Battlegrounds: The way to win

Battleground states, by definition, have lots different groups and regions that balance each other out to an almost-even split, especially when an election that looks as close as this one does. Small shifts in any can tip a whole state.

State-by-state, we'll explain the ways you'll be able to tell whether President Obama or Mitt Romney is doing well as the returns come in on election night -- the key regions, counties and groups to watch, and why.

Breakdown: Ohio

Dickerson on ways each candidate can win

It is no accident this is the Obama campaign's firewall and a key state in this race. Filled with the kind of working-class, middle-class voters that the Obama campaign is targeting, if Mr. Obama cannot hold on to the faster-growing boom states like Virginia and Colorado, then he has to make a stand in places like Ohio and Wisconsin. Ohio may not be a microcosm of the U.S. any longer but it is representative of America's industrial heartland - and a mix of large cities and suburban and small towns adds up to a closely divided state.

Signs Mr. Obama is doing well:

  • First and foremost, few things matter to Mr. Obama's chances more than the raw vote numbers around Cleveland. This is step one in the Democratic strategy for winning Ohio. Mr. Obama would like to take around 250,000 more votes out of the county than Romney does, and if he doesn't, it gets harder to make up that margin elsewhere in the state.

  • He's winning the Columbus area. This is something of a "swing" region. (And by region, here as elsewhere in these documents we mean the group of counties clustered in an area. Using regions allows us to get a better handle on overall trends than looking only at one single county.)

  • Nearby, he'd like to get above 52 percent in eastern Ohio, including the areas around Akron and Canton - these are usually Democratic areas where working class voters hold sway, and a key part of his coalition. Often, Democrats who only break even here can lose statewide.

  • He keeps it close in Northwest Ohio and gets big margins around Toledo. He's got to take 60,000 to 70,000 vote margins out of Lucas County (Toledo) and that may speak to how well the auto bailout played for him.

  • He's winning or even among working-class white voters.

  • Young people are turning out (they were 17 percent in 2008) and backing him with over 60 percent of the vote. For Mr. Obama in 2008 it was as much about vote swing as turnout among under-30 voters, he did six points better than Kerry. He has to again.

  • He's close - even if not winning - in the white vote. If he gets 45 to 47 percent he stands in good shape, repeating '08 levels. Lower and he'll have to make it up with turnout elsewhere.

  • African-American turnout is 10 to 11 percent of voters. Lower and he faces a problem.

Signs Romney is doing well:

  • He's winning the Columbus area. It's a swing part of a swing state; Obama won it, Bush won it.

  • He's winning college-educated voters back (Republicans lost them in 2008). Upper-income and college voters in Ohio don't tend to be as liberal in Ohio as they are on the coasts.

  • He's got solid majorities in suburbs around Cincinnati, especially Butler county (where Romney will aim for at least 60 percent or more) and the counties nearby. This is usually Republican territory as has been for a long time. Obama cut into it in 2008, but will have a hard time repeating that.

  • He's getting at or above 56 percent in small town and rural areas throughout the state. Obama held down typical Republican margins in such places last time. Romney should be able to run them back up again.

  • He's winning or even among white working-class voters earning under $50,000, which would be cutting Obama's margins slightly from '08. If Romney can win or hold even here it would be a sign his campaign has countered negative views of Romney among working class groups.

Next page: Florida

Breakdown: Florida

The largest of the competitive states, Florida is a quadrennial battleground for an assortment of reasons: a boom that filled it with a diverse array of people; because its major cities that aren't as reliably Democratic as the major cities of the northeast, and wealthy counties that aren't always as conservative as wealthy counties elsewhere...and because it has three district regions that balance each other out: the more conservative north and panhandle, the more Democratic south around Miami, and the middle, the famous I-4 corridor. Each plays an important role.

Here's how to know who is doing well - and who is falling short - as results come in:

Signs Mr. Obama is doing well:

  • He's doing very well with young voters under 30. You don't always think of young voters when you think of Florida, but they were a healthy 15 percent in 2008's electorate, and Obama won them two to one. By contrast if turnout drops or a tough economy (they're looking for their first jobs) and tough housing sector (buying that first house) turns them away, Obama could have difficulty.

  • He wins suburban vote. Yes, it is a broad definition. But he swung suburbs statewide (winning 53 to 45 percent) from Bush in 2004, and after the housing crash, they will be as good a test of his ability to win Florida as any.

  • He holds his own with seniors, perhaps drawing even. He lost seniors in 2008 but made up for it being even or better with all the other age groups. If he drops under 40 percent though, that's a bad sign for him.

  • He's getting at least in the high-50-percent range (57, 58 percent) in the Hispanic vote. Remember that the Hispanic vote in Florida isn't mainly Cuban-American anymore. Obama swung it a whopping 14 points from 2004 levels when Hispanics backed Bush. Obama will need at least that number, maybe better, now.

  • African-American turnout is high again: it was 13 percent in 2008 and it'll need to be at that level again.

