Should we be Lottocrats?

Introduction

In light of representative democracy’s failures, from an inability to effectively address pressing problems to the yawning economic inequality and deep polarization that it sustains, the need for reform is obvious and urgent. In the democratic theory literature, there is growing interest in the potential of lottery-based political institutions.¹ In this essay, we develop concerns about the critique of representative democracy that has given rise to interest in lottery-based alternatives, and assess the potential of these alternatives to solve the problems that ail representative democracies.

Our point of entry is Alexander Guerrero’s Lottocracy, which is among the most interesting, challenging, and provocative arguments for lottery-based political institutions.² More than any other work in the now voluminous literature, it offers detailed institutional descriptions of lottocratic institutions and engages with their real and presumptive critics. The position that Guerrero defends in the book is also, in an important sense, focal. Unlike most defenses of lottery-based political institutions, Lottocracy proposes to do away with elections altogether—what is on offer here is an alternative regime type, not a proposal for incremental reform or even a hybrid institutional regime. These qualities make Lottocracy an important touchpoint for thinking about the potential of lottery-based institutions.

Guerrero’s broad argument proceeds in two main steps. First, it contends that elections are a primary cause of modern democracy’s shortcomings. Second, it makes the case that a system of governance by randomly selected citizens, serving in issue-specific legislative assemblies, can be expected to make policy that would be preferable to what we can expect from representative democracies. This is both because selected citizens would not face the ostensibly perverse incentives created by elections and because those selected would form a descriptively representative body, which would bring a much wider variety of perspectives to bear than do elected chambers. Meanwhile, using single-issue chambers—devoted to the important topics facing a political community, from agriculture, crime, and energy, to housing, immigration, taxation, and so forth—would allow ordinary citizens, via expert testimony and three-year terms, to gain the knowledge needed to make competent decisions.

There is a lot to admire in this book, even for those who might be hesitant to embrace its ultimate conclusions. To pick just a few of the many important and novel argument threads that appear in it:

• There is a sharp and persuasive analysis of the challenges of relying on the media to provide citizens with the information that accountability-based mechanisms of governance require. The book articulates a key dilemma that democratic theory has not systematically grappled with: a privately funded media is likely to be overly focused on entertainment (at the cost of providing citizens with needed information), while a publicly funded media is unlikely to be sufficiently critical of government.³

• There is a trenchant critique of “hybrid solutions,” such as bicameral legislatures, with one chamber populated via popular elections, and the other via random selection.⁴ While recognizing their appeal, Guerrero shows how it can come to pass that “rather than ‘a best of both worlds’ situation, this kind of hybrid arrangement would likely result in a ‘worst of both worlds’ situation.”⁵ Among other difficulties he notes, forcing legislation to be approved by bodies populated in very different ways heightens the risk of a paralyzing status quo bias.⁶

• There is a nuanced and detailed response to Cristina Lafont’s criticism that, unlike direct and electoral democracy, lottocracy requires citizens’ “blind deference” to policy-makers, which is incompatible with self-government. The core of the response sets out important distinctions between kinds of deference: unqualified and irrevocable; instrumental vs. respect-based; to institutions vs. to specific persons; forced vs. by choice, and provides a superb closely argued conceptual analysis, ultimately rejecting the criticism. While Guerrero’s response is surely not the last word in this debate, it articulates a very thoughtful perspective that is novel to the literature.⁷

• There is a thoughtful and sensitive discussion of the ethical issues that arise when one presses for radical political reform. The book draws attention to the uncertainty that is inherent in such changes. This uncertainty is aggravated by the risk that a movement pressing for such changes will tend to vilify opponents in a way that will make the movement unreceptive to new information. In light of such concerns, Lottocracy endorses an ethic for revolutionaries that prescribes, among other things, a conscientious openness to dissenters and unwelcome information. It is to the book’s great credit that it takes such concerns seriously.

One of Guerrero’s explicit goals, and perhaps the book’s greatest contribution, is to spur a broader conversation about political reform in modern democracies. In that spirit, we want to press, first, a challenge to its criticism of representative democracy, and then, second, to its argument for lottocracy.

II. The Critique of Representative Democracy

The book draws a withering portrait of representative democracy. Citizens are ignorant and unable to hold representatives to account, representatives have an irredeemable short-term bias, and citizens are baited into a contest of vicious partisanship that is encouraged and heightened by an irresponsible press. Residents of the United States will, no doubt, find an awful lot that is familiar in this portrayal. We certainly do.

