I. Introduction
The state’s relationship to concentrated private power presents one of the most vexing dilemmas of modern politics. On the one hand, states are uniquely positioned to contest deeply entrenched social, political, and economic hierarchies. On the other hand, state policy written by and for dominant groups is one of the key mechanisms through which their dominance has long been maintained. How, then, can the state be transformed from a tool of concentrated private power into a weapon against it? The first aim of this essay is to show that this question cannot be ignored. The second is to outline a framework for answering it.
It is crucial to begin with this dilemma of public and private power, because opponents of the administrative state—including some represented in this volume—are apt to ignore it. Such critics are surely right to highlight the dangers of vesting too much power in a centralized bureaucracy removed from popular oversight and accountability. All too often, however, they neglect the dangers of vesting too little power in a centralized state. As I argue in Sections II and III, states are already so deeply embedded within the fabric of modern society that any attempts to retreat into the background will only exacerbate the problem, enabling dominant groups and powerful interests to further expand their social and economic advantages through decentralized means. States have often played a major role in creating and entrenching these advantages, of course, but they cannot simply withdraw at this stage, hoping to wash their hands of the problem. Their only choice is to develop new forms of action that challenge this concentrated private power.
The framework presented in Sections IV through VI answers this call. After showing how the threats posed by both public and private power can be understood to reflect the same underlying danger of state capture, I argue that orienting ourselves toward the goal of minimizing capture can help us answer three crucial questions. First, what general approach should states take to private power in each policy area? Second, which substantive policy tools are least susceptible to capture? And third, what procedural reforms can further insulate states from capture?
In responding to these questions, my primary aim is neither to defend a set of abstract normative foundations, nor to propose a concrete policy platform, but rather to provide a set of mid-range heuristics for the design of public and administrative institutions. These heuristics are meant to be compatible with many different foundational assumptions, and they are indeterminate between a similarly broad range of potential policy implications. Though I use certain policies as illustrations, that is, I do not defend any of them in comprehensive terms, and I do not claim that minimizing capture must be our only goal. Nevertheless, the framework that emerges yields powerful insights about the proper scope and optimal forms of state action, which may be useful to political actors with a variety of substantive convictions.
In turn, this framework demonstrates why opponents of the administrative state are wrong to suppose that abandoning its projects is the only solution to its pathologies. In fact, I argue, that is no solution at all. Assuming that there will be public power of some kind, one of its major imperatives must be to prevent itself from being captured by private interests—and that requires an active and capable administrative state. On the basis of broadly shared concerns about capture, in other words, I show that transforming the state into a reliable weapon against concentrated private power is both necessary and achievable. Grappling with the dilemma of public and private power may be difficult, but it is not impossible, and this essay highlights some promising paths forward.
II. What Is Concentrated Private Power, and Why Is It a Problem?
My first task is to explain why the state ought to be engaged in fighting concentrated private power in the first place, and I begin by defining the central concepts at work. Power, on my account, includes any of the conditions enabling agents to achieve their ends.Footnote 1 Not all forms of power must be possessed and used intentionally,Footnote 2 but those which are—such as wealth, social and cultural capital, institutional position, persuasive capacity, and so on—may be called power resources.Footnote 3 When held by actors outside of state institutions, these resources count as private, and any non-state agent or group with great power (relative to its peers) can thus be said to possess concentrated private power. Most obviously, these private actors include individual agents as well as organizations with the ability to act like agents, such as corporations or unions. In addition, however, my account also assigns concentrated private power to any group, class, or category of people whose members share common interests and possess asymmetrical power relative to members of other groups—even if they lack institutionalized collective agency. For instance, a racial or ethnic group can be said to have concentrated private power simply because its members possess more power resources, on average, than members of other groups.
Inevitably, there will be gray areas in classifying certain forms of power as public or private: What is the status, for instance, of state services that have been contracted out to private firms? This blurring of public and private realms is increasingly prevalent, and many regard it with suspicionFootnote 4—indeed, part of the reason I employ the concept of state capture is to help us grapple with such entanglements. Yet the existence of some overlap does not preclude all distinctions. Many forms of power are clearly public, including most legislative, judicial, administrative, and military institutions. Many others, meanwhile—such as the wealth of individuals, corporations, and interest groups—are not.
Defined as such, concentrated private power is not necessarily bad; in fact, some forms are indispensable to balance the concentrated public power of the state.Footnote 5 Yet concentrated private power can also be used to capture that public power, and this is, by definition, troubling. State capture occurs, on my account, whenever public power is made to serve the interests of some partial faction or private actor, at the expense of the public. All else equal, therefore, any instance of state capture is worthy of censure and resistance.Footnote 6 Yet capture can never be entirely prevented—indeed, we should expect it to be widespread in any realistic society—and the level of concern merited by any particular instance of capture will vary widely, depending on its severity. Capture can extend throughout many different arms of the state, for instance, or it may be confined to specific institutions. The public interest can be seriously undermined, or only mildly threatened. It is often initiated and maintained intentionally by some of those whose interests it serves, but this is not a requirement of the concept.
