Abstract
Lawmakers face strong political incentives to criminalize ever more conduct and punish it more harshly, even when adding a new crime or increasing sentence length may harm public safety. Strikingly, unlike in other areas of lawmaking, cost concerns and fiscal restraint play little or no role. That can happen because, although states pay for prisons, many costs of criminal law remain diffused onto local governments or foisted onto defendants; all these costs are obscured from view. The costs of imprisoning more people for longer don’t come due immediately, and states cut corners when the bills come due by underfunding indigent defense, prosecutors, and prisons. As a result, states get more criminal punishment than they’d be willing to pay for. Rights to counsel are routinely violated, and prisons are horrifically dangerous.
This Article considers how procedural tools that legislatures use to encourage fiscal restraint in other contexts can restrain criminal lawmaking. More specifically, the Article considers total spending caps, budgetary offset requirements, fiscal notes, sunset provisions, and prohibitions on unfunded mandates. To a very limited extent, some states already use these tools in criminal law. For instance, a very few states require themselves to budget for some portion of the estimated increase in prison costs when they pass carceral legislation. But even those states ignore the big-ticket expense of building new prisons or revamping existing ones. We propose procedural tools that can encourage legislators to think harder about the costs of criminal law, which may also help them consider hidden assumptions about the benefits of criminal law.
Introduction
Criminal law is extremely costly—for state governments who fund prisons; for local governments who police, prosecute, and jail; for people locked in jails and prisons; and for the families and communities who must carry on in their absence. What do we get for those extraordinary costs? Rarely do legislators who pass expansive criminal codes and harsh sentencing schemes that empower other system actors to cast a broad net pause to think hard about that question.1 Just about everyone proceeds as though criminal law were invaluable and that more were better.2 We aim to disrupt reflexive criminalization and carcerality to promote more deliberative lawmaking.3
State legislatures too often treat costs as though they were irrelevant to criminal law. Although most states require fiscal notes that estimate a bill’s cost before it can pass, many ignore those rules for criminal lawmaking. In our home state of Alabama, for instance, the legislature recently created a new felony prohibiting possession of firearms with “Glock switches”—devices that convert a pistol into a machine gun.4 According to the bill’s sponsor, no one throughout the legislative process mentioned how much the new crime would cost the State or counties to enforce,5 and its fiscal note simply said that it would cost “an undetermined amount.”6
With the significant exception of the need to ultimately provide subsistence funding to prisons, criminal lawmaking largely proceeds as though it were either free or impervious to cost.7 It’s neither. Indeed, even the price that states pay for prisons is in many instances too low to maintain minimal constitutional standards.8
Across many areas of law, legislatures temper their zeal using procedural mechanisms that urge fiscal restraint. These mechanisms range in the rigidity of their constraints. Total caps on spending growth impose a rigid constraint with room for tradeoffs only against other spending. Mandatory offsets require that new expenditures be paid for with either spending cuts or tax increases. Fiscal notes require legislatures to estimate a bill’s cost before passage but don’t require substantive action. Lastly, legislatures sometimes set new legislation to expire (or sunset) after a set period to revisit its cost-benefit tradeoffs later with the benefit of additional data.9 Unlike Congress, most states have to balance their budgets, so fiscal constraints necessarily play a role.10 But that role is far smaller in criminal law than it ought to be.
Our premise is simple: Legislatures ought to consider costs when they change substantive criminal law or sentencing law, as in other areas of legislation. Estimating cost takes some time and resources; that is a feature, not a bug. Carceral legislation has strong political tailwinds, so a process that modestly offsets those tailwinds and urges greater deliberation by increasing the salience of costs is good.11
A few states helpfully specify which costs of criminal law should be estimated during the legislative process—prosecutors’ offices, public defenders, police, jails, prisons, probation, and parole.12 Few require themselves to set aside money now for those future costs, however.13 Colorado’s legislature had required itself to appropriate funds to cover increased costs to state prisons for the first five years after passage of a new law—an excellent step toward forcing itself to account for at least some of the costs of carceral legislation.14 But Colorado recently repealed that fiscal accountability measure.15 Other states laudably require themselves to budget money now for future prison cost increases, but they leave a gaping hole by exempting the biggest ticket item: building new prisons or improving existing ones.16
Costs and decision-making responsibility are diffuse and frequently misaligned in criminal law.17 In general, states and municipalities pass criminal laws and set sentencing parameters, states pay for prisons, counties pay for prosecutors and jails, cities pay for police, and funding responsibility for indigent defense varies. Many states seek to shift costs onto people accused of crime, although that doesn’t work very well because the overwhelming majority of people charged with crimes in America are poor. The diffusion and misalignment make accountability difficult, and they afford state legislatures opportunities to evade fiscal responsibility for their choices. In discussing legislative tools of fiscal restraint below, we remain attentive to these issues of diffusion and misalignment.
