Category: Food Security

Qualitative research was (already) here . . .

You know, qualitative social scientists of various stripes have long complained of their marginalization in development.  Examples abound of anthropologists, geographers, and sociologists complaining about the influence of the quantitatively-driven economists (and to a lesser extent, some political scientists) over development theory and policy.  While I am not much for whining, these complaints are often on the mark – quantitative data (of the sort employed by economists, and currently all the rage in political science) tends to carry the day over qualitative data, and the nuanced lessons of ethnographic research are dismissed as unimplementable, ideosyncratic/place-specific, without general value, etc.  This is not to say that I have an issue with quantitative data – I believe we should employ the right tool for the job at hand.  Sadly, most people only have either qualitative or quantitative skills, making the selection of appropriate tools pretty difficult . . .
But what is interesting, of late, is what appears to be a turn toward the lessons of the qualitative social sciences in development . . . only without actually referencing or reading those qualitative literatures.  Indeed, the former quantitative masters of the development universe are now starting to figure out and explore . . . the very things that the qualitative community has known for decades. What is really frustrating and galling is that these “new” studies are being lauded as groundbreaking and getting great play in the development world, despite the fact they are reinventing the qualitative wheel, and without much of the nuance of the current qualitative literature and its several decades of nuance.
What brings me to today’s post is the new piece on hunger in Foreign Policy by Abhijit Banerjee and Esther Duflo.  On one hand, this is great news – good to see development rising to the fore in an outlet like Foreign Policy.  I also largely agree with their conclusions – that the poverty trap/governance debate in development is oversimplified, that food security outcomes are not explicable through a single theory, etc.  On the other hand, from the perspective of a qualitative researcher looking at development, there is nothing new in this article.  Indeed, the implicit premise of the article is galling: When they argue that to address poverty, “In practical terms, that meant we’d have to start understanding how the poor really live their lives,” the implication is that nobody has been doing this.  But what of the tens of thousands of anthropologists, geographers and sociologists (as well as representatives of other cool, hybridized fields like new cultural historians and ethnoarchaeologists).  Hell, what of the Peace Corps?
Whether intentional or not, this article wipes the qualitative research slate clean, allowing the authors to present their work in a methodological and intellectual vacuum.  This is the first of my problems with this article – not so much with its findings, but with its appearance of method.  While I am sure that there is more to their research than presented in the article, the way their piece is structured, the case studies look like evidence/data for a new framing of food security.  They are not – they are illustrations of the larger conceptual points that Banerjee and Duflo are making.  I am sure that Banerjee and Duflo know this, but the reader does not – instead, most readers will think this represents some sort of qualitative research, or a mixed method approach that takes “hard numbers” and mixes it in with the loose suppositions that Banerjee and Duflo offer by way of explanation for the “surprising” outcomes they present.  But loose supposition is not qualitative research – at best, it is journalism. Bad journalism. My work, and the work of many, many colleagues, is based on rigorous methods of observation and analysis that produce validatable data on social phenomena.  The work that led to Delivering Development and many of my refereed publications took nearly two years of on-the-ground observation and interviewing, including follow-ups, focus groups and even the use of archaeology and remotely-sensed data on land use to cross-check and validate both my data and my analyses.
The result of all that work was a deep humility in the face of the challenges that those living in places like Coastal Ghana or Southern Malawi manage on a day-to-day basis . . . and deep humility when addressing the idea of explanation.  This is an experience I share with countless colleagues who have spent a lot of time on the ground in communities, ministries and aid organizations, a coming to grips with the fact that massively generalizable solutions simply don’t exist in the way we want them to, and that singular interventions will never address the challenges facing those living in the Global South.
So, I find it frustrating when Banerjee and Duflo present this observation as in any way unique:

What we’ve found is that the story of hunger, and of poverty more broadly, is far more complex than any one statistic or grand theory; it is a world where those without enough to eat may save up to buy a TV instead, where more money doesn’t necessarily translate into more food, and where making rice cheaper can sometimes even lead people to buy less rice.

For anyone working in food security – that is, anyone who has been reading the literature coming out of anthropology, geography, sociology, and even some areas of ag econ, this is not a revelation – this is standard knowledge.  A few years ago I spent a lot of time and ink on an article in Food Policy that tried to loosely frame a schematic of local decision-making that leads to food security outcomes – an effort to systematize an approach to the highly complex sets of processes and decisions that produce hunger in particular places because there is really no way to get a single, generalized statistic or finding that will explain hunger outcomes everywhere.
In other words: We know.  So what do you have to tell us?
The answer, unfortunately, is not very much . . . because in the end they don’t really dive into the social processes that lead to the sorts of decisions that they see as interesting or counterintuitive.  This is where the heat is in development research – there are a few of us working down at this level, trying to come up with new framings of social process that move us past a reliance solely on the blunt tool of economistic rationality (which can help explain some behaviors and decisions) toward a more nuanced framing of how those rationalities are constructed by, and mobilize, much larger social processes like gender identification.  The theories in which we are dealing are very complex, but they do work (at least I think my work with governmentality is working – but the reviewers at Development and Change might not agree).
And maybe, just maybe, there is an opening to get this sort of work out into the mainstream, to get it applied – we’re going to try to do this at work, pulling together resources and interests across two Bureaus and three offices to see if a reframing of livelihoods around Foucault’s idea of governmentality can, in fact, get us better resolution on livelihoods and food security outcomes than current livelihoods models (which mostly assume that decisionmaking is driven by an effort to maximize material returns on investment and effort). Perhaps I rest too much faith on the idea of evidence, but if we can implement this idea and demonstrate that it works better, perhaps we will have a lever with which to push oversimplified economistic assumptions out of the way, while still doing justice to the complexity of social process and explanation in development.

Posted on by Ed 6

Future challenges, future solutions

On Global Dashboard Alex Evans discusses a report he wrote for ActionAid on critical uncertainties for development between the present and 2020.  Given Alex got to distill a bunch of futures studies, scenarios and outlooks into this report, I have to say this: I want his job.
The list he produces is quite interesting.  In distilled form, they are:
1. What is the global balance of power in 2020?
2. Will job creation keep pace with demographic change to 2020?
3. Is there serious global monetary reform by 2020?
4. Who will benefit from the projected ‘avalanche of technology’ by 2020?
5. Will the world face up to the equity questions that come with a world of limits by 2020?
6. Is global trade in decline by 2020?
7. How has the nature of political influence changed by 2020?
8. What will the major global shocks be between now and 2020?
All are fair questions.  And, in general, I like his 10 recommendations for addressing these challenges:
1. Be ready (because shocks will be the key drivers of change)
2. Talk about resilience (because the poor are in the firing line)
3. Put your members in charge (because they can bypass you)
4. Talk about fair shares (because limits change everything)
5. Specialise in coalitions (and not just of civil society organisations)
6. Take on the emerging economies (including from within)
7. Brings news from elsewhere (because innovation will come from the edges)
8. Expect failure (and look for the silver lining)
9. Work for poor people, not poor countries (as most of the former are outside the latter)
10. Be a storyteller (because stories create worldviews)
I particularly like #10 here, as it was exactly this idea that motivated me to write Delivering Development.  And #7 is more or less the political challenge I lay out in the last 1/4 of the book.  #9 is a clear reference to Andy Sumner’s work on the New Bottom Billion, which everyone should be looking at right now.  In short, Alex and I are on the same page here.
I have two bits of constructive criticism to offer that I think would strengthen this report – and would be easy edits.  First, I think Alex has made a bit of a mistake in limiting his concern for environmental shocks to climate shocks.  These sorts of shocks are, of course, critical (hell, welcome to my current job), but there are other shocks out there that are perhaps not best captured as climate shocks on such a short timescale.  For example, ecological collapse from overuse/misuse of ecosystem resources (see the Millennium Ecosystem Assessment) may have nothing at all to do with climate change – overfishing is currently crushing most major global fisheries, and the connection between this behavior and climate change is somewhat distant, at best.  We’ve been driving several ecosystems off cliffs for some time now, and one wonders when resilience will fail and a state change will set in.  It is near-impossible to know what the new state of a stressed ecosystem will be after a state change, so this is really a radical uncertainty we need to be thinking about.
Second, I am concerned that Stevens’ claim about the collapse of globalization bringing about “savage” negative impacts on the developing world.  Such a claim strikes me as overgeneralized and therefore missing the complexity of the challenge such a collapse might bring – and it is a bit ironic, given his admonition to “talk about resilience” above.  I think that some people (urban dwellers in particular) would likely be very hard hit – indeed, the term savage might actually apply to those who are heavily integrated into global markets simply by the fact they are living in large cities whose economies are driven by global linkages.  And certainly those in marginal rural environments who are already subject to crop failure and other challenges will likely suffer greatly from the loss of market opportunities and perhaps humanitarian assistance (look at contemporary inland Somalia for an illustration of what I am talking about here).  However, others (the bulk of rural farmers with significant subsistence components to their agricultural activities, or the option to convert activities to subsistence) have the option to pull back from market engagement and still make a stable living.  Opportunity will certainly dry up for these people, at least for a while, as this is usually a strategy for managing temporary economic fluctuations.  This is certainly a negative impact, for if development does nothing else, it must provide opportunities for people.  However, this sort of negative impact doesn’t rise to “savage” – which to me implies famine, infant mortality, etc.  I think we make all-to-easy connections between the failure of globalization/development (I’m not sure they are all that different, really, a point I discuss in Delivering Development).  Indeed, a sustained loss of global connection might, in the long run, create a space for local innovations and market development that could lead to a more robust future.
So to “be ready” requires, I think, a bit of a broadening of our environmental concerns, and a major effort to engage the complexity of engagement with the global economy among the rural poor in the world.  Both are quite doable – and are really minor edits to a very nice report (which I still wish I wrote).

Posted on by Ed 1

Perspective

I sat through an outstanding FEWS-NET briefing today at work – some of the material falls under the heading of sensitive but unclassified (SBU), which basically means I can’t give details on it here. However, the publicly-available information from the briefing (link here – click on the near-term and medium-term tabs) makes it clear that there are really bad things taking place in parts of the Horn of Africa right now that are likely to result in large areas being extremely food insecure, which FEWS-NET defines as:

Households face substantial or prolonged shortfalls in their ability to meet basic food requirements. Reduced food intake is widespread, resulting in significantly increased rates of acute malnutrition and increasing mortality. Significant erosion of assets is occurring, and households are gradually moving towards destitution.

