[LE: Regular commenter Lorenzo has written a post on what the role of central banks ought to be in setting
financial monetary policy goals. He starts with some historical examples, and continues on to contemporary central banks. Part of the issue (as I read his post) is that people must get some idea of where monetary policy is going from the Central Bank so that they can plan their affairs accordingly.
Lorenzo's piece starts over the fold.]
How stupid can a Central Bank be?
The short answer from history is: a central bank can be really, really stupid.
I am using ‘stupid’ in a technical sense: doing things that seriously adversely affect lots of people with no justifying benefits to any wider public good—that is, which show a lack of intelligence, understanding, reason, wit or sense. The actions may seem a good idea to the central bank at the time—due to perverse incentives, policy framings disconnected from economic reality or whatever—but in terms of wider public policy, they are (to varying degrees) disastrous. Central banks exist to serve, so how that “serving” is framed can make a great difference.
For example, hyperinflation is usually a deliberate attempt to inflate away government debt and/or generate revenue well beyond the willingness or ability to tax. It may be wicked, but it is not stupid in quite the above sense. (There are justifying benefits for decision-makers, without necessarily justified benefits.)
Beware of the French and central banks
Among stupid central banks, the all-time winner is the interwar Bank of France turning the gold standard into a doomsday device (pdf), helped by the US Federal Reserve, by building up its gold reserves without issuing money to match, so taking gold out of the monetary system, thus driving up the price of gold in the monetary system (and so the price of money, as such gold set the price of money) and thus driving down the prices of everything else. It and the Fed created the Great Deflation of 1929-32 we call ‘the Great Depression’ and so mass unemployment, the impoverishing of millions, the unravelling of much of (pdf) the world trade system, the fall of Weimar Germany and the rise of Nazism (followed by the Fall of France). It was a disaster of monumental proportions.
It was hardly the only disaster of central banking, however. Another (in)glorious episode also came from France with John Law’s Banque Générale gaining the right to issue paper money, which stimulated economic activity. The Regent, the duc d’Orleans, decided that if some paper money was good then even more paper money must be even better, leading to the truly spectacular Mississippi Bubble. This French disaster was based on the same logic (using that term loosely) as that which created the Great Deflation/Depression namely, “if some is better (some paper notes, some level of gold backing of the franc) then more is better and even more is better still.” One is reminded of the Abbe Sieyes dismissing the argument for bicameralism on the grounds that if the upper house agreed with the lower it was pointless and if it disagreed it was pernicious. Pernicious simplification passing itself off as sophistication: how very French. (Perhaps the baleful influence of Cartesian rationalism?)
By contrast, the Bank of England has a long history of considerable policy success, starting with vast improvement in management of government debt. The South Sea Bubble was rather less of a problem than the Mississippi bubble precisely because the Bank of England had disapproved from the beginning. While the Bank’s management of the gold standard over the two centuries up to 1914 suffered various bumps and problems, it had nothing to equal the aforementioned French disasters.
In our own time, the Bank of Japan’s management of the yen since the collapse of the bubble economy has come in for much criticism. However, the demographics of Japan make some of that criticism less clear-cut than is often suggested.
Even though some of the ECB’s problems are “built in”, there are also plenty of grounds for criticism for the European Central Bank (ECB), until recently with a French head (perhaps not encouraging; especially as the euro is effectively an artificial gold standard for its member countries).
A contemporary example of successful central banking is the Reserve Bank of Australia. It has run an inflation target since 1993 (pdf). Its website is very clear on its policy target. In the words of the Reserve Bank:
The Governor and the Treasurer have agreed that the appropriate target for monetary policy in Australia is to achieve an inflation rate of 2–3 per cent, on average, over the cycle. This is a rate of inflation sufficiently low that it does not materially distort economic decisions in the community. Seeking to achieve this rate, on average, provides discipline for monetary policy decision-making, and serves as an anchor for private-sector inflation expectations.
The minutes of its Board meetings are published two weeks after each meeting: this matters much less than that it has a clear monetary policy regime.
The Reserve Bank sees its role as providing an anchor for private sector inflation expectations and it does so by being upfront about its policy target. That it has an explicit target since 1993 is no coincidence: the experience of the severe 1992-93 recession where inflation was squeezed out of the Australian economy in a particularly costly way made it clear to policy-makers that being explicit about monetary policy was preferable. As had the problems with monetary policy in the 1980s:
In the early 1990s, the Reserve Bank did not enjoy the largely uncritical press it receives today.
