« You Want to Read This. | Main | Test Blog »

March 09, 2005

Owen on Whether Aid Works.

Owen asked whether we would like to have an informed discussion on whether aid worked, rather than my more normal wild accusations and rhetorical flutters.
OK, let’s do that, I’ll put aside my hyperbole and exaggerations and address myself specifically and directly to the point that he raises, the assertion that aid does increase economic growth as found in a paper by colleagues of his at the Center for Global Development.

As Owen describes it:

On average, aid worth one percent of national income increases annual growth in the recipient country over the medium term by about a quarter of a percentage point a year.

Because this higher level of income is maintained over time, $1 of aid leads to about $1.64 of additional wealth (measured as a net present value) for the recipients.

Or, viewed as an investment in the growth of developing countries, the average rate of return from aid is at least 13%.

It is true that, over the last thirty years, sub-Saharan Africa has received more than $200 billion in aid, and yet the continent has seen negative average growth, of about a quarter of a percent a year. But it does not follow that aid does not work. The analysis shows that, as a result of the aid, average growth was about 1 percentage point higher, each year, than it would have been without aid. In other words, Africans would be substantially worse off today, if it were not for the aid that they have received.

As the abstract from the paper decribes it:

Past research on aid and growth is flawed because it typically examines the impact of aggregate aid on growth over a short period, usually four years, while significant portions of aid are unlikely to affect growth in such a brief time. We divide aid into three categories: (1) emergency and humanitarian aid (likely to be negatively correlated with growth); (2) aid that affects growth only over a long period of time, if at all, such as aid to support democracy, the environment, health, or education (likely to have no relationship to growth over four years); and (3) aid that plausibly could stimulate growth in four years, including budget and balance of payments support, investments in infrastructure, and aid for productive sectors such as agriculture and industry. Our focus is on the third group, which accounts for about 53% of all aid flows. We find a positive, causal relationship between this �short-impact� aid and economic growth (with diminishing returns) over a four-year period. The impact is large: at least two-to-three times larger than in studies using aggregate aid. Even at a conservatively high discount rate, at the mean a $1 increase in short-impact aid raises output (and income) by $1.64 in present value in the typical country. From a different perspective, we find that higher-than-average short-impact aid to sub-Saharan Africa raised per capita growth rates there by about half a percentage point over the growth that would have been achieved by average aid flows. The results are highly statistically significant and stand up to a demanding array of tests, including various specifications, endogeneity structures, and treatment of influential observations. The basic result does not depend crucially on a recipient�s level of income or quality of institutions and policies; we find that short-impact aid causes growth, on average, regardless of these characteristics. However, we find some evidence that the impact on growth is somewhat larger in countries with stronger institutions or longer life expectancies (better health). We also find a significant negative relationship between debt repayments and growth. We make no statement on, and do not attempt to measure, any additional effect on growth from other categories of aid (e.g., emergency assistance or aid that might affect growth over a longer time period); four-year panel regressions are not an appropriate tool to examine those relationships.

I find those two descriptions to be subtly different. Owen that there is shown to be a correlation with long term growth, the authors quite specifically stating that they have only looked at a four year time period. I’m a little confused (and as above, we are trying to have a serious discussion here so please do leap in with corrections) as to how a study that specifically looks at short term results can be said to have found long term effects.

Leaving that little conundrum aside I have a further criticism. The paper specifically looks at aid that could or should have an effect in the four year period, including things like budget assistance and balance of payments support. Fine, but not something I find all that surprising, for think of this type of aid in a slightly different manner.

What is the difference between this type of aid and the government simply running a budget deficit? (In the short term, please, I realise that the long term effects are changed as it is not running up debt, assuming the aid is grants not loans.) Anything? Not so far as I can see, no. OK, do we know of any connection between running such deficits and short term economic growth? Well, yes, for a number of decades (1933 the General Theory came out wasn’t it?) it has been a commonplace of economic thought, that Keynsian deficit spending, pump priming, call it what you will, does indeed provide a short term boost in the growth rate.

The arguments are all over whether it provides a long term boost, something still causing cat fights amongst economists today. My view, for what little that is worth, is that  it does not provide that long term growth, for two reasons.

