Our Word is Our Weapon

As the G8 summit draws nearer, the policy announcements start to come in thick and fast. As ever, wait for the in-depth analysis by more sceptical minds rather than believing what the politicians say, but these do look like steps forward at least.

Firstly, there’s George Bush saying he’ll double aid to Africa by 2010. Of course, Congress might decide that they don’t feel that’s a particularly good idea - only this week the House decided to allocate $1.75bn to the Millennium Challenge Account rather than the $3bn Bush asked for. The original proposal was for the MCA to deliver $5bn a year, but at the moment it’s looking like it will eventually end up handing out between a quarter and a third of that. And then there’s the question of when the money will actually be delivered - the MCA was announced in 2002 but so far has only agreed to commit less than a billion dollars to a handful of countries.

To the extent that today’s announcement improves on that record, it is welcome, but to the extent that the extra funds are subject to the same subtractions and pitfalls, we shouldn’t get carried away. The BBC story says “Anti-poverty campaigners said Mr Bush’s pledge was insufficient, while South Africa said it was a step forward”, and I reckon both are right.

Secondly, the ‘Paris Club’ group of rich creditor nations has decided to write off a lot of the debt owed to them by Nigeria. Paul Mason has the details:

* 6bn arrears will be paid off
* A write off of 67% under “Naples Terms” has been agreed
* the rest will be bought back at a discount by the Nigerian government using oil revenues.

Within three years, a source close to the talks said, 100% Nigeria’s debt will be gone.

(Presumably that’s 100% of Nigeria’s debt to the Paris Club, rather than 100% of it’s debt to everyone) Lex Rieffel of the Brookings Institution has a good article giving the background here, and the Center for Global Development has a cornucopia of facts and figures on this page. Having not read any of it, I don’t know how much of a cut in annual debt service this will amount to.

Accepting for a moment the standard creditor position that any debt relief is an act of generosity, does the Nigerian government deserve such treatment? Possibly not, given their still massive corruption, but possibly so, given that they’re apparently somewhat less corrupt than previous regimes and that by appearing to reward such improvement this step may accelerate it. Or they may just rob the place blind again. I look forward to hearing Abiola Lapite’s views anyhow.

Commenter Paddy Carter points out this article by Matthew Lockwood and asks what I think. Well, I think it’s very good.

Lockwood succintly explains one critical reason why, even though achieving the MakePovertyHistory agenda is a necessary step for real development in Africa, it’s not sufficient. Politics in many African countries is polluted by corruption and patronage to an extent that makes Ireland in the Haughey years look relatively clean. In contrast to the bureaucrats who guided South Korea and the other Asian tigers through an incredibly prosperous few decades in the 20th century, the elites in some African countries are more tribal or even parasitical than developmental. As I understand it, the roots of the situation may lie in the ethnically, linguistically and religiously fragmented nature of many African countries, not helped in many cases by the blessing/curse of vast mineral wealth that encourages extractive rather than productive investment. Indeed, many African ’states’ seem to lack genuine ’stateness’ in the sense of the relative cultural homogeneity and administrative reach and effectiveness of European nation-states, so that politics is often about the clash of ethnic groups rather than ideas. At least, that’s the impression I’ve got from this distance, and anyone with better or deeper experience can correct me if they like.

So, if it’s difficult to see how Africa can achieve our levels of prosperity while some of their governments combine ineffectiveness and indifference, what does this say about the MakePovertyHistory campaign? Not necessarily that much, I think. Firstly, corruption and avarice may stop African countries becoming rich, but they don’t explain why they are so extremely poor, since there are countries out there today which are both more corrupt and better off than most of Africa. If the proceeds of aid, debt relief and better trade can be well spent, and the evidence suggests that they can be even in Africa, then the MPH agenda should still be supported.

Secondly, MPH is calling for more aid and debt relief, which in African countries has funded a huge expansion of healthcare and free education - and a healthy, educated population will be more willing and able to hold their governments to account than a sick, uneducated one.

Thirdly, donor governments and international organisations from NGOs to the IMF and UNCTAD are able to monitor where the proceeds of aid and debt are being spent - if there is evidence that aid is being wasted or is making things worse by funding oppressive regimes, it can be cut off.

