Princeton Election Consortium

Innovations in democracy since 2004

Outcome: Biden 306 EV (D+1.2% from toss-up), Senate 50 D (D+1.0%)
Nov 3 polls: Biden 342 EV (D+5.3%), Senate 50-55 D (D+3.9%), House control D+4.6%
Moneyball states: President AZ NE-2 NV, Senate MT ME AK, Legislatures KS TX NC

Polarized goggles

October 29th, 2016, 3:41pm by Sam Wang

While the journalistic herd runs after the Comey/email story, I am out of step with information that is more likely to stay true after the weekend. Today in the New York Times, I ask: despite the emotional nature of this year’s race, why have polls of the Clinton-Trump race moved so little?

The answer: polarization that took root in the mid-1990s, based on my previous examination of election statistics. Rapid communication made it possible for a national-level faction to form: the Trumpist wing of the Republican Party. This faction, first nurtured by Rush Limbaugh, Newt Gingrich, and others, has gotten ever more detached from traditional governance. With Trump, it has reached its culmination.

Today’s example of polarization is the news about Huma Abedin’s email, which it seems she accessed on a laptop she shared with her now-estranged husband, Anthony Weiner’s. Clinton supporters think it is a minor event with weird timing, which has blown up in the press. Trump supporters think it’s a game changer that can save their candidate.

One thing is almost certain: it will not affect Clinton’s lead substantially. Basically, the press needs a new direction, even if conditions haven’t really changed. I could imagine secondary effects downticket, though. Based on polls, Senate control could go either way. Even a small change in turnout could affect the outcome. Same story in the House, although partisan gerrymandering there seems likely to keep that chamber in Republican hands.

Worried? To improve your health, get out there and do something. Here in New Jersey, it’s a beauitful day – perfect for getting out the vote. Bucks and Montgomery Counties in Pennsylvania are especially juicy – 8th Congressional District, plus a tight Senate race. Farther away? Use this Sharon Machlis’s Competitive District app.

Tags: 2016 Election · House · Politics · President · Senate

89 Comments so far ↓

  • Jax

    Being from the UK it’s difficult to contribute, but I’m hoping sense will prevail and people will vote, and take their vote seriously!
    Thanks for the blog and analysis it’s been fascinating

  • Ed Wittens Cat

    awesome article, Dr Wang
    but given there are 2% less non-hispanic caucasians in the US every election cycle since 1988 doesnt the “conservatives” current strategy mean demographic doom for the GOP?

    • Bob Wallace

      College degrees either make people more liberal or give them job opportunities that the “White non-college” group in Sam’s graph don’t have. Probably some of each.

      The college educated percentage of the population is also increasing which should also erode the Republican base.

      Add to that the growing number of number of Americans who have ‘no religious affiliation’ and the less conservative (racial/environmental/homophobic) nature of young evangelicals and one sees all sorts of decline sources for the Republican base.

      As it is constituted the Republican party is an endangered species. It’s “breeding” significantly below replacement rate.

      And since the core value of the party is to have an America run by white Christian males it’s very hard for them to share power with those outside their group. Sharing would mean changing their goal.

      The question to me is how much will the current Republican party have to shrink until we see a reform movement that builds a new party either under the current or a different party name.

      In the interim we may see the corporate wing (the least ‘non-political’ part) move to the Democratic party because that is where federal power will concentrate.

      Most interesting times….

    • Josh

      On the surface, yes, it seems like a terrible strategy. And I largely agree with you and Bob. But there are some mitigating factors:

      -Overall, the US may be growing less white, so to speak, by 2% every cycle–but our electorate is not; some of these new non-white Americans are too young to vote, or are illegal immigrants, or are legal immigrants who can’t yet vote, or are new, voting-age citizens who choose not to vote. The upshot is that the electorate is becoming less white by more like 1-1.5% every 4 years. (The electorate was 85% white for Reagan in 1984 and will be about 70-71% white for this cycle, a change of 14% over 9 cycles.) It’s definitely a trend, but a slower one.

      -Many of these new non-white voters won’t automatically vote for a D candidate. Over the past dozen elections, for example, R presidential candidates have received roughly anywhere from 20 to 40% of the Latin-American vote, and non-trivial portions of the Asian-American vote as well. (Don’t forget that as recently as 2004 George W Bush won almost 45% of the Latin-American vote.) Incidentally, these are the two fastest-growing non-white demographics in the electorate.

      -Although highly educated whites, along with most non-whites, now tend to vote D, there’s been an opposing and almost equally large trend among lower and middle-class whites, especially those without much education–many of these voters would have supported D candidates in the past, but now overwhelmingly support R candidates at the national level.

      So basically, yes, if nothing changes–if the GOP doesn’t re-tailor its messaging, if its base continues to want the same kinds of candidates, if Democrats continue to do well with the same demographics–then over time, at least at the national level, it will become increasingly difficult for R candidates to win elections. That said, and assuming nothing changes, this built-in advantage for D candidates will probably continue to build slowly–that is, over decades, rather than in just a few cycles.

    • Ed Wittens Cat

      No Josh, sry ur wrong.
      Age is another axis of polarization and the olds will die off eventually– havent seen any data on the median age of Trump supporters but the rallies look pretty old.
      A tribe without reps cannot survive.
      And around 2040, the death cross– when declining number of whites crosses the ascending curve of hispanics.
      By 2020 there may be even less than 41% non-hispanic caucasians with conservative alignment.

