A lot or a little ?

by John Q on December 21, 2007

A dollar is not very much money. A billion dollars is a lot of money. Twenty billion dollars is an awful lot of money.

For most people reading this (though not for Bill Gates or for the billion or so people living on a US dollar a day or less), these statements should seem pretty obvious.

But all of these can be (and have been used as) different ways of measuring the same thing. If every Australian receives, or pays, a dollar a week, the total amount is very close to a billion dollars a year. And if you have a cash flow of a billion dollars a year, and your interest rate is 5 per cent, the present value of that cash flow (the amount of extra wealth you would need to generate the flow) is twenty billion dollars.

It’s easy to stretch this gap even further. A dollar a week is about fourteen cents a day. And, if we looked at the US (about 300 million people), or the entire developed world (around a billion people, depending on your definition), the total would be that much larger. Fourteen cents a day for everyone in the developed world has a present value of one trillion dollars.

The fact that the same flow of money can be presented in such radically different ways, and that each of them is appropriate in certain contexts, is one reason public policy debates get confused.

I’ll start with one illustration. It’s commonly observed that despite receiving over $500 billion in aid in the 50 years since the shift to independence, Africa is still poor and, on the whole getting poorer. The implication is that the aid must have been wasted or stolen, and of course, quite a bit of it has been. But an aggregate number over 50 years isn’t very helpful.

Much more useful in thinking about the likely impact of aid is the amount per person per week. With (very roughly) a billion people in Africa and a billion in the developed world, the aid that’s been given amounts to about $10 per person per year, or 20 cents per person per week on each side of the transfer. So, the implied complaint of the average Northerner to the average African can be translated “I’ve been giving you 20 cents a week for years now, and you’re still poor – you must have squandered my generous help”.

This doesn’t answer the question of why Africa has remained poor while so many in Asia have grown rich, or at least much better off. But unless you impute truly magical rates of return to aid, the question “why hasn’t aid made Africans rich” can be answered very simply “there’s never been enough to make a difference”.



chris armstrong 12.21.07 at 11:08 am

That’s an interesting point you’re making about the averaged ‘receipts’ of Africans over recent decades. I wonder if you could easily calculate how much the average citizen of a state like the UK has contributed weekly in official aid over a similar period? You’re right that the attitude of ‘we’ve given x already, and it has been squandered’ is pretty common. I guess these figures don’t touch the question of whether squandering has occurred or not, but it is interesting to see the ‘x’ represented in different ways.


Hidari 12.21.07 at 11:28 am

A very quick google search brings up this article. I have highlighted the points I find particularly important.

‘The evidence that aid can transform whole societies and lift millions out of poverty is unconvincing.

It can only speed up a process that is already happening.

When we see scenes of destitution from Africa we assume that we can change things by sending money.

But if aid could make Africa prosperous, it would have done so by now.

Despite nearly a trillion dollars of aid since independence in the 1960s, much of Africa is worse off now than it was then.

Much of that aid was spent by outsiders without consultation with Africans and with little understanding of Africa’s ways or needs.

….Aid creates and sustains unequal relationships. Talk of partnerships is false.

As one western diplomat in Africa put it to me recently: “We like it when they take ownership of the programme but we mean our programme. We don’t like it if they start having their own ideas.”

The aid business is an industry with its own dynamic.

Much of it is spent in the donor countries in the form of consultancies and goods.

The author lists a few things ‘we’ can do that would be better than aid, including:

‘The British government could … do something about arms flowing into Africa.

Weapons that kill in Africa’s wars may not be made in Britain but arms dealers base their operations in London and are largely unregulated.

And London has become the laundry of choice for money laundering because Britain applies one of the weakest banking scrutiny systems in the world.’

Dowden also points out that ‘we’ do not, of course give aid equally to every single country: ‘our’ aid is highly targeted and not always for the best. Giving aid to countries like S.A. or Botswana is simply a waste of time: they don’t need it. On the other hand, it can be argued that giving aid (as opposed to short term humanitarian relief) to ‘countries’ like the Congo or Sudan is also a waste of time; what these countries need is international pressure to stop the wars raging there.