Some readers may recoil at the idea of fiscally constraining criminal law insofar as they think that more criminal law is always good. But this perspective elides the difference between criminal law and public safety.18 Public safety is a critically important government function, and government officials should want as much of it as they can get. Criminal law can increase public safety, at least outside of prison. But so too does criminal law detract from public safety when we house incarcerated people in incredibly dangerous conditions where physical and sexual violence pervade. Criminal law removes huge numbers of people from their families and their communities. These are public safety harms that ought not be ignored. We suspect that greater consideration of the costs of criminal law might surface richer discussions of its benefits too.
It may be difficult to figure out in advance how much a new law will cost because of the diffusion of costs and responsibility in criminal law across various levels of government, and it may be at least as difficult to predict the benefits. But a decade after a legislature enacts a law, we can study those empirical questions and assign normative weight to the benefits; we can even compare criminal law to other tools that might accomplish the same objective. While we think it’s important for legislatures to consider costs before enacting criminal law, we think it’s also important to revisit those questions with the benefit of data. Sunset provisions in criminal laws can help promote such subsequent deliberation. Considering costs and weighing them against benefits several years after a law is enacted should seem entirely ordinary. But in criminal lawmaking that approach is far from ordinary because “more” is nearly always treated as better.19
This Article is driven by the insights that we shouldn’t get more criminal law than we’re willing to pay for and that budgets are supposed to align the cost we pay in taxes with the value of what we get from those tax dollars.20 As Rick Bierschbach and Stephanos Bibas put it, “in a bloated penal state, requiring criminal justice to compete against other objects of funding is a recipe for bringing costs and coercion into line.”21 We build on Bierschbach and Bibas’s work about scarcity as a tool to ration criminal law and Darryl Brown’s call for cost-benefit analysis in criminal law.22 Unlike those scholars, however, this Article turns to legislative process tools. Bierschbach, Bibas, and Brown rightly want criminal law legislation to account for a broad set of costs—including costs that criminal law inflicts on families and communities of those processed through the criminal legal system.23 Just because costs are not easy to monetize does not mean we should overlook them. We agree. And criminal law is rife with hard-to-monetize costs. Despite our fundamental agreement with this point, however, this Article focuses on the more modest project of structuring legislative decision-making to consider costs to the public fisc—both state and local.
A few scholars have considered, at least in passing, the role that a particular legislative fiscal restraint mechanism could play in criminal law.24 This Article is the first to consider the whole panoply of options and compare them to each other. It considers how the theoretical scholarship about these legislative tools maps onto criminal law and looks descriptively across the fifty states at how these tools are occasionally used in criminal law in some places.
There is plenty of room for skepticism about whether legislatures would adopt processes that fiscally restrain them in criminal law. Perhaps they like their systems better without such processes; otherwise, they would have adopted them already. Most haven’t. Instead, they muddle through the budget year by ignoring the future (and perhaps current) realities of their prison systems25 or rely on sentencing commissions to constrain the costs of the carceral state.26 We hope that more legislatures will subject criminal law to the fiscally constraining pressures to which they subject other legislation. But their choice not to do so is itself worthy of critique.
In sum, greater attention in the legislative process to the costs of criminal law and potential tradeoffs could promote deliberation and help restrain legislatures’ strongest carceral incentives and impulses. Some of the tools that legislatures use to constrain fiscal excess in other contexts could work in criminal law too.
This Article proceeds in three parts. Part I explains why the politics of criminal law yield steadily more carceral outcomes and how more focused attention to cost and its tradeoffs in legislative process might restrain some of legislators’ carceral impulses. Part II briefly explains how the costs of criminal law are spread across various levels of government and onto private actors. Part III analyzes the legislative procedural tools that states (and Congress) use to constrain legislation through fiscal pressure, and it analyzes how states can (and occasionally do) use these tools in criminal lawmaking. It does so with a close eye to avoiding simply shifting more costs onto local governments.