At the extreme end, examples of state capture include tyranny and apartheid, whereby nearly everything about the state is fundamentally organized around serving the interests of some at the expense of others. Yet the concept also covers more mundane cases of corruption and regulatory capture, in which a specific office or agency is turned away from its public purpose and made to serve a private or factional interest.Footnote 7 Though perhaps somewhat jarring, the extraordinary breadth of this concept has advantages, allowing us to see the commonalities in form among otherwise disparate phenomena.Footnote 8
The concept of interest plays a key normative role here, and in particular, the idea of the public interest will inevitably serve as the focal point for much contestation about the existence and relative severity of capture. Whether something counts as capture—and, if so, how problematic it is—will depend on whether, and to what extent, the public interest has been violated. In this essay, however, I refrain from saying much more about what the public interest actually is, because the broader theoretical framework I develop here is meant to be compatible with many different views on this question. Some might define the public interest in terms of a consequentialist conception of the common good, for instance, and others in terms of a deontological conception of justice, but both groups should have ample reason to oppose tyranny, apartheid, corruption, and so on. I hope to be persuasive to these and many other groups—each of whom can fill in their own, more specific views about the public interest—by leaving the normative foundations of my arguments deliberately underspecified.
Because we can identify particular instances of capture without settling upon a general definition of the public interest, meanwhile, this studied imprecision does not endanger the practical usefulness of my framework. As individuals seeking normative guidance, that is, we can be confident in identifying clear violations of the public interest, even if we remain hesitant to specify exactly what the public interest is in any detail. And as citizens deliberating with one another, similarly, we can agree on many concrete instances of capture without sharing an abstract conception of the public interest. Where imprecision or disagreement remains, finally—concerning the relative severity of different cases, perhaps—this may simply reflect the real uncertainty we face in the political world. Rather than claiming a level of certainty we cannot possibly possess, continuing to reflect and disagree along these lines is a valuable yet humbling exercise. It reminds us of the limits of our political knowledge while enabling us to expand, ever-so-slowly, our base of cumulative insight.Footnote 9
We are now better equipped to answer the question at hand: why must states fight concentrated private power? In the most general terms, the reason is that concentrated private power is dangerous.Footnote 10 More specifically, concentrated private power can be used to capture the state: though any partial faction or private interest could conceivably use state power to advance its ends at the expense of the public, those with the most concentrated power clearly pose the largest threat. This is not just intuitive speculation: capture of the state by those with highly concentrated private power has been the normal state of affairs throughout recorded human history.Footnote 11 Liberal rights, democratic procedures, and constitutional constraints all help to contain and resist capture, but every system has a breaking point, and in the absence of any restraints on the accumulation of private power, it is only a matter of time before some individual, organization, partial faction, or elite coalition becomes powerful enough to breach it, enabling them to use the state to pursue their interests at will. To protect against capture and abuse, therefore, we cannot simply limit the state’s power, defensively. Instead, we must also go on the offensive, taking proactive measures against the accumulation of private power—even though doing so inevitably carries risks of its own.Footnote 12
At the same time, of course, it would be absurd to seek the elimination of all concentrated private power: at the very least, I have emphasized, a flourishing ecology of diverse private organizations is necessary to check the abuse of state power. Thus, resisting state capture requires striking the right balance between public and private power, rather than subordinating one to the other. If the state is overwhelmingly powerful relative to all non-state organizations and private interests, it will be impossible to coordinate resistance to abuse and capture when it occurs. Yet as certain private interests and partial factions become more powerful relative to the state, public institutions become more vulnerable to having their public purposes hijacked.
In the rest of this essay, I adopt just such a balancing-act approach to the dilemma of public and private power. Before we proceed, however, it will be instructive to consider a salient alternative to that approach, which has proved tempting to a broad spectrum of democrats, socialists, and other radicals. This is to eliminate or at least mitigate the threat of concentrated private power by centralizing the functions performed by potentially dangerous non-state actors and subjecting them to the collective control of the people. Rather than fighting concentrated private power, in other words, the idea is to make private power public.
III. Why Not Make Private Power Public?
Perhaps the most prominent and far-reaching version of this suggestion belongs to a certain brand of socialism.Footnote 13 On the classic Marxist account, political power is largely epiphenomenal on material conditions. As long as citizens are barred from changing fundamental features of capitalist power relations, therefore—especially the private ownership of productive resources—democratic control of state policy can be little more than window-dressing. Even if it is exercised by elected leaders, that is, state power can never be legitimate unless the means of production are also subjected to collective decision-making procedures.
There is certainly something compelling about this idea, which follows a common account of political democracy to its logical conclusion. After all, nearly everyone agrees that in responding to the dangers of concentrated coercive power, it is clearly better to create a state with a monopoly on violence, and subject it to democratic oversight, rather than allowing warlords and private militias to compete over territory. Given this logic, many wonder why the same should not be true of economic and social power.
Traditional liberal replies to this suggestion have relied on an understanding of coercion as representing a categorically different kind of threat to individual autonomy than other forms of power.Footnote 14 In my view, however, decades of critique on a variety of fronts have rendered this response untenable.Footnote 15 The real problem with the idea of centralizing all economic and social power lies not with its refusal to make a principled distinction between coercion and other forms of power, but with its reliance on collective control to legitimize concentrated power.
