Republican Scott Brown’s 52%–47% victory in Massachusetts’s special Senate election on January 19 has shaken the political firmament. It’s a stunning achievement for a Republican to win in a state Barack Obama carried 62%–36% and that in the last four presidential elections has voted on average more Democratic and less Republican (61%–33%) than any other state. Such a result is even more stunning because Brown announced on December 28 that he would be the 41st vote—the decisive vote—against any Democratic healthcare bill.

It’s interesting to look over the numbers, nicely presented in this excellent New York Times graphic, to see how Brown won. The turnout, as it happens, was 1.3% higher than the turnout in the 2006 election, which included a race for governor, and in which more votes were cast than in that year’s Senate race. Western Massachusetts voted heavily Democratic, especially in Berkshire, Franklin, and Hampshire counties (which, if they were a separate state, would have been where Barack Obama had his highest-percentage state win in 2008—and would have been so even if the San Francisco Bay area were also a separate state); Hampden County, which is basically metro Springfield, went for Scott Brown. At the other end of the state, Cape Cod (Barnstable County) voted Republican, as it usually does in seriously contested races but did not in 2008.

The people who were the purported beneficiaries of the Democrats’ healthcare legislation did not come out in force to support it.

I will concentrate on the rest of the state, centered on Boston, which used to be known, to locals at least, as the Hub of the Universe. From the Hub, or from its inner beltway Route 128, half a dozen limited-access highways spread out, and I’ll look at the votes in major towns and cities in each of those corridors.

The first is what I would call, after a phrase in a recent David Brooks column, “the educated class” corridor. This runs west from Boston (and includes its northern gentrified wards) and Cambridge, roughly between Route 2 and Route 9 out past the Route 128 inner beltway. The following table shows cities and towns in this area, with the difference in percentage turnout from November 2006, the percentage for Scott Brown in January 2010 and John McCain in November 2008, Brown’s percentage gain over McCain, and the percentages of college graduates, blacks, Hispanics, and Asians.

This corridor accounts for almost all the cities and towns that voted for Coakley outside far western Massachusetts. Turnout, in general, was up or down only mildly over 2006. Brown ran better than McCain, but not much better in the most liberal areas. He fared very poorly in academic Cambridge and not much better in heavily Jewish Brookline and Newton. The major exception was Waltham, which still has white working-class neighborhoods; Brown carried it, and it was the only city or town in this group in which his gain over McCain was larger than the +14% he needed statewide in order to get 50%. “Educated class” voters may have been unenthusiastic about Martha Coakley and the excitement of the race did not prompt much of a turnout increase. But they stuck with the Democratic nominee and if other voters had reacted as they did, Scott Brown would have been defeated.

Now let’s consider the northern corridor along U.S. Route 3 and Interstate 93, which run northwest and north from Boston.

These cities and towns, with varying educational levels and relatively few minorities (but with a few interesting exceptions in each case), almost all gave Scott Brown the increase over McCain that he needed to win. The biggest exception was Medford, home of Tufts University. My sense is that the Route 3 and I-93 corridors have fewer academics and people who identify with universities than the Route 2 and Route 9 corridors, and also a much smaller Jewish population. More are in private-sector business rather than the professions. This corridor was carried pretty solidly by Scott Brown and contributed importantly to his victory.

Now let’s take a break from the corridors and look at the old mill and factory towns which have been losing population over the years. Many once had populations over 100,000 and now are well under. They are scattered around the state and don’t form a cohesive geographic unit.

A clear pattern: turnout down, in most cases sharply; Republican percentage up, in each case by more than Brown needed statewide to win; lots of Hispanics, some significant percentage of which must have voted for Brown (he couldn’t have gotten much below 30% in Lawrence, for example). Factory and mill town Massachusetts responded very differently to this election than “the educated class” Massachusetts. The people who were the purported beneficiaries of the Democrats’ healthcare legislation did not come out in force to support it—and seem to have in many, perhaps most, cases voted for the candidate who promised to oppose it.

Now let’s turn to Scott Brown’s home territory, the corridor running southwest of Boston along I-95.

Franklin is next door to Scott Brown’s home town of Wrentham, and his local appeal is obvious in the numbers here. But overall we are looking at pretty well-educated voters, who are private sector rather than university-oriented, and we see excellent showings for Brown.

Finally, the South Shore corridor along Route 3 as it heads from Boston to Cape Cod.

Education levels are somewhat lower than along the southern I-95 corridor, but Brown improved on McCain’s percentages and carried these towns about as impressively as he did there. Randolph, with its large black population, is the outlier in this region, with turnout down rather than up and voting Democratic rather than Republican; the high percentage of Asians in Quincy do not seem to have produced similar results. Five of these six towns—Milton, Braintree, Marshfield, Scituate, and Weymouth—along with nearby Pembroke had the highest percentages (between 33% and 38%) of residents in the state identifying their ancestry as Irish. It looks like the children and grandchildren of the residents of the Irish wards in Boston who elected Democrats like Speaker John McCormack (a House member from 1928 to 1971) voted in large numbers for the Republican this time.

Scott Brown’s victory has many fathers and mothers. But the picture is fairly simple. Massachusetts’s “educated class” corridor continued to vote heavily Democratic, as did the sparsely populated west end of the state. But the rest of the state swung sharply to Republican Scott Brown. Those who turned out in factory and mill towns voted for him by unusually high percentages, and he racked up big margins in the southern I-95 and Route 3 corridors and did almost as well in the I-93 and Route 3 northern corridors. These are all areas that political reporters seldom visit (except for perhaps driving up to New Hampshire from Logan Airport); they tend to hang around within striking distance of the Parker House bar in Boston and Harvard Square in Cambridge. But the real political story in Massachusetts—the upset victory that killed the Democrats’ healthcare legislation—took place elsewhere in the state.

Bottom line: The Democratic candidate held “the educated class” corridor, but failed to generate good turnout or high percentages in the factory and mill towns. The other corridors swung heavily to the Republican candidate. Under Barack Obama, who has chosen to live all his adult life in university neighborhoods, the Democratic Party has been reduced, at least in Massachusetts and in this special election, to “the educated class”—and not much else.

Michael Barone is a resident fellow at the American Enterprise Institute.

FURTHER READING: Barone specializes in analyzing elections and districts. He’s recently explained “Why Parties Last Longer in Britain,” reviewed “A Keystone Election” in Pennsylvania, and discussed the correlations between “Delayed Childbearing and Voting Behavior.” He has noted “An Immigration Tipping Point” in America’s decline in foreign-born population and explored “Republicans and Democrats: A Tale of Two Bases.” Barone also thinks the “GOP Should Push Education and Pro-Family Tax Reform” and that President Obama’s presidency is based on two “Mistaken Assumptions.”

Image by Darren Wamboldt/Bergman Group.