People stand in line as they wait to buy staple items and food near a supermarket in Caracas, Venezuela, on Sept. 14. (Henry Romero/Reuters)

Hunger is one of the world’s most pressing development problems. It should have been great news when the United Nations announced at the close of the Millennium Development Goals (MDGs) that 72 countries had cut their hunger rates in half since 1990. The problem? We haven’t decided what “hunger” is, nor is there agreement on how to measure it. This means that the United Nations’ old goal of cutting world hunger in half, and its new goal of achieving ‘zero hunger’ are open to all kinds of politics and manipulation.

The United Nations has honored Venezuela for winning the fight against hunger. That seems wrong.

The MDGs used two measures of progress toward the goal of halving the proportion of hungry people in the world. The first is the ‘prevalence of undernourishment’ (PoU) indicator produced by the U.N. Food and Agriculture Organization (FAO). The second is the percentage of children under 5 who are underweight, based on calculations provided by the World Health Organization and UNICEF.

Last year, the FAO held an award ceremony to honor countries that the FAO claims have met their MDG goal using these numbers. Countries such as Venezuela, Angola, Nepal, Egypt and Bangladesh were among those honored, with the FAO proclaiming that Venezuela and Egypt, among others, had reached the lowest category of hunger possible on their scale. These countries had brought the proportion of hunger down to 5 percent or less.

Some of these countries probably didn’t deserve the honors. For example, as the Economist says, there is a lot of evidence that hunger has gotten worse rather than better in Venezuela, a country in the middle of an economic meltdown. The numbers don’t reflect the evidence on the ground. Even under the FAO’s most recent data, Venezuela is still listed as having a hunger rate of less than 5 percent — among the lowest in the world. That suggests that there is something fundamentally wrong with the way that the PoU measures hunger.

The problem of unreliable measures of hunger is about to get worse.

However, U.N. measurements of hunger may be about to get more problematic. World leaders are meeting again this month at the U.N. Sustainable Development Summit to adopt the successors to the MDGs, which will be called the Sustainable Development Goals (SDGs). There is good reason to be worried that these new goals will ignore the serious problems with our measures. Instead of looking at the underlying problems in some of our hunger indicators or acknowledging that different measures reflect different basic understandings of what hunger is, the SDGs open the floodgates by proposing 14 indicators, corresponding to eight targets, aiming to “end hunger, achieve food security and improved nutrition and promote sustainable agriculture.”

Too many different and incompatible measures can mask what is happening rather than reveal it. What happens when measures don’t line up with one another? The graphs below show what happens when we look at just four of the proposed indicators for the SDGs side by side for the same country in the same time period. The colored bars represent how well or poorly a country does on measures of hunger. (Not all metrics are collected every year — see here for more information on where the measures come from).

The results are startling — different measures of hunger give wildly different impressions of the situation in a given country. This is in part because the proposed measures have different definitions of what hunger is and how best to measure it. But the differences are jarring, even if, for example, we focus on the three metrics attempting to estimate severe food insecurity (the orange, gray and yellow lines). There is just no agreement on who’s hungry in a given country.

It’s not clear how the United Nations proposes to reconcile these measures. Is it going to prioritize some measures over others? Who will decide this and on what basis? Could this allow states to play the numbers, picking and choosing the measures that are most flattering to them? For example, Angola could have a hunger rate of 62.4 percent or 29 percent or 14.2 percent, depending on which measure you picked. Thirteen percent of children under 5 in Venezuela have stunted growth, but the PoU indicator suggests they’ve nearly eliminated hunger.

The uncomfortable truth is that there is no consensus among the international community on what hunger is. When we say there should be “zero hunger,” what do we actually mean by that?  Do we mean zero hunger that is so bad that it’s leaving irreversible damage to the human body? Or do we mean that nobody should have their children go to bed feeling the sensation of hunger, even one night a week, because they are unable to afford to put dinner on the table?  Are we okay with parents who may be able to put food on the table but can’t afford meat, for instance, or fresh vegetables?  If we saw this occurring, would we say, “No, this is okay. This isn’t ‘real hunger?’ ”

The United Nations is trying to fix this with a plethora of incompatible measurements, but the underlying problem is that there is no international agreement on what hunger is. Until the international community can agree on the underlying question of what kinds of experiences with hunger are acceptable, and what kinds are morally and ethically unacceptable, we won’t know what ‘zero hunger’ is or how to measure it.

Michelle Jurkovich is an assistant professor of political science at the University of Massachusetts at Boston and a postdoctoral fellow at the Watson Institute for International & Public Affairs at Brown University.