County Change Predictions In This Year's Presidential Race

I like to track election returns down to the county level nationally, and down to the precinct level in my home state of Minnesota.  Particularly in the last three Presidential election cycles, the nation has become more polarized between urban/suburban and rural.  The national county maps during the Clinton years would have been far bluer than any of the past three elections, as white voters in the South and Appalachia have defected the Democratic Party en masse.  Even in 2008, when Obama won nationally by a decisive seven points, the national county map looked much more red than blue as Obama's strength was concentrated mostly in heavily populated areas while the countryside stuck with the party of the pachyderms.  This year, I anticipate that shrinking national support whether he wins or loses, Obama's county haul will be much smaller than in 2008.  Below is the list of counties state by state that I believe will switch to Romney in 2012, and perhaps in a couple of cases, counties that I believe with switch from McCain to Obama in 2012.....

Alabama--Obama loses Russell, Marengo, and Jefferson County (Birmingham).  Lower black turnout in an uncompetitive state will cost Obama these marginal Democratic counties.

Arizona--The same Democratic foursome that came through for Gore, Kerry, and Obama in 2008 will hang on for him in 2012.

Arkansas--Obama loses Woodruff County, which is the only majority white county in Arkansas that didn't go McCain in 2008.  One of the few states where Obama pretty much already hit the Democrats' basement last time.

California--More substantial Obama losses here given his wholesale domination of the state in 2008.  He loses Trinity, Butte, and Nevada Counties in northern California; Stanislaus, Merced, and Fresno Counties in the San Joaquin Valley; along with San Luis Obispo and Riverside Counties in southern California.  I'm guessing Obama narrowly hangs onto San Bernardino and San Diego Counties this year.

Colorado--Interestingly, most of the counties Obama won in 2008 in the state he won very decisively, so while I think his margins will be down substantially in many of them, I don't think his actual losses will be very substantial.  Right now I'm going with Las Animas and Ouray Counties as his only losses, but expect very narrow Obama wins in Larimer, Jefferson, and Broomfield Counties in suburban Denver.

Connecticut--Litchfield County in CT's northwest corner will move to Romney.

Delaware--No polling at all in Delaware to shed light on whether Biden's home state advantage is worth anything this year which it makes it very tough in regards to Kent County (Dover).  Given that Obama won it by 10 points in 2008, I'll predict he wins it again.

Florida--Pretty amazing how centralized the Democratic vote is in Florida as Obama won the state by a mere three points, winning only 15 out of 67 counties while doing so.  My bet is turnovers to Romney in rural Jefferson County in northern Florida, and in Flagler and Volusia Counties on the Atlantic Coast.  I'm betting Obama ekes out wins in the two primary Tampa-St. Petersburg counties of Hillsborough and Pinellas, but still loses the state.

Georgia--With the state uncontested this time (hard to believe Georgia was a battleground state in 2008), Obama will lose a ton of low-lying fruit, most likely including the Greater Atlanta suburban counties of Douglas and Newton and a bunch of rural Yellow Dog counties in the state's southwest and east-central sides, including Baker, Baldwin, Chattahoochee, Dooly, Peach, Twiggs, and Washington Counties.

Hawaii--Obama hangs onto all four counties in his home state.

Idaha--Certainly Teton County is gone and probably college town Latah County (Moscow) as well.

Illinois--The Illinois county map is likely to look a lot more like the Gore and Kerry county maps in 2012 than the 2008 favorite son Obama map, as a bunch of blue counties are likely to turn red.  In Chicagoland, expect Boone, McHenry, Kendall, and Grundy Counties to turn red, and that's being conservative and predicting that Kane and Du Page Counties stay blue, which is by no means certain.  Downstate will be even uglier, as I predict Romney to pick up the 2008 Obama counties of Bureau, Calhoun, Carroll, Cass, Coles, Henry, Kankakee, Macon, Mason, McDonough, McLean, Montgomery, Pulaski, Sangamon, Schuyler, Vermillion, and Warren Counties.  Even beyond that, there are another handful of downstate Illinois counties that I believe are on the bubble.

