Competitive need limits and the U.S. generalized system of preference.
Under the Generalized System of Preferences (GSP) developed countries offer preferential tariff treatment to imports from developing countries. The GSP's stated purpose is to stimulate industrialization of developing countries by giving their exports preferential tariff treatment. While the GSP has stimulated developing countries' exports, most analysts agree that it has not lived up to its original promise primarily because most developed countries offer limited GSP coverage (Baldwin and Murray, 1977; Karsenty and Laird, 1987; MacPhee and Oguledo, 1991; and Murray, 1977).
GSP coverage is restricted in several ways. This paper examines one restriction the United States employs known as a competitive need limit. It applies to GSP-eligible imports of a beneficiary's product when these imports exceed a certain dollar value or import share. GSP-eligible imports that exceed either threshold lose preferential tariff treatment. Thus, competitive need limits restrict the benefits of the U.S. GSP. Competitive need limits are one of many restrictions that the United States and other developed countries use in their GSP schemes. Others include import ceilings, value-added requirements, and exclusions for import-sensitive industries.
Analyzing competitive need limits is important for several reasons. First, in 1994, these limits applied to imports valued at $4.2 billion. The fact that competitive need limits affect a small number of products and countries magnified the impact on these imports. Just 10 competitive need limits covered more than half of affected imports in 1994. Workers and producers in developing country industries targeted by competitive need limits can suffer greatly.
Second, competitive need limits represent one of many restrictions that the United States uses to limit its GSP coverage. While the effect of any one restriction may seem minor, their combined impact is substantial. Without these restrictions, imports receiving preferential treatment under the U.S. GSP would have increased from $18.4 billion in 1994 to $29.2 billion, or by about 60%. Competitive need limits accounted for a large fraction of the U.S. imports denied GSP treatment.
Finally, analyzing competitive need limits helps identify who benefits from these limits. Competitive need limits apply only when import competition from a developing country is serious. Thus, domestic producers likely are the principal beneficiaries, especially when the import share of affected imports is high. In these cases, the developing country affected by a competitive need limit is likely to represent the only real competitive threat to a U.S. industry.
In the only previous study of the effects of competitive need limits, MacPhee and Rosenbaum (1989) find that competitive need limits reduce the market share of affected imports by an unweighted average of 23% and by a weighted average of 6% without significantly increasing imports from other GSP beneficiaries. This paper extends MacPhee and Rosenbaum's analysis in two ways. First, it estimates the effect that existing competitive need limits have on affected GSP imports and import shares. Second, it uses the most recent period for which data are available. MacPhee and Rosenbaum use data from 1976-1983. Substantial changes have occurred in the U.S. GSP since then, including a major revision in 1984. The list of beneficiary countries also has changed. A number of countries either have lost eligibility (Hong Kong, Korea, Singapore, and Taiwan) or have gained eligibility (Eastern European countries and the former Soviet republics), and one has received greater duty-free access (Mexico). These changes have altered the impact of competitive need limits, making a reappraisal worthwhile.
The completion of the Uruguay Round Agreement alters the effect of competitive need limits. This agreement cuts the average U.S. tariff rate by about a third, weakening the impact of the GSP and competitive need limits by lowering preference margins. Despite the lower preference margins, tariffs for many products imported primarily from developing countries will remain high enough to impede trade significantly. Examples include gold and silver jewelry, tropical hardwood, and telephone sets. These four products alone accounted for one third of the imports subject to competitive need limits in 1994.
II. COMPETITIVE NEED LIMITS AND U.S. TRADE LAW
The United States enacted the GSP in 1974 to stimulate development in under-developed countries. The GSP offers preferential tariff treatment to imports classified in some 4,400 eight-digit Harmonized Tariff Schedule (HTS) categories. In 1994, the United States gave preferential treatment to $18.4 billion in imports from over 150 beneficiary countries. While this figure is substantial, imports receiving preferential treatment represent only about 18% of all imports from GSP beneficiaries. Excluding imports that ordinarily enter the United States duty free reveals that imports entering the U.S. duty-free under the GSP represent only 27% of the imports that could receive preferential treatment.
