26 August 2013

Does it wash? A new Cochrane Review of WASH interventions and nutrition status of young children

Just out, a new review by Alan Dangour and colleagues at the London School of Hygiene and Tropical Medicine of interventions to improve water quality and supply, sanitation and hygiene practices and their effects of the nutrition status of under fives.

Their indicators are under 5 weight for age, weight for height and height for age. After a systematic, thorough and transparent search process they identify 14 studies with good evaluation designs.

These are very good researchers, so what did they find?

These are the authors' conclusions:

"This review provides evidence that some water, sanitation and hygiene (WASH) interventions (specifically solar disinfection of water, provision of soap, and improvement of water quality) may slightly improve height growth in children under five years of age. The quality of the evidence is generally poor and the overall estimates presented are based only on meta-analyses of data from interventions of relatively short duration (9-12 months) from only a small selection of possible WASH interventions. These estimates are therefore not applicable to the effect that wider WASH interventions may have on child nutritional status."

A bit depressing, no? Some interventions, may slightly improve height growth in children under five. The authors conclude that the interventions have too short a duration, and there are not enough good evaluations.

Lets look a bit closer at the height for age results:
  • 12 studies
  • 8 countries (4 in South Asia, 3 in Sub Saharan Africa, 1 in Central America)
  • 8 studies from urban/slum/peri urban settings, 4 from rural areas
  • 8 studies find a positive effect (2 fairly large peri urban South Africa, rural Cambodia and rural Ethiopia), 1 finds no effect and 3 find a negative effect (and this is with a control group to take care of differences in differences)
So, you are a policymaker--what do you make of this?

First, don't give up on WASH Interventions--the theory of change is very strong and this is only a small set of intervention types and settings.

Second invest in good intervention designs--some of the interventions worked, now find out why.

Third, for goodness sake, stick with the interventions beyond 12 months--why would you expect profound behaviour change in less?

Finally, fund a good evaluation--poor evaluations save money in the short run but corrupt the decision making process and squander resources in the medium run.

Do these results wash? Unfortunately they do. The authors end on a hopeful note--there are currently studies underway on interventions that have a chance to work and where an impact, if present, has a chance to be detected. 

Are you frustrated with such reviews? I am, but they can only survey what is out there. If they serve to stimulate the funding of better interventions and evaluations they will have served their purpose. Don't wash your hands of them--just be really careful how you use them.

19 August 2013

Indian Malnutrition--A Little Less Conversation a Little More Action

This is the week Elvis Presley passed away back in 1977. He famously sang "A little less conversation, a little more action".

These lyrics come to mind when trying to explain malnutrition in India. Everyone has their own ideas, and they tend to get called "enigmas".

In a short perspective paper prepared for the Indian Health Report (IHR), led by IDS partner the Public Health Foundation of India and out later this year, I review the different enigmas: is it the standards, women's status, agricultural disconnects, open defecation, weak health system etc.? 

The biennial IHR, aimed at a broad national and international policy and academic audience, will provide a periodic assessment of health in India, with each issue focusing on a specific area of health challenge. The inaugural issue on Nutrition will provide a rigorous analytical overview of the current trends, challenges, and puzzles related to maternal and child nutrition in India, and highlight the role of policy in improving a wide range of nutritional outcomes, especially at the state level.

As I said, in my contribution I run through the various enigmas. Some are more convincing than others in terms of the evidence, but the biggest enigma is why the Government of India is not more exercised about doing something about malnutrition.

As Elvis might say we're still waiting for Delhi to be All Shook Up about malnutrition in India.

16 August 2013

Some Reading Suggestions and Lists (but not for the beach)

Recently a senior policymaker wrote to me asking for some recommendations of IDS publications they should read. I spent a couple of weeks on and off thinking about the most dazzling things to recommend, but then gave up on that and remembered the wealth of material in the IDS Working Papers.

I picked 4 papers from the last 6 months. I share them with you. 