And even this analysis ignores the rigged ‘free market’ (systematically biased in favour of ‘the west’), the power impact of having local economies dependent on far richer ‘Western’ countries (‘(countries like) Mozambique, Tanzania and Uganda are already receiving around 50% of their budgets from aid. That makes them more dependent on Western aid donors than they were in colonial times.’), and the mainly disastrous effect of the IMF etc.


John Quiggin 12.21.07 at 12:27 pm

“But if aid could make Africa prosperous, it would have done so by now.”

As the post shows, this premise is false. A more truthful statement would be

“If the trivial amounts of aid we are prepared to give could make Africa prosperous, they would have done so by now.”


Planeshift 12.21.07 at 1:18 pm

We also need additional information to put the figures in context. How much money has Africa paid us in debt repayments, what has the been the value of the natural resources stolen/aquired at exploitative rates, what about the value we could hypothetically attach to the production of the rainforests in terms of oxygen and absorbing CO2 etc.


chris armstrong 12.21.07 at 1:22 pm

Could we suggest that for the next A Level Maths exam?


abb1 12.21.07 at 1:40 pm

Yeah, me too, my gut says: stop robbing them and they’ll do OK without any charity.


Barry 12.21.07 at 2:01 pm

How much of the money was deposited in one swiss bank account, and immediately moved to another?


P O'Neill 12.21.07 at 2:28 pm

What was Marshall Plan aid per head in Europe over its duration?


Random African 12.21.07 at 2:53 pm

There’s also another interesting question:

How much of it really qualifies as aid ?
Between the early 70’s and the mid-to-late 80’s, Somalia received billions of dollars from the USSR and then the USA but it was military aid and it was to fight Ethiopia.
Or one can talk about the expectations of anyone who cut a check to Zaire (two goals: helping him stay in power so that the commies won’t take over or support various anti-communist rebellions in neighbouring countries via him.


Quo Vadis 12.21.07 at 3:22 pm

“why hasn’t aid made Africans rich” can be answered very simply “there’s never been enough to make a difference”

I would have expected a rather different answer from a dedicated social democrat.


Dan Simon (troll) 12.21.07 at 3:44 pm

nd hr thght tht whn dvrtsrs s lns lk, “fr jst pnns dy…”, thy’r jst bng mnpltv nd dcptv. Lttl dd knw tht f thy’d bn strln cnmcs prfssrs pddlng d schms, nstd f, sy, nsrnc cmpns pddlng ndlss r xcssv cvrg, thn thr crfty wrdsmthng wld hv qlfd s Dp Thght bt Pblc Plcy…


stuart 12.21.07 at 3:48 pm

P O’Niell: As an example the UK had a population of about 50 million around 1950, and got $3.3 billion over 3 years of the plan. Which is 42c per week per person (inflation would convert that to $3.60 in current prices). Of course this was a country that already had full industrialisation, etc., and just needed to reconvert from war production mode and rebuild some of the southern towns/cities and infrastructure, rather than starting from scratch.

Of course the Marshall plan covered a lot of countries, but a number of them got only fairly small amounts, so no doubt if you took the time to work it out across the entire of Europe it would be lower. For example Turkey received 137 million over three years with a population of 21 million, which is only 4c per week per person (in 1950 dollars).


Walt 12.21.07 at 4:56 pm

Dan, I don’t know how to put this politely, but you’re a weirdo. John puts up a post that’s as uncontroversial as possible, and you find something to quibble over.


almostinfamous 12.21.07 at 5:42 pm

of course one has to qualify where in ‘Africa’ the aid has gone, as it is a large continent and not a single nation.


JLS 12.21.07 at 6:07 pm

About Africa.

first I think that corruption is not a problem like the lack of democracy.
In China there is corruption and no democracy and the country thriving like South Korea in the 80s.

