Global crop yields are not growing fast enough to meet the demands of a population that will reach 9 billion by 2050. This is the finding of a recent study published by Deepak Ray at the University of Minnesota’s Institute on the Environment (where I am a graduate research assistant), which is in response to previous studies estimating that crop production will need to double by 2050 in order to meet growing food demands. But if global population (currently at 7 billion) is expected to increase by 2 or 3 billion by 2050 — a 30 to 40 percent increase — why would crop yields need to double? Where is the extra demand coming from?

The main culprit is increasing meat consumption. As people move out of poverty and get wealthier, they consume more meat and dairy.

China, for example, has seen a rapid shift to the middle class in the past 20 years. In 1989 China produced about the same amount of meat as the United States, but now China’s meat production is almost twice the production in the U.S. Because meat-heavy diets require substantially more crops to produce than plant-based diets (it takes about 20–30 calories of feed to produce just one calorie of edible beef, for example, and 6–9 calories of feed to produce one calorie of chicken), growing global wealth has surpassed population growth as the bigger reason for increased crop demand. As higher proportions of crops go to animals first, people are only indirectly and inefficiently fed. On average, of every 100 calories we feed to animals we only get about 12 back in the form of meat and dairy. How many people could we feed with those calories if we eliminated that loss?

My colleagues and I set out to answer this question and found some surprising results, published recently in the journal Environmental Research Letters. Globally, 36 percent of all calories produced on croplands never become food for human consumption, but are instead used for animal feed. Differences by country are stark: In India less than 10 percent of crop calories are fed to animals; in China, it’s one-third; and in the U.S., 67 percent. (A bit more of the calories grown on global croplands are directed to biofuels, meaning that more than 40 percent of all the calories we grow never become food.)

If we used U.S. croplands to grow only crops for direct human consumption, we could feed more than 1.5 billion people. That means the U.S. alone could feed a staggering 1 billion more people on the calories that do not end up in the food system.

With all of the land, resources and investments that go into the U.S. agricultural system, my colleagues and I wondered: How many people could we feed with the calories we produce in the U.S. (if all were used for food), and how many people do we actually end up feeding with these investments?

Factoring in the 67 percent we use as animal feed, the conversion rate to meat and dairy, and the crops used for corn ethanol, the U.S. currently feeds only 524 million people (assuming a 2,700 calorie-per-day diet). Great Britain, Italy, Colombia, Ghana, India and Pakistan are among the 69 countries that feed more people per hectare of cropland. While these countries may not have higher yields than the U.S., they direct more of what they grow to people and are therefore able to feed more people per land area than the U.S.

If we used U.S. croplands to grow only crops for direct human consumption, we could feed more than 1.5 billion people. That means the U.S. alone could feed a staggering 1 billion more people on the calories that do not end up in the food system. When we look at a global scale, if we redirected all of the calories being used for animal feed and biofuels to direct human consumption, we could increase calorie availability by 70 percent, enough calories to feed an additional 4 billion people. But this would require us to drastically reduce meat and dairy products to those from only grass-fed or wild-caught animals. Also, the production of biofuels from human-edible crops would need to completely halt.

Those are pretty lofty and unrealistic goals. So we also investigated how less drastic changes could increase food availability. Because different kinds of livestock use feed at different levels of efficiency, switching from grain-fed beef to chicken and pork could allow us to produce more meat with the same feed crops. We found that if globally we chose to eat chicken or pork instead of grain-fed beef, there would be enough additional meat calories to feed 357 million more people. Or, if we directed all feed calories to the production of milk, eggs and cheese instead of meat, there would be enough additional calories to feed more than 800 million more people. This makes it clear that even small changes in diet can increase calorie availability.

Moving away from meat, or from beef to other products, would have ancillary benefits as well. Ruminants such as cattle and sheep have digestive systems that produce methane gas — a serious greenhouse gas with 25 times the warming potential of carbon dioxide over a 100-year period. If we ate chicken or pork instead of beef, we could reduce diet-related greenhouse gas emissions by more than 40 percent.

If the entire global population shifted to a meat-heavy diet tomorrow, we would need twice the amount of cropland currently under cultivation. Most cropland expansion in the 1980s and 90s came at the expense of diverse tropical rainforests in the form of deforestation. Given that global population is increasing and diets are changing, the number of people fed per cropland hectare must increase in order to meet the challenges of food security and prevent further deforestation. Luckily, as this research shows, even small changes in diets — ones that don’t even require us to completely give up things we love, like cheese — can increase the number of people we can feed on existing croplands.