Could Ditching Elections Save Democracy?

Lottocracy: Democracy Without Elections
Alexander Guerrero
Oxford University Press, $45

When democracy seems everywhere in crisis, it may sound paradoxical, to say the least, that the solution to our troubles is to scrap elections altogether. But that is precisely what political philosopher Alexander Guerrero proposes in his bold and illuminating book, Lottocracy: Democracy Without Elections. We should select political officials not by voting, he contends, but by lottery from among the entire adult citizenry.

As radical as it sounds, the idea, indeed the reality, of “sortition”—using random selection to select political officials—is nothing new. Nor is it the prerogative of any particular political persuasion. The Athenians used such a system more than two thousand years ago. The Trinidadian Marxist C. L. R. James celebrated this system when he argued, echoing Lenin, that “every cook can govern.” The idea has seen something of a popular revival in recent years thanks to the writing and advocacy of people like political theorist Hélène Landemore and Belgian historian David Van Reybrouck. And it has been put into practice in a variety of deliberative and citizens’ assemblies, including in Europe and the United States. What sets Guerrero’s analysis apart is that he has thought through how such a system might work in modern societies in exhaustive detail. The result is a landmark argument that must be reckoned with.

As radical as it sounds, the idea, indeed the reality, of “sortition”—using random selection to select political officials—is nothing new. Should it replace elections?

Guerrero spends much of the book putting flesh on the bones of the abstract idea of lottocracy, presenting a picture sufficiently well specified for meaningful comparison with real-world electoral democracy. In the rest of the book, he makes the case for the relative superiority of lottocracy and offers ideas about how we might get there from here. The book’s central claim is not that lottocracy is perfect but that, for all its flaws, it is still preferable to other political systems.

Of course, there are many ways to compare political systems. One might ask how well they comport with political equality: the ideal that everyone should have, at some level, the same say over policy. Or one might ask how well they offer opportunities for participation: the ideal that everyone be able to contribute to making policy. Guerrero contends that lottocracy does as well if not better than other systems on these criteria. But his primary interest is different: how well a political system solves problems, whether it delivers the objectively correct policy (which he thinks exists). While the capacity of a political system to solve problems—to, among other things, make people’s lives better—may not be a condition of a system’s counting as democratic, Guerrero is certainly right that it is something that we should want.

The argument is superbly detailed, even relentlessly thorough. Guerrero offers a response to just about every objection a reader might think of. But ultimately, the case is not convincing.

---

At the federal level, Guerrero’s imagined lottocracy looks roughly like this. Each year it would randomly select 3,000 new legislators and then assign them to one of twenty 450-member, single-issue, lottery-selected legislatures (call these SILLs), each charged with a different policy area, themselves more or less fixed in advance: a Water Quality and Water Access SILL, say. Legislators would serve three-year terms, with 150 veteran members cycling off each year and 150 new members joining.

During an agenda-setting phase, specific policy proposals would be advanced by “advocacy groups and citizens’ groups” and then selected by the 300 SILL members in their second or third year for closer consideration and discussion. During a subsequent learning phase, SILL members would hear from a slate of experts, advocates, stakeholders, and community members. The experts might be selected at random from a database, whereas advocates and stakeholders might be selected on the basis of “institutional and public support in the form of signatures and official group endorsements.” In any event, selection would be overseen by a “Learning Phase Structural Assembly” made up of veteran SILL members. After the learning phase, SILL members would deliberate in a more or less structured way and finally vote on legislation.

Would this system be an improvement over an electoral system? The answer depends on what type of electoral system we compare it to. Guerrero notes several pathologies of electoral systems like that of the United States, which uses single-member districts with plurality voting. A more attractive electoral alternative is a party-list system, in which voters select from a slate of party members, with the proportion of votes for a party determining how many from its slate are elected. In this kind of system, there would be less of a focus on the personal traits of individual candidates—and more of a focus on policy—than in the current U.S. system. There would be more than two major parties, and therefore less of the polarization that Guerrero sees as an affliction of our system. And legislatures would be more “descriptively representative,” meaning that they would more closely resemble the makeup of society at large.

