All aid is used for imports: get over it

All foreign aid is, in the end, used to pay for imports.

In a few moments we’ll discuss why this is true and why it matters. But let me offer two apologies.

To those for whom this is a statement of the bleedin’ obvious, sorry for wasting your time.  Please feel free to skip down to the reasons why this important.

And to those for whom this is a shocking revelation, I am sorry to puncture the bubble. This reality is not the result of a conspiracy between aid agencies and commercial interests but an unavoidable consequence of the fact that aid is provided from abroad.

In what sense is all aid used to pay for imports?

Suppose a rich country gives $100 million in aid to a poor country.   The recipient may be a person or organisation, or the government. Sometimes the aid is given as money, sometimes as goods and services (such as food aid, or technical expertise).  If the aid arrives as imported goods and services, then the aid has been used, by definition, to pay for imports.

But suppose the aid comes not as goods and services but as foreign currency.  The only value of that foreign currency is that it can be used to buy imports from abroad.

Of course the recipient might do something else with the foreign currency, like sell it or save it.  But in the end the foreign currency has to be used to pay for imports, or it is worthless.

Suppose you send $100 to an NGO in Nairobi.  That NGO might want to use your donation to pay its rent. The NGO has to sell the $100 and get Kenyan shillings to pay its rent.  The people who buy the $100 want the dollars at least as much they wanted the equivalent amount of Kenyan shillings. Why? So that they can buy something from abroad, either today or in the future (or sell the dollars on to someone who will).  Instead of using their Kenyans shillings to buy something locally, somebody at the end of the chain must be selling their shillings and using the $100 to buy something from overseas.  So the impact of your donation on the NGO has been that it can pay the rent (which of course is not an import); and the net impact on Kenya as a whole is that imports have gone up by $100.  The additional imports are of value to someone in Kenya, and because you originally gave the claim on those additional resources to an NGO, the NGO was able to use the value of those additional imports pay its rent.

The same argument applies to aid given as budget support to governments. We may expect that the recipient government will use the money to employ teachers and nurses, or to build roads.  None of that may be imported.  How does giving foreign currency to a government help it to employ teachers?  The answer is that we have given them something of value – a claim on imports – which they can trade in the economy to get the money they need to enable them to pay teachers.  But since the thing of value we have given them is a claim on imports, the government can only use it if somebody somewhere in that economy wants the imports enough to give up some local currency in return.  The aid is in the form of additional imports, even if the ultimate importer is not the original beneficiary of the aid.

It maybe that someone will save the foreign currency – perhaps the Central Bank will use it to increase its reserves, or a firm might keep some dollars in a safe. But these savings have value because the money can be used,  by someone somewhere, at some time in the future, to pay for an import.

This should all be uncontroversial – it is, after all, just another way of describing the national accounts identity.  Aid is a capital inflow; and an increase in capital inflows must be matched by an equal and opposite current account deficit.   So when aid goes up, either imports must increase, or if imports stay the same, exports needed to pay for imports must go down.

Why does it matter that all aid ends up as imports?

Here are some consequences

  • Aid to developing countries is a fiscal stimulus for rich countries.
    If the cause of the economic crisis was that rich countries were consuming too much relative to how much they were producing, the only possible way out of that is for rich countries to increase production faster than they increase consumption, and that means increasing exports.  Increasing aid does that.
  • The question of how much aid “ends up” in a developing country is a red herring.
    All aid “ends up” as goods and services imported from abroad.  The right question is whether they are the most valuable goods and services, and who benefits from them.
  • Effective aid must inevitably increase the current account deficit of recipient countries.
    Hare-brained schemes to “sterilize” the effect on the balance of payments could only work by preventing the aid from being used.  A worsened current account is the accounting counterpart of increased aid.  Unless you are a paid-up mercantilist, you’ll know that getting more imports for nothing, or not having to export as much for the same amount of imports, is a good thing.  (Of course, if the aid is used for things that increase the country’s productivity, that might increase exports and so improve the current account, but that is a second-round effect which might or might not offset the direct effect.)
  • The only significant difference between budget support that some donors give today, and the old balance of payments support they used to give, is the nature of the political dialogue that accompanies it.
    In the old days, donors were happy to be told that the money would be used for imports (what else could it be used for?).  Today, by calling it budget support, donors feel able to get more engaged in how the government allocates its overall resources.   It is weird that auditors seem to think this is a riskier business, when the economics is the same.

10 thoughts on “All aid is used for imports: get over it”

  1. Sterilization hair-brained?

    Surely, central banks do need to moderate the real exchange appreciation that results from large aid inflows? Yes it can be costly. And in the end absorption of the aid inflows will require some real appreciation. But too quick an appreciation would make life difficult for tradables producers. A group which in most aid dependant countries includes the rural poor–coffee, cocoa, cotton, tobacco farmers and the like.

    Owen replies: I should have made this clear. I don’t mean that all sterilization is a bad idea. Where aid is volatile, it makes perfect sense to smooth it to prevent gyrations in the real exchange rate, as you imply. But aid is the gift of a premature real exchange rate appreciation, and permanent, long-lasting increases in aid should feed through into changes in the real exchange rate and current account, otherwise it isn’t being used.

  2. Owen,

    Would you say that an additional consequence is that donors should always think about what imports are likely to displace?

    For example, if donors were giving a country free imports of food or clothing, it would be natural to think about the implications for the domestic food and clothing industries. If donors are giving dollars intended to pay the salaries of teachers and nurses, it might not be so natural to think about the implications for the domestic food and clothing industries, but depending on what the relevant import demand elasticities are, those aid dollars may still be spent on importing food and clothing.

    [I’m not sure about this myself; I think it’s a reasonable point, but I haven’t really thought through what it would mean in real world situations]

  3. Luis

    The effect on domestic producers is the real world counterpart of the appreciation of the real exchange rate.

