August 2017

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 31    
Blog powered by Typepad

« "Where do we come from? Who are we? Where are we going?" -- Paul Gauguin, 1897 | Main | 2010 Adam Smith Award -- Remarks Read on April 11th in Las Vegas »


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

I can't believe that Tyler Cowen wrote that post he wrote on Marginal Revolution. I think he must read a lot of new books and throw out the memories of old books.

Trying to make a big deal out of agreement between Austrians and Keynesians on the problem of uncertainty is a no brainer because everything depends on where you go from that point. Quite apart from intellectual errors like thinking that capital is homogeneous, there is the problem of regime uncertainty which the Keynesians address by saying you will get all the certainty you need from our helpful interventions. But we have long ago reached the point where most of the interventions are the problem, not the solution.

Uncertainty is a fact of life and maybe the role of good policy is to provide political and legal stability - not certainty but a reasonably stable institional framework where people can plan for the future with some expectation that they will not be shafted by the next political or administrative decision.

If the institional framework is badly broken it need to be fixed but as Roger likes to remind us, that can only be done by piecemeal, experimental steps to handle unintended consequences.

Wise words, Pete.

Krugman raises an interesting point, indeed. I don't think it has been addressed so far by ABCT scholars. Krugman claims that ABCT should predict unemployment during the boom. He sees the boom and the bust as symmetric deformations of the structure of production, both requiring reallocations of resources and causing unemployment.

Krugman is wrong about the boom, I think, because the lengthening of the structure of production during the boom is a gradual, marginal process. Its cause is the continued expansion of artificial bank credit over a period of time. It shifts entrepreneurial expectations gradually, but entrepreneurs have no reason to lose their "confidence" to speak keynesian. This creates a favorable labor market.

The cause of the bust is quite different. It is, fundamentally, a change of expectations; a massive, rapid one. All credit-financed projects cannot be completed simultaneously. So whatever the reason, people must find out at some point or other that many of their projects will fail. The way the bad news spreads is rapid, contagious (fire sales of assets), which is close to the keynesian story. This creates a burst of uncertainty. It takes time for a new set of expectations to form, for new asset prices to emerge, etc. During this process, entrepreneurs are "blind" and they know it (they know that they don't know future conditions). It is a situation of low confidence, high uncertainty, high preference for liquidity in keynesian terms. This creates an unfavorable labor market.

So my view is that the ABCT DOES describe correctly the boom, but does not generally focus on the bust. As for the keynesian story of the bust, it is perfectly compatible with austrian economics.

Gu Si Fang:

"ABCT should predict unemployment during the boom"

If demand increases in long-term capital goods, unemployment is reduced AND the structure of production is changed; if demand decreases in long-term capital goods, unemployment is increased AND the structure of production is changed.

There is no reason to expect that ANY change in the structure of production need increase unemployment: there are structural changes involving a fall in job vacancies (increased employment) and structural changes involving a fall in job positions (increased unemployment).

Peter -
Thanks for the interesting post, and the shout out.

I think it's important to note that for Keynesians it's not always a "broken joint". Lawrence Klein once wrote:

"It is wrong to think that Keynes’ system fails if it cannot predict pessimistic results. The pessimism is not inherent in this system; instead the determinants behind the system make it operate either pessimistically or optimistically, depending on the current state of affairs in economic and non-economic life."

For Keynesians, it's a "broken joint" only under certain conditions. Under most it's probably best thought of as "loose", and when it's "tight" there's no real conflict with the Classics. The point Keynes was making, though, is that there's no guarantee of the preconditions for the Classical perspective.

Anyway - I was wondering what you would suggest reading of Garrison's. I'm very under-read in the Austrian school, and Garrison is one that I'm particularly interested in when it comes to insights I can derive from Austrian economics (I'm sorry, you're not really going to hook me on the praxeology or epistemology).

Rafe -
I think you make too much of things like "capital homogeneity". Sure there's some homogeneity when ideas are first introduced, for the sake of easy explanation. They teach it that way at first because the model isn't driven by capital heterogeneity the way the Austrian model is. But don't make the mistake of thinking any of the insights would change by adding heterogeneity, or that Keynesians beyond their initial stages of instruction are somehow dedicated to an unrealistic capital homogeneity assumption.

Austrians do the same thing. In a recent Garrison lecture I watched, he spoke of five different types of capital. Yes, one type of capital is unnecessarily homogenizing. But only five is unnecessarily homogenizing too! It's all just a pedagogical tool - don't make too much of it.

I would agree that it's "what they do afterwards" that differentiates them. I would argue that what sets Keynesians apart is the liquidity preference theory of interest, which is the mechanism for providing the "loose link" that Boettke describes.

I don't understand the idea of the long-run rates untouched by monetary policy. Interest rates move in parallel, just look at FRED to be sure.

Correlation between fed funds and 3-year interest rates is very strong (R-squared 0.87 with a RATE3Y=a+b*FEDFUND equation). This means that to know the present monetary policy and to know the present structure of commercial paper, t-bills and constant maturity treasuries rates is almost the same thing, from 1m to 5y. One can phone Bernanke and know the whole structure of the interest rates listed above by just asking for the Fed rates.

So, either Bernanke looks at the interest rates and sets the target rate, or it affects interest rates at all maturities with his only policy instrument. The latter appears to be the correct explanation to me: a 1% cut in Fed funds is a 0.6% cut in 5y rates, i.e., a 3.3% compounded gift to 5y debtors.

But maybe I'm missing some VAR specification and some data modeling...

Like Cowen, I think Krugman is more or less correct.

Pietro unwittingly makes the point. He describes the process in such a way that aggregate demand rises in the boom and falls in the recession.

In the boom, real expenditure on capital goods rises, while real expenditure falls on consumer goods. (Of course, if expanding nominal expenditure can somehow increase the production of both capital and consumer goods in apparent conflict with scarcity, pushing outside the production possibilities frontier a la Garrison, then we are in "Keynesian" territory.)

