Here’s a moral dilemma that seems tragically timely, given the chaos surrounding attempts to deliver aid to Burma’s cyclone victims. There are 60 orphans at the Canaan Children’s Home in Buziika, Uganda, and their meal allotment has to be cut. What do you want to do: take six meals away from each of two kids, or 10 meals away from one? You have eight seconds to decide.

In this and similar moral dilemmas, efficiency (the total number of meals lost) is pitted again against equity (how evenly the burden of lost meals is shared among the children). You have to take away a total of 12 meals if two children share the loss, but only 10 (which would seem better) if a single orphan bears the entire burden. You have to decide whether to sacrifice efficiency (losing fewer meals) to equity (spreading the loss over more children).

Here’s another way to think about it. You are driving a truck to the Burmese cyclone victims. It holds 1,000 pounds of rice. The time it will take to deliver the rice to everyone in the Irrawaddy Delta village you are headed for means that 200 pounds will spoil. If you deliver the rice to people you meet en route, you will be distributing it to only half the population of the village, but only 50 pounds will spoil. Do you deliver the rice to only half the number of victims, maximizing the total amount of food provided (efficiency), or do you sacrifice 150 pounds to distribute it to more people (equity), giving rice to more people but also causing more rice to go to waste?

In a study reported online today in the journal Science, researchers posed the orphan dilemma to people while scanning their brains with functional magnetic resonance imaging (fMRI). Unlike most studies of the brain basis of ethical decision making ("neuroethics"), this one was grounded in reality: the volunteers’ choices would determine how many meals the research team actually donated to the Ugandan orphans. The volunteers knew this, which made the dilemma painful in the extreme. “Quite a few came out saying: ‘This is the worst experiment I’ve ever been in. I never want to do anything like this again!’,” said study co-author Ming Hsu of the University of Illinois’s Beckman Institute for Advanced Science and Technology.

So, which is more critical to our sense of justice, equity or efficiency? And how does the brain decide?

In the experiment, the volunteers (26 men and women, ages 28 to 55) first read short bios of the orphans. Then they watched a video on a computer screen, showing a ball rolling toward a lever. By moving the lever, they could steer the ball toward either of two depictions of the moral choices: photographs of the actual orphans who would be affected by that choice, with numbers for the number of meals that would be lost to those children if that option were chosen.

By an overwhelming margin, people chose to preserve equity at the expense of efficiency—lose a few more meals, but spread the burden among as many children as possible, rather than making one hungry child—whose imploring little face stared back at them from the screen--shoulder the entire loss.

According to the fMRI, different brain regions became active at different points in the decision-making. The insula, which is involved in processing emotions and the awareness of bodily states as well as (in some studies) evaluating fairness, was active when the volunteers wrestled with questions of equity. The putamen, which is activated during learning that brings rewards, lit up when people thought about efficiency.

Since equity won, it suggests that decisions about fairness are rooted in emotion more than in cold-eyed cost-benefit analysis. “That the brain has such a robust response to unfairness suggests that sensing unfairness is a basic evolved capacity,” Steven Quartz of Caltech and co-author of the study said in a statement. “The emotional response to unfairness pushes people from extreme inequity and drives them to be fair,” suggesting that “our basic impulse to be fair isn’t a complicated thing that we learn,” but an instinctive one.

And whoever said scientists have no heart? After the experiment, and based on the volunteers’ decisions, the team donated $2,279 to the orphanage.