Here is a new paper by Channing Arndt, Sam Jones, and Finn Tarp on whether aid leads to economic growth. The econometrics are done carefully, and it finds that aid inflows of about 10 per cent of GDP lead to an increase in economic growth of about 1 percentage point. (Reassuringly, this is also broadly consistent with a common sense calculation of the sort of effect that aid ought to have.) They also find evidence of bigger, more positive effects of aid, consistent with positive effects of aid on productivity.
I’m not a fan of these aid-growth regressions, because they are technically difficult to do well (see David Roodman’s article on the problems.) But they are important for one reason: they are a more systematic way of doing the popular “folk regression” offered by authors such as Dambisa Moyo and Bill Easterly. When Moyo and Easterly point out that countries that have had high levels of aid have also suffered from slow growth, they are implicitly pronouncing on whether there is a statistical relationship between aid and growth. But of course you would expect to see a lot of aid going to poor countries (rather as ambulances tend to be present at the scene of road accidents) so these simplistic comparisons do not tell us very much about the effect of aid on growth. The more careful question to ask is whether, other things being equal, aid leads to higher or lower growth, and that is what this kind of statistical analysis investigates. It is good to have confirmation that the folk regressions are wrong and that aid does, as best we can tell, lead to economic growth.
There are a few other interesting things about this paper:
- the paper uses the same data as the infamous and oft-cited Rajan and Subramanian paper which claimed that there was no effect on growth (which I criticised at the time here) and finds that, if the regressions are done more carefully, those findings were not correct;
- the effect of development aid on growth is probably understated by this analysis because it includes all aid (unlike the paper by Clemens, Radelet, and Bhavnani, which subtracts humanitarian aid and other aid which is not intended to lead to economic development and finds – as you would anticipate – much larger effects of aid on growth from the subset of aid that is actually intended to promote development);
- there is no sign of diminishing returns to aid in this analysis. (This is an unusual finding – generally studies have needed to include a diminishing returns term to generate a statistically significant relationship between aid and growth).
- the study uses donor-specific fixed effects (the only study to do so, as far as I am aware). I’m looking forward to looking at these in detail, as the estimates will give us an insight into which donors are the most effective.
(h/t Chandan)
Update: David Roodman, whom I regard as an authority on these matters, thinks that I am wrong and Bill Easterly is right.
10 Comments
Digital Scotsman · November 2, 2009 at 4:54 pm
Of course very important to defeat Moyo and Easterly on their own territory – but a more important question sis does Aid lead to poverty reduction.
Of course growth and poverty reduction are not always corellated.
And hypothetically if all of the aid were spent on Mercedes Benzes that wold still contribute to economic growth (providing they were bought domestically)
Roberto · November 3, 2009 at 2:06 am
10 per cent of GDP in aid inflows stir 1 percent in GDP growth? That is not really growth, come on. So you pour 800 million in a country to get an increase in GDP of 80 million… that is just aiding the aid industry, which is exactly Easterly point, now confirmed with econometrics!
If you give 800 million in cash transfers to all citizens in that same country, I can bet you will get at least twice that in GDP growth. It is called the multiplier… increasing demand provides increasing supply, which needs investments, and labor, and inputs, and so on. The good thing… there are already experiences demonstrating that.
Owen replies: Roberto, with respect, I think you’ve misunderstood. According to the econometrics – if you believe them – aid at 10 percent of GDP does not lead to a one-off increase in GDP of 1 percent of GDP (as you describe); it leads to an increase in the growth rate of 1 percentage point (so, for example, the growth rate goes from 2% a year to 3% a year). That means that after ten years of aid, GDP would be a bit more than ten percent higher than it would be without aid.
I agree that there is a lot of potential to make more use of cash transfers. The world’s largest social safety net programme is here in Ethiopia, and it seems to work pretty well. That is one of the things aid can be spent on.
Roberto · November 3, 2009 at 3:19 pm
OK… I understood, aid at 10% of GDP, increases GDP growth rate by one percentage point. That would mean that, as you point, at that level of aid, in ten years GDP would be 10% bigger. Well, China got more than that per year by opening its economy to the market. I still think that the paper does not “defeat” Easterly and Moyo, as Digital Scotsman said, but the opposite: it shows that aid is not efficient.
By the way… do you have some evaluations of how the cash transfer works in Ethiopia? Particularly, what its effects are at macroeconomic level?
Owen replies: Hi Roberto – These results absolutely do contradict Moyo’s claim that aid reduces the level of GDP. They show that aid increases the level of GDP. I am not aware of Easterly having claimed that aid reduces GDP, though in his book “Elusive Quest for Growth” he has a graph which shows growth falling as aid rises (figure 6.1). Unlike Moyo, Easterly knows that it would be too simplistic to conclude from this that aid reduces growth, though perhaps he wants to leave that impression.
Your point, if I understand you correctly, is that you would like to see a bigger effect. You might be interested to look at the CGD Working Paper which looks at the effect of the subset of aid that is intended to promote economic growth (e.g. excluding humanitarian aid). This paper does indeed find a much higher rates of return. This paper finds that an additional one percent of GDP in this kind of aid produces an additional 0.3 percentage points of annual growth over the four year period that they investigate.
