American Steel

Alarmist media reports on “Buy America” rules for steel used in US public infrastructure projects have emphasized the value of Canadian steel exports allegedly threatened, but have largely ignored the similar value of American steel imported by Canada. In fact, in the most recent month for which data is available (November 2008), Canada bought more steel from the US than it sold to the US.

The high degree of integration, and balanced trade, between the Canadian and American steel industries provides a compelling rationale for a Canadian exemption from “Buy America” requirements. Such an arrangement would not only allow continued Canadian steel exports to the US, but protect such exports from third-country competition for US infrastructure contracts.

Canadians should keep their eyes on this prize. Unfortunately, rather than working co-operatively and practically for an exemption, Canadian politicians (from both sides of the new Conservative-Liberal coalition) have been publicly lecturing Americans about their “international obligations” and the theoretical virtues of global free trade. This argument is not correct in the current economic context and certainly will not be very persuasive south of the border.

Free Trade vs. Procurement Policy

In a supply-constrained world where all capital and labour are fully employed, free trade may be beneficial because it allows each country to specialize in producing the product(s) for which it has a comparative advantage. (Even in this case, the distribution of benefits between countries or between capital and labour can provide rationales for regulating trade.)

However, in a demand-constrained world with excess capacity and unemployed workers, countries have a clear-cut interest in locating production within their own borders. The orthodox argument for free trade, which assumes full employment, is unlikely to convince Americans (or anyone else) at a time of rising unemployment.

As well as the theoretical argument for free trade, Canadian cabinet ministers and both national newspapers have cited the historical example of the Smoot-Hawley Tariff “causing” the Great Depression. But the Smoot-Hawley Act’s worst feature was that it did not provide (or accompany) significant fiscal stimulus. In effect, the US tried to use tariffs alone to grab a larger slice of a shrinking global economic pie.

By contrast, the “Buy America” provision of the American Recovery and Reinvestment Act must be evaluated in the context of the US funding proportionally more stimulus than other countries. The world economy needs significant stimulus from all (major) states. Mechanisms to prevent countries from “free riding” on stimulus financed by other governments – and thereby encourage them to contribute their own stimulus – may prove vital to a global economic recovery.

Today’s “Buy America” requirement is a relatively efficient mechanism. Unlike Smoot-Hawley, it would have no effect on imports for private ventures or personal consumption in the US. Instead, it ensures only that publicly-funded infrastructure will mostly be built with American-made inputs.

Whither a “Buy Canadian” Policy?

If the stated goal of increasing spending is to stimulate the domestic economy, it is reasonable for governments to ensure that a reasonable share of that spending remains within the domestic economy. The advent of stronger “Buy America” rules underscores the absence of any attempt at “Buy Canadian” policy in last week’s federal stimulus budget.

Indeed, the budget eliminated Canada’s few remaining tariffs on imported machinery and equipment while providing a 100% capital cost allowance for computers. Both measures seem designed to encourage corporate Canada to purchase more inputs from overseas.

In particular, offshore steel should be of as much concern in Canada as in the US. In recent years, Canada has exported about $1 billion of steel overseas and imported between $5 billion and $4 billion of steel from overseas. In 2007 and early 2008, rising shipping costs modestly reduced oversees steel imports.

But because of a recent collapse of shipping prices, increased infrastructure investment could easily spur an increase in steel imports rather than in Canadian steel production. In other words, there is a real risk of stimulus spending leaking offshore through Canada’s glaring steel trade deficit.

Canadian Exceptionalism

Canadian government officials and media pundits have implied that eliminating “Buy America” provisions is their goal, but that a specific Canadian exemption might be an acceptable compromise. In fact, “Buy America” rules that exempt Canada are the best-case scenario. Canada would be able to sell into the entire American market with less competition from third countries for government contracts.

If we had “Buy Canadian” policies, our bargaining position would be much stronger. In steel and other sectors characterized by relatively balanced trade with the US, we could offer Americans an exemption from our requirements in exchange for an exemption from theirs.

Even without any such bargaining, Canada has fared reasonably well. When the US applied tariffs to foreign steel in 2002, it exempted Canada and Mexico. This success was partly a result of steelworkers on both sides of the border being represented by the same union. It helped that this union was (and is) led by a man that The Toronto Star describes as “the most influential Canadian-born Democratic power broker in the United States.”

