Arnold Kling  

Some Comments on Recalculation

PRINT
The Depression that Wasn't... Tyler Cowen's Speech at APEE...

A friend writes to say that Recalculation sounds like an interesting theory, but he finds it hard to explain to others. Below are some thoughts.

1. Markets are decentralized. However, decentralization is so non-intuitive that we often resort to fictional characters in our narrative. There is the "Walrasian auctioneer" who brings about market-clearing prices. Another example is in How the Economy Works, where on p. 28 Roger Farmer writes,


Pareto is known to economists for asking the question: Could an omniscient planner reallocate commodities...in a better way than the free market?...

The planner is a fictitious character who symbolizes perfect knowledge.

For Recalculation, the fictional character is an economic boss. The boss has just discovered a big mistake--people who supposedly were doing useful work in fact have been laboring in vain. The allocation of effort needs to be re-thought. While the boss is thinking, lots of people just stand around idle. Being unemployed during a Recalculation is like being on a work site where the boss has temporarily run out of things to tell you to do.

2. Conventional economics makes much use of the phrase, "In equilibrium, ..." We say that "we know that in equilibrium, X has to happen." We do not talk much about the process of getting from disequilibrium to equilibrium. We talk even less about the things that go on while we are out of equilibrium. What makes Recalculation different, and awkward, is that it is explicitly not an equilibrium concept. In that sense, it has an affinity with other heterodox economic views that scorn equilibrium. Austrians who focus on dynamics and creative destruction are in that camp. But so are post-Keynesians who insist that Keynesian economics is a disequilibrium theory.

3. One of the reasons that nobody likes to do disequilibrium theory is that "out of equilibrium, anything can happen." Disequilibrium in one market can spill over into another market. Rather than converging rapidly and smoothly to equilibrium, markets might move farther away from equilibrium or overshoot equilibrium. The equilibrium concept lets you rule things out. Any disequilibrium story is less precise.

4. Both Keynesian and Austrian theories of business cycles start with a particular type of disequilibrium, relating to the supply and demand for capital goods, viewed in the aggregate. In the Keynesian story, there is an excess desire to hoard for the future. The capitalists lack the "animal spirits" to invest to satisfy this demand to provide for the future, and instead that demand is channeled into sterile money balances. In the Austrian story, the economy suffers from an excess of capital goods, probably brought about by central bank distortion of interest rates, misleading investors into believing that the demand for future consumption was higher than it truly was. Now that reality has sunk in, capital goods producers need to cut back in order to work off the excess.

Recalculation does not require that capital goods per se are in excess supply. Instead, I think of particular types of human and physical capital losing value. Suppose that we had a sushi fad that went out of control, so that lots of entrepreneurs built sushi restaurants and trained to become sushi chefs. When it turns out that sushi is not quite so popular, other uses need to be found for the restaurants and the chefs.

5. Reallocation of resources is taking place all the time, even when the aggregate unemployment rate is low. Taking the preceding example, the movement of resources into the sushi industry was a reallocation, also. So the question about Recalculation as a story for recessions is what makes them so sharp and severe. Why is adjustment to a boom any harder than adjustment to a bust? I have given two answers.

a. One answer is speed--that a boom tends to take place slowly, while a bust takes place quickly. Imagine that in any industry in any point in time, the short-run supply curve is kinked. It is steeply sloped to the right, and flatter to the left. If demand goes up, in the short run it takes a while for new supply to enter the market. At first, you see a big price response, and only gradually do you get a quantity response. However, if demand falls, it is easier for the industry to cut back on output and employment than to reduce wages and prices. Yes, this is a Keynesian story, and yes, it deserves further explanation.

b. Another answer is availability of information. In a boom, you know where resources need to move to. You don't know where resources need to move from, but that is not a terribly hard problem for markets to solve. Nobody has to leave their old job until they have already found a new one. On the other hand, in a bust, you know where resources need to move from, but you do not know where resources need to move to. People get laid off without knowing where to find a new job.

6. I do not think of the recalculation story as describing either a demand shock or a supply shock. It is a story about adjustment. As such, I think we ought to be agnostic about the effects of fiscal and monetary policy. The recalculation story does not rule out the possibility that they could speed the process of adjustment. However, it does not offer the sort of reassuring precision that is provided by Keynesian multiplier analysis or faith in a stable velocity of money. Other people want to talk about how more government spending or more money growth changes equilibrium. Recalculationists have to re-cast the question, because we do not think in terms of equilibrium.


