But are there as many votes there as Santorum thinks? The old steelworker House district where he was first elected in 1990 has been losing population ever since. And even in 2008, John McCain won non-college whites by a 58 percent to 40 percent margin.
Santorum would probably run better among this group than Romney, whose unforced errors and political tin ear have made him seem aloof.
But there's also a case to be made that Romney may run better among another, less noticed group -- affluent voters.
This year and in 2008, Romney's best showings in primaries have come in affluent areas. And polling seems to indicate that he does particularly well with affluent women.
Those are groups among whom Republicans have been slipping for more than a decade. In the 2008 presidential election, voters with incomes over $100,000 split 49 percent to 49 percent.
You can see the trend in the four suburban counties just outside Philadelphia. The first George Bush carried them with 61 percent in 1988. Since then, the Democratic percentage has been rising steadily, reflecting the liberal stands of affluent voters, especially women, on cultural issues.
Barack Obama carried them with 57 percent in 2008. You see similar patterns in the suburbs in most major non-Southern metro areas.
Santorum carried the Philadelphia suburbs in Senate races in 1994 and 2000. But in 2006, a dreadful year for Republicans, he lost them by 60 percent to 40 percent, a worse loss than McCain's.
Santorum would probably do better this year, with economics overshadowing cultural issues. But it's easy to imagine that Romney would run better in what seems to be his natural terrain.
Political analysts have been assuming that Democrats' gains among affluent voters are solid. But are they more solidly committed than non-college whites?
Both the "Santorum's stronger" and "Romney's stronger" theories seem plausible to me now; neither seems proven. I'll keep them in mind as the race continues.