July-August 2009

In times of recession, occupational sectors and professional groups naturally take stock. What might they have to offer the country? And what might the country have to offer them?

The College Board, on cue, has just published a Wake-up Call to the American People and American Educators. In a report entitled Coming to Our Senses, it argues that “our nation’s dominant position in the world order is at great risk ... . Across the globe, leaders have put their faith in education. They understand that economic growth rests largely on the quality of a nation’s human resources ... [and] many of us have been asleep at the switch.” It is critical, they argue, for the country to reduce college drop-out rates and increase the numbers of Americans who graduate by their mid-20s—and, obviously, to come up with the public spending that such an imperative requires.

It is not only the higher education sector itself that thinks this way. David Leonhardt, the New York Times’ leading business writer, has been offering his recipes for restoring growth, out of and beyond the recession. “When somebody wonders, What will replace Wall Street? What will replace housing? they can be given an answer,” he argued in a long piece on February 1, 2009.  It is education, the “lifeblood of economic growth.”

It would be wonderful for universities if this were true. Unfortunately it is not.  Instead, this argument misrepresents what degrees do for the individual. And it misrepresents what college graduates do for the economy at large.

Let me start with a little history. For anyone from Britain, like myself, the arguments cited above are pretty familiar. The UK just had a golden two decades when politicians thought they had the elements of a successful economic policy sorted out. They included banking, credit, the City of London, and globalization—but also education. Back in 2002, England’s then-Education Secretary Estelle Morris wrote an article for the Guardian newspaper, the national broadsheet most popular with educators, informing the world that “a one percentage point increase in the number of workers with higher education qualifications raises GDP by 0.5%.”  A tad over-specific perhaps, but this was indeed her government’s recipe for permanent economic success, adopted with enthusiasm by both Tony Blair and Gordon Brown. They believed, with the New York Times headline-writers, that “Graduates Equal Growth.”

As an officer of Western Europe’s most centralized government machine, England’s education minister runs all the nation’s public schools and community colleges and provides substantial amounts of its university funding. (Scotland, Wales, and Northern Ireland are autonomous.) So Morris and her colleagues were well placed to carry out their ideas. Government policy was dedicated to increasing university enrollments and increasing qualification levels (academic and vocational), and education spending rose rapidly.  

When everything fell apart last year, it suddenly turned out that funding more students wasn’t, after all, a recipe for guaranteed growth. It wasn’t even a recipe for relative protection from a global maelstrom. On the contrary, Britain is set to experience a particularly sharp downturn.

It is just as ridiculous to treat education as a sure-fire investment as it is to believe unreservedly in credit default swaps and securitized mortgages. Let me explain why—looking first at individuals and then at higher education’s public benefits.

The Private Benefits of Higher Education
When people talk about the benefits of education, they normally start with the private benefits. We have good data on how much more people earn, on average, for every year of education or for getting a particular credential (high school diploma, bachelor’s degree). So we can calculate what that extra amount represents as a return on the total educational spending or “investment” that preceded it (see box to right). That calculation includes not just the direct cost of the education—to the individual, the taxpayer, or other donors—but also the income people forego while studying.

Using this approach, one typically—though not always—ends up with figures that are solidly positive. In other words, “education pays.”  In the US, even going to college without graduating is associated with higher earnings, and a bachelor’s degree has very high returns, again on average, for those who earn it.

In her article, Estelle Morris seems to have assumed that each and every additional graduate from a UK university would earn exactly the same extra amount, over time, as the average for current UK graduates. Multiply up the difference between average graduate’s pay and the average for those without a college degree, and you soon get a pretty tidy sum—and a silver bullet for the economy.

But how might this actually work out there in the real economy?  The assumption is that by the time someone emerges from college, they have acquired a whole new set of skills. They have acquired, in economic terminology, a far higher level of human capital  than other adults without the benefit of a college education; and like other sorts of capital, this brings in extra income over long periods of time.

