April 2014

Sun Mon Tue Wed Thu Fri Sat
    1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30      
Blog powered by Typepad

« A Decade of Debt | Main | Can You Fill This Out? »


TrackBack URL for this entry:

Listed below are links to weblogs that reference Hayek on Deflation:


Feed You can follow this conversation by subscribing to the comment feed for this post.

When Hayek wrote on this in the 1920s and 1930s, he expressed doubts about whether monetary authorities could in fact halt a cyclical decline once begun. It was a variant of his knowledge problem. In Monetary Theory and the Trade Cycle, there is a line similar to his famous musings on "the curious task of economics...."

Before the Great Depression, central banks were not in the business of demand management, or counter-cyclcial monetary policy. They managed the gold standard, and (sometimes) acted as lender of last resort to banks.

In his history of the Fed, I believe Allan Meltzer observed 6 deflationary episodes. Only one ended in a depression.

I think that Austrian macro-economists should deal with this "secondary depression". Hayek have some interesting quotations but Haberler or Ropke can be of much help.

Last week I have written a short article in spanish on the topic (http://www.elcato.org/keynes-roepke-y-hayek-que-tan-diferentes-son-sus-ideas). I will try to translate it to english soon.

"I no longer think this is a politically possible method"

Operative word: politically

We can talk of the impacts of deflation, but if the political class is never going to let it happen, the question becomes, what other options are there?

Keynes and Hayek are not too different after all.

I am guessing - I'd love to write it out when I have the time - that Keynesian demand stimulation in a demand-side recession/depression is simply preventing what Hayek calls a (unnecessary) 'secondary depression'.

The role of government here is to provide stable conditions for trade to take place in and this includes stabilising NGDP or the growth of NGDP. This is what makes it possible for people to plan. The idea that demand ultimately comes from itself and the source of spending does not matter, is simply another way of stating that trade ultimately comes from itself: once (re-)started it continues.

Whether this is monetary or fiscal is rather irrelevant from an accounting perspective and the resulting misallocation is outweighed by the (re-)starting of trade/increasing of demand.

The question to ask about the argument for avoiding wage deflation is, where does it end? If no equilibrium money wage rates are to be allowed to decline because of flagging relative demand, very high inflation may be required to preserve general equilibrium. If, on the other hand, some money wages are to be left to decline, why not let them all do so?

This is very tricky territory for macro-economists to tread in. My own sense is that, while considerations of saving-investment equilibrium, or the requirements for keeping natural and actual interest rates in line, argue for changes in M to offset opposite changes in V, it is not at all clear to what extent any further degree of monetary "accommodation" is called for to limit downward changes in equilibrium money wage rates, e.g., in response to growth in the work force or relative changes in demand.

I've got more from Hayek on this topic here:


With more to come later.

After pointing out the disaster of returning the British pound to gold at par in 1925, Hayek writes this:

“The unfortunate decision taken in 1925 made a prolonged process of deflation inevitable, a process which might have been successful in maintaining the gold standard if it had been continued until a large part of the prevailing money wages had been reduced. I believe this attempt was near success when in the world crisis of 1931 England abandoned it together with the gold standard.”

Compare DeLong's non-sense.

DeLong is motivated by the desire to marginalize Hayek, as he's repeatedly proved.

He's not interested in understanding Hayek's policy thoughts, Hayek's economics, or the British policy debate environment in the 1925-1936 period.

In this passage at least, Hayek does not argue theoretically, but politically. He says that it would be impossible _politically_ to get the deflation. But he doesn't say theoretically that deflation would spiral the system into a greater depth of depression.

If I understand the argument of others -- sometimes made here --- that wants to "fight deflation" (bad deflation) it is NOT a political argument but a monetary THEORY.

George or Steve --- please explain to me what is going on here theoretically. I don't care about countering DeLong's efforts to marginalize that Greg raises, I want to know the truth of the position. First, what was Hayek's THEORETICAL position, and second, what actually is the right position independent of whoever said it.

Due to the political economy difficulties, I have come more or less to the older classical economic position that any supply of money is optimal given a flexible price level, so the critical question is to depoliticize the money supply by taking it as completely as possible out of the realm of public policy. Perhaps my position is completely wrong, it certainly might be politically unpopular, but I really don't care about the political popularity as much as limiting politics as much as possible.

It is important to understand the difference between a price level that is "fully flexible" in the sense meaning that there are no barriers to agents' freedom to buy or sell for prices that they deem appropriate, and one that is "fully flexible" in the sense meaning that it never fails to adjust to its G.E. value. Only the latter sort of flexibility--a flexibility which, despite contrary assertions by New Classical economists and some Austrians, is unlikely to be achieved in real markets--makes any pattern of money supply behavior as good as any other. The other, realistic sort of flexibility is such that certain patterns of money supply behavior are in fact likely to be better at keeping real variables at their "natural" or G.E. values than others. Considerations of Political Economy alone cannot alter this truth, though they might well suggest that the ideal behavior of M, whatever it may be, is not achievable in practice, and that we must therefore be on the look-out for the second best.

In fact I believe that a depoliticized money stock can be "second best" (and therefore, paradoxically perhaps, really first best) in the manner just described. But of course there are many different ways in which to "depoliticize" M, not all (and perhaps only one) of which has this merit. There's a vast difference, for example, between a frozen base 100-percent reserve regime and a constant base growth targeting) fractional-reserve alternative, as there is also between the last an a Sumnerian NGDP futures targeting regime. So calling for depoliticizing money still leaves us with plenty of homework to do.

Pete -- there can be different causes of below productivity norm deflation.

And below productivity norm deflation in different situations can be made worse or better.

