Forget those polls that showed Trump losing badly. Voters are saying they want change.
(Readers on computers and I-pads will see a box there on the upper right. You can "follow" me, which means that you get home delivery to your email box of new content in this blog, like a morning newspaper. It is easy. Just put in your email address. Don't worry, you can always un-subscribe instantly.)
Peter Sage Introduction of a Guest Post:
Yesterday the Quinnipiac Polling Institute released a poll saying Trump and Hillary are now tied in Florida, Ohio, and Pennsylvania--the states that determine who wins elections. My Democratic friends who were secretly hoping Trump would do well, confident he was unelectable, were cheering the wrong horse. Cruz may have been the unelectable one--the candidate people dislike for some reason (smarmy, extreme, self-righteous, creepy, selfish, ugly, hateful, unpopular, mean, uncooperative.)
Democrats should have feared Trump, not cheered him. Trump can win.
Trump is interesting. He is new. He says things in a way that seem genuine. TV audiences watch him so TV channels show him. People voted for gridlock but they don't actually like gridlock and Trump appears to be able actually to do something. As I have said, Hillary knows that even incremental change is hard. Obama made a joke of it showing the graphic of Hillary pushing the rock up the hill. Funny, but alas for Hillary, true.
Trump knows what to do in a fight. Imagine a prize fighter that we watched beat 16 opponents and made them look weak and foolish. Ranked opponents, the equivalent of governors and senators. Trump is gifted at finding some personality or issue element and using it to beat the person to the ground.
He is turning to Hillary. I just watched a pro-Trump ad with a mother of a man killed at Benghazi saying Hillary lied about the motivations of the mob, then cutting to Hillary laughing with a cackle,as if Hillary was laughing at the woman. Trump is defining Hillary as cold. As a liar. As a woman with an embarrassing laugh.
Today Thad Guyer, an attorney who represents whistleblowing employees and who looks closely at what kinds of messages and messengers influence judges and juries, says that not only could Trump win, he will probably win. The studies cited by Guyer are interesting and worth considering; they say history shows Trump will win. My own sense is that one can always find backtested correlations that "prove" things, e.g. the taller candidate always wins, the candidate born in an earlier month of the year wins, etc. So I question the "science" but I do not question Guyer's prediction of the election result: Trump is a very formidable candidate and uniquely suited to doing the two things he needs to do to win:
***Represent some sort of change from the status quo.
***Make Hillary Clinton even less likable than Trump by election day.
Thad Guyer Guest Post:
 |
| Thad Guyer |
Political
Climate Change Deniers
According to
tested statistical models, there is almost no chance for a Hillary Clinton win.
Climate change deniers are stereotypically Republicans. However, when it comes
to political science, Democrats appear to be big political climate change deniers.
In Salon’s “Donald
Trump Will Not Be President”, (May 5, 2016, http://goo.gl/0h7usd), the analysis
asserts that “history, polling data and demographics all point to a single
result” that Trump cannot win.
This
analysis, like so many others, relies on Trump’s unpopularity with minorities
and women, 2012 Electoral College demographics, etc. Glaringly absent from Salon’s
analysis is any references to three professors: Helmut Norpoth (Stony Brook),
Alan Abramowitz (Emory), and Ray Fair (Yale).
Omitting references to scientific modeling is the hallmark of climate change
deniers, political or otherwise. Pundits using statistical analysis don’t have
to agree with the models, but to have any credibility, they must address them.
Norpoth’s election model forecasted that if—if Donald Trump
wins the Republican nomination, there is little statistical chance that Hillary
Clinton or Bernie Sanders could win. (See, www.primarymodel.com). Well, the “if” is gone, and now Trump has
won, thus Professor Norpoth’s political science model gives him an 87%
certainty of defeating Clinton. The model is complex, but as to a Clinton-Trump
contest, it basically holds that the winner of the New Hampshire primary in one
party (Trump) will always defeat the loser of the New Hampshire primary of the
other party (Clinton). Whether the primary winner does so by a plurality or a
majority vote is not determinative. Political polls, presidential approval
ratings, unemployment and the economy, none of these are determinative. The
model has been wrong only once, in the 1960 Nixon vs. Kennedy contest where both
candidates won the NH primary, an error inapplicable to 2016 because Trump won
NH, and Clinton did not.