I. Why Cost Concerns Could Help Constrain Mass Incarceration
The carceral forces that drive mass incarceration in our politics are strong. But there is reason for some cautious optimism that putting carceral pressures into tension with fiscal conservatism might help constrain some of the worst carceral excesses, at least at some times and in some places. The carceral state is extremely expensive,27 and some might be willing to consider whether we’re getting anywhere near the benefit that we’re paying for.28 Would voters increase their taxes for more policing and prisons? That answer will vary across time and place, but the point is there’s reason to think that at least sometimes the answer will be no. And that answer may differ when the question is posed without making cost salient.
A. Pathological Politics
Criminal law is created by legislatures. Legislatures have strong incentives to appear tough on crime by expanding criminal codes and passing harsher sentencing laws.29 Important decisions about the scope of criminal law are sometimes made in “reaction to a single crime.”30 Consider Arkansas severely tightening parole after one person with a history of parole violations committed a high-profile murder.31 After a woman fabricated a report of a toddler who went missing along a roadway—already a misdemeanor—Alabama passed a new felony that applies when “the false report alleges imminent danger.”32 Politicians feel the need to address public sentiment that is shaped by cognitive biases and fear, amplified by media coverage.33
Although it would be politically dangerous for legislatures to pass too few criminal laws or not respond to perceived crime crises,34 legislatures don’t have much to fear from overcriminalization, as Stuntz explained.35 Prosecutorial discretion allows legislators to blame prosecutors if their voters perceive overreach.36 Indeed, diffusion of power across actors and branches of government in criminal legal systems makes it difficult for voters to know whom to blame for their discontent.37
Carceral legislation continues to pass in very high numbers—1,781 new punitive state criminal laws passed across the country from 2015 to 2018.38 But carceral legislation also fails at a reasonably high rate.39 And decarceral laws pass too, albeit far less often than carceral laws.40 To take a few examples of decarceration, in recent decades, several states such as Alaska, California, Colorado, Illinois, and Louisiana have reduced drug sentences, or decriminalized or even legalized marijuana.41 Decarceral interest groups, grassroots organizations, and the voices of individually affected citizens have traditionally been understood as quite weak at the state level,42 although recent evidence suggests that dynamic has shifted somewhat.43
Although we focus on state-level criminalization and legislation, local governments too drive mass criminalization. More than half of the criminalization of public order offenses comes from local rather than state governments.44 Offenses like panhandling and public urination come overwhelmingly from local rather than state criminal law.45 And because of the nature of local governments, local criminal laws can only add to state codes, not subtract from them.46 Although we focus our analysis on state legislatures, there is reason to think that it would be similarly important for local governments to consider the costs of criminal laws to avoid reflexive criminal lawmaking.
B. Cost as Constraint on Carcerality
Despite the pathological politics, there is reason to hope that in some places at some times fiscal pressures can restrain carceral policy.47 Fiscal prudence can lead legislatures to decrease sentence severity,48 such as by eliminating mandatory minimums.49 Fiscal pressure caused legislatures to make decarceral changes to sentencing law in a wide array of states across varying political terrain in the late 1990s and early 2000s—from Louisiana and Alabama to Connecticut, Oregon, and Washington—as the effects of the highly carceral ’80s and ’90s combined with economic downturn.50 Cost concerns were prominent in several states as a reason for abolishing the death penalty.51 Sometimes legislators may use fiscal responsibility as the way to justify what might otherwise be unpopular decarceral choices.52 In Alabama, for example, the legislature raised the dollar thresholds for theft crimes to reduce prison costs—a move that came after voters rejected a tax increase for prisons.53 Sentencing commissions are often responsible for estimating the costs to changes in sentencing law, and those estimates can limit or at least offset efforts to increase sentences.54 Although there is little reason to be hopeful about fiscal pressures constraining federal criminal law, there is reason for hope that the same effect can hold in states, nearly all of which must balance their budgets.55 Political pressures are lopsided in favor of carceral legislation, but costs and procedures that drive greater attention toward cost can help balance those pressures.56
Even in the Deep South, fiscal concerns have driven decarceral reform in recent years. We thus look to examples from Louisiana and Alabama. So too do we discuss what are perhaps the best-known examples nationally of the widespread use of fiscal constraint to confine sentencing—North Carolina and Minnesota.
