…And the winner is

Hillary Clinton, by a dominating margin. I’m speaking of last Monday’s debate, which she won roughly two-to-one over a collapsing Trump. His abysmal performance was the worst by any candidate in living memory.

As usually happens, the debate winner picks up momentum in the polls, and we see that happening now in both the national and key battleground state numbers. Huffington Post’s Pollster, the best of the poll averaging sites, shows she’s
back up to 4.5 points over Trump nationally. But readers know that the race will be decided in a few Battleground states, not the nation as a whole. Late yesterday we got our first look at Clinton’s post-debate position in five key swing states from Public Policy Polling. She’s moved back into the leads in all of them, most notably Florida and North Carolina, two Trump must-win states.

Many of us were confident Trump would not be an effective debater, but none could have foreseen his reversion to utter boorishness and confusion as his facade gradually slipped from the opening stages, began to crumble by the midway point, and totally disappeared by the final moments. He was at that point a quivering Orange mess, seething with barely-controlled rage.

It began with his dodging of questions on job creation. Then came his tax returns, and Clinton’s deft skill at making these into the issue they deserve to be. It went downhill from there. He was called out on his phony concession that President Obama was born in the U.S., and his refusal to answer the question why he pursued the Birther claims for five long years, when he was aware they were completely unfounded. When asked if he had anything to say to African-Americans, he firmly said “No”, and repeated it. Even a pro-Trump supporter heard the angry cry in a hundred ghettos across the land, and in some twenty million suburban households as well, very likely.

Already on the run, he reverted to his charge that Hillary doesn’t have the stamina to be President, doesn’t have a “Presidential look”, and doesn’t have the necessary (in his alternate-reality) “winning temperament”. He and his ratpack should count themselves lucky that she or the moderator did not at that point correct his poor understanding of English, and remind him that while there is no such thing as a “winning temperament”, there most definitely is a losing temperament.

Nearing the end, he had lost badly. But none of us could have imagined how he was about to self-destruct in the final minutes. Slapped in his puffy cheek with Clinton’s reminder of his misogyny, and her example of it, his public humiliation of Alicia Machado, his own Miss Universe pageant winner, Trump first tried to bully his way out of documented history, and, when Clinton was not intimidated, backed into his regular posture of petulant denial. Learning that the young woman, a struggling immigrant, had become a U.S. citizen, he gave up, crushed by not one woman, but two. Public humiliation in the eyes of 84 million-plus viewers. No Trump building was tall enough to escape his shame, no wall high enough to hide his weakness.

The Clinton team was understandably proud of her massive win, and the total success of the carefully laid Machado trap. They were even ready with a TV attack ad featuring Ms. Machado. It was at that moment simply a matter of waiting for the Friday and weekend polls to come in and see how well she had done.

There was absolutely no way they could have been prepared for what was to come.

The following morning, desperate to prove that his mega-fail the night before was simply a big lie put forth by the media under Clinton’s masterful direction, Trump attacked Ms. Machado on Fox News. Even his own supporters were caught flat-footed. How could he bring that terrible scene up, when he needed to get his message of tax-cuts for the rich and the horrors of inner-city neighborhoods back on track?

That was only the beginning of his descent into mindless rage this week. He simply would not, because, adolescent bully that he is, he could not let go. Clinton supporters and an estimated fifteen million undecided voters have now been treated to three full days of Trump’s winning temperament on full, shouted, incoherent display. Unlike the real and tragic one in Hoboken, his train wreck still dominates the news today.

Worse yet for the Trump cause, we’ve seen a massive surge in Latino interest in registering to vote after last Monday, in the key states of Florida, Texas, Arizona, and Nevada. Little did the self-centered Trump realize the nineteen year-old girl he’d publicly berated in 1996 had become a highly popular figure in the American Latino community.

And let’s be clear: the surge in Clinton’s numbers as of this morning do not reflect the additional voters who turn to her as a result of this latest round in the long-running Republican War on Women. By next Monday we’ll know better the damage he’s inflicted upon himself.

My guess is we’ll see Clinton’s well-organized advocates pushing hard on Machado and misogyny for at least another week, further disrupting Trump’s extremely shallow concentration as he vainly attempts to prep for the second Presidential debate a week from Sunday.

Would Clinton and her mighty surrogate team rather be talking about the goals she has for the country, and her well-considered plans to achieve them? Of course, especially since Trump has no plans except an economic plan that would bring about a recession while handing the richest 1% massive tax breaks. Trade? Trump’s “plan” is to try and get our trading partners to agree to new “deals” or risk the Donald bombing the s**t out of them.

Be sure that the Clinton team will continue to do their best to inform the voters on the pressing concerns and issues that confront the country. But the other side knows they dare not get into serious and complex issues when they must rely on a TV buffoon with the mind and emotional maturity of a fifth grader to explain the Republican position.

This is a sad state for our nation.

But sad or not, professionals and thoughtful citizens concerned with governing, economics, history, science, income distribution, education, public service, defense, foreign affairs, health care, climate change, civil rights, criminal justice, and women’s rights need to face the reality that Trump has created: this election will not be about issues, since only one side is prepared or willing to debate them.

Clinton vs. Trump uncertain, but poll junkies win big!

The terrific electoral vote mapping site 270towin.com has an excellent new page for seeing the updated national and state polls using simple pie charts. On one clean, uncluttered, easy to scan page, you’ll see how the accumulated recent averages of national and state polls stand as of that day. Each pie chart shows the current support for Clinton, Trump, and the “others”. You can start with the latest national numbers, then scroll across and down the page to see who has the largest slice of pie in each state. The folks at 270towin take the average of all available recent polls, so you need not drill down into the bottomless questions of poll credibility, etc. Below we’ll use the new page to gain some powerful state-by-state insights to this extraordinarily complex contest.

