Who will win the presidential race?

I've made my first stab at a prediction for the electoral college outcome for the US Presidential race, 2016. I use a roughly similar methodology as I did to accurately predict most of the Democratic primaries. However, since primaries are different from a general, the methodology had to be adapted.

For the primaries, I eventually used this methodology. I used results form prior primaries to predict voter behavior by ethnicity, in order to predict final behavior. That worked because primaries are done a few states at a time, and because all the people being modeled were Democrats.

It turns out that white people vary a lot across the country with how many per state are assholes. I think there is some variation among Hispanics as well, but African Americans are pretty consistent. So, here, I combined ethnicity with a "Romney Index" indicating how many people in a given state voted for Romney against Obama.
----LATEST PREDICTION HERE CLICK HERE----

I then put down the poll numbers, the averages of the last several polls, from RCP, where available. I then ranked the results to knock out states with no polls. I then took out the middle, which included swing states, close states, etc. to use only the 23 most distinct states for which there were data to produce a multi variable regression model using "white", "black", "hispanic", and "romney_index" as independent variables. The dependent variable was the poll value. In future iterations, that is what will change. I'll do a more refined version of that.

I then applied this formula to predict the breakdown between Clinton and Trump in the other ca. half of the states that are more ambiguous.

The multiple R-squared for this model was 0.952, so that's great. But, I was using only the values at the extreme, so I violated the law of homoscedasticity. But I don't care about no stinking homoscedasticity, because I have only one data set, am predicting only one election, and I am basically using the regression model as a fancy fill in the blank formula. The fact that the R-squared is so high is great, were it low, I'd be in trouble, but its actual value is not important.

I then took all the states where Trump gets over 50% of the vote and gave them to him. I then gave almost all the other states to Clinton, but I left out a few that were very close, to leave them as unknown. Even if all those unknowns go to Trump, however, the outcome is the same: Clinton wins. Trump loses.

I'll refine and revise again with more care given to the various parts of the model. I'd love to do this poll free, but not sure if that is possible.

screen-shot-2016-10-10-at-9-41-59-pm
The final output data are spewed onto 270 to win.

NEXT PREDICTION

More like this

Above is my latest electoral college projection. This uses the technique previously described. However, instead of using RCP averages for all polled states and then using extreme (non-tossup) states to develop the regression model, this method uses only polling from states with one or more recent…
As you know, I’ve been running a model to predict the outcomes of upcoming Democratic Primary contests. The model has change over time, as described below, but has always been pretty accurate. Here, I present the final, last, ultimate version of the model, covering the final contests coming up in…
Almost exactly 50% of the votes have been cast in the Democratic Party primary and caucus process. I've been updating a model to predict primary and caucus results all along, and the model has done fairly well. The most recent update, however, was a bit off. That update involved separating states…
I post some data analysis over at my other weblog. For example, today I looked at the relationship between food stamp usage and unemployment. The Census makes a lot of county-level data available, though it's often slapdash and disorganized. But using R I've constructed many data sets including…

Interestingly, RCP has Clinton at 260/Trump 165, and 538 has it as Clinton 333/Trump 204

My model killed 538 during the primary. This version is different in some ways but essentially the same idea. Before election day, I'll put out predictions of the percent distribution of popular votes between the two main candidates.

My go-to site for these things is Sam Wang's site. He shows Clinton 333, Trump 205 as the expected outcome. The differences between your map and his: All three of your toss-up states are some shade of blue on his map (FL is Likely Democratic, NC and OH are Lean Democratic), while IA is Likely Republican on his map (AZ, which he shows as Leans Republican, is more likely to turn blue at this point). Wang's numbers are based on medians of state poll aggregates, which minimizes his sensitivity to outlier polls.

Yes, Hispanics are a heterogeneous lot. Cubans have historically voted Republican (e.g., Mario Rubio and Ted Cruz). Most of the South Americans in the US, IME, are expat elites, and they also have historically leaned Republican. Mexicans, Central Americans, and people from Caribbean Basin countries other than Cuba are more likely to vote for Democrats. This year Mexicans are especially motivated to vote, thanks to Trump's comments, and that may make the difference in Arizona. And softening Cuban support for Republicans is likely to keep Florida blue. Those are the only two states where Hispanics are likely to affect the outcome.

By Eric Lund (not verified) on 11 Oct 2016 #permalink

My tossups are all blue as well, but I left the three weakest/closest blue states off since they weren't needed to give one candidate a win.

I know everyone is saying IA is likely republicans. Polls suggest this. But while my model uses polling dat to calibrate, it is not based on polling data. In so doing, this I beat the pants off 538 during the primaries, so I'm sticking to that this time too.

I have not broken down hispanics. I think the easiest way to do that is to separate out Cubans where possible.