In short, the concept of collective control is simply too slippery to be trusted with such a crucial task.Footnote 16 Consider, for one, that most people know very little about politics, and that even those with substantial knowledge reliably make decisions that affirm their group identities rather than an impartial reflective process.Footnote 17 We might imagine that under better circumstances, people could be relatively better informed and more civically minded, but no one—not even legislators themselves—can gain enough knowledge to make fully informed decisions about every political question, or entirely eliminate the possibility of motivated reasoning. Even if we assumed perfectly rational, impartial, and informed participants, finally, the possibility of preference cycles within a vast multidimensional issue space would still prevent the formation of a coherent popular will or truly collective decision.Footnote 18 As a result, what passes for collective decision-making in the real world is in fact pervasively structured by unequal power relations, as those with concentrated private power use their outsized resources to shape the political agenda—as well as media narratives and partisan attachments—to their advantage.Footnote 19
Many democratic theorists understand this problem perfectly well as it applies to elections. They recognize, in other words, that electing top-level political leaders is not sufficient for achieving truly collective control over centralized state power. As a result, most leave certain areas to individual discretion: even Marx recognized the wisdom of protecting personal property. Yet most democratic theorists also retain the ambition of bringing certain areas of common life under collective control, by moving beyond a minimalist conception of what is required by that ideal—calling for greater participation and deliberation, for instance, or a more thoroughgoing equality of opportunity for influence.Footnote 20
To put it plainly, I am not convinced. As I have argued elsewhere, these variations on the ideal of collective control suffer many of the same flaws as commonsense views grounded in elections, while introducing new problems of their own.Footnote 21 Yet further contesting these points is beyond my scope here. Instead, my aim is simply to provide a different way of determining which decisions ought to be centralized. Assuming that we cannot fully legitimize centralized political power by subjecting it to truly collective control, I frame the question in pragmatic and comparative terms, seeking to evaluate the dangers presented by concrete cases of centralization in relation to the dangers presented by concrete decentralized alternatives. And this perspective clarifies why our reasons for centralizing coercive power and subjecting it to collective control cannot be extended to the idea of socializing economic and social power.Footnote 22
At base, electoral democracy is an equilibrium in which no faction has a secure grip on power, such that all incumbents have strong incentives not to use coercive state power in brazenly partisan ways.Footnote 23 As a result, centralizing coercive power and subjecting it to the oversight of elected civilian leaders is far safer than allowing private actors to wield it independently. What electoral democracy cannot do, by contrast, is enable genuine collective self-rule.Footnote 24 As a result, our justification for centralizing coercive power and subjecting it to the collective control provided by electoral institutions cannot be extended to all forms of private power. Collective decision-making procedures cannot eliminate the danger of socializing ownership and control of all productive resources, media outlets, religious organizations, and academic institutions. Such measures would give leaders great discretionary power while leaving opponents with very few bases of power from which to contest incumbent action, thereby reducing the costs of incumbent abuse and threatening the fragile democratic equilibrium itself.
Nevertheless, if the state relinquishes all control over economic and social power resources, private actors are likely to accumulate them in dangerous amounts. After all, that is what gives rise to the dilemma of public and private power at the heart of this essay. In some policy areas, therefore, the safest option may be to facilitate decentralized competition with minimal state involvement, while in others, more active involvement will actually be safer. Determining which general approach to state involvement is appropriate in various circumstances is thus an important first step in employing the framework of resisting or minimizing state capture.
IV. Which General Approach to State Involvement Is Safest in Each Policy Area?
It is important to recall here that minimizing state capture is not the only goal we ought to pursue, and that state action might be justified for many different reasons. Purely on the grounds of economic efficiency, for instance, many people believe that a centralized state should provide certain public goods and regulate natural monopolies. Similarly, state-run public transportation systems may generate valuable environmental benefits, and universal provision of certain basic services may be desirable on grounds of justice. These are all potentially valid reasons for pursuing state action, which would have to be considered in any comprehensive practical judgment about the wisdom of these policies. Yet all raise questions about what counts as the public interest, and as noted, my arguments here aim to be compatible with many different accounts of what that is. I shall therefore set them aside, in order to focus more carefully on the particular questions raised by the dilemma of public and private power.
The sort of concerns I wish to consider here are internal to the demand to resist state capture, and they can push in either direction. State centralization can be quite dangerous in certain policy areas, and concerns about capture therefore lean toward decentralization in those areas. Yet as we saw in the case of coercive power, decentralized private ownership and control can also be quite dangerous, and when these dangers outweigh the danger of centralization, concerns about capture will suggest centralization instead. In practice, moreover, the options are not limited to “centralization” or “decentralization”; nor even to a continuous spectrum between the two. For instance, there are also dimensions of polycentricity and subsidiarity, while vast institutional diversity is possible within each dimension.Footnote 25
Setting aside such complexities for now, however, we may begin our discussion with two basic elements—a singular state and a pluralistic private sphere—which are typically related in one of four ways. First, there may be “public provision” by the state, such that private providers are rare or nonexistent—as with fire insurance, for instance. The second model may be called “public participation,” in which the state owns or controls one competitor in an otherwise decentralized market. Public education and postal services were often introduced to guarantee universal provision, for example, but private competitors are permitted in most places. Public radio and television broadcasters, meanwhile, typically aim to ensure that a certain type of content is provided, rather than universal access to these services.
The third model may be called “substantive regulation,” meaning that rather than participating in decentralized competition itself, the state seeks to shape the outcomes of the competition by making rules about how it must proceed. This is how crucial social functions like housing and finance are managed in most modern states, though in some cases the state may be a limited participant here as well—as with public housing projects or postal banking services. Indeed, the borders between these models can become quite fuzzy at times.
