International Trade of Multinational Corporations and Its Responsiveness to Changes in Aggregate Demand and Relative Prices

David John Goldsbrough

doi:10.5089/9781451972672.024.A005

Author: Mr. David John Goldsbrough

Publication Date:: 01 Jan 1981

eISBN:: 9781451972672

Language:: English

Keywords:: SP; exchange rate; price term; trade model; multinationals' Intrafirm trade; trade equation; trade decision; Manufacturing; Trade balance; Price elasticity

Download PDF (1.4 MB)

International trade flows are to a great extent determined by the location of the affiliates of multinational firms with worldwide production facilities. A great deal of this trade consists of direct intrafirm transactions between affiliates of the same firm. For example, in 1970 manufacturing exports from parent U. S. companies to their affiliates abroad accounted for over 20 per cent of all U. S. manufacturing exports. In the same year, intrafirm trade within the U. S.-controlled multinationals alone accounted for about 10 per cent of all world trade in manufactures.

Abstract

The importance of such internal transactions has been growing over time. This could have important implications for the response of a country’s trade to changes in economic activity and relative competitiveness. For instance, trade flows generated by the location decisions of a firm with large fixed investments in several countries may not respond as rapidly to shifts in relative prices as those of an independent producer that is unconcerned with the effect of its actions on the profitability of overseas affiliates. This paper reports on some attempts to test whether the response of such internal trade flows to shifts in aggregate demand and relative prices differs from the response of more conventional trade. A simplified version of the Fund’s World Trade Model was fitted to data on the transactions of U. S. multinational companies and their overseas affiliates, and the results were compared with those obtained for conventional trade flows.

The paper first discusses the size of intrafirm trade flows and variations in their relative importance across countries and industries. The next section considers possible implications of such trade for the response to relative price changes. There follows an outline of simple export and import equations that can be fitted to data on conventional and intrafirm manufacturing trade in order to test whether significant differences exist in the parameters for the two types of trade. Final sections discuss the results and summarize the major conclusions. An appendix discusses some of the economic problems in measuring price elasticities in international trade caused by the use of artificial transfer prices in intrafirm transactions.

I. Relative Importance of Intrafirm Trade Flows

The most reliable statistics on intrafirm trade come from the U.S. and U. K. authorities. Table 1 gives details on the importance of the internal trade of U. S. multinationals in overall U. S. manufacturing trade. Intrafirm trade accounts for nearly one fifth of total trade and seems to be growing over time. If intrafirm trade with minority-owned affiliates and in third-party goods channeled through the firm were also included, then the share would rise to over one fourth of total U. S. manufacturing trade. No data are available on the magnitude of intrafirm trade by foreign-controlled companies with investments in the United States, but if this is in the same proportion to inward direct investment as the U. S. multinationals’ intrafirm trade is to outward direct investment, then it would account for another 4 to 5 per cent of total U.S. manufacturing trade. This implies that close to 30 per cent of all U. S. manufacturing trade is handled through the internal transactions of multinational firms. In the United Kingdom, intrafirm trade of both home and foreign-controlled multinationals accounted for 26 per cent of total exports in 1970, compared with 22 per cent in 1966.¹ No figures are available for the import side of U. K. trade. Therefore, intrafirm trade constitutes a substantial, and growing, proportion of merchandise trade in both the United States and the United Kingdom.

Table 1.

Share of Intrafirm Trade of U. S. Multinational Corporations in Total U. S. Manufacturing Trade¹

(In millions of U. S. dollars and in per cent)

Source: United States, Tariff Commission (1973)

^¹The data on intrafirm trade cover only trade between a U. S. parent and a majority-owned foreign affiliate (MOFA), that is, an affiliate in which a single U. S. firm or its other affiliates owns at least 50 per cent of the voting interest. Intrafirm trade is here defined to include only trade in products manufactured by the parent or affiliate. For instance, exports to the affiliates by unrelated U. S. firms are often channeled via the U. S. partner, but this trade is excluded from the above figures (if included, the shares of intrafirm trade in total trade would rise by about 5 per cent).

Table 1.

Share of Intrafirm Trade of U. S. Multinational Corporations in Total U. S. Manufacturing Trade¹

(In millions of U. S. dollars and in per cent)

Type of Trade		1966	1970
EXPORTS
	Total U. S. manufactured exports	21,227	31,742
	Intrafirm manufactured exports from U. S. parents to overseas affiliates	4,050	6,831
	Share of intrafirm trade in total manufactured exports (in per cent)	19.1	21.5
IMPORTS
	Total U. S. manufactured imports	16,893	30,795
	Intrafirm manufactured imports from U. S. parents to overseas affiliates	2,197	4,827
	Share of intrafirm trade in total manufactured imports (in per cent)	13.0	15.7

Source: United States, Tariff Commission (1973)

Table 1.

Share of Intrafirm Trade of U. S. Multinational Corporations in Total U. S. Manufacturing Trade¹

(In millions of U. S. dollars and in per cent)

Type of Trade		1966	1970
EXPORTS
	Total U. S. manufactured exports	21,227	31,742
	Intrafirm manufactured exports from U. S. parents to overseas affiliates	4,050	6,831
	Share of intrafirm trade in total manufactured exports (in per cent)	19.1	21.5
IMPORTS
	Total U. S. manufactured imports	16,893	30,795
	Intrafirm manufactured imports from U. S. parents to overseas affiliates	2,197	4,827
	Share of intrafirm trade in total manufactured imports (in per cent)	13.0	15.7

Source: United States, Tariff Commission (1973)

The only data available for a wider sample of countries give information on the share of intrafirm trade in the exports of large multinational corporations only, thus overstating the relative importance of such trade by excluding the smaller multinationals and the purely national firms. However, the data illustrate the wide variations in the importance of such internal trade for firms based in different countries (Table 2).

Table 2.

Importance of Intrafirm Trade in the Exports of the World’s Largest Multinational Corporations, 1972

Source: Buckley and Pearce (1977).

^¹Belgium, Luxembourg, and the Netherlands.

Table 2.

Importance of Intrafirm Trade in the Exports of the World’s Largest Multinational Corporations, 1972

Country		Number of Firms in Sample	Share of Intrafirm Trade in Parent Companies’ Total Exports
			Per cent
Sweden		5	58.7
United States		59	54.8
Germany, Fed. Rep. of		10	51.1
Benelux countries¹		3	47.1
Canada		2	43.8
Italy		3	42.2
United Kingdom		27	35.5
Switzerland		3	28.0
France		3	13.2
Japan		17	7.8
	Total	132	40.8

Source: Buckley and Pearce (1977).

^¹Belgium, Luxembourg, and the Netherlands.

Table 2.

Importance of Intrafirm Trade in the Exports of the World’s Largest Multinational Corporations, 1972

Country		Number of Firms in Sample	Share of Intrafirm Trade in Parent Companies’ Total Exports
			Per cent
Sweden		5	58.7
United States		59	54.8
Germany, Fed. Rep. of		10	51.1
Benelux countries¹		3	47.1
Canada		2	43.8
Italy		3	42.2
United Kingdom		27	35.5
Switzerland		3	28.0
France		3	13.2
Japan		17	7.8
	Total	132	40.8

Source: Buckley and Pearce (1977).

^¹Belgium, Luxembourg, and the Netherlands.

The most striking feature of these percentages is the position of Japan, whose large multinational corporations rely on intrafirm trade for a much smaller portion of their exports than do U. S. and German multinationals, more than half of whose exports are accounted for by transactions internal to the firm. This is largely a reflection of the small Japanese direct investments in overseas affiliates at the end of 1972 (about $8 billion, compared with $13 billion for the Federal Republic of Germany and $90 billion for the United States). The multinationals based in France and Switzerland also depend to a relatively minor extent on exports to overseas affiliates, although this conclusion is less reliable as it is based on observations for only a few firms. The large variations in the importance of intrafirm trade for multinationals of different countries may have an impact on the reaction of overall trade flows to shifts in prices and exchange rates, if significant differences occur in the response to relative price changes of intrafirm as opposed to conventional trade.

Large variations can be seen in the importance of intrafirm trade between different industries—for example, in the United States (Table 3). The share of intrafirm trade for the transport equipment industry is much larger than the share in any other major industry, primarily because the 1965 Automotive Products Trade Agreement between the United States and Canada, which provided for free trade in vehicles and components, subject to certain protective restrictions, and led to a large increase in the trade of automobile products. Since most of the Canadian automobile plants were subsidiaries of U. S. firms, this has largely taken the form of intrafirm rather than interfirm trade owing to the increased integration of the Canadian subsidiaries and their U. S. parents.