My model at this time does not consider differential motivation or turnout. I probably won't. That, I think, is the wild card. So, I'm predicting what people will do based on basic identity not getting too fancy, with an overlay of Republicanosity, and that's it .... then we see what they actually do!

Where I live, most south-of Mexican Latin Americans are refugees from the ex pat elites!

I think Mexicans and SA/CA refugees swamp the elites.

My understanding is that Cubans are not close to 100% Dem, but many hold out for Republicans, but it is still not a huge number. But enough to matter.

It will be interesting to see who comes closest you or Nate Silver. Currently 538 has it (as of 3 hours ago) at

Hillary Clinton 334.5
Donald Trump 203.4

By Doug Alder (not verified) on 11 Oct 2016 #permalink

I will be closer, but not necessarily THIS prediction. I'll be refining.

The calibration here comes from the uncritical use of RCP poll averages, which cover most states. My next iteration will use only reliable and recent polls, and only after inspecting the last few to correct for obvious trends, if any. That will certainly be a smaller set of states, but a more uniform sample across ratios of T:H.

That will be used to calibrate the ethnic identity and R:D numbers. I may also use a better R:D number than Romniosity. Though I'm pretty happy with romniosity.

In the end, I'll be providing numbers per state of voter choice (as a Trump:Clinton ratio), which 538 will not likely provide.

It turns out that white people vary a lot across the country with how many per state are assholes. I think there is some variation among Hispanics as well, but African Americans are pretty consistent. So, here, I combined ethnicity with a “Romney Index” indicating how many people in a given state voted for Romney against Obama.

That paragraph alone is all kinds of brilliant, and sadly, just about sums up the current state of the electorate in America. +1 internets to our host.

By metzomagic (not verified) on 11 Oct 2016 #permalink

Needless to say, while *not* having a racist, misogynist f#ckwit like Trump as the POTUS is definitely a good thing for not only the U.S. but the world, Hills will still be 'working' with a Rethuglican majority in one or both houses, and be able to accomplish absolutely nothing of any worth.

And then two years from now, because she will have been able to accomplish nothing, there will be the usual
wingnut backlash and even *more* Rethuglican a'holes will be elected to both houses. Rinse and repeat, ain't it amazing that America still manages to function? Well, at least on a superficial level.

By metzomagic (not verified) on 11 Oct 2016 #permalink

One very interesting phenomenon in all of this; somebody has finally been able to break the religio-economic-nationalistic-narcissistic chains that have bound the typical Republican to their party's authoritarian standard bearer. The horror of having an American version of Saddam Hussein using unlimited power to violate whomever he felt like has finally sunk in to at least a few Republican soccer moms and their spouses, and they are repulsed by it. The GOP has over the years always seemed to have appealed to something like an inner Narcissus drive in their flock, but that drive, which until now had been nearly completely impervious to reason or argument, now seems to be getting over-come by some sort of socio-biological drive to protect one's family from predators.

So Republicans won't change their behavior out of concern about climate change that can kill millions, but the words of a powerful tyrant hoping to abuse and exploit their daughters with impunity is something that at least a few can finally get their TV addled minds around. Interesting.

SteveP, don't worry, they'll fall in line and not do that either, by the end of the week.

#11
...don’t worry, they’ll fall in line...

Which line?

By cosmicomics (not verified) on 11 Oct 2016 #permalink

He meant, 'They'll toe the line', right, Greg? :^)

By Brainstorms (not verified) on 12 Oct 2016 #permalink

#13
Understood, but which line? Trump is not a normal candidate and this is not a normal election. States like Georgia, Utah, and Arizona are in play. Some Republicans are so alienated by Trump that they'll vote for Clinton or Johnson. Some, who normally do vote, won't. Yes, some Republicans who've disavowed Trump will fall in line, but the words they've said in disavowing him, that he's not qualified to be president, will persuade some voters not to. This election isn't just quantitatively extraordinary, but qualitatively, and there isn't just one line.

Apparently, the campaign has given up on reaching out to undecided voters and will instead focus its resources on mobilizing Trump's limited group of supporters. This is further evidence of disarray and delusion because the Trump campaign has largely ignored the lessons of turnout and will not be able to do what's necessary to make sure that Trump supporters get to the polls. (My belief is that Trump's organizational deficiencies will more than make up for the effects of Republican voter suppression.)

Furthermore, as it's now clear that Trump will lose, donors are more reluctant to use their money on him. And what was previously a trickle of negative stories that received scant coverage, has now become a stream of front page news.

For a Republican Party, whose policies and leaders have been rejected by its base, falling in line is no longer a simple alternative. It's damned if I do, damned if I don't.

By cosmicomics (not verified) on 12 Oct 2016 #permalink

Depends on what the other end is wrapped around, I guess...

By Brainstorms (not verified) on 13 Oct 2016 #permalink