Despite these advantages over the current U.S. system, Guerrero thinks that even a party-list system would do worse than his imagined lottocracy on problem-solving. The source of the problem, he argues, is simply that voters are ignorant. First, the problems that our society needs to solve are enormously complex, requiring expertise to understand that most voters simply don’t have the time to acquire. I suspect that most readers of this magazine, for example, would not be able, without the help of a specialist in international economics, to understand, let alone explain, what effect tariffs might have on the dollar’s status as a reserve currency and then to evaluate whether this was a good or bad thing. Thus, even if representatives had an incentive to follow the preferences of the public, doing so would typically not lead to good solutions. Second, voters often can’t tell whether their representatives are solving the problems, so there will be no electoral incentive for representatives to solve them (and, one might add, no incentive for voters to elect representatives with any relevant policy expertise themselves). Given such free rein, representatives will be “captured” by the well-monied and powerful, in which case they again will not solve the problems, at least when the solutions run counter to what the powerful want.

One way capture works is through promises of sinecures: rewarding elected representatives with high-paying or powerful positions after they leave office, for example. Of course, randomly selected legislators could be captured in this way too, so Guerrero suggests a number of measures that a lottocratic system should take to prevent it. But it’s not clear why these measures couldn’t also reduce this kind of capture in electoral systems. Guerrero’s case for the superiority of lottocracy thus appears to hinge on the ability of random selection to eliminate the kind of capture that is intrinsic to elections themselves.

This sort of capture operates in many ways. One is the control of the media, which influences what information voters hear, how they understand policy issues, how they frame the choices they face in an election, and how they rank which issues are most important to them. Suppose we grant, for the sake of argument, that voters do vote for whatever party the powerful tell them to, thanks to this and other mechanisms of capture. Why think that members of a lottocracy won’t be equally captured in their views? Guerrero’s hope is that the learning phase—when SILL members hear from experts, advocates, stakeholders, and community members—will cure their ignorance and deliver the system from capture.

The central problem with elections, Guerrero contends, is simply that voters are ignorant, leading to perverse incentives and weak problem solving.

The prospect is doubtful. For one thing, the experts themselves might be captured—a worry that Guerrero considers but does not fully assuage. For another, lottocratic governance would itself be subject to capture during the agenda-setting phase. Who is going to have the time, money, and organization—let alone sheer awareness that there is a Water Quality and Water Access SILL—to put together a proposal for the lottocratic legislature to consider? Special interests, most likely. An industry threatened with stricter regulations might use its resources to flood the zone with proposals that distract from stricter regulations, for example.

Things get worse. Guerrero imagines that each SILL would be guided in its deliberation by a poll of those few citizens who somehow are able to take a week off of work and other responsibilities to pay attention to the five day-long discussions of the final five proposals. Again, if the powerful can, in effect, buy off the general public to support a particular electoral party, then why can’t the powerful mobilize a (again, presumably quite small) group to pay attention to the review of proposals for the Water Access and Water Quality SILL and support what they favor? No one but the powerful, one worries, would be minding the store.

Now suppose that you are a politically engaged person, who cares not only about water access and quality, but also about all the other nineteen topics of public concern, who is suddenly faced with the transition to such an agenda-setting process. You don’t have the freedom to take a week off to hear the final five proposals for the Water Access and Water Quality SILL, and it would be impossible for you to hear from more than one SILL, let alone all twenty. All the same, you want to influence the process toward what you take to be better outcomes, or at least to keep the powerful from influencing the process away from them. Presumably, you would want to fund an organization that shares your general priorities, which could formulate agenda proposals and resource at least some of its members to participate in, or at least give feedback about, the agenda-setting process, not just in one, but in all twenty of the SILLs. What we are describing, the sort of organization that you would be desperate for, is more or less something that Guerrero claims would be far less important without elections, perhaps to the point of eventually withering away: namely, a political party.

So much for avoiding capture. When it comes to problem-solving, another important question is whether randomly selected representatives will be up to the job, or rather less up to the job than the sort of people who tend to become elected representatives, who are, among other things, disproportionately lawyers. I accept Guerrero’s four basic egalitarian assumptions about people’s abilities. First, anyone can, given the requisite time and training, acquire expertise about how policies will affect people’s lives, given an accurate understanding of what people’s lives are like. Second, anyone can, given sufficient time and training, acquire skill in understanding what other people’s lives are like, even if one’s own life is quite different, on the basis of hearing their testimony about what their lives are like. Third, anyone can, given sufficient time and training, acquire skill in judging how these effects on people’s lives will affect their interests. Finally, anyone can learn to give due weight to the interests of everyone affected in judging which policies should be adopted.