    My view is that sustained, long-term predictable aid is the gift of a premature appreciation of the real exchange rate, which should be embraced.

    The problem comes when volatile or temporary aid from capricious donors drives up the real exchange rate, putting out of business exporters and firms competing with imports, and then disappears, leaving the country with neither the aid to pay for the imports they need and want, nor the domestic industries needed to export or to substitute for imports.


  4. Owen,

    Thanks for the response. I think that’s a sensible answer, but I’m not so sure that the “premature appreciation of the real exchange rate” is such an unmitigated gift. Even if aid was sustained and predictable, I think it’s worth asking how a particular aid recipient’s economy will adjust to premature appreciation, but this raises some difficult questions about when and how economies adjust in a beneficial fashion, and when adjustment can entail things getting worse. (and also some questions about the relationship between the exchange rate, exports and growth). I don’t even have a clear enough grasp of the question to be able to say much about it, let alone sketch any answers; I think adjustment is one of the most poorly understood areas of economics (not just in the context of developing countries; it is also evident in our poor grasp of how the economy is responding to the current financial crisis). I’m really not sure how a real world donor might go about thinking through the likely forms of adjustment to premature appreciation in any given country.

    (I’m sure you’re as familiar with the lit. on Dutch Disease and such like as I am … nonetheless you might like this: Session IV here, has some interesting discussion.)

  5. Interesting post. Some thoughts, though; though the foreign currency must eventually become imports (as you say, completely uncontroversial though some countries have accumulated massive reserves for exchange rate stabilisation purposes, as HK did), the degrees of separation from the initial inflow of cash and the eventual outflow of imports has an economic importance.

    If the inflow is immediately converted to imports (e.g. through assistance in kind) it has an economic effect solely through the value and/or usefulness in production of the goods purchased (which is not necessarily a bad thing: as early as 1954 or so, Michal Kalecki was arguing that imports of capital goods are necessary for economic development in the modern world). If however, the inflow is used to generate goods and services domestically in the first instance, with the foreign currency sold to someone else, then there is an additional benefit of stimulating the production process locally; similarly, if it is used to make changes to the production process directly, it may have significant effects in support of the local economy.

    Ultimately, that all foreign inflows fund imports is almost irrelevant to the aid effectiveness debate: the real problems lie when the immediate impact of the money coming in is negligible or negative. And I’m not sure anyone is claiming that budget support is better for the balance of trade: the benefits lie in the ability it gives to a Government to pursue its own development objectives with fewer restraints, which assumes that this will allow for a more efficient and developmental process for using and allocating aid.

    Ranil: Thanks for taking the time to respond. I think we agree that the fact that all aid ends up as imports is less interesting than the question of which people or institutions are given the value of the claim on those imports; and the question of what constraints donors put on how those resources are used (eg. in the extreme, by providing the aid in kind).

    I don’t however agree with you that spending money on domestic goods and services is an “additional” benefit compared to imports. Your phrase “the inflow is used to generate goods and services domestically in the first instance” is, I think, misleading. It is possible that if the beneficiary of the aid uses its value to purchase non-tradables (e.g. teachers) then there will be a change in the composition of domestic demand. (This shift to non-tradables is, of course, the underlying cause of the appreciation of the real exchange rate that occurs in some countries when aid increases). But looking at the country as a whole, total domestic demand has increased by exactly the value of the imports that the aid finances, so there is no impact on total demand for domestic production (since the increase in domestic demand is exactly matched by an increase in imports). The composition of demand may change, as a result of the aid being received by particular individuals or institutions, but there is no change in the aggregate.

    Put another way, in a situation in which an increase in aggregate demand would stimulate a supply response of the kind you describe, developing countries could do this themselves through fiscal or monetary expansion. Whether or not they receive foreign aid makes no difference, since it does not affect aggregate demand for domestically produced goods and services.

    I do agree with you that if the aid is specifically used in ways that increase domestic productivity then this may well have long run effects on growth.

    I am not sure that you are correct that the Government receiving budget support can “pursue its own development objectives with fewer restraints”, at least when compared to the old balance of payments support. In the old days, a government simply had to produce receipts for the oil or machine parts that it had imported with the foreign currency it was given: donors had no say over how the government allocated its resources. Today, the government faces a barrage of commentary, advice and restrictions from donors on how they should use the budget support – all predicated on the change of name, even though the economics is no different. Budget support may give governments more flexibility than donor-managed projects (though much of this is an illusion, given that projects are almost as fungible as programme aid), but they have arguably less flexibility than in the days of balance of payments support.

  6. Owen, thanks for responding.

    Your point re: budget support is well taken. I wasmaking a comparison to project support, but what you say re: it’s comparison with BoP support is true.

    As far as demand stimuli go, I again agree that this would best be done by domestic expenditure; that said, it is still possible through aid, though dependent on multiplier effects and consequently weakened. the big issue is whether the production process can be changed using aid, and we agree on this one, it seems. Thanks again,


  7. Great article and great discussion.

    I think a large part of aid that isn’t given in the form of money is being left out of the discussion. Specifically I’m thinking of human resources and loans. Coincidently, these are the two things I believe developing countries need more of, that is – high skilled labour & access to capital.

    All the volunteers and development workers on the ground have the potential to build systems and empower residents living within the country. This money paid would normally go into the same pockets, the difference is that the work these people do is now in another country rather than in the west where they came from.

    On the loans, I feel its fairly self-explanatory, the business/individual that receives a loan is now able to invest in opportunities. The money that comes in goes back and there is no net change in national accounts (save the interest).

    Maybe this all falls under whats already been said, but I feel its worth mentioning.

Leave a Reply

Your email address will not be published. Required fields are marked *