Princeton Election Consortium

A first draft of electoral history. Since 2004

Sometimes life comes at you fast

September 16th, 2016, 7:15am by Sam Wang


It looks like the Presidential state-poll snapshot is heading for a near-tie. This should become evident in the PEC analysis by the time of the first debate. I believe this will be a temporary situation. It will take at least until after the first debate on September 26th to find out.

National polls currently show Clinton ahead by only 1.0 ± 0.6 % (median ± estimated SEM, 7 pollsters with at least some post-Phlegmghazi respondents). However, our state poll-based analysis moves more slowly. I expect the Meta-Margin to keep on moving toward Trump for at least a week as state poll medians catch up. To get an idea of where the PEC analysis will head in the next few days, see electoral-vote.com, whose main map displays the most recent single poll for each state and therefore is noisier than my calculation – but more up-to-the-moment.

I estimate that if we had up-to-date data in all states, in an election held today the Presidential outcome would be extremely close, approximately like the map below.


Click the map to create your own at 270toWin.com

If the election were today, it would be a nail-biter. But as Glenn Thrush at POLITICO points out, there are lots of ways for Clinton to recover.

Today, I will go over some reasons that may make Clinton supporters (and Trump opponents) feel concerned.

1. National polls have narrowed. Seven polls that include post-September-11 respondents show Clinton ahead by 1.0 ± 0.6 % (median ± SEM, 7 polls), quite close. Compare that to the last week of August, which showed Clinton ahead by 4.0 ± 0.9 % (10 polls).


The narrowing of the race seems to be driven by a break in undecided voters toward Trump. Since early August, Donald Trump has gained about 4 percentage points and undecided voters have decreased by a similar amount. At about 42% support, Trump is nearing his previous highs for the year.

Until now, undecided voters have been running about 4 points ahead of where they were in 2012. Many of those undecideds were probably disaffected Republicans unwilling to support Trump, a radical candidate unlike any that the U.S. has seen in over 100 years. Trump is not saying a lot relative to his usual inflammatory and divisive statements. It appears that this approach of keeping relatively quiet is bringing home the Republican base – and increasing his favorability rating. We will see how long he can keep it up.

If you were to extrapolate this trend linearly, by Election Day, 52 days from now, Donald Trump would win the popular vote by 10 percentage points…and there would be negative 5% undecided voters. This is obviously impossible. But silliness aside, it is within the realm of possibility for Trump to take the lead, at least temporarily.

I still expect Clinton’s lead to increase again, on the grounds that she has led all year. This is basically the concept of regression to the mean. Previously, I noted that the national Clinton-vs.-Trump margin in 2016 has averaged 4.5 percentage points. The standard deviation is 2.2 points, comparable to the four Presidential elections from 2004 to 2012. Such a small standard deviation indicates very high stability, consistent with the intense voter entrenchment of the last 20 years. If conditions swing back toward Clinton, especially after the first debate on September 26th, I can imagine the stories now: “Clinton shows renewed vigor” or something like that.

Despite the lack of big ups and downs this year, it is important to remember that the large number of undecideds and minor-party voters provides a source of uncertainty. We have to wait to see how they shake out. As we see with today’s conditions, recent changes have benefited Trump.

It is usual for support for minor candidates to fade in the home stretch. Johnson and Stein get strong support among young voters, a group that would otherwise be expected to tilt strongly Democratic. A big hope for Hillary Clinton must lie in bringing these voters back. Maybe Clinton should bring out Bernie Sanders, Elizabeth Warren, or some of the entertainment celebrities who showed up at the DNC.

Technical note: The Presidential model’s Bayesian prior does contain the possibility of a big swing in undecided and minor-party voters. First, the prior’s standard deviation was set to 3%, somewhat larger than recent elections. Second, it was set to have “fat tails” by using the t-distribution with 1 degree of freedom instead of the usual bell-shaped distribution. Under the latter assumption, a swing of 2 standard deviations would occur almost 15% of the time. If Trump wins, it was still in the model as an unlikely but not impossible scenario.

2. State polls are still narrowing. The advantage of national polls is that they provide information quickly, within days. But national polls are less accurate than state polls, and worse, do not reflect Electoral College mechanisms.

This is where the Meta-Margin comes in. The Meta-Margin acts like a regular Clinton-versus-Trump margin, but it is based on all state polls. It measures the amount of across-the-board swing in opinion that is needed to tie up the Electoral College. However, it is dependent on the availability of state polls, which dribble in slowly.

The coming 1-2 weeks will probably feature a steady drip, drip of state polls favorable to Trump. Likewise, the Meta-Margin will move toward zero over the coming week, and could even favor Trump at some point. Do not interpret that as new information. That is just the same information that we already know now, expressed at the level of individual states.

3. Are Trump supporters undercounted in polls? In a phenomenon called the Bradley effect, it used to be that poll respondents tended to vote for minority-group candidates a bit less than their answers would indicate. This idea has been used to suggest that additional Trump supporters are lurking in the wings, but are unwilling to admit their possible racism to a human interviewer.

I doubt that there is any hidden bonus for Trump. The Bradley effect was never more than about 2 percentage points, and in any event, it disappeared a few decades ago. A research finding from Morning Consult earlier this year suggested that Trump’s support was lower in live-interviewer polls than in computer-conducted “robopolls” has been the same – but later analysis has not confirmed this result. That original report probably arose by chance or had deficient statistical methods.