Why are democracies performing so poorly? Lottocracy’s account is subtle, and we cannot summarize every aspect of it here. The crux of the argument, however, proceeds as follows:⁸

1. Modern democracies are large.

2. Partly as a result, the problems that they must contend with are complex.

3. Given that citizens are extremely unlikely to have a decisive effect on policy choices, they have little reason to invest in the knowledge that would be needed to address those problems effectively.

4. Indeed, citizens typically know too little to hold representatives accountable, which gives representatives the freedom to use their power for rent-seeking.

5. Meanwhile, even if citizens did manage to get representatives to be responsive to their policy preferences, such responsiveness would reflect citizens’ lack of information, which would itself be problematic.

6. Therefore, modern representative democracies cannot deliver a form of responsiveness that meaningfully answers to the interests of citizens.

The book, thus, identifies the fundamental inability of elections to provide meaningful forms of accountability as the central problem facing modern representative democracies. This problem is exacerbated by three important and connected difficulties. First, electoral systems stoke vicious partisanship, pitting citizens against one another.⁹ Second, electoral systems incentivize representatives to focus on short-term performance, often at the expense of addressing long-term problems.¹⁰ Third, electoral systems tend to select demographically unrepresentative officials, which leads them to consider an artificially limited set of issues and viewpoints.^11,12

Since the problems that undermine electoral accountability on this account are not tied to particular features of a given electoral system (such as the way that it funds campaigns), but instead relate to the potential of electoral systems in a modern context, they imply that one should not expect to make significant headway via piecemeal reforms to the electoral system (e.g., campaign finance reform, lobbying reform, etc.).¹³ This is why the book calls, instead, for a regime type—lottocracy—that foregoes the use of elections altogether.

III. An Empirical Puzzle: Evidence of Accountibility’s Importance

The critical argument taps into important, widely shared intuitions, but it also creates a puzzle that goes to the heart of the argument. The puzzle concerns the relationship between the critical argument and empirical evidence about the performance of electoral democracies.

A large body of literature shows that extending voting rights to a group (even a minority) leads to public resources being dedicated to their interests. Consider a few influential examples:

• A study of Brazil shows that lowering the burden that illiterate citizens face in casting ballots led to a substantial increase in spending on redistributive programs, including on healthcare for the poor, and, consequently, to a significant reduction in the incidence of low-weight births.¹⁴

• A study on the 1965 Voting Rights Act in the United States provides evidence that removing literacy tests led to substantial increases in government spending in areas with larger black populations. This increased spending is then linked to increased educational enrollment for black teenagers, as well as improvement in the quality of their educational experience.¹⁵ Another study shows that the Act also led to a substantial (11%) reduction in child mortality for black infants in counties that were “covered jurisdictions” under the law, compared to infants in counties that were not.¹⁶

• A study of the United States provides evidence that women’s suffrage led to significant, and virtually immediate, increases in public health spending and, as a result, substantial reductions in child mortality.¹⁷ There is also evidence that this extension of voting rights led to significant improvements in the education of children, particularly for those from more disadvantaged backgrounds.¹⁸

• A study on immigration in Belgium and Switzerland provides evidence that the extension of voting rights to immigrants led to substantially higher redistributive spending in areas with larger immigrant populations.¹⁹

Even in quite imperfect representative democracies, with citizens who are far from fully informed, we have, then, striking evidence that the extension of effective voting rights to a group of citizens immediately and dramatically increases the public resources devoted to their problems.

Here, then, is the puzzle for the book’s critical argument: Evidence about the significance of extending the franchise for policy outcomes suggests that citizens’ views importantly affect representatives’ behavior. This, though, is difficult to understand if “the central mechanism of electoral accountability simply doesn’t make sense given modern political conditions of size, complexity, and durable voter ignorance.”^20,21

Now, it is important to note that Lottocracy allows that “modern democratic governments do many things well” and says that “it would be a serious mistake to think that electoral representative democracy is a disaster.”²² But this acknowledgement of the achievements of representative democracy belies the rhetorical frame of the book’s broader critical account, which is in terms of a catastrophic failure:

like other forms of technology, political systems have failure conditions: conditions under which they can no longer function to accomplish their central purposes. Think of a car submerged in water, or a heart without blood. I argue that commonly present conditions of the modern world constitute failure conditions for electoral representative government.²³