Fourthly, the state bureaucrats in the Asian tigers were indeed quite skilled at intervening in the right way at the right time, but don’t forget that they were also helped by a favourable external environment. The West - America in particular - had a specific interest in seeing these neighbours of Communist China thrive, and they did everything they could to make that happen, including giving the Asian countries lots of aid, access to their markets, and the freedom to choose their own economic paths. MakePovertyHistory goes some way towards giving Africa a similarly accommodating external environment.

Fifthly, the rest of us have no right to complain about corruption in Africa if we actively abet it. The UK, for example, seems unwilling or unable to deal with firms who bribe foreign officials. And despite signing up to the UN convention against corruption, which includes the requirement that countries will freeze, confiscate and repatriate assets accumulated through corruption, the UK shows no signs of being willing to actually ratify it and be held up to its standards. Here’s one particularly odious example of our fake concern about corruption:

last October the Nigerian government finally abandoned its three-year effort to get UK government assistance in recovering $1.3bn in state funds, known to have passed through some 23 British banks, stolen by former president Abacha. Britain failed to provide the Nigerians with any of the key documents they requested.

Do the serious flaws in some African governments undermine the argument MakePovertyHistory makes against trade liberalisation imposed from outside, as Tim argues elsewhere? Well, while this is at least one of the better arguments against trade protection, I don’t think it’s necessarily decisive. Their corruption does give us a right to withold aid or tell them exactly how to spend it, but it doesn’t give us the right (or the ability) to set their trade policy for them. Also, I think it’s getting harder for governments, even corrupt ones, to enact trade policies that are clearly against the interest of their people, as there’s a lot more scrutiny of developing country trade policies today (from the media, from NGOs, from foreign governments and international institutions), and a lot more advice on trade poilcy should they wish to avail of it (from the same sources).

Paul Mason’s ‘Newsnig8t‘ blog is worth keeping an eye on in the run-up to the G8 summit. Paul’s a reporter for the BBC’s Newsnight and is doing an admirable job tracking all the main issues and the myriad talking heads trying to make themselves heard. One particularly good post is this one on conditionality.

Sorry for not responding to comments here and posts elsewhere, but I’ve been away on holiday. I’ve got the rest of this week off work and amongst other things I intend to work through the backlog so hopefully there’ll be a bit more activity here over the next few days.

Trade liberalisation in poor countries is a qualitatively different ball game from trade liberalisation in rich countries. One reason is that poor countries are often very dependent on the revenue from trade taxes, and when they liberalise they find it very hard to replace this income with revenue from other sources. Here’s some recent evidence.

Over the last 25 years, have countries actually managed to offset reductions in trade tax revenues by increasing their domestic tax revenues? For high-income countries, the answer is clearly ‘yes.’ For middle-income countries, there is also evidence of significant recovery: there are strong signs that this has been in the order of 45–60 cents of additional domestic tax revenue for each dollar of trade tax revenue, with apparently full recovery when separately identifying the episodes in which trade tax revenues fell. For low-income countries, however, recovery has been far from complete. At best, they have on average recovered no more than around 30 cents of each lost dollar. Since many of these countries also face an intense need to enhance revenue to provide sustainable finance for poverty relief and development, and may also face revenue pressures from other sources, the auspices for the prospect and impact of further trade liberalization are troubling.

There’s a new litmus test which you can use to quickly establish whether somebody has anything worthwhile to say about development in Africa. If they say something like

In the last four decades Africa received around $450 billion (or a similarly massive sum) in aid. The result? Africa got poorer. Therefore aid doesn’t work.

then you can switch off the television or, if they’re talking to you face to face, just walk off without a a word, because they’re actually subtracting from the sum of human knowledge and should not be encouraged.

Why is this kind of statement so stupid? Well, two reasons. Firstly, there’s a logical fallacy, in that it assumes that because Africa received aid and got poorer, the aid caused the poverty. But quite a lot of other things were going on at the same time, so the aid need not have caused the poverty - it could even have relieved a potentially much worse situation. And in fact that’s exactly what did happen.

The second reason is that all that $450bn is described as ‘aid’. This statement is usually dragged up to criticise today’s aid-giving, and the implication is that today’s aid is no different from that $450bn. Well, that’s not true. Much of that ‘aid’ was simply bribes the rich countries paid to friendly African regimes not in order to promote development but in order to promote their foreign policy. The ‘aid’ was not expected to be spent as aid, nor was it. So pointing out that it didn’t benefit Africa says nothing about the effectiveness of aid today.

Now that the Cold War is over and Africa is increasingly democratising, there’s much less of a market for this kind of ‘aid’. Aid today is much more likely to be intended and spent for the purposes of promoting development. This is probably one of the main reasons why aid seems to be getting more effective, not less (see page 3 here).

Nevertheless, I am forced to employ this litmus test again and again these days, as more and more people trot out versions of this fallacy. Who do we have to thank? Well, quite a few of them seem to draw their inspiration from William Easterly, a distinguished development economist formerly of the World Bank who is a long-standing ‘aid sceptic’ (to over-simplify for a moment). He has been appearing in the media of late as a kind of anti-Jeff Sachs, arguing that an aid-funded ‘big push’ won’t do much to help Africa. And in amongst a lot of sensible stuff, he keeps trotting out this fallacy. Now, I get the impression that he brings it up not in order to say that we should not give aid, but in order to highlight the importance to ensuring the effectiveness of aid spending that we are very careful indeed about who we give money to and what for, since it is easier, as it was in the past, to simply waste it or give it to crooks.

Well that’s true, but that’s not how his figures are being used being used, and I wonder how he feels about lending legitimacy to a host of cranks who oppose aid per se.

While manually deleting mountains of quarantined comment spam, I accidentally deleted a half dozen recent genuine comments. Sorry about that. I’ve tried restoring them, but for some reason it only worked a couple of times, and now I’m heartily sick of Movable Type’s idiosyncracies so I’m giving up on it.

[edited 30th June 2005]
The free-market ‘International Policy Network’ (what grand titles these people give themselves) says its new study shows that “foreign aid does more harm than good”. But it cherry-picks only the scanty evidence that supports the IPN case, and distorts even that to produce the desired result.

The full study, “Aid and development: will it work this time?” is here, and its author is Fredrik Erixon of the Swedish think-tank Timbro. His methods are not exactly sophisticated: claiming to analyse the “historical impact and effectiveness of aid”, his first gambit is to show us the graph below and declare that it proves there is no linear or even positive relationship between aid and growth.

There’s two things wrong with this kind of analysis. First, since income is the numerator of one indicator (GDP per capita growth) and the denominator of the other (Aid/GNI), it would be surprising if there was a linear relationship. It might be more revealing to compare growth with aid per capita (I’ll come back to this later).

Second, it’s perfectly reasonable for some aid to be negatively correlated with growth, since for humanitarian reasons aid might go up in the bad times only to fall when a country is growing strongly again. This does not mean that aid is bad for growth - indeed, one interpretation of the graph could be that the growth in aid halted and then reversed the decline in growth rates. But that would be reading far too much into a deliberately simplified picture - nobody pretends that aid is the only significant determinant of growth rates, except, it seems, when they are trying to prove that it is no good at all.

Erixon’s next step is to point to William Easterly’s analysis of links between aid, investment and growth in poor countries between 1965 and 1995, which found that only 6 out of 88 countries “experienced a significantand positive effect of foreign aid on growth”. True, but again this is hardly surprising when you look at Easterly’s study - he just compared thirty-year averages of aid/gdp and investment/gdp, and since some aid will always go to current expenditure rather than investment, and some (like humanitarian aid) will tend to be negatively correlated with growth, and since a particularly large proportion of Cold War era aid was simply used to pay off friendly dictators, the absence of a positive relationship is hardly surprising.

But what’s most interesting is that Easterly admits as much, saying: “I do not intend here to make a general statement about whether foreign aid is effective … It could be that in any given country that there was an adverse shock like a drought that caused investment to fall and aid to increase”. Yet Erixon not only doesn’t inform us of Easterly’s caveat, he dishonestly chooses to draw precisely the conclusion that Easterly refused to - that aid is ineffective.

Next, Erixon moves on to look at studies that attempt to isolate the effects on growth of aid from other significant variables. Well, he looks mainly at one study (by Peter Boone), which is nearly ten years old and which has since been superceded and roundly dismissed by some (Hansen and Tarp say of Boone that at one point “he decides, in passing, to discard the result that aid does have an impact on investment when his full sample is used”).

