This paper addresses aspects of the Uruguay Round of particular concern to Africa. It covers four areas. First is preshipment inspection. This is an aspect of the Round about which little has been said and yet is one of some significance to African negotiators. Second is the erosion of preferences: the fear that, as most-favored-nation tariffs came down, the degree of preference that Africa received in markets of the Organization for Economic Cooperation and Development (OECD) would be reduced. This was perhaps the major concern expressed by African commentators. Third is the nontariff barriers affecting African exports and the abolition of the Multifiber Arrangement from an African viewpoint. Fourth is the effects of the liberalization of agriculture on food prices.

This paper addresses aspects of the Uruguay Round of particular concern to Africa. It covers four areas. First is preshipment inspection. This is an aspect of the Round about which little has been said and yet is one of some significance to African negotiators. Second is the erosion of preferences: the fear that, as most-favored-nation tariffs came down, the degree of preference that Africa received in markets of the Organization for Economic Cooperation and Development (OECD) would be reduced. This was perhaps the major concern expressed by African commentators. Third is the nontariff barriers affecting African exports and the abolition of the Multifiber Arrangement from an African viewpoint. Fourth is the effects of the liberalization of agriculture on food prices.

Preshipment Inspection

Preshipment inspection (PSI) is the practice of asking private companies to inspect OECD goods that are destined for certain African markets before they are dispatched.1 It is, perhaps, like a private sector equivalent to customs activities, designed to ensure the quality of the goods and that the revenue collected from tariffs approximates what should be collected. This practice has raised a number of complaints and concerns among industrial country exporters (see Table 1): for example, at a practical level, it imposed extra costs on them; it imposed delays on exports; it lacked natural justice in not having any appeal; it ran the risk of companies having to disclose confidential information; it lacked transparency, accountability, and agreed standards; it was applied in a discriminatory fashion, and there were weaknesses in the ways PSI companies verified prices. There were also more philosophical points: it interfered with the freedom of contract between buyers and sellers, and it infringed on the sovereignty of nations. The last two are essentially unfounded, for much commercial law has those effects. But the first set of points were serious and legitimate concerns for the industrial countries.

Table 1.

Exporter Complaints Against Preshipment Inspection

article image
Source: P. Low, Preshipment Inspection Services, World Bank Discussion Paper No. 278 (Wahington: World Bank, 1995).Note: ECE = Economic Commission for Europe; ICC = International Chamber of Commerce; PSI = preshipment inspection; and USITC = United States International Trade Commission.

At the outset of the Round there was some uncertainty about the legitimacy of PSI activities. Switzerland and Germany had acted against the preshipment inspection companies, and other industrial countries had wondered whether they should also do so. The issue was put on the Uruguay Round agenda by Indonesia, but thereafter became an issue quite closely associated with the African states. Thus the very fact that we have an agreement on preshipment inspection, and an agreement that I shall argue meets many of these worries that the industrial countries have, can be counted in some sense as a victory for Africa, for it quite clearly legitimizes the practice, so long as it is conducted subject to various controls and regulations.

In return for this legitimacy, the user countries, that is, the importing (African) countries, have to undertake a number of obligations about the way in which their preshipment inspection companies will behave (Table 2). First, preshipment inspection should be applied in a nondiscriminatory fashion. That does not mean that every consignment has to be treated identically but that the rules that determine how a consignment is treated are objective and are uniformly applied. Second, government requirements on PSI must satisfy national treatment so far as laws, regulations, and requirements are concerned. That is not exactly the same as national treatment under the GATT, but it is something approximating it. Third, the PSI agreement states that goods should generally be inspected in the country of export rather than the country of manufacture, and, fourth, that inspection should be designed to discover whether the goods meet standards that have been agreed between the seller and the buyer. A fifth requirement is an obligation to have transparency in PSI practices; that is, to ensure that the laws, regulations, inspection procedures, and so on, are all published. Sixth, there is provision for the protection of confidential information. This means that PSI companies are committed not to ask for information that is not pertinent to their contracted duties, and that they are responsible for the maintenance of the confidentiality of the information they are given. Seventh, there are procedures for handling conflicts of interest within the PSI companies. Eighth, to counter delays, there are agreed inspection dates, and a commitment that the report by the preshipment inspection company will be made within five days of its receiving the final paperwork. For exporters, a critical part of the agreement concerns methods of price verification. The agreement states that the price comparison should be for identical or similar goods offered for export from the same country of exportation at or about the same time, under competitive and comparable conditions of sale, in conformity with customary commercial practices. In other words, price verification is very narrow. The PSI companies cannot be asked to find the lowest knockdown price that some other country is selling to some other partner. The comparison has to be pretty much specific to the contract that is being reviewed in terms of timing and size, for example. Finally, there is now an appeals procedure and a brief addendum about a derogation. Countries can set consignment size thresholds below which there will not be inspections unless there are suspicious circumstances.

Table 2.

User Country Obligations Under the WTO Agreement on Preshipment Inspection

article image
Source: P. Low, Preshipment Inspection Services, World Bank Discussion Paper No. 278 (Washington: World Bank, 1995).

Does the agreement match the aspirations of the proponents and critics of PSI? Fairly well. It is true that delays will probably still continue somewhat and that there will still be costs to permitting the preshipment inspection. But according to work done by the U.S. International Trade Commission (USITC) these costs are not generally very large. On the positive side, the appeals procedure, the protection of confidential information, the transparency, and the agreement over standards are all steps forward. In addition, discrimination is now explicitly forbidden and price validation must be “relevant” as well as honest. So, all told, the agreement largely meets the industrial countries’ concerns. On the other side, the developing countries now have recognized rights to undertake preshipment inspection that was, for them, an important goal in the negotiations.

Preference Erosion

Figure 1 demonstrates one important point about preferences. When one is talking about trade policy, at least in some rather loose sense, trade creation (imports replacing costly domestic outputs) is a good thing and trade diversion (preferred imports displacing cheap imports from nonpreferred sources) is a bad thing. Tariff preferences divert trade: they shift trade from the lowest-cost supplier to a less low-cost supplier. When preferences are eroded, we are essentially undoing that trade diversion. Trade moves back toward the sort of pattern that it would have had under free trade. This is an important perspective on the erosion of preferences. Preferences are, on the whole, distortionary. They may benefit the recipients of the preferences, but in terms of world welfare, they are not smart policy.

Figure 1.

Figure 1 applies to the market for a particular good and has prices on the vertical axis and quantities on the horizontal axis. I represents the demand for imports by the curve (Dd - Sd) domestic demand minus domestic supply. That is, imports are what you need to buy from the rest of the world when your domestic suppliers have satisfied the part of domestic demand that they are willing to meet at the ruling price. I assumes that the rest of the world comprises two parts. First, a very large part that can supply any amount of the good, at least so far as this market is concerned, at a fixed price, Pw, the world price. Second, there is a small preference-receiving country whose supply curve to the market is Sp. It slopes up to show that higher quantities can be obtained only in return for higher costs (higher prices). The case I have in mind is of a good that the EU is buying from the rest of the world, but on which it gives preferences, say, to the small ACP countries.

Imagine that we start off with an initial tariff that everybody pays. I have called tT the Tokyo Round tariff before preferences were granted. The world price is Pw, and so the internal price in our market is (Pw + tT). At this price, the demand for imports is MT. But how much of this is imported from the ACP countries? That depends on the price they receive when they sell. This is Pw, the world price, because in order to sell in the internal market at (Pw + tT) they get Pw and the government gets tT. The ACP supply curve tells us how much they will supply at Pw; and I have called that amount Amfn. The remaining imports (MT - Amfn) come from the rest of the world.