1) Capacity constraints.  The growth spurt provided by boosting demand lasts until the economy begins to hit capacity constraints or bottlenecks. At this point growth in nominal GDP becomes inflation, not growth in real GDP. There is a further answer at this point, generally known as supply side reforms....in order to remove the bottlenecks, changes need to be made to the underlying structure of the economy. We have noted in the advanced economies that things like privatisation, competition, markets red in tooth and claw work to do this, paternalistic government exercises do not....and also that supply side changes without the demand boosting part also seem to increase growth, both in terms of the possible trend rate and also in actual real growth in GDP. Why not try the supply side changes anyway, adding the aid if necessary? A plan that is 180 o divergent from what is usually proposed.

2) The aid is being spent by governments. Even in the best run states in the world there is little (no?) evidence that government spending boosts growth rates more than an equivalent amount of private spending. Indeed, there is much evidence that private spending on any particular sector provides more bang for the buck than public. Yes, even in education, sewers, water supply and the rest. We also need to add that these are not the best run states in the world, are in fact amongst the most corrupt and badly ruled.

Before anyone leaps in and castigates me for not caring about the poor, allow me to point out that I’m all in favour of growth in poor countries, not least for the entirely selfish reason that it will be good for myself. I am, after all, one of those who think that wealth is something that we create, the more of it the better, not one of those who think it is a fixed amount that has to be shared.
No, my complaints about how aid is raised and then spent are that I simply do not think it is being done in the most efficient manner possible. I want the supply side changes to be made, want it to be private sector, more efficient, actors who allocate it, want the money not to flow through bureaucratic and corrupt channels.

(Very off point but think about tarrifs and trade protection for a moment. We should not have tariffs, quotas or protections against poor country imports for the simple reason that we will be better off if we do not. There’s a view (which I wholeheartedly reject) that poor countries should be able to protect themselves from cheap goods from the rich countries. Despite my rejection of that view I’d even put up with the idea that poor countries should be allowed tariff barriers against goods from the OECD countries....as long as they stopped having tariffs on goods from each other. I still haven’t heard an argument that bolsters Mali having tariffs on imports from Angola, or Tanzania.)

What I find really annoying about Owen’s argument is this idea that yes, sub -Saharan Africa has been shrinking even while getting all this aid, and that the truth is it would have been shrinking even faster without it. There are some countries that get aid of 5-8% of GDP, so by applying the logic in the paper, they should be growing at 10-15% or something, assuming that there is no natural growth at all, only that driven by aid. That economies are still shrinking would imply that the people running those countries are so incompetent that their rule provides a normal growth rate of minus 15%. If that were true then the most obvious thing to do to reduce poverty is to clear out those rulers, try to bring a little competence to the whole process, don’t you think?

It’s OK, the numbers in that paragrah are exaggerated...for I don’t believe the premise in the paper. Deficit spending does provide a short term boost to GDP growth, as we know, Keynes showed that. It doesn’t provide a long term boost for the reasons given in the supply side argument, in the medium term inflation rises. And no, I really don’t think that a study looking only at the effect of such pump priming does much to boost the story of aid providing long term growth, that four year cut off point meaning that they are only measuring the first part, not the second.


March 9, 2005 in Make Poverty History | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Owen on Whether Aid Works.:

» More on aid effectiveness from Our word is our weapon
To his credit, Tim Worstall responds to criticism from myself and Owen Barder of his previous pronouncements on aid with a detailed post taking issue with this paper from the Center for Global Development, put forward by Owen (who works... [Read More]

Tracked on Mar 10, 2005 12:37:26 AM


Tim has this wrong.

1. The phrase I used was "medium term" - which is in my view the right term for a 4 year time horizon. (Remember Nigel Lawson's "Medium Term Fiscal Strategy"?) I did not use the phrase long-term which Tim attributes to me, not because aid does not have long term effects but because they are harder to measure, and the study that I quote does not attempt to quantify them.

2. Tim's analogy with an expansionary fiscal policy is not useful, for at least one simple reason. The crowding out effects, which reduce or eliminate the long term effect of an expansionary fiscal policy, come about because the Government has to borrow or print money to finance the deficit, and/or raise taxes in the future to repay the borrowing. This offsets the expansionary effect of the policy. But in the case of aid-financed financed expenditure, the recipient Governments don't have to borrow or tax to pay for it. It is a transfer from abroad. So there is no crowding out effect to eliminate the long term growth effects.

3. Tim says that economc growth comes from the supply side. He will be pleased to hear that much of the aid is precisely aimed at tackling capacity constraints in developing countries, raising the supply potential of the economy as Tim suggests. That is what investments in education, health, and infrastructure do.

4. The public-private choice is one for the recipient government to make. Many do, in fact, choose to deliver the investment and services through the private sector.

Tim adds:
1) Accepted. My error.
2) Sorry if I wasn’t clear enough. I was trying to state that the effects on growth will be, in the short/medium term, exactly the same as an expansionary fiscal policy...which you agree with. I am therefore not surprised at an increase in spending of 1% of GDP creating growth. I am unconvinced that the financing of the expenditure is the only reason that an expansionary fiscal policy has little or no effect on long term growth. A great deal more convinced that it is the trend growth rate, the capacity for growth, that is, the supply side, which influences long term growth.
3) Investments in education and health were deliberately excluded from the study.
4) Quite. I don’t want corrupt incompetents to have that choice with our money. I want them to be forced into the more efficient delivery mechanism. What they do with their money is up to them.

Posted by: Owen Barder | Mar 9, 2005 5:26:24 PM

The aid looked at in the study includes budget support and balance of payments support, a large part of which is in fact spent on education and health.

I'm somewhat unclear where Tim's argument has got to now. His starting point seemed to be that supply side improvements, not increases in aggregate demand, increase economic growth; and aid can at best temporarily increase aggregate demand and so cannot increase growth in the long run.


a. aid financed expenditure does improve the supply side; and

b. there is robust evidence that aid leads to increases in economic growth sustained for at least 4 years.

Is Tim going to offer some evidence for his view? Or should we take his word for it, despite the analytical arguments and empirical evidence to the contrary?

Finally, Tim's description of recipient country governments as corrupt incompetents is ignorant prejudice. There are some corrupt politicians in the world, both in poor countries and rich countries. But the gross generalisation of this remark is more than an exaggeration: it betrays a deep contempt for the truth, and for fellow human beings doing the best they can in very difficult circumstances.

Posted by: Owen Barder | Mar 9, 2005 7:39:50 PM

Good to see a proper debate on this issue here. A few points on your reponse, Tim:

1. As Owen pointed out, the study is mainly about the short-term impact of short-impact aid. However, particular results from the battery of statistics in the paper indicate that short-impact aid has long-term positive growth effects ("lagged aid is never negatively correlated with current growth, suggesting that indeed the effects of short-impact aid last beyond the current period") AND that 'long-impact' aid (e.g. investment in health and education) has long-term positive growth effects ("the results in Table 5 do suggest an important long-run positive impact on growth from long-impact aid, but this study does not intend or attempt to quantify it"). So the study is saying that aid is definitely good for growth in the short-term and very probably good for growth in the long-term.

2. Your point about capacity constraints seems to have it backwards. As Owen says, aid can do and does reduce capacity constraints, by funding investment in physical and social infrastructure. That's a major role of public spending in any country. There's more to 'supply-side' reform than privatisation, y'know.

3. "Even in the best run states in the world there is little (no?) evidence that government spending boosts growth rates more than an equivalent amount of private spending. Indeed, there is much evidence that private spending on any particular sector provides more bang for the buck than public." Firstly, you're just making unsupported assertions again. Secondly, a lot of aid finances public goods. Your old friend 'basic economics' should tell you that the private sector tends to be pretty bad at investing in these. Thirdly, the poorest countries lack private sector investment in large part because of a lack of public investment in infrastructure, health and education, which is precisely what aid is meant to pay for.