In a footnote, Erixon also cites research by that well-known fount of knowledge the Cato Institute, and remarks that he knows of “only one empirical study claiming that aid has definitely had a positive effect on growth” - this one.

Correction: that’s not quite accurate. Erixon actually says that he knows of only one study claiming that aid has definitely had a positive effect on growth and that the effectiveness is not determined by the policy conditions in the recipient countries.

Well, that’s truly an astonishing claim, because I’m just an amateur and I know of several. One that has received a lot of coverage - but not enough for Erixon to notice, apparently - is this one by the Center for Global Development, which deliberately separates out the humanitarian aid that should be negatively correlated with growth and finds a strong, significant effect of aid on growth in the short term (quite apart from its long-term effects through improved health and education).

In fact, directly contrary to what Erixon and the IPN would have us believe, studies that point to a negative or zero effect of aid on growth are now in the minority. A very recent survey of the literature by Mark McGillivray finds “overwhelming evidence that aid increases growth and other poverty-relevant variables”. McGillivray says that the “the clear, unambiguous finding of practically all empirical studies conducted over the last seven or eight years” is that “Aid now appears to work in the sense that per capita economic growth would have been lower in its absence”.

Note: McGillivray also lists twenty-five studies (the oldest from 1998) apart from the CGD one that specifically “suggest that aid works in countries irrespective of the quality of policy regime”.

McGillivray says that the new optimisim on aid in recent years may be down to either or both of an increased effectiveness of aid or improved analytic methods.

So the question must be: is Erixon unaware of this “overwhelming evidence” (in which case he’s not really competent), or is he just choosing not to tell us about it (in which case he’s being dishonest)?

But wait, there’s more: Erixon says that “many scholars” have observed that in Africa “the savings ratio has actually fallen when aid has increased”. According to his footnote, these “many scholars” consist of the five authors of a ten-years-old IMF working paper and nobody else. But once again, he is contradicted by more recent research, namely this detailed World Bank study, which says:

Cross-country time-series data from Africa suggest that higher foreign aid tends to reduce national saving (that is, a large fraction of aid is consumed). But this evidence may reflect the fact that aid to poor countries increases at times of adverse income shocks, when saving is lowest … Scrutiny of the countries that have moved from low to high savings rates reveals that increases in foreign aid are positively associated with takeoffs of both private and national saving.

(Emphasis added)

There’s a good deal more to read in Erixon’s paper, and it’s not all bad - there’s some interesting detail on the differing fortunes in the 1980s and 1990s of Kenya, Tanzania, Uganda and Botswana, for example. But there’s also another blatant attempt to mislead: we are meant to conclude that Botswana’s stellar growth performance in the last thirty years is pretty much entirely down to good policy choices, in stark contrast to Kenya and Tanzania, who had a terrible time of it over the same period and were “both large recipients of aid”. Now, I’d say it’s quite likely that Botswana did make better policy choices than Kenya and Tanzania, but is it right to conclude that aid played no part in Botswana’s success and a big part in Kenya and Tanzania’s troubles? The graph below should tell you: it compares aid per person in Botswana, Kenya and Tanzania between 1960 and 2001 (the most recent year for which I’ve got OECD aid data and population figures available).

Yes, unless I’ve got these figures completely wrong, then Botswana received much higher levels of aid per person for about thirty years. Looking at aid as a proportion of GDP will tend to downplay this, since Botswana was throughout a significantly wealthier country. Over the forty years, Botswana received almost five billion dollars in today’s terms, and a graph (below) of aid per capita compared to GDP per capita growth shows a far more ambiguous (some might say even positive) relationship than that suggested by Erixon.

So no, I’m very far from convinced that aid “does more harm than good”.

Another damn regression analysis to chew over, but with an interesting focus:

Does aid contribute to human development other than by increasing growth? In doing so, is aid more or less effective in poorer countries (those with low levels of aggregate welfare)? This paper addresses these issues, assessing if there is cross-country aggregate evidence for an effect of aid on welfare levels. We posit that aid can enhance human development by financing public expenditures that increase welfare indicators. Using quantile regressions, we report evidence that aid is associated with higher human development (the Human Development Index) and lower infant mortality (both indicators of aggregate welfare). Where there are differences across quantiles, aid is more effective in countries below the median of the welfare distribution, i.e. with lower levels of human development. Insofar as aggregate welfare is (inversely) correlated with poverty, we find evidence that aid can make a positive contribution to alleviating poverty, and that the effect appears to be greater in countries with lower levels of human development indicators.

