Does Exposure to International Trade Justify Relaxed Antitrust Treatment of Mergers
International Trade and Antitrust
Does Exposure to International Trade Justif y Relaxed Antitrust Treatment of Mergers
William James Adams 0
0 This Article is brought to you for free and open access by Northwestern University School of Law Scholarly Commons. It has been accepted for inclusion in Northwestern Journal of International Law & Business by an authorized administrator of Northwestern University School of Law Scholarly Commons
Part of the Antitrust and Trade Regulation Commons, and the International Trade Commons Recommended Citation
Does Exposure to International Trade
Justify Relaxed Antitrust Treatment
William James Adams*
When industries are exposed to foreign competition, relaxation of
antitrust law in general, and of antimerger law in particular, may be
justified in two ways. First, it may be argued that the ability to compete
with foreigners requires possession of market power. One variant of this
argument stresses the desirability of market power itself. Domestic
enterprises must neutralize restrictive practices abroad-whether they are
inspired by foreign governments or merely tolerated by them-if such
enterprises are to enjoy their "natural" comparative advantages. A
second variant of the argument emphasizes the growth of minimum efficient
scale in manufacturing, due in no small measure to the rising importance
of invention and innovation. In this variant, the market power acquired
by domestic firms, through merger or restrictive practices, is thought to
be an unfortunate, but necessary, by-product of achieving productive and
dynamic efficiency. Several recent studies,1 including one published in
this Journal,2 have questioned the wisdom of this first argument, and
* Professor of Economics and Adjunct Professor ofLaw, The University of Michigan; author of
RESTRUCTURING THE FRENCH ECONOMY: GOVERNMENT AND THE RISE OF MARKET
COMPETITION SINCE WORLD WAR II (1989).
1 See Caves, Effects of Mergers and Acquistions on the Economy: An IndustrialOrganization
Perspective,in THE MERGER BOOM (L. Browne & E. Rosengren eds. 1988); W. ADAMS & J. BROCK,
THE BIGNESS COMPLEX (1987).
2 Adams & Brock, The Bigness Mystique and the Merger PolicyDebatea An
InternationalPerspective, 9 Nw. J. INT'L L. & Bus. 1 (1988). Adams and Brock argue as follows: the governments of
several rich market economies have tolerated and even promoted major mergers since World War II;
retrospective evidence suggests that these mergers failed to augment the profitability or the
productivity ofthe combining enterprises; as a result, the foreign experience does not contradict the wisdom
of restoring stringency to American regulation of mergers-of returning to the treatment that
existed before the implementation of the Reagan guidelines.
Northwestern Journal of
International Law & Business
consequently I shall ignore it here. Instead, I shall focus on a second
argument: that the existence of foreign competition eliminates both
actual and potential reduction of competition that might otherwise result
from market conduct, or from market positions, challenged by antitrust
authorities. In this view, competition remains a desirable policy
objective; yet it is achieved not through antitrust intervention but through the
market mechanism of international trade.
To what extent does failure to take account of foreign competition
result in overestimation of the anticompetitive consequences of domestic
mergers? Many studies suggest that across manufacturing industries,
market power (measured by long-run profitability) varies negatively with
exposure to imports.3 Studies of particular industries also suggest that
across industrialized nations, seller-concentration varies negatively with
exposure to imports.4 As a result, the thoughtful framer of antitrust
policy must ascertain whether foreign sellers account for major fractions of
American markets, and whether American producers sell considerable
fractions of their output in foreign markets which they fail to dominate.
If the findings in each case are affirmative, then American levels of
producer-concentration may systematically and substantially overestimate
the levels of seller-concentration in markets supplied by American
enterprises. For this and other reasons, they may overestimate the
anticompetitive consequences of mergers.
To calibrate the exposure of American industries to foreign trade, I
shall compare the American situation with the situation in France.
Among industrialized nations, France is often considered to be relatively
protectionist in matters of international trade. Hence, my interest is in
determining how exposed American enterprises are to international trade
in comparison with their French counterparts.
Using information gathered by the Organization for Economic
Cooperation and Development ("OECD"),5 it is possible to examine the
ratio of imports to consumption, and the ratio of exports to output, in
identically-defined French and American industries. The information
reported herein relates to 1982, the year the first Reagan merger guidelines
took effect.6 In manufacturing as a whole, the ratio of imports to
domestic consumption ("import ratio") was 25% in France and 9% in the
United States. The ratio of exports to production ("export ratio") was
26% in France and 8% in the United States. In other words, the
incidence of international trade on domestic manufacturers was three times
as great in France as in the United States.
The importance of trade can vary, of course, among industries
within the manufacturing sector. To what extent do figures on the entire
manufacturing sector represent accurately the situation in particular
industries? The OECD data permit decomposition of manufacturing into
twenty-eight industries, a level of aggregation lying between the two- and
three-digit levels of the United States Standard Industrial Classification
Looking first at imports, in twenty-seven of the twenty-eight
industries,8 the French import ratio exceeds its American counterpart. The
average gap between the two ratios is eighteen percentage points.9 Table
1 shows the distribution of value-added among manufacturing industries
classified according to import-exposure. Nearly two-thirds of American
value-added, but less than one-tenth of French value-added, is generated
in industries with import ratios below 10%. On the other hand, nearly
half of French value-added, but only one-hundredth of American
valueadded, is generated in industries with import ratios of 30% or more.