Table 3.

Share of Intrafirm Trade by U. S. Multinationals in Total U. S. Trade, by Industry, 1970¹

Source: Fund calculations using data from United States, Tariff Commission (1973).

^¹As in Table 1, intrafirm trade includes only trade between a U. S. parent and a majority-owned foreign affiliate.

Table 3.

Share of Intrafirm Trade by U. S. Multinationals in Total U. S. Trade, by Industry, 1970¹

	Intrafirm Exports from United States		Intrafirm Imports to United States
Industry Group	Value	Share of total exports	Value	Share of total imports
	Million U. S. dollars	Per cent	Million U. S. dollars	Per cent
All manufacturing	6,831	21.5	4,827	15.7
Transport equipment	2,142	32.8	2,733	43.0
Nonelectrical machinery	1,632	20.6	400	12.9
Electrical machinery	509	16.9	425	16.2
Chemicals	813	20.3	203	16.2
Paper and allied products	144	13.0	439	28.4
Food products	349	13.5	76	2.1
Textiles and apparel	78	10.8	104	4.4
Primary and fabricated metals	253	6.7	37	0.8

Source: Fund calculations using data from United States, Tariff Commission (1973).

^¹As in Table 1, intrafirm trade includes only trade between a U. S. parent and a majority-owned foreign affiliate.

Table 3.

Share of Intrafirm Trade by U. S. Multinationals in Total U. S. Trade, by Industry, 1970¹

	Intrafirm Exports from United States		Intrafirm Imports to United States
Industry Group	Value	Share of total exports	Value	Share of total imports
	Million U. S. dollars	Per cent	Million U. S. dollars	Per cent
All manufacturing	6,831	21.5	4,827	15.7
Transport equipment	2,142	32.8	2,733	43.0
Nonelectrical machinery	1,632	20.6	400	12.9
Electrical machinery	509	16.9	425	16.2
Chemicals	813	20.3	203	16.2
Paper and allied products	144	13.0	439	28.4
Food products	349	13.5	76	2.1
Textiles and apparel	78	10.8	104	4.4
Primary and fabricated metals	253	6.7	37	0.8

Source: Fund calculations using data from United States, Tariff Commission (1973).

^¹As in Table 1, intrafirm trade includes only trade between a U. S. parent and a majority-owned foreign affiliate.

Other sectors in which intrafirm trade is an important part of total U. S. trade are the machinery (electrical and nonelectrical) and chemical industries, whereas such trade is relatively unimportant in the food products, textiles, and metals industries—partly because multinational corporations have a larger share of the market in the former industries, but even within the group of multinationals there are substantial industry-by-industry variations in the importance of intrafirm trade in total trade. Most cross-section investigations of the phenomenon have concluded that intrafirm trade tends to account for a larger share of a firm’s total international trade in the larger firms and the more research-oriented ones. For instance, Buckley and Pearce (1977), using the same sample of multinational firms as in Table 2, showed that the 60 most research-intensive multinationals (in terms of research and development expenditure as a percentage of total sales) had intrafirm exports amounting to over half of total exports, whereas a similar group of the least research-intensive multinationals had intrafirm exports accounting for less than a fifth of their total exports.²

These wide industry variations in the importance of intrafirm trade indicate that the commodity composition of such trade differs substantially from that of conventional trade. Thus, it is difficult to interpret aggregate import equations that compare the responsiveness to price and demand changes by intrafirm and conventional trade, since any differences in the estimated parameters may be caused by commodity composition rather than the internal or external nature of the transactions. To help overcome this aggregation problem, a number of equations were estimated for both conventional and intrafirm U.S. imports for particular disaggregated industry groups.

II. Response of Intrafirm Trade Flows to Changes in Relative Prices

A number of reasons why the reactions of intrafirm trade to relative price changes might differ from those of conventional trade flows are discussed in the following paragraphs.

TRANSFER PRICES

The accounting prices used in a multinational firm’s internal transactions, which are the prices recorded in the trade statistics, affect the geographic distribution of its profits and tax liabilities. The firm, therefore, has an incentive to manipulate these prices so as to minimize its overall tax and tariff burden. Consequently, reported prices do not necessarily reflect either true market prices or the opportunity costs of production on which the firm bases its location decisions. Horst (1971) has shown that a firm with plants in an importing and an exporting country will maximize its global after-tax earnings by always choosing either the smallest or largest possible transfer price, depending upon whether the relative differential in corporate tax rates is less or more than the import tariff on the product. Firms also have an incentive to manipulate their accounting prices to transfer profits from those countries that impose restrictions on profit repatriation by overstating transfer prices on intrafirm imports and understating the price of intrafirm exports.

Very little empirical evidence exists pertaining to the actual extent of overstatement and understatement in transfer prices. Lall (1973) cites a study of intrafirm imports into Colombia that found considerable use of overpricing as a means of evading controls on profit repatriation. It was estimated that the weighted average of overpricing ranged from 100 per cent for a large group of pharmaceutical products to 25–50 per cent for groups of chemical, electrical, and rubber products. Such examples are not confined to developing countries’ imports. The Great Britain Monopolies and Restrictive Practices Commission (1973) estimated that intrafirm imports of certain pharmaceuticals into the United Kingdom were overpriced by as much as 150 per cent. However, the pharmaceutical industry is probably an extreme example, and industries with more homogeneous products are likely to have less scope for introducing variations between transfer and “arm’s length” prices.

As firms have a definite incentive to report intrafirm prices for tariff and trade statistics that are different from the real prices on which their location and trade decisions are based, the measurement error in observed prices is much greater for intraindustry than for conventional trade flows. A detailed examination of the econometric implications of the use of artificial transfer prices in intrafirm trade is given in Appendix I. The major conclusions are

(1) If the measurement error caused by the divergence of observed transfer prices from true prices is uncorrelated with the true price (i.e., as for conventional measurement error) or is positively correlated (i.e., if there is a percentage markup of transfer prices over true prices), the estimated price elasticities will be biased downward.

(2) If there is a negative correlation between the measurement error and true prices (i.e., if there is a percentage markdown of transfer prices from true prices), the estimated price elasticity may be biased upward or downward, depending upon the relative size of the covariance and variances of the measurement error and the true price. However, the downward bias will again predominate, provided that the understatement in transfer prices is not too large. (See Appendix I for a more exact definition.)

Therefore, the transfer price phenomenon will generally cause a downward bias in estimated price elasticities.

EFFECTS OF INTERNALIZATION OF INTERNATIONAL TRADE TRANSACTIONS ON PRICE RESPONSIVENESS

The previous discussion has shown why estimation of price elasticities for intrafirm trade is more difficult than for conventional trade, and why the estimated price elasticities might be expected to be biased downward. There are also reasons why the actual price responsiveness of intrafirm trade may be lower, because transactions have been internalized. The trade flows generated by the affiliates of a multinational firm with large fixed investments in several countries may not respond as rapidly to relative price shifts as those generated by independent producers that are unconcerned with the effects of their actions on the profitability of affiliates in other countries. Such a firm may divide the world market in a particular product into distinct geographic areas and allocate a specific market area, which may consist of the host country’s domestic market alone, to each affiliate, in order to restrict competition between different affiliates. For example, Hogan (1966) has estimated that almost half of U. K. manufacturing subsidiaries in Australia were subject to some such restraint on their exporting.³

A more subtle, but perhaps equally important, reason for the lower price responsiveness of intrafirm trade may be that the integration of plants in different countries into a multinational company has an effect on the type of products produced in each plant. The integrated plants produce goods tailored to the firm’s particular manufacturing and distribution needs, and these goods have fewer close substitutes than the more standardized alternatives purchased on the open market. For example, consider the case of a vertically integrated multinational firm requiring a very specific input for its production processes (e.g., a specialized electronic component). If the fixed costs involved in the production of such inputs are substantial and the demand curve is fairly steep, as would be expected with such specialized requirements, there may be no single price at which an independent supplier could recoup its cost of production, even though the total benefits from the product may exceed the total costs of its production. The only products available on the open market would then be the more standardized alternatives, supplying a larger market so as to spread the fixed costs. However, the purchasing firm has the alternative of integrating backward to produce the specialized input itself, in which case it would be able to expropriate the consumer surplus and cover the costs of producing the input by acting as a discriminating monopsonist with its own subsidiary supplying the input.⁴ Thus, a switch from purchasing inputs on the open market to relying on intrafirm supplies would involve a switch from standardized to more specialized inputs. Since the specialized inputs have few close substitutes, intrafirm trade in them will be less responsive to changes in relative prices. As the need for inputs with highly specialized characteristics occurs more often in the technology-intensive industries, intrafirm trade is much more prevalent in such industries.

