NGDP Targeting and a Small Open Economy

IHS and the Mercatus Institute had meeting about monetary policy in San Diego on June 25th.   I was fortunate to attend.

Scott Sumner was interviewed by David Beckworth for Macro Musings.   I suppose we were the live audience.   (As I write, the interview isn't up yet.)

In the question and answer period, Sumner was asked whether it would be possible to test out Nominal GDP targeting in some smaller country--say Kenya--rather than hold out for the Fed or the ECB.  Sumner said that he was not sure that NGDP targeting is appropriate for a small open economy.   The problem is that they may be too specialized in producing a commodity with an unstable world price.

When I spoke to him later, he said that for countries in that situation it is probably better to stabilize nominal labor income.   Of course, Nominal GDP is hardly ideal even for a country like the U.S. either.   Changes in indirect business taxes, for example,  could create problems.   A stable growth path for something like total labor compensation might be better  for the U.S. too.

But I would like to explore the small open economy issue a bit more.   While I can imagine scenarios where shifts in commodity prices might cause problems, I think that the problem isn't really specialization in commodities with unstable world prices.

For example, suppose everyone in a country is a coffee farmer.   The country stabilizes the growth path of nominal GDP, so regardless of world coffee prices or local coffee production, nominal income from coffee sales grows at a stable rate.

How is this possible?   If world coffee prices rise, the value of the currency rises enough so that the domestic price of coffee remains the same.    Nominal incomes remain the same and imported goods become cheaper.

If, on the other hand, coffee prices fall, the value of the currency falls enough so that the domestic price of coffee is unchanged.  Nominal incomes are the same, but imported goods become  more expensive.

Complete specialization in producing a commodity with an unstable price looks like no problem--other than menu cost of all of the shops changing the prices of  imported consumer goods.

Now, lets add a bit more realism.   Who is operating the shops full of imported goods?   What about haircuts?   What about home construction?

If the world price of coffee rises, at first pass, the currency rises in value so that the domestic price of coffee is the same.    Nominal income for the coffee growers and nominal income in the nontraded sector is unchanged.   The imported goods in the shops are cheaper.   If the income and substitution effect for nontraded goods exactly offset, then that is all.   In other words, if the increase in the demand for nontraded goods due to the higher real incomes is equal to the decrease in quantity demanded due to their higher relative price compared to imports, then the distribution of nominal income and the allocation of resources remain unchanged.

However, if the income effect is greater than the substitution effect, the currency must rise more than in proportion to the world price of coffee so that the domestic price of coffee falls, reducing nominal income in the coffee sector.   The profitability of coffee falls a bit, freeing up resources to provide more nontraded goods.   If nothing else, somebody is going to have to handle the increased volume of the imported goods coming in.   Nominal income in the nontraded sector increases.

The less pleasant scenario for this nation of coffee growers is a decrease in the world price of coffee.   The currency falls in value and, at first pass, the domestic price of coffee is the same as is nominal income in the coffee sector.   Imported goods in the shops become more expensive.   If the income and substitution effects exactly offset, nominal incomes in the nontraded sector are unchanged, but real incomes fall just as they do for coffee growers.  There is inflation of consumer goods prices--particularly the imported ones.  If income effect is greater than the substitution effect, the currency will fall enough to raise the domestic price of coffee, making coffee growing more profitable, expanding the demand for labor and other resources no longer needed in the nontraded good sector.

The situation where the substitution effect is greater than the income effect is a bit inconsistent with the conventional terminology of "nontraded goods sector."    The analysis is no different from a situation where there are import competing industries.   If the world price of coffee rises, the currency rises, and the now cheaper imports result in lower demand and lower nominal incomes in the import competing sector.   The currency, therefore, must rise less than in proportion to the world price of coffee, such that domestic coffee prices and nominal incomes in the coffee sector rise an amount that offsets the decline in the prices of import competing goods and the resulting decrease in nominal income in that sector.   Profitability in the coffee sector rises drawing resources from the import competing sector into the production of coffee for export.  Of course, import prices are cheaper, making the effect on real incomes in the import competing sector somewhat ambiguous, but real income rises in aggregate because of the improved terms of trade.

If the world price of coffee falls, the currency decreases in value.   The demand for import competing products rises, resulting in higher prices and higher nominal incomes in that sector.   The decrease in the value of the currency is dampened then, so that there is a decrease in price and nominal income in the coffee sector that offsets the increase in spending and nominal income in the import competing sector.   The increased profitability of the import competing sector creates an incentive to pull resources away from coffee production to the production of goods for domestic consumption.   With the higher prices of imported goods, real incomes in the import competing sector are ambiguous, though aggregate real income falls with the less favorable terms of trade.

