Sunday, June 15, 2014

Student Loans III: The Redistribution Question


We’ve been talking about student loans, and the fact that the federal government is turning a multi-billion dollar profit on them (especially law-school and other graduate loans).  These profits may be the equivalent of a progressive tax on borrowers.  Is that a reason to support the current design of the program?  Assuming that the program actually is progressive (admittedly, an unanswered question, as we saw last time), I would say yes.  I’ll argue co-blogger David Gamage should say yes, too, but that many other folks would disagree.

              Standard public finance economics analysis would suggest that government shouldn’t use its own market participation as a way of redistributing wealth.  The reasoning is what I’ll call the “heavy buckets” theory, but most people follow Kaplow & Shavell in terming it a “double distortion” argument.  Let’s say you have to schlep some water from the river to your house.  Would you rather use a heavy steel bucket or a nice light plastic one, assuming both hold the same amount of water?  That’s a no-brainer.

            The double-distortion argument is roughly the same.  The claim is that any time we redistribute income, we will change people’s incentives to work, as the expected after-tax returns for earning more are diminishing.  So whether government imposes a pure transfer, or just charges more for the brownies at its bake sale, either way the after-transfer value of my salary is lower.  That’s the water in the bucket.  The bake sale, though, is the heavy bucket, because it has a second drag on the economy: it also discourages me from eating brownies. 

            Anyway, that’s the standard story, but Gamage has a different one

He points out that the thing we use for the supposedly “pure” transfer---the individual income tax---has some extra distortions of its own.  People change their behavior to minimize the tax they face, in ways that have nothing to do with how hard they work.  They might demand payment in bitcoins to make their earnings hard to trace, or claim more in depreciation deductions, or decide to become self-employed.  All of these things can be drags on the economy, too. 

        Gamage focuses on alternative sources of tax revenue, but we can extend his point to governments as buyers and sellers.  The bake sale doesn’t cause any of the unique distortions of the income tax.  By bringing in money at the bake sale, the government can reduce its reliance on the income tax.  If rates are lower, people have less incentive to do all the behaviors that are unique to income-tax reduction.     

            Critically, not all of these distortions are equally socially costly.  As a rule of thumb, the  social cost of a distortion rises with the square of the effective “tax.”  That means that a small profit at a bake sale, although costly, can trigger much bigger income-tax savings. 

            We can maybe extend this same line of reasoning not just to selling, but also to buying (although here the analysis is more complex, because the government must first raise the money to pay for stuff).  When the government buys labor from poorer communities, perhaps it should overpay, as a way of reducing the need to redistribute income-tax revenue to those communities.  And, by the same logic, it should underpay professionals.  Taking this one step further, we could potentially help to justify labor law as a tool for reducing the distortiveness of the income tax.

            Anyway, we’ve come a long way from student loans.  Bottom line: if the student loan business is profitable, and progressive, that should count as a reason to keep it.  And quite possibly we should think about ways to make it more progressive than it is now.

| Permalink

TrackBack URL for this entry:

Listed below are links to weblogs that reference Student Loans III: The Redistribution Question:


Maybe I should add a caveat: two big questions are whether the government enterprise really makes economic profits, and if it does it fair and square. If government employees work the bake sale but charge their salaries to the Treasury Dept. and leave out the opportunity cost of capital, then their apparent profit of $.50/brownie might be an actual loss of $.80/brownie.Or, it the government sells the brownies on credit to homeless people and has no bad-accounts reserve in its profit calculation, that's another reason an apparent profit might be an actual loss.

Also if the reason the government makes a profit at $1.50/brownie is because it bans private competitors or regulates them discriminatorily (e.g., by requiring that they keep 5 years of backups of their computer files while only keeping government files for 6 months), then the government operation is questionable.

So a good question is whether either of these applies to the government's involvement in student loans. I can imagine either a profit or a loss at the end of the day.

Posted by: Eric Rasmusen | Jun 18, 2014 12:58:12 PM

Post a comment