The problem of Africa are :
a) a low level of education, especially technical level.
b) High birth-rate, start to decline but still very high.
c) and not enough work, an African farmer work about 800 hours a year, in Far East people work 2000 hours a year (in manufacturing, in farm much more I think).


notsneaky 12.21.07 at 6:52 pm

John is right that the amount of aid to Africa has been small. But this is a bit like just using one observation to draw an inference when you can use a whole bunch. The ‘aid doesn’t do much for growth’ argument is more than just the ‘if it did it would have’ line in #2. Different countries in Africa (and around the world) received different amounts of aid. So while on average it might have been 20 cents a week, there are places where it was much higher and places where it was much lower. So the question to ask, is there any difference between the economic performance of those places with higher aid per person per week and those with lower (controlling for other stuff and the obvious reverse causality). And the answer seems to be still ‘no’ (*)- this is essentially the William Easterly line (who BTW, is not opposed to aid, just how it’s given out and administered. And also argues that aid just cannot be a panacea for the economic malaise of Africa).

(*) – It might be a weak ‘yes’ if you buy the Burnside and Dollar result that aid has been somewhat effective in places with better economic policy.

Oh, and debt repayments. Why should they be subtracted off? Unless we count the initial loans as ‘aid’ as well. In fact in some cases aid has been done through debt forgiveness or well-below market interest rates on loans.


Kenny Easwaran 12.21.07 at 6:54 pm

I thought the population of Africa was actually much lower, like 500 million – Nigeria, South Africa, and Egypt are the only countries with more than about 30 million I believe. At any rate, this doesn’t change the basic point of the post – 40 cents a week is hardly more likely to take someone out of poverty than 20 cents a week.

But you still might think that 40 cents a week per person, if spent on the right sort of infrastructure and governmental reform and the like, could have a much larger impact. It’s just that the money hasn’t been spent that way, and has instead stunted agricultural growth and the like.


Kenny Easwaran 12.21.07 at 6:58 pm

Alright, Wikipedia says “over 900 million” and this site says “over 680 million”. I think I got my numbers by adding up the populations of countries in 2000 according to some fact book I found several years ago – it’s possible that some of those numbers were wrong, that I missed a few countries, and that the population has grown substantially (at a 3% growth rate per year, that is actually quite significant over 7 years).


Random African 12.21.07 at 7:02 pm

700 millions.. DRC, Ethiopia, Sudan, South Africa, Tanzania, Morocco, Kenya and Algeria also have over 30 millions and Uganda is getting close.

Actually, 700 millions is subsaharan Africa, according to http://esa.un.org/unpp/ the figure is 900 millions for the whole continent.


Timothy Burke 12.21.07 at 7:03 pm

John, it matters somewhat how the aid has been distributed. Fifty million dollars paid to one institution or person is a rather different prospect than ten dollars to a great many people. The expectations one might have about the latter would be nil, but you might well expect some consequences from the investment of a large sum in one project, location, individual, government, etcetera.

I’d also second the point that it doesn’t necessarily follow from the observation that the total aid paid has not been very much that:

a) the aid has produced no outcomes


b) much more aid would produce favorable outcomes

On a), for example, if I gave an extended rural African family enough money for school fees for all the children in one generation from the beginning of their education to whatever final degree they wish to pursue, with the proviso that this would not cover education pursued outside of their home nation, the total cost would be very very small. But the impact on this family might be enormous.

On b), however, if I gave $500 million in a single year to a national government, no strings attached, no conditionalities, no reporting required, it’s entirely possible (indeed likely) that the entire sum would end up being largely an economic boost for Swiss bankers and real-estate owners in southern France. That wouldn’t change if I increased that amount to any sum you care to name.

On the other hand, equally, if I fund a highly specific project for $50 million over ten years, and put numerous conditionalities on it, and hire a bunch of non-Africans to carry out the project, it’s also likely to do no lasting good except for the development industry itself. Again, regardless of how much money I put into it.

