Jared Bernstein summarizes an awesome debate that is going on in the economics world and blogosphere right now:
There’s a very interesting, albeit down-in-the-weeds, analytic debate brewing around a confluence of recent publications. Tim Geithner’s new book defends the interventions of the Treasury Department he led to reflate credit markets (and I worked with the team on this back then). Mian and Sufi’s new book, reviewed here by Bin Appelbaum, argues that Treasury got it wrong by not recognizing the extent to which debt burdens were restricting growth and intervening in ways to write off more debt: “The fact that Secretary Geithner and the Obama administration did not push for debt write-downs more aggressively remains the biggest policy mistake of the Great Recession.”
Dean Baker has long argued the problem was not just the debt overhang but the wealth effect’s sharp shift into reverse when the housing bubble burst. That’s similar to Main/Sufi except it implies that even had you forgiven the debt, consumption still would have tanked. Brad DeLong articulates an “all-of-the-above” theory, suggesting each of these analyses gets at one part of the problem but you need all of them to understand what happened.
Here’s what I think. It comes from a line Matt Yglesias once wrote, reflecting on the Reinhart/Rogoff view that balance-sheet recessions that result from a leveraged bubble bursting tend to be particularly hard to grow out of. Matt said something to the effect of: “it’s like trying to drive out of a deep fog–you have to go very slowly or risk crashing. However, if you have strong fog lights, you’re much less restricted.”
“Fog lights” in this context mean you have to do everything you can to get the system back up. You have to reflate the credit system through both liquidity (as in the TARP) as well as Mian/Sufi-style principal reductions and cramdowns of mortgage debt that cannot realistically be serviced without sustained pain. The administration did a lot of the former and little (not none) of the latter.
Why not? From where I sat, and I didn’t see everything from my seat for sure, it was less about protecting creditors per se than the belief that breaking contracts was an interventionist step too far (though the fact that this principal was less active in the auto-bailout suggests a protection that extended more to white than blue collar industry).
It was also argued at the time that the administration tried to get Congress to write legislation allowing bankruptcy judges to allow cramdowns on primary residences but the members uniformly said no. You ask me, I think we didn’t push hard enough, again, based on the perceived sanctity of debt contracts.
Dean’s point means that debt forgiveness and revived credit flows must be met with deep fiscal stimulus that lasts as long as needed (he also emphasizes the demand-zapping role of the US trade deficit). No question that accelerating the deleveraging cycle will hasten the recovery, but as long as unemployment remains high, job growth slow, and real wages flat, demand will be seriously constrained by the negative wealth effect, even as debt forgiveness dampens that dynamic.
With the private sector still licking its wounds, absent committed stimulus there’s no reason to expect deleveraging, or even aggressive monetary policy, to trigger the growth needed to reach escape velocity. In fact, you might expect to see a pattern as in the figure below, which plots households’ debt service as share of disposable income against real, year-over-year GDP growth. Deleveraging would have surely been faster with more debt forgiveness, but without doing more to offset the crashing wealth effect, we may have still been in the slog you see there: basically, a return to trend growth before the output gaps were closed.
To be clear, the administration did, in fact, respond quickly and effectively with precisely this stimulative “fog light” but the headlights dimmed before the fog cleared. Why? Because of a precipitous pivot to deficit reduction. And, of course, aggressive stimulus was not accompanied by aggressive debt reduction.
So I’m with Brad—all of the above. And let’s keep it real: the problem was not only that we didn’t do all of the above. It’s that even when we did the right things, we didn’t stick with them long enough. The important thing now is to try to learn from our mistakes, and I for one am thankful to all of these authors for continuing to plumb these deep waters.
I was taken with Mian and Sufi’s argument until I read Professor Delong’s take:
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.
I wish that the politics/policy world had taken homeowner debt relief seriously. The proposed cram down legislation, in addition to dozens of other plans to help homeowners, could have improved millions of lives. Having the Delong historical perspective would have made the arguments for those policies that much stronger. With such a myopic and dysfunctional Congress, it seems to me that helping homeowners (not to mention housing finance reform) never stood much of a chance.