A closer look at North America’s steel trade provides a further explanation of why Americans have little incentive to target Canadian steel:

Canadian Trade with the US from Iron and Steel Mills

($ million; NAICS Code 3311)





















 Nov. 2008





Canadian Trade with the US in Products from Purchased Steel

($ million; NAICS Code 3312)





















 Nov. 2008





Total Canadian Steel Trade Balances

($ million; NAICS Codes 3311 and 3312)
















 Nov. 2008



*Excluding December 2008, for which data is not yet available

Typically, Canada buys more from American iron and steel mills than the US buys from Canadian mills. This pattern reversed dramatically in several months of 2008 (Canadian mills ran a remarkable $250-million surplus in June and July).

But these months were an aberration rather than a trend. American mills have since regained their surplus. Limiting such trade would worsen the US trade deficit.

Turning to steel products made from purchased steel, Canada typically sells more to the US than we buy from it. Since Canada makes some of these products with steel imported from the US, this trade hardly poses a threat to the American steel industry. (Indeed, in the most recent month for which data is available, the US actually ran a trade surplus with Canada in products made from purchased steel.)

The main story is North American integration. In contrast to glaring offshore steel trade deficits, steel trade balances between Canada and the US are very small relative to the volume of steel trade. Canada’s net surplus with the US has typically equalled only 2% of the value of total Canadian steel exports to the US. Even in 2008, it did not exceed 10%.

The balanced, integrated nature of Canada-US trade in steel provides a compelling rationale for a Canadian exemption from “Buy America” requirements. Canadian officials and editorialists need to start telling this story rather than hectoring Americans about the theoretical virtues of free trade and the evils of Smoot-Hawley.

UPDATE (Feb. 2): Around the same time that I posted this commentary, Paul Krugman posted essentially the same economic justification for “protectionism” to address the international free-rider problem associated with fiscal stimulus. Do I dare say “Great minds think alike” or would that just invite critics to add that “Fools never differ”?

UPDATE (Feb. 3): Quoted by The Toronto Star

UPDATE (Feb. 5): An abridged version of this commentary is printed in today’s National Post (page A17).


  • Please tell me that this is not your proposal:

    a) seek an exemption for Canadian steel makers from US protectionism.
    b) let other exporters suck it up.
    c) let Canadians consumers suffer the consequences of a ‘buy Canadian’ policy.

  • The Pittsburgh Steelers just won the Super Bowl, which must be a good omen for us.

    Your summary of USW’s policy is fairly accurate. While I appreciate that you may not share our concern for North American workers, even from a theoretical/global perspective, we need to address the international free-rider problem with respect to fiscal stimulus.

    On point (c), consumers would not be directly affected by a “Buy Canadian” policy for government procurement. Taxpayers might be exposed to somewhat higher costs, but these could be explicitly limited to a given percentage (as the “Buy American” proposal does). If a major purpose of the infrastructure spending is to stimulate the Canadian economy, then it makes sense to spend a few extra dollars to keep the money within Canada.

  • As you correctly point out, local procurement preferences are not trade measures in any conventional sense, and do not impose any restrictions on the trade in goods across international borders. Rather “Buy American” rules simply condition public procurement spending to foster local economic development – a key purpose of present stimulus plans.

    Contrary to much of the current commentary on present US proposals concerning steel procurement, the right of state, provincial and local governments to adopt local preferences is specifically preserved under both NAFTA and WTO rules. It is long past time for Canada take advantage of this right to follow the US precedent with a “Buy Canadian Act”.

    But the right to impose local preferences doesn’t mean that such policies are sound in every case, particularly in an integrated north american market. As you point out, there are good reasons for Canada and US to collaborate on devising a rational approach for managing North American steel trade. It may even make sense for them to consider broadening the frame of economic reference to include other nations.

    The important point is that a managed, not de-regulated trade program is needed to optimize the beneficial economic gains from public procurement, particularly in the present context. Simply abandonning such key decisions to the market makes no more sense for the industrial sector than it did for banking or financial services.

  • Erin, you seem quite proud of the USWA-driven exemption of Canada and Mexico from American restrictions in steel trade. Do you think it’s a good thing that US-based vested interests lobbying Washington directly is the best way to advance Canada’s interests?