In the "Austrian" recession, real expenditure on capital goods fall and real expenditure on consumer goods rises. (Of course, if falling nominal expenditure causes real output to fall, then we are in "Keynesian" territory, though, the monetary equilibrium school will _correctly_ point out that the "secondary depression" is fundamentally monetary and the Keynesian account of it is so confused that it is best described as a "Keynesian diversion.")

Pietro's statement that the boom is gradual and the bust sudden is more sound. But I think this is exactly where the Austrian Business Cycle theory goes wrong. The malinvestment can only be due to monetary disequilibrium. The real excess supply of credit is the other side of the coin of the excess supply of money. It is people holding excess money balances that they haven't got around to spending yet. It is people holding excess balances whose real value has yet to return to equilibrium by decreases in its purchasing power.

Similarly, if there is a true "austrian" recession, sans secondary depression, then why is it so rapid? If nominal expenditure for capital goods continues to rise, but just more slowly than consumer goods, so that resources are stripped away from capital goods industries, why is this process sudden?

None of this means that malinvestment cannot be part of the inflationary processes, it is just that the gradual boom and sudden bust is something else. For example, there is a gradual nominal expansion, and perhaps a sudden nominal collapse. The nominal expansion isn't all malinvestment. The real distortion is created by the nominal expansion because prices don't rise enough to keep real balances equal to demand to hold them. Most of the nominal expansion in money, credit, stock prices, and the like just is reflecting higher prices of the capital goods that reflect equilibrium saving and investment.

Anyway, back to Pietro's point about gradual and sudden. If the change in the allocation of resources is mostly about what industries expand, then unemployment will show little change. The economy is growing! Some industries grow faster than others. New employment opportunities appear in some sectors rather than others. New entrants in the labor force find jobs in some sectors or others. The composition of output is changing, without any industry shrinking.

If, on the other hand, the change in the composition of demand is so great that some industries must contract absolutely, then specific capital goods will abandoned. There will be layoffs. And more than usual unemployment. (Kling makes this point sometimes, though I think in a confused manner.)

If the boom was very large, so that consumer goods industries must actually shrink to free up resources for capital goods industries (rather than grow more slowly than optimal so that capital goods industries can grow faster than they really should,) then there would be structural unemployment in the boom.

In a static economy, this would always happen. But in the real world, industries are growing all the time. And so, monetary disequilibrium can distort the direction of growth rather than force a absolute contraction of consumer goods production.

And, of course, there is no reason why the bust cannot occur in the same way. That is, consumer goods grow faster than usual, and capital goods grow less than usual. New entrants into the labor force get jobs in consumer goods industries rather than capital goods industries. The demand for capital goods appropriate to shorter processes grow more rapidly. For longer processes, more slowly.

If the bust is sudden, so that output must shrink in capital goods industries, then there is unemployment, abandoned capital goods, and the like.

The more gradual "bust" is a very likely scenario if it occurs by the monetary authority simply failing to accelerate inflation--that is, if price inflation is allowed to catch up, return real balances to the desired level, and so allow the real supply of credit to return to equilibrium and the market interest rate to return to the natural rate. If nominal expenditures on consumer goods industries grow more rapidly than on capital goods industries. If capital goods industries cannot maintain past rates of growth because they cannot compete for resources with the even more rapid growth of consumer goods industries.

P.S. While heterogeneous capital goods is more realistic, I think the role in the debate is a counter-argument against the claim that the accumulation of capital goods in the boom allow a permanent increase in production and consumption. While that is an effect, it doesn't mean that there is no misallocation of resources. Capital goods are heterogeneous.

For example, we have lots of single family houses right now. That is real wealth. The resources that would have been needed to produce housing are freed to producing other things. But that doesn't mean that resources don't need to be shifted from construction to other industries. It doesn't mean that there wasn't waste. But the houses that exist are a benefit that partially offsets the other, more valuable goods that were sacrificed.


I would guess you probably don't have the "right" reading of praxeology and instead have the caricature version that often floats around the internet supported by disembodied quotes from Mises that don't place the positions in their proper philosophical context. But this is not the place to discuss that. Instead, you ask about Garrison. Well, if you go to his website he has most of his articles linked. Personally, the two articles that had the biggest impression on me were an essay on Intertemporal Coordination and the Keynesian Vision published in HOPE, and his essay Time and Money: The Two Universals of Macroeconomic Theorizing from the Journal of Macroeconomics. Both were published in the 1980s.

But given your interest in Keynesian type arguments about the problem situation of economic actors, I would suggest O'Driscoll and Rizzo's The Economics of Time and Ignorance. You might also benefit from a reading of the various reviews of the book in Critical Review (Davidson's rather nasty review, but followed by the more sensible reading by individuals such as Jochen Runde (Cambridge University).

I know epistemology and ontology are not necessarily your thing, but a good Keynesian (just as a good Austrian) should be familiar with the works of Shackle (especially Epistemics and Economics) and Tony Lawson (and the social ontology group) even if in the end you decide that is not how you want to pursue economic studies.


And I would recommend Roger's book *Time and Money* Daniel. (You could even read my book too, as it's highly complementary to Roger's, though less direct of a comparison of Austrian and Keynesian approaches.)

I've been wanting to read Time and Money, Steve. Thanks for the article suggestions too, Peter. Those are probably easier to start with. Although I lead with the similarities in the framing of the problem in my blog post, that actually isn't my primary interest. The other day I was listening to a Garrison lecture and I just thought "Geez - this looks like it could easily be merged with a Modigliani-type IS-LM model". Modigliani would add thinking about liquidity preference to Garrison, and Garrison would bring insights about the capital structure. They're different, but they're not inherently contradictory, and one of the things I've appreciated about the Austrian school is their thinking on the capital structure. That just got me thinking. I don't know if anyone has tried that sort of synthesis before, but it seemed fruitful to me.

There are lots of little things as well. A lot of New Keynesianism is driven by lots of asymmetric information type models. Decentralized information drives a lot of Austrian insights, so what happens when Austrians explicitly consider asymmetries and idiosyncracies in the distribution of information the way New Keynesians have? Perhaps they already have considered these sorts of things and I'm just not aware. Regardless, it always seems like we're talking past each other when I get lectured in an Austrian comment section about decentralized information, as if I'm not aware of how important that is.