Now let us give the ACP preferences, so that they no longer have to pay the tariff. When the good sells at Pw + tT internally, all of that amount passes back to the ACP exporter: the government does not take a cut anymore. Now that the ACP countries face price (Pw + tT) they are happy to produce some more, and so they expand their sales to AT. The total amount of imports and the internal price are exactly the same, as before: the latter is still (Pw + tT) because, on the margin, supplies still come from the rest of the world (ACP cannot meet the whole demand for imports) and if the price is the same, so is demand. All that has happened is that we have moved from MT - Amfn being supplied by the rest of the world, and Amfn being supplied by the ACP countries, to a smaller supply from the rest of the world (MT - AT) and a larger supply by ACP (AT). That is trade diversion. It is wasting real resources so far as the world is concerned, because the rest of the world can produce this good at cost Pw but it costs the ACP countries more than that to produce units beyond Amfn. Trade of (AT - Amfn) is diverted and the lost welfare is area IJK—the sums of the amounts by which ACP costs exceed world costs for the units between Amfn and AT- NOW let us have the Uruguay Round and reduce the MFN tariff to tu. The internal price falls to (Pw + tu) and the demand for imports expands. But the fall in internal price also reduces the price that the ACP suppliers receive. Hence they cut back, moving sales from AT to Au. They get a lower price from the EU market because there is less protection and they were previously beneficiaries of EU protection. This change in exports (AT - Au) is the most obvious consequence of preference erosion. But it is not a cost to the world—just the opposite. Instead of paying extra to get the ACP to supply Aj of imports, we now pay extra only to get them to supply Au. The trade diversion implied by preferences has gone down.

In addition to the change in the quantity of exports that the ACP countries send to Europe, they also lose some export revenue. The price they receive within the EU has fallen, so instead of earning (Pw + tT) on their exports, they earn only (Pw + tu). This is essentially a cut in the income transfer they receive from EU preferences. Preferences essentially allow the preferred exporters to collect the tariff revenue on their exports: when tariffs fall, so do the revenues they collect.

It is important to keep in mind these two effects. Preferences are usually justified as an incentive for industrialization: that is, the benefits are held to be a result of the quantity changes that preferences induce. They are not usually justified just in terms of a straight income transfer, for an income transfer to the producers of a particular set of goods is often held just to encourage them to be inefficient. Thus it is usually declining exports rather than declining prices that people are concerned about when they worry about preference erosion.2

I am not keen on preferences as a means of aiding African countries’ development. Nonetheless, should I not be concerned that they might lose some exports as a result of the Uruguay Round? In fact, not really, for in the end it turns out that, at least so far as the least developed countries in Africa are concerned, there are virtually no losses. The fact is that preferences never were very much use for increasing exports from most preferred countries. They are subject to very tight rules of origin. There are considerable exclusions, the United States excludes textiles and clothing, for instance. There are limits on the amount that may be sold duty free. Preferences are unilateral concessions that can be withdrawn by their donors at any time. It turns out that only relatively small proportions of the trade between the OECD countries and the preference-receiving countries actually qualified for preferential treatment. Figures from the United Nations Conference on Trade and Development (UNCTAD) indicate that in 1992 only 18 percent of exports from GSP-receiving countries to OECD countries qualified for GSP treatment. Probably an even smaller percentage of them actually used those preferences, because it required extra paperwork to make use of them.

All these factors discourage developing country entrepreneurs from investing to exploit their markets access. The returns are frequently constrained and almost always too risky. Thus, by and large, preferences have failed to foster industrialization, and, therefore, if they are lost, that is not such a disaster.

It is also the case that the margins of preference are generally very small. For example, within the EU, something like half the imports from the ACP countries already face a most-favored-nation tariff of zero. In other words, there is no preference to be had because the MFN tariff is zero. The WTO Secretariat suggests that of sub-Saharan Africa’s exports to the EU, 56 percent have MFN tariffs of less than 3 percent. And, if you have to fill out a lot of paperwork and bear a lot of uncertainty just to get a preference of 3 percent, then it is quite unlikely that you will bother.

