Arnold Kling  

Responses to Two Critics

PRINT
Sumner on Why Friedman Was Too... Basic Economics Is Intuitive...

I have been making the following claims about macro.

1. Major surges in unemployment are due to Recalculations. A sudden, large shift in demand across industries can occur in such a way that employment falls sharply in the receding sector(s). Meanwhile, it takes a long time for potentially expanding sectors (a) to realize that they are going to be the winners and (b) to accumulate the human and physical capital to expand. It is as if a central planner has decided to shut down some firms that no longer are producing valuable output but has not yet figured out where to redeploy the resources or how to do the necessary retraining and relocation.

I should emphasize that for economy-wide adjustment one of the most subtle and difficult shifts in demand is a shift in the form in which people wish to save for the future. If we go from wanting to accumulate capital in the form of houses to wanting to accumulate capital in other forms, that is a very difficult Recalculation for the economy to carry out.

2. Inflation rates are very sticky. A major change in the trend inflation rate requires a regime change, which changes people's expectations. That process of changing expectations takes years.

Under (1) and (2), money does not matter. In the short run, the economy is going its own way, regardless of monetary policy. Higher M leads to lower V, and vice-versa. In the long run, a significant change in the rate of money creation causes a similar change in the rate of inflation. However, the lag is long and the effect on the rate of Recalculation is small and of indeterminate sign (I can think of reasons why it could be helpful or harmful).

Point (1) can be traced to Clower and Leijonhufvud. Point (2) is not far from Fischer Black, who also saw general price behavior as determined by convention and who saw financial markets taking steps to offset changes in the central bank balance sheet. In the sense that I see monetary growth having its effects with long lags, rather than with short lags or leads, I am harkening back to Milton Friedman and rejecting what Paul Krugman calls the "dark age" macro of the past thirty years.

Nick Rowe attacks point (1).


Mackerel can't stop the inexorable logic of Say's Law. If you have an excess demand for mackerel, and so does everyone else, you are stuck. Unless you are prepared to go out on a fishing boat, the only way to get more mackerel is to buy it, and you can't, because you are on the long side of the mackerel market.

Money can stop the inexorable logic of Say's Law. If you have an excess demand for money, and so does everyone else, you can't get more money by selling more goods. But you can get more money by buying less goods. Of course, in aggregate these attempts will fail, but that doesn't stop individuals buying less goods, just like confessions in Prisoner's Dilemma.

Thanks to Mark Thoma for the pointer.

Rowe, like many economists, has a very hard time imagining anything other than a frictionless economy with instant movement to general equilibrium. In that (mythical) economy, a shift in demand between industries cannot possible lead to unemployment. In order to explain unemployment, these economists posit a shift in demand toward a non-produced good, and they leap on money as an example.

I am explicitly attacking this tradition, which I associate with Keynes, of making a big deal of money as a store of value. In this view, as long as the supply and demand for money are in balance, there can be no unemployment. In my view, the supply and demand for money could be in balance, but if the supply and demand for different forms of capital (houses vs. small businesses) is out of whack, there can be unemployment.

What is known in the macro literature as the "real business cycle" model also assumes frictionless markets. That is why I want to differentiate Recalculation from standard RBC. However, Recalculation is not a monetary story, and in that sense it is a "real" story.

On to a critic of (2). Bryan quotes Scott Sumner.


The most expansionary monetary shock in U.S. history, by far, was the 1933 dollar devaluation and the decision to leave the gold standard. During the first four months of this policy, the WPI rose by 14 percent and industrial production soared 57 percent, regaining half the ground lost in the previous 3 ½ years...Other easily-identified monetary shocks, such as the 17 percent decline in the U.S. monetary base between late 1920 and late 1921 had an immediate and severe impact on both prices and output.

Beware of proof by selective example. Some thoughts:

1. The monetary regime in 1920-21 was different than today's regime. It could be that relative to the gold standard, prices had risen way too much in 1919, and everybody knew it. That would have made it easy to bring prices back into line.

2. As to the 1933 episode, what was the long-term impact on general wages and prices? In the short run, the wholesale price index (WPI) can be dominated by commodity prices, which are volatile. Today, in order to gauge the trend of inflation, economists use broader price indexes, and they remove changes in food and energy prices in order to focus on "core inflation." I wonder how "core inflation" behaved during the episode in question.

3. I can do "proof by example" going in the other direction. Consider how long it took for inflationary expectations to rise from the early 1960's to the late 1970's. Consider how long it took for inflationary expectations to fall from 1980 to 2000, even though the "regime change" under Paul Volcker was sharp and highly publicized.


