Over the past couple of months I've been having an ongoing discussion with a gentleman by the name of Mark Peters. He’s a well-read amateur economist with a strong free market inclination, so we have much in common. In fact, the discussion has been particularly interesting because we do share so much similarity in economic policy preferences, and I actually find myself engaged in a discussion with someone who’s arguing economic policy slightly from my right, which is most unusual.

Now as much as we agree on the basics, we’ve managed to run on to some considerable length about some of the details, and in our most recent exchange Mr Peters recommended a book called Economics in One Lesson, by Henry Hazlitt. The following is my reply, and though it is tailored to our previous discussion in particular, it may be of some interest to a wider audience, so I’m posting it here.

I will be posting all our previous exchanges in the archives in the next day or so, to provide some context for all this.

Thanks again to Mr Peters, who has kindly granted his permission for me to re-post his previous messages on this site.

------------------------------

Mr Peters,
I had been interested to continue or previous discussion, and thought the best preparation for doing so would be to read the book you recommended by Hazlitt. I’ve read most of it, and I find myself agreeing with the great majority of what he has to say. I was a bit surprised by finding the bulk of the book so agreeable to my opinions, as you had recommended it to me based on it being a fuller explication of your point of view on a few of the matters we disagreed on.

I had only a few objections to the book, leaving aside its obvious and unavoidable appearances of age: many of his chapters were dedicated to fighting off arguments you don’t hear much of any more, while Kerry and Gephart and company offer somewhat more subtle arguments which are equally preposterous and equally destructive. But it doesn’t seem to cover any really new ground (for me, anyway). I have a modest amount of college economics background, and the book clearly is aimed at an audience which (a) doesn’t; and (b) believes a bunch of silly things about protectionism, minimum wages, and general intervention in the markets. I am also a staunch fiscal conservative, and I found that for me the book was essentially preaching to the converted, and doing so without much new supporting detail.

I also thought it was surprising to read several approving references to Wealth of Nations in the Hazlitt book you recommended, as I thought you had said you found classical economics too riddled with errors to be useful.

I recall that you protested that you’re not a libertarian, but Hazlitt certainly was. I hope you’ll forgive me for continuing to refer to your essential argument as being libertarian (small “l” anyway) in what follows.

It’s been my observation that libertarianism is primarily a philosophy, and that like most other philosophy studies, discussion of it centers on logic and the absence of any internal contradictions. I said before that I think that’s what libertarianism has going for it primarily: it is one of the only systems for which its application never results in contradictory results. Unfortunately it does this by maintaining a simple premise and insisting on its universal applicability, which produces results that are admittedly self-consistent, but sometimes a trifle simplistic. But part of the problem with employing a simple argument and insisting on its universal applicability is that it invites refutation of the whole system simply by disproving one of its elements (or citing one instance where its application produces unacceptable results). I really don’t claim to have “refuted” libertarianism, of course, in any logically rigorous way. But I’ve identified enough outcomes which seem like weaknesses or errors to me, to say that I remain therefore unconvinced of the system as a whole. I think consequently the philosophy will never be dominant in the US, though it will likely continue to influence fiscal conservative policy—and that generally in a good way.

I agree that our main differences here are that you’re using philosophical arguments and I’m using economic arguments. I think it’s possible to carry on the discussion that way without both parties pointlessly talking past one another, because I agree with you that good economics depends on good philosophy. But I’d take your assertion one step further and point out that philosophy is just a group of ideas until it’s experimentally confirmed, and in this field economics is the experimental science. Aristotle, for all his obvious genius, was philosophically convinced (which was good enough for everyone till Newton) that a heavy object must fall faster than a less weighty object. Philosophy diverges from truth if it is never checked by experiment.

I recently pointed out to another correspondent that there is a difference between policy and philosophy, and that it’s important to know when to argue which. Economics is neat in that good philosophy (free capitalism) almost always results in good policy. But some of what I’ll discuss below is explicitly a policy decision, since we would ask the taxpayers of the United States whether and how to invest their money, and the outcomes will have significant economic consequences to total societal wealth. In these cases, I would suggest that allowing a philosophy argument to dictate an economically suboptimal policy is a poor decision.

If you want to argue that in cases of government intervention, the invisible opportunity cost is more expensive than the benefit, we can at least discuss that sensibly. But if you would argue that there should be zero cost to a program which provides a clear benefit, just because it’s the government implementing it, then you’re making an argument that doesn’t talk to me as an economist. To that I’d suggest that the argument is not well suited to a policy discussion.

Now, less generally, about the book:

I agree that minimum wages create inflation and unemployment; that tariffs hurt the consumers and economy they purport to protect; that redistribution of wealth (beyond a very minimal safety net, which as a taxpayer I do find myself willing to pay for) is fundamentally unjust. Our main disagreements in our previous series of essays eventually were reduced to some fine points, at what I called the extreme edges of capitalism—exactly where things get a little messy and I start to doubt the workability of a monolithic libertarian solution. I was hoping to see those particular cases treated in the book, since you had recommended it as a more elaborate treatment of your main thesis (which, if memory serves, we were debating primarily via monopolies and trust-busting activities by the government). I was therefore a little surprised to see that monopolies are not defended in the book anywhere that I could see. It’s not in the index, it’s not in any of the chapters I read in their entirety, and I didn’t happen across it in the chapters I skimmed. If it’s in there somewhere, please point it out to me.

I did find one specific point in the book I disagreed with—again, endangering the logical validity of the whole all-or-nothing premise. This didn’t surprise me, as I had expected to find one or two around the edges of the argument, even though I agreed with the great bulk of it. In the chapter “Credit Diverts Production,” Hazlitt essentially makes the argument that government loans always distort markets and should never be made. I disagree, having myself benefited from the loan guarantees provided by the Federal government for me to attend college—and, I hope, having benefited society in some small way in return. As I think that observing a single instance where libertarian philosophy produces suboptimal results sufficiently justifies my skepticism of the system as a rigorous basis for economic policy, I want to examine this case of Federally guaranteed student loans with some particularity.

There is a clear benefit to society for this program: a more-educated workforce will certainly produce economic benefits compared to a less-educated workforce (more inventions, more scientific basic research, a cure for the common cold to allow more days spent at work, whatever). More people with more knowledge allows more wealth to be created than the converse. I hope that idea is sufficiently self-evident I don’t need to really elucidate it further, but I hope you’ll tell me if we disagree here.

The trouble is that college is expensive, and college students don’t normally have much in the way of established credit or demonstrable ability to repay a loan from an independent financial institution. The government’s provision of a guarantee for a loan I could otherwise not have gotten has allowed me to finish a degree, get a good job, help both myself individually and therefore also the economy in general, and (incidentally) pay back the loan to the bank with interest.

Now I would expect that your main objection to this is in opportunity cost—what else society (meaning collectively the taxpayers who essentially all chipped in to guarantee my loan) could have done with the money. As it was only a Federal guarantee, and the Federal government itself didn’t issue me the money, the opportunity cost is initially with the bank who lent me the money at lowish interest rates. But the bank’s loan to me always yields a higher interest rate than (say) Treasury securities, while still being nearly risk-free. And if the bank really doesn’t need to add more low-risk, low-yield securities to its portfolio it can (and often does) sell the loan to another debt service agency who is actively seeking to add such assets to their portfolio. So it’s essentially voluntary on the part of the lending institutions, at zero cost to the US Treasury until someone defaults.

Yes, some people default on these loans, but we’ll get to that in a minute. So far we’ve established that many students are helped and therefore the economy benefits; and that the banks making the loans aren’t really out any kind of opportunity cost since they can rebalance their portfolios easily if they want to. I’m interested in considering how Hazlitt’s argument of “crowding out” more efficient and creditworthy loan applicants works in this case.

Hazlitt illustrates this argument by example of a pair of farmers who want to get a loan to buy a new tractor, or possibly an entire farm. One farmer has good credit, by having “efficiently” managed his affairs; the other farmer, presumably by being “inefficient,” has poor or borderline credit. Under the bank’s normal credit standards, the second farmer will not qualify for a loan. The Federal government guarantees his loan on his behalf, and the bank is thereafter willing to make the loan to either farmer.

Now Hazlitt’s point, if I understand it correctly, is that this has the potential to crowd out the efficient farmer, since there are only so many farms and so many tractors to go around, and now the less efficient farmer has an equal shot to be the one to get it. I was frankly shocked to read Hazlitt suggest that only farmer A or farmer B can get the tractor, but not both: our whole economic system creates wealth by creating things, and this capacity is essentially infinitely expandable in the long run. I think Hazlitt makes a fundamental error here; tractor production is not a zero-sum activity.

I think this makes no sense when talking about buying a tangible item like a tractor, but it makes some sense admittedly with respect to buying farmland. Now the important aspect of this is that he suggests what is really being lent is the farm or the tractor itself, rather than just the money, since there are only so many farms to go around. Now that we’ll have all these inefficient farmers getting the loans to buy one of the fixed populations of farms out there, while relatively fewer efficient farmers can get theirs, capital is therefore less efficiently employed. The efficient producers are crowded out.

So, to return to our particular modern example, how does this “crowding out” phenomenon hurt the “efficient” population of students with good credit? It makes their less-efficient peers more able to get loans to finish college, yes; but this doesn’t really prevent the students with good credit from also finishing college in most cases. The worst it might do would be to render a good student with good credit equally likely to get into and finish at Harvard, as the good student with bad credit (though I recall reading that Harvard considers “an applicant’s ability to pay” as one of its criteria when considering who to accept). And once everyone’s enrolled into whatever college they’ve chosen and which has accepted them, nothing hampers the good students from outcompeting their classmates (making valedictorian, therefore getting that fabulous job, etc) regardless of what everyone’s credit score is.

So there may be some small amounts of crowding out in some pretty unusual cases in the student loan guarantee program. But we’re looking for enough hidden opportunity cost to override the benefit to society provided by the program as a whole, and the small crumbs of opportunity cost we may have uncovered in the “crowding out” argument don’t make it. If those crumbs are enough to convince you to tank the entire program, then I respectfully redirect you to the philosophy vs policy discussion above.

Now for the defaulters, and the case for opportunity cost here is much stronger. The default rate is (not unexpectedly) pretty high for these programs, though it has dropped below 6% for the past several years (see this graph from the Department of Education), aided by low interest rates and an improving process of loan selection by the Federal government. This selection process is evidently based on certifying an entire school as an institution whose students are eligible for Stafford loans, and disqualifying entire schools which have too high a default rate. From just a cursory Googling I turned up an article about this, here, just as one example.



I’d agree that the early 1990s default rates according to that first graph above were unacceptable, and even current default rates are expensive: a 5.4% default rate on a loan program issuing $48B per year in loans (refer to the 2004 OMB budget in the Dept of Education section, here) amounts to a budgetary expense of $2.6B annually. While this undoubtably a lot of money, it’s important to keep it in context: it comes out of a Federal budget of $2298B (OMB again), and goes to support an economy worth $9837B in 2002 (The Economist Pocket World in Figures, 2003 edition, pg 222). By contrast, we spend $748B a year on retired folk via social security and medicare. Although I think those programs are horrible and not affordable, again we have to talk policy from where we are right now; and the cost for the student loan program is trivial in comparison. I think that in order to keep the US university system, and by extension American workers in general, the highest quality and most productive in the world, the $2.6B annually is affordable. Surely at least 0.02% of our annual GDP can be ascribed to having an excellently educated workforce.

So I saved the strongest argument against for last and came up with some comparative figures which (to me) make the student loan program look reasonable and affordable. I guess I would prefer to have hard figures as to what specific incremental contribution the loan programs make to the economy, but I am not aware of any immediately and the case seems persuasive to me even without it. It seems to me that the “productivity miracle” the US has experienced the past decade cannot be wholly unrelated to the education of our workforce.

As much as I’ve enjoyed our discussion (and would welcome another of your consistently thoughtful replies), I conclude that I still find rigorous libertarianism to be kind of troubling at its edges. I think it’s a good guide to policy, but I don’t find the philosophy itself sufficient to blindly use it as policy. I hope my tedious and detailed treatment of this one particular case can illustrate why.

Regards,
JKS.
10/17/2003

Return to the index for this subject.