In this superb* post, Owen Barder lists and dismantles 13 common fallacies about aid and development. Two highlights for me:

Fallacy 5. Extra aid in Africa could not be absorbed: it would just be wasted.
Aid to Africa per person halved in the 1990s. There is no reason to believe that if twice as much aid could be fairly well spent then – and the evidence is that, on average, it was – that it couldn’t be well spent again now. If donors made a serious effort to make it easier for recipient countries to use aid productively, much more could be used to benefit the world’s poor.

Fallacy 6. We’ve already pumped billions of dollars into Africa and it we have nothing to show for it. So we know it doesn’t work.
Jeffrey Sachs tells a story about a village fighting a forest fire. They send their fire engine, which tries valiantly to douse the flames. But the fire continues to burn. Should the village conclude that fire fighting is ineffective, or should they conclude that they need more fire engines? There is no question that some aid money has been wasted – especially the politically motivated aid to badly governed countries – but that only makes it even more remarkable that on average aid is so effective. The aid that is well spent must be making a huge difference, given that it more than offsets the mistakes

*And not just because it says nice things about me.

Sorry if you’re seeing some comment spam on this site - the MT Blacklist filter blocks hundreds of comments a day, but in the last couple of days some spam has been sneaking through. I can’t do much about it during the day, but I’ll try to delete any that has got through each evening.

Oh alright then, I’ll do the book meme thing, but this isn’t going to be very good.

Number of Books I Own
About three hundred, (of which about fifty are Tintin and Asterix books). That doesn’t sound like many, I know, but it’s more than most people I know of my age - we like reading, we just don’t seem to buy our own books. I must have borrowed hundreds more from friends and libraries.

Last Book I Bought
Night Flight for the Little Red Train, by Benedict Blathwayt
Last book for me? The Evolution of the British Welfare State, by Derek Fraser.
Last book bought I intend to actually read through rather than just refer to? The Autobiography of Malcolm X

Last Book Read
Last book read and finished: The 1929 Crash, by J. K. Galbraith.

I almost finished ‘Brick: A World History’, but got tired of lifting it.

Five Books That Meant A Lot To Me
Catch 22, by Joseph Heller
V for Vendetta, by Alan Moore and David Lloyd
Revolution in the Head, by Ian MacDonald
The Quiet American, by Graham Greene
We Wish To Inform You That Tomorrow We Will Be Killed With Our Families, by Philip Gourevitch

Close runners-up include the Last Temptation by Nikos Kazantzakes, Picture This by Joseph Heller, Woody Allen’s Complete Prose and Matt Groening’s Life in Hell. In terms of economics, politics or sociology, I’ve read any number of brilliant articles or chapters, but I can’t think of anything that stands out as a great self-contained book.

Passing this on to
I suppose I’m meant to pass it on to five bloggers, but quite apart from a reluctance to spread what’s basically a chain letter, I don’t actually know five bloggers. So I’ll pass it on to two I know quite well, Tom and Martin, and to James R MacLean, because I know his answers, if he even takes it up, will be a lot more interesting than mine.

The Globalization Institute’s campaign against aid for schools, hospitals and better transport links in Africa continues apace, with a new post delightedly reporting that “British say foreign aid is wasted”. As usual, they’re twisting the truth to fit their version of reality.

The post is based on a poll for the Telegraph carried out by YouGov. The full findings are here and are discussed in this Telegraph article.

Contrary to post author Alex Singleton’s claim, the poll findings don’t actually reveal anything about British people’s views of how aid is currently spent. Rather, those surveyed were asked how confident they were that any extra aid sent to Africa would be “spent wisely, rather than being wasted or finding its way into the pockets of criminals and corrupt governments”. In response, 1% said they were “very confident”, 10% fairly confident, 41% not very confident, 42% not at all confident, and 7% didn’t know.

