Malawi success and donor fallibility

Election Poster for Bingu wa Mutharika

On the Oxfam blog, Max Lawson has an excellent guest post telling the story of how Malawi has used an extensive programme of fertilizer subsidies to generate seven years of economic growth, reduductions in poverty and child deaths.

Max cites a forthcoming paper by Andrew Dorward and Ephraim Chirwa (ungated version here).  Dorward and Chirwa argue that:

Malawi’s agricultural input subsidy programme addresses a low maize productivity trap that leads to food insecurity and poverty, and constrains economic growth and, paradoxically, diversification out of maize and agriculture. This low productivity trap arises as a result of severe seasonal credit constraints affecting very large numbers of poor, food deficit farming families, together with thin and high risk, high margin input and maize markets. The key successes of Malawi’s subsidy programme arise where it relieves both affordability and profitability constraints to increased staple crop productivity from increased input use, and in doing this both raises land and labour productivity and improves food security for large numbers of poor households through some combination of increased real wages and reduced food prices.

The only part of Max’s post that I disagree with is his remark that  “we should leave our economic theory at the door and instead focus on what works empirically.”  As Jonathan points out in the comments, economic theory tells us that government intervention may be an appropriate response to market failures.  While recognising the success of the programme so far, we should not stop asking whether the same results could be achieved more cheaply and more sustainably with some other, even better approach.

A more relevant challenge is: why did some donors oppose this programme, and what have we (and they) learned from that error?

Dr Bingu wa Mutharika fought and won the 2004 election on a platform of guaranteeing food security. HIs proposals for a targeted subsidy was overturned by the Malawi Parliament in favour of a universal subsidy, which was introduced in 2005.

Election Poster for Bingu wa Mutharika

Election Poster for Bingu wa Mutharika

Donors are – on paper – committed to respecting government ownership and supporting the governments’ development programme.  Yet despite clear national commitment, endorsed in a democratic election, donors generally opposed the introduction of fertilizer subsidies, consistent with the World Bank’s position throughout the 1980s and 1990s. The donors argued against the government’s proposed scheme because they thought it would be too expensive; it was insufficiently targeted on the poor; it would undermine private sector development; and because they doubted the capacity of the government to implement it.

When Malawi introduced its programme in 2005, the IMF and the US Government opposed it outright, on the grounds that it would damage the private sector. The World Bank, EU and UK Department for International Development adopted a more nuanced position: they argued that instead of a universal programme there should be “smart subsidies” which should be tightly targeted to reduce the costs, and that the programme should include an explicit exit strategy.  DFID eventually supported the programme after extracting an agreement from the government that it would use private fertilizer suppliers.  Some of the Scandinavian donors and UN agencies supported the programme from the outset, partly influenced by the apparent success of a local Millennium Villages Project.

The apparent success of the Malawi fertilizer subsidies is primarily a story about the Malawi government, not donors; though the scheme could not have been afforded, especially through the 2008 price hike, without donor funding.  But it does give rise to two questions about donor policy and behaviour.

First, are donors still labouring under too simplistic a view of the role of government in the economy? Donors continue to be sceptical of agricultural subsidy programmes (which is rank hypocrisy, given the subsidies they provide their own farmers).  This seems to be partly because we have an insufficiently rich analysis of the nature of the market failures and how they are best addressed; and partly because donors still suffer from the sustainability delusion, which requires them to oppose perfectly sensible government policies and programmes for which there is no identifiable exit.  If the UK government were only allowed to implement inherently time-limited policies there would be no National Health Service.

Second, how should donors reconcile their own views of a policy with their commitment to respect country ownership? Donors are committed to support developing countries’ own development strategies.   But what happens if they disagree either with the thrust of those policies, or with particular details?  Should they refuse to finance them? Should they act as “critical friends”, identifying the shortcomings of the policies and seeking to get them changed?  Should such opposition be private or public? How is that consistent with respecting country ownership? If they do try to change the policy how are they held to account when – as was apparently the case in Malawi – they are wrong?