    • WildIrish

      Josh, you are forgetting one major factor from this election. The “nasty women” who have left the R party in droves, or remain R’s but will vote D now. This is bigger than just HRC vs. DJT in 2016. The GOP didn’t stand up for women against DJT, and they should have. Those R women won’t forget any time soon, and many won’t be going back to voting R now that they have seen the light. Women are 51% of the voters, and not likely to diminish over time.

    • Gelatinous Cube

      As Josh suggests, whether the Republican Party is demographically doomed depends on whether it continues to set itself against virtually ethnic group other than white Christians (and, to an extent, Jews). If they can escape that ethnic cul-de-sac, I think the GOP can continue to exploit cultural conservatism and family values in the service of its other core constituency: anti-tax, anti-regulation business interests.

    • Bob Wallace

      Josh, IIRC there is data that shows that if someone votes once they have tendency to vote in future elections.

      In 2008 and 2012 African-American turnout was very high. Most likely a lot of first time voters who will be return customers. And more new AA voters are likely to show up this year due to anger over Republican voter suppression attempts.

      From what I can see it looks like Trump is going to bring a lot of Hispanic voters to the polls for the first time.

      2% less white every cycle + much higher voter participation on the part of non-whites.

      And the non-white factor will become a player in some large electoral college vote states. Texas (38 votes) could soon flip. Arizona (11) is teetering. Georgia (16) could flip in 2020 or 2024.

      Montana (3), North Dakota (3), Oklahoma (7) and similar states are going to be the states that stay white-majority longest.

      Not every black, Hispanic, no longer religious, college graduate person is going to vote Democratic. But they tend to at a higher rate than they vote Republican.

      Unless Republicans can come up with a widely admired candidate, a Dwight Eisenhower, I suspect their slide into oblivion is going to be more rapid than 1.5% per year.

      If we look back at the elections when Republican presidents were elected post Eisenhower it runs like this –

      Nixon = 0.7% of the popular vote
      Reagan = 9.7%
      Bush 1 = 7.7%
      Bush 2 = -0.5%

      Democrats have won by similar low popular vote margins. It wouldn’t take a lot of movement from one side of the dividing line to the other to make large and long term changes.

    • Josh

      Thanks for the thoughtful responses. I’ll respond to a few.

      -I think one of the consequences of this level of polarization is that who parties nominate as candidates for the presidency simply doesn’t matter as much. Barack Obama in 2008 was “a watershed moment in our nation’s history” and Hillary Clinton in 2016 is “a deeply flawed, crooked demon”…and they will each probably win between 50% and 53% of the vote in their first general elections. So the idea that the GOP needs a “Dwight Eisenhower” to right the ship misses the point, I think–they COULD nominate the second coming of Ike and Ike II probably still wouldn’t get more than 52% of the vote.

      -I’m not so sure that all these “‘nasty women’ leaving the party in droves” are really leaving the party…or that there really are droves of them. Again, our country is–wait for it–so polarized that very few people are actually switching sides (a recent survey had close to 80% of GOP women still strongly supporting Trump). Some of the remaining 20% of GOP women may sit this one out or vote for a third party; perhaps a small number (likely suburban, well-educated) will vote for Hillary. I suspect that if the GOP nominates anyone more electable than Donald Trump in 2020, many or most of these same nasty women will vote for him or her.

      -Some minority groups vote at very high levels relative to the general electorate–black voters are a good example of this. Conversely, however, other minority groups lag the general electorate in terms of their engagement/enfranchisement; I recall as recently as 2008, something like less than 50% of voting-age Hispanics in Texas were registered to vote. Sure, voting once is a predictor of voting again, but the idea that many new non-white Americans will magically become enfranchised and consistently vote (and vote D!) sounds a bit overly optimistic to me.

      -Finally, call me crazy, but it’s hard to take seriously the notion that indisputable proof of the GOP’s supposed age problem is that it looks like everyone in a clip of a Donald Trump campaign event might be kinda old.

    • Ed Wittens Cat

      Apolos Josh, but i said pretty old, not kinda old.
      Like Dr Wang says, data over drama.
      One axis of polarization is age– millenials are trending dem while boomers and silents are trending repub–>
      I was making an empirical observation– most Trumpkins observed at rallies (except for Pepe le Frog & the alt right) are middle aged or above. Maybe its because these people dont have jobs, and maybe they dont have jobs because they dont have college education? i dunno.
      But the widening chasm is the college-education gap– college education is increasing in US, but not among the olds.

    • Sam Wang

      I believe much of that trending arises from the fact that younger voters are ethnically more diverse. Millennial whites are not more progressive than their older counterparts. See this Politico article and this research by Vincent Hutchings.

      Therefore if we want a small number of dimensions for describing demographic change in partisan affiliation, the dimensions are (a) race, (b) religion, and (c) education. Though now that I think of it, the rate of going to college has changed. I wonder if that is enough to estimate how much of this effect could arise from people self-sorting.

    • Ed Wittens Cat

      I had thought i was more cynical by now, but im still a starry eyed educationist i guess…
      like ur discussion with AL on the racial divide of evangelicals.
      Old people do not go to college in large numbers, while attending college brings young people into contact with racially diverse students and professors.
      this is all i meant, going forward–
      “while the age profile of the GOP is growing older more quickly than that of the country.”

    • Sam Wang

      Your explanation of what college accomplishes is plausible. I’d like to think that what I do for a living has lasting effects on people.

      However, it is also the case that (a) information can be used to reinforce a person’s pre-existing views, and (b) a person could want new ideas and perspectives and therefore choose to attend college, which then amplifies a pre-existing tendency. Think of it as an interaction between innate and environmental mechanisms.