Several factors limit the scope of the U.S. GSP. Legal statute partially or entirely excludes many goods produced by import-sensitive industries (such as textiles, apparel, watches, footwear, and handbags) from the U.S. GSP. These statutory exclusions, which cover many principle imports from GSP beneficiaries, prohibit GSP treatment under any circumstances. A crude estimate based on data from the U.S. International Trade Commission (USITC) indicates that excluded products account for about 50% of beneficiary imports subject to tariffs.
Beneficiary imports not excluded by statute are referred to as GSP-eligible. However, even these imports may be denied GSP treatment under certain circumstances: (i) Importers may fail to claim GSP benefits. (ii) Imported products may fail to meet a value-added requirement used to prevent transshipment through beneficiaries. (iii) Most important for the analysis here, imports of certain products from specific beneficiaries may be "too competitive," which subjects them to competitive need limits. According to the USITC, competitive need limits denied GSP benefits to imports with an average annual value of $10.6 billion between 1988 and 1994. Without these measures, beneficiary imports receiving GSP treatment would have increased by 68% during the period while the share of total beneficiary imports receiving GSP treatment would have increased from 14.3% to 24.2%.
Competitive need limits place a ceiling on either the value or import share of a product imported from a beneficiary country. Once this ceiling is reached, imports of the product from the beneficiary lose duty-free treatment under the GSP. In 1993, the ceiling for the value of imports was $108 million (this ceiling increases annually by the growth rate of U.S. GDP) while the ceiling for a beneficiary's import share was 50%.
In 1984, policymakers introduced additional limits known as reduced competitive need limits. These limits apply to imported products from particular beneficiaries that are sufficiently competitive and go into effect at lower thresholds. Reduced competitive need limits are applied when imports from a beneficiary that are deemed sufficiently competitive exceed a certain dollar value (roughly $40 million in 1993) or import share (currently 25%). In 1993, 33 products with imports valued at $460 million were subject to reduced competitive need limits.
Competitive need limits may be waived under certain circumstances. Since 1984, all such limits have been waived for the 30 or 40 countries designated as least developed beneficiaries. This is a token measure since imports from these beneficiaries are small and unlikely to reach any competitive need limit. Limits also may be waived in cases where a product is not produced domestically. In addition, the U.S. President may waive competitive need limits if a beneficiary's import share exceeds 50% but total imports of a product are below a de minimis limit ($12.3 million in 1993). Limits also can be waived if a petition for a waiver is filed with the U.S. Trade Representative (USTR) and granted by the President.
The U.S. GSP program expired on July 31, 1995, and has not been renewed because the Budget Enforcement Act of 1990 requires new programs to operate on a "pay as you go basis" to ensure that the budget deficit is unaffected. The USTR currently is seeking to have the GSP renewed for a one-and-one-half year period retroactive to July 31, 1995.
III. COMPETITIVE NEED LIMITS IN PRACTICE
Tables 1 and 2 present data on the U.S. GSP and competitive need limits for the period 1988 through 1994. Table 1 summarizes U.S. GSP coverage and competitive need limits for current GSP beneficiaries. The table shows that total imports from current beneficiaries nearly doubled between 1988 and 1994 while GSP-eligible imports nearly tripled. These trends raised the percentage of beneficiary imports eligible for GSP treatment from 19.9% to 28.1%. Eliminating GSP-eligible imports that did not receive duty free treatment reveals that duty free GSP imports from current beneficiaries increased by 192% during the period. The ratio of duty free GSP imports to total beneficiary imports increased from 11.7% in 1988 to 17.7% in 1994. Rising imports from beneficiaries such as Malaysia, Thailand, Indonesia, and the Philippines help explain these trends.