Indiana--No other state swung as far from red to blue between 2004 and 2008 than Indiana, and in 2012 I expect to swing at least halfway back to where it was.  So of the 15 Obama counties in 2008, I expect he'll lose Porter, Starke, Madison, Spencer, and Vanderburgh.  Tippecanoe County (Lafayette) is on the bubble, but I'll go with Obama eking it out.

Iowa--Another state where Obama maxed out Democratic performance in 2008 and has nowhere to go but down this year even as most indicators suggest he'll win the state overall.  Going tier by tier from northwest to southeast, here are the 2008 Obama counties that I believe Romney snags this time.....Emmet, Kossuth, Palo Alto, Franklin, Hamilton, Hardin, Delaware, Carroll, Greene, Benton, Audubon, Iowa, Warren, Louisa, Adams, Union, and Clarke.  There's several more on the bubble.  I think Obama's biggest losses will occur in the counties south of Des Moines and Iowa City, where the culture is closer to Missouri.

Kansas--Even in the big Democratic year of 2008, Obama won only three out of 105 counties in crimson red Kansas.  That will drop two in 2012 as he'll lose Crawford County.

Kentucky--It's gonna be bloody.  Frankly I was surprised eight counties held out for Obama in 2008.  But I think this year even most of those counties collapse as Hancock and Henderson Counties in contested Indiana media markets will flip to Romney this year, along with the college town of Lexington's Fayette County, and three of the last remaining blue counties in east Kentucky (Wolfe, Menifee, and Rowan).  Only Jefferson County (metropolitan Louisville) and the die-hard Democratic holdout Elliott County in East Kentucky will stick with Obama this year.

Louisiana--Higher black turnout helped Obama win East Baton Rouge and Caddo Parishes in 2008.  I think lower black turnout in 2012 will flip them back to Romney, but that will probably be it although Tensas Parish on the Mississippi River Delta in northeastern Louisiana is on the bubble.

Maine--There will probably be a few losses here, particularly in northern Maine which Romney is targeting for its' single electoral vote.  Penobscot, Washington, and Somerset Counties all seem like credible losses.

Maryland--Only one of the seven 2008 Obama counties in Maryland is vulnerable in 2012, but the one (Kent County) is a near certain loss.  Interestingly, the 2008 returns suggested some demographic changes were likely underway in Frederick County, and if I'm reading that right, it's very plausible that Frederick County will be one of the few McCain counties that turn blue this year.

Massachusetts--We've narrowly hung onto increasingly conservative Plymouth County in the state's southeastern corner through recent election cycles but I think 2012 is the year we lose it.  Some other 2008 Obama counties (Obama won all 14 MA counties) are liable to be close, but Obama should be able to hang on.

Michigan--Polls show a mid-single-digit advantage for Obama in Michigan compared to 2008's 16-point Obama edge, meaning there's a ton of territory likely to be ceded to Mitt Romney.  On the Upper Peninsula, I expect Romney picks up Ontonagon, Iron, Delta, and Schoolcraft Counties.  South of the bridge, there's even more low-hanging fruit, and I expect to see Alpena, Arenac, Berrien, Cass, Clare, Clinton, Gladwin, Gratiot, Iosco, Jackson, Kent, Leelanau, Lenawee, Mason, Oceana, Ogemaw, Presque Isle, Roscommon, and St. Clair Counties all flip to Romney.

Minnesota--Obama did well in western Minnesota in 2008 because he contested media markets in North Dakota that serve northwestern Minnesota.  Since those media markets are going ad free in 2012, I expect considerable county losses for Obama, including Marshall, Pennington, Red Lake, Polk, Aitkin, Pine, Grant, Stevens, Pope, Lincoln, Murray, Olmsted, and Watonwan.  Polls are all over the place regarding the strength of Obama's advantage, so it's entirely possible Obama could cede even more territory here.

