The submerged prison state: Punishment,private interests,and the politics of public accountability

Abstract

Much of penal policy is “submerged,” in the sense that Suzanne Mettler uses the term. There are networks of rules and regulations that link public funds, services, and institutions to various private interests with far reaching consequences. These networks are largely a stealth presence in the lives of citizens and their subterranean status, I argue, warps the wider politics of punishment. Resources are circulated along this network in such a way revenue is generated for some, costs cut for others, all in the shadow of public law. To the extent that this kind of redistribution lacks citizens’ consent and approval, it also represents a potentially undemocratic development. Here, I show how the obscured visibility of these public–private connections distorts public attitudes about, and public support for, the US prison state. The final pages draw out the normative implications of that distortion.

Introduction

The criminal law is simply one of many avenues used to secure public goods related to punishment. ¹ Working alongside the criminal legal system is a skein of rules, codes, and procedures that pursue similar ends by other, often more bureaucratic means. Municipal codes are used to target drug and sex-work markets, for instance, and the tax code buttresses a sprawling industry of private prisons and private prisoner transport services (Beckett and Herbert, 2010; Eisen, 2017; Gunderson, 2022). The world of “criminal justice” is increasingly a world beyond criminal law. A world of plea bargains and bail bonds, not trials. A world of actuarial justice, where general criminal behavior is managed through financial incentives (Feeley and Simon, 1992; Harcourt, 2007). Some see these policy developments as pernicious and worry about their potential to widen the net of penal institutions (Beckett and Murakawa, 2012; Harcourt, 2007). ² Others accept this transformation and seek to rationalize an otherwise haphazard collection of rules and procedures through, for example, coordinating organizations modeled on the US Sentencing Commission or Federal Reserve Board (Barkow, 2019; Natapoff, 2018; Slobogin, 2021).

Central to this form of statecraft is the use of third parties to achieve policy objectives. The most familiar version of this relationship is the privatization of public services; that is, the contracting out of the provision of public goods to private organizations. For many, this calls to mind government contracts with Aramark to provide food to prisoners or with CoreCivic to administer immigration detention facilities. ³ However, there is a less familiar, but equally pervasive use of third parties to implement public policy in the domain of punishment—and it's the form I explore here. A number of policies covertly steer rather than overtly direct by adjusting the financial incentives for action or inaction. Consider the difference between a state government contracting with a nongovernmental organization (NGO) to provide job training to parolees and that same government passing a tax credit for businesses that employ workers with a felony record. The effect is the same, perhaps, but the line of accountability that links policy enactment to outcome is more opaque for the latter. Similar incentives not only are used to influence the actions of private companies and NGOs, but also to shape the behavior of landlords, employers, social workers, family, romantic partners, friends, and neighbors. ⁴ To borrow a term from Suzanne Mettler, these kinds of policies are “submerged” (Mettler, 2011). ⁵

Below I argue that submerged penal policies are a hidden but significant presence in the lives of citizens and their sub rosa status, in turn, warps the wider politics of punishment. Resources are circulated along this network of rules, procedures, and regulations in such a way that revenue is generated for some, costs cut for others, all in the shadow of public law. In an analysis of legislation, judicial cases, policy regulations, and survey data, I explore this increasingly important face of “penal levies” (Garland, 2020). Vitally, efforts to transform these submerged policies generate a distinct kind of politics that alienates the public, entrenches private interests, inhibits organizational learning, and expands the size of the carceral state.

I start by locating the amalgam of these penal policies, what I call the submerged prison state, in existing scholarship, then turn to its defining political dynamics. Politicians have a strong incentive to pursue submerged policies over direct spending programs and, once in place, private interests that benefit from these various and sundry policies are well-placed to defend them. Given these dynamics, I sketch two underappreciated features of the submerged prison state. First, I suggest a simple focus on direct spending obfuscates the scale of the total social investment in punishment. Second, I show how submerged policies redistribute wealth on the sly—and, importantly, redistribute in a way that likely diverges from public preferences. In the absence of a clear line of responsibility, citizens have difficulty attributing blame to public officials. Worse, as argued in the final pages, there is a broader civic cost to the use and extension of these policies: a passive public. Unaware both of how government is acting and how justice is actually administered, a passive public develops attitudes toward both that undermines trust, distorts perceptions of fairness, and erodes accountability.