The fuzziest boundary of all, however, is that between substantive regulation and the fourth model, “formal regulation,” which describes less active state involvement in a decentralized sphere. Some might be tempted to call this a “laissez-faire” or “free market” model, but these labels are seriously misleading in this context. No matter how “private” or “unregulated” it seems, every sphere of life in modern societies is pervasively structured by the state’s background enforcement of private and criminal law. Terms like “laissez-faire,” “free market,” and “unregulated” occlude this reality, giving the impression of a stark binary between areas of social life where the government “intervenes,” and areas where private actors choose “freely.” By contrast, the distinction between formal and substantive regulation enables us to discuss real variation in the state’s general stance toward private actors across different areas—that is, how specific its policies are, how frequently they change, and how intentionally they seek particular outcomes—without falsely implying that proactive forms of involvement are necessarily more consequential for the distribution of power than the enforcement of “basic” legal structures that everyone simply takes for granted.
Even when they are written in general terms, stable over time, and not clearly intended to achieve particular outcomes, for instance, small differences in rules governing contracts and torts can still have enormous distributive implications.Footnote 26 Indeed, the very existence of a system of private law has immense distributive implications as well, and cannot simply be assumed as a neutral background.Footnote 27 The question we must ask, then, is not whether or how much a state ought to intervene within a pre-existing private sphere, but rather how it ought to structure private interactions. More specifically, the choice between public provision, public participation, substantive regulation, and formal regulation turns out not to be like choosing a point on a quantitative scale of state involvement, such that more involvement always brings greater risks of capture. Instead, the safest option in any case depends on particular features of the policy area or set of social functions under consideration.
Take the question of health care provision. Given its great complexity and central social importance, no reasonably advanced modern state engages only in basic formal regulation of actors in the health care industry. Most employ fully public provision, while the others rely on public participation and/or very significant substantive regulation. If we assume that minimizing state involvement will always reduce the overall risks of capture, then, we would expect some form of substantive regulation to be the safest available model. Yet a comparison between European-style public provision and the model of substantive regulation employed by the United States gives us reason to doubt this supposition, for the latter gives powerful private actors far more substantial opportunities to ensure that state policy serves their interests.
To begin with, the very idea of structuring the health care system around private insurance corporations was originally devised and implemented at the behest of the American Medical Association, over widespread protests that it would predictably serve the interests of doctors above everyone else.Footnote 28 The most significant revision of that system, meanwhile—the 2010 Affordable Care Act—was written in close consultation with insurers and other industry giants, whose profits have soared in the years since it was passed.Footnote 29 This model of substantive regulation within an otherwise decentralized market is almost certainly the major reason that health care costs are so much higher in the United States than in other comparable countries.Footnote 30
Of course, systems of socialized medicine are also subject to all sorts of pathologies, and it is not my purpose here to defend any particular way of organizing a national health care system. The point is simply that a centralized system with more active and obvious state involvement is not, by definition, more prone to capture than a decentralized system with a state that apparently stands back from the fray, serving merely to police the competition of private actors. Indeed, the unmistakably proactive nature of the state’s involvement in a system of public provision makes its activities salient in ways that may help to insure it against capture.
If everyone uses public transit, public education, or public health care services on a regular basis, the quality of service provision will be very salient for everyone, and the state actors responsible for this provision will have relatively little room to extract rents. If the role of the state in structuring a private market is hidden behind an impenetrable thicket of regulations and licensing rules, by contrast, the state’s responsibility for poor provision and other undesirable outcomes delivered by private actors will be far more opaque.Footnote 31 This lack of clear accountability channels is a major enabling factor for capture.Footnote 32 In fact, Jon D. Michaels claims, this is precisely what has happened in recent waves of privatization, which have given certain private interests exclusive access to the material, cultural, and legal support of the state, while weakening democratic oversight of the broadly public functions they continue to serve.Footnote 33
One final dynamic worth noting is that systems of universal provision make it more difficult to use health care as a bargaining chip or particularistic reward for certain factions, which can help to entrench particular individuals or parties in power. Far beyond the particular context of health care, in fact, universality can often be a powerful weapon against capture, eliminating the discretionary power that state agents would otherwise use to extract rents and perpetuate their own factional interests.Footnote 34 This is another reason to expect that public provision may be less prone to capture than regulation of decentralized markets.
Again, this is not the place to stake out a comprehensive position on health care or any other deeply complex policy question. All four models of public involvement in the private sphere have advantages and disadvantages from the perspective of resisting state capture, and these concerns must then be weighed against others we might have—about efficiency, fairness, and so on—which are external to this framework altogether. In many policy areas, formal or substantive regulation may be the safest approach, as well as the best overall. The point of lingering on the health care example, rather, is to show that we cannot assume that less active, salient, and intentional forms of state involvement are always less prone to capture. Instead, we have to evaluate which approach is safest—public provision, public participation, substantive regulation, or formal regulation—on a case-by-case basis.
This case-by-case evaluation will focus on the policy area itself, of course, but it may also involve more contingent features of the context. As long as significant independent sources of power remain, for instance, some degree of state ownership may be acceptable, but the more the balance of power between state and non-state actors shifts, the more dangerous it will be to put additional resources in state hands. Meanwhile, the effectiveness of accountability through elections and civil society will affect the relative safety of centralization projects, such that wealthier, more educated, more engaged, and less polarized societies will likely be able to accommodate greater levels of state ownership or management, while those divided along regional, ethnic, racial, or religious lines may face unique risks with particular policies.
