Tuesday, June 14, 2011

Kling's PSST

Arnold Kling continues to contrast PSST with AD-AS. In his view, AD-AS is mistaken and the reason for high unemployment is PSST--Patterns of Sustainable Specialization and Trade.

In my view, Kling's PSST is a factor that determines potential income, the productive capacity of the economy. In the absence of monetary disequilibrium, that is what determines the actual level of output. If prices and wages were perfectly flexible, then the real quantity of money would always be equal to the demand to hold real money balances and there would be no monetary disequilibrium. At the same time, real expenditures on output would be equal to the productive capacity of the economy. Firms would sell what they could produce, and real output would equal potential income.

While I have a relatively optimistic vision of the market process, so that in the absence of monetary disequilibrium, entrepreneurs are introducing better patterns of sustainable specialization and trade, raising potential income, I think error is an inevitable part of the process.

For example, suppose there is a complicated pattern of production associated with domestic car production, and it turns out that a better way to get cars is to build office buildings, sell them to foreign investors, and purchase foreign cars. Shifting from domestic car production to constructing office buildings is difficult.

But I see this as normal. It is called creative destruction. Some parts of the economy grow (constructing office buildings for foreign investors) and other parts shrink (domestic car production.) Old jobs are destroyed and new ones are created. This process of shifting the pattern of employment often results in structural unemployment.

Kling is a student of Solow, and I have read Solow dismissing this sort of thing. I didn't read him as saying it didn't happen at all, just playing it down as a source of unemployment. If one assumes that the process of creative destruction and the associated structural unemployment (including technological unemployment) is continual and constant, then perhaps it is sensible to claim that potential income remains on a three percent growth path more or less all the time, and that the natural unemployment rate stays at five percent (or something like that.) Any fluctuations in the growth of real output and in the unemployment rate around those levels would be due to fluctuations in aggregate demand. Certainly, it is possible.

But I have never taken such an extreme view. It has always seemed to me that the process of creative destruction should be expected to be irregular so that potential output will grow more slowly and then more rapidly and that structural unemployment will be sometimes higher and other times lower. And while Solow might be a bit dismissive of such factors, my impression is that it is now the dominant approach in mainstream macro. The "problem" isn't that output deviates from a constant growth rate or that unemployment deviates from a stable target level, it is rather that output and unemployment deviate from what they would be with perfectly flexible prices. Since there is no monetary disequilibrium with perfectly flexible prices, it seems to me that mainstream macro made some progress.

Of course, if prices and wages are not perfectly flexible, then an increase in the demand to hold money, not matched by an increase in the nominal quantity of money, results in a different problem than PSST. The patterns of specialization and trade maybe be there, but firms cannot complete those trades because they lack buyers, and they lack buyers because the real quantity of money is too low relative to the demand to hold money.

The fundamental proposition of monetary theory is that an individual can always adjust money holdings to the amount demanded by reducing expenditures out of income, but the economy as a whole can only adjust money holdings to the amount demanded by changes in interest rates, real output, the price level (including wages) or else a change in the nominal quantity of money. Changes in the real quantity of money, either by changes in the nominal quantity or else changes in prices and wages, is what is needed for real output to remain at capacity.

While the imbalance between the demand to hold money and the quantity of money is the essential problem, simple aggregate demand and aggregate supply analysis tells the story as well. Money expenditures on output change, and unless the price level (and wages) adjust enough, real output shifts away from potential.

Nothing in the AD-AS "paradigm" requires that potential income remain constant. It is just that shifts in nominal expenditure can cause deviations of real output away from potential. From a deeper perspective, this occurs because the changes in real output are one means by which the real demand to hold money adjusts to the real quantity of money.

Kling's macroeconomics fails due to an inability to grasp monetary disequilibrium. His arguments amount to a set of criticisms against what seem to me to be straw men. AD-AS is wrong because potential income isn't constant. Monetary disequilibrium is wrong because velocity isn't constant. And so on. But AD-AS and monetary disequilibrium don't require that potential income or velocity be constant.

Not only is Kling's macroeconomics wrongheaded, his microeconomics is problematic. I gave a simple example of an error in coordination. The pattern of specialization and trade involving domestic car production was over extended because there was a better alternative involving the production of office buildings to exchange for foreign cars. The result was losses and unemployment in the domestic auto industry, and profits and expanded production and employment opportunities in construction of office buildings.

In Kling's micro, the problem is supposedly a decrease in the demand for one thing and an increase in the demand for nothing. For example, there seemed to be a sustainable pattern of specialization and trade had involved the production and purchase of many new single family homes. Now people don't want to buy as many new family homes, and no entrepreneur has discovered what they do want to buy now. And so, we must wait until entrepreneurs discover what people do want to buy and start to produce it.

