---

Behind the Changing Numbers on Doha Round Global Welfare Gains
On Friday, July 21, the Economic Policy Program hosted a public luncheon discussion on the subject of economic modeling and the changing numbers on trade liberalization and global poverty.  David Orden of the International Food Policy Research Institute (IFPRI) gave a short introduction to the subject, Antoine Bouët of IFPRI presented the findings of his recent research brief, How Much Will Trade Liberalization Help the Poor? Comparing Global Trade Models, and Will Martin, Lead Economist with the Development Research Group at the World Bank, gave a response.  An hour-long Q&A session followed with the 70-plus participants in attendance from the Washington policy community, the embassies, NGOs, the private sector, and U.S. government agencies.

A common complaint heard during the Uruguay Round of trade talks (1986-94) was that negotiators were “negotiating in the dark,” but the advent of more powerful computing has meant that the Doha Round of WTO negotiations has been the subject of intense scrutiny and research using the very latest techniques in economic modeling and simulation analysis.  A number of high-profile studies have sought to quantify potential welfare gains resulting from a likely agreement and the proportion that would go to poor countries. 

The main empirical tool for these studies, the multi-country computable general equilibrium model of the global economy, is a sophisticated and complex tool of analysis but often appears as a “black box” from which the results are either oversimplified or difficult to understand.  Policy-making processes now demand economic modeling, but the results should be expected to inform, rather than settle, trade policy. Instead, studies which should have been treated as helpful signposts to the impacts of trade reform have instead been turned into a stick with which to beat liberalization by its opponents.  Widely varying estimates on the number of people worldwide that would be lifted out of poverty — from 72 million to 440 million — and the predicted increase in overall global welfare — from 0.3 percent to 3.1 percent, a difference of a factor of more than 10! — have cast doubt on the value of such exercises and, more significantly, helped feed a growing trade pessimism skeptical of the gains that trade could mean for the poor. 
A graph of the projected gains from various Doha modeling exercises between 1999-2006 shows a clear trend of diminishing optimism (though projected benefits are still positive). 
According to the speakers, there are four main reasons why there is so much divergence between economic models: differences in the data included in the models; different behavioral parameters; different theoretical assumptions; and differences in the experiments conducted.  As computing power has increased and data have improved on such matters as bound versus applied tariffs, regional trade agreements, preference programs, etc, models have been updated to include these elements. 
In general, the earlier studies did not take a lot of already existing market access into account, and the fact that the world is already more globalized than had been assumed five years ago means that there is less to be had by way of potential new welfare gains from more open trade.  Different behavioral parameters in the models and different theoretical assumptions — such as assessments of the dynamic relationship between trade openness and global factor productivity — have also contributed to the widely varying results, and the different “scenarios” tested by modelers — many of them tracking the falling level of ambition in the WTO negotiations — have also contributed to the declining numbers.  
Paul Krugman’s “dirty little secret” remains: the welfare gains from trade liberalization are small relative to the overall size of the global economy (0.7% of GDP in recent a recent World Bank study).  However, this should not be discounted: these gains would ultimately be distributed progressively, raising the incomes of the world’s poorest countries by a greater amount than their current share of global income.  There are also many likely gains from trade that are not yet being effectively measured in current modeling exercises, and many additional steps such as the provision of “Aid for Trade” to boost countries’ supply-side capabilities. 
With the collapse of the Doha Round talks, these questions will likely take on increasing urgency in the United States, Europe, and across the developing world as concern grows that further trade liberalization may be now be slipping out of reach for a generation.