GSP-eligible imports affected by competitive need limits increased by 114% during the period, reflecting the overall increase in beneficiary imports. As a percent [TABULAR DATA FOR TABLE 1 OMITTED] of beneficiary imports receiving duty free treatment, however, eligible GSP imports subject to competitive need limits actually fell from 34.9% to 25.5%. Nonetheless, beneficiary imports receiving duty free treatment would have increased by $4.7 billion in 1994 without competitive need limits and related restrictions.
Table 1 provides data for current beneficiaries and excludes imports from four countries "graduated" from the U.S. GSP in 1989 (Hong Kong, Korea, Singapore, and Taiwan). It also excludes imports from Mexico, which lost its beneficiary status after the North American Free Trade Agreement went into effect in 1994. Including these countries in the analysis greatly increases the impact of competitive need limits because their U.S. exports were subject to a large number of those limits.
Table 2 presents data for the 10 GSP beneficiaries most seriously affected by competitive need limits in 1994. The table shows that the impact of competitive need limits is heavily concentrated. Imports from the 10 countries listed in table 2 account for 97% of all imports subject to competitive need limits in 1994. Imports from just three countries (Brazil, Malaysia, and Thailand) account for two-thirds of all imports subject to competitive need limits.
Table 2 also includes information on the average preference margins lost as a result of competitive need limits. Overall, the average margin is 6.2%, slightly higher than the average U.S. tariff rate. Combining these margins with the other import data in table 2 identifies Thailand, Brazil, and Indonesia as the countries most seriously affected by competitive need limits.
IV. EMPIRICAL ANALYSIS OF COMPETITIVE NEED LIMITS
Data for the empirical analysis of the competitive need limits in effect in 1994-1995 are from a variety of sources. An annual [TABULAR DATA FOR TABLE 2 OMITTED] USTR press release reports GSP imports subject to competitive need limits. For each competitive need limit, the press release reports the value and import share of affected imports at the eight-digit HTS level. Because the United States switched to the HTS tariff classification in 1988, data defined according to the HTS are not available for earlier periods. U.S. tariff rates at the eight-digit HTS level for imports subject to competitive need limits are from USTR (1991). Elasticity estimates, total imports, and domestic shipments defined at the eight-digit HTS level are from the USITC's periodic GSP commodity digests.
Gauging the impact of existing competitive need limits involves two methods. The first method compares import values and shares before and after competitive need limits are imposed. This ex post approach assumes that any changes in import values or shares is attributable to competitive need limits. The second method uses trade and elasticity data to compute the ex ante effects of competitive need limits. Using this information, the analysis computes measures of trade creation and trade diversion, which provide an estimate of the competitive need limits' impact.
Consider the ex post approach. It provides limited information because only about 40% of current competitive need limits were imposed during the period for which data are available (1988-1993). The other competitive need limits were imposed prior to 1988. Thus, calculating the ex post effect is not possible. Of the cases for which data are available, an even smaller percentage involve competitive need limits in effect for at least two years, the period used to assess the impact of these limits. All told, the sample includes 45 of 175 cases. The 45 cases cover imports [TABULAR DATA FOR TABLE 3 OMITTED] valued at about $0.9 billion in 1993, or about a quarter of the value of all imports subject to competitive need limits in 1994-1995.
Table 3 reports the results from the ex post analysis. Consider first how competitive need limits affect import shares. The average import share declines from 52.7% in the year prior to imposition of a competitive need limit to 40.6% two years after imposition of a limit. This is a significant drop, but it is misleading because import levels in smaller cases are affected more severely than are import levels in larger cases. This is apparent in the smaller decline in weighted average import shares, which are computed using import values as weights. Results for the median change after two years reinforce the conclusion that import shares in smaller cases are particularly hard hit. The median change after two years indicates that competitive need limits reduce import shares by more than 50%.