Comments and Sharing


CATEGORIES: Macroeconomics




COMMENTS (19 to date)
Boonton writes:

Arnold,

In previous posts we've been arguing about two post World War recessions, 1919 which was supposedly an example of good Austrian gov't policy and post WWII which Austrian sympathizers claim is an example but I think is really the opposite.

Anyway wouldn't this be a classic area to look for a recalculation recession? In WWII you had millions who were either sucked up by the armed forces or worked industries that had retooled to supply war goods to the gov't. Suddenly not only does the war end but it ends in a way that requires a lot fewer war goods than thought (the secret weapon of the atomic bomb made a D-Day style invasion of Japan unnecessary).

So here you have a lot of goods and capital everyone thought would be employed for a while more suddenly not needed. To me this should be the ideal place to look for a recalculation recession. While you do find a minor post WWII recession it hardly seems very dramatic. Which suggests either that:

1. Recalculation recessions are very minor disequilibrium events quickly destroyed by a market model that drives the economy towards equilibrium.

2. Recalculation recessions are ideally handled by Keynesian fiscal and monetary stimulus to keep the idle workers with sufficient demand to keep the economy stable until they find where they need to reallocate too. To use your image of a boss who has run out of things to tell his workers to do, his workers stay on the clock anyway!

The other question is since the essence of a recalculation recession is that everyone knows where the capital and labor has to *leave* but no one knows where it has to *go*, then how does that make the current recession (or the Great Depression for that matter) a recalculation recession? While unemployment in some areas like finance and homebuilding and mortgages is higher the fact remains the recession seems remarkably broad based when supposedly its all about over investment in one particular area. Yes I know there's spillover unemployment....if the mortgage company goes bust the bagal shop accross the street may go bust too....but I don't see a recession clearly based in one sector of overinvestment.

Pedro writes:

5b is very intuitive. Never thought about about it that way, and it sounds relatively easy to incorporate into a search model of unemployment. As far as the cause of the need for readjustment, why not let demand and supply shocks account for different episodes? Or do you have something else in mind?

wlu2009 writes:

Agree about 5b. This seems a pretty simple, elegant answer to the question Krugman has posed. I wonder what he'd have to say about it.

Pedro writes:

A search model would use an assumption of equilibrium always, but I think would retain all of the features you're talking about.

Combining this with some kind of monetary disequilibrium theory, like Austrian or Yeager, you'd have real shocks resulting from monetary policy (or policy's response to other shocks, a la Sumner), and a model of readjustment following a sub-prime crisis in which the central bank's response is itself sub-prime - thus causing the further complications that Sumner has been talking about.

Doc Merlin writes:

So recalculation is RBCT with the addition of transaction costs and asymmetric uncertainty?

This makes sense to me. I particularly liked point 5.

ThomasL writes:

When I was talking about this to someone yesterday, we brought up technical sophistication as well.

If I was a highly skilled X, it would take time to become a highly skilled Y, even if I knew what Y was and that I wanted to do Y--which I probably don't (cf. 5b).

That would make this something of a 5c.

Do I throw away my years of specialization for something new? I may have to, but I'll probably wait until I'm certain X isn't coming back, or I simply cannot wait for it any more. Over time my skills in other areas have atrophied or been left undeveloped in favor specialization in X, so a switch won't be easy to do.

This is kind of like Cowen's point that nominal declines become real declines the longer the recession goes on; but with the Austrian twist that the decline dates the start boom, not the bust.

The summary version is that I think the more sophisticated and specialized a workforce is, the worse a problem recalculation poses. Recalculation means throwing away the years of training and experience behind you, to start down a road that will take years of training and experience to arrive anywhere.

Sean A writes:

If I understand the theory correctly, the boom, e.g. the housing bubble, was due to a "fad." This doesn't seem to offer very much description to the formation of a bubble. In that sense, it seems pretty compatible with the "animal spirits," explanation. Correct me if I'm mistaken here.

Lord writes:

That information is the profitability of the remaining industries. What makes so much recalculation necessary at the time of the bust is the failing industries were the largest most profitable industries in the economy. It was the very concentrated narrowness of the economy being based on those few industries that leave the remaining ones unable to counter their loss directly leading to the recession. Most of the remaining ones were really only profitable off the funds generated by those now failing industries. Still, there are some industries distant enough from the failing ones and large and profitable enough that will eventually take over the lead.

The problem may not be the boom or bubble but too few bubbles, bubbles that don't have other bubbles to compete with them, allowing them to grow too large, too dominate, with no others to take their place once they bust.

