September 25, 2016

Is any historical model predicting a Hillary win?

To keep folks from getting too emotionally invested in the debates and media polling as we head down the home stretch, it's worth asking what the long-term perspective suggests.

Political scientists say that the consensus of researchers is that campaigns don't matter -- debates, polls, ads, rallies, etc. -- and that the general election is a referendum on the incumbent party in the White House. There is either a demand for more of the same, or a desire for changing the guard, and this mood about the past four years has already formed by the time the general campaign season begins, so that its myriad happenings will not shift the mood by Election Day.

This standard view rarely gets a hearing in the media, whose profits come from micro-obsessing over the speeches, polls, gaffes, etc. Nor is the public very interested -- there's nothing exciting about historical models, when there's so much relatable stuff going on every day during election season, and you want to be part of the buzz in social settings.

However, this year is different because most of the media and nearly half the country want to allay their fears that Trump could actually win the election. If the normally overlooked historical models were predicting a victory for Hillary, then for once they too would be endlessly discussed and dissected, as yet another source of confirmation that "Trump can't win (he just can't)".

An honest political scientist came out way back in the early stage of the primaries and said Trump is all but a guaranteed win. Helmut Norpoth was basing this off of the performance of his primary model, which has correctly predicted the winner of the popular vote since 1996, and retrospectively makes the correct call back to the beginning of the data-set in 1912, aside from 1960 (which was a coin flip in the popular vote).

The model is a simple theory of the electoral pendulum: it awards points to the incumbent party after one term, penalizes the incumbent party after two terms, and favors the stronger candidate in their respective primary contest. After Trump won New Hampshire, South Carolina, and cleaned up on Super Tuesday, Norpoth felt comfortable to call it then. Hillary lost the key early battle of New Hampshire, and was saddled with the burden of running for a third consecutive term of the incumbent party.

Another political scientist whose model has successfully predicted the outcome for even longer -- back through 1984, and retrospectively back to 1860 -- has taken far longer to admit that his model, too, is predicting a win for Trump. He's a hardcore Hillary supporter, and was holding out for nearly a year in an attempt to soothe his cognitive dissonance. But even they have had to concede that Crooked Hillary's candidacy is doomed.

Allan Lichtman's 13 keys to the Presidency is starting to make the rounds in the media now, although he discussed it with The Fix (WaPo) back in May, and even as far back as last December in a faculty research profile at his university. It asks 13 questions about the past four years -- economic trends, foreign policy successes, scandals, etc. -- and about the candidates themselves -- incumbency, hero status, facing third-party splitters, etc. If 6 or more of the 13 go against the incumbent party, they are out, and the opposition party is in.

By these measures, it was already clear last December that the Democrats would not keep the White House, although you could have debated a point here or there. But certainly by mid-May, when Lichtman was interviewed for The Fix, it was guaranteed. He was still delusionally insisting that it remained an open question whether or not there was a serious contest for the nomination in the incumbent party, which hurts the party in the general -- yeah, that whole Bernie phenomenon was all in our imaginations. This guy had his head buried deeper in the sand than the Cruz Cult.

At any rate, both of these respected academics with solid track records are predicting a win for Trump. Neither has mentioned alternative models that also have solid track records, both predictive and retrospective, which are however predicting the Democrats to go three in a row. Lichtman especially ought to be hyping those up, if they existed, since he's so desperate for Trump to lose.

If there are any political science nerds out there, let us know.

Otherwise, take comfort in what the long-term, big-picture data are predicting, and don't obsess so much over what the micro-term, info-overload of BIG DATA are predicting. All signs point to the incumbent party not holding on for another term.

6 comments:

  1. Like you, I want Trump to win. I also think it likely that he will win. However, being the deep pessimist that I am, I think its possible that the liberal-left backers (Soros types) will pull some kind of shenanigan to derail Trump. The democrats and their backers are getting awfully panicky lately. This potentially makes them dangerous.

    I'll say this. If Hillary does win, we are screwed big-time.

    ReplyDelete
  2. Not to jinx it, but we've dodged plenty of Establishment bullets for over a year.

    First they tried to ostracize him at the opening debate (pledge), then vilify him with the initial attack about fat pigs, etc.

    Then they tried to caricature him in the media.

