Reviewing Ryan Avent’s Review of Amir Sufi and Atif Mian’s House of Debt: Friday Focus: May 16, 2014

For reasons I do not completely understand, Amir Sufi and Atif Mian’s House of Debt is being sold as a “consumer” alternative to the “banking panic” theory of the crash of 2007-2009. See, for example, Ryan Avent:

Ryan Avent: Free exchange: The opposite of insurance: A new book argues that household debt, not broken banks, fuelled the recent recession…. The notion that financial crises are transmitted to the broader economy by the “bank-lending channel” has dominated… policymaking…. Atif Mian… and Amir Sufi[‘s book]… challenges this orthodoxy. The real cause of the post-crisis slump, they argue, was… the run-up in household debt that came before, which became an anchor on consumption when home prices subsequently collapsed….

By analysing the relationship between consumption, employment, home equity and other debts and assets by county, Messrs Mian and Sufi find that consumption fell most sharply in counties that experienced the biggest drop in residents’ net worth. This, they argue, would not be the case if the primary restraint on consumption was damage to the banks…. In counties where net worth fell most, jobs were slashed not by small companies, which depend most on banks, but big ones—because sales were slumping…

It would have been polite, I think, for Ryan to have here noted that my sometime coauthor Dean Baker has been beating this housing-wealth-and-consumption drum since 2004…

But to the argument–to Amir and Atif, and Dean, and Ryan–what I want to do is to respond: “yes, but…” and “not really…”

As any monetarist will tell you (whether you want them to or not), an economic recession is an episode in which less money is supplied than the economy’s decision-makers wish to hold at full employment. In such a case, the consequence is a “general glut”: a critical mass attempt to cut their total spending below their incomes as they try to build up their money balances with no counterbalancing mass trying to spend more than their incomes, but such my spending is your income the result is depression. An excess demand for money accompanied, by virtue of Walras’s Law, by an excess supply of pretty much everything else. This process can be interrupted only if the system “prints” the extra money decision-makers want to hold. The banking system can print “inside” money–if it dares, and if decision-makers trust it enough to classify transferable banking-sector liabilities as means of payment. The central bank can print “outside” money. Or we can wait, mired in depression, until price levels fall enough to raise real balances supplied to a sufficient level.

The key to depression, in my view at least, is not a decline in any component of spending: it is the decline in one component of spending below its decision-makers’ incomes while the normal economic channels are blocked that would normally lead some other component of spending to rise above its decision-makers’ incomes.

We can see this from 2005-2007. Residential investment as a share of GDP collapses. But as residential investment collapses, does GDP shrink? No. The grey zone of GDP shrinkage starts only in early 2008, even though residential investment had already fallen from 6.2% to 3.7% of GDP beforehand. From 2005-2007 those who had financed residential investment did not decide to stop funding residential investment and start building up their cash balances. They decided to stop funding residential investment and use their money to fund something else–net exports, or additional government purchases, or business investment.

And then, from 2007-2009, collapsing home values significantly impact household wealth and lead them to cut their consumption spending back below their household incomes. But why is it that this reduction in money flow into the shopping malls as consumers cut spending below income to build up their cash balances is not properly compensated by an increase in money flow elsewhere as foreigners, governments, businesses draw down their gross cash balances, and as bankers draw down their net cash balances?

And to understand that we need to look at the banking system. We already know about the size of the negative shock to spending from the construction sector. Mian and Sufi do an excellent job of documenting the size of the negative shock to spending from the household sector via the housing equity-underwater mortgage-wealth channel–and in the process do, I think, successfully demolish Tim Geithner’s claims in his Stress Test that aggressively refinancing mortgages would not have materially helped the economy.

But to explain a deep, long, persistent downturn you need more than a big shock. You need: (a) a big shock, and (b) forces that make that shock persistent, and (c) an absence of damping mechanisms elsewhere in the system, and (d) an absence of recovery mechanisms elsewhere in the system. Mian and Sufi, I believe, do an excellent job of tracing the household wealth channel via (a) and (b)–how the overleverage followed by the collapse of the housing bubble delivered not juts a construction shock but a consumption shock, and how the failure to resolve underwater mortgages made the consumption shock persistent. (And I reflect that Chrysler and GM got resolved with their top executives fired and their shareholders and option holders zeroed out; the New York banks got fed huge amounts of money while their top executives kept their jobs and their shareholders and option holders got made large whole, and underwater homeowners got nothing.) But to explain (c) and (d) we need more than Mian and Sufi (and Baker) can give us: we need the banking, monetary, and fiscal policy stories as well.

Thus when Ryan Avent writes:

Shifting focus from banks to households, in turn, leads [Mian and Sufi] to very different remedies, the most radical of which is to replace many loans with equity-like contracts in which lenders share losses with borrowers…

Again I want to say: “yes, but…” As Larry Summers, Paul Krugman, Joe Stiglitz, and Laura Tyson all like to say: you don’t have to fill a tire through the leak. restoring consumer spending via successfully rewriting mortgage contracts to rebalance household balance sheets would have been a wonderful thing to do. It would still be a wonderful thing to do. But it was and is not the only thing to do to get us out of our current mess.

But let me endorse Ryan Avent’s endorsement of Mian and Sufi’s bottom line:

What is needed, they argue, is to make debt contracts more flexible, and where possible, replace them with equity. Courts should be able to write down the principal of mortgages as an alternative to foreclosure. They recommend “shared-responsibility mortgages” whose principal would decline along with local house prices. To compensate for the risk of loss, lenders, they reckon, would have to charge a fee equal to 1.4% of the mortgage, or receive 5% of any increase in the value of the property…

All I can say is that I thought that this was the system that we had. I thought–from the Great Depression era history of the HOLC and the RFC, from the 1980s history of the Latin American debt crisis, from the 1990s history of the RTC, from innumerable emerging-market crises, et cetera, that we understood very well that this is what we should do. Whenever the financial system got sufficiently wedged we resolved it–we turned debt into equity, and we crammed losses down onto debt holders whose investments were ex post judged to have been ex ante unwise.

And from my standpoint the true puzzle is why Bernanke, Geithner, and Obama were so uninterested in pulling out the Walter Bagehot-Hyman Minsky-Charlie Kindleberger playbook and following it in housing finance from 2009-2014. Did they read no history?

And let me qualify Ryan’s conclusion:

[But] while debt may be dangerous for the borrower, it is the opposite for the lender. Savers prefer the safety and predictability of a debt contract, which is why they accept lower returns on bonds over time than on equities. Many debt contracts exist primarily to satisfy this desire for safe assets—most notably bank deposits…

This is true only as long as the economy is not overleveraged. When it is overleveraged–when debt is underwater–the mismatch between cash-flow and control rights means that even lenders should prefer to “resolve” and transform debt into some more equity-like claim.