Economists tell a lot of stories about how a fall in dollar spending
can hurt the real economy. And we should
be forced to tell these stories–we
shouldn’t take it for granted that a fall in spending causes a fall in output,
we shouldn’t assume
the can opener of general gluts.
Why bother explaining the root causes of gluts when so many
think they can see a glut by just looking out the window? Because gluts conflict with one of the best
ideas economists have ever had: That surpluses–of workers, of unsold homes, of
cars rusting on the lot–push down prices and move the product out the door.
Surpluses set in motion a process that ends the glut: Just
watch the last half hour of a garage sale.
Any force strong enough to fight against the power of prices
should be a strong force indeed, strong enough for all to see. But when economists talk about the “frictions”
that keep gluts alive, we usually talk about “sticky prices” and “sticky wages”
and cultural norms and public sector unions and a few other forces. Strong forces, yes, and forces I believe in, but
stronger than creative destruction and supply and demand? For years on end?
Here’s my favorite friction, one that exists by force of
contract, not because of worker sociology: Debt. Debt in the household, debt in the firm,
and–for state and local governments at least–government debt. Irving Fisher beheld debt destroying a
deflating economy, and he wrote an excellent paper
about it in Econometrica v.1 back in
1933–a theory of depressions better than anything I’ve seen in Keynes’s General Theory.
In a world of debt, a modest fall in sales threatens
bankruptcy while a big fall in revenues sets off a fire sale of assets. Massive leverage means fewer dollars are
adjustable; more dollars are tied up already.
Miss a few credit card payments and you’ve entered a new social class;
miss a dozen mortgage payments and the police will be there to escort you out
of your home. Those are some sticky
prices.
Perhaps that’s part of the reason that real
retail and food service sales have fallen about 15% below their pre-crisis
trend: It’s a part of cash flow that a family can adjust without the police
showing up at the door.
One question I wish I had a solid answer to: Does private
sector debt have big negative externalities?
Fisher reminds us that when national income is unstable, the answer is
at least “Maybe.”
And finally, I’d like to thank David for his kind welcome yesterday. Glad to be aboard.
READER COMMENTS
The Sheep Nazi
Sep 11 2012 at 12:21pm
Dr. Jones: welcome aboard. Just a quick question: why is private-sector debt (as opposed to public-sector debt) uppermost in your mind right now?
Doug
Sep 11 2012 at 1:16pm
And yet the state’s official policy encourages debt over equity in the capital structure. One because profit that pays debt is exempt from corporate taxes, and two because regulatory capital frameworks give a lower charge to (particularly rated) debt than they do to low beta equity.
JoeMac
Sep 11 2012 at 1:32pm
Garett,
Your story doesn’t explain two things. First, why don’t debt contracts have clauses that adjust for nominal changes, like inflation and NGDP? That would entirely solve the problem.
Second, concerning Fisher story, isn’t each person’s debt somebody else’s asset? If so, an increase in debt will be counteracted by an increase in somebody else’s asset.
effem
Sep 11 2012 at 3:27pm
Welcome!
I tend to agree with you. What I find most fascinating is actually the feedback loop between leverage and macroeconomic policy.
The more you believe that Fed policy will achieve stable NGDP, the more incentive you have to leverage up any business with >NGDP revenue growth and semi-fixed costs. After all, debt is very effective at aiding wealth creation, provided the “sticky price” problem never becomes an issue.
As Greenspan and the Fed got better at keeping recessions short and shallow, the financial system leveraged up. This isn’t surprising – greater confidence in the Fed = greater payoff to leverage. However, in 2008 the Fed wasn’t quick enough to ensure a shallow recession, mostly because of an oil-led inflation problem. Thus, we had a blow up.
However, that blow up was cut short as the Fed kicked back into gear to stabilize NGDP.
The proponents of NGDP targeting and the like have to explain why this won’t simply lead to greater and greater leverage thus creating a monstrous risk that if NGDP can’t be stabilized at some future point (inflation problem, natural disaster, political change, etc.) you have the mother of all liquidations.
My personal belief is that the greatest single oversight of economists’ is that the market adjusts for economic policy. If you smooth NGDP, we will lever it up!
Garett Jones
Sep 11 2012 at 4:34pm
@JoeMac:
1. Negative externalities would be one possible reason—you don’t index because indexing mostly helps other people, not the parties to the contract. In high-inflation countries indexing of mortgages is reasonably common, so it’s not impossible to index debt.
2. As Fisher notes, bankruptcy causes disruption: A supply side cost. So deflation only helps the creditor if she gets repaid. But also, as in the “debt overhang” literature, when firms or families are close to bankruptcy, they radically adjust their spending patterns, again causing supply-side disruptions. Negative externalities.
