August 21, 2002
Go To College!

If you're listening to this, are under 35, and haven't been to college: go to college. If you have children who are thinking of not going to college: convince them to go to college.

Harvard economists Claudia Goldin and Larry Katz have called the twentieth century, "America's century of education." During that century the United States widened its lead over other industrial economies by creating the universal high school, and by developing a large and flexible system of colleges and universities. They believe that this educational expansion greatly increased the skills and adaptability of America's labor force throughout the twentieth century, and was a large part of the reason that America was and is richer and Americans more productive than people anywhere else in the world.

But in the past few decades we've seen a slowdown in the growth of education in America. This slowdown has been accompanied by a sharp rise in the difference between the earnings of young workers who have and have not been to college. We economists are scratching our heads: since the--economic--benefits of going to college have become so large, why aren't more people going?

One possibility is that people do not realize that the earnings gap--even after controlling for every other factor that labor economists can think of--between those who have and have not been to college now approaches fifty percent. So I'm telling you: in today's world, every American who can should go to college. You can barely afford not to.

I'm Brad DeLong.

Posted by DeLong at August 21, 2002 05:09 PM | Trackback

Email this entry
Email a link to this entry to:

Your email address:

Message (optional):


Do you (or Golding and Katz) have any numbers on rates of education for the US and other countries during the last century? Or better yet, skills comparisons that might demonstrate that school has some useful effect? I'm not at all sure that much if any of America's economic strength can be attributed to universal public education or directly to its universities. Universal education seems to me to be as strong or stronger a social and political value in other countries as in the US. Besides, staying in school until the age of 22 or longer can just as easily be seen as a drain on productivity unless you can show that education directly contributes to greater productivity later in life.

I think a strong case can be made that direct state funding of academic research has had an impact on the American economy and that other nations have failed to place a similar priority on research in universities. Among other things, this open-ended funding enables the US to attract many of the best graduate students from abroad, despite the byzantine visa system the INS maintains.

But - and I say this as someone who attended public and private schools in the USA from Grade 4 until my Master's degree - America's secondary and undergraduate schools are on the average just pitiful. I am not at all convinced that the years most Americans have spent in school were time well invested from a social point of view. Few if any job skills are taught after grade 6 and none of my schools ever gave me much in the way of depth or insight into the world.

As individuals, even if your Bachelor's degree is worthless, you have to have one in order to qualify for a decent job. In that sense, college is a good investment, but that can easily be a consequence of degree inflation. It certainly doesn't demonstrate a causal link between education and productivity.

Perhaps you have cause and effect reversed. Perhaps the increased benefits of going to college are a consequence of less emphasis being placed on education? I worked once for a company that required a college degree of all of its employees, regardless of their job or level in the firm because this was (they felt) the only way they could guarantee that their employees were fully literate. If more people went to college, the effect might well be to drive firms to require Master's degrees for skilled work, forcing capable people to invest still more time and money in degrees they neither need nor want.

Posted by: Scott Martens on August 21, 2002 06:02 PM

How much of the earnings gap though is purely due to the signalling effects of education? In other words, if 100% of Americans had a college education, would everyone earn what today's graduates due (i.e 50% more than today's non-college people) or would todays' graduates salaries fall as they became less exclusive, so in the end college graduates (i.e. the entire population) would only earn 25% more than today's non-college graduates?

Posted by: Matthew Turner on August 21, 2002 11:55 PM

What about the rapidly increasing cost of tuition and/or some families' inability to pay for college? Is this a big problem? If it is, is there anything that can or should be done about it?

I once worked for a Senator who wanted to make college tuition 100% tax deductible. I asked his policy director whether the increased demand would increase tuition. His answer was that the economics of education is very complicated and that there are so many other factors increasing costs that the tax deductibility would have neglibible consequences (I'm pretty sure that he didn't know what he was talking about, though I have no idea whether he was right or wrong).

Posted by: Bobby on August 22, 2002 12:36 AM

The senator could have been right. Remember that the headline "tuition" figure is not always particularly representative of the actual cost paid by the majority of the students. Making college tuition deductible would amount to a subsidy to the richest parents of college students, which would be recouped by increased tuition and (potentially and partially) recycled into increased financial assistance; the colleges would be able to undo the redistributive effects of the subsidy with ambiguous overall effects on the cost of college for the representative student.