    • Ed Wittens Cat

      truedat– Complex adaptive systems dynamics of societies–
      In the 21st century there is a ferocious push to acquire college degrees (perceived as the only way to succeed in the current economy) so that exacerbates the GOPs problems– and also, we talked about this before, Trumpism is distinctly anti-intellectual– the GOP “eggheads” like George Will have left the building.

  • Sebastian

    The so-called polarization is applicable to only certain types of candidates: corporate clone candidates. Clinton and Trump are virtual (mutated) clones of Obama and McCain/Romney. The populace is not the origin of the Blue D & Red R polarization, the candidates are the origin, and behind them … the de facto Business Party, with its two wings.

    Replace Clinton with Sanders … and there would be no such polarization. Poof. The country would be overwhelmingly for the populist New Dealer.

    It’s not that the books are cooked, or the fix is in, with the populace … but with the candidates. The CC candidates are superficially polarizing. I think it’s important to note that this polarization is a mirage seen only when the populace is forced to look, choose, act through the lens of CC candidates only.

    Everyone wants to whip up on the populace for being calcified. No, it’s the choices that are calcified. And the choices are absurdly unpopular and artificially limited, that is, phony. So to is the supposed polarization phony, once one gets below skin deep.

    The Business Party narrative is not very wide and is very carefully controlled with enormous sums of money and other carefully constructed levers of power. The age of the internet has allowed this control to both intensify and calcify. It has become blindingly obvious and very brittle. Sanders almost cracked it. Four years from now – three, really – could be very interesting, when the ostensible polarization should be on its way to the dustbin. Eight years at the most, one would think.

    • Michael Coppola

      Sorry, but the country would most certainly not be overwhelmingly supportive of a self-described socialist running for POTUS.

      The polarization may be a top-down phenomenon, but is absolutely real. I live and work with many many people who sincerely believe all of the right wing talking points, especially the ones that are demonstrably false.

    • Harry

      I know what you are saying sounds good, and seems to be steeped in some kind of “outsider” doing good philosophy. But it is all “blue sky” stuff. If there was really someone who truly was this independent, non beltway type who would it be? Why hasn’t someone like that form a third party that’s been successful? Even George Washington and the founding fathers, who we consider revolutionaries,were wealthy property owners. The type of person you describe could never lead. Why? Because he would need a congress that believed in him or her to get anything done. So in actuality you need to build a congress first, before a so called “independent” thinker comes to Washington. Also, even parties that seem to be independent of business at first, soon learn that you need money to run a campaign…and where do they turn…business. In order to influence votes..where do they turn.. you got it. So, although you think the government is run by big business, and to some extent it is, in the final analysis they are just people trying to make the system work. If you really want the change you are speaking of, we need to change to a parliamentary system of government. That’s not going to happen anytime soon.

    • Josh

      What you are missing is that our current choices are “absurdly unpopular”, as you put it, precisely *because* of polarization. If people didn’t already know exactly how they felt, and were completely inflexible with regard to choosing political parties, they wouldn’t feel so deeply negative toward candidates from opposing parties.

      If you still can’t see what I mean, just think about it this way: if our current choices really are that bad, why is it that somewhere between 90 and 95% of voters will still cast a vote for one of them?

    • Kate H

      “The country would be overwhelmingly for the populist New Dealer.”

      Another hallmark of polarisation is the belief that “the country would be overwhelmingly” for *my* partisan choice, and that other choices are aberrant.

    • JamesE

      Gee, I thought that the reason Clinton and Trump are the major party candidates is because they got more primary and caucus votes than their opponents. Clinton got 4 million more than Sanders and Trump got 6 million more than Cruz.

    • Taylor

      Sen. Sanders is a good man and would probably make a very good president, but if he had won the Dem nomination, the GOP would have been constantly screaming “SOCIALIST,” and then in a more subtle way, they would have made sure to remind everyone that Sanders is Jewish. It’s very sad to say, but yes, that would have been a problem. A friend of mine canvassed for Gore in the Midwest in 2000. People told him outright that even though they liked Gore, they refused vote for him because he chose Lieberman as his running mate — and that was long before Lieberman went crazy.

    • Robert

      Sure, let’s say you elect Sanders. Better pray the Dems take congress or it’ll be four years of nothing happening. I’ve always admired the energy and enthusiasm of the Bernie folks and always wondered why they would repeatedly gloss over this simple reality. If a moderate Dem peddling semi-GOP ideas (Romneycare anyone) can’t move congress, a full throated socialist would only meet with greater resistance.

    • Linda

      I’m not sure that you understand that Sanders is also a corporate senator. Despite his calls for change, he’s supported almost every bill on war financing, voted for financing of several planes the pentagon did not want ( they were partially built in Vermont), sought to increase cocaine penalties to the same level as crack rather than lowering crack penalties. He does have a perfect record on some things- he’s never hired a woman or a non-white male into a leadership position.

    • FlyingSquirrel

      I’m not too sure about that. While Clinton and Trump are both more corporate-friendly than Sanders, the electoral map hasn’t changed a whole lot since 2000. If you look at the candidates nominated across these last five election cycles – Gore, Kerry, Obama, Clinton / Bush, McCain, Romney, Trump – there are some noticeable differences in personality and policy there, but the shifts have mostly been on the order of a few points here and there.

      After 2004, I remember thinking that perhaps the map didn’t change much because Gore and Kerry were somewhat similar candidates in terms of strengths and weaknesses – i.e. smart policy wonks who weren’t great communicators or “salesmen.” But seeing that same map one election after another with only a small number of states changing hands makes me think that it really is more about party identification at this point.

      Maybe a Sanders vs. Trump election could have really scrambled the two parties’ coalitions. But maybe not.

