AGOA: THE UNINTENDED CONSEQUENCES OF LIBERAL RULES OF ORIGIN

Transcription

1 AGOA: THE UNINTENDED CONSEQUENCES OF LIBERAL RULES OF ORIGIN Lawrence Edwards (University of Cape Town) Robert Lawrence (Peterson Institute of International Economics, Harvard Kennedy School & NBER)

2 Main Points: Rules of origin matter, especially for the poorest countries. For the most part, they are unnecessary there are other ways of ensuring that preferences are not abused. Liberalizing rules of origin can also have unintended consequences and as a result, while liberal rules may generate trade, they do not necessarily lead to development.

3 Rules of Origin and GVC s Entering Global Value Chains: A Challenge for the Poorest Developing Countries. Even When Given Preferences, Current Rules of Origin are Often Binding Constraints: If the costs in a T-Shirt are 70% fabric, 30% sewing, a 35% percent value-added rule can prevent a least developed country from participating in a GVC.

4 AGOA: The Exception that Proves the Rule. Africa Growth Opportunities Act: S-Saharan countries were given duty-free access to the United States market. The least developed, and only the least developed AGAO countries were given a rule of origin allowing the use of non-originating fabric. All they needed for eligibility was a passport. i.e.us inspection that there were genuine production acitivities

5 AGOA Rules of Origin.

6 /2009 billion M2 Result: Rapid surge in US imports from 2000, almost entirely under LDC Provision. Though some declines after expiration of MFA Apparel imports from AGOA countries according to Import Programme (bill SME) Other Commercial availability based Cashmere and merino wool sweaters Lesser developed country provision From regional fabric from U.S. or African yarn Apparel made of U.S. yarns and fabrics Non-preference imports

7 $ thousands LDC Countries: Major AGOA Beneficiaries US imports of clothing and textiles by AGOA countries under all import programmes Lesotho Madagascar Kenya Swaziland Mauritius South Africa Botswana Malawi Ethiopia Tanzania Ghana Cameroon Note contrast between LDCs and South Africa and Mauritius

8 AGOA at Ten: Virtues Trumpeted in 2010 Ron Kirk USTR: a substantial increase in two-way U.S.-Africa trade since 2000, with African countries now exporting to the United States a more diverse range of value-added products, powerfully demonstrates the link between trade and economic development.

9 Many Studies Confirm: AGOA Worked Collier & Venables (2007) AGOA apparel provision a positive and significant effect. Frazer and Van Biesebroeck (2007) Overall AGOA apparel exports increased by 53 percent with stronger impacts on products with high initial levels of protection. Portugal-Perez (2008) impact of 96 percent for 22 countries eligible for the third-country fabric provision, and 303 percent for the top 7 beneficiaries (Olarreaga and Ozden 2005). AGOA exporters captured some of the tariff rents created by the preferences. No evidence of differential effects in taking advantage of AGOA based on measures of corruption or institutional quality (Frazer and Van Biesebroeck 2007).

10 Poster-Child Lesotho - Remarkable Growth in Exports from 2001 Least developed country, mountainous, landlocked, impoverished. Major increases in apparel exports to the US - up from $110 million in 2000 to $467 million in 2004 Largest exporter of clothing from S- S Africa. Over 80 % of manufacturing jobs Employment (mainly women) in apparel industry increases from to from 2000 to 2004

11 But the Glass is also Half Empty. Limited horizontal diversification Few products are exported (Lesotho: 60 to 118 to 84). Just one buyer -- the US retailer the GAP -- accounts for almost 40 percent of overall output. Limited Productivity Growth. Limited Skills Acquisition. Limited Spillovers to other exports. AGOA LDC countries have not enjoyed similar growth in clothing exports to other countries, especially EU or South Africa Limited vertical diversification Mainly cut-make and trim (CMT) products with little diversification into higher value-added clothing Few backward linkages into the rest of the economy & very few locals in management positions. Some organizational learning in trade networks, but still only control assembly slice of supply chain. Concerns Each Time Rule Expires:

12 No Real Diversification of clothing Exports

13 Low Priced Products! 25 th Percentile of US Imports.

14 EXPORT GROWTH: Almost all Intensive Margin Lesotho: 92% Intensive Margin! Mauritius and South Africa: No Growth.

15 Our Explanation: We argue that these results are the predictable consequences of the incentives created by AGOA preferences and the MFA. Encouraged LDCs to specialize in products with little value addition and a high fabric content. Made exporters especially vulnerable when MFA ended. Induced exporters to use relatively expensive fabrics that discourage backward linkages. Provided little benefit to more developed countries like South Africa and Mauritius. Rules of Origin Matter: Can help but also hinder. Trade does not equal development!

16 Early Evidence. European Performance: Rules of Origin matter

17 US Clothing Protection: Three phases. 1990s. Tariffs and MFA Quotas AGOA Duty-free plus Rule of Origin. 2005: Expiration of MFA. Each, in theory will influence the composition and the volume of imports

18 In China s Shadow

19 (1) MFN Tariffs. Do Not distort quality or sourcing decisions. Apply to all producers equally

20 But in addition, quotas may increase attractiveness of fabric intensive products. The fixed fee is higher percentage of low value added products. Unconstrained beneficiaries will specialize in low quality and fabric intensive products. (2) Quotas. Operate like a specific tariff a fixed fee per unit exported. Will induce constrained countries to upgrade to higher quality products. So for example China will move into higher quality and higher unit value products. (Ship the good apples out). (The fee is a smaller percentage of higher prices, will reduce relative prices of higher quality products) So unconstrained countries will move into the lower quality products (both value-added and fabric)

21 AGOA for LDCS: No Tariffs, Third-country fabrics. (a) Tariff removal makes previous production more attractive expand the intensive margin. (b) Fabric benefit: encourages low value-added more fabric intensive products. the intensive margin. So use expensive fabrics but add low value.

22 AGOA with Rules of Origin.(South Africa and Mauritius) No tariffs but non-ldcs must use either US or local fabrics. In principle depends on fabric tariff rates relative to the overall tariff preference. If these are high, fabric intensive products would not benefit much from AGOA. If low, could still be some benefit,

23 Example: Preferences promoted lower labor-added quality and fabric intensive products. Most Fabric-Intensive Medium Fabric-Intensive l=0.8 l=0.6 Quality Parameter China Fabric Labor cost Lesotho RULC: (wa/w*a*) Fabric Labor cost Lesotho advantage (%) -23% -23% -23% -38% -38% -38% Scenario 1: Tariff USA MFN Tariff China US Price Lesotho US price Lesotho advantage (%) -23% -23% -23% -38% -38% -38% Scenario 2: MFA MFA Shadow Price China US price Lesotho US price Lesotho Advantage % 6% 1% -2% -14% -18% -21% Scenario 3: AGOA Preferences China US price Lesotho Cost Lesotho Advantage % 27% 21% 17% 3% -2% -5% Scenario 4: MFA removed China US Price Lesotho Cost Lesotho Advantage % -8% -8% -8% -25% -25% -25%

24 Implications of Theory Apparel quotas under the MFA induced non-binding quota restricted countries to export low value added apparel products based on the intensive use of fabric. AGOA preferences raised the value and range of apparel products exported by beneficiary countries with relatively strong growth in imports of products facing high tariff preference margins and from countries eligible for the third-country fabric provision. Finally, the end of the MFA induced quota restricted countries to shift towards lower quality fabric intensive products and varieties within each product category. The removal of quotas was therefore particularly detrimental to all AGOA countries as it encouraged entry by previously quota restricted countries into those products in which AGOA recipients were specialized. The AGOA LDC countries were protected by their fabric preferences. Other AGOAS countries were not.

25 Main Results. Do find that AGOA worked to generate more exports Confirm that AGOA-LDC Countries produce low valueadded, fabric intensive products. However, do not find that as a result of AGOA these countries shifted into more fabric-intensive products. Do find an impact of ending of MFA on China and others shifting towards low unit value products. Also a rise in fabric intensity and increased use of cheaper fabrics.

26 Empirical Analysis: Approach 1 US Import responses to preferences: Triple difference-indifferences approach ln IMP cpt D2001* Dldc 1 D2001* DAg 2 c * DAg c cntry / prod Data: cp prod / year pt cpt HTS 10 digit data from Foreign Trade Division of the U.S. Bureau of the Census, Approx 1500 time consistent HTS codes 41 AGOA countries (27 Apparel eligible), 213 countries in total Merge with 21 NAICS 6-digit VA/sales for USA c

27 AGOA Preferences: Strong export response in value and product range, but gains eroded by end of MFA Dependent variable ln IMP ln IMP Import dummy Sample All Positive value product lines All Years Marginal impact LDC preference relative to Other AGOA 13.9% 16.8% 351.3% 281.5% 0.9% 1.2% Marginal impact Other AGOA relative to non-agoa -9.2% -8.4% -11.8% 17.6% -1.1% -0.9% Regression coefficients DApp_start ct *DApp c *DAg c 13.0% 15.5% 150.7% 133.9% 0.9% 1.2% (34.44) (26.20) (29.62) (19.66) (21.80) (20.18) DAg_start ct *DAg c -9.7% -8.8% -12.5% 16.2% -1.1% -0.9% (31.66) (23.61) (2.98) (3.46) (33.62) (21.56) N Dummy variables product/year country/product product/year country/product product/year country/product Notes: Robust t-statistics presented in parentheses. Estimates are robust to heteroskedasticity.

28 MFA and Price Analysis: Expectations MFA : AGOA countries specialise in low priced, low quality, fabric-intensive apparel varieties AGOA, post 2004: Exacerbates incentive for LDC AGOA to produce fabric-intensive Post MFA: Entry by China and other Quota constrained countries into those products (across product shift) and varieties (within-product) exported by AGOA countries

29 Empirical Analysis: Approach 2 Price based analysis: Estimate fabric-intensity using price equations Testing AGOA on fabric-intensity t it t it it ct ct st it tariff q e pva pf p l ) ln(1 ln ln ln ln ln ict t ct hs it naics t ct ct ct c ct b ct ft c ft b ft ic ict tariff Pcompete usppi e GDPwork pva Dldc D pva Dldc pva pf Dldc D pf Dldc pf p l ) ln(1 ln ln ln ln ) ln 01 ( ) ln ( ln ) ln 01 ( ) ln ( ln ln 4, 7 6,

30 Use Fabric Prices to Estimate Fabric Intensity

31 MFA: Within-Product estimates of fabric-intensity Table 5: Regression to determine relative fabric-intensity of AGOA apparel exports under the MFA. AGOA relative to Marginal impact All countries AGOA relative to quota constrained quota constrained, full period Period Column (1) (2) (3) ln(gdp/capita), PPP 0.126*** 0.152*** *** ln(pf) 0.272*** 0.214*** 0.185*** DAGOA x ln(pf) 0.383*** 0.293*** Dquotacntry x ln(pf) 0.141*** 0.246*** ln(pva) 0.237*** 0.284*** 0.441*** DAGOA x ln(pva) *** *** Dquotacntry x ln(pva) *** *** ln(e) *** *** *** ln(pcompete) 0.037*** 0.038*** 0.045*** ln(us ppi) 0.135* *** ln(1+t) *** *** *** N Fixed effects country/product year country/product year country/product year

32 US$ mill MFA IMPACT: Almost all AGOA apparel exports to US in product lines where Chinese Quotas are binding Lesotho apparel exports to US according to binding and non-binding 2003 Chinese quota categories Unbound Bound, fill rate >=90%

33 Large Within (and Between) Product Price Variation in Apparel Figure7: Unit values and level of development: Top Apparel product exported by Lesotho to US in 2004 (Women s or girls other pullovers of cotton, knitted). HS : Women's or girls' cotton pullovers, 2004 Madagascar India Bangladesh Cambodia Ethiopia Malawi Lesotho Kenya Tanzania China Indonesia Thailand Sri Lanka Mauritius Namibia Swaziland Botswana South Africa Hong Kong Macao Korea, South UAE Nigeria ln(gdp/capita PPP)

34 Dramatic Decline of Chinese (and Indian) prices in those products Lesotho (and AGOA) export (withinproduct) Import weighted average price relative to Lesotho (using Lesotho exports as weights) Bangladesh China India AGOA less Lesotho

35 Combined With Structural Shifts by China to Low Price Products (Across product shifts) Structural shifts in the composition of imports, import weighted US average unit value ($) per SME 4.0 China India Bangladesh Lesotho LDC AGOA excl. Lesotho

36 Lessons Trade preferences do work: incentives matter. Rules of origin matter: Stringent rules of origin requirement preclude developing countries from participating Trade preferences are not a panacea or a substitute for more comprehensive development strategy. Preferences (waiver on domestic fabric requirement) discouraged additional value-addition in assembly and stimulated use of expensive fabrics that were unlikely to be produced locally. Specialization exposed countries to intense competition from quota constrained countries when MFA was removed