Policy, Planning, and Research WORKING PAPERS Trade Policy Country Economics Department The World Bank June 1989 WPS 216 Price and Quality Effects of VERs - Revisited A Case Study of Korean Footwear Exports Jaime de Melo and L. Alan Winters Voluntary export restraints on exports of Korean footwear re- sulted in significantly higher prices to the restricted markets and a mild downgrading - rather than the expected upgrading - in quality. The Policy, Plarnning, and Research Complex distributes PPR Working Papers to disseminate the findings of work in progress and to encourage the exchange of ideas among Bank staff and all others interested in development issues. These parers carry the names of the authors, reflect only their views, and should be used and cited accordingly. The findings, tmerpretauions, and conclusions are the authors' own. They should not be attributed to the World Bank, tts Board of Dluctors, Its management, or any of its member countries. Plc,Planning, and Research Trade Policy The application of index numbers to disaggre- VERs were associated less with quality gated Korean footwear exports during 1974-85 upgrading and if anything - for exporters suggests that binding voluntary export restraints whose objective was foreign exchange eamings (VERs) led to significant price increases but not rather than profit maximization - were possibly to the upgrading of quality predicted in earlier associated with quality (.owngrading. theoretical analyses. The pattem of quality changes across Unlike most previous empirical studies of markets seems not to have varied significantly VERs, the results de Melo and Winters present during the VER period, so adjustment costs may in this paper were derived from exporters' rather have been enough to prevent a change in product than imponers' data. Thus they provide a more mix toward restricted markets. direct test of the reactions of exporting countries subject to VERs. The empirical results, the authors conclude, confirm that as a result of VER-type restrictions Drawing their comparisons from multilateral on shoes, rents accrue to the exporting country indices (adjusted for changes in product and (and some rents to the importers, if imponing is country mix), the authors found that the price of not a competitive activity) - but shoes do not Korean footwear exports to the United States show the improved quality that automobiles do (and to a lesser extent to the United Kingdom) under similar restrictions. rose significantly in the years when VERs were binding (when quotas were filled). This paperis a product of the Trade Policy Division, Country Economics Department. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Maria Ameal, room N8-073, extension 61466 (36 pages with tables). The PPR Working Paper Series disseminates the findings of work under way in the Bank's Policy. Plannitng, and Rcsearch Complex. An objective of the series is to get these fimdings out quickly, even if presentations are less than fully polished. The findings, interpretations, and conclusions in thcse papers do not necessarily represent official policy of the Bank. Produced at the PPR Dissemination Center Price and Quality Effects of VERs Revisited: A Case Study of Korean Footwear Exports by Jaime de Melo and L. Alan Winters Table of Contents 1. Introduction 1 2. VERs on Korean Footwear Exports 4 3. Index Numbers for the Measurement of Export 14 Prices and Quality 4. Multilateral Comparisons 9 5. Bilateral Comparisons 15 6. Conclusions 23 Footnotes 25 Bibliography 27 Appendix 1: The Korean Footwear Export Data 28 L. Alan Winters is of the University College of North Wales. This paper is part of the research project RPO 672-40. We thank Taeho Bark for helping us collect the data. We thank Paul Brenton and Julie Stanton for many helpful suggestions and extremely valuable assistance. Helpful comments from Bela Balassa and particularly from Bee Aw and Mark Roberts on an earlier draft are also acknowledged. 1. Introduction It is expected that when a binding voluntary export restraint (VER) negotiated between two parties is for broadly defined product categories, the price of the restricted products will rise in the restricted market. It is also expected that producers in the exporting country that entered the VER will shift production from less to more expensive product categories (Falvey (1979), Rodriguez (1979)). Thus a quality upgrading effect is predicted by a VER. Less upgrading is expected when restriction is by an import quota because import quotas are usually extended on a multilateral basis and commodity categories subject to quotas are often less loosely defined than categories subject to VERs. 1/ Much empirical support has been produced showing that exporters facing binding VERs have changed their product mix shipped to the restricted market towards higher priced varieties. Fur highly differen- tiated products such as autos, hedonic price regressions have indicated that a statistically significant increase in the price of autos sold in the restricted market was attributable, at least partly, to an improvement in quality where quality is measured in characteristics space (Feenstra (1986), (1988), de Melo and Messerlin (1988)1. For more homogeneous products like footwear, index number comparisons that control for changes in product mix among the constrained products have also been claimed to indicate quality upgrading after the VER came into effect (Aw and Roberts, 1986). 2 To our knowledge, all evidence on quality upgrading has relied on the analysis of data referring to imports in the constrained market. This has allowed for comparisons between constrained and unconstrained suppliers to the restricted market; but this, however, is not the issue addressed by the theory. Data on imports to the restricted market do not allow for comparisons between a particular exporter's sales to different markets, which is the phenomenon on which theory offers a prediction, namely that exports to a constrained market will be upgraded (presumably more rapidly than exports to unconstrained markets). This paper measures quality and price effects of VERs using exporting country data. Oulr application is to Korean exports of footwear to various industrial countries over the period 1974-86, a period which offers both instances and removals of VERs. We approach the measurement of price and quality effects through the use of Tornqvist numbers previously applied by Aw and Roberts (1986, 1988) to US imports of footwear. Our results indicate the existence of binding VERs -- as reflected in price increases in the restricted markets when the VERs were in effect -- but quality downgrading. We also provide an interpretation for the observed downgrading. The paper is organized as follows. In the remainder of the introduction, we summarize briefly why a VER may (or may not) be expected to provoke upgrading. In Section 2, we recall tne advantages of index numbers over unit values to measure price and quality shifts We also recall the assumptiQns under which index numbers are likely to be good indicators of price and quality shifts and describe the formulas upon which rests the evidence in the paper. Section 3 describes briefly the VERs into which Korea entered for its exports of footwear. Our evaluation of price 3 and quality shifts starts with multilateral comparisons in Section 4. The evidence is for 14 categories of footwear exported to 7 destinations. Because more disaggregation is preferable, in Section 5, we extend our comparisons to 26 footwear categories. However, at this level of disaggre- gation, we can only make bilateral comparisons because the multilateral data contain too many missing observations. The evidence in Sections 4 and 5 should therefore be viewed as complementary. Conclusions are in Section 6. The theoretical prediction that a VER will lead to quality up- grading rests on the assumption that "quota tickets' are freely traded among quota recipients. 