Why We Should Let a Lottery Decide Our Government

Politicians are often older, wealthy, and white. Could randomly selecting average citizens lead to better representation?

Hands holding ballots
The Walrus

In ancient Athens, citizens—free adult men—would gather in the assembly, thousands at a time. At the peak of Athenian democracy, in the fourth and fifth century BCE, every citizen had the right to speak in this legislature on issues of the state, which ranged from finances to military matters and even the impeachment of generals. After these discussions, they could vote on final decisions, either with a show of hands or by secret ballot.

In this popular image of Athenian direct democracy, we often overlook another institution: rule by lot. Members of the council, responsible for the administration of Athens, were selected randomly from a group of willing and eligible candidates to serve for a year-long term. Members were tasked with setting the assembly’s agenda along with other administrative duties, such as managing the finances and foreign affairs of the city-state. Only motions that had been approved by the council could be discussed by the larger assembly. The juror-judges of political courts, too, were randomly selected; among other prerogatives, they could strike down laws that had been passed by the assembly.

The use of lot—also called sortition—embodied the democratic principle of rotation: Athenians regarded elected office as an oligarchic or aristocratic institution because not everybody had an equal chance of occupying it, while sortition allowed citizens to take turns ruling and being ruled. Through most of the fourth and fifth centuries BCE, the state even paid office-holders so that ordinary citizens, otherwise constrained by economic need, could freely participate in politics. By some calculations, between one-third and one-half of all citizens in fourth-century-BCE Athens could have been part of the council during their lives.

If the gathering of citizens in one large assembly seems unrealistic to us, then assigning political office by lottery may seem even more so. Yet political activists and scholars across Western democracies have proposed that we bring back this ancient institution. Instead of opposing elections for anti-democratic reasons, advocates of sortition are claiming that elections aren’t democratic enough.

To answer the question of whether an institution is “democratic” is to aim at a moving target: it depends what the meaning and stakes of democracy are. But some contemporary critics say that elections should be about more than handing power to whichever politicians the majority of people support at any particular moment. Democratic decision making should reflect the perspectives and concerns of ordinary people.

In electoral systems today, politicians tend to be overwhelmingly white, wealthy, and male compared with the increasingly diverse constituents they represent. In Canada, the forty-second Parliament was touted as the most diverse in the country’s history, but as of March 2019, only 27 percent of MPs were women. Immigrants and Indigenous peoples are similarly underrepresented relative to their presence across the country. Meanwhile, certain professions are overrepresented: in the 2015 federal election, most candidates had backgrounds in business, politics, or law, according to a CTV News analysis. Despite its increasing diversity, the House of Commons is still an unrepresentative sample of Canada’s citizens, and that has implications for policy making. Can representatives effectively “represent” the interests of their constituents without sharing their life experiences, concerns, and points of view? And, if not, could our democracy be strengthened by bypassing elections and directly involving citizens through a system of sortition?

Brett Hennig, who codirects the Sortition Foundation—an organization that advocates for sortition and provides random-selection services—presents an alternative to elections. Randomly selecting members of legislative assemblies would, he argues, enable “descriptive representation”—meaning representatives would mirror the preferences of their communities and their demographics across categories such as class, race, and gender. Pointing out that the US Congress has become a “club for millionaires,” Hennig asks, “What kind of legislation would be produced if the politicians did represent the US population at large and around 15 percent of them were receiving food stamps to feed themselves and their families?” In the Canadian context, the “affordability crisis” looms over millennials, who struggle to find affordable housing. According to a study by an organization that advocates for younger Canadians, millennials will take almost a decade longer to buy homes than the generation before them. However, they are represented by MPs who, with an average age of fifty-three in the forty-second Parliament, often do not face the same sources of anxiety.

