By Bob Hockett
In a characteristically thought-provoking post on the Fed’s monetary easing policies earlier this week, Professor Dorf suggests that in a certain sense, the wisdom with which Chairman Bernanke’s Fed has managed post-crisis credit and liquidity conditions is a fortunate accident – one that we might not, strictly speaking, have earned through legislation. There are two mutually complementary reasons.
First, what many take for the traditional anti-inflationary rationale behind central bank independence (as manifest both in the term structure of Fed Board appointments and in the Fed’s substantive mandate) is not implicated in the current credit environment. And second, there is no inherent, structural reason to expect politicians to be biased toward tight money policies as there is to expect them to be biased toward loose money policies, hence no reason to suppose that the Federal Reserve Act’s framers meant to insulate the Fed from politics – that is, to render it more or less independent – for any reason other than inflation-prevention. As Professor Dorf points out, the two reasons jointly suggest that the Fed’s now filling-in for a derelict Congress by using expansionary monetary policy as a substitute for missing fiscal policy is more happy accident than it is anything that we as a polity have acted through White House and Congress to assure.
It seems to me that Professor Dorf is quite right, but it might be helpful nevertheless to add a few supplementary observations concerning the evolution of central bank theory, practice, and legislation in recent decades. For if what I’m about to report is correct, then there is at least one important sense in which current Fed policies, even in the very forms they’re now taking, can be said in effect to have been legislated – or at the very least legislatively ratified.
Here’s what I mean, proceeding in order from theory, to practice, to legislation.First as a matter of theory, the traditional ‘sound money,’ anti-inflationary rationale for central bank independence can be, and has indeed long been in the process of being, generalized and refined along two dimensions. The generalization and refinement begin by noting that inflation has a contrary – deflation – which is just as apt to occur over time as inflation. Indeed there are structural reasons, to which I’ll return, for expecting so, and these are even deeper seated than are the structural reasons of which we typically think when we think of the anti-inflation mandate alone. The generalization and refinement then conclude by recognizing that inflation and deflation – and, indeed, hyperinflation and hyperdeflation – can and do tend to occur in more than just consumer goods and services markets. They happen in financial markets as well. Indeed a financial asset price bubble just is a species of hyperinflation, and a bubble-following bust is symmetrically an asset price hyperdeflation, as argued here.
Now, asset price inflations and deflations, via the so-called ‘the wealth effect,’ sometimes spill into the real economy to induce counterpart consumer price inflations and deflations – often with devastating effect. Frequently, in fact, the latter bring on what the great American economist Irving Fisher dubbed ‘debt-deflations.’ They do so any time that the antecedent asset price bubble has been significantly fueled – as most if not all of the worst of them are – by cheap credit, in a manner that incents even rational market participants to ‘leg’ the proverbial ‘spread’ between low borrowing costs and high capital gains appreciation rates. When variable prices of the latter sort ultimately drop in the bust, and the fixed debt obligations incurred in financing the speculative asset purchases do not, victims are left with potentially crippling private debt overhang. Growth- and employment-underwriting consumer expenditure in consequence takes a significant hit, and protracted slump is the end product. Such is the substance of Fisher’s regrettably forgotten ‘debt deflation theory of great depressions.’So what does this have to do with central banking theory? The key point, in my view, is that asset price hyperinflations and debt deflations are what I call ‘recursive collective action problems.’ In a classic collective action problem, multiple individually rational actions aggregate into a collectively damaging outcome. Such problems take on recursive properties when the rational actions in question are rational responses to antecedent actions of the same type. Where the recursive element is present, a collectively ‘damaging’ outcome can become downright calamitous. For recursive processes lack tolerable stable equilibria; things either grow worse and worse or they settle into unbearable patterns – as that other celebrated 20th economist, J. M. Keynes, in effect noted when diagnosing what he dubbed ‘liquidity traps’ and ‘paradoxes of thrift.’