For example, you can create BPN deflation by returning to an international gold standard at an out of date prior parity (e.g. Britain, 1925).

Or, you can set up a pathological banking system and mismanage monetary policy in the midst of a post-boom, secondary deflation creating bust -- turning a minor secondary deflation into a system wide monetary collapse and mega deflation (USA, 1930s).

And note well, a domestic deflation in not unrelated to what is being done in other markets and with other currencies, e.g. the debt problems and gold & monetary policies of France, Britain, Austria, Germany and the U.S. in the 1914-1938 period.

All of these different variations matter to Hayek in thinking about monetary and fiscal policy -- theory gets applied to very different "deflation" situations with very different empirical & market mechanism conditions.

Note also, there are many different possible "cures" for the different pathological deflations -- Hayek's theory gives reasons to believe some will help, and others will make matters worse.

It depends on what the causal / historical background is, what "cure" is most appropriate -- while some others may never be appropriate.

Does anyone know why Hayek (or anyone) thought that a monetary policy that creates surpluses of labor for those workers whose real wages should rise or remain stable is a good way to get workers whose real wages should fall to accept nominal pay cuts?

Is the rationale that if everyone must take pay cuts (even those whose real wage are rising due to deflation of product prices,) then those who need to take real pay cuts won't feel so bad?

Or is it about an assumption that the problem is the false beliefs of labor union bosses? We must teach union bosses that not all wage cuts represent exploitation by employers? Or that even we must force the union bosses to tell the workers they represent that not all wage cuts represent an intensification of exploitation by employers?

As best I can tell, the "politically impossible" point is that the proposal is for policy makers to generate a level of aggregate money expenditures such that current levels of money wages cause high unemployment in nearly all sectors. We need to teach all the workers that sometimes wages fall, by making all of them take pay cuts if they want full employment, because sometimes some real wages should fall. Well, that argument doesn't fly "politically."

Well, I think such a policy would be evil. And so, I would vote out any policy makers that do it. So, I would help make it politically impossible.

Now, if we had a gold standard, and the relative price of gold rose, and this required just about all workers to take nominal pay cuts, then maintaining the gold standard wouldn't be evil. But if the whole point of the exercise is to teach the workers a lesson by imposing unemployment on lots of workers--particularly workers in sectors that should be expanding and getting higher real wages--well, that pretty bad.

To me, if some sectors require lower real wages, the least bad approach is to maintain an appropriate macroeconomic environment, but with surpluses of labor in those sectors requiring wage cuts until nominal wages fall the proper amount.

Jerry, unless Meltzer's conclusion (about only one deflation episode ending in depression) depends on a distinction between "depression" and "recession," it puzzles me. Both 1920-21 and 1930-33 would appear to fit the bill; had Meltzer not written before it I might include 2008-9.

I hasten to add that "ending in" is admittedly problematic: deflation isn't of course a cause of depression, understood to mean below-natural output. Instead deflation and depression are (in the episodes in question) contemporaneous consequences of a collapse in demand. But recognizing this doesn't make Meltzer's omission of 1920-21 seem any less puzzling, to me at least.

In the context of this discussion of downturns and deflations, I'm wondering if any, here, has read Nouriel Roubini's forecast about the Chinese economy?

See, "China's Bad Growth Bet"


He concludes that the artificial rates of capital investment of various sorts in China, induced by the Chinese government -- partly through direct government expenditure, and significantly through interest rate and credit manipulation by the government and central bank -- has resulted in a growing and unsustainable "over-investment" situation.

If one reads his analysis and substitutes for "over-investment" the more Austrian-preferred terms of "mal-investment" and misdirection of resources, it may appear that China will be facing a significant readjustment and resource reallocation process at some point in the future (Roubini suggests after 2013).

And he anticipates that will include signifiant declines in prices of capital assets due to their being pushed to unsustainable levels during the boom period.

Any thoughts by anyone?

Richard Ebeling


I know the argument, I am asking a question about Hayek's use of the term political, which I would flip on its head.

As Larry White recently put it:"Hayek's explicit policy norm (in P&P) was in favor of preventing secondary deflation. The letter to the editor is consistent with his favoring the necessary amount of deflation to return to consistency with the gold standard, but no more."

I am interested in the "no more" and also the argument that Selgin makes about central banking and the knowledge problem (associated with matching money supply and money demand) in his first book.

So if Selgin is correct, then isn't the problem that efforts to fight deflation turn into efforts to reinflate the bubble. So there is no conflict between P&P and MT&TC.

So to Bill, I would just ask whether he worries symmetrically about violating the "no more" clause as he does about the "evil" consequence of deflation. Given public choice issues, what is more likely to result?

If Hayek is correct, that the necessary recalculation is politically unpopular, isn't the opposite that we are going to get reinflated bubbles? If so, then wouldn't the "best" policy be simply to freeze the money supply and focus all policy efforts on eliminating all restrictions on the flexibility of prices (including wages).

BTW, look at the Facebook thread that John Papola started on this, I think the comments both by Larry White and Mark Thornton are particularly useful in thinking about these questions.


Brad DeLong "the scholar" is up to his old game of selectively deleting comments to his blog posting material which establish facts that conflict with DeLong's fake "scholarship", this time deleting several comments with directly quotes from Hayek laying out Hayek's actual views on Britain, deflation, and all that -- i.e. quotes correcting DeLong's mischaracterizations and misunderstandings & showing that Hayek's views were utterly different that those attributed to him by DeLong.

The man really is a disgrace -- an enemy of understanding, an enemy of scholarship, an enemy of truth.

Amazing that a scientific society would let this man retain professional standing ...


Let's distinguish what Hayek actually wrote on money and prices in his work on cycles from what he in the 1970s recalls was his position. ("I do not think that was ever what I argued.")