    On the topic of using “Buy Canadian” policies as a bargaining chip, do you honestly think it would have been a good idea to build that kind of momentum within Canada and then try to turn on a dime when you can cut a deal with the US?

    I think it’s one thing to be a nationalist, and another to be a globalist, but Canadians going to bat for US protectionism is just plain weird.

  • As we have made reference to on this Blog several times, the American Stimulus package will have a great effect on the Canadian economy.

    Given the limited spending, anti-worker, regressive, pro market nature of our own feebly inept stimulus package, the Americans are most likely not to thrilled about our free rider stance.

    The whole space within the test, “Made in USA”, has plenty of room to exclude Canada. However there was a price to open up that space, and our Harper government failed to capture that space, and we are now in a diffcult space. We have no bargaining power.

    And instead of realizing this we have the likes of the Stockwell Day caught up in the “lost world of the Reform party”, yelling and screaming at the US, in some whining tantrum with the EU.

    Not only do have these tories lost the ticket to get into the game, they are now pissing off the gate keeper.

    Thankfully we have the Steelworkers, the Paperworkers, the Autoworkers, and public sector workers who can most likely help ease our pain within this space.

    Its bad enough the tories are demanding our autoworkers take a pay cut, they are now seeking to wreak havoc own our Steel industry.

    Somebody quickly shut up Stockwell Day, before he costs thousands of Steelworkers their jobs.


    Good article Erin, I lost you email address (cpu died) could you send me a copy of he longer document regarding this.

  • P.S. I meant to underline the statement

    our stimulus package is mainly perceived by the US as a free rider stimulus package, which means we blew it.

    Flaherty can package his budget nicely for the Canadian public but the American are not idiots.

    I am not saying our stimulus would have greatly affected the US economy, but you can bet your ass, given the political overtones of such an anemic stimulus pissed off Obama and friends.

  • THis whole meandering within the ideology of the right, practiced so proficiently by Haper, is going to start costing us dearly.

    For Iggy not to do something about this lame assed budget will be the cause fo a lot of despair for workers across the country.

    We should have been spending proportionately more than the Americans, as the pay off would have been tremendous as we would not have had to worry about this whole potential redefining of “Made in USA” which at one point also included Made in Canada due to the integrated nature of our markets. Yes they are still integrated, but there is room for this policy to separate the wheat from the chaff along nationalistic lines.

    We might have saved a few billion on this budget but it is going to cost us many billions in “free rider” retribution. Game theory predicts it, do the math if you like.

    sorry I am ranting

  • This isn’t the same argument as Krugman’s. Krugman concedes the comparative advantage argument for free trade. Krugman therefore must concede that stimulus packages with buy domestic provisions are, or at least can be, inefficient. Indeed, he admits suboptimality. But he insists that we consider the possibility that the argument for demand management is so strong, government should spend in a downturn even if governments can buy more with the same spending at other times.

    The argument here, on the other hand, suggests that buy domestic provisions lead, or at least can lead, to optimality.

  • I have a hard time accepting Krugman’s concession, but that is another debate on labour standards and a whole array of measures.

    So pushing that aside for the moment, I do think one could argue somekind of hybrid optimality for an exclusionary policy on Canada from the mad ein USA policy, that would make Krugman’s claim false.

    It focuses on Canada-USA, not North America as I have not factored in Mexico yet.

    Basically the higher the degree of integration between two economies, with a concurrent degree of homogeneity in labour market regulations one could argue a global maxima exists above the domestic case which Krugman argues. It is the correlation between economic output and integration, and demand management that I would rest this hypothesis. A large mature, integrated relationship such as Canada-USA, with consideration to demand management being considered by Obama would result a situation where inefficiency overtakes output gains versus demand leakage. It is a bit difficult to draw up here, but I think it is clear. THe more you try and fix demand leakeage with a made in USA policy, given the nature of our relationship brings about inefficiencies at least fro a time, that is until capital flight can resolve the situation in the longer run. But that disruption in the short to medium run would make any such demand policy highly disruptive.

    This also is highly dependent on the fact that we Canadians have a small open economy and a majority of the so called leakage stimulus, is actually fictitious leakage, as most of it goes back south.

    So in the case of Canada, I would argue against Krugman’s optimality of demand management and made in USA.

  • The gains from trade, in standard theory only hold in the context of the full employment of resources. In fact Krugman knows this it was partially the basis for his faux prize.

    I think Paul T. is onto to something when he says that the US understands that Canada and much the rest of the world’s stimulus amounts to a beggar thy neighbor policy.

    A strong position would be that Canada had implemented a stimulus package in excess of the rest of the world (proportionally) and therefore the buy American policy was parasitic.

    Imagine the case where the US state adopted the laissez faire position. In that case there would not be any stimulus domestic or otherwise. Why should US citizens pay the bill for the rest of the world?

    The only thing the amendments to the stimulus package do is insist that the extra money being spent benefits Americans. The argument being that no stimulus is an option and as such that none would consider the “no-stimulus” option as being protectionist So how does a stimulus option become protectionist regardless of the conditions? The conditions are simply whatever money the government injects ought to be directed toward domestic stimulus. That is the conditions are not that general tariffs be erected only that new money from the GOVERNMENT be consumed domestically.

  • It is good to hear that domestic procurement rules are exempted from NAFTA. However, until recently didn’t progressives mostly argue that NAFTA should be torn up? It seems strange that we would point to the exemption in a conservative-mandated deal and say “Yes, well, they got THAT right. Behold, sound trade policy! Oh, and investor state is still bad.”

    I think everyone is wading in too far on mixed messages, views that superficially contradict, two-word slogans that are “in code”, and cozying up with people who are sincerely protectionist. I predict this will not end well. People will support the free-trade message because it is clearer, more consistent, and more closely adheres to the high school lessons about why the Depression lasted so long.

  • There is nothing mixed here. The gains from trade are contingent upon full employment–that is standard trade theory . Stimulus is an addition to what the private sector left to its own calculus will do. How can SUCH stimulus be counted as protectionist? Protectionism is interventionism and stimulus is interventionism. That is… the muddy water.

  • I haven’t come across this claim that the comparative advantage argument only holds when an economy is at potential. Makes no sense to me. Growing bananas in Canada instead of trading for them is idiocy regardless of the level of Canadian employment or output at any given time. Moving supposedly idle labour and capital inputs into banana growing still has the opportunity cost of not being moved into something that doesn’t involve constructing unnecessary greenhouses.

    That said, I’ll grant that Krugman seems to think operating below potential DOES excuse govt spending with a low expected return on investment given that he thinks governments may indulge domestic monopolies instead of more competitive foreign bidders in such times.

  • Fine show me a standard trade model in which there is 7 % unemployment and excess capital and which demonstrates greater gains from trade then could be generated by achieving full employment at home.

    Moreover, some version of your resource wasting example occurs under free trade. Raw logs trucked out of BC, milled in the US, then exported over-seas, and then re-imported as furniture back into Canada at Vancouver and then sent by rail across Canada to eastern markets.

    And none of this answers the 800 billion dollar question of why an optional program of stimulus that is new (interventionist) money with local procurement rules should be viewed as anti-free trade and a policy of *no* stimulus would be called–besides irresponsible–what?

    The real issue, it strikes me, is the underwhelming, under-thought, over-porked, and light-weight stimulus package AKA the Cons budget.

  • Brian, when the economy is at full potential, protectionism arguably shifts resources from comparative-advantage industries to comparative-disadvantage industries. When the economy is below full potential, protectionism arguably shifts resources from unemployment into industries that may have a comparative disadvantage. But having resources sub-optimally employed is better than having them unemployed.

    Krugman articulates the case for protectionism in the short-term. He stops short of endorsing it because it could be difficult to reverse when the economy returns to potential in the long-run: “there is an economic case for it, but also that there is a strong political economy case (which I consider dominant) against acting on that economic case.”

    Bananas are an extreme example of comparative advantage because they require a specific climate. Buy America applies not to bananas, but to steel. Comparative advantage is a more nebulous concept because steel can be produced efficiently anywhere that has iron and water. (In fact, one of Canada’s more successful steel mills is in Regina, far from either natural input.)

    To the extent that steel production is subject to economies of scale, temporary support for a country’s steel industry (such as Chinese subsidies or Buy America stimulus) can conceivably give it an enduring advantage by shifting it down the cost curve.

Leave a Reply

Your email address will not be published. Required fields are marked *