I’d like to suggest two ways in which donors can better respect country ownership. First, where they have an opinion about a policy, they should produce publicly their analysis and evidence, to allow this view to be discussed as part of the public debate, rather than exert political and economic power behind closed doors.  Second, there should be a version of the Salisbury Convention in aid: if a government is pursuing a policy for which it has an explicit mandate in a reasonably democratic election, the donors should not try to undermine it.

UPDATE: Smart commenters below ask two questions.  First, is it premature to say this has been a success, until we have a year of bad rains?  Second, were the donors as hostile as my blog post suggests?  If you have insight into either question, please leave it in the comments below.

23 thoughts on “Malawi success and donor fallibility”

  1. Something most analysts forget: until the 80s, agricultural input subsidies and credit schemes were economic orthodoxy. There were failures and abuses, but subsidies were not abandoned because of evidence, but because of an ideological shift. The same shift that made us deregulate banks and ask for school and health service fees. Economics is the dismal science because even now, the lure of ideology is often stronger than sound evidence.

    Moreover,less errors of judgment would be made if donors stopped supporting governments and instead started supporting the poor, pro-poor policies and administrations.

  2. The World Bank had funded an earlier pilot subsidy program under the previous ruler, Muluzi, referred to in Malawi as a starter pack. And, a leading agricultural thinker for the World Bank in Malawi, Stephen Carr, actually was an important force behind the use of subsidized fertilizer… this is all to say that the World Bank was very supportive of the program even before Bingu was.

    Malawi had suffered greatly from famine. It wasn’t my experience that donors in the country were at all opposed to the subsidies (except, perhaps how they were politically manipulated, esp. the season before the election). Maybe the rhetoric back in their home countries was anti-subsidy, but in Malawi, donors didn’t want to see more famine.

    To answer your first question, then, I’d say that it’s not simplicity of view that makes donors reticent to support government intervention in the economy. In fact, it’s the cynical view that doing so will entail corruption, which happened even in the stellar case of Malawi’s fertilizer subsidy.

  3. Interesting post Owen. Four comments:

    1) I hate it when people talk about “economic theory” as if there is one economic theory – meaning that the version of economic theory that they prefer is the one true god. Sam’s comment covers that I guess.

    2) The proposal that donors should make their analysis public is intriguing – and perhaps in these days of transparency, wikileaks etc. it won’t be long until donors are unable to keep their analysis private.

    3) Your salisbury doctrine suggestion is interesting, but of course the devil is in the detail. Who will decide whether a government has an “explicit mandate in a reasonably democratic election”? What standards will be used? Global ones? Regional ones? National ones?

    4) How about – in an effort to support a broader notion of country ownership – extending the proposed salisbury doctrine so that when donors’ analysis suggests that policy Y in country X is a bad idea for poverty reduction, but donors do not object to that policy – let’s say for geo-political/strategic reasons – donors share their analysis to add to the public debate?!

    Hmmm, my point 4 (and owen’s suggestion re the salisbury doctrine) still rests partially on the assumption – one that I’m not very comfortable with, and which Sam also touches on – that there is such a thing as objective analysis and evidence which is not shaped by political views.

    Disclaimer: I currently work for DFID. These thoughts are mine, not DFID’s.

  4. Kim – the subsidies have not been proven to be resilient against the conditions that led to previous famines in Malawi: drought. The largest elephant in the room with the subsidy programme is the generally favourable rains the country has seen since the last major shortfall (04-05).

    It was drought that brought the previous famines, and it isn’t clear that fertiliser is going to do much the next time there is a drought. So instead of investing in long term drought-resistant strategies like irrigation, the government has preferred a complete political win (each bag of fertiliser is a vote in the next election).

  5. Matt – I don’t know that I’ve said anything to the contrary earlier, but I will say that drought is a condition under which famine can occur — but it’s not a necessary (and sometimes it’s an insufficient) condition for famine. See esp. Lofchie 1975, but also the more famous Sen 1999. And, in Malawi, it’s not just drought, but also floods that can wash away crops (which the innovation of irrigation doesn’t fix).

    And, to put all of this in context, the fertilizer subsidy is not a new thing with Bingu, or even with Muluzi. In fact, the former dictator, Hastings Kamuzu Banda, had also provided fertilizer to smallholder farmers. As we might expect, his distribution of the fertilizer was even more politically charged than what we saw with Bingu.