Lord writes:

One of the problems with recalculation is profitability is a function of interest rates and though some industries were unprofitable at existing rates (or at any rates given the level of fraud involved), doesn't mean they aren't profitable at lower rates, or indeed, still aren't the most profitable available. Then there is little need for reallocation of resources, only for recalculation by the Fed for rates to reach their natural level. Certainly the losses and fear of losses is real, housing prices will slide further in real if not nominal terms, equity extraction is unavailable, and inventory levels are elevated, yet rates may fall enough to return it to growth without fraud, and if not to the leading industries, still sizable contributors. Recalculation may be more of wealth than income.

Russell writes:

Is there a role for behavioural economics in the recalculation story? Here, I'm thinking about loss/risk aversion helping to explain why booms/busts move asymmetrically.

In boom times, loss aversion operates to dampen (relatively) optimism, which stretches out the boom time. But when the bust comes loss aversion accelerates the process as folks hurry out of the relevant market to avoid losses.

fundamentalist writes:

"Suppose that we had a sushi fad that went out of control..."

Sushi restaurants are examples of excess capital goods because the building and equipment are capital goods. ABCT says the business cycle take place primarily, but not exclusively in capital goods.

And you have to ask what makes the sushi fad possible? Interest rates should act as a break on such fads. If they don't, it's because the central bank is flooding the country with money. The bank may not be the cause of the boom, but it is a necessary enabler. Modern Austrians place too much emphasis on the central bank because it is the main cause of artificially low interest rates. But as Hayek noted in "Monetary Theory and Trade Cycles" booms can start without a lowering of interest rates, but they can't continue without it.

As for unemployment, Hayek pointed out that almost all employment today requires capital. In the boom, new capital is being created so more people can work. But much of that capital is being wasted, as in the sushi fad. When that waste becomes visible in the bust, capital is lost, or as Kling says, loses its value. Less capital means that fewer people can be employed.

As a simple example, imagine a ditch digging company that has five shovels and five employees using them. One day a truck runs over 3 of the shovels and destroys them. Let's assume that the shovels are very expensive and can't be immediately replaced. The ditch digging company would have to lay off 3 workers until it could save enough to buy more shovels. In the real economy, capital is wasted in the boom. As a result, workers who depended on that capital lose their jobs. They will have to wait until savings rise enough to create new capital in other industries that will make their employment possible.

Boonton: "In WWII you had millions who were either sucked up by the armed forces or worked industries that had retooled to supply war goods to the gov't. Suddenly not only does the war end but it ends in a way that requires a lot fewer war goods than thought..."

Yes, the end of WWII brought about huge reallocations of resources. But keep in mind what Austrians consider the goal of production to be: it is consumption. If resources are misallocated, that misallocation is in relation to what consumers want now and in the future. There is no abstract measure of allocation; only consumer wants applies. In WWII, the allocation of resources to the war effort was the misallocation. Consumers had huge pent up demand. The reallocation that took place after the war, and for which most businessmen planned during the war, was correcting the misallocation that took effect during the war. The "depression" from misallocation took place during the war, but was hidden due to price fixing and to the fact that gdp includes government spending. The depression actually happened in the private sector where people became poorer. Unemployment was hidden because all of the unemployed were drafted and many killed in the war. The reallocation after the war went smoothly because 1) it was planned for during the war so the change wasn't completely unanticipated and 2) it correctly predicted the pent up demand for consumer goods.

Russell: "Is there a role for behavioural economics in the recalculation story? Here, I'm thinking about loss/risk aversion helping to explain why booms/busts move asymmetrically..."

I think there is room for behavioral econ in the ABCT. There are many theories of the business cycle and they're all true; they are true at different times during the cycle. The ABCT is the grand theory; the other theories explain peculiar moments in the cycle.

Boonton writes:

fundamentalist

And you have to ask what makes the sushi fad possible? Interest rates should act as a break on such fads. If they don't, it's because the central bank is flooding the country with money. The bank may not be the cause of the boom, but it is a necessary enabler.

Why is it then that interest rates when the bubble was happening in real estate were higher than they are now? In bubbles like the dot-coms or real estate returns were in excess sometimes of 20%-50% per year! Tell me how interest rates are supposed to be a break on that unless you're advocating the central bank force the economy into a major recession!

As for unemployment, Hayek pointed out that almost all employment today requires capital. In the boom, new capital is being created so more people can work. But much of that capital is being wasted, as in the sushi fad. When that waste becomes visible in the bust, capital is lost, or as Kling says, loses its value. Less capital means that fewer people can be employed.

