Dawn Patrol: Above the Fog
Scott Sumner Does the Lord's Work

What Have We Unlearned from Our Great Recession?

Jan 07, 2011 10:15 am, Sheraton, Governor's Square 15 American Economic Association: What's Wrong (and Right) with Economics? Implications of the Financial Crisis (A1) (Panel Discussion): Panel Moderator: JOHN QUIGGIN (University of Queensland, Australia)

  • BRAD DELONG (University of California-Berkeley) Lessons for Keynesians
  • TYLER COWEN (George Mason University) Lessons for Libertarians
  • SCOTT SUMNER (Bentley University) A defense of the Efficient Markets Hypothesis
  • JAMES K. GALBRAITH (University of Texas-Austin) Mainstream economics after the crisis:

My role here is the role of the person who starts the Alcoholics Anonymous meetings.

My name is Brad DeLong.

I am a Rubinite, a Greenspanist, a neoliberal, a neoclassical economist.

I stand here repentant.

I take my task to be a serious person and to set out all the things I believed in three or four years ago that now appear to be wrong. I find this distressing, for I had thought that I had known what my personal analytical nadir was and I thought that it was long ago behind me

I had thought my personal analytical nadir had come in the Treasury, when I wrote a few memos about how Rudi Dornbusch was wrong in thinking that the Mexican peso was overvalued. The coming of NAFTA would give Mexico guaranteed tariff free access to the largest consumer market in the world. That would produce a capital inflow boom in Mexico. And so, I argued, the peso was likely to appreciate rather than the depreciate in the aftermath of NAFTA.

What I missed back in 1994 was, of course, that while there were many US corporations that wanted to use Mexico's access to the US market and so locate the unskilled labor parts of their value chains south, there were rather more rich people in Mexico who wanted to move their assets north. NAFTA not only gave Mexico guaranteed tariff free access to the largest consumer market in the world, it also gave US financial institutions guaranteed access to the savings of Mexicans. And it was this tidal wave of anticipatory capital flight--by people who feared the ballots might be honestly counted the next time Cuohtemac Cardenas ran for President--that overwhelmed the move south of capital seeking to build factories and pushed down the peso in the crisis of 1994-95.

I had thought that was my worst analytical moment.

I think the past three years have been even worse.

So here are five things that I thought I knew three or four years ago that turned out not to be true:

  1. I thought that the highly leveraged banks had control over their risks. With people like Stanley Fischer and Robert Rubin in the office of the president of Citigroup, with all of the industry's experience at quantitative analysis, with all the knowledge of economic history that the large investment and commercial banks of the United States had, that their bosses understood the importance of walking the trading floor, of understanding what their underlings were doing, of managing risk institution by institution. I thought that they were pretty good at doing that.

  2. I thought that the Federal Reserve had the power and the will to stabilize the growth path of nominal GDP.

  3. I thought, as a result, automatic stabilizers aside, fiscal policy no longer had a legitimate countercyclical role to play. The Federal Reserve and other Central Banks were mighty and powerful. They could act within Congress's decision loop. There was no no reason to confuse things by talking about discretionary fiscal policy--it just make Congress members confused about how to balance the short run off against the long run.

  4. I thought that no advanced country government with as frayed a safety net as America would tolerate 10% unemployment. In Germany and France with their lavish safety nets it was possible to run an economy for 10 years with 10% unemployment without political crisis. But I did not think that was possible in the United States.

  5. And I thought that economists had an effective consensus on macroeconomic policy. I thought everybody agreed that the important role of the government was to intervene strategically in asset markets to stabilize the growth path of nominal GDP. I thought that all of the disputes within economics were over what was the best way to accomplish this goal. I did not think that there were any economists who would look at a 10% shortfall of nominal GDP relative to its trend growth path and say that the government is being too stimulative.

With respect to the first of these--that the large highly leveraged banks had control over their risks: Indeed, American commercial banks had hit the wall in the early 1980s when the Volcker disinflation interacted with the petrodollar recycling that they had all been urged by the Treasury to undertake. American savings and loans had hit the wall when the Keating Five senators gave them the opportunity to gamble for resurrection while they were underwater. But in both of these the fact that the government was providing a backstop was key to their hitting the wall.

Otherwise, it seemed the large American high commercial and investment banks had taken every shock the economy could throw at them and had come through successfully. Oh, every once in a while an investment bank would flame out and vanish. Drexel would flame out and vanish. Goldman almost flamed out and vanished in 1970 with the Penn Central. We lost Long Term Capital Management. Generally we lost one investment bank every decade or generation. But that's not a systemic threat. That's an exciting five days reading the Financial Times. That's some overpaid financiers getting their comeuppance, which causes schadenfreude for the rest of us. That's not something of decisive macro significance.

The large banks came through the crash of 1987. They came through Saddam Hussein's invasion of Kuwait. Everyone else came through the LTCM crisis. Everyone came through the Russian state bankruptcy when the IMF announced that nuclear-armed ex-superpowers are not too big to fail. They came through assorted emerging market crisis. They came through the collapse of the Dot Com Bubble.