    I think we’re on the same page in saying that the subsidy has been a political win for the ruling party (as have the various cash transfer schemes going on in the country). What I challenge in Owen’s post is the notion that donors [in Malawi] were against the subsidy. Sure, here in the US, we talk a lot of trash on subsidies (while our farmers continue to rely on them), but if you talked to a USAID official in Malawi in the 2004-05 period, I don’t think they would have raised a challenge to the subsidy program.

    In my own discussions with Stephen Carr (who, again, worked for the WB), it was clear that he had lobbied for the application of fertilizer by smallholder farmers for years.

  6. Kim – I agree that drought is neither necessary nor a sufficient condition for famine.

    You argued that country donors supporter the fertiliser subsidy because they were concerned with avoiding famine in Malawi. I was just pointing out that the fertiliser subsidy doesn’t do much to avert the probability of drought-driven famine. True, irrigation doesn’t help with flooding, but neither does fertiliser!

  7. I don’t know Malawi, but I used to work in Nicaragua in a FAO fertilizer programme, doing field research and promotion for sustainable fertility management. The programme got seriously attacked during an evaluation as “plots lost” were excluded from the analysis of the ideal dose of fertilizer. Indeed: there was no financial nor economical benefit from fertilizer use when including the plots lost due to drought in the calculations. This links up with the point made by Matt: stability is in the short run a necessary condition for development, and general budget aid seems to provide stability by keeping the government in power for election after election.

  8. Two comments on the comments:
    1. We may have to wait to see whether the fertiliser subsidy would have a direct positive impact and make a difference during a drought year, but it is debateable whether its unique or even its main intended goal is actually to avert famines (though these were a powerful trigger in its establishment), rather than for example to enhance overall food security and reduce poverty. These are quite different aims, though of course they may overlap, and the capacity of the programme to achieve each should be evaluated separately. On the other hand, seven years of better agricultural output, not to mention the indirect effects of the programme on agricultural wages, on risk-taking and on food prices, have strengthened the resilience of small-holder farmers. Breaking the low productivity trap and reducing poverty is itself a way to counter famines, and arguably on the long term a better one, even if fertiliser subsidies should be found to have a scarce direct role on the causes of the famine, be they drought or floods.
    2. The fact that poverty-reduction measures such as input subsidies or social safety nets are also instrumentally used in political campaigns is not an argument against them or their effectiveness. In fact if politicisation helps put such measures on parties’ and candidates’ (i.e perspective governments’) agendas, then maybe we should pragmatically welcome such instrumental use. Something that politicians in the rich world, by the way, are not alien to. The risks of favouring clientelistic relationships too, cannot be an argument against the the subsidies themselves, and should be addressed separately and countered through specific insititutional mechanisms.

  9. Responding to comments on the impact of drought on food production in Malawi I think that it would help to note that Malawi is in the wettest period which it has experienced for almost six hundred years (evidenced by the dramatic rise in the level of Lake Malawi). In consequence there have only been two years in the past 50 in which a shortage of rainfall had an impact on national maize production and one year (2000/01) when excess rainfall led to waterlogging (not flood) which seriously reduced production. Naturally there have been districts which have a shortage of rain in a year when others have had an excellent crop but a recent Ministry of Agriculture table plotting rainfall against national maize production revealed virtually no significant relationship whilst a graph which plotted fertiliser availability against national production produced a much closer match. It is of course true that in any country the use of fertiliser cannot avoid serious crop loss in the face of a real drought on the other hand there is plenty of evidence that strong, well nourished plants are more able to withstand shorter periods of stress better than those which are poorly nourished.
    Malawi is blessed with very little in the way of significant rivers which run downhill sufficiently to provide large areas of gravity fed irrigation. The alternative is to use motorised pumps to drive water up-hill from the lakes in order to irrigate. With a long and expensive transport line for its fuel it is well appreciated that using oil based fuels to pump water on to maize fields is guaranteed to lose money. The alternative is foot operated pumps in favourable areas and there is an increasing number of these but the potential for this form of irrigation is strictly limited. For the foreseeable future the great bulk of the nation’s food supply will come from the 2.4 million farm families who rely on rainfed crop production. Their dominant constraint is a shortage of soil nutrients. Efforts to restore these by organic methods have had limited success in a country where only a tiny fraction of farmers own livestock and where the density of the population precludes any kind of improved fallowing to restore fertility. In consequence farmers have to follow the whole rest of the farming world and restore nutrients with fertiliser until scientists come up with some alternative way of producing food under conditions of permanent cultivation. As nobody has been able to identify a cash crop for 75% of the smallholder population they have little cash with which to buy fertiliser and good seed. Without the subsidy the country would sink back to widespread hunger and a 22% child death rate which used to be its lot.
    The writer lives in a village in Malawi and has been working with Malawian smallholders for the past 23 years