Analogy: Falling home prices are good news for the young, but the debt contracts disrupted by the price collapse sure set off a wave of churn.
Garett Jones
Sep 11 2012 at 4:49pm
@TSN:
It’s the nature of blogging: No need to blog topics in order of importance. I’ve written about the looming public sector debt crisis before (in a symposium with Arnold) and will in the future.
But the economics of private debt contracts are still underappreciated, especially the difference between ex ante and ex post incentives. The battle within the firm, the conflict of incentives between shareholders and bondholders, is a great topic.
Another reason: private debt is much larger than public debt, and what we learned in this financial crisis is that a substantial portion of so-called private debt is actually public debt.
Examples of de facto public debt:
1. All bonds issued by the biggest banks doing substantial business in the U.S.
2. Bank deposits (via FDIC)
3. Mortgage debt guaranteed by Fannie and Freddie
4. Short term commercial paper whenever there’s a panic.
And these are just examples we know of today. Part of what we learned from the crisis is that the list of de facto public debt grows in only modestly predictable ways.
So today, right now, some private party is issuing debt that would be guaranteed by the federal government in the event of a panic. They are making commitments on behalf of the U.S. government, perhaps without even knowing it…or worse, perhaps they do know…
I think that’s worth reflecting on, for both macro and public choice reasons…..
Joshua Wojnilower
Sep 11 2012 at 6:59pm
Garett,
Great post and nice seeing the focus placed on private debt instead of public debt, for a change.
In terms of negative externalities, I think household/private debt may be a case where on a micro level the effect is small but on a macro level can be very large (which fits with my understanding of Fisher/Minsky). The large negative externalities arise because the forced sale of assets (e.g. homes) by some causes the assets of other households to fall without a corresponding drop in liabilities. If these households were also highly levered, the corresponding effect can be further bankruptcies and asset sales. The large negative externalities arise in this scenario because aggregate private debt is high and an asset that accounted for a significant share of wealth and debt across the population was being sold/falling. I think the externalities were further exacerbated by relatively unequal distribution of debt within the economy.
Looking forward to future posts.
Harrison Searles
Sep 11 2012 at 11:27pm
If debt is a price then what is it the price of? Debt seems to me here just to be an ex post cost of borrowing rather than an ex ante signal that guides present action in the direction of equilibrium.
Joshua Wojnilower
Sep 11 2012 at 11:39pm
@JoeMac,
You are correct that each person’s debt is somebody else’s asset, but this obscures Fisher’s point. When the bank makes a loan, it creates both an asset and a liability. However, it does not create the necessary assets to cover interest costs. So although the bank’s assets (loans) increase in correspondence with rising liabilities elsewhere, the funds generally don’t exist to cover all liabilities. If assets must be sold to pay off liabilities, causing asset prices to fall, a further dislocation may occur (at least in an accounting sense since banks can avoid mark-to-market). Expanding the story further, in the recent crisis the distribution of assets and liabilities was heavily skewed. When a number of households were unable to payback their debts, banks’ assets were suddenly insufficient compared to liabilities, which called into question their solvency and thereby reduced liquidity in short-term funding markets.
To sum up, as Garett rightly points out, debt is incredibly sticky. The fall in assets raises the real debt burden, which causes the private sector to further restrict spending. This reduces aggregate income, which leads to further asset sales and a vicious debt deflation cycle has begun.
Peter H
Sep 12 2012 at 12:42am
Welcome as well!
To the specific question of externalities in private debt, I can think of a couple:
One that nobody’s mentioned are legal fees and court costs. When you file for bankruptcy, or file a suit to collect a debt, that imposes a cost on non-party taxpayers in the administration of that suit. The filing fees are generally not nearly enough to recoup that cost. Also some cost to taxpayers in the administration of liens, and to employers in administering garnishments. I’d say it’s arguable whether or not the attorneys for suits/bankruptcies are impacted as positive externalities, but as non-parties to the initial transaction, I’d say there’s a strong case to be made.
Lien enforcement sales I also think are arguably an externality as they impose a sale of an asset into the market, depressing the value of competing assets for other sellers. But I think this is not a well-formed case of an externality, for the reasons I’m elaborating below.
I don’t think changing consumption patterns due to indebtedness should be listed as externalities of the debt, but just as effects. One reason being that they’re just counterfactuals, not concrete consequences of the debt. That is, because you had debt X, Y was a specific tangible consequence that followed. The inward motion of a consumption frontier more generally may be a result of the indebtedness, but it doesn’t imply any one specific event Y that is made to happen, where Y impacts a 3rd party as Garret said in his first reply to JoeMac.