    Then Lyin' Ted set him and Carson up in the first caucus.

    Then ol' Dubya was trotted out to condemn Trump before South Carolina.

    Then the Pope, then the sitting President, then Barbara and HW Bush, then the Speaker of the House, then this person, then that person.

    Then we were going to get up-staged at the Convention. Then there was going to be a delegate revolt. Then there was going to be a splitting of the Party. On and on and on.

    The only major coup that the Establishment scored was that rat Kasich coercing Democrat voters in Ohio to turn out for him in the primary, or else. But he didn't resort to rigging the machines, or electronically intercepting the data and altering it.

    In the general, he won't be able to force Democrats to vote against Trump -- they're already going to. And those that are not voting partisan, cannot be bullied -- see Kasich's losses in Appalachia, including Mahoning County, which is filled to the brim with Trump Democrats.

    I'm not saying we don't have to raise these issues and threaten the Establishment in case they try to fight dirty. But so far we've dodged all sorts of bullets, and the highest ranking members of Trump's party, including the RNC, are on our side.

    Look at the Democrats if you want to see the Establishment getting away with whatever the hell it wants. Bernie wasn't as popular on their side as Trump was on ours.

    With the polling, rallies, etc., all showing how well we're going to do, they can't plausibly steal entire states like Ohio and think no one will notice it's bullshit.

    ReplyDelete
  3. Random Dude on the Internet9/25/16, 8:19 PM

    Voter fraud and rigging is definitely an issue but Trump and Roger Stone have brought enough attention to it to where although it may not be stamped out everywhere, people, especially Trump voters, are alert to it. I think this will definitely pay off for Trump because many attempts to commit voter fraud will likely get foiled in advance or during election day. Thanks to social media, any shenanigans can be caught, distributed, and shared. Otherwise we'd see Ohio-type scenarios in multiple states.

    As you mentioned before, Trump has successfully fought off dozens of attempts by the establishment to take him down and there's no reason to think he can't do it for the final run. He has his high energy, dedicated base to thank for this. It's really his campaign to lose at this point and by now, anything they would have on Trump has already been revealed and discussed several times over. Think about how many Trump employees, former Miss America contestants, etc. have been badgered by the media to denounce Trump and yet they haven't.

    ReplyDelete
  4. Not to mention that we are holding more cards than they are when it comes to skeletons in the closet -- way way way more.

    Success is relative, so even if they dug up one legitimate concern about Trump, there would be ten on the Clintons' side, and of greater severity. Net result: Trump wins.

    ReplyDelete
  5. I actually hope they do try to "go there" with Trump, to the extent that he counter-punches them into disgrace and oblivion. Although that may take place after he wins the election, and doesn't lose support for negative campaigning.

    Bill is remembered as a philanderer, but not as a sexual abuser and rapist. We could saturate the airwaves and the internet for the final month with ads about Juanita Broaddrick et al. whom Bill has molested and raped. Piano music or dead silence in the background, uncomfortable tone of voice, ending in weeping.

    Not only will what remains of Bill's legacy be ruined, those women will also voice righteous anger over how Hillary psychologically abused them afterward so they'd shut up and not ruin the Clintons' ambitions.

    The Haiti relief scam of the Clinton Foundation, expressed by Haitians of all social levels, would kill their image as philanthropists.

    Even the Democrat partisans are already tired of the Clintons -- Bill was a dud at the Convention, and everybody hates Hillary. If they want to go nuclear, we will wipe out what little remains of their legacy.

    We could also see attack ads on the broader Clinton network -- George Stephanopoulos, Sidney "Birther" Blumenthal, etc. Make sure that after the Clintons are crushed, nobody in their crumbling empire is allowed to survive either.

    ReplyDelete
  6. If you know of any good Republican Presidential Primary polls at the state level, I can help here.

    I've done a meta-analysis of Republican Presidential Primary polls at the state level. They underrate Trump support 77 out of 94 times. Across those 94 polls, average predicted Trump support was 3.2% lower than actual Trump votes!

    Problem is, my p value is 0.26; there's a 26% chance that this is all coincidence.

    ReplyDelete

You MUST enter a nickname with the "Name/URL" option if you're not signed in. We can't follow who is saying what if everyone is "Anonymous."