But how do we know they have these extra skills and that the skills are economically valuable? Answer: because they earn more. In other words, the job market rewards them—and why would employers pay more if a graduate weren’t worth the extra?

Except that, unfortunately, correlation and causation are not the same. There are three reasons why we cannot simply assume that a year in college, or a college degree, automatically makes everyone first more skilled, and then more productive, and so more prosperous. The first is that education is about sorting people out, not just about teaching and learning. The second is that not all degrees are created equal. And the third is the nature of the job market.

Employers use education to put people in rank order. This is perfectly sensible of them: going to college is not a random event.  Anywhere in the developed world, kids who are academically successful are more likely to attend college than those who are not. Clever students, including those from ordinary working-class homes, by and large earn college degrees. (It is among less-academic students that family wealth and background make an enormous difference.) So while students may indeed learn useful things in college, their college education is not the only thing that makes them different, as a group, from non-attendees.

Employers are well aware of this. In fact, they tend to see potential employees as having spread-out abilities—some learned, some innate—distributed much like the classic bell curve or normal distribution.  People commonly perceive things as being distributed that way; for example, when grading term papers, academics generally find that some are very good, some are very bad, and most are clumped in the middle and quite hard to separate.

Figures 1-3 show this curve has applied to employers’ perceptions, over time, of young workers.  Remember—nothing here implies that intelligence and ability actually are distributed this way, let alone that they are in some sense “innate.”  What matters is that employers perceive the labor market this way. They think there are some potentially great hires, some catastrophic ones, and a lot somewhere between the two. They also think that, on average, college graduates are likely to be smarter and more reliable than those who have dropped out along the way.

Figure 1 is a representation of the population of America in 1930 or Western Europe in 1950. Only a very small proportion of young people attended university, equivalent to the small shaded part under the curve. As an employer, you would have seen those graduates as belonging at the “highly able, highly skilled” end of the distribution. But there weren’t many of them. The non-graduate population contained a lot of very able, hard-working people, and you would have been perfectly happy, and indeed would have had no choice, but to hire from there too.

Several decades later (Figure 2 represents the American or Japanese population in the 1970s, say, or the British in the 1980s), things look very different. A substantial proportion of young people now attend college and graduate, and many will, by definition, be drawn from what employers perceive to be the large middle hump in the ability distribution. In theory, it is perfectly possible that many graduates do not belong on the right-hand side of the distribution but are scattered randomly all over it. However, this is not a very plausible (or convenient) conclusion for an employer to reach.





It is much simpler to see Figure 2 as a good approximation of reality. As an employer, while accepting that there may be some very able non-graduates, you don’t waste energy trying to find them. You assume that graduates are mostly from the top end of the distribution—the shaded area again. If you have a desirable, well-paying job on offer, you recruit only college graduates. Even if they are not the world’s greatest brains, graduating successfully shows they are apt to be reliable, conscientious, and hard-working.  Graduates duly earn more. But only some of that is because of their college-learned skills. Some of it is because they had above-average skills and abilities before they entered college, and some is because the graduate “badge” put them in the employer’s line of vision.

Move now to Figure 3, which reflects something close to modern reality in the U.S., the wealthy parts of Asia, and much of Europe. Among elementary school students, almost every child aspires to go to college. And indeed, a very high proportion of young people attend college, though a fair proportion does not graduate. Finally, more and more jobs are open only to graduates. At the same time, on the left-hand side of Figure 3, we can see that a proportion of the young are still not attending college (or, often, finishing high school). Few employers will be very interested in them, given all the graduates they have to choose from and their assumption that non-graduates are likely to be less able and less conscientious than graduates.



Looking at the world this way suggests something else important. All graduates are not equally valuable to employers. As I noted earlier, how much a degree is “worth” depends a great deal on what someone studies. The amount that graduates earn may be higher than non-graduates, on average, everywhere in the world, but this conceals enormous variations within the graduate population. In Britain, returns to degrees have already dipped badly for specific groups, especially those majoring in the liberal arts or attending low-status schools.