This story is inconsistent with nonsatiation--more is better. The usual microeconomic story is that people don't want additional houses because they want to buy something else instead. Perhaps they want new houses less, and so, they just purchase more of whatever consumer goods and services they weren't buying before because houses were more valuable.

The notion that nothing substitutes for additional houses is inconsistent with the usual shape of indifference curves. With ordinary indifference curves, they should have been purchasing even more houses and fewer other goods before. (Very roughly, what are we saying about the marginal utility of other consumer goods before? It suddenly drops to zero at anything more than current leves of consumption?) Sure, you can draw linear indifference curves. Perhaps the marginal utility of every other good but houses drops to zero at anything beyond the current level of spending. But how likely is that?

Kling, then, is focusing on a rather unusual scenario where people were satiated with consumer goods and services. Purchasing new homes involved no sacrifice of consumer goods and services for them, because they want nothing else beyond the current amount they are consuming.

But usually, the next step is to point to leisure. Perhaps some people have every consumer good or service they desire, but surely they value leisure. So a sound microeconomic account would be that people no longer want to buy houses, and have nothing else they want to buy, and so they choose to enjoy more leisure. They cut back on labor. (Of course, the more usual scenario would be that they purchase fewer houses and more other consumer goods and more leisure.)

So, fewer new houses are produced, and people who would have been buying them now take it easy. They retire early, take more days off, fewer family members work, and the like. And so, in this scenario, production and employment falls at least partially because there are increased quits, increased job vacancies, labor shortages, less labor provided to production, and so less output. Firms should be cutting back production, while complaining that they cannot find enough workers to maintain output at past levels.

This is possible, of course, especially if we drop the assumption that there is no added demand for other consumer goods. A sudden drop in the appeal of a single consumer good should be expected to result in a lower supply of labor. However, in the recent recession, there was a drop in the number of quits and job vacancies. Firms in nearly every sector complained that they had to reduce production because of weak sales, not because of too few workers.

That quits and job vacancies exist during a recession may be inconsistent with some naive notion that all unemployment is due to deficient expenditures on output in aggregate (read monetary disequilibrium,) that is just a straw man. The question is whether some, and perhaps quite a lot, of unemployment during a recession is due to monetary disquilibrium.

Let's consider another possibility. Suppose our households that are satiated with consumer goods continue to work, and choose to save. They don't want any consumer goods today, but perhaps some will appear in the future that strike their fancy. This appears to fit in very well with the Kling scenario. People were purchasing single family homes, but now they don't want them and so they save the money they would have spent on the homes. Entrepreneurs have yet to discover what they do want to buy yet. Those who were working to produce the homes are unemployed, waiting for entrepreneurs to discover what those who no longer want to buy the houses want instead.

The notion that we have unemployment because people choose to save is a traditional question in macroeconomics. Still, it should have a simple microecomic answer. If people want to save more for whatever reason, this should result reduction in the production of consumer goods, but an expansion in the production of capital goods. Continuing with the single family housing assumption, people purchase fewer single family houses, and so there is less residential investment and instead there is more investment in capital goods--equipment and factories.

Of course, the firms who are purchasing the capital goods will only do so if they expect to be able to sell consumer good in the future. And what consumer goods will those no longer buying the single family homes want in the future? What a difficult problem for entrepreneurs to solve. As before, they must determine what it is that those who were buying single family homes want now. One might even say that savers and investors are different people with different motives, right? Sounds right out of vulgar Keynesianism 101.

Of course, not only are savers and investors different people with different motives, so are the buyers and sellers in every market. What coordinates every market is prices, and the relevant price for saving and investment is the interest rate. The solution to the problem where satiated individuals no longer want to buy new single family homes, don't want to buy anything else, and instead want to save and wait until some new good they do like is introduced by entrepreneurs, is a lower interest rate.

How does a lower interest rate solve the problem? It causes a decrease in the quantity of saving supplied and an increase in the quantity of investment demanded, so that saving and investment are again coordinated. If the individuals who no longer want to purchase houses are truly satiated and want no more of any existing consumer goods, a decrease in the interest rate will hardly reduce their quantity of saving supplied. But the lower price creates a signal and provides an incentive for other people who are not satiated to expand their current purchases of consumer goods. As for capital goods, there are any number of consumer goods currently being produced that at least some people will want in the future, and the interest cost of their production is now lower.