How to set up the web page

On the new 270towin polls summary page look for the SORT button (on a forest green background) just above the “National” pie chart on the lower left of the top row of the pie charts display. Above the SORT button is a check box to show the 3rd party and undecided breakout in each pie chart. Check the box and wait for the page to auto-refresh.

Now, select another useful option.  To the right of the green SORT button you’ll see a set of radio button choices  Choosing “Competitiveness” will bring the Battleground and other closely-contested states up toward the top of the array display, saving scan time. Select that button and wait for the page to auto-refresh again.

Now we’re ready to put this new 270towin tool to work.

Tip: As you scan down the page, watch the gold/green/gray slices of the Battleground state pies

The new all-polls-on-one-page illustrates how the combined third party/undecided “other” voters are making the estimation of “likely voters”, and the final November win/lose percentages so difficult this year. (Note that 270towin’s “other” is the same as my “3PU” group name.) In a recent post I’ve dwelt at some length on estimating the likely voting behavior of the 3PUs, so we need not belabor the problem again here. My “net-net” conclusion was that, when all the votes are in on November 9th, “slippage” from the 3PUs will be due to half the Stein, Johnson, and (some) undecided voters deciding that Clinton is safer for the country than Trump. These last-minute decisions will be the 2016 version of the expected traditional “natural” movement away from the third parties as voters decide they want their vote to count.

How can we use the pie-charts page to gauge the potential impact of the expected wave of last-minute voter bloc decisions, especially in the key swing states that will actually decide the national election?

To answer this question for a given state, you’ll first need to add the percentages and subtract from 100 to deduce the Green Party’s voter preference share in each pie chart. In Nevada for example, add Clinton’s 41%, Trump’s 43%, Johnson’s 8%, and the undecided (as of today) 7%. Subtracting from 100% yields Stein’s current share, 1%, and thus, a total of 16% 3PUs.

In past elections where there was a significant third party and/or a large proportion of undecideds, roughly half of the total 3PUs have gone to each of the two main candidates. So, if that happens again this year, who cares about the gold/green/gray “others” pie slices in 270towin’s terrific one-page array?

You do, if you’re a Progressive or Democrat.

Remember please, my conclusion a few days back was that very few of these 3PU switcharoo votes will go to Trump:

Trump has all the supporters at this point he’s likely to get. He has so poisoned himself in terms of character, competence, and temperament that very few 3PU voters are going to change their mind in the polling booth and vote for him. The Libertarians are uniformly opposed to him as are the Greens, so he can, at best, hope for a few undecideds to take him up on his “Whaddya got to lose?” offer.

Unlike in previous elections, where we’d expect the undecided group to split roughly evenly between the candidates, I suggest that the electorate this year is so exceptionally angst-ridden we could see the undecideds splitting 40% for Clinton, 30% for Trump, and 30% failing to vote. Combine this breakout with my assumption that half of the 3Ps will switch to Clinton while the other half stick with their 3P candidate, and we have a wholly different take on the likely final results in November than the simplistic reading of the current polls by the TV commentators would indicate. In a phrase, I’ll call this the “November surprise” of 2016.

Is a “November surprise” a real possibility?

Yes, as can be seen when we do a state-by-state “flip-ability” estimate for the key Battleground and other closely-contested states.

Assuming our last-minute voter behavioral analysis model is basically correct, let’s scan the array of state pie charts on the new 270towin pie charts page. When we do, we see how, in state after state, a switch of just a small portion of the 3PUs (shown in the gold, green and gray slivers at the top of each chart) to Clinton in a few states would decide those states for the Dems. If this same “November surprise” effect occurred in just a few of the red-leaning states, it would turn the overall national contest from the so-called “dead heat” the TV media is selling to a Democratic rout of Trump and his Republican fellow travelers.

Again, let’s use Nevada for a test case. Above we saw that the projected tally based on polls is 41% for Clinton, 43% for Trump, with a total of 9% for the 3Ps and 7% undecideds. My analysis therefore is half of the 9% goes to Clinton, while the other half toughs it out. Meanwhile of the 7% undecideds, I’ll call 3% to Clinton, 2% to Trump, and 2% not voting, That would give Clinton 41 + 4.5 + 3 or 48.5% total to Trump’s 43 + 2 or 45% total, with 4.5% staying with the third parties and 2% not voting. From a 2 point deficit using today’s polls, Clinton pulls out a comfortable win on November 8th.

Now let’s examine the other truly key Battleground states using the new 270towin array.

  • If just one-third of North Carolina’s 15% 3PU voters decide at the end to be safe with Clinton, she flips the state’s 15 Electoral Votes (EVs)
  • Ditto with Ohio, where a quarter of the state’s 18% 3PUs would give her a squeaker, but a win is a win…
  • And in Iowa (!), now being Trumpeted as a likely win for the Orange windbag — a 25% Clinton final decision among the large 20% of “other” voters would move Iowa into the Clinton column
  • A 25% last-minute decision for Clinton among Florida’s 12% 3PUs would let her take Florida by a margin larger than President Obama’s in 2012.

And it may not end there. Look at solid red Texas (where Clinton is actually now leading in the registered voter polls): a last minute decision for Clinton by just 25% of Texas’ huge 27% “other” category would give her an even-money chance of nabbing the state’s 38 EVs. In Georgia, the same scenario is possible, if a third of the state’s 15% “other” voters opt at the last minute for Clinton — not inconceivable considering the Atlanta urbanites’ distaste for Trump, coupled with an expected surge in African-American turnout. In Arizona, the story is even worse for the Republicans, since just a twenty percent last-minute decision for Clinton among the state’s enormous 24% “other” bloc would flip this red state to blue.