2/ Implicit in these theoretical studies is not only that perfect competition prevails among firms, but also that firms can costlessly alter their product mix (Falvey, 1979; Rodriguez, 1979) or that firms are specialized in one type of product (Chang, 1984). Under these assumptions, firms will alter their product mix so as to achieve equal unit rent of each variant supplied. Since per unit rent will initially be higher for the higher priced (quality) product, the VER will induce quality upgrading. For relatively undifferentiated products with low start up costs like footwear and textiles, it is safe to assume that production will continue to take place under competitive conditions so that market structure is exogenous with respect to the imposition o2 VERs. However, even if there are no adjustment costs and 'quota tickets' are transferable among exporters, the perception of foreign exchange scarcity (perhaps in part because of VERs) may lead to an attempt at minimizing the loss of export earnings loss rather than maximizing profits. 3/ Bark and de Melo (1987) have studied the case of export minimization loss with and without quota ticket transferability. In that 4 case, foreign exchange losses will be minimized when marginal revenues (MRs) are equated implying that the export composition will shift towards the high quality (low quality) product if MRH is greater (smaller) than MRL. Representing marginal revenues in terms of import demand elasticities and initial conditions, Bark and de Melo show that, for most elasticity ranges, the conditions for upgrading under competitive quota ticket trading are the conditions for downgrading under a quota ticket allocation (with no transferability) to minimize foreign exchange earnings loss, and vice- versa. One must therefore conclude that the direction of product mix change in response to a VER cannot be determined unambiguously. 2. VERs on Korean Footwear Exports 4/ Korean exports of footwear provide an interesting case study of the price and quality implications of VERs for at least two reasons. First, unlike many other commodities subject to VERs, VERs on footwear intensified during a relatively short period of time, between 1977 and 1982. After mid-1981, the intensity of protection on footwear trade subsided either because non-tariff measures were not renewed (US) or because the restrictions ceased to be binding (UK). Hence, footwear offers a unique experimental design with restricted observations surrounded by unrestricted ones. Second, Korea is among the largest footwear producers in the world. Third, in 1985, footwear exports were the fifth largest Korean export earners with 87 percent of domestic production exported. Table 1 sumnarizes the main restrictions on Korean footwear exports for the major importers of Korean footwear and the relative importance of each country in total Korean footwear exports. By far the most important purchaser of Korean footwear exports is the US which nego- Table 1: MAJOR RESTRICTIONS ON KOREAN FOOTWEAR EXPORTS Share of Korean Footwear Country Export. (1978) Footwear Type Exporters NTM !) Yearu USA 62.8 Non rubber Korea, Talwan VER July-77-June S1 UK 4.9 All footwear Kore, Taiwan VER Jan 1979- Poland Canada 4.4 All footwear All MFN Quota 1977-85 Japan 16.1 Leather All Value quota 1962- Norway 0.7 Rubber boots Korea, T:iwan YER 1979 Sweden 1.9 Rubber boots Korea, Taiwan VER and quotA 1970-t4 Leather and All Non-MFN quota 197C-77 plastic Australia 1.0 All footwear FNl Value Import 1974 quota EEC / 8.7 All footwear All Posteriori 1978- import surveillance NTM : non-tariff "asure. Export share excludes UK. Source: adapted from Hamilton (1989). 6 tiated a VER (also known as an Orderly Marketing Agreement -- OMA) with Korea for four years starting in July 1977. The second most important purchaser is Japan which has maintained a value quota on an MFN basis for imports of leather footwear, at least since 1952. There is no information indicating a change in Japanese policy during our sample period (1974-86). Furthermore, for countries with a value quota like Japan and Australia, one would not expect an inducement to change product mix. The UK footwear industry negotiated bilate-.al VERs with the major suppliers of British footwear imports since 1979, but the evidence shows that quotas have not been filled since 1982. 5, Like the UK, Canada followed in the footsteps of the US, but Canadian quotas were extended to all countries on an MFN basis. It is therefore unclear whether these restrictions should be expected to have an effect on Korean exports, especially in view of the small size of the Canadian market for Korean exporters. Finally, Norway and Sweden extended restrictions to Korea (and Taiwan) mainly for rubber boots. In view of the pattern of Korean footwear exports, and the relative importance of trading partners which established NTMs, we aggregate Norway and Sweden with the rest-of-the world (ROW) and disaggregate Korean footwear exports with respect to seven partners: US, Japan, Unrestrained 61 EEC, UK, Canada, Australia, ROW. In this EcountryN classification, we expect VERs to affect differentially the US, the UK and perhaps Canada when the NTMs were in effect. 3. ,ndex Numbers for the Measurement of Export Prices and Quality Though frequently used for price comparisons, unit value indices are biased because they fail to correct for changes or differences in the 7 mix of sub-products within any product heading of the trade bundle. In particular, a change in the mix of differently priced sub-products will result in a change in the product-level unit value index even if there is no change in any price. In our study of Korean footwear exports, one might expect to observe price increases when the VERs on Korean footwear exports became binding. If, however, there were also a simultaneous change toward the shipment of more (less) expensive shoes to the restricted market, the unit value index would overstate (understate) the true price change. It is generally agreed that unit values which are constructed at the industry level, and which consequently treat all commodities as perfect substitutes, lose reliability with aggregation. These drawbacks or unit-value indices used in most previous studies of VERs [e.g. Yoffie (1983)] can be remedied with the use of index numbers. Relying on the theory of index numbers developed by Caves, Christensen and Diewert (1982), and applied by Aw and Roberts (1988) to the study of US footwear imports, we construct multilateral price indices to compare price and product indices for Korean footwear exports to different countries. Let k be an index over footwear categories, c an index over countries, and i and j refer to footwear bundles, where i and j may refer to a country c, or a time period t. Assume that the price per unit of service of bundle i is equal to the average price per unit of quantity (unit value), Pi, divided by the level of quality, Ai. As shown by Caves, Christensen and Diewert (1982) the product-mix-adjusted Tornqvi&2 multilateral price index defined bys (1) ln P 1 E(S + §k)(ln Pi - ln Pk) J 2 (S + k)(ln PJ In k 8 allows for transitive cross-sectional and time-series comparisons. In (1), Skand ln Pk are the hypothetical cost and price shares (derived by taking a logarithmic mean over all commodities in the sample) against which the bilateral comparisons are made. Furthermore, with the above definition of quality, one can show that the change in unit value between any two ob- servations is the sum of quality change and the Tornqvist price index, i.e. (2) ln Pi - ln PJ - ln Ai - ln AJ + ln P; As recognized by Aw and Roberts (1988), multilateral comparisonb can be severely affected by missing observations since the construaction of P* requires observat %ns on the price of exports for each footwear cate- gory, to each country and for each year. If the mean share for a footwear category is large, an imputed price for a missing commodity-country flow may receive substantial weight in the overall index. Hultilateral comparisons are therefore likely to require a relatively higher level of aggregati.n to minimize the number of missing observations. For the detection of quality change, aggregation is a drawback since it limits the scope for detecting quality upgrading. (The scope for upgrading is reflected in the ratio of the highest to the lowest commodity unit value and the shares of trade in these commodities). On the other hand, time-series comparisons derived from bilateral indices are less affected by missing observations for two reasons: first, prices need not be inputed if a commodity is not expL- :ed in adjoining years; second, commodity-country combinations for which certain observations are zero usually have a small share in total export value and hence receive little 9 weight in the bilateral Tornqvist price index. This index, ln Pij, is de- fined bys (3) in Pij 1 YSk +k k)(ln Pk lnP) where i and j refer to adjoining time periods and Sk, is, as before, the share of total export revenue for observation i from commodity k. 7/ In Section 4 we rely on multilateral comparisons for Korean footwear exports aggregated over 14 footwear categories, 8/ and in Section 5 we carry out bilateral comparisons for a subset of countries with footwear exports disaggregated over 26 categories of footwear. 4. Multilateral Comparisons We start with a comparison of unit value, unit price a:nd unit quality indices for the seven countries purchasing Korean footwear. Due to the potential biases caused by missing values, we have been obliged to aggregate Korean footwear exports into 14 commodity categories. Details of aggregation and data preparation procedures are given in the appendix. Our fabricated commodity categories (see appendix) are distributed as follows (codes, followed by number of footwear categories in parenthesis): Rubber (6401,2); leather, including artificial (6402,10); wood and cork, gaiters, etc. (6403,1); parts (6405,1). 91 Because the US VER with Korea extended over a 4 year period starting at mid-year, we annualized the quarterly data from mid-year to mid-year points. Thus our sample extends from 1974:3 to 1986s2 and an annual observation for, say, 1978 extends from 1977s3 to 1978s2. 10 Unit values appear in table 2, Tornqvist prices in table 3 and the quality index in table 4. Because all data are reported from July to June (e.g. data for 1975 is from July 1974 to June 1975), the restricted period corresponds to 1978, 1979, 1980 and 1981 and unrestricted data to all other years. Even at this level of disaggregation, unit values are relatively poor indicators of prices. For example, the ratio of the highest to lowest unit value index in 1986 is 1.96, whereas the corresponding Tornqvist price ratio is only 1.23, indicating that there were substantial differences in the product mix of the 14 commodity categories between countries. Before looking for price and quality effects during the VERs, it is interesting to use the properties of the multilateral price and quality indices to check informally whether high priced (high quality) Korean shoes were destined to high income markets. For quality-adjusted price indices, the pattern is clear and as expected: the highest price bundle is destined to the highest-income country at the beginning of the period with a narrowing in the price gap through time as one would expect from a pattern of narrowing income differentials. The quality index shows little systematic variation for any country except the US which shows strong upgrading effects. However, it occurs after th& OMA, period rather than within it. How misleading unit value indices are as indicators of price changes is seen by comparing the difference in average growth of the unit value index for the US during the restricted and unrestricted years (2 percentage points), with the growth of the quality-adjusted price index for the same years in table 3. The unit value indices show small differences in average growth during the restricted and unrestricted periods whereas the quality-adjusted price index indicates a doubling of the growth rate TABLE 2 UNIT VALUE INDICES (REL. TO US=1.O IN 1977) YEAR JAPAN EEC UNITED CANADA USA AUSTRALIA REST ENDING KINGDOM OF JUNE WORLD 1975 0.52 0.48 0.47 0.48 0.66 0.41 0.55 1976 0.49 0.54 0.45 0.55 0.84 0.43 0.58 1977 0.57 0.67 0.54 0.72 1.00 0.51 0.76 1978 0.61 0.81 0.58 0.75 1.11 0.51 0.82 1979 0.71 0.96 0.86 0.96 1.31 0.73 0.82 1980 0.89 1.22 1.09 1.15 1.59 0.75 1.27 1981 0.96 1.21 1.15 1.22 1.60 1.13 1.48 1982 1.01 0.94 1.05 1.23 1.72 1.12 1.40 1983 1.09 0.97 0.89 1.28 1.66 1.00 1.17 1984 1.11 1.05 0.99 1.13 1.70 0.95 0.84 1985 1.20 1.06 1.03 1.31 1.89 1.08 .90 1986 0.99 1.19 1.11 1.52 2.06 1.20 1.06 AVERAGE GROWTH RATES FOR RESTRICTED PERIOD (1978 TO 1981) JAPAN EEC UK CANADA USA AUS REST 0.144 0.163 0.218 0.144 0.128 0.239 0.198 AVERAGE GROWTH RATES FOR 1975-76 AND 1982-86 JAPAN EEC UK CANADA USA AUS REST 0.023 0.057 0.022 0.102 0.105 0.046 0.015 AVERAGE GROWTH RATES FOR 1975-86 JAPAN EEC UK CANADA USA AUS REST 0.067 0.095 0.093 0.117 0.113 0.116 0.082 TABLE 3 TORNQVIST PRICE INDICES (REL. TO US=1.0 IN 1977) YEAR JAPAN EEC UNITED CANADA USA AUSTRALIA REST ENDING KINGDOM OF JUNE