Still, there is a question of how much time and training is required. Consider the fact that 54 percent of Americans read at a sixth-grade level. In two years, they might be brought to an eighth-grade level. But that hardly seems sufficient for the sort of expertise that they would need to do what Guerrero asks of them, which is not only to understand in depth, but also to independently evaluate, the testimony of experts about a massively complex topic. And then, given the three-year term limits, the process would start all over again, losing whatever institutional memory and expertise had been developed. There is a simple solution, of course: just allow for much longer terms. But this raises other problems—more on that below.

As for the learning phase, there is a puzzle, by Guerrero’s own lights, as to why SILLs should feature the input of stakeholders, advocates, and community members at all. By hypothesis, they too are ignorant and captured. If they say that a certain policy would set back their interests, why should better-informed SILL members pay them any mind? Why not listen only to policy experts? Presumably, the answer is that these non-experts can provide relevant testimony about what their lives are like. If SILLs were perfectly descriptively representative, it’s hard to see why such testimony would be necessary, since SILL members themselves could offer the relevant perspectives without needing to bring in outside voices. But setting that issue aside, the only way for this picture of the relevance of non-expert testimony to work is for Guerrero to assume that people are not ignorant about what their lives are like and that nothing money can buy will lead them to misrepresent what they know about it.

One problem with these assumptions is their implausibility: if media indirectly influences citizens’ views in myriad ways, surely one of those ways is how citizens narrate and represent their lives to themselves. We are not Cartesian islands; our self-knowledge is far more open to social influence than Guerrero appears to imagine. Worse still, Guerrero reduces the role of advocates, stakeholders, and community members in a lottocratic society to communicating  what their lives are like—rather than, say, communicating their judgments about what policy should be and why. More on this later.

Guerrero sees descriptive representation as a significant advantage of lottocratic bodies over elected bodies, which tend to overrepresent dominant groups and social elites. In the United States, for example, white Americans make up 74 percent of Congress but only 58 percent of the population, while the median member of Congress has a net worth more than twelve times that of the median U.S. household. But it is unclear how far any practical system of random selection would go in generating descriptive representation—a possibility that the above line of reasoning, about the relevance of non-expert testimony, suggests Guerrero is aware of. He rejects a system of compelled service, persuasively arguing that “a person can be compelled to participate, but it is much harder to compel them to participate in good faith and to the best of their abilities.” His system relies on carrots rather than sticks, incentivizing service with “perhaps a floor of $100,000 per year, with anyone whose regular income was above the floor being paid 1.1 times their normal yearly income” along with efforts “made to accommodate family and work schedules (including providing relocation expenses and legal protections so that individuals or their families are not penalized professionally for serving).”

But even with all that, there are surely many people who would prefer not to disrupt their lives with three years of lottocratic service. It isn’t uncommon for people to take a pay cut to do work that they enjoy more, and I daresay that many people would enjoy their work more than three years of hydrology lectures on the Water Quality and Water Access SILL. Some people might not want to be in the political spotlight or would look anxiously on being expected, among other things, to explain their decisions to 449 other legislators, as well as the public at large. Some people might simply not want the responsibility and strain of making momentous decisions for other people. In an effort to assuage these concerns, Guerrero at one point observes:

What they were worried about before coming into power—raising a child as a single parent while working double-shifts to make ends meet . . . helping their brother re-enter society after spending time in prison, keeping their small business open . . . revitalizing their town in response to factory closings, caring for their elderly parents—will be largely the same as the concerns they will have on the other side.

But what if they raise a child with the help of a grandparent who cannot move? How they will be able to help their brother re-enter society or care for their parents or revitalize their town if they have moved to the other side of the country? How will their small business survive if it has to be shuttered for three years? Guerrero expresses the basic problem when criticizing Landemore’s view: “self-selection in politics results in dramatic skew along every demographic dimension we are likely to care about.”

---

For all these reasons, Guerrero’s main argument—that lottocracy beats electoral democracy in terms of problem-solving ability—is not convincing. But he also has notably little to say about whether lottocracy beats epistocracy, or rule by experts, despite the fact that many of the considerations he offers in favor of lottocracy would seem to point in that direction. After all, if we can identify the true experts to give input to lottocratic legislatures, why not cut out the middleman—the rest of us—and let them decide how to govern? When it comes to building a bridge, it would be odd to rely on what a randomly selected group of people made of engineers’ presentations rather than to leave it up to the engineers themselves. Why try to create experts out of ordinary people rather than rely on existing experts?