The conduct of monetary policy in the 80s was fundamentally incoherent, unsuccessfully pursuing multiple objectives and shrouded in a veil of secrecy.
Without a policy commitment to price stability, the Australian economy lacked a nominal anchor.
(Does any of this sound familiar, by chance, to American readers?)
The success of the Australian economy since then has provided strong evidence for the good sense of this approach of a clear monetary policy regime via an explicit target. But there is also no mystery about why being explicit has been a successful approach. The point of money is to facilitate transactions by massively decreasing transaction costs. Not only are the search costs that barter imposes avoided by use of money, but there are a range of problems with barter than using money eliminates or greatly ameliorates, thereby greatly facilitating transactions.
If people have reasonably accurate expectations of how (money) prices in general will go, they can make arrangements (including contracts) based on those expectations. As Canadian economist Nick Rowe points out, inflation targeting in Canada came out of pressure from the private sector. They wanted reliable expectations about prices so as to set wage contracts.
Sudden, unexpected changes in prices can leave these arrangements misaligned with actual prices. If, for example, that results in changes in the terms of labour—the ratio of labour costs to the price(s) of what the firm sells—so that wages become seriously over-priced (in normal, somewhat imprecise, economic speak, “real wages have risen”) then firms will stop hiring, workers may be sacked, firms may collapse (i.e. they absolutely stop hiring and all their workers lose their jobs). It is not good to have significant, unexpected downward shifts in price movements, since that essentially guarantees that the terms of labour will rise unexpectedly. (So unexpected disinflation can have similar effects to deflation.)
Which is what happened at the beginning of the Great Recession in the US. When uberblogger Matt Yglesias calls it a “huge failure of central banking” he is absolutely correct. To put it another way, serious expectation failures were imposed on the US economy, resulting in a dramatic drop in transactions. (That the Federal Reserve decided to surreptitiously disinflate as a financial crisis—the sub-prime crash—was building made things much worse: including the financial crisis, providing some reprise [pdf] of the Great Depression.)
How did this happen? Have a look at the US Federal Reserve website. There is no statement about what the specific aim of US monetary policy is. The US Federal Reserve provides no explicit anchor for expectations in the economy. So, the US Federal Reserve can decide to disinflate—to significantly reduce the inflation rate—and there was no warning for private agents that this was happening. To act in this way is to actively degrade the level of information in the economy and so misdirect expectations.
This is deeply stupid in both theory and practice. There is no economic gain from changing monetary policy surreptitiously, there are only unnecessary costs. Australian policy makers found this out the hard way in 1992-93. They learnt the lesson and have moved on. But, alas, almost no one takes what Australia does seriously: we are too small, too far away, too “lucky”, too “colonial”. Europeans and Americans tend to be deeply parochial people, seeing themselves as the measure of all things, and so are rather bad at learning from the policy experience of others.
Australian policy-makers, facing what had been a long-term downward trend in our terms-of-trade (the ratio of the prices of what we sell compared to the prices of what we buy) and conscious of the tyranny of distance, have undertaken a series of reforms over three decades which have, overwhelmingly, paid off. Not only are we the country where the Great Moderation has not ended; in fiscal balance and public debt we are in a very different world from the US, Japan, the Eurozone and the UK.
In the US, by contrast, people try to “read the signals” of the Fed in ways reminiscent of how Kremlinologists used to try and glean Soviet policy from indirect signals. US Federal Reserve Deputy Chair Janet Yellen’s statement in a recent speech:
I believe that the Federal Reserve qualifies as one of the most transparent central banks in the world.
is a stunning piece of parochial blindness. (Nick Rowe is even more pointed about it.)
In the same speech, Deputy Chair Yellen tells us:
Inflation picked up significantly over the first half of this year, with the price index for personal consumption expenditures (PCE) rising at an annual rate of about 3-1/2 percent–a pace that is well above the level of 2 percent or a little less that most Federal Open Market Committee (FOMC) participants consider consistent with the Federal Reserve’s dual mandate for price stability and maximum employment. In contrast, PCE inflation averaged less than 1-1/2 percent over the preceding two years.
The members of the FOMC have views: the Federal Reserve has no explicit target. So, there is no explicit anchoring of nominal (i.e. in money terms) expectations in the US economy.