What will drive 20 to 40 percent of 3PUs in these key states to change to Clinton? The deep Clinton ground game in most of them will be a huge facilitator, but mostly it will be the realization by voters of the likely dangerous consequences of allowing Trump to win. For most of these last-minute switchers, it will start and end with the Supreme Court, and the prospect of as many as three Trump appointments, turning our country into a fascist paradise for the super wealthy and their corporations.

Yeah, I understand: my “November surprise” scenario sounds too good to be true. And it may well be. Admittedly, my handy-dandy formula for estimating the average final split percentages in the state results will be “off” for any given state. And my estimate of 50% of the 3P votes going at the end to Clinton may be too high. Ah, but it could also be too low: if the fear of Trump in the Oval Office continues to grow, we’d see perhaps 70% of the combined Green/Libertarian voters switching. Considering the large proportion of 3P voters in today’s polls, a 70% average move of these voters to the Democrats would seal Ohio for Clinton, in addition to her wins in Florida, New Hampshire and North Carolina. Game over, Orange guy; back to the penthouse and your other reality show.

A warning for the faint of heart

We cannot predict with any certainty how large a percentage of the 3PU/”other” folks will, in the end, switch to Clinton in any given state. Nonetheless, IMO the argument that a) there will be a significant degree of “others” slippage in these final weeks and b) that the great majority of the switcheroos will go for Clinton are sound hypotheses.

But understand the nature of the last-minute decision process: whatever slippage comes Clinton’s way will not be evident until the votes are counted. It means the contest is likely to be a nail biter all the way to the end, with the results not fully known until the early hours of November 9th.

We also need to remember that the “likely voter” models of the various pollsters would be wrecked (statistically) if as few of 5% of Republicans stay home in disgust or protest. If an additional say, 5% of Trump’s core voters decide he’s hopeless beaten and don’t bother voting, that would further deflate the likely voter estimates for the Republicans. And if, say 5% more of his supporters finally realize what a ruthless, truthless con man he is, we can envision a shocking tidal wave of switches to Clinton. (I’m thinking primarily of the non-college white women here, especially the ones who have a daughter.) The combination of these three sub-bloc behaviors on Election Day would not only defeat Trump: they’d have a profound impact on the Republican candidates for the House and Senate. But we’ll deal with that prospect in a future post.

Ironically, should the former Sanders supporters ultimately decide not to vote, the impact on Clinton would not be significant, since these hard-care voters are now supporting the Libertarians, the Greens, or are milling about in the undecided category. The larger point is that likely voter numbers are suspect this year, and could well have a bearing on my estimates above.

Understand as well that an “October surprise could throw off my “November surprise” estimates. But it’s not clear what such a surprise might be, nor can we say it would help Trump more than Clinton (as the TV talking heads and pundits mostly seem to assume.) I’d bet we’ll see more leaked emails, and that they won’t significantly affect the results. The possibility of a major external terrorist attack cannot be ruled out, of course. But such an event could turn out to be to Clinton’s benefit, as scared voters opt for the more stable, experienced, and competent candidate. My personal favorite October surprise.is that after two debates, or even just one, Trump whines the debates are rigged, and walk away from another face-off. That would definitely cost him permanently in the polls.

No way around it; there’s weeks of uncertainty ahead! But here’s a silver lining to keep you calm. Note that I’ve only listed eight of the states that could conceivably be in play, using my behavioral analysis of the “others” category in these states. Just one red-to-blue flip among these states due to pro-Clinton 3PU slippage would effectively block Trump from any path to 270 EVs.




Follow-up — Polling bla bla

As I tried to explain yesterday, panic by Clinton supporters is premature:

Check out Clinton regains momentum and read Perspective on the polls. And this deeper analysis, also from Stu Rothenberg: Clinton’s lead is bigger than it looks.

Note that none of the recognized pollsters and polling experts are using my “soft support” modifier concept, described yesterday, at least so far. This isn’t surprising — we’ve not seen an election where, out of disgust, a significant percentage of a candidate’s professed “likely” backers might in the end simply refuse to come out and vote for him (or her).

Be wary also of articles or TV talking heads claiming that 80% or more of voters “have made up their minds”. Either or both our flawed candidates could still drive allegedly “committed” supporters away, with more stupid lies or insulting comments by Trump, or newly-leaked emails by her. This election is a war of attrition, folks.

Finally, for today’s edition, see this stunning experiment by the NYT Upshot team. The same Florida polling sample was given to five different pollster teams, and they came back with a range of answers, from Trump up one percent to Clinton up 5% — a six point spread within one state sample! Without spinning or obfuscating, the NY Times crew showed us the deep flaw in our assumption that polls can be trusted.

The tightening — dead heat or simply more media muddle?

Watching the latest rounds of Pundit Pinball over the past two weeks, with the media’s death-spiral emphasis on emails, Putin mania, Deploramania, Clintonian health panic, Trump’s I-was-against-it-before-I-was-for-it-before-I-was-against-it babble, the “I’ve got a secret plan to defeat Isis, but the useless Generals I’ll be inheriting from useless Obama next January haven’t told me what it is yet” gambit, and the “I’ve got more Generals and Admirals than you do — so there, fat boy!” counter thrust, one could be excused for giving up on this election altogether.

Thanks to the Washington Post for what will likely prove a vain attempt to call for an end to the “which email was that?” witch hunt. Not that any press or TV news organizations are going to suddenly become professionally responsible again anytime soon.

And thanks even more to Kim Jong Un, for blowing a five kiloton hole in his own country, and reminding us that the election is about very real and very scary issues.

And then, just when I thought things would be getting back to normal, meaning another week of the campaigns sniping at each other over concerns of personal competence and character instead of debating the issues, we learn from General Powell what he really thinks, and not just about Trump, and the media are off to Kandidate Kiddyland again.