It seemed that they understood risk management thing and that they had risk management thing right. In the mid 2000s when the Federal Reserve ran stress tests on the banks the stress was a sharp decline in the dollar if something like China's dumping its dollar assets started to happen. Were the banks robust to a sharp sudden decline in the dollar, or had they been selling unhedged puts on the dollar? The answer appeared to be that they were robust. Back in 2005 policymakers could look forward with some confidence at the ability of the banks to deal with large shocks like a large sudden fall on the dollar.

Subprime mortgages? Well, those couldn’t possibly be big enough to matter. Everyone understood that the right business for a leveraged bank in subprime was the originate-and-distribute business. By God were they originating. But they were also distributing.

I thought about theses issues in combination with the large and persistent equity premium that has existed in the US stock market over the past century. You cannot blame this premium on some Mad Max scenario in which the US economy collapses because the equity premium is a premium return of stocks over US Treasury bonds, and if the US economy collapses then Treasury bonds' real values collapse as well--the only things that hold their value are are bottled water, sewing machines and ammunition, and even gold is only something that can get you shot. You have to blame this equity risk premium on a market failure: excessive risk aversion by financial investors and a failure to mobilize the risk-bearing capacity of the economy. This there was a very strong argument that we needed more, not less leverage on a financial system as a whole. Thus every action of financial engineering--that finds people willing to bear residual equity risk and that turns other assets that have previously not been traded into tradable assets largely regarded as safe--helps to mobilize some of the collective risk bearing capacity of the economy, and is a good thing.

Or so I thought.

Now this turned out to be wrong.

The highly leveraged banks did not have control over their risks. Indeed if you read the documents from the SECs case against Citigroup with respect to its 2007 earnings call, it is clear that Citigroup did not even know what their subprime exposure was in spite of substantial effort by management trying to find out. Managers appeared to have genuinely thought that their underlings were following the originate-and-distribute models to figure out that their underlings were trying to engage in regulatory arbitrage by holding assets rated Triple A as part of their capital even though they knew fracking well that the assets were not really Triple A.

Back when Lehman Brothers was a partnership, every 30-something in Lehman Brothers was a risk manager. They all knew that their chance of becoming really rich depended on Lehman Brothers not blowing as they rose their way through the ranks of the partnership.

By the time everything is a corporation and the high-fliers' bonuses are based on the mark-to-model performance of their positions over the past 12 months, you've lost that every-trader-a-risk manager culture. i thought the big banks knew this and had compensated for it.

I was wrong,

With respect to the second of these--that the Federal Reserve had the power and the will to stabilize nominal GDP: Three years ago I thought it could and would. I thought that he was not called "Helicopter Be"n for no reason. I thought he would stabilize nominal GDP. I thought that the cost to Federal Reserve political standing and self-perception would make the Federal Reserve stabilize nominal GDP. I thought that if nominal GDP began to undershoot its trend by any substantial amount, that then the Federal Reserve would do everything thinkable and some things that had not previously been thought of to get nominal GDP back on to its trend growth track.

This has also turned out not to be true.

That nominal GDP is 10% below its pre-2008 trend is not of extraordinarily great concern to those who speak in the FOMC meetings. And staffing-up the Federal Reserve has not been an extraordinarily great concern on the part of the White House: lots of empty seats on the Board of Governors for a long time.

With respect to the third of these--that discretionary fiscal policy had no legitimate role: Three years ago I thought that the Federal Reserve could do the job, and that discretionary countercyclical fiscal policy simply confused congress members, Remember Orwell's Animal Farm? Every animal on the Animal Farm understands the basic principle of animalism: "four legs good, two legs bad" (with a footnote that, as Squealer the pig says, a wing is an organ of locomotion rather than manipulation and is properly thought of as leg rather than an arm--certainly not a hand).

"Four legs good, two legs bad," was simple enough for all the animals to understand. "Short-term countercyclical budget deficit in recession good, long-run budget deficit that crowds out investment bad," was too complicated for Congressmen and Congresswomen to understand. Given that, discretionary fiscal policy should be shunted off to the side as confusing. The Federal Reserve should do the countercycical stabiization job.

This also turned out not to be true, or not to be as true as we would like. When the Federal funds rate hits the zero lower bound making monetary policy effective becomes complicated. You can do it, or we think you can do it if you are bold enough, but it is no longer straightforward buying Treasury Bonds for cash. That is just a swap of one zero yield nominal Treasury liability for another. You have got to be doing something else to the economy at the same time to make monetary policy expansion effective at the zero nominal bound,

One thing you can do is boost government purchases. Government purchases are a form of spending that does not have to be backed up by money balances and so raise velocity. And additional government debt issue does have a role to play in keeping open market operations from offsetting themselves whenever money and debt are such close substitutes that people holding Treasury bonds as saving vehicles are just as happy to hold cash as savings vehicles. When standard open market operations have no effect on anything, standard open market operations plus Treasury bond issue will still move the economy.