No doubt, the idea offends our most basic democratic impulses. Guerrero suggests it is immediately ruled out by a commitment to political equality. And if we take that to mean that everyone should have an equal chance of being a decider, it might well seem that epistocracy must fare worse than lottocracy. In an epistocracy, after all, only those people with the capacity to become experts—the right mix of innate talents and life background—would have chances of being a decider.

Lottocracy would give everyone an equal shot at being selected, but some people would pay to serve more than others—scrambling equal likelihood of serving in government.

But are the cases really so different? Guerrero envisions a lottocratic system with three-year term limits, but he offers no good reason for this constraint. Why not employ random selection at age eighteen for a life term, which would give each person an equal chance of being a decider? With more time to serve, those selected for service would have more time to develop their expertise. Guerrero might object that life appointments would lead to legislatures that aren’t descriptively representative in terms of members’ life experiences—someone who has served in government from age eighteen has a very different life experience than someone who has labored on a farm. But Guerrero himself assumes that testimony from stakeholders, advocates, and community members can convey at least as much, if not more, relevant information about life experience to lottocratic legislators. Alternatively, it might be thought that term limits are needed to guard against post-selection capture. But again, Guerrero himself assumes that we can prevent such capture by imposing certain safeguards.

The point is that his arguments are just as consistent with this kind of lottocratic system as with the system that he presents, and this system doesn’t seem recognizably more democratic than epistocracy. It might fare better than epistocracy on some interpretation of political equality—the interpretation that requires everyone have an equal chance to become a decider—but by guaranteeing that the vast majority of citizens would never be able to exercise control over government, it too seems to offend our democratic impulses.

Returning to the comparison between lottocracy and electoral representative democracy, Guerrero argues that “there are no plausible considerations of political morality that electoral representative democracy respects that lottocratic systems do not respect.” In particular, he contends, lottocracy does just as well, if not better, in terms of three measures of political equality, as well as opportunities for participation.

Regarding political equality, he claims, first, that lottocracy gives people an equal chance of serving as representatives, whereas electoral democracy does not. Your chances of being elected are much higher if you possess qualities that attract votes: charisma, extroversion, megalomania, or whatever we might think those qualities are. That much seems uncontroversial. But does lottocracy without compulsory service really give people equal chances to serve? Again, some people will find the costs of serving a three-year term too high and not worth bearing, despite all of Guerrero’s carrots. A person with inherited family wealth, who lives off a trust fund, is more likely to agree to serve than the caretaker of an ill grandparent or the sole proprietor of a small business who cannot afford to leave the workforce for several years. Lottocracy would give everyone an equal shot at being selected (assuming a well-designed system for bona fide random selection), but some people would pay to serve more than others, scrambling equal likelihood of serving in government.

A second way to measure political equality is to consider influence over policy. Neither lottocracy nor electoral representative democracy gives equal influence in this sense, Guerrero argues, because in both systems representatives have much more power to determine policy than non-representatives. Indeed, in the sense of direct influence, only representatives have such power, Guerrero claims. So it’s a tie. Still another measure of political equality is whether all citizens enjoy equal influence over who the representatives are. This is true under lottocracy, of course, because random selection ensures that no one has any influence over who the representatives are. And while elections that accord with the principle of one-person, one-vote might seem to give equal influence over who the representatives are, Guerrero notes that, in practice, vast inequalities in indirect influence—including through control of the media—shape how people cast their votes. So on this score, he concludes, lottocracy comes out ahead.

But these second and third comparisons are not as convincing as they may seem. Yes, the wealthy have vastly more indirect influence than the poor over who gets elected. As we saw earlier, however, there are channels for indirect influence in lottocracy too. It seems highly unlikely that the learning phase could be insulated from such indirect influence. And then there are the opportunities for capture at the prior agenda-setting phase.

Moreover, non-representatives do have a kind of direct influence on policy under electoral systems: namely, by having a direct say over who the representatives are. The electoral power to choose representatives is power to determine policy. This point is stressed in a splendid and convincing recent counterpoint to arguments for lottocracy, Cristina Lafont and Nadia Urbinati’s The Lottocratic Mentality: Defending Democracy Against Lottocracy. As Lafont and Urbinati succinctly put it, “By choosing some political programs and parties over others, [voters] shape the political/ideological space within which the elected representatives must operate until the next election.” Which party comes into power has significantly different, and largely predictable, consequences for policy. As Guerrero himself observes:

For a moment, the generalist legislature is for the war and a bigger military, against abortion, against more public school funding, in favor of subsidies for oil drilling, for more police on the streets, for state support of religion, against environmental regulation, against immigration, for significant corporate tax breaks, against a minimum wage. In the next moment, nearly everything is reversed.