It is not as if people in the Federal Reserve are not aware of the problem. The current Federal Reserve Chair Ben Bernanke co-authored an article 12 years ago that proclaimed:
The Fed needs an approach that consolidates the gains of the Greenspan years and ensures that those successful policies will continue—even if future Fed chairmen are less skillful or less committed to price stability than Mr. Greenspan has been.
We think the best bet lies in a framework known as inflation targeting, which has been employed with great success in recent years by most of the world’s biggest economies, except for Japan. Inflation targeting is a monetary-policy framework that commits the central bank to a forward-looking pursuit of low inflation—the source of the Fed’s current great performance—but also promotes a more open and accountable policy-making process. More transparency and accountability would help keep the Fed on track, and a more open Fed would be good for financial markets and more consistent with our democratic political system.
As our research on the use of this approach around the world documents, successful inflation targeting requires that the central bank and elected officials make a public commitment to an explicit numerical target level for inflation (usually around 2%), to be achieved over a specified horizon (usually two years). Equally important, the central bank must agree to provide the markets and the public with enough information to evaluate its performance, and to understand its reasoning when policy and inflation deviate from the long-run goal–as they inevitably will at times.
Too bad he has not followed his own (excellent) advice since becoming Chair of the Fed. (If that is because he does not have agreement within the Federal Reserve that just underscores the lack of an explicit target, of any clear monetary regime, and so the lack of any clear anchoring of expectations.)
In the absence of an explicit, open monetary policy regime we instead have a ludicrous focus on instruments—such as ‘Quantitative Easing’ (QE) or “Operation Twist”. In the absence of an explicit monetary policy regime, these actions are self-defeating as they do nothing to change expectations because they provide no reliable, continuing framing for such expectations. Lars Christensen puts it well:
The real problem with QE is not that the money base is increase[d], but that is done in a completely random fashion without any clear framework.
Quite so. As Christensen points out, when the monetary policy regime is clear, people often pay little or no attention to the instruments by which it is brought about since the implications thereof are clearly framed. In the absence of such policy regime clarity, waving those instruments around is much ado signalling nothing; nothing that anchors (or, in this case, shifts) expectations. And expectations matter. (Including expectations of inflation and spending.)
Expectations and actions
We have no knowledge of the future, we merely have expectations about the future based on inferences from the past. Including our experience of the patterns and structures of reality (both physical and social). A clear monetary regime provides such pattern and structure.
Economist Frank Knight famously distinguished between risk and uncertainty:
Uncertainty must be taken in a sense radically distinct from the familiar notion of Risk, from which it has never been properly separated. … The essential fact is that ‘risk’ means in some cases a quantity susceptible of measurement, while at other times it is something distinctly not of this character; and there are far-reaching and crucial differences in the bearings of the phenomena depending on which of the two is really present and operating…. It will appear that a measurable uncertainty, or ‘risk’ proper, as we shall use the term, is so far different from an unmeasurable one that it is not in effect an uncertainty at all.
To put it more simply, uncertainty is risk that is immeasurable, not possible to calculate. But (calculable) risk and uncertainty are about expectations: about looking forward. All action, being future-directed, is about expectations, is framed by expectations. Indeed, requires expectations.
If we cannot frame sufficient expectations, we have no basis on which to act. If the level of uncertainty is too great over some range of action to even frame sufficient expectations on which to base actions, we will direct our actions elsewhere. Uncertainty is fundamentally based on information, because our ability to calculate risk is a function of the information available. (So, expected risks are a function of information.)
The centrality of expectations to action, btw, is why law and economics analysis by people such as Richard Posner has been so fruitful; law is all about creating pattern and structure to frame expectations. (Posner blogs here.) This work uses the two key insights of C20th economics—transaction costs matter and information and action are not separable.
Just because something cannot be calculated does not mean we will not frame expectations to cover that uncertainty: it just means that such expectations cover more than is directly inferable from such information as we have. Economic “confidence”—including business confidence—is, to a large degree, how what cannot be calculated is being framed in a given time period: whether it is being framed positively or negatively and how strongly so. This confidence will be in part based on various indicators but, by uncertainty’s non-calculable nature, cannot be definitively or wholly so. The wider the range of uncertainty (provided it is not too wide to frame any expectations to base action on at all), the more unstable confidence is likely to be, because the greater the possibility of new information changing how uncertainty is being framed. (John Maynard Keynes’ notion of animal spirits represents a rather clumsy formulation of this—since he was a noted theorist of probability, I put down said clumsiness to too much Bloomsbury.)