III. A Simple Model of Conventional and Intrafirm Manufacturing Trade

This paper aims principally to discover whether there are significant differences in the responsiveness of conventional and intrafirm manufacturing trade to changes in relative prices and real aggregate demand. This is done by fitting a model of trade flows onto data for conventional and intrafirm trade separately and then testing whether the estimated parameters are significantly different for the two types of trade. The trade model used is a simplified version of the Fund’s World Trade Model.

However, before discussing the model, it is necessary to describe the type of data to be used in the estimating equations. The only adequate time-series data on trade by multinational companies is for trade by U. S. majority-owned foreign affiliates (MOFAs), but these data are not ideally suited for the present purpose, since they include some trade that is not truly intrafirm trade. For instance, a MOFA’s exports to the United States include some trade destined for firms other than the affiliate’s parent company. However, the vast majority of total manufacturing trade by MOFAs, particularly for exports to the United States, consists of intrafirm trade (Table 4). Intrafirm trade accounted for 70 per cent of all MOFA exports, with the share rising to almost 90 per cent for trade with the United States. Therefore, it is reasonable to use the time series on trade by MOFAs as independent variables in the intrafirm equations.

Table 4.

Share of Intrafirm Trade in Total U. S. Majority-Owned Foreign Affiliates’ (MOFAS’) Manufacturing Trade, 1970¹

(In millions of U. S. dollars and in per cent)

Source: United States, Department of Commerce, Bureau of Economic Analysis (1972), Table 3, p. 23.

^¹The figures on intrafirm trade are lower than those given in Table 1 because the survey on which they are based covers only a sample of overseas affiliates, whereas the figures in Table 1 are total estimates.

The other major drawback of the data is that time series are available only for U. S. manufacturing imports from MOFAs and for exports from MOFAs of U. S. multinationals in each of the major industrial countries. No time series have been collected on U. S. manufacturing exports from the parent company to its overseas affiliates and on total imports by the affiliates in each of the industrial countries. Therefore, the trade model can be fitted for only one side of the trade flows—namely, imports to the United States and exports to all destinations from the other major industrial countries.⁵

The import and export volume equations that are used in this paper correspond fairly closely to those of the Fund’s World Trade Model (Deppler and Ripley (1978)) and follow the approach suggested by Armington (1969). It is assumed that the demand for manufactured goods in a country can be specified separately from the demand for other goods and services. This in turn requires the assumption that the marginal rates of substitution within the group of manufactured goods is not affected by the quantity of nonmanufactures consumed.⁶ Therefore, the demand for manufactures can be seen as a two-stage decision process. In the first stage, which takes place outside the trade model, the total demand for manufactures is determined by the level of income and the aggregate price of manufactures relative to the price of other goods. The second stage, which is the subject of the trade model, determines how much of the total demand for manufactures is to be satisfied from domestic sources and how much from imports. However, if there is a large number of supplying countries in the model, the demand equation to be estimated would contain a large number of relative price terms. A more manageable specification is obtained by making use of the composite commodity theorem derived from the separability assumption already made, which makes it possible to write the price for inputs, PM_i, as a weighted average of the prices of the countries that export to country i, with the weights equal to the market share of each exporting country. So the demand for imports can be written as

M_i = α_o(AVG_i)^α1 (P_i/PM)^α2

where M_i represents the volume of imports into country i; P_i represents the domestic price of manufactures; PM_i represents the weighted aggregate price of manufactured imports; and AVG_i represents an index of total real demand for manufactures in the importing country, being a weighted average of domestic demand for final and intermediate manufactured goods (see Appendix II for detailed description of variables). This model is then estimated for conventional and intrafirm imports* separately.⁷

Simplifying assumptions are also required to reduce the number of relative price terms in the export equations. Here it is assumed that the elasticity of substitution between manufactured imports from any two exporting countries is the same for all countries competing in a given market; that elasticities are constant over time; and that market shares are independent of the size of the market. With these assumptions and aggregating over the various markets for each country’s exports, the basic export equation is derived⁸

X_i = α₀(FM_i)^α1 (RXPM_i)^α2

where X_i denotes the volume of manufactured exports from country i, FM_i denotes a weighted aggregate index of demand in the markets for country i’s products, and RXPM_i denotes an index of export prices for country i’s manufactures relative to a composite index of competitors’ prices (see also Appendix II). Before discussing the test results, it should be noted that the estimating equations used in this paper, although based on those in the Fund’s World Trade Model, differ in a number of important respects:

(1) No price equations are estimated. Instead, prices are treated as exogenous in the volume equations.

(2) The trade matrix used in the various weighting schemes cover only 10 industrial countries, as opposed to 14 in the Fund’s World Trade Model.⁹

(3) Because of the nature of the available intrafirm trade data, only equations on imports by the United States and equations on exports from the other major industrial countries could be estimated.

(4) The model used here is estimated on the basis of annual data, whereas the World Trade Model is estimated on the basis of semiannual data.

IV. Results

AGGREGATE U. S. MANUFACTURING IMPORT EQUATIONS

Separate equations have been estimated for U. S. manufacturing imports from MOFAs and for conventional imports. In addition, equations have been estimated for total imports, to assist comparison with other studies and to illustrate the implications of not distinguishing between the two types of trade. The exact form of the equations estimated is

\begin{matrix} L N (M) = & B_{0} + B_{1} L N (AVG) \\ + \sum_{i = 0}^{n} α_{i} L N (P / PM)_{- i} + t + CAR \end{matrix}

where t denotes a linear time trend and CAR is a dummy variable used to represent the effects on trade of the 1965 U. S.-Canadian Automotive Products Trade Agreement. Two alternative specifications have been adopted for the lag structure of the relative price term: a simple version with relative prices lagged one and two years, and a version with a quadratic lag structure, over four years, estimated using the technique developed by Almon (1965). The two specifications gave similar results (Table 5).

Table 5.

Test Results for Aggregate U. S. Manufacturing Import Equations¹

^¹Figures in parentheses are t-statistics; ^* and ^** represent significance at the 5 and 1 per cent levels of confidence, respectively. D-W denotes the Durbin-Watson statistic. The relative price terms have U. S. domestic prices in the numerator, so that the theoretically expected sign for the price elasticities is positive. See Appendix II for the exact definition of variables used. The period of estimation is 1962–76.

Table 5.

Test Results for Aggregate U. S. Manufacturing Import Equations¹

					$\sum_{i = 1}^{3} α_{1^{*}}$
	Constant	LN(AVG)	LN(P/PM)₋₁	LN(P/PM)₋₂	LN(P/PM)₋₁	Time Trend	CAR Dummy	R²	D-W
Imports from MOFAs (M^′)	7.56	0.91^*			−3.59	0.076^**	0.331^**	0.961	1.72
	(32.5)	(2.12)			(1.75)	(5.05)	(5.02)
Conventional imports (M)	8.88	1.02^*			3.00^**	0.101^**	−0.037^**	0.999	2.80
	(210.9)	(6.11)			(8.07)	(20.12)	(0.95)
Total imports (M^′+M)	8.97	1.02^*			3.48^**	0.108^**	0.024^**	0.999	2.73
	(147.9)	(4.49)			(5.80)	(14.70)	(0.45)
Imports from MOFAs (M^′)	7.57	0.90^*	−1.66	−1.30		0.074^**	0.329^**	0.950	1.74
	(32.7)	(2.29)	(1.32)	(1.76)		(4.98)	(4.34)
Conventional imports (M)	8.87	0.92^*	0.54	2.01^**		0.098	−0.002	0.999	2.72
	(189.2)	(5.46)	(1.29)	(5.25)		(19.19)	(0.03)
Total Imports (M^′+M)	8.97	0.94^*	0.59	2.54^**		0.105^**	0.049	0.998	2.67
	(150.7)	(4.75)	(1.08)	(5.31)		(16.25)	(1.23)

Table 5.

