Why a Democracy Needs Uninformed People

Uninformed people catch a lot of flak in society, whether they’re sitting on decision-making committees, choosing a new PTA president, or voting in a national election. Political commentators often daydream of a fully engaged and 100 percent knowledgeable electorate. New research, though, suggests these know-nothings may be more vital to democracy than anyone has given them credit for.

Researchers at Princeton have discovered that the least informed among us may have a crucial role in tempering the most opinionated minorities. Surprisingly, they started off by studying fish.

Iain Couzin, the lead author on the paper published this week in the journal Science, is actually a professor of ecology and evolutionary biology, not a political or social scientist. His hypothesis about uninformed individuals grew out of a model of animal behavior that tries to capture how, for instance, schools of fish or flocking birds make group decisions.

“One of the key pieces that we’ve been able to elucidate with those types of models is that individuals both influence and are influenced by those with whom they interact,” he said. “Even though they have relatively simple, relatively local interactions, this can give rise to the beautiful coordinated patterns of activities that we see in nature.”

Couzin became interested in how animals potentially manipulate each other within these groups. Animals in a given herd have different preferences — the thirsty ones want to head for a watering hole, for example, while the hungry ones hope to set out for a feeding ground. How does the group decide? Do the hungriest individuals sway everyone else? In nature, there’s a high cost to not reaching consensus. If the group doesn’t make a single decision, it loses many of the advantages of traveling en masse.

To study how this decision-making process works, Couzin and his colleagues applied findings from a series of mathematical models in designing experiments with golden shiner fish. These fish are naturally drawn to the color yellow (Couzin has no idea why this is, although similar phenomena are common in nature). The researchers used this color bias as a stand-in for a “strong opinion” among the fish. In experiments where a minority of fish was trained to swim toward a yellow target, and a majority toward a blue target, the minority swayed the whole group more than 80 percent of the time.

Then the researchers added “uninformed” fish to the mix, and a curious thing happened.

“Adding those individuals dramatically changes the outcome of group decision-making,” Couzin said. “They inhibit the minority and support the majority view, and this allows the majority to be heard and that view to dominate.”

Individuals with no particular preference — they could go to the blue target or the yellow one, the watering hole or the feeding ground — essentially dilute the influence of the vocal minority.

“We thought, ‘Wow, that’s kind of interesting,’” Couzin said, “because you don’t normally think that adding uninformed individuals to decision-making processes would have that sort of democratizing effect.”

The finding would be even more interesting if it applied as a general principle to human behavior. Humans are much more complex than animals. But we share the same feature Couzin and his colleagues were studying in fish: the ability to influence and be influenced by each other. We also have the capacity for strong opinion.

The researchers took this idea and applied what they’d learned from the golden shiners to a series of simulated models of human behavior derived from the social sciences. Sure enough, the concept seemed to hold.

“As we add in uninformed individuals, not much changes,” Couzin said. “Then we suddenly get to this threshold where suddenly just adding a few more uninformed individuals completely changes the group dynamic, and instead of the minority winning, the majority does.”

The researchers will need to do a next generation of controlled experiments with actual people (students, that is) to further develop the finding. But if the concept holds, it has the potential to knock down a bit of conventional wisdom about how people make group decisions — that is, that uninformed people are easily swayed by the loudest voice in the room, enabling extreme minority views to spread.

There is such a thing, however, as too many uninformed individuals. If you have 20 of them in a room and only one or two people with an actual opinion, the process breaks down. Information, Couzin says, gets drowned out in the noise. He adds this caution: remember that strongly opinionated people in the minority will have good ideas, too. So while we’re celebrating the potential power of the uninformed in bringing us all closer to democracy, it’s worth remembering that democratic consensus may not necessarily settle on the best idea.