VoxEU Column Development International trade

Who benefits from aid for trade?

One of the few areas where multilateral trade talks are making progress is the so-called Aid-for-Trade Initiative designed to remove frictional barriers to trade such as in transportation, communication and energy infrastructure. This column discusses research suggesting that both donors and recipients benefit from the aid. Aid-for-Trade, however, seems to best promote the exports of middle-income countries rather than, for instance, sub-Saharan African ones.

Donors are widely believed to use aid as a means to foster their own commercial interest (e.g., Berthélemy 2006, Hoeffler and Outram 2011). Nevertheless, assessments of the Aid-for-Trade initiative have largely ignored so far that exporters in the donor countries may be among the main beneficiaries.1 Furthermore, there is limited theoretical guidance with respect to the relative strength of the effects of Aid-for-Trade on recipient versus donor exports. Suwa-Eisenmann and Verdier (2007) survey the recent literature, summarising that “aid flows may affect trade flows, either because of the general effects they induce in the recipient country, or because aid is directly tied to trade, or because it reinforces bilateral economic and political links (or a combination of all three)”. Considerable ambiguity persists, however, on how the major transmission channels affect either donor exports or recipient exports – not to speak of the relative effects on trade in opposite directions.2

Aid-for-Trade financing infrastructure

The largest part of Aid-for-Trade aims at financing better economic infrastructure such as transportation, communication and energy supply (see Figure 1a).3 The argument that improved infrastructure provides an important stimulus to recipient exports commands wide support in the literature. It is often neglected, however, that donor exports, too, could be promoted by better infrastructure. One may even suspect that donors target Aid-for-Trade by selecting infrastructure projects that serve primarily their own export interests. Another observation may point to selfish donor motives when providing Aid-for-Trade. The regional distribution of total Aid-for-Trade is skewed towards Asia (see Figure1b), where many developing countries tend to be closely integrated into the trading patterns of donor countries. The smaller share of Aid-for-Trade given to African countries is striking from a needs-based perspective, considering that facilitating trade is essential for Africa’s economic development.

Figure 1. The composition of aid for trade from all DAC donors, 1990-2010 period


(a) By category

Notes: Trade Policy and Regulations (CRS Code 331); Economic Infrastructure, consisting of Transport and Storage (210), Communications (220), and Energy Generation and Supply (230); Building Productive Capacity, consisting of Banking and Financial Services (240), Business and Other Services (250), Agriculture (311), Forestry (312), Fishing (313), Mineral Resources and Mining (322), Industry (321), and Tourism (332).

(b) By region

New research

Against this backdrop, we assess the effects of Aid-for-Trade from the donors of the OECD’s Development Assistance Committee on the exports and imports of recipient countries to and, respectively, from the donor countries (Hühne et al. 2013). We estimate nested models based on an aggregated version of the gravity model of Anderson and Van Wincoop (2003) covering the 1990-2010 period.4 This approach allows us to identify significantly different effects of Aid-for-Trade on the trade flows in opposite directions by means of a Wald test. Specifically, we would expect significantly stronger effects on the exports of recipients, compared to the imports of recipients, if Aid-for-Trade was unaffected by the self-interest of donors.

Total Aid-for-Trade proves to be effective in promoting trade in both directions. In quantitative terms, a doubling of total Aid-for-Trade would imply that recipient exports increase by about 5% and recipient imports increase by about 3%. The Wald test indicates that the effect on recipient exports is stronger than the effect on recipient imports. Hence, the results for total Aid-for-Trade do not support the skeptical view that donors grant Aid-for-Trade primarily to promote their own export interests.

This also holds when disaggregating total Aid-for-Trade into its three sub-categories.5 Interestingly, the smallest Aid-for-Trade category aiming at trade facilitation with respect to trade policy and regulations appears to be particularly effective. In quantitative terms, the estimates indicate that a doubling of Aid-for-Trade in this category would be associated with a 10% increase in recipient exports. Doubling this category would involve fairly limited donor funds; it could be achieved, for example, by re-directing about 10% of Aid-for-Trade in economic infrastructure.

We gain additional insights by re-estimating the empirical model for various sub-samples of recipient countries. Most strikingly, Aid-for-Trade seems to have missed its objective to overcome supply constraints exactly where these tend to be most severe. The significantly positive effects on recipient exports for the overall sample do not hold for the low-income group. By contrast, the effects on the imports of the low-income group prove to be significantly positive. Compared to the low-income group, it is more likely that Aid-for-Trade stimulates exports of middle-income countries. The Wald tests indicate significantly stronger effects on the exports, relative to the imports, of middle-income countries.

East Asia and Latin America benefit more than sub-Saharan Africa

The findings on selected regions point into the same direction. While Aid-for-Trade may be needed most in sub-Saharan Africa to strengthen the recipient countries’ integration into world markets, Aid-for-Trade is more effective in promoting the exports of east Asia and Latin America than the exports of sub-Saharan Africa. At the same time, it is mainly for sub-Saharan Africa that Aid-for-Trade is positively associated with higher imports from the donor countries.

