Romney/Obama Projection

This is not especially sophisticated, but here’s a quick projection of how the county-by-county breakdown for Romney and Obama might look on election day in Wisconsin (based on the recall election and the 2008 election).  As it happens, I set it up to result in an exact tie.  My hope is that this will be useful for people to follow along on election day…. if one candidate or the other is outperforming the other in a clear majority of the counties, that will might be a good indicator as to how the race is shaping up.

Who do you think is more likely to outperform the totals below?  Are there any counties that jump out at you as having unrealistic results?

COUNTY Obama Romney Romney Edge Obama% Romney%
ADAMS  4787 4723 -64 50.3% 49.7%
ASHLAND  5114 2923 -2191 63.6% 36.4%
BARRON  9588 11662 2073 45.1% 54.9%
BAYFIELD  5625 3707 -1918 60.3% 39.7%
BROWN  54753 65716 10963 45.4% 54.6%
BUFFALO  3044 3526 482 46.3% 53.7%
BURNETT  3454 4582 1129 43.0% 57.0%
CALUMET  10426 15451 5025 40.3% 59.7%
CHIPPEWA  13506 15824 2318 46.0% 54.0%
CLARK  5475 8081 2606 40.4% 59.6%
COLUMBIA  15339 14012 -1328 52.3% 47.7%
CRAWFORD  4123 3448 -676 54.5% 45.5%
DANE  198870 84083 -114787 70.3% 29.7%
DODGE  16982 26708 9725 38.9% 61.1%
DOOR  8311 8645 333 49.0% 51.0%
DOUGLAS  14134 7960 -6174 64.0% 36.0%
DUNN  10040 10063 23 49.9% 50.1%
EAU CLAIRE  27150 23295 -3855 53.8% 46.2%
FLORENCE  937 1595 658 37.0% 63.0%
FOND DU LAC  20169 31949 11780 38.7% 61.3%
FOREST  2080 2311 231 47.4% 52.6%
GRANT  11799 10364 -1435 53.2% 46.8%
GREEN  9941 8429 -1512 54.1% 45.9%
GREEN LAKE  3325 6246 2920 34.7% 65.3%
IOWA  6976 4890 -2087 58.8% 41.2%
IRON  1616 1716 100 48.5% 51.5%
JACKSON  4566 4251 -315 51.8% 48.2%
JEFFERSON  18422 24316 5894 43.1% 56.9%
JUNEAU  5304 5904 599 47.3% 52.7%
KENOSHA  38266 33860 -4406 53.1% 46.9%
KEWAUNEE  4661 6021 1361 43.6% 56.4%
LA CROSSE  32000 25884 -6116 55.3% 44.7%
LAFAYETTE  3865 3824 -42 50.3% 49.7%
LANGLADE  4044 5969 1925 40.4% 59.6%
LINCOLN  6973 7653 680 47.7% 52.3%
MANITOWOC  17591 23578 5987 42.7% 57.3%
MARATHON  29298 37160 7862 44.1% 55.9%
MARINETTE  8725 11159 2434 43.9% 56.1%
MARQUETTE  3479 4325 846 44.6% 55.4%
MENOMINEE  901 219 -682 80.4% 19.6%
MILWAUKEE  294192 163699 -130493 64.2% 35.8%
MONROE  8202 9897 1695 45.3% 54.7%
OCONTO  7891 11026 3135 41.7% 58.3%
ONEIDA  9732 11194 1462 46.5% 53.5%
OUTAGAMIE  40240 48756 8516 45.2% 54.8%
OZAUKEE  17671 37097 19426 32.3% 67.7%
PEPIN  1666 1932 266 46.3% 53.7%
PIERCE  9301 10150 849 47.8% 52.2%
POLK  8806 11980 3174 42.4% 57.6%
PORTAGE  20485 15990 -4495 56.2% 43.8%
PRICE  3621 4204 583 46.3% 53.7%
RACINE  48169 51105 2936 48.5% 51.5%
RICHLAND  4231 4008 -223 51.4% 48.6%
ROCK  43830 30642 -13187 58.9% 41.1%
RUSK  3016 3889 873 43.7% 56.3%
SAUK  16026 14049 -1977 53.3% 46.7%
SAWYER  3951 4582 631 46.3% 53.7%
SHAWANO  7950 11558 3608 40.8% 59.2%
SHEBOYGAN  24726 36171 11445 40.6% 59.4%
ST. CROIX  17358 24457 7099 41.5% 58.5%
TAYLOR  3343 5756 2413 36.7% 63.3%
TREMPEALEAU  6482 6163 -319 51.3% 48.7%
VERNON  7246 6530 -715 52.6% 47.4%
VILAS  5392 8014 2622 40.2% 59.8%
WALWORTH  19383 28870 9487 40.2% 59.8%
WASHBURN  3994 4793 800 45.4% 54.6%
WASHINGTON  21474 55804 34331 27.8% 72.2%
WAUKESHA  73145 167130 93984 30.4% 69.6%
WAUPACA  10308 14676 4368 41.3% 58.7%
WAUSHARA  4874 6822 1948 41.7% 58.3%
WINNEBAGO  40050 43441 3392 48.0% 52.0%
WOOD  17585 19584 1999 47.3% 52.7%
Total 1450000 1450000 0 50.0% 50.0%


