I have not had enough time to look closely at the election returns in all target states, much less all the 50 states. But my initial take on the returns in two neighboring states, Ohio and Indiana, shows an interesting contrast. Ohio was a target state in 2004 and 2000; Indiana wasn't but was this time. Indeed, from the 1970s up through 2004, Indiana has been in national elections much more Republican than the neighboring Great Lakes states. That changed in the 2006 off-year elections, when Democrats picked up three seats by impressive margins, and polls in the 2008 cycle indicated that Barack Obama was competitive there. As a result, the Obama campaign used its huge resources in money and manpower to make Indiana a target state.
Obama also carried Ohio 51 percent to 47 percent, a 3-percentage-point Democratic improvement over 2004, when George W. Bush carried it 50 percent to 48 percent (I am using the figures from Dave Leip's Election Atlas in all cases here). That's a 3-percentage-point Democratic gain and a 3-percentage-point Republican loss. Obama got 1 percent more popular votes than Kerry; McCain got 10 percent fewer popular votes than Bush. Overall turnout, as now reported, was down 4 percent. (I think this figure is a little fishy; with some votes still unreported; the numbers show overall turnout down between 31 percent and 44 percent in Marion, Tuscarawas, and Clinton counties.) It rose, at least according to these figures, in 13 of 88 counties, with the highest percentage rises in exurban counties where the population has been growing fastest (Delaware County outside Columbus, Warren County outside Cincinnati.)
McCain actually ran ahead of Bush's percentage in six counties along the Ohio River—southern-accented coal country, where Obama did very poorly in the Democratic primary—and also, curiously in Clark County (Springfield), which switched from Gore to Bush in 2004 after voters were bombarded by letters from readers of the British left-wing Guardian urging them to vote for Kerry. The Guardian evidently has a lasting effect. Republican percentages were down negligibly in the steel country around Youngstown and Warren (Mahoning and Trumbull counties) and not by much in Cleveland's Cuyahoga County and adjacent suburban counties. But if the coal-and-steel country didn't switch away from Republicans, voters in northwest Ohio's auto-centered economy did. And Columbus's Franklin County continued on its recent march away from its ancestral Republican Party.
Overall, however, the Ohio 2008 numbers look a lot like the Ohio 2004 numbers. A modest shift, large enough for a Democratic victory, but nothing looking like an earth-shattering realignment.
Indiana is quite different. George W. Bush carried Indiana 60 percent to 39 percent in 2004; John McCain lost it 50 percent to 49 percent in 2008. The Democratic percentage rose and the Republican percentage declined by a whopping 11 percentage points. Turnout was up 11 percent (these look like final figures), a large number in a state with relatively low population growth (above the regional average but well below the national average). Obama's popular vote was an enormous 41 percent higher than Kerry's (and compare that with Obama's 1 percent gain in Ohio). McCain's popular vote was 9 percent below Bush's in 2004 (almost the same as the 10 percent figure in Ohio).
The biggest turnout increases in Indiana, in percentage terms, came not in industrial counties where the economy has been lagging but in the ring of relatively affluent suburbs around Indianapolis and also in two university counties, Monroe County (Indiana University) and Tippecanoe County (Purdue). Turnout was down in some small-factory-town and rural counties. In the two counties with the largest black percentages, it was up an impressive 17 percent in Marion County (Indianapolis) and only 10 percent in Lake County (Gary).
The Democratic percentage, unsurprisingly given the statewide numbers, increased in all 92 counties, with some of the largest increases in the suburban counties around Indianapolis (including heretofore arch-Republican Hamilton County, the most affluent of them) and the smallest increase in southern-accented rural counties in southern Indiana, where Obama did not run well in the Democratic primary.
Lesson: Organization matters. I was not sure how much the Obama organization could deliver in actual votes. The answer turns out to be a lot. The Indiana results are very impressive, as are those, which I have not studied as closely, in Virginia and North Carolina, which like Indiana weren't target states in 2004. But organization counted for less—seems to have made a much more marginal difference—in states where the two parties competed in organization before, like Ohio. The Bush and Kerry campaigns, and their allies, were already doing something in 2004 that was much like what the Obama campaign did in 2008 in Indiana. The Kerry Ohio organization did an excellent job of turning out black voters in the central cities. The Obama campaign did a little better, and it seems to have fanned out into other, less heavily Democratic areas. The Bush organization in Ohio in 2004 did an excellent job of turning out votes in all kinds of areas, as showed up in a comparison of 2000 and 2004 results. The McCain campaign may have been on these efforts but was less effective. In Indiana you see almost no evidence of a Republican organizational effort.
The contrast between Ohio and Indiana leads me to wonder about the conclusions we draw about different demographic groups from the national data. The difference between responses in target states and in nontarget states is so striking that we should perhaps examine them separately.
The 2004 Bush and Kerry campaigns showed that organizational efforts could make a vast difference, more than I would have thought in the 1990s, when turnout was much lower. The 2008 Obama campaign, though it did not raise turnout hugely nationally, shows that organizational efforts can make a huge difference, especially when there's no appreciable organizational effort on the other side. The Obama campaign took something of a risk. If the national numbers had stayed where they were in the first half of September, with McCain slightly ahead, the Obama organizational efforts in Indiana and North Carolina (though not, probably, Virginia) would have gone for naught. But the Obama managers had enough money and manpower to gamble, and they won big and impressively. More to come....