Tuesday, March 14, 2017

The Trade Deficit, National Income Accounting and Aggregate Demand

Peter Navarro has been using the national income accounting identity to argue that policies that reduce the trade deficit must increase the production of goods and services in the U.S.   There have been various efforts to dismiss his claim as absurd.   Interestingly, he appears to be following in Keynes' footsteps by being a little confused about the relationship between identities and equilibrium conditions.   But the best way to interpret Navarro's claim is to see it as a very simple Keynesian approach.   On this view, aggregate real output in the U.S. is driven by aggregate demand for goods and services produced in the U.S.

U.S. GDP is a measure of the production of goods and services in the U.S.   The textbook accounting identity is Y = C+I+G+X where Y is GDP, C is consumption, I is investment, G is government spending, and X is net exports.   Net exports are exports minus imports.   If there is a trade deficit, then X is negative, so it subtracts from C+I+G.   The bigger the trade deficit, then, the smaller is GDP.   A smaller trade deficit, or better yet, a larger trade surplus, implies a larger GDP.    Navarro claims that this means that policies that encourage exports or discourage imports will raise GDP.

The argument that this is just an identity points out that the C, I, and G refer to goods and services categorized as consumption, investment, and government in the U.S. regardless of where produced.   Imports are subtracted to get the goods and services produced in the U.S.   Exports are added to account for the goods produced in the U.S. that are not categorized as U.S. consumption, investment, or government because they are accounted for as C, I or G in some other country.

However, I think this is an unfair characterization of Navarro's view.   Navarro's point is that GDP is equal to spending by U.S. residents on domestically produced  U.S. output plus foreign spending on domestically produced output.   The process for calculating GDP is to add up all spending by U.S. residents on different categories of goods and services and then subtracting off spending on foreign produced goods.  This provides spending by U.S. residents on goods and services produced in the U.S. Spending by foreigners on all types of U.S. output is added back.   The result is total spending on U.S. goods and services.   This is nominal aggregate demand.

Navarro's theory is the simple Keynesian one that the various categories of spending are determined, and if it becomes more costly to purchase foreign goods, then spending will be shifted to purchase domestically produced output.   For example, if total spending on consumer goods and services by U.S. residents is $12 trillion, and $2 trillion of that would be spent on foreign consumer goods and services with relatively open markets, a policy of banning all imports would result in U.S. consumers spending $2 trillion more on consumer goods or services produced in the U.S.    The same would be true for investment--purchases of newly-produced capital goods by U.S. firms as well as purchases by various levels of government.   Therefore, a smaller trade deficit (or larger surplus) would be associated with more aggregate demand in the U.S.

Navarro emphasized the possibility of an increase in exports.   Trump's threats to impose higher taxes on imports from Mexico would result in Mexico agreeing to purchase more U.S. products in order to be allowed to continue to export to the U.S...Assuming that spending by U.S. residents on consumer goods, capital goods, and government in aggregate is unchanged, this increase in the demand for U.S. goods by Mexicans increases total spending on domestically produced U.S. goods and services.   Again, the trade deficit is lower and aggregate demand is higher.

In the simplest Keynesian model, output is solely determined by aggregate demand.   If trade policy can expand exports and reduce imports, these "deals" that reduce the trade deficit (or expand a surplus) will increase aggregate demand and so production in the U.S.

This Keynesian model is most plausible when there is unemployment of labor and excess capacity in most sectors of the economy.    The increase in demand both results in more production and more employment.

In a world with near universal price floors, and a quantity of money such that equilibrium prices are below those fixed prices, this Keynesian approach would apply.    If the floors instead apply to wages, the result would be similar.    Of course, rather than hoping for "better" trade deals to raise aggregate demand, it would be possible to repeal the price floors or else increase the quantity of money.

Fortunately, the U.S. does not have anything like universal price floors.   However, in the short run at least, it seems that prices and wages do not immediately adjust so that real aggregate demand equals productive capacity.   In fact, they appear to have some perverse momentum that is only gradually dissipated.   Prices and wages continue to rise even when market conditions suggest they should fall.   Only very gradually does the rate of increase of prices and wages slow.

It is possible then, that during a recession, tough trade negotiations would boost aggregate demand and hasten a recovery.    This is not an argument that reduced imports and expanded exports will permanently increase output and employment.   In other words, Trump's approach might have made sense in 2008 and 2009, but no longer.    The Fed has already started raising interest rates to restrain the growth of spending on U.S. output.   Policy that reduces imports and expands exports would reduce the trade deficit, but it would it would have no impact on aggregate demand or employment.  It would instead just result in higher interest rates.