Test Results for Aggregate U. S. Manufacturing Import Equations¹

					$\sum_{i = 1}^{3} α_{1^{*}}$
	Constant	LN(AVG)	LN(P/PM)₋₁	LN(P/PM)₋₂	LN(P/PM)₋₁	Time Trend	CAR Dummy	R²	D-W
Imports from MOFAs (M^′)	7.56	0.91^*			−3.59	0.076^**	0.331^**	0.961	1.72
	(32.5)	(2.12)			(1.75)	(5.05)	(5.02)
Conventional imports (M)	8.88	1.02^*			3.00^**	0.101^**	−0.037^**	0.999	2.80
	(210.9)	(6.11)			(8.07)	(20.12)	(0.95)
Total imports (M^′+M)	8.97	1.02^*			3.48^**	0.108^**	0.024^**	0.999	2.73
	(147.9)	(4.49)			(5.80)	(14.70)	(0.45)
Imports from MOFAs (M^′)	7.57	0.90^*	−1.66	−1.30		0.074^**	0.329^**	0.950	1.74
	(32.7)	(2.29)	(1.32)	(1.76)		(4.98)	(4.34)
Conventional imports (M)	8.87	0.92^*	0.54	2.01^**		0.098	−0.002	0.999	2.72
	(189.2)	(5.46)	(1.29)	(5.25)		(19.19)	(0.03)
Total Imports (M^′+M)	8.97	0.94^*	0.59	2.54^**		0.105^**	0.049	0.998	2.67
	(150.7)	(4.75)	(1.08)	(5.31)		(16.25)	(1.23)

The results show that there is a striking difference in the response of conventional and intrafirm U. S. imports to shifts in observed relative prices. The relative price terms in the conventional import equations are significant at the 1 per cent level of confidence in both the simple, two-period lag and in the quadratic form, whereas the price terms in the intrafirm import equations are insignificant and of the wrong sign. The test derived by Chow (1960) has been used to determine whether the differences in the estimated price coefficients for the two types of trade resulted from their derivation from underlying populations with significantly different responses to relative prices. The test involves estimating a pooled regression of both the conventional and intrafirm trade observations with the coefficients of the price terms constrained to be equal for the two types of trade, but all other coefficients being allowed to vary as before, that is,

\begin{matrix} [\begin{matrix} M \\ M^{'} \end{matrix}] = [\begin{matrix} α_{0} \\ α_{0} \end{matrix}] + α_{1} [\begin{matrix} LN (AVG) \\ 0 \end{matrix}] \\ + {α^{'}}_{1} [\begin{matrix} 0 \\ {LN (AVG)}^{'} \end{matrix}] + α_{2} & [\begin{matrix} LN (P / PM) \\ LN (P / PM)^{'} \end{matrix}]_{- 1} \\ + α_{3} [\begin{matrix} LN (P / PM) \\ LN (P / PM)^{'} \end{matrix}]_{- 2} + α_{4} & [\begin{matrix} t \\ 0 \end{matrix}] + {α^{'}}_{4} [\begin{matrix} 0 \\ t \end{matrix}] \end{matrix}

The sum of the squared residuals from the constrained regression are then compared with the sum of squared residuals from the unconstrained regressions, to test whether the constraint is significant.¹⁰ The test showed that the relative price coefficients for the two types of trade were significantly different at the 1 per cent level of confidence.

By contrast, there seems to be no significant difference in the response of the two types of trade to aggregate demand, nor in the time trend coefficient. The estimated parameters for these terms are fairly close to those obtained in the Fund’s World Trade Model. The CAR dummy for the U. S.-Canadian Automotive Products Trade Agreement is significant only in the intra-firm trade equation, which is the expected result, since most automotive products exported from Canada to the United States are from affiliates of U. S. companies.

EQUATIONS FOR DISAGGREGATED U. S. MANUFACTURING IMPORTS

It has already been shown that the composition of intrafirm trade differs from that of conventional trade. Therefore, to show that the difference in the price responsiveness of the two types of trade in the aggregate equations is not solely due to a difference in commodity composition, several equations have also been estimated for particular product groups of U. S. imports. The same model is used as for the aggregate import equations, except that the relative price terms now refer to the domestic price of the particular product group relative to competing imports in the same product group. As before, the import prices for intrafirm and conventional trade are weighted averages of export prices in the major supplying countries, with weights equal to the share of the supplying country in each type of trade. However, export prices (or unit value) data could be found only for Canada, France, the Federal Republic of Germany, Japan, and the United Kingdom for three industries only: chemicals, electrical products, and transport equipment (see Appendix II). Fortunately, these five countries account for the majority of U. S. imports in those particular product groups (supplying 64 per cent, 69 per cent, and 93 per cent of total 1970 U. S. imports of chemicals, electrical products, and transport equipment, respectively).

The results are given in Table 6. As for the aggregate import equations, the period of estimation is 1962–76. Because the equations using an Almon quadratic lag for the price terms gave similar results to those using the simple price term lagged one and two periods, only the latter are reported. The results for imports of chemical and electrical products are similar to those of the aggregate equations, with significant price elasticities for conventional imports, but the intrafirm imports show no significant response to relative price changes. The Chow test for equality of the price coefficients showed that the price responsiveness of the two types of trade were significantly different at the 1 and 10 per cent levels for imports of chemical and electrical products, respectively.¹¹ Indeed, for these two industries, the whole model fits the data on conventional trade flows much better than for intrafirm trade, where only the time trend variable is significant at the 5 per cent level. Again, this illustrates the problems that arise from fitting a model derived from consumer demand theory to intrafirm trade flows, which are basically determined by the internal location decisions of the multinational firm. The implications of this for models of total world trade that combine conventional and intrafirm trade can be seen from the total import equations. The estimated price elasticities for these two industries are lower and less significant in the equations for total imports than in the equations for conventional imports alone. Once the intrafirm trade is removed from the total, the consumer demand model gives a better picture of the effects of relative prices on trade flows.

Table 6.

Equation for Disaggregated U. S. Imports¹

^¹Figures in parentheses are t-statistics; ^* and ^** represent significance at the 5 and 1 per cent levels of confidence, respectively. D-W denotes the Durbin-Watson statistic. The relative price terms have U. S. domestic prices in the numerator, so that the theoretically expected sign for the price elasticities is positive. See Appendix II for the exact definitions of variables used. The period of estimation is 1962–76.

Table 6.

Equation for Disaggregated U. S. Imports¹

				Price Terms
Dependent Variables		Constant	Demand LN(AVG)	LN(P_i/PM_i)₋₁	LN(P_i/PM_i)₋₂	Time Trend	CAR Dummy	R²	D-W
CHEMICALS
	Imports from MOFAs $(M_{1}^{'})$	0.17	1.02	−0.46	−1.00	−0.041^*		0.919	2.10
			(1.45)	(0.33)	(0.76)	(1.91)
	Conventional imports (M₁)	1.63	0.69^**	0.45^**	0.92^**	0.099^**		0.999	2.42
			(6.89)	(3.20)	(6.03)	(33.12)
	Total imports $(M_{1}^{'} + M_{1})$	1.82	0.64^**	0.06	1.17^**	0.092^**		0.999	2.27
			(6.33)	(0.41)	(6.70)	(30.50)
ELECTRICAL PRODUCTS
	Imports from MOFAs $(M_{2}^{'})$	1.83	1.83	−1.69	0.87	0.146^**		0.941	1.93
			(1.36)	(0.73)	(0.39)	(3.03)
	Conventional imports (M₂)	1.54	2.16^**	1.73^*	0.02	0.148^**		0.982
			(3.22)	(2.19)	(0.02)	(6.53)
	Total imports $(M_{2}^{'} + M_{2})$	1.63	2.11^**	1.19	0.28	0.151^**		0.986	1.50
			(3.42)	(1.55)	(0.35)	(7.88)
TRANSPORT EQUIPMENT
	Imports from MOFAs $(M_{3}^{'})$	0.66	1.59	5.04	8.49^*	0.150^**	1.33^**	0.987	1.96
			(1.13)	(1.07)	(1.87)	(5.05)	(4.65)
	Conventional imports (M₃)	1.65	0.72	0.61	1.79^*	0.189^**	−0.085	0.994	2.70
			(1.33)	(0.84)	(2.47)	(10.80)	(0.78)
	Total imports $(M_{3}^{'} + M_{3})$	1.88	0.90	1.49	2.97^*	0.192^**	0.40^**	0.988	1.80
			(1.41)	(1.11)	(2.39)	(8.02)	(3.22)

^¹Figures in parentheses are t-statistics; ^* and ^** represent significance at the 5 and 1 per cent levels of confidence, respectively. D-W denotes the Durbin-Watson statistic. The relative price terms have U. S. domestic prices in the numerator, so that the theoretically expected sign for the price elasticities is positive. See Appendix II for the exact definitions of variables used. The period of estimation is 1962–76.

