The Effect of MFA Quota Removal on Apparel Exporters: Kenya, Tanzania and Uganda. February 2005

The Effect of MFA Quota Removal on Apparel Exporters: Kenya, Tanzania and Uganda February 2005 Çağlar Özden 1 DECRG World Bank 1 * Development Research Group (DECRG), The World Bank, 1818 H Street, NW, Washington, DC 20433; Email: cozden@worldbank.org. The views expressed here are those of the author and should not be attributed to the World Bank. 1

Executive Summary The objective of this study is to assess the impact of the MFA quota removal on Kenya, Tanzania and Uganda. Although all three countries are among the beneficiaries of the unilateral trade preferences granted by the United States under the African Growth and Opportunity Act (AGOA), apparel exports are especially important for Kenya. Producers in Kenya exported close to $190 million worth of apparel to the US in which comprised 75% of all exports to the US. While many countries in Africa, Latin America and the Caribbean enjoy preferential access to the US and the EU markets for their apparel exports, many countries in South and East Asia were facing rather restrictive trade barriers in the form of MFA quotas. WTO members agreed to slowly relax and finally eliminate these quotas during the Uruguay round negotiations. The final and most restrictive group of quotas were removed on 1/1/2005, leaving tariffs as the only trade restrictions in place. Many large exporters that were not restrained by quotas and enjoyed preferences, such as Kenya, are quite concerned about the considerable erosion of these preferences and the damage this might inflict on their fledgling apparel sectors. We first examine the overall patterns of the US apparel imports, most importantly by comparing the exports of quota facing countries in South and East Asia with the exports of preference receiving countries in Latin America, the Caribbean and Africa. The total apparel exports from quota facing countries increased in volume but declined in market share to 53% in 2003. More importantly, we show that around 65% of their exports face quotas but only 2/3rd of the quotas are binding. In short, slightly over 20% of US imports entered under binding quotas in 2003. It is the relaxation of the quotas on these exports that will have the largest impact on all countries involved. The performance of exporters from AGOA countries is one of the most notable developments in the apparel market during the last years. Total apparel exports increased from AGOA countries increased from $360 million in 1996 to $1.5 billion in 2003. the main beneficiaries are Lesotho, Kenya, Madagascar and Swaziland. Unfortunately, these countries compete at the lower end of the quality ladder where price competition is more intense and likely to increase after quota removal. 2

When we analyze the impact of MFA quota removal on Kenya, we find that East and South Asian countries have a smaller overall total market share in the categories that Kenya exports. The main competitors of Kenya are other preference receiving countries. However, a larger portion of existing Asian exports enter under binding quotas in Kenya s main export categories, compared to an average category. In the final section, we perform product specific analysis for Kenya. The most important observation is the concentration of exports in few categories. Despite a relative diversification in recent years into knitted and synthetic-fiber categories, cotton notknitted items still account for over 75% of apparel exports of Kenya. When we observe the largest export categories, we notice that market shares of quotafacing countries have been in decline in categories in which Kenyan exporters have been most successful. These are also the categories in which the share of exports entering under binding quotas have also been in decline. Thus, we an predict the continuing success of Kenyan exporters. The same is also true for certain categories that the Kenyan exporters recently entered, such as knitted sweaters (both cotton and synthetic fibers). We finally look at the Kenyan export prices for specific categories and observe that these have also been improving which implies relative quality upgrading. This is another positive signal for the future. 3

1. Introduction The objective of this paper is to assess the impact of the MFA quota removal on Kenya, Tanzania and Uganda based on tracking the evolution of individual countries exports to the US over the last 7 years. These three countries are all beneficiaries of the African Growth and Opportunity Act (AGOA) which grants their exports quota-free and duty-free entry into the United States, their main destination by a very wide margin. Even tough they enjoy identical trade preferences, Kenya, Tanzania and Uganda have very different apparel export profiles. Kenya exported close to $190 million worth of apparel to the US in 2003, up from $27 million in 1996. As a result, apparel became the most important export category for Kenya at 75% of total exports in 2003to the United States. Tanzania and Uganda, on the other hand, are not major exporters of apparel. Tanzania was exporting $8 million worth of apparel as late as 1998 and this was close to 25% of its total exports. Despite the AGOA preferences, apparel exports collapsed in later years and totaled only $1 million in 2003. Uganda barely exported any apparel until 2002 and the volume reached $2 million in 2003 which is only 4% of total exports. For the first three quarters of 2004, the latest time frame for which we have data, total apparel exports from Kenya were close to $200 million, for Tanzania $1.5 million and for Uganda slightly above $3 million. AGOA preferences explain only part of the story about the rapid increase especially in Kenya s exports to the US. These preferences are valuable as long as other competitors are facing restrictive tariff barriers. In the case of apparel, these were the MFA quotas. Many countries in South and East Asia, which have strong comparative advantage in apparel manufacturing, have faced rather restrictive quotas (in addition to tariffs close to 20%) on their exports during the last three decades. WTO members agreed to slowly relax and finally eliminate these quotas during the Uruguay round negotiations. The final and most restrictive group of quotas were removed on 1/1/2005, leaving tariffs as the only trade restrictions in place. 4

