Fire Congress, Hire a Jury

Excerpts from the article:

In 2018, Ireland held a referendum on whether to repeal the Eighth Amendment, a constitutional near-total ban on abortion that had stood for thirty-five years. The country voted yes, by a two-to-one margin. What broke the deadlock was not a charismatic politician, not a viral campaign, not a court ruling. It was 99 randomly selected citizens.

The Irish Citizens’ Assembly, formed in 2016, pulled those 99 people from the general population the way you’d pull a jury: by lottery. They weren’t activists or lawyers or pundits. They were teachers, plumbers, retirees, students. Over five weekend sessions, they heard from constitutional scholars, medical experts, advocacy groups on both sides, and women who’d had abortions. Then they voted, 79 to 12, that the amendment should not be retained in full.

The politicians had been too terrified to touch the issue for decades. The randomly selected citizens did the hard deliberative work in five sessions over several months.

If you asked most people what defines a democracy, they’d say elections. Voting. Choosing your representatives. This conflation is so widespread that it’s practically a reflex: democracy is elections.

The Athenians who invented democracy would find this answer bizarre.

In Athens, from roughly the fifth century BCE onward, most public offices were filled by sortition, which is just the fancy word for random selection. The Boule (a council of 500 that set the Assembly’s agenda) was chosen by lottery. Most magistrates were chosen by lottery. The massive citizen juries that decided court cases were chosen by lottery using purpose-built randomization machines called kleroteria. Elections existed, but they were reserved for a narrow category of roles that required specific expertise, like military generals. For everything else, the Athenians believed that random selection was more democratic than voting. (Yes, Athenian “citizens” excluded women, slaves, and foreign residents. The definition of who counts was terrible. The mechanism of selection is what matters.)

The reasoning is simple. Elections, by their nature, select for a certain kind of person: someone wealthy enough to campaign, well-connected enough to build coalitions, charismatic enough to win media attention, and ambitious enough to want power in the first place. That’s a highly filtered subset of the population, an aristocracy in everything but name.

In recent years, more than half the members of Congress have been millionaires. The median net worth of a U.S. senator is around $4.4 million, according to recent disclosure analyses. Meanwhile, the median American household sits at around $192,000. Congress looks like America the way a country club looks like a surrounding neighborhood.

This isn’t an accident. Running for Congress costs money. A competitive House race now routinely exceeds $2 million in spending; a Senate race can top $50 million. To raise that kind of money, you either need to be wealthy already or be very, very good at asking wealthy people for favors. The result is a candidate pipeline that systematically filters out anyone who isn’t affluent, well-connected, or willing to spend years building a donor network.

And once you’re in, you stay. Congressional incumbents who seek reelection often win in the mid-to-high 90s, cycle after cycle. That’s a tenured position with name recognition and a rolodex (i.e. not a competitive job market).

Sortition eliminates this entirely. If legislators are chosen by lottery from the adult population, with stratified sampling to match demographics like age, gender, geography, and income, the resulting body looks like the country. Like its neighborhoods, its workplaces, its kitchen tables. Not like its country clubs.

Okay, here is a number that should appall you: in 2013, a leaked DCCC orientation schedule for new House members recommended four hours a day of fundraising “call time.” It also set aside an hour for “Strategic Outreach” (i.e. breakfasts, meet-and-greets, press) on top of just two hours for committee work and floor votes. Because federal law prohibits fundraising calls from government offices, members walk to party-affiliated call centers near the Capitol to do this. There are buildings in Washington, D.C., whose primary purpose is to house rows of cubicles where sitting members of Congress dial for dollars.

Alright, let's assume you knew nothing about earthly politics and then imagine two machines.

Machine A works like this: you have a pool of 330 million people. You run a contest where the winners are determined by who can raise the most money, who is most telegenic, who has the best connections, and who most wants the most power. The winners are locked into a building where they spend half their time calling the richest people in the country and asking for money, then vote on policies they barely have time to read, under instructions from party leadership that threatens to cut off their funding if they dissent. They stay in this building for decades, because almost every cycle, the contest just confirms whoever won last time. The machine produces a body that is disproportionately white, male, wealthy, fucking old, and drawn from law and business. Approval rating: 15%.

Machine B works like this: you take a stratified random sample of 300 people that statistically mirrors the population. You give them a very nice year’s salary, cancel their other obligations, provide them with balanced briefings from competing experts, and have trained facilitators run structured deliberations. They serve for three years, then go home. They owe nothing to donors, face no reelection pressure, and have no party leadership telling them how to vote. They focus exclusively on one policy area, so they can learn it deeply.

Which machine would you expect to produce legislation that reflects the interests and judgment of the people being governed?

I think the answer, for most people, is extremely obvious. Machine B sounds better. The reason we don’t use Machine B is inertia, not logic.

The claim that “random people can’t govern” is a prejudice dressed up as common sense. Every time anyone has tested it, the randomly selected citizens have performed at least as well as their elected counterparts, and usually better. They listen more. They deliberate more carefully. They’re less captured. And they look like the people they’re supposed to represent.

Maybe the problem with Congress isn’t that we chose the wrong people. Maybe the problem is that we chose them at all.