In P&P, he suggested that MV be kept constant. To me, it is not entirely clear if that is a long-run target or a suggestion for counter-cyclical monetary policy. He certainly didn't explain how one would go about accomplishing it.

Around the same time, in his June 1932 Preface to the English translation of his 1929 essay, Monetary Theory and the Trade Cycle, Hayek wrote as follows. "But whatever may be our hope for the future, the one thing of which we must be painfully aware at the present time ... is how little we really know of the forces which we are trying to influence by deliberate management; so little indeed that it must remain an open question whether we would try if we knew more."

Hayek clearly was skeptical about demand management or the efficacy of counter-cyclical monetary policy. He suggests a knowledge-problem along with unintended conequences. To say the least, he is ambivalent on the question.

Again I want to remind everyone that, under a classical gold standard, central banks did not engage in activist countercyclical policy. They managed the gold standard.

It is no accident then that Hayek turned to Henry Thornton's Paper Credit, which analyzed monetary policy during suspension. But note importantly that Thornton wrote under the presumtpion that the gold standard would be restored. Monetary policy during suspension is not the same thing as monetary policy of fiat monetary system. In modern parlance, the anticipation of resumption anchors expectations.

Hayek grapled with a real problem in real time. We do not have the knowledge to conduct countercyclical monetary policy. His concerns would resurface in Milton Friedman's AEA presidential address on the long and variable lags.

What we have here is a classic old-man Hayek muddling or conflating of different historical periods and different policy problems.

Here's the problem. What Hayek is talking about here is NOT a post-boom/bust secondary deflation, it is the policy /causal situation of Great Britain faced SEVEN years after the British gold standard deflation of 1925, with politically privileged, internationally overpriced union labor:

"What I did believe at one time was that a deflation might be necessary to break the developing downward rigidity of all particular wages."

In other words, Hayek is giving a deeply confused account of the situation discussed here:


And here:

The unfortunate decision taken in 1925 made a prolonged process of deflation inevitable, a process which might have been successful in maintaining the gold standard if it had been continued until a large part of the prevailing money wages had been reduced. I believe this attempt was near success when in the world crisis of 1931 England abandoned it together with the gold standard.”


Yes, Meltzer was saying there was no Great Depression. 1920-21 was a sharp contraction. The economy recovered on its own. Meltzer pesents evidence that it was the real balance effect that turned things around. The Fed didn't act until after the nadir.

"isn't the problem that efforts to fight deflation turn into efforts to reinflate the bubble."

I'm not sure why this follows.

After the turning point of the bust, an effort to contain deflation and satisfy changing demand for money balances doesn't imply supplying money in the sectors and in the volume required to "reinflate the bubble".

Law professor Glenn Reynolds on marginalization:

"Whenever anyone is effective, the lefty apparatchiks try to spread the idea that it’s somehow impolite to associate with them, rather than attack them on substance. That’s been their MO forever, but it’s become quite obvious and thus far less effective lately."

This has forever been the MO of academics & intellectuals going after Hayek -- Naomi Klein, Paul Krugman, Brad DeLong, and many others in the current period, Paul Samuelson, JM Keynes, JK Galbraith, Melvin Reder, and many others in prior years.

It's a version of the strategy Mises' confronts in the first few chapters of his great _Human Action_.

When Hayek (the "early" Hayek or the "later" Hayek) spoke about the deflationary process in Great Britain after 1925, he was referring to a monetary contraction to restore the foreign exchange value of the Pound to its prewar level, the same way the British government had intentionally did after 1815.

Mises had referred to the "non-neutral" effect on the structure of relative prices during a monetary contraction as much as during a monetary expansion in "The Theory of Money and Credit" (1912, 2nd ed., 1924). And in his 1925 lecture on "The Return to the Gold Standard," he made a point of referring to the "negative" effects that had resulted from Czechoslovakia's decision to "reverse" the earlier price inflation from monetary expansion by contracting that nation's money supply.

In 1931 and 1932, when Austria had gone off the gold standard and the government attempted to "reflate" the price level, Mises went out of his way to say that any attempt to "reverse" this depreciation in the purchasing power of the Austrian schilling had to be done soon and BEFORE all prices had adjusted to a higher price level consistent with the increased domestic monetary supply. Otherwise, a reversing monetary contraction would involve a wrenching adjustment to bring prices in general down, again, to the previous lower level.

In his writings during and after the Second World War, Mises insisted that any return to the gold standard should be at the post-inflationary purchasing power of the monetary unit (the higher price level induced by years of monetary expansion), and not through a monetary contraction to restore the pre-inflationary level of prices -- precisely because of the difficult and disruptive non-neutral adjustment process on the structure of prices and wages to successfully restore the earlier, higher, purchasing power of the monetary unit.

The problem of a (price) deflationary process following the beginning of the Great Depression in 1929-1930 was, I would suggest, a different problem in both Mises' and Hayek's eyes than the intentional monetary contraction resulting in price deflation undertaken by the British in the second half of the 1920s.

(By the way, there were market-oriented, Austrian-"friendly" economists in Great Britain who attempted to make a reasonable and well-argued case for returning the Pound to its prewar level. Such people as LSE professor T.E. Gregory in his book, 1924 book "The Gold Standard Before, During and After the War" [which is an excellent "primer" on the nature, workings and policy value of the gold standard as is his later work, "The Gold Standard and Its Future" (1935)] and his 1925 short book, "The Gold Standard, One Year After.")