It's little things like that where I think "we aren't really as contradictory as we make ourselves out to be - we just emphasize different angles from which we can view the problem".

Epistemology and ontology definitely do interest me, but it's something I inevitably dabble in rather than really dive into. I'll be applying to PhD programs in the fall. I'm hoping I can read more deeply than I have the opportunity to now once I get into that.

Also - thanks for being so gracious in your link to my blog. I can sometimes be terse when I discuss the Austrian School on mine, but I really do have a great deal of respect for it.

As a Garrisonmania suggestion, considering that the comovement issue (consumption falls or rises with investments) has been discussed by many, there is a 1993 (I guess) paper, available on his site, about overconsumption and forced savings in the theories of Mises and Hayek. Although that material is also in Time and money, that article for me is the starting point to assess most of Krugman's and Cowen's (and Woolsey's, Tullock's...) cricitisms of ABCT.

The standard version of ABCT with reduced consumption during the boom originated with Hayek and is so at odds with reality that although I believe ABCT is right I would never say that. :-)

For what concerns praxeology, I quote Boettke: praxeology is not absurd, although less powerful than most think. However, although hyperrationalist interpretations of Mises are for bloggers, they are not the layman's fault: Rothbard and Hayek sometimes made the same mistake. The problem is that Mises is read in terms of Theory and the other 50% of his epistemology, History, is downplayed. Theory is a jackknife, not a miracle machine.

Prof. Woolsey,

I think I didn't want to say what you claim I said. Besides, I'd want to express some technical opinions about ABCT which, although influenced by my favourite living economist (Prof. Garrison), may not be obvious to all because it's my own reconstruction.

Very long comment. Sorry.

"Pietro unwittingly makes the point. He describes the process in such a way that aggregate demand rises in the boom and falls in the recession."

My point is that *real* GDP rises more than sustainably feasible, the fact that NGDP rises is true, too, but not central. The economy is overheating in real terms during the boom, producing beyond capacity: it's Y/P > PPF.

"In the boom, real expenditure on capital goods rises, while real expenditure falls on consumer goods. (Of course, if expanding nominal expenditure can somehow increase the production of both capital and consumer goods in apparent conflict with scarcity, pushing outside the production possibilities frontier a la Garrison, then we are in "Keynesian" territory.)"

Real overproduction is not keynesian, it is misesian (in Hayek there is nothing like that): malinvestment and overconsumption are a movement beyond the sustainable production possibility frontier, not a free lunch of an economy working below it's potential due to market failures. Only movements within the PPF are keynesian (or MET, or credit channel).

I would describe the boom in three chronological steps, in which the first one lasts a money circulation period (one month?) and is separated for expository purposes only:

phase 1: money injected as credit causes an investment boom.
phase 2: money originally injected as credit circulates as wages and causes a consumption boom. (underlying hypothesis: C and I go beyond the sustainable PPF, otherwise phase 2 would cause a halt to investments).
phase 3: binding production constraints halt the simultaneous booms, prices may start rising, consumption and investment will no longer be able to comove, the boom ends independently on monetary policy (the delay between phase 3 and 2 is not due to monetary circulation, is due to the time structure of production: overproduction is possible, but not forever).

"In the "Austrian" recession, real expenditure on capital goods fall and real expenditure on consumer goods rises"

Here I have three things to say:

1. It is true that price rigidities (MET) can push the economy below its PPF.
2. Not only price rigidities can do this: also credit deflation (as in Strigl, "Capital and consumption") and financial accelerators of various kinds. MET is a special case of a more general class of accelerators which make the recession harder than neoclassically ("movements along the PPF") necessary: all kinds of financial frictions can do the same job as price rigidities.
3. Even if the bust had no decelerators (the original Austrian "below PPF" accelerators are credit destructure and panic coordination problems), real expenditures in consumption should *fall* and not rise during the recession, because the recession is the discovery of the squandering of capital, i.e., it is an inward movement of the PPF. In reality, part of the recession is not a movement of the PPF, of course. There are frictions, and rigidities are one of them.

"Pietro's statement that the boom is gradual and the bust sudden is more sound."

It wasn't me. It was Gu Si Fang. Anyway, of course it is.

"But I think this is exactly where the Austrian Business Cycle theory goes wrong. The malinvestment can only be due to monetary disequilibrium. The real excess supply of credit is the other side of the coin of the excess supply of money. It is people holding excess money balances that they haven't got around to spending yet."

I'm not going to say that ABCT explains the severity of all the recessions, it is not true (as recognized by Mises 1931, C. Phillips 1937 and Rothbard 1963 for the Great Depression, when they pointed out the role wage rigidities). But the recession happens when productive unsustainability is revealed, it is a concept that can do without nominal frictions (which, however, exist).

As there is no exposition of ABCT which satisfies me on this point, the best being Garrison's - as usual - which is too aggregate, I make an example.

Let's assume that I've undermaintained old fixed capital for five years. I've moved workers toward the production of new fixed capital (which will require a lot of saving which will not be available on time and which adds to the consumption stream during the overconsumption phase), and toward the production of consumption goods. I can do this because fixed capital is durable, and as long as old and new fixed capital does not physically deteriorate, the boom can go on. When however the surge in the demand for saving arrives but demand for consumption remains strong, the recession begins. During the recession I have to reduce consumption, invest in the old and new fixed capital to save part of it, and scrap the unusable (for lack of complementary capital goods, i.e., savings) submarginal part of it.

This would be a problem also without frictions. Frictions only add severity and length to the readjustment. The countercyclicality of consumption which Krugman and Cowen criticize only exists in Hayek's writings, but not in Mises's or Stigl's or Garrison's. Hayek's theory of forces saving is the standard account of ABCT, but it is theoretically indefensible. Unfortunately, more sophisticated versions to my knowledge are way too complex. What really happens to the economy during overproduction and how real resources come to bind at the onset of the depression is at least to me still a mistery.

"If nominal expenditure for capital goods continues to rise, but just more slowly than consumer goods, so that resources are stripped away from capital goods industries, why is this process sudden?"