On cue, along comes the Birther nastiness, as if we needed another dose to go with our pneumania pills. I’ll save you the time and cut to the chase. Friday we saw Trump’s casual concession, tossed off during a promo event for his new hotel, that, yes, the President is really an American. Offered without any mention of the five years of racist madness he led and fostered. Offered without any apology to our President and our people. Offered in fact with another lie about Clinton, and hence, a justification in Trump’s adolescent mind, where he believes his own lies, for his five year attempt to sabotage the legitimacy and effectiveness of the Obama administration.

Then we get another forty-eight hours worth of home-grown terrorism, if that’s what it was. Another two days down the media gutter, during which we’ll learn nothing of any substance about the most momentous election facing us since 1964. Media muddle, indeed.

Please, can we get back to the polls?

Spoiler alert: Clinton and Trump are not “tied”.

Given the way the TV and online media hype every poll result that comes in, readers may be forgiven for having that impression. I’ll say this for hopefully the last time this cycle::

  • Poll results may tell us how voters would vote if the election were held now, but they are not a “prediction” of what will the results will be seven weeks from today
  • “How voters would vote now” is itself a misleading phrase, my apologies: so many voters in this 2016 contest will probably be making up their minds as they enter the polling station, every “this minute” poll would probably be significantly wrong
  • Some polls are more accurate than others; it’s your job to learn which tend to be accurate over (long stretches of) time — if you don’t care to do this, and my sympathies are with you!, then you should simply not listen to “news” reports based on the latest poll
  • Similarly, many polls tend to be biased over time toward the Republicans and a smaller number toward the Democrats — as I’ve suggested in a previous post, go to 538’s polling organizations ratings table to see a given polling organization’s credibility
  • State polls are the ones that count, since they assess today’s voter opinions at the state level, where, thanks to our obsolete Electoral College system, Presidential contests are won or lost
  • But for technical (not “conspiratorial”) reasons, state polls lag national polls, meaning the latest national poll result may not reflect the current preferences in a state, or a group of states
  • Polls of even the same voter population can change dramatically in a short period of time, due to events, surprises, discovery of new information, or an external threat to the country.

OK, Mr. Truthiness, all well and good, but WHO’S LEADING?

Given the above, we don’t actually know who’s ahead in the “if we voted right now” case. As far as November 8th is concerned, however, Clinton is still ahead. According to Dr. Wang, Clinton has an eighty-five to ninety percent chance of winning more than 269 Electoral Votes. In the national polling, of the approximately 80% of declared Trump or Clinton backers nationally, she has about 41% to his 39%.

Loyal readers and true Patriots, bear with me, please. Political polling in the U.S. is really complicated. The producers of TV news programs are not going to spend the hours required to explain why. 99% of voters don’t really care to know. So here are some tips to get you through the coming six weeks:

  1. Until roughly the final two weeks before Election day, polls should be treated as one would any other information resource. Obsessing over them is a waste of time.
  2. In every election, a significant number of voters remain undecided or seemingly committed to a third party until the last moment, roughly five percent. Right now, I’m betting this number is at 29%, give or take a few percentage points. No, I can’t prove this, since the polls are more unreliable than in prior years.
  3. The debates may or may not persuade folks that Clinton’s competence and real qualifications outweigh Trump’s bluster and snake-oil salesmanship, but we probably won’t know the full polling impact of the three debates until the middle or third week of October.
  4. There probably will be more revelations of damaging emails, or of Trump’s shoddy business and financial practices, and we won’t know what impact, if any, these have on the voters’ choice process, but I’m betting any such impacts will prove to be minor.
  5. For all the wise-sounding back and forth by the TV and press commentators, we don’t know who’ll actually turn out to vote, or mail in their ballot. “Likely voter” polls try their best to screen out those registered voters who will, or may probably not vote, but this is not an exact science. This year, with roughly two-thirds of the electorate intensely disliking one or both major candidates, it’s especially challenging to predict who will actually show up to vote, and how relatively firm they are at that deciding moment about their essentially negative choice.

The complications of “soft support” mixed with two independent parties

At least seven — I’m guessing nine — point’s worth of “decided” voters are “soft” on their seemingly chosen candidate. Trump and Clinton are, people tell the pollsters, the two most disliked candidates in modern history. We know that many of Trump’s and Clinton’s backers prefer their candidate as the lesser of two very poor choices. It’s reasonable to assume therefore that, conservatively, perhaps 5% of Trump’s stated supporters and 4% of Clinton supporters fall into this “soft” category. These queasy supporters could easily abandon their stated choice in the final days, and simply stay home. How they’ll actually vote, if they do, is a once in a lifetime headache of enormous proportions for pollsters.

Now, add in the Libertarian and Green supporters, who poll nationally at roughly nine and four percent respectively. Include the truly undecided seven percent, and my back-of-the-envelope estimate is that the true national numbers today are 37% Clinton, 4% “soft” Clinton, 34% Trump, 5% “soft” Trump, with 9% Johnson, 4% Stein, and 7% truly undecided bringing us up to 100% of likely voters. Clever readers will note my estimate of twenty-nine percent of soft supporters and undecideds would suggest that the nationally quoted 80% combined committed Clinton and Trump voters is actually just 71%. Yes, that is what I’m saying. Note that this says just 71% are hard-core committed to a major candidate. If my reading of the inherent softness and confusion among those polled is correct, it means that the 29% will decide the contest in November.