WORLD 1975 0.63 0.57 0.57 0.55 0.70 0.56 0.60 1976 0.58 0.66 0.55 0.65 0.88 0.61 0.69 1977 0.63 0.77 0.59 0.79 1.00 0.75 0.78 1978 0.62 0.83 0.62 0.78 1.05 0.75 0.86 1979 0.76 1.04 0.93 1.05 1.29 1.00 1.00 1980 0.99 1.29 1.15 1.21 1.55 1.12 1.28 1981 1.02 1.36 1.23 1.28 1.59 1.32 1.45 1982 1.07 1.13 1.08 1.28 1.56 1.19 1.44 1983 1.15 1.03 0.91 1.17 1.37 1.16 1.22 1984 1.09 1.12 1.02 1.22 1.40 1.05 1.16 1985 1.20 1.16 1.12 1.34 1.46 1.25 1.22 1986 1.15 1.29 1.21 1.48 1.52 1.39 1.30 AVERAGE GROWTH RATES FOR RESTRICTED PERIOD (1978 TO 1981) JAPAN EEC UK CANADA USA AUS REST 0.136 0.154 0.216 0.135 0.127 0.157 0.168 AVERAGE GROWTH RATES FOR 1975-76 AND 1982-86 JAPAN EEC UK CANADA USA AUS REST 0.020 0.044 0.008 0.078 0.051 0.057 0.030 AVERAGE GROWTH RATES FOR 1975-86 JAPAN EEC UK CANADA USA AUS REST 0.062 0.084 0.084 0.099 0.079 0.094 0.080 13 during the restricted period when compared with the unrestricted period. From equation (2), this small difference in unit value growth during the restricted years coupled with the large price increase implies much less upgrading during the restricted than during the unrestricted years (table 4). However, a look at the average growth rates of the Tornqvist price and quality indices across countries suggests that the expected price increase in the restricted markets (US and UK) was also significant elsewhere. Likewise, the slowdown in quality growth during the restricted period was not confined to the US. To check further whether indeed price increases and quality decreases were significant during the restricted years, we regressed the price and quality indices in tables 3 and 4 on time and country dummies with separate dummy variables for the restricted periods for the US (1978-81) and the UK (1979-81). The results were (country and year dummies omitted): (4) p*- 0.089 DUS + 0.080 DUK R2- 0.99 (1.93) (1.57) DOF - 64 (5) Q - -0.12 DUS + 0.01 DUK R2 - 0.99 (-2.43) (0.27) DOF - 64 The results from the price equation indicate a significant price rise for the US but not for the UK although a specification of (4) and (5) in natural logarithms (not reported here) gives a significant acceleration of price increase for the UK but not for the US. The results from the quality index regression also suggest a significant slowdown in quality growth for the US during the VER-period but no change for the UK. 10/ Thus the fixed- effect regressions further support the direct inferences drawn from tables TABLE 4 QUALITY INDICES (REL. TO US=1.0 IN 1977) YEAR JAPAN EEC UNITED CANADA USA AUSTRALIA REST ENDING KINGDOM OF JUNE WORLD 1975 0.83 0.84 0.83 0.86 0.94 0.73 0.93 1976 0.85 0.83 0.81 0.84 0.96 0.70 0.85 1977 0.90 0.87 0.92 0.91 1.00 0.69 0.97 1978 0.97 0.98 0.95 0.96 1.06 0.68 0.95 1979 0.93 0.92 0.93 0.91 1.02 0.73 0.82 1980 0.90 0.94 0.95 0.95 1.02 0.67 0.99 1981 0.94 0.89 0.93 0.95 1.00 0.86 1.02 1982 0.95 0.83 0.97 0.96 1.10 0.94 0.97 1983 0.94 0.94 0.97 1.09 1.21 0.86 0.96 1984 1.01 0.93 0.97 0.92 1.22 0.90 0.72 1985 1.00 0.92 0.92 0.98 1.29 0.86 0.73 1986 0.86 0.93 0.92 1.03 1.36 0.86 0.81 AVERAGE GROWTH RATES FOR RESTRICTED PERIOD (1978 TO 1981) JAPAN EEC UK CANADA USA AUS REST 0.012 0.009 0.003 0.012 0.001 0.066 0.020 AVERAGE GROWTH RATES FOR 1975-76 AND 1982-86 JAPAN EEC UK CANADA USA AUS REST 0.002 0.012 0.014 0.022 0.053 -.008 -.019 AVERAGE GROWTH RATES FOR 1975-86 JAPAN EEC UK CANADA USA AUS REST 0.006 0.011 0.010 0.018 0.034 0.019 -.004 15 3 and 4, namely a significant price rise in the US when the VER was binding accompanied by a quality decline. It should be noted, however, that rubber products were not subject to a quota. Therefore, the quality index may be picking up substitution from covered to non-covered products (e.g. covered leather to noncovered rubber). To the extent that the VER was on higher- valued footwear, substitution towards non-covered items would show up in quality downgrading. In the next section, we look for further confirmation of this result by relying on more disaggregated bilateral comparisons. 5. Bilateral Comparisons As pointed out in section 2, the detection of quality change is likely to be better the more disaggregated the commodity set since commodities are assumed to be homogeneous within each category. Our multilateral comparisons in section 4 were unavoidably restricted to fourteen commodity categories. As a corroboration of our earlier results, therefore, it is useful to conduct similar calculations ca a data set with greater commodity disaggregation. A necessary coro'lary of further commodity disaggregation, however, is reduced country coverage and bilateral rathar than multilateral comparisons. Our bilateral comparisons are for the USA, Japan and West Germany (FRG): the three countries with fewest missing observations. Furthermore, West Germany has the added attraction that, with the exception of EEC surveillance, which is generally held to have had little effect, she had no restrictions on footwear imports over our sample period. We now analyze price and quality effects of VERs with bilateral indices. These indices are attractive because their weights are more relevant to the country under consideration, but at the expense of reducing 16 the degree of cross-country comparability. Our results with bilateral indices are directly comparable with those calculated by Aw and Roberts (1986) who calculated bilateral indices for Korean-US footwear trade using US import data. The results of the bilateral calculations are summarized in table 5. Two commodity sets are reported: all footwear; 11/ and US-restricted commodities. The former classification contains 26 commodities, while the latter contains a subset of 21 of them. Average growth rates of price, quality and unit value indices are reported for the years of the US restrictions and the remainder of the sample period. Several comparisons are possible for the data reported in table 5. Compa-4ng the OHA and non-OMA periods for the US reveals a very considerable rise in the rate of price increases in the former period, but a decline in the rate of quality change. This is consistent with our previous finding. Further evidence of the effect of the OMA on the restricted footwear types may be gleaned by comparing the indices for the restricted and all-footwear commodity classifications. The difference between OMA and non-OMA price rises is larger for the restricted commodities, but the difference in quality growth is the same. The lower quality changes in the all-footwear classification suggest that while substitution has been occurring from cheaper to dearer footwear and more strongly in the restricted than in the unrestricted set, the process was not affected during the OMA period. The interpretation of the US data becomes more complex once account is taken of the results for the FRG and Japan. They display stronger price rise differentials than the US and the FRG also reports some quality upgrading. Thus comparing across countries seems to suggest Table 6: PRICE AND QUAUITY CHANGES: BILATERAL INDICES Tornqvist Price Index Quality Index Unit Valuo Index Period!