Cars submerged in water and hearts without blood do not do many things well. Consistent with the diagnosis of catastrophic failure, the implication that the book urges is that we have little to lose by trying something radically different, parallel to a patient undergoing heart failure who has the option of a risky transplant. Yet, if, as the micro-level evidence suggests, electoral accountability generates many important benefits even under quite non-ideal conditions, then we have more to lose in trying an untested alternative than we would if the simile of cars submerged in water and hearts without blood were apt.²⁴

The key issue, to be sure, is not a rhetorical mismatch but a substantive one. It is unclear how one could reconcile the idea that modern democratic governments manage to do many things well with the claim that voter ignorance destroys the prospects for meaningful accountability. Evidence of effective accountability directly undermines the book’s key critical argument, suggesting that—somehow—that argument has gone awry. Without some rival account (which Lottocracy does not offer), and in the face of the kind of micro-level evidence noted above, it is difficult to avoid concluding that democratic governments do many things well precisely because meaningful (if also highly imperfect) accountability is not only possible, but actually operative.

IV. Heterogeneity in the Performance of Representative Democracies

We now turn to heterogeneity in the performance of existing representative democracies, which—as we will see—also suggests reasons for skepticism about the book’s central critical argument. There is enormous heterogeneity across all of the areas where the book anticipates failures of institutional performance. Consider, as one example, vicious partisanship, which the book argues is “endogenous to elections.”²⁵ The empirical literature shows, though, that Americans did not have negative feelings about those in the other party as recently as the early 1990s and that the kind of vicious partisanship, or affective polarization, that the book’s criticism hinges upon only emerged in the United States beginning around 2008.²⁶ This makes it difficult to see vicious partisanship as an inescapable feature of even two-party systems, much less of representative democracy more broadly.

Beyond the United States, the evidence that vicious partisanship is not an inescapable feature of modern representative democracies is even stronger. An influential recent paper shows that the United States has had the largest growth of any democracy on this measure over the last several decades.²⁷ Meanwhile, several democracies (including Germany, Japan, and Norway) show levels of affective polarization roughly similar to the United States prior to the growth that began in the 1990s. It is hard to square this empirical variation with the book’s conclusion that while some reform of the electoral system might make a difference on the margins, “the ingroup/outgroup dynamics will emerge in pretty much the same way, given the other background conditions, as long as there are elections.”²⁸ Similar points could be made with respect to the book’s other worries about representative democracy, for democracies display extremely heterogeneous performance across practically all areas relevant to institutional performance, including the level of capture or corruption, the levels of economic inequality that they support, the level of responsiveness to citizens, and so forth.

Importantly, many of the reasons for variation in the performance of representative democracy appear to be institutional. While each of these areas touch on large literatures that we cannot review here, and much remains unknown,²⁹ there is evidence about the relationship between electoral systems and various important outcomes, including polarization,³⁰ economic inequality,³¹ corruption,³² and policymaking responsiveness.³³ Furthermore, institutional innovations developed in the literature could likely further enhance the performance of representative democracies—for example, random constituencies could mitigate the tendency of representative democracies towards affective polarization³⁴ and calibration of district magnitudes under proportional representation could improve accountability.³⁵

But if vicious partisanship (for example) can be mitigated with proper institutional design, then the argument for lottocracy that proceeds by comparing its ideally designed form to forms of representative democracy that feature significant levels of affective polarization seems off the mark. The argument does not establish the superiority of lottocracy, for it fails to show that lottocracy has an advantage over well-designed forms of representative democracy. Moreover, if variation in performance among electoral democracies is in part due to underlying differences in institutional design, this speaks against the view that more modest reforms are unlikely to significantly alter such performance. More broadly, because the critical argument generates expectations that are inconsistent with the empirical record—for example, that all representative democracies should be mired in affective polarization—evidence of significant variation in actual performance, like evidence of electoral accountability, raises concerns about the argument’s plausibility.