This election is the most difficult to poll or “model” than any in our modern history

We’ll deal with the softies in a bit, after we consider the independent Parties and “truly” undecideds:

  1. Expect that with no Gary Johnson in the debates, his roughly nine percent at the moment of the national polling averages will fall to as low as five or less percent of the final total. Your Truthteller’s magic cigar box says recognizing at the end how crucial this contest is for our country, about forty percent of his voters will vote for Clinton, while no more than a few percent will vote for Trump. Bitter to the end, perhaps fifty-five percent will doggedly stick with him.
  2. As for the Greens, perhaps half of Stein’s three to four percent will switch to Clinton, and exactly zero will go for Trump.
  3. The bigger mystery is how the presently truly undecided seven percent or so will split, since we won’t have data until the late evening of November eighth. The smart approach I suggest is to assume two points to Clinton and one to Trump, with the remaining four sitting it out.

How do I expect the 29% to break?

Potentially almost a third of likely voters could change their choice in the four weeks after the debates. Yikes! We got here thanks to two highly distrusted major candidates and two independents whose only actual contribution would be to help the totally unqualified Trump get elected. Nasty. Unprecedented. Frightening. Impossible to model. And yet, the very same massive pool of 29% “effectively undecided” voters that makes this election so difficult to analyze is the 29% that offers us a strong hope for a positive outcome.

How could this be?

Because, patient reader, Trump is unlikely to gain more than a few additional supporters in the final seven weeks before November 8th. And a very substantial proportion of the 29% are going to realize the prospect of a disastrous Trump presidency is not a risk worth taking. What looks like a tooth-and-nail fight for Clinton right now could actually be a landslide in the making..

Let’s break out the 29% and test this hypothesis:

Trump — I expect 3% of his softies to sit out the election, and 2% to pray intensely and then vote for him, while he picks up 2% of the undecideds; he’ll get zero from Stein, no surprise, and zero from Johnson (whose supporters came to the Libertarians to avoid the bitter nightmare of Trump).

Clinton — I’m confident she’ll keep three of her four percent soft supporters, with the one percent staying at home (but certainly not switching to Trump); she’ll pick up 4% from Johnson, 2% from Stein, and 3% of the 7% truly undecided.

The final tally on the back of my envelope?  Trump at 38%, Johnson at 4%, Stein at 2%, with Clinton at 49%, and 6% not voting. No, she wouldn’t win a national majority, but she would trounce Trump in most of the Battleground states, making for a very large Electoral Vote majority. On the safe side, I’ll go all-in: Clinton will win the Presidency if she takes Virginia and Pennsylvania.

This outcome will be the result of the (completely warranted) fear of Trump, not of any sudden outpouring of trust in Clinton. So be it; we’ll take it. Instead of talking about the “enthusiasm gap”, media mavens should be discussing the “riskiness factor”.

But what about the issues?

Thanks to the disruption of the Trump candidacy, the serious issues of governance have been shoved out of the arena this year, so we’ve heard next to nothing about the challenges that face our country. Trump’s clever methods of grabbing the spotlight with bluster, lies, personal and cultural insults, provocation and controversy have forced the Clinton campaign to try and respond to and cope with this whirlwind of disinformation, instead of presenting her policies and having a rational debate.

The Clinton team was fully aware of the kind of nasty fight they would face if Trump became the nominee, and prepared for it by at least publishing her policies and positions in detail on her website. They’re there for all to see, clearly presented, free of BS and boilerplate. But very few voters have taken the time to read this information. Trump has done nothing like this, since he knows his supporters don’t care about the issues, and since he doesn’t want to be pinned down on anything specific.

In spite of this, your faithful Truthteller will discuss the serious issues soon, as we approach the first debate. We can but hope…

Labor Day shortakes

The Trump campaign team is having a spin-fit, and the TV media are still reporting from Silly City re: the election polls. Here’s a brief update.

How to understand the polls

After we published our long explanation of the polling problems of late August, the Fivethirtyeight.com site published a handy list of a baker’s dozen things to keep in mind when reviewing the polls. Readers will find some useful suggestions in addition to the ones I covered.

Likely vs. Registered

Back for a moment to my post of yesterday — to see how changing from “registered” to “likely” voters can radically impact the poll averages, take a look at these two polls from the same pollster, Reuters/IPSOS, both reported the same day, September 1st. The first one is their latest likely voters poll and the second is their registered voters poll. As you can see, the likely voters poll says Trump and Clinton are virtually tied, nationally. The second poll, representing all registered voters, shows Clinton with a six point lead over Trump!

The poll averaging folks live in a quantitative swamp, friends. It’s not that hard to see why the lightweights of broadcast journalism choose to avoid the nastiness of analytics, and content themselves with spouting the most-repeated headline of the day.

Spoiler alert!

Our analysis of expected voting patterns is not quite ready for prime time, but we’ve reached a point where we can say that, in this election, many of the most revered assumptions by pollsters and pundits alike will not prove accurate. We’re confident that the legendary “90 %” who pollsters always say have made up their mind by Labor Day will be shown to be incorrect, come November 8th. The number today is significantly less than 80%. What may seem to be a minor difference to non-poll and election analysts will have a major impact on the final results. We see a 95-plus percent chance that Clinton will win with an Electoral College landslide, and with a popular vote margin large enough to win the House along with re-taking the Senate.

We’ll be publishing a series of posts before the first debate discussing exactly how Clinton will win. That will give us a further seven weeks, of course, to be proven totally wrong. Stay tuned!

Polls Panic!

The state of play

This week saw an excellent reminder why polling and political science experts universally agree that pre-Labor Day polls are not to be fully trusted. If one was an innocent, and listened only to the TV hosts and commentators, it would seem as if Clinton’s strong post-convention, mid-August 8 point lead over Trump had fallen precipitously to about 2 percent.

