That “Massive New Study” Says Nothing About Economic Anxiety

By James Kwak

Last week, the Washington Post summarized a draft paper by Jonathan Rothwell of Gallup on the demographic correlates of support for Donald Trump. As various people have noted, the headline was a bit over-the-top:

Screen Shot 2016-08-15 at 5.49.49 PM

The “widespread theory,” of course, is the idea that Trump supporters are, at least in part, motivated by economic anxiety—an idea that sophisticated columnists like Matt Yglesias like to make fun of, as I discussed recently.

The article itself, as many people have noted, is considerably more circumspect than its headline. (Note to those who don’t know: Headlines are written by editors, not the people on the byline.) This is the summary near the top of the article:

According to this new analysis, those who view Trump favorably have not been disproportionately affected by foreign trade or immigration, compared with people with unfavorable views of the Republican presidential nominee. The results suggest that his supporters, on average, do not have lower incomes than other Americans, nor are they more likely to be unemployed. [Actually, according to the paper, they are more likely to be unemployed, but that’s not particularly important.]

Yet while Trump’s supporters might be comparatively well off themselves, they come from places where their neighbors endure other forms of hardship. In their communities, white residents are dying younger, and it is harder for young people who grow up poor to get ahead.

The paper itself is more circumspect still. Here’s an excerpt:

Higher household income predicts a greater likelihood of Trump support overall and among whites, though not among white non-Hispanic Republicans. In other words, compared to all non-supporters or even other whites, Trump supporters earn more than non-supporters, conditional on these factors, but this is partly because Republicans, in general, earn higher incomes, and the difference is no longer significant when restricted to this group. …

On the other hand, workers in blue collar occupations (defined as production, construction, installation, maintenance, and repair, or transportation) are far more likely to support Trump, as are those with less education. … Since blue collar and less educated workers have faced greater economic distress in recent years, this provides some evidence that economic hardship and lower-socio- economic status boost Trump’s popularity.

Before we go further, let’s make sure we understand exactly what this paper does and does not show. For the most part, it’s based on a probit regression of the likelihood a person will support Trump (that’s the dependent, or left-side variable) on a long list of variables for that person (employment status, religion, etc.) and a long list of variables measured for the area in which that person lives (share with BA degree, share of manufacturing jobs, etc.). For each variable, there is a regression coefficient that shows the impact of that variable on the likelihood of supporting Trump, and then an indication of whether that variable is statistically significant. For example, in model 1, looking at all people, being unemployed increases the chances that someone will support Trump by about 5%, which is significant at the 99% level.

There are two reasons why this paper says less than readers might think. The first is that many of the right-side (explanatory) variables are highly correlated. When you have highly correlated explanatory variables, you can get wildly inaccurate results. Let’s say you are trying to figure out what factors determine the number of words in a child’s vocabulary. In your model, you include age, since kids learn more words as they get older. You also include grade in school, since they learn more words the longer they spend in school. Do you see the problem? Age and grade are almost perfectly correlated; you’re basically using two variables when there is only one in real life—so the actual results of your model will be highly volatile. You might find that age is significant but not grade; or vice-versa; or that both are significant. If both are significant, you might conclude that both have a positive impact on vocabulary: that is, fourth graders know more words than third graders, but within any grade, older kids know more words than younger kids. That sounds plausible—but it would be a mistake. When explanatory variables are highly correlated, results are extremely sensitive to outliers. If you have one older kid in fourth grade who knows lots of words, you could get a positive coefficient on age; but if that one kid doesn’t know very many words, you could get a negative coefficient.

How does this apply to this paper? The individual explanatory variables include, among other things: employment status (e.g., self-employed); religion; “works for government”; sex; marital status; “works in blue collar occupation”; union member, non-government; race and ethnicity; highest degree; and household income. The regional explanatory variables include: share of college graduates; share of manufacturing jobs; median income; share of white people; and white mortality rate. All of those variables are obviously correlated with income, particularly highest degree. So we have the same problem described above—too many variables for the amount of variance in our sample—which produces arbitrary results. (One way to think about this is that you could use a bunch of those variables to predict household income pretty accurately, at which point the household income variable itself becomes unnecessary.)

