Until quite recently, the Rome headquarters of the Food and Agricultural Organization (FAO), the United Nations' flagship food agency, was draped with enormous banners that read: "1,000,000,000 live in chronic hunger and I'm mad as hell."
The banners were filled with black-and-white images of painfully malnourished children, and the overall effect was arresting, even if the one-billion figure seemed suspect. A number starting with a "b" had much more impact than one starting with an "m."
On Tuesday, the FAO rolled out its new estimate for the number of "chronically malnourished" people in the 2010-2012 period and – guess what? – it is no longer the nice, round one billion. It is 870 million, still shockingly high but an improvement of 130 million over the previous figure.
Before you cheer, the new figure does not necessarily mean that 130 million have dropped off the hunger rolls since the FAO removed its "mad as hell" banners earlier this year. The FAO explained that it is using new measurement methodology, meaning the old estimate and the new estimate cannot be compared. The new measurement says the number of hungry has dropped from one billion in 1990 to 870 million today, not from the time the banners were removed.
The FAO is thrilled, because it means that the UN's target of halving the undernourished in the developing world by 2015 – the so-called Millennium Development Goal – is suddenly within reach. Only a year ago, with food prices surging, achieving that goal was unthinkable.
But if the one-billion figure was suspect, why not the new one?
Indeed, the FAO gets most, if not all, of its food data – from crop production to consumption – from national governments. The data is not always reliable. Some governments simply lack robust data-collection systems. Others have been known to pad certain figures. A country experiencing severe hunger might be tempted to exaggerate the number of hungry citizens, all the better to get more aid. Crop production figures are notoriously unreliable. The FAO learned that a certain African country magically reported a 400 per cent increase in cereal production one year. Magic it was.
The FAO admits that the data it collects from governments is not always iron-clad. So it is doing more checking. For example, the FAO takes the UN's new estimate for the population of Bangladesh, one of the world's poorest countries, into account when generating its hunger data. The UN believes Bangladesh's population is about 10 per cent lower than the official estimate.
It is also measuring "chronic" undernourishment as opposed to "episodic" undernourishment. The latter can happen during extreme price spikes, like the one that happened in 2007 and 2008, triggering food riots in dozens of countries. But since spikes by definition are short-lived, they do not belong in a long-term estimate, the FAO argues, and it has a point. The high prices generated more food production rather quickly.
Other changes in the methodology include new thresholds for caloric undernourishment and better estimates for "food losses," that is, the losses that occur during the farm-to-fork distribution phase.
The new methodology is no doubt an improvement even if a cynic would argue that it serves the UN's purposes to roll out figure that give the impression that the three main UN food agencies are finally making some progress on the war on hungry after a few scary years, when food commodity prices went wild and the threat of mass hunger was real.
The new methodology is something the FAO had to do for its own credibility – the one billion hungry figure was just too convenient. Hunger remains a huge problem, especially in sub-Saharan African, where the number of hungry continues to rise. Better number-crunching must make it less of a crisis than it was a few years ago. The question is: What will the FAO's next methodology bring?