4. "There are some countries that get aid of 5-8% of GDP, so by applying the logic in the paper, they should be growing at 10-15% or something, assuming that there is no natural growth at all, only that driven by aid." You seem to have got the wrong end of the stick there somehow, but fortunately the CGD paper helps out here:

"while average growth in Sub-Saharan Africa was -0.23%, it would have been 0.57% lower, or -0.80%, had the region received world average short-impact aid flows (2.70% of GDP) instead of the amount it actually received (5.33%). Conservatively, then, aid raised average annual per capita income growth in Sub-Saharan Africa by half a percentage point or more between 1973 and 2001. That is, while Africa's growth performance has been disappointing, it would have been much worse in the absence of aid. Here, as elsewhere, we take no account whatsoever of any additional impact on long-run growth due to long-impact aid. Half a percentage point is thus a very conservative estimate of the total impact of aid on growth in Africa".

Correct me if I'm wrong, but I think that's all of your objections taken care of.

Posted by: Jim | Mar 9, 2005 11:32:28 PM

What if the same sums were spent in buying off the political lobbies in the advanced countries who obstruct any possibility of e.g. scrapping agricultural subsidies, scrapping import quotas and levies, etc.? What if the same sums were spent in the advanced countries on research on tropical diseases, or on educating youngsters from aid-recipient countries? It's implausible that handing over cash to the governments of misgoverned hell-holes does more good to the populations than the two suggestions above would do. Are such hypotheses investigated when Aid is considered?

Posted by: dearieme | Mar 10, 2005 11:03:54 AM


You are right that it would be a good idea to scrap trade subsidies. You probably know the relative sizes: agricultural subsidies to farmers in industrialized countries cost about $325 billion a year; total global aid is around $50 billion a year. Unfortunately, it is not clear that more "paying off" the farmers would help: three quarters of the agricultural subsidies are already in the form of "direct payments" to farmers (which enable the farmers to go on producing and selling a loss).

The World Bank estimates that removing all subsidies would result in an additional $250 billion earned each year by the world's agricultural sector -- of which nearly $150 billion would accrue to low- and middle-income countries.

Very large investments are already being made on research in tropical diseases; though we could do more. As it happens, my day job is thinking about aid could be used even more effectively to accelerate R&D into diseases such as malaria and HIV. Have a look at this item if you want to know more about this.

But you are mistaken to stigmatized all developing countries as "misgoverned hellholes". Some poor countries have bad or corrupt governments; many do not. There are some very good governments, including in Africa, doing a difficult job. It is not easy to deliver health services with a budget of just $6 per head per year. We should - and we are - use aid to try to help goverments do a better job (which is one reason why so much of the aid budget is spent on much derided "consultants"). We should try to ensure that aid is not used to prop up bad governments (as it has been in the past - for example, to keep Mobuto in power in Zaire).

But whatever you think about those governments, the evidence is that aid works to reduce poverty. 20,000 people day each day of avoidable conditions as a result of their poverty. We can, and we should, be doing more.

Posted by: Owen Barder | Mar 10, 2005 12:20:07 PM

Thanks for the considered reply, Owen. But who is this "we" who should be doing more? I hope that you're not suggesting that this is a good use of public funds i.e. taxpayers' money? It is surely a classic case where there is no respectable justification for using the coercive power of the state to extort money from the taxpayer to hand over to third parties. Leave it to individuals, say I (knowing that one of my family used to give a preposterously large part of his income this way: OK by me; it was his money).

Posted by: dearieme | Mar 10, 2005 3:37:51 PM

Actually, this is a classic case where we should be using the coordination ability of the state, because reducing povery is a global public good.

Leaving aside the moral issues, which you might argue are a private concern, there are sound public policy reasons for intervention to reduce poverty.

There are enormous risks to all of us if we continue to allow the poorest in the world to fall behind, to our economic and political stability, from migration, the spread of disease, organized crime, deterioration of the environment, drugs and perhaps terrorism. Conversely, there are enormous potential benefits to us from having more affluent trading partners and the mutual benefits of globalisation. These are all public good benefits (or risks) which each individual does not have individual incentive to tackle - each person's own contribution will not make much difference, so they are likely to conclude that they may as well free-ride on everyone else's contributions.


Posted by: Owen Barder | Mar 11, 2005 2:15:50 AM