Trade Barriers and the Collapse of World Trade During the Great Depression.

• Author: Madsen, Jakob B.

Jakob B. Madsen [*]

Using panel data estimates of export and import equations for 17 countries in the interwar period, this paper estimates the effects of increasing tariff and nontariff trade barriers on worldwide trade over the period 1929 to 1932. The estimates suggest that real world trade contracted approximately 14% because of declining income, 8% as a result of discretionary increases in tariff rates, 5% owing to deflation-induced tariff increases, and a further 6% because of the imposition of nontariff barriers. Allowing for feedback effects from trade barriers on income and prices, discretionary impositions of trade barriers contributed about the same to the trade collapse as the diminishing nominal income.

1. Introduction

   The contraction in world trade during the first phase of the Great Depression stands out as the strongest adverse shock to international trade in modern history. From 1929 to 1932 world import and export volume in the industrialized nations decreased about 30%. However, it is not well understood which factors were responsible for the collapse. The factors that have been highlighted in the literature are declining demand, escalating tariff and nontariff trade barriers, increasing bilateral trade agreements, and international exchange rate policies. The importance attributed to each of these factors has often been controversial. Pollard (1962, p. 200) for instance argues that for Britain the "fall in total foreign trade as a proportion of home production was a part of a secular trend, and may well not have been caused by the tariff as such." By contrast, Khan (1946, p. 246) claims that, within 12 to 18 months, UK nominal imports of manufactures from most of Europe and the United States were reduced by something like 60% "as a result of the tariff." Similarly, Saint-Etienne (1984, p. 29) argues that "by the mid-1930's, international trade had become, in large proportion, barter trade" as a result of the tariffs and nontariff barriers.

   Empirical studies have examined the effects of trade restrictions on incomes to explain the declining trade for individual countries. The studies of Crucini and Kahn (1996) and Irwin (1998) find that the tariffs were influential for the U.S. imports and exports. In his study of nominal imports to the European countries, Friedman (1974) quantified the effects of trade barriers on nominal imports, and ultimately income, by means of dummies in periods of significant tariffs and nontariff barriers. He found that trade restrictions had a significant impact on trade in a few countries. However, as Friedman himself acknowledges, the weakness of this approach is that strong and weak forms of trade barriers are restricted to impact equally on imports in the estimates. Coupled with the small sample problems that plagued his estimates, the trade-barrier dummies were unlikely to effectively have captured the effects of trade barriers on imports, which explains substantial variations of the estimated effects of the tariff s and the nontariff barriers across countries. The study of Eichengreen and Irwin (1995) is probably the most extensive analysis of trade flows in the interwar period. Using 561 cross-sectional bilateral trade flows over three periods (1928, 1935, and 1938) they estimate a gravity model of trade patterns. They relate the value of bilateral flows to national income, population, distance, contiguity, trade and currency block indicators, and exchange rate variability, to examine the effects on trade of trade and currency blocks, and exchange rate variability. They observe a declining marginal propensity to import and export during the Depression, which

   they attribute to quotas and other binding trade restrictions, but do not formally test their importance.

   This paper seeks to estimate the contribution of income, tariffs, and nontariff barriers on world trade during the Depression using panel data for 17 countries over the period from 1920 to 1938. The panel data approach enables the assessment of the influence on trade of nontariff barriers from estimates of import and export functions, by using as an identifying assumption that the nontariff barriers were to some degree simultaneously imposed and relaxed across the industrialized nations during the interwar period. The estimates and extensive evidence from the literature suggest that this identifying assumption is valid (section 3). In section 4 the changes in world trade in the interwar period are decomposed into income effects, tariff effects, and nontariff barrier effects. The trade effects of the tariff changes are furthermore decomposed into deflation! inflation-induced tariff changes and discretionary tariff-induced changes. Because a significant fraction of import duties were specific (Liepmann 1938; C rucini 1994), and therefore denominated in fixed nominal values, tariff rates were automatically pushed up by declining import prices in the first years of the Depression. Overall the decomposition shows that 41% of the collapse in world trade from 1929 to 1932 was due to discretionary escalations of trade barriers, and 59% as a result of falling nominal income, assuming that the decline in prices and output were independent of the increasing trade barriers. However, as discussed in section 5, because nominal income was influenced by trade barriers, the discretionary impositions of trade barriers had stronger trade effects than suggested by these figures. Section 6 concludes the paper.

