Saturday, October 13, 2012

Berry & Bickers: Obama will lose; say hello to President Romney

Berry & Bickers  electoral map forecast

Two professors who have a forecast model with an accuracy rate of 100% for the past 8 presidential elections believe that the next president will be named Mitt Romney.

Michael J. Berry, of the University of Colarado, Denver, and Kenneth N. Bickers, of the University of Colorado, Boulder, developed a forecast model based on various factors at state level, which has been successful in forecasting the votes per state in the past 8 elections:

Berry & Bickers success record

Their forecast for the 2012 election is a Romney win, as follows:

The Berry & Bickers Model shows that voters punish unemployment more harshly under Democrats than under Republicans:

While many forecast models are based on the popular vote, the Electoral College model developed by Bickers and Berry is the only one of its type to include more than one state-level measure of economic conditions.

In addition to state and national unemployment rates, the authors looked at per capita income, which indicates the extent to which people have more or less disposable income. Research shows that these two factors affect the major parties differently: Voters hold Democrats more responsible for unemployment rates while Republicans are held more responsible for per capita income.

Accordingly -- and depending largely on which party is in the White House at the time -- each factor can either help or hurt the major parties disproportionately.

Interesting analysis, and very interesting model.

I bet there a  gazillion progressive mathematicians who are burding the midnight oil to prove – in the next 2 to 3 weeks – that this time round the Berry & Bickers Model is just plain wrong.


  1. It's not over 'til its over, Ligneus.

  2. I like how they have a "100% accuracy rate" yet the chart shows they've been anything but - 100% accuracy means they called every state correctly every single time. Clearly they have not.

    1. 100% accuracy is refering to the outcome of the election, not the amount of states judged correctly.

    2. It is not right to characterize the Bickers study as being "anything but" accurate. the 100% of course is for the overall outcome but pay close attention to the Electoral split. Bickers and Berry are QUITE CLOSE on their splits and they are only off by 20-something electoral votes in each election. In this case even if their error margin of the 20 something electoral votes is maxed out in this election, Romney still is forecast as a clear winner. So its important to pay attention to a study like this that doesnt depend on fickle humans but on immutable economic fact that cannot be swayed by a poll or a newscast. And most importantly, the Berry/Bickers model has in fact been absolutely correct in the outcomes even with the uncertainties in the electoral split.

  3. Nate Silver's model on the other had, still lists Obama as having a 63% chance of winning.. down a lot from the first debate, but still better then these guys. (No poll so far has shown Obama to be in danger of losing Minnesota)

  4. Scott, Silver's model gives a very high 60% plus probability to Obama winning, which troubles me a bit given the catastrophic first debate, and the scrambling of the Dems since then to recalibrate their campaign.

    I would think the chances should be roughly 50-50 right now, with Tuesday's debate the key.

    If Obama loses, game over. If he ends up with a draw, he is in trouble because his Plan A has not worked and he has not sold a Plan B during the debate.

    The only hope for Obama is a crushing defeat of Romney, which - based on his performances on TV without cues - is a very unlikely event.

    Right now I would bet that Romney will squeak into the White House.

  5. Latest news from Realclearpolitics:

    2:21PM EDT October 14. 2012 - The race between President Obama and Mitt Romney is closer than ever, including the measure that will decide it: The Electoral College.

    Obama leads in states with 201 electoral votes, according to state polling averages compiled by the RealClearPolitics website. Romney leads in states with 191 electoral votes.

    It takes 270 electoral votes to win the presidency.

    Two-and-half weeks ago, RealClearPolitics gave Obama a lead of 265-191 in the Electoral College, in part because Ohio was considered a "likely Obama" state.

    Now Ohio is again a tossup.

    The reason: The debate.

    1. Guys... You published the Berry/Bickers data and even with data showing 100% accuracy over the last 8 elections you are quibbling over RealClearPolitics and poll numbers. It is in the Liberal interest to show that there is a chance for Obama to win again to get their base out to vote and polls by the media are used to good effect to do this. Polls are a political tool as you know like anything else.

      Berry and Bickers on the hand have no political agenda. They are objectively reporting data and crunching it using opinion-neutral science. There is no human opinion in their result. THAT is very important. Did you see their scientific paper that illustrates the actual methodology they used? There is a reason it has been 100% accurate, not 92%, not 87.6% but 100%.

      I believe that history will show that Berry and Bickers are once again correct. Look at Ohio now, the "polls" show Romney actually two points ahead of Obama. Typically when the polls show a neck and neck result it typically means that the republicans are actually a little farther ahead because of that suppressed republican respondent result.
      But no matter and I am not trying to start any argument. Just pointing out that polls are useless.

      Recall if you remember, that the Carter/Reagan race showed Carter obliterating Reagan in the polls... And Reagan won with a massive landslide. That was because during that race, democrats were so completely polarized by Reagan who spoke his mind that they used the polls visciously to try to dispirit voters who might vote for Reagan. They wanted them to feel there was NO hope so dont bother. The same technique is being used now in favor of Obama. It is a decades old technique and in this election will fail miserably in my opinion. Obama is truly an ACCIDENTAL PRESIDENT. I said this on other national blogs where I participate and stand by it.
      -Marc Dantonio



Thank you for commenting; come again! Let us reason together ...

Random posts from my blog - please refresh page for more: