Are we using the right methodology to count the number of hungry people in the world? Are we really going in the right direction to end hunger?

Children queue for food at a refugee camp on the Burma-China border.
Children queue for food at a refugee camp on the Burma-China border, February 2015. Photograph: Jianan Yu/Reuters

This year, as the millennium development goals (MDGs) expire, the United Nations and the international NGO community are rallying around the conclusion that this has been “the most successful anti-poverty movement in history”. Poverty has been cut in half, they tell us, and hunger – the other big goal of the MDGs – has taken a serious hit, falling narrowly shy of the target. It’s a powerful story, and provides compelling evidence for those eager to convince us that the global economic system is basically on the right track; that whatever we’re doing, we need to do more of it.

But for anyone who has been following the hunger numbers, the UN’s conclusion is a little surprising. After all, until very recently they were reporting that the number of hungry people in the world had been steadily rising over the past two decades, not falling. When the world’s governments first set up the World Food Security pledge in 1996 with a vow to cut the number of people suffering from hunger in half, there were 788 million hungry people in the world. In 2009, according to their reports, there were 1,023 million, or about 30% more.

This trend has long been a thorn in the side of the United Nations, although they’ve managed to make the story a bit more palatable by shifting the goal posts. When the millennium declaration was signed in 2000, the goal was now to halve the proportion of hungry people rather than the absolute number, which made it easier to achieve. The MDGs also pushed the base year back to 1990, which allowed them to claim China’s progress during that decade even though it had nothing at all to do with the millennium campaign. Even so, it was clear to all that global hunger was getting worse, not better.

Then, at the end of 2012, the news changed. With only three years to go before the expiry of the MDGs, the UN’s Food and Agricultural Organisation (FAO) announced an “improved” methodology for counting hunger. And the revised numbers delivered a rosy tale at last: while 23% of the developing world was undernourished in 1990, the UN was pleased to announce a reduction down to 15%. The goal still wasn’t in reach, but at least the millennium campaign could finally claim some progress.

How did the FAO manage to transform a steadily rising trend into a steadily falling one? The technicalities are a bit opaque: it has to do with retroactively revising estimates about food supplies, and new assumptions about access to calories. The FAO has given its reasons for the switch: it states that when the 2009 and 2010 hunger numbers came out, they showed a “dramatic increase” in the number of hungry people due to the food price crisis that had been spurred by reckless financial speculation and sparked riots across much of the developing world.

“The jump in the number of undernourished, posited to have occurred in 2009, led commentators to voice concerns about the reliability of the FAO method to estimate the number of hungry. These concerns culminated in the request to FAO by the Committee of World Food Security to organise a technical roundtable to discuss the FAO measures of undernourishment.” The revisions ensued and the new numbers reported that undernourishment was constant – and even decreasing slightly – during the food price crisis.

The FAO has come under fire for the decision to change the methodology of a 25-year longitudinal study just three years before its conclusion, which is bad practice by any scientific standard. But in addition to this, a number of scholars have raised questions about the definition of hunger itself, claiming that it “gravely underestimates” the scale of the problem. The FAO counts people as hungry only when their caloric intake reaches rock-bottom levels, “inadequate to cover even minimum needs for a sedentary lifestyle” or “minimal activity” for “over a year” (around 1,800 calories per day). The problem is that most poor people don’t live sedentary lifestyles; in fact, they are usually engaged in demanding physical labour, so in reality they need much more than the FAO’s caloric threshold. The average rickshaw driver in India, for example, burns through about 3,000-4,000 calories per day.

The FAO itself recognises this flaw. It concedes that “ideally, undernourishment should be assessed at the individual level by comparing individual energy requirements with individual energy intakes,” but says that this approach is just not feasible when it comes to gathering data.

So what happens if we measure hunger based on the calorie needs for a population engaged in normal or “intense” activity? Using those standards, between 1.5 and 2.5 billion people are hungry, according to the FAO, which is two to three times as many as the millennium campaign would have us believe. And the numbers are rising, not falling, even according to the new methodology. And if we take China out of the equation, things look even worse. 73% of the progress that the UN claims against hunger comes from China, mostly in the 1990s, as a result of aggressive land reform. China’s gains mask the fact that most other developing countries have seen a net increase in the number of hungry people, even according to the FAO’s most conservative definition.

Other problems with the methodology? It only counts calories, so people who have serious deficiencies of basic vitamins and nutrients (which even the FAO admits affects 2.1 billion people worldwide) are not counted as undernourished. People who suffer from parasites, which inhibit food absorption rates, also fall through the cracks. And people who are hungry for periods less than a full year are for some reason not counted as hungry, as long as they consume enough calories to keep their hearts pumping.