As this discussion suggests, then, the contribution of the framework articulated here—which aims at minimizing the overall risk of state capture—is not only to help us decide which general model of public-private interaction is safest in a given policy area. In particular, it helps us answer two further questions, in each case generating generalizable strategies for the design of public and administrative institutions. The first concerns the form substantive policy should take. Given that regulation of the private sphere is required, for instance, many choices remain about how to implement that regulation, and a focus on state capture suggests several heuristics for making those choices. Meanwhile, the second question concerns the procedural tools that might be able to further insulate state action from capture. I address each in turn.
V. Which Substantive Policy Tools Are Least Susceptible to Capture?
The framework of resisting state capture is grounded in a constant awareness of the threats of capture that are posed by those with concentrated private power. In considering the design of substantive policy, therefore, it aims to minimize three key risk factors for this outcome: the amount of discretion allotted to state actors, the number of levers for influencing policy that are open to private actors, and the independent capacity of private actors to perpetrate capture. In turn, this mandate generates several design principles for capture-resistant policy, notably including transparency, simplicity, and adversarialism. As usual, however, there are no silver bullets, and these design principles will apply differently across different circumstances.
Each of these design principles can be observed in the idea of “structuralist” regulation developed by K. Sabeel Rahman.Footnote 35 According to Rahman, early Progressive-era thinkers like Louis Brandeis and John Dewey sought to democratize the nascent administrative state, so that it could curb the concentrated power of big business without succumbing to the forms of capture that its laissez-faire critics rightly identified. Yet these democratic approaches to administration were eventually superseded by expert-driven or “managerialist” models of regulation during the New Deal era—which then set the tone for the subsequent development of the American administrative state.Footnote 36 Managerialists did not ignore capture entirely, Rahman acknowledges, but their primary concern was to insulate bureaucrats from political conflict, on the supposition that expert administrators would better realize long-term common interests if not forced to cater to the short-term interests of politicians. And while that may be true, he points out, insulation from all political influence also opens these expert bodies to other forms of capture.Footnote 37 On Rahman’s telling, the result of New Dealers’ managerialist confidence was a regulatory state that did manifest a number of pathologies, and when world events tested public confidence in experts during the 1970s, the neglected Progressive alternative was overlooked in the course of a sharp turn back to the laissez-faire pole with the onset of neoliberalism.Footnote 38
Rejecting this binary between capture-prone “managerialism” and an uncompromising laissez-faire alternative, Rahman posits the early Progressives’ ideas as a third way.Footnote 39 Instead of abandoning the project of regulation altogether, more specifically, he argues that we should follow Brandeis and others in elaborating “structuralist” approaches to regulation that are more resistant to capture.Footnote 40 And though his book primarily addresses financial regulation, the lessons of his broader narrative can be applied to many areas where substantive regulation is the appropriate model of state involvement.
For managerialists, on Rahman’s account, the goal is to fine-tune market dynamics, and this process thus places significant trust in experts to get the details exactly right. Aside from the risk that experts might make poor decisions for entirely innocent reasons, this level of discretion also creates a serious threat of capture—and far from resolving this threat, insulation from all political scrutiny tends to exacerbate it. A managerialist approach to certain regulatory tasks requires extremely detailed knowledge that must often be provided by regulated firms themselves, creating what Wendy Wagner calls “information capture.”Footnote 41 More generally, the close working relationship that is required for this style of managerialist regulation—as well as the broader social background and cultural affinities that regulators share with those they regulate—can result in what James Kwak calls “cultural capture.”Footnote 42
On the alternative approach Rahman proposes, regulation would focus less on fine-tuning market outcomes and more on formulating “bright-line” rules and broader changes to the underlying structure of financial markets.Footnote 43 On the one hand, this “structuralist” model exhibits greater humility about the capacities of regulators to do the fine-tuning work. On the other hand, it also exhibits more skepticism about the tendency of “free” markets to produce optimal results. It accepts the broader social value of finance—that is, allocating resources more efficiently than feasible alternatives—but because of the high systemic risks of innovation in such a crucial field, it puts the burden of proof on financiers to show that novel products are worth whatever dangers might accompany them, rather than on regulators to show that they are too dangerous to allow. In other words, a simple and transparent mandate serves to minimize the number of opportunities available to industry actors for influencing the proceedings.
Crucially, it also enables regulators to take an adversarial stance toward those they seek to regulate. Where managerialist regulation aiming to exercise fine-grained control requires close collaboration between state and industry actors, a structural approach guards the public nature of state power more jealously. Indeed, hypervigilance toward industry actors is embedded within the substantive mandate for structural regulation, which seeks not only to make defensive regulatory policy that is relatively capture-resistant, but also to go on the offensive by reducing the independent capacity for capture possessed by industry actors.
In other words, at least part of the job of regulators is to undermine the structures making capture of future regulators more likely. Capture is more likely where industries have united interests, for instance, or when particular companies wield immense power on their own, and less likely when there are diverse interests vying for a regulator’s attention.Footnote 44 Thus, structuralist regulation should try to prevent these concentrations of private power from existing in the first place. As examples, Rahman mentions the separation of commercial and investment banking, limits on bank size, and aggressive antitrust enforcement—all of which have clear potential benefits from the perspective of minimizing capture.Footnote 45 Of course, these proposals may also have drawbacks not considered here: again, my aim is not to offer a comprehensive defense of any particular policy, but simply to illustrate the sort of tools that state actors might use if they are seeking to enable administrative action while minimizing the danger of capture.
