Cirera, X., Winters, L.A. 2015. Has Aid for Trade helped African economies achieve structural transformation? GREAT Insights Magazine, Volume 4, Issue 6. December 2015/January 2016.
While some changes have occurred in patterns of exports from Africa, it cannot credibly be attributed to Aid for Trade, nor are there signs that AfT has helped switch employment from agriculture to industry.
The volume of Aid for Trade has increased more than tenfold in the past twenty years with the objective of accelerating economic development in developing countries. Structural change is a necessary part, if not the core mechanism, of development. Hence, in this article, we ask: ‘Has Aid for Trade (AfT) helped African economies achieve structural transformation?’ Our conclusion is ‘unfortunately, not as far as we can see.’
Structural change entails the reallocation of employment from low productivity “traditional” sectors, such as agriculture, to high productivity “modern” sectors, such as manufacturing and tradable services. Such reallocation increases average labour productivity and hence average incomes. International trade plays a key role in structural change as it allows countries to transform their production structures towards products based on their comparative advantage, without having to effect an equal change in consumption (as would be necessary if there were no trade). Many Sub-Saharan African (SSA) countries’ exports are concentrated in minerals; mining is a high productivity activity but it typically employs very few people. Since mineral exports tend to discourage other exports because they appreciate the exchange rate, a focus on mining tends to reduce average economy-wide productivity and thus not entail positive structural change.
Aid for Trade is development assistance aimed explicitly at improving conditions for international trade – for example, by improving infrastructure and trade policy or by helping producers to meet exporting standards. If it were effective, we might reasonably hope that it would foster income-increasing structural change. Because we have such weak data on labour productivity and the sectoral composition of employment, we adopted a two-part research strategy: first, we asked if structural change was likely to occur as a result of AfT, by testing how AfT affected SSA’s costs of doing international trade and its volume of trade; second, we traced this through to actual structural change.
Trade patterns have changed, often for the better
According to the OECD-DAC database, disbursements of AfT to SSA countries have grown from US$ 0.78 billion in 1995 to US$ 7.5 billion in 2010. The flows increased strongly for nearly all African countries; the main recipient was Ethiopia, which absorbed around 9 percent of flows to the region, followed by Tanzania, Ghana, Mozambique and Uganda. Decomposing AfT by sector, around 60 percent was concentrated on transport and storage, (35 percent, with 27 percent to road transport only) and agriculture (24 percent), followed by energy (13 percent) and the financial sector (6 percent).
As a first descriptive exercise, we looked at the structure of SSA countries’ exports in 1995-2000 and in 2005-2010, disaggregating into three broad sectors: agriculture, extractive industries and manufacturing/industry. In 15 out of the 44 countries, the largest sector changed between these two periods, with 9 of these cases changing from agriculture to industry, albeit often by not very much. The largest changes were evident in the three countries that had major resource discoveries over this period, which pushed exports from the extraction industries into first place.
In a second exercise, we considered exports disaggregated by the approximately five thousand products distinguished in the Harmonised System Classification. We asked whether changes in trade patterns at this level tended to reinforce comparative advantage or work against it. Measuring comparative advantage with Balassa’s Revealed Comparative Advantage (RCA – the ratio of the share of a product in a country’s exports to the share of that product in world exports, so that a value exceeding one indicates comparative advantage), one very clear result emerged. In every country except one, the products for which the country had a revealed comparative advantage in 1995-2000 showed declines in RCA by 2005-2010 more often than increases. This might suggest that, on average, SSA countries were shifting their exports away from traditional products, exactly the sort of export diversification that AfT hoped to achieve.
Unfortunately, we cannot confidently attribute these shifts to AfT. We have no idea which sectors AfT would affect (and neither in many cases did policy makers, as AfT was used to make changes in conditions that pertained to all products). Hence, all we have to go on is the observation that while AfT increased, trade patterns diversified. SSA’s aggregate exports and AfT have risen together although this does not prove a causal relationship.
AfT flows do not appear to explain changes in trade costs and in trade flows…
AfT boosts international trade by reducing trade costs. With a suitable set of assumptions one can infer trade costs directly from actual trade: you start by working out what the level of trade between each pair of countries would be if there were no trade costs; you would then ask what costs would need to be imposed on trade in order to reduce the volume from the predicted levels to the levels actually observed. We did this and found no sign that the changes in trade costs over 1995-2010 implied by this method are related to AfT.
In a more direct approach we also used data on (a) how long it takes to export and import goods from each SSA country, and (b) the cost of exporting a 20 foot container from each country. Measured AfT flows have virtually no explanatory power for the evolution of either of these measures of trade costs. The only exception is a hint that AfT devoted to policy development might help reduce the time needed to clear customs (measure a). We also looked for the effects of AfT directly in international trade data. We asked whether the evolution of SSA countries’ exports and imports either in total or disaggregated by partner were related to their receipts of AfT. This exercise generates large numbers of statistical results and occasionally one or two appear to be statistically significant. Overall, however, as far as we can discern, AfT played no significant role in shaping either SSA countries’ aggregate or bilateral trade levels over 1995-2010.
…and have no discernible impact on the structure of the labour force
Our final econometric exercise was to ask whether countries’ allocations of labour between agriculture and non-agriculture could be related to AfT. We calculated the labour force splits as averages over successive three-year periods between 1995 and 2010, although as we explain in Cirera and Winters (2014), deriving data as simple as those required rather heroic assumptions. Perhaps unsurprisingly, given the above, we also find no relationship running from AfT flows to the allocation of SSA countries’ labour forces.
It would be easy to conclude from the above that AfT had no effect on structural change or even on international trade, and hence, has all been a waste. While our analysis clearly challenges simple-minded assertions that, ‘of course, investing in reducing trade costs must be beneficial’, concluding that AfT had no benefit is premature.
First, the challenges of constructing the data necessary to make these tests are formidable, not least in quantifying AfT itself. It is the donors who attach the moniker ‘Aid for Trade’ to a flow, rather than the recipients who could do so in the light of what AfT is actually spent on.
Second, there is clearly great heterogeneity across countries in terms of trade, comparative advantage and structural change. Our search for general results is therefore competing with a plethora of country-specific factors and circumstances; our inability to find such results may just reflect the low power of our tests (using relatively small amounts of relatively weak data) rather than the absence of effect. It is possible, for example, that while some AfT has diversified international trade, commodity price increases have offset the potentially positive impact of this on overall economic structure
Third, structural change is not an explicit objective of AfT. Indeed, given policy-makers’ and donors’ imperatives not to be seen as making mistakes, AfT may be devoted to helping existing exporters – i.e. reinforcing existing patterns of trade and production – rather than stimulating new ones. Such an outcome is reinforced by the political power of incumbent exporters who will lobby for AfT to be spent on things that benefit them.
We have two recommendations from this research: first, we need better data and more analysis if we are to get more definitive answers on the effectiveness of AfT and particularly better impact evaluations where specific and focused interventions are supported; second, if we feel that AfT should be oriented towards structural change, programmes should be chosen and managed with that objective in mind. However, while those governments that are intent on development should make more space for structural change, AfT is not necessarily an ideal tool for doing so. Working to reduce the frictions on countries’ current baskets of exports and imports can be useful as well.
This article was first published in ICTSD Bridges Africa, vol.4, no.5, May 2015. It is based on a longer paper Xavier Cirera and L A Winters (2014) Aid for Trade and Structural Transformation in Sub-Saharan Africa, Commonwealth Trade Policy Discussion Papers, No 2015/01, London, Commonwealth Secretariat.
About the authors
Xavier Cirera is Economist, World Bank.
L. Alan Winters is Professor of Economics, Sussex University.
Photo: Nyirefami Limited, which employs 28 people in a flour milling and processing plant. The UK’s Department for International Development is an example of promoting business partnerships with the aim of building prosperity. Credits: Ed Hawkesworth/DFID/CC.
This article was published in GREAT Insights Volume 4, Issue 6 (December 2015/January 2016).
L. Alan Winters