Perhaps, though, the appearance of effective accountability in some representative democracies and the heterogeneity in their performance primarily reflect differences in favorable background conditions and should not give us solace about representative democracy’s potential when those conditions are not in place. Lottocracy states explicitly that its argument is contextual, such that the appropriate form of government depends on the relevant background conditions. But, even setting aside the evidence that some of the variation in empirical performance of representative democracies is institutional, that assertion as we will see below, only adds to the puzzle. The conditions that the book identifies as “failure conditions” for representative democracy and so as important for the case for lottocracy, are (1) the size of modern political communities, (2) the complexity and technical nature of the problems that they face, and (3) the extent of divisions between citizens (racial, ethnic, economic, etc.).³⁶ The thought is that while representative democracy may be able to function effectively in certain environments, many modern representatives face these “failure conditions,” which preclude their ability to operate effectively. Can reference to these conditions mute the concern that reforming representative democracies may have more potential than the book allows?

Take size first, since it is the dimension that the book theorizes most explicitly. The key idea underlying the book’s critical argument is that, given the size of the political community, citizens have insufficient incentives to acquire the knowledge that they would need to vote in an informed manner. This, though, would be true even in the Scandinavian democracies: in Finland (population, 5.5 million), Sweden (population, 10 million), and Norway (population, 5.5 million). The size argument is certainly consistent with the possibility of effective collective choice based on elected representation when it comes to internal governance in small groups like neighborhood associations or even parliamentary chambers (e.g., in electing chamber leaders who have agenda control powers). Members of such bodies would likely recognize that their votes could reasonably end up being decisive and find it worth their time to invest in detailed knowledge about their representative’s behavior (though, perhaps, direct-democratic practices may be even more suitable for such bodies than internal elected representation). But when it comes to modern states, it would seem that they are all sufficiently large that the book’s critical argument should apply. If size is a failure condition, we should expect all modern democracies to display similarly abysmal performance.

The complexity of problems that are faced by modern electoral democracies may, and likely does, exacerbate the already weak incentives of citizens to become informed. Does this further precondition of the argument imply a narrower scope for the argument—one that would allow us to say that it applies to, for instance, the United States, but not Finland? It is hard to see why, for all modern political communities face complex problems, ones whose solutions cannot be easily intuited by ordinary citizens paying little attention to global or economic affairs. Again, when it comes to modern states, it would seem that they all face sufficiently complex problems that the critical argument should be expected to apply.

This brings us to the third condition—divisions between citizens. While all modern political communities wrestle with divisions between different groups of citizens (racial, religious, ethnic, economic, etc.), the book focuses most explicitly on economic inequality. One natural way to interpret its arguments—given the pervasive incidence of such inequalities—is as having a universal or near-universal scope of application. When the book discusses inequality as a “failure condition,” it laments the high levels of inequality in every region in the world,³⁷ suggesting that the argument may, in fact, have a global remit. Indeed, it is hard to read the book’s criticism of electoral representation as anything other than a general critique of modern electoral democracy, given that it begins by saying that a “main” aim is to “present a case against what is the heart of almost every modern political system: the use of elections” and goes on to say that, under modern conditions, elections are “no longer able to support and enable a thriving, healthy polity.”³⁸

Still, given the considerations we detail above, one may think that the best version of the argument would only claim applicability when it comes to democracies that face particularly high levels of economic inequality. Sustaining that as the implied scope limitation, though, is difficult. One reason is that the book’s critical argument does not explain the level of inequality that will be sufficiently problematic to require lottocratic political institutions. In several places, the book works with the United States as an example, but neither offers nor invokes a sustained analysis to show why the argument would apply in the U.S. but not in other major democracies—after all, all modern democracies face significant economic inequality. Is there really a qualitative difference between the U.S. and Germany or the United Kingdom on this score?

Perhaps the primary difficulty with going this route, though, is that it effectively presumes that economic inequality is simply an exogenous given—that is, not itself a reflection of institutional choice. Alas, there is considerable evidence of historical variation in levels of inequality, and there are reasons to believe that part of this variation is driven by institutional and policy decisions. For instance, in the early 1980s, at the onset of the Reagan revolution, the Gini coefficient in the U.S. was approximately the same as that of Switzerland, and the difference in the Gini coefficients between the United States and Canada was one quarter of what it is today. Or, to take a different case, the U.S. had far more inequality than the United Kingdom in the late 1960s, just about the same in 2000, and again far more in recent years (see the graph here). Clearly, neither absolute nor relative levels of inequality are fixed.³⁹