    • Greg

      I don’t agree with the original post, but I’m inclined to think that Sanders, Biden or Kaine would be ahead at least 7-8 points right now, and possibly a few points more than that. I think Sam’s right about polarization. But I also think that the degree of polarization in this election is exacerbated by the flaws of the two candidates. Many in the public — absurdly in my opinion — view Clinton and Trump as equally bad. Also, people are hungry, even desperate, for change. Clinton represents the status quo like no one else. Trump supporters are willing to look past a lot with Trump because he says he will change things. (They overlook the fact that change can be horrific.) Out of this desire for change, and because his economic message resonates, I think plenty of people would have looked past the oppo research on Sanders too. Basically, if the Dems had nominated any competent politician not named Clinton, I don’t think any serious person would be doubting the eventual outcome at this stage. The Senate would be in the bag, and we’d be talking about their chances at winning the House majority.

  • M. Leo Cooper

    Still over a week to go, and plenty of time for new developments. Maybe even a couple of November surprises. We’ll see. I don’t know that Trump really truly wants to be president, and perhaps the prospect of it will tip him over into some of his trademarked self-destructive behavior.

    • TeddyVienna

      I’ve wondered that, too. Despite his denials, it seems clear that “Trump TV” is the endgame. So would it ruin his plans if he won?

      But you’d think he would back off by now, just to make absolutely sure he doesn’t win.

      If elected, he won’t serve four years. No way. He’ll either be impeached or just get bored and step down.

    • Michael Coppola

      I’m pretty sure that Trump is planning to do Trump TV if he wins. How else could his administration disseminate the truth to the masses? Can’t expect the rigged media to do it.

  • GC

    Sam, while changes in communication and travel since the early republic are responsible for some of this, I think you miss Madison’s (and before him, Montesquieu’s) point about the chief benefit of a federal republic: the many states provide a bulwark against consolidation (and tyranny) by retaining all powers not explicitly delegated to the federal government via the Constitution.

    The Supreme Court’s expansive interpretation of the Constitution, federal income taxes, the Seventeenth Amendment, Federal Reserve, “Chevron deference,” etc., etc. all contribute to the federal government’s (especially the executive branch) ever-increasing involvement in our lives, usually at the expense of the states.

    I bring this up because consolidation in D.C. means that national elections become these pitched “do-or-die” battles since so much is at stake — much more, in fact, than the Framers intended.

    • Scott Matheson

      I’d dispute your last paragraph – I’d say elections were higher pitched at the beginning of the republic, with more at stake than our situation now (and the founding fathers were more comfortable with high stakes democracy than the current generation).

      When the political question of the day is “shall we have a hostile relationship with empire A or empire B” when hostilities with either could result in anything from attacks by allied Native Americans well east of the frontier to loss of territory to the complete disappearance of the state, the stakes are higher than any question we face today short of “total nuclear war?”.

      And moving forward we had the Jacksonian era when among other things the franchise itself, not barriers to its exercise, were up for debate, and then the midcentury debate over slavery and its associated territorial integrity questions; it’s only once we reach the post-reconstruction era that the social contract is settled enough for elections to be considered low stakes, and the expansion of federal involvement at this point is probably not entirely coincidence.

  • Shawn Huckaby

    Such a terrific article, and already being shared widely, at least by me!

    Looking at the demographic trend lines on the chart above, it certainly points toward increasing irrelevance for the party placing all of their electoral eggs in the red basket.

  • Mark Lawrence

    Sam, your analysis of the cause of the polarization of politics is interesting, but do you have hard data to back up your contention that the right wing media is to blame? There has been a resurgence of right wing parties in many Western countries in the last 20 years. In many cases this seems to be driven by fear of immigration. Look at Geert Wilders, or Le Pen, or the Brexit supporters. Racism was a very big part of the Brexit movement.
    Now, Limbaugh and his ilk obviously represent a nativist and racist viewpoint. I’m not sure if changes in media are as important, per se.

    • Jay Sheckley

      Polarization is measurable and Sam and others have shown the phenomenon on various charts and discussed here and on his Woocasts.
      By blame do you mean cause? By ilk do you mean allied? Your chosen phrasing is perhaps unhelpful.
      Dr Wang is a scientist and statistics expert. I’ve been meaning to ask him how many decades he’s been working to clarify the meaning of numbers.
      Today’s chart speaks for itself.

    • Phoenix Woman

      Remember that before he worked for Rupert Murdoch, Roger Ailes was doing much the same thing for Jack Welch at NBC.

  • Hem13

    This is the perfect example of salient factual analyses which makes PEC such an invaluable resource. I have voted absentee ballot from Australia and made financial contributions to Clinton’s campaign and a few Progressive candidates down ticket.
    I hope all those who appreciate the assiduous efforts and sound reasoning of Prof. Wang and his colleagues will heed his sagacious advice to get involved in support of your preferred candidates.

  • Harold Bridges

    Sam, I think you have missed the big picture here. Poll stability arises out of political polarization. The big fact behind political polarization is the rise in income inequality beginning the in 1970’s, but accelerating with the Reagan Revolution, i.e. lower taxes on the rich and anti-unionism for the workers. From that point on the working class faced stagnant or declining real income. Up until about 1980 the median income rose in lockstep with gains in productivity, but from that point on the upper income segment succeeded in sequestering virtually all of the productivity gains.

    It’s a material world. Voters certainly have feelings, but in the aggregate and in the long term, we can expect them to be driven by their economic interests or, at least, their own understanding of those interests.