Now once attention has been called to it, it is easy to see the structure just outlined at work in any inflation or deflation – including those that afflict financial markets and broader macroeconomies. While prices are rising and credit remains virtually limitless and cheap as it did for much of the late 1990s and early 2000s, it is financially rational for market actors to borrow and buy, then to borrow and buy more as previous iterations of the process drive prices yet higher without correspondingly raising the cost of credit. Subsequently, when confidence in the soundness of assets evaporates and the ‘run’ or the ‘fire sale’ begins, it is symmetrically rational for market actors to rush to avoid being last out the door. There is nothing that any one market participant can do to arrest the upswing or downswing, after all, so her best response is simply to join in the race and work hard not to lose. But the rationality of each in these settings is the futility of all.
Recursive collective action problems of this sort, as I’ve argued elsewhere, are pervasive in financial markets and macroeconomies. What can – what do – we do about them? To solve a collective action problem requires a collective agent – some agent able to act in the name of all. That agent must act as to render no longer rational such decisions as, when aggregated, bring on the calamitous outcome. In the case of an asset price hyperinflation, that is done by closing the spread between low borrowing costs and high capital appreciation rates – that is, by raising borrowing costs, taxing away capital gains at increasing rates, or both. In the case of a corresponding asset price hyperdeflation and ensuing debt deflation, the collective agent symmetrically does the contrary. It acts to lower borrowing costs, stimulate asset price growth, or both. In extreme cases, it must act to pare down private debt too – either directly, by facilitating loan modifications or bankruptcy cramdowns, or indirectly, by pumping up asset prices and thereby trimming the relative reach of the private debt overhang.
Now if we look at the evolution of central bank practice over the past century or so, I think we can spot what amounts to an implicit recognition on the part of central bankers that they are the requisite ‘collective agents’ upon whom I have just called. On the upswing side of the business or credit cycle, central banks have long acted to counter consumer price inflation, of course, but now have begun looking for ways to curb financial asset price hyperinflation as well. The old ‘lean versus clean’ debate that used to divide American central bankers from their counterparts at the Bank for International Settlements in Basle, in other words, appears at last to have been won by the ‘leaners’ – those who, like Fed Chairman Martin nearly five decades ago, view the job of the central bank as that of ‘leaning against the wind.’ You’ll find evidence of the shift every time you read or hear the words ‘macroprudential regulation’ or ‘systemic risk regulation’ in connection with the word ‘financial.’ The name of the game now is to preempt financial asset price hyperinflations before they commence, just as it’s long been to act similarly in respect of gathering consumer price inflations. You can read more about these developments here and here, in two papers stemming from work I have done for the New York Fed and the International Monetary Fund in recent years.
On the downswing sides of the business and credit cycles, central bankers have also long acted to counter consumer – and asset – price deflation. Indeed that’s precisely what ‘cleaning’ refers to when one speaks of ‘leaning or cleaning.’ Walter Bagehot famously called attention to the cleaning role, under the now canonical ‘lender of last resort’ rubric, back in the late 19th century. And even Fed Chairman Greenspan believed in an energetic ‘cleaning’ role, his having pursued it a bit overzealously post 9/11 being thought by many to have played a critical role in fueling our recent housing price bubble. The best way to view what Chariman Bernanke’s Fed has been doing since 2008 – and what its Japanese counterpart has been doing even longer and its European counterpart is beginning to do now – is, accordingly, as a simple and straightforward intensification of what central banks long have done after busts. From easing credit and monetary conditions through commercial paper or treasury security purchases as was done in the 19th and 20th centuries, respectively, central banks are now moving toward easing conditions yet further – and propping up particular asset class prices in particular wealth-effect-salient markets like housing – by purchasing additional classes of financial asset like mortgage backed securities. I for my part have proposed yet more incremental, fine-tuning extensions of the Fed-as-market-actor role in the name of credit-modulation.