Finally, I note that Trump’s support in the primaries was very close to what polls indicated. In all probability, polls will do fine this year.

Having said that, because state polls have not yet caught up with national polls, it is useful to see what would happen if one candidate got a boost in state polls. Over in the right sidebar are links you can click if you want to see what would happen if Trump had 2% more support than state polls currently indicate. That probably gives a good picture of the true state of the race today. Conversely, once the Meta-Margin gets closer to zero, the regression-to-the-mean concept suggests that a Clinton recovery would become likely. At that point you can click on the Clinton +2% link to see where things might head.

4. What about downticket races? Four indicators seem to move roughly in unison: the Presidential race, the Senate, the House generic, and the current President’s approval ratings. This is consistent with the nationalization of politics that has emerged in the last 20 years. After the national party conventions, all four measures moved toward Democrats. Now all four measures are moving toward Republicans.

This coordinated movement does not seem entirely consistent with the idea that undecided voters who moved to Trump were disaffected Republicans. Shouldn’t they have shown up all along in polls as supporting downticket Republicans? This could be a quantitative issue of how much each indicator moves – for example, the Senate aggregate has not moved as much. Or there’s some other mechanism at work, such as differential rate of survey response.

>>>

The close race that we see today is not likely to last. Odds favor a swing back toward Democrats, but this is not certain by ant means. It will take a little while to find out – maybe a week, maybe longer. A good number to watch is the median of national polls.

Many of you might want to have an impact on what happens in November. The most effective use of your time and money is to work on races that are closest to the edge, i.e. in the 20-80% probability range. In the right-hand sidebar, The Power Of One Vote lists states where individual voters have the greatest potential to influence (a) Senate control, and (b) the Electoral College outcome (click through for that). Those are good places to campaign. Depending on your political preference, you can also contribute using the ActBlue/NRSC links on the left.

Tags: 2016 Election · President

146 Comments so far ↓

  • SpecialNewb

    Once again this entire week everyone in leftblogistan was saying ‘Sam has HRC at 90% we got this.’ There is going to be quite a freakout today, no doubt.

    • Burlywug

      Well, 86% is still quite good, but maybe this will help keep complacency from setting in. Overconfidence in results is how you end up with a Brexit.

    • JMV Pyro

      I’m just glad this is happening now. Hopefully it scares enough people out of their complacency.

    • AP

      I think Prof. Wang’s comment about the possible reversal of the meta-margin in the next few days was meant to prepare us for numbers much closer to 50% as state polls roll in. As of yesterday, there was a 27% spread between PEC and 538, with all the other major forecasters in between. In the meantime, the authors of the predictions discuss the philosophy and structure of the models rather their past performance. What does it mean when a model swings from 99% to 1% for a candidate, and then back again?

    • Sam Wang

      Yes, this was my point.

  • Kevin King

    No freak out for me. Not yet, anyway. Besides, we’ve had weeks of generally positive Trump national conversation and a pile on on Clinton. And still, what we have is a tie basically, after all that. Let’s see if it stays that way

  • Ron Petracca

    I’m interested in PEC’s thoughts on how Presidential approval factors in. I think PEC tries to avoid intermediate outcomes in its statistical modeling, but Clinton’s current performance is lagging behind Obama’s fairly strong ratings. Logically, it seems as if voters who approve of Obama’s performance would be difficult for Trump to win, and, given Obama’s all out push for Clinton, likely to end up with her. If we see Presidential approval as a key for Clinton, and view last weekend’s health related issues as causing an aberrational blip akin to Obama’s subpar first debate, then there is all the more reason to think the sustained pattern of a small but solid Clinton lead will reemerge. The danger I see in my line of thinking is that it becomes a kind of unskewing of reality, I don’t like the data therefore I explain it away.

  • Jwa

    How do/did undecided affect the prediction before; Were they expected to break a certain way or effectively a tacit 50-50 split on the probability curve?

    People usually reach a decision before they are aware of it. Since it seems trumps support came from undecideds and now it more closely matches previous years, is this possibly a case that these undecideds were just historical republican voters that were still convincing themselves they were ok with Trump as long as he was the republican candidate?

  • Frank

    In 2008 during this time, this is the week in September when Obama fell behind McCain in the electoral college polls. It wasn’t until 9/20 that he would take the lead again (270+), and after then he didn’t fall below 270 up to election day.

    Well, so far, Clinton has not fallen below 270. We have four or so more days to go. I have no evidence of a correlation, but as we entered this week, I predicting Clinton would start falling in the electoral college and she has, but not as far as Obama in 2008 did.

    • JMV Pyro

      Right, I remember that. What caused that bounce anyway? Was it Palin-mania or something else?

    • Alex P

      JMV Pyro, it was the collapse of Lehman Brothers and their quite different responses. McCain said the fundamentals of the economy are strong and Obama didn’t say anything foolish.

    • pechmerle

      The big — and decisive — event of that week in September of 2008 was the Lehman bankruptcy, followed instantly by the biggest banking crisis since the Great Depression. I have a job in which analysis of economic behavior plays a role, and I knew right away that McCain was toast from that point forward. McCain’s clueless response to the banking crisis certainly didn’t help him. Nor did the disastrous Palin choice. But it was the Lehman bankruptcy/banking crisis that was the decisive nail in the coffin of the in-party holding onto the presidency that year.

