Countries that want to use preferential trade agreements to boost trade with Africa should re-examine the rules of engagement. New evidence shows that certain rules underlying preferential trade agreements are drastically hindering their intended benefits. In fact, in a World Bank Policy Research Paper  and an article forthcoming in The World Bank Economic Review, we find that relaxing those definitions could increase the agreements’ benefits by four times more than just removing tariffs.
Economists generally agree that it is a good idea for wealthier importers – such as the US and EU – to exempt their poorest trading partners from tariffs through preferential trade agreements. In theory, this gives low-income countries a leg up by helping them sell more products overseas. In the modern trade environment, these agreements can help low-income countries join global supply chains, one of the few ways for low-income countries to industrialize in today’s competitive markets. This stepping stone is particularly important now, because wages are growing fast in Asia, giving lower-wage African countries another chance at developing and diversifying exports of manufactures and services. Two examples of agreements that strive to boost African exports are the US’ Africa Growth Opportunity Act (AGOA), a preferential agreement that provides low-income African producers preferential access to the US market, and the Cotonou preferences that apply to certain EU imports.
The problem lies with defining how this preferential treatment should work. Preferential agreements use so-called “rules of origin” to determine which products qualify for duty-free entry into the member countries’ markets. They are judged on criteria of being “sufficiently transformed” or “originating” in the beneficiary country. For instance, a sweater exported by Madagascar must have been made with a certain amount of local yarn or labor to qualify for duty-free access to the EU. The idea behind this requirement is logical. It prevents what is called “trade-deflection,” or the opportunity for a low-income country’s trading partners to get duty-free access to the US or EU markets through a “back-door” route.
But the devil is, in fact, in the details. As it turns out, the precise form for these origin requirements can have quite a large impact on trade. Specifically, we found that a seemingly minor relaxation of requirements under the AGOA regime led to a significant change in apparel exports from African countries: exports to the US took off, while exports to the EU did not.
This natural experiment unfolded as follows. During the period of our study, apparel products from qualifying countries needed to meet any of the following requirements to gain duty-free access to the EU market: They had to have (1) been made from yarn originating in an African country or in the EU; or (2) undergone a “double-transformation” process, which meant the product had to have transformed from yarn to textile to apparel (yarn→textile→apparel) within the beneficiary country.
Beginning in 2000, under AGOA, apparel from qualifying countries gained duty-free access to the US market. They had to (1) have been assembled in one or more AGOA-eligible countries, and (2) contain US fabrics made from US yarn (or African fabrics, to a certain percentage), the latter being called a “triple-transformation” process (cotton→yarn→textile→apparel).
As the graph below illustrates, in the first part of the study period, 1994-2000, exports from low-income African countries to both the US and Europe followed similar growth patterns. In 2000, upon the implementation of AGOA, exports to the US began to grow at a faster pace.
The key moment came in 2001, when the US relaxed the triple-transformation by enacting a “Special Rule” that conferred duty-free access to African apparel regardless of the origin of fabric (cotton, yarn, textiles) used to produce it. In effect, this change meant that meeting origin requirement under AGOA only required a single-transformation process (fabric → apparel). From this point on, African apparel exports to the US grew quickly. Exports to the EU, however, mainly stayed flat until they dropped in 2001 due to a political crisis that hit Madagascar, the largest exporter to the EU at the time. We estimated that this “Special Rule” increased African exports to the US by 168 percent, or four times as much as the export growth that can be explained by the reduction in tariffs. Moreover, we also found that much of the growth took place at the extensive margin – that is, countries developed new types of products, not just higher volume of the same goods.
Source: de Melo and Portugal-Perez (2013) 
The Special Rule, however, did not have a uniform effect among beneficiary countries; some performed better than others. This result begs the question: why were some African countries so much more successful at taking up preferences and at experiencing higher export growth in apparel? A possible explanation lies in countries’ business environments. Some country settings may be more conducive to attracting foreign investment in apparel plants and to diminishing fixed costs. Figure 2 shows the Doing Business (DB) ranking of African exporters benefiting from the Special Rule and their apparel export growth during AGOA. Indeed, on average, countries that ranked best in the DB indicator experienced higher growth in apparel exports during AGOA.
Source: de Melo and Portugal-Perez (2013) 
These results suggest that high-income importers, such as the US and EU, would best serve their poorer trading partners by relaxing the stringency of rules-of-origin requirements. More importantly, in the near future, African countries will be looking to export to fast-growing, middle-income countries such as Brazil and China that are starting to grant preferential access to less-developed countries. These countries should resist adopting the costly rules of origin devised by the EU and the US and go for simpler ones instead. For instance, countries could eliminate proof of origin for preferences below a threshold (say 3%-5) and/or adopt a single requirement, such as a minimum domestic value content, which could be set at a lower threshold for less-developed countries. Or they could take inspiration from ASEAN , where a single rule requiring that 40% of the final value of the product originate among members applies across the board.