  1. Wow. Just wow.

    Remember how clever Ryan was with finding the missing Prosser votes before you discount this outright, folks. He’s got some talent in this area.

  2. Ryan Morgan says:

    All I did here was look at the % of votes Walker got from each county (X% of Walker’s votes came from Milwaukee country, Y% from Waukesha county, etc.) and the same for McCain and take the average of those 2 for each county. Then do the same for Barrett and Obama.

    Turnout in the last couple presidental elections has been about 2.9M (for the 2 major candidates… no one else really matters… sorry Recess Supervisor) 🙂 so assume that turnout persists and solve for how many votes in each county that translates to.

    Looking a little closer, I honestly don’t know if some of these numbers are realistic.

    Can Romney really add 8+% more Waukesha county voters above and beyond what Walker got in the recall? Turnout was so high for the recall, seems like a tall order.

    By the same token, can Obama really come within 3% of his total votes from Dane county in 2008, in the midst of Obama-mania? He’s not quite as popular as he was then (understatment), even in Madison.

    Both of those will be tough, but in the past couple elections, betting on higher than expected turnout in Dane and Waukesha was definitely the right side to err on. We’ll see if it continues. If one and does an the other doesn’t, that will probably determine who our president is.

  3. Ryan Morgan says:

    Also, sorry Cindy for plopping a big 74×6 spreadsheet on your front page. 🙂

  4. J. Strupp says:

    Holy crap.

    My head just exploded.

  5. Recount Wisconsin, 2012. Paging Jim Baker.

  6. Just remember, we’re likely not going to know who won on Wed. morning b/c of all the provisional ballots in Ohio. Expect to be hearing about this crap for a while yet.

  7. If Romney wins WI (possible), VA (likely), CO (fairly likely), and 1 of NH or IA or NV (likely he will win at least 1), he doesn’t need OH.

  8. What do you make of this, Ryan?

    It could be desperation (i.e. a sign that internal polls show Ohio, WI, and IA aren’t budging). I see Rasmussen now shows a tie nationally.

    Regardless, I think a “pre-mortem” of sorts is on topic. Entering 2012, “Generic Republican” had pretty good odds of defeating Obama and taking enough of the swing Senate races to turn the Senate. The Senate was practically handed to the GOP on a silver platter.

    Now the presidential race is going down to the wire (with Obama having a critical electoral college advantage), and the Senate looking like it may be a net 0 change (and perhaps even a Democratic pickup). So the Establishment got their presidential candidate, and he’s at best a 40% chance of winning Tuesday. And the social conservatives* got their candidates in the Senate races (Mourdock, Akin, etc.) who cost the GOP sure wins. The status quo favors Obama and the Democrats when we reach the fiscal cliff (will John Boehner REALLY risk his Speakership by holding up tax cuts for the middle class in order to push through a complete extension of the current rates?). It’s a mess of the GOP’s own making. The question is how they fix it in 2014.

