February 25, 2016

"The bottom line is that the primary model, using also the cyclical movement, makes it almost certain that Donald Trump will be the next president..."

Says Stony Brook University polisci professor Helmut Norpoth, whose "primary model works for every presidential election since 1912, with the notable exception of the 1960 election. These results give the model an accuracy of 96.1 percent."
“When I started out with this kind of display a few months ago, I thought it was sort of a joke.” Norpoth said referring to Trump and Sanders, as many alumni in the audience laughed. “Well, I’ll tell you right now, it ain’t a joke anymore.”

As the presentation continued, laughter turned to silence as Norpoth forecasted a 61 percent chance of a Republican win in the general election.
I know some people have trouble accepting Trump as President, but I just want to say that I don't accept "forecasted" as a word.

59 comments:

Mike Sylwester said...

"polisci" is not a word.

traditionalguy said...

Happy Days are almost here again, and the Dems can go sit on it, says Fonzi.



Curious George said...

I wonder how many of my idiot lefty friends will actually move to Canada?

Beach Brutus said...

The model was probably correct for 1960 too, but for Daddy Joe and LBJ buying votes in Illinois and Texas.

Tank said...

Curious George said...

I wonder how many of my idiot lefty friends will actually move to Canada?


Zero.

What do I win?

robother said...

Back off, Ann, the guy is a SCIENTIST. He knows you always yell "fore!" when predicting a result that his well-educated audiences won't like.

traditionalguy said...

Anybody need a job Wall Building? Applications being will soon be taken in Texas, New Mexico, Arizona and California.

tim in vermont said...

I am curious how he accounts for back test overfitting. Somebody said if you give me three independent variables, I will draw you an elephant, give me the fourth, I will draw you the tail.

Barry Dauphin said...

How many undocumented workers will be laying bricks for the wall?

Amexpat said...

Anybody need a job Wall Building? Applications being will soon be taken in Texas, New Mexico, Arizona and California.

If Mexicans are going to pay for it then it's only fair that they build it as well. One idea would be for each of the 12 million illegal aliens to leave a cinderblock on their way out.

Mark said...

Curious, we are all waiting for Rush to honor his claim that he would move to Costa Rica if the ACA passed and was implemented.

Nice attempt at distraction. Why don't you ask your guy before forecasting about others?

Bushman of the Kohlrabi said...

Althouse is no longer predicting a Hillary win?

This is news.

Alexander said...

Ideally, let Mexico pay for it and let American labor build it. It's gonna be a YUUUUUGGGEEE employment project, get the young men out working doing some honest labor. Really gonna help the working class. I love the working class. They're incredible. And they've been kept down by groups in Washington who sold them out for foreign interests. Started in 1965, been a definite downhill ride for the pillars of America since 1973 at the latest. Not gonna happen anymore.

But back to the wall - Obama, he thought he knew shovel-ready. Get out of here! Trump does, though. So we're gonna start with building up a wall around America, and then we're gonna build up America. Build her up till she's great again. It's gonna be fantastic. Just fantastic!

Trump 2016! Choo-choo!

YoungHegelian said...

Yeah, I've pretty much resigned myself to a Trump presidency, too. At least, we'll have a real smoke-show for a First Lady...

I'm also going to go bold & make my own prediction: Trump will peel off a surprising number of black votes, especially black men. Ya see, Democrats think that they get a lot of black votes for ideological reasons. In the past, yes, but not now. Now, it's just habit, and when the Repubs run some ultra-rich Bostonian honky who wears magic underwear, well, what's he got to do with the day to day life of some black guy? Why break old habit?

But, Trump, Trump is different. Trump is rich as shit, lives in a penthouse in his own building in Manhattan, tells anyone who gets in his face to fuck off, hangs with the rich & famous, & surrounds himself with USDA certified grade A prime boo-tay. You know what call that if you're a young black man? A role model.

Alexander said...

I'd like to say that young black men would recognize that bringing in masses of latins fucks them royally, first and foremost.

But... experience and observation in the real world forces me to conclude: yeah, what YoungHegelian said.

Nonapod said...

Think back to when you first became aware of Trump, perhaps going back to the 1980s. If, at the time, someone said to you "One day that dude's gonna be the President of the United States.". What would your reaction be?