The ex post results for the value of imports confirm the above analysis. On average, imports increase slightly from $18.4 million in the year prior to the imposition of a competitive need limit to $18.5 million two years after the imposition of a limit. While the average value of imports increases despite the imposition of competitive need limits, the median change in the value of imports shows a decline of about 44%. These results leave little doubt that competitive need limits have a greater impact at lower import levels. This finding is consistent with MacPhee and Rosenbaum's results.
Pomfret (1986) provides an alternative method for estimating the effects of competitive need limits. His partial equilibrium approach yields an ex ante estimate of the increase in imports produced by eliminating a discriminatory tariff. International economists traditionally divide this increase into two components: trade creation and trade diversion. Trade creation measures the increase in a country's total imports resulting from a discriminatory tariff cut while trade diversion measures any shift in imports between sources. Import shifts occur because discriminatory tariff cuts increase the competitiveness of some imports at the expense of others.
In Pomfret's model, import demand takes the form:
[Mathematical Expression Omitted]
[TABULAR DATA FOR TABLE 4 OMITTED]
where M is total imports; [M.sub.c] and [M.sub.n] represent imports that are and that are not subject to competitive need limits, respectively; and [P.sub.c] and [P.sub.n] represent the prices of imports that are and that are not subject to competitive need limits, respectively. The parameter [Alpha] is the elasticity of import demand while q is the ratio of competitive need limit imports to total imports (i.e., [M.sub.c]/M). The parameter [Alpha] denotes the elasticity of substitution between imports that are subject to competitive need limits and imports that are not and is defined by the following equation:
([M.sub.c]/[M.sub.n]) = B[([P.sub.c]/[P.sub.n]).sup.[Sigma]].
Assuming the elasticity of export supply ([Epsilon]) is the same no matter where exports are produced, the value of trade created by removing a competitive need limit (TC) is given by
TC = ([Alpha]([Epsilon] + 1)/([Epsilon] - [Alpha]))
[M.sub.c] (dt/(1 + t))
where t is the initial tariff rate and dt is the reduction in the tariff rate resulting from eliminating a competitive need limit. The value of trade diverted from other exporting countries (TD) is given by
TD = ([Epsilon]([Epsilon] + 1)(1 - q)([Sigma] - [Alpha])/(([Epsilon] - [Sigma])([Epsilon] - [Alpha]))) [M.sub.c](dt/(1 + t)).
The total effect on imports subject to competitive need limits is equal to the sum of trade creation and trade diversion.
Computing trade creation using Pomfret's approach requires using estimates of the elasticities of import demand and export supply. The GSP digests report import demand and export supply elasticities as being "low," "moderate," or "high." Operationalizing these estimates involves assigning values of 0.5, 1.5, and 5 to the low, moderate, and high USITC categories, respectively. Consultation with a USITC trade analyst suggested these values. Elasticity estimates are available for about 60% of the cases. For the remaining cases, sensitivity analysis provides a rough measure of the impact of competitive need limits. The sensitivity analysis involves three different elasticity combinations. In the base case, the import demand and export supply elasticities are set to 1.5; in the most restrictive case, both elasticities are set to 5; in the least restrictive case, both elasticities are set to 0.5.
Computing trade diversion using Pomfret's model also requires using an estimate of the elasticity of substitution. Such estimates do not exist at the level of disaggregation employed here, but Leamer and Stern (1970) show that one can express the elasticity of substitution as
[Sigma] = [Alpha] - [Beta]
where [Beta] is the cross price elasticity of demand between imports affected and unaffected by competitive need limits. While estimates of the import demand elasticity ([Alpha]) are available in the majority of cases, no estimates exist for the cross price elasticity of demand ([Beta]), so sensitivity analysis becomes useful. The analysis sets cross price elasticity to 1.5, 5, and 0.5 respectively in the base, most restrictive, and least restrictive cases.
Table 4 reports the results from the ex ante analysis. In the base case, competitive need limits reduce affected imports by $382.0 million, or by about $2.6 million per case. The median reduction is much smaller and stands at only $0.3 million. The average percentage reduction in imports per case is 11.7% with the median reduction being 8.4%. These differences suggest that the reduction in imports attributable to competitive need limits is highly concentrated. A case by case analysis confirms this; just 10 cases account for about 66% of the reduction in total imports.