But suppose the interest rate falls so low that people who are saving for whatever reason simply accumulate larger money balances. Interest rates don't fall further because people would rather hold money than lend. In other words, the zero nominal bound on interest rates interferes with the ability of interest rates to coordinate saving and investment.

Again, this is a pretty standard macroeconomic issue. In the situation under consideration, those who would have purchased homes are instead accumulating money balances. They continue to work and save, intending to spend the money they are accumulating when entrepreneurs come up with some new consumer good that strikes their fancy. While interest rates have fallen and may have led to some people to consume more than they would have and some firms to invest more than they would have, still, these expansions in the demand for goods fail to match the decrease in the demand for housing. Only when some new good is introduced, while those who used to produce the houses become employed.

From a monetary disequilibrium perspective, all we can say is... exactly. There is monetary disequilibrium.

Suppose that the quantity of money is fixed, and the real quantity of money must rise to match the demand. As the prices and wages fall, real money balance increase. If it is outside money, then this raises real wealth. Those holding money balances (which is everyone) are wealthier. Some of them who are not satiated with all consumer goods begin to purchase consumer goods. This is a reduction in saving. While those satiated individuals who have nothing they can do with their income since they aren't purchasing new homes, and are saving by accumulating money balances will presumably just hold these added real balances, other people will consume more. Firms will invest more to produce the consumer goods desired by those who are not satiated.

If all money is inside money, the situation is more difficult. As before, the lower price level (including wages) expands real balances, but the effect on real wealth are exactly offsetting. Those holding money are creditors, and like all creditors, the are wealthier due to the lower price level. Unfortunately, there are matching debtors who are poorer. For a pure inside money, interest rates must fall. If the monetary order is based upon a zero-interest hand-to-hand, inside money, then the nominal interest rate cannot fall below zero (or rather the cost of storing this currency in a vault somewhere.)

If the price level has some anchor and it is expected to return to that level at some future time, then a lower price level now results in higher expected inflation and so a lower real interest rate. Once the real interest rate is sufficiently low (perhaps negative,) then as above, the lower real interest rates generate current consumption by those who are not satiated as well as investment to produce those existing consumer goods in the future.

Unfortunately, if the price level is not anchored, the situation is more problematic. The worst scenairo is where lower prices cause expectations of falling prices. The result would be a higher real interest rate. Almost as bad is a scenario where a central bank tries to keep the price level rising at a constant rate, say 2 percent, from its current level. That means that real interest rates can never be less than minus 2 percent.

If, on the other hand, the nominal quantity of money can rise, then monetary disequilibrium can be corrected without any decrease in the price level. Prices and wages can remain stable or even grow at their previous trend. If there is a shift in the composition of output away from things that are no longer in demand, such as single family homes, to whatever goods those who are not satiated want to buy, then like in all such situations, productive capacity may be depressed and catch up as appropriate capital goods are constructed and workers are trained.

Can an expansion in the quantity of money result in higher expenditures on output? Yes, and in any number of ways. The fact that there are some people who are satiated, but still want to work and save, in no way prevents an expansion in the quantity of money from correcting monetary disequilibrium and returning real expenditure to the productive capacity of the economy.

So, PSST is one way of describing the market process and creative destruction. However, Kling's analysis goes beyond that and heads into really basic fallacies that are inconsistent with scarcity and the market processes that return the economy to full employment in the long run. How do changes in interest rates and prices and wages bring real expenditure back to the productive capacity of the economy? That people are just waiting for entrepreneurs to come up with something new doesn't exactly fit in.

Sunday, June 12, 2011

Cochrane Again Suggests that Money Doesn't Matter.

Cochrane claims:
Now, of all the stories we've heard to explain our sluggish recovery, how plausible is this one: “Our big problem is the maturity structure of Treasury debt. If only those goofballs at Treasury had issued $600 billion more three-month bills instead of all these five-year notes, unemployment wouldn’t be so high. It’s a good thing the Fed can undo this tragic mistake.” That makes no sense.
It makes perfect sense.

Cochrane's argument is that with nominal interest rates on short term government debt very close to zero, it is the same as money.     That is plausible enough, and it means that if the Treasury funds the national debt with T-bills, then it is expanding the quantity of money.    If the Fed creates base money (currency or bank reserves) through open market operations using short term government bonds, it isn't creating money, but rather changing the composition of the quantity of money.   If the Fed creates base  money using open market operations with long term bonds, it is creating money.

It is hard to believe that Cochrane hasn't thought through these implications.    So, what he is really saying is that the quantity of money doesn't matter.