Table 6.

Equation for Disaggregated U. S. Imports¹

				Price Terms
Dependent Variables		Constant	Demand LN(AVG)	LN(P_i/PM_i)₋₁	LN(P_i/PM_i)₋₂	Time Trend	CAR Dummy	R²	D-W
CHEMICALS
	Imports from MOFAs $(M_{1}^{'})$	0.17	1.02	−0.46	−1.00	−0.041^*		0.919	2.10
			(1.45)	(0.33)	(0.76)	(1.91)
	Conventional imports (M₁)	1.63	0.69^**	0.45^**	0.92^**	0.099^**		0.999	2.42
			(6.89)	(3.20)	(6.03)	(33.12)
	Total imports $(M_{1}^{'} + M_{1})$	1.82	0.64^**	0.06	1.17^**	0.092^**		0.999	2.27
			(6.33)	(0.41)	(6.70)	(30.50)
ELECTRICAL PRODUCTS
	Imports from MOFAs $(M_{2}^{'})$	1.83	1.83	−1.69	0.87	0.146^**		0.941	1.93
			(1.36)	(0.73)	(0.39)	(3.03)
	Conventional imports (M₂)	1.54	2.16^**	1.73^*	0.02	0.148^**		0.982
			(3.22)	(2.19)	(0.02)	(6.53)
	Total imports $(M_{2}^{'} + M_{2})$	1.63	2.11^**	1.19	0.28	0.151^**		0.986	1.50
			(3.42)	(1.55)	(0.35)	(7.88)
TRANSPORT EQUIPMENT
	Imports from MOFAs $(M_{3}^{'})$	0.66	1.59	5.04	8.49^*	0.150^**	1.33^**	0.987	1.96
			(1.13)	(1.07)	(1.87)	(5.05)	(4.65)
	Conventional imports (M₃)	1.65	0.72	0.61	1.79^*	0.189^**	−0.085	0.994	2.70
			(1.33)	(0.84)	(2.47)	(10.80)	(0.78)
	Total imports $(M_{3}^{'} + M_{3})$	1.88	0.90	1.49	2.97^*	0.192^**	0.40^**	0.988	1.80
			(1.41)	(1.11)	(2.39)	(8.02)	(3.22)

^¹Figures in parentheses are t-statistics; ^* and ^** represent significance at the 5 and 1 per cent levels of confidence, respectively. D-W denotes the Durbin-Watson statistic. The relative price terms have U. S. domestic prices in the numerator, so that the theoretically expected sign for the price elasticities is positive. See Appendix II for the exact definitions of variables used. The period of estimation is 1962–76.

The results for the transport-equipment equations run contrary to those of the other product groups, with the estimated price elasticity of intrafirm trade being much higher than the price elasticity of conventional trade. This is largely due to the effect of the 1965 U. S.-Canadian Automotive Products Trade Agreement, which liberalized U. S.-Canadian trade in automobile products by virtually eliminating trade barriers between the two countries. When there have been few barriers to trade, the response of trade flows to any relative price changes has been very large. As virtually all Canadian automobile plants are subsidiaries of U. S. companies, most of this price-sensitive trade also happens to be intrafirm trade. This can be seen from the coefficients on the CAR dummy variable, which represents the effects of the Agreement and is very large and highly significant in the intrafirm equation but completely insignificant in the conventional import equations. Therefore, the high price elasticity of intrafirm imports of transport equipment is mainly due to this specific trade agreement rather than to the special nature of intrafirm trade.

AGGREGATE EXPORT EQUATIONS

The major problem in estimating export equations for intrafirm trade is that data on intrafirm exports from U.S. affiliates in the major industrial countries is available only for the period 1966–76.¹² Therefore, any equations estimated over the period are necessarily less reliable than the earlier import equations, for which a longer time series is available.

The basic export equation estimated is

LN(X_i) = α₀ + α₁ LN(FM_i) + α₂ LN(RXPM_i)_-1 + α₃ LN(RXPM_i)_-2

where X_i denotes the volume of intrafirm or conventional exports from country i and FM_i denotes a weighted average index of demand for country is exports, with weights equal to the share of intrafirm or conventional exports, respectively, in a particular market. Similarly, RXPM_i represents an index of the price of country i’s exports relative to a weighted index of competitors’ prices, with the weights again depending on the share of country i’s intrafirm or conventional exports in each market (see Appendix II).

The results for the United Kingdom, Belgium, France, the Netherlands, and Switzerland, shown in Table 7,¹³ are similar to those for the earlier U.S. import equations. With the exception of the United Kingdom, the price elasticity of intrafirm exports is generally lower than that for conventional exports. Again, with the exception of the United Kingdom, the model has a poorer fit for intrafirm exports from U. S. affiliates in these countries than for conventional trade. However, the small number of observations (11) on which the estimates are based makes it difficult to come to very firm conclusions. The Chow test for differences in the price elasticities for the two types of trade shows significant differences for only Switzerland, Belgium, and, to a lesser extent, the Netherlands. There is no significant difference between the estimated price elasticities of conventional and intrafirm trade for the United Kingdom and France.¹⁴

Table 7.

Simple Equations for Conventional and U. S. Majority-Owned Foreign Affiliates’ (MOFAS’) Manufacturing Export Trade¹

^¹The figures in parentheses are t-statistics; ^* and ^** indicate significance at the 1 and 5 per cent levels, respectively. D-W denotes the Durbin-Watson statistic. The period of estimation is 1966–76. See Appendix II for the exact definition of all variables used. As the exporting country’s price index is the numerator, the theoretically expected sign of the price terms is negative.

Table 7.

Simple Equations for Conventional and U. S. Majority-Owned Foreign Affiliates’ (MOFAS’) Manufacturing Export Trade¹

Dependent Variable		Constant	Demand LN(FM_i)	LN(RXPM_i)₋₁	LN(RXPM_i)₋₂	R²	D-W
united kingdom
	MOFA exports $(X_{1}^{'})$	0.50	0.78^**	−0.62		0.949	1.46
			(7.74)	(0.93)
	Conventional exports (X₁)	6.26	0.34^**	−0.39		0.796	2.01
			(4.06)	(1.39)
belgium
	MOFA exports $(X_{2}^{'})$	−5.69	1.42^**	−0.60		0.834	1.90
			(5.64)	(0.40)
	Conventional exports (X₂)	2.19	0.74^**	−1.05^*	−1.52	0.993	1.91
			(24.51)	(2.83)	(4.13)
france
	MOFA exports $(X_{3}^{'})$	−4.35	1.20^**		−2.16	0.990	1.86
			(19.87)			(1.58)
	Conventional exports (X₃)	0.13	0.98^**	−1.07^**	−0.41^*	0.999	2.94
			(127.80)	(6.71)	(2.22)
netherlands
	MOFA exports $(X_{4}^{'})$	−8.55	1.79*		−1.04	0.992	2.52
			(2.71)		(0.88)
	Conventional exports (X₄)	0.95	0.88^**	−0.77^**	−0.34	0.997	1.82
			(40.80)	(3.53)	(1.62)
switzerland
	MOFA exports $(X_{5}^{'})$	3.40	0.28	1.40		0.430	1.46
	Conventional exports (X₅)	2.78	0.66^**	−0.65^*		0.932	1.53
			(5.37)	(1.86)

Table 7.

