Nominal GDP Stability and Coordination

A nominal GDP level target requires changes in the demand to hold money be accommodated by changes in the nominal quantity of money.   In my previous post in response to the Wagner and Viteel paper, I argued that there is no presumption that these changes in the quantity of money occur with no impact on the flow of expenditure through the economy or the allocation of resources.   Rather, the argument is that such changes are entirely appropriate and coordinating as long as the quantity of money increases such that it matches increases in the demand to hold money.   Rather than imagining some kind of perfect neutrality, the proper comparison is to an increase in the demand to hold some other type of financial asset.    Given that changes in the quantity of money, even if accommodating changes in the demand to hold money, involve some change in the pattern of expenditures and allocation of resources, why is it desirable to keep nominal GDP stable?

First, I would point out that there are some changes in the demand for money where the needed change in the allocation of resources is small.   The thought experiment of a reduction in consumption and increase in the demand to hold money, matched by an increase in the quantity of money that is used to fund additional investment involves a major shift in the pattern of expenditure and allocation of resources.   Suppose instead that households sell off financial assets to increase money balances and this is accommodated by a bank (commercial or central) making open market purchases, creating money by purchasing financial assets.   It is possible that the issuer of money would purchase the exact financial assets sold, resulting in no change in the flow of expenditure or the allocation of resources.   Of course, if a private bank purchases some other sort of financial asset (or makes a commercial bank loan,) the pattern of finance would be changed somewhat.   Only with stringent and unrealistic assumptions (which are not uncommonly made in finance theory,) would this have no impact on what particular investment projects are ultimately financed.     In my view, it is at least roughly similar to a household's choice to sell off bonds issued by corporation A with one more year to run to maturity, and purchasing bonds issued by corporation B with five more years to maturity.   Perfectly neutral, not likely.   But how neutral would be a deflation in prices and nominal incomes in response to a sale of financial assets to fund increased money balances?

Rather than continuing on to consider various scenarios regarding shifts in the demand and supply of money, I will use a different approach to explaining why stable spending on output is desirable.   First, consider a two consumer good economy.   A change in preference results in an increase in the demand for apples and a reduction in the demand for oranges.   I must admit that I find it most natural to consider this scenario as an increase in spending on apples and a reduction in spending on oranges.   Perhaps this is based upon a simplistic transfer of the reasoning of a single household with a budget constraint constructed from a given nominal income.   The household remains on the budget constraint and so shifting along it involves less spending on one good and more spending on the other.   Less spending on oranges and more spending on apples.

 Regardless, I would construct the market experiment as the demand curve for apples shifts to the right while the demand for oranges shifts to the left.   The price and quantity of apples rise.   The higher price of apples results in a move up the supply curve of apples.  Resources are pulled into apple production by higher prices and profits.   Meanwhile, the price and quantity of oranges fall.  The lower price of oranges results in a move down the supply curve of oranges.   Resources are pushed out of the orange industry by lower prices and profits, and perhaps even losses.

If, as is usual, the long run supply elasticity is greater than the short run elasticity, profits and prices will fall in the apple industry, with profits returning to normal and any long run increase in price reflecting higher rent on resources specific to apple production.   Similarly, any losses in orange production must disappear and prices will recover.  Any long run reduction in prices will reflect lower rents to resources specific to orange production.   These shifts from short run to long run supply can be represented by short run supply curves shifting--to the right for the apple industry and to the left for the orange industry.

Perhaps this is just my naive microeconomist thinking, but surely this is the way the market process is supposed to work?

However, it is clear that the same reallocation of resources and production would be possible if the demand for oranges fell, and the demand for apples stayed the same.   In this simple two good world, it might seem to make no difference..   The demand for oranges falls.   The price and quantity of oranges fall.  The lower price results in a shift down the supply curve for oranges.   Lower prices and profits, perhaps even losses, push resources out of the orange industry.   Where do the resources go?   To the apple industry.   The supply curve for apples shifts to the right.   The result is a shift down the apple demand curve, with the result being a lower price and higher quantity of apples.   The long run adjustment from the less elastic short run supply to the more elastic long run supply can be represented by further shifts in short run supply.   The short run supply of oranges shifts to the left and the short run supply of apples shifts further to the right.

