Printer Friendly

Product variety, trade, and growth.

Since the early 1980s, theoretical work in international trade has incorporated models of monopolistic competition.(1) Much of my own research has used these models to estimate the effects of trade barriers and economic growth on trade patterns.


One trade policy that dramatically affected the characteristics of imports was the "voluntary" export restraint (VER) that limited the sale of Japanese autos in the United States. The VER began in April 1981 and limited the number of cars that could be exported by Japanese firms; the VER thus created an incentive for Japanese firms to upgrade their export models. In the following years, many Japanese models became heavier, more powerful, and added air conditioning, automatic transmissions, and other features as standard equipment. I estimate that fully one-half of the nominal increase in prices of Japanese imports from 1980 to 1985 reflects the upgrading of their characteristics.(2)

Correcting the price increases for the quality upgrading is one step toward inferring the "pure" increase in prices explained by the VER. The other step is to estimate what would have happened to Japanese auto prices in the absence of any trade restriction, To determine this, I use the prices of Japanese compact trucks imported into the United States. These were not subject to the export restraint, but did have an increase in their tariff from 4 percent to 25 percent effective August 1980. By incorporating this tariff change into a hedonic regression for trucks, I can estimate the quality-adjusted price change that would have occurred under free trade. I find that the export restraint raised prices by $1000 by 1983 and 1984. In later years the export restraint was loosened, so the price effect fell.

The tariff imposed on imports of Japanese trucks allows us to evaluate the effect of trade policy on new product varieties. After the imposition of the tariff, several U.S. companies began producing compact trucks that were very similar to existing Japanese trucks. Using econometric methods, I find that some Japanese trucks were sufficiently different, or distinctive, to provide a substantial welfare gain. However, the American models produced only a small additional welfare gain, since they were similar to the Japanese models.(3)

In joint work with James A. Levinsohn, I identify the close substitutes of a product in terms of its characteristics in order to compute demand elasticities.(4) The firm uses these own and cross-price elasticities to determine profit-maximizing prices under Bertrand competition. The optimal prices are related directly to the "distance" (that is, similarity or difference) to neighbors: as the distance increases, so does the price.(5) We find that distance is very significant, especially in explaining the high prices of certain luxury cars.

In future research, we intend to examine how the VER on Japanese autos affected pricing behavior. In particular, American models with Japanese neighbors should show some increase in price under the quota. This would shift demand toward the Japanese competitors, who would have to respond by increasing their own price. Thus, the quota acts to support collusion among the firms, as has been demonstrated theoretically. An empirical model of neighbors and oligopoly pricing will allow this hypothesis to be tested.

The upgrading of imports under quota restraints has occurred in many industries in addition to autos. In research with Randi Boorstein, I have estimated the increase in quality of steel imports as the difference between the increased value of such imports and the increase in an exact price index of steel products. I find that quality upgrading in steel imports during the first year of the 1969-74 quota explains one-half of the nominal increase in the unit value, although less of the increase in later years.(6)

Changes in this measure of import quality are explained by shifting relative consumption weights within the import bundle, and there is a welfare cost to this distortion in consumption. The welfare cost can be seen as the cost of the quota as compared to a price-equivalent ad valorem tariff, which would not lead to quality change. Again, this welfare cost can be measured by a comparison of index numbers: for 1969, the welfare cost of upgrading was in the range $3.5 million to $7 million.(7)

New product varieties also can be incorporated into the index number measurements. New varieties in particular can be used when estimating U.S. import demand. We know that the estimated income elasticity of U.S. import demand exceeds unity, and also exceeds the income elasticity of foreign demand for U.S. exports. These estimates imply that equal growth in the United States and abroad automatically will lead to a worsening in the U.S. trade balance. Some authors have suggested, however, that the high U.S. income elasticity is caused by the omission of new product varieties from import price indexes. Then the rising share of imports in the United States is attributed to a high income elasticity in the demand equations.

To correct for this, it is possible to incorporate new product varieties directly into the price indexes.(8) Including new product varieties lowers the price index by an amount that depends on their share in expenditure, and on the elasticity of substitution. In practice, new product varieties can be measured by using data on new supplying countries, and treating products as differentiated across countries. Applying this technique to several U.S. imports, I find that the corrected price indexes rise significantly slower than conventional measures. The corrected indexes can explain a part--but not all--of the high income elasticities in the import demand equations.


Modifying index numbers for new and disappearing product varieties is one way to implement empirically the recent models of endogenous growth, particularly growth resulting from the expansion of intermediate inputs.(9) Just as new varieties will lower the price indexes, they will raise the quantity index, reflecting the contribution of new input varieties to productivity. Again, the extent to which the new input varieties affect the quantity index depends on their share in expenditure, along with the elasticity of substitution. If there are many different kinds of inputs, then we also can determine their effect on GNP. Provided that the primary resources are adjusted for profits in the production of the intermediates, this function has constant returns to scale while holding the ranges of intermediate inputs fixed, and is increasing in these ranges.(10)

In joint research, I have applied these quantity indexes to explain the productivity of industries in South Korea.(11) In this case, it is difficult to measure new inputs available to the firms directly, but an indirect method is available. Each of the firms belonged to a chaebol, or business group, that consists of firms with some common ownership and strong vertical links. For each of these groups, we can observe the addition of new input-producing firms over time, and can presume that these firms supplied preferentially to other businesses in the group. Provided that exactly the same inputs were not available earlier, this should enhance the productivity of the purchasing firms. We find that in industries dominated by chaebol that experienced substantial new entry over 1983-6, productivity was higher than otherwise. The results support the hypothesis that the new entrants to the chaebol conferred a productivity boost to the member firms.