Figure 2 presents the average preference margins for all exports to the EU for the set of least-developed African countries. It is measured in a slightly unusual way: for example, for three African countries, the average tariff paid by exporters is between 2 and 1¾ percentage points lower than the average tariff paid by other exporters, for one country, between 2 and 2¼ percentage points lower, and so on. Most countries get average preference margins of less than 3 percent within the EU, and these averages are much lower in Japan and the United States. Four percent is the largest average preference margin that we came across.

Figure 2.
Figure 2.

Preference Margins for African Exports in EU Markets, by Exporting Country

Moreover, when we come to the Uruguay Round, not only are we talking about small tariff preference margins, but many of the preferences were offered on agricultural goods for which preferential quantitative restrictions are more common. The Uruguay Round protected most of those rights of preferential access. Indeed, the tariffication whereby nontariff barriers were converted into tariffs has actually made duty-free access more valuable in some parts of agriculture, because “duty-free” did not mean “free of all variable levies.” Now that these have been converted into tariffs, which are then remitted on preferred exports, the value of preferences is increased. The average reduction in the preference margin as a result of the Uruguay Round was 23 percent in the EU, 9 percent in the United States, and 15 percent in Japan. This means that even for the two most preferred countries, with average margins exceeding 4 percent, the new average margin is about 3 percent—a loss of competitiveness of about 1 percentage point.

To summarize, preferences never did much in the first place. Lots of them were not affected by the Uruguay Round; those that were, suffered very small changes. The largest average change you might find in the rents to exporters is losses of ½ of 1 percent of the export revenue.

When it comes to the quantities, the story is, in fact, even less dramatic. Preliminary work in the World Bank has analyzed tariff change item by item for least-developed countries. This is more straightforward than for developing countries as a whole because least-developed countries are generally granted totally duty-free access to OECD markets and for any amount that they can sell. If we look up the changes in MFN tariff rates and the volumes of trade occurring in a base year, and then make some assumptions about elasticities, we can estimate the effects of the Uruguay Round on trade.

We find that, as a group, the least-developed countries will lose something like $6 million worth of exports to the EU, and $3 million worth of exports to Japan, but gain $52 million worth of exports to the United States. The reason for the last figure is that developing countries do not receive preferences on textiles and clothing exports to the United States, and the liberalization of the Multifiber Arrangement plus the reductions in the United States’ MFN tariffs will allow them to compete more effectively, not against third-country suppliers, but against U.S. producers.

For the least developed countries in Africa, we estimate approximately a $4 million gain in the United States, a $5 million loss in the EU, and a $6 million loss in Japan. Figure 3 graphs these data in terms of percentages of total exports, excluding textiles and clothing. It also estimates the effects of abolishing all preferences. The first two columns of each block are very small. By far the largest is a loss approaching 1 percent in Japan, but Japan is a small market for African countries’ exports. The overall loss is on the order of about 1/5 of 1 percent of exports.

Figure 3.
Figure 3.

Effect of Uruguay Round on African Exports

Source: Azita Amjadi, and Alexander J. Yeats. Nontariff Barriers Africa Faces: What Did the Uruguay Round Accomplish and What Remains to Be Done? Policy Research Working Papers 1439 (Washington: World Bank, 1995).

One should not rely too much on the precision of these figures, but the general message is quite clear. Preferences were not doing much for you. Do not shed too many tears that they have been eroded.

The Multifiber Arrangement

The Multifiber Arrangement, rather like the agricultural preferences, is essentially a quantitative preference. It has the effect of keeping, say, Korean and Chinese textiles and clothing out of the OECD markets. That has two effects, each of which rebounded to the advantage of African countries. First, it has opened up market “niches” in the OECD markets, which African suppliers might be able to fill, and second, it has encouraged Hong Kong, Korea, Thailand, and so on, to sell those goods that they could not sell in the OECD elsewhere, including African countries.