It is these considerations that suggest to me that nominal GDP targeting might well be appropriate for a small open economy specializing in the production of a commodity with an unstable world price.   What is Sumner's concern?

Consider a situation where our small open economy has a giant copper mine or maybe a giant diamond mine.    The product makes up approximately all exports and a substantial portion of GDP, but directly generates little employment.

If the world price of copper increases, then the value of the currency increases, the domestic price of copper is unchanged and imported goods are cheaper.    If the income effect is greater than the substitution effect for nontraded goods, then the price of copper rises somewhat less so that the domestic price of copper is less and so nominal income generated by copper falls, making it slightly less profitable to produce so that fewer workers are needed and they can be shifted to the nontraded sector where prices and nominal incomes rise.   Of course, with the assumption that there are very few copper miners anyway, and the other resources useful for copper mining might not be very useful in other endeavors, this adjustment in relative nominal incomes might provide what is superficially the correct signal and incentive to reallocate resources, but there just is not much reallocation possible.  There has just been a pointless inflation in the nontraded goods sector.

If the world price of copper falls, this problem is even more apparent.   The value of the currency decreases.  At first pass, the domestic price of copper is unchanged.   Imported goods are more expensive.   If the income effect is greater than the substitution effect, the value of the currency decreases by less, the domestic price of copper actually rises a bit, nominal income in the copper industry rises, and nominal income in the nontraded sector decreases.   This provides the signal and incentive to shift resources from the nontraded sector to copper production.   But if copper production generates few employment opportunities, the result is that there is really just a pointless deflation of prices and wages in the nontraded sector.   Compounding the pain in the nontraded sector, there is substantial consumer price inflation due to higher import prices.

Inflationary recession in most of the economy, while the copper mine earns more nominal profit.   If the copper mine were privately owned, this would be a political disaster.

If there are import competing industries, these problems are exacerbated.   With an increase in the world price of copper, the value of the currency rises, with nominal income rising in the copper sector while falling in the import competing sector.   While this provides a good signal and incentive to shift more resources to copper production, by assumption that happens to a minor degree.  Again, there is mostly just a pointless deflation of prices and wages in the import competing sector.

If the world price of copper falls, the currency falls in value. Prices and nominal income in the import competing sectors increase, while the domestic price of copper and nominal income in the copper industry decrease an offsetting amount.   While this provides the proper signal and incentive to shift resources from copper production to import competing industries, by assumption, there is little opportunity for such an adjustment.   The result is just an unnecessary inflation in prices and wages in the import competing sector.   Of course, rising import prices imply consumer price inflation anyway.

If there are other export industries along with copper, for example, fruit, an increase in the world price of copper and the resulting increase in the value of the currency will reduce domestic prices and nominal incomes in these other export industries.   While this would provide an appropriate signal and incentive to shift from the production fruit to copper, again, the possibility for such a reallocation of resources is limited by assumption.   With a decrease in the world price of copper, the reduction in the value of the currency will result in higher domestic prices and nominal incomes in other export industries.

Consider a scenario where the copper mine is on a distant offshore island.   The mining is done by a foreign multinational with expatriate workers from other parts of the world.   Leaving aside any income the government collects from this enterprise, does it make any sense to include the nominal output of this operation when determining an appropriate monetary policy for the mainland? It would seem more appropriate for monetary policy to stabilize nominal GDP for the mainland while ignoring what is happening in what is effectively a foreign industry.  

Sumner's suggestion that total labor compensation be stabilized would probably help solve the problem where an export generates a substantial part of GDP and little employment.    But more generally, the problems I see with nominal GDP targeting in this context involves the specificity or substitutability of resources in production of various goods.   My coffee example assumed labor and resources could be shifted between coffee and other products-nontraded, import-competing, or other exports.    My copper example assumed that this was nearly impossible.

I believe this is related to the notion of an optimal currency area.   The example of the copper producing island causing pointless disruption on the mainland makes this plain.  The island and mainland do not make an optimal currency area.   Regardless of its geographical location, however,, the same issues apply.   Regardless of the location of the copper mine, perhaps it is better to stabilize the growth path for nominal production for the rest of the economy (nominal GDP less final copper output.)

But nominal GDP targeting, does not, in general, result in problems when countries are specialized in the production of a commodity with an unstable price.   In the extreme, where all that is produced is such a good, it works quite well.   And it also works even better when resources can be shifted between the export sector, the nontraded sector, the import competing sector, and other export sectors.

Now, it might be that stabilizing the growth path of nominal labor compensation would do ever better.   But nominal GDP targeting would work better than stabilizing the exchange rate or consumer price inflation.