Can an Algorithm Help Solve Political Paralysis?

10 min read


[responsivevoice_button rate=”1″ pitch=”1.2″ volume=”0.8″ voice=”US English Female” buttontext=”Story in Audio”]

Can an Algorithm Help Solve Political Paralysis?

Dave Johnson refuses to eat fish farmed near Minneapolis. The retired construction worker, who lives in a rural area 150 miles north of the Twin Cities, claims “tourists’ garbage, salt runoff and fertilizer” make the metropolitan area’s farmed fish inedible. But for Johnson, the “divide between city people and rural people” is about much more than fish: it’s about how individuals relate—or don’t relate—to the environment. “People in cities are busy living in the fast lane, while rural communities are really attuned to what’s going on in nature,” explains Johnson, who says he voted for Donald Trump in 2016 and links the country’s bitter urban-rural divide to bad policy making: “A lot of the politicians at the state capital pass laws because there’s peer pressure,” he says. “But they don’t really know what life is like up here—and they don’t care about us.”

Shona Snater cites a different kind of divide. Instead of geography, the 31-year-old soil-health organizer in southeast Minnesota says “corporate interests [are] dictating how politicians think and act,” especially when it comes to the environment. “There’s progress at the grassroots level,” she adds. “But when farmers are using half of the fertilizer they used to use, that’s terrifying for agribusinesses.” A registered Democrat, Snater says she believes both of the dominant U.S. political parties can be “bought out.” She blames corporations for supporting “terrible policies” that favor “economies of scale” and leave many people behind. “Small and midsize farmers say, ‘These politicians aren’t representing my best interests,’” Snater says, noting that Minnesota has been losing dairy farms at a rate of nearly one a day.

Such frustrations around the usual mechanics of policy making are rife throughout much of the U.S. In a 2018 Pew Research Center survey of 10,000-plus Americans, 75 percent said trust in the federal government has been shrinking. And when the survey asked who Americans trusted to “act in the best interests of the public,” the majority of respondents revealed they had the most confidence in scientists and the military—and the least in elected officials. Brett Hennig, an advocate for alternative ways of conducting democracy, says he can understand why.

“If you believe an ideal democracy involves informed deliberation among a representative group of people, the current electoral system fails on both counts,” says Hennig, who has a Ph.D. in astrophysics. He believes something called “citizens’ assemblies” offer a better way to elicit policies in line with people’s real interests—with a little help from an algorithm.

Hennig explains citizens’ assemblies using simple logic: society is made up of people who are young and old, rich and poor, and mostly in between, so decisions governing it should more directly involve a group proportionally representing these kinds of characteristics. But because many ordinary citizens may lack technical knowledge of the issues at hand, citizens’ assemblies invite these individuals to make decisions in a “deliberative environment”—in which they can consult experts to “reduce the effect of biases, misleading information and ignorance” when learning about a problem and assessing possible solutions, Hennig says. From there, these citizens collaboratively craft recommendations for policy makers to consider.

Philipp Verpoort, a Ph.D. candidate in physics at the University of Cambridge in England, is another scientist who advocates for citizens’ assemblies. “Everyone’s talking about the three P’s: pessimism, populism and polarization. And we’re at this point where people split up into groups, don’t trust their politicians, and nothing gets done,” he says. “But when people see a decision being made by people like them, they trust it.”

To turn their theories to practice, Hennig, Verpoort and their colleagues co-direct the Sortition Foundation—a nonprofit organization that offers “selection and stratification services” for citizens’ assemblies and similar deliberative bodies. The foundation (which is financed largely by payments received for its services) has supported about 20 such projects among the roughly 200 that various organizing bodies have hosted around the world, according to Verpoort and Hennig. One of the most famous efforts, which the Sortition Foundation was not involved in, was held in Ireland, where a 99-person assembly prompted the nation in 2018 to repeal a law that had effectively banned abortion. But much of the buzz today surrounds climate change and the U.K.: after a series of smaller city-based gatherings in 2019, 110 citizens were selected earlier this year for the country’s Climate Assembly UK project. Through this deliberation, participants were invited to recommend ways the U.K. government could meet its legally binding target of net-zero greenhouse gas emissions by 2050.

After meeting in person and virtually over six weekends between January and May, the participants recently published their final report of policy recommendations, which range from levying a tax on frequent air travelers to investing in low-carbon public transportation.

Using an Algorithm to Build a Scale Model of a Society

With a mathematician’s precision, Hennig explains how an algorithm he created generated the 110-person “mini public” from the U.K.’s population of 67 million. The process began by sending invitations to 30,000 households from the nation’s postal database. Hennig says completely random selection would have skewed the responses toward people with higher incomes (who are more likely to have the time and money to participate). So 20 percent of the sampled individuals were randomly invited from the “most deprived areas,” and 80 percent were chosen at random from every region. To further reduce the effects of income-related selection bias, participants were promised a small stipend and travel reimbursements.

