David R. Henderson  

Maskin's Failure

A certain laxity in word choic... Should Financial Literacy Be P...


In particular, if I'm a bank and I'm making risky loans, I have an incentive, if I can, to make those loans using other people's money, in other words to make highly leveraged loans. But when I do that I don't take into account the fact that if the loan goes bad it's not just me who fails but all of my creditors. In other words I'm creating systemic risk by making these loans on borrowed money. That's an externality which I have no incentive as a bank to take into account. And the only way that that externality will be taken into account is if someone else, a regulator for example, puts a limit on leverage. That's a simple lesson from economic theory that was ignored six years ago. I hope it won't continue to be ignored.

This statement is by Eric Maskin, a past winner of the Nobel Prize in economics. He made it at a recent conference at Mercatus. I'm assuming that the author of the article in which this statement is quoted, Tim Cavanaugh, quoted him correctly.

Maskin is wrong. He has not identified an externality. It's true that lenders who are borrowing other people's money are subjecting those other people to risk. It's not true that that constitutes an externality. Those other people--creditors--know that and can price the risk accordingly in the terms of the loan. I'm not saying they will. They might make mistakes. But mistakes are not an externality per se.

It's also possible that federal deposit insurance will warp incentives by bailing out the creditors. Creditors, knowing this, will not be as careful in lending to banks. But then that's an externality created by government. (It's one, by the way, that even FDR was uneasy about in 1932, before he was president. He wrote that deposit insurance would "lead to laxity in bank management and carelessness on the part of both banker and depositor." Not surprisingly, given how spacey FDR was, that didn't stop him from signing a law creating deposit insurance the very next year.)

So Maskin's comment, "the only way that that externality will be taken into account is if someone else, a regulator for example, puts a limit on leverage" is false. The way you would get the externality taken into account is by removing the warped incentive that deposit insurance creates.

His mistake reminds me of one that Tyler Cowen made last month. Tyler wrote a post titled, "What is the market failure in data storage and protection at the retail level?" and tried to identify the market failure. He did not succeed. As I wrote on his site:

I see how your proposal could make sense, but I don't see how this is a market failure. The company and the customers bear the costs and there are market transactions between them. So it's hard to see where the externality is.

One of the participants at the event contacted me and told me that, indeed, Tim Cavanaugh did quote Eric Maskin accurately. Here is the link and you can go to the 36:00 point (approximately) for the relevant quote.

Comments and Sharing

COMMENTS (15 to date)
Gene Laber writes:

Excellent points. The externalities argument is grossly overused in discussions of economic issues.

Nick J writes:

I don't have the full quote he gave so I'm guessing a bit here, but I don't think you're giving him full credit. Here's how I usually see this externality described:

Say there are three financial institutions, A, B, and C.

Say that A borrows from B, and B borrows from C.

A invests the money in a risky manner. I agree with you that B prices the interest rate on its lending accordingly.

But C doesn't see that B is borrowing from A, either because it's illegal to see it, or because it's too complicated to trace out the full network of financial relationships with everyone they do business with.

So when A makes risky investments, it's exposing both B (who correctly prices that risk) AND C, who is indirectly exposed but unable due to informational constraints to price it correctly.

david writes:

There's an externality from the perspective of capital allocation because it is impossible to contract the bank into optimal behaviour. This is not imposed by a government, but there is a restriction on the bank's ability to sell a pledge of good investment behaviour to a customer who wishes to buy it nonetheless.

the costs of this restriction can be substantial, the full cost of not trusting any bank is the same as the social loss that would be incurred if fractional reserve banking were illegal and all banks were full-reserve banks. Fractional reserves reduces the loss but does not eliminate it (by definition)

Greg G writes:

David, you've made a reasonable enough argument for why you prefer the definition of eternality that you do. It's worth noting though, that a financial crash like the one we had affects everyone, not just those with investments in banks and mortgages.

Either way, which definition of the jargon people prefer is less interesting to me than the policies they prefer. Assuming we have deposit insurance (which is and was the reality after all) do you think we would have been better off or worse off if regulators had permitted significantly less leverage in the financial industry?

rvman writes:

You are right, there is no "externality", here. An externality occurs when a voluntary transaction between "A" and "B" harms "C". Here, if "B" knows or has reason to believe that "A" is taking these risks, he'll price that risk in. The problem is that this isn't what actually happened. "A" told "B" (and in many cases believed) that "A" was applying a set of restrictions "X" that resulted in a certain level of risk. "A" and "B" may or may not have properly estimated the risk from "X" restrictions, but it doesn't matter. "A" did not apply the "X" restrictions, which inflated risk well beyond that which it represented to "B" it was using. This is fraud. And fraud is very much a case where government oversight is generally accepted as efficient.

MG writes:

A Nobel prize winner should have been (at least) as careful as Nick J to recognize that his example does not necessarily represent an externality -- a condition in which the true costs can not be internailzed -- without assuming additional conditions ranging from fraud to moral hazard which corrupts the pricing of lending money to risky lenders. He is even more wrong in saying that the "only" way to internalize the true costs of default is to regulate more or differently.

Nick writes:

rvman, the externality here is that if A borrows from B, then B is at a greater risk of default if A defaults. Then if party C had lent to B, he is now at greater risk of not having that money paid back to him because B might default.

Daublin writes:

I don't believe these statements are inconsistent. Even if you correctly price the systemic risk into higher interest rates, it's still an externality. It's still something that everyone in the market is suffering from, and that no individual agent is in a good position to do anything about.