    The other factors, such as changes in communication technology, play a subordinate role. Other countries that experienced the same changes in technology, such as Germany, have not experienced political polarization comparable to the US. But then, while the income share of the upper 1% in the US grew from 9% to 24% during roughly this period, the share of the German 1% in their own national income grew only from 8% to 9.5%.

    I realize that the focus of your piece is the current election, but I think that discussions of political polarization that treat it as a first cause, not requiring explanation, are a disservice.

    • Sam Wang

      I rather doubt that communication technology plays a subordinate role in the transmission of political ideas. Note that the United States has larger physical distances than in those other nations.

      Also note that the U.S. has a rich variety of measurements of party uniformity, geographic voting patterns, and subgroup measurements, allowing a closer look at the time course of polarization than any other nation. They point toward a series of events that led to a sharp uptick in the stability of elections increased sharply in the mid-1990s. Your story is interesting and of importance, and I have written this season about the 1964 election and about the 1932-1964 run-up (see my podcast with Eric Schickler). These slow trends do not have the right time course to be directly relevant to what we have seen over the last 20 years.

    • Periwinkle

      If other countries haven’t experienced the shift that the US have, we might ask if there is any difference in their broadcast news.

      The US in an outlier in its minimal support for PBS and NPR. I am aware of the idealogical reasons for this, but the outcome is that most other first world countries provide much many more resources to their public broadcaster. Am I right in thinking that most Americans have probably watched more hours of BBC programming than PBS?

      Public broadcasters can still behave badly – producing low-budget “reality” TV shows, executing cheesy ratings stunts, or acting as a propaganda arm for the sitting government. But at their best, they are a standard to measure commercial stations against, and even at their worst they are an example of where not to go.

      I think that Sam Wang is right, that US politics are uniquely polarised at present. I don’t have hard numbers for this, but I also think that the reason other countries have not divided badly despite the same technology being available is because of greater support for public broadcasting in those countries.

      I can think of some experiments to test this hypothesis, but they wouldn’t be ethical to run.

    • Igor

      This is an interesting idea. Sam’s point is still valid, though. Perhaps one other important thing of difference between other countries and US is the organization of political system. It would be interesting to see what would happen if US was a parliamentary democracy.

    • Sam Wang

      Consider governments that allow nationwide proportional representation. Israel comes to mind – and there we have multiparty representation.

    • SoddingJunkMail

      Mr Bridges:

      “It’s a material world. Voters certainly have feelings, but in the aggregate and in the long term, we can expect them to be driven by their economic interests or, at least, their own understanding of those interests.”

      This is a tidy thought, but if true, how to explain the large swaths of rural and evangelical voters who consistently vote against their economic self interest?

      I suspect that you’re projecting your voting motivations onto the general public, not observing how they actually behave.

  • Ed Wittens Cat

    I feel like im caught in a time loop…i hadnt thot abt Trende or Cost or the rightosphere in 4 years…so i checked their TLs– they are all still doing the exact same thing! Which is pushing the same magical thinking and flawed reasoning as in 2012.
    Its like they learned nothing from Romney’s defeat.
    And Nate is still wedging extra uncertainty into his analysis…
    plus ça change, plus c’est la même chose

  • Matt McIrvin

    The reporting on polls is amazingly bad–that one ABC/Washington Post poll that showed a 10-point drop in Hillary Clinton’s support is getting more press than every other poll in the world combined, because it shows something none of the others do.

  • anonymous

    It is still possible for candidates to outperform their expectations determined by polarization. See Republican Larry Hogan in Maryland, Republican Charlie Baker in Massachusetts, or Democrat Evan Bayh in Indiana. Polarization is not electoral destiny.

    • BFreeman

      From where I sit, as a Massachusetts citizen, it seems to me that Baker’s success was more likely the result of his opponent’s weakness. For example, consider the fact that his opponent was Martha Coakley, a Democrat who was also responsible for losing a Senate seat to the Republicans in 2010. The election of Charlie Baker was the second time Coakley ran for statewide office and lost to a Republican. In between those two contests, the people of Massachusetts elected both Democrats Elizabeth Warren and Ed Markey to the Senate.

      In other words, it seems to me that the issue was not Governor Baker outperforming beyond the expectations of a Republican in Massachusetts so much as it was a single weak candidate losing both races for the Democratic party.

  • Matt McIrvin

    How does the polarization hypothesis square with the higher number of undecideds this year? Are these decided but shy voters, or decided non-voters? There are twice as many as in 2012, which I would not expect if polarization is higher.

    • Sam Wang

      A consequence of polarization is poor approve/disapprove ratings for both candidates. In such a circumstance, one might expect more undecided/Johnson/Stein/McMullin voters.

    • Some Body

      OK. So then, why so few undecideds in 2012?

      And as for this latest (non-)scandal, I’d expect it to cut Clinton’s lead more than trivially, but mostly because newly-encouraged Republicans (and Republican leaners) would flock to Trump.

  • Matt McIrvin

    I do think Clinton’s lead is almost certainly going to be dropping from here on in, because the Comey story may be about nothing, but it’s nothingness structured with a lot of the kind of who-knew-what detail that lets reporters role-play Woodward and Bernstein. That will keep generating at least a week’s worth of headlines which are not about Donald Trump. Every time in this campaign that that has happened, whenever it’s not about Trump being awful, the numbers slide toward a partisan mean.

    The thing that will probably save Clinton is just that there’s not much time.

    • EDYS

      Matt, I hope you are greatly underestimating the ire this has generating among Clinton supporters. I have voted early. I have NEVER voted early. I am spending Election Day in NH driving people to the polls. I have NEVER been part of ANY candidates GOTV campaign. Not even Obama. Who I adored. Much has been made of Trumps supporters, but zero attention has been given to Hillary’s. We are pissed, engaged, and very very active.