In 2017, Louisiana passed comprehensive criminal law reform so that it would no longer have the highest per capita incarceration rate in the United States.57 Prison population growth had been driven by a combination of harsh laws with long sentences and restrictions on parole. Even after release, long supervision terms and escalating caseloads for probation and parole officers created a toxic feedback loop such that by 2015 most of the people admitted to Louisiana prisons were there for violating supervision conditions.58 A package of ten bills raised dollar-amount thresholds on property crimes, reduced sentences for nonviolent crimes, and expanded parole eligibility, among other things.59 Even the Louisiana District Attorneys Association supported the reforms, arguing that they would improve public safety and fairness, while also saving money.60 Some of these reforms were unwound in 2024 despite the costs that abolishing parole and increasing sentences will ultimately carry,61 but fiscal pressure prompting decarceration in Louisiana is noteworthy nonetheless.
In 2015, Alabama’s prisons were operating at 195 percent of capacity, and lawmakers worried that they would need to spend hundreds of millions of dollars to avoid federal courts’ ordering prisoner releases as in California.62 In response, lawmakers reduced sentences for lower-level offenses and expanded community supervision.63 The reforms were projected to save $380 million by 2021 and result in 4,243 fewer people in prison.64 By 2018, the prison population had shrunk by 5,400 people, saving an estimated $170 million.65 Nonetheless, because Alabama did not make these sentencing changes retroactive, the aging prison population offset some of the savings that otherwise would have accrued: Daily prison costs per inmate more than tripled between 2000 and 2022.66 As with Louisiana, we do not read this example to suggest that fiscal pressures always lead lawmakers to decarcerate. To the contrary, in the past five years, Alabama approved $1.3 billion to build two new prisons; when the $1.3 billion proved sufficient to build only one prison, Alabama authorized bond funding for the other.67 Nonetheless, as prisons continue to consume a greater share of the State’s budget in the coming years, fiscally motivated reform may again follow.
In North Carolina, fiscal restraint played a significant role in sentencing reform in the late 1980s and for much of the 1990s68—an extraordinarily carceral time across the country.69 In 1973, North Carolina had the highest per capita incarceration rate in the country.70 Once federal litigation ultimately forced the State to address prison overcrowding, the process began in which the State ultimately took “a fiscal planner’s approach to sentencing.”71 North Carolina’s legislature created a sentencing commission that it tasked with developing categories of crimes and a new sentencing structure, meant in large part to control and manage prison costs.72 The State sought to balance the budget while still trying to transform its public education system—a notable instance of favoring education spending over prisons.73 In 1990, North Carolina’s legislature authorized $75 million in bond funding for prison expansion; it then asked voters to approve another $200 million in bonds for prison expansion.74 Although voters passed that second bond proposal, the narrow margin of passage indicated to legislators that they had reached the extent of voters’ willingness to increase spending on incarceration.75
Legislators’ perception about voters’ willingness to increase prison spending manifested in exchanges between the legislature and sentencing commission. When a subcommittee of the sentencing commission developed a proposal estimated to require $1 billion in new prison construction with an annual operating budget of $1.5 billion, the commission rejected the plan because the price tag was much too high to garner legislative approval.76 Instead the commission sent the legislature a much more modest proposed sentencing structure that would require more than 3,000 new prison beds immediately and 15,000 new beds over the next ten years.77 That new plan still required too much new prison spending, so the legislature tasked the sentencing commission with coming up with a plan that didn’t require prison growth.78 This sentencing reform plan that did not require any new prison bed construction passed the legislature by resounding margins.79 In North Carolina, the desire to staunch the bleeding from corrections spending to protect other spending priorities played a substantial role in revamping its sentencing law.