Here, in the early 1930s, the argument was that the prior boom years had brought about a misdirection of resources and capital mal-investment. Mises, in his 1931 lecture on "The Causes of the Economic Crisis," insisted that he was not calling for or saying that "all" wages and prices" had to reduced -- a general policy of price-wage deflation. Rather, what was needed was market-based individual price and wage flexibility for the structure of prices and wages to be re-coordinated to store market-clearing, and sustainable price-wage relationships in the economy.

Part of Mises' argument in some of his writings from this period that if, in fact, scarce capital resources had been misdirected by credit expansion and interest rate distortions, some of the "real" capital" of the society would have been wasted ("consumed") for investment purposes now seen to be unsustainable in the post-boom period.

The society would be poorer and the marginal productivity of part of the labor force would now be lower due to the "loss" (the misdirected "wasting") of part of the capital supply. Hence, market-clearly wages in some sectors of the economy would have to decline to bring markets into balance to reflect this change in the supply of capital goods with which workers worked. (See his 1931 article on "The Economic Crisis and Capitalism.")

All of this was intensified, no doubt, due to the "elasticity" of the currency (both multiplicatively expansionary and contractionary) under fractional reserve banking. (See Hayek's clear explanation of this process in his 1937, "Monetary Nationalism and International Stability.")

Both Mises and Hayek recognized, and pointed out, that following the "break" in the boom individuals may desire to hold larger cash balances (greater "liquidity") precisely because of the financial difficulties and general greater uncertainties in the downturn stage of the business cycle.

Mises, however, believed that there was no way to compensate for this increased demand for money held, due to the inability of "injecting" new money or credit into the economy (certainly under central banking) without super-imposing possible new misdirections of capital and labor on those that still needed to be corrected for.

Mises, too, argued like Hayek (though in his own ways of expressing things), that any monetary authority lacked the omniscience to manage the monetary supply in a way that would adapt and adjust it "just right" to reflect any changes in the demand to hold cash balances. That is, to "neutralize" changes on the "money-demand" side so as to leave all "real" relationships in that pattern that would be "as if" money, as the intermediary in all transactions, was itself having no impact on the economy. (That is, a "neutral" money.)

Hence, the monetary central planners lack the knowledge, wisdom, and ability to maintain some hypothetical "monetary equilibrium" in the face of money-demand and money-supply changes in the economy. Restoration of "monetary equilibrium" in the market had to work through and rely upon changes in the structure of prices and wages that cumulatively brought with it a change in the value or purchasing power of the monetary unit.

Richard Ebeling

What Hayek is saying in the passage beliw is what he said about the motivation for Keynesian money policy in the post-1925 period -- downward rigidity of wages leads to inflationary money policies to accommodate or ameliorate it.

Hayek writes,

"the developing downward rigidity of all particular wages which has of course become one of the main causes of inflation."

Even in the quoted passage, Hayek is talking about relative prices and wages. He is even clearer on this in his theoretical work. There is absolutely no idea of "all" wages needing to adjust.

Hayek and Mises distinguished between expected and unexpected inflation and deflation. That distinction is the source of the distinction between "good" and "bad" deflation.

What the anti-deflationists refuse to recognize is that sometimes adjustments in relative prices and wages will be accompanied by deflation. (George Selgin raised the issue earlier in the thread.)

Further, attempts to prevent any deflation will in some circumstances distort relative prices. As Richard pointed out, models of monetary equilibrium are not up to the complexities of actual economies. Thye require more knowledge than policymakers can possess.

Hayek and Mises saw this before the profession could fully comprehend the problem. Friedman also was unable to get the message across to the profession. Brunner and Meltzer explain this very well.

What Hayek is saying is that deflation is not part of the cure for a malinvestment boom. He once hoped that deflation might break the downward stickiness of wages, but he later realised that this was a pipe dream.

Deflation driven by productivity gains is relatively harmless and possibly beneficial, but downwardly sticky wages resist deflation whatever its cause and encourage constant inflation by monetary authorities. Hayek would normally oppose deflation driven by changes in the supply and/or demand for money, but he didn't speak up because he hoped it might break the a developing political economy that he believed would promote further malinvestment booms. Hayek acknowledged this was naive and misled untold numbers about his views on monetary policy.

We can complain till the cows come home that people like DeLong do not understand Hayek on monetary policy, but it is mostly Hayek's fault. When it really mattered, Hayek failed to communicate his true views about monetary policy and deflation; instead, he apparently obfuscated matters for seemingly tactical reasons. Hayek's calculation backfired.

Some commentary from a paper I recently presented at ASC drawing from Hayek, Friedrich A. 1979. Unemployment and Monetary Policy: Government as Generator of the “Business Cycle”. San Francisco, CA: Cato Institute.


I find myself in an unpleasant situation. I had preached for forty years that the time to prevent the coming of a depression is during the boom. During the boom nobody listened to me. Now people again turn to me and ask how we can avoid the consequences of a policy about which I had constantly warned. I must witness the heads of governments of all Western industrial countries promising their people that they will stop the inflation and preserve full employment. But I know that they cannot do this. I even fear that attempts to postpone the inevitable crisis by a new inflationary path may temporarily succeed and make the eventual breakdown even worse. (Hayek, 1979, p.3)

From the paper:

"On monetary policy, at least if tied to a fiat currency under the auspices of a central bank, Hayek (1979, 17-18) retreated or retrenched from his strong anti-price stabilization position and from what was perceived, wrongly according to White (2008), to be his rigid liquidationist views of the 1930s. In the presence of an on-going boom, the way to prevent the boom from becoming an uncontrollable inflationary spiral, was, increases in the quantity of money must be stopped or at least reduced to the “rate of growth of production”, … but “(i)t does not follow that we should not stop a real deflation.” He continued to maintain that deflation was not the “original cause of the decline in business activity,” but came to believe, that the effects of the secondary deflation may be worse than warranted by the money induced misdirection of production and would provide “no steering function.” He thus argued absent significant institutional monetary reform, “Though monetary policy must prevent wide fluctuations in the quantity of money or the volume of the income stream … (t)he primary aim must again become stability of the value of money.” To paraphrase, in normal times there is a need to get back, a la Friedman, to a more or less automatic monetary framework. While such a policy would not entirely eliminate cyclical misdirection of production, the consequences of such a policy would, per White (1999, 118), be “too small to worry about.” Where policy deviated and generated a boom-bust, then, to prevent “liquidity crisis or panics” there is a need “to ensure convertibility of all kinds of near-money into real money” For this, “the monetary authorities must be given some discretion” (Hayek 1979, 18)."