That differences in economic inequality reflect underlying institutional differences is a longstanding theme of scholarship in comparative politics.⁴⁰ Indeed, Alfred Stepan and Juan Linz, reviewing influential literature on this topic, write that “a key determinant of inequality is the institutional structure of a society” and pick out the U.S.’s unusual constitutional structure, with its many veto points, as a key reason for its higher levels of inequality over time.⁴¹ Further work in political science offers evidence that larger growth in inequality in the U.S. in recent years can be partly attributed to the strong status quo bias of its institutions.⁴²

The more that there is an institutional explanation for levels of inequality, the less it makes sense to say that some countries are inside the scope of the argument and some not due to their prevailing levels of inequality. This does not mean, of course, that the United States is just a small institutional tweak away from being Norway. But it does seem to speak against treating levels of inequality and social division as fixed for the purposes of debates about institutional reform.⁴³ (It bears noting that treating the levels of inequality in different societies as exogenously given and then choosing political institutions to match those circumstances would seem to require embracing geographical or cultural determinism, neither of which is particularly palatable.)

Pausing to take stock before turning to our critical discussion of lottocracy, we would like to draw out three key implications from the discussion in Sections III-IV. The first is that the critical argument seems to paint with too broad of a brush. Whether or not the central critical claims that Lottocracy makes about representative democracy accurately reflect the state of affairs in certain existing democracies, interpreting them as identifying limitations of representative democracy that are unavoidable in modern political communities appears to be in tension with important micro-level evidence about electoral accountability, as well as the heterogeneity of performance observed in existing representative democracies. The problems that the book identifies with representative democracy as a regime type may, thus, be more plausibly thought of as the problems of its poorly performing exemplars.

The second implication is that arguments in democratic theory need a more sophisticated model of how representative democracies operate, for such systems appear, in at least some cases, to be able to generate high-quality outcomes, even when voters are poorly informed about questions of public policy.⁴⁴ Yet, many of the literature’s criticisms of representative democracy, including those mounted by Lottocracy, imply that this is impossible.⁴⁵ A more sophisticated model would allow us, among other things, to say something more nuanced about the conditions under which we might expect poor performance from representative democracies, which should, in turn, illuminate discussions about reform.

This brings us to the third implication, which is, broadly, that if the critical argument is overdrawn, then the case for pursuing institutional reforms within the broad framework of representative democracy is stronger than the book allows. It is difficult to see how to avoid this inference. One could try to block it by arguing that representative democracies will only perform well under certain conditions—conditions that are simply not present in the countries that the book means to target. Yet, as we have seen, at least with respect to the conditions that Lottocracy highlights, it is not clear that one can identify conditions that, at once, are exogenously fixed and can successfully differentiate, say, the U.S. and the U.K. from Norway and Finland.

If that is right, then the only way to block the inference about the relative strength of the case for pursuing institutional reforms within the broad framework of representative democracy would seem to be to insist that even the better performing democracies are failing in a way that requires them to abandon elections and institutionalize lottocracy. While that strikes us as a hard bullet to bite, whether it is right or not ultimately depends on a comparison of the likely performance of well-designed versions of representative democracy and lottocracy. The relevant comparison is not ultimately how we should expect a well-designed lottocracy to perform relative to one of the high-performing existing democracies, but—rather—how a well-designed lottocracy would compare to a well-designed representative democracy, which may include substantial institutional improvements even relative to today’s best performing democracies.⁴⁶

Indeed, once we agree that reform is necessary, we need to know which of the mutually exclusive ideals (i.e., representative democracy, lottocracy, or some alternative regime type) should guide our reform efforts. It remains possible, for all that we have said, that a well-designed lottocracy would perform better than a reformed representative democracy. An inquiry into whether it would, or not, is necessarily speculative, but such a comparison is important, and it requires a detailed picture of what may be expected from lottocracy. This is one reason why Lottocracy is a valuable book—it goes a long way toward clarifying that picture. We turn to that picture, and our concerns about it, in the following section.

V. Lottocracy

One way to respond to these concerns about the critique of representative democracy is to allow that well-designed representative democracies could avoid many of the liabilities described in the first part of the book, but to argue that lottocracy could perform better still. Evaluating a hypothetical alternative is unavoidably rife with uncertainty and judgment. But even as nobody can opine with certainty on what such an experiment would bring, particularly when it comes to specific equilibrium patterns of behavior, it is important to identify and debate possible reasons for caution. Indeed, this is the spirit in which the book puts forward the proposal for lottocracy, inviting the reader to “think about whether it is a good idea, or the beginning of a good idea, or whether we might shape it into a good idea.”⁴⁷

Taking up that invitation, we will articulate some reasons for worrying that, given the absence of electoral accountability, such a system would be particularly prone to corruption.⁴⁸ Guerrero certainly recognizes that this is a concern that one might have about the proposal and devotes considerable attention to addressing it.⁴⁹ His hope is that corruption can be reasonably controlled through a combination of mechanisms, including initial random selection, high salaries for those selected, legal punishment of those engaged in corruption, and regular rotation through positions of power. How susceptible, then, would the resulting system be to corruption?