To summarize, people are dying due to food insecurity in the Horn of Africa right now, and it is going to get a whole lot worse for the next 6 or so months.
The briefing was very well run and presented, and the question session afterward was generally quite informative.  FEWS-NET is a remarkable tool – I think it is probably the best food insecurity assessment tool in the world right now – and I am engaged with thinking about how to make their assessments and projections even more accurate.  So I had a sort of technical disconnect from the meaning of the data during the briefing – to me, the numbers were data points that could be parsed differently to better understand what was actually taking place.
I returned to my desk, head buzzing with ways to reframe some of the analysis, but before I could get to writing anything down, an email came in telling me that the wife of one of my closest friends had passed away from ovarian cancer.  She was 41, and leaves behind my friend and their very young son.  For some reason, in that moment all of my data points became people, tens of thousands of mothers, fathers and children whose loss was beyond tragic.
That was it for me. I logged out, walked out of the office, and went to get my oldest daughter out of preschool an hour early.  Somebody needs to parse the data, to reframe and retheorize what we see happening in places like the Horn of Africa so we can respond better and reduce the occurrence and impact of future events.  But not me, not today.
Tomorrow, maybe.

Posted on by Ed 2

What else we don't know about adaptation

RealClimate had an interesting post the other day about adaptation – specifically, how we bring together models that operate at the global-to-regional scales with an understanding of current and future impacts of climate change, which we feel at the local scale. This post was written from a climate science perspective – and so focuses on modeling capabilities and needs as related to the biophysical world.  In doing so, I think that one key uncertainty in our use of downscaled models for adaptation planning is huge – the likely pathways of human response to changes in the climate over the next several decades.  In places like sub-Saharan Africa, how people respond to climate change will have impacts on land use decisions, and therefore land cover . . . and land cover is a key component of local climate.  In other words, as we downscale climate models, we need to start adding new types of data to them – social data on adaptation decision-making, so that we might project plausible future pathways and build them into these downscaled models.
For example, many modeling exercises currently suggest that a combination of temperature increases and changes in the amount and pattern of rainfall in parts of southern Africa will make it very difficult to raise maize there over the next few decades.  This is a major problem, as maize is a staple of the region.  So, what will people do?  Will they continue to grow maize that is less hardy and takes up less CO2 and water as it grows, will they switch to a crop that takes up more CO2 than maize ever did, or will they begin to abandon the land and migrate to cities, creating pockets of fallow land and/or opening a frontier for mechanized agriculture (both outcomes likely to have significant impacts on greenhouse gas emissions and water cycling, among other things)?  Simply put, we don’t really know.  But we need to know, and we need to know with reasonably high resolution.  That is, it is not enough to simply say “they will stop planting maize and plant X.”  We need to know when this transition will take place.  We need to know if it will happen suddenly or gradually.  We need to know if that transition will itself be sustainable going forward, or if other changes will be needed in the near future.  All of this information needs to be part of iterative model runs that capture land cover changes and biogeochemical cycling changes associated with these decisions to better understand future local pathways of climate change impacts and the associated likely adaptation pathways that these populations will occupy.
The good news* is that I am on this – along with my colleague Brent McCusker at West Virginia University (see pubs here and here).  Between the two of us, we’ve developed a pretty solid understanding of adaptation and livelihoods decision-making, and have spent a good bit of time theorizing the link between land use change and livelihoods change to enable the examination of the issues I have raised above.  We have a bit of money from NSF to run a pilot this summer (Brent will manage this while I am a government employee), and I plan to spend next year working on how to integrate this research program into the global climate change programming of my current employer.
Long and short: climate modelers, you need us social scientists, now more than ever.  We’re here to work with you . . .
*Calling this good news presumes that you see me as competent, or at least that you see Brent as competent enough to make up for my incompetence.

Posted on by Ed 0

Satellite Sentinels: We can do better than this (but it won't be as sexy)

The Satellite Sentinel Project released a report the other day that detailed what appears to be violence in the villages of Maker Abior and Todach in the Abeyei region of Sudan.  The imagery in the report is fairly standard DigitalGlobe 60cm stuff – and nothing fancy has been done to it to enhance analysis – it’s not clear if the imagery is even georectified, though given its largely illustrative use it probably doesn’t matter.  In the images are clearly burned buildings, and what certainly appear to be fortified areas where the Sudan Armed Forces are moving in equipment, fortifying defenses and improving storage facilities.  They claim to have imagery related to a parallel buildup of forces on the South Sudan side of the border.
But what do these images really tell us that good, on-the-ground intelligence does not?  Nothing.  In fact, I would argue that these images might be leading to unwarranted conclusions . . . or the Satellite Sentinel Project needs to do a much better job of explaining how the imagery enhances their conclusions.  For example:

  • How are the structures on the South Sudan side of the border representative of military buildup? Do they share a construction or layout with other known military encampments? Or is this conclusion completely supplied by on-the-ground intelligence?  If the answer is the latter, what exactly to these images add to the analysis?
  • How are the burned structures in Maker Abior and Todach linked to the military buildup in the subsequent pictures? There is no imagery of an attack in progress – and there will likely never be this sort of smoking gun evidence from this project. Data is gathered irregularly, and often at fairly wide intervals – so what you will end up with are a lot of before and after photos that can only be explained by on the ground intelligence.  In this case, it seems the on-the-ground intelligence has provided (at best) a weak link between this buildup and whatever happened in Maker Abior and Todach . . . but in presenting the imagery in this sort of a sequence, it appears that the evidence for the connection is much stronger than the data allows.