    *I hate to call them Tea Party since the Tea Party had its origins in economic issues.

  9. I’m skeptical. If Romney were up by 1 in Wisconsin, then he’d be campaigning there, Iowa, and New Hampshire, rather than trying to get 20 in a long-shot bid in PA. Heck, if he’s really up in CO and VA, then WI and NH would be enough.

    We’ll find out Tuesday night, I guess. However, the raw data from the polls has been pretty consistent. Even the maligned Gallup’s registered voter totals were in line with the others. The difference is in how each side is interpreting the results (i.e. the likely voter screens). The 2008 electorate won’t show up (nor will the 2010), but if something close enough to it does, Obama wins. If something close to the 2004 electorate shows up, Romney can win if enough stars align.

    Wisconsin has truly become a swing state. How a state can give Walker a bigger mandate than he got in 2010 and just 5 months later even think about putting Baldwin in is quite incredible.

  10. Personally, I think the lesson from the 2010 recall was not a win for Walker, rather an over reach by the public unions to force a recall instead of waiting for the next election cycle.
    WI is now outside the margin for error (538 projects a win by 4.4%) and I think Baldwin will lose unless BO can win by more than 4% .
    Question for Ryan how has the demographic profile of WI changed since 2000.
    The battle for the future of the GOP will begin shortly after 11/6, this election shouldn’t have been close.

  11. I think Romney’s hope is that Sandy (which took away early voting days) might lead to lower turnout in Philly, which makes the state much more winnable.

    Romney was here this morning and Ryan had 3 stops in Wisconsin, so I think they are fighting hard here.

    Bottom line: Romney has to win 1 of OH, WI, and PA. OH is the closest, but WI and PA seem doable too, so why not contest all 3 hard and see what happens?

  12. Demographics in WI really haven’t changed much. A little older…. 50-59 year olds are now the biggest group. The city of Milwaukee shrunk and the suburbs grew. Race is mostly unchanged… a few more hispanics.

    Your tone on the “battle for the future of the GOP” is strange. It is very difficult to defeat an incumbent president, especially one who won by over 7% when they were elected. Romney-Ryan is a strong ticket. Along with Rubio, Christie, Susana Martinez, etc. the future of the GOP is strong.

  13. Romney might not win Wisconsin, but I’m pretty dang confident that it is going to be decided by 2 points or less, one way or the other. The exercise above shows that both sides have tough but plausible paths to get close to the 1.5 million votes they need to win.

  14. Ryan, Ronald Reagan and Bill Clinton beat incumbents in economic conditions that were better than what we have now. The difference is that from Day 1, Obama’s campaign knew the strategy. They did it in 2008 in the primary. Hillary Clinton actually “won” the popular vote in the primaries. Obama knew how to run up the delegate score through targeted campaigning and GOTV. He’s done the same with the electoral vote in 2012.

    More bad news for Romney. Rasmussen now reports Ohio is a tie. Romney had been up by 2 points before Sandy.

    It could be statistical noise, but all the other polls show Obama up.

    I think it’s time to admit that Romney lost a perfectly winnable election, and the GOP threw away a perfectly attainable Senate majority.

  15. “I think it’s time to admit that Romney lost a perfectly winnable election, and the GOP threw away a perfectly attainable Senate majority.”

    Why the heck would we do that before the votes are counted? Even a liberal who doesn’t make any adjustment for the consistently Democratic-favoring samples in the polls like Nate Silver gives Wisconsin gives Romney about a 20% chance to win (Intrade shows in the mid 30% range). Either of those better than the Packers’ chances of winning the Super Bowl. Should we call the Packers’ season over as well then?

