Free Trade in North America: Some Observations
Paul Wonnacott* Alan R. Holmes Professor of Economics Middlebury College at Conference on Free Trade @ Ten and NAFTA @ Five McGill University June 5, 1999
John McCallum's background paper comes out about right. To date, the U.S.-Canadian FTA deserves two cheers. It has worked well, although some of the expected long-run results have yet to occur.While trade flows have increased in roughly the way that might have been anticipated following the FTA,[1] overall Canadian productivity apparently has not -- at least not yet. This is very important. The point of freer trade is not simply to increase trade, but to improve living standards. Trade can contribute to higher incomes in two ways -- by changing the composition of production between sectors according to comparative advantage, and by increasing productivity within sectors. In the run-up to the free-trade negotiations, productivity improvements were seen as the key to large Canadian gains.
For supporters of free trade, the difficult puzzle is that a number of studies indicate that the U.S.-Canadian labor productivity gap in manufacturing has increased since the free-trade agreement -- although at a very much slower rate than in the previous decade. Because productivity is so important, I want to comment on it, even though my observations will partly overlap those of John McCallum.
One answer to the productivity puzzle may lie in difficulties in the data, which are considered in detail by Daniel Trefler. He concludes that there are, in fact, major data problems in comparing trends in U.S. and Canadian productivity; the basic U.S. and Canadian productivity data, which have been at the center of the debate, are, in his words, "completely incomparable" (p. 5, italics his). When calculated correctly, Canadian labor productivity in manufacturing grew 0.6% more per year than in the United States in the 1989-95 period -- a narrowing of the productivity gap that he attributes entirely to the FTA (Trefler, p.1).
A second study that throws light on the productivity puzzle is that of Donald Daly, who focuses on differences between large and small plants. For large plants -- those with over 500 employees -- Daly finds that Canadian productivity has improved more rapidly than overall U.S. manufacturing productivity -- a result "in line with the studies showing larger gains from freer trade" (p. 12). The Canadian productivity lag lies in the smaller plants. While this conclusion is somewhat reassuring to those who expected large Canadian gains from free trade, it is also somewhat disconcerting, given the contributions which smaller firms can make to economic dynamism and employment.
A third approach to untangling the productivity puzzle is to look at sectoral trends in productivity. In a paper delivered last weekend at the Canadian Economics Association, Daniel Schwanen presented evidence that Canadian productivity would have increased by 0.4% more per annum since the late 1980s if Canada had had the same sectoral composition as the United States.
One of the most important sectors, of course, is the automobile industry. Because of the AutoPact of 1965, the U.S.-Canadian productivity gap had been substantially closed by the end of the 1970s, and recent Canadian improvements in productivity in this industry have been similar to those in the United States. However, productivity increases in automobiles have lagged behind industry in general. Because autos are more important as a percentage of Canadian than U.S. manufacturing production, the result has been a drag on overall Canadian productivity compared to U.S. productivity.
The sectoral issue is reinforced by the fact that Canada has lagged in ultra high-tech industries where U.S. productivity growth has not only been spectacular, but seems to be accelerating. Canada has probably been hurt in the competition for high-tech industries by the migration of the computer-related industries toward the South and West, most notably to California and Washington states. This seems to me to be rather close to a pure economic accident. Geography was not the cause of the collapse of Wang Laboratories, nor of the relative decline of the computer companies based in New York, Minnesota, or Michigan. At the starting gate, the MIT and Stanford areas were essentially equal; geography counted little. But, even though the industry was footloose at the beginning, it is less so now; there are strong benefits in going where the action is. And now the action is, most importantly, in the San Jose-San Francisco area.
Even though I would classify the migration of the computer industry as primarily an historical accident, there has also been a broader, more general shift of the center of gravity of the U.S. economy to the south and west, away from the Chicago-Boston-Baltimore triangle which is so close to the major Canadian manufacturing regions. [For example, upstate New York population and labor force were lower in 1998 than in 1990 (Deitz and De Mott, 1999.)] In other words, the center of Canada's population is now further away from the core of the North American economy than it was several decades ago. This may be a continuing negative factor for the Canadian economy.