  10. Pingback: Stephen Carr weighs in on Malawi fertilizer subsidy debate « haba na haba

  11. Pingback: Of Seat Belts, Modernity, Subsidies, Declaration and More Things for PhD Students to Agonise Abouts, Welsh Myths « Waylaid Dialectic

  12. Pingback: New year, new start « Bottom Up Thinking

  13. This isn’t the first time that the Goverment of Malawi have been proved right in providing a universal service, against opposition from sceptical donors. Throughout the 80s and 90s it resisted pressure to introduce health user charges in public facitilies and therefore maintained a universal entitlement to free health care. As a result health care utilisation, as measured by outpatient attendance rates has been one of the highest in Africa. Perhaps this has been another factor contributing to Malawi’s relatively good performance in reducing child deaths.

  14. Owen,

    On the two questions you raise in the update here go the views of a former DFID economist that spent the period Sep 2005 to Aug 2008 in Malawi:

    (1) As Stephen Carr’s lengthy response suggests the vulnerability of Malawi to rainfall in recent times has been grossly overstated. The volatility of fertilizer and maize seed policy has been a larger determinant of maize production failures. Improved seed varieties and the basal fertilizer dressing result in a maize plant that is more resilient to lower rainfall.

    (2) On donor politics I think some of your statements are inaccurate. Firstly, Norway was the only Nordic donor in Malawi at the time. They, together with the EU, only committed to supporting the programme following DFID’s announcement of their support. Secondly, the IMF had learnt from the maize sales scandal of the early 2000s of the reputational risk of meddling in microeconomic issues. You certainly won’t find any evidence of their opposition to the subsidy in official documentation. Even if you had access to unpublished correspondence I think you would find they deferred to the views of the Bank and DFID on the matter. As long as the subsidy’s costs were properly budgeted for and managed they were happy, leaving the debate on efficiency, sustainability and impact to the budget support group of donors.

  15. Apologies for the inarticulate comment.

    There was neither a Nordic country supporting it from the outset nor the IMF was arguing against it. This policy was good national politics (“a complete political win” as Matt put it above), it was a clear Government commitment and nobody was going to stand in its way. But neither did anybody get as excited with blanketting the country with Nitrogren as Jeff Sachs would have liked us to do.

    Budget support, which had only resumed with the election of the new government a year earlier, kept flowing in at an increasing rate (some additional funding in 2005 was even justified on the grounds of there being a funding gap for seeds for the subsidy). Without it, the subsidy would not have been feasible.

    In 2006 direct support for the subsidy itself was also provided with the aim of improving it at the margin: allowing private sector retailers to compete with the parastatal in provision of the fertilizer; giving more choice to farmers; and improving monitoring and evaluation (including funding Dorward and Chirwa to produce the first evaluation that has led to subsequent papers, including the one you cite).

    1. Thanks: that’s very clear. That seems to dispose of the view which I reported that donors had opposed the subsidy. I stand corrected and I’m grateful to you for making this clear.

  16. I sympathise very much with Owen: the “story” doing the rounds is that the donors opposed the subsidy. It is apparently not enough to read the main stories on the internet to know the facts.

    Sam: that’s one of the things I love about blogs. You can find out more from people who have first hand knowledge

  17. @sam, @bsanchez

    I think bsanchez sums it up pretty well. Definitely the World Bank was less than enthusiastic when I interviewed them in 2007- they were fixated on the greater use of private providers of fertiliser (despite their inadequacies) and this became a performance criteria in both their PRSC and in the donor budget support framework. But to say donors in the country were outright opposed or trying to stop the subsidy would indeed be an exaggeration.