Abolishing the MFA is going to undo these two effects. It is quite possibly going to reduce the supply of clothing to African consumers and hence drive up the prices they face. Second, to the extent that African producers are selling in OECD countries they will face extra competition. Work that is being done in the World Bank3 suggests that there is a policy issue here that needs to be addressed. All such numbers are very doubtful, but as far as we can tell, the direct costs of production for clothing are rather higher in Africa than they are in some other low-wage countries—see Figure 4. The message is that the real problem that African clothing producers face lies in the cost of fabric, which is the major cost in producing low-quality clothing. This problem might arise from the efficiency with which the fabric is cut, but more likely, we suspect, it resides in tariffs and customs procedures, and so on, for importing the goods.

Figure 4.
Figure 4.

Direct Cost for the Production of Casual Long Sleeve Shirt

Source: Azita Amjadi, and Alexander J. Yeats, “Nontariff Barriers Africa Faces: What Did the Uruguay Round Accomplish and What Remains to Ee Doner?” policy Research Working Papers 1439 (Washington: World Bank, 1995).

Against the difficulties that I have just noted one must remember that the MFA always threatened to put a cap on any successful exporter. The way the MFA worked was, of course, that you had free trade until you got too good at it, at which point the OECD governments would suggest negotiating a friendly little quota with you. Just as I said about preferences, this is no basis for an industrialization drive. Entrepreneurs will not invest if they think they will be hit as soon as they get good at something. Thus, by removing this threat, the abolition of the MFA, albeit somewhat delayed, is going to be a significant benefit to Africa in the long run.

The critical point to recognize in all this is that the abolition of the MFA is moving us from a managed market to a market that is free. The people who benefit from free markets are those who are flexible enough to expand output. At present, African producers face rather few actual, as opposed to potential, barriers to clothing exports. Indeed, they benefit from barriers on other producers and, because of these, are able to stay in business even if they are allowing difficulties to emerge in getting decent fabric into the firms where the shirts can be produced. In the long run, African governments must address these supply problems not only to expand where the MFA might once have stopped them, but to hang on to their existing market shares.

Food Prices

A big worry for African countries about the agricultural liberalization intended in the Uruguay Round was that by cutting export subsidies, domestic protection in the OECD would drive up the price of food on world markets. For poor countries that imported a lot of food this could be serious, and during the Round commentators were talking about possible increases of 10–15, or even 20 percent in some food prices. These estimates, however, were predicated on the assumption that the liberalization of agriculture achieved in the Round would be at least as deep as foreshadowed in 1991 by Mr. Arthur Dunkel, the former GATT Director General—namely, 36 percent cuts in protection levels, 21 percent cuts in quantities of subsidized exports, and 20 percent cuts in domestic support. However, in terms of the erosion of industrial countries’ agricultural protection, the Round was very very much less successful than that. Agriculture has not been liberalized very much at all. That is bad news for the world economy and for the countries that export those goods. It is not, however, bad news for the countries that import food, because the corollary of weak liberalization is only small increases in world food prices.

Table 3.

Terms of Trade Effects of the Uruguay Round

article image
Source: I. Goldin and D. van der Mensbrugghe, “An Assessment of Economywide and Agricultural Reforms,” Chapter 2 in The Uruguay Round and the Developing Economies, ed. by W. Martin and L.A. Winters, World Bank Discussion Paper No. 307 (Washington: World Bank, 1995).

Goldin and van der Mensbrugghe4 estimate the effects of the Uruguay Round on a selection of food prices. Table 3 gives four sets of figures: the world price in 1992, the World Bank’s forecast of prices in 2005, the percentage difference between those two figures, and, finally, Goldin and van der Mensbrugghe’s estimate of the effect of the Uruguay Round. The Uruguay Round effects are small, both absolutely and relative to what we think was going to happen in prices anyway.

