Pecuniary aspects of self-interest in bilateral aid

Development Policy16 Jul 2013Milad Zarin-Nejadan

It is estimated that donors receive a return from development aid of 50 to 80 cents for each dollar given. But so far, studies have not taken the complex links between aid and export into consideration. To gain a full insight into the financial effects of aid on donor countries, a new economic model needs to be elaborated.

The primary objective of official development aid (ODA) is to favour sustainable economic growth and reduce poverty in recipient countries. In some cases, aid also serves to promote institutional reforms and humanitarian goals. This altruistic vision of aid contrasts with the reality of self-interest on the donor side. The pursuit of political goals (influence over recipient and international image) through bilateral aid is the most visible aspect of this self-interest. In a recent contribution to this blog, Anders Riel Müller interestingly argues that development aid is self-interest-led, because its focus tends to differ depending on the donor’s own past development experience.

A more common feature of self-interest in ODA is however of a pecuniary nature. Donors tend to keep an eye on the impact of aid on their own economies, especially through exports to aid recipients. This is a built-in characteristic of tied aid, but it is also true for donors practicing untied aid. In recent years, against the backdrop of the debate around aid efficiency abroad and harsh budgetary conditions at home, governments have increasingly focused on commercial benefits for domestic firms when presenting development aid budgets to their legislative bodies. For instance, in Switzerland, with only 0.45% of Gross National Income devoted to ODA (far below the 0.7% threshold set by the UN), it has become difficult to increase, or even maintain, aid budgets just by invoking aid’s beneficial effects on recipient countries. While some donors quantify and publish estimates of this domestic return to aid (Switzerland being an example), a thorough evaluation of its presumed effect is often lacking and, when available, confined to academic circles.

Aid, whether tied or untied, tends to have a positive impact on exports both in the short run (e.g. through budget constraint) and in the long run (e.g. through economic growth). Yet, the relationship between the two variables is far from straightforward and one cannot exclude a positive or negative effect running in the opposite direction. Exports can influence aid positively if the donor tends to ‘reward’ countries already importing its goods and services. The relationship can be negative if exports to the recipient country are interpreted by the donor as an inverse indicator of the degree of need for aid. Last but not least, the relationship between aid and exports should ideally be studied in a multilateral setting, due to substitution and complementarity effects.

Given the theoretical ambiguities involved, the issue of the relationship between aid and exports can only be settled on empirical grounds. However, comprehensive quantitative studies examining the complex relationship between aid and the export performance of the donor are quite rare. Worse, empirical studies on this subject are mostly accounting-oriented rather than economic in the sense that they usually measure the positive impact of aid on exports by focusing on procurements associated with projects financed by aid. A survey of such studies for various donor countries shows that the return from aid lies usually somewhere between 50 and 80 cents for each dollar given.

Not only do these studies have a narrow focus, but they also suffer from serious shortcomings. Most importantly, they are subject to a measurement risk. First, to the extent that aid is untied or partially tied – which has in fact increasingly been the case in the OECD countries since the 1990s – these studies may overestimate the beneficial effects of aid on exports. One cannot rule out the possibility of the recipient country importing from the donor anyway, with or without aid.

Second, such studies could well underestimate the effect of aid on exports in so far as they do not take into account the impact of aid on the recipient country’s growth, which is often the main motive for giving aid. If aid exerts a positive impact on growth, the developing country’s long-term capacity to purchase goods and services from the donor country is automatically enhanced. Another long-term mechanism is based on ‘goodwill’ generated by aid over time, i.e. the positive predisposition – in the recipient country toward the donor – resulting from cumulative development aid.

Third, accounting-based studies implicitly assume that causality runs unilaterally from aid to exports. But better export performance can also give rise to higher (or lower) aid. So the direction of causality is far from being established at the theoretical level.

Failure to take into account the various complex links between aid and exports might give rise to spurious results and lead to erroneous conclusions. It is therefore necessary to elaborate and estimate a theoretically-founded econometric model in order to test the significance and measure the magnitude of the presumed relationship between aid and exports.