This is a very interesting question. Hayek said that it is impossible to consume now what will be available for consumption tomorrow, although it is always possible to save and invest now what is not consumed today: it's the time of production asymmetry. Frictions (among which price rigidities, I repeat, are only a part of the problem) add to the problem. I would consider, however, speed as a positive thing: it is better to have a 3 month crisis then a 5 years one, even for similar cumulative losses. The ideal would be a one week crisis: 100% unemployment for a week and then business as usual. This is impossible, however, because the capital structure need be fixed before reverting toward the sustainable path. Fixed capital is to be reshuffled, consumption need to change, workers need to be retrained and reallocated. I would allow several months for this, and for the panic and the ensuing coordination problems, even in the best case. This is a point I never considered, thanks.

There can be, however, a slow recession, which implies no readjustment of capital: Greenspan called it a soft landing. We saw them in 1921, in 192 and 1927, and in 1990 and 2000. Problems weren't solved, just postponed, but the bill turned out to be expensive.

Can there be a slow recession, i.e., a slow building up of demand for capital because of capital undermaintaince during the boom which causes a slow reduction in employment, capital utilization, output, monetary multipliers and consumption which does not impede recovery? Can time be bought for readjustment?

I believe that the secondary deflation is at least in part unavoidable. It adds to the pain, but there are no painkillers available. Friedman and Schwartz said that the expansion in the '20s made the banking system more vulnerable to shocks. There is no boom without an overextension of monetary multipliers, financial leverage, risk... and there is no recession without a painful return to normalcy. If banks need to deleverage, also some good investments will be starved, if banks are forced not to deleverage by nominal expansion, the inherent fragility of their position at the onset of the boom remains (and moral hazard adds to the problem). It is a pity that the financial structure has never played a key role in ABCT... and it is largely compatible.

"Anyway, back to Pietro's point about gradual and sudden. If the change in the allocation of resources is mostly about what industries expand, then unemployment will show little change. The economy is growing!"

This is Krugman's argument. It assumes, however, that production with capital goods (roundabout) and production with hand-to-mouth technologies are equally productive. The lack of saving stops the production process. There is no boom in consumption because there is a tug-of-war between consumption and demand for credit. No one can produce without capital, and if consumption weren't curtailed, there would be no funding available for the restructuring of production.

I will make an easier example. Krugman (Brookings, 1998) said that devaluation may have helped the Japanese economy by fostering net exports. By fostering carry trade, however, devaluation would just syphon domestic savings out of the country. With which funds were japanese going to restructure their economy, if investing in T-bills was better than investing in Mitsubishi bonds?

"For example, we have lots of single family houses right now. That is real wealth. The resources that would have been needed to produce housing are freed to producing other things. But that doesn't mean that resources don't need to be shifted from construction to other industries. It doesn't mean that there wasn't waste. But the houses that exist are a benefit that partially offsets the other, more valuable goods that were sacrificed."

Houses do not require complementary factors of production, and this make them unlikely to be submarginal, although I read of 1$ houses somewhere in Detroit. The problem is the sunk cost of having wasted so many resources producing things of so little usefulness. If they were oil drills instead of houses, and had no saving for refineries, we also would have submarginal fixed capital.

PS I corrected a dozen mistakes. There may be another couple of dozens.

Krugman and Cowen are clearly not right about Austrian Business Cycle Theory relying on Keynesian ideas. If we assume they were, however,shouldn't they admit that, chronologically, Keynesians are relying on Austrian ideas? Mises' THEORY OF MONEY AND CREDIT, MONETARY STABILIZATION AND CYCLICAL POLICY, THE CAUSES OF THE CRISIS, and THE CURRENT STATUS OF BUSINESS CYCLE RESEARCH AND ITS PROSPECTS FOR THE IMMEDIATE FUTURE; and Hayek's PRICES AND PRODUCTION were all published BEFORE Keynes wrote the GENERAL THEORY. It would seem that if there is serious similarity of views, one would have to say that Keynesians are relying on the Austrians. Wouldn't Krugman and Cowen have to admit that we are all Austrians now?

Shawn -
Ya, I think they overplayed the "reliance" issue. I don't think Austrians are "relying" on Keynes any more than Keynes relied on them. The point is, Krugman is arguing that Austrians abandon the logic that they share with Keynesians when it comes to busts. Keynesians therefore are more consistent than Austrians in the sense that they haven't simply abandoned the dynamic of boom times when it comes time for a bust.

I'm not saying I agree with Krugman and Cowen on that - I'm not informed enough on this to say. But I think that was the point, not that either theory is derived from the other.


You are probably right that that was their point. I was using Boettke's language. However, Cowen does say this:

"Austrians slip Keynesianism in through the back door. Implicitly, they associate booms and slumps with rising or falling aggregate demand — utterly unaware that their own theory doesn’t actually make room for such a thing as aggregate demand to exist, or at least to affect overall employment. So Austrians are basically Keynesians in denial."

How can it be true that Mises and Hayek associated booms and slumps with rising or falling aggregate demand if they didn't have a place for it in their theory?

ABCT is incomprehensible if interpreted in terms of aggregate demand. It is all about changes in the composition of demand. As soon as one says "the money supply," the aggregate demand tribe reinterprets the meaning to fit into their own framework.

As many have noted, the genesis of the crisis is a key issue and much of the debate. Bill Woosley reinterprets the crisis in terms of monetary disequilibrium.

The crisis is brought on by real factors: projects quite literally run out of capital and cannot be completed. Abandoned, uncompleted homes in Las Vegas are a good metaphor. Workers cannot be employed at any price on those projects. (The price of raw land in Las Vegas actually fell to 0 at one point in some areas.)

In that sense, there is no symmetry between the expansion and the crisis. There is no crisis in the upturn as Pietro explained.


I think you hit on what is a central issue, the matter of aggregate analysis. If Krugman and Cowen have caught a sticky wicket here, it is in noting that ignoring aggregate analysis and only focusing on reallocation does not explain why employment does not fall in the consumption sector during the boom. Indeed, while of course Garrison does have both investment and consumption rising during the boom, this does require some extra effort to explain, And, indeed the bottom line remains how one explains an aggregate increase in unemployment from a bunch of micro reallocations.

