Leonhardt appears to be in the advocacy corner for creating circumstances under which more students can attend university. He has seized on a couple of research papers that he thinks support his position. But I am not sure that they do, based on his reporting of the results of those papers.
Leonhardt poses a number of persistent and legitimate questions about university education.
How much money should taxpayers spend subsidizing higher education? How willing should students be to take on college debt? How hard should Washington and state governments push colleges to lift their graduation rates? All of these questions depend on whether a large number of at-risk students are really capable of completing a four-year degree.What both papers appear to support is the relatively uncontested evidence that completing university has a material financial benefit, to the tune of a 20% income premium over those who do not attend university. But that doesn't answer the questions being asked.
The income premium is only earned by those who complete college, not those who enroll and don't complete. There is a real cost in terms of postponed income when attending university. There are loans that might have to be taken out. Lost income and future interest payments might be well worthwhile if the degree is completed and the credential premium earned. But only if completed.
While elite universities have six-year graduation rates in the high nineties, for most others the rates are down in the sixties and lower. You can't just count the benefit to those who graduated. You have to count the costs incurred by those who enrolled but didn't complete. Those costs can be crippling and take decades to pay off. Without taking into account both costs and benefits to the entire population of university attendees, either the researchers or Leonhardt have put their thumb on the scale to arrive at the predetermined conclusion that attending university is an ineffably good thing.
Leonhardt's original questions are good ones. There is demonstrable evidence that completing university increases personal productivity (and other good life outcomes). But what we need to know is what can government properly and effectively do to make it as easy as possible for everyone who can gain productivity from the university experience to do so while not incurring any costs for people who can't benefit from university (because they fail to complete). Until you unlock the issue of being able to accurately predict those who can complete university and properly and fairly distinguish them from those who can't, then you aren't going to make much progress. Currently we make it as easy as possible for everyone to attend university and we celebrate those who graduate and count their benefits from attending. We ignore the costs of failing to graduate. Costs to the attendee and the costs to the taxpayers who is on the hook for loans that will never be repaid.
And lurking under all this is Leonhardt's disdain for two-year colleges/community colleges. Places where you can increase your productivity without, ideally, incurring the crippling costs associated with the structured four year program. My interpretation is that Leonhardt suffers from that common prejudice among the clerisy that if you can give people the outward attributes of the middle class (university education, home ownership, etc.) then the increase in personal productivity will magically follow.
As many of the commenters to his article are pointing out, productivity flows from the behavioral attributes (self-control, self-discipline, perseverance, commitment, etc.) that enable successful completion of university. It is those attributes which lead to increased productivity, not the credential. Until you have a way of fostering those successful behaviors, you will just be wasting individual and taxpayer money by flushing ever more bodies through the higher education mill.
We want to find the easiest, cheapest and fastest way to enable people to increase their own personal productivity at any stage in their life. University might be the answer for some. Nothing in this reporting of the research tells me that they actually know how to figure who would benefit.