Comments and Sharing





COMMENTS (17 to date)
Taras Smereka writes:

I partially disagree with you on point 1. Consumer preferences change fluidly over time, and normally markets should be able to clear fast enough for that to occur without massive changes in unemployment. There is generally a catalyst that introduces staticism into the economy, preventing the market from clearing. Things like regulation, wage and price controls, and monetary policy.

Mitch Oliver writes:

Consumer preferences change fluidly over time, and normally markets should be able to clear fast enough for that to occur without massive changes in unemployment.

I think the key to the Recalculation is that when we see sudden, large unemployment it is because consumer preferences changed drastically and suddenly. Markets were unable to keep up with the changes, and the newly demanded good is not yet clear.

ThomasL writes:

Taras,

"Normally" that may be the case, but Arnold is focusing on business cycles, and those are not normal because the downside shift happens so rapidly. I'll use a microcosm of boom and bust, Cash for Clunkers, as an example. Consumer preference was stimulated tremendously, and then the stimulus withdrawn just as quickly. The USG claimed the effort was going to create 60k jobs. If that were true, what are those 60k people doing now that auto sales figures are far, far lower? Did the market adjust smoothly? Personally I doubt they were ever hired and the USG lied, but if they were, bursting bubbles aren't smooth, it is almost impossible to believe those 60k new workers are needed at this level of demand, and it takes time to retrain auto workers to do something else.

It also takes time for it to be come apparent what that /should/ do instead. We had a boom in housing and finance, we had a six week micro boom in cars, but what is next? Nobody knows, and until they know what the next big thing is, and realize /it/ is in fact /them/, they don't know whether they even want to hire all the displaced construction, financial, and auto workers; and the workers don't know if they are suited for the, as yet undiscovered, new jobs.

Ironman writes:

Picking up Arnold's thoughts on selected examples....

1. The monetary regime in 1920-21 was indeed very different from today. The high inflation that preceded this period was caused by the Fed's decision to monetize gold that had been shipped to the U.S. by Europeans for safekeeping during World War I. After the war was over, they wanted it back....

2. Core inflation data for the period would certainly be nice. I'm only aware of data going back to 1957 - it would make for an interesting project for an enterprising grad student to take it back further.

3. Let's not forget that there is very likely a demographic component to inflation, which would explain much of what we've observed in general inflation trends in the U.S. from the 1960s through today. Keep in mind too, that demographic factor today could well be a primary trigger for Arnold's economic recalculation hypothesis.

Don Lloyd writes:

Arnold,

"I am explicitly attacking this tradition, which I associate with Keynes, of making a big deal of money as a store of value. In this view, as long as the supply and demand for money are in balance, there can be no unemployment. In my view, the supply and demand for money could be in balance, but if the supply and demand for different forms of capital (houses vs. small businesses) is out of whack, there can be unemployment."

The only true demand for money (for holding) is for its use as a medium of exchange. While everyone has some money, and maybe a lot, that is serving as a store of value, this is only because of a lack of preferable alternatives, and not a demand for money. If a MMF yielding 10% with no transaction costs, and completely liquid, convertible to money at the click of a mouse, were available, then why would anyone hold money as a store of value any longer than necessary?

Regards, Don


Dan writes:

Arnold,

This is an interesting debate. I suspect you won't really begin to change economists' minds until you have some kind of rudimentary model that can be tested with numbers.

Felix writes:

As an outsider to the economics field, I am, as usual, confused. Arnold's points seem to be in the realm of "sky, blue - pope, Catholic".

Calling an economy an adaptive system doesn't seem like a stretch. That adaptations take time doesn't seem altogether impossible. That it takes time for people to switch from being home-loan sellers to being ... what?!? ... shouldn't be a surprise.

Dan suggested putting numbers to a model. Sounds about right, though we do have quite a few numbers from the real deal already.

It seems to me that both of Arnold's points are riffs on the same thing. Call it a "great Recalculation", but it's just "great" in the sense of being bigger than "normal". And the key word in point 2 is "very" - as in "very sticky". What's "very"? Relative to what other stickiness?

In each case, is there a way to measure the stickiness/calculation speed? If stickiness is bad, is there a way to minimize it? Should it be minimized? After all, delays in action can be quite handy to have around if the action is a bad thing. Consider how a bit of stickiness last fall could have avoided some silly actions. :)

If Arnold's overall point is that you can't help an economic system solve a mackerel supply/demand calculation by diddling with money, then why is anyone contending otherwise?