  • Kristin Johnson

    Sam, I’ve been following you for years, although I’ve never been good at math. Most of your analysis goes right over my head. However, one website recently caught my eye, and I wondered what you thought of it. It’s called Pollyvote.

  • smartone

    Sam a question – what about ground game? For the first time in modern politics we have a candidate who is being significantly outspent on ground game.

    Will this have an effect depressing actually likely Republican voters in swing states on election day?

    • Kevin King

      I suspect Sam would reply that’s what the +2% button is for. I have a feeling this will make a difference in a close race, but really the ground games disparity is just speculation, as far as i can tell.

  • Charles

    Wasn’t it like a month ago that the Trump campaign was in shambles, morale was in the toilet, top Republicans were acting like he didn’t even exist, some were even suggesting that he drop out and be replaced, etc.? It seems like he kept doing all the wrong things and is inexplicably reaping benefits. I can’t see any of this gain lasting, it’s based on absolutely nothing.

  • Sean

    I’m not entirely convinced he will actually show up at the debates. He has to know that won’t help him at all. He’s weak when it comes to facts and knowledge about world affairs etc. He’s all about rhetoric which won’t really work in a debate setting.

    • Olav Grinde

      I’m not so sure about that. I remember listening to that famous Carter–Reagan debate on the radio. Jimmy Carter was beating the GOP hopeful hands down, and he had just decisively won a key debating point.

      So what does the foremost Republican actor of the 20th century do? He utters his famous words, utterly devoid of content:

      “There you go again.”

    • bks

      Please believe me when I tell you that is not how the presidential debates work. If he appears strong and Hillary appears weak, the facts will not matter. Ronald Reagan was a borderline illiterate and was trounced by Carter on the facts, but he looked presidential.

      Could Trump say anything stupider than he has already said?

    • Olav Grinde

      @BKS, I totally agree. Case in point: pull up one of the Gore–Bush debates on YouTube. Turn off the sound – and ask yourself who looks more Presidential?

      I hate to say it, but it wasn’t Gore. He came across as a distant college professor. Whoever coached George W. Bush did a brilliant job of betting him to use his body language.

      In our Post-Factual Era, where Trump is whipping up voter fear and anxiety to frenzied levels, almost half the country wants a strong bully. They reassure themselves that Trump is, after all, their bully – he will simplify the world, “making American great again” to the detriment of our enemies and rivals.

      How wonderful it must be to imagine the world as a simplistic, zero-sum game!

    • Olav Grinde

      * getting him

    • Sean

      What’s interesting to me is that his loudest supporters are definitely not his kind of people if you know what I mean. I suppose that’s always been the case with the GOP though, people voting against their own self interest.

      I’m not saying Hilary particularly cares about your average blue collar worker either but I live in West Virginia and there are literally unemployed coal miners who think that they will be going back in the mines the day he comes into office.

    • Richard

      Trump has extreme overconfidence bias. And he loves attention and love being contentious. I think a debate is catnip that he won’t want to pass up.

    • Asher Steinberg

      John Wayne was the foremost Republican actor of the 20th century; Ronald Reagan was the foremost Republican of the 20th century who was an actor.

    • Olav Grinde

      @Asher Steinberg: I beg to disagree!

      I concur that Ronald Reagan may well have been merely a B-movie actor; however, his Oscar-worthy roles took place long after he left Hollywood…

      Surely Reagan’s performance as the charming, reassuring, and grandfatherly American President make each and every John Wayne movie effort pale by comparison.

  • we_are_toast

    There’s something that continues to bother me. The national and state polls continue to be correlated.
    We have a unique test case here where the Clinton campaign has spent hugely on TV ads in battleground states, while Trump has spent very little. This seems to indicate that TV ad campaigns are pretty much worthless.
    So what is moving voters? It appears to be the news cycle which appears to be heavily slanted against Clinton while normalizing Trump.
    If this is the case, then the polls are not measuring voter preferences as much as they’re measuring the bias in news coverage, and it also means it is the media who will determine the next president.

    • Michael Coppola

      “This seems to indicate that TV ad campaigns are pretty much worthless.”

      Trump is sui generis. I would not conclude that his ability to get his message out in non-traditional ways is something that could be replicated by any other candidate, nor would I assume that his success without advertising means that Clinton’s advertising has not been effective. For all we know, Trump might be up by double-digits absent Clinton’s ads.

  • Keith Romig

    The thing that probably is most dispiriting right this minute is the generic House margin for the Democrats has slipped to under 3%, well short of the margin needed to flip control of the chamber.

  • Sophia

    I was mentally prepared for this election. I know that so much craziness would happen with Trump as a nominee. There will be so much back and forth. As Obama says, it is never as bad as you think and never as good as you think. I suspect there is more to come. This is not fact or science but rather a psychological insight. I will bet that Hillary has an incredible team of psychologists who will give her the one thing she needs to say that will make Trump do or say the wrong thing at those debates. This is how she won her senate seat. She was in a debate and her opponent came towards her and everyone felt protective of her. They loved it when she cried while campaigning against Obama. When she genuinely shows vulnerability people come to her. Plus I believe that there is someone out there holding onto something they have on Trump. I really do. I think the press is aware they created this monster and they will take it down.