There are two possible reasons why this might be true.   One reason is that nominal expenditures on output don't increase with the quantity of money.   That requires that the demand to hold money passively increases with any change in the quantity of money.    Instead of the plausible version of the liquidity trap which Cochrane has described, (open market operations with zero yield government bonds don't effect nominal expenditures because it is just a change in the composition of money,) Cochrane is proposing a much stronger version.   Money doesn't matter.

The second reason is perfect market clearing.   In that scenario, the prices and wages always adjust so that the real quantity of money equals the demand to hold money, and at the same time, the level of real expenditures is made to adjust to the real productive capacity of the economy.   Increasing the quantity of money will increase nominal expenditures on output, but not real expenditures on output, much less actual production or employment.   The only result will be higher prices and wages.

What then are we to make about his further claim:
Unemployment isn't high because the maturity structure of U.S. government debt is a bit too long, nor from any lack of “liquidity” in a banking system with $1.5 trillion extra reserves.

Because the banking system has more reserves than they are legally required to hold, and further, much more than they have held in the past, it must be that there would be no excess demand for reserves if real output and employment expanded substantially.   What?

However, there is an element of truth in Cochrane's argument.   While his hardcore Keynesian "money doesn't matter" arguments are absurdly wrong, and his unstated assumption that if more nominal expenditures on output would help, prices and wages would be lower already, is almost nearly as bad, there is an alternative explanation.

As he states:
QE2 distracts us from the real microeconomic, tax, and regulatory barriers to growth
From a continuous market clearing perspective, it must be that any reduction in real output or increase in the unemployment rate is due to supply-side factors.   Prices and wages are always such that on the whole, firms sell what they produce and people work the amount they want.   Output can only be low because people are unable or willing to produce, never that they cannot sell for a lack of buyers.    Unemployment can only rise because people are unable or unwilling to work, never because they can't find jobs.

However, there is more to the story.   It is conceivable that a pro-business political agenda would raise business confidence, reduce the demand to hold money, and raise investment demand--money expenditures on capital goods.   As sound businesses reduce their money holdings, (which would include selling off their holdings of short term government bonds) and purchase capital goods, spending on output would rise.   As sound businesses seek to borrow more from banks, banks would use their excess reserves to make loans.   The banks would be lowering their demand to hold base money  and that would expand the quantity of money in the form of deposits available for households and firms.

It is true that past experience suggests that the existing quantity of base money far exceeds what is necessary for money expenditures and real output to be substantially higher than any plausible estimate of productive capacity.     The "problem" is that the demand to hold money is exceptionally high.   That includes the demand by banks to hold reserves, but also the demand by firms and households to hold money, including the quasi-money short term government bonds.   If that preference were to change, and firms, households, and banks should choose to hold less money, then money expenditures on output would rise.   Of course!

My view is that the Fed's job is to keep money expenditures on a stable growth path.   If poor tax and regulatory conditions cause an increase in the demand to hold money, the Fed should expand the quantity of money enough to keep money expenditures on target.   And if better microeconomic  policies result in a lower demand to hold money, then the Fed should reduce the quantity of money.  

Cochrane's approach of claiming the Fed can do nothing and calling for pro-business policies to reduce the demand to hold money is...inaccurate.   I favor pro-market regulatory and tax reform, and I suspect that Cochrane and I would have large areas of agreement regarding such reforms.   But I also favor a Fed policy that adjusts the quantity of money (including short term government debt when appropriate) however much is necessary to match the demand to hold money and keep money expenditures on a slow, steady growth path, even if actual tax and regulatory policy result in reduced business confidence.

Sunday, June 5, 2011

100% Reserve Banking--Some History from George Selgin

George Selgin, writing in the new blog, Free Banking, has an interesting post on 100% reserve banking.   He argues:

...every significant 100-percent bank known to history was a government-sponsored enterprise, which depended for its existence on some combination of direct government subsidies, compulsory patronage, or laws suppressing rival (fractional reserve) institutions. Yet despite the special support they enjoyed, and their solemn commitments to refrain from lending coin deposited with them, they all eventually came a cropper.

Selgin discusses the history of 100% reserve banks in Amsterdam, Venice, and Barcelona.    That government action in the early modern period replaced the international fraction-reserve banking system developed by Italian money changers in the middle ages wasn't news.   Exactly how transparent were these efforts to get money in the hands of the government was news, at least to me.

In the comment section of that post, Selgin points to his working paper on the role of goldsmith's in the rebirth of fractional reserve banking in the 17th century.     He reviews the textbook stories of fraud (which I have repeated for years in my money and banking classes.)  