Out of the 30,000 people invited, nearly 2,000 accepted and completed an online survey indicating seven characteristics: their gender identity, age, ethnicity, educational attainment, location, description of their residence as urban or rural, and level of concern about climate change. Hennig applied his algorithm to those 2,000 respondents to select 110 participants who would proportionally represent the U.K. with respect to those seven categories.

To begin the process, the algorithm went through the survey answers and randomly and repeatedly selected from the “hardest category to fill,” Hennig says. He adds that it was initially (and unsurprisingly) hardest to find respondents among those who reported they were “not very concerned” or “not at all concerned” about climate change. The algorithm continued by selecting for the “most in need” demographic—“determined by the maximum of the ratio between number of people still required for fill a category quota and the number of people available for selection in that category,” Verpoort explains—until it drew 110 names. In the end, though the Climate Assembly UK Web site cautions that the algorithm “slightly over sampled” certain demographics, the selected group’s distribution of the seven categories very closely matched that of the U.K.

But Jonathan Mattingly, a Duke University professor of mathematics who is also using algorithms in an attempt to help fix democracy, takes issue with that framework. “Who’s determining which categories inform decision-making? Maybe it’s your family income before you turned 16, or whether your parents were immigrants,” Mattingly says, referring to the seven characteristics used by Hennig and Verpoort. “How do we know which factors account for clarity, empathy and decisiveness?”

Geoff Bedford, a participant in a “citizens’ jury” (a smaller kind of citizens’ assembly) on climate, in the city of Leeds, England, also expresses concerns about how such assemblies are run. He points specifically to the expertise process. “Most [participants] seemed to rely solely on the opinion of so-called climate scientists,” says Bedford, who doubts that humans cause climate change. The participants, he adds, “don’t have time to read the conclusions of [the experts’] reports and, understandably, accept sound bites from the mainstream media.”

Verpoort and Hennig counter that, for practicality’s sake, most assembly organizers prioritize experts whose research is peer-reviewed—and the vast majority of such research confirms anthropogenic global warming. But to make sure people like Bedford were still a part of the conversation, another algorithm was used during Climate Assembly UK. “Once you’ve got a diverse group of people representing society together in a room, you don’t want all of the climate skeptics to end up on the same table and talk to nobody else,” Verpoort says. To avoid this clustering, he created an algorithm to divide the 110-person sample into tables of eight individuals, among whom the seven categories—age, location, and so on—were proportionally represented.

Can Citizens’ Assemblies Work in the U.S.?

Even if citizens’ assemblies prove effective elsewhere, some doubt they would work as well in the U.S.—and that the country’s federal and local governments would take the resulting recommendations very seriously. But Crystal Chissell, vice President of operations and engagement at the nonprofit environmental organization Project Drawdown, says she is “absolutely” convinced that ordinary Americans can help craft actionable policies. She points to Cincinnati, where hundreds of residents participated in developing the Green Cincinnati Plan of 80 evidence-based strategies to reduce the city’s carbon emissions by 80 percent by 2050. “‘Ordinary people’ are very savvy when it comes to seeing the co-benefits of climate solutions,” Chissell says, explaining that many simply want policies benefiting their health, well-being and finances while also addressing climate change.

One of the U.S.’s prominent champions of people-powered policy making is James Fishkin, a professor of communication and political science at Stanford University. He is known for developing Deliberative Polling, a process that—like citizens’ assemblies—uses stratified random sampling to gather a representative group of citizens, facilitate informed deliberation among them, and highlight “actionable priorities” for governments.

In the past, Fishkin’s polls have been used to advance policies such as sustainable energy choices in Texas, Nebraska and Vermont. And he says that America in One Room, a recent national gathering of 523 citizens (a stratified, random sample that was representative of the American electorate in terms of attitudes and demographics) showed they can also build understanding across political ideologies. “When a candidate wants to win an election, they deploy one-sided arguments just to generate outrage,” Fishkin says. “But when you empower people to go beyond tribalism and consider an issue under the best possible conditions—in an evidence-based discussion with multiple perspectives—there’s more policy consensus and less gravitation toward extremes.”

Johnson, the retired construction worker, may offer some proof of that theory. He recently attended the Rural Climate Dialogues, a series of Minnesota-based “citizens’ juries” that were hosted by the nonprofit Jefferson Center, which was one of the earliest adopters of such juries in the U.S. After taking part, he says he now firmly disagrees with Trump’s position on climate change and better understands its effects. “To hear how climate change is behind polar bears losing their homes, our changing forests and the heavy rains that keep my construction buddies from working—I just learned so much,” Johnson adds.

Snater, the soil-health organizer, also says the Rural Climate Dialogues opened her mind to new perspectives. Although she says she has spent much of her life researching climate change, Snater believes the jury helped her better understand how the phenomenon relates to the on the ground experiences of her neighbors. “This woman [from the dialogues] told a story about running out of fuel and living in a cold house for a couple of weeks with her kids,” she says. “It was a real moment where I realized, ‘Not everyone is in the same position I am.’”

You May Also Like