For example, British males who major in a liberal-arts subject such as literature or history would today on average (though not always) be better off financially over a lifetime if they went straight into the workforce after completing high school. We know this from looking at the earnings of people with the same high school grades who did that, However, their peers with engineering degrees will do much better, financially, than either of these groups.

It also matters where one studies. You earn more—again on average—than other graduates if you go to a highly selective institution, particularly if you go on to advanced academic or professional education and even more if is a world-renowned university (Harvard, Oxford). Of course, the flip side of above-average returns in some places is below-average returns elsewhere.

Granted, future earnings are not the only reason to study. Moreover, people with good grades who could go to college and don’t are also very atypical today, so we are not really comparing like with like. But the data underline how misleading it is to paint college as a certain route to riches.

In societies where having a degree opens the door to the best jobs, people naturally want degrees. We owe it to all young people to provide an education that gives them a fair chance of learning and making economic progress; everyone deserves a shot at getting a good job. It is a scandal and a disgrace when schools fail their pupils. But we also need to recognize that, if there is reasonably open access to education, then a country will naturally generate an over-supply of both skills and degrees.

As Paul Barton pointed out in Change last year (January/February 2008), many jobs simply do not require college degrees in any substantive sense. Even before financial services imploded, the occupations that were growing the fastest in absolute numbers—which is what matters when you are seeking a job—were not in investment banking, or IT, or design. They were working as a nursing aide, or in a big retail outlet like Wal-Mart, or as a janitor or cleaner. The same is true in France, Britain, Germany, and Japan.

Studies by British academics show repeatedly that over a third of the current workforce is educationally over-qualified for the jobs they hold. A recent survey found that only 17 percent of workers considered that a degree-level qualification was required to do their job. And China and Korea are currently grappling with huge numbers of unhappy graduates for whom there are no “college-graduate jobs” available.

In summary, education helps people get ahead. In a recession, no one will take much persuading that college is one of the best purchases that they, as individuals, can make. But what college does is put you in there with a chance; a college degree suggests that you may be better positioned than other people competing for the same jobs who don’t have the degree.

That is a very different thing from being a way for countries to deliver economic growth overall. Is there anything at all to the argument that it does?

Education and the “Public Good”
Students (and universities) sometimes argue that they “deserve” to be subsidized because of the economic benefits they provide society with. Their skills, they argue, make them more productive, which means that they will pay more taxes in the future when they go to work and earn their higher salaries. Taxpayers should therefore be happy to finance them now. And without the skills they acquired in college, the argument continues, not only would they be much less productive (and poorer), but so would the people who work with them. In other words, their college-acquired skills make the whole economy more productive and raise everyone’s income.  

The trouble with this argument is that the only evidence we have for all this productivity gain is the high wages graduates are getting. Between 1998 and 2008, that was exactly the argument that was used for the financial services industry. It must be adding huge amounts of value to the economy, because otherwise why would employers be so stupid as to pay such enormous salaries on Wall Street and in the City of London? We now know that a lot of that was bubble. And yet when people argue that college generates productivity and growth, it is people’s current earnings they are working with.

A detailed rate-of-return analysis, looking at which sorts of degree pay most, suggests that the fastest way to boost growth and end the recession is to turn every faculty into a law school. That seems implausible, even if we don’t want to go as far as the rebels in Shakespeare’s Henry VI (they had a simple political program: “The first thing we do, let’s kill all the lawyers” [Henry VI (1) Act IV sc. 2]). Yet if you are going to argue that “education generates growth” because of the salaries that educated people earn, then you have to believe that the market pays what people are worth—and that lawyers are worth a lot more than teachers.

This does not mean that nobody learns anything useful at college. But try a thought experiment. Imagine that the country’s universities and community colleges suddenly shrink. There will still be doctors trained, and engineers, and some research scientists—but most of the high school graduates who currently go on to four-year bachelor’s degrees will complete college-track courses and  then head straight into the workplace.  