We're living in bizarro times.

tim in vermont said...

Curious, we are all waiting for Rush to honor his claim that he would move to Costa Rica if the ACA passed and was implemented.

Do you have a link Mark? Or are you just slinging bullshit you heard over at DailyKos without questioning it? The answer is no. Rush said though that rich people like him could always just fly to Costa Rica for care that the rest of America couldn't get. I could be wrong though. More likely you are a liar.

Dude1394 said...

"Anybody need a job Wall Building? Applications being will soon be taken in Texas, New Mexico, Arizona and California. "

Now THAT is a useful public works shovel ready project. Heh..

John Tuffnell said...

The science is settled. The 97% chance has been FORECAST by a MODEL.

Don't be an election science denier.

Bay Area Guy said...

Obviously, I love this, because it will cause Hillary acolytes to fret just a smidgen.

However, for the same reason I find future modeling suspect in the field of global alarmism, I mean, global warming, I find political modeling suspect as well.

Nobody knows, what shoes will drop for the Donald over the next 8 months for the Donald -- and we know Hillary will have several hundred million to find those shoes.

But, more importantly, when any College Professor says something positive about the Donald, it is a good day.

Learnin' to Love the Donald.

Dude1394 said...

Think back to when you first became aware of Reagan, starring in that Bedtime for Bonzo movie, perhaps going back to the 1950s. If, at the time, someone said to you "One day that dude's gonna be the President of the United States.". What would your reaction be?

We're living in bizarro times.

We always have been. Go back and look at the founding fathers, all kinds of folks. I for one am SO happy ( I was happy about Romney as well ) to have a non-lawyer-politician in the running. I am ecstatic.

Limited blogger said...

Will the wall have the 'nicer' side facing us or Mexico?

Quaestor said...

"Forcasted" is pretty nasty, nevertheless Wiktionary has it as a verb. Merriam-Webster agrees that it's specious.

Alexander said...
This comment has been removed by the author.
Alexander said...

Our side will be a gold-plated base overlain with wrought iron 'TRUMP' worked into a motif that is nothing if not awesome.

Their side will soon be nothing but shitty murals

Bushman of the Kohlrabi said...

The Reagan/Trump comparisons fall flat. Yes, there were attempts to disparage Reagan due to his acting career. However, the attempted smears never received any real traction. Reagan had been a two term Governor of California and had a well known conservative track record prior to running for President. None of this applies to Trump.

PB said...

Curious George said...

I wonder how many of my idiot lefty friends will actually move to Canada?

Zero, because Canada will put up a fence to stop them!

Hagar said...

The difference between predict and forecast presumably is that predict has a connotation of personal judgment while forecast implies some sort of scientific model - like the one that called for this to be an unusually wet winter here in River City.

But I agree with the professor and cast my vote with hers.

rehajm said...

As the presentation continued, laughter turned to silence as Norpoth [forecast] a 61 percent chance of a Republican win in the general election.

That made my day.

Smilin' Jack said...

I just want to say that I don't accept "forecasted" as a word.

From Merriam-Webster:

Full Definition of forecast

forecast also fore·cast·ed fore·cast·ing

transitive verb
1
a : to calculate or predict (some future event or condition) usually as a result of study and analysis of available pertinent data; especially : to predict (weather conditions) on the basis of correlated meteorological observations
b : to indicate as likely to occur
2
: to serve as a forecast of : presage

Writ Small said...

From my reading of the article, this analysis is based on two primary input variables correlated with a third output. Input one is the relative success in early primary states of each of the candidates. The second input is whether the candidate is in a party that previously held the white house or not and over what time period. Because the D's have held it for the last two terms and the incumbent is not running, the R's have a large statistical advantage. The single output variable appears to be overall, national percent of the vote (popular vote percent). I assume a straightforward multiple linear regression model was used.

Because Hillary is having a close race with Bernie in the early states, which implies weakness, and Donald has won 3 of 4 of the early primaries against a large field, which implies strength, the model "predicts" a sizable win for Trump (nearly 55% to 45% of the popular vote). The 96.1% chance of Trump beating Hillary apparently takes the popular vote split from the model and estimates the odds the electoral map agrees with the popular vote winner.

