Dodgy, dodgy, statistics

Of course, we\’d like to know whether official aid helps. And whether it crowds out private sector investment. So we get these numbers from The Guardian:

On average, in these countries aid flows increased threefold between 2000 and 2010, but there was no sign that official assistance crowded out private investment. Far from it, foreign direct investment increased fourfold over the same period. Growth averaged 5.5% between 2000 and 2011, an impressive performance given the meltdown in the global economy that followed the financial crisis of 2007.

Glory Be! Private sector aid follows public! Quick, right that cheque on the taxpayers\’ account!

Err, hang on a bit.

Africa is in the midst of an economic boom. Foreign Direct Investment [FDI] in Sub-Saharan Africa grew from US$9bn in 2000 to over $62bn in 2008, with Ernest & Young capital investment forecast of $150bn by 2015.

You mean that those countries that got that lovely aid got less private investment than those that did not?

This is evidence that there is no crowding out how exactly?

BTW, the general growth rate in sub-Saharan Africa over this period was 5.4%.

We\’ve not got any great evidence in favour of official foreign aid here, have we?

11 comments on “Dodgy, dodgy, statistics

  1. Tim

    You are right that the statistics which Larry Elliott cites do not show that aid works (any more than the statistics which Dambisa Moyo cites show that it does not work). You cannot simply make pronouncements about receiving aid and investment or growth and conclude that one causes, or prevents, the other.

    But that perfectly reasonable point does not get you to your headline ‘dodgy, dodgy statistics’. The numbers quoted by Elliott, sourced to ONE, sound broadly accurate to me, even if Elliott then tries to put more weight on them than they will bear.

    The statistical problem is that you would EXPECT aid to go mainly to countries that are, for other reasons, growing more slowly than average. (Nothing wrong with giving aid to countries that need it most). So the correlation between aid and growth could be positive or negative, whether or not aid works.

    Here is a paper that does it properly. Peer reviewed and all that (and without revealing too many secrets, about to win a prestigious econometric award):

    When you do the statistics properly, they show that aid does indeed lead to economic growth.

  2. The correlation is modest, the effects vary a lot depending on where you spend the aid, and high levels of aid help less and less.

    And so you have proven your point, OB?

  3. @’Bloke in France’

    Absolutely: the effect depends on a lot of things (including how and where the money is spent) and the study I cited does indeed suggest diminishing returns. I don’t agree that the effect is ‘modest’, however – the average returns are pretty good once you take account of all these effects.

    My own views are somewhat less positive than this particular study implies. If you are interested, they are set out here:


  4. @ Owen Barder
    (i) NO, you would, unless you were an idiot or thnk that DfID is populated by idiots. expect aid to go to the countries where it would generate most growth
    (ii) why do you wear lipstick for your website photo?

  5. @john77

    (1) it not clear that the utility function that DFID’s political masters are seeking to maximise is the total growth generated by aid. So DFID staff, who are neither idiots nor political decision-makers, allocate aid to achieve a more complex utility function than the one you describe;

    (2) if DFID were merely trying to maximise the growth impact of aid, which it is not, that would not necessarily be inconsistent with spending aid in poor and slow-growing economies. It is possible that the marginal returns to aid are higher in poor countries and/or slow growing economies than elsewhere. So even if they agreed with your simplistic and one-dimensional view of their objectives, they might well do as suggested as in my original comment with which you purport to disagree.

    I won’t dignify your personal remarks about my appearance with a response.

  6. @ Owen Barder
    The clue is in the name: Department for International Development. Their remit is to reduce the sum of human unhappiness by fostering economic growth in “developing countries” because poverty is the leading cause of hunger, sickness, unsafe practices that cause injury, disregard for the environment. The DfID’s programmes are multi-dimensional including education and health as well as the very occasional direct investment in agriculture or manufacturing but these are a means to an end of reducing want which can be measured in terms of increased GDP/head. So my single data-point is a summary of their multi-dimensional targets.
    DfID’s current political master is called David Cameron. He has shown no sign of having a more complex utility function for DfID (that is complex enough) and has rejected calls for a more selfish attitude even at the cost of popularity and votes.
    On its website, DfID says “We know economic growth is the most effective way of raising incomes and permanently lifting people out of poverty”
    The UK has a small budget for DfID – Cameron is trying to raise it to 0.7% of GDP – so the best use of it involves allocating it to those projects where it will have greatest impact. QED

  7. @john77

    a. DFID’s objective is not only economic growth – for example, they are concerned with humanitarian relief, sustainability, rights, protection of minorities and marginal communities and the absolute poorest.

    b. if the concern were growth alone, and the utility function were simply to maximise the total sum of all growth created, then presumably DFID would spend ALL aid in the one country in which aid was predicted to have the biggest impact on growth. (The normal models of the effects of aid on poverty suggest this one country would be India.) In fact the government spreads aid across many countries, which suggests that their utility function is not to maximise the simple sum of all global growth: they are concerned with its geographic distribution too. So even if they are concerned with growth, the utility function they are maximising is not the simple sum of that growth across the world.

