Are the Prohibited Subsidies Categories Unsound?

This is from law professor Alan Sykes:

The prohibited subsidies categories are unsound

As noted, import substitution subsidies and export subsidies are automatically “specific” and prohibited under WTO rules (ASCM Article 3). Neither prohibition makes much economic sense.

Import substitution subsidies are those “contingent (in whole or in part) upon the use of domestic over imported goods” (ASCM Article 3.1[b]). A government might provide, for example, that farmers receive a cash payment for the purchase of domestically produced farm equipment but not imported farm equipment. Such a subsidy can disadvantage imports to be sure, distorting resource allocation, and also creating a negative terms of trade externality, as described in Section 1. The difficulty is that a nearly perfect substitute exists that is not prohibited—namely, a subsidy to the domestic producers of farm equipment. If the domestic producers do not export, it is possible to precisely replicate the market equilibrium under an import substitution subsidy with a properly calibrated domestic producer subsidy. What is the logic of prohibiting one but explicitly allowing the other (via GATT Article III[8][b])?

Export subsidies, also as noted in Section 1, will often produce positive net international externalities, even if they distort overall resource allocation. Likewise, when they arise in an environment where international trade volumes are “too small” relative to the free market ideal due to trade barriers, export subsidies can improve global resource allocation by expanding trade toward its efficient level. Once again, the logic of treating them as a prohibited category is not apparent.

What I like about the existing prohibited subsidies categories is they target subsidies used to favor domestic firms over foreign ones, either in the home market or in foreign markets.  Other subsidies might do the same thing, but these do it the most clearly and egregiously.  I assume the SCM Agreement drafters were thinking along the same lines to some degree when they created this category.

With regard to import substitution subsidies, the local content requirement aspect already violates GATT Article III:4, for reasons that I think are widely accepted.  As result, SCM Article 3.1(a) may not be that crucial, although it does provide a stronger remedy in the form of tighter implementation deadlines.  Basically, it's just an additional provision making clear how problematic local content requirements are.  (If you want to protect domestic producers from import competition, you need to do it transparently, with tariffs.)  Such subsidies could be reconfigured as domestic payments, meaning they would not violate Article III:4 because they fall under Article III:8(b), but they could still be challenged as actionable subsidies, and countervailed if there are exports.

For export subsidies, like other forms of intentionally favoring your own producers over their foreign competitors, it's an international relations nightmare, with proliferation unavoidable.  One country does it, and then in response other countries do it, and all you are left with is taxpayers subsidizing producers, causing massive economic distortions.  But governments know they have trouble resisting it -- even though it makes them worse off -- when others might do it, so they all commit not to do it.

More generally, producer subsidies are most likely to be protectionist or trade distorting; subsidies to consumers -- not that I'm advocating them! -- are a much better way to accomplish your policy goals without causing trade friction.  And consumer subsidies seem less likely to violate the SCM Agreement, which is probably not an accident.

I had further thoughts on all this a while back in this article.