How many people are hungry? David Steven
November 4, 2011 | More on Economics and development, Key Posts | 6 comments
The good news: poverty is in retreat. The bad news: hunger isn’t. That’s the headline finding for the first Millennium Development Goal , which aims to halve the proportion of people living on less than $1.25 a day and the proportion of people living in hunger between 1990 and 2015.
Great strides have been made on poverty, as I explained in a recent post, with the proportion of the poor projected to fall to 14.4% of the population of developing countries, from 41.7% in 1990. But what about hunger?
According to the UN’s 2011 assessment of the MDGs, the news is not good. In 1990, 828m people were hungry or 20% of the population of developing countries. Progress has been very slow since then:
The proportion of people in the developing world who went hungry in 2005-2007 remained stable at 16 percent [837m people], despite significant reductions in extreme poverty. Based on this trend, and in light of the economic crisis and rising food prices, it will be difficult to meet the hunger-reduction target in many regions of the developing world.
But hang on a minute. Why is the UN trotting out data for 2005-2007? That’s before the global food crisis, which hit at the same time as the financial crisis and has been just as slow to go away.
Food prices hit rock bottom in 1999, but then rose quickly with vicious increases in 2007 and 2008 (20% and 18%) and 2010 and 2011 (17% and 28%) as illustrated in the chart below. Yet we’re still relying on data from five years ago to estimate hunger.
The UN reported ‘dire’ news of a spike in its 2009 and 2010 MDG reports, with an estimate of more than 1 billion people hungry by 2009. But then it backed off in 2011, simply reporting the old data (which, oddly and without explanation, had been revised up slightly for all years, including 1990).
What gives? The problem is that our data on hunger are extremely patchy and rely on assumptions so heroic that I am left wondering if we are currently able to say anything useful about global hunger at all.
Here’s how it works at the moment. The target on food was set at the 1996 World Food Summit and predates the MDGs. The UN Food and Agriculture Organisation (FAO) has been measuring progress since 1999. Its measurement system has four steps:
- First, it estimates the minimum energy requirement for each member of a population, based on a complex algorithm that takes age, sex, weight, and height into account (with an added allowance for pregnant women).
- Second, it works out how much food is available for human consumption, based on a country’s Food Balance Sheet, which provides an estimate for how much food is available for each person, and how many calories, protein, and fat that food contains.
- Third, it uses household surveys – where they are available – to estimate how evenly food is distributed.
- Finally, it plugs numbers into a formula that estimates the proportion of the population below the minimum energy requirement cut-off point.
There are many problems with this methodology including:
- Micronutrients are left out entirely even though they play a huge role in nutrition, especially for the normal development of young children.
- We often don’t really know how much food a country has available – Food Balance Sheets are poor at capturing data on non-commercial food production; and at estimating how much food is being used for animals, is being stored in reserves, or is wasted.
- People aren’t that good at estimating their own food intake when asked in a household survey (let alone that of family members), while countries do surveys infrequently, if at all. As a result, estimates of the distribution of food are, at best, educated guesses. Moreover, and as far as I can tell, FAO has not been good at documenting which countries have surveys, or from when.
- It’s hard to know how much food people need – especially as they’ll become less active as their food gets scarce (thus burning less energy).
- In countries where many people go hungry, a large proportion of the population hovers just above and below the minimum energy requirement, making estimates of hunger extremely sensitive to small changes in underlying assumptions.
- Perhaps most importantly for the current crisis, price is not fully considered. As staples become more expensive, presumably the poor consume less and the distribution of food changes markedly. Hunger may therefore grow much faster than suggested by the fall in average food consumption. Surveys have no hope of capturing the impact of volatile prices.
FAO’s figures have faced sustained criticism for at least a decade. Back in 2004, a somewhat huffy note from a FAO statistician defended its methodology as ‘the best available’ and dismissed various ‘methodologically incorrect’ alternatives.
More recently, however, the damn has burst, with FAO sent back to the drawing board in 2010, by the Committee on World Food Security. We are promised revised statistics that will improve modelling of the impact of price increases and income shocks, strengthen food balance sheets, integrate more household surveys, and include micronutrients and other factors in the mix.
But in the meantime, the presentation of data is suspended. Estimates for the number of undernourished people in 2009 and 2010 have been withdrawn, and no figures for 2011 have been prepared.
In the midst of the first ever global food crisis, in other words, the lights have been turned off. 837m people were probably hungry four to six years ago. Maybe. That might have gone up above a billion, or perhaps it didn’t. Hunger is either resurgent or it isn’t.
Of course, leaders are using the old figures without too many scruples. Here’s Ban Ki-Moon from just last month:
Even before the food crisis began, eight hundred million people were going to bed hungry at night. Now, a staggering nine hundred and twenty-three million people suffer from chronic hunger and under-nutrition.
But I think it’s pretty clear that, until the FAO comes back with new and better data, the ‘correct’ answer to the question ‘how many people are hungry?’ is – ‘we simply don’t know’. Apparently, the matter will be considered at an International Scientific Symposium in Rome in January next year.
Let’s hope FAO pulls its finger out and we don’t go that much longer without any data.
I worked at the FAO for a year and remember all sorts of arguments – in both good and bad faith – about this. It was at the height of the food price crisis, and FAO was becoming known, often with good reason, as a reliable source for food stats. And yet, as you argue here, the most important single statistic was so deeply compromised that it was almost meaningless. Indeed, I know from informal chats that at least some well-informed folk at the top viewed these numbers as vital political tools rather than (necessarily flawed) attempts at objectivity.
That said, FAO was put in an impossible position that will be familiar to many technocrats: political leaders (i.e. its funders and the governments it is meant to be accountable to) wanted immediate headline figures, not fancy explanations of uncertainty. But at the same time, the historical defunding process was continuing: I doubt there was scope for extra number crunchers, let alone …
[continued from above]
… support for surveys and the like.
(Interestingly, rich-country governments were also in the process of pushing the Organisation towards a focus on passive stats production rather than active intervention. That, however, is another tale entirely.)
Personally, I think FAO is broadly quite good on stats, though it can fall prey to the same oversimplification practised by, well, almost every other public body. And if GD or anyone else can provide better metrics, there are still many good-faith FAOists that will receive them gratefully.
i show the first scenario for the hungry people. Read By this blog, know for the people for the hungry people.
Personally, I think FAO is broadly quite good on stats, though it can fall prey to the same oversimplification practised by, well, almost every other public body. And if GD or anyone else can provide better metrics, there are still many good-faith FAOists that will receive them gratefully.
Personally, I think FAO is broadly quite good on stats, though it can fall prey to the same oversimplification practised by, well, almost every other public body. And if GD or anyone else can provide better metrics, there are still many good-faith FAOists that will receive them gratefully.
Hello there, just became aware of your blog through Google, and found that it is really informative. I am gonna watch out for brussels.
I will appreciate if you continue this in future. Numerous people will
be benefited from your writing. Cheers!