Simple Equations for Conventional and U. S. Majority-Owned Foreign Affiliates’ (MOFAS’) Manufacturing Export Trade¹

Dependent Variable		Constant	Demand LN(FM_i)	LN(RXPM_i)₋₁	LN(RXPM_i)₋₂	R²	D-W
united kingdom
	MOFA exports $(X_{1}^{'})$	0.50	0.78^**	−0.62		0.949	1.46
			(7.74)	(0.93)
	Conventional exports (X₁)	6.26	0.34^**	−0.39		0.796	2.01
			(4.06)	(1.39)
belgium
	MOFA exports $(X_{2}^{'})$	−5.69	1.42^**	−0.60		0.834	1.90
			(5.64)	(0.40)
	Conventional exports (X₂)	2.19	0.74^**	−1.05^*	−1.52	0.993	1.91
			(24.51)	(2.83)	(4.13)
france
	MOFA exports $(X_{3}^{'})$	−4.35	1.20^**		−2.16	0.990	1.86
			(19.87)			(1.58)
	Conventional exports (X₃)	0.13	0.98^**	−1.07^**	−0.41^*	0.999	2.94
			(127.80)	(6.71)	(2.22)
netherlands
	MOFA exports $(X_{4}^{'})$	−8.55	1.79*		−1.04	0.992	2.52
			(2.71)		(0.88)
	Conventional exports (X₄)	0.95	0.88^**	−0.77^**	−0.34	0.997	1.82
			(40.80)	(3.53)	(1.62)
switzerland
	MOFA exports $(X_{5}^{'})$	3.40	0.28	1.40		0.430	1.46
	Conventional exports (X₅)	2.78	0.66^**	−0.65^*		0.932	1.53
			(5.37)	(1.86)

V. Conclusions

This paper shows that a significant proportion of total manufacturing trade takes the form of internal transactions by multinational firms, and that the importance of such intrafirm trade varies substantially from country to country, being particularly important for the trade of the United States and the United Kingdom and relatively unimportant in, for example, Japan. There are reasons for expecting the price responsiveness of such trade to be lower than that for conventional trade, both because intrafirm trade is more likely to concern products that are specific to the company, with fewer close substitutes, and also because the affiliates of the same multinational company are likely to compete less vigorously than independent firms that are unconcerned with the effects of their actions on the profitability of plants in other countries. It has also been shown that, because of the transfer-pricing phenomenon, price elasticities estimated for intrafirm trade would be expected to have a downward bias.

Traditional models of world trade generally combine intrafirm and conventional trade flows. This paper has estimated separate import and export equations for each type of trade in order to test whether significant differences occur in their responsiveness to relative price changes. A simplified version of the Fund’s World Trade Model was adapted for this purpose. The results, particularly for the U. S. import equations, for which longer time series are available, show that observed price elasticities are significantly lower for intrafirm than for conventional trade. This is also generally true when the imports are disaggregated by industry group. In addition, the trade model used, which is derived from consumer demand theory, has a poorer fit for the data on intrafirm trade than for conventional trade. This is not surprising, because the pattern of intrafirm trade is essentially determined by the locations of the affiliates of multinational firms.

The observed differences in the price responsiveness of intrafirm and conventional trade may also have implications for external adjustment. The growth in the share of manufacturing trade that is internal to the firm may lower the price elasticity of trade flows, especially between the industrial countries, where most intrafirm trade is concentrated. In addition, the wide variations in the importance of intrafirm trade in different countries affect the relative speed of adjustment of their manufacturing trade to price changes.

APPENDICES

I. Econometric Implications of Using Artificial Transfer Prices in Intrafirm Trade

If the true model of intrafirm imports (small letters denoting log terms and all variables measured as deviations from means, to eliminate the constant term) is

m = απ + βy + e

but observed transfer prices, p, differ from the true prices, π, by a component, v, so that

p = π + v

then the model actually estimated is

m = α(p − v) + βy + e = αp + βy + (e − αv)

If all the ordinary-least-squares assumptions continue to hold, except that the error term is now correlated with the price variable, then the expected value of the estimated price elasticity will be

E(α) = α − αb_vp·y

where b_vp·y denotes the partial regression coefficient of p in the auxiliary regression of v on y and p (see Griliches (1957) for a more detailed discussion of the terminology). This auxiliary regression is

v = bp + δy + u

v = b(π + v) + δy + u

therefore,¹⁵

\begin{matrix}  \end{matrix} \begin{matrix} b_{vp . y} & = \frac{b_{vp}}{1 - r_{yp}^{2}} \\ = \frac{cov (v, p)}{var (p)} \cdot \frac{1}{1 - r_{yp}^{2}} \\ = \frac{cov (v, v + π)}{var (v + π)} \frac{1}{1 - r_{yp}^{2}} \\ = \frac{cov (π, v) + var (v)}{cov (π, v) + var (v) + var (π)} \cdot \frac{1}{1 - r_{yp}^{2}} \end{matrix}

and therefore,

bias = E (α) - α = \frac{- λα}{1 - r_{yp}^{2}}

where

λ = \frac{cov (π, v) + var (v)}{cov (π, v) + var (v) + var (π)}

therefore, if

cov (π, v) ≧ 0, then 0 < λ < 1 and E(α) is biased downward; if
−vari(v) < cov(π, v) < 0

then 0 < λ < 1 and E(α) is still biased downward; if

−var(v) − var(π) < cov(π, v) − var(v)

then λ < 0 and E(α) is biased upward; if

cov(π, v) < − var(v) − var(π)

then λ > 1 and E(α) is biased downward.

In the conventional measurement error case, cov(π, v) is usually regarded as zero, so 0 < λ < 1 and E(α) is biased downward, with the magnitude of the downward bias being greater, the larger is the variance of the measurement error relative to the variance of the true price. But in the case of measurement error owing to transfer prices, generally a systematic relationship exists between the measurement error and the true price, so that cov(π, v) ≠ 0 and the following cases must be distinguished:

(a) If there is an incentive to overstate transfer prices, then we would expect cov(π, v) > 0 (for instance, if transfer prices were a percentage markup on true prices). In this case, the estimated price elasticity will again be biased downward in absolute amount, and if $λ > (1 - r_{yp}^{2})$ , the bias will be large enough to result in an estimated price elasticity of the wrong sign.

(b) If there is an incentive to understate transfer prices, so that cov(π, v) < 0 (for instance, if transfer prices were marked down to some fraction of true prices), the estimated price elasticity may be biased upward or downward, depending on the relative sizes of the covariances and variances of π and v. To investigate this further, it may be recalled that it has been shown that the estimated price elasticity is biased upward only if

− var(v) − var(π) < cov(π, v) < − var(v)

Dividing through by var(π) yields

\frac{- var (v)}{var (π)} - 1 < \frac{cov (π, v)}{var (π)} < \frac{- var (v)}{var (π)}

− (1 + c) < b_vπ < − c

where

c = \frac{var (v)}{var (π)}

and b_vπ is the regression coefficient of π in the regression of v on π, that is,

v = b₀ + b_{v π} · + E

It is known that b_{v π} > −1 (otherwise a rise in the true price would mean that the firm would reduce its transfer price—an unlikely move). Therefore, it is only in the range of −1 < b_{v π} < −c that an upward bias in estimated price elasticities will result. For instance, if the variance of the true price were twice the variance of the error term, so that c = ½, the firm would have to follow a policy of changing its transfer price by less than half the amount of the true price change before an upward bias in the estimated price elasticity would result.

II. Data Sources

All data are on an annual basis. All time series in index form have 1970 as the base year.

Data for Equations on Aggregate U. S. Imports

M′ = Manufacturing imports into the United States from majority-owned foreign affiliates (MOFAs) of U. S. multinationals, taken from the Survey of Current Business (United States, Department of Commerce), various issues. The data are based on annual sample surveys, which are then extrapolated to obtain universal estimates from census benchmark years, the latest of which was 1966.¹⁶ Estimates for years prior to 1966 are derived from an earlier benchmark year (1957) and have been adjusted proportionately to the 1966 base to give a consistent time series. This was then deflated by the index of import prices for MOFAs’ trade, PM′.

M = Conventional manufacturing imports into the United States, equal to total manufacturing imports less imports from the MOFAs. The series was deflated by the index of import prices for conventional trade, PM.

PM′ = Estimated price index of manufactured imports into the United States from MOFAs. This is a weighted average of the export-unit-value indices of the major industrial countries exporting to the United States¹⁷ with the weights sm_j equal to each country’s share in total MOFA manufacturing imports into the United States, that is,

LM = ({PM}^{'}) = \sum_{j} {sm}_{j} \cdot LN ({XPM}_{j})

The export unit value indices (in dollar terms), XPM_j, are taken from the Fund’s World Trade Model.

PM = Price index of conventional manufactured imports into the United States. This is calculated in a similar manner to PM′, except that the weights are each country’s share in conventional U. S. manufacturing imports.

P = Index of domestic prices of manufactures in the United States, taken from the Fund’s World Trade Model.