Interestingly war allows us to test this theory from a slightly different angle. Suppose the sushi stores were not a fad but represented a real, long term change in our eating preferences. But a dastardly terrorist organization bombs 40% of the sushi places in one day! That would be equiliviant of the market suddenly waking up one day and discovering 40% of the sushi places were total wastes of money. Yet employment after such tragedies doesn't fall, it rises. With less capital before, the return on putting people to work building capital goods increases which results in more employment. Yes total output falls, which is why you don't boost an economy by war..

In WWII, the allocation of resources to the war effort was the misallocation. Consumers had huge pent up demand. The reallocation that took place after the war, and for which most businessmen planned during the war, was correcting the misallocation that took effect during the war.

I think this veers dangerously towards the mythical perfect knowledge the rational expectations schools are justly mocked for advocating.

First: The war ended sooner than expected. The secret weapon of the atomic bomb was a surprise. A very expensive and bloody D-Day style invation of Japan was averted.

Second: I don't doubt that far sighted businessmen planned civilian work before the end of the war but I think you're making too much of plannings ability. The US was in a major depression for nearly a decade before the war. Why wouldn't businessmen assume post-war demand would be along the lines of pre-war? There was real fear that actual fighting would take place on American soil. While the end of WWII might seem obvious to us today, I don't think it was so clear in 1942 or 43. Maybe once the tide started turning, but still I don't think the lack of a major Austiran depression caused by the gov't reallocation of resources can be explained by hyper-intelligent businessmen 'planning' the economy so it could outwit Hayek's theory!

The reallocation after the war went smoothly because 1) it was planned for during the war so the change wasn't completely unanticipated and 2) it correctly predicted the pent up demand for consumer goods.

If #2 is true then the economy wasn't poorer during the war. Clearly increased earnings are economic growth...even if your earnings take the form of deferred compensation. How could consumer demand be increased without real growth?

fundamentalist writes:

boonton: "Why is it then that interest rates when the bubble was happening in real estate were higher than they are now?"

Absolute interest rates don't matter; what matters is the rates relative to the point in the cycle. Interest rates are always high at the end of the boom as demand for loans accelerate. Rates are low in a depression because demand for loans are low and because the feds are trying to stimulte the next boom.

And there is a lag between changes to interest rates and effect on the economy. Low interest rates stimulate investment in capital goods, but not immediately. It can take years to have an effect. And in the same way, high interest rates don't immediately shut down borrowing and investing; again, it can take years. The main problem with stimulation via low interest rates is that when the damage appears, it's too late to do anything because of the lags.

"The war ended sooner than expected."

By a few months. But manufacturers of weapons knew for a long time that the end was coming and had begun making plans to switch to consumer goods.

"Why wouldn't businessmen assume post-war demand would be along the lines of pre-war?"

They very well may have assumed that. But even if they did, what were they supposed to do, not try to compete for it? Should they have decided that there won't be enough demand so they should just shut down there businesses? No. They would try to compete and become the firms that survived.

"While the end of WWII might seem obvious to us today, I don't think it was so clear in 1942 or 43."

Probably, but what about 1944 and 1945?

"How could consumer demand be increased without real growth?"

It didn't increase. The war artificially suppressed it.

Boonton writes:

Probably, but what about 1944 and 1945?

So let's see, the gov't basically was running almost the entire industrial economy but 24 months is more than enough planning time to seemlessly transition out of that with hardly any unemployment.

But Herbert Hoover asks Henry Ford (asks, no law passed!) not to cut wages and we get 25% unemployment and the Great Depression.

It didn't increase. The war artificially suppressed it.

I pointed this out on the Depression that wasn't thread. Austirans start to sound like Marxists when it comes to gov't production. Marxists would sometimes try to assert that the better living conditions in capitalist societies were illusions. People were buying rock'n'roll and fish tailed cars not because they wanted those things but because capitalists fooled them into wanting those things with clever advertising and control of the media! Austirans, it seems, would have us believe that right after Pearl Harbor Americans didn't want bombs dropping on Japan and didn't want to see Hitler's stadiums turned into powder. They wanted steak dinners and toy bb guns!

Production is production and during WWII there was a massive amount of production and employment. Yes war aritficially suppressed consumer demand but nonetheless consumers earned rights to consume (imagine a person who wasn't drafted and didn't work or save during the war, he would have no 'war bonds' to cash in to buy stuff in 1949). Trying to depict WWII as a period of some type of Keynesian Depression that's hidden from all statistics is about as sane as a Marxist describing the Western World as some type of depressed, seething revolution waiting to happen with a Matrix-like illusion of prosperity on top.