So far, there has been no specific discussion of markets for resources, such as labor.  In the first scenario, the increase in the demand for apples results in a nigher imputed demand for labor while the reduction in the demand for oranges results in a lower demand for labor   While there is no guarantee that the two effects are exactly offsetting, there is some offset.   At least some of those workers pushed out of the orange industry can find new jobs in the apple industry.   Leaving aside a substitution from labor to leisure, all of the labor must shift from oranges to apples.  (That is my usual assumption, of course, real business cycle theory emphasized just this substitution to leisure that I usually ignore.)

On the other hand, in the second scenario, one key reason the supply of apples increases is because lower resource prices, including wages, results in a lower cost of production for apples.   The process requires a surplus of labor.   Both the apple and orange industry must cut the wages of all of their workers for there to be a return to equilibrium.   Compared to the shift in demand in product markets, the market process in the second scenario puts an additional burden of adjustment on resource markets, including labor markets.  In the end, just as in the first scenario, the workers are all shifting from the orange industry to the apple industry.

More importantly, if apples and oranges are but two products in an economy made up of thousands of products then the adjustment  process will inevitably impact other markets.   Substitutes and complements in production and consumption will be impacted, as well as the outputs of apples and oranges.   Still, in the second scenario, the reduction in wages and lower costs will impact all markets, raising their supplies and tending to result in lower prices and higher quantities too.   If their relative prices are to remain unchanged, the demands for each and every good in the economy must fall at least a little.   At least, they must do so if the entire adjustment is to be made with no increase in the demand for apples.

Now, back to the two good scenario.   It would be possible for the reallocation of resources to occur solely through an increase in the demand for apples, with the demand for oranges remaining constant.   The production of oranges would be curtailed by a decrease in the supply of oranges as resources are pulled into apple production.   With labor markets, this would involve a shortage of labor and higher wages.  The higher cost of production is what would cause the lower supply of oranges.   In a many good world, the result would be decreases in the supplies of all goods and at least slight increases in the demands for everything other than apples.  Worse, outside of the world of the Walrasian auctioneer, there might well be some overshooting of apple demand.

Wagner-Viteel seem to argue that increases in the demand for money should not be accommodated by increases in the quantity of money because it is better to coordinate a shift in demand solely by pushing resources out of firms with reduced demand, while surpluses of resources lower resource prices and so costs and thereby increase the supplies of other goods.   As explained above, it could happen that way, but why is it better to require these adjustments in prices and wages in all product and resource markets?

The Mises problem is directly a need to contract the production of capital goods and expand the production on consumer goods.    Why shouldn't this be accomplished by a simultaneous increase in spending on consumer goods (apples,) pulling resources in that direction, along with a push of resources out of capital goods (oranges?)     Why would it somehow be better for this reallocation to occur solely by pushing resources out of the production of capital goods so that surpluses of resources push down costs and so increase the supplies of consumer goods?

The Schumpeter problem has two parts.   Consider the introduction of a new product.   Why shouldn't the demand for the new product rise (apples) and the demand for the old, heading-to-obsolescence product (oranges) fall?  Why not have resources pulled to the innovative product and pushed out of the obsolete product?   While it would be possible to coordinate this solely by a reduction in demand for obsolete products.....

Well, it is actually hard to imagine how that would be possible.   The shift in expenditure would seem pretty much essential.   And, of course, Schumpeter's recession occurs when there are few (or no) innovations.   Given this simple thought experiment, what is the problem?   While there might sometimes be no added demand for new products, neither would there any decreases in demand for existing products due to their becoming obsolete.  

