Now at Publish What You Fund

Effective January 2015, I have left the Africa Governance Initiative and now work at Publish What You Fund, the global campaign for aid transparency.

Posted in Uncategorized | Leave a comment

Two posts on Ebola

I spent October-December 2014 in Sierra Leone working on the Ebola response. This article in the Guardian describes some of what my team was trying to do. My own description of the work is here.NERC

Posted in Liberia, Sierra Leone | Leave a comment

Rice prices and West African production

Steve Wiggins and Sharada Keats at the Overseas Development Institute have just put out a briefing on ‘The end of cheap rice’. They argue that the high rice prices of 2007-08 that I have written about before marked the transition to a world of sustained higher rice prices. The reasons are technological (yields are running out of headroom in Asia), climate-related (downward pressure on yields) and demographic (population growth in rice-consuming countries). Asian governments have accentuated this transition by export bans (in India) and domestic buying programs (in Thailand).

Higher prices for rice mean a transfer of income from rice consumers in cities to rice producers in rural areas.  The largest net importers are in coastal West African countries and the largest exporters in south-east Asia. Rice farmers in West Africa could benefit from this, provided they are net sellers (many are not, relying on sales of rubber or other cash crops to get through the lean season). So is rice production in West Africa increasing to meet the demand?

First, let’s look at production in the 4 Mano River Union countries, plus Ghana. With some ups and downs, it has increased gradually over the past twenty years. Guinea is the leading producer, at almost 1.5 million tonnes, but Liberia and Sierra Leone have also recovered to their pre-war production levels.

rice production

Next, let’s look at imports. These are also substantial. Côte d’Ivoire is the largest importer by a long way, importing roughly twice as much rice as it produces. Liberia’s imports are roughly the same its production; only Guinea and Sierra Leone are anywhere close to self-sufficiency. (Big disclaimer: the data on FAOStat only go to 2009, and if you live in this part of the world you soon learn to distrust official statistics, because statistics bureaux are poorly resourced and a lot of food flows across borders without being recorded).

rice imports

Across the region, the overall picture seems to be that production and imports have been increasing at similar rates. Consumption is fairly consistently twice the level of production. That may have changed in the last few years, for which FAOStat doesn’t record import data, but the evidence from the production figures, and more recent balance of imports data from Liberia, is that there has been no major change since prices surged in 2007.

rice consumption

What might change this picture? Liberia is hosting a CAADP conference this week, which might shed some light. There are a number of widely-touted remedies, none of which are without problems. First, building roads makes it easier for farmers in rural areas to get their rice to market, but also makes it easier for imports to penetrate the rural areas (Côte d’Ivoire has about the best roads in the region, but also the highest level of imports). Second, removing import tariffs on rice, as Liberia has done, keeps the cost of imported rice lower for the urban population, but reduces the incentive to buy local rice – something that the Liberian government tried in 1979 with disastrous consequences,  riots that led to a coup, that are still remembered by the President and Minister of Agriculture (who were both in government at the time). Local content requirements, like the Nigerian government’s insistence that locally baked bread contain at least 10% cassava flour, don’t really work for rice, since it only needs hulling before it is consumed. As for improved seed and fertiliser, like the celebrated Nerica varieties (New Rice for Africa), these certainly seem to have helped production in Guinea, but are slow to take hold elsewhere, as there is hardly a public extension service and very little agro-dealer penetration in rural areas, unlike Southern or East Africa.

For now at least, we can conclude that the increase in West African production that Wiggins and Keats  anticipate has not materialised yet. Maybe the production surge, when it comes, will come from land-locked countries, like Mali and Niger, where imports are relatively more expensive, and demographic pressures even more intense than the coastal cities.

Posted in Agriculture, Côte d'Ivoire, Food, Ghana, Liberia, Rice | Tagged | Leave a comment

The New Alliance for Food Security and Nutrition: a necessary step, or a corporate land grab?

The past weekend saw a nutrition summit in London. The conference recognised what economists and development practitioners have known for a long time: hunger and malnutrition are usually caused by poverty, limited healthcare and bad feeding practices, rather than an absolute shortage of food.