  • Ed Wittens Cat

    Hillary’s choice of VP alienated me fo’ sure.
    Incredibly selfish and arrogant– she didnt want to be overshadowed by Corey Booker or Julian Castro.
    Have Kane drop out for psych reasons, grab the news cycle, pretend to deliberate for a week or so and have Booker or Castro step in– a page out of Trumps playbook– saturate the news cycle.
    That said, she utterly deserves to lose– she could have been grooming Castro or Booker for a 2024 run to benefit the party.
    Republicans are reluctantly coming home to Trump
    but bluebrain genetic tendency tells me young Democrats may not the same.
    Im sure not going to work GOTV this time around.

    OTOH all is not lost– OFA won 2012 for Obama.
    The Dream Team turned out “invisible” voters–
    https://www.washingtonpost.com/politics/how-the-obama-campaign-won-the-race-for-voter-data/2013/07/28/ad32c7b4-ee4e-11e2-a1f9-ea873b7e0424_story.html
    Project Ivy may do the same, and turnout invisible voters for downballot–
    http://swampland.time.com/2014/02/24/project-ivy-democrats-taking-obama-technology-down-ballot/

    Do equivalent “invisible” voters exist on the GOP side?
    I’m doubtful because of redbrain genetic tendency– much easier to lather up the uneducated angry white men and “deplorables” (eg: KKK officials) and normalize racism and xenophobia via media equal coverage — plus all the GOP voters that are going to vote are already out there yelling “USA! USA! USA!”

  • Andrew

    Thanks for a very useful post (and the links are really great)! Recently Trump has toned down his rhetoric in his stump speeches and stuck to the standard Republican talking points:

    http://www.rollingstone.com/politics/features/how-donald-trump-lost-his-mojo-w438162

    It looks like this has worked, by toning things down, he has given permission for Republicans to vote for him who don’t think of themselves as white supremacists.

    As Sam says, my guess is that the rest of this movement in polls is all differential survey response, Clinton supporters are depressed by the news cycle and aren’t as enthusiastic to answer calls from pollsters.

    Hard to say how this plays out in the polling booth, but the Democrats best weapon is still the R. candidate himself, if he feels the need to express himself freely Clinton will go back up. (But he isn’t stupid, he can read polls the same as anyone else.)

  • Hugh J. Martin

    HRC did this to herself. Yes, millions of Trump supporters are racists. But millions more who support Trump or may be undecided have been left behind by the economy and the culture. In the basket of deplorables comment they heard a wealthy politician dismissing them and their lives in a humiliating way.
    Then HRC followed up by collapsing. Anyone who saw the video has reason to be concerned. Heat stroke – a possibility that reporters seem unaware of – and pneumonia can each be extremely serious. Her campaign’s effort to spin the cause as no big deal just fed back into many people’s belief that she cannot be trusted.
    This is starting to remind me of the primaries. Bernie Sanders would make a speech and his face would turn red because he was so mad at the world. But for some reason HRC could not put him away until late in the primaries.
    I don’t know why everyone thinks she will easily win the debates. She keeps stepping on herself.
    I say all this as a Clinton supporter. She should find Trump easy to beat, but instead she keeps finding new ways to make it harder for her to win.

    • Andrew

      I entirely disagree that HRC did this to herself. Only a damn idiot would go and declare “Hey I’ve been working so hard I’ve given myself pneumonia!” to the press, especially the way they have treated her.

      Yes, collapsing hurt her, but much more damage has been done by the media and their “Clinton rules” for reporting (http://www.vox.com/2015/7/6/8900143/hillary-clinton-reporting-rules). If you are in doubt, see Matt Lauer’s treatment of Trump’s outright lies about Iraq, or his foundations illegal donations to a prosecutor, or the 6 foot tall statue his foundation purchased for him. Then, imagine what the narrative would be if HRC had done any of those things.

    • Andrew

      P.S. Apologies for my “damn idiot” remark — came out more strongly than I wanted in writing and hit “Submit” without editing.

  • Latichever

    The pneumonia may be a blessing in disguise.

    It lowers the bar for her debate performance.

    Any credible performance now will give her a boost.

  • Marty Schiffenbauer

    Trump can win as long as the campaign focuses on the personal not the issues. You hardly ever hear or read anything about the stark differences between Trump & Clinton on climate change (Trump thinks it’s a Chinese hoax); abortion (Trump wants to appoint Supreme Court Justices who will overturn Roe v. Wade); Obamacare (Trump wants to abolish it, leaving millions without health insurance); tax reform (under Trump’s tax plan the super rich get the biggest tax breaks). And the list goes on. Let’s hope the debates refocus the campaign from Hillary’s emails and Trump’s refusal to release his tax returns to his far right positions on so many issues.

  • CalStateDisneyland

    In short, Republican voters are voting Republican. A certain portion of nominally Democratic voters are exercising their right to vote for the Ralph Nader of their choice.

  • A New Jersey Farmer

    Just as many Republicans are returning to Trump, the debates and Hillary back on the trail will bring Democrats back to her. Trump’s birther comment and retraction are evidence that he can implode at any time. In the end, Clinton has the more popular policy answers.