The submerged prison state

A large body of law and society scholarship over the past 25 years has converged on a shared portrait of the modern welfare state. The choice has never been between a free and regulated market, we’re told, but different forms of state management. The “rollback” of the welfare state over the past half century has been accompanied by the rollout of new forms of statecraft. Prisons are part and parcel of this transformation. New institutions and new modes of governance have been constructed out of the old (Cruikshank, 1999; Feeley and Simon, 1992; Garland, 2001; Harcourt, 2011; Hörnqvist, 2020; Soss, 2002; Wacquant, 2009).

The rise of the submerged state is a central through-line in this story. There is an increasingly unwieldy conglomeration of federal, state, and local policies that work by providing incentives, subsides, or payments to private organizations or households to encourage or reimburse them for conducting activities deemed to serve a public purpose. Much of this is done through the tax code and civil law. In contrast to direct spending, tax expenditures are “departures from the normal tax structure… designed to favor a particular industry, activity, or class of persons.” They can take the form of tax deductions, tax credits, preferential tax rates, tax deferrals, or outright exclusion of income from taxation. The classic examples used by scholars in this area are tax deductions for home mortgage interest and charitable contributions. Tax expenditures are about three-quarters the size of all discretionary direct expenditures and far exceed the national deficit (Howard, 1995).

What I’ll refer to as submerged policy is defined by its methods, not by its aims. It uses the tax code and liability rules to “nudge,” rather than explicitly direct, the target group into particular actions or behaviors that forward wider state interests. These interests might be as basic as encouraging citizens to save for retirement, to name one, or secure their property against natural disasters, to name another. There have been searching examinations of these policies in contexts that range from healthcare to education to housing (Sunstein and Thaler, 2003). While some have turned to criminal justice, the scope, scale, and consequence of these policies remain underappreciated and undertheorized (Thurston, 2020). From policing, to bail, to the siting of prisons, to “re-entry” programs—this technique of governance steeps through civil society with far reaching effects.

A host of important studies over the last decade have drawn attention to the proliferation of pathways that now join civil and administrative law with prison confinement, and case studies are legion. Parole revocation, to name one prominent pathway, is now largely treated as an administrative affair. Parolees are subject to myriad rules, regulations, and requirements, many of which potentially entail re-incarceration for noncompliance. Despite this, revocation hearings are not treated as part of the criminal legal process and do not require criminal due process standards. ⁶ Most infractions within jails and prisons, to name another pathway, are also adjudicated by administrative procedure. And these hearings have similarly severe consequences—reductions in good time, the imposition of solitary confinement, among others (Calavita and Jenness, 2014; Rudes, 2022).

Scholars have used a variety of terms to characterize this more administrative face of the carceral state: “shadow,” “hidden,” “covert,” among others. While the visibility of particular regulatory processes is an important theme that I’ll return to, the class of phenomena I examine here—submerged policy—doesn’t cleanly reduce to these existing frameworks. ⁷ The submerged prison state is both more narrow and more broad. It's more narrow than the “prison beyond the prison” described by Gottschalk, Beckett and Murakawa's “shadow state,” or Feeley and Simon's “new penology” (Beckett and Murakawa, 2012; Feeley and Simon, 1992; Gottschalk, 2015). It doesn’t simply mark administrative and civil pathways to prison or new, “net widening” carceral entanglements (electronic home monitoring, for instance). ⁸ Submerged policies are a broad form statecraft that works by using financial incentives to induce third parties to pursue public crime control objectives.