/ I US JAP FRO US JAP FRO US JAP FRO I . Restricted k/ R I 0.140 0.196 0.156 0.023 0.014 0.070 0.160 0.210 0.227 Comnoditios I (0.134) (0.141) (0.168) (0.053) (0.031) (0.049) (0.137) (0.173) (0.212) (21 Categories) I U I 0.010 0.048 0.02S 0.066 -0.024 0.007 0.078 0.023 0.0o2 I (0.007) (0.036) (0.022) (0.058) (0.008) (0.017) (0.065) (0.044) (0.039) w All Commoditioes R 0.121 0.147 0.122 0.014 0.001 0.088 0.134 0.149 0.156 (26 categories) U I 0.034 0.029 0.04U 0.057 0.012 -0.027 0.092 0.040 0.019 Figures In parenthese are for 20 commodity categories (exclusion of outdoor shooe of artificial leather). / U a unrestricted period 1975/7 TO 1977/6 and 1981/7 to 1986,6 R a restricted period 1977/1 to 1981/6 y categorlos restricted by the US VER (se appendix for definitions) / all categories excluding parts, wood and cork, gaiters (se appendix for deflnitions) 18 negative price effects of the OMA. In fact, such a conclusion is probably too extreme, for the figures in the top part of table 5 are strongly influenced by one particular commodity: outdoor sports shoes of artificial leather. Over the sample period the prices of Korean exports of these shoes to the USA, FRG and Japar. grew fairly rapidly and by broadly similar amounts, and in each case displayed one year of very substantial increase (over 40X). Unfortunately, however, the years concerned differed -- 1976, 1978 and 1980 respectively, so that whereas for the US the spurt fell in the non-OMA period, for the others it fell in the OMA period. Moreover, these footwear are generally held to have benefited from a large shift in fashion in the late 19709 (jogging, etc.), so that the weight attached to the large price increase was 11 for the US but 91 and 511 for the FRG and Japan respectively. When the comparisons are made for 20 commodity categories excluding outdoors sports shoes of artificial leather (figures in parentheses in table 5), the differences between countries are considerably narrowed. The increase in the rate of US price increases in the OMA period exceeds that of Japan and equals that of Germany, while the decline in the rate of increase in quality in US falls both absolutely and relatively to the other countries. Table 6 shows the considerable coherence between our results and those of Aw and Roberts (1986). They were able to distinguish seventy commodity categories in the US import data and so considered the two restricted sub-groups separately: non-athletic leather footwear (401 of relevant imports in 1976/77) and other footwear (60Z). On price increases, the two studies both find average increases of around 121 to 13Z per annum during the VER and of around 41 over the non-VER years. Our later data 19 Table 6: COMPARISON WITH AW AND ROBERTS (1986) de Melo Aw and Roberts a/ and Winters (1989) Rate of Increases Non-Athletic All Footwear in Price Leather Other Restricted and Quality Footwear Footwear by OMA b/ Prices 1978 to 1981 .117 .132 .140 (.138) 1974-1977 and .021 .058 .037 1981-1982 (.036) 1974-1977 and -- -- .010 1981-1986 (.009) Quality 1978-1981 .055 .018 .023 (.014) 1974-1977 and .024 .105 .102 1981-1982 (.094) 1975-1977 and .021 -.011 .054 1981-1982 (.033) 1974-1977 and -- -- .008 1981-1986 (.065) Value Share 1976/1977 .398 .602 1.00 a/ Calculations are for calendar years. b/ Calculations are for July to June. Calendar-year calculations in parentheses. 20 suggest that even the non-VER increase is perhaps an overestimate of the underlying natural rate of change of footwear prices. On quality, the two studies differ a little more, but not drama- tically so. De Melo and Winters find relatively less upgrading during the VER than at other periods, while Aw and Roberts suggest the opposite, at least when the anomalous year 1974 is excluded. However, the upgrading findings are not particularly robust. Aw and Roberts have six observations of binding constraints: non-athletic leather footwear in 1978 and 1979, and other footwear in 1978, 79, 80 and 81. They find detectable upgrading in the first three cases and downgrading in the last three (other footwear during 1979-81). Moreover, their finding of upgrading coincides with the three observations with the largest price changes in their sample. This may be of significance because the commodity classification used by Aw and Roberts -- TSUSA -- classifies footwear by both type and price; thus some of the movement between TSUSA headings which is classified by Aw and Roberts as quality change merely involves movement between one price category to another, as "footwear, leather, work, welt, not over $2 per pair* to "footwear, leather, work, welt over $2 per pair, not over $5 per pair". 12/ This could, of course, still reflect improvements in quality, but it could also arise from VER-induced price rises with no change in nature at all. Our own data, while coarser in classification, are not subject to this problem, and thus might well be the more reliable indicators of quality as a result. 13/ Return now to table 5. Comparing the figures for all commodities with those for the restricted commodities alone, it appears that for Korean exports to the US, the growth of both prices and quality was greater for the restricted than for the unrestricted commodities while the VER was in 21 effect. During the unrestricted years, on the other hand, the prices of the unrestricted footwear appear to have grown faster. This again lends credence to the view that the VER caused some increase in the prices of the commodities it affected. There is no such indication for quality changes, however. A final comparison is made in table 7. This looks at the increase in the quality-corrected price index for Korean exports to the USA, Japan and the FRG using US bilateral weights. The figures for Japan and the FRG reflect the price changes those countries faced during 1974-86, but with the relative importance of each commodity in the aggregate determined by its importance in the bundle of exports destined for the US. Thus while they are not appropriate for commenting on Korean exports to Japan and the ERG, they are proper comparators for the rate of incrase in the prices of exports to the US. The new figures for Japan are virtually identical to those in table 5 and thus corroborate our view that the US experienced unusually large price increases for restricted products during the VER. Those for the FRG, on the other hand, shift considerably and end up looking fairly similar to the US. data. This casts some doubt on the hypothesis that the US experienced exceptional price rists, although it must be admitted that this result probably owes somethiztg to the way in which missing values were constructed. The FRG had consiferably more zero flows (80) than the US (4) or Japan (24). The notional price change for such a flow was equated to the change in the price of exports of the commodity concerned averaged over all destinations, an average in which the US usually figured very prominently. When FRG weights are used this does not matter much, because the price of zero trade flows receives approximately zero weight, but when US weights are used and the US data have influenced 22 Table 7: PRICE AND QUALITY CHANCES: BILATERAL INDICES (Average Annual Orowth Ratee) US Weights Tornqvi_ t J Tornqvist Prles Index Price Index 2/ Pero0d I US JAP FRO US JAP FRO Rostrlctedk/ R .188 .114 .160 .140 .172 .188 Commodities (21 Categorles) U . 