  • Tom_b

    If Johnson and Stein are pulling some young voters, Clinton ought to use surrogates to underline that Johnson, as a libertarian, would not be inclined to increase the minimum wage or increase government funding for college. Nor would he be inclined to use government to do what needs to be done on climate change. Stein is a “vaccine-denier”. That by itself raises a red flag, given the government’s important role in health care. We have enough anti-science politicians.

    • Bernard

      Hillary isn’t going to win by tearing down Johnson and Stein. If highlighting the obvious negatives of an opponent were sufficient, she’d be up 10 points on Trump. Ultimately, Hillary need to find her voice to make an affirmative case for her presidency that is distinct from (a) I’m less bad than the rest, and (b) read these policy papers.

    • Sridhar

      My theory is that the debate benchmark set at 15% for third party candidates is probably also driving some of the poll numbers for Stein and Johnson. In this regard if Sept 26 comes and goes and only the main candidates debate there could be a possible movement of these folks back to Trump/Clinton giving us a clearer picture of the race. In 2012 all 3rd party candidates combined had a 1.75% share of total vote cast. The state of society today is not much different than 4 years ago to warrant such a dramatic shift towards them. The last poll in OH was especially generous to them as in 2012 Johnson/Stein combined only had 1.25% of the vote in Ohio

    • Sridhar

      Just a follow up to my previous post regarding debates and third party candidates.
      This from POLITICO today

      “[T]he Board determined that the polling averages called for in the third criterion are as follows: Hillary Clinton (43%), Donald Trump (40.4%), Gary Johnson (8.4%) and Jill Stein (3.2%). Accordingly, Hillary Clinton and her running mate, Tim Kaine, and Donald Trump and his running mate, Mike Pence, qualify to participate in the September 26 presidential debate and the October 4 vice-presidential debate, respectively. No other candidates satisfied the criteria for inclusion in the September 26 and October 4 debates. The criteria will be reapplied to all candidates in advance of the second and third presidential debates,” the Commission announced.

  • Richard

    Political science “fundamentals” models predict a close race (see e.g. Sabato’s website for examples of these models). A highly partisan electorate sets the condition for a close race. Two candidates whom are both widely reviled (whether rightly or wrongly) make it difficult for either to pull away from the other. Hence, we have a close race. The odds favor Clinton, but of course her winning is not certain. And when a race is close, a big gaffe, or poor GOTV, or an October surprise may tip the balance.

  • Evan

    Dr, Wang:
    I’m a stats guy, so I’m surprised to hear myself saying this, but doesn’t this analysis, as exemplified by your statement that “I still expect Clinton’s lead to increase again, on the grounds that she has led all year,” just treat the election as a statistical process rather than as a moving, breathing marketing-driven popularity contest? When do events like a shock to the system (i.e., a terrorist attack or a debate debacle), or even subtler movements drive the race beyond the predictive ability of a stochastic process? And how do you reconcile the pure quantitative analysis with your call to get involved in the down-ballot races – how does involvement in the process impact the quantitative analysis, especially if you’re making ex ante predictions about where things will most likely move?

  • Joeff

    What I find interesting (and reassuring) is that the “internals” still seem to favor Clinton–to take the most obv example, white college-educated voters, normally solidly R, are tipping D. Apart from non-college whites (especially men), other subgroups strongly lean D. Who will vote for T, who didn’t vote for Romney in 2012?

  • PECismyoasisofsanity

    The third week of September is always the busiest week of the year for hurricanes. Many storms suddenly appear, gain strength and threaten the US. But then they fizzle out.

    This is the third week of September.

  • whirlaway

    It is odd that there are no polls from PA for a long time. Last poll was 09/08 and we can see why a week is a looooong time in politics!

    • Sridhar

      Pennsylvania is most likely not close enough to be in play at this time if the campaign schedules are analyzed.Neither Trump or Pence is scheduled to be in my state till at least the first debate. Mrs Clinton has a speech in Philadelphia on Monday. Time will tell if the state is still competitive

    • DebbieR

      I heard that Clinton internal polling has seen a significant PA drop and she now leads by only 2%. This comment was made by a guest on CNN, can’t confirm accuracy of that statement.

  • Roger

    I wonder if the tightening in polls will make the news media start treating Trump more seriously and more critical.

    So far they seem to treat him much less critical, more like a entertainment celebrity than someone running for President.

    There were signs of that this week, and if it continues, it would be bad for Trump on many fronts. For example: Trump nor his campaign seem capable of handling critical questions.

    Trump can never let someone get the last word or seem to be in control, so he would likely not be able to handle it.

    His campaign manager lost her cool to shout, “are you calling him a liar” just this week, to a reporter who was pressing her about Trump’s income taxes.

    His daughter hung up on a reporter during planned interview about Trump’s maternity/family leave policy, because she felt it was unfair that the reporter asked about a prior comment of Trump’s, “pregnancy is an inconvenience for business.”

    • TeddyVienna

      Conspiracy theory: NBC has a secret deal to the exclusive rights for a reality show on a Trump presidency. That would explain Lauer and Fallon’s reverential treatment.