The Washington Post writeup remained blissfully unaware of this problem:

After statistically controlling factors such as education, age and gender, Rothwell was able to determine which traits distinguished those who favored Trump from those who did not, even among people who appeared to be similar in other respects.

This is the argument that the statistical significance of the income coefficient means that, among people who are otherwise identical, higher income does have an effect (pro-Trump, in this case). But as explained above, that’s a fallacy. Multicollinearity, as this statistical problem is called, means that individual coefficients are unreliable. The model as a whole may predict support for Trump pretty well, but you have no way of knowing which variables are doing the predicting.

That’s the first problem with this paper: we can’t trust the coefficients. The second problem is one of interpretation. Even if we accept for a moment the coefficients on the explanatory variables, the paper says nothing about why people actually support Trump; it’s just a long list of correlations.

So imagine this simple world. There are 100 people. 50 are poor and 50 are rich. In each group, one half (25 people) vote based on their feelings, such as economic anxiety. The other half vote based on their interests. So the electorate looks something like this:

Screen Shot 2016-08-15 at 9.42.21 PM

Of the people who vote their feelings, let’s say economic anxiety does increase support for Trump. So Trump gets 15 of the people in the Feelings/Poor box but only 10 people in the Feelings/Rich box. For people who vote their interests, however, income is positively correlated with Trump support, since he has promised to cut their taxes. So Trump gets 20 of the people in the Interests/Rich box but only 5 people in the Interests/Poor box (because the other 20 realize that Hillary Clinton’s policies will be better for them).

Now our exit poll looks like this:

Screen Shot 2016-08-15 at 9.47.06 PM

Trump gets only 40% of the poor voters, but 60% of the rich voters.

That’s what the Gallup paper shows, and that’s what the Washington Post editors used as their headline: rich people prefer Trump, so economic anxiety is a myth. But I constructed this outcome using a model that explicitly incorporated economic anxiety as a factor (in the Feelings row, Trump does better with poor people). In other words, the economic anxiety story is consistent with a study showing that, on average, rich people prefer Trump.

The lesson is very simple, and it’s one that everyone knows before becoming a poll-reading pundit: People make decisions for different reasons. Something can be an important factor—here, it gives Trump a 20-point advantage among half the population—but get outweighed by some other important factor. Or, to put it in sophisticated language, you can’t use income as an instrument for economic anxiety, because income affects Trump support through other channels (in this example, because some rich people realize that Trump’s tax cuts will be good for them). This is really just the same mistake that Matt Yglesias made yesterday with race and age.

As should be obvious, I think that economic anxiety is a reason why some people support Donald Trump. I can’t prove it from poll breakouts, or from the Gallup paper, because this type of hypothesis can’t be proven or disproven with that type of data. That’s the one thing you should remember the next time someone argues that some demographic statistics show why some politician is popular.

24 thoughts on “That “Massive New Study” Says Nothing About Economic Anxiety

  1. This is true enough but kind of misses the point. “Economic anxiety” as used in discussion of Trump doesn’t really have anything to do with feeling nervous about money. Leftists use it to mean something like false consciousness; in the neutral press it means, so far as I can tell, nothing at all, functioning merely as a way to talk about Trump’s supporters in a way that fits their self-image (hard-working normal guy being taken advantage of by amorphous outside forces) instead of one that doesn’t (geriatric racists). Both versions of “economic anxiety” are premised on the assumption that Trump voters have been uniquely harmed by the structural changes to the economy in the last twenty years or so and this study does provide some evidence that this is not the case

  2. Well lets throw in the fact that each word has 3 interpretations, then age or grades become irrelevant yet the discussion continues on. Now people have every right to be concerned about their “economic anxiety”. Once you financialize the economy with high interest rates, you have gone off track, It’s the gvts job, the Feds job to keep the cost of living so low that these “anxieties” are kept to a minimum rather than being the last straw which breaks the camels back.
    As soon as these big bang___ anxieties are resolved, the new problem becomes the same old living in the past and how to control the resolution to keep the status quo on track. Life’s goals in this country never steer onto the right track, rather they swerve back and forth trying to remain on course and then fall back on insurance co’s to pick up the tab of the wreck. This in turn raises the cost of living and adds another straw to the camels back, which should concern everyone yet gets blindly swept under the carpet as the hoover is being repaired.