1. Revenue Reform and Statebuilding in Anglophone Africa
Moore, M. IDS Working Paper 428
I like this paper because it is on a highly relevant topic for the next 10 years—how do we transition from aid to domestic resources in SSA? The transition will not be smooth. This paper finds that tax revenues have not increased that much in response to investing in tax authorities and that a more formalised tax sector gives larger private sector companies more opportunities to set the rules.

2. From Poverty Traps to Indigenous Philanthropy: Complexity in a Rapidly Changing World.

Jarrett, S. IDS Working Paper 425
I like this one because it discusses indigenous philanthropy, an issue that is not yet so important, but will become increasingly so in the next 10 years.

3. Rent Management – The Heart of Green Industrial Policy.

Schmitz, H., Johnson, O. and Altenburg, T. IDS Working Paper 418

A paper that cleverly draws on many different literatures to serve as a guide to governments wanting to bring about a smooth transition to renewable energy

4. Gender Equality and Economic Growth: Is there a Win-Win?

Kabeer, N. and Natali, L. IDS Working Paper 417
A review paper that concludes that gender equality is more likely to contribute to economic growth than economic growth to gender equality. Of course the interesting question is does the type of economic growth generated by gender equality lead to greater gender equality? This paper focuses us on one dimension of the quality of growth.

Key Readings for Students
In addition, IDS has for the first time published our Reading Lists for our Master's courses. This is pretty standard practice in the US but in the UK we have been more cautious. No more.  If you want to copy our reading lists, go ahead--sincerest form of flattery and all that.

So, lots of reading, but if you are spending serious time on the beach I can recommend Sharp Objects by Gillian Flynn and The Bat by Jo Nesbo. 

13 August 2013

Call that a Narrative? This is a Narrative! The need for researchers to get beyond Crocodile Dundee

In the research and policy worlds, narratives duel.  Your narrative will do battle against mine.  However, there are too few judges of these narratives, only crazed academic knife fights.  So it is refreshing to find two recent papers that assess narratives in a dispassionate way.

The first paper is by Lisa Smith and is coming out as an IDS Working Paper and focuses on the Great Indian Calorie Debate.

The puzzle is that calorie availability in India, as measured by existing surveys, shows a steady decline over the past 15 years. Various reasons --or narratives--have been put forward for this decline: lower physical activity levels, greater poverty which is constraining purchases, and food inflation doing the same thing. 

Lisa's paper takes a new angle, she asks "what if the surveys are missing food eaten away from home?".  Using existing data, she goes through a series of analyses that explore the prevailing narratives --casting doubt on each--and then examines her own. Her analysis is thorough and compelling--it looks to me like she is on to something.  Could this great debate really be settled by something as ordinary as a flaw in a questionnaire design? I think there is definitely merit in this argument.

The second paper is by Stefan Dercon (Prof at Oxford and also DFID Chief Economist) and has just been published in the journal Agricultural Economics. In the paper Stefan laments the simplistic narratives that are put forward for agriculture in sub-Saharan Africa. 

Is agricultural growth really necessary for economic growth? Is it really the most poverty reducing kind of growth? Is raising smallholder productivity really the best way of getting this economy jolting, poverty reducing growth? 

The article points out that in many contexts these assertions may hold true, but that it is highly unlikely that they do in all contexts (sub-Saharan Africa has hundreds of different agroecological contexts). And by asserting they hold true in all contexts, then we are crowding out the kind of analysis that allows some nuance into the debate. Is agricultural development really the most cost-effective way of advancing economic growth and reducing poverty and hunger? Is smallholder investment the best way to go about this? Even if we assume that, on average, probably yes (note, that apart from some work by Shenggen Fan and others at IFPRI from about 7 years ago for Uganda we don't have the evidence on cost-effectiveness) nevertheless the averages are surely blinding us to many new opportunities for thinking about (and investing in) agriculture in Africa. Truth be told, the evidence base on African agriculture is too weak to find answers to these questions. Agricultural research funders need to be braver in testing their own assertions.  

Of course Lisa and Stefan have introduced their own narratives into the debates, but they have done so after giving a fair hearing to the competition--and that is all too rare.