Katzenstein and Waller come the closest to describing submerged phenomenon, but they’re largely focused on the net effect of these policies, not their internal functioning or politics. They introduce the term “state seizure” to turn our attention to processes that tax poor families to fund the state's project of mass incarceration and poverty governance. The authors rightly call our attention to a slew of nauseating, vulturous practices. To take one, they note a Texas-based company that provides “video visitation” services in over 2600 correctional facilities. Rather than expanding access, this “service” is largely financed by poor families. Seventy percent of the contracts require video visitation replace in-person visitation, and families are made to pay about a dollar per minute to talk to their loved ones (Katzenstein and Waller, 2015: 639). Recently, Page and Soss, along with Lara-Millan, have written persuasively about fines, fees, forfeitures, and premiums as part of a vast network of revenue generation—the rise of a predatory form of capitalism that strips resources from the least well-off (Lara-Millan, 2021; Page and Soss, 2021).

Collectively, these works offer a helpful corrective to accounts that otherwise ignore the increasing significance of administrative and civil law to the prison system. However, they also leave open the broader puzzle of how these policies emerge, who defends and contests them, and their wider consequence for democratic politics. By wedding law and society scholarship on the carceral state with the political science literature on the submerged state, a clearer picture comes into focus.

Submerged politics

All elected officials, if they wish to remain in office, need to find ways to represent their constituency, to respond to the needs of those who elected them and those votes they would like to attract. In the United States, the challenge for Republicans and conservative Democrats, in particular, is to do so without expanding the visible state. As a consequence, they prefer submerged state policies primarily because they make it possible to deliver goods and services to core constituencies while avoiding the creation of new direct spending programs or enlarging the bureaucracy (Faricy, 2016; Mettler, 2011: 17). Submerged policies are also a way to deliver benefits to stigmatized populations like prisoners without risking public backlash (Faricy, 2016; Thurston, 2020). Tax incentives for churches to provide prison-based programming can be sold as promoting private philanthropy. The Work Opportunity Tax Credit which provides tax credits to employers that hire workers with a criminal record can be billed as “promoting employment” rather than “providing a handout” to an undeserving population. And so on.

It should be unsurprising that submerged policies have become a preferred vehicle for policymaking. Whatever the domain—punishment, housing, healthcare—institutional features of Congress and state legislatures make these policies easier to enact than other direct spending programs. Tax expenditures face fewer veto points in Congress because the House Ways and Means and Senate Finance committees serve as the equivalent of authorizing and appropriating committees, not to mention that expenditures rely almost exclusively on third parties for service delivery (Howard, 1995: 441). Ruthie Gilmore describes the importance of the municipal finance sector in the push to use Lease Revenue Bonds (LRBs) in the buildup of California's “golden gulag” during the 1980s. They pushed LRBs as a solution to the problem of how to expand a politically popular program (prisons) without running up against the politically contradictory limit to taxpayers’ willingness to use their own money to defend against their fears. Unlike General Obligation Bonds, LRBs do not have to be placed before the voters in general elections, and on approval by the legislature, they can be quickly organized and issued. In less than a decade, the amount of state debt for prison construction expanded from $763 million to $4.9 billion, an increase from 3.8% to 16.6% of the state's total debt for all purposes (Gilmore, 2007: 97f).

These political incentives have had a profound impact on the spread of submerged policies in the domain of crime and punishment. The aggregate of these penal policies—the submerged prison state—is not the product of one, coordinated program. Nor is it designed to be efficient. It arose out of politics, and its design reflects the interests, strategies, and compromises of those who exercise political power. Over time these policies have cultivated the support of agents that defend them vigorously. From social service providers, to private employers, to property owners, these new policies have given rise to new stakeholders. Agencies, interest groups, and individual private actors seek to maintain the rules and have a vested interest in resisting any efforts to contract that system (Mettler, 2011: 16–17).

