Aid for trade and the quality of exports

ABSTRACT

This article studies how aid for trade (AfT) affects the quality of recipient countries’ exports. It shows that the quality effect is most discernible for AfT for assistance in trade policy: a 50% increase in the value of AfT received in this category is associated with a 0.5–1% increase in the quality of exports to the donor and other OECD countries. On average, the actual AfT received for assistance in trade policy leads to a 2% upgrade of the recipient country in the quality ladder of all developing countries. Around half of this quality effect is driven by the quality improvement of continued products in continued markets (intensive margin), and the other half by the quality upgrading of new products in continued markets and existing products in new markets (extensive margin).

KEYWORDS:

• Aid for trade
• exports
• quality
• quality ladder

JEL CLASSIFICATION:

• F35
• L15
• O19
• O24

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Using non-stationary panel estimators, Nowak-Lehmann et al. (2013) find negative but insignificant impact of general aid (official development assistance which includes AfT) on recipients’ exports to donors.

² There are three problems with country-country-product-year-level regressions. First, because of the inter-sectoral nature of a great part of AfT (e.g. technical assistance on trade policy) even the most disaggregated AfT statistics cannot be cleanly aligned with products or industries. Second, such disaggregated level regressions are computationally very demanding especially with high-dimensional fixed effects. Third, results of regressions at this level may be driven by the bias from the fact that some observations are assigned a large weight in the sample simply because some countries happen to export a great number of varieties to some markets.

³ Lee, Park, and Shin (2015) find that WTO members are more likely to receive AfT than non-members. However, we do not include WTO membership dummies for countries or a dyadic dummy indicating the common WTO membership of importing and exporting countries because they would be absorbed by country-year fixed effects here.

⁴ The Broda, Greenfield, and Weinstein (2006) sample reports country-product-specific elasticities of substitution, but the numbers of countries reported for different products are extremely unbalanced, ranging from 1 to 73.

⁵ The classification is accessed via OECD website at:http://www.oecd.org/dac/aft/Aid-for-trade-sector-codes.pdf.

⁶ See Bourguignon and Sundberg (2007) for a general discussion on the possible causes for the mixed effects of AfT found in reduced-form analysis.

⁷ See Wagner (2003), Helble, Mann, and Wilson (2012), and Pettersson and Johansson (2013), among others, for more detailed discussions on ‘tied aid’ in trade context.

⁸ The normalization is constructed as .

⁹ Note that the upgrade is less statistically significant for AfT in economic infrastructure judging from the t-statistics in Table 1.

¹⁰ The conclusions we have here are not qualitatively sensitive to the measure of AfT in terms of lagged periods.

Additional information

Funding

This work was supported by the National Science Foundation of China under Grant No. 71373156.