2. Tariffs and the Pattern of World Trade in the Interwar Period

   Before estimating the influence on world trade of trade barriers and income, a casual graphical analysis of the world macro tariff rate and the pattern of world trade is undertaken. Figure 1 displays the macro import tariff rates for the most important trading nations and the import weighted tariff rate for 22 countries in the interwar period, subsequently referred to as the world tariff rate. [1] The macro tariff rates are estimated as import duties divided by import value. The figure shows that the world macro tariff rate almost doubles from 1929 to 1932, a period that was associated with the two major events in tariff policy history. The first shock was the passage of the Hawley-Smoot Tariff Act in 1930. The second shock was the passage the Abnormal Importation Act in November 1931 and the Import Duties Act in February 1932 by the British Parliament (Friedman 1974, p. 26). These shocks led to widespread worldwide reactions according to Jones (1934) and Friedman (1974). In his detailed taxonomy of trade ba rriers, Jones (1934) finds that the introduction of the Hawley-Smoot Tariff Act of 1930 led to concerted worldwide retributions against U.S. exports and escalations of trade barriers that were not specifically targeted at U.S. products. On the basis of detailed studies of major trading nations he concludes that the Hawley-Smoot Tariff had "very definite effects upon the commercial policies of the principal trading nations of the world and upon the general development of the principles of commercial policy throughout the world" (pp. 1-2). The Hawley-Smoot Tariff was an important catalyst for worldwide escalations of trade barriers because the United States, which was the greatest creditor nation at that time, withdrew many of its international loans and did not make new loans available, and therefore forced deficit countries to lower their imports.

   Figure 2 shows the world nominal trade flows between trade blocks and nontrade blocks in the interwar period, where the trade flows are measured by exports. The estimates are based on the 22 countries in Figure 1. The following trade/currency blocks are considered: The Sterling block, the Reichmark block, and the Gold block. This block classification follows the classification of Eichengreen and Irwin (1995), and the countries contained in each block are listed in the notes to Figure 2.2 From 1920 to 1939, 66.6% of world trade was between non-trade blocks, on average. The curves have been standardized to have a mean of 100 over the whole period. The figure shows that the gain in world trade throughout the 1920s was lost within the first four years of the Depression, when nominal world trade declined more than 50%. The two curves show significant comovements between trade and nontrade blocks, which suggests that changes in country-specific tariffs and nontariff and trade barriers were not crucial determinants for the changes in trade flows in the interwar period. This visual impression is consistent with Eichengreen and Irwin's (1995) results that trade between trade and currency blocks did not gain significantly in importance, at the expense of other countries, during the Depression. Kitson and Solomou (1995) report a similar finding. It is also consistent with the evidence of Woytinsky and Woytinsky (1955, p. 80) that the distribution of international trade among continents did not change over the period 1928 to 1938. Perhaps it was realized that countries would not gain much from country-specific trade restrictions. Gardner and Kimbrough (1990) demonstrate that countries have little to gain from country-specific tariffs and that all trading partners are affected by country-specific tariffs, not only the nations that are targeted.

3. Estimates of Imports and Exports in the Interwar Period

   This section estimates import and export equations using pooled cross-section and time-series data for 17 important players in the world market during the interwar period, to disentangle the effects on world trade of income, tariffs and nontariff barriers, and exchange rate variability. [3] The panel data nature of the estimates not only overcomes the small sample problems that are associated with single country estimates, which use annual data, it also enables a quantification of the effects on trade of the imposition of nontariff barriers.

   The following equations for export and import volume are...