Posted by: Daniel Davies on August 22, 2002 04:41 AM

From comparing my experience in the British system to that of my fiancée in the US system, I'd support the contention that college offers the best bet for US students, even at current tuition costs, but I do worry about the devaluation of the US undergraduate degree, particularly because it's accompanied by an emphasis on graduate degrees that rather suits universities, since they can brazenly charge the equivalent of a small mortgage in tuition. And since turkeys don't vote for Christmas (or Thanksgiving), I can't see there being the will to reinvigorate undergraduate courses.

Posted by: nick sweeney on August 22, 2002 06:05 AM

I have to agree with the point of view that most American high schools are genuinely awful. We could probably get more productivity gains by focusing on teaching people basic skills and advanced skills in certain areas (I'm not talking job-specific -- I'm talking algebra, logic, elementary science). And while I think that some sort of postsecondary education is important, I don't think that four-year institutions are for everyone. For many -- perhaps most -- the better alternative is job-specific training.

Posted by: Paul on August 22, 2002 07:42 AM

It seems to me that the ROI on a college education is immense. If I assume that the real discount rate is about the same as the average wage growth (these may not be good approximations, but they seem semi-reasonable, but it allows me to get rid of the net present value calculation, which I'm too lazy to do at the moment...), if we say that the total cost of college is $20,000 a year, (which seems on the high end to me... many colleges charge more than that nominally, but usually mostly to leave room to cut it back with scholarships) and the wage of a high-school grad is $30,000 a year, then four years of college costs a total of $200,000. If both the HS and college grad work from 22 to 65 (43 years), and the college grad gets $35,000 a year, then during the 43 years of working, the college grad makes $215,000 more than the HS grad, paying $15,000 more than the total oppurtunity cost of going to college in the first place.

This argument is based on a rather high college tuition and high wage for a high school graduate working in the four years the college grad goes to college, and a rather low differential between the wages. I don't plausibly see how going to college could be a bad investment, since realistically, I would guess that the payoffs were higher.

As for the argument that more college grads == lower wages for college grads, I think that's 100% true and a good thing. Unless there's some way that more college grads would lower total productivity, this would mean that wages for non-college grads would also rise as the pool of people seeking low-wage jobs shrinks. It seems college education is a triumph for productivity and distribution.

If that was way, way, way too simplistic, which it probably was, someone please explain why I'm wrong. I'd be grateful.

Julian Elson

Posted by: Julian Elson on August 22, 2002 10:38 AM

I posted at Demosthenes's house. Summary:

Increased debt load during undergraduate years discourages people. Change in financing.

College/non-college is be a marker of no-problems/problems, in addition to being a cause.

As often with economics, lumping is happening. Most of the richest of the rich go to college, most of the poorest of the poor don't. But there's a lot of overlap. (Suppose you're averaging 10-60,000 for no school and 20,000 on up for school. With lots of people making more than 200,000, there will a major difference in averages even though there's a lot of overlap in the 20-60,000 range.

The BA degree is a joker here. In and of itself it has essentially no cash value except in a few odd fields.

Last, not mentioned over there, the non-college sector includes a very large number of legal and illegal immigrants who were brought here specifically to take low-paying jobs. It's not like their other alternative was to go to Harvard.

In short, the general truth isn't necessarily instructive in the individual case -- which is a truism, but significant here I think.

An addition point though, is that at the same income level college-educated have better self esteem than not, and often certain perks such as less-hostile management practices. "Hidden Injuries of Class"

Posted by: on August 22, 2002 02:50 PM


It's wrong because it treats college educations as homogenous -- "Getting a degree" may not help you as much as "getting a degree in a field that will be 'hot' five years from now" -- ie, the classic example of the Starbucks employee with a French lit degree. *Some* college students are doing very well, and they are driving up the averages. (For example, teachers don't make all that much money, while certain business majors do.)

Having said that, I promise to actually sit down with some numbers later this evening. I reserve the right to retract this argument at any point without feeling like a fool.

Posted by: Paul on August 22, 2002 02:53 PM
Post a comment

Email Address:



Remember info?