009 .06 .01. .010 .089 -.030 All Co_inoditles d R I .116 .180 .116 .121 .169 .162 (26 categories) u I .0o3 .038 .052 .034 .013 .006 * U a unrestricted period 19 R a restricted period b/ categories restricted by the US OMA (see appendix for definition*) S/ all categories excluding parte, wood and cork, gaiters (see appendix for deflnitions) I/ Annual (calendar year) gf Quarter 3 - quarter 2 year 23 some of the prices, then one must expect the FRG figures to be biased towards the US ones. 6. Conclusions The application of index numbers to disaggregated Korean footwear exports during 1974-86 suggests that the period of binding VERs led to significant price increases, but not to the quality upgrading predicted by early theoretical analyses. The distinctive feature of our results is that, unlike virtually all previous empirical studies of VERs, they are derived from exporters', rather than from importers' data. Hence the results in this paper provide a more direct test of the reactions of exporting countries subject to VERs and to the theoretical analyses of quality upgrading which refer to the exporting countries behavior towards restricted markets. 9 Our comparisons drawn from multilateral indices allowed us to show that, after adjusting for product and country mix changes, the price of Korean footwear exports to the US (and to a somewhat lesser extent to the UK) rose significantly in the years when the VERs were binding (i.e. when quotas were filled). These results confirm and extend earlier work by Aw and Roberts (1988) on US footwear imports, but are more pertinent to testing the price and quality effects of VERs insofar as disaggregated exporters data to different sources provide a better control for product differentiation (etc.) than equally disaggregated importers data from several sources. These results were further confirmed by more disaggregated bilateral calculations. The multilateral results for product mix change suggest that the VERs were associated, if anything, with less quality upgrading during the 24 period of the VERs, a result also confirmed by bilateral comparisons. However, as noted in the introduction, quality downgrading in response to a VER is plausible if exporters have foreign exchange earnings objectives rather than profit maximization. Since the pattern of quality changes across markets did not appear to vary significantly during the VER period, it is also possible that adjustment costs were sufficient to prevent product mix change towards the restricted markets. To conclude, the analysis in this paper confirms empirically that VER-type restrictions result in rents that accrue to the exporting country (some further rents may be realized by importers if importing is not a competitive activity). However, the evidence for Korean footwear exports does not suggest any quality upgrading as a direct consequence of VERs, as, for instance, in autos. 25 Footnotes 1/ Quotas in the MFA and in American dairy products are quite precisely defined whereas VERs on autos, steel and footwear have been negotiated on broader categories of products -- i.e. uumber of autos, number of pairs of leather shoes. 2/ With adjustment costs, profit maximizing firms will equate per unit rents after deducting adjustment costs. If effective, non- transferability of quota tickets (prevalent among ASEAN countries (see Hamilton, 1986) will preclude quality mix adjustment. 3/ Yoffie (1983) documents that Taiwanese and Korean negotiators for the footwear negotiations with the US were primarily concerned with minimization of foreign exchange earnings loss. As indicated by Bark and de Melo (1988), the allocation of quota tickets for Korean footwear (and other products subject to VERs) is consistent with Yoffie's observation. 4/ This section draws on Hamilton (1989). 5/ See Winters (1988). An import quota was imposed on Taiwan in August 1977 and was converted into a VER in January 1981; the VER with Korea was instituted in January 1979. In 1985 the British footwear industry let the VERs with Taiwan expire as it was manifestly over-shipped. 61 West Germany, France, Italy, Netherlands. Other EC countries (e.g. Ireland) had bilateral restrictions. 7/ Expression (2) also holds for the bilateral index with ln Pij instead of ln P* 8/ Aw and Roberts (1988) multilateral comparison used 13 commodity groups. 9/ At this level of aggregation, the number of missing observations (pairs of country-commodity flows with zero values) is reduced to 27 out of a total of 1,092 observations. l0/ We also tested whether individual dummies for each of the restr!cted years (for both the US and the UK) were significant as a group and found them not to be statistically significant. We also ran regressions including dummy variables for Canada during the restricted period, yielding no discernable effect of the NTH. 11/ We exclude rubber overshoes (6401.3), women's shoes of artificial leather (6402.26), shoes of wood and cork (6405.0), gaitirs, spats, etc. (6406.0), all of which have very small and erratic exports, and parts (6405.0), for which upgrading is not an issue. 26 12/ In private communication, Aw and Roberts confirmed that their breakdowns were for broadly defined price categories (less than $2.50 per pair and over $2.50 per pair) which, to some extent, would minimize the possibility of measuring as quality upgrading, what is, in effect, a price rise induced by the quota. 13/ Our use of July to June data years allows us better to capture the effects of the OMA, which operated from July 1977 to June 1981. How- ever, to compare with Aw and Roberts we repeated our calculations on calendar years. This made very little difference to our US results, and hence to table 5, but it did affect those for the FRG and Japan having the effect of increasing their rate of price increase over the restricted period. Aw and Roberts define their restriction as starting with 1978 data expressed relative to 1977 values. 27 Bibliography Aw, B. and M. Roberts. 1986. "Estimating Quality Change in Quota Constrained Markets: The Case of US Footwear," Journal of International Economics, vol. 21, pp. 45-60. Aw, B. and M. Roberts. 1987. *Price and Quality Level Comparisons for US Footwear Imports: An Application of Multilateral Index Numbers," in R. Feenstra, ed., Empirical Techniques in International Trade. Cambridge: MIT Press. Bark, T. and J. de Melo. 1987. 'Export Mix Adjustment to the Imposition of VERs: Alternative Allocation Schemes," Weltwirtschaftliches Archiv, Band 123, pp. 668-78. Bark, T. and J. de Melo. 1988. "Export Quota Allocations, Export Earnings and Market Diversification," World Bank Economic Review, vol. 2, no. 3, pp. 341-8. Caves, D.B., L. Christensen, and W. Diewert. 