  3. There is plenty of Global Moral Compass Annie, you just have to know where to look, like way in the past. At one time they took the most natural hostile land (Russia) and put the most hostile people(Americans) together to find that they ended up with millennials of civil war.
    At the same time they put the most civil people (Russians) on the most natural forgiving land (America) and everything was peaches and cream…………… Except for the growing worldly civil unrest which threatened to consume the peaches and cream.
    So the middle ground was taking the most natural forgiving land(America) and placing the most hostile people (Americans) together, and placing the most civil people in the most natural hostile land, and let the experiment begin.

  4. I agree with the post 100%. What I don’t get is that, even taken at face value, the paper didn’t really claim what most pundits said it did (–the post title is snarky, not serious BTW.

    Many of the same pundits who exalt data-driven analysis either didn’t read or didn’t understand the basics of the paper. Or else engaged in some really egregious motivated reasoning.

  5. “Rich versus “poor” is a grotesque over-simplification, but it fades a bit in the face of another issue. Even as rank speculation,this whole thing falls apart based on the assumption that people who at present have more money have less “economic anxiety.” I would have to see convincing evidence for that. After all, if we assume they’re not all trust fund babies like Donald Trump, how did they acquire that money? I suggest it has much to do with “economic anxiety” expressed through going to college, starting a business and/or living in corporate cultures. Maybe the “rich” just have more economic anxiety in general. Maybe not. This study and article do nothing to address that.

  6. No foothill, it doesn’t, yet we do know who and how they acquired the money, think lots of dry power and the competition of insanity for money. The big question now is, what can we do about it?

  7. I see this interesting analysis as missing the point. Trump’s stated solutions for economic anxiety would not improve the lot of those who are voting out of economic anxiety because the overwhelming majority of them have not been subjected to wage competition from immigrants or suffered through the China trade shock. Kwak simply does not speak to this most important aspect of Rothwell’s analysis. To be sure, Trump may have healthy majorities of those living in communities shocked by Chinese imports, but they don’t make up (I think Rothwell estimates) more than 14% of his supporters.

    So Kwak misses the point here: why would people support Trump if his concrete policies would not help them. The answer of course is that they are voting to assert their sense of ontological belonging qua white people and to assert the value of their “racial” capital–meaning, as Kwak should understand, that white people should never be passed over for a job, promotion, bid or government allocation to anyone “foreign” which of course means anyone without European heritage.

    White nationalism as a purely symbolic statement is what Trump is promising. This is how he intends to allay people’s economic anxieties. This is clear from the New York Times video of his rallies.

    Trump also seems to be drawing disproportionate support from those areas
    where the white mortality rate is not falling in line with trends elsewhere in the wealthy Western world. But this makes support for Trump all the more surprising because he clearly has no solution other than a wall to keep out drugs to speak to the opioid, heroin and alcohol abuse problems in this country. All that Trump is offering people here is again white nationalism as a solution to the health crises afflicting their community while Clinton and Kaine promise to follow through on the work begun by Vivek Murthy.

    most Trump supporters have not been hurt by trade (Rothwell, I think estimates only 14% have)

  8. PCA, followed by [logistic,pro it] regression? While it would not dis-aggregate the factors, it would get rid of the multicolinearity. The factors would gather together variables with high |r^2|, and at least different factors could be contrasted.

  9. Trump will spend the next 2+ months trying to convince people that he will amend his Muslim ban to tolerate only the half-dozen Muslims that would fall on an IED to save white Americans. He’ll also clarify that he will whip up anti-Latino hatred not just to help with his Trump University case in front of Judge Curiel but also for the sake of white Americans as a whole.

    He’ll repeat horror stories of white cops killed by black activists, black people killed by illegal aliens, and white people killed en masse by Muslims. Then he’ll promise to humiliate China and Japan in trade and currency negotiations to the point of there never being a chance for another Copenhagen or Paris agreement were he elected.

    And he’ll hope for both a massive terrorist attack organized from the failed state of Libya and for a damaging wikileaks.

    Yeah, I guess his campaign is in some important way about economic anxiety.

  10. Since we already proved that hope always ends in failure, I guess that ” both a massive terrorist attack organized from the failed state of Libya and for a damaging wikileaks.” won’t be happening, thank’s fer the input.