Signs Romney is doing well:

  • He does well with seniors. Romney is looking to get 53 to 55 percent or better. Under that could be a problem. More than one-fifth of the electorate, there is some question of how Medicare will play. Most didn't back Obama in 2008.

  • He wins boomers. The single biggest age-group shift toward Obama in 2008 was among those 45-64. Romney will try to win them back. Polls indicate he can. This is a swing group now in part because of the economy, and in part because of the Medicare issue as some of the older voters in their ranks approach retirement.

  • He shifts the Hispanic vote toward even. That won't be easy, but it would be a very big boost to his chances. If he can do this among Latino men - who swung even more toward Obama than Latino women in 2008 - maybe he can.

  • He holds down Obama's margins in Miami-Dade area. The area went 63 percent for Obama. There are so many votes there, if that moved toward or under 60 percent Romney would be in good shape statewide.

  • He gets near 60 percent among upper-income ($100,000+) voters. There are plenty in Florida - one-fourth of voters - and they shifted toward Obama in 2008. He needs to shift them back.

  • He narrowly wins or gets even among moderates. This may seem obvious. But they went heavily for Obama and Kerry won them as well. Inroads here would be very positive for Romney.

Florida geography notes:

The importance of vote margins in Miami-Dade cannot be overstressed. Obama wants 130,000 to 140,000 votes more than Romney here. Less than that and he's in some trouble.

We hear a lot about the I-4 corridor but let's split it into two parts, the Tampa and Orlando/Central coast area, as we do for making estimates and in the exit polls. Tampa is a classic swing area: it went 54-45 for Bush, then 51-49 for Mr. Obama. It's population has changed with the influx of retirees, middle-class families, and Latinos from Puerto Rico and other parts of Central America as their heritage. It is around 16 to 17 percent of all voters.

The Orlando area, stretching on to the Central Atlantic coast, also swung to Obama from Bush, and also showed big population changes over the last decade. But the tough economy and the housing bust will surely have this in play again. If either Romney or Obama wins both these regions they'll likely win the state. It's a little larger than Tampa area and worth a bit more, around one-fifth of all voters.

The South (Miami) and the North (Panhandle and Jacksonville) cancel each other out to some extent. Republicans look to get 60 percent in the latter, Democrats need 60 percent in the former. Watch whether Obama wins the Miami area by anything near or under 60 percent. If he does, Romney is likely to win.

Next page: Virginia

Breakdown: Virginia

Virginia has changed from a southern-profile, reliably-Republican state into a competitive one thanks to expanding Northern suburbs of more-moderate, upper-income voters who'd trended toward Democrats -- but can and will vote Republican, too. Obama captured these along with large African-American turnout in 2008 but often overlooked was his relatively good performance, for a Democrat, in smaller towns and rural areas. That'll be hard for him to repeat, and this will invariably be closer. Those northern suburbs and exurbs will probably tell the tale.

Signs Mr. Obama is doing well:

  • He's winning at least 65 percent to perhaps 70 percent in the northern state, DC-area suburbs containing Fairfax county, Arlington and Alexandria. Winning isn't enough, he needs big wins. Less and he's got an uphill battle. Simple as that.

  • He wins or is close to even (around 49 percent or better) in the outer-DC exurbs of Prince William, Loudoun and nearby counties. This has been swing territory in recent elections.

  • In the same vein, he needs to run about even with college graduates and the affluent. They're a big part of this region, and he'd like to get mid-40 percent of upper-income voters to duplicate his 2008 success.

  • He wins the Richmond area. Another swing region where he did well in 2008, but Republicans who win statewide win it.

  • He's getting about 53 percent in the Tidewater region around Norfolk. He needs a win here to offset what will likely be worse performance in the western region than he had in 2008.

  • He gets into the mid-50 percent range of the women's vote. The campaign has targeted women, as everyone knows, and he'll need large margins to make up for what may be larger losses among men. Which leads to...

  • He hangs on to at least some (mid-40s) of the men's vote. Men voted for him in 2008, but aren't in his camp this time based on recent polls. Keeping that loss manageable, in the mid-forties, might be enough.

Signs Romney is doing well:

  • Same rule in the northern D.C.-area suburbs. If gets even in the high-40s or close to 50 percent he's on his way.

  • He wins the outer-D.C. exurbs of Prince William, Loudoun and nearby counties, with low-to mid-50 percent of the vote. He needs this region, even if marginally.

  • He wins or is very close in the Richmond area. Neither candidate can really afford to lose this area by a wide margin.

  • He wins handily among the upper-income voters, especially those in the northern part of the state. A return to Bush 2004 levels among $100,000+ voters, which would put him into the high-50 percent range, could be crucial.

  • He wins boomers and younger adults age 30-49. These age groups showed the biggest swing to Mr. Obama in 2008 but are back in play.

  • He runs up big margins in the men's vote; the opposite effect of the keys for Mr. Obama, above. White men, especially, moved toward Mr. Obama by double-digits in 2008, and are poised to move back to the Republicans; that could offset any lead Mr. Obama has among women.