Nutrition is also getting global attention through the ‘New Alliance for Food Security and Nutrition’. The New Alliance was formed at a G8 summit in May 2012 and brings together governments and multinationals in food and farming. The New Alliance has come in for some stiff criticism from NGOs recently, particularly in the UK, which has prompted me to take a critical look at it.

I should start with a disclaimer: I have worked on agriculture for developing country governments, a consulting firm and an NGO, but I am probably closer – by experience or inclination – to the backers of the New Alliance than its detractors. I also plead guilty to spending more time in offices in capital cities than with farmers. I have never been  comfortable with this balance, and if I go back to working on agriculture in the future I hope to do so in a more practical capacity and spend less time on policy. Nevertheless, I wanted to get these issues clear in my mind.

The case for the New Alliance is that we need private sector capital and expertise to reduce hunger and nutrition. After nearly 50 years of large-scale aid to agriculture, with disappointing results in nutrition at least, that seems like a reasonable proposition. After all, in Africa, food production has failed to keep pace with growing populations; and in India, where food production has kept pace, hunger and malnutrition has barely fallen in recent decades. (The second question has been investigated extensively by Amartya Sen, and more recently Banerjee and Duflo’s ‘Poor Economics’ touches on it). In both of these regions, the private sector has been largely absent from food production and delivery, though it continues to play a major role in growing cash crops.

Second, let’s consider the case against. There are two arguments here: that the ‘New Alliance’ is dominated by Western corporations who have no business in developing agriculture; and that it is a cover for land grabbing by investors at the expense of small farmers. George Monbiot recently made this critique, and while his characterisation of it appears extreme, it is shared by many European NGOs, and no doubt many in Africa as well.

The first argument – the New Alliance is dominated by multinationals – is, in my view, only a relevant argument against the New Alliance if you are ideologically hostile to multinationals in general. I can’t speak for all farmers, but most whom I have met, including in historically statist Ethiopia, don’t mind whether they buy their seed or fertiliser, or sell their produce, from or to a corporation, a government or a cooperative. What matters is the price, the quality of the product (if buying) or timeliness/reliability of payment (if selling). Our team in Ethiopia found that farmers with as little as hectare of land were willing to pay up to three times as much for privately produced seed as for seed distributed by the public seed companies, because of its higher quality. We also found that when cooperatives act like businesses and offer farmers the best price, as in France, they thrive; when they become agents for the state or prone to corruption, as in much of Africa, they become irrelevant.

The second argument feels more salient, since land grabbing is undisputably going on in Africa – indeed, Liberia, where I live now, was the first independent African country to lease land to foreign investors, the Firestone Rubber Company, in 1926. The six countries that have agreed cooperation frameworks for the New Alliance, beginning with Ethiopia, Ghana and Tanzania, have all welcomed foreign investors into agriculture in recent years, with mixed results. Their governments are typically enthusiastic (see this example from Nigeria), but that is not enough, since governments may not necessarily have the interests of their people, or at least not all smallholder farmers, at heart.

Looking for a balanced perspective on the New Alliance, I found two reviews useful: from the ONE campaign and the Future Agricultures Consortium. ONE has scrutinised 80 letters of intent, as outlined in the cooperation frameworks for the first six countries. They find a wide range of investors and countries represented, including European, US and African companies; seed and fertiliser sellers like Syngenta and Yara; and output buyers like Mars and Cargill. None of these are primarily land buyers, or investors. Indeed, input providers seem to be getting involved in the New Alliance primarily because they want to sell to the millions of smallholders who offer their biggest market. That seems fair to me as long as nobody forces farmers to buy from them. The ONE campaign also points out in this article that the letters of intent are dominated by companies who want to buy from African farmers, rather than sell to them.  

However, the letters of intent don’t seem to be specific enough to rule out land grabbing, or at least, not to rule it out without the deep community consultation and engagement that is needed to make it ‘free, prior and informed’. For this, a briefing from the Future Agricultures Consortium is instructive, because it looks at an established project that has now been ‘adopted’ by the New Alliance, Tanzania’s Southern Agricultural Growth Corridor. Their finding is more nuanced: the New Alliance is vulnerable to the same imbalances of power and unequal access to markets that already affect farmers all over Africa. If the investments concentrate on technology, like seeds, they may miss the most important constraints: infrastructure and legally guaranteed (i.e., titled) access to land.