I’m not a great believer in these statistical relationships. But people who claim that there is an inverse correlation, in which aid reduces growth, are themselves employing a highly simplistic form of this form analysis; what these papers show is that their simple analysis is wrong.
George · November 3, 2009 at 6:11 pm
Good to reference Roodman here. Is it too much to quote his conclusions in a review of these studies? IMHO he’s spot on. I believe his background is maths first, than economics/econometrics.
“Each of the papers examined here embodies a set of choices about model specification and data [variable for “aid”, timespan of study, periodisation etc…] All limit the scope of a strict interpretation of the results. A question of great importance for the literature is, how many of such implied assumptions can be dropped without harming the conclusions?
The results reported here suggest that the fragility found in Easterly, Levine, and Roodman (2004) for Burnside and Dollar is the norm in the cross-country aid effectiveness literature.
Does this mean that the various stories of aid effectiveness should be summarily dismissed? Are recipient policies, exogenous economic factors, and post-conflict status irrelevant to aid effectiveness? Are there no diminishing returns to aid? Is helping the neediest countries a hopeless task? No. There can be no doubt that some aid finances investment, and that domestic policies, governance, external conditions, and other factors these authors study influence the productivity of investment. Why then do such stories of aid effectiveness not shine through more clearly? The reasons are several.
Aid is probably not a fundamentally decisive factor for development, not as important as, say, domestic savings, inequality, or governance. Moreover, foreign assistance is not homogeneous. It consists of everything from food aid for famine-struck countries to technical advice on building judiciaries to loans for paving roads. And much aid is poorly used—or, like venture capital, is good bets gone bad. Thus the statistical noise tends to drown out the signal. Perhaps researchers will yet unearth more robust answers to the fundamental questions of aid policy. Or perhaps they have hit the limits of cross-country empirics. Either way, robust, valid generalizations have not and will not come easily. Despite decades of trying, cross-country growth empirics have yet to teach us much about whether and when aid works.”
Owen replies: Thanks, George. You can never have too much Roodman, in my humble opinion. As I hope I have made clear in my posts on this subjects, and in the discussion, I share his view that cross country growth regressions don’t tell us very much at all. But that means that they also don’t tell us what the aid sceptics claim to see in the data: that aid doesn’t work.
The aid sceptics want to have us believe that, although there is much micro evidence of individual aid projects and programmes being successful, that a big macro perspective tells us that there is no meaningful effect of aid on the rate of economic development. (This claim became known as the macro-micro paradox.) But if we believe that aid-growth regressions don’t tell us anything very much, or that to the extent that they do, the statistics seem on the contrary to suggest that aid leads to higher economic growth, then there is no macro-micro paradox, and the main claim of the aid sceptics is false.
Ayana Haile · February 21, 2015 at 4:48 pm
Debate 1: Aid has no direct Impact on the Economic growth. The two main determinants of economic growth are Investment and Export. AID can increase saving if managed well, the increase in saving increases investment. the Increase in investment has two impacts: The increase in GDP and Increase in Export if the Investment is Export Oriented. In this case AID has not direct Impact, even though have an impact through investment.
Debate 2: some argues that AID promotes investment such that GDP= f (I, G, X-M, T), where T may indicates domestic revenue, but G indicates Government Expenditure comes from T and Foreign finance particularly AID. in this case AID has direct impact on GDP, when the government uses foreign Finance as fulfilling Budget deficit through Loans, grants, etc.
Ayana haile, Consultant and Economic Development Advisor
[email protected]
[email protected]
Addis Ababa, Ethiopia
It’s endless…. « Swahili Street · October 30, 2009 at 10:15 am
[…] Barder points us to recent research here, which confirms what he terms “common sense”. Of course aid […]
From Poverty to Power by Duncan Green » Blog Archive » Aid, growth and love; North Korea; people power; sleepless physicists; Afghan fudge and a warmer world map: links I liked · November 2, 2009 at 6:50 am
[…] Barder reports on a new paper that finds that on average, aid promotes economic growth. I share Owen’s doubts about the fixation with regressions, but it’s worth noting that […]
Global Development: Views from the Center » Blog Archive » Yes Bill, No Owen: Why I Still Doubt Aid-Growth Regressions · November 5, 2009 at 4:55 pm
[…] bloggers who spent time at CGD, looked over a new paper and (at least provisionally) reached opposite conclusions about whether aid has at long last been shown to boost economic growth on average. Also […]
All the Views Fit to Print « Waylaid Dialectic · September 24, 2010 at 10:29 am
[…] that there is evidence that aid leads to higher rates of economic growth. For example, here and here). It is true that the methodological issues associated with aid growth regressions mean that such […]
The impact of life-saving interventions on fertility – David Roodman · December 6, 2016 at 5:33 pm
[…] per woman, has fallen much faster than mortality since 1950 in developing countries. A “folk regression” of the fertility trend on the mortality trend would conclude that saving one life prevents […]
Comments are closed.