AVG = Index of real demand for manufactures in the United States, taken from the Fund’s World Trade Model. This is an arithmetically weighted average of output in manufacturing and real final domestic demand for manufactures, with the weights equal to the share of manufactured imports going to intermediate and final demand, respectively.

CAR = This is a dummy variable to represent the U. S.-Canadian Automotive Products Trade Agreement. It takes a value of 0 up to 1965 and 1 thereafter.

Data for Equations on Disaggregated U. S. Imports

$M_{i}^{'}$ = Imports into the United States from MOFAs in industry i (chemicals, transport equipment, and electrical products, respectively) from the same source as M′ (defined previously), deflated by the index of MOFA import prices for that industry, PM_i.

PM_i = Price index of conventional imports into the United States in industry i. Again, this is a weighted average of the supplying countries’ export price indices, with weights equal to each country’s share in conventional imports into that industry. However, export price indices by industry could be derived for only Canada, France, the Federal Republic of Germany, Japan, and the United Kingdom. The sources for these export price indices (or export unit values) are as follows:

Canada: No disaggregated export price indices were available. Indices of Industrial Selling Prices for particular subcategories, weighted by the share of each subcategory in Canada’s export trade, were used to construct suitable industry indices. The Industrial Selling Price indices were taken from the Canadian Statistical Review (Statistics Canada), various issues.

Federal Republic of Germany: Export price indices were taken from Preise und Preisindizes für die Ein und Ausfuhr, Fact Series 17, Issue 8 (Statistiches Bundesamt).

France: Export unit value indices were calculated from the export volume and value indices given in Annuaire Statistique de la France (Institut national de la statistique et des études économiques), various issues.

Japan: Export price indices from Price Indexes Annual (Statistics Department, Bank of Japan), various issues.

United Kingdom: Export unit value indices from the Monthly Digest of Statistics (Central Statistical Office), various issues.

Each of the price indices was converted into dollar terms by adjusting for exchange rate movements.

${PM}_{i}^{'}$ = Price index of MOFA imports into the United States in industry i. This is calculated in the same manner as PM_i, except that the weights are now equal to the particular country’s share in U. S. imports from MOFAs in industry i.

P_i = U. S. wholesale price index for industry i, taken from the Monthly Labor Review and the Handbook of Labor Statistics, (United States, Department of Labor, Bureau of Labor Statistics), various issues.

Data for Equations on Exports

$X_{i}^{'}$ = Manufacturing exports from country i by MOFAs of U. S. multinationals, taken from the Survey of Current Business (U. S. Department of Commerce, Bureau of Economic Analysis). These are deflated by the manufacturing export price index for that country, in dollar terms.

X_i = Conventional manufacturing exports from country i, equal to manufacturing exports less exports by MOFAs. These are deflated by the manufacturing export price index, in dollar terms.

RXPM_i = Index of export prices for conventional manufactures from country i relative to a composite index of competitors’ prices, that is, LN(RXPM_i) = $LN ({XPM}_{i}) - \sum_{j} {sx}_{ij} \cdot LN ({PM}_{j})$ where XPM_i denotes the index of export unit values of manufactures from country i (in U. S. dollars), PM_i denotes the aggregate price index (in U. S. dollars) of conventional manufactured imports in market j, and sx_ij denotes the share of country i’s conventional manufactured imports that goes to market j in the base year 1970.¹⁸

${RXPM}_{i}^{'}$ = An index of export prices for MOFA exports from country i relative to a composite index of competitors’ prices. This is derived as for RXPM_i except that the weights now refer to the share of country i’s MOFA exports, which go to each market j. (Data on the destination of MOFA exports from a particular country are generally available only for regional destinations, covering a group of markets. The shares within a group have been estimated on the assumption that they are distributed proportionately to total manufacturing exports, obtained from the Fund’s World Trade Model.)

FM_i = An index of foreign market demand for conventional manufactures from country i, that is,

{FM}_{i} = \sum_{j} {sx}_{ij} \cdot M_{j}

where sm_ij denotes country i’s share of country j’s market for conventional manufactured goods and M_j, denotes an index of the volume of manufactured imports into country j.

${FM}_{i}^{'}$ An index of foreign market demand for MOFA exports from country i, derived in a similar manner to FM_i Again, the full-share matrices sm_ij could not be calculated from available data and had to be partially estimated from the import-share matrix for total manufactures used in the Fund’s World Trade Model.

The market share for MOFA trade, on which the various weighting schemes are based, were calculated partly from data contained in the Special Survey of U. S. Multinational Companies, 1970 (United States, Department of Commerce) and partly from some unpublished data collected for the Special Survey and provided by the International Investment Division, Bureau of Economic Analysis, U. S. Department of Commerce.

REFERENCES

Almon, Shirley, “The Distributed Lag Between Capital Appropriations and Expenditures,” Econometrica, Vol. 33 (January 1965), pp. 178–96.
- Search Google Scholar
- Export Citation
Armington, Paul S., “A Theory of Demand for Products Distinguished by Place of Production,” Staff Papers, Vol. 16 (March 1969), pp. 159–78.
- Search Google Scholar
- Export Citation
Buckley, P. J., and R. D. Pearce, “Overseas Production and Exporting by the World’s Largest Enterprises: A Study in Sourcing Policy” (unpublished, University of Reading, Department of Economics, September 1977).
- Search Google Scholar
- Export Citation
Chow, Gregory C., “Tests of Equality Between Sets of Coefficients in Two Linear Regressions,” Econometrica, Vol. 28 (July 1960), pp. 591–605.
- Search Google Scholar
- Export Citation
Deppler, Michael C., and Duncan M. Ripley, “The World Trade Model: Merchandise Trade,” Staff Papers, Vol. 25 (March 1978), pp. 147–206.
- Search Google Scholar
- Export Citation
Great Britain Monopolies and Restrictive Practices Commission, “Chlordiazepoxide and Diazepam” (Her Majesty’s Stationery Office, London, 1973).
- Search Google Scholar
- Export Citation
Griliches, Zvi, “Specification Bias in Estimates of Production Functions,” Journal of Farm Economics, Vol. 39 (February 1957), pp. 8–20.
- Search Google Scholar
- Export Citation
Hogan, W. P., “British Manufacturing Subsidiaries in Australia and Export Franchises,” Economic Papers, the Economic Society of Australia and New Zealand, New South Wales and Victorian Branches (July 1966), pp. 10–27.
- Search Google Scholar
- Export Citation
Lall, Sanjaya, “Transfer-Pricing by Multinational Manufacturing Firms,” Oxford Bulletin of Economics and Statistics, Vol. 35 (August 1973), pp. 173–95.
- Search Google Scholar
- Export Citation
Lall, Sanjaya, “The Pattern of Intra-Firm Exports by U. S. Multinationals,” Oxford Bulletin of Economics and Statistics, Vol. 40 (August 1978), pp. 209–22.
- Search Google Scholar
- Export Citation
Porter, M. E., and A. Spence, “Vertical Integration and Different Inputs” (unpublished, University of Warwick, 1977).
- Search Google Scholar
- Export Citation
Strotz, Robert H., “The Empirical Implications of a Utility Tree,” Econometrica, Vol. 25 (April 1957), pp. 269–80.
- Search Google Scholar
- Export Citation
Strotz, Robert H., “The Utility Tree—A Correction and Further Appraisal,” Econometrica, Vol. 27 (July 1959), pp. 482–88.
- Search Google Scholar
- Export Citation
United Nations Conference on Trade and Development, “Restrictive Business Practice: Studies on the United Kingdom of Great Britain and Northern Ireland, the United States of America and Japan,” Report No. TD/B/390 (New York, 1973).
- Search Google Scholar
- Export Citation
United States, Department of Commerce, Bureau of Economic Analysis, Special Survey of U.S. Multinational Companies, 1970 (Washington, November 1972).
- Search Google Scholar
- Export Citation
United States, Tariff Commission, “Implications of Multinational Firms for World Trade and Investment and for U. S. Trade and Labor: Report to the Committee on Finance of the U. S. Senate and its Subcommittee on International Trade on Investigation No. 332–69, Under Section 332 of the Tariff Act of 1930” (Washington, February 1973).
- Search Google Scholar
- Export Citation

Mr. Goldsbrough, economist in the Special Studies Division of the Research Department when this paper was prepared, is currently an economist in the African Department. He is a graduate of Cambridge University and received his master’s degree and doctorate from Harvard University.

