"Buy American" debate

Over at the New York Times "Room for Debate" blog, a number of commentators take up the "Buy American" provision in the stimulus. There are some interesting points on both sides:

• Jagdish Bhagwati: "The buy-American provisions unravel previous trade agreements unilaterally and in violation to the concessions we made."
  
  
• Ha-Joon Chang: "Some people worry that this will lead to a 1930s-style all-out trade war. But in the short run, there is actually no danger of that. Now we have the World Trade Organization, the European Union and many other regional trade agreements that limit protectionism. Of course, in the longer run, if veiled protectionism continues, we run the risk of making a mockery of these agreements and destroying the global trading system....the solution to this problem should not be an adherence to the principle of free trade, which is not workable in practice anyway, but instead to establish a new international agreement that allows a transparent, forward-looking and time-bound protectionism as well as more infant-industry protection for developing countries. In other words, by allowing more protectionism now in a controlled way, we will be able to preserve the international trading system better in the longer run."

• Anne Krueger: "The buy-American measure in the stimulus package would do little, in part because few imports are used in construction projects. But the signal that it would send to other countries would invite protective measures to the detriment of American exports and employment. Once protectionist measures are adopted, they are difficult to remove. In the long run, choking off through protection the integration of the world economy reduces productivity and prospects for future growth of all economies. It does not make any sense to sacrifice longer term growth prospects for measures that, even in the short run, offer very little prospect except for the very few at the cost of many others."
  

• Robert E. Scott: "When the government buys steel for a bridge, for example, it has several objectives. Minimizing costs is one, but when the economy is in recession, there is added incentive to stimulate domestic employment. And when steel is purchased from a domestic producer the workers’ wages generate further spending, which supports yet more jobs in the domestic economy."

Bhagwati and Kreuger are ardent free trade supporters, while Chang and Scott support selective protectionism, particularly for developing countries.

I agree with Kreuger that this won't have a huge economic impact and I agree with Chang that this probably won't cause a trade war, given our international institutional structure. Chang is also right about the dangers of "veiled protectionism", but I disagree that infant industry protection is the right solution.

Scott's argument breaks down if there's a big difference between domestic and foreign components. As I've said before, if we pay a lot more for inputs to construction projects, we can't do as many projects and can't hire as many workers. My feeling is that this is a gift to domestic steel producers. It could ultimately end up costing jobs and impeding the stimulus.