Protectionism and Tariff Wars

Economists and philosophers have argued for centuries about the costs and benefits of free trade versus protection. Scarce, however, are early discussions of the macroeconomic consequences of protection, and fewer still are arguments sufficiently well-articulated to distinguish them from their microeconomic antecedents. During the interwar period, however, the flow of intellectual history and commercial policy changed course.

John Maynard Keynes (1883–1946), one of the founding fathers of modern macroeconomics, changed his views dramatically. As late as 1923 he was as ardent about free trade as Adam Smith (1723–1790), but by the early 1930s he began advocating tariffs as a remedy for British unemployment. Keynes argued that a tariff would be stimulative in the aggregate if wages were downwardly rigid, the exchange rate was fixed, and labor was underemployed. In response, Lionel Robbins, T. E. Gregory, Arnold Plant, J. R. Hicks, and others had collaborated on a book that claimed that classical theory remained operative under conditions of unemployment. Keynes was unmoved. In 1951 Hicks conceded a great deal, concluding that free trade had been called into question by the theoretical changes brought about by Keynes.

MACROECONOMICS OF THE GREAT DEPRESSION

The debate on the macroeconomic effects of commercial policy continues to this day, in part driven by a desire to search for a complete explanation of the Great Depression. The focus of macroeconomists on this episode is easy to rationalize: from the period 1929 to 1933 real gross national product (GNP) in industrialized countries fell by 18 percent and export volume fell by 35 percent. The combination of these two facts—the dramatic worldwide slump and the greater relative decline in trade than in income—has led economists to seek international explanations for the economic trauma of the 1930s.

The escalation of tariff levels during the late 1920s and particularly during the 1930s is one such explanation. Many accounts of tariff history cite the infamous Smoot-Hawley Tariff Act of 1930 as an accelerant to the evolving move toward higher international levels of protection. However, some of what appears to be retaliation may in fact be a domestic response to falling output under the assumption that duties would lift the economy out of recession, or at least help the most ailing industries. Senator Reed Smoot (the chief architect of the 1930 tariff bill in the United States) had considerable conviction that the U.S. tariff increase would help: "How can we escape the conclusion that the tariff has been a stabilizing influence? In the scramble of all nations to dispose of surpluses [the tariff] has saved American producers from a deluge of cheap foreign goods" (Smoot 1931, p. 181).

The figures below show the path of U.S. output and tariff levels from 1900 to 1940 along with aggregates of the same variables for major U.S. trading partners (in order of importance for U.S. imports, the trading partners are: Canada, the United Kingdom, France, Germany, and Italy). The correlation between U.S. output and either U.S. or foreign tariff levels is strongly negative; the same is true of the correlation between foreign output and the U.S. tariff. The only correlation that is positive is between foreign output and foreign tariffs, and this is true only during the period 1920 to 1940.

Taken at face value, the correlation analysis suggests that both the United States and its major trading partners were harmed by U.S. tariff increases, whereas major U.S. trading partners benefited from the imposition of their own tariffs at the expense of the United States. In a nutshell, this is the general ambiguity of the channels through which tariffs impact aggregate economic activity. Theoretical models and sophisticated empirical exercises offer some hope of sorting through this ambiguity.

Trade theory generally predicts that free trade is optimal because it results in a more efficient allocation of the factors of production to their respective comparative advantage, while also providing consumers with price signals that reflect the opportunity costs of production. Tariffs, in contrast, encourage firms to continue to produce goods that could be produced more cheaply abroad. Moreover, the allocation of resources to protected sectors leaves other sectors of the economy with fewer inputs, driving up their costs. The fact that international trade and output collapsed at a time when global tariff levels escalated is often cited in textbooks as evidence in favor of this classical view that free trade is better for macroeconomic performance. Macroeconomic models built upon a similar theoretical structure to traditional trade models will, not surprisingly, produce qualitatively similar results.

OPTIMAL TARIFFS

There is latitude in trade theory for a rise in the tariff to increase domestic income or welfare. In fact, there are multiple mechanisms that could work in this direction. The most famous is the optimal tariff argument that states that for a sufficiently large country, a tariff increase from a position of free trade may actually increase both output and welfare. The first article to discuss optimal tariffs and retaliation was Harry Johnson's (1953). The logic is similar to that of a monopolist charging a price above marginal cost in order to raise profit levels. The major trading partners of the United States were certainly a large enough block to exert an influence over the terms of trade with the United States. One possible explanation for the positive correlation between the foreign tariff levels and foreign output during the interwar period is that the United States was sufficiently small on the world stage and therefore was harmed by its own tariff, but its trading partners were sufficiently large that they benefited from the imposition of their tariff.

Another theoretical argument that delivers a positive impact of a tariff increase on output is the "expenditure-switching" effect. By raising the relative price of imports compared to exports (i.e., foreign goods relative to domestic goods), the tariff induces a shift in expenditure away from foreign goods toward domestic goods, raising domestic demand. At the level of an individual good, this result follows without much controversy, but when applied at the macroeconomic or world level, consideration must be given to feedback from the tariff to foreign demand for domestic products.

The fragility of this theoretical result is well exposited by Jonathan Ostry and Andrew K. Rose (1992). Moreover, they conducted an empirical analysis using data from the Organization for Economic Cooperation and Development from the 1970s and 1980s to find that tariffs appear not to have a statistically significant impact on the level of output, the trade balance, or the price of domestic goods relative to foreign goods. Corroborating this evidence in historical context, Douglas Irwin (2000) conducted a very careful study of the United States in the late nineteenth century and found that although growth was rapid under the high tariffs that existed at the time, the tariff is not needed to account for that growth experience.