Clearly, the French exposure to imports is as pervasive as it is intense.
Turning to exports, the French export ratio exceeds that of the
6 During 1982, the United States dollar tended to be strong, while the French franc tended to be
weak. As a result, ratios of imports to consumption may be abnormally high in the United States.
Similarly, ratios of exports to output may be abnormally low in France and abnormally high in
France and abnormally low in the United States. The Merger Guidelines adopted by the
Department of Justice in 1982 and 1984 are set forth in 4 Trade Reg. Rep. (CCH) 13,102-13,103.
7 In order to facilitate the classification ofestablishments by economic activity, and to promote
the collection, presentation and comparability of data from those establishments, the SIC was
formulated as the statistical classification standard underlying all establishment-based economic statistics
classified by industry in the United States. The SIC covers the entire field ofeconomic activity and
defines industries according to the composition and structure of the economy, and it is used to
promote the comparability of establishment data describing various facets of the United States
economy. See OFFICE OF MANAGEMENT & BUDGET, STANDARD INDUSTRIAL CLASSIFICATION
MANUAL (1987). Similarly, the International Standard Industrial Classification of All Economic
Activities ("ISIC") was adopted by the United Nations to provide a framework for the international
comparison of national statistics classifying establishments based upon economic activity. See U.N.
DEP'T OF ECONOMIC & SOCIAL AFFAIRS, INTERNATIONAL STANDARD INDUSTRIAL
CLASSIFICATION OF ALL ECONOMIC ACTIVITIES, U.N. Doc. ST/STAT/SER.M/4/Rev. 2, U.N. Sales No. E.
68, XVII.8 (1968).
8 All save footwear (ISIC 324).
9 The average import ratio was 28% in France and 11% in the United States. The gap differs
positively from 0 at the .001 level of statistical significance in a two-tail t-test.
Northwestern Journal of
International Law & Business
United States in twenty-seven of the twenty-eight industries. 0 On
average, the export ratio is 26% in France but only 7% in the United
States.1 1 Table 2 shows the distribution of value-added among
manufacturing industries, classified according to export-exposure. Only one
French industry, but twelve American industries, display export ratios of
less than 5%. On the other hand, eighteen French industries, but only
one American industry, display export ratios of 20% or more. Unlike
their French counterparts, most American industries sell the
overwhelming bulk of their output at home. Even if foreign markets function
competitively, their relevance to American enterprises is too tenuous to affect
the degree of competitive pressure they experience.
French industries are heavily exposed to international trade, while
American industries are not. Nevertheless, it is interesting to note that
industries with relatively high import ratios in France (by French
standards) tend also to show relatively high import ratios in the United States
(by American standards). For example, the printing and publishing
industry (International Standard Industrial Classification ["ISIC"] 342)
exhibits a relatively low import ratio in both countries, while the
footwear industry (ISIC 324) exhibits a relatively high import ratio in both
countries. Exceptions to this synchrony do exist-the textile industry
(ISIC 321) shows relatively great exposure to imports in France but not
in the United States-but, statistically speaking, French and American
import ratios are positively correlated. 2 French and American export
ratios are also positively correlated. 13 Furniture and fixtures (ISIC 332)
illustrates a relatively low export ratio in both countries; machinery
(ISIC 382) illustrates a relatively high export ratio in both countries; and
miscellaneous petroleum and coal products (ISIC 354) illustrates the
occasional discrepancy of ratios (relatively high in France, average to low
in the United States).
In Tables 1 and 2, industries are defined very broadly. Each
industry contains many economically distinct lines of business. If industries
are defined more narrowly, does one observe greater variation among
industries in import and export ratios? To answer this question, one must
rely on data reported at the national level. The disadvantage of these
data is the incongruity of French and American schemes of industrial
classification. Unlike the OECD information, reported according to
ISIC for both countries, the national information cannot be compared at
10 All save tobacco (ISIC 314).
11 The gap differs positively from 0 at the .001 level of statistical significance in a two-tail t-test.
12 At the .05 level of statistical significance in a one-tail t-test.
13 At the .01 level of statistical significance in a one-tail t-test.
the industry level. On the other hand, the OECD's information is
derived from these national data; the national figures are certainly of higher
Table 3 shows the distribution of American manufacturing
industries, narrowly defined, by exposure to imports and exports. Although
several such industries are heavily exposed to imports (see Table 4), and
others sell heavily in foreign markets (see Table 5), the numbers in each
case are small in relation to the total number of industries observed.
