WTO success: No trade agreement but no trade war

Ralph Ossa

11 June 2015



At first glance, there does not seem much to celebrate as the WTO is commemorating its 20th anniversary this year. Most obviously, the WTO has so far failed to deliver any significant multilateral trade liberalisation showing only minimal progress towards completing its Doha Development Agenda. Moreover, its role as the leading forum for international trade policy cooperation is increasingly eroded by the proliferation of regional trade agreements to which governments are turning instead.

Big picture

However, concluding from this that the WTO is a failure would clearly be premature. After all, the WTO’s success does not only depend on how well it promotes trade talks but also on how well it prevents trade wars. And its track record seems much better in this regard. Most notably, trade policy cooperation did not break down in the wake of the recent financial crisis even though this was a real concern given the precedent of the 1930s. There was nothing resembling the Smoot-Hawley Tariff Act and the world trading system emerged from the crisis largely intact.

A casual look at the data already suggests that the WTO’s success at preventing trade wars is likely to far outweigh its failure to promote trade talks. While the average tariff applied during the trade war of the 1930s was around 5o% (Bagwell and Staiger 2002), the average tariff applied by WTO members today is only around 9%. Adding to this the result that even ideal trade negotiations would not eliminate all tariffs since governments are subject to lobbying pressures (Grossman and Helpman 1995), it becomes clear that global trade policy cooperation has already come quite far.

Best guess

In recent research (Ossa 2014), I move beyond such casual observations and provide a more rigorous quantitative analysis of the successes and failures of global trade policy cooperation. In particular, I develop a framework to simulate fully efficient trade talks and fully escalated trade wars and use it to study the effects of moving from applied tariffs to these best-case and worst-case scenarios. My application focuses on the six major players in recent trade negotiations (Brazil, China, EU, India, Japan, US) and a residual Rest of the World. It uses data on 33 industries spanning the agricultural and manufacturing sectors for the year 2007.

In my preferred specification, I estimate the possible real income gains from perfectly efficient trade talks to average 0.3%, and the possible real income losses from fully escalated trade wars to average -3.2%, both expressed relative to the factual situation in 2007. Converting this into monetary values by using manufacturing and agricultural value added in 2007, this implies that the failure of the WTO to promote trade talks costs up to $26 billion per year and the success of the WTO at preventing trade wars is worth up to $340 billion per year (Figure 1). To put these numbers in context, a move to autarky would cost the world $1.461 trillion per year according to my analysis so that a trade war would eliminate around 23% of the gains from trade.1

Figure 1. Successes and failures of the WTO

These numbers are best interpreted as rough estimates of orders of magnitude which have to be taken with a large grain of salt. The reason is simply that they are obtained from a theoretical model with numbers which abstracts from many features of the global economy. Having said this, the applied methodology has a number of attractive features which make them the best guesses available from the academic literature to date. In particular, the model takes a unified view of trade policy, can be parsimoniously matched to the data, and predicts non-cooperative outcomes which are broadly consistent with historical experience, as I will now explain.


The model features inter-industry trade based on comparative advantage, intra-industry trade based on product differentiation, and governments which are subject to pressures from industry interest groups. As a result, it can take a unified view of trade policy according to which governments use tariffs to manipulate their terms-of-trade, shift profits away from other countries, and protect politically influential industries. These forces feature prominently in the theoretical trade policy literature and there is a growing body of evidence suggesting that they also matter empirically (Maggi 2015).

Using recent advances in quantitative trade theory, this model can be parsimoniously matched to the data using only observed trade flows and tariffs as well as trade elasticity estimates. The trade elasticities measure how much trade flows respond to changes in trade costs and are important parameters in the analysis. The model predicts that trade follows a so-called gravity equation which is an empirically successful relationship relating bilateral trade flows to incomes and trade costs.

In my preferred specification, the predicted trade war tariffs average 63%. While this might appear high initially, it is broadly consistent with the trade war tariffs of the 1930s which are usually reported to be around 50% (Bagwell and Staiger 2002).


Overall, the formal analysis therefore confirms the conjecture that the WTO’s success at preventing trade wars far outweighs its failure to promote trade talks. While this should not be taken as a reason to delay institutional reform, it should also not be forgotten when assessing the overall performance of the WTO.


Bagwell, K and R Staiger. (2002), The Economics of the World Trading System, Cambridge, MA: MIT Press.

Grossman, G and E Helpman (1995), “Trade Wars and Trade Talks”, Journal of Political Economy 103(4): 675-708

Maggi, G (2015), “International Trade Agreements.” Chapter 6 in Helpman, E., K. Rogoff, and G. Gopinath (eds.), Handbook of International Economics, Volume 4. North Holland.

Ossa, R (2014), "Trade Wars and Trade Talks with Data", The American Economic Review 104(2): 4104-46.


1 These monetary values all refer to 2005 US dollars. I calculate them by multiplying the real income changes reported in Panel B of Table 4, Panel B of Table 6, and page 4114 of Ossa (2014) with appropriately aggregated manufacturing and agricultural value added numbers obtained from the World Bank’s World Development Indicators (WDI). Since WTO member countries accounted for 96.7% of world GDP in 2007 according to WTO estimates, I include all countries from the WDI dataset in my Rest of the World for simplicity.



Topics:  International trade

Tags:  WTO, trade policy

UBS Professor in Economics of Globalization and Emerging Markets, University of Zurich