    • SLH

      I too had never volunteered for a political campaign until this year. My wife and I are calling registered Democrats here in Arizona. I figure if I’m doing it for the first time, there must be many others.

    • Sam Wang

      Comments like SLH’s demonstrate the polarized nature of politics: on the Democratic side, this news is seen as a positive motivator to get out and vote.

  • bks

    I predict that one year from now Comey will not be the Director of the FBI. Probability > 99%.

  • Amitabh Lath

    Can someone please rigorously define white? The definition seems fluid and unhelpfully broad. At some point Irish and southern european catholics were not, but now they are.

    What about Arabs, Persians, Turks? And peoples who hail from the region between Russia and Iran, bordering the Caspian sea (ie, the “real” Caucasians)? Are they white?

    In the absence of a proper definition of this cohort, everyone conjures up a “typical” white person, probably a caricature that fits with whatever point they are making.

    I expect we will eventually break this overly large group into manageable subsets like “Minnesota Scandinavian” and “Appalachian Scotch-Irish” etc.

    • Sam Wang

      Yeah, many immigrants get absorbed into “white.” For example, it has been argued that the political effects of boom in the Hispanic population will be offset by their absorption into the “white” category.

    • bks

      There are no rigorous definitions in natural languages and precious few in the life sciences. That’s one reason why it is the job of rhetoric to convince people, not that of logic.

      In physics photons are interchangeable, but in biology each cell is unique.

    • Olav Grinde

      I sometimes describe myself, especially when speaking to a right-winger with a racist bent, by saying: “I am Afro-European – just like you.”

      Meaning I am European by birth, but that my ancestors migrated out of Africa, perhaps Olduvai Gorge, some tens-of-thousands of years ago. Unfortunately, Afro-European is not an ethnic category recognized by the American Department of Homeland Security.

      And while we are on the subject of “race”, when is an American “black” person no longer classified as “black”? Half? Quarter? One-eighth? I am puzzled…

      I recall a great Doonesbury strip that appeared right after Barack Obama was first elected. It showed two soldiers, one black and one Caucasian, sitting on a hillside in Iraq, and the latter insisting on including Obama in his own tribe says: “He is half white, you know.”

      I do think census reports should have an additional option:

      – Human being. Never mind the details!

    • Amitabh Lath

      BKS, ok maybe rigorous is not possible but the point remains that “white” is an overly broad category that impedes analysis. And growing this category by adding Hispanics as Sam suggests might happen, will not help.

      As an example, one could say that among South Asians there is some (very small) support for Trump. But if you broke that down into Pakistani, Indian, Nepali, Bangladeshi, Sri Lankan, etc. you find that most (all?) of that support comes from Indian Hindus originating from the north-west part of the country, who believe Trump will punish Pakistan.

      In the absence of such sub-group information you might conclude that somehow this small fraction of South Asians were with him on taxes or charter schools or anti-LGBT social issues. But that would be incorrect.

    • Davey

      The question is whether fine-tuning our definition of racial groups would provide useful data. I’d think that the more granular we go, the less reliable the results would be…and we’d still be totally dependent on respondents to understand and self-designate appropriately.

      Also, I wonder if our dependency on political analysis through ethnicity is about to crumble anyway. According to PEW, 7% of the nation is now multi-racial. At what percent of multiraciality in the population does the resulting polling data become too corrupted to tell us anything significant about voting habits by enthnicity?

    • Sean Patrick Santos

      This relates to an issue I’ve wondered about from time to time. In interpreting polling data, we often assume that if self-reported race correlates with a political belief, that this is because race plays some causal role in the formation or rejection of that belief. This is probably true in most cases, but it’s also quite possible for someone’s culture and political beliefs to affect the way that they identify themselves racially.

      For instance, there was a Pew report on multiracial identity in U.S. with some interesting conclusions. It found that people with mixed white and American Indian background (defined by whether they could identify at least one parent or grandparent from each race), the majority considered themselves to be white, not multiracial or American Indian, and they also tended to identify with Republicans more than Democrats. One possible explanation is that living in a conservative white community both discourages people from identifying themselves as non-white, and fosters conservative political beliefs.

  • Bill Moore

    Great article in NYT today Dr Wang. And such a simple but telling graphic. Well done sir.

  • Michael Tiemann

    Sam, I thought it would be easy enough for me to put my hands on the data for a different take on the chart you published above. But it’s not.

    The different take on the chart would be to show a colored dot product of the democratic/republican leanings (colored purple at the 50/50 split, then desaturating to light blue/light red at 60/40, then saturating to deep blue/deep red at 80/20) vs the percentage of the electorate represented by the demographic line. We all know the USA is getting less white, but by how much over time? By showing how the various populations are increasing/decreasing, and by showing also a quantitative lean for each line over time, we can get a better feeling of what the future holds.

    Note that for exclusive groups that add to 100%, the lines can be stacked. For non-exclusive groups (white evangelicals may or may not be whites with college degrees), the lines can float along whatever percentage they represent.

    • Sam Wang

      To do what you are asking, one would need a complete-ish list of demographic categories over time, including their political preference and their fraction of the population.

      For the NYT chart I just picked out major groups and changing groups. For instance, note that I left out Latinos.

    • Michael Tiemann

      Indeed. I didn’t say the data was easily got. But rather that if it could be got, would greatly improve the explanatory power of the chart (without adding too much graphical clutter).