If confidence is how we frame uncertainty—and so is the more unstable the wider the range of uncertainty—then it is a central obligation of a central bank to narrow such uncertainty, to give a clear monetary policy framework for private agents to frame their actions. The point of money, after all, is to facilitate transactions. The point of monetary policy can hardly be to either lessen money’s ability to facilitate transactions or to lower the trend level of transactions. On the contrary, the “dual mandate” of stable prices and low unemployment is all about facilitating transactions; which is much more likely to be achieved by a an explicit monetary regime, especially in adverse economic circumstances.
What Greenspan got wrong
This is where criticism of Greenspan’s tenure as Federal Reserve Chair should be directed. Criticising him for various asset bubbles is misguided. Stable macroeconomic conditions in an expanding economy will tend to lead to expansion in credit and other forms of capital (and so rises, even surges, in asset prices): both because rising incomes lead to larger sums being available to be lent or spent on assets and because assessments of risk will tend to fall, making credit cheaper and pushing up asset values (due to increased confidence in the future). There are also the effects on asset prices of (pdf) innovation generating uncertainty about growth of income from assets. There are some fascinating issues regarding prudential regulation, the undermining of prudence and the effects of land rationing (including its ability to create land value surges and attract credit to housing), but they are not monetary policy matters.
Besides, the various asset boom and busts, and other economic shocks, were dealt with considerable success in terms of the maintenance of macroeconomic stability during Greenspan’s tenure. (A useful partial list of such shocks and difficulties is provided here.)
Greenspan’s gnomic central banking policy persona was much more problematic. It was basically “trust me” framing of an, at best, implicit policy regime; which turned out to be self-fulfilling as confidence-in-Greenspan (including his ability to “manage” the Fed) framed expectations that prices would be kept relatively stable in ways which would not too adversely affect output and which gained strength the more things continued to turn out that way. In practice, what he was doing was keeping growth in GDP in money terms (nominal GDP or NGDP) stable.
Two points about this: first, what happened when Greenspan left? Bad things, it turned out, because no new basis for confidence was generated to replace him nor any guarantee that his implicit policy regime would continue. On the contrary, the implicit policy regime became something quite different while the myth of the master-lever puller generated the belief that it was all about pulling the “right” lever, of what you did with the instruments, rather than framing expectations (even if only implicitly). This was profoundly misguided: confidence in Greenspan, in what Greenspan was about with the levers (his implicit policy regime), was at least as important as which levers he pulled (provided it was congruent with that confidence and implicit policy regime) because that confidence and implicit expectations framed the lever pulling: said framing and actions then led to stable growth in NGDP. (To use Nick Rowe’s Chuck Norris theory of central banking, Greenspan was the ultimate Chuck Norris of central banking—the central banker was the message.)
It can be fixed
Second, the Reserve Bank of Australia (RBA) with its explicit target, did better. Yes, it ran a slightly higher inflation premium, but that may well be justified given the differences between the US and Oz economies. But it also oversaw less fluctuation in output. In terms of facilitating transactions, competent bureaucrats (or econocrats in Oz wonkspeak) running an explicit monetary policy target providing a clear anchoring of nominal expectations did better than the maestro. This is because the RBA provided explicit anchoring of expectations (so greater narrowing of uncertainty) as a basis for confidence (positive framing of the remaining uncertainty) rather than merely implicit anchoring of expectations derived from confidence in the incumbent central banker. (Any claim that the Oz economy is persistently less pressured by major economic shocks than the US economy, which is almost twelve times its size, is not very plausible.)
To put it another way, the RBA reduced uncertainty in a way that generated confidence; Greenspan generated confidence by positively framing a higher level of uncertainty based on implicit rather than explicit expectations. The former proved to be superior: both in less fluctuations in output and in being a far more robust monetary regime—it had no “succession problem”.
By delivering narrowing of uncertainty via an explicit monetary regime (specifically, one of inflation targeting), the Bank of Canada and RBA minimised the danger of systematic expectation failures in part because a plausible central bank target can be self-fulfilling, as people acting on the basis of such expectations help to make them come true. (The Swiss central bank’s exchange rate target is currently providing a nice example of how credible price targeting by a central bank can operate without the bank having to intervene: the market equivalent of the parent or teacher “don’t make me come over there!” threat.) The US has no such explicit central bank target: which is the first and greatest problem of US monetary policy.