Now, I do think there are ways around this, including some matters that have been mentioned, such as the asymmetry of booms and declines and so on, although some of this ultimately relies on looking at crashes through glasses that are at least partly colored by Keynesian or Minskyan tinges, as Arnold Kling notes is helpful in resolving these problems.


We agree on the centrality of the aggregation problem. That was a major theme of Leijonhufvud's reinterpretation of Keynes.

Let me just add that there's no reason in principle, as Pete's original post implies, that some Keynesian insights might not be a helpful addition to an "Austrian" story that comprehensively deals from boom to turning point through the full bust.

Some Rothbardians who comment here have tagged my views as "Keynesian." To them, this is a way of insulting me. To me, it's neutral. It might or might not be right, but it doesn't bother me to have that label applied. If the economic explanation is right, it's right no matter who originated it or what we call it.

It does seem like the aversion to Keynesian ideas in some Austrian quarters is simply a result of the belief that "Keynesian" must necessarily mean one supports government intervention, hence we can't have that! Of course it need not, as Pete's description of the actor's problem situation suggests. Markets can solve Keynesian problems. Or at least that's a proposition worth debating.

In any case, let's make sure we're using "Keynesian" to mean something more than "you think we need more government." So far, so good for the most part in this thread.

I agree with Pietro M. that there is an Austrian "bust" which is separate from the "secondary depression". Certainly there is a secondary depression, but it's not the whole story.

During the boom there is confidence in the future and expectations of high future earnings. In the bust this confidence is dashed. There is a rapid change in expectations. There is no such rapid change of expectations in the boom, it is gradual because of the gradual effects of capital change.

I think it is easy to understand this if we introduce the theoretical concept of the "anti-bust". The "anti-bust" is the true mirror image of the bust.

Consider a country that is about to be invaded by a hostile neighbour. The citizens all believe they will be imprisoned or killed very soon. So, they all go out and have a huge party. Everyone spends their money on consumption, capital goods become worthless. Then the plans of the potential invader go wrong. The citizens then realise that they have a future and their discount rates return to normal very quickly. In such a situation there would be widespread frictional unemployment because the consumption goods industries would quickly contract, though it would be a period of growth. This is the anti-bust.

I think that in the Austrian bust and the secondary depression frictional effects (such as searching, heterogeneity-of-skills and heterogeneity-of-capital) are necessary.

While most business cycles appear to be at least somehat asymmetric and almost all bubbles/crashes are, one exception to the latter is the Japanese housing market, which went up far more rapidly during the 1980s than the long and gradual glide down since it peaked in 1991, which glide has not really come to an end yet, despite appearing to briefly plateau a couple of years ago. What exactly was and still is going on in that market remains one of the larger mysteries out there.

Minsky addresses the good old question of the state of confidence, which is a topic of enduring interest to Post Keynesians. At least one Austrian has taken a serious interest in the state of confidence, namely, me. Although I did not use the term, it was the topic of my 2002 book on Big Players. There I make the argument that expectations will be more prescient when policy and money supply are stable in some sense and less prescient in the opposite case. I particularly examine the consequences of "Big Players" and blame them for encouraging herding and irrational waves of optimism and pessimism. Keynesian policies create a Keynesian economy. My argument does mix Keynesian and Austrian arguments. In particular I take Keynes's Chapter 12 very seriously.

If I am on to something, then we should pay more attention to the idea that the epistemic efficiency of the macroeconomy is endogenous to the policy regime. I think Keynes viewed "capitalism" as inherently epistemically flawed regardless of policy regime. Once you've got a stock market that matters, then irrational and fundamentally exogenous waves of optimism and pessimism can kick the economy around. I think many Austrians have tended to view "capitalism" as epistemically flawless except for the malinvestment or overinvestment (depending on who is writing) caused by "artificial" credit expansion. I think we should ask how alternative policies and institutions influence the epistemic efficiency of markets.


Good points, especially about the way in which policies and institutions influence expectations and the operation of markets. (I prefer avoiding "efficiency.")

I don't understand this at all. Hayek argued that there was no necessary link between changes in investment and unemployment rates. More importantly, neither Krugman nor Cowen seem to be cognizant of capital structure, and particularly time aspects. It's not reasonable to critique ABCT if you don't understand it, and I think they don't.

1. Someone on this thread is confusing RBCT and ABCT. RBCT isn't about the aggregate demand or supply. It is the changes in the market as a result of supply changes. ABCT is about mispricing of interest rates due to a monopoly on money issuance.

2. I don't accept Krugman's argument, but because a shift towards long term increases in production ability naturally causes a decrease in labor. Labor isn't something magical, it is just a production factor, and long term labor contracts are much like long term capital.

3. During the austrian boom both labor and capital demand increase because people are trying to increase long term production above market equilibrium.

Interesting thread. It seems that the ABCT still elicits a great deal of discussion.

But I don’t understand Krugman’s point. You cannot have unemployment in the boom faze because you have workers bided away from other (potential) jobs. ABCT starts with full employment after all. It is not only about businesses that lose workers, there are also businesses that never open or don’t extend because they cannot afford the prices and there are many of those. As a matter of fact it is exactly the latter type of businesses that will absorb the freed labor and other production factors after the bust.

And I don’t understand the need for symmetry. It seems to me just a fancy need for a graphical representation. The bust can be quick, as were many before the Great Depression and the one in the ’80, or very long, as the Great Depression, Japan, almost all the last fifty years in France (joking) … No need for symmetry or asymmetry or other catchy words. In the boom there is also productive work, something not denied by ABCT, the bust does not have to be equal with the boom, it does not have to last just as long and so on.

And again about this secondary depression thing. There is no such a thing as a secondary depression, it is only a nice term useful in expositions of the theory, an intellectual construction, but in reality you cannot determine when and where that “secondary depression” starts, or ends, or even what it is, not in real time any way and even in retrospect there will be many versions of it, depending on the persons that try to indentify it and their personal experience during the boom, and the bust.