    • Sam Wang

      That would have been better. A great project for the people at The Upshot, who are paid full-time to do this kind of thing!

  • Ken Barish


    Going back to Harold Bridges’ comment on the correlation between income inequality and political polarization. This hypothesis should be relatively easy to confirm or disconfirm, looking at data since WWII. It should also be possible to determine the relative contribution of inequality and rapid communication of ideas to polarization. Has any study been done?

    • Sam Wang

      In the GOP primary, Trump voters had slightly lower income than Kasich/Rubio/Cruz voters, but only by a tiny amount. Also see this Vox piece, which indicates racial resentment as a stronger factor than income.

      Note that this kind of question cannot generally be answered by simple correlations, since variables often track together (income and race for instance). Often multiple regression is used, but that is also a blunt instrument and an outdated tool for analysis – its uncertainties are quite large, for one thing. There are other approaches such as regression trees…I haven’t kept up with this.

    • Ed Wittens Cat

      Ken it’s not about income inequality– it’s about SES inequality (Socio-Economic Status).
      Cultural conservatives have been denigrated as racist, uneducated, bigoted– and ritually mocked by the organs of culture that exploit them…television and film.

  • A

    Is gender equality making men feel discriminated against and thus more likely to prefer Trump?

  • Dave

    Sam.What is your take on how the latest Comey/FBI/ emails event will move the polls in the presidential race?

  • Ken H

    Professor Wang, you need to take your astute analysis in today’s NYT piece to its logical conclusion, i.e., the supplanting of our form of government by a right wing dictatorship.

    The polarization and stalemated government fostered by today’s instant, ubiquitous communications technology eventually will fatally undermine the principles of representative democracy established by Madison and the other founding fathers. The common thread between the Republican impeachment of President Clinton (on Constitutionally dubious grounds) after Clinton’s resounding re-election in 1996; the explicitly avowed Republican strategy from day 1 of the Obama administration to flatly obstruct all of his initiatives; Trump’s bleating that this 2016 election is rigged; McCain’s recent vow in Pennsylvania to block any Clinton nominee for the Supreme Court; and House Republican vows to incessantly investigate Clinton if she wins (and certainly if she does not), is nothing less than unilateral Republican nullification of the preceding preceding presidential elections. Without all sides accepting the legitimacy of elections, which Republicans and Trump supporters show no signs of doing, our form of government cannot survive.

    Whenever Trump or a like-minded apostle eventually is elected President (which will happen at some point because of continually stalemated government caused by extreme polarization), he/she will fabricate an excuse to become a dictator in order to “save” our country and mask the failures of his or her policies. The inevitable next step then will be refusing to cede power, much as Putin and other dictators have initially gained power through the ballot and then consolidated it into despotic regimes.

    That is where the Republicans and polarization are taking this country — pulling down our house around us. Please follow through with your analysis and publicize where this country is headed.

    • David

      I believe that a half day of study of American political history (especially Congressional history) will make current political behavior seem less apocalyptic.

      I’d also mention that 1942–1972 were atypically civil. Yet that is our (mine anyway) baseline.

    • David

      1800–1875 were particularly nasty and brutal…

    • alurin

      … culminating in civil war. So I think it’s reasonable to fear increasing levels of disrespect for political norms and institutions (incivility is the least of our problems, though perhaps a leading symptom).

    • OJ Nordhagen

      Who was the president that had a full week of a drinking party after he had won the election back in the old days anyway? I must say, I don’t envy you folks having to live through the present situation. The rest of us around the world are shaking too. I did my best (US born & absentee voter). May your God be with you.

  • TC

    Dr. Wang has polarized this site by his heavy-handed moderating…

    • Sam Wang

      It is true that I shape the discussion.

    • Chuck

      Because unmoderated comment reads are always so enlightening.

      Oh, wait…

    • James Orr

      x-axis is mislabeled on the Presidential Vote Margin graph. “10” should be “12” Not that anyone is likely to be misled.

    • Dave Rodland

      Please. Expression of our polarization is one of the primary functions of the Internet, outside of cat gifs. (Inside of cat gifs … help me out, Groucho?) Heavy-handed moderation can encourage civil expression, polarized or not – which is why this is the only site on the Internet where the comments section is worth reading in its own right. That alone is an immense accomplishment, particularly this season, and one for which I believe we are all grateful.

  • elkern

    I know why I hate the Republican party:
    – Reagan:
    – Voodoo Economics
    – Bloody Central America policy
    – who really was in charge anyway?
    -Bush I
    – should have kept Bush II out of politics
    -Gingrich & 1990’s Congress:
    – Impeachment? Really?
    – Bush II
    – Cheney (everything he ever did or said)
    – Invasion of Iraq, enabled by lies
    – killed/maimed 100K – 1M people
    – wasted $1-3T
    – made permanent enemies of millions of Iraqis
    – turned our Republic to Empire
    – undermined international Law
    – squandered global goodwill toward USA
    – AGW denial
    – Republican Congress since 2008
    – highest, nay, ONLY priority was to thwart Obama
    – blocked Infrastructure investment
    – won’t even consider filling Supreme Court vacancy (won’t do their job)
    – et-f’ing-c

    I know that millions of (US) Americans hate the Democratic party as much as I hate the GOP. Can anyone here explain WHY, in like, maybe a list like I’ve got, above?

    • Chomper

      I can’t speak for everyone, and I kinda agree with some of your reasons for hating the Republican party. Here’s why I hate the Democrats though:

      – Social Welfare programs that enslave swaths of people to a gov’t of nannyism
      – Racist Social Justice programs that foster more separation among the populace and assume minorities are incapable of reaching the same levels of achievement
      – Failure to appoint neutral, non-partisans to judicial positions
      – Backing an electoral system that blatantly disenfranchises its own voters

      that’s off the top of my head.

