Will Rogers, Bernie Sanders, and Donald Trump

I’ve been thinking for a while about why anger is so politically popular right now, on both the left (Sanders) and the right (Trump etc.)—even though the country as a whole seems to be doing pretty well (low unemployment, low inflation, low crime, not really at war …). It almost seems like most people think that, despite those good numbers, the group they’re part of is at real risk of being left behind. There wouldn’t be any mystery if just a few groups felt that way, but the weird part is that pretty much all groups seem to feel that way at the same time.

After putting in some thought, I’ve got an answer. I’m sure it’s not the whole explanation, but it’s one I hadn’t heard before, so I thought I’d share. It’s related to the Will Rogers phenomenon, which is based on the quote “When the Okies left Oklahoma and moved to California, they raised the average intelligence level in both states.” That is, Will Rogers was (jokingly) suggesting that the Okies were dumber than average for Oklahoma, but smarter than average for California, and thus when they moved they inceased the average intengence in both states. (This statistical quirk also explains stage migration in medical testing.)

I’m wondering if a similar phenomenon has happened in a bunch of areas in America lately. Let’s start with college.

Here’s the toy example I came up with when thinking this through. Imagine a population of 100 people, sorted by ability 1 to 100. Of these 100, the top 30 go to college; the rest don’t. (Perfect meritocracy: achieved. Yay toy hypotheticals!)

Further, imagine they’ll all get jobs, sorted by desirability 1 to 100. Employers can’t measure ability, but can see whether someone goes to college (the cynical among you will have no trouble with the realism of this assumption). Thus, employers randomly assign the 30 best jobs to people who go to college and randomly assign the 70 worst jobs to people who didn’t.

What are the average outcomes in this toy world? Well, all college graduates will get a job somewhere between 71 and 100 in desirability; 85 on average. People without college will get a job between 1 and 70; 35 on average. People with a college degree are guaranteed a very good job and people without a college degree have a shot at some decent-but-not great jobs.

But what happens if colleges start accepting twice as many people? Then jobs 41—100 go to college graduates. This isn’t great for people who didn’t go to college: suddenly, they no longer have access to the mid-tier jobs they used to get. Their average job falls from a 35 to a 20. And it’s also not great for college grads: they are no longer guaranteed a pretty good job. Their average job falls from a 85 to a 70. Both groups are much worse off than they were, even though the number of good jobs available to the 100 people as a group hasn’t changed. Indeed, the effect is so large that both groups might still be worse off than they were even if the jobs had all gotten a bit better. (Like, by 5 points or something).

Is everyone worse off? No, there’s one group that’s benefited from the change: the people with ability between 40 and 70. This group of people now get to go to college when they wouldn’t have before. On average, their outcome has shifted from 35 to 70. Except (if ability is hidden from the people as well as the employers) those people don’t know who they are. All they know is that they’re in college, and the outcomes for college grads aren’t what they used to be. Everyone can look back at past outcomes and say that people in their group used to have it better.

Is this model anything like reality? I think so. Of course, in reality, jobs aren’t handed out randomly to all college grads. But you could add in a bunch more groups (“graduates of highly-selective schools with decent grades” etc.). And, just like in the model, the people who have benefited from increased college enrolment largely don’t know it (no one tells someone “we only admitted you because we decided to increase enrolment 10%”).

I’m fairly convinced that this happened with college, but for it to explain much of the anger, it would have to apply to much more than just college. And I think it does.

The same phenomenon also applies to grade inflation—more people get A’s, so getting A’s doesn’t help as much. Same for getting advanced degrees, or going to good elementary schools, or working really hard, or picking the right major, or basically anything else people do to get ahead in the educational system. A similar dynamic seems to play out in the workplace, where more people are willing to work long hours/make sacrifices to get ahead—and so making those sacrifices doesn’t cause people to stand out as much.

In many, many areas, people can correctly say “it used to be that [X] basically guaranteed a good outcome and, even if you did [not-X], you had a shot at a decent outcome. These days, even if you [X], you might not have a good outcome, and if you [not-X], then you’re really in trouble.” For many X’s, that statement is true.

Explaining the anger

Let’s go back to the college example and think through how this can explain the anger on all sides. Let’s think about three different people:

  • Alice finished high school and never went to college. She looks around at the world and says “The world may have gotten better for other folks, but for people like me—people who didn’t go to college—the modern world has gotten a lot worse. It used to be that you could graduate high school and still get a decent job as a factory worker, or a secretary, or at the bottom rung of a company, and then work your way up. But these days, most of the decent jobs require a college degree, and the ones that don’t are dead-end jobs with no future. Even if the world has gotten better for other people, it sucks for people like me. That’s unfair, and someone should do something about it!”

  • Bob went to a good-but-not-great state university. He looks at the world and says “The world may have gotten better for other folks, but for people like me, the modern world has gotten a lot worse. It used to be that you could go to a decent state school, pay your way through by working some summer jobs, and then get out and basically be guaranteed a middle-class, white-collar job. But these days, a decent college education is no guarantee of anything. The good jobs are reserved for the people with fancy degrees, we’ve all got way more debt than we used to, and a decent fraction of my graduating class will end up working at Starbucks as their first job as a college grad. Even if the world has gotten better for other people, it sucks for people like me. That’s unfair, and someone should do something about it!”