Furthermore, the industries included in these tables are not, by and large,
the targets of strict merger regulation; 14 nor would they be under the
preReagan guidelines. International trade does not obviously tend to
impinge heavily on industries especially prone to major mergers. In most
instances, recalculation of concentration ratios to account for foreign
competition, at home and abroad, would not alter the apparent wisdom
of curbing a major American merger. 5
If international trade should be taken into account in the regulation
of mergers, the place to do it is in countries like France where exports
and imports impinge heavily and pervasively on manufacturing
industries.16 In fact, it is more than possible that the failure of major mergers
in Europe can be attributed to the persistent impact of market forces,
transmitted through international trade, on European producers. After
all, productive inefficiency would not have resulted in negative profits
had the European national champions enjoyed substantial amounts of
market power. It is precisely because so many American producers are
14 It is beyond the scope of this article to correlate across American four-digit manufacturing
industries the frequency of merger activity and the exposure to either imports or exports. On the
other hand, it is interesting to examine the number of indictments for criminal violation of the
Sherman Act in industries heavily exposed to international trade, as identified in Tables 4 and 5.
(Data on indictments taken from . CLABAULT & M. BLOCH, I & 2 SHERMAN ACT INDICTMENTS
1955-1980, at 688, 1053-70 (1981). Between 1955 and 1980, 617 such cases were brought. Of these,
just seven concerned the 23 industries of Table 4; and just another seven concerned the 12 industries
of Table 5. These results are not surprising: most Sherman Act indictments involve conspiracies to
fix prices in a domestic market, and such conspiracies are unlikely to develop where potential
conspirators face substantial import competition in the domestic market or sell large fractions of their
output in foreign markets.
15 Although it involved prosecution for monopolization rather than for merger, United States v.
Aluminum Co. of America (Alcoa), 148 F.2d 416 (1945), illustrates the point. Assuming the
relevant product market was primary aluminum ingot, Alcoa was a literal monopolist if imports were
excluded by definition from the market; with imports included, Alcoa's market share remained
above 90 percent.
16 Import and export ratios for narrowly defined French industries, confirming the broad impact
of international trade on French manufacturing, appear in W. ADAMS, RESTRUCTURING THE
FRENCH ECONOMY: GOVERNMENT AND THE RISE OF MARKET COMPETITION SINCE WORLD WAR
II ch. 4 (1989).
naturally or artificially protected from competition through international
trade that vigor must be restored to United States regulation of mergers.
0 or more 28 28 100.0 100.0
Source: ORGANIZATION FOR ECONOMIC COOPERATION AND
DEVELOPMENT, INDUSTRIAL STRUCTURE STATISTICS, 1984 (1986).
Note: Import ratio is (100)(M)/(Q-X+M), where M is imports, Q is
production, and X is exports.
Distribution of Value-Added in Manufacturing by Exposure
to Exports, France and the United States, 1982
Distribution of American Manufacturing Industries by
Exposure to International Trade, 1982
Source: U.S. Department of Commerce, Bureau of the Census. U.S.
Commodity Exports and Imports as Related to Output: 1982 and 1981, tables
1A and lB.
Note: Four-digit manufacturing industries importing (column 1) or
exporting (column 2) $10 million or more of merchandise during 1982. Import
ratio is imports c.i.f. as percent of new supply (domestic output plus imports).
Export ratio is exports f.o.b. as percent of domestic output.
American Manufacturing Industries Heavily Exposed to
Source: U.S. Department of Commerce, Bureau of the Census. U.S.
Commodity Exports and Imports as Related to Output: 1982 and 1981, table
Notes: Four-digit manufacturing industries for which (1) the import ratio
(imports c.i.f. as percent of new supply, where new supply is domestic output
plus imports) was 30 percent or more, and (2) imports were $10 million or more.
As in source, some listed industries contain parts of certain unlisted industries.
American Manufacturing Industries Especially Prone to
Turbines and turbine generator sets
Lace and net goods
Medicinals and botanicals
Pulp mill products
Aircraft parts and auxiliary equipment
Sewing machines and parts
Measuring and controlling devices nec
Petroleum and coal products nec
Milled rice and by-products
3 See Jacquemin, de Ghellinck & Huveneers, Concentrationand Profitabilityin a Small Open Economy, 29 J. INDUS . ECON. 131 ( 1980 ) ; Pugel, Foreign Tradeand US Market Performance,29 J. INDUS . ECON. 119 ( 1980 ) ; and Jenny & Weber, Profit Rates and Structural Variables in French ManufacturingIndustries,7 EUR. ECON . REV. 187 ( 1976 ).
4 See Adams , Producer-Concentrationas a Proxyfor Seller-Concentration:Some Evidencefrom the World Automotive Industry, 29 J. INDUS. EON . 185 ( 1980 ).
5 ORGANIZATION FOR ECONOMIC COOPERATION AND DEVELOPMENT , INDUSTRIAL STRUCTURE STATISTICS, 1984 ( 1986 ).
Source : U.S. Department of Commerce, Bureau of the Census. U.S. Commodity Exports and Imports as Related to Output: 1982 and 1981 , table 1A .
Notes: Four-digit manufacturing industries for which (1) the export ratio (exports f.o.b. as percent of domestic output) was 30 percent or more, and (2) exports were $10 million or more . Industry 3829 includes industry 38244 . 9: 589 ( 1989 )