Although rising living standards are built on rising productivity, the tie between productivity and real wages is not tight in an open economy. Productivity gains may go either to the producing country (in terms of higher incomes) or to the importing country (in terms of lower prices), or a combination of the two. While the composition of Canadian output -- relatively large in autos but small in high tech -- has acted as a drag on Canadian productivity, its effect on relative Canadian real incomes has been much less clear. Automobiles -- and particularly automobile assembly -- are a high-wage sector, providing good incomes for many Canadians with moderate skills. Although there are lots of millionaires in Seattle, most of the gains in the high-tech industries have gone to buyers, in the form of precipitously falling prices. Incidentally, high tech is one place where protection simply doesn't work. A country may get along with a protected automobile industry that is ten years out of date, but producing computers that are ten years out of date is a waste of time.
In terms of the important issue of assured market access, one of the notable NAFTA contributions came in the wake of the 1994-95 Mexican crisis. Mexico responded with devaluation and tight monetary and fiscal policies, rather than the trade and capital controls so often used by developing countries in trouble. Although the crisis of 1994-95 was extremely painful to Mexico, it was conspicuously shorter than the prolonged crisis of the 1980s. By 1996, the Mexican economy was growing again -- by more than 5%. NAFTA probably contributed to the speed of the recovery, although the Mexican economy also benefited from the reforms of the 1980s. There is just one negative note in this favorable picture: NAFTA may have increased the pre-crisis complacency of Mexican officials, by reinforcing expectations of a continuing capital inflow.
Finally, I note that one of the fears of the critics of a free-trade agreement has conspicuously failed to come true. Canada didn't hitch its economy to a falling star. Since 1982, the performance of the U.S. economy has been very strong. Let us -- in all three countries -- hope that this strength continues.
Footnotes* I have benefitted from discussions with Donald Daly, Jeffrey Schott, Daniel Schwanen, and Ronald Wonnacott.To Text.
1. Although Treffler (1999), p 2 concludes that "the recent explosion of Canadian trade . . . was not primarily driven by specifically FTA tariff cuts." He notes (p. 28) that "Most of Canada's increased trade was in industries that had no tariffs in 1988." Clearly, the long, strong expansion of the U.S. economy in the 1990s was a major cause of the increase in U.S. imports from Canada, Mexico, and elsewhere.To Text.
BibliographyBergsten, C. Fred, and Jeffrey Schott, "A Preliminary Evaluation of NAFTA." Testimony before the Subcommittee on Trade, Ways and Means Committee, US House of Representatives, Sept. 11, 1997.
Daly, Donald J., "Small Businesses in the Canada-U.S. Manufacturing Productivity Gap." Toronto: Schulich School of Business, York University, xeroxed, rev., March 1999.
______, B.A. Keys, and E.J. Spence, Scale and Specialization in Canadian manufacturing. Economic Council of Canada, 1968.
Deitz, Richard and Mike De Mott, "Is Upstate New York Showing Signs of a Turnaround?" Federal Reserve Bank of New York, Current Issues, May 1999.
Fuss, Melvyn A. and Leonard Waverman, Costs and Productivity in Automobile Production. Cambridge University Press, 1992.
McCallum, John, "Two Cheers for the FTA." Royal Bank of Canada, xeroxed, June, 1999.
Schott, Jeffrey, "NAFTA: An Interim Report." Washington: Institute for International Economics, xeroxed, Dec. 1997; forthcoming in a volume to be published by the World Bank.
Schwanen, Daniel, "Trading Up." Toronto: C.D. Howe Commentary #88, March 1997.
Trefler, Daniel, "The Long and Short of the Canada-U.S. Free Trade agreement." University of Toronto, Institute for Policy Analysis, xeroxed, rev., Feb. 1999.
Wonnacott, Ronald J. and Paul Wonnacott, Free Trade between the United States and Canada: The Potential Economic Effects. Harvard University Press, 1967.
Beginning of Paper
Return to List of Papers
Return to Econ Faculty page