It is no accident that politicians are likely to be older, wealthier, and better educated than the average citizen. Running for office requires time, money, and connections—resources that many people don’t have. Moreover, research suggests that women tend to wait to be asked to run for office rather than express political ambition first. Research also suggests that party nomination committees are less favourable to women candidates. Institutionalizing the ancient principle of rotation in offices, would have uniquely modern benefits: it would ensure that decisions reflect the concerns of the average voter, who does not have to court interest groups for campaign donations or the political support necessary for reelection, and it could ensure that women and minorities have seats at the table.

Limiting the impact of money and vested interests may also improve the quality of political discourse. Currently, large amounts of money are spent on broadcast and social media that spread sound bites of arguments and attacks. According to Facebook’s Ad Library, almost $10 million was spent on Canadian political ads on that platform between June and the federal election. Some interest groups spent hundreds of thousands of dollars in less than three months. Recent legislation—passed amid public concern about foreign interference and fake news in elections—forces digital platforms to disclose political advertisements and their funding sources. But, as anyone who reads the news might notice, there is a more general problem beyond the most egregious cases of public manipulation: today’s elections are plagued by low-quality political discourse. British sociologist Colin Crouch describes the process of campaigning as a “controlled spectacle, managed by rival teams of professionals expert in the techniques of persuasion, and considering a small range of issues selected by those teams.” With the prevailing power of business interests and the constant noise of social media and paid advertising, citizens’ participation in politics is often reduced to ticking off a box once every couple of years. But sortition promises another way.

In the fall of 2003, electoral reform was on the political agenda in British Columbia. In an innovative move, the provincial government decided to gather a group of citizens—a citizens’ assembly—to make recommendations on the future of electoral reform. Two hundred randomly selected people from each BC district’s electoral rolls received an invitation to take part. Chosen participants would gather over the course of the year to learn about electoral systems, consult with members of the public, then make a recommendation—whether to keep the existing first-past-the-post system and, if not, what system to replace it with. For each day of participation, assembly members would receive a $150 honorarium. From the pool of volunteers who responded to the invitation, 160 participants were chosen after adjusting for age and gender. This group was composed of one man and one woman from each of seventy-nine electoral districts, as well as two Indigenous people.

The citizens’ assembly was supported by a paid staff, which included administrators and academics. The academics gave lectures on BC politics and on legislative systems, which were followed by small-group discussions facilitated by graduate students. However, decisions were largely in the hands of the assembly members. They created their own set of norms and guidelines for discussion, clarified the values that mattered most to them in an electoral system, and debated the feasibility of each system in the context of British Columbia. After extensive debate, the assembly seemed to narrow down its options to two electoral systems: Single Transferable Vote (STV), a system in which voters can rank their preferences on a single ballot, and Mixed Member Proportional (MMP), a system in which each voter casts one vote for a single-seat constituency and another for a political party. Finally, on “decision weekend,” members held one last debate before taking a vote. An overwhelming majority chose to recommend STV to the citizens of British Columbia—a decision that was then put to a binding referendum.

The good news from the British Columbia citizens’ assembly is that ordinary people, even those without extensive political experience and background knowledge, are indeed capable of engaging in political debate and decision making. Of course, candidates who agreed to participate were more likely to be curious about politics and (at least to some extent) confident in their ability to engage with the material. The members of the assembly generally possessed little initial knowledge about electoral systems, but after the learning phase, they were able to thoughtfully debate the issue without accepting what experts said at face value.

The bad news is that the assembly’s recommendation did not gain wider traction. Voters were not sufficiently informed about the assembly’s proceedings, the evidence it examined, and its reasons for recommending STV. In the end, the proposal was twice defeated at the polls (once in 2005 and again in 2009). In 2006, the Ontario government initiated a similar citizens’ assembly on electoral reform. Again, the people of the province did not accept the Ontario assembly’s recommendation (for MMP, in this case).

Despite the mixed results of these early Canadian experiments, many countries, including the UK and the Netherlands, have followed suit in creating randomly selected assemblies. In 2017, for example, Ireland’s citizen’s assembly recommended repealing the country’s restrictive abortion laws and legalizing same-sex marriage, issues long considered divisive in Irish politics. This time, both policy changes received green lights in the subsequent nationwide referenda.