    As with all these things, and as Owen points out, the further you get from the field the more theory driven and reactionary the response. Certainly I haven’t heard anything from the bank centrally celebrating this success apart from a relatively mealy mouthed box in the WDR on agriculture. In itself this doesn’t matter, but it does when this means that lessons for other countries are not being learned and shared, a key role of the Bank and others. It is one thing not to actively oppose a policy you don’t like, it is another to recognise its success, and to look at where that can work elsewhere (and where it can’t).

    Interestingly a similar story in the same country, the DFID and Global Fund sponsored programme to increase healthworker salaries for 50% has been another incredible success-

    which can and should have been translated to another ten countries with huge healthworker shortages by now, but has not arguably because of a donor fear of being seen to support salaries, despite salaries and other recurrent costs being absolutely critical in health and education. Once again this is due to owen’s sustainability delusion.

    Finally, by saying in my orginal post ‘leave your economic theory at the door’, what I should have said really is treat your economic theory more like science than philosophy- a set of hypotheses to be tested against empirical realities, rather than an a priori set of truths. Jonathan’s comment about credit highlighted that clearly for me- a theoretically attractive alternative which would not work in the Malawi context.

  18. For Sam and Max. This is an unusual situation in which two apparently opposing statements are both true and it depends at what point in time you look at them as to what answer you get. When the Malawi government mooted its intention of providing a large subsidy on fertiliser to a major part of the farming population there was opposition from the British, USAID and the World bank. It is this which found its way into the world press. Then came the efforts to bring about a change of position. It did not happen overnight. In the case of the Brits it involved a presentation in the House of Commons, a visit to Malawi for five days of a Shadow Minister, detailed discussions in London with the Scientific Advisor to the Minister of Overseas Development (Sir Gordon Conway) and finally a visit to malawi by Sir Gordon which seems to have been the event which put a seal on the change to real support for the subsidy. In the case of USAID and the Bank it again required detailed discussions in Malawi and Washington to bring about a grudging acceptance of the government initiative. So there was opposition and there was subsequently support.

  19. Pingback: Big farm, big deal « Bottom Up Thinking

  20. I’m surprised no one has mentioned the role of prices in this discussion. My memory is fuzzy so please correct me if I’ve got the following wrong. I attended a session at the EWB conference in Toronto in which Gary Toenniessen (Rockefeller Foundation) showed the history of fertilizer subsidies in Malawi. From my recollection, he said that past subsidies led to sharp declines in prices and incomes whereas the current market is being buoyed by the crisis in Zimbabwe. If Zimbabwe gets its act together, does the Malawi success story disappear?

  21. Bill, I am not sure what prices and incomes you are referring to. I guess maize prices and farmer incomes. As outlined in our paper, the Zimbabwe maize situation has had a serious impact on Malawi maize prices in only one season, 2007/8, when there were arond 300,000MT exported to Zimbabwe, seriously raising the price of maize in Malawi (informal cross border trade between Mocambique and Zimbabwe may have been a minor upward influence on Malawi prices in subsequent years, but I think this would have been relatively small). The 2007/8 and subsequent high maize prices were very unfortunate. As our paper argues, the major benefits of the programme arise when maize prices are lowered (I think that you are implying the opposite?). This is because the majority of Malawian smallholders are maize deficit producers (ie they are overall buyers in the market and therefore benefit from low prices which raise their real incomes) and therefore if the subsidy stimulates low prices this has real income and economic growth benefits as well as food security benefits. Malawi is curently enjoying low maize prices. This does cause problems for the relatively small numbers of maize surplus producers (but should encourage diversification out of maize production – a long standing and important development objective in Malawi). These issues are discussed extensively in the paper.
    It is also worth saying that although the detailed micro-economic theory of Malawi’s rationed and targeted subsidies differs from conventional theory on general price subsidies (see, these price effects are critical in conventional theory on subsidy impacts and benefits. Falling food prices (relative to incomes), and diversification out of staple crops, have been core and universal processes where economic development has occurred. These are pertinent observations in the context of unstable and upwardly trending international cereal prices ….

Leave a Reply

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