Regionalism ... Trade Creation or Trade Diversion?

All those of you who have taken Econ 101 know about the welfare gains associated with the engagement in free trade and consequently why global economic integration is considered as a good thing in a strict economic sense because this will make countries better off, measured as an overall aggregate. However, there is as always a big difference between reality and theory and although the textbooks shows us how the world would be a better place as a result of global free trade under for example WTO provision we are not quite there yet. Indeed some of the big and salient issues of for example agricultural subsidies are still posing some troubles at the recent trade round under WTO in Doha. This entry is not meant to pick on the WTO but merely to point to the fact that alongside the very real multilateral and global efforts in terms of global free trade we have seen since the end of World War II under most notably the WTO we have also under the same WTO provisions seen a notable surge of PTA (preferential trade agreements) such as NAFTA, MERCOSUR, EU, ASEAN, CAFTA, etc which all are allowed as exemptions to the WTO principle of non-discrimination and most-favored-nation (MFN). The underlying impetus for the co-existence of WTO and this myriad of PTAs is that the creation of for example NAFTA in 1994 can be seen as a step in the right direction towards global multilateral free trade and not as a replacement. But is this true? Under the assumption that free non-discriminatory trade is good how does this correlate with the existence and creation of PTAs? To estimate this international economists use the concepts of trade creation and trade diversion as a result of the creation of PTAs.

Trade Creation - This is a very intuitive outcome of the creation of a PTA and is quite simply operationalized as the increase in the volume of trade (and FDI flows) between the members of the PTA as a result of the removal of trade barriers towards and between a limited number of states. 

Trade Diversion - This concept relates to the distorting nature of a PTA or as it is noted in a WTO framework its discriminatory nature in terms of trade and (potentially also) investment flows. As such, trade diversion is operationalized as the amount of trade created among members of a PTA on the expense of trade between members of the PTA and members outside the PTA. So if country A augments trade with country B as a result of the entering into a PTA but simultaneously decreases trade with country C outside the PTA this last effect would be coined as trade diversion.  As such, the dictum goes that while there is certainly notable trade creation as a result of the creation of a PTA this trade effect has to netted (deducted) relative to the trade diversion.

In the end we are ultimately looking at the net effect of trade creation and trade diversion when we try to asses the benefits and foundation of the creation of a PTA. So what is the evidence you ask? Well, this brings me to the sources in this entry. 

In a very elaborate empirical study based on gravity models of trade the Australien Productivity Commission (PDF) finds that trade diversion often exceedes trade creation in the context of the creation of a PTA.

Contrary to popular conception, the main benefits to a country from preferential (or any) trade arrangements come, not from export opportunities (since these cost domestic resources), but from their ability to provide import needs at lower real
resource cost. This is ultimately what can improve a country’s living standards.

Both theoretical and empirical work has highlighted that PTAs can boost trade among members, but this is often at the expense of non-members. So whether it is beneficial overall to join a PTA depends on the cost structures in partner countries, compared with the cost structures in third parties. If a preferential trade arrangement
diverts a country’s imports from a low-cost third party to a higher-cost preferential trade partner, it can be made worse off. Thus, the best chance a country has to gain from preferential trading arrangements would be to choose a trading partner with the lowest (quality-adjusted) costs. But then the economic rationale for liberalising
preferentially, rather than unilaterally across-the-board, is lost.

The theoretical literature reviewed in this paper has always noted the possibility of losses from net trade diversion, and the empirical work undertaken here suggests that this trade outcome is more common than other recent empirical analyses have suggested. The hypothetical (often computable general equilibrium) analyses of PTAs have sometimes made assumptions that have skewed the findings in favour of economically beneficial outcomes.

So, PTAs are not efficient carriers and conveyors of welfare gains in terms of trade? Well, the story does not quite end here. Consequently, a recent paper by Dean A. DeRosa from the Peterson Institute takes on the study mentioned above and draws a somewhat different conclusion.  

(from the abstract)  

Do preferential trade arrangements (PTAs) create or divert trade? An Australia Productivity Commission study assesses large PTAs such as NAFTA, the European Union, Mercosur, as well as smaller PTAs such as the Andean Group using gravity models and Tobit regressions and suggests that major PTAs are trade diverting while minor PTAs are trade creating. Using an augmented Rose gravity model, the author challenges the APC’s results by first measuring the 20 PTAs in the APC study and then expanding to include nearly 46 PTAs in total. While the author does not fault the APC’s methodology, he concludes that most current PTAs are trade creating and that there is no uniform evidence of trade diversion.

(from the actual paper; PDF!)  

(...) to explore the robustness of the APC study results, the present paper has undertaken an independent empirical analysis of the trade impacts of the same PTAs considered by the APC study and a substantially larger number of current PTAs identified by the WTO. Using a variant of the Rose (2004) gravity model augmented by somewhat more up-to-date data than utilized by the APC study and applying a variety of econometric techniques, inclusive of the Tobit method employed by the APC investigators, the present analysis finds that the majority of PTAs in force today are trade creating rather than trade diverting. This is true not only on an intrabloc basis but also on an overall trade basis. In sum, the present econometric results are diametrically at odds with the APC study findings but nearly squarely in line with most of the gravity model literature on PTAs.

There is room for a some nice term papers here I would say.