Norpoth’s model
is not alone. The predictive models of professors Fair and Abramowitz come to
the same conclusion. Prof. Fair uses econometric models correlating economic
factors with historical election outcomes. Like Norpoth, his model finds that
“items like the burning issues of the day, the identities, personalities and
speeches of the candidates and the strength of their campaign organizations”
etc. all lack determinative value. (See, NYT, “Trump and Sanders Test Economic
Model Predicting a G.O.P. Win”, Mar. 16, 2016, http://goo.gl/61gaRp).
Despite the denier article title, nothing in
it “tests” the accuracy of the modeling. Norpoth’s conclusion is also mirrored by Prof.
Abramowitz’s model based on incumbent presidential job approval
ratings, economic indicators, and the rarity of any party winning a third
consecutive White House term. (See NBC, “Can These Stats Predict Our Next
President?”, http://goo.gl/3M5waj; see also, The Hill, “Economic Models Predict GOP White House,
Even with Trump”, April 4, 2016, http://goo.gl/idg6dN).
Nate Silver, who
has a track record as one of America’s most accurate political statisticians, published--two
years ago— “Clinton
Is Polling Like an Incumbent, And That Could Help Her in 2016” (FiveThirtyEight.com, Mar 26, 2014, http://goo.gl/OY1IeX). Although conceding that his analysis
might be contradicted by Norpoth’s model, Silver posited that a combination of three
likely factors might nevertheless give Clinton the win: (1) the absence of any
serious primary challenger, (2) an early successful end to the Democratic
primary, and (3) a drawn out Republican primary.
All three never happened, all
are against Clinton. How did Silver get
it so wrong? The answer is Silver’s “The
Party Decides” theory, in which his model factors in “endorsements” from party
elites, a factor the professors don’t use. See, Business Insider, “Nate Silver:
We basically got the Republican Race Wrong” (May 4, 2016, http://goo.gl/qJk3YL).
 |
| Now Dead Even where it Matters |
We can assume that if the
models of Norpoth, Abramowitz and Fair are correct, then we should soon be
seeing a rebalancing of the lopsided polls showing Clinton annihilating
Trump. In fact, in just one week of
becoming the Republican nominee, that’s exactly what we’re seeing. See,
USAToday, “Quinnipiac Poll: Clinton, Trump 'dead even' in Florida, Ohio, and
Pennsylvania”, (May 10, 2016, http://goo.gl/aMuOOO):
"Six
months from Election Day, the presidential races between Hillary Clinton and
Donald Trump in the three most crucial states, Florida, Ohio and Pennsylvania,
are too close to call." Critics of Quinnipiac’s poll cite its assumption of
above average white voter turnout, but the polling company defends that
assumption for 2016 based on skyrocketing white primary turnout nationwide.
We won’t have to
wait until November to make our first assessments of the accuracy of the
Norpoth, Abramowitz and Fair models, especially in comparison to liberal pundits
who still forecast a landslide for Clinton, if not a near universal bipartisan
rebuke of Trump.
But we may have to wait until the July 25th
Democratic Convention, the date Sanders says he will stop bloodying the
frontrunner if the super-delegates can’t be persuaded to dump her. For until Hillary is crowned the nominee, we
will not know if up to half of Sanders’ supporters nationally will vote Trump
as they said they would in West Virginia.
See, The Hill, “Exit polls: Nearly half of W.Va. Sanders backers would
vote Trump” (May 10, 2016, http://goo.gl/AUufyW).
We will then find out if under Norpoth’s NH “Primary Model”, Bernie has mortally
wounded Hillary by denying her the NH win.
We can then debate whether Democrats allowed a self-inflicted mortal
blow by allowing a registered “independent” to hijack our party. (See
PolitiFact, “Is Bernie Sanders a Democrat?”, Feb. 2016, http://goo.gl/bCcHlx).
No comments:
Post a Comment