11 August 2013

Review of Why Nations Fail. Guest Blog by my Father in Law

My Father in Law, Dr. George Lambrakis a career US diplomat who worked in Guinea Bissau, Lebanon, Vietnam, Laos, Swaziland and Iran amongst other places, has been telling me how much he enjoyed reading the Acemoglu and Robinson book Why Nations Fail. So I asked him to do a review for this blog, which he kindly did.  Here it is.  It is interesting to get a diplomat's view of a development story. Enjoy.  
"Why Nations Fail: The Origins of Power, Prosperity and Poverty” is a  memorable new study on how nations succeed or fail to develop by economist Daron Acemoglu and political scientist James A. Robinson. As a retired American Foreign Service officer specializing mainly in politics I have accepted the usual theories of economic development as a given, and grist for my colleagues specializing in economics to mill. This book is an eye-opener.“

Forget about modernization breeding democracy from successful economics. Forget about authoritarian regimes moving faster than democracies toward sustainable development. Forget about the “ignorance theory” that you just need to teach effective development to leaders of less developed (not “developing”) countries. Forget even geopolitical or cultural theories that seek to explain success or failure. Think, rather, of the need for “inclusive” political institutions to exist or develop if innovative economic institutions are to emerge and not be strangled at birth by “extractive” elites.

Inclusive political institutions permit and reinforce inclusive economic ones if not blocked by extractive institutions created by kings, emperors, dictators, aristocrats and other agents of political exclusion or repression. These by their very nature react against and try to smother the sustainable economic development that requires Schumpeter's “creative destruction” - since innovations necessarily disrupt comfortable monopolies and other institutions controlled by the ruling elites.

This is not to say that economic development does not frequently make some progress under extractive rulers. It just runs out of steam, as the authors indicate through innumerable – and impressively researched – examples from all parts of the world (with emphasis on Latin America and Africa) and through millenia of history. The Soviet Union or Communist China can achieve much catch-up growth utilizing other peoples' economic innovations, but they eventually run into a blank wall by resisting the creative destruction that will hurt the leadership's selfish interests – unless they are forced one way or the other to reform.

Inclusive institutions grow from largely contingent small differences like Britain's 1688 “Glorious Revolution” among national societies (e.g. the British or Dutch as opposed to most others) that take advantage of great “historical turning points” (economic disruptions) such as the medieval Black Death, the opening of Atlantic trade, or the Industrial Revolution.

There is no recipe for building inclusive institutions. But history indicates (p.460) the need for 1) “a degree of centralized order” to prevent chaos, 2) “preexisting political institutions that introduce a modicum of pluralism” and 3) the “presence of civil society institutions that can coordinate the demands of the population” - especially empowered media.

Today's Brazil and Botswana are good examples. Fujimori's Peru and a variety of chaotic or dictatorial countries in Africa and places such as Afghanistan and others in Asia are bad examples.

Explaining how history operates is what this book is about. It does not really concern itself with how creative destruction can in the short term hurt the unemployed poor, before improving their lot in the long term.

However, the authors do take a crack at foreign aid in the final part of their summarizing chapter.  Aid agencies do not pay enough attention to the countervailing power of extractive elites in the countries whose poor they are trying to help. Conditional aid sounds good – but recipient governments get the same aid even if they do not meet the conditions, because their people are so poor. Unfortunately, such wastage is compounded by the plethora of agencies and middlemen who also take their cuts for overhead along the way, leaving only 10 to 20 per cent of donations for the poor – though even that is better than nothing, especially if targeted, as for schools.

For me, Acemoglu-Robinson's argument is a convincing one as far as interpreting history goes. The issue becomes how to apply it in practice. 

For example, there is little doubt that much if not most of the world's progress has taken place under extractive governments who, using this book's terminology, were playing catch-up following innovations produced by others – or perhaps themselves at an earlier time – who enjoyed inclusive institutions. The authors themselves support this interpretation when they refer to the “iron law of hierarchy” by which groups overthrowing one extractive regime often seize its institutions to profit themselves from their extractions – as many post-colonial regimes did in Latin America, Africa, the Middle East, and Asia.