The reality is that much of her perceived decline was due to her large post-convention bounce finally wearing off. Some would add that the roughly 2 point “bonus” she earned when Trump stupidly attacked the Kahns is gradually disappearing, too, but this analyst is convinced that a substantial number of voters from all parts of the voting population are never going to forget the calm, courageous Mr. Kahn asking Trump if he had indeed ever read the U. S. Constitution. I suggest a 1-2 point edge for Clinton was “baked in” for the Democrats going forward by Trump’s thoughtless, irrational response. (And we still don’t know if Trump has read the foundational document of our country.)

The actual national consensus number, folks, is about a 5 point lead for Clinton as of this morning, Sunday, September 4th. (FWIW, my own adjustments put Clinton at seven or eight percent, at least, over Trump.) Much more important, Clinton has maintained her lead over Trump in all but one small Battleground state, namely Iowa, with its six Electoral Votes (and its combination of few minority voters and large population of uneducated white voters, who are Trump’s core constituency). Clinton is thus very, very far ahead of Trump in the predicted Electoral Vote tally, even with this week’s media gnashing of teeth and talk of “tightening”.

The real question today is not what Clinton’s poll lead is; if audiences understood quantitative reality, they’d realize that the big headline right now would be asking how Trump reaches 270 EVs, when he looks to get less than 200 come November 8th.

What’s in a political poll (super-compressed version)?

Readers will be advised here to recall the old question about what goes into sausage. It ain’t pretty. But sausages, at least, can taste really good, unlike poll results, which are doomed to leave at least half of us with a bad taste in our mouth.

The largely undiscussed truth here is that a number of polls have been shown to have a consistent Republican bias year after year, just as a few others show a tendency to inflate the Democratic vote. (To learn about each polling company or organization, see the last link in my previous post, to Fivethirtyeight.com’s excellent, ranked, and really comprehensive list of polling organizations and partnerships.)

For example, Rasmussen, Emerson, IBD/TIPP, and the LA Times/USC polls all have demonstrated a small to ridiculous Republican bias over time. On the Dem side, YouGov and Monmouth results tend to favor the blue candidates. FYI, in both cases, these biases are typically due not to some conspiratorial plot, but to various sampling and technical differences.

And, lest the above put you off polling altogether, let me be clear: as can be seen in the Fivethirtyeight.com rankings, the vast majority of polls do not show any long-term bias.

There is another difference in polls that is critical around Labor Day. Poll samples tend to include all registered voters during the year build-up to the Labor Day before the election. It’s easier to find these, and easier means less costly for the polling organization. Around Labor Day, many of the leading pollsters begin sampling likely voters, a smaller part of the population than the registered folks. Obviously, your poll result will be more accurate if it reflects the views of the people who actually intend to vote. But assembling a statistically valid large sample of likely voters is way harder than just taking random names out of a state database of registered voters. $$ more difficult, which explains why the polling folks avoid this type of survey until Labor Day, and why some simply don’t ever bother changing over to “likely” from “registered”.

Without going into mind-numbing detail, polls of likely voters tend to slant Republican, not because the pollsters have “programmed” that tendency; it just comes with the real difficulties to building the sample group.

Leaving us with another problem. A number of online-only polls may mix likely and registered, and may even be biased by Party affiliation, since these polls have no reliable way to know who the respondent really is. Give me an online legion of Young Republican students and I’ll show you a poll that has Trump winning by a mudslide. I meant “landslide”, sorry.

Poll averaging trauma

What all this means is from roughly August through early September the two leading poll averaging sites Huffington Post’s Pollster and Real Clear Politics are (necessarily) mixing Clintonian apples and Trumpster oranges. The two sites use all the generally accepted polls to compute their current leaders and losers. There are numerous arguments as to why the two main poll averaging sites would include the polls that have a known bias; I won’t regurgitate the pros and cons of this “maximum inclusiveness” approach here. Let’s just say it reflects the tried and true dictum of safety in numbers.

Averaging the hundreds of election poll results is further complicated in these (still) early days of the campaign by the infrequency of polling, especially for the “real” type of poll, that is, where respondents are called on the telephone by a trained survey technician. Less poll results means shakier average numbers, right? Add to this the infrequent results dribble in singly or in small batches over the course of the week.

The averaging sites don’t simply combine all available results and divide. That would be acceptable if one were, say, computing the average rainfall in various parts of the country in the most recent week. Voter opinions are not so simple. They evolve. So as you’d expect, the averaging sites will tend to give more “weight” to the more recent results.

But, and there are many “buts” in this no man’s land of human behavioral statistics, probability dictates that, in some weeks the averaging sites will get a group of poll results that, collectively, just happen to be of the biased type, and which share the same bias.

This is what happened over the past few days, when IBD/TIPP, Rasmussen, and the LA Times/USC were the only polls to report new national results. (Note that the other major recent poll result was from the excellent Reuters/IPSOS poll. As luck would have it, this week the poll averagers seem to have decided to switch from IPSOS’ “registered voter” results to its “likely voter” number. The result was an apparent four point drop in Clinton’s lead!)

What not to worry about on Monday, your last official beach day of this summer

As of this holiday weekend, the averages appear to have lowered Clinton’s averaged national results by a point or three. But this is a temporary blip; she still leads Trump nationally by four to six points, or more. Let’s further remind ourselves that President Obama soundly defeated Romney by achieving a four-point edge in the popular vote. And of greater importance, Clinton has a very strong lead in projected Electoral Votes, the only race that matters.

I’m not suggesting Democrats and Progressives relax at this early stage, far from it: an enormous effort is needed to turn out the progressive and independent vote and take back the House of Representatives. But panic is not in order, unless one is backing Trump, and looking at the Truthteller Electoral Vote map.

2016 EVMAP 001 - DY0812 REVd

Best sites for poll results

Of the several dozen sites and blogs your Truthteller checks frequently, here’s my shortlist for people who want to keep up with the race but who don’t have much time.