These are major issues that the project should be thinking through carefully.  Inadvertent misrepresentation of events on the ground will greatly damage not only this project’s legitimacy, but indeed any efforts to use remotely sensed data to identify/verify events on the ground in this region.
Please note: I am NOT suggesting that there is no violence in the region, or that what is happening isn’t hugely problematic.  However, I want our interpretations and responses to be based upon clear evidence, not loose circumstantial data strung together into potentially flimsy arguments about what has happened, and what might happen next.
So, what can we do about the problems in this region with this sort of data?  Well, for one thing the project might think about how to use its considerable remotely-sensed imagery resources to fill some significant gaps in data and interpretation about the political economy of natural resources in this region. Abeyei has a long history of conflict between different groups using natural resources for their livelihoods – especially conflicts that occur when pastoral/semipastoral groups move their cattle through agricultural areas, damaging fields (this is a thin distinction – really, most everyone in this region makes a living through a mixture of pastoralism and agriculture. The question is which group’s crops are impacted by the other’s cattle.).  This may be one of the most significant challenges facing this region – how to address this ongoing challenge, especially once there is a border dividing the transhumance routes these different groups have used to move their cattle to new watering and feeding areas.  Given the potential impact of a border on these routes, and therefore access to needed natural resources, we’ve already seen the Dinka to the south and the Messiriya to the north laying out territorial and resources claims far in excess of any previously recognized situation.  It is nearly impossible to adjudicate these claims because, as my colleague David Decker at the University of South Carolina – Sumter has argued, there is very little literature on the political ecology of this region.  The bulk of our understanding of natural resources, livelihoods and political economy that we do have are derived from colonial accounts more than a half century old.  With good intelligence, some serious on-the-ground research and the mobilization of people like David, and the integration of satellite imagery of the region that we can use to analyze (no more pretty pictures, just serious analysis) things like land cover, soil moisture, biomass, etc. we might at least create a stopgap for this knowledge gap that can then enable a settlement in this area that meets the widest range of livelihoods needs possible, lowering the potential for future conflict.

Posted on by Ed 1

On Math, Climate Change and Food Security

Idiot Tracker has a post on food security that uses food security as a means of focusing the reader on the challenges that climate change are likely to present in the near future.  In short, the argument goes that climate change will negatively impact our future agricultural productivity, making it difficult to increase that productivity as our population grows.  If we do hit nine billion people by mid-century (barring cataclysm this seems to be the minimum number we will hit), the author calculates that we will need to come up with 14.5 trillion calories per day, and notes that climate change is likely to present significant barriers to meeting this need.
I agree . . . in a general way.  We are losing huge amounts of arable land each year to soil degradation, and we are running out of productive places in which to extend new farms that do not create really problematic ecological tradeoffs (like massive deforestation that speeds climate change).  Climate change is likely to force the transformation of entire agricultural regimes in otherwise sustainable areas – for example, by changing temperatures and precipitation such that most strains of maize will have difficulty germinating in Southern Africa in a few decades.  This is all a very big deal.  But this post is also very, very thin on support for its argument.
As the post does not present any hard data, including how the 14.5 trillion calorie per day figure was derived, I cannot be sure if the author did any real math on our current production or the likely loss of caloric production that might occur under any number of likely climate scenarios (a problem unto itself, at global circulation models are much better for temperature than they are for rainfall, and there are few regional circulation models that can correct this problem – see the fascinating recent work of FEWS-NET on modeled versus empirically-measured patterns of precipitation in East Africa).  All of these might create significant error bars around likely future caloric production.  Further, I cannot tell if the author has considered whether or not crops will migrate as their ecological zones shift – surely farmers that previously could not raise a certain crop will start to take it up as the local environment allows and as other producing areas fall out of favor.  We know that some ecosystems will at least start to migrate if corridors for such movement are available – and agricultural systems are just another form of (heavily managed) ecosystem.  As cropping areas shift, what will the net caloric impact be?  It is not enough to say that we will lose a lot of calories when maize stops germinating in southern Africa.  We will need to get a net figure by calculating in all of the new areas in which maize will germinate.
Of course, such math only works at the global scale, and issues of hunger have very little to do with global production – hunger is local, shaped more by the intersection of markets, the environment, politics and society.  So noting that maize will germinate in new areas does nothing for the people in southern Africa who will be without maize.  However, we have to obtain another net figure: the lost calories from maize versus the new calories from new crops that people can grow, but chose not to before.  This may still total a net decrease in calories (indeed, it probably will), but this is not the same as simply subtracting maize from the equation.
Finally, what of plants that are edible, but that we currently choose not to eat?  The clearest analogy, to me, is the evolution of seafood here in the US.  I like to explain to my students that these new, exotic fish that are showing up at restaurants are the species that no self-respecting chef would touch two decades ago.  But when you wipe out the cod, you start getting creative.  And don’t get me started about tilapia.  It’s the rat of fish.  Seriously, it likes murky, stagnant water.  It will grow anywhere.  There is nothing I find funnier than hearing a server say “we have a very nice tilapia today.”  Yeah, I’d love to pay $20 for the swimming pigeon, thanks!  That said, people do eat tilapia and all sorts of other hilarious species because they are hungry and willing to pay.  So what new species of plant and animal will we be willing to eat a decade from now?  Three decades from now?  This is hard to predict, but I’ll bet quite a lot that we will find new species to exploit and offset even more of this caloric loss.
Despite all of this, I do think we face significant food challenges in the next three to four decades.  These will be felt very unevenly around the world, but they will be felt in significant ways.  To figure out what these impacts will look like, and who will experience them, requires that we carefully think through not only the exposure of crops to climate change impacts, but also the sensitivity and adaptive capacity of the agroecological system to those impacts.  It is only when we understand how such systems are likely to respond that we can begin to really plan for the challenges ahead.

