Some numbers on buy American

This Policy Brief (pdf) from the Peterson Institute for International Economics looks at the number of jobs that the Buy America campaign will create in the US steel industry.

The Policy Brief estimates that the additional US steel production fostered by the Buy American provisions in the the US House of Representatives Bill will amount to around 0.5 million metric tons. This in turn translates into a gain in steel industry employment equal to roughly 1,000 jobs. I guess the good news from that is that only 1000 jobs will be lost elsewhere in the US economy.

The reason the job impact is small because steel is very capital intensive and thus not many workers are needed in steel production. In fact, only around 150,000 workers - out of a total US workforce of around 140 million - are employed in the US steel industry as a whole.

The draft bill before the US Senate stipulates that, in addition to steel and iron, all manufactured goods used in projects financed by the stimulus plan must be produced in the United States. In this case the Policy Brief estimates that Buy American requirements will create a job gain of roughly 9,000 jobs. At this point you should note that given the size of the US labour force, as a percentage, the numbers being talked about are very small.

The Policy Brief goes on to say

In response to the Buy American measures, other countries would likely choose to echo US legislation by further restricting the ability of foreign firms to bid on public contracts. Such action—applied to lucrative new projects covered by their own stimulus programs—would raise additional barriers to US manufactured exports.

[...]

Moreover, foreign countries might extend their retaliation list to other US manufactured goods, especially if the Senate version of Buy American provision becomes law, since it covers all manufactured goods. In particular, foreign countries might cut off purchases of US products for public projects.

To deal with this possibility the authors of the Brief identified 12 major US trading partners and looked at government procurement spending in each of those countries as a percent of GDP. Applying that proportion to US exports of goods and services to each country, they estimate what share of those countries’ direct and indirect imports of US goods and services are the result of government procurement. The total value is around $104 billion. In their view, at least a small share of those exports are "at risk" of echo or retaliation measures. But even if as little as 1 percent of those exports were in fact lost by echo or retaliation behaviour, the resulting employment loss in the United States would be around 6,500 jobs. In an extreme case that 10 percent of those exports are lost, as many as 65,000 jobs could be lost.

The Brief summaries their results so far as

The negative job impact of foreign retaliation against Buy American provisions could easily outweigh the positive effect of the measures on jobs in the US iron and steel sector and other industries. The difference is that jobs lost would be spread across the entire manufacturing sector, while jobs gained would be concentrated in iron and steel and a few other industries.

Over the longer term, of course, the US economy will adjust and new jobs will be created so that the overall effect on the number of jobs will be zero. But if you take the above results as an indicator of the short term effects of Buy America, they don't look good. What is clear is that the US, and the world, will be poorer as a result of such protectionist measures.

But the Policy Brief goes further. It points out that with a Buy American requirement, the prices charged to public agencies would likely be higher for US iron and steel and other manufactured products. Higher prices would mean that fewer roads and schools could be built with the stimulus money. Of course such higher prices will also hurt steel using export industries such as heavy machinery. On the other hand, prices might fall for foreign steel sold by countries where the steel industry depends on exports to the United States, such as Mexico and Canada. And effect of this could be that private US buyers might in turn switch their purchases to those foreign producers. Depending on the size of the switch, the jobs created by the Buy American provisions in the short term could be significantly reduced by a loss of sales to private business in the United States.

None of this makes Buy American look any better.