Generalizing from Rahman’s approach, then, we can note its strong resonances with the more general Progressive spirit of antimonopoly—perhaps most memorably encapsulated in an essay by Brandeis called “The Curse of Bigness.”Footnote 46 In this spirit, we might replace a backward-looking, criminal model of regulation, whereby only violations of pre-existing rules can trigger regulatory discipline, with a forward-looking, prophylactic model. The idea is that by the time serious harm has been done, it is often too late, because the wrongdoers will have already captured or co-opted those who could have punished them. Instead, we must understand bigness itself as a threat, because it provides the kind of market power and organizational capacity that enables anti-competitive behavior and the capture of state institutions. More urgent in the long run than punishing particular cases of wrongdoing, on this model, is preventing the conditions that would enable future wrongdoing to go unpunished. Competition is a crucial tool for resisting capture, in other words, but protecting it often requires active state involvement. Once again, the dilemma of public and private power cannot be ignored.
VI. How Can State Action Be Procedurally Insulated from Capture?
The motivating question of this essay, in brief, concerns how the state can safely fight the dangers posed by concentrated private power, given that empowering it to do so inevitably carries dangers of its own. So far, I have focused largely on substantive policy—considering the general approach the state should take to the private sphere across various policy areas, as well as the kind of tools it should give to regulators and bureaucrats, if we want them to be resistant to capture. No matter which substantive policies we pursue, however, it is also crucial to consider how they are implemented. Any power that is given to state actors inevitably carries certain dangers, and in order to reduce those dangers as much as possible, we must also develop a suite of procedural strategies to keep state actors tethered to the public interest.
The basic framework of liberal democracy, anchored by competitive elections, is the most obvious example of such a procedural strategy. Unfortunately, it is also far from sufficient: diverse forms of capture plague even the most “advanced” liberal democracies. Indeed, it seems like the administrative state poses special problems in this regard: as the other essays in this volume demonstrate, independence, legislative oversight, and executive control all open the door to various forms of capture. In this section, therefore, I explore how institutionalizing oppositional expertise and direct popular oversight within administrative agencies can democratize the administrative state from within, providing additional layers of protection against capture and thus helping us resolve the dilemma of public and private power.
A. Institutionalized oppositional expertise
As I have suggested throughout, adversarial competition can often serve to weaken the grip of concentrated private power, but whether it does so depends entirely on the way this competition is structured. In regulatory contexts, for instance, practices of lobbying predictably favor those with the strongest incentives to influence policy, the greatest capacity to coordinate, and the largest stockpile of resources. Unsurprisingly, the results are often contrary to the public interest. Yet the proper solution is not to eliminate lobbying, as many are tempted to suggest: even if this quixotic feat could be accomplished, the activity of lobbying serves too many important informational purposes in a complex and pluralistic society to be abandoned entirely. On the other side of the spectrum, a similar problem afflicts the proposal of George Stigler and other public choice theorists, who insist that the only possible solution to regulatory capture is to eliminate regulation itself.Footnote 47 As we should recognize by now, this extreme posture does not represent a serious attempt to grapple with the dilemma of public and private power.
Rather than forsaking lobbying as inherently unfair, or regulation as invariably corrupt, I argue that administrative procedures can best be protected from capture if we transparently embrace their adversarial nature, while leveling the playing field among groups with a clear interest in the outcome. The idea is to make rulemaking fairer, not by stifling or concealing competition between groups who seek only to protect their interests, but by equalizing the terms on which that competition will inevitably take place—that is, by supporting disadvantaged groups and institutionalizing opposition at every stage.
As counterintuitive as it may sound to abandon neutrality in this way—putting our fingers on the scale to support competitors with the fewest resources—this sort of proposal should not be wholly unfamiliar. It is widely recognized that stable democratic institutions depend on strong, well-organized opposition parties, and most people also recognize that antitrust and other forms of antimonopoly can contribute to healthier economic competition. The idea of institutionalizing oppositional expertise simply extends the same logic to the administrative state. Of course, the question of which interest groups require remedial support will always be somewhat contentious. Yet in the context of the administrative state, it is clear that regulated industries have the strong incentives to manipulate policy, as well as the most concentrated organizational capacity, so it is not difficult to see why it is their opponents who need additional support as a corrective.
Scholars of regulatory capture have generated a variety of proposals that fit this general model. One is to place “contrarians” of various stripes within regulatory agencies themselves.Footnote 48 Another is to invite challenges from third parties, and reimburse their costs when their challenges are successful.Footnote 49 Still another is to give oversight to courts,Footnote 50 or to super-regulatory watchdog agencies.Footnote 51 Yet the details of the proposals are not particularly important to my broader argument. The key point is that regulators must defend their decisions as consistent with the public interest, against adversaries with an institutional mandate to highlight where those decisions may actually serve the private or factional interests of a corporation, industry, or other lobby group. Because of their embrace of an adversarial approach, such strategies do have significant potential to disrupt cycles of capture within regulatory agencies. Yet over time, those with concentrated private power will predictably figure out how to manipulate the new rules to their advantage, and those rules may gradually lose their ability to prevent capture.
In my view, therefore, institutionalizing opposition will work best when supplemented by avenues for direct participation by ordinary people. To be sure: widespread citizen participation is no panacea, and as I argued in Section III, there is no simple connection between popular participation and collective control over legitimate outcomes.Footnote 52 Indeed, it is not without reason that managerialists see insulation from political pressure as an important tool for resisting capture. Yet popular participation takes many forms, and it is perfectly consistent to insulate bureaucrats from the corrupting influences of legislative horse-trading and hyper-partisan mass politics while creating more promising levers for popular influence.