The quota relaxation and elimination process caused much turmoil and anxiety in the global trading regime over the last decade, especially as we got closer to January 1, 2005 deadline. Their removal will naturally have a significant impact on many developing countries that were facing restrictive quotas in certain categories. Many other developing countries are also quite concerned about the impact of quota removal since they enjoy preferential access into the US and the EU markets. Many large exporters, such as Kenya, are concerned about the considerable erosion of these preferences and the damage this might inflict on their fledgling apparel sectors. In the following paragraph, we outline a strategy to measure certain effects, especially relying on the US trade flows and policy data. In the next section, we describe the two data sources (import volume/price/tariffs and quota utilization) used in this study and how we combined them. Then we describe the apparel import patterns of the United States since 1996. We specially identify the exports patterns from three groups of countries: countries that have preferential market access, such as Kenya; countries that face tariffs, such as Italy; countries that face both tariffs and quotas, such as China. Then we analyze the overall import volume by quota status: no quotas, binding quotas and non-binding quotas. Next is the analysis of imports from AGOA countries which is followed by the export performance of Kenya, Tanzania and Uganda. The next two sections look at the tariffs on imports and prices of apparel imports, respectively, especially by different exporting country categories. Then we analyze the impact of MFA quota removal on these three countries. The analysis is carried out at the aggregate level and we especially look at what percent of Kenya s exports overlap with exports from quota-facing countries and are in binding-quota categories. Aggregate analysis is followed by sectoral analysis. We focus on several sectors which are either historically important or potentially promising categories for Kenya. We look at the share of exports from quota-facing countries as well as exports entering under binding-quotas in each category to comment on the impact of quota removal. 5

2. Data The United States International Trade Commission (ITC) provides detailed data on bilateral trade flows between the US and its partners. More specifically, we have unit volume, price and tariff data at the 8-digit product category level and for each preference program (such as NAFTA, AGOA or CBI) for each trading partner. This enables us to track the evolution of individual countries exports to the US over time (in volume, price and quality), especially as the quotas are relaxed and finally removed. The second source of data is Office of Textiles and Apparel (OTEXA) which provides the data on quota levels and their utilization across product categories and countries which face them. The data provided by OTEXA is compiled using a different classification regime which is quite different than the ITC trade flow classification categories. Furthermore, the quota classification show significant variation across countries and, to a certain extent, over time. One of our innovations is to provide a concordance between the two classification systems so that we can analyze quota restrictiveness and utilization data together with trade flow data. All of the graphs and tables in the paper were based on the data from these two sources. 3. Apparel Imports of the United States Total apparel imports 2 of the United States have reached $62.5 billion in 2003 up from $37.7 billion in 1996. Over this time period, the growth rate of apparel imports has been slightly above the growth rate of aggregate imports. Although it is the largest import category after motor vehicles, mechanical and electrical machinery and oil, apparel comprises only 2% of the total US imports. Despite this relatively small share, the trade policies faced by apparel imports are some of the most fiercely debated issues in the global arena. Due to relatively labor intensive production methods and low technology 2 For the purposes of this paper, apparel includes categories 61 and 62 of the HTA classification which are officially defined as woven or knitted articles of apparel and clothing accessories. 6

requirements, many developing countries have comparative advantage in apparel and related products In addition, for some countries, these are among the most important export categories. On the other hand, domestic producers are highly organized and geographically concentrated in developed countries. They have been able to mobilize resources to lobby effectively against liberalization of trade restrictions on apparel imports over the years. As a result, over the last three decades, apparel and textiles sectors in developed countries were protected through special quota arrangements, called the Multi-Fiber agreement (MFA) as well as relatively high tariffs averaging around 20%. Each importer, mainly the US and the European Union, established its own MFA regime with country and product specific quota limits which could be adjusted as the economic and political conditions changed. Over time, the MFA regime became a rather complicated arrangement and imposed significant transactions and administrative costs on parties involved. Many developing countries, especially in South and East Asia, had their apparel and textile exports subjected to restrictive quotas by the US and the EU. On the other hand, some other developing countries not only were exempted from these quotas but were actually granted preferential access through special unilateral and reciprocal trade programs. In the case of the United States, the first program to be implemented was the Caribbean Basin Initiative (CBI) in 1983 and modified in 1990. The main benefit was the tariff and quota free treatment granted to apparel imports (subject to certain rules of origin requirements) from the eligible countries. The Caribbean Basin Trade Partnership Act (CBTPA), passed in 2000, extended the benefits of the original program considerably. In essence, the new rules provide NAFTA-equivalent treatment for certain items (mainly apparel) which had partial preferences under the original CBI and were excluded from duty-free treatment under the GSPTwenty four countries in the Caribbean and Central America are currently eligible (I hope I this is correct). In 2004, five Central American countries (Costa Rica, El Salvador, Guatemala, Honduras and Nicaragua) and the Dominican Republic signed the Central America Free Trade agreement (CAFTA) 7

with the United States. This agreement does not change the status of most exports from these countries but makes the market access privileges permanent. The next program to be implemented was the Andean Trade Preferences Act (ATPA) which extended unilateral preferences to Bolivia, Colombia, Ecuador, and Peru. Enacted in 1991 as part of U.S. efforts to reduce narcotic production and trafficking, it was modeled after the CBI and has similar eligibility requirements and product coverage. ATPA was renewed in 2002 as the Andean Trade Promotion and Drug Eradication Act (ATPDEA) and expanded to include tuna, leather and footwear products, petroleum products, and apparel again subject to restrictive rules of origin similar to the CBI. For example, if apparel is assembled from U.S. fabrics, no quotas or duties apply, but if local inputs are used, duty-free imports are subject to a cap of 2 percent of total U.S. apparel imports (increasing to 5 percent in equal annual installments). The final unilateral program implemented by the US is the African Growth and Opportunity Act (AGOA), passed in 2000, which offers beneficiary Sub-Saharan African countries duty-free and quota-free market access for essentially all products. AGOA excludes textiles but extends duty- and quota-free treatment for apparel made in Africa from U.S. yarn and fabric. If regional fabric and yarn are used, there is a cap of 1.5 percent of U.S. imports, increasing to 3.5 percent over eight years. African LDCs are basically exempt from all rules of origin for a limited period of time. The relaxation of the rules of origin requirements helped apparel exports to significantly expand from countries such as Kenya and Lesotho. Currently, there are 37 sub-saharan countries eligible for AGOA preferences, including Kenya, Tanzania and Uganda. However, only eight of these countries are exporting any significant amount of apparel to the US. These are Lesotho, South Africa, Kenya, Mauritius, Madagascar, Swaziland, Namibia and Malawi (see Olarreaga and Ozden (2005) for an initial analysis of apparel exports of AGOA countries). A key feature of all of these programs is that, by legislation, they are in effect for certain period of time and need to be renewed by the Congress periodically after deliberations. 8

Apart from these developing countries that receive unilateral market access, the US has signed bilateral free-trade agreements with a number of countries. These are Mexico and Canada (under NAFTA), Israel, Jordan, Singapore, Chile and as mentioned before Costa Rica, El Salvador, Guatemala, Honduras, Nicaragua and the Dominican Republic (under CAFTA) and Australia. There are several negotiations continuing, including, South Africa, Andean Countries and Panama. Apparel exports from all of these countries also enjoy duty and quota free access to the United States market. As a result of these highly discriminatory trade policies, apparel exporting countries to the United States may be divided into three categories: (i) Countries that are exempt from all restrictions due to bilateral, regional or unilateral trade preferences. These are FTA countries (such as Mexico, Canada, Israel etc.), Central American and Caribbean Countries (through CAFTA and CBI/CBTPA), Andean Countries (through ATPA) and African countries (through AGOA) (ii) Countries that are subject to quotas and tariffs. These are mostly developing countries in South and East Asia. The largest exporters are China, Hong Kong, Vietnam, Indonesia, India, Philippines, Korea, Bangladesh, Thailand, Taiwan, Sri Lanka, Turkey, Cambodia, Pakistan and Malaysia. (iii) Countries that are subject to tariffs, but not to Quotas. These are mostly developed OECD countries that specialize in higher-value items. Among the largest exporters are Italy, Japan, France, United Kingdom, Australia and Portugal. 9

Apparel Imports of the US - Different Exporters Apparel Imports of the US - Different Exporters 70 70% Billions $ 60 50 40 30 20 10 Percentage of Total Imports 60% 50% 40% 30% 20% 10% 0 1996 1997 1998 1999 2000 2001 2002 2003 0% 1996 1997 1998 1999 2000 2001 2002 2003 TOTAL PREFERENCE RECEIVING QUOTA FACING DEVELOPED QUOTA FACING DEVELOPED PREFERENCE RECEIVING Graph 1. Apparel Imports of the US from Different Exporters The figure above shows the value of exports from these three groups of countries and their market shares in the US. Quota facing countries increased their exports from close to $22 billion in 1996 to over $33 billion in 2003 but their market share declined from 58% to 53%. Preference receiving countries exported around $12 billion in 1996 which was 31% of total US apparel imports. Their export volume increased to more than $22 billion in 2000 and a had a peak market share of 38%. Since then, the volume stayed the same but the market share went down to 35% due to increasing apparel exports from quota-facing countries. Exports from developed countries stayed stable between $4-5 billion in this time frame, corresponding to roughly 10% of total imports of the US. In conclusion, the apparel import data of the US reveals the following: (i) (ii) there is a steady increase in the US imports of apparel at the same as the rest of the imports, the main beneficiaries have been the preference-receiving countries which experienced rapid growth during the last decade but this seems to be slowing down 10

(iii) (iv) quota-facing countries increased their export volume, lost market share but the decline seems to be ebbing developed countries which do not enjoy preferences but do not face quotas maintained their market share 4. MFA Quota Phase-out The most important change in the history of the MFA occurred during the Uruguay Round Negotiations that witnessed the creation of the World Trade Organization (WTO) and active participation by developed countries. One of the key agreements pertained to the gradual elimination of the quotas created under the MFA. The first, second and third quota elimination (referred to as "integration" in official documents) took effect on January 1st 1995, 1998 and 2002 respectively. The most restrictive and important quotas were left for the final integration stage which was scheduled for January 1st, 2005. Furthermore, in each stage, existing quotas were expanded gradually. As a result, even before they were eliminated, quotas for many countries and/or products ceased to be binding. Graph 2 represents the distribution of apparel exports to the US by quota status. Apparel Imports of the US by Quota Status Apparel Imports of the US by Quota Status 50 80% 40 60% Billions $ 30 20 Percentage 40% 10 20% 0 1997 1998 1999 2000 2001 2002 2003 0% 1997 1998 1999 2000 2001 2002 2003 No Quotas Binding Quotas Non-Binding Quotas No Quotas Binding Quotas Non-Binding Quotas Graph 2. Apparel Imports of the US by Quota Status 11

(a) The largest portion of apparel imports of the United States are actually not subject to any quotas. This category includes all apparel exports by preference receiving countries and by developed countries, groups (i) and (iii) from the previous section. Furthermore, not all apparel categories of countries in South and East Asia (group (ii) above) face quotas, even tough a large of their exports do so. Imports exempt from any form of quotas increased from $26.5 billion in 1996 to $41.70 billion in 2003. Their overall share in US imports increased from 60% to 67%.This growth is due to increase in exports from preference receiving countries as well as the elimination of some quotas during this period. (b) The second group of exports enter under binding quotas. We refer to a quota as binding if the fill-rate is higher than 80%. Recall that each country-specific quota category covers multiple HTS 10-digit trade categories. We apply the same quota-category specific fill rate to all HTS-categories that is covered by it. Then we calculated total trade volumes by quota status. Volume of imports entering the US under binding quotas were between $12 and $14 billion during 1996-2003. Since the total imports were increasing more rapidly, the share under binding quotas declined from 28% in 1996 to 21% in 2003. (c) The third group by quota status is the exports that enter quotas that are non-binding, i.e. where the quota fill-rate is below 80%. This group increased from $5.5 billion in 1996 to $7.5 billion in 2003 while the share stayed stable around 12%. In some studies, the distinction between binding and non-binding quotas is ignored. This is mainly due to the difficulty of merging the quota data (which records trade by unit volume) with the trade flow data (which records trade by value). However, the distinction is crucial as can be seen. Close to 1/3 rd of imports that are covered by a 12

quota are actually entering the US under non-binding quotas which means those quotas do not really impose significant restrictions on the exporters (other than some paperwork) and their removal will not have much material impact on the trade patterns. In the section analyzing the impact of the MFA quota removal on Kenya, Tanzania and Uganda, we will revisit the issue of binding vs. non-binding quotas. 5. AGOA Countries In the previous section, we mentioned the African Growth and Opportunity Act (AGOA). AGOA was signed by President Clinton in May 2000 and quickly became one of the most high-profile unilateral preference programs implemented. The aim of AGOA is higher levels of trade and direct investment in support of positive economic and political developments throughout Sub-Saharan Africa (US Congress [2000]). As of December 2004, 37 countries were designated to be eligible for the AGOA preferences. AGOA has several unique provisions: (i) it covers items such as apparel, footwear, handbags, luggage etc with significant importance for most developing countries, and (ii) the Apparel Provision which, in essence, relaxes the rules of origin requirements 3, a standard feature of most preferential trade agreements 4. In addition, Special Rule for Lesser Developed Countries 5 allows duty-free access for the apparel made from fabric from anywhere in the world. As of December 2004, all apparel exporters, except South Africa and Mauritius, qualified for this special rule that basically eliminates rules of origin requirements. The apparel sectors in AGOA countries differ country by country. South Africa and Mauritius have mature apparel industries while the other countries are relatively new 3 Such rules are important elements of bilateral agreements such as NAFTA and unilateral programs such as CBI. 4 Under AGOA II, apparel produced with regional or U.S. made yarn and fabric (at least 85% of value added) have duty-free and quota-free access to the U.S. market. There is a cap of 3% of total U.S. imports, growing to 7% over an 8-year period, but this cap is far from binding. Apparel exports under AGOA provisions are currently less than 1% of total US imports in these sectors. 5 Those with per capita GNP below $1,500 in 1998. 13

to the sector. In some countries, South East Asian parent companies have strong influence in ownership and management of the firms; this is the case in Lesotho, Kenya and Malawi (Salm [2002]). One of the consequences of AGOA has been this rapid foreign investment by East Asian companies in these countries. In most cases, the raw materials are imported from Asia and the majority of the finished apparel is exported to the US, taking advantage of the absence of rules of origin requirements. The trade data indicates that a small group of apparel exporting coutnries have been the main beneficiaries of AGOA preferences so far and there has not been significant change in the exports of any other product from the eligible countries. These countries, in order of declining apparel exports to the US, are Lesotho, Mauritius, South Africa, Madagascar, Kenya, Swaziland, Namibia and Malawi. These eight countries account for 98% of apparel exports to the US from the 37 eligible countries. Apparel Exports from AGOA Countries to the US Billions $ 1.75 1.50 1.25 1.00 0.75 0.50 0.25 0.00 1996 1997 1998 1999 2000 2001 2002 2003 AGOA AGOA excl. S.Africa & Mauritius Graph 3. Apparel Exports from AGOA Countries to the US South Africa and Mauritius are the only two countries that have been apparel exporters historically. The other six countries are relatively new entrants and their success, in many respects, are due to the valuable preferences granted under AGOA. For example, exports from Lesotho increased from $65 million in 1996 to close to $400 in 2003. for Madagascar, Kenya and Swaziland, the growth is even more dramatic from 14