The implication is that the New Alliance needs to incorporate governance factors, for example make sure that smallholder farmers have title to the land they farm. However, what of the large areas of land that are communally owned, used for grazing, or sparsely inhabited? Not every patch of land has a single owner, nor should it. Should we just leave this land as it is?

Here I think we need to be realistic and accept that if we want to feed a population of 9 billion without cutting down what remains of the rainforest, we have to accept that existing farmland and under-used savannah will have to be used more intensively. In sparsely populated regions, like eastern Zambia or central Tanzania, extensive commercial farms may be the best way to do that. In densely populated Rwanda or Ethiopia, intensification seems a much better prospect. Here I think the multinationals have an important role to play, as long as they don’t crowd out traditional, low-impact methods of intensification.

The agricultural revolution of the 19th century rested on the brutal (genocidal?) colonisation of Argentina, Australia, Brazil and the US Great Plains. The green revolution in Asia in the 20th century relied on unsustainable use of irrigation and fossil fuels. I am hopeful that with our improved education, communications and technology, the coming agricultural revolution in Africa will be fairer, greener and more sustainable than any before. On balance, I think the New Alliance can contribute to that.

Posted in Africa, Agriculture, Australia, Development, Ethiopia, Ghana, Rwanda, USA | Tagged , , , , , , , , | Leave a comment

Is Liberia really the second most ethnically diverse country in the world?

According to this survey from the Harvard Institute for Economic Research, it is. (Uganda comes first). The Washington Post reports this week, though the survey data is old and the original article was published in 2002.

Why is Liberia, a country the size of Ireland with 4 million people, so ethnically diverse? Most Liberians reckon there are around 15 ethnic groups in Liberia (though they usually call them tribes), not counting settlers like the 19th century Americo-Liberians or 20th century Lebanese. None of them dominates numerically: the largest linguistic groups are probably the Kpelle and Bassa at around 15-20% each. So you can see why Liberia would appear diverse on the measure of diversity chosen, which is the probability that two randomly selected people come from different ethnic groups.

Certainly, this ‘fractionalisation’ index is an appealing way of measuring diversity, and a lot more satisfying than counting. For example, Papua New Guinea and Australia both have many hundreds of ethnic groups; but in PNG there are many large indigenous ethnicities, whereas most Native Australian communities are tiny and most Australians are descended from white Europeans or, more recently, East Asians. So PNG is, for most practical purposes, more diverse than Australia.

Liberia may be diverse on this index, but this diversity may be less salient than in other countries in the region. In Nigeria or Kenya, for example, which region the President comes from matters a great deal, and Nigeria has developed an informal ‘rotation’ system between ethnic groups. In Liberia, on the other hand, the most important political distinction is whether the President is descended from American settlers or not; almost all of them have been, but this is as much an economic/class distinction as an ethnic one. President Sirleaf is a case in point: her father was a Gola chief and her mother half-Bassa and half-German, but by virtue of her education, she has been assimilated into the Americo-Liberian settler class, and if asked to define her ethnicity she would probably just say ‘Liberian’.



Indeed, I suspect that ethnicity has become more fluid in Liberia since the civil war began in 1990, because a lot of people fled the rural areas and were thrown together in refugee camps or the melting pot that is Monrovia. Many Liberians also spent time in Ghana and Nigeria; more recently migrants from Ghana and Guinea have come to work here, joining the large number of Lebanese who came in several waves during the 20th century. Liberia’s constitution is explicitly racist and bars ‘non-Negro’ races from citizenship; but in most other countries, the Lebanese would long since have become Liberian.

Lastly, I note that the Harvard researchers asked people how they perceive their ethnicity, which is morally appealing but can introduce all sorts of subjective bias. One is that most Liberians prefer the word ‘tribe’ to ‘ethnicity’, which gives a politically correct aid worker like me a minor panic attack. Another is that how you self-define depends on who does the asking. A Liberian might self-define as Bassa to another Liberian, a Liberian to a Nigerian and an African to a white European. Similarly, in Italy people usually describe themselves by their hometown or region (‘Modense’, ‘from Naples’, etc), but once outside Italy they are almost always ‘Italian’ and sometimes ‘European’.

