CIEM and Star—Vietnam Project, 2003, An Assessment of the Economic Impact of the United States—Vietnam Bilateral Trade Agreement, Vietnam Central Institute of Economic Management and USAID.
Deardorff, A., 1984, “Testing Trade Theories and Predicting Trade flows,” in Handbook of International Economics, Vol. 1, ed. by R. Jones and P. Kenen (Amsterdam: Elsevier Science Publisher).
Feridhanusetyawan, T., 2005, “Preferential Trade Agreements in the Asia-Pacific Region,” IMF Working Paper 05/149 (Washington: International Monetary Fund).
Fukase, E., and W. Martin, 1999, “A Quantitative Evaluation of Vietnam’s Accession to the ASEAN Free Trade Area (AFTA),” Policy Research Working Paper No. 2220, (Washington DC: The World Bank).
Soloaga, I., and A. Winters, 2001, “Regionalism in the Nineties: What Effect on Trade,” North American Journal of Economics and Finance, No. 12, pp. 1- 29.
Thornton, J., and A. Goglio, 2002, “Regional Bias and Intra-regional Trade in Southeast Asia,” Applied Economic Letters, Vol. 9, pp. 205- 208.
Wang, Z., and A. Winters, 1991, “The Trading Potential of Eastern Europe,” Centre for Economic Policy Research, Discussion Paper Series, No. 610, UK.
Wei, S., and J. Frankel, 1997, “Open versus Closed Regional Trade Blocs,” in Regionalism versus Multilateral Trade Arrangements, ed. by T. Ito and A. Krueger (Chicago University Press).
ANNEX I Gravity Model Specifications
The first specification of the gravity model is as follows:
The coefficient of the RTA2 dummy, β4, measures the amount of trade between two members of the same RTA in excess of that between two countries at least one of which is not a member of that RTA.
The second specification is as follows:
The coefficient of RTA1, β6, expresses the amount of trade between one RTA member and one non-RTA member relative to that of two non-RTA members. The coefficient on the RTA2 dummy, β5, now represents the extra amount of trade between two RTA members relative to that of two non-RTA members. Finally, the difference between the RTA2 and RTA1 dummies represents the level of trade between two RTA members relative to that of one RTA member and one non-RTA member. Because the second specification makes it possible to disentangle the effects of trade among members and non-members of RTAs, it provides an indication as to whether RTAs in the region may have been trade creating or trade diverting.
While the coefficients of the dummies of the first specification measure the extra amount of trade between two member countries of a regional group compared with “any other pairing,” the second specification is able to further distinguish within the “any other pairing” group the extra trade between a member of the regional group and a country that does not belong to that group. For example, in the case of a dummy for ASEN, the first specification could measure the extra amount of trade between Vietnam and Indonesia because of their common membership to ASEN, relative to any other pair of countries of which at most one belongs to ASEN (e.g., Vietnam and Albania, or Albania and Angola), but without distinguishing among them. The second specification, however, can distinguish the pair Vietnam and Albania (i.e., when at least one country is part of ASEN), from a pair of countries where neither belongs to ASEN, such as Albania and Angola.
In the first specification (Table 7., Column i,), the RTA2 dummy takes value 1 when both countries belong to the same RTA and 0 otherwise (e.g., either when both countries are nonmembers of RTAs, or when only one country in the pair is a member of an RTA). In the second specification (Table 7., Column ii), RTA1 takes value 1 if the pair includes one country that belongs to the tested RTA and one that does not, and zero otherwise. An example regarding ASEN RTA is shown in the matrix below.
|Colsumn (i)||Column (ii)|
|ASEN 2||ASEN 2||ASEN 1|
|Colsumn (i)||Column (ii)|
|ASEN 2||ASEN 2||ASEN 1|
Prepared by Patrizia Tumbarello (APD).
This specification is in line with that developed by Wei and Frankel (1997). It differs from other gravity models in the literature (Wang and Winters, 1991; Frankel, 1997; and Thornton and Goglio,2002), in that it uses a gravity equation that permits gauging whether RTAs’ effects on trade within their membership may have occurred at the expense of trade with nonmembers.
The ASEN members are: Brunei Darussalam, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines, Singapore, Thailand, and Vietnam.
APEC comprises: Australia, Brunei Darussalam, Canada, Chile, China, Hong Kong SAR, Indonesia, Japan, Korea, Malaysia, Mexico, New Zealand, Papua New Guinea, Peru, the Philippines, Russia, Singapore, Taiwan Province of China, Thailand, the United States, and Vietnam.
The estimated change in trade when a dummy variable is equal to one is calculated as the exponential of the estimated coefficient minus one (i.e. [exp (1.649)-1=4.2] and [exp (0.806)- 1=1.2], respectively).
Frankel (1997) found that ASEN countries trade 6 times the amount that would be traded by two otherwise similar countries and concluded that ASEN is one of the most significant trading areas in the world. See also Wang and Winters (1991)
A negative coefficient in ASEN 1 or APEC 1 would have pointed to possible trade diversion.
This estimated effect is significantly lower than the corresponding estimate in the first specification of regression 5. This is because the APEC 2 coefficient in the first specification measures trade between two APEC countries relative to any other pair of similar countries, while in the second specification the comparison is with a pair composed of one APEC andone non-APEC country.
The fitted values are based on the second specification of the gravity model that allows for a separation of trade-creating and trade-diverting effects.
The predicted values reported in Table 8. are based on regression 1 (Table 7., coulum ii) for a number of reasons. Coefficients on the regional dummies in regressions 2, 4 and 5 are not significant, and therefore it is not possible to use these specifications to assess the possible undertrade/overtrade of Vietnam with its partners. Regression 6 (Table 7.) does not provide a good enough fit for the purposes of this exercise, in that the ASEN dummy, together with the other standard variables of the gravity model, explains only about 20 percent of Vietnam’s trade flows with its neighboring countries.
The data used in the gravity model refer to the year 2002 (i.e., only three years after Japan granted MFN status to Vietnam).