This year's Ohio Republican Presidential Primary was truly contested. In 2008, the Ohio Primary occurred roughly a month after Super Tuesday. McCain had for all practical purposes wrapped up the nomination by then. Huckabee was still in it, but there was nowhere near the level of contesting that there was this year. Ohio GOP primary turnout in 2012 as compared to 2008 was greater. 1,196,566* votes were cast in this year's Ohio GOP primary. That's 113% of the total votes in the Ohio GOP primary in 2008.
Romney barely edged Santorum in a squeaker, 38.0% to 37.0%.
As has been the pattern in previous contests this year and as was expected in Ohio, Romney won the metropolitan areas while Santorum won the rural/small town areas. The possible exception to this pattern in Ohio was Toledo (if you consider Toledo and its surroundings to be metropolitan). Santorum narrowly beat Romney in Toledo's Lucas County and tallied sizable margins in the counties immediately surrounding Lucas (Fulton, Henry, Wood and Ottawa).
Though Santorum lost the big metropolitan areas (Cleveland, Columbus, Cincinnati) to Romney, it appears that Santorum was able to cut into Romney's metropolitan margins. This was especially true in Franklin County where Columbus is located. Romney bested Santorum in Franklin County 40.7% to 36.1%. While that's certainly a win for Romney, it wasn't equal to Romney's drubbing of Santorum in Cleveland's Cuyahoga County 48.7% to 29.6% and in Cincinnati's Hamilton County 48.9% to 29.3%.
Santorum, meanwhile, dominated the more rural areas of Ohio. It was almost enough for him to win - almost, but not quite.
Here's what's interesting in comparing the Ohio GOP primary vote in 2008 with 2012. Turnout in 2012 was higher everywhere, but it was a tad higher still in the metropolitan areas. If you add up the total 2012 vote in the "Cleveland area" counties (Cuyahoga, Geauga, Lake, Lorain, Medina, Portage, Summit), the "Cincinnati area" counties (Butler, Clermont, Hamilton, Warren) and the "Columbus area" counties (Delaware, Fairfield, Franklin, Licking, Madison, Pickaway, Union) you get 596, 457*. That's 49.8% of the total Ohio vote in 2012. If you do the same for 2008, then you get 514,655 votes - 48.6% of the total 2008 Ohio vote.
So, while turnout increased across the board from 2008 to 2012, the share of the vote coming from "metropolitan" areas of Ohio increased from 48.6% to 49.8%.
What this increase in "metropolitan" share of the vote suggests to me is not necessarily that the Romney campaign made an effort to turn out metropolitan voters generally. Instead, it suggests to me that the Romney campaign had a larger, more organized and more effective canvassing effort than did Santorum. Romney voters tend to be more suburban/urban than Santorum voters. If the Romney campaign did a better job at identifying and turning out its likely voters (wherever they may reside), then a good (albeit correlative) piece of evidence confirming it would be an uptick in the "metropolitan" share of the primary vote from 2008 to 2012. There is indeed an uptick.
In Michigan, Romney's endorsements and support from local and statewide elected officials gave his campaign the ability through the officials' campaign apparatuses to dominate Santorum's campaign in the early voting/absentee campaign. Romney's 2-1 margin in the early vote (some 16% of Michigan's total vote) provided for Romney the cushion he needed to eke out a victory. In Ohio, Romney's more organized campaign appears to have done a better job than Santorum's seat-of-the-pants campaign in identifying its likely voters and turning them out. When the margin of victory or defeat is 1 percentage point, organization is all the difference.
Of course, money helps an awful lot too.
*At the time of this writing, 81% of Medina County's precincts have reported. I have extrapolated Medina's as well as Ohio's total vote accordingly.