The long-term effects of North Carolina’s focus on fiscal planning as a way to manage the carceral state are somewhat less clear, but it does seem that at several moments since the mid-1990s the legislature made more modest carceral changes than it otherwise would have because of cost concerns.80 When North Carolina passed a recidivism law, a gun enhancement law, and increased penalties for rape in 1994, the bills that actually passed carried a cost of only 10 percent of the bills as initially proposed.81 And in the 1990s when prison populations swelled, there is reason to think that fiscal concerns slowed prison growth in North Carolina.82 So too does the North Carolina experience demonstrate the need for continued attention to fiscal restraint and continued projections.83 Although the commission projected a fairly small increase in the cost of a new recidivism law, prosecutors increased charging in response to the new recidivism tool, which caused costs to exceed projections.84
Minnesota passed new presumptive sentencing guidelines and capped prisons at 95 percent of their capacity.85 Under that structure, either system actors needed to limit their use of the prisons or the State needed to keep paying to fund more prison construction.86 Although Minnesota built more prisons to increase their capacity somewhat, it also substituted violent offenders for property offenders in existing prisons.87 In other words, it made tradeoffs within criminal law and opted to swallow the bitter pill of spending somewhat more as well. But folding fiscal pressure into the discussion meant that incarceration grew less in Minnesota than elsewhere.
The Model Penal Code’s sentencing provisions are structured such that sentencing commissions will “create guidelines projected to yield sentenced populations that will fit the capacities of existing (or funded) correctional resources.”88 This approach tracks successes from North Carolina and Minnesota.89
Our discussion focuses on the role of cost considerations for state legislative decisions in criminal law, although others have sought to infuse cost considerations to slow the carceral machinery in other ways. Missouri provided its judges with information about the cost of different potential sentences that they might impose, leaving it to the judges to decide whether and how to consider that information.90 (Perhaps unsurprisingly, the primary criticism of the approach was that legislatures, not the judiciary, should account for the role of cost in criminal law.)91 In Philadelphia, District Attorney Larry Krasner required his prosecutors to state on the record the cost of potential sentences in each case92—a different route to a similar structure as Missouri to make cost information available to judges. One scholar proposed providing prison population statistics to line prosecutors to inform their plea-bargaining choices.93 One of us has proposed requiring prosecutors to disclose cost information about their cases to better inform voters.94 Although these mechanisms vary in their implementation, all seek to make criminal law more cost sensitive.
II. Diffusion of Costs and Misalignment of Costs and Benefits
One of the complicating factors well-known to criminal law scholars is how criminal legal systems diffuse costs across different levels of government—state, county, and city.95 That ability of legislators to diffuse costs and foist them onto other levels of government, or even onto defendants, complicates the administrability of deploying fiscal constraints to restrain state criminal lawmaking in ways that will require careful attention below.
As a rough approximation, states pay for prisons; prosecutors and jails are largely funded by counties; and policing occurs at the city level.96 Indigent defense sometimes gets funded at the state level and sometimes at the county level.97 In some instances, states pay for some of the costs of prosecution such as line prosecutor salaries, and grant funding from the federal or state government sometimes offsets some of the costs to local government.98 In short, the point that the costs spread across multiple levels of government remains under any of these structures.
Moreover, actors across the system make decisions in which other agencies or even other levels of government bear the costs.99 That’s true for state legislators, prosecutors, and police. The most well-known misalignment of costs and decision-making authority is the “correctional free lunch” where counties pay for prosecution but states pay for prisons.100 When prosecutors seek prison time, they’re attempting to spend the state’s money rather than their own.101 When prosecutors seek only jail time, they’re spending their county’s money, although they’re likely spending the sheriff’s office budget rather than their own.102
California significantly realigned these incentives after the Supreme Court ordered the State to release a substantial number of incarcerated people to alleviate terrible prison overcrowding.103 The State’s response to the Court was to place individuals convicted of their first felony for lower-level offenses under county rather than state supervision and to shift post-incarceration supervision responsibility from state parole to county probation departments.104 While the State provided funding to counties for this change, the legislature initially provided only a fraction of the funding necessary to jail each incarcerated person, shifting a significant portion of post-conviction costs onto local governments.105 Counties were free, under California’s new system, to “choose from a wide variety of . . . alternatives” to incarceration.106 By shifting financial responsibility down to the local level, where important criminal law choices are made, and by passing other decarceral reforms around the same time period, California reduced its incarceration rate.107 Although decarceration from a place of mass incarceration is a success story, California’s decarceration has been racially uneven, with the incarceration rate for White adults decreasing more than that for Black or “Latinx”108 adults.109 So too has decarceration varied widely from county to county.110