Posted on by Ed 0

Necessary adjustments – but quant and qual still meet

The other day, I posted about the convergence between my own qualitative findings on the food security outcomes of food price instability and those of Marc Bellemare, Chris Barrett and David Just: that, at least in various parts of Africa, such instability was most likely to impact the middle and upper income cohorts more than the lower income cohorts of a given population.  However, I jumped too quickly in assuming that their dataset included rural and urban households – as Marc pointed out on his blog, they used a panel of rural household surveys.  So my initial argument about convergence does not hold up, as they did not consider the urban context in their work.
This is not to say that I am backing away from my assessment of the vulnerabilities of urban populations to this sort of challenge – I stand by it, having seen it, if only anecdotally, in towns and cities in Ghana over the past 13 years.  Urban populations are generally much more dependent on markets for their food supply than those living in rural areas (though this is not always true), and therefore price instability does create significant livelihoods uncertainty that is very difficult to manage, especially for the urban poor.  I therefore stand by my argument that we need to be keeping a close eye on the relative impact of price volatility on urban and rural populations, as the impacts of such volatility is likely to have very different impacts on these groups.
But recognizing that Bellemare et al’s work only addresses rural outcomes is not a problem for my argument about what I am loosely calling temporary deglobalization as a strategy for managing price instability (and price increases) – indeed, I think it strengthens the argument because it means that their dataset is now commensurate with mine, which was also rural.  As I argued in an extended comment on Marc’s blog:

The rural farmers most hit by price instability are those most integrated with global markets – the ones least able to deglobalize, as it were, when things get uncertain . . . Meanwhile, the bottom 60% is not as engaged with markets in which price volatility matters, and therefore can back away from them in terms of how they use their crops. In my work in Ghana, I found very few true cash crops (in the area I was working). Instead, some crops were treated like “cash crops” in years where price conditions and farm outputs of staple crops were favorable, and as staple crops when either prices were not favorable (including periods of volatility) or outputs of other staples used for subsistence were not adequate to meet household food needs. (Note: in many cases, the treatment of a crop as “cash” or “staple/subsistence” was highly gendered as well). The real difference between the rich and poor (relative terms in the Ghana sample) is the overall livelihoods strategy – one strategy (seen among the wealthier) is much, much more engaged in production for local markets, while the other (seen among the poorer) hedged market production with significant subsistence production (again, highly gendered). In years of volatility (or really in the face of most shocks), the market-oriented livelihoods were simply less resilient than the more diversified livelihoods strategies of the poorer households.

Or, as Marc himself noted in his response to my post on his blog:

[The wealthier] households tend to be hurt by price volatility because they are producers and therefore net sellers of most of (if not all) the seven commodities retained for analysis (i.e., coffee, maize, beans, wheat, teff, barley, sorghum).

So this means that the “temporary deglobalization” argument is not merely a rural-versus-urban argument, but one that can separate households in the same rural community.  This, I think, strengthens one of the arguments I was making in my original post:

  • Demanding that rural producers orient themselves toward greater and greater integration with global markets in the absence of robust fallback measures (such as established, transparent microinsurance and microsavings initiatives) will likely extend the impact of future price instability further into the poorest populations.
Posted on by Ed 1

Where Quant and Qual meet: On speculation, price instability and food insecurity

UPDATE: Marc Bellemare pointed out some issues with this post, which I have addressed here.  These issues, though, strengthen the argument about strategic deglobalization . . .

§§§§§§

There have been an interesting series of blog posts going around about the issue of price speculation in food markets, and the impact of that speculation on food security and people’s welfare.  Going back through some of these exchanges, it seems to me that a number of folks are arguing past one another.
The most recent discussion was spurred by a post on the Guardian’s Global Development blog by John Vidal that took on the issue of speculation in food markets.  In the post, Vidal argues that food speculation is a key driver of price instability on global food markets, which results in serious impacts for the poorest people in the world – a sort of famine profiteering, as it were.
The weakness of this post, as I see it, are twofold.  First, it doesn’t take the issue of price arbitrage seriously – that is, how speculation is supposed to function.  Aid Thoughts, via one of the comments on Vidal’s post, takes Vidal to task for this.  As Aid Thoughts/the commenter point out, the idea behind speculation is to pull future price impacts of shortage into the present, stimulating responses to future shortages before they occur.  Thus, a blanket condemnation of speculation makes very little sense from the perspective of one who wants to see food security enhanced around the world – without speculation, there will be no market signal for future shortage, creating a world that addresses shortages in a reactive instead of proactive manner. This is a completely fair critique of Vidal, I think.
However, neither Vidal nor those responding to him actually address the evidence for significant market manipulation, and the intentional generation of instability for the purposes of profiteering.  This evidence first emerged in a somewhat anecdotal manner in Fredrick Kaufman’s “The Food Bubble: How Wall Street starved millions and got away with it.”  In this article, Kaufman uses a fairly limited number of informants to lay out a case for the intentional manipulation of wheat markets in 2008.  It is an interesting read, though I argued in an earlier post that it suffers from trying to be a parable for the pervasive presence of complex investment vehicles in the modern world.  And in the end, its findings can hardly be called robust.
Though Kaufman’s argument might, by itself, be less than robust, it received a serious empirical boost from the International Food Policy Research Institute (IFPRI) in the fall of 2010.  In a discussion paper that remains underreported and under-considered in food security circles (trust me, it is difficult to get anyone to even talk about speculation in program settings), Bryce Cooke and Miguel Robles demonstrate quantitatively that the dramatic price rises for food in 2008 is best explained by various proxies for speculation and activity on futures markets.  Now, we can argue about how large an impact that activity had on actual prices, but it seems to me that Cooke and Robles, when taken in concert with the Kaufman piece, have demonstrated that the speculation we see in the markets right now is not merely a normal market response to potential future shortage – indeed, the Food and Agricultural Organization (FAO) of the United Nations has been arguing for months that there are no likely supply issues that should be triggering the price increases we see.  In other words, while it is foolish to simply blame price arbitrage for food insecurity, it is equally blind to assume that all of those practicing such arbitrage are doing so in the manner prescribed in the textbooks.  Someone will always try to game the system, and in tightly connected markets, a few efforts to game a market can have radiating impacts that draw in honest arbitrage efforts.  There is need for regulatory oversight.  But regulation will not solve all our food problems.
But this all leaves one last question unanswered: what is the impact of price instability, whether caused by actual likely future shortages or by efforts to game markets for short-term profits, on the welfare of the poor?  Vidal, Kaufman and many others assume that the impacts are severe.  Well, maybe.  You see, where matters (again – yep, I’m a geographer).  In a very interesting paper, Marc Bellemare (along with Chris Barrett and David Just) demonstrates that, at least in Ethiopia:

contrary to conventional wisdom, the welfare gains from eliminating price volatility would be concentrated in the upper 40 percent of the income distribution, making food price stabilization a distributionally regressive policy in this context.

This finding may be a shock to those working in aid at first glance, but this finding is actually intuitive.  In fact, in my book (out tomorrow!) I lay out a qualitative picture of livelihoods in rural Ghana that aligns perfectly with this finding.  In Bellemare et al, I would bet my house that the upper 40% of the population is that segment of the population living in urban areas and/or wealthy enough to be purchasing large amounts of processed food.  Why does this matter?  This is the segment of the population that typically has the most limited options when food prices begin to get unstable.  On the other hand, the bottom 60% of the population, especially those in this cohort living in rural areas (it is unclear from the study how much of an overlap between poor and rural there is in the sample, but I am betting it is pretty high), has a much more limited engagement with global food markets.  As a result, when food prices begin to spike, they have the ability to effect a temporary partial, or even complete, disengagement from the global market.  In other words, much as I saw in Ghana, this study seems to suggest that temporary deglobalization is a coping strategy that at least some people in Ethiopia use to guard against the vagaries of markets.  Ironically, those best positioned to effect such a strategy are the poorest, and therefore they are better able to manage the impact of price instability on food markets.
In short, I would argue that Marc’s (and his co-authors’) work is a quantitative empirical demonstration of one of my core arguments in Delivering Development:

2. At globalization’s shoreline the experience of “development” is often negative. The integration of local economies, politics, and society into global networks is not the unmitigated boon to human well- being presented by many authors. Those living along the shores of globalization deal with significant challenges in their lives, such as degrading environments, social inequality that limits opportunity for significant portions of society, and inadequate medical care. The integration of these places into a global economy does not necessarily solve these problems. In the best cases such integration provides new sources of income that might be used to address some of these challenges. In nearly all cases, however, such integration also brings new challenges and uncertainties that come at a cost to people’s incomes and well- being. (pp.14-15)

I’m not suggesting Marc endorses this claim – hell, for all I know he’ll start throwing things when he sees it.  But there is an interesting convergence happening here.  I’m glad I met Marc at a tweet-up in DC a few weeks ago.  We’re going to have to talk some more . . . I see the beginning of a beautiful friendship.
In summary, while efforts to game global food markets do exist, and have very serious impacts on at least some people, they do not crush everyone in the Global South.  Instead, this instability will be most felt by those in urban areas – in the form of a disaffected middle and upper class, and a large cohort of the urban poor who, lacking alternative food sources, might be pushed over the brink by price increases.  The policy implications are clear:

  • We need to be watching the impact of price increases on urban food insecurity more than rural insecurity
  • Demanding that rural producers orient themselves toward greater and greater integration with global markets in the absence of robust fallback measures (such as established, transparent microinsurance and microsavings initiatives) will likely extend the impact of future price instability further into the poorest populations.
  • We need to better understand the scope of artificially-generated instability and uncertainty in global food markets, and establish means of identifying and regulating this activity without closing price arbitrage down entirely.
Posted on by Ed 9

Liveblogging Dead Aid (Chapter 4)