Huffington Post’s Pollster is the recommended first stop when checking the latest national polls for Clinton versus Trump. Then you’ll probably want to compare the Huffington Post team’s average to the average as computed by the folks at Real Clear Politics.

You’ll definitely want to check the Real Clear Politics daily listing of poll results page every few days or so, just to be aware when the major polling organizations publish an update.

See the innovative New York Times election polls analysis site for a quick but deep update on the Trump/Clinton race. Scroll down to see the current results in the key Battleground states. Minimum hassle.

Check out 538’s current forecast page for their consensus view of the current results (based on Huffpost). And don’t miss 538’s polling organizations ratings for their excellent poll ratings list (must see!)

Speaking of the polls, I’ll discuss my take on the “tightening” in my next post. Be prepared to  cry, if you’re a Trumpster.

What’s that? No, not the “softening”; I’ll leave it to others to debate that nonsensical non-issue. Donald Trump is still hardened, thanks to the newly-installed Mercer-owned threesome now pulling Trump’s strings.

Caught Between The Devil And A Deep Blue Sea

Panic! Much tearing of hair on “Morning Joe”! Anxious searching for the emergency exits on the Trump campaign bus! OMG! We’re all gonna die!

Is it due to the unexpected strength of Hillary’s recent speeches as she closes out the Democratic Primary? Or perhaps to the way her party is rapidly coming together behind her, and already performing as well as either of the two great Obama coalitions in ’08 and ’12? Or is it the decision of numerous prominent Republicans to support the “flawed” Clinton instead of their own totally unfit, unthinking, self-centered candidate?

No, the present collective Republican Establishment dismay is due to polling numbers:

  • A 17 point spread in national favorability of the Democratic Party over the Republican Party
  • Early battleground (“swing”) state polls indicating a potential Electoral College rout of Trump come November 8th.
  • A devastating, highly respected poll showing 55% of Americans and 65% of women would never vote for Trump
  • Recent indications that Trump’s support among less-educated whites, the sole demographic he has not insulted and alienated, is beginning to erode

And the most devastating number of all:

  • A roughly 15 point drop in Trump’s national popularity in the 30 days since he was declared the presumptive Republican nominee, putting him 12 points behind Clinton, with nearly five months of rough campaigning ahead.

It all suggests an approaching Clinton tsunami, a veritable deep blue sea to wash away the Orange devil and his crazy agenda, and swamp the rest of the Republican ticket as well.

Poor Paul Ryan, Mitch McConnell, Reince Priebus, and the rest of the Republican parade of do-nothings who just weeks ago disavowed their “principles” and readily lined up behind the loudest know-nothing in U.S. history.

Chill, guys! It’s early days, yet. Four and a half months to go. Plenty of time for your big Orange-otan to change his insulting, self-centered ways and act Presidential. Don’t you get it yet? All you’re asking is for him to act. He’s a TV star, right?

And even the Donald can manage to learn at least one real speech in all that time. Or, if he can’t, or more likely, if he won’t, you could try voice-over at his rallies and lip-syncing.

And let’s not forget the plane. That massive symbol of being above everyone else, with a sexy bedroom and golden toilets, the very essence of Trump’s essential world-view. Think of the impact on voters when that behemoth lands in hotly contested places like Fairbanks, Jackson Hole, and Charleston, West Virginia. Just watch those thoroughly impressed folks come out in November and vote Republican like they always do.

OK, so maybe he can’t pull off the scam of his career with those moves. Well, maybe he won’t need to! We have just one word for all you naysayers:

The debates!

OK, two words. But just think how Donald will shred the oh-so prepared and polished and professional Clinton with his proven, crowd pleasing, national TV network-worshiping technique of simply saying “Lyin’, crooked Hillary!” to each and every statement she utters. Even her name. Even her opening greeting, “Good evening Mr. Trump, I’m looking forward to debating you on the really important issues facing our country.” “Lyin’, crooked Hillary!” He can shout it back at the moderators each time they ask him a question. “Lyin’, crooked Hillary!”

Just imagine! People across the land will soon be shouting at their TV screens in unison: “Lyin’, crooked Hillary!” each time she opens her mouth. For an hour or more, the fifteen percent of American likely voters who adore the Orange windbag will be able to scream their heads off and put “that woman” firmly in her rightful place. The nerve of Hillary Clinton, thinking any strong man is obligated to debate her!

Um, wait a sec, people.

Maybe relaying on the debates isn’t the answer, either.

Maybe yelling your determination to defeat Senator Clinton come Hell or high water isn’t going to help.

After all, you’re already staring into the Orange face of Hell, and the high blue water is on its way…

Dumb-bo The Flying Elephant

Donald Trump seems to be inordinately proud that his campaign organization is about one tenth the alleged size of Senator Clinton’s.

The thinning crowds in his rallies seem to eat this up, but it’s clear they have no concept of building a team to execute a billion dollar national election campaign. We only hope the Trump brigade’s expectations are low, because the eighty or so people reputed to be working for his campaign seem to be incapable of doing even five percent of the work done by Clinton’s.

The specifics are telling.

The Trump team has set up no visible ground game in the all-important battleground states. He has no allies of any note or weight. His endorsers are little more than Trump sound alikes: dull brown paper bags filled with hot air. The staff are not raising any money, it appears; maybe they are among the roughly 400 people in America who still believe the Donald is “self-funding”. As a result the Chancellor of Trump University has found himself having to chase for money all over, but especially in the safer red states. With apparently no cash on hand to speak of, his staff has managed to produce and air very little TV or online attack advertising. This from a TV celebrity who claims to be an expert on working that medium.

The staff seems particularly weak at the numbers game. They have the great white elephant set to fly to states where he doesn’t have a prayer of contending like California and New York, thus taking away campaign visits from states like Michigan, Pennsylvania, and Virginia, states he absolutely must win to have a shot at 270 or more electoral votes.