Mississippi--Higher black turnout flipped a handful of counties to Obama in 2008 that I expect will be flipped back to Romney in 2012, including Oktibbeha, Panola, Copiah, Chickasaw, and Pike.  Yazoo County is on the bubble, but I think there's a demographic shift there that will keep it blue.

Missouri--Obama managed to keep Missouri incredibly close in 2008 winning only nine of its 115 counties.  It won't be at all close this year and I expect he'll lose at least four of those nine counties in 2012, including Buchanan, Jefferson, Washington, and Iron.   Even Boone and Ste. Genevieve Counties are no sure thing to stay blue this year with the toxic political environment in Missouri.

Montana--With a hotly contested Senate race liable to keep turnout high, Obama probably won't surrender as much turf as he ordinarily would after his impressive 48% showing in Montana in 2008, but I still see losses in Cascade, Gallatin, Lake, Lewis and Clark, and Rosebud Counties.

Nebraska--Polls are very confusing in Nebraska this year, with some showing Romney barely winning by double digits, but if that was the case Obama would be more strongly contesting the Omaha-centric Congressional district.  My guess is three of the four Nebraska counties that went Obama in 2008 flip to Romney this year, including Lancaster, Saline, and Douglas.

Nevada--Early voting figures show a closer race than the double-digit Obama blowout in 2008, which is no surprise.  I'll guess Obama narrowly hangs onto Washoe County but definitely loses Carson City.

New Hampshire--Polling indicates a New Hampshire result comparable to 2004, so Obama will definitely not be winning all 10 of the state's counties this year.  Rockingham, Belknap, and probably Hillsborough seem likely to go, but I'll guess Obama narrowly hangs onto Carroll County along with the six Kerry counties.

New Jersey--Hurricane Sandy is obviously a huge wild card here, but only two of the 14 Obama counties from 2008 seem vulnerable.  I'm narrowly betting on Somerset County to flip back to Romney but expect Salem County definitely does.

New Mexico--Obama's big 2008 win seems close to a new Democratic baseline in New Mexico with its soaring Latino population, but I still think Luna and Hidalgo Counties flip to Romney with Los Alamos and Valencia Counties hanging on for Obama.

New York--Again, Hurricane Sandy is a wild card here in the greater New York City area and could potentially prove game-changing in Nassau and Suffolk Counties on Long Island, even though I believe Obama will hang on in both. The county losses are more likely to occur upstate, including Chautauqua, Madison, Niagara, Orange, Oswego, Seneca, Warren and Washington Counties.

North Carolina--The consensus opinion is that Obama's most unlikely state in 2008 will revert to form and go Romney in 2012, but the magnitude of the losses could well be less than expected at the county level.  I still expect flips to Romney in Bladen, Caswell, Gates, Hyde, Martin, Richmond, and Watauga Counties.

North Dakota--This state, contested by Obama in 2008 but most certainly not in 2012, is hard to measure the extent to which it will revert to form, particularly in the eastern part of the state, but I'll go with Eddy, Grand Forks, Mountrail, Nelson, Towner, and Traill Counties flipping to Romney, six of the 13 Obama counties last time.

Ohio--A very fascinating situation here in the mostly hotly contested state in the country, where I expect fierce polarization between north and south to form, with coal counties like Jefferson and Belmont almost certainly gone, and probably Monroe as well.  I'm leaning towards Cincinnati's Hamilton County flipping back to Romney as well even as Dayton's Montgomery County narrowly hangs on for Obama.  On the other hand, I could actually see Obama picking up a few counties in northern Ohio that went McCain four years ago where the auto bailout is popular.  Seneca and Huron Counties seem like the best bets.

Oklahoma--Obama won no counties in 2008 and won't win any again in 2012.

Oregon--Polls indicate a steep drop for Obama in Oregon this year, far more than I would have expected.  A few of the 2008 Obama counties will be casualties, including Jackson, Marion, and probably Wasco Counties.