The study also addresses the question of whether the effectiveness of Aid-for-Trade varies between different types of donor countries, drawing on Berthélemy (2006) who rated donors to be ‘altruistic’, ‘moderate’, or ‘egoistic’. Indeed, the effect of Aid-for-Trade granted by altruistic donors on recipient exports proves to be significant and quantitatively important.6 A doubling of total Aid-for-Trade by altruistic donors would imply an increase in recipient exports by 8-10%. The quantitative effect on recipient exports is stronger for the group of altruistic donors, compared to the group of egoistic donors. In contrast to what one might have expected, however, we find no evidence that donors classified as egoistic by Berthélemy (2006) have used Aid-for-Trade mainly to promote their own export interests. Rather, the coefficients on Aid-for-Trade are highly significant -- independent of whether recipient exports to egoistic donors or recipient imports from egoistic donors are taken as the dependent variable. Moreover, all three donor groups have in common that the Wald tests do not reveal any significant differences in the strength of the effects of Aid-for-Trade on either recipient exports or recipient imports.


All in all, recipients benefited at least as much from Aid-for-Trade as the donors themselves. This contradicts the skeptical view that donors grant Aid-for-Trade primarily to promote their own export interests. This conclusion holds when disaggregating Aid-for-Trade into its subcategories. The quantitative effect of Aid-for-Trade on recipient exports is relatively stronger for the group of altruistic donors, but even the group of egoistic donors does not appear to have promoted mainly their own exports.

All the same, our empirical analysis points to important limitations in the effectiveness of Aid-for-Trade. Strikingly, the significantly positive effects on recipient exports do not hold for the low-income group of recipient countries. Aid-for-Trade rather promotes the exports of middle-income countries, most of which are probably less dependent on aid to overcome supply constraints. Similarly, Aid-for-Trade is more effective in promoting the exports of east Asia and Latin America than the exports of sub-Saharan Africa – even though the need for Aid-for-Trade seems to be most pressing in large parts of sub-Saharan Africa.

Policy implications

Regarding policy implications, the effectiveness of Aid-for-Trade may be increased by shifting the focus of Aid-for-Trade from projects in infrastructure and production sectors toward support in the field of trade policy and regulations. Potentially, donors could achieve much more for the recipient countries’ integration into global trade patterns by raising a relatively small of amount of technical assistance with respect to trade facilitation, compared to spending the same amount in infrastructure or production sectors. This result underscores the importance of ‘finding the path to an Agreement on Trade Facilitation in Bali’, stressed by Pascal Lamy at the Fourth Global Review of Aid-for-Trade in July 2013. The donors should also redress the skewed distribution of Aid-for-Trade in favor of Asia and at the expense of sub-Saharan Africa. However, targeting Aid-for-Trade more strongly in line with regional needs may be insufficient to ensure greater effectiveness of Aid-for-Trade. It would require refined country-specific criteria to identify where pressing need coincides with local preconditions for an effective use of Aid-for-Trade.


Adam, CS and DL Bevan (2006), “Aid and the supply side: Public investment, export performance, and Dutch disease in low-income countries”, World Bank Economic Review 20 (2), 261-290.

Anderson, JE and E van Wincoop (2003), “Gravity with gravitas: A solution to the border puzzle”, The American Economic Review 93(1), 170–192.

Berthélemy, J-C (2006), “Bilateral donors’ interest vs. recipients’ development motives in aid allocation: Do all donors behave the same?”, Review of Development Economics 10 (2), 179-194.

Helble, M, CL Mann and JS Wilson (2012), “Aid-for-trade facilitation”, Review of World Economics 148 (2), 357-376.

Hoeffler, A and V Outram (2011), “Need, merit, or self-interest – what determines the allocation of aid?”, Review of Development Economics 15 (2), 237-250.

Hühne, P, B Meyer and P Nunnenkamp (2013), „Who benefits from aid for trade? Comparing the effects on recipient versus donor exports”, Kiel Working Paper 1852, Kiel Institute for the World Economy.

OECD and WTO (2011), “Aid for trade: Showing results”, Policy Brief.

Pettersson, J and L Johansson (2013), “Aid, aid for trade, and bilateral trade: An empirical study”, Journal of International Trade and Economic Development 22 (6), 866-894.

Suwa-Eisenmann, A and T Verdier (2007), "Aid and trade”, Oxford Review of Economic Policy 23 (3), 481-507.

1 Most of the empirical literature on aid and trade considers aggregate aid flows and trade in just one direction – either from donors to recipients or from recipients to donors. Helble et al. (2012) and Pettersson and Johansson (2013) are notable exceptions.

2 For instance, Adam and Bevan (2006) argue that export-depressing Dutch disease effects tend to be dominated by positive supply-side effects when looking beyond the short term.

3 It should be noted that the relevant aid categories existed already prior to the AfT initiative of 2005, even though donors have pledged to commit additional funds since then.

4 Control variables include a proxy of market access for each recipient country, as the weighted sum of distance-related trade costs and market opportunities in relation to all donor countries. As expected this variable typically enters significantly positive in our estimations.

5 Major results also hold when extending the lags for the AfT variables in order to mitigate endogeneity concerns.

6 Note that reverse causation, i.e., more intensive trade relations between the donor and the recipient driving the donor’s allocation of AfT, is unlikely for the altruistic group. Hence, the results for the group of altruistic donors suggest that endogeneity problems are not particularly serious.

5,879 Reads