Back to the Main ET Page

On the “Extravagant” Predictions of Benefits from the Uruguay Round

Arvind Panagariya

In June 1993, Peter Sutherland succeeded Arthur Dunkel as the Director General of GATT and finally brought the Uruguay Round (UR) to a close. One of the arguments he used to persuade member countries to conclude the Round was that, according to the numerical predictions by researchers, the Dunkel Draft promised welfare gains worth hundreds of billions of dollars.

Rubens Ricupero, Secretary General, UNCTAD, has recently questioned these predictions, characterizing them as “extravagant”. Many observers, familiar with the UR Agreement and the studies that made the extravagant predictions, share the Secretary General’s view.  Indeed, economist Jagdish Bhagwati had warned even as the Round was approaching closure that appeals to overly optimistic predictions could undermine future trade liberalization.

I have consistently maintained that trade liberalization, whether multilateral or unilateral, is beneficial to the liberalizing countries.  Yet, in scholarly work, it is important to limit claims to what is justified by careful analysis and evidence.  It is this cardinal principle that the researchers who offered overblown estimates of the benefits of the Uruguay Round violated leisurely.

The tendency to overstate the benefits and understate the costs of the UR Agreements is best illustrated by the study, The Uruguay Round and Developing Countries, edited by distinguished economists Will Martin and Alan Winters and published by the World Bank in October 1995.  The study consists of thirteen essays and brings together prominent researchers from some of the most important multilateral institutions including the World Bank, IMF, OECD and GATT/WTO.

The UR Agreement embodies well over a dozen wide-ranging agreements.  Of these, only four lend themselves to straightforward numerical analysis: Agreement on Agriculture, Agreement on Textiles and Clothing (ATC), Agreement on Market Access in industrial products and Agreement on Trade Related Aspects of Intellectual Property Rights (TRIPs).

            Martin and Winters note at the beginning of their executive summary that though the Agreement on Agriculture achieved a great deal in terms of defining rules for agricultural trade, it achieved “little in terms of immediate market opening.”  The agreement did bring about some reductions in export subsidies and introduced minimum imports through tariff quotas.  But the unambiguous message of the executive summary is that its net impact on agricultural protection would be negligible.

            In industrial products, the agreement to phase out the multi-fibre Agreement (MFA) under ATC constitutes a major accomplishment.  As for other products, with rare exceptions, developing countries chose tariff bindings well above their applied rates.  Therefore, their commitments can generate virtually no impact, positive or negative. Commitments by developed countries amount to a reduction in their import-weighted average tariff from 6.3 to 3.8%.  As any trade economist will tell you, in a realistic analysis, this small reduction from an initially low level of tariffs cannot lead to large gains.  This is even truer when there are other high trade barriers in existence, as in agriculture.

            Finally, the TRIPs agreement promises to lower the world welfare by extending monopoly rights of patent holders to the entire world for 20 years.  It is also expected to result in a transfer of income from developing country consumers to innovators who are located mainly in developed countries.  Indeed, working under the even-handed leadership of Arthur Dunkel at GATT, Arvind Subramanian, now at IMF, had estimated the losses from TRIPs to developing countries to be several billion dollars.

            The upshot of this analysis is that, among the UR commitments that lend themselves to numerical analysis, only the removal of MFA and TRIPs could generate significant effects.  Under no reasonable scenario can they add up to benefits of $200 billion for the world and $50-90 billion for developing countries.  The predictions for developing countries are especially troubling since the gains to them from the removal of MFA are muted by the loss of quota rents and the contribution of TRIPs is negative.

            A close look at the essays reveals that the larger numbers were generated by building into the models liberalization that did not even take place.  For instance, the essay by Ian Goldin and Dominique van der Mensbrugghe went on to predict a welfare gain of $48 billion from liberalization in agriculture alone.  Similarly, though tariff bindings chosen by developing countries were above applied rates, they got counted as real liberalization.  On top of these factors were piled up unrealistic model structure and parameter values.  Finally and mysteriously, the impact of TRIPs was uniformly excluded from the welfare calculations.

            To be sure, one of the more careful essays, by Glenn Harrison, Thomas Rutherford and David Tarr, did make less extravagant predictions: $93 billion for the world and a modest $18 billion for developing countries.  But these estimates were not included in the executive summary and, predictably, never picked up by the press.

            Why have these studies escaped the critical assessment they deserve in popular press?  The key reason is that they employ gigantic computable general equilibrium models, which are often opaque even to an applied general-equilibrium theorist such as myself.  Frequently, only the author knows the key assumptions that drive the results and sometimes not even he.  The large dimensions are generally not necessary to assess the broad impacts of policy changes but are, nevertheless, fashionable among the model builders.

            Sadly, given the appetite of the press and politicians for numerical estimates and the publicity they readily offer researchers, these models are here to stay.  The lesson for large developing countries such as India, therefore, is that they should significantly boost their own native capacity to conduct research on issues relating to multilateral trade negotiations.  The experience with UR negotiations shows that the ability to negotiate in one’s own interest can be seriously hampered by pressures generated by research motivated by the agendas of rival nations.

With the Seattle Round in the offing, the game of extravagant predictions is bound start all over again. The new Director General of WTO, Michael Moore, himself an author of many carefully written books, should take lead in discouraging this game.  He should also strengthen research at WTO, which is the natural institution for the analysis of issues relating to multilateral trade negotiations.

Economic Times, August 25, 1999