"Events of the period under analysis clearly undermine these empirical judgments. As recognized by Leijonhufvud (2008, p. 1), “Operating an interest-targeting regime keying on the CPI, the FED was lured into keeping rates far too low far too long. The result was inflation of asset prices combined with a general deterioration of credit quality. This, of course, does not make a Keynesian story. It is rather a variation on the Austrian overinvestment theme.” The misdirection of production caused by money creation in a growing economy can be significant and can set up conditions that can cause a threat of a massive secondary deflation. Further research, both historical and theoretical, is urgently needed to re-access what is actually meant by secondary deflation and what are the consequences of such a secondary deflation. Is a secondary inflation a solvency cum money and debt deflation problem (Bordo and Lane 2010) which would negatively impact the money supply and money spending stream, a money demand/liquidity problem reducing the money spending stream through velocity changes and monetary aggregate changes from a buildup of reserves held by financial institutions (Bordo and Lane 2010 and Salerno 2011), or a Keynesian income constrained reduction in aggregate demand and spending (Garrison 2003 and O’Driscoll Jr. and Rizzo 1985)? Is secondary deflation a necessary part of the correction process (Salerno 2011 or early Hayek) or per the later Hayek and many monetary disequilibrium theorists, something that must be prevented to keep a minor bust from turning into a prolonged crisis and depression. Whatever the answer, it should be clear that even if a secondary deflation, however defined, must be prevented, a near tripling of the monetary base (see figure 5) was policy response overkill. This policy response to the current situation has set up future monetary conditions that may be very difficult to unwind without significant inflation and/or a continuing boom-bust pattern; Hayek’s “tiger by the tail” (Garrison 2009)."

Please, oh please: "Fed," not "FED." It's not an acronym. The Fed's apologists seem to get this; its critics can afford to do no less.

Maybe "FED" stands for Fiduciary Exchange Devaluer.

John P. Cochran,

Well done.

Monetary policy is now asymmetric. Rising asset prices are not inflationary, but falling asset prices are deflationary. Of course, the focus of policy was on propping up dodgy bank balance sheets.

Richard Ebeling, I think, is arguing what I was hoping to suggest, especially his point: "the monetary central planners lack the knowledge, wisdom, and ability to maintain some hypothetical "monetary equilibrium" in the face of money-demand and money-supply changes in the economy".


That's right Pete, they don't. But, again, what exactly does it mean to do what you want, which is to "do nothing?" First you have to specify what that means and then we need comparative analysis: are the net costs of "doing nothing" greater than or less than the net costs of missing the ME mark?

Lord knows I've never suggested the Fed can maintain ME. I wrote a whole chapter in each of my books about why they can't for just the reasons Richard says. But in the world of the second best, under some circumstances, it might be the case that instructing the very imperfect central bank to try might be less bad than telling it to "do nothing" (assuming we even know what that means). I just don't see how you can dismiss that possibility a priori given what we know about the costs of price level adjustments.

And now for my usual disclaimer: nothing the preceding is an endorsement of any specific policy of the Fed (no caps George!) over the course of the last three years.

This is a very interesting topic. I agree with Prof. Horwitz that to do nothing requires a specific definition of what that means. In a certain way one may ask how doing "nothing" is not actually doing "something" as well.

What will give us the biggest distortion is, I think, the relevant question. Trying to imitate the market and miss or do "nothing" and miss as well? There's no first best.

If I may use the term in this context, to me it seems that in a situation like the one presented in the question what the Fed (no caps as well) needs to do is to have a good alertness/understanding of what is going to happen in the market and how his 'doing' or 'not doing' will play and evaluate what is the least worst course of action for each particular case. Ultimately, besides the fine tuning provided, the good central banker needs to have a certain kind of vision of the future market situation. The Fed chais is not chosen for his technical skills, but for his understanding of the market beyond the technical results. At some this point the Fed officials stop to be optimizers and need to become a certain a kind of "entrepreneur" as well: see what will bring less distortions.

If, for example, it is given that a certain utility (i.e. electricity) is provided by a state monopoly, should it do "nothing" or "something" when there is a shock in the market or a change in its demand? If it should do something when demand increases, why that doesn't apply to a monetary authority as well. We want the state monopoly to do nothing or at least try to imitate to the best of its limited possibilities what a private firm would do?

The entrepreneur uses observed prices and his estimation of what will happen to the market; objective data and subjective assessment. Shouldn't the Fed officials do a similar exercise? (Taking the Fed as given).

I'm leaving aside Public Choice/Incentive and knowledge problems just to present the ideal central banker only concerned with providing stability.


Lee Kelly makes an important point.

Hayek's "conflicting" policy recommendations turn on changing assessments of policy relevant empirical realities, and not so much on any changes in Hayek's scientific understanding.

Those whose purpose is driving by a politically motivated desire to marginalize Hayek intentionally exploit the ambiguity between Hayek's contingent policy & empirical judgments and Hayek's rather unchanging scientific vision.