California similarly realigned costs in juvenile criminal law, significantly increasing the extent to which counties bore the financial burden of their juveniles being housed in state facilities.111 Realigning incentives led to a huge decrease in imprisoned children without an increase in crime.112
In recent decades, state lawmakers have increasingly turned to fines and fees, ostensibly hoping to shift costs of criminal law from the public fisc onto defendants.113 But these fines and fees are a microcosm of the diffusion of costs and revenues. For instance, in Alabama, revenue paid toward fines is typically destined for the State General Fund, while revenue collected pursuant to fees is split among scores of specialized pots of money under the jurisdiction of counties, clerks, district attorneys, state agencies, and even a nonprofit theme park that employs costumed historical interpreters to teach schoolchildren about the Revolutionary War.114 A typical conviction for felony drug possession in Alabama might result in more than $1,000 in fines and fees.115 If the person who owes that money cannot pay all at once, judges often allow them to pay in installments of $25 or $50 per month.116 If ninety days pass and the person fails to make a payment, a “collections” fee of 30 percent will be assessed on top of the original debt, increasing the debt balance to $1,300.117 Once the collections fee is assessed, any money the person pays is kept by the District Attorney’s office (75 percent) and the clerk (25 percent) until the full $300 collections fee is collected.118 Revenue collected after that is disbursed according to a default priority system that places the patriotic theme park ahead of the State General Fund, which is itself ahead of the Crime Victim’s Compensation Fund. Victims who are owed restitution come last in the default priority schema.119
By shifting to fines and fees instead of general fund revenue to cover some of the costs of criminal law, state legislatures sometimes avoid the hard questions of whether criminal enforcement is worth taxpayers’ money—perhaps intentionally.120 Greater use of fines and fees also implicitly pressures prosecutors to increase enforcement,121 at least if the prosecutors’ office wants adequate funding.122 In 2021, court fines and fees comprised 15 percent of local DA funding.123 Because some of these fees accrue only upon conviction, this structure amounts to partially paying prosecutors on contingency124—an approach in substantial tension with prosecutors’ duty to serve as ministers of justice.125
III. Legislative Procedures to Constrain Fiscal Excess
In other contexts, states use a variety of tools to fiscally constrain or restrain legislation. These tools vary from hard constraints such as caps on the total spending increase in the budget each year to softer tools meant to increase the salience of costs such as requiring fiscal notes that estimate the cost of legislation before it passes.126 Because of the way that criminal law diffuses costs across state, county, and even city governments,127 tools meant to constrain the legislature from burdening local governments or increase the salience of such burdens are especially worth exploring here.
For the most part, states don’t use the tools in criminal law that constrain fiscal excess in other legislative contexts. There are, however, some noteworthy counterexamples, such as states that require the legislature to estimate the increased cost of incarceration that a new criminal law would yield before passing it. A statute that Colorado recently repealed required pay-as-you-go funding for expansions of criminal law—preventing new criminal legislation unless the legislature paid for the costs of additional imprisonment that it would generate.128 This Part considers the limited extent to which states currently employ these tools of fiscal restraint in criminal law and how these tools may offer greater promise of constraining mass incarceration through fiscal cross-pressuring.
Ultimately, our point is to urge legislators to consider the costs of carceral criminal legislation that otherwise benefits from significant political headwinds.129 Much like Darryl Brown’s proposal for cost-benefit analysis in criminal law,130 this Article seeks to encourage active deliberation about the costs and benefits of criminal law. But unlike Brown’s proposal, this Article looks at ways to structure legislative deliberation about the scope of criminal law and its penalties to weigh its costs against its benefits rather than focusing on cost-benefit tradeoffs for police and prosecutor deliberation in individual cases.131
Even if we improve decision-making processes, legislators will still make some combination of policy and political judgments.132 Legislative tools of fiscal restraint certainly do not (and do not purport to) divorce politics from lawmaking. And no decision-making structure could force legislators to balance concerns in just the way we would wish.