I have a different problem with the quote, though, and it's the casual assertion that regulators just never thought about systemic risk before. That's the complete opposite of the truth. Banks were pressured to take on a stock of mortgage securities because regulators thought those securities were reliable.

Banking regulators have been thinking about systemic risk since at least the advent of deposit insurance. They're just not necessarily very good at it.

Rick Hull writes:

Nick J explains the intuitive unease that we have regarding shaky financial foundations, and that very case is worth exploring strictly in terms of externality and the need for regulation.

The problem is one of counterparty risk. Every credit transaction is subject to counterparty risk -- namely that the borrowing party will refuse or be otherwise unable to repay the loan and extinguish the debt. This is not an externality because every lender chooses to incur this risk, and good lenders price for it. Lenders which do not price for risk are not a problem for regulators to solve; markets themselves weed out bad lending practice. We cannot make risk go away by fiat.

I think I have the same position as Henderson, just using different words, which may be useful. With this understanding, are there any broad externality arguments or questions remaining?

wd40 writes:

Whenever the the price is fully or partially determined ex ante and not fully conditional on the individual's behavior, an externality will arise. An easy example can be found in insurance. The insurance company knows that 1/2 the people will park on a street when it is cost effective to park a few blocks away where the car is less likely to be stolen. But suppose that the insurance company cannot determine either ex ante or ex post who these people are. Then there is moral hazard, a type of externality, as the cost of careless behavior by the insured and the car being stolen is shifted to the insurance company and possibly to the other policy holders. Even though the company on average is compensated, the careless individual has imposed a negative externality. Now in the car example, one can determine where the car was parked and not cover the theft or raise the insurance rates in the future. But when a bank goes bankrupt, such methods are not so readily available.

Banks were pressured to take on a stock of mortgage securities because regulators thought those securities were reliable.

Exactly right. Banks didn't make 'subprime' loans (maybe one or two for every hundred home loans) prior to a spate of legislation and regulation in the 1990s, and that nettled the 'community organizers'. Who launched a political assault on congress and regulatory agencies to make lending standards less onerous to low income people (and people without down payments or good credit reports).

In fact, banks that wanted to expand often found themselves collaborating with those community organizations to pressure regulators for allowing such expansion into Too Big To Fail institutions. I.e., far from being the only ones able to prevent the risky behavior, the regulators caused it. Maskin has it exactly backwards.

Calomiris and Haber are very good at detailing it in their recent Fragile by Design: The Political Origins of Banking Crises and Scarce Credit

michael pettengill writes:
The way you would get the externality taken into account is by removing the warped incentive that deposit insurance creates.

So, given the money market funds were not FDIC insured, and regulated on the basis that depositors fully understood the risks, the money market funds being invested in bad debt was not something that would shock "depositors" into panic, so the funds that broke the buck were simply a sign of poorly run businesses and failed individuals who failed to keep cash on hand and thus failing to delay withdrawals until the funds were available.

I wish Democrats had refused to compromise and fought for tough terms upto nationalization of insolvent banks, leading to the Bush administration rejecting those terms, and that leading to a month of money market fund panic runs and money market funds freezing withdrawals, to see whether unregulated uninsured banking is better than regulated insured banking in stabilizing the economy.

My guess is the money market funds would have triggered the kinds of ripple effects that were common in the 19th century and in 1907 and 1930.

Then Democrats would have gotten the laws requiring nationalization of any financial institution that was insolvent that had financial obligations to the public.

But what a great test that was never let run its course....

Charlie writes:


In response to Tyler's post and in this one you seem to be assuming that all market failures are externalities. That's a strange definition. The most famous market failure paper of all time Akerlof's Market for Lemmons has no externality.

Tyler: "so I was thinking what kind of market failure might be present."

DH: "I see how your proposal could make sense, but I don’t see how this is a market failure. The company and the customers bear the costs and there are market transactions between them. So it’s hard to see where the externality is."

Well, I would say the Market for Lemmons is a market failure without an externality. Maybe you disagree.

vikingvista writes:

Was there any externality to the Federal government bailing out Bear Stearns then later choosing to not bail out Lehman?

Was there any externality to the Federal government forcing most of the largest commercial banks, whose CEOs considered themselves solvent, to take bailout funds they did not want so that they all appeared insolvent?

Was there any externality to forcibly transferring record peacetime levels of capital to the control of some of the nation's most infamous liars, vote buyers, and financial ignoramuses?

Was there any externality to rewarding the most oblivious deniers of the impending financial crisis the authority to fix it?

Mark V Anderson writes:

David - I listened to Maskin's comments on the update link you attached. Maskin did indeed say essentially what you quoted him as saying.

Just goes to show you that the Nobel Prize doesn't mean you won't say nonsense at least occasionally. I've heard similar from many pro-regulators, but it is a little shocking to hear it from an esteemed economist. In a free market economy it's the one who stands to lose the money that is the proper regulator of the process, so it is clear in Maskin's example that it is the creditors who should be controlling how the bank lends out its money. If they cannot control this, maybe the law that needs to be changed is one that gives creditors more control. Giving a third party control has so many more problems (as the Great Recession shows us), that it should be clear to any economist that this is a third rate solution. Of course I've never thought that the issuers of Nobels are all that smart themselves.

Comments for this entry have been closed
Return to top