And I see no reason to let DeLong and others off the hook -- their purposes are not driven by truth seeking, and they routinely and intentionally misrepresent and suppress knowledge that doesn't serve their primary malevolent purpose.


But let's return to John Cochran's concrete case: triple the monetary base? There was no price deflation, conventionally defined. There were, however, large losses on the assets owned by banks. Should central banks expand the monetary base in order to try to reflate asset values?

No Jerry, of course not. That's why I put my disclaimer at the end. To be clear:

The ONLY possible Fed action for me that was justified was to provide funds for healthy banks at the peak of the crisis in September of 2008. Boosting asset values, grabbing new powers, QE1, 2, 3, 4 or whatever are not defensible. I've said this here and in print multiple times.

My question to Pete was as much theoretical as anything else. I just find it strange that he seems to want to avoid the type of analysis he's done so well his whole career: the comparative question.

I will complete agree that an "active" Fed will be far from perfect. But let's not pull the reverse of the "ipso facto" fallacy and claim the imperfections of an active Fed justify *without examination* "doing nothing." Doing nothing has costs too.

I am also not arguing that the Fed doing something is a priori better. I'm just saying it's not a priori worse.

Scott Sumner's argument is that paying interest on reserves created a monetary contraction, that accounts for most of the economic downturn and most of the unemployment problems.

When we are talking about "doing something" about deflation, sometimes what we are advocating is that something not be done that is being done.

With regard to the comparative point that Steve makes - I wrote a relatively friendly review of Sumner's NGDP targeting proposal (should be online soon, if anyone wants a copy now email me), not because I think it is better than "doing nothing", but because I think it might well be better than "CPI targeting".

I'm constantly surprised at how quickly people want to take the theoretical claim that under a desired monetary regime we might expect the banking system to maintain MV stability, to make an inference/accusation that this is a defense of present central bank attempts to do so. Let's all be clear that Hayek's argument should always be viewed as a critique of central banking.

And lets give greater attention to the differences between policy debates and regime debates.

But even if we agree that farm subsidies should be repealed, it doesn't mean we can't make an informed judgment about whether subsidised prices are likely to be above or below their free market ones, and try to nudge policy in that direction.

Given that central banks exist, we can either focus efforts on campaigning for their abolition, or try to push them in the direction of least harm. This is merely a question of whether we're willing to make second-best policy prescriptions (and thus engage with the current policy debate).

No one is arguing that we *know* what these prescriptions should be, but provided we feel comfortable labeling certain policies "bad" (of which there is no shortage), we can in large part focus on *un*doing things.

Personally I think a "Hayek Rule" would be a marginal improvement compared to the status quo, and that's an intellectual argument that is worth making. It doesn't mean I support the institutions through which a Hayek Rule would be made, nor the knowledge assumptions required to think it would be perfect.

One question and one statement.

The question regards the 1920-1921 crisis. Friedman and Schwartz showed that prices went down from 250% to 150% of pre-war levels in 18 months. The economy was on a gold standard, I don't remember if inflation was achieved by increasing the multiplier given an amount of outside money or increasing the monetary base by printing paper money.

In both cases, however, there was no way to avoid deflation. The banking system either was pyramiding normally, but upon an artificially high monetary base, or was pyramiding too much over a normal amount of (mostly) golden monetary base. In both cases, either by reducing the monetary multiplier to a sustainable level, or by removing paper outside money from circulation, a deflation in the money supply would have ensued.

The fact that it ensued and had no remarkable economic effects except for a little more than one year strikes me as evidence that deflation in normal times is not to be feared because it has just a minor and short-term impact on real variables.

Of course, if it can be avoided it should be avoided, but if to avoid it either an excessive and unsustainable monetary multiplier or an ever growing monetary base are required, maybe it's better to let the deflation runs its course.

The statement regards the 1929-1941 depression.

There are more theories of economic depressions than historical depressions. Depressions are emotional issues and have been given excessive importance in economics research. One of the most harmful and irrational outcomes of this sentimentalism was the spread of Keynesian theories since the late '30s.

One of the last theories in the theoretical interpretation of the Great Depression has been the Cole/Ohanian explanation which gives a central role to Hoover, Roosevelt, government-sponsored unions and government-sponsored cartels. These are real phenomena, not monetary ones. This theory has also been the preferred explanation by Austrians: 1937's book by Chester Phillips and 1963's book by Rothbard are an example. Also Mises said something regarding unions in the late '30s in "on the manipulation of money and credit", without developing a detailed explanation.

One of the first theories, on the other hand, was the financial accelerator theory of Irving Fisher, which has become a kind of religion since Bernanke's (and other's) scolarship in the '80s.

The uniqueness of the Great Depression probably depends on Cole and Ohanian's factors. Thus, we shall not care about money taking the '30s as an example. Of course, without deflation price rigidities are a minor concern, but the source of price rigidities are a much more important factor (and easily solvable problem) than deflation per se, which in some cases it can be unavoidable (for reasons well established in the credit channel literature: if banks need to reduce leverage and improve liquidity after a boom frenzy, they will have to reduce they monetary multiplier, too).

What we know is that the crisis lasts about one year without government intervention during the bust, that the conditions for a large crisis would hold without government intervetnion during the boom, that financial elements can do the recession worser and longer, up to several years.

Aren't we talking about a problem that would exist without union and cartel-oriented microeconomic and stabilization macroeconomic policies?

What does doing nothing mean? I suggest that even when the Fed is very activist, it still does nothing, because it controls much less than it thinks. The latest crisis is a good example. The Fed has been trying to boost inflation for three years. It has succeeded only recently.