an average $20 million each to around $200 million each. Similarly, Namibia and Malawi are new entrants with almost no imports in 1996 to $40 million in 2003. The preceding graph in Figure 3 depicts the apparel exports of AGOA countries to the US. The total apparel exports of AGOA countries to the US increased from $360 million in 1996 to $1.5 billion in 2003. South Africa and Mauritius accounted for close to 2/3 rd in 1996 but their share had dropped to 1/3 rd even tough their exports had more than doubled. Apparel exports went from 10% to 30% of total exports (excluding oil and precious metals) from AGOA eligible countries. In conclusion, the AGOA experience can be summed up as the following: AGOA countries enjoyed a rapid increase in their exports, but there is only a handful of them. These are Lesotho, Kenya, Madagascar and Swaziland. Namibia and Malawi are new successful entrants. Tanzania seems to have failed in this market while the future of Uganda is too early to forecast. 6. Export Performance of Kenya, Tanzania and Uganda Kenya, Tanzania and Uganda exhibit three completely different patterns in terms of their apparel exports to the US. In 1996, only Kenya was exporting apparel to the US and the volume was $27 million. During the following years, Kenya managed to take full advantage of AGOA preferences and increase its exports to $190 million. It is frequently quoted as one of the most remarkable success stories of AGOA. Tanzania s apparel exports were $4 million in 1996, increased to $8 million in 1998 but declined sharply since. In 2003, the total volume was less $1 million. Uganda, on the other hand, basically exported no apparel to the US until 2003 during which the total approached $2 million. The countries experience is presented in the following Graph 4. 15

Apparel Exports of Kenya to the US Apparel Exports of Tanzania and Uganda to the US 200 10 160 8 Millions $ 120 80 40 Millions $ 6 4 2 0 1996 1997 1998 1999 2000 2001 2002 2003 0 1996 1997 1998 1999 2000 2001 2002 2003 Tanzania Uganda Graph 4. Apparel Exports from Kenya, Tanzania and Uganda to the US A more illustrative graph would be the share of apparel in total exports to the US as presented in the Graph 5. In 1996, apparel constituted around 25% of their total exports to the US for both Kenya and Tanzania, but the ratio was zero for Uganda. Over the years, apparel became more important for Kenya and reached 75% of total exports to the US in 2003. On the other hand, the percentage of apparel in total exports to the US declined to 4% for Tanzania in 2003 which is also the same level for Uganda. The data shows that apparel became a very significant export sector for Kenya as the exporters used the AGOA preferences to the full extent. On the other hand, Tanzania was at a similar starting position in 1996 but failed to take advantage to the same extent and its apparel export volume declined sharply both in terms of absolute and relative volume. It is too early to make any judgments for Uganda. After years of absence in this sector, exports started to appear in 2003 and it is too early to tell whether if this is a permanent or temporary event. 16

Apparel as a Share of Total Exports to the US 80% Percentage 60% 40% 20% 0% 1996 1997 1998 1999 2000 2001 2002 2003 Kenya Tanzania Uganda Graph 5. Apparel as a Share of Total Exports to the US for Kenya, Tanzania and Uganda 7. Tariffs In addition to highly restrictive quotas, apparel exports are subject to relatively high tariffs. While the tariffs imposed by the US on other manufactured goods are generally below 5%, the average MFN tariffs on apparel and textiles products are close to 20%. Of course, the preference receiving countries are exempt from these tariffs (as long their products satisfy rules of origin requirements) while the developed countries and quotafacing developing countries face these additional barriers. Ave ra ge ta riffs Fa ce d By Diffe re nt Ex porte rs 20% 15% Percentage 10% 5% 0% 1996 1997 1998 1999 2000 2001 2002 2003 TOTA L QUOTA FACING PREFERENCE RECEIVING DEVELOPED AGOA Graph 6. Average Tariffs Faced by different Countries on Apparel Exports to the US 17

The graph above presents the average tariffs paid on their exports by different country groups. The tariffs paid by quota-facing developing countries and developed countries remained mostly stable since 1996, around 17% and 18%, respectively. The tariffs on the exports from preference-receiving countries, on the other hand, declined continuously from an already low level of 6% to 3%. As a result, the average net tariffs on the total imports of the US is 11% in 2003, down from 14% in 1996. Finally, the most dramatic decline was experienced by the AGOA countries whose average tariffs declined to 3% in 2003 from 19% in 1996. This is another reason why the exports to US increased so rapidly 6. The same pattern is true for Kenya, Tanzania and Uganda as we can see in the net graph. All three faced tariffs near 20% in 1996 and this declined to around 1% in 2003 as they use AGOA preferences. Thus it is reasonable to assume that the existing exporters in these countries, especially Kenya, learned quite rapidly how to satisfy the rules of origin requirements of AGOA and will continue to utilize them in the future. Average Tariffs faced by Kenya, Tanzania and Uganda 25% 20% Percentage 15% 10% 5% 0% 1996 1997 1998 1999 2000 2001 2002 2003 Kenya Tanzania Uganda Graph 7. Average Tariffs Faced by Kenya, Tanzania, Uganda on their Apparel Exports to the US 6 Note that the tariffs on the exports of AGOA countries are not zero. This is due to the fact that the rules of origin on certain exports could not be met and tariffs had to be paid. 18