I don’t think the question of the amount of aid really does much of anything to answer why aid in Africa has had or lacked a particular impact. It does clarify just how little a burden aid has been on donors, however–they could certainly afford to give a great deal more. But I think there are plenty of studies of how development functions at the ground level in Africa that demonstrate that it’s not amounts of money that are the problem.


Random African 12.21.07 at 7:05 pm

So the question to ask, is there any difference between the economic performance of those places with higher aid per person per week and those with lower (controlling for other stuff and the obvious reverse causality). And the answer seems to be still ‘no’

Has such a study been done ? If yes, any recommendations ?


John Quiggin 12.21.07 at 7:10 pm

Another useful comparison is to look at the Marshall plan relative to donor GDP. Google produces Niall Ferguson quoting the Marshall Plan as an average of 1.1 per cent of US GDP (more in the peak years, I guess) compared to US aid of 0.2 per cent of GDP now.

It’s really the Marshall Plan that created the persistent and false belief among Americans that a big share of their taxes is spent on foreign aid.


Tom T. 12.21.07 at 7:37 pm

Much more useful in thinking about the likely impact of aid is the amount per person per week.

I don’t see how this is useful in any way. Look at the examples from the linked article. How much would Norway have had to spend on fish-processing plants in Kenya, per person per week, for that aid to have been a success? How much should the EU have spent on Kenyan highway repair, per person per week, to overcome the failure to budget for maintenance, the political non-cooperation, and the highway bribery that the article says have prevented that road from being economically successful?

For that matter, why should the success or failure of a project in Kenya be judged based upon the number of people on the entire continent?

Note too that John Q’s calculation does not take into account the fact that Africa’s population has quadrupled during the period in question, but of course that’s an argument only on the margin.


John Quiggin 12.21.07 at 8:02 pm

“For that matter, why should the success or failure of a project in Kenya be judged based upon the number of people on the entire continent?

Indeed, the total volume to Africa of aid, however it’s measured, is irrelevant to the question of why particular projects succeed or fail. But that’s something you should take up with the authors of the article, not with me.


Suther 12.21.07 at 10:27 pm

Well, the Iraq war is running ~$5/wk per person (US). May be that’s too low, which is why the war wasn’t going well until recently. I think we should spend $10/wk … that sounds right.

Foreign Aid, $1.22 per person per week sounds right too. Also, I wouldn’t have any of it wasted or sent to Switzerland, that seems like a bad use of the money. That seems very easy – I’ll just buy another fair trade cup of coffee at starbucks this week and everything will be better.


foxmarks 12.22.07 at 12:00 am

The implied premise, that some amount of aid will make a person or continent rich, is false. The success of aid is entirely dependent upon the recipient, his industriousness and the cooperation of his neighbors.

Give my circle of friends 20 cents a day per person and we’ll make millions of it. Give our millions to a circle without rights, law, or skills and they’ll make twenty cents out of it.

Being honest, reliable and autodidactic are free. But you can’t package such things and ship them to the dark continent. Nor can one so eloquently wring one’s hands about another not doing enough for himself.


notsneaky 12.22.07 at 12:36 am

random african,

the above mentioned Burnside and Dollar paper does essentially that, and in addition interacts aid with policies:


Also on what variable to use – I’m not sure aid/person is better than aid/gdp (or more precisely, (aid/person)/(gdp per capita). If the purpose of aid is to help countries grow by spurring investment and if you think there’s diminishing returns to capital then it should be that aid/person has higher impact in poorer countries. If aid does what it is supposed to do.


Tom T. 12.22.07 at 1:41 am

Re: 24. But the authors of the article don’t suggest that the total volume of aid to Africa is relevant to the question of why particular projects succeed or fail.


Moz 12.22.07 at 2:20 am

The second important point is that “repayments” extracted from Africa greatly exceed the “aid” flowing in. It’s all very well giving 20c/week, but when you’re similtaneously demanding $1/week for the privilege… who wins again?


terence 12.22.07 at 4:12 am

NotSneaky et al.

Burnside and Dollar is essentially dead. Everyone more or less agrees that Easterly, Levine and Roodman killed it.