I agree with Barkley Rosser in that overproduction is intuitively and theoretically a more complex object than the basic (and indefensible) Hayekian story of forced saving and reduced consumption.

As far as I know, what happens to the economy when there is overproduction has never been analyzed in details. Garrison talks of triangles pulled at both ends, but it's a metaphor, not a theory. He also talk of money illusion by part of workers, which can cause overproduction. Hayek and Machlup occasionally talk about undermaintenance of capital, increased exploitation of capital goods, shifts from long-term to short-term production at the top of the boom (Ricardo effect). However, to my knowledge this is a work in progress, not a full-blown theory.


The speed of recessions is quite relevant. If the recession is slow then why shouldn't the labour market adjust to it as it occurs? In Japan this may have been what has happened.

The Great Depression wasn't one long depression, it was a set of specific events starting with the Crash of '29. Later there were recessions caused by regime uncertainty and other exacerbating factors.

The secondary depression is not supposed to be necessarily "identifiable" through statistics. It refers to the bust caused by the rise in demand for money that is not satisfied. It's a theoretical differentiation.

Pietro M,

You bring up a lot of interesting points.

In Garrison's theory there is the "broken backed" Hayekian triangle. In that situation there is a lot of investment in long term capital because of the low rate of interest. Also, because of expectations of future wealth there is a lot of "derived demand" for consumption goods. So, the consumer goods industry is doing well. But, between these two there is a lack of capital. This isn't overproduction however.

Overproduction occurs when the PPF is pushed outside what would be its normal boundary. This is problematic, as many have pointed out. As you say, it's not really a completed theory.


I think that contractors quit building housing because they didn't expect to be able to sell them at prices that would cover their costs. Discovering this, they stopped construction and left unfinished homes idle.

I don't think that they had ready buyers but couldn't complete the homes because they were stripped of capital.

I am still confused about the nature of this "circulating capital" that is short? Is it money? Is it money in the hands of certain entrepreneurs? Is it the resources that those certain entrepreneurs want to buy? Is it labor? Intermediate goods? Nonspecific capital goods?

Assuming that the boom has resulted in a failure to maintain or replace capital goods, so that when the boom ends, output must fall because of less capital, the _obvious_ solution is to devote existing resources, including labor, in repairing these capital goods or replacing them with new ones.

In this case, there is a change in the sort of capital goods produced (presumably) and perhaps a decrease in the production of consumer goods.

But none of that matters.

The "problem" is that resources, including labor, are stripped away from those particular sectors that had over exptended.

Some parts of the economy shrink, and others expand.

I think most of you would do well to focus on the theory that that maintaining the malinvestments require every accelerating inflation. Suppose that isn't done. Suppose inflation is maintained, and so the malinvestments get liquidated. What happens then? Why are the malinvestments liquidated?

The sectors that were underproducing pull resources away from sectors that whose profitability would require the accelerating inflation.

P.S. Whatever it is that prevents real output staying on the PPV, (which usually is to the right so that most of these discussions involve how far to the right over time,) _is_ the problem.


The point Jerry makes about contractors literally running out of capital is compatible with the point you make about their concluding that they couldn't cover their costs by completing their projects, and so abandoned them. Your point is the flip side of his.
Builders were dependent on banks to supply a steady flow of credit so that they could pay their construction crews, buy more supplies, reinvest in tools, nails, etc. This was also true for other firms up and down their supply chains. When the banks demanded more collateral and higher loan rates, and then supplied fewer loans, the builders were sometimes left with insufficient capital to finish their buildings. Hence the layoffs and uncompleted homes and commercial buildings.

Btw, someone mentioned IS-LM earlier in the comments. Have any Austrians tackled this? Roger Garrison or Steve Horwitz (#28)? A search turns up some mainstream critisicm (Axel L. is among the critics), but have the Austrians pounded some nails in the Hicksian facade of the Keynesian edifice, to use a horrible building analogy?

Mr. Steele, I second your comment with the following...

It is very easy to beat up a complex theory if one does not understand it at all. It is an easier thing still to beat up a theory when one does understand it but decides to lie about it.


My point _was_ that the houses weren't completed because no one wanted to buy them, not because the contractors couldn't obtain enough credit to complete them. Even if you can find situations where contractors wanted to borrow but could find no lenders, that still doesn't mean that the problem was a lack of funds to lend to the contractors. Instead, it was foolhardy contractors willing to spend the lenders money to complete a house on the hope that things would turn around so they would be able to sell the house at a price that would allow them to pay the money back. The lenders are finally a bit more realistic and tell the contractors no, the demand for your product has collapsed. We are cutting you off.

I am claiming that the causation was something like--investors no longer willing to lend to investment banks by purchasing their commercial paper. Investment banks no long able to take over loans from mortgage originators because they cannot fund the mortgage backed securities. Mortgage originators not able to lend to potential homeowners because they cannot sell the loans. Potential homeowners not able to borrow to buy houses. The demand for houses fall. The price and quantity both fall. Construction firms produce less. And that includes stopping in the middle of the process of building some of the houses.

Your theory would be that contractors can't get credit and so they stop producing. There are shortages of homes. People want to buy them, but they aren't produced because the contractors couldn't get financing to produce them.

Perhaps recessions are caused by financial accelerators or decelerators. I am not sure I understand what that means.

But I am pretty sure that if there is a real problem, it must have to do with a shortage of the resources contractors (or whomever) need to use, not with the money they need to buy them.

In the end, costs are opportunity costs. The reason that it isn't profitable to continue building houses is that some of the resources--land, labor, and capital--needed for the houses can be used to produce things more valuable that the houses. The real demand for something else is pulling those resources away.

I believe it's helpful to abstract away from capital goods versus consumer goods, the heterogeneity of capital, etc. Consider the simple asset market experiments done by Vernon Smith and replicated by many others. Even though subjects are made aware of the fundamental value (FV) of the asset, when traded it typically exhibits a classic bubble and crash, i.e. it starts below FV then slowly rises above FV and finally crashes very quickly. The reason is that many traders pursue (rationally) a strategy of momentum trading. Additionally, there are those who pursue speculative strategies, i.e. they are betting on the presence of momentum traders. That is, not all traders are value traders. The resulting dynamic gives us asymmetric asset prices, and the very same dynamic is at work in business cycles generally.

