Grasping Reality on TypePad, by Brad DeLong

1. The size of the shock arising from what appeared ex ante to be relatively small and manageable balance-sheet vulnerabilities.

2. The failure of the tools that the Federal Reserve and other North Atlantic central banks had thought would allow them to build firewalls between financial disturbances and the real economy of production, spending, and employment—and the consequent magnitude of the economic crash of 2008-2009.

3. The slow and incomplete nature of the subsequent recovery. Of employment, production, productivity, price levels, and unemployment rates, only the last is in a range that reasonable forecasts as of 2007 would have anticipated.

I know that I, at least, have been surprised at every step of this chain. In the mid-2000s I was looking for financial tail risks to come from Wall Street banks having sold unhedged foreign-exchange puts via derivatives and thus transformed dollar-denominated into euro-, yen-, pound-, and renminbi-denominated debts. In the context of a possible substantial downward realignment of the value of the dollar, that seemed to me to be a plausible source of trouble.

Truly mammoth and unmanageable financial shocks were, I believed, almost surely confined to countries with balance sheets suffering severe currency mismatch (Obstfeld (2012)). Subprime was not a plausible source. Yet Eichengreen and O’Rourke (2009, 2010, 2012), tracking the downturn in real time, made a convincing and so far unrefuted case that the magnitude of the initial shocks coming from the financial sector and impacting the real economy of spending, production, and employment were at least as large as those of the Great Depression.

In 2008-9 I was confident in the strength of the monetary policy tools that had enabled the Federal Reserve to substantially damp any real-side economic impact arising out of the financial disturbances of 1987’s Black Monday on the stock market, 1990’s S&L crisis, the 1994 collapse of the Mexican peso, the 1997 East Asian crisis, the 1998 trifecta of Korea and Russia and LTCM, the 2000 collapse of the dot-com bubble, and 9/11.

And back in 2010 I was expecting a bounce-back along the lines of the post-1982 recovery carrying the economy near to if not all the way back to its pre-criss trend. Large expansions in government purchases made irresistible by secularly low borrowing costs, quantitative easing on an unprecedented scale, helicopter money, credible forward guidance that returning nominal GDP to its pre-crisis path—I did not know which of these four would work, but I was confident that they all would be planned out, that those that seemed most promising would be attempted, and that at least one would succeed.

How far down into the lower tail of outcomes, both in terms of our Bayesian beliefs back in 2007 of the risks we faced and in terms of the true distribution of risks of which we were appalling ignorant, have we been carried?

It is largely a matter of taste.

Perhaps we would have assessed ourselves as facing large enough uncertainties that this would place us in the 5-10% lower tail—think that the past decade has been the kind of thing we expect to see once every century or two.

Perhaps we would have assessed ourselves as having greater knowledge and facing lesser uncertainties, such that finding ourselves in a place like this as being in the 1%-2% lower tail—think that this is the kind of thing that we really did, especially after the Great Depression, have the understanding and tools to head off, and that such an adverse combination of shocks, propagation mechanisms, and policy responses should only happen every half-millennium.

Thus the answer to the question of “did macroeconomic policy play a different role in the (post-2009) recovery than in other post-World War II recoveries?” has to be “yes”. Of course the role of policy is different. It could not be otherwise. What we have seen over the past decade has been and in many ways is still a very different phenomenon than the other post-World War II business cycles.

But let me take the organizers’ orienting question to be somewhat more subtle. Back before 2007 or so we had a sense of how macroeconomic policy—banking and regulatory, fiscal, and above all monetary—reacted to the course of the business cycle in attempting to head off trouble, in attempting to deal with shocks and crises, and in attempting to clean up the mess after the nadir. We followed some kind of implicit policy rule. From this perspective, the major orienting questions seem to me to be:

1. Have we, since 2007 or so, kept following the implicit policy rule that we then understood ourselves to be following?
2. Have the policies we have implemented had the effects and consequences that we used to expect that they would have?
3. Was that pre-2007 policy rule the right policy rule to be following?
4. What now do we wish that our pre-2007 policy rule had been?
5. Why haven’t we made larger mid-course corrections as we have been surprised by the outcomes that the world has dealt us?