In the most restrictive case, imports fall by $550.9 million as a result of competitive need limits, or by an average of $3.8 million. The median decline again is much smaller at $0.5 million. Imports decrease on average by 16.8%, with the median reduction being 15.6%. In the least restrictive case, imports decline by $329.2 million or 7.2%.
In all three cases, the ratio of trade creation to trade diversion exceeds 4 and ranges as high as 15. This finding shows that the principal beneficiaries of competitive need limits are not other GSP beneficiaries or other exporting nations but rather U.S. import competing industries.
The analysis here estimates that imports subject to competitive need limits decline by $330 million to $550 million annually as a result of these limits. This decline represents less than four percent of duty-free GSP imports in 1994 and less than one percent of all beneficiary imports. Although the aggregate effect of competitive need limits is small, this effect is highly concentrated, producing large annual declines in a small number of cases.
Estimates suggest that competitive need limits currently have a small effect on the total flow of imports from beneficiary countries. Total beneficiary imports decline by less than 1%, and GSP-eligible imports decline by four percent at most. While the aggregate effect of competitive need limits is small, the effect in particular cases is by no means insignificant. The estimated average reduction in affected imports is between 10 and 17% with total affected imports falling by as much as $550 million annually. In a significant number of cases, competitive need limits reduce import shares by more than half despite the fact that preference margins are relatively small. In the 10 leading cases, imports decline by an average of $25 million annually.
The results also suggest that U.S. import competing firms gain the most from competitive need limits. Estimates of trade diversion are small in comparison to estimates of trade creation, indicating that competitive need limits do little to reallocate trade to less developed beneficiaries or other exporters. Because competitive need limits apply only in cases where developing countries have exhibited competitive strength, the benefits they create for U.S. producers directly undermine the comparative advantage of developing countries. In this sense, competitive need limits clearly violate the original intent of the GSP.
GSP: Generalized System of Preferences
HTS: Harmonized Tariff Schedule
USITC: United States International Trade Commission
USTR: United States Trade Representative
Baldwin, R., and T. Murray, "MFN Tariff Reductions and LDC Benefits under the GSP," The Economic Journal, 87:345, 1977, 30-46.
Karsenty, G., and S. Laird, "The GSP, Policy Options, and the New Round," Weltwirtschaftliches Archiv, 123:3, 1987, 262-295.
Leamer, E., and R. Stern, Quantitative International Economics, Allyn and Bacon, Boston, 1970.
MacPhee, C., and V. Oguledo, "The Trade Effects of the U.S. Generalized System of Preferences," Atlantic Economic Journal, 19:4, 1991, 19-26.
MacPhee, C., and D. Rosenbaum, "The Asymmetric Effects of Reversible Tariff Changes under the United States GSP," Southern Economic Journal, 56:1, 1989, 105-125.
Murray, T., Trade Preferences for Developing Countries, John Wiley and Sons, New York, 1977.
Pomfret, R., "The Effects of Trade Preferences for Developing Countries," Southern Economic Journal, 53:1, 1986, 18-26.
United States International Trade Commission (USITC), Operation of the Trade Agreements Program 40th-45th Reports, Washington, D.C.
United States Trade Representative (USTR), A Guide to the U.S. Generalized System of Preferences, Washington, D.C., 1991.
Assistant Professor, Department of Economics and Business, Lafayette College, Easton, Pa., 18042. This is a revised version of a parer presented at the Western Economic Association International 70th Annual Conference, San Diego, Calif., July 7, 1995.
|Printer friendly Cite/link Email Feedback|
|Publication:||Contemporary Economic Policy|
|Date:||Oct 1, 1996|
|Previous Article:||Unification of China's foreign exchange rates.|
|Next Article:||Consequences of minimum wage indexing.|