Does this poll mean the British public think we should be spending less on aid to Africa, as the Globalization Institute would prefer? No. In fact, a majority (52% of the total, 57% of those who expressed an opinion) agreed with another statement put to them that “Africans can solve their problems but only if they receive financial and other assistance from rich countries”. The Telegraph report glides over this finding, while the GI doesn’t mention it at all. Other findings supporting the MakePovertyHistory campaign were that a huge majority agreed that it was the responsibility of “a partnership between African and rich countries” to solve African problems, and that almost twice as many supported the statement that “most or all” of Africa’s large debts should be cancelled.

Lastly, the most significant thing about the poll for the Telegraph is the question which wasn’t asked. In a previous YouGov poll for Channel 4, 85% of people overestimated the proportion of national income we actually spend on aid, and when the true amount (0.34% at the time) was revealed, twice as many thought it was too little as thought it was too much. Surely asking this question is essential for getting any true picture of the public’s attitude to aid? Yet the Telegraph didn’t want it asked. I wonder why.

Tim Worstall has said some crazy things about development in the past, but this might just take the biscuit.

Writing in the Guardian, George Monbiot criticised enforced trade liberalisation in poor countries on the grounds that

As the World Bank’s own figures show, across the 20 years (1960-80) before it and the IMF started introducing strict conditions on the countries that accepted their loans, median annual growth in developing countries was 2.5%. In the 18 years after (1980-1998), it was 0.0%.

Tim responds:

There�s just one tiny problem with this view (apart from it being totally wrong). Under the Lome Convention and successors, since 1975 the poor countries have had preferential access to EU markets without having to provide free or preferential access to their own in return .. Precisely in that time period that the poor countries did not have to open up their internal markets median growth was zero.

There’s just one very big problem with this view: the Lome convention was made fairly irrelevant in the 1980s and 1990s by the structural adjustment programs imposed on poor countries by the IMF and World Bank in exchange for vital loans, programs which usually involved significant trade liberalisation. That’s why the 1980 to 1998 period is well known to everybody except Tim as one of unprecedented trade liberalisation in poor countries. For anyone writing about development to claim that during this time poor countries didn’t open their markets is, to use Tim’s word, “breathtaking”.

The Globalization Institute says that there is no need for public spending on transport in developing countries because the private sector developed an extensive rail network in Britain in the early 19th century while it was still, by today’s standards, a developing country.

So what they’re basically saying is that there’s no significant difference between the transport economics of Britain of the early 19C and, say, Ethiopia. This is completely wrong. Why?

• Wealth: 19th Century Britain was wealthier than Ethiopia is today. According to Angus Maddison, the UK in 1820 had a per capita income of $1,706 in 1990 US dollars, compared to a 2001 per capita income in Ethiopia and Eritrea of $660. So Britain (assuming the same level of income as the UK) was about 2.6 times richer.
• Size: Ethiopia is four times the size of the UK.
• Population density: In 1820, Maddison says the UK had a population of 21.2m people, translating (assuming land area didn’t change much) into around 88 people per sq km. The population density in Ethiopia today is around 53 people per sq km.
• Terrain: Compared to Britain, Ethiopia has a pretty unforgiving terrain for railway-building, what with a massive mountain range slap bang in the middle.
• Climate: Britain has and had in 1820 a temperate climate, in which building a railroad probably seemed like a refreshing outdoor activity. In contrast, Ethiopia can get really hot and working out in the desert or mountains is consequently a lot more difficult.

So no, once you actually compare the two countries I don’t think the private sector would be quite as keen to build railways all over poor, massive, sparsely populated, mountainous and hot Ethiopia as they were in the rich, small, dense, relatively flat and temperate Britain of the early 19th century. And Ethiopia isn’t such an extreme example, either: look at most African countries and they have similar issues especially in terms of wealth, density, terrain and climate (though many are tropical rather than arid, which brings its own problems). And guess what, it turns out that in many cases the private sector isn’t that keen, so government does have to stump up the money. Of course, being swivel-eyed ideologues the Globalization Institute really hate this, which is why they feel the need to talk such complete shite all the time.

All of which is just a simple illustration of what should be the simplest lesson in all of development economics: not all countries are the same. In particular, poor countries tend to be different from rich ones. You’d think that would be obvious, but it’s obviously not obvious enough for some people.