But the previous three years of Fed straining to create inflation will bear fruit over the next three years as the economy recovers and businesses borrow more. The result will be procyclical, as Fed policy has always been.

Monetarists tend to think that the Fed can do no wrong. The truth is that it can do no good; it can only restrain itself from doing greater harm during an expansion, but it cannot undo the harm it has done. Only the market can do that.

Didn’t Hayek recommend in “Monetary Theory and the Trade Cycle” a fourth generation of monetary theory? The third generation was fixation on general price levels. The fourth generation would focus on the effects of money on relative prices instead of a general price level.

Hayek’s fourth generation theory should guide the Fed. Without free banking discipline, fractional reserve banking requires an institution like the Fed to limit the growth of credit. But the general price level is a terrible guide, as every Austrian has demonstrated. The Fed tried targeting money growth in the early 80’s and that failed, too.

The Fed needs an index to keep track of the relative prices of consumer and producer goods. I think Wainhouse offered a good start toward such an index in his 1984 paper.

Steve raises important issues. There are really two separate policy/institutional issues that need to be addressed: 1. what is the best policy response given current institutions once a crisis has occurred?; 2. How can the institutions be reformed to reduce the probability a next crisis?
Hayek in the 1970s simultaneously did both in different settings - contrast Unemployment and Monetary Policy where he primarily deals with 1 and Denationalization of Money which addresses 2.
As a guide to examining what was actually done recently relative to #1 I suggest a very interesting and important paper published in the Independent Review, “Ben Bernanke versus Milton Friedman: The Federal Reserve’s Emergence as the U.S. Economy’s Central Planner”, by Jeffrey Rogers Hummel. Hummel provides, without explicitly mentioning the term, the intellectual foundations for John Taylor’s criticism of recent policy as a “Mondustrial Policy”. Hummel builds his case by illustrating the significant differences in “approaches to financial crisis” between the Bernanke approach and a Friedman approach. In addition to exposing the theoretical foundation of a what can be reasonably interpreted as a misguided and dangerous policy, Hummel provides a very detailed almost step by step use of this type of policy in response to the major events of the recent crisis. A must read for anyone interest in the minute details of how and why the Fed’s balance sheet expanded so significantly and how much of what was done did not and does not show explicitly in ‘regularly’ reported monetary aggregates, their sub components or Fed balance sheet reports.
Hummel argues the differences have been rarely noticed, perhaps because Taylor’s warning went unheeded within mainstream commentators, but he correctly summarizes the impact as “those differences resulted in another Fed failure – not quite as serious as the one during the Great depression, to be sure, yet serious enough – but they have also resulted in a dramatic transformation of the Fed’s role in the economy. Bernanke has so expanded the Fed’s discretionary actions beyond controlling the money stock that it has become a gigantic, financial central planner
A couple of other Hayekian insights from the 70s all from Unemployment and Monetary Policy relative to #1 and deflation.
“if I were responsible for monetary policy of a country, I would certainly try to prevent an impending deflation (that is , an absolute decrease in the stream of incomes [emphasis mine]) by all suitable means, and would announce that I intended to do so. This alone would probably be sufficient to prevent a degeneration of the recession into a long-lasting depression.”
“But although I recognize that a general reduction of money wages is politically unachievable, I am still convinced that a required adjustment of the structure of relative wages can be achieved without inflation only through the reduction of money wages of some groups of workers, and therefore must be achieved.” Both p. 16.
“We must certainly expect the recovery to come from a revival of investment. But we want investment of the kind that will prove profitable and can be continued when a new position of fair stability and high-level employment have been achieved. Neither subsidization of investment nor artificially low interest rates [emphasis mine] is likely to achieve this position. And least of all is the desirable (i.e. sable) form of investment to be brought about by stimulating consumer demand.” p 42

The last paragraph (p. 42) quoted by John Cochrane repeats Hayek's long-standing position. It is well-articulated in the letter he co-signed to the Times, answering the call of Keynes et al. to stimulate consumption.

Hayek's reassessment of what to do about deflation is as much politics as economics. It is not necessarily consistent with his theory of wage determination. It may not even be internally consistent.

In the wake of the recent crisis, it may be that most wages needed to fall in nominal terms (though certainly not to the same extent). Then Hayek is confronted with a choice between permitting deflation or achieving the needed fall in real wages (again to different degrees) through inflation.

The reason for my assessment is that capital was malinvested and, hence, to some extent lost. Labor is a complementary factor and thus the demand for labor fell. Pete will recognize the analysis as Hayekian and embodying Mill's 4th Fundamental Proposition with Respect to Capital.

Folks, I'm sure everyone here is familiar with the interview in the late 1970's when Hayek was asked what he thought of Milton Friedman. With a grin on his face and a slight chuckle, Hayek explained that in some ways Friedman was still a Keynesian of sorts. Then he went on and explained what he meant. The bottomline of Hayek's remarks revealed that even in old age, he never truly bought into the entire enterprise of macro-economics. At his core, he was a microeconomist, and though he differed on some of his core epistemological groundings, as a practical economist, he never diverged too far from Mises.

In this light, I'm not sure Hayek took himself all too seriously on any second-best proposal that he might have discussed in an offhanded way when it came to central banking.

With the volumes of paper Hayek spent deconstructing constructivist rationalism, it is difficult for me to accept that he actually took any proposals about monetary rules under central banking (even his own) all too seriously. In the end, I think he would say that there is very little hope that any of them will work according to plan, but some may be more politically tractible than others.

I more or less would liken his view to an apocryphal story about the Titanic's eager crew, where they suggest to the ship's captain that they might be able to go down below decks and patch the leak. The captain pauses for a minute, and then the eager crew reminds him that if they don't try something the ship will sink and they will all go into the cold ocean. In private resignation knowing that the ship is headed to the bottom, the captain tells the crew to go down there and give it a good college try.