This discussion indicates how an increasing range of inputs in the endogenous growth models can be implemented empirically. An alternative formulation is to consider quality improvements in intermediate inputs, as in the "quality ladders" model.(12) Under this approach, new generations of inputs replace the old: computers are a good example. Using characteristics data, we can estimate the declining price per unit of services of each generation, but for many other goods these data would not be available. I believe it will still be possible to implement the "quality ladders" approach, if we are willing to make the same simplifications used in the theory.

In particular, the utility function adopted in this approach is linear in the generations of each product, where newer generations receive a higher welfare weight. The linearity implies that the price of each generation is an accurate measure of its utility (or quality) per unit. Then a quality index can be constructed by applying relative quantity weights across the prices of each generation: this is identical to the "product mix" index that has been used to measure upgrading of imports under quotas. Using this index, changes in product quality over time and across countries can be measured.

In current research, I am applying this technique to U.S. imports from South Korea and Taiwan. It appears that Korea is the leader among these two countries, being the first to export new products or generations to the United States. After some lag, these products are transferred to Taiwan, which then exports to the United States. Analyzing this pattern more systematically will provide an empirical counterpart to the "quality ladders" model.

1 This terminology is taken from E. Helpman and P. R. Krugman, Market Structure and Foreign Trade, Cambridge, MA: MIT Press, 1985. The authors provide a comprehensive treatment of monopolistic competition and international trade models.

2 See R.C. Feenstra, "Voluntary Export Restraints in U.S. Autos, 1980-1: Quality, Employment, and Welfare Effects," in R. E. Baldwin and A. O. Krueger, eds., The Structure and Evolution of Recent U.S. Trade Policy, Chicago: University of Chicago Press, 1984, pp. 35-59; and R. C. Feenstra, "Quality Change Under Trade Restraints in Japanese Autos," Quarterly Journal of Economics 103, 1 (February 1988), pp. 131-146.

3 R. C. Feenstra, "Gains from Trade in Differentiated Products: Japanese Compact Trucks," in R. C. Feenstra, ed., Empirical Methods for International Trade, Cambridge, MA: MIT Press, 1988, pp. 119-136.

4 J. A. Levinsohn and R. C. Feenstra, "Identifying the Competition," Journal of International Economics 28, 3/4 (1990), pp. 199-216.

5 R. C. Feenstra and J.A. Levinsohn, "Distance, Demand, and Oligopoly Pricing," NBER Working Paper No. 3076, August 1989; revised as "Oligopoly Pricing and the Characteristics Approach," November 1991, mimeo.

6 R. C. Feenstra and R. Boorstein, "Quality Upgrading and Its Welfare Cost in U.S. Steel Imports, 1969-74," in E. Helpman and A. Razin, eds., International Trade and Trade Policy, Cambridge, MA: MIT Press, 1991, pp. 167-186.

7 For other examples of upgrading and its welfare costs, see R. C. Feenstra, "How Costly Is Protectionism?" Journal of Economic Perspectives 6, 3 (Summer 1992), pp. 159-178.

8 R. C. Feenstra, "New Goods and Index Numbers: U.S. Import Prices," NBER Working Paper No. 3610, February 1991; revised as "New Product Varieties and the Measurement of International Prices," August 1992, mimeo.

9 See, for example, G. M. Grossman and E. Helpman, Innovation and Growth in the Global Economy, Cambridge, MA: MIT Press, 1991.

10 R. C. Feenstra and J. R. Markusen, "Accounting for Growth with New Intermediate Inputs," NBER Working Paper No. 4114, July 1992.

11 R. C. Feenstra, J. R. Markusen, and W. Zeile, "Accounting for Growth with New Intermediate Inputs: Theory and Evidence," American Economic Review (May 1992), pp. 415-421.

12 See G. M. Grossman and E. Helpman, Innovation and Growth in the Global Economy, chapters 4 and 12, and also P. S. Segerstrom, T. C. A. Anant, and E. Dinopoulos, "A Schumpeterian Model of the Product Life Cycle," American Economic Review 80, 5 (December 1990), pp. 1077-1091.
COPYRIGHT 1992 National Bureau of Economic Research, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Feenstra, Robert C.
Publication:NBER Reporter
Date:Sep 22, 1992
Previous Article:Inflation dynamics.
Next Article:The economics of aging.

Related Articles
International Trade and Investment.
Food safety and trade mutually reinforcing, says ERS.
International Trade and Investment.
International trade and investment.
International Trade and Investment.
International Trade and Investment.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters