arrow_back

How much foriegn "drain" is there in government spending?

3/09/2013 12:00:00 PM
Tweetable
One of the concerns often raised about federal contractors is whether the money will go to employ americans, or be outsourced to other countries. The conventional logic is that if the government hires a foreign contractor, then americans loose the full amount paid to them, plus any multiplier effect that money might have had.

To be sure, in the short run there are some valid Keynesian reasons to think this way.  There are frictions and sticky prices, markets in disequilibrium, and (as Nick Rowe pointed out to me) Walras' Law is sometimes violated. But, in a long-run, neo-classical world, this thinking is wrong-headed.

The reason is a basic story of comparative advantage. If the government hires an american to make something, then that american is unable to spend that time and those resources to make something else. If the government is hiring an american who charges a higher price to make that thing than the price a foreigner is offering, then we know that the American is not being employed in his comparative advantage--the government has taken someone away from a task at which they are more productive and diverted them into something at which they are less productive. Domestic GDP actually declines relative to the case where the government hires the cheaper foreign contractor, as does total taxpayer wellbeing.

Now, of course when there is a shortfall in aggregate demand, the story changes a bit. Since there are lots of involuntarily unemployed Americans, we can hire one to produce something without displacing them from any more productive pursuits. Doing anything is always more productive than doing nothing. This logic is where Keynes fathomed up his coal mine story, and where Krugman's alien invasion parable came from.

We can state this more abstractly. In an economy with an aggregate demand shortfall, prices are not meaningful indicators of comparative advantage. The true price of hiring an unemployed american in a liquidity trap is essentially zero, automatically less than or equal to whatever price a foreign worker could possibly charge. In this regard, the amount of foreign drain from government stimulus efforts depends not just on the fraction of each dollar we spend on foreign goods and services, but also on the degree of discrepancy between equilibrium and actual prices.