In How It Works, Fairbanks takes a look at flophouses in Philadelphia with a careful, critical ethnographic gaze. He describes how everything from recovery house licensure, to zoning laws, to tax credits, has enabled a cottage industry of informal addiction recovery houses to flourish. To receive general assistance (food stamps, cash aid) for drug and alcohol dependence, to take one of his examples, one must attend intensive outpatient treatment, followed by a “step down” to outpatient treatment. The municipally mandated treatment requirement has given rise to a host of providers in North Philadelphia which serve the welfare population on a fee-for-service basis. Attempts to fight this regulation are scattered and, unsurprisingly, are fiercely resisted by providers. The concept of “recovery,” Fairbanks notes, serves as a single site for multiple regulatory projects, each of which has a stake in maintaining the status quo (Fairbanks II, 2011).

The influence of private interests is magnified in the context of submerged penal regulations. In contrast to, say, questions about the morality of the death penalty, most people don’t get exercised about tax policy or civil liability rules. For those with a financial stake, however, the story is different. They are energized, active, and informed participants in their specialized issue areas, and they know that their goals are crucially dependent on precisely those fine details of the tax code that cause most folks’ eyes to glaze over. ⁹ These rules and procedures are valuable to them, and they have every incentive to mobilize their political resources to get what they want. As a result, they are normally the only source of political pressure when submerged policies are being discussed (Faricy, 2016; Howard, 1995; Mettler, 2011: 32; Moe, 1989: 269).

These core political dynamics are cause for concern. Political failure is worse than any one policy failure, whether it be with prisons, policing, or crime; political irrationality is a kind of a “meta-problem,” it's a problem that inhibits our ability to collectively solve all other problems. A lack of public visibility and salience, policy attribution problems, and powerful interests with an incentive to maintain the status quo lead to a broken system of policy adoption. To these criticisms I’ll add that, unchecked, submerged policies promote civic decay. Many citizens lack basic information about how these regulations function, and public officials fail to provide it, leaving the public unable to express its preferences in any meaningful way (Mettler, 2011; Ramirez, 2021).

Below, I suggest these policy dynamics entrench an unwieldy prison state that hides its true scale, redistributes wealth, and erodes public knowledge and trust. Using legislation, court decisions, survey research, and data on policy implementation, let's consider each dimension in turn.

Scale

Overall criminal justice spending rose from under 10 billion in the early 1970s (and about 50 billion in today's dollars) to over 200 billion by the early 2010s. This spending has largely moved in lock-step with state and local revenues. During peak expenditures during the late 1990s, states dedicated 7% of all spending to criminal justice, and 4% to corrections; for counties, about 8% to criminal justice and 2% to jails (Pfaff, 2017: 94–97).

We don’t have similar aggregate figures for tax expenditures in the penal sphere, yet it's still possible to glimpse their scale. Tax expenditures for security measures on commercial properties alone, for instance, stack up to the height of billions. Section 179 of the Tax Cuts and Jobs Act (TCJA) of 2017 explicitly allows security deductions, to take one recent example, not to mention the more than 1000 business improvement districts (BIDs) in the United States that spend about one-fifth of their budgets on policing and security through self-imposed property taxes. Earlier work has shown how these funds subsidize a private security force that, as a whole, is now substantially larger than public police departments (Sklansky, 1998).

Tax expenditures needn’t be so large in scope; they also can be smaller and more targeted. In Washington, DC, for instance, the Private Security Camera Rebate Program creates a tax rebate for residents, businesses, nonprofits, and religious institutions to purchase and install security camera systems on their property and register them with the local police (Stein, 2016). When thinking about scale, focusing on county, state, or federal budgets only tells a part of the story. The DC program just mentioned, for example, issued about 7500 rebates between 2016 and 2020. The cost to the District—about 5 million dollars—doesn’t capture the total resources being spent on the program. The program encourages residents to spend at least as much of their own funds on the purchase of equipment and private subscription plans (Amazon's “Ring”, SimpliSafe, to name two that are popular at the time this article was written), in addition to spending time to install and monitor that technology. A proper discussion of scale would not just look the incentive, itself, but the total social investment steered by the incentive.