Do we really believe that, in this situation, national wealth and average salaries will all plummet to the levels implied by the falling wages of today’s non-college workforce? If so, then what it is about today’s average college education—with its huge classes, lack of contact with professors,  often cursorily corrected term papers and tests, and general grade inflation—that is working such miracles?

Globalization
But hasn’t globalization changed everything? As the College Board puts it, “As the world has arisen, new global competitors have emerged with a compulsion to excel. ... China is not alone. Canada and the Russian Federation display the same compulsion, as do Japan and Korea.” And then there are Great Britain and “Norway, Ireland, Belgium, Denmark, Spain and France”—their list, not mine. “A compulsion to excel elsewhere in the world has transformed education globally; it is time we developed that same compulsion,” the Board abjures us.

It is worth remembering that we have been here before, many times. Back in the 1950s, the successful launch of Sputnik convinced the US that it was falling disastrously behind the USSR and led to a huge program of science and math reform in schools. In the 1980s, Japan was the feared rival—and not just for Americans. The French concluded that they had to catch up with Japan’s high school graduation rates and poured enormous effort into doing so.

It is not obvious what this did for the French economy, which has, for the last fifteen years, had unemployment rates well above the OECD average. Meanwhile the Japanese themselves were on the cusp of a decade-long recession from which the economy has never fully recovered. You might think Japan was no longer anyone’s top choice for an economic scare story. But no:  “Australia, Japan and the United Kingdom are close on our heels” warns David Leonhardt in his New York Times plea for more and more education.

It is perfectly true that rich countries have more educated people. They have jobs that pay well, and they can also afford lots of teachers, schools, colleges, and campuses. But the idea that pouring education in at one end produces wealth at the other is simply not borne out by the facts. It may look that way when you compare Pakistan or Ethiopia with Korea, Canada, or the US—high levels of illiteracy and poverty versus universal and lengthy education and wealth. Yet there is no obvious statistical relationship between a country’s growth rate (or income per head) and the proportions of the population entering higher education or education spending.

This is true whether you look at the world as a whole or compare like with like—developing countries with each other, or countries within the “rich” group. The World Bank’s studies show that some of the poorest and economically least successful developing countries (such as Egypt) have had very high levels of university participation. Within the group of developed nations, the richest per head—Switzerland—has the lowest university participation rate and has done for years.

Policy-makers and the media often whip themselves into hysteria over their scores on international attainment surveys. This was evident in America in the 1980s, Britain in the 1990s, and Germany in the 2000s. Yet in the 1980s results, children in the countries that were to grow fastest in the following decades scored no higher than those in countries heading for sluggish growth or recession. Indeed, 1980s America was on the brink of enormous technological innovation and renewed economic dominance. In the most recent international TIMSS surveys, American children scored rather well—and comparatively much better than in the past. But I would strongly advise against using that as a reason to bet on the 2030 US stock market.

As we have seen, education is generally excellent news for the individual who has it; and the number of parents who don’t care about their children’s future is very small. So as countries get richer, people spend more on education, and pressure governments to increase school spending and student loans, to help their children have a good chance of doing well in life. Having a well-educated population is attractive to both employers and investors, but the causal chain goes at least as much from economic growth to more education spending as it does the other way.

As for the Chinese, let us get a sense of proportion. One favorite statistic is the numbers of engineers pouring out of Chinese universities, which is typically put at 650,000 a year for China compared to 70,000 for the US. But China is extremely large: over a billion people compared to the US’s 300 million, the UK’s 60, or Germany’s 80. If China is to be a developed country, it will need more engineers (and doctors and pilots) in total than the US or the whole of Europe.  Meanwhile, to see how far China has to go, just look at its output in terms of science journal citations (see Figure 4).



Education and Inequality
Everything I have said so far applies globally. There is, however, one more-localized phenomenon. In some countries, including the US, there has been a marked increase in income inequality in recent decades.  American workers with just a high school diploma are less well off compared to graduates than they used to be and have less of an advantage than before over those who drop out.  