  11. @skunk – what a plan, opposite of The Prime Directive from the Star Trek series – share superior technology with people who have been living for millennia without ever inventing a broom. And watch who wins?

    Can’t publically discredit that kind of plan…why?

  12. See, no one wants to lay claim to inventing a “hoover”…

    Guess you are not going to get out of my face, skunk, by just me asking…?

    So in response to the failure to prevent 911, a spy network was set up to keep me “safer” that depended on emptying out my bank account in order to pay for the authorized erection of The Patriot Act (rob Peter to pay Paul)

    – except the internet, wifi, and cloud can never be secured enough to keep me safest from skunks in my face…?

    Wait, what?!

    Wow – ever follow the people driving away from their workplace each day where they are programming how driverless cars will “drive”…? Worst pricks on the road – won’t merge, speed, tailgate, text and drive…so are they programming OUT all their driving habits or programming them in to the “driverless” car…?

    Countdown to Facebook censors closing my account for “public” posts like this one:

    “….If you can wrap your head around the possibility that “the devil” has been repeating this pattern out of contempt for humanity (he was sent here to be our caretaker through our INFANT stage of spiritual development), then you will find more courage than you knew you had to level the playing field, so to speak…what a prick…

    Page 759 – The Urantia Book – “The Caligastia scheme for immediate reconstruction of human society in accordance with his ideas of individual freedom and group liberties, proved a swift and more or less complete failure. Society quickly sank back to its old biologic level, and the forward struggle began all over, starting not very far in advance of where it was at the beginning of the Caligastia regime, this upheaval having left the world in confusion worse confounded.”

    200,000 years and counting….”

  13. Obviously Kwak is a very smart man, but this seems to be a big question-begging exercise.

    So to take Kwak’s example: why would 60% poor people voting on their feelings of economic anxiety (15 of the 25) vote for Trump? Why do the poor feeling economically anxious favor Trump? And why is that 60% of the poor made up overwhelmingly of whites? What is it that Trump is saying or dog whistling or promising that allays their feelings of anxiety? If it’s not white nationalism that is serving as anti-anxiety medication for the economically anxious white poor, then what is it? Is it his promise that he’ll spit on the cosmopolitan liberal elites who look down on the poor? And is that appeal supposed to be not tied up in white nationalism?!

  14. Academics and analysis aside – Trump’s core constituency, (completely off the socalled mainstream media radar) is the anti muslim or slightly more nuanced and smaller antijihadi voter. Some are dim, some are bright, “some are born to endless light, some are born to sweet delight”. Pan cultural, pan class this constituency is finished with the idiocy of coddling, or holding any respect for the barbarian psychopathic lunacy of islam. This constituency believes there is no peaceful islam, that there are no good muslims, and that the Muslim freaks and shaitans mercilessly butchering unarmed and innocent civilians blowing up airports, trains, and malls, hacking pregnant women with Machete’s, exhibiting a disturbing penchant for beheadings, burning, drowning alive and other grotesque forms of murder, practicing systemic and endemic perpetuation and proselytizing of misogyny, and generally relegating all women (including their own sisters, mothers, and wives) to pets and chattel), throwing acid in the face of fiveyearold girls for the great sin of attending school or exposing their elbows, and insanely imagining in their dim evil minds that the rest of the earths’ humans would imagine or ever peacefully tolerate, or accept a global jihadist caliphate – (collectively muslims and whatever is that thing called Islam) – are evil pschopathic massmurderers. This constituency, this anti muslim, jihadi, anti Islam demographic (which again – no one appreciates) is broad and wide and deep, composed of factions and elements of redneck Amerika, white supremist and other racist elements, various coteries, cliques, and klans of wingnut Amerika – and also progressives who have actually studied the militant vengeful wrathful god composition and pile of ca ca known as the the koran, and the hateful bloodthirsty, supremist teachings of the slave owning, and woman abusing massmurderer Mohamed. Also various gangs of various sorts and cultures and races, and classes, and individual criminal enterprises, – even predatorclass factions whose profits are won beyond warmaking industries, – are done coddling, appeasing, respecting, or recognizing the deseased persverion that is islam, and all its muslim followers and practionioners. Brexit was rooted in the same. There is no religion in islam. No peace. No tolerance. No love. No future unless your a male holding the mysoginist fantasy and fiction of a paradise enjoying the delights of 72 brown eyed virgins, or chats and stroles with this or that pantheon of socalled prophets. The real trump supporters (and there numbers are not appreciated or measured) are anti muslim, anti islam, and/or anti jihadi or some combination or variation thereof. Socalled politicians speaking to freedoms and protections and appeasing or shielding the barbaric mass murdering insanity of jihadist islam, islam in general, and all muslims. These massmurdering barbarian, and mysoginists perverts have relinquished and shattered any claim or hope for religious freedom or any protections or privileges as a socalled religion.
    This Trump constituency – the unknown unknown Trump supporters- right or wrong, ignorant or educated, decent citizen, or hardened criminal – fiercely reject, renounce, and repudiate any and all appeasing, protecting, shielding, or excusing, and and all tolerance or acceptance of the barbaric mysoginist, massmurdering, persverion and satanic freakshow that is islam, and/or jihadi islam, and all muslims. Trump is the only candidate any where in the socalled 1st world speaking to the primary concerns of these disparate constituencies.