When one starts to aggregate these submerged policies more broadly, these policies may be hidden, but, make no mistake, they are costly. A tax “break” or “credit,” on the books, is no different from a direct transfer of those funds to the target group. ¹⁰ Even so, tax expenditures are rarely mentioned in the same conversation as direct expenditures. One hears about the amount spent on prison construction or the amount spent on private police, but less about tax benefits of siting of prison on a particular parcel of land or the tax credits claimed by commercial properties to fund those private security services (Eason, 2017; Howard, 1995; Sklansky, 1998). More concretely, by operating as a Real Estate Investment Trust (REIT), private prison companies receive favorable tax benefits. In 2015, GEO Group paid a mere $7.4 million in international, federal, and state income taxes on its $1.84 billion in revenue (Eisen, 2017: 123). ¹¹

The tens of thousands of these rules, procedures, and regulations, large and small, need to be cataloged and evaluated—an enormous task. Only then can local, state, and federal governments have a road map for their repeal or modification. To be clear, the fact that investments in the penal sphere are so immense does not mean that they are unsustainable. Despite its tremendous size, spending on crime and punishment is not so much of a drain on budgets that decarceration is a necessity, as some reformers hoped after the 2007 economic crash (Berk, 2011; Pfaff, 2017; Zimring, 2020).

Redistribution

It's not simply that submerged policies increase the total “size of the pie” of investments in the prison system; they also alter the way that pie is divided and who pays for what slice (Faricy, 2016). Submerged policies generate revenue for some and cut costs for others. While not perfect, it's helpful to think about the distributive effects of these regulations in terms of taxation—even if they’re not written in the tax code. The central thrust of Natapoff's discussion of the misdemeanor system in Punishment without Crime riffs on this melody. The petty-offense process, she notes, has “quietly become a regressive feature of American tax policy.” It actively extracts revenue from an ever-widening pool of low-income people in order to fund the operations of both private and public criminal justice institutions (Natapoff, 2018).

The tax breaks given to land-owners to site prisons, the credits given to employers, the legal protections for landlords, the notification requirements for those on parole: benefits tend to flow from the least well-off to those that are in a relatively better position. Consider the Work Opportunity Tax Credit (WOTC) mentioned in passing earlier. Few, I suspect, would object to the goals of the WOTC and other government bonding programs—increasing employment opportunities for those returning home from prison and other disadvantaged workers. Even here, however, one can see distributional consequences. The vast majority of these tax breaks goes to large corporations and, even then, in only a few sectors characterized by high-turnover and low wages. While eligible, few community businesses actually use the credit even though they form a larger part of the tax-base that funds the program (Collins and Donovan, 2018; Hamersma, 2011).

To the extent that politicians discuss these submerged policies publicly, they usually imply that they facilitate widely held goals, such as reducing crime, strengthening families, or promoting employment. In some cases, this portrayal is reasonably accurate. In others, like the WOTC, it's more ambiguous. However, most submerged policies exacerbate inequality. They shower their most generous benefits on the well-off, and they generate detrimental side effects that adversely impact those less well-off.

Sometimes the re-distributive consequences of the tax code and liability rules are more opaque. The legal tools used to insulate police departments, private prisons, half-way houses, addiction clinics, and so on, from lawsuits, for instance, raise a similar set of questions. ¹² Worse, in a number of instances payouts to victims of neglect by private organizations come from public coffers. In 2006, Virginia began contracting with Armor Correctional Health Services. The company worked in only three of the state's prisons at that time, but by 2014 Armor took over as the only private healthcare company contracting with the Virginia Department of Corrections. Armor began to have problems in 2012 when women in the Fluvanna Correctional Center for Women brought suit. After several women died, many others came forward saying that they were not given medications they had been prescribed and that healthcare workers at the prison regularly lost prisoners’ medical records. The case dragged on for years until it was finally settled in 2016 for $1.5 million dollars—a sum ultimately paid by Virginia, not Armor. ¹³

These examples underscore a more general point. Submerged policies can favor different parts of criminal justice system (away from job readiness and social work, e.g., to surveillance and policing) and entire sectors of the economy over others, at government expense. It picks particular winners and losers. And as with other submerged state policies, the winners tend to be finance, insurance, and real estate (Mettler, 2011: 25).