B. Randomly selected citizen oversight juries
One such mechanism involves random selection for office—an ancient practice known as “sortition” that is rapidly winning converts among theorists and reformers alike.Footnote 53 So far, however, this contemporary discussion is largely dominated by a single use case: deliberative mini-publics composed of randomly selected citizens.Footnote 54 More generally, it assumes that lotteries are primarily valuable as an alternative to representative elections, serving to interpret the popular will more accurately or authentically.Footnote 55 In my view, this narrow focus obscures the full potential of sortition, directing reformist energy away from more fruitful goals. Though many of its advocates tend to understand sortition as an extension of deliberative democracy, thereby rejuvenating representation or renewing popular sovereignty, the most astute analysts have understood that its distinctive potential lies in its capacity to prevent the capture of public institutions.Footnote 56 Rather than trying to replace existing legislative institutions with bodies of randomly selected citizens, I have argued elsewhere, the most promising use for sortition lies in “citizen oversight juries” that screen administrative decisions for signs of capture.Footnote 57
This entails giving randomly selected citizens real power—unlike many existing experiments with deliberative mini-publics, whose “advisory” recommendations have often been ignored by real legislatures.Footnote 58 Yet it would also avoid many of the serious objections to more radical “lottocratic” proposals that would give direct legislative power to randomly selected bodies. For one, the idea of replacing representative bodies faces extraordinary practical hurdles in most countries, and even if such obstacles could be overcome, the resulting system would likely face serious legitimacy issues. Perhaps most concerning, it would still depend heavily on agenda-setting procedures that are hidden from view, and thus would hardly be immune to the kinds of capture plaguing representative institutions. Understanding sortition as one among many methods for preventing capture and corruption, by contrast, suggests a role for ordinary democratic citizens that is realistic yet crucial. Without denying the importance of political elites in policy formation, it casts ordinary people as guardians of the public interest.
Why might this help to prevent capture? On the one hand, sortition introduces an element of randomness into political procedures that frustrates the attempts of powerful private actors to corrupt the public nature of those procedures. Despite their real advantages, election and appointment are also subject to certain forms of manipulation, which predictably render them biased toward the interests of those with concentrated power. By inserting a “blind break” which interrupts these channels of influence, a truly random process thus levels the playing field.Footnote 59
On the other hand, those randomly selected for oversight roles will be ordinary, coming from outside the established halls of power. Given that wise political judgment often requires significant expertise and experience, of course, this requirement has obvious drawbacks. Indeed, that is one reason not to follow sortition’s more radical proponents in demanding the complete replacement of elections with lotteries. At the same time, political naïveté has distinct advantages as well. It is precisely because most participants would begin the process with low political knowledge and few ideological commitments that their deliberations could provide a distinctive check on career politicians and other official experts, whose judgments are intimately tied up with ideology and other identity commitments.
On the specific model I propose, moreover, ordinary citizens would be tasked not with general-purpose legislation but with oversight on specific issues. They would not be expected to generate an entire agenda, but simply to make yes-or-no judgments about particular cases. Oversight juries might be convened on an ad hoc basis in response to administrative decisions of a certain type or significance level—such as redistricting decisions, for instance, or merger approvals and defense contracts worth more than a billion dollars. Meanwhile, longer-standing councils could meet for several months to conduct a more general review of an administrative agency, or scrutinize interactions between legislators and lobbyists.
In each case, crucially, the process would be structured to include institutionalized opposition, so that the agency would be forced to defend its decisions as serving the public interest, while various adversaries would attempt to convince the jury otherwise. The jury would then decide to approve the policy, reject it, or send it for further review elsewhere—though in theory, juries would not need to reject policies very often, because the threat of exposure posed by this process would disincentivize capture from occurring in the first place. Meanwhile, it would not require specialized knowledge beyond what the agency and its opponents could provide. As with the ordinary citizens in trial juries, therefore, members of oversight juries could be sufficiently trained through this adversarial process itself. This would give supposedly neutral facilitators and experts far less power to set the agenda or otherwise influence the proceedings, thus making juries far less susceptible to capture.
Nevertheless, a system of citizen oversight juries would still depend on the existence of organized groups to raise challenges, and countervailing experts to make arguments against agency policy. Ordinary citizens serve as neutral arbiters in contests among various experts and elites, representing a generic public interest. Yet clearly, ordinary people may also have distinctive interests of their own, which will not be well represented by any of these elites. In addition, then, it is also crucial to enable ordinary people to organize themselves to press their own priorities, forming groups that can recruit and employ such countervailing experts. The idea, in other words, is to create leverage points for popular mobilization: footholds in the policymaking process that encourage oppositional organizing outside the state.
C. Leverage points for popular mobilization
More specifically, my proposal is to structure rulemaking so that members of disadvantaged groups have opportunities to exert meaningful leverage, outside of the electoral cycle, on issues that are relevant to their interests. These formal entry points into the policy process could then serve as focal points for mobilization and organization in civil society, in turn enabling ordinary people to meaningfully counter the extra-procedural influence that well-resourced groups are routinely able to exercise with at least some measure of genuine political power.
