of the voters were evangelical, and Romney won 57 percent of their vote. Meanwhile, in Vermont, 27 percent were evangelical, and Romney edged Santorum 39 percent to 33 percent. In neither state was the religious motivation question asked. So while in the Bible Belt Romneys evangelical share maxed at 30 percent, in New England it would appear that evangelicals are more open-minded.

Romney’s messy policy positions encourage the narrative that his difficulties are ideological rather than religious. The two almost certainly do feed on each other, but logic also demands that we take the voters at their word on the religious alignment question. No other survey question that consistently draws a 25 percent or greater response produces such a sharp separation between the leading candidates.

Romney now has a prohibitive advantage in the delegate hunt, and he can take reassurance that “evangelicals” and the “very conservative” in state polls suggest strong solidarity for the eventual GOP nominee. The real questions are how much blood and treasure will be spent getting to Tampa — and how enthusiastic the evangelical base will be in November.

Eric Schulzke writes on national politics for the Deseret News. He can be contacted at eschulzke@desnews.com.