      More seriously — yeah, the media have been ridiculous. Trump could propose conscripting all immigrants for a ground strike on behalf of Putin in Ukraine, and Clinton would still lead the papers if she sneezed. (Sneeze-gate!)

  • Anthony

    Liberals/Progressives really need to get their act together and vote responsibly instead of thinking their moral “protest vote” is going to move the policy needle now or 4 years down the road.

    The number one argument they use is that they don’t care if Trump wins because he will have a Democratic check in the Senate. Yeah, very unlikely that Trump wins and the Democrats take the senate. Can you guys imagine how much damage Trump can do with a Republican House and Senate? With Republicans judges in the lower courts say goodbye to voting rights for minorities. With Supreme Court judges say goodbye to LGBT rights and campaign finance reform/citizens united. I could go on and on, but I will take DECADES to recover (if we ever do) if Trump is elected with a Republican Congress.

    • Ed Wittens Cat

      What is happening is the news media beginning to understand their part in the rise of Trump.
      But it may be too late.
      Todays press conference is overreach for Trump– its made birtherism front and center when he just wanted to move on, and is energizing african-americans.
      But the media is covering black caucus speakers in depth right now.
      Now Hillary needs to engergize millennials– let Kaine fall on his sword and make Booker or Castro VP.
      The difference is liberals will vote for Nader– while conservatives value a winner “uber alles”.

  • A

    We are looking at something very unique, and I don’t think it’s Sam’s fault that models are having a hard time dealing with what Trump represents.

    The media’s destruction has brought about a complete failure on the part of those like CNN to adequately address the Trump phenomenon.

    A large part of our electorate just wants to see a wrecking ball in action.

    The justification of “I did it for the money” allows Trump to say and do anything he likes–people say that he’s doing it for the ratings, so it’s okay.

    We are the victims of a total lack of education, a total lack of respect for facts and reality.

    We want fantasy and so we get candidates like Trump who peddle it for a living. This isn’t new, he is just the culmination of decades of lies being sold to the most vulnerable and ignorant people.

    But I think that this election is going to look very different than those in the past.

    And I think unless some very substantial changes and efforts are made by those who oppose Trump–we are going to see a shocking and horrifying result here.

  • Ed Wittens Cat

    Neither Stein or Johnson qualified for the debate.
    Perhaps the fade happens now.
    https://twitter.com/ZekeJMiller/status/776861822024638465
    Hillary needs an issue to energize millenials and youth.

  • anonymous

    If I have your permission, then I am going to go ahead and panic now.

    • Sam Wang

      Sure. Then find some competitive Congressional
      Districts near you to work off the stress by getting out the vote.

    • Phoenix Woman

      She’s already starting to recover. The Ipsos poll came out today and she’s firmly on top.

      A lot is going to hinge on whether the press finally has stopped grading Trump on a curve. (From what I hear, the individual reporters have wanted to stop coddling him months ago, yet their bosses wouldn’t let them.) Judging from how they handled his openly mocking them over the past two days, the media may finally be willing to call him out on his lies.

    • Joel W.

      Between the birther retraction this morning, in which he erroneously blamed Hillary for the birther movement, and the implied threat to Clinton later in the day (disarm the Secret Service and see what happens), I don’t think today was a particularly strong one for Trump. Is it too much to ask that his mistakes be reflected in public opinion a few days down the road?

    • Greg Gross

      Well, perhaps we can take some comfort that in the last coupla’ days, The Donald has: (a) belittled a minister for reining in his HRC rant in her church; (b) said that HRC started the birther movement while reversing himself and saying POTUS was in fact born in the US, after a fawning press was tricked into enduring a Trump hotel promo; and (c) suggested that HRC’s secret service agents disarm in order to “see what happens” to HRC. That may nudge the undecideds away from His Orangeness a bit.

  • David Auerbach

    A hopeful sign: The NYTimes used the word ‘lie’ in a headline about Trump. I’m pretty sure they haven’t before (nor even about Bush). They’ve often documented the lies but the word often escapes them.

  • W.Dow Rieder

    Something weird is going on with electoral-vote.com and New Mexico. That’s been pretty consistently blue on PEC and most other sites, but electoral-vote is showing T up by 10. Huffpost pollster doesn’t list any state polls at all there that I can find. Anyone have an idea what’s up?

    • W.Dow Rieder

      Whoops–just found the polls on Huffpost. The graph was blank because there were too few of them. One from May and two from August, showing C up by 8, 9 and 14. Still no idea where electoral-vote is getting that T + 10.

    • Bela Lubkin

      See electoral-vote.com/evp2016/Pres/Maps/Sep13.html#item-13. They incorporated the 50-state results from Ipsos. That 38C:48T NM result is the only one within their time window.

    • coolchien

      Ipsos’ NM results have been very spiky. The latest is Trump +5, but sample size only 141 people. IMHO, best to ignore it. The entire States of Nations poll have been problematic, e.g., it shows Vermont as red this Friday–”high confidence Trump but unknown margin as there are too few respondents”. I think this entire exercise in 50 states polls is border line polling malpractice

    • Sam Wang

      It is not clear that subsamples can be interpreted with much confidence. Consider ignoring that result?

    • Jeremiah

      They had quite a bit of heartburn concerning the Ipsos polls as noted here: http://www.electoral-vote.com/evp2016/Pres/Maps/Sep13.html#item-13
      They erred on the side of including them even though they note that New Mexico looks very suspicious and they were internet based only.