    • elkern

      Chomper –

      Thanks, I guess. Frankly, it seems awfully weak, and not even accurate; kinda confirms my worst fears.

      1. Social Welfare programs “enslave” us? Social Security, Medicare, etc, condemn us to living in chains, with no more legal rights than a chair? Read some real history.

      2. Can you be more specific about “Racist Social Justice Programs” and how they exacerbate segregation? Like when the Feds forced States to desegregate schools & golf courses?

      3. “Neutral Justices” like those appointed by Republican Presidents? Piffle.

      4. You seem confused about which Party is intentionally disenfranchising voters. Did you miss the case last year where the Courts overturned some Florida’s (or ?) new restrictive voting laws after one of the REPUBLICAN legislators who sponsored it admitted it was designed to suppress African-American/ Democratic turnout?

      The things I listed have led me to HATE (not a word I use lightly) the (modern) Republican Party because:
      – in a Democracy, we’re all responsible for the actions of our Government. I don’t want all that blood on my hands.
      – Denying Global Warming, refusing to invest in infrastructure, and intentionally squelching the economy (under Obama) have made the future worse – all for short-term gain by the .1%.

      Do the things you mentioned really lead you to hatred? They seem pretty fuzzy to me. I suspect that there’s more there, perhaps a bit more subconscious. For one thing, you say you “hate the Democrats”; perhaps it’s just sloppy writing, but it implies that you hate the PEOPLE who are Democrats, and not just the Party. (Note that I’ve been careful to express hatred for the GOP, not Republicans).

    • Sam Wang

      Independent of policy preferences, I think a fairly serious problem is the disintegration of governing norms, whether it be government shutdowns, impeachment and Congressional investigation to achieve policy goals, or failure to give deference to the executive on judicial appointments. Here’s a list for this year alone.

      These and other events over the last 20 years seem to comprise a fairly radical remaking of how government operates. I think it is not broadly appreciated what has occurred, cumulatively speaking. One could make an argument in favor of these…but they amount to changing the nature of the national government.

    • elkern

      Chomper – I want to apologize for getting a bit too heavy in my 1st response to you. I really do appreciate the fact that you bothered to Reply to my (sincere) question; we – Left & Right – need to have more sincere dialog across political lines, and my smackdown of your post is not the way to encourage that.

      As you may have guessed, I’m furious about what’s happened to the GOP. As a greying salaried WASP (age, econ, and ethnic clades), I’d be a natural Republican by now, but – as Reagan said of the Democratic Party – the Party has moved away from me.

      In my view, right-wing media (Fox “News”, etc) warped the psyche of millions of Americans, by feeding their fears and unleashing anger. Limbaugh led the way, dropping any pretense of civility or factuality.

      Anger is exciting, and people were drawn to that. I believe that the rightwing propaganda from Fox, etc, created a wave of ill-defined anger, and focused it on the Democratic party.

      Your first two points are probably a real key to the way that propaganda worked. Many Leftists say it’s all about Racism; that’s part of it, but that’s an over-simplification. Wage stagnation since 1980 is a real problem, especially for non-degreed white males. Their status has declined – RELATIVE to women, non-whites, and people with (advanced?) degrees.

      To me, it’s obvious that the problem is that Wall Street has captured almost all of the gains in productivity since 1980. But millions of Americans have been convinced that it’s the Blacks, Mexicans, China, etc, who have “stolen” the jobs which once made it possible for (white) men to support a family on an hourly wage.

      Anger is natural, but dangerous. It’s an easy handle for manipulating people. So it’s important to be clear about the roots of our anger.

  • Joan Van Tassel

    What a great discussion! Thank you for your NYT article, Dr. Wang — it was a very interesting and useful framing.

    As I read the comments about the educational gaps in the electorate this cycle, I was reminded of Neal Stephenson’s book, Anathem. The novel described a future society where the educated lived apart from the general population (genpop, as it were), in scientific monastery-like learning/ living environments. Oh, and there was lots of free love and sex there too, as one might expect.

    Members of the genpop had “sacked” the intellectuals’ monasteries several times, so the scientists had constructed their defenses and ultimate made themselves safe. The larger populace did request scientists for help with specific projects and problems, which the scientists usually provided.

    In my view, having a “college degree” is not really the point any more. Universities are the gateway through which people learn languages (statistical and textual analysis, and artistic resonance) that allow them to make and express conclusions about the nature of reality. All these perspectives require education beyond the capabilities of all but a few high schools.

    I teach adult college students (average age 33) in online courses. Graduate students typically need advanced writing support — most already speak “modern” (or post-modern!). But the undergraduates don’t. They’ve reached a point in the workplace where the way they gather information, process and analyze it, and then express themselves does not allow them to function at the level they would like to reach.

    Much of my job involves exposing them to these alternative frameworks of information-seeking and communication. It is challenging and exceptionally rewarding, because these students know they missed the golden ring and many are alternatively puzzled, threatened, excited, and wary of the new world they have entered.

    I expect that DJT’s limited vocabulary and oft-repeated simplistic “solutions” comfort people who, unlike my students, cannot or will not open up their minds to post-19th century conceptual universes. I suspect they are quite thrilled to use Twitter and text on their phones — they have been accepted into the computer world via their mobiles!

    I hope we will have a chance to develop and diffuse educational systems and programs that begin to address the “conceptual divide” that
    is really the parent of the “digital divide.”

Leave a Reply to Ken H (Cancel)