With respect to the fourth of these--that no American government would tolerate 10% unemployment: I thought that American governments understood that high unemployment was social waste: that it was not in fact an efficient way of reallocating labor across sectors and response to structural change. When unemployment is high and demand is low, the problem of reallocation is complicated by the fact that no one is certain what demand is going to be when you return to full employment. Thus it is very hard to figure which industries you want to be moving resources into: you cannot look at profits but rather you have to look at what profits will be when the economy is back at full employment--and that is hard to do.

For example, it may well be the case that right now America is actually short of housing. There is a good chance that the only reason there is excess supply of housing right now is because people's incomes and access to credit are so low that lots of families are doubling up in their five-bedroom suburban houses. Construction has been depressed below the trend of family formation for so long that it is hard to see how there could be any fundamental investment overhang any more.

It is always much better to have the reallocation process proceed by having rising industries pulling workers into employment because demand is high. It is bad to have the reallocation process proceed by having mass unemployment in the belief that the unemployed will sooner or later figure out something productive to do. I thought that American governments understood that.

I thought that American governments understood that high unemployment was very hazardous to incumbents. I thought that even the most cynical and self-interested Congressmen and Congresswomen and Presidents would strain every nerve to make sure that the period of high unemployment would be very short.

It turned out that that wasn’t true.

I really don’t know why. I have five theories:

  1. Perhaps the collapse of the union movement means that politicians nowadays tend not to see anybody who speaks for the people in the bottom half of the American income distribution.
  2. Perhaps Washington is simply too disconnected: my brother-in-law observes that the only place in America where it is hard to get a table at dinner time in a good restaurant right now is within two miles of Capitol Hill.
  3. Perhaps we are hobbled by general public scorn at the rescue of the bankers--our failure to communicate that, as Don Kohn said, it's better to let a couple thousand feckless financiers off scot-free than to destroy the jobs of millions, our failure to make that convincing.
  4. I think about lack of trust in a split economics profession--where there are, I think, an extraordinarily large number of people engaging in open-mouth operations who have simply not done their homework. And at this point I think it important to call out Robert Lucas, Richard Posner, and Eugene Fama, and ask them in the future to please do at least some of their homework before they talk onsense.
  5. I think about ressentment of a sort epitomized by Barack Obama's statements that the private sector has to tighten its belt and so it is only fair that the public sector should too. I had expected a president advised by Larry Summers and Christina Romer to say that when private sector spending sits down then public sector spending needs to stand up--that is is when the private sector stands up and begins spending again that the government sector should cut back its own spending and should sit down.

I have no idea which is true.

I do know that when I wander around Capitol Hill and the Central Security Zone in Washington, the general view I hear is: "we did a good job: we kept unemployment from reaching 15%--which Mark Zandi and Alan Blinder say it might well have reached if we had done nothing." That declaration of semi-victory puzzles me.

Three years ago, I thought that whatever theories economists worked on they all agreed the most important thing to stabilize was nominal GDP. Stabilizing the money stock was a good thing to do only because money was a good advance indicator of nominal GDP. Worrying about the savings-investment balance was a good thing to worry about because if you got it right you stabilized nominal GDP. Job 1 was keeping nominal GDP on a stable growth path, so that price rigidity and other macroeconomic failures did not cause high unemployment. That, I thought, was something all economists agreed on. Yet I find today, instead, the economics profession is badly split on whether the 10% percent shortfall of nominal GDP from its pre-2008 trend is even a major problem.

So what are the takeaway lessons? I don’t know.

Last night I was sitting at my hotel room desk trying to come up with the "lessons" slide.

The best I could come up with is to suggest that perhaps our problem is that we have been teaching people macroeconomics.

Perhaps macroeconomics should be banned.

Perhaps it should only be taught through economic history and the history of economic thought courses--courses that start in 1800 back when all issues of what the business cycle was or what it might become were open, and that then trace the developing debates: Say versus Mathis, Say versus Mill, Bagehot versus Fisher, Fisher versus Wicksell, Hayek versus Keynes versus Friedman, and so forth on up to James Tobin. I really don't know who we should teach after James Tobin: I haven't been impressed with any analyses of our current situation that have not been firmly rooted in Tobin, Minsky, and those even further in the past.

Then economists would at least be aware of the range of options, and of what smart people have said and thought it the past. It would keep us from having Nobel Prize-caliber economists blathering that the NIPA identity guarantees that expansionary fiscal policy must immediately and obviously and always crowd-out private spending dollar-for-dollar because the government has to obtain the cash it spends from somebody else. Think about that a moment: there is nothing special about the government. If the argument is true for the government, it is true for all groups--no decision to increase spending by anyone can ever have any effect on nominal GDP because whoever spends has to get the cash from somewhere, and that applies to Apple Computer just as much as to the government.

And that has to be wrong.

So let me stop there and turn it over to Scott Sumner.

Well Worth Reading