What about the other element of creative destruction?   Improved efficiency for producing existing goods    Suppose there is improved efficiency in producing apples.   There is little doubt that this should be associated with more apple production.   However, the direct effect is an increase in the supply of apples, shifting the supply curve to the right.   Output and profits rise, prices fall.

What happens to spending on apples depends on the elasticity of demand.   If demand happens to be unit elastic, then spending on apples remains the same.   If total income and expenditure is held constant, then the demand for oranges remain the same.

If the demand for apples is inelastic, then spending on apples fall.   Some of those involved in the apple industry must earn lower nominal incomes.   However, what this means is that some of those buying apples want to use part of their added real incomes to purchase other goods.  In a two good scenario, that is a higher demand for oranges.  Prices and profits rise in the orange industry as does the production of oranges.  Where do the resources come from?   Out of the apple industry.   Demand rises in the various industries where people want to buy more.  While "demand" remains the same in apple industry, spending falls.   Isn't that exactly what provides the proper signal and incentives?

On the other hand, if the demand for apples were elastic, then the increase in supply of apples results in more spending on apples.   That means that the lower price of apples motivates some of those purchasing other goods to spend more on apples.   In the two good scenario, that is less spending on oranges.  Prices, profits, and production fall in the orange industry.   This pushes resources out of the orange industry into the apple industry, where total spending and incomes are rising.

At least qualitatively, these adjustments appear appropriate.   Again, there is no guarantee that there would be no net change in the demands for resources, such as labor, there would surely be at least some offset.

And, of course, this is exactly how nominal GDP targeting works.

(Schumpeter's theory of the business cycle very much depends on credit demand and the quantity of money being driven by innovation.   More innovation results in more credit demand, an increase in the quantity of money and so nominal GDP.   The "problem" of recession occurs when innovation slows or stops.  Whether this could occur with a constant quantity of base money or even some broader measure of money hardly matters to Market Monetarists.   Our approach constrains the process of creative destruction to operate in an environment of stable nominal GDP.)

Anyway, we had nearly a half-century of Hayek and Mises, and their followers arguing that trying to generate "full employment" by increasing the quantity of money is a bad idea.   If there is a need to reallocate resources from capital goods production to consumer goods production, then trying to keep employment unchanged in capital goods industries is a bad idea.   Similarly, trying to maintain employment in industries producing obsolete product is a bad idea.   Trying to maintain employment in industries where productivity has increased and demand is inelastic is a bad idea.

Targeting total employment or unemployment is a bad idea.

Similarly, trying to maintain the real output of capital goods or obsolete products or import competing industries or housing or.... anything, is a bad idea.     If bottle necks limits expansion in some industries while decreases in output occur rapidly in shrinking industries, then real output will fall.   Nominal GDP level targeting does not try to keep aggregate real output from falling in that circumstance or any circumstance.   In particular, it does not propose increasing the quantity of money and nominal GDP to reach some target for real GDP.

Targeting real GDP is a bad idea.

 Nominal GDP level targeting is not the unemployment rate targeting.  Nominal GDP is not real GDP targeting.   Isn't that obvious?

Of course,  I don't favor keeping nominal GDP constant in a progressive, growing economy. (I don't favor Hayek's "ideal" policy.)  I favor a stable growth path for nominal GDP, with the growth rate based upon the estimated growth of potential output.   At any point in time, the level of nominal GDP is given and the above analysis applies.

I need to explain why this growing trend for nominal GDP is not especially disruptive, and more importantly, not more disruptive relative to the alternatives--including constant nominal GDP.  I'll get to that.

Returning to Wagner and Veetil, I need to explain why accumulating money balances due to uncertainty is saving and as such, implies a reduction in the natural interest rate.   In other words, it is appropriately coordinated with other households and firms by an adjustment in market interest rates.    There is nothing about a temporary increase in the demand for money due to uncertainty that makes reduced nominal GDP somehow coordinating and an expansion in the nominal quantity of  money discoordinating.