    • Jeremiah

      @Bela Lubkin sorry – didn’t see you had noted this also.

  • Big G

    I have a question about voter screens and how much that has to do with this shift. In one particular poll, recent CBS/Times, they have Trump losing white college grads by 11 (surely a historic low for a Rep. with this group, and 20 points worse than Romney). He’s winning white non-college grads by 28 (close, but slightly better than Romney). Yet he’s winning overall whites by 11, and he trails Clinton by 2. I can’t figure out how he can be doing so much worse with the largest demographic group, but better overall than Romney; certainly I can’t believe he’s doing significantly better with blacks and hispanics. As near as I can tell, this poll assumes non-college white out-voting college whites 3 to 2, which is opposite to last election and against recent trends. Similarly it must assume a big drop in minority turnout. When a poll publishes cross-tabs like this, do they ever show your their assumed voter makeup? They must be assumed a very very different makeup of the electorate from 2012, which you would think they should have to say.

  • Scott

    Calm yourselves by remembering Romney’s meteoric rise in the polls after Obama’s disastrous performance in the first debate of the 2012 election. That race reverted back to the norm, as Sam expected.

  • Norman Ohio

    You’ve already lost most of your credibility with this pre-emptive article. You’ll lose the rest of it when Trump wins in November. You’ll be arrested in January for criticizing Trump all through this election.

    • Sam Wang

      Laugh while you can, and come on back on November 9th.

    • Jeremiah

      Arrested for criticizing the President – how American!!

    • Michael Levinsohn

      Please leave this site to serious inquiries. There are plenty of places to go for what you’re looking for.

    • Paul Griner

      Arrested? So, a President Trumps’ first action would be to somehow suspend the Constitution? Not surprising. But depressing that US citizens can’t even think through some basic civics, or, worse, despise them

    • Matt McIrvin

      This particular comment is an interesting reminder of the stakes in this election.

  • Michael

    Is there reliable data on the minimum percentage of any one, or a combination of, the African American, Latino, and Asian demographics a candidate would need to win the presidency?

  • coolchien

    Hi Sam,

    I wonder if you are familiar with David Rothschild et al’s work on the mythical swing voter. The gist of it is that they believe a large portion of the swing in polls as results to campaign events are not swings in voter intention but rather fluctuation in polling participation. This would bolster your view on reversion to mean as well.

  • Amitabh Lath

    How does this volatility compare to post ’96 Erikson and Wlezien data? Does the assumption of increased recent stability that you use in the November projections still look ok?

    • Sam Wang

      Yeah, everything is within normal limits so far. Violations would be if national margins or meta-margins stayed at or below 0% for a while. Zero percent (a tie) is approximately two sigma.

      I actually think we are only around 1 sigma at the moment, given my assumptions…will check later…in Boston at the moment

  • Nathan

    Just curious: any evidence of a regular “teacher effect” in polling? Following Frank (above’s) suggestion of a similar pattern in September 2008 and the idea that voter swings are often about non-response, I imagine the biggest group suddenly overwhelmed into no longer responding to surveys in September might be the 3 million + teachers in the USA, who probably lean Democratic. I’d love to see evidence to test this! Is it out there?

  • Trump+Democratic Congress?

    Hello: I have no serious math training. A friend told me that this is the place to come to found out the odds of something happening in a US presidential election.

    My preferred outcome is Trump winning the presidency while the Republicans lose the Senate and the House. Is there any credible scenario where that could happen?

    • Sam Wang

      No. The Presidential and national House vote are highly correlated. If Trump wins the election, surely Republicans will hold the House.

  • WRGerman

    Sam, who exactly _are_ the majority of undecideds? I have a tough time imagining that African-Americans, Latinos, Asians and women comprise the majority of these.
    Demographics matter more every year, and that point was brought home to me in 2012, when President Obama only got 39% of the white vote.

  • Ed Wittens Cat

    I realize this is probably the Aspergers talking but I completely dont understand why we cant discuss red/blue brain hypothesis and the core of “unpersuadables”.
    number dems >> number repubs
    but
    repub enthusiam >> dem enthusiasm
    Trump (with the aid of the media) has normalized white nationalism, racism, xenophobia, and gun-fever for his base. His base is EXCITED abt voting for him.
    Clinton has alienated the youth/millenial part of hers, the only part with equivalent enthusiasm to repubs.
    The DNC leaks, the choice of VP, Clinton is treating this segment of her base as if we were republicans–that we will “come home” if we have no choice.
    This demographic has proven historically that it will vote for a Nader.
    Thats why she may lose– the enthusiasm gap– and thats why we dont care.
    Its stupid to try to bully a critical section of the base– Trump tried with GOP intellectuals– but they just defected, and there arent enough of them to matter.
    But youth carried the election for Obama.
    Sending Obama out as a surrogate may not work.
    its not possible to force enthusiasm.

    • Partha Neogy

      I agree that Clinton is a cautious candidate and has opted for a traditional rather than an innovative/disruptive strategy (choice of VP, etc.). Millennials may not be enchnated with her, but sitting out this election or voting for Trump would exact a terrible price over the next four years at least, probably much longer. Sometimes the choices presented are not the best: only better and the worst possible.