This proposal differs from many popular proposals for encouraging citizen participation, which are generally designed to gather information and work out compromises among equally situated interest groups, not to accommodate coordinated mobilization and popular pressure from below.Footnote 60 In practice, such participatory forums are often co-opted by powerful interests with the means and the incentive to coordinate and exert pressure.Footnote 61 In order to meaningfully level the playing field, then, leverage points must be carefully designed to prevent this kind of co-optation. And again, an oppositional orientation is crucial: these forums should invite contestation while providing support for those groups with fewer resources.
One way to accomplish this is to adapt existing proposals for institutionalizing oppositional expertise, like the idea of rewarding groups for successful challenges. This has been analogized to the “tripartism” between employers, unions, and the state, where the union represents a source of countervailing power that is supported by the legal framework. A more speculative possibility is raised by plebeian theorists of democracy, which entails excluding elites entirely from certain participatory forums, as with the institution of the tribunate in ancient Rome.Footnote 62
Both proposals are more promising than much of what passes for participatory democracy. Admittedly, however, both are still quite limited in their ability to overcome deep-seated asymmetries of power. On the one hand, administrative tripartism enables people without many resources to contest bureaucratic policy on a more equal footing, but is unlikely to generate widespread participation among those whose interests are at stake. On the other hand, plebeian institutions may generate mass mobilization and identification with popular leaders, but they provide little incentive for healthy epistemic practices or long-term organizing. Ordinary people participate, on this model, but only as an undifferentiated mass.
In order to match the consistent pressure on agencies and other state actors that wealthy elites are routinely able to apply, ordinary people must somehow acquire a comparable capacity for coordinated and collective action, over sustained periods, in defense of specific goals. That requires organizing large numbers of people—the only concrete advantage ordinary people have over wealthy elites—within a structure that accommodates long-term strategic planning.Footnote 63 Of course, states have only a limited role to play in facilitating this kind of organizing, but the priorities I have suggested—that is, providing forums where organized groups can contest policy decisions, and actively supporting those with fewer resources—are an important start.Footnote 64
As always, of course, many objections might still be leveled against each of these policy ideas, and it has not been my intention here to offer a comprehensive defense of any of them. Rather, the purpose of my discussion is simply to illustrate what it might mean to orient our political action toward the goal of minimizing capture. In the context of designing procedural protections against capture in the administrative state, this means employing oppositional expertise and direct popular oversight within agencies—and indeed, these strategies reinforce one another. On the one hand, the adversarial institutions recommended by scholars of regulatory capture will be most effective at resisting capture if they engage popular participation in the ways I have suggested. On the other hand, popular participation will be most effective at resisting capture when it is employed in adversarial rather than collaborative contexts, and when it is specifically targeted at supporting countervailing power. Though these principles may be realized in a variety of different ways, both are necessary if the administrative state is to take seriously its mandate of fighting power with power.
VII. Conclusion: Can We Ever Trust the Administrative State?
Contemporary critics of the administrative state sometimes suggest that the project of using centralized state power to provide public services and regulate the private sphere should largely be abandoned. Tracing the origins of this project to the frightening visions of social engineering articulated by elitist (and often racist) Progressive-era reformers, they condemn its track record as anti-democratic. Looking to the future, they warn of ever-increasing prerogative for unaccountable experts operating at ever-greater distance from the concerns of ordinary people. And though the picture they paint is often one-sided—neglecting more democratic visions of administration, as well as the important democratic goals many agencies have achieved—these critics point to very real dangers with centralized administrative governance.
This essay has tried to acknowledge these dangers, while insisting that the solution cannot be to abandon the project of the administrative state altogether. On the one hand, I have argued, the public power of the state is not the only threatening kind of concentrated power. Private power can become dangerous as well—both in its own right, and because of its tendency to capture the state. If we eliminate the state’s capacity to contest the growth and concentration of private power, crucially, we actually increase the risk that the state’s public power will eventually be captured by powerful private actors. Critics of the administrative state who seek to resolve its problems by getting rid of its major functions are simply refusing to grapple with this dilemma of public and private power, and the cure they propose is almost certainly worse than the disease.
On the other hand, I have claimed, it is entirely possible to respond more productively to the real dangers of administrative governance, by making it more robust to capture. First, we can choose a general approach to state involvement across different policy areas that minimizes the risk of capture within each domain. As noted, this will not always entail decentralization or a withdrawal of the state, and in fact may sometimes require more active or intensive state involvement. Second, we can design substantive policies that minimize the discretion given to state actors while restricting private actors’ formal opportunities and independent capacity for capture, by employing principles of transparency, simplicity, and adversarialism. Third, we can procedurally insulate state action from capture by institutionalizing countervailing power and popular oversight at specific junctures that are especially prone to capture.
Clearly, these sketches of an approach to reforming the administrative state fall far short of a comprehensive policy platform. As noted throughout, my aim in this essay has not been to defend a particular normative theory, nor to prescribe a particular set of institutions. Appealing to broadly shared concerns about the danger of state capture, rather, I have sought to provide a set of mid-level heuristics for designing administrative institutions. By design, these heuristics are open to a range of interpretations, meaning that those with different foundational assumptions and substantive intuitions will inevitably come to different conclusions about the particular policy program they entail. At the same time, the framework that emerges is not infinitely flexible. Of this much, at least, we can be certain: abandoning the project of the administrative state is a dead end. Instead, we can and must learn how to transform it into a reliable weapon against concentrated private power.
