The J-curve and NAFTA: evidence from commodity trade between the US and Mexico

Abstract

The North American Free Trade Agreement (NAFTA) was predicted to have a substantial impact on the US–Mexico trade, especially on specific importing and exporting industries. In this article, we use annual industry-level export and import data from 1962 to 2004 to discern both the short- and long-run effects of real exchange-rate depreciation on the Mexico–US trade balance, as well as the effects of NAFTA on this trade. We find that peso depreciation has a positive long-run effect on 24 of 102 Mexican industries and a negative short-run effect on 19 of 102 industries. Only a small fraction (7 of 102 industries) show any support for the J-curve hypothesis. NAFTA has had a significant effect on a significant number of the industries, however.

Acknowledgements

Valuable comments of two anonymous referees are greatly appreciated. Remaining errors, however, are ours.

Notes

¹ Indeed, estimating import and export demand functions separately forms another body of the literature. Examples include Tegene (1991), King (1993), Alse and Bahmani-Oskooee (1995), Charos et al. (1996), Truett and Truett (2000), Du and Zhu (2001), Agbola and Damoense (2005), Berument and Dincer (2005), Love and Chandra (2005) and Narayan and Narayan (2005).

² This new tradition of testing the J-curve at commodity level began by Bahmani-Oskooee and Wang (2008), who tested the phenomenon between the US and China.

³ For a more detailed explanation of the bounds testing approach, see Bahmani-Oskooee and Tanku (2008).

⁴ For some other trade models and issues in this journal, see other studies that have used aggregate trade data to address different issues related to the trade balance–Briguglio (1989), Rahman et al. (1997), Vamvoukas (1999), Miljkovic et al. (2000), Kyereme (2002), Dar and Amirkhalkhali (2003), and Dutta and Ahmed (2004).

⁵ For the list of industries, see Table 1.

⁶ Developed cinematographic film; Fish, in airtight containers, N.E.S. and frozen; Leather; ores and concentrates of nonferrous; Other crude minerals; Other electrical machinery and apparatus; Paper and paperboard; Pottery; Power generating machinery, other parts thereof; Sugar confectionery, Sugar preparations, excluding chocolate; Universals, plates and sheets of iron; Vegetables, roots and tubers, fresh or dried; Wool and other animal hair.

⁷ Two additional industries, Fruit, preserved and Organic chemicals, have a significant coefficient, but are shown in Table 2 not to be cointegrated.

⁸ Note that the income variables carry significant coefficients in almost half of the cases.

⁹ Two industries: Stone, sand and gravel; and Special textile fabrics and related show significant NAFTA effects but fail both cointegration tests.