Unlike in ancient Athens, most citizens are not likely to ever be randomly selected for office, given the longer terms of office and larger population sizes we are working with today. So, even if there are potentially positive outcomes of sortition systems, there are also considerable challenges—citizens’ assemblies might make “better” decisions, but they can’t fully capture the egalitarian ethos embodied by “one person, one vote.”

Randomly selecting political representatives could give us the decisions that average citizens would make if they had the chance to learn and think deeply about an issue—but this is a crucial limitation of sortition: it relies on counterfactual judgment—would haves and what ifs—and risks making people who have a limited sense of ownership over those judgments feel more alienated. In the case of the BC referendum, the citizens’ assembly’s limited communication with the public contributed to the tension between “considered” and “raw” judgments. Citizens had to decide whether to go with their own relatively uninformed views, their raw judgment, or defer to others who had thought it through. In such cases, it’s hard to make a case for simply privileging the considered judgment of the representatives, even if they are a microcosm of the population—we can imagine the backlash that might result if a small group of randomly selected citizens, with limited accountability to the public, were to make binding decisions on high-stakes policy issues. This lack of accountability is, after all, a common criticism of the Senate.

Sortition might lead to better representation of a diverse public, but even when it comes to the interests of women and minorities, the system faces problems of accountability. Demographic traits such as gender and race can indeed serve as proxies for the wide range of perspectives in the population, but one member of a minority group can hardly be expected to represent or effectively advocate for the varied experiences of all members. Moreover, women and minorities might find it hard to advocate for their interests in a small, consensus-oriented group setting. In the BC citizens’ assembly, even though women made up half the members, calls for women’s representation did not gain as much traction as those for rural interests did. And the emphasis on a shared, common good dissuaded participants from advocating for the interests of particular groups. These assemblies are part of a wider conversation, and in this conversation, there should also be a place for social movements: the reasoned deliberation of disinterested citizens should not come at the cost of the voices of those who have stakes in the issues being debated. In theory, sortition doesn’t rule out the role of bottom-up movements. For example, the British environmental movement Extinction Rebellion lists a citizens’ assembly on climate and ecological justice as one of its demands.

Given concerns about accountability and practical viability, some advocates of sortition defend a “sortition chamber” that would coexist with the elected chamber in legislative systems. Arash Abizadeh, a political theorist at McGill University, suggests that we should replace another powerful but hardly accountable institution, the Senate, with randomly selected citizens. He argues that there is a place in our democracy for the free deliberation of representatives who do not have to worry about following party lines or getting reelected. The Senate represents this idea, but sortition could realize it more fully. Other advocates of sortition argue that the mandate of randomly selected officials should be narrower, restricted to particular issues such as electoral reform. Real-life examples of citizens’ assemblies in Canada, Ireland, and elsewhere have been shorter term and focused on specific issues, making participation more accessible.

Regardless of how it is implemented, randomly selecting citizens to take part in deliberation and decision making seems to be an attractive tool for revitalizing democracy. But effective communication between selected citizens and the wider public would be key. In contrast to the Canadian cases, the success of the Irish citizens’ assembly’s proposed reforms may be partly due to the effective publicization of the assembly’s proceedings: the proceedings were livestreamed and covered by national media outlets. That being said, the controversial nature of the issues being debated—abortion and same-sex marriage—likely also contributed to the heightened attention.

We need to help all citizens, whether selected for office or not, be better equipped to engage with politics. If the problem is that there is a division of labour between professional representatives and ordinary citizens, then making ordinary citizens into our representatives will help. But we should also question that division itself. If paying citizens to take time on weekends to participate in politics worked for the citizens’ assemblies, what other measures could we use to allow ordinary citizens to participate in politics? Instead of only relying on proxies and counterfactuals of what citizens would do if they had the time and energy, we can try, as much as possible, to create the conditions for that time and energy—for everyone.

Kate Yoon
Kate Yoon is a graduate student at the University of Oxford. She writes about the intersection of politics and philosophy.