    • Ed Wittens Cat

      One more thing– I think cloud connectivity lends itself to increasing polarization– the pivot to the middle may not work anymore– pretty obvious that Trump is just headfaking pivots that fool no one except the media– which is held in babylonian thrall to Trumps every word.
      red/blue polarization visualizes to me as analogous to simulated annealing– bulwarks of unpersuadables on each side.

    • Commentor27

      Partha,

      I’d suggest that anyone who truly dislikes both candidates decide who to vote for based upon who’d they rather pick Supreme Court justices.

  • Mark

    I love reading this site and the majority comments which are very thoughtful!

  • Graham

    Dear Sam:

    In my travels and encounters this year, I have met and noticed numerous individuals who in past elections were politically apathetic and could not have been dragged to watch a presidential debate, yet this presidential election cycle they are completely engaged in the presidential election. Many of them are Trump supporters (primarily white) who have never before been politically active, yet now they zealously post on Facebook, attend rallies, etc., in support of their chosen hero. They are looking forward to the presidential debate as sort of a heavyweight prizefight when in past years they would never have even tuned in to watch a presidential a debate.

    My question is, what effect, if any, will this phenomenon have on 2016 presidential polling and predictions? Can it be measured statistically and, if so, is it statistically significant?

    Thanks, and I love your site.

  • Jaymes Winn

    Looking at this NBC article: http://www.nbcnews.com/meet-the-press/clinton-losing-key-millennial-support-nationally-key-states-n650076

    She’s bled a lot of millennial support between August and today, yet I can’t think of any significant things that would cause them to drop their support of her (I could be misreading what they care about).

    What do you think of the idea that millennials mostly started going third party because they assumed Clinton had the election won? In early-mid August she had those gargantuan post-DNC leads, I can see them going “oh well she doesn’t need me, I don’t have to hold my nose, I’ll vote Johnson/Stein and actually feel good.”

    If that’s the case, I think they might be back if they notice the race being very tight again.

    • MarkS

      The third-party-voting millenials that I talk to are mostly concerned that the person they vote for represents their values as closely as possible. The practical outcome of their votes is not their highest priority.

    • Commentor27

      MarkS, I guess young people have to make mistakes to learn from, don’t they?

  • DonC

    I wonder if the GOTV efforts will affect polling accuracy? Obviously at the national level not at all, but in battleground states like Florida it might be significant. Clinton has invested heavily in GOTV and Trump has not done much. Assuming that GOTV moves the needle on turnout, you’d expect the electorate to be slightly different in 2016 than in 2012. Given that pollsters have to make decisions on turnout when normalizing their samples, and that assuming the electorate in 2016 would mirror 2012 would be the way to do that, the polls could be underestimating Clinton.

    My SWAG at why the national polls show a narrowing race is that whites in heavily Republican states have returned to the fold. Not entirely sure that college educated whites in swing states — for example white women in the Philadelphia suburbs — will move in the same direction.

  • Remi

    So the cake isn’t baked after all.

  • JayBoy2k

    Having paid attention to PEC for a number of cycles, I still believe that the cake is baked for a Clinton victory in November. She is at 1 to 3 % point ahead at the bottom of her range… and that is coming from someone who will not be voting for her. I am adjusting my thinking a little on the Senate races and would love to hear data and thoughts about ticket spitting since that could be a factor this time around.

  • Abe Fisher

    Sam–can you comment on the analysis contained in this article? Basically it suggests that the main stream media, looking for a horse race in order to boost ratings, has an incentive to do whatever they can to make the election look closer than it truly is…thoughts?

    http://www.dailynewsbin.com/opinion/yes-hillary-clinton-is-still-winning-and-yes-the-media-is-lying-to-you/26054/

    • A

      I read this article as well as a similar Forbes article by a pollster who said there has been little “real” movement in weeks between Trump and Clinton.

      Both have been polling consistently in the mid 30s to low 40s with Clinton slightly ahead, and the only time it changes slightly where she appears to go down a bit in those margins and Trump rises slightly, it makes it look like a big change when there is none.

      Is that really all this is?

      If so, I am a fool for believing the hype.

    • Jaymes Winn

      Even as a Clinton supporter, I’ve found it beneficial to steer clear of Daily News Bin, Blue Nation Review, ShareBlue, PoliticsUSA, Crooks and Liars, etc. These are pure echo chambers with no substance.

      Talking Points Memo, Daily Kos, etc. are some liberal blogs where you get analysis with from the left’s perspective, but they also admit when Clinton makes a mistake or that she might lose Ohio or something.

    • Kevin King

      Echoing Jaymes, I’m a liberal and a Clinton supporter, but I would take anything from the Daily News Bin with a grain of salt. They and other places like Blue Nation Review and some others are more or less propaganda sites. I don’t even read Daily Kos or Talking Points Memo very often. The Washington Post and The Atlantic have had excellent coverage this year. This site has meaty analysis, and personally I like to read FiveThirtyEight’s coverage as well.

    • Jeremiah

      http://www.electoral-vote.com/ has some of the best coverage I’ve read.

  • Michael Levinsohn

    Sam, how much, if any, weight do you give to the concept of “differential response,” the idea that when the news is bad for a candidate, their supporters tend not to respond to pollsters?