Even in recession, restrictions on imports can easily backfire.   That is because exports are not fixed.  For a "small" country, this might not be a problem.   But if the U.S. restricts imports, the result could easily be a recession in other countries, which will result in fewer U.S. exports.  The foreigners demand fewer U.S. goods when their economy suffers recession.   This results in reducing income to those producing the exports, which likely would result in reduced demand for both domestically-produced goods and services, and imported goods.    Casual empiricism suggests that triggering a recession in the U.S. is the best way to reduce U.S. imports and so the trade deficit!

For a small country, the amount that it imports from the rest of the word is small, and so reductions have little impact on the world economy.   For such a country, the only problem is "retaliation," where foreign governments restrict imports from the small country and so reduce its exports.  

As a Market Monetarist, I believe that the monetary regime can and should target a stable growth path for nominal GDP--spending on domestically produced output.   What I think is the most reasonable interpretation of Navarro, that trade policy can reduce trade deficits and increase aggregate demand which might hasten the recovery of output and employment in a recession and avoid the need for disinflation, is unnecessary and inappropriate.    Under the current monetary regime, better monetary policy by the Fed would make this unnecessary.

Regardless, thinking of our current situation, and the long run across many business cycles, it is not sensible to have a trade policy that is about expanding aggregate demand to hasten recover from a recession.   The long run analysis must take into account that a trade deficit is matched by a net capital inflow.   The result is a lower natural interest rate and more investment.    That is, the demand for capital goods, including domestically produced capital goods, is higher than it would be if the trade deficit was lower.

If some policy does manage to reduce the trade deficit, the result may well be more domestic production of goods and services that would have been imported and more goods produced for export, but there would be fewer capital goods produced.   The result is a change in the composition of demand and the allocation of resources, but there is no increase in aggregate demand.

Worse, the long run effect of any restriction in the production of capital goods is a lower growth path of productive capacity and so future real GDP will be lower than it otherwise would have been.

And that points to the fundamental determination of trade deficits--the relationship of domestic saving and investment to world saving and investment.   It would be great if foreign governments would reduce barriers to trade and this allowed for increases in U.S. exports.   However, this could easily result in more U.S. imports, leaving the U.S. trade deficit the same.   Americans would earn more from exports and use the extra income to purchase more imported goods.  That is the key reason to export goods and services--getting imports in exchange.    But that also ignores international investment.

If the world interest rate is below the interest rate that would coordinate U.S. national saving with U.S. domestic investment, then foreigners will be motivated to invest in the U.S. to obtain higher returns.   The amount of this net capital inflow must be matched by a trade deficit.  Foreigners will export goods and services to the U.S., not to buy U.S. goods and services and take them home with them, but rather to buy capital goods in the U.S. or else claims to capital goods--like stocks or bonds.

It is important to understand that this is not foreigners buying up a fixed quantity of assets. The U.S. can and does produce additional capital goods to sell to foreigners.   U.S. workers and firms are building a plant for Volvo in the South Carolina lowcountry.   These new capital goods increase future production in the U.S.  A reduction in the trade deficit, then, means that increase in future production does not occur.   That doesn't mean that production doesn't grow over time.   It probably would grow, but its level at any future time would be lower than if there had been a larger trade deficit, net capital inflow, and investment.

The foreign investors expect a return, and the reduction in the growth path of real GDP would not be entirely a decrease in the growth path of real GNP.   Less will be produced in the U.S. in the future than otherwise, but less income will be paid out to foreign investors.   Still, this is unlikely to benefit U.S. workers--even foreign owned capital goods are a complementary factor of production which would tend to raise their incomes.

It is incomes that foreigners earn from U.S. investments that leads some to describe a net capital inflow as an increase in U.S. indebtedness to the rest of the world.   While foreigners could hold deposits in U.S. banks or purchase corporate bonds or U.S. government bonds or even accumulate paper currency, they in fact also purchase equity--stocks--or make direct investments.   If everything goes as planned they can earn income from all of these investments and even repatriate the money they invested at some future time.  But foreign equity investments in the U.S. are not debts of any U.S. resident.

Now, I favor a reduction in the U.S. budget deficit, shifting it to a modest surplus.   This would result in an increase in U.S. national saving.   The result should be a slightly lower trade deficit in the U.S.   Further, I favor social security privatization, shifting from a pay-as-you go system to a fully funded system.   That will increase U.S. private saving.   This also would tend to reduce the trade deficit.    I favor ending excessive regulation of business in the U.S. (A goal I share with Trump.)  This should increase domestic investment in the U.S., which would tend to raise the U.S. trade deficit.

So, I think Navarro isn't just confusing an identity with an equilibrium condition.  He may seem to.   And maybe he is a little confused.   But I think the best way to understand his view is that he is making a simplistic Keynesian argument--acting as if production is always constrained by demand while ignoring key role of productive capacity.   My Economics 201 back in the day emphasized the simple Keynesian Cross.  Advanced Macroeconomics (first year graduate school) was a bit more sophisticated.   Perhaps it was different at Harvard in Navarro's day.  But I suspect he is using a very simplistic approach that most economists understand quite well.