1982. "Multilateral Comparisons of Output, Input, and Productivity Using Superlative Index Numbers," Economic Journal, vol. 92, pp. 73-85. Chang, E.T. 1984. "The Effects of Quantitative Controls on Imports of Footwear from Korea and Taiwan," University of Wisconsin, unpublished manuscript. Falvey, R. 1979. "The Comparison of Trade Within Import-Restricted Categories," Journal of Political Economy, pp. 1142-65. Hamilton, C. 1986. "ASEAN Systems for Allocation of Export Licenses Under VERaW in C. Findlay and R. Garnaut (eds.), The Political Economy of Manufacturing Protection: Experiences of ASEAN and Australia. Sydney, pp. 235-47. Hamilton, C. 1989. 'The Transient Nature of 'New' Protectionism: The Case of International Trade in Footwear," Weltwirtschaftliches Archiv, forthcoming. Melo, J. de and P. Messerlin. 1988. "Price, Quality and Welfare Effects of European VERs on Japanese Autos," European Economic Review, vol. 32, pp. 1527-46. Rodriguez, C. 1979. 'The Quality of Import and the Differential Welfare Effects of Tariffs, Quotas, and Quality Controls as Protective Devices," Canadian Journal of Economics, vol. 12, no. 3, pp. 439-49. Winters, L.A. 1988. "A Case Study of UK Footwear Trade Restraints,. mimeo. Yoffie, D. 1983. Power and Protectionism. Columbia University Press. 28 Appendixt 1/ The Korean Footwear Export Data The data, kindly provided by Taeho Bark of the Korea Development Institute, was collected on a quarterly basis from Korean Customs data publications. The data was initially collected for the period 1970 to 1986 for 12 export destinations and a residual wrest of the worldO (ROW) group. The data used in the main body of the paper was constructed by: (1) aggregating Panama, Norway, and Sweden into the ROW group and Italy, France, Netherlands, and Federal Republic of Germany into a category, unrestricted EEC; (2) annualizing the quarterly data flows. Prior to annualizing the data, an examination of the decumulated data revealed a fairly large number of typographical errors (e.g. positive value flow and zero price or negative entries) which amounted to 5 percent of the observations. After resolving these and other data inconsistencies, many errors remained in the data for 1973 (and earlier years). Because the Koreans changed their data classification in 1973 (and again in 1977) we decided to drop the observations prior to 1974. In the reduced data set covering 1974 to 1986, about 3 percent of the observations remained suspicious after correction for obvious errors (e.g. a decline in magnitude of one of the variables from quarter to quarter accompanied by increases in the other two). Next we used the standard Korean Trade classification (Revised 1977) and the original descriptions of the earlier classifications for 1973 and 1976 to obtain our own derived concordance (see Table Al) with 31 1/ We are extremely grateful to Paul Brenton for preparing the data set. 29 commodity groups. These data, when annualized, showed that Korean footwear exports are concentrated, in a relatively small number of commodity groups. Given the importance of zero entries in the multilateral indices, for the multilateral comparisons we aggregated the 31 commodity categories (new code column) to 14 categories (final code column). The 26 categories used in the bilateral comparisons are the 31 listed in the new code (table A2, column 3) less the 5 footwear categories listed in footnote 9. 30 Table Al: AGGREGATION OF FINAL COMMODITY CLASSIFICATION New Code Final Code Description 6401 FOOTWEAR WITH OUTER SOLES AND UPPERS OF RUBBER OR ARTIFICIAL PLASTIC MATERIAL 1100 1100 Boots 2000 9999 Other 3000 4500 6700 9000 6402 FOOTWEAR WITH OUTER SOLES OF LEATHER OR COMPOSITION LEATHER; FOOTWEAR (OTHER THAN IN 6401) WITH OUTER SOLES OF RUBBER OR ARTIFICIAL PLASTIC MATERIAL Footwear with uppers of leather 1100 1000 Other specified 1200 1300 1400 1500 1600 1700 1700 Outdoor sports shoes 1900 1900 Footwear n.e.s. Footwear with uppers of artificial leather 2100 2222 Other specified 2200 2300 2400 2500 2600 2700 2700 Outdoor sports shoes 2900 2900 Footwear n.e.s. 31 Table Al: AGGREGATION OF FINAL COMMODITY CLASSIFICATION (continued) New Code Final Code Description Footwear with uppers of fabric 4100 4100 Shoes for work, jungle, mountaineering and similar laced boots 4200 4200 Indoor sports shoes 4900 4900 Other Footwear with uppers of plastic and other footwear 3000 9999 Footwear with uppers of plastic and other footwear 6403, 6404. 6406 FOOTWEAR WITH OUTER SOLES OF WOOD OR CORK OR OF OTR MATERIALS AND QATJ'E1RS, SPATS, ETC. 6403.0000 6403.9999 Footwear with outer soles of wood or 6404.0000 cork of or other materials and 6406.000 gaiters, spats, etc. 6405 PARTS OF FOOTWEAR 0000 0000 Parts of footwear COUMTRY AGGREGATION 1. Japan 2. Unconstrained EEC -- France, W. Germany, Italy, Netherlands 3. Norway/Sweden 4. UK 5. Canada 6. USA 7. Australia 8. ROW -- Rest of World, Panama The final code referred to above is that assigned to the aggregated commodity groups and is to be found on file for these data. 32 Table A2: CONCORDANCE BETWEEN THE 1973, 1976 AND 1977 REVISIONS OF THE KOREAN TRADE CLASSIFICATION 73 Code 76 Code 77 Code New Code Description 6401 FOOTWEAR WITH OUTER SOLES AND UPPERS OF RUBBER OR ARTIFICIhL PLASTIC MATERIAL 12 120 0200 1000 Boots 1/ 15 130 19 190 20 200 0300 2000 Shoes 2/ 80 0500 30 300 0400 3000 Overshoes 40 430 0601 4500 Sandals and zori 3/ 50 500 0602 0699 60 600 0701 6700 Slippers 4/ 70 700 0702 0799 11 900 0100 9000 Footwear n.e.s. 5/ 90 0800 6402 FOOTWEAR WITH OUTER SOLES OF LEATHER OR COMPOSITION LEATHER; FOOTWEAR (OTHER THAN IN 6401) WITH OUTER SOLES OF RUBBER OR ARTIFICIAL PLATIC MATERIAL Footwear with uppers of leather 11 110 0100 1100 Furskin boots 12 120 0102 1200 Walker, mountain shoes and similar lace boots 13 130 0103 1300 Women's boots 14 140 0104 1400 Men's (and boys?) boots 6/ 15 150 0105 1500 Men's shoes 16 160 0106 1600 Women's shoes including high heels 33 Table A2: CONCORDANCE BETWEEN THE 1973, 1976 AND 1977 REVISIONS OF THE KOREAN TRADE CLASSIFICATION (continued) 73 Code 76 Code 77 Code New Code Description 17 170 0107 1700 Outdoor sports shoes 19 190 0108 1900 Footwear n.e.s. 7/ 0110 0199 Footwear with uppers of artificial leather 21 210 0201 2100 High (furskin?) boots 8/ 22 220 0202 2200 Walker(?), mountain and similar lace boots 9/ 23 230 0203 2300 Women's boots 24 240 0204 2400 Men's (and boys'?) boots 10/ 25 250 0205 2500 Men's shoes 26 260 0206 2600 Women's shoes (inc. high heel) 27 270 0207 .700 Outdoor sports shoes 28 280 0208 2900 Footwear n.e.s. 111 0299 Footwear with uppers of plastic 31 310 0300 3000 Footwear with uppers of plastic 12/ 32 320 33 330 34 340 36 350 37 360 39 390 34 Table A2: CONCORDANCE BETWEEN THE 1973, 1976 AND 1977 REVISIONS OF THE KOREAN TRADE CLASSIFICATION (continued) 73 Code 76 Code 77 Code New Code Description Footwear with uppers of fabric 41 410 0401 4100 Shoes for work, jungle, mountaineering and similar laced boots 42 420 0402 4200 Indoor sports shoes 43 430 0499 4900 Footwear n.e.s. 13/ 47 470 Other footwear 90 900 0500 