In the end, I feel these labels matter a lot less than the meanings we give them. The mark of a decent society is how we deal with the people we find there, and the people who choose to move there. In this, as in many things, Liberia is making good progress by the standards of its troubled history; and a constitutional review has just begun that may provide a chance to do better still.

Posted in Africa, Australia, ethnicity, Kenya, Liberia, Papua New Guinea | Leave a comment

Feedback in development

Cross-posting an article I wrote with my colleague Dan Hymowitz. Would love to see your comments. The idea is that development workers and governments don’t get direct feedback like businesses do, because our customers are rarely paying for their product, which makes it all the more important to get feedback in other ways.

Posted in Uncategorized | Leave a comment

Fitness landscapes and getting from one peak to the other

A few weeks ago, I described the aspiration of many Liberians for their country to develop a broad-based economy and create jobs. In this post I want to explore how they can get there.

Liberia’s economy has traditionally combined a large subsistence farming sector with a small, ‘enclave’ sector consisting of mines and plantations. Many of these were abandoned and eventually destroyed by civil war, although small-scale rubber tapping, logging, and panning for gold and diamonds continued.

Since the end of the Liberian civil war in 2003, many old plantations and mines have been reactivated, and new ones are being developed. So far, it looks like Liberia is recreating its pre-war development model, plus a large service economy in the capital Monrovia that is fuelled by heavy aid spending and remittances.

For the first decade recovering from war, the results have been good. Economic growth has averaged 8% a year, doubling GDP per capita. Achieving an economic transformation will be much more difficult. Economic transformation means moving from a single-sector economy to a structurally diverse one, and creating jobs. Liberia’s resource-based growth model requires a lot of capital, but doesn’t create many jobs. For jobs, you need manufacturing industry, which Asia has a lot of, but Africa very little.

One way to visualise this challenge is by looking at what biologists call a fitness landscape. Imagine a set of agents (organisms or economic actors) crawling over a landscape, looking for food, mutating and reproducing. The ones with the most useful mutations get more food and reproduce faster. As a result, they climb the fitness peaks in the diagram below. In a simple model, the actors can only move one step at a time. In a more elaborate one, they can also make occasional jumps. (I am drawing on Beinhocker’s ‘Origin of Wealth’ for this description of fitness landscapes; an interesting private sector example is here).


Liberia has managed to free itself from the trough of civil war and is crawling up a slope that leads to a resource peak. However, we suspect that the summit of the resource peak will be lower, and the chances of falling off will be greater, than some of the other peaks nearby. The objective of economic transformation is therefore to get onto a different, probably higher, peak.

What is the track record of countries that have tried to do this? It’s hard to find any. There are plenty of countries who are climbing up the ‘resource exporter’ peak, many of them Africa. There are others who are climbing up the ‘manufacturing’ peak, especially in Asia. But it’s hard to think of examples of countries who went part of the way up the resource peak, and then developed a manufacturing base as well.

There are two principal ways in which a country that is climbing up the resource peak can diversify its economy. The first is to maintain the resource dependence, but use the tax revenues thus generated to create a large public sector and create jobs in service industries that feed on the resources. Botswana seems to have done this in Africa, and Algeria may be another example. Jamaica, Nigeria and Angola are less successful examples.

The second is to shift from resource dependence to manufacturing, and thus diversify the economy. In Asia, it seems that Malaysia and Indonesia have done this: both are oil producers, but have also developed diverse manufacturing industries. I can’t think of good examples in Africa. In fact South Africa and Russia may have climbed down this peak a little in the last 20 years, as the opening of formerly autarkic economies led them to concentrate more heavily on resource exports.

I am looking for papers that investigate how to get from one peak to another, or data sets that help illustrate it.  Anyone who has any pointers, please let me know. As an alternative, I am considering using the World Development Indicators data to get the value added in different sectors to identify growth trajectories, which may shed some light on which countries have managed this transition, and which have not. Has anyone experience in using WDI for this sort of analysis, and have a sense on the accuracy (or otherwise) of the value added data?

Posted in Africa, Development, Economics, Liberia | Leave a comment