The myth is that merchants carried about bags and chests of gold until they began to store them in the royal treasury.   Wicked Charles I seized the money.   Then, people paid to store coin with goldsmiths.   Soon the goldsmiths began to secretly lend out some of the stored money that was gathering dust.    Fractional reserve banking was discovered!

One problem with that story is that it ignores the banking system developed by Italian money changers that operated all over Europe.   Selgin provides some evidence that the Italians took over the business from the Jewish money changers centuries before.    Perhaps it should be no surprise that English goldsmiths were similarly stepping into a well developed business model.

Apparently, the rule was that if a "depositor" placed coins in a tied or sealed bag, then these were to be stored.    Any loose coins "deposited" were the property of the "bank," which was under the general obligation to keep its promises to pay money in the future.

I was also interested to discover that the gold "stored" by in the royal treasury in the reign of Charles I was metal waiting to be coined.   It was borrowed, but most was repaid immediately (due to the outcry) and the rest repaid promptly.    Interesting episode regarding the government monopoly on coining full bodied commodity money, but not too relevant to the development of banking.  

Saturday, June 4, 2011

New Monetary Economics

My perspective on monetary economics was strongly influenced by Robert Greenfield and Leland Yeager's "A Laissez-Fair Approach to Monetary Stability."     In that paper, they described what they called the "BFH payments system."

Tyler Cowen recently claimed that the "New Monetary Economics" is alive and well.    He said:
The standard view is that Fischer Black, Bob Hall, Neil Wallace, and Eugene Fama wrote a few creative papers on monetary theory in the 1980s (for Black the 70s), but that the embedded monetary scenarios were “too weird” and the line of research did not prove fruitful.  Even in “free banking” circles the “New Monetary Economics,” as it was called for a while (NME), wasn’t always taken very seriously.
The connection between Greenfield and Yeager and the New Monetary Economics is that BFH is short for Black-Fama-Hall.   What might better be called the Greenfield-Yeager payments system was inspired by ideas that had been introduced by Fischer Black, Bob Hall, and Eugene Fama.

At first pass, Greenfield and Yeager's proposal was for the "dollar" to be defined in terms of a broad bundle of goods and services and for the medium of exchange to be a system of competing money market mutual funds.    As they, and others (including me) explored this system, the centrality of checkable money market funds shifted to the background while the role of "indirectly convertibility" came to be seen as central.   Checks (or other dollar-denominated monetary instruments, would be redeemable in gold or some other settlement medium, equal in market value to the market value of the bundle that defines the dollar.

Despite the shift away from any focus on market pricing for mutual fund shares early on, it remains true that much of the inspiration for the scheme involved thinking about a monetary order without any hand-to-hand currency.    To me, one important legacy of the "New Monetary Economics" is a shift away from seeing zero-nominal-interest hand-to-hand currency as the core concept of money to a view that sees it as a possibly useful appendage to the monetary order.   Money pays interest, well, except for hand-to-hand currency, where that exists.   The opportunity cost of holding money is the difference between the interest rate earned on money and other assets.   Of course, if there is no nominal interest rate paid on currency, it is a little different.  (One error of the "New Monetary Economics" was to assume that if none of the peculiar characteristics of zero-interest hand-to-hand currency exists, then the resulting payments system is no longer a monetary order.)

Another element of the "New Monetary Economics" that has influenced me is the notion that in equilibrium, anyway, it is the medium of account that determines the price level.    And further, if the actual price level is  different from whatever clears the market for the medium of account, then there are going to be serious macroeconomic difficulties.      For example, if gold serves as the medium of account, then the dollar price level depends on the supply and demand for gold.   To the degree that there is a demand for gold for coins or reserves, then that impacts the price level much like demands for dental work or jewelry.    Various sorts of financial instruments must adjust in price or quantity to equilibrate the supplies or demands for them.

Of course, the gold standard is long past, but this approach leads to a focus on the supply and demand for the monetary base.   That is what defines the dollar and serves as medium of account.    All the variety of other sorts of financial instruments that play monetary roles must adjust in price or quantity to equate supply and demand.     But once one accepts this perspective, there is no particular reason to assume that the demand for the monetary base is proportional to nominal expenditure on final goods and services.   Why shouldn't there be substitutions between base money and other sorts of financial instruments that provide similar services?

Finally, it was concerns with the operations of indirect convertibility that lead me (and others) to shift towards index futures convertibility.     Because targeting the growth path of money expenditures on output has advantages over targeting some measure of the price level (say the price of a broad bundle of goods and services,) the new monetary economics lead me to "quasi-monetarism."

HT to Kurt Schuler at Free Banking.