The RBA’s target of stabilising growth in prices (P) over the business cycle means that if output (y) falls, then there is more scope to let prices rise, which improves the terms of labour for firms (since their prices rise before wages do), lessening the impact of the fall in output. (That economic reform and structural changes mean that more Australian labour income is paid in ways responsive to economic conditions also helps cushion employment levels against changes in output.) Conversely, if output surges, then that means that there is scope for less rise in prices, both evening out movements in prices over the business cycle as per the RBA’s target and passing on benefits of output growth to wage-earners, reducing upward pressure on wages and allowing employment to grow faster. In effect, the RBA has been stabilising growth in Py (since more y means less P and less y means more P) and Py = NGDP. So, the RBA has basically been doing the same as Greenspan was, but with an explicit monetary policy regime rather than an implicit one, and so more effectively.
In arguing for an explicit monetary policy regime, I am not arguing for a monetary policy rule. A target is of the form “the Bank will operate in what way it deems appropriate to achieve that target”. Indeed, part of the anchoring of expectations within the Australian economy is the credibility the RBA has that it will operate as required to achieve its declared target.
A rule is of the form “the Bank will act as the rule determines”. First, that presumes the optimum rule is already known so retards learning in the system, beyond learning how to operate the rule. Second, a rule can result in de-stabilising expectations if circumstances become such that acting as the rule requires results in sharply diverging results from previous experience. (Pause here for nod to Goodhart’s law.) A target can be adaptive in a way that a rule is not.
So, yes to explicit anchoring of expectations; no to rule-based roboticism in central banking.
The ECB has a (price stability) target that does not respond to circumstances (at least, not outside Germany) while running an “artificial gold standard“: a sort of bastardised roboticism that anchors expectations about price but destabilises expectations about spending (i.e. NGDP) and debt. The countries that had problems with the gold standard during its 1873-1895 deflationary period included Greece, Italy, Spain, Portugal (scroll down to version in English): the same countries which are now having problems with the ECB’s tight money policies.
There is a fascinating debate about the optimum monetary policy regime. That Christina Romer, President Obama’s original Chair of the Council of Economic Advisers (CEA), has come out in favour of targeting nominal GDP, supported by Paul Krugman, is important in the building pressure toward the US Federal Reserve adopting such a monetary policy regime. Something which is apparently being considered inside the Fed. That stabilising NGDP growth was what Greenspan was, in effect, doing is certainly a strong argument for making it the explicit target.
But the first need is to have an explicit monetary policy regime. This the US Federal Reserve does not have, and (short of a new “Greenspan”: that is, an icon-for-confidence with attached implicit policy regime) any US economic recovery is likely to be anaemic until it does.
On which point, if you look at the US Federal Reserve website, you will see reference to the seven Governors of the US Federal Reserve, but only five pictures. That is because two positions are vacant. Which is the clearest single instance of economic incompetence by the Obama Administration. If President Obama can get people appointed to the US Supreme Court, he can get them appointed to the US Federal Reserve Board. If he could be bothered to seriously try to do so.
It is stunning that such key positions have been left vacant for so long. Any Australian Government that left two positions on the RBA Board vacant for any length of time (let alone months) would be savaged in the economic media and by its political opponents: it is remarkable that there seem to be little or no political costs to the Obama Administration for its amazing, and culpable, negligence.
Disastrous monetary policy destroyed Hoover’s Presidency (Herbert Hoover was a highly travelled, intelligent and open-minded progressive, we should remember), and it is likely to take down Obama’s as well, unless there is strong economic growth between now and the election. If FDR is Obama’s inspiration, it was FDR changing monetary policy that began the economic recovery from the Great Deflation/Depression: the key difference between the Hoover and FDR Administrations was not the New Deal (pdf) but monetary policy. Without serious economic growth, Obama is likely to be, like Hoover, a one-term President. (Not that any Republican alternatives are advocating the right policy mix: but that will not stop Obama losing, as incumbent Administrations get blamed for economic conditions.)
Alas, being “inside the policy loop” does not make you policy smarter: Christina Romer’s public advice on economic policy was better before and after she was CEA Chair than during. The illusion of special knowledge (fostered by the reality of extra information); the pressure to conform, to be a “team player”; the institutional comfort of a lack of an accountable “bottom line”: they all help create official stupidity. And when central banks become stupid, they can be really, really stupid in thoroughly disastrous ways, insulated as they are from the consequences of their actions.
What does the US need? An explicit monetary policy regime. When does it need it? Now.