Public trust and knowledge

In addition to distributive consequences, the submerged prison state comes with another, admittedly more speculative kind of cost—democratic erosion. In addition to complicating attributions of blame and credit, submerged policies can undermine public trust and knowledge.

To start, submerged policies encourage a passive public. That is, they do little to instill in their beneficiaries an awareness that they have utilized public benefits (Mettler, 2011). Likewise, those that receive little in return for their contribution to the tax base are unable to identify, and evaluate, how public funds are being spent. In the absence a clear line of responsibility, citizens have difficulty attributing blame to public officials (Ramirez, 2021). This matters for programs that run the gamut from domestic violence shelters to Pell Grants for former prisoners (Johnston and Wozniak, 2021a). Citizens see a government that does little about crime and punishment and so they expect little of it. And while lay people might not be eager to provide input into all political decisions, they do want to know that they could have input if they ever so desired (Hibbing and Theiss-Morse, 2002). Unaware both of how government is acting and how justice is actually administered, a passive public develops attitudes toward both that undermine trust, distort perceptions of fairness, and complicate accountability.

Diminished trust and engagement have knock-on effects for prison reform efforts. For one, it encourages punitive responses to social harm. Noting the role of trust, Nicola Lacey usefully offers a comparative perspective on reform. Moderate penal policies and low imprisonment rates in a country, she writes, are associated with high levels of trust and legitimacy, low levels of income inequality, strong welfare states, professionalized as opposed to politicized criminal justice systems and consensual rather than conflictual political cultures. ¹⁴

For another, as the severity of a given crime increases the need for trust becomes even higher to elevate expert-backed responses to that crime. Bringing restorative justice practices, to take one demanding approach, to cases of murder, rape, and assault require a well of social capital much deeper than, say, bringing those same practices to elementary school discipline (Sered, 2019). As a number of observers have noted, violent crime (and the fear of violent crime) is likely the single largest stumbling block to prison reform efforts (Gottschalk, 2015; Miller, 2016; Pfaff, 2017). One of the shared lessons of both procedural justice and street-level bureaucracy scholarship is that the visible state, and perceived fairness of process, often matter more than actual outcomes (Tyler, 2006; Zacka, 2017). By obscuring state policymaking, particularly with rehabilitative and restorative alternatives to incarceration, the well of that trust is difficult to fill.

There is also a more general consequence for legal socialization (Tyler and Trinkner, 2017). The interactions between individuals and criminal legal institutions (police, prisons, courts, postrelease services) constitute a kind of hidden civic education that encourages or discourages active citizen participation and shapes attitudes about justice (Justice and Meares, 2014; Lerman and Weaver, 2014). For those returning home from prison, for instance, that hidden curriculum is an assembly of object lessons in powerlessness. Reuben Miller has recently described the phenomena of “carceral citizenship.” This term, Miller explains, marks how “third parties operating within the public and private sphere are empowered to manage, correct, sanction, and care for the carceral citizen.” Leaving prison, he writes, there “is a new power dynamic. Relationships, for formerly incarcerated people, now began from a position of need. You need the apartment or the job with the boss who mistreats you or the relationship with the lover you no longer care for because you have nowhere else to go” (Miller, 2021; Miller and Stuart, 2017). The warp and weft of these transformed social ties, in no small part, is structured by submerged policy. A social world that appears to be “pre-given” for those returning home from prison is better described as made.

