Trade liberalization and export sophistication in Vietnam

ABSTRACT

This paper empirically examines the impacts of trade liberalization policy on the sophistication level of Vietnam's exports from 2001 to 2010. The export sophistication measure proposed by Hausmann and co-workers is computed by using the disaggregated trade data. By descriptive analysis, this indicator reveals that Vietnam's export structure was similar to that of Indonesia and the Philippines and then became much more similar to Thailand after Vietnam's accession to World Trade Organization (WTO). In addition, this paper econometrically analyzes the effects of trade liberalization on Vietnam's industry-level export sophistication with the additional consideration of its WTO accession in 2007. This paper suggests that tariff reductions have a positive impact on the sophistication level of Vietnam's industry exports. Trade liberalization has a stronger impact on the nonmanufacturing sectors than on manufacturing sectors. However, the results also imply that the WTO membership does not have any additional effects on Vietnam's industry export sophistication.

KEYWORDS:

• Exports
• trade policy
• tariff liberalization
• Vietnam

JEL CLASSIFICATIONS:

• F13
• F14
• O24
• O53

Acknowledgements

I am indebted to the two anonymous referees for their insightful comments. I also wish to thank Craig Parsons, Eiichi Tomiura, Kiyotaka Sato and Keiji Nagai for their useful comments. The usual disclaimer applies.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. Following WTO terminology, in a free trade area (FTA), trade among members is duty free but members set their own tariffs on imports from nonmembers.

2. Cambodia, Lao PDR and Myanmar have the same schedule as Vietnam. However, the other members of ASEAN (Brunei Darussalam, Indonesia, Malaysia, Philippines, Singapore and Thailand) committed to the full elimination of tariffs in intra-ASEAN trade by 2010. According to ASEAN, import duties on 99.65% of all tariff lines under the CEPT–AFTA had been eliminated by 1 January 2010. For the newer ASEAN members (Vietnam, Cambodia, Lao PDR and Myanmar), 98.96% of their tariff lines had been brought within the 0%--5% range (joint media statement of the 42nd ASEAN economic ministers' meeting in Da Nang (Vietnam), 24–25 August 2010. Viewed at: http://www.asean.org/news/asean-statementcommuniques/item/joint-media-statement-of-the-42nd-asean-economic-ministers-aem-meeting-da-nang-vietnam-24-25-august-2010-2).

3. For more details, see http://vietnam.usembassy.gov/uploads/images/LyF8aTkwwk2Jr3-pvrFMKA/bta-crsrpt020909.pdf

4. http://www.mof.gov.vn/portal/page/portal/mof_vn/1371620/1371622/1371631/45916532?p_ page_id=2202417&pers_id=45917882&item_id=46123480&p_details=1 (in Vietnamese)

5. http://www.trungtamwto.vn/node/4353 (in Vietnamese)

6. http://www.mof.gov.vn/portal/page/portal/mof_vn/1539781?pers_id=2177088&item_id= 83889508&p_details=1 (in Vietnamese)

7. http://www.moit.gov.vn/vn/tin-tuc/5109/le-ky-chinh-thuc-hiep-dinh-thuong-mai-tu-do-viet-nam–han-quoc-(vkfta).aspx (in Vietnamese)

8. For more details, see the statistics at http://www.gso.gov.vn/default.aspx?tabid=720 (in Vietnamese).

9. Trade Policy Review Report No.WT/TPR/S/287, 13 August 2013. Viewed at https://www.wto.org/english/tratop_e/tpr_e/s287_e.pdf.

10. Bank (2011), Taking Stock - An Update on Vietnam's Recent Economic Developments, prepared for the Annual Consultative Group Meeting for Vietnam Ha Tinh, 8–9 June 2011. Viewed at: http://siteresources.worldbank.org/INTVIETNAM/Resources/TakingStockEng.pdf.

11. Lall, Weiss, and Zhang (2006) also develop a similar index which is called the ‘sophistication level of exports.’ They also argue that it is not a specific technology level, but that it captures many other factors affecting export location, and care is needed in interpreting results. However, if Hausmann and co-workers proposition, ‘countries become what they export,’ is really true, the key determinants of export sophistication should be a fruitful avenue for academic research and important implications for policy makers in developing countries.

12. I wish to thank an anonymous referee for suggesting this test.

13. The source of tariff data for this study is the World Bank's World Integrated Trade Solution database (WITS).

14. Two Chow tests (Chow 1960) were performed to detect the possibility of structural change, starting in 2007 when Vietnam become a WTO member, as well as regime change between manufacturing sectors and non-manufacturing sectors. The results of both tests indicate that there were structural changes in the model starting in 2007, and between manufacturing and nonmanufacturing sectors. This enables me to include dummy variables for the WTO accession and manufacturing industries. Although the WTO accession is likely the most important among Vietnam's trade liberalization schemes, it is still true that there may be other confounding effects operating around 2007.

15. PRODY_i,t is a log-transformed variable. Tariff rates are approximated as τ_i,t ≅ ln(1+τ_i,t ) and world trade growth is also approximated as .

16. Vietnam's FTA partners include the other ASEAN member countries, Japan, Korea, China, Australia, New Zealand and India. Following Hummels and Klenow (2005), I compute the intensive margin as follows:

where i denotes the Vietnam's FTA partners, k denotes the ‘rest of the world’ by including all exporters to Vietnam except for i. S_ijt is the set of products s in which the exporter i exports to Vietnam in year t. S is the set of the four-digit ISIC commodities. IM_ijt measures the intensive margin between i and j in time t by comparing exports from i to j in time t with exports from the rest of the world to j in time t in the sectors s in which i exports to j in time t. IM_ijt is positive and can be above or below 1. The extensive margin of Vietnam's imports from its FTA partners is also calculated using this approach. However, it does not change over time and remains at the maximum value of 1. Thus, it was not included in the regression equation.

17. In this paper, I implement the xtabond2 user written command in STATA 10 (Roodman 2006) to estimate the model.

18. The OLS, FE and GMM estimates are summarized in an appendix.

19. For more details, see the Trade Policy Review No. WT/TPR/S/287 at https://www.wto.org/english/ tratop_e/tpr_e/s287_e.pdf.