To qualify as an externality in this context, there’s a 3 part test I’d apply.
1. The transaction imposes a material consequence on parties who did not participate in the transaction, which material consequence could reasonably be foreseen as having a nontrivial probability at the consummation of the transaction. I.e., it did something to someone not involved and it wasn’t an absurd chain of events like a butterfly causing a hurricane.
2. The material consequence of (1) is concrete and measurable, at least in theory. So for pollution, you could at least theoretically measure the before/after particulate matter in an area.
3. The non-participating parties of (1) are specifiable and identifiable. If you wish to specify the set of all market participants in a very large market, or all persons on Earth, then that transaction had better be a REALLY big deal. So for example the invention of the lightbulb or transistor plausibly had positive externalities for all persons on Earth. A single default on a consumer loan does not.
The reason I’m harping on this is to keep the definition of an externality meaningful. An externality is not shorthand for a bad thing happening to someone. When an economist tells me something is a negative externality on the standard definition of the term, that is usually a good argument to bring in the coercive force of the state to stop the externalizing. And I don’t take that coercive force lightly.
Luis Enrique
Sep 12 2012 at 9:45am
have you seen SRW’s post: The Stickiest Price
http://www.interfluidity.com/v2/910.html
(sorry if you cited and I missed it).
follow up posts good too
http://www.interfluidity.com/v2/1548.
David Beckworth
Sep 12 2012 at 12:58pm
Garett, I am on board with the sticky debt idea and have advocated something like it before. Steve Randy Waldman has had some good things to say about it, see here for example: http://andolfatto.blogspot.com/2012/04/ngdp-targeting-some-questions.html?showComment=1335600504385#c3293738639515635388
I would also note that the Fed’s failure to keep nominal incomes along their pre-crisis expected path is part of reason this sticky debt problem is so pronounced: http://macromarketmusings.blogspot.com/2012/05/dereliction-of-duty.html
Silas Barta
Sep 12 2012 at 4:32pm
I’ve never been a fan of relaxing the “stickiness” of the debt price, and especially not as an economic recovery policy.
Cold-hearted as it is, the entire point of such debt contracts is to provide one party a steady stream of payments, independent of the venture’s success magnitude (unless it fails and goes into bankruptcy).
So a pretty fundamental part of taking on debt is that it makes the debtor less liquid and less able to ride out times of uncertain or reduced demand. Unfortunately, borrowers seem not to appreciate this, but we shouldn’t indulge their misunderstanding.
Furthermore, “ability to ride out times of low/uncertain demand” is an economic good for a producer, and the price of that good provides a vital signal for the economy — so toying with this price is bad for the same reason that blurring or shoving underground *any* price is bad — it ruins economic calculation. (Here, it impacts how robust businesses should make their plans.)
Borrowers need to accept that they are receiving a valuable good (upfront funds) in return for the loss of another valuable good — the ability to tolerate spans of low sales. If borrowers fail to appreciate the severity of this cost, then we should let overleveraged businesses fail until the lesson is clear.
It bothers me to see institutions large and small miss this: they think that, “hey, if we can get a higher return than the cost of the loan, go for it” — neglecting that they’re also betting that the future will be similar enough to the present for their business plans to work. No one should get free “protection from uncertainty”, whether under the banner of “NGDP targeting” or debt forgiveness, or aggregate demand goosing, or bailout of vital industries, etc.
Joshua Wojnilower
Sep 13 2012 at 10:38am
@Harrison
“If debt is a price then what is it the price of?”
Debt is the price of some future payment streams that can include interest and principal amounts. One can see the price of debt in various asset markets (e.g. Treasuries, MBS, corporate debt).
Carl Lumma
Sep 13 2012 at 4:03pm
I had this argument with David Glasner over a year ago. See here and here
Glasner: The distinction that I am making is, I think, a fairly standard one in economics, namely between the resource allocation effects of a change and the distributional effects. If the preferences of the parties between whom wealth is being redistributed have identical preferences, then the final equilibrium would be identical to the initial equilibrium except for the shift in consumption between the parties. That is a pure redistribution effect.
me: …there is a factor of 10 (the money multiplier) operating on this wealth transfer — ten units of wealth are lost to lenders for every one gained by savers … that might produce substantial real effects without any need to invoke things like menu costs (which seem to me rather contrived, and must have plummeted over the last ten years…)
-Carl
Carl Lumma
Sep 13 2012 at 4:20pm
@JoeMac: Why weren’t steam engines developed in the middle ages? That something hasn’t happened (yet) is a poor argument against it.
Comments are closed.