    I’ll bet you anything Romney wins FL and VA. He should win CO too. It will come down to if he can pull of one upset, in OH, WI, or maybe PA. Given there have been polls that have showed OH and WI tied, it wouldn’t really be that shocking if he pulled it off, especially given how fired up conservatives in both states seem to be.

  16. You are probably right on the Senate. Barring an upset that I don’t forsee, there will be 4 close races… WI, VA, MT, and IN. Republicans need to win all 4 to get to 50 Senators. But given Romney will win and have coat-tails in at least 3 of those states and the 4th features a popular multiple term ex-governor against a fringe lefty, there’s not one of them that you’d say is not winnable.

  17. Going back to my grid above, where I think R/R will outperform is in Paul Ryan’s district… there’s a lot of people in Rock, Walworth, Racine, and Kenosha counties that regularly vote for Ryan but voted for Obama in 2008. Since these people have actually checked the Paul Ryan box before (and might like the idea of a hometown guy as VP of the country), I’d expect more of those voters to swing to the Republicans. So if Romney can stay “on schedule” in the other counties, there’s probably some boost for him in SE Wisconsin that could put Romney over the top.

    Of course, the big ones (Milwaukee, Dane, Waukesha) are so powerful that they will drive the results more than anything else….

  18. Now THIS is interesting. Early voter turnout is down over 15% in Obama’s best county in Ohio relative to this time in 2008.

    At a certain point, don’t actual voting statistics have to matter more than polls?

  19. Another helpful resource for election night… a map of when the polls close

  20. “Turnout in the last couple presidental elections has been about 2.9M (for the 2 major candidates… no one else really matters… sorry Recess Supervisor)”

    HAHA! You know what, I stopped counting when I moved to South Dakota. Anyone who says “one vote can make a difference” doesn’t see the margins by which Republicans win out here.

    I’ll be the first person to back you up on your assertion. 🙂

    While the “bad model” meme is popular among conservatives right now (and understandably so – if the models are right this race is over) what’s interesting is that a lot of these polls with which the GOP takes exception are hitting the ideological ID labels dead on (conservative-moderate-liberal). Those identifiers have been rock solid for the better part of 20 years running, even through the wave elections in 2006/08/10. So in order to hit the party ID label, the polls right now would have to oversample conservatives even vis-a-vis the numbers obtained in 2010 exit polls in swing states. And of course, exit polls don’t measure likely voters or registered voters – they measure actual voters. Hard to find a better sample than people who actually show up to vote. Do conservatives have a bigger turnout advantage now than in 2010? Gosh, I find that scenario a hard one to comprehend, especially considering what House and Senate projections look like this year compared to two years ago.

    Anyway, I’m with Nate Silver, insofar as Romney supporters basically have to have church-like faith that all of these pollsters have somehow missed the mark. When one aggregates all these polls, Romney’s outside the MOE in Ohio, Wisconsin, Minnesota, and Pennsylvania and he’s inside the MOE in states he absolutely has to win (Virginia, North Carolina, Florida).

    Possible that they’re all wrong? Yes. Probable that they’re all wrong? No.

  21. Obama’s up by 6 in OH and 2 in FL in the latest WSJ poll. It is looking more and more like a landslide. It turns out Sandy was the October Surprise since it made him look good. Just imagine if Bush were president. Why, NYC probably would have had to cancel tomorrow’s marathon, and people might be waiting hours for gasoline.

  22. “Possible that they’re all wrong? Yes. Probable that they’re all wrong? No.”

    This misses the point. The national polls show a tie race (Obama is up .1% in the latest RCP average). the state polls results are saying Obama is really up 4 or 5 points nationwide (he has to be if Obama is winning in lean-R states like OH and VA).

    So we have a bi-modal distribution. At least 1 of the two is wrong. It’s either a very close race or a clear Obama win. There’s support for both views. Personally, when you look around at things like this, I think it is clear which one is correct if you use common sense. But we’ll find out soon enough.