  • Carol went to a fancy Ivy League school. She looks at the world and says “The world may have gotten better for other folks, but for people like me, the modern world has gotten a lot worse. It used to be that you could work really hard, earn your way into a good college, and then be safely guaranteed an upper-middle-class, professional life. Back then, you could start working at a company and enjoy a comfortable career there until you retired with a gold watch; if you didn’t want to scrape and struggle in the business world, you could find a job as an academic or working for the government. But these days, getting into a top college is just the start. I’ll graduate with six figures of debt, and have to keep striving, proving myself, and constantly applying for jobs since I can never count on staying anywhere more than a few years. And if I decide that I want to be an academic or work for the government—well, these days that’s just as competitive as anything else. Maybe more! Even if the world has gotten better for other people, it sucks for people like me. That’s unfair, and someone should do something about it!”

And here’s the weird thing: Alice, Bob, and Carol might all be right—the world really has gotten tougher for “people like them.” But that’s not because of any unfairness, it’s because there are more Carol’s and Bob’s in the world, and fewer Alice’s. It’s the Will Rogers phenomenon in action.

But people don’t see that. They just see that the world has gotten better “on average” but worse for people like them. And, understandably, that makes them angry. At least that’s my current theory.

7 thoughts on “Will Rogers, Bernie Sanders, and Donald Trump”

  1. That actually makes a lot of sense. I know a lot of people have been talking in recent years about how higher college attendance rates could be having a number of unintended trickle down affects, but I don’t think I’ve ever seen anyone connect it to general anger and discontent (at least not in a clear A to B because of C sort of way).

    On a side note, I’ve also notice a slowly growing trend encouraging people to move back in the other direction, taking on blue collar jobs that were once seen as low paying and undesirable, partly because of the glut of people going to college and seeking white collar jobs. Maybe this stuff will eventually self-correct.

  2. I kind of like it, though a few of the details in your three tales of woe seem orthogonal. It didn’t used to be the case that people emerging from college carried such crippling debt. And the practice of staying in the same job for life did used to be more common (or at least, it is perceived as having been more common). I’m not sure I can fit either of these observations into your model.

    But it all does fit nicely into the idea of revolution as something that happens when conditions start to improve — changes happen that move a few people from just below the threshold of success to just above it. You would think that at least those people would be happy, but instead it means they are at the bottom of their class instead of at the top. Meanwhile those who would have been on top anyway see these newbies eating away at their security from below, and those who would have been on the bottom anyway see themselves being left behind even though in fact they may be closer to the threshold than ever before.

    1. Yeah, I agree not all of the details I mention are caused by Will Rogers phenomenon that is the focus of the post. My point in bringing up those details was just that each (reasonably) plausible story to tell about how their subgroup has gotten worse off. And when they combine that with the background knowledge that things have—on the whole—gotten better, well, that can lead to the conclusion that their group has been uniquely and unfairly left out of the general progress. What this misses is that every group has an argument about why their group has gotten worse, and that there’s nothing at all inconsistent about every subgroup having gotten worse at the same time that the total group has gotten better (as counterintuitive as that is).

      In fact, I would not only say that this may be true, I tentatively think that it is. I think that, over the last ~50 years, things have gotten worse for college grads, and have gotten worse for non-college grads, and have gotten better for the group containing both college grads and non-college grads. And that’s the Will Rogers phenomenon in action.

  3. Hm, interesting concept. I think there’s probably some truth to this. The question is, why would this be happening? Are employers becoming more selective due to an oversupply of college grads? Or are more people going to college as a result of employers becoming more selective? What do you think is the root cause of the entire situation?

    1. I don’t think the explanation is nearly that complicated. I think more people want to go to college because going to college is really beneficial: the value of of a college degree in increased earnings is over $300,000, well more than the cost of tuition.

      And colleges want to admit more people because they have a mission to educate, and want to educate as many people as they can—especially given that the education they provide is so valuable (on average).

      All of this has a lot of positive effects. The downside, though, is increased anger: people don’t think “my earnings have increased $300,000 because I was admitted to college when I wouldn’t have been”—they don’t even know that they wouldn’t have been admitted. Instead, they think “I worked hard and went to college, but being in college doesn’t get you what it used to. How unfair!”

  4. Working in higher education (though not in the US) you definitely hear the complaint that too many students go to university, that standards are being lowered to let in more people, but never will you EVER hear someone say “I shouldn’t have been allowed to go the university because I’m not cut out for it”.

    It’s similar to people who get really excited about overpopulation – you don’t ever hear them suggest that their parents shouldn’t have had THEM. Other people are always expected to make the sacrifice for the sake of the deserving (that is, our genial generic interlocutor).

    1. I’m not sure how directly you were responding to the post. To be clear, I am not arguing that “too many students go to university.” (I do think university becomes too much of a signaling game, that sometimes students would be better off spending their tuition money differently, and that college-alternatives might have some promise, but none of that was the point of the post.) I definitly do not think that standards should be raised in a way that prevents people who can go to college right now from attending in the future.

      What I do think is that expanding the number of people who go to college has changed average outcomes. One way it’s changed average outcomes is that the average college grad has a harder time finding a good job; another is that the average non-college grad has a harder time finding a decent job. I don’t think this is a bad thing, on net. In fact (because of the Will Roger phenomenon), it is perfectly consistent with things having gotten better all around.

      But I do think it explains at least some of the anger our country is currently experiencing, as people see “people like them” who aren’t doing as well as they used to.

Leave a Reply

Your email address will not be published. Required fields are marked *