The most recent skirmishes in the futile war of cross country regressions and aid are (Sorry in a hurry can’t hyperlink):

It Works

Hansen and Tarp (but with diminishing returns)
(try: http://www.nottingham.ac.uk/economics/credit/research/papers/cp.99.10.pdf)

Sanjay Reddy (a DESA paper) works if you separate good quality aid from that which is poor.

Clemens and Radlet
If you disaggregate aid from humanitarian assistance there is a positive relationship.

David Fielding and a few others
Aid works when measured against social indicators

Fielding again, this time aid works if you measure it vs real GDP per capita.

It doesn’t
Rajan and Subranathan
(Who, appear, if their econometrics is good, to knock over quite a few of the positive studies above)

This Roodman study is also interesting:

As I said above, this debate always strikes me as kind of futile as we know, thanks to a few big successes, that aid can work. The real question ought to be what distinguishes successful aid from that which fails.


Spike 12.22.07 at 5:00 am

Also bear in mind that probably the biggest chunk ($50 billion) of that $500 billion went to aid to Egypt as part of an American subsidy for their making peace with Israel. Egypt isn’t rich, but its a middle income country with a standard of living significantly above the deep poverty you get in sub-Saharan Africa, which is what people are generally thinking of when they consider African aid.

And of course, the vast majority of that $50 billion to Egypt was in military aid that helps to prop up the police state they have going there…


notsneaky 12.22.07 at 6:01 am

moz, what “repayments” are you talking about?


a 12.22.07 at 7:27 am

“Give my circle of friends 20 cents a day per person and we’ll make millions of it.” How do I invest?


JohnTh 12.22.07 at 12:32 pm

Although the analogy is fair, the right thing to do would be to measure it over the average population over the period, which was much lower. Also, could an argument be made that sub-Saharan Africa has grown much richer, but it has invested its GDP increase in more people rather than a higher GDP per person? Typically in European economic history a high population growth has been associated with prosperity, hasn’t it? Isn’t it only in the last 200 years, in some places, that greater wealth has been accumulated per person rather than in the form of higher population? It would be interesting to get the views of any passing economic historian on this


notsneaky 12.22.07 at 6:09 pm

terence, you’re right, but it still should be read as a sort of intro into the whole thing. I was going to say something like “but it (B&D) has been widely criticized” but was in a hurry to write and run to do Xmas shopping so I didn’t. Thanks for the links!


Random African 12.22.07 at 6:27 pm

yeah thanks for the links.


Michael B Sullivan 12.22.07 at 8:34 pm

Here’s a question I always have on this dollar a day stuff: how is that calculated?

It seems on the face of it to me to be ridiculous. A dollar isn’t enough money to buy enough food to live on (much less buy clothing and shelter to keep the elements away, or many other bare necessities). Obviously, one billion people aren’t starving to death right now, or next month we’d have a population one billion lower than we do today. So I’m going wrong somewhere above, and the figure is repeated enough that I assume that there’s some validity behind it.


Pete 12.22.07 at 9:16 pm

#37: Food is cheaper in Africa. But you could survive on $1/day of rice even at US supermarket prices. Everything else, people make or scavenge for themselves, or among their families.


abb1 12.22.07 at 10:10 pm

I got the impression that it’s not so much food but water, clean water.


Michael B Sullivan 12.23.07 at 12:36 am

#38: But it’s purchasing power parity, right? It’s a dollar’s worth of food, not an actual dollar. So the difference between prices in the USA and Africa are supposed to be taken into account.

You’re probably right that you could get enough calories to survive on per day, at $1, for the very cheapest foods. But literally nothing else. No clothing, or shelter, or anything.

The issue of making or scavenging for themselves is more or less what I’m getting at. If I’m a farmer, mostly subsisting, and I make enough food to eat, and then sell a small remainder for $1 per day, how is that counted? Am I in the $1 per day bottom billion, or not?


terence 12.23.07 at 1:14 am

Michael & Pete,

Because of purchasing power parity conversions (and inflation adjustments) the WB’s $1 a day poverty line means the following:

To be under this line you will have to be consuming less per day than you could have consumed in the United States, in 1992, with US$1.08*.