The use of tax expenditures and civil law, as opposed to direct expenditures, to fund the expansion of the prison state magnifies these effects. Local, state, and federal policies delivered in this way are largely hidden from public view and are harder to attribute to a particular authority or a coherent policy platform. They’re simply experienced as arbitrary rule.

Publicity

The ideal of democratic responsiveness is predicated on the assumption that the citizenry can hold political authorities accountable. This requires judgments about who to credit and who to blame. Consequently, the method of policy delivery—not simply its content—matters. Unchecked submerged policies cultivate a passive public that, at best, ignores and, at worst, reinforces the worst excesses of the US criminal legal system.

However, it doesn’t have to be this way. It's not that tax expenditures and adjustments to liability rules, simpliciter, are worse than their direct-spending counterparts. Experimental work on lay attitudes suggests support for otherwise identical social programs is generally higher when such programs are portrayed as being delivered through tax expenditures than when they are portrayed as being delivered by direct spending. Importantly, the authors of the study find that support for tax expenditure programs which redistribute wealth upward drops when citizens are provided information about those redistributive effects. Both results, the authors continue, are conditioned by partisanship, with the opinions of Republicans more sensitive to the mechanism through which benefits are delivered, and the opinions of Democrats more sensitive to information about distributive consequences (Faricy and Ellis, 2014).

This highlights a simple, but an important point. The submerged prison state need not remain obscured from view. Opportunities exist to expose it. ¹⁵ And exposing it can matter for public support, which is a key backstop for reform programs that hope to endure through periods of heightened public concern about crime (Miller, 2016). Likewise, publicizing elements of the submerged state can precipitate the withdrawal of that public support and bolster social movement organizations. In our moment, one can see this at work in movements to abolish cash bail, divest from private prisons, limit “e-carceration” (e.g. electronic home detention, geo-fencing), and rein-in local prosecutors (Bazelon, 2020; Eisen, 2017; Schenwar and Law, 2020; Walker, 2022).

By distancing themselves from those who implement public policy, public officials may believe they can avoid blame for potential failures. However, recent work in moral psychology suggests a more complex story. The use of private agents to implement public policy, once exposed, can result in citizens assigning more blame to government policymakers in situations involving, for example, wrongful deaths or civilian casualties in war. Beliefs about the motivations of actors matter for blame attribution (Ramirez, 2021). This fits with recent public opinion research that shows a general distaste for the privatization of punishment (Burkhardt, 2019b; Enns and Ramirez, 2018). ¹⁶ The result, as divestment campaigns have learned, is a useful opportunity to leverage public pressure to constrain the submerged prison state (Eisen, 2017: 6). Politics, of course, extends to visibility itself. The private prison industry, to take one prominent example, is well aware of the potentially damaging effects of public opinion on their bottom line and acts to minimize or contain potential harm (Gunderson, 2022). ¹⁷

In the domain of punishment, however, some argue that the situation is more complicated. The idea that “offenders should help pay for the cost of their punishment” is popular among voters, and drawing attention to these costs might actually increase support for maintaining the status quo and undermine reform efforts. As a result, we’re told, one should be cautious transplanting the lessons of submerged state policies in domains like education and healthcare to criminal justice (Thurston, 2020: 290). This echoes other scholars, like Rachel Barkow (2019), that contend a disinterested, or detached, public provides insulation against capricious and vindictive lay attitudes. ¹⁸ Likewise, scholars in this vein argue that worries about democratic erosion are less pressing if policies, or policy outcomes, track what the public actually wants. Or, better, what the public wants when its preferences are formed under good conditions.