In a study of extraordinary scope, Michael A. Clemons and Jeffrey G. Williamson (2001) examined the linkages between tariff policy and economic growth from 1865 to 1990 across a group of countries. They found that growth and tariffs are negatively associated after World War II, consistent with the findings of Jeffrey D. Sachs and Andrew Warner (1995); this suggests that tariff increases are harmful to the initiating nation. However, they also found a positive association in the period before World War II, suggesting exactly the opposite interpretation. They attributed the switch in sign to changes in the world economic environment (lower tariffs and transportation costs and higher economic growth generally in the period after World War II).

The type of reduced-form evidence marshaled in these papers is sometimes criticized because the economic model that forms the basis for the investigation is evaluated indirectly. The result may be ambiguity as to what effects are being identified. Are the effects of tariffs truly being isolated from the effects of other economic shocks that impinge upon the aggregate economy? The answer is generally uncertain.

An alternative quantitative approach involves writing down a specific economic model restricted by data and then using a simulation method to infer the implications of the model for various macroeconomic outcomes. Two studies have employed this type of analysis to estimate the impact of tariffs and retaliation in the context of the U.S. depression during the 1930s. These studies place the quantitative impact of the tariffs at about 2 percent of U.S. GNP, but claim different signs for the effect. Mario J. Crucini and James Kahn (1996) emphasized the role of tariffs in distorting capital accumulation and input choices (many traded goods were intermediate inputs into final production) and got a negative impact. Basically, by reducing the efficiency of factor allocations, the tariff war shifted the world production frontier inward, reducing output at home and abroad. Barry Eichengreen (1989) argued that U.S. tariff increases shifted expenditure toward U.S. goods and away from foreign goods, giving rise to a positive impact. His analysis assumed that nominal wages are fixed, and he ignored the negative supply-side impact of tariffs due to production inefficiencies. As a result, an increase in the U.S. tariff, by increasing the price level, would evoke—at fixed nominal wages—a positive supply response because the real product wage falls. Under a unilateral tariff increase, demand also rises due to expenditure switching. Under retaliation by the foreign country the demand-side effect is neutralized, but a foreign supply expansion follows by the same logic described above. Thus, one set of modeling assumptions attributes a role for tariffs in accounting for the economic slump; the other attributes a role for tariffs in preventing things from getting even worse.

The advantage of the simulation approach is that the channels through which tariffs affect various economic variables is transparent; a disadvantage is that models typically have broader economic implications that are rejected by the data, limiting the confidence one has in their predictions. Thus, whether one places a great deal of theoretical structure on the exercise or not, in practice, the results are debatable.

Another practical difficulty in estimating the size or even the sign of the tariff's impact on the macroeconomic level is the fact that so much else is changing at the same time. In the U.S. case alone, there was a monetary contraction, massive bank failures, crop failures, and a stock market crash. Sorting out the causal role of tariffs in the context of a multitude of alternative causes has proven a daunting empirical task, and much remains to be done.

An important role for academic economists is to sift through an extensive body of evidence, possibly taking quite distinct theoretical approaches, and arrive at an informed scientific opinion on the matter at hand. There does appear to be considerable academic agreement on the role of commercial policy during the Great Depression. In a survey of economists and historians conducted by Robert Whaples (1995), two-thirds of both groups generally agree with the statement: "The passage of the Smoot-Hawley Tariff exacerbated the Great Depression." Another 20 percent agree with (unspecified) provisos when asked the same question. It is quite likely that similar numbers would prevail on the general question of the merits of free trade relative to protectionism. One is tempted to conclude that the combination of inconclusive empirical evidence on the macroeconomic effects of commercial policy combined with the strength of the theoretical arguments in favor of free trade has led to agreement among most economists on the merit of free trade.

SEE ALSO Import Substitution;Protection Costs; Tariff Theory.

BIBLIOGRAPHY

Crucini, Mario J., and Kahn, James. "Tariffs and Aggregate Economic Activity: Lessons From the Great Depression." Journal of Monetary Economics 38, no. 3 (December 1996): 427–467.

Eichengreen, Barry. "The Political Economy of the Smoot-Hawley Tariff." In Research in Economic History, vol. 12, ed. Roger Ransom. Greenwich, CT: JAI Press, 1989.

Irwin, Douglas. Against the Tide: An Intellectual History of Free Trade. Princeton, NJ: Princeton University Press, 1996.

Irwin, Douglas. "Tariffs and Growth in Late Nineteenth Century America." World Economy 24 (January 2001): 15–30.

Johnson, Harry. "Optimal Tariffs and Retaliation." Review of Economic Studies 21, no. 2 (1953): 142–153.

Ostry, Jonathan D., and Rose, Andrew K. "An Empirical Evaluation of the Macroeconomic Effects of Tariffs." Journal of International Money and Finance 11, no. 1 (February 1992): 63–79.

Sachs, Jeffrey D., and Warner, Andrew. "Economic Reform and the Process of Global Integration." Brooking Papers on Economic Activity 1 (1995): 1–118.

Smoot, Reed. "Our Tariff and the Depression." Current History (November 1931).

Whaples, Robert. "Where Is There Consensus Among American Economic Historians? The Results of a Survey on Forty Propositions." Journal of Economic History 55, no. 1 (March 1995): 139–154.

Mario J. Crucini