9000 Other footwear 6403 FOOTWEAR VITH OUTER SOLES OF WOOD OR CORK 10 100 0000 0000 Footwear with outer soles of woord or cork 6404 FOOTWEAR WITH OUTER SOLES OF OTHER MATERIALS 00 000 0000 0000 Footwear with outer soles of other materials 6405 PARTS OF FOOTWEAR 11 110 0101 0000 Parts of footwear 19 190 0102 21 210 0103 29 290 0104 30 300 0199 40 400 0200 51 510 0300 59 590 0400 90 900 0500 6406 GAITERS, SPATS, ETC. 10 0000 0000 0000 Gaiters, spats, etc. 90 35 Table A2s CONCORDANCE BETWEEN THE 1973, 1976 AND 1977 REVISIONS OF THE KOREAN TRADE CLASSIFICATION (continued) Notess 1/ In the notes to the 73 classification the descriptions of the groups 12, 15 and 19 all mention boots containing furskins, whilst for the apparently identical groups in the 76 classification (120(1/2 boots), 130(ordinary boots) and 190(waders)) furskins are not mentioned. In the 77 classification there is a group (0100) described as "footwear containing furskins". The entrires in this group are very small or zero, whilst the entries under 12, 13, and 19 in 73-75 and 120, 130, and 190 in 76 are all relatively large. Thus we conclude that 12, 13, 19, in 73-75 are compatible with 120, 130 and 190 in 76. In the 77 classification there is only one apparently compatible group (0200) entitled wboots". Thus we combine the three groups in the earlier classifications. This leaves a problem with group 11 in 1973 which is described as "high boots containing furskins' and group 0100, described above, in the 77 classification. In both cases the entries are small or zero and so oe group them together with "other footwear" (see below). It was deemed important to keep the 'boots" group separate since this type of footwear was excluded from the OHA negotiated with the US in 1977. 2/ All three classifications contain a group entitled "shoes with soles and uppers of rubber or plastic material' (20,200,0300) whilst the 73 and 77 classification show an additional group "rubber shoes". In both cases the flows are very small or zero (and so all entries may have been zero in 1976) and so we combine to obtain a single group "shoes". 31 In the 73 and 76 classifications sandals and zori are split into two groups; "of rubber' (40,400) and 'of plastic" (50,500). In the 77 classification in addition to these two groups (0601,0602) there is a further group "of other" (0699). We assume that the latter were formerly classified in either of the two groups of sandals and zori, and hence obtain a single group by combining all the relevant groups of sandals and zori. 4/ The same issue as above applies to slippers. 5 See note 1. 6/ The 73 and 76 classifications explicitly mention "mens and boys", while the latter is excluded from the description of the 77 group. Without further information we assume that the groups in the different classifications are identical. 7/ In the 77 classification there are two additional groups not present in the earlier classifications ('shoes for indoor sports" and 'work shoes"). We assume that such footwear was previously classified under "footwear n.e.s." 36 Table A2: CONCORDANCE BETWEEN THE 1973, 1976 AND 1977 REVISIONS OF THE KOREAN TRADE CLASSIFICATION (continued) 8/ The description of this group in the 77 classifiction simply reads 'furskin boots" whilst in the earlier classification it is described as "high boots'. There are no other possible matches for the 77 group. 9/ The description of the 77 group omits "walkers' which do not reappear elsewhere in the classification. We assume that if flows of such footwear are positive after 1977 they are classified to this group are not to 'footwear n.e.s.'. 10/ As in note 6 above. ll/ The 77 classification contains an additional group 'shoes for indoor sports' which cannot be matched to any of the groups in the earlier classifications. We assume that such footwear was previously classified under 'footwear n.e.s.'. 12/ The 77 classification allows for only one group of plastic footwear, while in the earlier classifications there are seven. This appears to reflect the very small flows of such footwear. Hence we are obliged to sum the more detailed groups in the 73 and 76 classifications. 13/ The 73 and 76 classifications include a group 'shoes with uppers of canvas' which is not present in the 77 classification. Hence we assume that such footwear is now classified under 'footwear n.e.s.'. The new Code referred to above is the code we assign to groups following concordance. PPR Working Paper Series Title Author Date Contact WPS194 Efficient Debt Reduction Jeffrey Sachs May 1989 L. Chavarria 33730 WPS195 How Has the Debt Crisis Affected Commercial Banks? Harry Hulzinga May 1989 L. Chavarria 33730 WPS196 A Review of Alternative Debt Strategies Eugene L. Versluysen May 1989 E. Hadjlgeorgalis 33729 WPS197 Differentiating Cyclical and Long-Term Income Elasticities of Import Demand Fernando Clavljo May 1989 K. Cabana Riccardo Faini 61539 WPS198 Equity In Unequal Deductions: Implications of Income Tax Rules In Ghana and Nigeria Chad Leechor May 1989 A. Bhalla Robert Warner 60359 WPS199 Private Sector Assessment: A Pilot Exercise In Ghana Samuel Paul May 1989 E. Madrona 61712 WPS200 Women and Development: Objectives, Framework, and Policy Interventions T. Paul Schultz April 1989 J. Klous 33745 WPS201 WPS202 A Cross-Section Analysis of Financial Policies, Efficiency and Growth Alan H. Gelb WPS203 Optimal Commodity Taxes In the Presence of Rationing Nanak Kakwani Ranjan Ray WPS204 The Impact of Labor Costs on Manufactured Exports In Developing Countries: An Econometric Analysis Luis Riveros May 1989 R. Luz 61762 WPS205 What Determines National Saving? A Case Study of Korea and the Philippines Sang-Woo Nam May 1989. R. Luz 61762 PPR Working Paper Series Title Author Date Contact WPS206 The Effects of Single-Sex Schooling on Student Achievement and Attitudes In Nigeria Valerie E. Lee May 1989 C. Cristobal Marlaine E. Lockheed 33640 WPS207 Occupational Training Among Peruvian Men: Does It Make a Difference Ana-Maria Arriagda May 1989 C. Cristobal 33640 WPS208 Effective Primary Level Science Teaching in the Phtilippines Marlaine E. Lockheed May 1989 C. Cristobal Josefina Fonacier 33640 Leonard J. Bianchi WPS209 Can the Industrial Countries Return to Rapid Growth? Space Seminar International Economics Department and International Economic Analysis and Prospects Division WPS210 Notes on Cash-Flow Taxation Roger H. Gordon WPS211 Coffee Pricing Policies in the Dominican Republic Panos Varangl^ May 1989 D. Gustafson 33714 WPS212 Beyond the Debt Crisis: Alternative Forms of Financing Lessard WPS213 Conditionality and Debt Relief StlJn Claessens Ishac Diwan WPS214 Macroeconomic Adjustment and Labor Market Response: A Review of the Recent Experience In LDCs Peter R. Fallon Luis A. Riveros WPS215 Adjustment and Income Distribution: A Counterfactual Analysis Francois Bourguignon William H. Branson Jaime de Melo May 1989 M. Ameal 61466 WPS216 Price and Quality Effects of Vers--Revisited: A Case Study of Korean Footwear Exports Jaime de Melo June 1989 M. Ameal L. Alan Winters 61466