To Don's point about money as a store of value: I'm trying to figure that out, too. If money were instantly convertible at zero cost to other things, then, yeah, you could store value in other things just as well as in money. But why would you? Both stores are equivalent. And since there will never be instant, zero-cost conversion, does it make sense to try to get the speed and cost of conversion down as close to zero as is possible? At what expense? :) Maybe there's a sort of Platonic-ideal speed and cost of money/thing conversion. Maybe the best thing is to get the real speed and cost of conversion as close as possible to that Platonic ideal. Maybe an economic system is constantly trying to calculate and approach that ideal.

andy writes:

Money can stop the inexorable logic of Say's Law. If you have an excess demand for money, and so does everyone else, you can't get more money by selling more goods. But you can get more money by buying less goods. Of course, in aggregate these attempts will fail, but that doesn't stop individuals buying less goods, just like confessions in Prisoner's Dilemma.

That's not true for money - the mackerel analogy is incorrect. If everyone wants more money, the price of money will rise (deflation occcurs), which directly causes every holder of money to be more wealthy - which makes everyone to have "more money". No problem. Sorry. It does NOT stop logic of Say's law.

Prices must change and they are somwhat sticky. Which is probably what causes unemployment.

fundamentalist writes:

"During the first four months of this policy, the WPI rose by 14 percent and industrial production soared 57 percent..."

Four months is not nearly long enough for monetary policy to have that kind of effect. Lag times are much longer, in the years. Something else must have caused the WPI increases. I would bet on the fact that prices had fallen enough in the previous year that people started buying again and businesses began investing.

Arnold is right. If you want to be selective enough, anyone can find a historical example to match his ideas no matter how nutty. Finding one example that appears to corroborate your theory doesn't take much intelligence or cleverness and it certainly doesn't prove anything.

Don Lloyd writes:

Felix,

"To Don's point about money as a store of value: I'm trying to figure that out, too. If money were instantly convertible at zero cost to other things, then, yeah, you could store value in other things just as well as in money. But why would you? Both stores are equivalent...."

You've just claimed that there is no point in making an interest bearing investment while money is depreciating due to price inflation, for example.

A store of value makes no claim about the value being constant. In fact, the exchange value of every store of value, including money, changes every second.

Regards, Don

winterspeak writes:

ARNOLD: Suppose the private sector wants to perform the difficult recalculation of going from saving less of its dollar earnings to more of its dollar earnings. How can it go about that? This is SUCH a complicated question! Where can it get the money to save from? There are *so* many sources of dollar bills in today's world.

Suppose the Govt massively increased taxes and began running surpluses. What effect would this have on inflation? Hint: inflation comes from too many dollars chasing too few goods.

ANDY: People's debts are nominally denominated. If they save more, and get deflation, their real debt burden rises. This is the opposite of what they are trying to do.

Todd writes:

Winterspeak, are you saying that by paying down their debt, people increase their debt burden?

Bob Murphy writes:

Arnold wrote:

3. I can do "proof by example" going in the other direction.

But wait a second. You make the sweeping claim that the Fed can't influence nominal GDP in the short run, because people take years to change their expectations and hence offset any increase in M by reducing V.

All Scott Sumner has to do is come up with one counterexample to show that you are wrong. If you agree with any of his examples (which you are disputing, I grant you), then it doesn't matter how many examples of your claim you make back at him.

Greg Ransom writes:

Felix, in economics if you can't turn the sky or the pope into a
fancy math construct, the sky and the Pope don't exist, and you won't publish or get tenure if you produce work premised on the fact that they do exist. E.g. most economists produce work that insists that no more than a single capital good exists ..

Felix writes:

As an outsider to the economics field, I am, as usual, confused. Arnold's points seem to be in the realm of "sky, blue - pope, Catholic".

Noah writes:

Isn't the "Recalculation" model basically the same as Mankiw's "Sticky Information" model (imperfect competition, inattentive agents, sticky inflation)?

And doesn't money end up mattering a great deal in that model?

Greg writes:

It seems to me the unidentified variable in all this talk is credit.

If all transactions were in cash all the time, prices could fall faster and unemployment would be less. But since everyone has debts they have to pay any fall in prices leads to increases in debts.

Both the 1930 and current depression were related to debt levels. I'm not sure you can have true deflation without debt. I say true deflation because fall in prices is not necessarily deflation. A huge contraction of the money supply seems only to be possible when a large portion of that money supply is debt.

winterspeak writes:

TODD: I'm saying that the private sector cannot increase its net savings unless another sector decreases its net savings.

If the private sector tries, and is not supported by another sector decreasing it's net savings, and you get deflation, then yes, the real burdens private sector will get higher.

This is all true as a matter of accounting.

The Government can and does absolutely influence the quantity of net private sector savings, and can do this at any instant, by pretty much any amount. It just needs to do this through fiscal policy, it cannot do it through monetary policy (as Sumner seems to think).

Comments for this entry have been closed
Return to top