I believe this is the essential truth that Keynes was getting at: there is herd behavior and momentum trading is a very real and often profitable strategy. And presumably this an evolved strategy, a left over vestige of our "animal spirits." That's the end for me and Keynes, the rest is gobbledygook, as Krugman would say.

Austrians hit upon the monetary influences, and rightly so. This can be seen in Vernon's experiments: bubbles are most extreme when the money supply is increasing.

The question then is which set of institutions mitigates these fluctuations (that is if we think that the fluctuations are inefficient, and I think we do). I think the primary institutional failure is central banking, and in that I am an Austrian. For my dissertation, I'm running a set of simulations and experiments to try and figure out how different banking regimes, including free banking, effect the asset bubbles we see in the lab.

Thanks, Jerry. I see what you mean, "efficiency" might not have been the best word to use in my comment.

To Bill Woolsey's questions (and others').

I dealt with the crisis and the scarcity of capital on pp. 109-11 of Economics as a Coordination Problem. For a fuller treatment, I direct folks there.

The Coordination Problem is among types of capital. Too much durable and long-lived capital is produced, with too few resources left for circulating capital (raw materials) and to pay wages of workers.

The coordination problem comes about because prices, particularly intertemporal prices, are distorted by monetary shocks.

The narrative is not about aggregate demand, or aggegate excess supply or demand. It is about the maldistribution of resources due to distorted prices.

"Austrian" capital theory is not peculiarly Austrian. Its roots are in classical political economy and British monetary theory. Hayek and Mises repeatedly emphasized that point.

It is an irony of intellectual history that the meaning of the classical tradition had been lost and distorted, especially at Cambridge in the 1930s. In Keynes, it was unrecognizable.

To go back to my Las Vegas example, Bill Woolsey is just wrong. Homebuilders don't stop until the money runs out. A year ago I witnessed homes in foreclosure and being auctioned on one side of the Blue Highway coming out of Vegas, while new ones were being framed across the road. Presumably the banker for the former builder was more stressed than the banker for the second builder.

Neither the banker nor the builder can recover their investments unless projects are completed. The accumulated investments raises the projected return on the last increments. Hayek dealt with this phenomenon in "Investment that Raises the Demand for Capital."

The last cycle has been made particularly complicated by two factors which ABCT normally neglects: complex finance and international trade.

Maybe it would make things clearer to shift the focus on the financial structure rather than on the production structure, and abstract from capital theory.

The results don't change, but while there are no operational tools to investigate the structure of production, for the former there is a lot of material.

#1: monetary policy is a cost-socializing technology which reduces the costs and risks of investment, causes bubbles and pushes for a type of overconsumption called "equity extraction": consumption out of unreal wealth, i.e., capital consumption (Machlup described the same process).

#2: moral hazard makes irresponsibility and recklessness privately rational, and the bubble economy is the result. This is the transmission mechanism in its most abstract form: cost socialization begets coordination problems in terms of bad incentives and false information.

#3: at the certain moment either (3a) the monetary drug is no longer available or insufficient, maybe because of inflationary fears (exogenous credit crunch), (3b) the financial structure is no longer capable of transmitting monetary stimuli because it is broke (endogenous credit crunch, which, as the famous rapper Freddie H sang "That credit crunch ain't a liquidity trap, just a broke banking system"), or (3c) the economic structure is no longer capable of sustaining overconsumption (real resource crunch, whose result is normally a scarcity of circulating capital, like raw materials, that in fact have skyrocketed the first 12 months of the recession, and are rising now that someone ventures to talk of a recovery).

In ABCT there is (3a) (the Fed takes away the punch bowl) and (3c) (the economy collapses because of capital structure problems), but there is no explicit analysis of (3b), i.e., systemic risk.

I think that a breakdown of the financial system due to an unsustainable financial structure, such as excessive leverage, monetary multipliers, reliance on foreign credit, reliance on liquidity, maturity mismatch (30y mortgages vs 3m commercial paper!) is a more apt description of the recent crisis than problems in the capital structure.

However, it is logically isomorphic to ABCT! I just used a financial theory language instead of a capital theory one.

However, malinvestment has been evident in at least three markets: automotive, finance and real estate. I would add imports because if US consumers saved some part of their incomes, they wouldn't need to import capitals from abroad, and thus they wouldn't need to import present goods at the present and past levels.

An unsustainable financial structure is not an aggregate demand problem: it's the legacy of a past moral hazard problem which causes a dearth of capital and a crisis. There is no path to a new equilibrium which does not pass through a credit crunch, the repayment of debts, massive deleveraging, exposure to maturity mismatch, and thus a reduction in investment levels.

Now that the Fed has succeded, at least temporarily, in stopping the adjustment process, has the US economy become structurally sound? No. No amount of aggregate demand stimulus can eliminate a structural problem, whereas it can add to the very cause of that problem, i.e., policy moral hazard.

Bill Stepp: I have a fairly extensive discussion of the problems with IS-LM as seen from an Austrian perspective in my *Microfoundations and Macroeconomics* book.

There is more in Hayek's macro than is imagined in the impoverished world of cartoon versions of "ABCT", designed to slander Hayek and others, rather than to advance understanding.

Hayek talks about changes in risk behavior across the business cycle, he talks about the interrelations of international money and finance with cycle phenomena, etc., etc.

Krugman and Cowen and others are out to slander Hayek on false grounds. They are after competent explications or the advance of understanding.

Pietro, Hayek touches on most of the themes you mention in his TCofL, in his MT&TTC, and his MN&IS.

Pietro, Hayek gets the financial side of the trade cycle, and it is at the heart of his work, both early and late, if with few example or little complexity of real world illustrations.