Finally, we should emphasize that the tariffs on apparel imports of the US and the EU will stay in place even after quotas are eliminated. Preference-receiving countries that are concerned about the consequences of quota removal will continue to have considerable advantage over quota-facing countries. Currently, there are no unilateral plans by the US government or any international negotiations to lower MFN tariffs on apparel. In conclusion, the tariff data tells us that (i) Developed and quota-facing countries face rather high and stable tariffs (ii) Preference receiving countries continue to enjoy a large advantage due to continuing tariffs which will be maintained in the future (iii) AGOA countries, including Kenya, Tanzania and Uganda, managed to take advantage of the preferences and face rather low tariffs. 8. Prices One of the frequently ignored issues in empirical work in international trade is that trade policies should be mainly evaluated by looking at their effect on prices. However, this is rarely implemented empirically, mainly due to data limitations. Luckily, the US trade data has very detailed and disaggregated unit value and quantity data for apparel imports. Using the price data, we can gain valuable insights into the effects of policy changes on the welfare of exporting countries as well as the US. Graph 8 presents average unit prices of total apparel imports for different groups of countries. The average is calculated by taking weighted average of unit prices for each 10-digit HTS category for each country (or group of countries) in a given year. Average unit price of total imports was $127 in 1996 and it continuously declined to $99 in 2003. These are nominal prices and the decline in real terms is much higher. The decline is due to lowering of trade barriers as well as increased competition in the global apparel market. The main beneficiary of this 20% price decline, naturally, are the American consumers 7. 7 The role of the exchange rate fluctuations is an important and yet difficult issue to identify. There is a vast exchange-rate pass-through literature that aims to identify what portion of exchange rate fluctuations are 19

300 Average Unit Prices of Exports by Different Exporters 250 200 Dollars $ 150 100 50 0 1996 1997 1998 1999 2000 2001 2002 2003 TOTAL QUOTA FACING PREFERENCE RECEIVING DEVELOPED AGOA Graph 8. Average Unit Prices of Apparel Exports by Different Exporters The average prices of developed countries exports are much higher than the other groups of countries, mainly reflecting the large quality gap and product mix (such as suits and dresses which tend to be more expensive). Nevertheless, there is rapid decline there as well, from $276 in 1996 to $224 in 2003. Quota-facing developing countries command, on average, 20% higher prices than preference-receiving countries. There are multiple explanations for this gap. First, due to quota restrictions, these countries might be preferring to export higher quality items since the quota imposes a per-unit tax rather than an ad valorem tax and it is much lower in percentage terms for higher value items. The average prices of apparel exports of quota-facing countries declined from $115 to $88 in 2003. The decline is less sharp for preference-receiving countries, from $82 to $75 passed on to the consumers in the importing country. If the exporters in a given country have significant pricing power, they can pass their higher costs to the American consumer when their currency appreciates. However, the impact of exchange rate fluctuations on prices seem to be less important in the apparel case. It is one o the most competitive sectors and most exporters, especially in developing countries, are unlikely to have significant pricing power. Furthermore, the prices seem to be in a smooth secular decline while exchange rates exhibit relative volatility. 20

during the same time frame. The decline is partially due to relaxation of quotas which increased competition. It is also the case that some of the preference receiving countries moved up the quality ladder as they find it difficult to compete with Asian exporters on price. AGOA countries receive the lowest prices for their exports, around $65 during 2002-3, implying their products are of relatively low quality. The price competition is more severe at the lower end of the quality ladder and the abolition of apparel quotas is likely to have more negative effects on the profits for exporters concentrating in this range. The average prices of the exports from Kenya, Tanzania and Uganda exhibit similar patterns to other AGOA countries as seen in the next graph. Average prices of exports from Kenya fluctuated between $53 and $60 dollars during 1996-2000. One positive issue is that Kenya did not experience a decline in export prices as other countries did even tough its exports are at the low end of the price range 8. Exports from Tanzania experienced a sharp decline in price; the sharp decline in total exports is partially due to decline in prices. Export of Uganda, the new entrant to the market, managed to command prices that are also at the low end of the range. 80 Average Prices of Exports from Kenya, Tanzania and Uganda 60 Dollars $ 40 20 0 1996 1997 1998 1999 2000 2001 2002 2003 Kenya Tanzania Uganda Graph 9. Average Unit Prices of Apparel Exports by Kenya, Tanzania and Uganda 8 There can be a variety of explanations for this quality upgrading, capturing higher prices due to lower tariffs from AGOA etc. It is difficult to identify the exact reason with the existing data. 21

In conclusion, the price data is revealing that (i) (ii) (iii) (iv) (v) there is a secular decline in prices received by all exporters, developed countries receive prices much higher than all countries, mainly due to quality differences, quota-facing countries receive higher prices than preference-receiving countries but their prices have declined much more and the gap narrowed, AGOA countries are among the lowest in terms of export prices, mainly due to low quality of their exports, this includes Kenya, Tanzania and Uganda Kenya also receives low prices but did not face a similar decline, while Tanzania experienced a very large decline. 9. The Impact of MFA Quota Removal on Kenya, Tanzania and Uganda The previous sections identified certain patterns for these three countries: (i) Kenya s are exports have been growing rapidly and, as of 2003, it is the most important export category to the US; Tanzania exports, which were quite low to begin with, almost disappeared; Uganda is a new entrant (ii) Regardless of the volume of their exports, all three countries utilize substantial tariff preferences granted under AGOA, as can be seen through the average tariffs on their exports (iii) All three countries export low-price and low-quality items 22