So far, all the Trump team, its feuding, bumbling bosses and their windbag candidate have accomplished since he became the presumptive Republican nominee is to manage a fall of fifteen points in his national popularity versus Senator Clinton.

Paul Manafort may soon have to import some of his close friends from the Russian military intelligence command, to get the campaign marching in proper goose-stepping order.

It’s rough, folks. We hear that an insider in the campaign organization said that of the roughly 80 campaign staffers, apparently half are assigned each day to help wash down his Jumbo Jet and polish the gold-plated faucets and toilets.

To experienced politicos and campaign managers, this may seem dumber than dumb. But they forget that the big fat plane is really Trump’s only campaign asset. He swears his name is his brand, and since he cannot fly one of his towers to places like Fresno, he simply must have his plane fueled with Trump BS and ready to go.

But Trump is kidding himself. The man lives in his golden cage penthouse, isolated from reality and protected from the truth and anyone who dares disagree with him. He only comes down from his perch to fly around the country attacking some people while thoroughly boring the rest.

He seems unaware that the Trump brand is rapidly becoming poison, as the details of the Trump U scam come to light. By the end of October, after the NY Times and Washington Post finish digging up his nasty past and many failures, he may well need to repaint his polluting plane with his real family name, “Drumpf”. Or “No Longer Trump”. or “Once and Never King”.

But we needn’t be harsh. Drumpf’s relatively few campaign people have done their best, we’re sure.

But, still, a 15 point drop in his national popularity in 30 days?

Go figure. No, really, Donald, go figure.

Why 2008 is not really comparable to 2016

We’re firmly persuaded that the only value of national voter preference polls is to give the TV anchors and pundits and political journalists something to yammer about. Keep those ratings up! Sell those newspapers and magazines!

Many others have written extensively about the inherent weaknesses in even the most highly respected polls, so we won’t belabor them here.

We’re also critical of using exit polls of voter choices from four or eight years ago as a 2016 predicting tool. Exit polls may help explain why a particular voting group (“demographic”) voted for or against a given candidate or position then, but they are basically useless now, when years of events and developments have made the landscape of choices very different.

Nate Cohn of the NY Times published a solid analysis Thursday of the weaknesses with exit polling as a predictive tool. Cohn’s well-developed takeaway point: “demographics” change over time. As an example, he shows how the White middle and working class demographic is growing faster than we realized. From this he suggests it’s a bit of good news for Trump, who is desperate for support from this less-educated white demographic. He needs them to attend his endless boring rallies, for starters, but he needs them even more because he’s now alienated most of the other demographics.

My ad hoc takeaway regarding the analytical frailty of the 2008 and 2012 exit poll numbers is different:

How a given demographic votes can also change over time.

Here’s a specific example of my out of the box (or perhaps out in left field!) thinking. We know from 2008 exits that Obama had a hard time attracting votes in this key (to Trump) less-educated white demographic. So, facing “crooked Hillary”, if this demographic is larger (as a percentage) of the voting population, then yes, it seems Trump will appear to be likely to benefit in November.

Ah, but consider:

In 2008, many, many white working class voters mobilized to stop Obama from becoming the first Black President. We know that a deep strain of anti-Back racism lives on among these white voters. But there’s no rational reason to assume all of this hard-core right wing category will be so energized to come out against a known white candidate, however much she may be disliked personally.

How many of these folks will simply throw up their hands at this year’s Presidential choices and stay home? We estimate that at least seventy-five million eligible American voters will not cast a vote. Some of these, probably a smallish number, will be the die-hardest of Sanders’ supporters. They will be few because by five months from last week, the country is going to know in agonizing detail not just what a fraud Trump is, but how he could do great harm to the nation with just one single Supreme Court appointment.

Some stay-at-homes will be core Republican conservatives and white evangelical Christians, who will see by November that he is going down, waaay down. They’ll be able to stay home without the slightest twinge of guilt.

If history repeats, the majority of these no-shows will be male: females tend to vote much more than men. And what the old exit polls cannot do is guess how many of these women will also stay home this year.

We estimate a possibly substantial proportion of Republican women will decline to vote. Trump has attacked them every time he’s attacked other women, and these women of principle will want to make him pay for it. Oh, and those who claim these Party loyalists will turn out to save the skins of the down-ballot Republican candidates may want to think again, in our view: many of these women will be equally disgusted with those men who feel forced to support Trump.

As for the assumed near-universal hatred of Clinton by Republican less-educated white women, that notion has yet to be tested in a historic year when women of every political stripe have the chance to vote for the first female President in U.S. history.

Look more closely at this less-educated white demographic, at the half that’s female, and who actually show up or mail in a ballot. How many of these frustrated, angry, fearful women are likely to vote for a grasping, self-promoting, windbag as opposed to a woman? Not merely a woman, but a woman with a more than 40 year history of public service, service that focused on the needs and hopes of other women, their daughters, and their families. A woman who by her grit and wit and unceasing good will has now surpassed her own husband?

The males who do most of the poll work and polling punditry have no clue how compelling this lure of history will loom to millions of women, and to a goodly number of pro-female men, too, by the first week of November.

Of the 140 million we estimate will go to the polls, we predict it’s the Clinton voters who will be energized this year, not the Trump supporters. And the energy won’t be confined to females from all the demographics. The execrable Trump is motivating many millions of white and other males to help ensure that he never sets foot on the White house lawn, much less enters its portals as our Commander-in-Sleaze.

My advice is to be wary of reading too much into either hyped national polls that don’t focus specifically on the swing states, or assuming that old exit pools describe a new reality.

Damn! If we only had an entrance poll we could rely upon…