Assessing Quality and Impacts of Major Free Trade Agreements

The presentation made in this session covered the issues noted below. The presentation verified and assessed the quantitative impact of FTAs using econometric methods.

The impact of FTAs on trade flows is examined using descriptive and econometric analysis.

The trade creation and trade diversion effects of FTAs are estimated using a gravity model approach.

The Urata Presentation used the two methods of analysis noted below to analyze the impact of FTAs on trade flows, and in particular, the trade creation and trade diversion effects of FTAs.

Descriptive analysis was used to determine the importance of intra-regional trade among FTA members based on the two indices of relative trade shares and trade intensity index.

Econometric analysis based on a gravity model was employed to estimate the trade creation and trade diversion effects of FTAs based on overall trade data and industry trade data.

Descriptive analysis yielded the following two points.

1. Trade creation effects were observed in NAFTA, AFTA, Mercosur, and CER.
2. The importance of intra-regional trade in overall trade increased in the EU, NAFTA, AFTA, Mercosur, and CER.

Econometric analysis based on a gravity model yielded the following points.

1. Analysis of overall trade data indicated that FTAs in total had a trade-creating effect.
2. Analysis of specific FTA treaties indicated that the outcome varied according to FTA.
3. Analysis of industry trade data indicated the following. In the case of the EU, trade-creating effects were observed in the following industries: foodstuffs, plants, and animal products, apparel, and transportation equipment. In the case of NAFTA, trade-creating effects were observed in the following industries: foodstuff, plant and animal products, and transportation equipment. Trade diversion effects were observed in the EU, NAFTA, Mercosur, and were not observed in AFTA.

The following three conclusions were obtained from the two methods of analysis.

1. The Mercosur is most closed, and the EU and NAFTA are relatively more closed than AFTA, or CER. Other FTAs appear to be too recent to show substantial impacts yet.
2. One limitation of the study is the exclusion of such factors as foreign direct investment which would affect trade.
3. Future research agenda includes panel data analysis.

Professor Scollay responded to the Urata Presentation with the following discussions.

The FTA dummy coefficient in gravity models needs to be handled with greater precision. Specifically, accurate analysis may not be possible unless due consideration is given to differences in the implementation periods of FTAs and differences in the features (such as the scope and time profile of liberalization) of FTA treaties.

When using relative trade shares and trade intensity indices to determine whether FTAs generate changes in existing trends, it is probably necessary to distinguish between new and old treaties.

What approach should be taken to reflect the varying levels of liberalization (differences in existing trade impediments, rules of origin, and trade facilitation measures) and time profiles in individual FTAs in the analysis?

Wouldn't impact on trade differ by the scale of FTA?

Macroeconomic conditions must be taken into account when analyzing the impact of FTAs.

The following questions were received from the floor.

Some FTAs are between neighboring countries, while others cover long distances. What impact does distance have on FTAs?

How do you treat the specific features of individual FTAs, such as provisions on foreign direct investment and preferential margins?

The following responses were given to the above questions.

The model contains a distance variable. Therefore, the question of geographic distance is taken into consideration.

I agree it is necessary to examine in detail the features of individual FTAs. However, because analysis based on specific data is difficult, in this study we used dummy variables.