In other words fricking poor**.

Also, because it is a consumption poverty line rather than an income poverty line, anything you consume from the subsistence economy is monetarised and included in your consumption bundle. This process is inexact of course – particularly at measuring the value of public goods. But the salient point remains – if you’re under this line you are very, very poor indeed.

If you go to the UNPD’s International Poverty Centre’s website there are some good briefing papers on this but, one of the many flaws of the WB poverty lines is that they were simply created by averaging the poverty lines of some low income countries. They were never tied to a fixed set of calorific needs. Attempts to construct genuine needs based poverty line come up with (please get the exact numbers from the aforementioned website) something like 90c / day for absolute basic calorific needs and $1.50 day for absolute basic basic needs. A line based on health outcomes (using the Preston Curve) produces a figure of about $2 a day (possibly significantly higher depending on calculations).

Former WB economist Lant Prichard has a (in my opinion) sensible approach to categorising poverty.

Under $1 / day – destitute
Under $2 / day – extreme poverty
Under $10 / day – poverty (the $10 figure is arrived at by averaging rich country poverty lines)

*Might want to double check here in a hurry – presents to wrap and I could have the exact number wrong.
**So poor that poverty depth below this line would have to be minimal as you couldn’t fall much lower without being dead. And, for this reason, it won’t take a huge improvement to bring people over the $1 a day line.


Michael B Sullivan 12.23.07 at 10:33 am

Thanks, Terence!

So, $1.08 in 1992 is $1.60 today, according to some random inflation calculators that I Googled up. $1.60 a day doesn’t have quite the same ring as dollar-a-day, but it makes a lot more sense in terms of how the bottom billion survives.


foxmarks 12.23.07 at 4:51 pm

33: I can send you a list of addresses, but remember, since rights and law operate here, you’ll only be paid a return on your share in the total effort. Expect about 5%, long-run inflation-adjusted. I’m keepin’ the bulk of the fruits of my genius. Otherwise, why wouldn’t I just eat your twenty cents?


Jim 12.23.07 at 8:01 pm

P O’Neill and Stuart, I estimated here (based on a possibly dodgy back of the envelope calculation) that average annual Marshall Plan aid per capita to the four largest European countries was about three times the aid per capita to Sub-Saharan Africa in 2002, adjusted for inflation.


quantitative today 12.23.07 at 10:05 pm

I would be interested to see an example of a $1.60 per day budget. If based on PPP, it could of course use easily researched developed-world prices (e.g., at a nearby store). One would have to allow for a suspension of housing construction codes and other enforcers of high costs.

At the margin, lower income will be engender increases in do-it-yourself activity, that is, a displacement of money-mediated division of labor. In terms of quality of life, how can one properly account for services provided by oneself, and (working out from there) services provided by family members, informally bartered with neighbors, more-formally bartered with more distant parties, etc.? Almost all profound poverty is the poverty of the rural poor. Regarding the central question of food, how can one properly account for one’s own (family’s, neighbors’…) garden production? There must be documents describing methodology, but the bare numbers have outrun the information needed to fully interpret them.

Considerations like the above suggest how $1.60 per day might be more tolerable than it seems to wealthy urbanites — while still being appallingly low.


Alex 12.24.07 at 1:11 am

Any figure on how much private aid has gone to africa in the same time period?


terence 12.25.07 at 5:47 am


Off the top of my head – considerably less. The OECD DAC website might have the actual numbers.


Craig Howard 12.26.07 at 4:12 pm

Much of the discussion here seems to assume that the aid to Africans was meant as income. In that case the 20 cents a week figure does seem laughable. But the aid was not meant as welfare payments, it was intended to improve infrastructure and provide some capital where there was none.

That it was wasted or mis-targeted is beyond dispute — that it was a negligible amount is laughable.

Comments on this entry are closed.