There are a number of problems with using this methodology against this particular race. I would point out two. One, Trump's polarizing nature and very high personal negatives are ignored. If, as some predict, Trump's negatives put a ceiling on his total support even as his core support is very strong, he may begin to lose when the field narrows to one or two opponents. On the other hand, a bandwagon effect may convert skeptics. The point is the model does not account for what makes Trump unique. It is fair to say that, at the very least, other historical presidential candidates with similar early win patterns to Trump were nowhere near as controversial, and that casts doubt on model correlation.

The second problem is with the other input variable: the fact that the D's are running to hold the presidency for a third straight term. Pulling that off would be highly uncommon. George Bush Sr. was historically rare, and so the Republicans would seem to have a large inherent advantage. What makes this year unique, however, is the complete lack of unity on the R side. Some are wondering if the party is going to go the way of the Whigs. The good professor's model does not consider relative party unity, even as it seems obvious that party unity and determination is one normal output of losing the prior two cycles that helps the party to win.

That said, I am not arguing Trump is a long shot to win the nomination or the presidency. I put more faith in betting markets, which have Trump at a greater than 50% chance of winning the nomination at this point and Rubio in the mid-forties. Unfortunately for Ted Cruz, the markets no long see a credible path for him, and put him under 5%.

Virgil Hilts said...

Someone at work gave me 3:1 odds so I bet him that Trump would be president. The places that post odds seem to really think Trump has little chance.
I think he does, and I think Ann thinks he does, and I believe in the "Preference Cascade" concept first touted by Ann a few weeks ago (and by Instapundit today).
I still will not admit (even to myself) that I want Trump to win.

mikee said...

As it is universally accepted that the 1960 election was stolen by the Democrats through ballot box stuffing, I would argue that the model fits ALL elections which were conducted more honestly than that one.

Curious George said...

"Mark said...
Curious, we are all waiting for Rush to honor his claim that he would move to Costa Rica if the ACA passed and was implemented.

Nice attempt at distraction. Why don't you ask your guy before forecasting about others?"

Oh, Mark, you trusted your typical lefty sources, who like you, are full of shit. Bad move.

Rush never said he would move to Costa Rica. He was speaking to a person representing a medical group who said they would set up a practice their for surgical vacations if private practices were outlawed here in the US. Rush said he would take advantage of something like that himself.

Too bad. SO sad.

rhhardin said...

Forecasted is fine, forming a past according to a regular rule, and makes it clear it's not a noun in the position it's in, which is the force driving the author to that form.

David said...

No forecasted? I have backcasted a few times in my life and believe me that's a word. English vocabulary is such a tangle.

Simon said...

I'm not ready to throw in the towel. I acknowledge that the math is daunting and that if Rubio doesn't get out—barring some fundamental defect in my spreadsheet or a change in the race dynamics that falsifies its underlying assumptions—it's hard to see Cruz winning, but we have an opportunity here. It should not be sloughed off lightly, and certainly not for a populist boob like Trump.

Writ Small said...

"Forecast" is built from the word "cast".

If you are OK with the sentence, "The fisherman casted his net," you will have no problem with "forecasted" either.

traditionalguy said...

Poor Rush Limbaugh cannot wait for Cruz to pull out. He is hogtied with most of his audience wanting Cruz like they were taught, but Rush wants a victory over the DEMS worse. So he just mentions that over and over again as a signal that he is fine with Terrible Trump, the imperfect one.

Let's get this over for Limbaugh's sake.

tim in vermont said...

Forecasted is fine, forming a past according to a regular rule, and makes it clear it's not a noun in the position it's in, which is the force driving the author to that form.

I guess so is "loose" for "lose," after all it rhymes with "choose." Still forcasted is going to clink on a lot of ears the way it does mine. I guess we can say "drinked" too, or "deers" for the plural of "deer," or "oxes" instead of "oxen" because, after all, we say "exes" for ex spouses.

I worked with an Indian programmer once who said asked why we say "creation" and "deletion," but not "updation." I said because we weren't as smart as him.

tim in vermont said...

Professor Helmut Norpoth’s forecast presentation took place Monday evening in the SUNY Global Center in Manhattan, which was organized by the Stony Brook Alumni Association.

OK, so it is the political science department of a public university. I am sure they have taken great care to maintain diverse viewpoints to better prepare their students for the real world... right?