  • He gets back to 60 percent-plus margins in the small towns and rural areas. Like with North Carolina nearby, these were the kinds of places McCain fell well short of Bush's marks in 2008 - nearly splitting them with Mr. Obama - but Romney is positioned to do much better, and needs to.

Next page: Colorado

Breakdown: Colorado

Mr. Obama's Colorado win in 2008 did more than just flip a longtime Republican state. It looked like part of a trend that might reshape the Democratic coalition; one that took advantage of the demographic changes in the booming west, pulling together upper-income and college-educated voters whose ranks had grown in Colorado, younger people (Mr. Obama won all groups under 65) along with middle-class whites and Hispanics.

In 2010, the Democrats' Senate win in a Republican wave year gave them continued hope that this coalition was durable... but today that coalition looks fragile at best, undermined by a weak economy and doubts among a lot of those suburbanites who backed him in 2008.

Geography notes:

The population is mostly around Denver, so that's where the action is, but, as elsewhere, this story is about margins, not swing.

Democrats need big wins in liberal-leaning, affluent Boulder county: Mr. Obama needs upwards of 70 percent of the vote and a margin of about 80,000. Similarly, in Denver, he needs around 75 percent and around 140,000  to 150,000 vote margin. Less than that is a good sign for Romney.

Adams, Arapahoe and Jefferson counties, suburbs east and south of Denver, may tell the tale. They were the closest 2008 counties in the region. Mr. Obama won the area with 55 percent in 2008. Bush beat Kerry in the region 51 to 48 percent. If Romney leads in the area, he'll probably win.

Don't forget the "front range" out around Pueblo and El Paso Counties (Colorado Springs). Romney will win the area but he's aiming closer to 60 percent or above.

Here's what to watch as you see the returns:

Signs Mr. Obama is doing well:

  • He's holding the gains he made in 2008 among college-educated and higher-income voters. This is part of the suburban boom story of Colorado:

  • Mr. Obama gained 12 points over Bush's margin among those earning $100,000+, and they were 1/3 of the Colorado electorate. He would like to keep at least some of that edge, but it is in jeopardy, possibly on tax and economic issues.

  • Mr. Obama needs at least two thirds of the Hispanic vote, which he should be able to get, and it is at least 13 percent of voters. He may not be able to withstand a turnout drop with this group.

Signs Romney is doing well:

  • The size of the over-65 vote is closer to 14 percent than 12 percent. This was the only age group McCain won in Colorado in 2008; they break for Romney today. If their turnout increases that's a good sign for Republicans.

  • Romney manages to get over 40 percent of the Hispanic vote. That would exceed McCain's number, though polls suggest it won't be easy to do.

  • College graduates are going for Romney. They have been in recent polls. Mr. Obama won them in 2008, which was a six-point gain over Bush in 2004. They're a swing group now, and make up more than half the Colorado electorate. Upper-income ($50,000+) white voters are also a marginal swing group.

Next page: Wisconsin

Battleground: Wisconsin

Before 2008 Wisconsin had been one of the closest Presidential states each cycle, even though Democrats usually managed to squeak out wins. Then Mr. Obama won it comfortably, bolstered by younger voters as well as boomers, solid wins among white voters - including, importantly, working-class whites. Now this, as much as any state, is a test between two formidable turnout machines.

Signs Mr. Obama is doing well:

  • He wins in the Green Bay area. Mr. Obama won it in 2008 and Kerry survived losing it narrowly. A lot of swing votes here.

  • He's got low- to mid-60 percent in Milwaukee. This is where any Democrat needs to run up a margin of over 150,000 votes in the state's largest city.

  • He's got at least 60 percent in the region around Madison. This is solidly Democratic territory, including the area around the university, but again he needs to run up the score.

  • Turnout is up in Democratic areas like Milwaukee and Madison. It's an obvious point, perhaps, but there's been so much political activity in the state in recent years, one question will be how well the organizing on both sides is working, or whether any fatigue has set in. A very competitive Senate race may help as well.

  • He gets at least 1/3 of the vote in the Milwaukee suburbs. That doesn't sound like much, but it's important to at least make a showing in this heavily Republican area.

  • He holds about even with the white vote. He won it easily in 2008, but this year a closer election means breaking even could be enough. John Kerry lost it narrowly and won the state narrowly.

Signs Romney is doing well:

  • He wins in the Green Bay area. It's important for Republicans when they win Wisconsin, and it is a winnable region for Romney. That's especially true if his margins in heavily-Republican suburban Milwaukee aren't that strong.

  • He runs up big margins in the suburbs of Milwaukee. Watch Waukesha county, which McCain won with over 60 percent, and Romney should exceed that.
  • The small cities and rural counties, which together comprise a third of the state, go for him. Unlike a lot of states, these are not reliably Republican in Wisconsin. Obama won these kinds of areas in '08 but Romney can certainly win them, as Bush did.

  • He wins the white vote. Doing so would force the Democrats to really rely heavily on base turnout to win.

  • Young (under 30) vote is down, under 20 percent, or it moves toward Romney, or both. Last time under-30s were 22 percent of voters and 2/3 voted for Obama.