    Tim adds: Yes….but.

    That same observed fact, spreading the aid around the world, can be explained by a different utility function. Imagine, just imagine, that the purpose of aid is to purchase political friends for Britain by corrupting politicians. You would also spread the cash around in this manner of that were your purpose: bribing the maximum number of people with the limited cash available.

    And I’m entirely unwilling to insist that there is none of this in it…..that study about how ODA committments increase to countries who get themselves voted on to the Security Council as rotating members….

  8. @ Owen Barder
    Humanitarian relief is carried out, with more money and less waste, by the DEC and a range of specialist charities who do not have to deal with government bureaucracy.
    DfID’s first goal is to reduce extreme poverty which is a prerequisite for any of its secondary goals. Their list of targets differs greatly from yours, starting with eradicating extreme poverty, followed by primary education, gender equality (too high up the list in my view), health (including child mortality which is more important than the number of women legislators) and seventh environmental sustainability which can only be a concern once they have overcome starvation.
    *You* say their concernms are “humanitarian relief, sustainability, rights, protection of minorities and marginal communities”s and, last “the absolute poorest” whereas DfId put those first and exclude four out of your first five. Are you calling them liars?

  9. @ Owen Barder
    Your second implies massive ignorance – you must think I am stupid.
    Producing a list of projects that will generate the greatest benefit per £ spent is not going to be all in country A followed by all those in country B, followed by all those in country C etc
    It will. however, be weighted towards those countries or regions within countries that have the greatest shortage of capital for investment since these will, generally, offer the highest return on investment (however you measure returns whether in money terms or reduction in child mortality or …); these areas also tend to be those that are poorest so the GDP growth rate will be most affected.
    India is NOT the country with the least access to capital for investment so there must be questions about your model.
    Rational private donors give to charities that create sustainable, rather than “sustainable”, improvements: simple irrigation systems, converting manure and sewage into fertiliser, small-scale hydro-electricity harnessing streams and small rivers etc.

  10. @Tim – Agreed. Revealed preference is that the government’s utility function is not to maximise the simple sum of global growth. That’s the point I’ve been making all along.

    @john77 – I am now unclear if you are making statements about what you think DFID’s objective function is, or ought to be. There was no implied ordering in my list of objectives, nor did I intend it to be complete. The debate began with your assertion that you “expect aid to go to the countries where it would generate most growth”. I assume that on reflection you would now agree that this is not so.

    Furthermore my original point – to which you took exception – was that one would expect aid to go to countries that are poor or growing slowly (which is why you might not see a strong simplistic correlation between aid and economic growth). This is no more controversial than saying that you would expect medicines to be administered mainly to sick people or that fire brigades tend to work in and around burning buildings. That point is not inconsistent with your opinion (that aid should go to countries where it has the biggest effect on growth) – just as medicines are likely to have the biggest effect on sick people. So your view about where aid should go is not, as you seem to think, a rebuttal to my point that aid tends to go to poor and slow growing countries.

    Finally, I find your throwing around of insults (lipstick, idiots, massive ignorance etc) a bit wearing.

  11. @ Owen Barder
    You replies are full of suggestio falsi
    When did I say “country” instead of “countries”?
    NO. You are talking nonsense: Tim is arguing about the rate of growth AFTER, not BEFORE, the aid is disbursed. It might, or might not (the criteria are unrelated) be appropriate to distribute Aid to those countries that had low economic growth in the past: what Tim is looking at is the growth *after* the aid is disbursed.
    So I should have said places? Maybe – but Tim was illustrating his argument on “crowding out” (which, incidentally I find”not proven”) with statistics on countries which you followed up on a “country” basis. It’s a bit rich to condemn me for discussing the question under the terms you agreed before I read the post!!
    The massive ignorance is what you ascribed to me by putting forth your previous argument: yes, it is a bit wearing. As is your assumption that I don’t know what DfID stands for and cannot read its website. More suggestio falsi: I asked whether you thought the DfID was staffed by idiots. That is one reason why I got annoyed. If you want to avoid criticism, stop treating those putting alternative arguments as idiots. We can read and read that DfID sets out, mostly reasonable, aims that are NOT what you say they are.
    You claim to be unclear as to what I claim by quoting data from DfID website. OK I claim that you need to state either you claim that they are lying or that you accept that they are telling the truth and you are wrong.
    Tim’s cynical comment doesn’t work for Cameron whose support for the Aid budget goes against all sensible political calculations and who seems incapable of bribing foreign politicians.
    You are just plain wrong. Aid should generally go to those areas that are (among the) poorest but AFTER they have received the aid, their growth should be (among the) fastest. If the post-aid growth is below-average, there is something seriously wrong.

Leave a Reply

Name and email are required. Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>