Pennsylvania--Most of the southwestern Pennsylvania coal counties already flipped to McCain four years ago, but Cambria and Carbon will be the last two coal counties to make the red transition this year.  Beyond that, I suspect Elk County to flip and Obama to barely hang on to a number of the new counties he picked up in 2008, including Berks, Chester, and Dauphin Counties, as well as Bucks County which has long been Democratic turf but will be very narrowly Democratic turf this year.

Rhode Island--Obama wins all five counties like last time.

South Carolina--Most of the counties Obama picked up in 2008 due to high black turnout will revert back to red, including Barnwell, Calhoun, and Colleton Counties.  I'm on the fence about McCormick County but expect Obama squeaks it out.

South Dakota--Without any high-profile races at the top of the ticket, South Dakota is a wild card, particularly its Indian Reservation counties where turnout is dramatically lower when there are no tight races.  I'll predict losses in Brown, Brookings, Lake, Miner, Minnehaha, and Moody Counties, reverting to something close to the 2004 map.  It should be said, however, that Bennett and Mellette Counties, just outside of Indian reservations, are seeing Native American population growth so impressive that it's only a matter of time until they flip.  Not sure this will be the year it happens though.

Tennessee--Having hemorrhaged a full 30 counties between 2000 and 2008, there's not much left for Democrats to give, but the last two Yellow Dog holdouts for Obama last time--Jackson and Houston Counties--are almost certainly gone this year, and I suspect heavily black Hardeman County in west Tennessee goes too since I'm expecting low turnout.

Texas--The demographics are changing quickly in many of these Texas counties, so I'm not anticipating too many losses, but I'll bet on Brewster, Jefferson, Kenedy, and Reeves Counties turning this year, with Harris, Bexar, and Val Verde Counties on the bubble tilting Obama.

Utah--The Mormon factor will loom incredible large for Romney here and will almost assuredly flip 2008 Obama counties Grand and Salt Lake.  His very close Olympic association with Summit County (Park City) probably means that will turn red again too despite demographic changes.

Vermont--The wholesale transformation of Vermont to indigo blue tells me that all 14 counties stay with Obama, even though Essex County is the weakest link and would be the one to go if any do.

Virginia--Obama cleaned up in Virginia in 2008 and is sitting on a lot of territory that he won't win again, including the counties of Buckingham, King and Queen, and probably Montgomery, along with the independent cities of Chesapeake, Harrisonburg, Hopewell, Radford, Staunton, and Winchester.  I've gone back and forth on whether I think Obama hangs on to win Loudoun County in northern Virginia, and am now leaning towards him barely hanging onto it.

Washington--Polls are all over the map with Washington state so this is a very tough call, but I'll go with blue-to-red counties that include Clallam, Clark, Klickitat, Skamania, Wahkiakum, and Whitman.  If the most recent Washington poll showing Obama leading by only seven points is right, then even more counties will flip to Romney.

West Virginia--A "full Oklahoma" here.  The remaining seven Obama counties in West Virginia will all flip to Romney in 2012.

Wisconsin--Obama won a mind-blowing 59 of Wisconsin's 72 counties in 2008, his highest number of county wins in the nation.  It will be nothing like that in 2012 as Obama is likely to lose about 10 points of his 2008 support, and with it will go the counties of Barron, Brown, Burnett, Calumet, Chippewa, Clark, Jefferson, Juneau, Kewaunee, Langlade, Manitowoc, Marathon, Marinette, Marquette, Monroe, Oconto, Oneida, Outagamie, Pierce, Racine, Rusk, Sawyer, Shawano, Washburn, Waupaca, Waushara, Winnebago, and Wood all seem like probable losses to me.

Wyoming--Albany County, home of the college town of Laramie, flipped to Obama in 2008.  It will go back to Romney this year.

Overall, a pretty substantial list of losses, leaving Obama with county victories more on par with Gore and Kerry's victories than his own 2008 baseline.  Pretty much everybody expected he'd lose much of the territory he picked up in 2008 but the fact that Obama can lose as much territory as I have listed yet still be favored to win the election really tells the story on how concentrated the Democratic vote has become in America's population centers, more today even that the 1984 Mondale-Reagan blowout.  I don't see this changing either.