As the presentation continued, laughter turned to silence as Norpoth forecasted[sic] a 61 percent chance of a Republican win in the general election.

Rick said...

Donald Trump has a 97 percent chance of defeating Hillary Clinton and a 99 percent chance of defeating Bernie Sanders in the general election, according to Norpoth’s formula.

As the presentation continued, laughter turned to silence as Norpoth forecasted a 61 percent chance of a Republican win in the general election.

If Trump has that high a chance to beat the Dems but Republicans are only 61% then Trump's chances in the primary aren't that great. Why is this particular data point - the most relevant right now - omitted?

Molly said...

Writ small has a clear explanation. Another model (coming from a completely different direction -- using macro economic variables to predict the outcomes, by Yale Economist Ray Fair) also shows about a 54-46 Republican vote share. Neuroscientist Sam Wang aced predictions in the 2012 elections, and he is following Obama net approval (now at -1) which I think does not bode well for Democrats.

But the Iowa Electronic Markets -- where people can put real money on t heir opinions, whether those opinions come from dreams, hunches, hopes, or complex models -- has a pretty pro-Democrat outlook with Dem vote share in the 50s and Rep in the 40s. A second IEM market shows not the expected vote share, but the probability of a victory -- that market has Dem with a 65% chance of winning (and Rep a 35% chance).

AJ Lynch said...

Impacted is a word I dont accept as a verb.

Saint Croix said...

Donald Trump will not release his tax returns. Do you know why?

He's not a billionaire. He's never been a billionaire. The man is a pathological liar.

One author reported that Trump was just a millionaire. Trump sued him for $5 billion dollars for defamation and lost. Why did he lose? Trump refused to show his tax returns.

How bad is it to brag about money that you do not have? Very embarrassing. His net worth is a joke. Probably his penis is a joke too! But definitely his net worth is a joke. Poor, poor Donald Trump. Poor, poor, poor Donald Trump!

"There's a bombshell in his taxes." Yeah, a stink bomb!

Char Char Binks said...

Saying "forecasted" garners disdain and makes the speaker seem unfitted for serious consideration.

Saint Croix said...

"My net worth fluctuates and it goes up and down with markets and with attitudes and with feelings," he said. "Even my own feelings."

What happens when the balloon pops, jackass?

Saint Croix said...

I think it might be time to read this book.

Michael K said...

'You know what call that if you're a young black man? A role model."

Yup. I think Trump will get 30% + Democrat votes, especially from blacks and Latinos.

All the glee about Trump's tax returns will turn out to be a big bust because most of those Trump supporters (and I'm not yet one) don't care. THEY DON'T CARE !

Romney cares. Jon Corzine cares. NOBODY else cares !

Michael K said...

Saint Croix, that's the Global Warming instruction book.

Saint Croix said...

Saint Croix, that's the Global Warming instruction book.

No, there's no bubble (yet) in wacky environmentalism. But the Trump phenomenon reminds me so much of the internet bubble, and the real estate bubble, and the tulip bubble in Holland. The tulip bubble was really insane. There was no intrinsic value at all! And people went crazy.

The only real question is when the bubble will pop. Before the election, or after?

Fritz said...

Statistical models work until they don't.

cyrus83 said...

I will agree that "forecasted" isn't acceptable. How many times with the root verb is "casted" acceptable over "cast"?

Diamondhead said...

These things are true until they're not. Remember Missouri always picking the President? I'm trying to think of a time one of the major parties has nominated someone unacceptable to a large portion of that party. I won't vote for someone who says Bush lied to get us into Iraq...just won't. I'm not a majority in this party but there may be enough to deny Trump the presidency.

Henry said...

If his name was anything other than Helmut Norpoth, no one would give him the time of day.

eddie willers said...

I still will not admit (even to myself) that I want Trump to win.

Me neither.....but I will giggle uncontrollably if he does.

Unknown said...

The model actually did work for the 1960 election, but that election was stolen by Democrats in Illinois and Texas.

Drago said...

In Joseph Kennedys defense, he was sufficiently frugal to only pay for a close win and not a landslide.

When stealing an election its important to not waste all of your ill-gotten gains from boot-legging and other illegal activities.

That kind of behavior might lead several of your male progeny to whack a skirt or 2.

Tom said...

This is similar to the Election of 1830