Does free trade lead to insufficient diversification?

The theory of comparative advantage, and the theory of increasing returns, both predict that free trade brings specialization. Michigan and Tennessee produce automobiles, but Delaware does not. I have heard free trade critics suggest that regions end up especially vulnerable and overspecialized.

Yes in economics virtually everything is possible, once you recognize that Giffen goods and upward-sloping demand curves cannot be ruled out. But I am not so worried about free trade leading to excess specialization.

First, individuals can buy equities from multiple regions or countries. There does appear to be an inefficient "home bias" when it comes to investing, but surely free trade is not at fault here. If anything, free trade should encourage investors to think more globally and to diversify more.

Second, individuals specialize, in response to trade, because specializing brings higher returns. When investors expand a firm or line of business they internalize risk-return trade-offs.

Third, we must consider global risk. Let us say that protectionism helps a country avoid a declining industry. The world as a whole does not gain. Those economic problems are simply shouldered by another country, and in a less efficient way, relative to free trade.

Fourth, risk is borne by individuals, not by countries per se. Having many different industries in a country does not by itself create safety for individual investors or workers. Who cares if France has its own movie industry, if you are losing your job in computer software? Contrary to what Dana Rodrik suggests, it is not necessarily economically safer to live in a large country.

I can see a potential political version of the argument. Free trade, by inducing specialization, might make tax revenue more volatile in a country. Still, tax revenue should be higher in absolute terms. And if revenue volatility is a problem, I would prefer fiscal responsibility over protectionism as the proper medicine.

I am indebted to Randall Parker and Arnold Kling for sharing some of their email correspondence on this topic with me.