The impact of the MFA quota phase-out on burgeoning apparel exports of Kenya, the disappearing exports of Tanzania and the recently appearing export of Uganda needs to take into account who their competitors are. More specifically, we need to determine the degree of competition that Kenya, Tanzania and Uganda face from East and South Asian countries who no longer face MFA quotas as of January 2005. 80% Market Share By Country Groups in Kenya's Export Categories Percentage 60% 40% 20% 0% 1996 1997 1998 1999 2000 2001 2002 2003 QUOTA FACING PREFERENCE RECEIVING DEVELOPED Graph 10. Market Share of Various Country Groups in Main Apparel Export categories of Kenya The graph above illustrates the share of different country groups only in the categories in which Kenya exports to the US. This is calculated by taking the average of their shares in each 10-digit HTS market category weighted by Kenya s exports in that category. The aim of using these weights is to give more significance to categories that are important for Kenya. For example, in 1996, quota-facing countries had 63% share in Kenya s export categories while preference-receiving countries and developed countries had 30% and 7% share, respectively. We should note these are similar to these country groups shares in total US imports of apparel, which were 58%, 31% and 11% respectively. However, we see a rapid reversal over time as the Kenyan exports increase rapidly. By 2003, the share of quota-facing countries have declined to 47% and the preference receiving countries and developed countries shares went up to 42% and 12% respectively. In other words, other preference-receiving countries and developed countries increased in importance as competitors in categories where Kenya exports the 23

most. The implication is that MFA quota removal might have less of an impact on Kenya today than it might have in 1996, when quota facing countries were more important competitors. The next issue is how binding the quotas are in categories in which Kenya s exports are concentrated. In an earlier section, we had noted that, in 2003, 67% of US imports entered without quotas (mostly from developed and preference receiving developing countries), 21% were imported under binding quotas (fill-rates above 80%) and 12% came under non-binding quotas. In 1997, these rates were 60%, 28% and 12%, respectively. In order to evaluate the competition for Kenya s exports, we calculate the following. For each import category in a given year, we calculate the percentage of imports that are entering the US under (i) no quotas, (ii) binding quotas and (iii) nonbinding quotas. Then we take their averages weighted by exports of Kenya in that category. Note, the percentages cited in the previous paragraph for the aggregate US imports were calculated by using the US-imports in that category as the weights. In 1997, in categories where Kenya s exports were concentrated, only 42% of US imports entered without quotas while 43% entered under binding quotas and 15% under non-binding quotas. In other words, Kenya s exports in 1997 were concentrated in categories where a large portion of US imports entered under binding quotas when compared to overall imports of the US. In 2003, the situation had somewhat changed. 61% of imports in Kenya s export categories entered without quotas while 28% entered under binding quotas. These imply that the threat of quota removal decreased considerably for Kenya in the preceding 7 years. However, Kenya s exports were still more concentrated in categories where larger percent of US imports entered under binding quotas compared with other countries. Although the threat of quota removal subsided considerably for Kenya, it is higher than it is the case for an average country. 24

Market Share by Quota Level in Kenya's Export Categories 80% 60% Percentage 40% 20% 0% 1997 1998 1999 2000 2001 2002 2003 No Quotas Binding Quotas Non-Binding Quotas Graph 11. Market Share by Quota Level in Main Apparel Export categories of Kenya We constructed the same graphs for Tanzania and Uganda as well. Before proceeding, though, we need to note that the low level exports from these countries makes analysis more difficult. Tanzania follows the same pattern as Kenya and all three countries have very similar levels in the last two years. Tanzania s and Uganda s exports were concentrated in categories where higher portion of the US imports were entering under binding quotas. This level declined the last years but is still above the US average. Market Share by Quota Level- Tanzania Uganda 80% 80% Percentage 60% 40% 20% 0% 1997 1998 1999 2000 2001 2002 2003 60% 40% 20% 0% 2002 2003 No Quotas Non-Binding Quotas Binding Quotas Graph 12. Market Share by Quota Level for Tanzania and Uganda 25

This section s conclusions can be summarized as follows: (i) Kenya s exports are concentrated in categories where other preferencereceiving countries have higher market shares (as opposed to quota-facing countries) when compared to total US imports (ii) On the other hand, although a smaller portion of Kenya s exports compete with quota-facing countries, these exports tend to be more in categories with binding quotas. Although this concentration declined over time, it is still above the US average. (iii) Tanzania and Uganda follow similar patterns. The implication of these results is that the severity of the impact of the MFA quota removal is going to depend on several factors. The most important factor will be the supply response of quota-facing countries when these restrictions are removed. Even though quota facing countries in South and East Asia do not have a large share in Kenya s main export categories, more of their existing exports are entering under binding quotas. If their export supply response is relatively muted after quota removal, which is unlikely given the response seen so far for categories where quotas were eliminated, the impact on Kenya might not be as large. On the other hand, if the supply response is strong, then Kenyan exports might suffer more than other exporters. Furthermore, Kenyan exports are concentrated in low-price low-quality items where price competition is more intense and profit margins are much lower. and this makes the problem more acute. 10. Product Specific Analysis In this section, we provide product specific analysis with the aim of providing further insights on the effects of MFA quota removal. The focus is on Kenya only since Tanzania and Uganda do not export large volume of apparel and it is difficult to perform a thorough product-level analysis. Our first observation about Kenya is that its apparel 26