Both objections, on closer inspection, are unconvincing. As I’ve argued elsewhere, among the reasons this view is misguided is that institutions are rarely as “insulated” as advocates claim. Effective insulation from public opinion or partisan machinations requires a number of empirical conditions that are unlikely to be present, such as public trust in experts and second order insulation (insulating the institutions that determine the rules and procedures of that institution). Not only are these factors rarely all present, particularly in the United States, these conditions are subject to decay over time. As a result, the label “insulationist” grants too much to its advocates; their position is probably better described as technocratic. In practice, this means the actual choice is not between insulation and integration, but between different forms of integration with public opinion. Haphazard public involvement, or decayed forms of insulating institutions, risk producing worse outcomes than either strict insulation or unmediated lay participation (Berk, 2021; Dzur, 2012).

In addition, it would be a difficult case to make that both the scale and the distributive impacts described earlier track public preferences. What public opinion data we do have on punitiveness and prison costs suggests a general distaste for heavy financial investments in incarceration and strong support, with some caveats, for social services. ¹⁹ This appears to hold across a number of social programs that directly assist former prisoners. ²⁰

One might harbor a broader objection to publicizing submerged policies: the concern for democracy, in this domain, is misplaced. The demands of the public might depart from the requirements of a morally defensible theory of justice. Changes in public attitudes have directly impacted prison policy adoption over the last half-century, for example, sometimes for the better, sometimes for the worse (Enns, 2016). Fair enough. There are a number of domains where the gap might be particularly significant—at least under existing conditions. ²¹ However, changing attitudes is possible. Enns, among others, has noted that how crime is portrayed in the news matters for the tone and direction of public response. When news media use thematic frames, focusing on the social factors that influence criminal activity, for example, instead of episodic frames emphasizing violence, public attitudes move in different directions. ²² Even partisanship doesn’t appear to be an insurmountable barrier (Johnston and Wozniak, 2021b).

Moreover, there is a distinct and potentially offsetting kind of civic harm created by the use and extension of submerged policies in criminal justice. When used to sidestep the public, this tool of statecraft becomes a form of moral abdication. It binds a democratic polity to “a condition in which critical reflective moral judgment is rendered impossible or undermined to the point at which the only available choice is to obey someone else” (Cordelli, 2020). That is, it degrades the lay public's capacity to reason about justice—a capacity that is built from participation, not abstention (Tyler and Trinkner, 2017).

Conclusion

Stepping back, we’ve seen how networks of submerged penal policies link public funds, services, and institutions to various private individuals and organizations with far reaching consequences. They are largely a hidden presence in the lives of citizens and their subterranean status facilitates their steady encroachment on large swaths of social life. This mix of rules, codes, and procedures creates troubling incentives for politicians and for private interests. Collectively, I’ve argued these policies constitute a “submerged prison state” that not only obscures its true scale, but redistributes resources on the sly and erodes democratic accountability.

The full scope of these policies has yet to be fully cataloged and assessed, and here I’ve only gestured at their long reach. What's brought into focus, however, is an underappreciated style of statecraft and a few key implications of its use. More research remains to be done.

Still, two broader lessons are already evident. The first is that particular policy tools exhibit distinctive patterns, and those distinctive patterns matter for the growth and maintenance of the prison system. Law and society scholarship has gone to great lengths to assess the prudence of particular regulations, liability rules, and privatization more generally. Those regulations that promote employment, for example, in contrast to those that create a debt trap for fathers trying to get back on their feet (e.g. Miller, 2021; Lara-Millan, 2021). We need to look not only at how these policy tools shape the incentives of third parties, but the incentives that drive the adoption and maintenance of the tools themselves.

The second lesson concerns democratic politics. As crime and punishment objectives are increasingly pursued through alternative means, and when those other means stay submerged, they threaten the basic foundations of public accountability. The obscured visibility of public–private connections distorts public attitudes about, and public support for, penal policies. This is in addition to the extractive, exploitative forms of seizure that many submerged policies typify (Katzenstein and Waller, 2015). A number of authors have called to extend advocacies for progressive taxation to the fiscal costs of punishment. ²³ While laudable, that call might be productively paired with a conversation about the public's role in shaping, constraining, or rolling-back the submerged prison state.