Boonton writes:

And in regard to interest rates, we still have the same issue. Returns on a 'bubble good' are huge. Returns in the rest of the normal economy are modest. A 5% interest rate (let's say) is fine for acting as a break for most parts of the economy. In the dot-com boom, though, 5% was nothing compared to the 100%+ returns people were making. But trying to pop the bubble with accross the board high interest rates is putting the breaks on everything.

liberty writes:

"I do not think of the recalculation story as describing either a demand shock or a supply shock. It is a story about adjustment. As such, I think we ought to be agnostic about the effects of fiscal and monetary policy. "

I was with you until here. I do not think we can be agnostic about monetary policy because artificially lowering interest rates is going to drive investment into longer term investment--yes it may speed reallocation, but probably not toward the right things, since we were likely already over-long-term-invested since that was what fueled the boom that preceded the bust.

I also do not think we can be agnostic about fiscal stimulus since the government cannot know what the right investments should be for the economy--how could it?--and if government cannot know how to spend stimulus money well, and if it will tend to crowd out the reallocation that the private sector would help to do (the boss figuring out what to tell the workers to do) once prices had fallen enough, then this will be a mistake as well. Fiscal policy will tend to slow the market's reallocation, and soak up more of the economy into the public sector, which will tend to be inefficient.

Boonton writes:

liberty,

I think your post is the gist of the beef Krugman and mainstream economists have with Austirans. You let your model deviate from the classical assumption of rational allocation, full employment on the way up but you enforce that on the way down. You let the economy blow irrational bubbles but in the crash you assume all is working just fine to efficiently correct errors.

Bubbles, though, fight against gravity but crashes are going with gravity. IMO the collapse is just as likely to take down capital that was not irrational nor wasteful.

To put it in recalculation terms: Let's say right now the economy has realized that some home construction workers, finance people, real estate agents and associated capital was misallocated and is trying to think of where that should go. While we wait, though, there's a lot of jobs that are going to be lost just because everything is 'on hold'.....jobs that would be maintained if the recalculation was instantanous. Stimulus is fine to the degree that it maintains that level of employment while allowing the unemployed the time to see where the economy takes them. Your fear is applicable if the gov't overstimulates or tries to target stimulation towards sectors that need a healthy dose of 'recalculation'. Of the stimulus done so far only a few targetted programs might fall into that category (I'm thinking cash for clunkers, the $8K home buyers program, some mortgage assistence programs and such). No one, though, thinks they are significant enough to do anything but slow the fall down a bit. In other words, real estate is still pretty lousy even though the $8K helps a bit.

" It is steeply sloped to the right, and flatter to the left. If demand goes up, in the short run it takes a while for new supply to enter the market. At first, you see a big price response, and only gradually do you get a quantity response. However, if demand falls, it is easier for the industry to cut back on output and employment than to reduce wages and prices. Yes, this is a Keynesian story, and yes, it deserves further explanation."

You're describing what physicists call latency, or phase lag, in demand response. There is a reasonable argument to be made that market prices perform a function of synchronizing the phase of production and consumption cycles (whose frequency changes slowly on the time scale of each cycle).

Tom Grey writes:

No specific Industrial market is ever "in equilibrium" -- meaning no changes are planned in the next time period (month? quarter? year? week? day?); nor even that no unplanned changes will be made in the relevant "in equilibrium" period.

However, in a non-recession, the industries going down, are roughly balanced by the industries going up. And even within a stable industry, like retail, while Wal-mart adds stores and revenue, and mom & pop stores close, most firms are in non-equilibrium.

The housing bubble pop means almost all industries are shedding workers, now, and there's no big industry / service to absorb (enough of) them; health care is going up, along with gov't.

The easy money led to a bubble in asset prices, which has disappeared -- so much so that the economy can absorb huge amounts of new money without much additional inflation, at least until house prices are some 90% or more of their peak (or in some 80-90% of the state/ metro markets).

Gov't spending stimulus is certain to be less effective than tax cut stimulus for the same additional deficit -- it takes longer to enter the economy, it has higher admin costs, it's an attempt to choose 'winners' which gov't is lousy at doing.

The Bush 2008 tax rebates would have reduced the problem more, if they had been big enough. Notice how when gov't spending isn't "enough" stimulus, there is a call for more spending -- but not so much a call for more tax cuts, that weren't enough to fill the house bubble pop.

Austrians need to have active gov't programs, to reduce gov't, ready to support when times are bad.

Comments for this entry have been closed
Return to top