So have we actually legislated this still evolving collective agential role for our central bank? Is the Fed statutorily authorized to act as a collective agent of the kind I’ve described? In two mutually complementary senses, I think that it is. To see why, note first this entailment of the central bank’s role as I have just sketched it: the ‘collective’ agent that I am describing is a countercyclical agent. The central bank’s role is to buck trends and act contrary to deeply rooted, otherwise inescapable business and credit cycles. It is to be cyclically contrarian. It’s to act bearishly during bullish times, and to act bullishly during bearish times. It’s meant to steer us clear of, or bump us out of, intolerable equilibria – underspending, underemployment equilibria during busts and ensuing debt deflations, and overspending, overborrowing equilibria during inflations.
But now note this further entailment of my characterization of the central bank’s role: the countercyclical agent must by definition be a countermajoritarian agent. That which fuels cyclical movement, after all, is majoritarian action – aggregated individually rational actions. For the good of all, then, the central bank must act contrary to all, precisely because its bailiwick – the financial, monetary, and macro- economies – is structurally susceptible to recursive collective action problems per which each person’s acting rationally is all people’s acting self-defeatingly. In this sense, the case for a countermajoritarian central bank might be even stronger than that for countermajoritarian courts. After all, crowds seem merely susceptible to ‘rushes to judgment’ on various matters, but seem downright doomed by the structure of reciprocal action itself to financial hence macroeconomic boom and bust cycling. So have we legislated this – the central bank’s institutional contrarian role? Have we deliberately conferred a countermajoritarian function upon the Fed in the way that we have on the federal courts – by insulating its key personnel from political pressure and by formulating its mandate in sufficiently capacious terms? I think that we have – both implicitly and, to a degree, even explicitly.
Note first that the Fed’s enabling legislation charges it with acting to maintain both ‘stable prices’ and ‘maximum employment.’ What do those phrases mean? Well, long before these two sides of the so-called ‘dual mandate’ had both found their way into the Fed’s enabling act, economists, policy analysts, and members of Congress alike had come to understand that stable prices are maintained by tightening credit and monetary conditions in the face of inflationary pressures, and loosening the same in the face of deflationary pressures. They also understood that ‘maximum employment,’ at least as a central bank mandate, is pursued through the same aforementioned loosening of credit and monetary conditions, and that ‘going too far’ in this direction could under some conditions imperil the successful execution of the price-stability-maintenance function.
The Fed has in consequence viewed its role for decades as maintaining a balance between maximum sustainable employment on the one hand, price stability on the other hand. It has acted, in other words, to arrest declining employment so as to prevent its morphing into indefinite underemployment equilibrium on the one hand, and to arrest increasing prices so as to prevent their morphing into consumer price hyperinflations on the other hand. This was the case well before the current wording of the mandate was formulated in the late 1970s, and the latter can accordingly be viewed, along with the enhancing of Fed independence during the 1950s and its continuation ever since, as implicit Congressional ratification of the Fed’s gradually developing countercyclical, hence countermajoritarian, role.
But wait, you might say. Given that ‘stable prices’ used to be taken, in effect, to allude only to consumer price inflation, can we plausibly take that phrase to embrace debt deflations and financial asset price bubbles and busts too? I don’t see why not. Since only the word ‘prices,’ not the word ‘inflation’ or the phrase ‘consumer prices’ is used, and since it has always been understood that there is at least some relation between the financial and ‘real’ economies such as enables dysfunction in the former to spill into the latter, it would seem that the Fed’s – along with other central banks’ – gradually coming to appreciate that financial markets are as prone to destructive yet modulatable cycling as are consumer goods and services markets was foreseeable enough by those who introduced the ‘stable price’ and ‘maximum employment’ mandates, and presumably embraced accordingly by Congress. Why not, then, read the Federal Reserve Act as both (a) ratifying developments that had occurred in central bank practice as informed by developing theory by 1977 when the Act was amended, and (b) embracing developments that occur along that same trajectory since then?
If this is sound thinking, then we are in a certain sense indeed currently enjoying the monetary policy that we deserve. We are not, however, enjoying the fiscal policy that we deserve. We won’t have that till we get the Congress that we deserve – an eventuality which, thankfully, the Senate took measures to obtain for us this week with the ‘nuclear option.’