And I will focus on the United States alone. The United States alone is already much too big a topic. I will spend a very little time on banking and regulatory policy as an arm of macroeconomic policy. I will spend a little more time on fiscal policy. And I will spend most of my space trying to think about monetary policy.

My principal guesses—for I am not at all sure that I have gotten this right—are:

1. Back before 2007 our policy rule as far as monetary policy was concerned consisted of setting interest rates in a countercyclical fashion according to a near-consensus Taylor-type feedback rule. Deviations from this were allowed to achieve other goals, like assessing growth headroom created by technological innovations, and supporting constructive fiscal policy actions.

2. There was, in addition, a willingness at the zero lower bound to go the extra mile with respect to open market operations and quantitative easing along the lines recommended by Friedman (1997).

3. But there was not a willingness to engage in “helicopter money” along the lines of Bernanke (1999, 2002).

4. And there was not a willingness to make credible promises to engage in stimulus that would be seen as “irresponsible” after the recovery was complete, as recommended by Krugman (1998).

5. Back before 2007 our policy rule as far as fiscal policy was concerned was that fiscal policy was to be set on “classical” terms. The focus was to be on rightsizing the state, levying taxes efficiently, and achieving long run fiscal balance.

6. Countercyclical fiscal policy was to be restricted to automatic stabilizers. Why? Because fiscal authorities could not operate discretionary countercyclical policies in a timely fashion, and were in any event not competent to choose appropriate policies. This reservation (fiscal automatic stabilizers aside) of fiscal policy to central banks was to hold even at the zero lower bound, for monetary policy was more than powerful enough to do the job.

7. Back before 2007, our policy rule as far as prudential banking/regulatory policy was one of benign neglect. The continued existence of a substantial equity return premium implied that there was a substantial market failure involved in the difficulty financial markets had in mobilizing anything like society’s total potential risk bearing capacity. Largely unregulated financialization and experimentation might well produce substantial benefits in making the mobilization of such risk bearing capacity easier. And there were few risks generated by such benign neglect of systemic financial risk because of the power of the Federal Reserve’s monetary policy tools.

8. Back before 2007, our policy rule as far as banking/regulatory crisis management was: In the event of a financial crisis, the Federal Reserve, the other banking regulatory agencies, and if necessary the Treasury—after seeking Congressional approval—should follow Walter Bagehot (1873), as interpreted and recommended by Kindleberger (1979). The Bagehot Rule is: In a financial crisis, (1) lend freely, (2) to institutions that are solvent but illiquid, but (3) lend at a penalty rate. No institution that receives financial support should, after the crisis is over, feel that it came out of the situation well. No institution should feel that the positions they adopted that created systemic risk still look like they were worth running. No executive should get rich as a result of the government’s lender-of-last-resort activities. And if an institution is (a) insolvent but (b) systemically important? Resolve it immediately, zeroing out the equity, options, and mezzanine holdings of stakeholders.

9. The effects of the policies that have been followed have been in the ballpark of what expectations had been, but have nevertheless been disappointing.

10. The economy has been undershooting Federal Reserve—and market—expectations for a decade, but has not been undershooting expectations by a gross amount in any one year. The policy levers are connected. The linkages are just on the weak side, and the slippage is on the high side of what was thought likely.

11. In retrospect, that pre-2007 policy rule was not the right policy rule.

12. There is not even a near-consensus on what the right policy rule would have been.

13. In large part our failure to make larger mid-course corrections toward the “right” policy rule is the result of our disagreements over what would have been the right policy rule. Thus the next move needs to be with the head.

14. If the Federal Reserve is going to continue to retain plenary countercyclical stabilization policy power, it needs—either via central banking custom or via explicit legislative powers, or both—more and better tools.

15. If fiscal authorities are going to share some responsibility for countercyclical stabilization policy, they need to step up their technocratic game.

16. In any event, better and stronger fiscal automatic stabilizers would potentially be a great help.