exports are highly concentrated even tough there has been diversification over the years. The table below lists the major export categories for Kenya at the HTS 6-digit level for the years 1996, 2000 and 2003. D E T T I N K D E T T I N K T O N the US HTS code Description 1996 2001 2002 2003 COTTON ITEMS 610462 WOMEN'S OR GIRLS' TROUSERS 1 19 2,393 3,449 610510 MEN'S OR BOYS' SHIRTS 192 2 150 2,497 610610 WOMEN'S BLOUSES, SHIRTS 0 1 153 2,669 610910 T-SHIRTS, SINGLETS 329 37 687 2,573 611020 SWEATERS, PULLOVERS 118 0 6,390 15,855 SYNTHETIC FIBER ITEMS 610230 WOMEN'S OVERCOATS, ANORAKS 0 0 1,654 2,621 610343 MEN'S OR BOYS' TROUSERS 0 16 1,758 2,070 610443 WOMEN'S OR GIRLS' DRESSES 485 0 227 144 610463 WOMEN'S OR GIRLS' TROUSERS 0 0 849 2,653 611030 SWEATERS, PULLOVERS 0 55 5,676 8,749 COTTON ITEMS 620342 MEN'S OR BOYS' TROUSERS 8,677 20,129 29,668 37,147 620462 WOMEN'S OR GIRLS' TROUSERS 267 36,681 51,274 80,512 620520 MEN'S OR BOYS' SHIRTS 11,921 2,437 3,704 4,933 620630 WOMEN'S BLOUSES, SHIRTS 383 100 534 796 620442 WOMEN'S OR GIRLS' DRESSES 952 516 1,036 549 620452 WOMEN'S OR GIRLS' SKIRTS 101 66 818 875 620192 MEN'S ANORAKS, WINDBREAKERS 999 0 12 12 SYNTHETIC FIBER ITEMS 620343 MEN'S OR BOYS' TROUSERS 0 461 1,726 2,394 620463 WOMEN'S OR GIRLS' TROUSERS 0 260 6,129 2,431 620530 MEN'S OR BOYS' SHIRTS 199 1,036 187 128 620640 WOMEN'S BLOUSES, SHIRTS 1 325 1,950 1,994 620444 WOMEN'S OR GIRLS' DRESSES 426 0 3 175 TOTAL 25,051 62,141 116,978 175,226 Table 1. Export Values ( 000$) in Select Apparel Categories from Kenya to 27

In 1996, almost all of the exports were in cotton not-knitted categories and the bulk of these items were men s trousers and shirts. Kenya exported almost no knitted items or not-knitted synthetic-fiber apparel. This pattern continued until 2001 when AGAO was fully implemented. Significant diversification occurred during 2002-3. The share of cotton not-knitted apparel decreased to 71% in 2003 despite the fact that their total volume increased to $124 million from $23 million in 1996. Exports of knitted items, especially pullovers etc, increased rapidly during these two years and reached 25% of total exports. No-knitted apparel from synthetic fiber also increased rapidly and reached $7 million in 2003, tough this is below the 2002 level of $10 million. Initial concentration and relative diversification of Kenyan apparel exports is more apparent if we were to look at the composition at the 10-digit level. Although there are over 1300 apparel categories at the 10-digit level in the HTS classification system, Kenya had exports exceeding $100 thousand only in 30 categories in 1996. Furthermore, top 20 categories accounted for 95% of its exports. Exports were significantly diversified by 2003, but they were still concentrated compared to other exporters. In 2003, top 20 categories represented 80% of exports and over 90 categories had exports over $100 thousand. In short, along with rapid increase of apparel exports during 2002-3, we also observe entry into different product categories by Kenyan exporters. More importantly, products in these categories tend be of higher price and quality. This has been one of the main effects of AGOA. In this section, we analyze what the impact of MFA quota removal would be in some of the main export categories of Kenya. We picked seven representative 6-digit categories from the table above. Two of these categories are knitted items (cotton and synthetic-fiber sweaters) where Kenya is a new entrant. Three of them are not-knitted cotton items (men s shirts, men s and women s trousers) which are historically the main export categories. The final two are not-knitted synthetic-fiber categories (men s and women s trousers) which are relatively small but have future potential. In the graph 28

below, we show the market share of quota-facing countries in these categories in the US over time. 75% Market Share of Quota-facing countries (%) 60% 45% 30% 1997 1998 1999 2000 2001 2002 2003 611020 - cotton sweaters 611030 - syn. Sweaters 620342 - men's cotton trousers 620343 - men's synt. trousers 620462 - women's cotton trousers 620463 - women's synt. trousers 620520 - men's cotton shirts Graph 13. Market Share of Quota Facing Countries in Select Apparel Categories of Kenya Quota-facing countries have the highest shares in men s cotton shirts and the lowest share in men s cotton trousers. The former was the largest category for Kenya but declined in importance over time. The latter is the second largest export category for Kenya. It is hard to show causality but the decline in Kenyan share in men s cotton shirts and increase in men s cotton trousers might be due to opposite evolution of exports of quota-facing countries. Women s cotton trousers is the largest export category for Kenya (close to half of total exports) and quota-facing countries maintained a steady market share of slightly below 50%. Cotton and synthetic-fiber sweaters are new categories in which Kenyan exporters have been quite successful recently. In both categories we see that the market share of quota countries have been in decline which partially explains the relative success of Kenya. Thus, Kenyan exporters are likely to be successful in these 29