  23. Understood. Personally, I don’t see much value in national polls that are surveying people in 40 states whose outcomes are beyond question to predict behavior in the ten states whose outcomes might be in question. I would just ask the people in those states, since they’re my target anyway. But I know where you’re coming from.

    Beyond that, I find it interesting that you adhere to a perception of Ohio and Virginia as leaning Republican that not only isn’t supported by polling but isn’t supported by their recent electoral histories in statewide races. Ohio and Virginia long ago started their move from lean R towards toss-up status. Same goes for Florida.

    I’d spot you that if we were talking about 1992. But a lot has changed in those states since then, and none of it in a particularly Republican direction.

  24. Randy in Richmond says:

    KPOM, I know you are the nullifying cynic of negativism on this site but you overlooked Gravis, CNN, and Mason Dixon who all have Romney up in Florida–in the same polling period.

    The latest poll taken in Ohio, by Rasmussen, is a dead heat tie.

    In Wisconsin, Rasmussen, a tie. That We Ask America is an online or robocall polling.

    In Virginia, Romney up a tad on RCP.

    I live in Virginia. You could not be more wrong.

    In 2009 Republican Bob McDonnell was elected Governor with 59% of the vote. In 2010 Republicans picked up 3 seats in the House of Representatives.

    In 2011 in the Virginia Senate Republicans picked up 2 important seats to gain the majority. In the Virginia House of Delegates Republicans picked up 9 seats to maintain a larger majority.

    Our current Senate race is a pickum between 2 former governors.

  25. @Ryan, clearly the likely voter screens are the difference, and a lot of pollsters will be proven wrong (we just don’t know which ones yet). However, the only way Romney wins is if the best assumptions about the electorate prevail, and then it is just barely. Right now I don’t see anything to doubt the projections of an Obama victory. I just don’t see all the stars aligning the way they would need to. Romney going into PA sounds like a sign of desperation. His internal polls must be trending toward the national polls. There is really no other way to interpret it. Why Karl Rove, Michael Barone et. al are going out on a limb predicting a Romney blowout is beyond me. The facts are clear. If anyone is going to run away with this election, it’s the president.

    I would not be surprised at all if Obama wins CO, OH, FL, VA, WI and IA, ending up with 332 electoral votes.

  26. Randy in Richmond says:

    Looking beyond the election regardless of the winner, it is shaping up that a large part of President Obama’s legacy will be established by what happened prior to, during, and after the Benghazi attack. Sure, it’s been ignored by the MSM and the can kicked down the road beyond the election by the administration but as Obama’s long time pastor Rev. Jeramiah Wright famously said, ‘the chickens are coming home to roost’. This is not going away. Recent findings and discoveries are all bad for the Obama administration. I’ve talked about the inebtness of this administation for years and this will become the whipped cream with cherry on top of what’s gone before. Time will bear this out.

  27. there’s plenty of ways to interpret Romney moving into pennsylvania. The most likely is that Romney thinks even though pa is normally a few points to the left of oh, but the combination of Obama’s bazillion attack ads over the summer along with the auto bailout makes oh a better state for Obama than normal while pa has seen huge increases in unemployment recently which makes it a worse state for Obama. Also, Obama doesn’t Have the turnout machine in Philadelphia that he does in Cleveland… Didn’t think he’d need it…..(some have suggested not contesting pa until the last week was an intentional move on Romney’s part for precisely this reason) So since Romney probably only needs one or the other, why not hedge Your bets and try for both… It’s not like Romney has a shortage of money at this point.

  28. Pew Research is showing Obama up by 3, and critically, at the 50% mark. Their previous poll had this a tie.

    This could be a sign that Nate Silver was right, and the state polls were more accurate than the national polls. Rasmussen has had the race a tie for 3 days since he shifted to a sample of 1000. His final poll is tomorrow. It will be interesting to see if his surveys today showed a shift toward Obama.

  29. BrkfldDad says:

    Ramussen is out. Romney 49, Obama 48.

    CNN has it tied, but Democrats are oversampled by +11, that’s not gonna happen.

    And Pew is Dems +6