After a few days off (a sort of sherbet for the mind, as it were), I’m back with Chapter 4 . . .
p.48: The chapter starts with a strong diatribe about the ubiquity of corruption in Africa.  First, it depends on where you are . . . and when you are.  Ghana in 1997 was run with small bribes.  Ghana now is navigable without much, if any, bribery – and a new generation of public servants is more efficient and transparent than ever.  Which leads to my next point . . . in the last chapter, Moyo warned against arguing that African culture somehow prevented development from taking root, and demanded we move past surficial explanations.  Here, however, she never interrogates why corruption happens – inadequate salaries of public servants, huge financial demands on the employed by extended families that lack access to social safety nets, etc.  By leaving this discussion out, Moyo is implying that Africans are inherently corrupt – and she is not moving past the surficial to interrogate causes.  Aid does not cause corruption to happen – aid is what is stolen when corruption exists.
p.50: Moyo is making staggeringly sweeping statements about how aid leads to corruption, arguing against the view that increased civil servant salaries reduces corruption.  She offers no evidence, just armchair psychology.  But there is evidence . . . that increased salaries help.  I’ve seen it myself, in Ghana.  It is not a magic bullet, but her dismissal of this corruption reduction tactic is unconscionable.  She’s just tossing away arguments that don’t fit her narrative.
p.51: Er, this isn’t Moyo’s fault (except that she is using it as evidence), but a study statistically examined the correlation between an ordinal scale of perceptions of corruption and economic growth?  Are you joking?  Do you know how many variables you’d have to control for to even begin to make that sort of analysis meaningful?
p.52: Wow, this is all sorts of loose correlation . . . OK, let’s say that 25% of all World Bank lending ever has been misused (as she claims).  First, is misused the same as stolen?  No – sometimes it was rerouted to other projects that were over budget, and might have had some productive outcome.  You have to capture that before you claim how much aid has actually been lost.  Second, this statistic does not really support the claim “vast sums of aid not only foster corruption – they breed it.”
In fact, let’s do some quick math here.  The World Bank had been making loans for 63 years at the time Moyo was writing.  Let’s say that an average of 110 countries a year received those loans (a low estimate, for sure), we have 6930 country/year data points.  Divide the $525 billion in total loans made by the Bank across this time, and we find out the average loan per data point (country/year) is  . . . $75 million.  Sorry, but this is not vast, by any stretch.
But let’s get concrete.  Ghana’s 2009 GDP was $29 billion.  That same year it pulled in $7.8 billion dollars in revenues.  Its net aid receipts were $1.2 billion.  Yeah, that’s a lot of money, but still only 15% of Ghana’s total revenues.  In the scheme of things, aid is not the big slush fund Moyo is trying to make it seem.
p.53: Holy crap, if you are going to point out we lend to corrupt governments, you might want to talk about why . . . and bring a real discussion of geopolitics to the table.  We lent to Mobutu because we feared the communists – everyone knows that.  So the problem wasn’t aid, it was the geopolitics driving bribes in the form of aid.
p.54: The section is title “Why give aid if it leads to corruption?”  Well, mostly because the links are pretty unclear, and because you’ve done nothing in this chapter to link them meaningfully.
To her credit, though, she is quite right about the agencies and how they value the size of the portfolio of lending, not the outcomes.  The World Bank has long been accused of this, and there is enormous pressure in every agency to get the budget spent on something . . . lest the budget be reduced next year.  However, USAID just took a huge step toward addressing this with Shah’s call for independent, transparent and publicly-available impact assessments of all projects.  Really crap projects will soon be visible to the public, and those responsible for them will be held to much greater account if this comes to pass.
p.55: Moyo has no idea what she is talking about on the Malawi food corruption issue.  As a result, she misapplies it to her larger argument that we lend regardless of corruption.  The issues of corruption in Malawi in 2002 had nothing to do with the food insecurity of the country that year – that was driven by the removal of a seed/fertilizer subsidy program at the insistence of the US and World Bank (who saw it as a market distortion).
p.57-58: And we are further into territory for which she seems to have no real understanding . . . the problem of government accountability is not really driven by aid.  The argument that aid reduces the need for taxes – and so the middle class and the population more generally could care less what the government is doing is astonishingly Western-biased (and neoliberal as hell).  The lack of responsiveness preceded aid, and persists because the state tends to lack the capacity to do anything for much of its population.  If anything, you could argue that aid has failed to improve state capacity such that the citizenry might feel bought in . . . but aid is not eroding civil society.
p.59: Mother of God, aid is what people are after when they try to take over a country?  Really?  Hell, even her example argues against this – Sankoh wanted the DIAMOND MINES, not aid.  She undermined her own argument – who the hell edited this book?
p.61-63: Well, yes, aid can be inflationary, causing problems for exports.  This is a problem that should be addressed.
p.64: Yes, inadequate absorptive capacity (the ability of a country to take up income of any sort and use it productively) can be a huge challenge in aid, and lead to waste and fraud.  But how often is it a huge challenge?  Note what I observed above – average annual World Bank lending, per country per year, is only $75 million.  That’s not a huge amount of money.  Absorptive capacity examples are much clearer in contexts where oil comes online quickly . . . which is why I am a bit concerned for Ghana at the moment.
p.66: OK, I’m getting worn out here by the overgeneralized, unsupported statements: “Aid engenders laziness on the part of African policymakers.”  Really?  All of them?
But what is the source of frustration here?  Keep reading, and you find this:

Because aid flows are viewed (rightly so) as permanent income, policymakers have no incentive to look for other, better ways of financing their country’s longer-term development.  As detailed later in this book, these options, like foreign direct investment and accessing the debt markets, offer more diversified and greater prospects for sustainable development.

This sounds a hell of a lot like an investment banker pitching a fund . . . oh, wait . . . she’s an investment banker.  Assuming Moyo believes that this really is the best way to go, it strikes me as remarkable how unreflexive she is about her own background and biases.
p.68: Oh, hubris: it seems that nobody has ever thought of an alternative to aid.  Really?  There is a lot of stuff in the later postdevelopment literature, all kinds of efforts to reimagine capitalism . . . now, we can argue about whether or not these are viable alternatives, but at least explore them before we run to the capital markets!
This is deeply frustrating – I like a controversial argument, but I also like a well-framed and supported argument.  We have the first part, but the second is completely absent thus far.