This is exactly what is meant, Lafont and Urbinati remind us, by the saying that “elections have consequences.”

In electoral representative democracy, then, non-representatives do influence policy, precisely through their votes. Granted, they do not have the same influence over policy as representatives, and most will not have the kinds of outsize influence that the wealthy and powerful can exert in more indirect and less predictable ways (which we should, of course, work to reduce). But in a lottocracy, citizens who are not representatives—which is to say, the vast majority of citizens, roughly 99.997 percent at any given time—have no direct influence over policy at all. It is not clear how depriving almost everyone of direct influence over policy, while vesting it in the hands of a fraction of a percent of the population, helps the cause of equality. This is to say nothing of the idea that because voters influence which platforms will be implemented through their votes, representatives, in implementing those platforms, can be seen as doing their will, in which case what we have described as the greater influence of representatives over policy might be seen, in a way, as the influence of the voters themselves.

Against the suggestion that lottocracy deprives the vast majority not selected, or otherwise unable, to serve on a SILL of any opportunity for influence over policy or meaningful participation, Guerrero claims that “there would be many other avenues for political participation in a lottocratic system” beyond “serving as a representative on a SILL or other lottocratic body . . . just as serving as an elected representative is only one way in which ordinary citizens might participate under electoral systems.” In particular, he imagines “there will be many of the exact same avenues of influence” as in electoral systems, such as “protesting, petitioning, speaking, writing, assembling, organizing, creating and working within political parties and organizations, donating money and time to political issues that one cares about,” with the goal of influencing SILL members or speakers or affecting public opinion at large.

We might count ourselves lucky if even the highly flawed system that Guerrero decries can survive in the near term.

The question, though, is why SILLs should be receptive to such attempts at influence, by Guerrero’s own assumptions. Insofar as SILL members are doing their problem-solving job in the way that Guerrero imagines, they should treat the input of advocates, stakeholders, and community members, and other non-experts merely as sources of information about what their lives are like. Any other input they might give would, by hypothesis, simply reflect ignorance or capture. Thus, their role as providers of such information might be, in principle, entirely supplanted by empathetic social scientific studies conducted by third parties. This picture might reasonably be said to give non-representatives some degree of influence, but I don’t think that this is what most people have in mind by political influence. As Guerrero concedes, it is “important” that “democratic participation isn’t just about providing epistemically valuable input.” What people have in mind by political influence is something like their judgments about policy having an influence on policy. And it is unclear that there is a route for this in a lottocracy functioning in the way that Guerrero believes it must to effectively solve problems.

The upshot is that Guerrero’s argument for lottocracy’s improved problem-solving capacity is at odds with his claim that it costs us nothing by way of political equality and participation. In fact, lottocracy promises improved problem-solving only to the extent that it vests all power in a minuscule fraction of the people and insulates them from the corrupting influence of the will—as opposed to the autobiographical reports—of the rest.

---

There is something poignant, in the present moment, about Guerrero’s criticism of electoral democracy. We might count ourselves lucky if even the highly flawed system that Guerrero decries can survive in the near term, without collapsing into something that electoral democrats and lottocrats, for all their differences, agree does not count as a democracy—or even a constitutional order governed by the rule of law. If we get through the current crisis with a window for institutional change, it seems to me that our efforts should be focused on authoritarianism-proofing whatever structures are still recognizably democratic.

Guerrero does consider how we might get from our electoral-representative “here” to his lottocratic “there.” But the immediate problem is more how we get from “here” today, perhaps at this point the “here” of yesterday, to “here” tomorrow—how we see to it that what we have achieved, in Lincoln’s words, “shall not perish from the earth.” My money is not on pressing for the sort of radical transformation that Guerrero imagines, in part because the effort is almost certainly destined to fail in the face of the interests arrayed against it; in part because even if a lottocracy were somehow to be established, we have little basis to know whether it could perpetuate itself, cultivating its own allegiance; and in part because the survival of institutions depends largely on established understandings—a sense that certain things simply are not to be done.

Dispensing with elections is one of those red lines it seems we cannot afford to consider crossing. The risk is that nothing but brute power and opportunism would take their place.

Independent and nonprofit, Boston Review relies on reader funding. To support work like this, please donate here.

The post Could Ditching Elections Save Democracy? appeared first on Boston Review.