Table 4 illustrates the effects of these price rises on the Central African Republic, assuming that quantities do not respond at all. Column 2 gives the value of net exports of each of the products in 1990 to 1992. Column 3 values these net exports at the prices that we project in the absence of the Round, and Column 4 repeats this at the prices projected allowing for the Round. The difference—the effect of the Round on Central African Republic’s food net export bill—is given in column 5. The numbers are tiny: in total, half a million dollars. That is, the Uruguay Round will worsen Central African Republic’s terms of trade for food by $0.5 million, or 0.039 percent of GDP.

Table 4.

Effects of World Price Changes Due to Agricultural Liberalization in the Central African Republic

(In millions of U.S. dollars, unless otherwise noted)

article image
Source: World Bank calculations.

Q’P(WR)- Q’P(NR)- 2002.00

The cost of the agricultural liberalization to the Central African Republic is around four ten-thousandths of the GDP. It is real. If you took four ten-thousandths of GDP away from me I would feel pretty aggrieved. But it is by any standard small: much smaller than all the other things that are happening to the economy, and small enough that if only one gets one’s own policy right by improving small things, like customs administration or the distribution of the agricultural inputs, one could easily offset it.

Table 5 repeats this exercise for every net food importer among the least-developed countries in Africa. The figures are small throughout. For example, Egypt, is estimated to lose 11-thousandths of GDP. To be sure, this is a real cost, not a transitional one, in the sense that it will not go away by itself. But it is a small cost, a small shock, that it is easily handled by better economic management; it is a very very small shock in a rather noisy system.

Table 5.

Estimates of the Effects of the Uruguay Round on Net Food Importers’ Terms of Trade

(In millions of U.S. dollars, unless otherwise noted)

article image
Source: World Bank calculations.

Q’P(WR) - Q’P(NR) - 2002

What conclusions do I draw from this? I do not say that we need not worry about these numbers, but I do conclude that to put effort into seeking compensation for it is a waste of time. Everybody suffers shocks of this size in the course of everyday economic life. The answer is to offset them through the benefits of liberalizing your own markets and domestic policy stances.


This paper has considered briefly four aspects of the outcome of the Uruguay Round of interest to sub-Saharan Africa. The first represents a clarification of the trading rules—on preshipment inspection—that Africans had desired. The remainder comprise two areas where the costs to Africa were much less than feared—preference erosion and food prices—and one that laid longer-term gains against shorter-term competitive pressure. What is conspicuously missing is any unambiguous gains. The reason is that African countries generally stayed on the margins of the Uruguay Round’s main business—exchanging market-liberalizing concessions. The Africans had relatively little to gain from other countries’ concessions, for they already have fairly good access to OECD markets, but they had large amounts to gain from liberalizing their own markets for imports. That they generally chose not to reap these gains sets them apart from very many other developing countries. Thus, while the Uruguay Round represents something of a shot in the arm for many developing countries, for most African countries it represents more of a self-inflicted shot in the foot.


This section is based on the paper by P. Low, Preshipment Inspection Services, World Bank Discussion Paper No. 278 (Washingtoon: World Bank, 1995.)

Note: The author is grateful to Merlinda Inges for help with certain sections of the paper. The views expressed are the author’s; they do not necessarily reflect those of the World Bank or its member governments.


This is rather perverse, for in most cases one suspects that the straight income transfer is worth more than the profits on the additional exports. Note, it is the profit (surplus) on exports that affects welfare not the total revenue, because the extra exports incur costs of production.


T. Biggs, and others, Africa Can Compete! World Bank Discussion Paper No. 278 (Washington: World Bank, 1994).


I, Goldin and D. van der Mensbrugghe, “An Assessment of Economywide and Agricultural Reforms,” Chapter 2 of The Uruguay Round and the Developing Economies, ed. by W. Martin and L.A. Winters, World Bank Discussion Paper No. 307 (Washington: World Bank 1995).