Posted on by Ed 1

Liveblogging Dead Aid (Chapter 3)

And the beat goes on . . . ladies and gentlemen, Chapter 3.
p.29: Well, so much for starting brightly.  She has grossly oversimplified Diamond (which is hard to to, y’all) to argue that a country’s wealth and success depend on geography and topography.  Er, no, that would be a form of environmental determinism.  Diamond was writing an anti-racist history of the world, explaining how the conditions that would eventually result in the ability of some groups to colonize others, etc., was enabled by environmental and geographic situations – but Diamond does not simply erase colonialism from the equation, he is trying to set the stage for how it came about.  You could argue that he has a somewhat environmentally determinist take on the causes of colonialism, maybe . . .
Oh, and for Diamond’s purposes, Africa was not resource-rich . . . it lacked easily domesticable crops and animals when compared to other world regions.  The whole discussion of squandering natural riches on page 30 is a total non-sequitor in the context of Diamond.
Note: I really don’t love Diamond’s book . . . and I am defending it here.  Ugh.
p.30: OK, the geographer in me just screamed.  I can’t blame Moyo for this – it is all about Collier, who along with Sachs and a few others in the field of economics is slowly resurrecting environmental determinism (or at least geographical determinism) with their damn correlations between coastline, endowment of natural resources, and economic growth.  The connections between these three issues are so complex that any analysis that simply divides countries into three categories (resource poor/coastline, resource poor/no coast, resource rich) is going to over-aggregate different relationships and causes into gross oversimplifications and false correlations.  Further, the damn N for these analyses is going to be less than 20 for one or more categories (less than 60 countries in Africa, folks).  I mean, you can run non-parametric stats on this sort of thing, but for the love of God, why?  Just do the qualitative work, dammit.
p.31: Moyo seems to have completely and utterly missed the reason why colonialism had such a brutal impact on African development.  Sure, artificial countries were not great.  And the inherited governmental structures after colonialism often caused problems.  But this sort of thing only really mattered after independence.  By then, these places had been completely restructured into sources of primary materials for the industries of the Global North – infrastructure, agricultural innovation, etc., all of it was aimed at enriching someone else and ensuring the colonized never developed any economic power of their own.  This led to the perpetuation of colonial relationships by other means after independence (neocolonialism), and I have little doubt this is way more important than the borders or governmental structures when we try to understand the growth trajectories of Africa since independence.  Either she is stunningly ignorant of her own country’s history, or this is a very disingenuous reading of African history.
p.32: Wonderful, Paul Collier postulates that the more ethnically divided the country, the more likely the prospect of civil war.  In other news, people with guns are more likely to shoot one another.  How much more likely?  Is this a cause unto itself, or a variable mobilized to political ends that can be better explained by another variable (I’m looking at you, Rwanda)?
p.34: If you are going to use Botswana as an example of a place where growth and development were facilitated by good institutions (which it was), you still have to contextualize the huge growth numbers by noting the GIANT DIAMOND MINES in the country.  I’m just sayin’.
p.35: Nondiagnostic diagnoses make me crazy.  “Africa’s failure to generate any meaningful or sustainable long run growth must, ostensibly, be a confluence of factors: geographical, historical, cultural, tribal and institutional.”  Again, no kidding.  This is meaningless.  Of course, it also discounts her previous example of Botswana having meaningful economic growth. Or Ghana. Or South Africa.  In other words, her whole statement is an overgeneralized negative that doesn’t hold up to scrutiny (or, in fact, her own argument from a page ago).   Next part of the diagnosis: “No factor should condemn Africa to a permanent failure to grow.” I don’t know of anyone making that claim.  If we were, we wouldn’t really bother with development, would we?  We’d just give up and walk away . . .  And the final part: “for the most part, African countries have one thing in common – they all depend on aid.”  Er, and colonialism (except maybe Ethiopia, and then mostly on a technicality.  And don’t tell me about Liberia – for God’s sake, we carved the place out to resettle freed slaves).  And colonialism has a lot to do with what CAUSED the situations we now address with aid.
I cannot, for the life of me, understand how she is ignoring this.
p.40: Yes, I am skimming a bit here.  That first bit really killed me.  But here I can give her some credit for hammering the “democracy gives us development” crowd – at least that portion of the crowd who thinks the relationship is simple.  It is not, of course, and some of the new thinking on this examines how, for example, governments can make difficult decisions that balance needed reforms/changes and their electoral interests.  But sadly, much of the mainstream writing on the subject tends toward the simplistic.
p.42-43: OK, I am now uncomfortable with what seems to be a bit too much lauding of dictatorships.  Yeah, they produce great growth numbers, but growth is a means to an end . . . improving the human condition.  Dictatorships tend to create large tradeoffs in quality of life that seem, on balance, to have negative impacts on their populations.  Not a lot of Chileans think back on Pinochet as the good old days, you know?
p.44: Moyo is quite right – the timing of aid, and inappropriate aid, can do much more harm than good.  For example, having food aid arrive nine months after a famine (not all that uncommon), just as the new harvest comes in, crushes local food prices (oversupply of free food drives prices of locally-grown crops) and re-impoverishes the local farmers.  But this is not an inherent problem of aid – this is about timing, something people are well aware of, and trying to address.  Further, Moyo’s complaint about celebrities bringing mosquito nets to the continent, and thereby putting local producers out of buisiness – while valid – steps outside her definition of aid (government-to-government transfers) that she laid out earlier in the book.  Apparently her terms of reference are not stable.  Super.
p.46: Moyo does not know what I feel in my heart of hearts, despite her claims – I do think aid can work.  Her evidence against it has to do with aid’s impact on various economic indicators.  But this is just means to an end, and does not capture many of the benefits of aid in a clear manner (reduced illness means a better quality of life, and might be partially captured in a growing GDP via the extra days the individual can work . . . but maybe not very clearly).  This isn’t to say that aid is perfect.  Hell, I wrote a book arguing that we don’t really know what it is we are trying to fix in much of the world, so I have my issues with aid and development.  I just want an honest reading of their impacts and drawbacks.

Posted on by Ed 1