Mr. Smith W. Allnutt of the International Investment Division, Bureau of Economic Analysis, U.S. Department of Commerce, provided some unpublished statistics from the Special Survey of U. S. Multinational Companies, 1970 that were needed to calculate the weights used in the demand and price terms of the estimating equations.

See Trade and Industry (U. K. Department of Industry) (April 13, 1972), Table 3, p. 50.

Similar results are obtained by Lall (1978). A possible explanation of why intrafirm trade is more important in technology-intensive industries is discussed later.

A number of other examples of such restrictions on the export trade of overseas affiliates of multinational companies are given in United Nations Conference on Trade and Development (1973).

The economics of the specialized versus the standardized input are discussed in greater detail in Porter and Spence (1977).

Export equations for conventional and intrafirm trade are estimated for the United Kingdom, Belgium, France, the Netherlands, Switzerland, Canada, and the Federal Republic of Germany.

This is the “separability” assumption of consumer demand theory. See, for instance, Strotz (1957, 1959).

This paper is concerned with testing for significant differences in parameters for conventional and intrafirm imports, using the same model. More fundamental objections could be raised to the whole concept of applying a model based on consumer demand theory to intrafirm trade, which is essentially determined by decisions taken within the firm. Some attempts have been made to construct a completely different model for intrafirm trade, based on a choice-of-location framework, but these have not succeeded to any great extent.

See Deppler and Ripley (1978), pp. 161–62 for a more detailed description of the derivation.

Belgium, Canada, the Federal Republic of Germany, France, Italy, Japan, the Netherlands, Switzerland, the United Kingdom, and the United States. The World Trade Model covers these same ten countries plus Austria, Denmark, Norway, and Sweden.

The F-test for this is F(R^*,R) = [(SSR_r − SSRun)/R*]/[SSRun/R] where SSR_r and SSRun are the sum of squared residuals in the restricted and unrestricted regressions, respectively; R* is the number of restrictions; and R is the number of degrees of freedom in the unrestricted model. For this test, the result was F(2, 20) = 13.4, which is significant at the 1 per cent level of confidence.

The F−statistics on the test for significant differences in the price coefficients were

Chemicals: F(2,20) = 8.03, significant at the 1 per cent level

Electrical products: F(2,20) = 2.52, significant at the 10 per cent level

Transport equipment: F(2,18) = 10.01, significant at the 1 per cent level

A breakdown by country of origin of manufacturing exports by U. S. majority-owned foreign affiliates (MOFAs) is available only from 1966.

Equations were also estimated for Canada and the Federal Republic of Germany. However, the price terms in both the intrafirm and conventional export equations were generally insignificant and of the wrong sign. No equation could be estimated for Japan because exports from U. S. affiliates located there were negligible.

The F-statistics were

United Kingdom: F(2,14) = 0.21, not significant

Belgium: F(2,14) = 3.70, significant at the 5 per cent level

France: F(2,14) = 1.44, not significant

Netherlands: F(2,14) = 2.64, significant at the 10 per cent level

Switzerland: F(2,14) = 10.37, significant at the 1 per cent level

Strictly speaking, the argument here should be in terms of probability limits, because the expressions concern products and ratios of moments of random variables, but these terms have been avoided for ease of exposition.

See R. David Belli, Smith W. Allnutt, and Howard Murad, “Property, Plant, and Equipment Expenditures by Majority-Owned Foreign Affiliates of U. S. Companies: Revised Estimates for 1966-72 and Projections for 1973 and 1974,” Survey of Current Business, Vol. 53 (December 1973), pp. 19–32, especially pp. 21–22, for further details.

Canada, the United Kingdom, Japan, Belgium, France, the Federal Republic of Germany, Italy, the Netherlands, and Switzerland.

Again, the markets included in the weighting scheme are the United States, Japan, Canada, the United Kingdom, Belgium, France, the Federal Republic of Germany, Italy, the Netherlands, and Switzerland.

Other Publishers

Save
Cite
Email this content

Share Link

Copy this link, or click below to email it to a friend
Email this content
or copy the link directly:

https://www.elibrary.imf.org/view/journals/024/1981/003/article-A005-en.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

IMF Staff papers: Volume 28 No. 3

Author: International Monetary Fund. Research Dept.

Volume/Issue:: Volume 1981: Issue 003

Publisher:: International Monetary Fund

ISBN:: 9781451972672

ISSN:: 1020-7635

Pages:: 208

DOI:: https://doi.org/10.5089/9781451972672.024

Keywords
I. Relative Importance of Intrafirm Trade Flows
II. Response of Intrafirm Trade Flows to Changes in Relative Prices
- TRANSFER PRICES
- EFFECTS OF INTERNALIZATION OF INTERNATIONAL TRADE TRANSACTIONS ON PRICE RESPONSIVENESS
III. A Simple Model of Conventional and Intrafirm Manufacturing Trade
IV. Results
V. Conclusions
APPENDICES
- I. Econometric Implications of Using Artificial Transfer Prices in Intrafirm Trade
- II. Data Sources
REFERENCES

Table 7.

Simple Equations for Conventional and U. S. Majority-Owned Foreign Affiliates’ (MOFAS’) Manufacturing Export Trade¹

Dependent Variable		Constant	Demand LN(FM_i)	LN(RXPM_i)₋₁	LN(RXPM_i)₋₂	R²	D-W
united kingdom
	MOFA exports $(X_{1}^{'})$	0.50	0.78^**	−0.62		0.949	1.46
			(7.74)	(0.93)
	Conventional exports (X₁)	6.26	0.34^**	−0.39		0.796	2.01
			(4.06)	(1.39)
belgium
	MOFA exports $(X_{2}^{'})$	−5.69	1.42^**	−0.60		0.834	1.90
			(5.64)	(0.40)
	Conventional exports (X₂)	2.19	0.74^**	−1.05^*	−1.52	0.993	1.91
			(24.51)	(2.83)	(4.13)
france
	MOFA exports $(X_{3}^{'})$	−4.35	1.20^**		−2.16	0.990	1.86
			(19.87)			(1.58)
	Conventional exports (X₃)	0.13	0.98^**	−1.07^**	−0.41^*	0.999	2.94
			(127.80)	(6.71)	(2.22)
netherlands
	MOFA exports $(X_{4}^{'})$	−8.55	1.79*		−1.04	0.992	2.52
			(2.71)		(0.88)
	Conventional exports (X₄)	0.95	0.88^**	−0.77^**	−0.34	0.997	1.82
			(40.80)	(3.53)	(1.62)
switzerland
	MOFA exports $(X_{5}^{'})$	3.40	0.28	1.40		0.430	1.46
	Conventional exports (X₅)	2.78	0.66^**	−0.65^*		0.932	1.53
			(5.37)	(1.86)

Total MOFA exports to all countries		12,458
	Intrafirm trade	8,703
	Share of intrafirm trade in total MOFA exports to all countries (in per cent)	69.9
Total MOFA exports to the United States		4,106
	Intrafirm trade	3,664
	Share of Intrafirm trade in total MOFA exports to the United States (in per cent)	89.2

Total MOFA exports to all countries		12,458
	Intrafirm trade	8,703
	Share of intrafirm trade in total MOFA exports to all countries (in per cent)	69.9
Total MOFA exports to the United States		4,106
	Intrafirm trade	3,664
	Share of Intrafirm trade in total MOFA exports to the United States (in per cent)	89.2

Topics

Countries

Series

About

Resources

Abstract

I. Relative Importance of Intrafirm Trade Flows

II. Response of Intrafirm Trade Flows to Changes in Relative Prices

TRANSFER PRICES

EFFECTS OF INTERNALIZATION OF INTERNATIONAL TRADE TRANSACTIONS ON PRICE RESPONSIVENESS

III. A Simple Model of Conventional and Intrafirm Manufacturing Trade

IV. Results

AGGREGATE U. S. MANUFACTURING IMPORT EQUATIONS

EQUATIONS FOR DISAGGREGATED U. S. MANUFACTURING IMPORTS

AGGREGATE EXPORT EQUATIONS

V. Conclusions

APPENDICES

I. Econometric Implications of Using Artificial Transfer Prices in Intrafirm Trade

II. Data Sources

Data for Equations on Aggregate U. S. Imports

Data for Equations on Disaggregated U. S. Imports

Data for Equations on Exports

REFERENCES

Same Series

Other IMF Content

Other Publishers

Asian Development Bank

Inter-American Development Bank

International Labour Organization

The World Bank

Share Link

Table of Contents

Headings

Metrics