There will still be actors with enormous private stake in the shape of public policy, including large corporations whose profits depend on the shape of public policy. Meanwhile, there will be lawmakers with the discretion to shape such policy. We worry that the absence of electoral accountability significantly increases the likelihood that officials would find ways to monetize their power and special interests would find ways to enable them to do so.

Here we describe one such possibility. Imagine that we are on the transportation legislature and so are charged with reforming and improving our system of transportation. The ideal system of transportation would be much more accessible than our system. We’d have clean and efficient trains and light rail everywhere. All roads would be well-lit, would have effective guard rails, and would be lined by pedestrian-friendly sidewalks, with lit crosswalks, and so forth. If we sit down and deliberate just about transportation, we can surely imagine fantastic solutions to the crumbling mess that is the status quo. The same is true in other areas: healthcare, space exploration, defense, and so forth.

The problem, of course, is that the preferred solutions will typically be expensive. Given budget constraints, we will face cross-issue trade-offs. Furthermore, even setting aside budget constraints, many of the proposals will be fundamentally incompatible with one another. No doubt the environmental committee won’t be thrilled about the proposals that come from the space exploration legislature and vice versa. Lottocracy explores the possibility that conflicts posed by budget constraints be handled by a budgeting body (whose members may be randomly selected), which would make “specific allocation decisions based on these [single-issue] proposals.”⁵⁰ Given that it will be impossible to fully fund the proposals endorsed by single-issue legislatures, some such body would be necessary, and its members would have substantial discretion to shape proposals as participants prefer.

One worrisome implication of this relates to competence. The book’s defense of lottocracy on epistemic grounds is crucially tied to the expectation that important decisions are going to be made by citizens who would accumulate a certain amount of knowledge in a given area by being exposed to expert testimony and serving fairly long terms. However, if much of the consequential work will need to happen in a single, general body—which will inevitably need to dice up proposals and fund small bits of one and another—these decisions will themselves require expertise that far outruns the posited modest expectations. We may be stuck with impressionistic and under-informed citizens making crucial decisions, after all.

Our bigger fear, though, is that the trick of dividing power between participants in a large number of legislatures, which seemed like it might limit corruption, essentially unravels here. We will ultimately have a single body with the power to make massively influential decisions across the whole domain of policymaking, while being unconstrained by considerations of electoral accountability. The members of this body will be too powerful to avoid the attention of special interests. If, for example, the body is considering a defense appropriation bill that includes a proposal for a $5 billion fighter jet contract, on which the future of Lockheed Martin hinges, we should expect the company to try to influence the decision-makers by creating hard to resist inducements, influencing expert testimony, and so forth. Even if many citizens discharge their duty honestly, there will be others for whom power corrupts—citizens for whom the once-generous $150,000 salary that the book proposes suddenly seems like quite small potatoes, indeed.

Lottocracy argues that such influence trading can be prohibited and disincentivized through threats of punishment. It also suggests that it will be “easy to detect” payoffs from special interests, “particularly if they are ‘paying out’ to everyone who voted a certain way on a given proposal.”⁵¹ There is some ambiguity in how to understand this claim. If it means that we can figure out which votes have been inappropriately influenced just by seeing who votes with a special interest, this seems to understate the degree of disagreement that one should expect. Was I bought? Or did I just have a different considered opinion than you? It can often be hard to say. In an environment of reasonable disagreement, it would surely be unfair to presume that everybody who votes with a special interest was bought. And, in any case, we may reasonably expect that, for most votes, there will be special interests with huge private stakes on both sides of the question.