In an influential recent book, The Race Between Education and Technology, Harvard’s Claudia Goldin and Lawrence Katz argue that this is because there has been an undersupply of college-level skills in the US economy and that most other developed countries have not experienced this because they increased their education spending and attainment.  “Because education continued to advance in Europe precisely when it lagged in the United States, wage inequality should have increased far less in Europe than in the United States. That is precisely what happened,” they argue (p. 329).

Goldin and Katz’s book contains a wealth of fascinating analysis of the role of education in American history, but anyone from a European country, reading that page, would—like me—do a double-take. Are they serious?  And where is this “Europe” place?  (It apparently excludes the UK but takes in the rest of the continent.) Education is not run by EU bureaucrats in Brussels but by individual nations. To date, there is little substantive convergence among their very different university systems in spite of “Bologna Agreement” rhetoric. European countries also differ a good deal in whether, when, and in what ways there have been increases—or decreases—in wage inequality over the last thirty years.

The comparison between the US and “Europe” is central to Goldin and Katz’s argument that more education is needed to address American inequality, but it simply doesn’t work. The idea that, for example, “educational supplies” grew rapidly in Germany after 1980, and hence it experienced less growth in inequality than the UK, is simply odd. Germany has an apprentice system which remained pretty much unchanged throughout the period in question, while university enrollments were at a lower level than the UK’s in 1970 and grew more slowly during the three decades thereafter.  German universities were also, as the country was well aware, in a parlous state.  (Much of Germany’s good scientific research takes place in free-standing institutes.) You simply cannot find a close correlation between European countries’ policies for supplying education and skills and changes in their wage structures.

So is education basically irrelevant to how we think about growth? No. Education does matter for the economy. But in developed countries, the way in which it matters most is under threat from an obsession with sheer numbers.

A growing economy is one of creative destruction. Companies emerge, and vanish, at speed because success is so closely tied to innovation. The remarkable thing about America is the way in which, over the last twenty years, it has produced new ideas and new products that have changed the world and has reasserted the economic pre-eminence that, in the 1980s, it believed lost.

The innovators in our societies are, on the whole, highly educated and the product of very high-quality universities. If your best people are really good, they will create an innovative, dynamic economy that generates jobs for everyone. So if you are thinking about education in economic terms, then what a country needs is, above all, to create and protect universities that are at the cutting edge of research and theory.

World rankings of universities all use different measures, each with their strengths, weaknesses and peculiarities. But they all agree that the best institutions are concentrated in just a few developed countries, all of which encourage competition between universities. Figure 5 summarizes the distribution, using two rankings that are very different in design and quite similar in outcome. The Times higher education rankings use one of the largest numbers of different measures, while Taiwan’s Higher Education  Evaluation & Accreditation Council uses only publications of scientific papers in core science citation index journals.



The US comes out on top, of course, but the absences are equally striking. The EU’s poor performance reflects the near-total absence of both Italy and Germany, both of which once had great universities. Tiny Switzerland—not a member of the EU, and with an unusually low college participation rate—nonetheless boasts several of the world’s top institutions. And China and India barely register.

President Obama wants everyone to go to college. This is an estimable goal: as he and everyone knows—including any high school drop-out—education is critically important for individual success. Moreover, college can still open people’s minds and feed their souls. But as a solution to recession?  Forget it.


Resources

Leadbeater, Charles and Wilson, James. The Atlas of Ideas (London: Demos, 2007).

Goldin, Claudia and Katz, Lawrence K. The Race Between Education and Technology (Cambridge, MA: Belknap Press of Harvard University Press, 2008).

The College Board. Coming to Our Senses: Education and the American Future, 2008.

Wolf, Alison. Does Education Matter? Myths about Education and Economic Growth (London: Penguin, 2002).

Read Comments  |   Submit Your Comment