Correlation Doesn’t Mean Causation

As Clarion Call readers know, the Pope Center and some other reform-minded groups have been arguing that higher education has been badly oversold in America. In short, we contend that the on-going campaign to get more and more young people into college has led to the erosion of academic standards, watering down of the curriculum, credential inflation in the labor market, and other maladies. The nation has produced a higher education bubble, much like the housing bubble.

The case that we have oversold higher education is, in my view, so solid that it comes as something of a shock to read a paper like “The Undereducated American” by Anthony Carnevale and Stephen Rose. It argues that the nation needs to “produce” far more college graduates than it presently does. That seems utterly contrary to everything we know about the effects of turning college into a mass consumer product; it’s as if someone said that doctors should go back to the practice of bleeding the sick as a treatment.

Nevertheless, the paper has received a lot of favorable comment.  New York Times columnist David Leonhardt crafted a splashy article around it, contending that its data refute what he termed the “elitist” case that for many people college is a bad use of time and money.

So I carefully read “The Undereducated American” to see if the authors had indeed refuted the idea that college has been oversold. Not at all. There is much that’s wrong with the paper. In this Clarion Call, I will try to expose its most glaring sins of commission and omission.

In a nutshell, here’s the paper’s argument. The U.S. is “underproducing” college graduates, proved by the fact that there is a wide earnings gap between college grads and those who aren’t. Since that gap has grown over the last several decades, it must be the case that the demand for college-educated workers exceeds the supply. The best solution is to increase the supply of college graduates (currently 66 percent of high school graduates enroll; the authors say we must raise that to at least 86 percent), which will simultaneously lower the earnings gap (the authors say that 46 percent is the right “premium” for college, down from 74 percent) and make the economy more productive.

The result would be beneficial for everyone: “The wages of high school-educated workers will rise by 24 percent, those with Associate’s degrees will rise by 15 percent, while the wages of those with Bachelor’s degrees will rise by 6 percent.”

But can we really pull ourselves up by the bootstraps just by expanding college completion? Would the additional students brought into the higher ed system add so much in productivity that GDP will increase by $500 billion annually, as Carnevale and Rose state?

Here we run up against the first great sin of omission. The notion that increasing years of formal education is an elixir for the economy has been around for quite a while—and has been subjected to considerable scrutiny. Refuting it was the central theme of Professor Alison Wolf’s 2002 book Does Education Matter? (which I  reviewed here). She showed that rising “educational attainment” is neither necessary nor sufficient for economic progress. Some nations that have pushed hard to increase it have gone nowhere economically; conversely, others that don’t have the fetish of educationism have done very well.

In a recent interview, Wolf said, “The idea, which I have to say has affected large numbers of politicians, that you can just give people a university certificate and hey presto, they’ll earn this amount more and the country will be x-amount richer has always seemed so bizarre to me that I have to pinch myself that so many apparently rational people believe exactly that.”

You will look in vain for any reference to Wolf or other writers who have argued that it’s just not true that the more college graduates a nation “produces” (yes, that’s how the paper refers to college education, as if people were fungible chunks of raw material), the more prosperous it becomes.

Evidently, the authors think that the data they present, showing that people with college degrees earn substantially more than people without them is so conclusive that they don’t need to deal with the counter-argument. The counter-argument, which critics like Rich Vedder, Charles Murray, and I have been making for years, is that financial success correlates with college graduation but is not necessarily caused by it.

That is because people who have higher levels of skill and ambition are far more likely to succeed on their own capabilities. We know that quite a few extremely successful people did not graduate from college. Some never attended at all. It’s a logical error to think that college education is a necessary condition for or causes high earnings. Some people will attest that their college courses did little or nothing to help them in their careers. Journalist John Stossel is one; he denies that his Princeton education had anything to do with his career. Another is the author of this Pope Center piece.

Carnevale and Rose try to clinch their argument by showing that people with degrees do better financially even when they’re working in jobs that clearly have nothing whatever to do with college studies. For example, waitresses and cashiers who have been to college earn more on average than do people in the same line of work who haven’t. They claim that the “Bachelor’s degree premium” for waiters and waitresses is 34.3 percent; for cashiers it’s 55.6 percent, and for hairdressers and cosmetologists, it’s a huge 69.1 percent.  Those data lead them to conclude that college is a good investment even if you end up in a job that high school dropouts can learn.

It’s something of a breakthrough when the “college for everyone” advocates acknowledge that we have so saturated the labor market with college-degreed people that we now find them working in jobs like that. But what about the authors’ attempt to turn that to their advantage by saying that their higher earnings proves how beneficial college is?

Their attempt actually undercuts their case. Since no one (yet) takes college courses such as Table Waiting 101 or Telemarketing Sales 237, the fact that college-educated waitresses and telemarketers earn more on average is powerful support for the argument that it’s the natural advantages in ability that matter, not the courses or degree. While it is true that on average, waitresses and telemarketers with college credentials earn more, it is not true that they’re being paid a “premium” for having them.

Employers don’t have separate pay scales for workers with and without degrees. The higher average earnings for those with degrees reflect the correlation between having gone to college and better work habits. Employees who are punctual, reliable, and cooperative are more apt to earn raises; those who are not punctual, reliable, or cooperative are more apt to get fired or leave. Those who have been to college are more likely to have the good traits, but college isn’t a causal factor.

(Andrew Gillen’s post on this question gives quite a few more reasons why waitresses with degrees are apt to earn more, none relating to any college learning.)

Besides, the “investment” in a college degree looks very poor even if you become one of the best-paid waitresses, hairdressers, or cosmetologists. The benefits don’t warrant the costs and the so-called “premium” over the earnings of less capable workers is irrelevant. 

Now let’s ask what sorts of students we would lure into college if Carnevale and Rose could somehow engineer the big expansion they want. They say that there are many young people who are in the top half of their high school classes who aren’t going to college. That may be true, but that doesn’t mean that they have much academic aptitude or interest.

To get a good idea about the level of the marginal college student today, read In the Basement of the Ivory Tower by “Professor X,” a book I recently reviewed. The author relates his struggles to teach English to students who are barely literate. Few learn much and it’s hard to see how their courses increase their human capital and make them more productive workers.

It’s impossible to see how putting more students like that through college (and expanding the numbers would mostly mean enrolling still weaker ones) will lead to the benefit of higher earnings for everyone.

Nor would we even get the narrowing of the wage gap that the authors regard as so important.  Here’s why. The more people who have college degrees, the more credential inflation spreads. Credential inflation forecloses people who don’t have college degrees from jobs they could perform, thus forcing them to compete for jobs in the shrinking part of the labor force where they aren’t discriminated against. If anything, that exacerbates the wage disparity.

There is much more to criticize in “The Undereducated American” but I’ll close with a practical objection.

Carnevale and Rose are apparently aware that many of our current college students are academically weak and ill-prepared for college because they write of the need to “improve the quality of graduating high school seniors.” That’s a gigantic understatement. The problem, of course, is that we have been trying to improve public K-12 education for more than 30 years and the results have been pathetic. The authors offer no means for achieving the elusive goal of making K-12 education effective. Wishful thinking won’t get us there.

I’m reminded of a cartoon. Two math professors are looking at a blackboard filled with equations. One points with his chalk and says, “And then a miracle happens….” Thinking that our K-12 system will somehow change so that many more students graduate from high school who are prepared for and interested in true higher education is to expect a miracle.

Carnevale and Rose are right in saying that many Americans are undereducated, but the time and place to fix that problem is grade school. Trying to “produce” college graduates from young people who have scant interest in or aptitude for serious academic work won’t accomplish anything.