So how do those who wish to help the poor avoid simply filling the coffers of extractive elites? The authors warn against trusting common theories of development. But they cannot suggest a viable alternative apart from looking for countries that have somehow obtained the prerequisites for inclusiveness through accidental “contingencies” of history.

Perhaps in a follow-up study the authors might try to fathom how the favorable contingencies in, say, Britain's history might be reproduced in other countries suffering under extractive regimes today. They make a promising start on this quest when they analyze how Botswana's relative poverty protected  it from colonial or other extractive invaders and how “Lula's” rise transformed politics in Brazil."

02 August 2013

Development Research: The Shape of Things to Come?

A colleague of mine outside of IDS recently asked me for my opinion on development research--what was the future demand going to look like and what would need to happen to meet that demand? Of course these are just my half-baked opinions, but I thought you might find them interesting.

Dear, x, you asked about the shape of future research, here are my views on why we do research, how we do it, what we do it on and how it is used and evaluated.

Why are we doing the research?
I think 2 types of problems will increasingly drive the research agenda: (1) common problems and (2) collective problems. Common problems are development issues that all countries—rich and poor—are facing. Examples include dealing with inequality, connecting citizens with the state, tacking obesity, care provision to the elderly. Collective problems are things that affect everyone and require collective action. Examples include: climate, tax flows, migration, and illegal trade in drugs, arms, people.

How do we do the research?
The Common and Collective research problems will require big changes to the research infrastructure.
  1. Partnerships. For the Common problems, we will need more N-S-E-W kinds of collaborations for comparative work—the partners will be more traditional research partners and will be issue driven. For the Collective problems, the partnerships will be more strategic and more networked—to get a different perspective on a problem that is bigger than any single nation. The partners will tend to be less traditional "development" partners, as many of these issues go way beyond self defined development issues: security forces, businesses, religious groups—all will need to be engaged in various ways
  2. Funding. For the Common problems, we will need funders to tear down the walls between domestic and international issues. For Collective problems, we will need funders to take the long view—10 years or so and for them to back the researchers--invest in the network and the broad goals, but let them get on with it and respond to uncertainties as they emerge.
  3. Journals. The common and collective problems are difficult for journals that are fragmented along domestic/international lines and along disciplines. They need to become more issue driven and less discipline driven.
  4. Research uptake. Both types of problems will entail thinking differently about consumers of research and ways to help them consume it in the right way. The Common research problems will require finding research users who are open to experiences beyond their borders. The Collective problems will require the identification of research users who realise that their actions have consequences beyond their borders, consequences that will eventually come back to bite them unless addressed.
What to do research on?
Everyone will have their favourite issue. But in the Common problem space, non-communicable disease is an issue that is expanding rapidly and one for which we have few good public policy interventions. The challenges of rapid urbanisation, disaffection with elected politicians and harnessing of mobile technologies for more sustainable development represent other common problems. On the Collective side, I listed some issues above. I think we need to add hunger reduction to the list of collective action problems. It is increasingly clear that unless you are a very large country with a massive domestic market (e.g. Brazil, China) it is very difficult to reduce hunger at the national level without a great deal of help from the international architecture (e.g. regulation of biofuels, biotech, trade, intellectual property rights, international land and water acquisitions).
So what? How do we assess the impact of research? 
Fuelled by the recession and the need to demonstrate VFM (almost before the research has begun), research funders have tried to force the issue by insisting on an increasingly narrow set of tools for defining evidence that is good enough to be used to influence policy. While I welcomed the introduction of RCTs and Systematic Reviews into the development toolbox 4-5 years ago, in my view, their advocates have gone way too far and have overplayed their hand. Over and over I see excellent research being blanked out of decision making because it does not fit the narrow definitions of RCTS and Sys Revs. Bizarrely, senior researchers within funding organisations are advocating these methods in a one-eyed way without themselves ever having done one! I can only hope the pendulum will swing back to the middle--and soon.