The claim could also be understood as insisting that it will be easy to detect the giving or the receiving of payoffs directly. Our sense is that this optimism is similarly unwarranted. Was my friend’s cousin really the best-qualified candidate for Lockheed Martin’s new executive position? Was my new position in the government relations office of a company upstream or downstream from Lockheed Martin justified by the expertise that I acquired during my years of service on the budgeting body? Were they anyway going to make a large contribution to my favorite charity? Were they always going to build the planes in my hometown? Extended network relationships, with their “six degrees of separation” properties, magnify such problems. If I am crude enough to sign a contract that specifies that I will vote in a given way for a check of a certain amount, then, sure, I might be found out and punished. But we should expect people, and corporations, ordinarily to be more effective in pressing their interests. The underlying concern is that lottocracy shares many of the tendencies towards corruption that endanger strong performance in representative democracy, but simultaneously lacks the primary tool—electoral accountability—that helps prevent its worst manifestations.⁵²

Lottocracy stresses the need to make fair comparisons between representative democracy and lottocracy. Yet, in the discussion of representative democracy, it views the corruption generated by private money as irremediable, quoting Samuel Issacharoff and Pamela Karlan, who argue that “‘political money, like water, has to go somewhere’—it will always find its own level.”⁵³ As we have just argued, though, there are surely non-campaign-related ways for money to influence outcomes in lottocracies. The book’s optimism that such attempts could be relatively easily detected and addressed in lottocracy seems to belie its commitment to making fair comparisons between regime types.⁵⁴

Lottocracy may, then, require more from anti-corruption measures than they can reasonably be expected to deliver. But, in fact, even if one could make those measures effective at policing corruption, it is far from clear that that would be, overall, desirable—especially for a lottocracy. Consider, for instance, the possibility of making revolving-door influence outright illegal, as the book recommends. If it is illegal to work in an area in which one had legislated and members of the budgeting body effectively oversee the whole economy, what are they supposed to do when they complete their service? And, would the group of people willing to serve in the face of broad restrictions on ex post private sector employment plausibly be descriptively representative? The more strictly one enforces anti-corruption measures, the more difficult it will be to get people to participate. Yet, if participation is not attractive, this threatens our capacity to assemble a truly representative governing body.

Indeed, even bracketing these anti-corruption measures, we are skeptical that it will be possible to form descriptively representative legislatures—a key part of lottocracy’s epistemic appeal, as the book imagines it. Consider the herculean efforts that go into getting a representative group of people to participate in weekend-long deliberative polls—James Fishkin even arranges for someone to milk one of his participant’s cows!⁵⁵ Yet even after all of this effort, the response rate remains relatively low. It is quite likely, then, that even if the participants match the broader voting-age population with respect to some salient demographic categories, they differ with respect to other, less observable, characteristics. This raises the concerns that often go under the heading of the “demographic objection” in recent democratic theory.⁵⁶

In light of the difficulties scholars have faced in constructing representative deliberative polls, it seems challenging, indeed, to get a representative group to spend, as the book’s plan requires, three years of their life in government (moving their families, putting other projects on hold, and so forth). One should expect that many people would not wish to devote a substantial portion of their life to deliberating about politics and living in the nation’s capital, and that those who would are far from randomly distributed within the broader population.⁵⁷ If this analysis is correct, however, not only will it be enormously difficult to assemble such a body, but our ability to do so will be inversely related to the enforcement of anti-corruption measures.

VI. Conclusion

Disagreements aside, it is clear that many democracies are failing in myriad important ways—failing to address pressing problems, stoking social conflict, generating vast inequality, and so forth. We urgently need a robust discussion about the best path forward. Lottocracy importantly contributes to that discussion—it is imaginative, bold, forcefully argued, and highly provocative. In laying out a comprehensive case for lottery-based political institutions, it provides a critical touchstone for current debates at the heart of contemporary democratic theory. Yet, as we have explained here, there are reasons to worry both that its dismissal of electoral institutions may be hasty and that lottery-based institutions will face problems precisely because of the absence of electoral accountability. As lottery-based institutions take an increasingly central position in reform debates, we hope that this essay invites a more sustained reckoning with the critical challenges to their justifiability.

Acknowledgements

The first version of this essay was presented at PPE’s annual conference in November of 2024. We thank panel participants and audience members for an illuminating discussion. Alex Guerrero not only invited us to participate in that panel, but also generously provided detailed comments on a written version. We appreciate his willingness to do so. George Klosko also helpfully commented on an earlier version of the essay. Finally, we are grateful to Free & Equal’s referees, whose careful comments helped us to improve the piece.

Competing Interests

The authors have no competing interests to declare.