    This constituency (right or wrong, good or bad) is done with Islam or jihadi Islam and all socalled Muslims.

    Now academics and analysts may disagree with this constituency on any number of grounds and polls and predatorclass lobbies dispositions. But “Ye shall know the truth. And the truth will set you free”.
    The Anti Musim vote will hurl Trump to the presidecey

  15. Well Annie, I know you are not going to “get out of my face” so my life can return to normal until this gvt defaults, so it is, or, would seem to be, what it is. And yes, the devil’s and satan’s tenticles spread far and wide, and are based on the competition of insanity for money. This has never changed. Now, can you tell me something that I don’t know?

  16. I really want to make sure I am not missing Kwak’s point. I get that the data don’t allow us to determine positively how important white nationalism is to Trump’s supporters. I get that many of Trump’s supporters may just be reflexive Republicans raised that way or of course attracted to his tax policies. But Kwak has conceded quite a bit to this Gallup poll at least conceptually. Again take Kwak’s example. Only 5 of the 25 poor who are interest voters support Trump. But this concedes the Rothwell finding–those who would plausibly stand to benefit from his core issues of deportation and protectionism make up a small part of his supporters.

  17. Some of the benefits to getting rid of “white nationalism” – leave it to wild dog packs to get rid of those pesky Social Security collectors:

    @Tony – it is not a “religion”, is it? But then while people such as skunk keep “meh-ing” about their complicity in “that is the way of the world” and delight in the game of bringing in a new wave of immigrants to deal with the previous wave herded into the bankster’s version of “the American dream”, resistance is not futile :-).

    Simple – get outta my face.

    With whitey dead and gone and the battle for “good jobs” and home ownership heats up between Aztec heart eaters and Arab beheaders, wonder which group the IRS will go after then…and using which “immigrants” who got here first to do it….?

    I predict that unity will prevail because the Aztecs already ADMIRE and are envious of the control Arabs have over their women.

    Never under-estimate the power of misogyny as a unifier in “religion”…

  18. Religion is a circle which has no end, that was in your face, face. Unlike my Federal Institutional check which hangs framed on the wall, still uncashed naturally.

  19. Tsk tsk, “hillbilly” still has not caught on to the WORD game – the point is to make it sound as if you are saying something so profound that it can be assumed it would be above the heads of the hoi polio….and it is HOW you say it – remember the chastising people would get, and the permanent censorship off boards if they wrote in all caps

    “IT IS TORTURE”…..?

    Well, here is the next round of censorable (is that a word?) all caps, hair on fire internet soul shaming from yours truly…



    That is not an economy that needs a cadre of specialists with PhDs to explain the complexity and genius of the heist to stupid Americans.

    You wanna get “religious” about it? Think of it as Judgement Day. Heck, if the pot of hooligans can be stirred up enough to do what was done to Jesus in the name of “religion”….where can “protection” be found for anyone who gets in your face about you NOT believing what they believe…?

    I do not believe in driverless cars being made possible by the biggest prick drivers on the road right now – the “elite”…

    I do not believe in anything they say.

    I just want them outta my face.

    Burn the Patriot Act – preferable with them holding it….

Comments are closed.