In later years Hayek comes close to saying that Keynes' advocacy of inflation may have been the only politically viable response to Britain's discoordination problem caused by the misguided return to gold at par in 1925.

At other times, he suggests that Britain's labor price problems were almost resolved at the time Britain abandoned the gold standard.

What we need to consider is that Hayek didn't have the same knowledge we have today about the size of French gold reserves, and all sorts of other macro data from that period which we take for granted today.

Hayek's views changed dramatically after learning the data presented in Friedman's Monetary History -- and we've learned more in all sorts of ways in the years since.

I have a simple question for those who advocate that the Fed do nothing facing an increase in the demand of money.
Let's say that base money demand increases by 10%. And let's say that the Fed expand the balance sheet by either 8% or 12 %, failing to match the 10% due to knowledge problems.
In that scenario we would have, ceteris paribus, a 2% deflation or a 2% inflation.
Is that better or worse than having a 10% deflation doing nothing?

I find O'Driscoll's arguments in this thread puzzling.

Am I missing something? In my view, if the Fed had not increased the money base a vast amount, the result would be that the quantity of the media of exchange would have contracted to a tiny fraction of its pre-crisis level, and the result would have been a Great Depression scale contraction in the flow of money incomes.

(As an aside, I don't believe that it means that output would remain permanently below capacity. And, with strong efforts to fight price and wage rigidity, maybe a full recovery could occur in a few years. So?)

If the Fed had not begun paying interest on reserves, then the size of the needed expansion in the monetary base to keep the quantity of money that nonbanking firms and households use from falling would have been less. I don't really know how much less. But given that interest rate policy, nearly all of the increase in the monetary base was necessary.

I find worries about the base puzzling. The Fed gets weekly figures on the quantity of money. To me, offsetting changes in the money multiplier is not a problem.

Now, offsetting changes in broader money demand (or velocity) _is_ fraught with difficulty. The broader measures of the quantity of money have expanded, and I think they should have expanded much more. I do think that is problemantic. How much more should they have expanded? When exactly should they contract? GDP figures come out quarterly with a lag. (I am not entirely convinced by Sumner's claims that changes in the quantity of money impact the stream of money expenditures with no appreciable lag.)

But the monetary base? Sure, it is pretty much certain that it will need to contract alot in the future. At that time, the expansion of bank lending will need to be dampened a lot compared to the exansion that would occur if the monetary base were maintained at its current high level.

I understand that there were (and maybe are) solvency issues with the banking system. I think it is absurd to think that the banks are holding large amounts of reserves because the assets that they sold off to the Fed would have made them insolvent if the Fed hadn't bought them. Even if they would have been insolvent and the Fed's asset purchases bailed them out, that isn't a motivation to hold reserves now. No, we are left with the .2 percent interest paid on reserves. And that means that banks hold reserves rather than T-bills.

(I would rather the banks hold T-bills and favor negative interest rates on reserve balances.)

I know it can't be true, but when I read O'Driscoll, I get this strange feeling of money and credit confused. I know the money multiplier account of the quantity of money is a bit simplistic, but there is a lot of truth in it.

Bill Woolsey writes that he finds my arguments "puzzling." What I find puzzling is Woolsey's puzzlement. He drops in periodically and expresses amazement that folks at CP actually make Austrian arguments. Hard to believe.

I also don't understand why he singles my comments out since they are in conformity with most of the other 40 comments in this thread.

For instance, consider the argument made by Leionhufvud (quoted by John )Cochrane. If you engage in CPI targetting, you have no reliable gauge of the ease or tightness of monetary policy. If the real rate rises, you will be holding nominal rates too low and misallocate capital all over the place. Regardless of the movement in the prices measured in the CPI (which is overweighted toward the weak housing sector).

Woolsey also appears to be locked in a time warp with respect to the growth of the monetary aggregates. They're growing and have been since the end of the recession. Go to FRED and take a look at M1.

And the problem is not just how much is on the Fed's balance sheet, but its composition. The Fed can't realiztically sell a lot of the junk, so it will need to try to restrain growth in the aggregates by increasing what it pays on reserves. Whether it was good idea to start paying interest on reserves, the Fed is stuck with it now.

As a blackboard exercise, Bernanke can control money growth by paying interest on reserves. What will the rate need to be once we get a normal recovery? If he miscalculates, we could get either high inflation or another recession.

Pietro M.

Rothbard was dismissive of the power of unions in the early years of the great depression and through the 1920s because unionised employees accounted for such a small share of all employment – less than 10%.

Rothbard attributed the high real wages to an employers high wage cartel sponsored by Hoover, but he is weak on explaining how this is wage cartel is enforced.

Ohanian explains the start of the great depression by focussing on late 1929 when there was a 40% collapse in manufacturing hours worked inside a year.

Ohanian argues that Hoover demanding high wages is return from protecting the large manufacturing employers from the threat of unionisation.

This was some sort of cartel enforcement technology. There was no similar drop in hours worked in agriculture in the great depression and a large drop in the real wages of agricultural workers.

The federal government was 4% of GDP back then in 1929 and 1930 with few regulatory powers so I am not sure that Hoover carried as big a stick as either Rothbard or Ohanian suggest.

The Ohanian-Cole writings on the New Deal and the inter-war depression in the UK are excellent explanations for regulation and unemployment insurance stifling employment, labour supply and productivity growth and prolonging depressions.

Real business cycle explanations of the prolongation of the great recession and the much larger relative drop in hours worked in the USA such as by Ohanian in JEP deserve close attention.

The comments to this entry are closed.