Greg Ransom:

I read Hayek's economics books and many others, except TCofL and TRtS and part ot the MC&F. I don't remember explicit references to these topics (except MN&IS and international monetary flows: my previous sentence on the neglect of international trade was wrong although I know of no full integration of ABCT and international economics up to now, but there's a lot of material I have to read on this).

As far as I remember:

1. Hayek analyzes the role of money in terms of M and occasionally says that what really matters is MV.
2. He sometimes mentions the role of higher order monetary aggregates (credit money) without much detailing, in the works in the '30s.
3. There is a genial reference to time inconsistency (40 years earlier than Kydland & Prescott!) in MN&IS when he says that a central bank which promises to print money ad libitum will destabilize the economy, but it is just half a page.
4. He sketches a theory of the non-neutrality of changes in the relative liquidity of heterogeneous monetary instruments in TPToC, which can be seen of the skeleton of a more general theory of money, but it is not developed (the second book of TPToC has never been written, so only 3 chapters exist regarding monetary theory).
5. In later monetary works, the underlying theory is more difficult to recognize, especially in TDoM in which price stabilization seems the goal of the monetary constitution. I must admit to be deeply puzzled by this.

There was a post on your blog claiming that Hayek talked about a financial leverage cycle like the present one, but I could not find any bibliographical reference to his writings regarding this.

PS I must admit that I have troubles reading Hayek because it is often difficult to clarify the problem he was analyzing and the premises from which he started. Mises was a much clearer thinker and writer, but Hayek tried to tackle so many fundamental issues that he must be the starting point for every further advancement. There are innumerable topics in which he gave some fundamental insight which often has been neglected in the subsequent literature (caveat: I'm not omniscent), but an insight is not a full-blown theory, and that's what I was referring to.

In contrast to what Austrians like Jerry O'Driscoll rightly emphasize, that is, the impact of changes in financial balance sheets on the price system and the resulting reallocations, the variables in the model behind the ABCT are highly aggregate (capital goods are homogenized as 'time' which is an aggregate measure due to Böhm's legacy still ruling the roost in modern Austrian circles) and the model is deterministic. The core 'prediction' that any money-induced traverse by necessity reverses can only be maintained by diminishing the role of redistributions of wealth and income during the traverse. Mises is explicit: to the extent that money reallocates endowments there is no need for mean reversion (since the mean drifts). Mises simply assumes that this effect is neglible. In fact, a more humble or Hayekian point would be to admit that we have little knowledge about the impact of money on relative prices. Money's impact is non-neutral AND NON-PREDICTABLE! I think the EMH of modern finance is much closer to this Hayekian perspective. The fact that Austrians are discussed more often together with Minsky and Keynes in the aftermath of crisis should worry Austrians, at least as long they believe that crises are exceptions to the rule that Smithian and Schumpeterian horses outpace government stupidity. Thus, whereas Austrians correctly focus on recalculations (which can be framed by standard neoclassical devices), I do not see the point in referring to the ABCT over and over again. There is also no reason to ban concepts like aggregate demand: after all, Wicksell and later Mises and Hayek had changes in the LEVEL AND structure of sprending in mind! The expansion of Wicksell's 'ideal bank' is a change in the consolidated balance-sheet is a change in the level of spending. As I understand Hayek the ABCT is about the impact of level changes in demand on the structure of production (the transmission mechanism characterized by Cantillon effects).

A related topic appeared in The Economist last week about how the recession has caused macro text writers to revise their books with a view to incorporating topics generally ignored in previous editions. Here's a link:

Among the topics to be included is "the idea of leverage and how it contributed to the crisis."

That's all well and good, but text book writers should drop the fiction, which I well remember reading in a principles text as an undergrad, that economics has nothing to do with accounting, and therefore students of economics can ignore it. But how can you teach the subject of leverage without having students learn at least some rudimentary accounting?

My guess is that the new texts are going to ignore this and introduce confusion instead of clarity.

The article mentions the "liquidity trap" and asset bubbles. Keynesians would say that the former never went away and shouldn't have been ignored. Austrians would say the same thing about bubbles. The debate continues....

I think that folks are making this all far too difficult for themselves....

People are delving very far into capital theory and the details of the structure of production. Now, that is all very good, and certainly more work should be done in this area. There are many points that aren't satisfactory.

However, for ABCT very detailed capital theory isn't really needed. The principle points are, simply, that cantillion effects exist and that the structure of production is distorted. It doesn't matter that much *how* the structure of production is distorted.

Yes, it may be that the interest rate/time aggregate system is inadequate. It probably is the case that future expectations of income cause consumption to grow rather than fall in the boom. There may be forced saving. It maybe that we should think of a boom as "pushing out" from the sustainable PPF rather than changing the shape of the Hayek triangle.

But, there is absolutely no chance that all that these complications cancel out and invalidate the ABC theory. That would be an astronomically unlikely coincidence.

But this claim is asserted in English words.

Where is your model? ;-)

"there is absolutely no chance that all that these complications cancel out and invalidate the ABC theory. That would be an astronomically unlikely coincidence."

Mr. McBride: In your comment you wrote that "I'm running a set of simulations and experiments to try and figure out how different banking regimes, including free banking, effect the asset bubbles we see in the lab." It might be worthwhile to consider an experiment where Money is defined as specie (the classic American Constitutional standard) and credit, including bank notes, as the promise to pay specie; and the market participants may demand either specie or bank notes in payment and the issuers of credit are free to determine what specie reserves they will hold against redemptions. The asymmetry in asset prices that Vernon Smith found exists only in an experimental world where money and credit are indistinguishable as methods of payment and stores of value. What remains remarkable about the brief historical experiment that the United States conducted under the Constitutional gold standard from the Grant through McKinley administrations is the consistency of the purchasing power of Money in the midst of investment booms and busts. It is an anomaly that both conservative and liberal Keynesians have found distressingly at variance with their theories.

What a load of tripe. Austrians are the only pure economists out there. There's no way that they're just self-hating twiddle-finks that some people suggest they are. It's all about the deductive logic my friend, not empiricism. That doesn't lie, ever.


This is being discussed at great length on Bill Woolsey's blog.

The comments to this entry are closed.