A recent guest essay in the New York Times, “The Worst People Run for Office. It’s Time for a Better Way” made a seemingly radical suggestion—improve our democracy by doing away with elections and choosing our leaders randomly.
Could this actually work?
Two kinds of evidence speak to that question: experience and experiments.
Although not many examples of randomized leadership exist, it has been tried and found effective. Many leaders in ancient Athens chosen by lot were deemed capable. Juries offer another example, one that provides both historical and contemporary context. While the final selection of jurors ensures that they have characteristics that make them suitable to the task at hand, lawyers build juries from randomly created pools of jurors. Jurors exercise a form of democratic leadership and generally do so very effectively.
Leadership has been explored in experiments that use public goods games in which participants have the option of contributing to a public vs. a private account. Individual players keep the returns that are private, while the group shares the public returns. The “common good” in this set-up is unambiguous—it is simply the outcome that maximizes total group earnings, which occurs when all players contribute the maximum to the public account. Yet it doesn’t always emerge because it is in the interest of individual participants to “free ride” off of others and stick their money in their private account, while it is in the interest of the group as a whole to have all participants contribute to the public account.
What might leadership look like in this situation?
An effective leader is someone who increases contributions to the public good relative to the contributions to the public good received in a leaderless world.
Suppose the experiment is set up to offer a chosen leader the ability to communicate with others, to encourage them to contribute to the public account. Does that encouragement work, and does it work if the leader is chosen randomly? It turns out that randomly chosen leaders are effective at garnering support for the public good. In short, they are good at leading. Randomized leadership generates an increase in contributions relative to the leaderless experiment.
What have we learned?
First, effective leaders are persuasive, and their ability to communicate with the group matters. They successfully make the case that group members will be better off if they follow the leader’s advice.
Second, for randomized leadership to work, it is important for group members to know that the system for selecting leaders is fair—not rigged—ensuring that any capable group member has an equal shot at being the leader.
Third, when the election is based on merit, elected leaders are more successful at obtaining increased contributions to the public account than randomly chosen leaders. The case for merit is extremely important. If the group members believe the election is based on merit, they are more likely to follow the leader’s advice. Partway through the experiment described above, each person specified a “platform” regarding contributions. After reading all platforms, group members voted for a leader. In this way, they established the merit of their chosen leader.
When people believe they may be chosen in the future, and they have a fair chance at being chosen in the present, they are willing to listen to and follow reasonable advice. Under these conditions, 18th-century Scottish economist and philosopher Adam Smith said leaders are motivated to act in a “praiseworthy” manner and people look up to such leaders “with a certain degree of esteem and respect.” But importantly, Smith worried a great deal about what happens when the random selection of leaders breaks down, when leaders are beholden not to everyone, but rather to supporters in a small group or “faction.” As he famously put it, these leaders lead “all for ourselves and nothing for other people.”
When a leader is chosen in a system that requires enormous sums of money to succeed, as happens in the U.S. electoral system, it is no longer true that anyone might lead. And the electorate no longer believes the election is based on merit. When the sums of money are huge, the electorate realizes that those involved are beholden to donors and special interest groups. If people conclude that the election is rigged in favor of the extremely rich, they may well be more apt to follow a randomly chosen leader than a rigged election leader. Perhaps it is time to introduce randomization into our leadership institutions, to establish more praiseworthy leadership.
Dr. Sandra Peart is Dean of the University of Richmond Jepson School of Leadership Studies and E. Claiborne Robins Distinguished Professor in Leadership Studies. Her areas of expertise include leadership ethics and ethics and economics.