We assume throughout the Article that state legislatures’ policy goal is public safety.133 Nonetheless, the policy questions are complicated because criminal law is not synonymous with public safety, and harsher criminal law will not always increase public safety.134 Criminal law can advance public safety.135 But criminal law also detracts from public safety. Policing hurts and kills people.136 Incarcerating people in incredibly dangerous jails and prisons exposes them to a profound lack of public safety—assaults, sexual assaults, an environment of intense fear and insecurity, and all too often, death.137 These instances of violence and death don’t go away just because they happen behind jail or prison walls.138 And unless we’re talking about people who will die in prison, those safety harms that people suffer while incarcerated may make them far more dangerous to people outside of carceral facilities once they are released.139
These procedural tools are meant to pose tradeoffs between spending in one area versus another or between spending and taxes. They thus can be navigated by cutting important social safety net programs to fund criminal law, but that tradeoff will often be unwise. Dollars spent on early intervention for at-risk children140 and other programs that support community organizations will likely do far more for public safety than criminal law does, especially in its current state of vast overuse.141
Similarly, financing criminal legal systems in part through financial levies like fines and fees rather than having to pay for them from the public fisc might seem like a way to get more criminal law for less, but fines and fees are deeply counterproductive. Fines and fees collect little net revenue,142 and they seem to increase crime.143
Although a balance sheet showing “accounts payable” on fines and fees nationally would suggest that state court systems alone could collect billions, the reality is far different.144 One study found that counties in Texas and New Mexico were spending forty-one cents for every dollar raised through fines and fees, with one county spending $1.17 to collect each $1.00 in revenue.145 By contrast, the Internal Revenue Service spends far less than one cent to collect each dollar of tax revenue.146 One of the most expensive (and perverse) means of pursuing revenue is jailing debtors,147 which is often done without a serious review of their ability to pay and costs money rather than generating it.148 In this way, fees, which are often pitched as a way of offsetting the costs of criminal law, can increase governmental expenses. On top of that, even short periods of detention can destabilize a person’s life, resulting in job loss and increased likelihood of rearrest—counterproductive outcomes that, among other things, increase costs for the state.149 Criminal defendants are overwhelmingly poor,150 so it ought not be surprising that with remarkable consistency across the country, people who owed debt they could not pay explained to researchers, “You can’t squeeze blood from a turnip.”151
Those facing criminal justice debt also sometimes feel pressed to commit crimes like selling drugs, stealing, and engaging in sex work—ways of earning money to satisfy their debt.152 Higher cumulative debt from fines and fees, perhaps unsurprisingly, increases recidivism.153 It thus makes little sense for legislatures to avoid paying the costs of the criminal legal system by trying to shift them onto defendants instead.
Several of the tools we discuss require the legislature to have information about the anticipated cost of a change in criminal law—to the state, to local government, or to both. The quality of the information is important.154 Sentencing commissions are a far better source of data than are interest groups or legislators.155 In several states, sentencing commissions have been providing cost information, and that information has helped shape criminal reform.156 Moreover, to address the inherent challenges in predicting costs and benefits of a law ex ante, we suggest that sunset provisions can help facilitate consideration of the costs and benefits of a newly enacted criminal law in a way informed by several years’ worth of experience.
With these aims of structuring legislative decision-making to better consider costs and thus facilitate deliberation in criminal law, we turn to the particular mechanisms—opportunities they offer and challenges they pose. Starting with tools that provide the least flexibility to legislatures and moving toward those that provide the most, this section considers the role that total spending caps, mandatory budget offsets, and fiscal notes could each play in criminal lawmaking. So too do we address unfunded mandates, although we think that tool inapt in this context. We then consider how sunset provisions can interact with the challenges of predicting costs and benefits. In each section below, we begin by considering the nature of each tool and then consider how it could (and occasionally does) apply in criminal lawmaking.
A. Total Spending Caps
A substantial number of states have overall limits on the extent to which their taxes and expenditures can grow from one year to the next, known as TELs, although some of these can be overridden by a supermajority vote of the legislature.157 Studies conflict as to whether TELs effectively reduce state spending growth.158 The more successful TELs tend to allow a certain measure of growth from year to year based on growth in the state’s population and inflation.159 Some TELs adjust to permit less state spending if the state devolves responsibility for some functions to other levels of government.160 TELs work best when they measure whether the budget actually balanced at the end of the year rather than whether the budget is projected to balance ex ante.161
Limiting growth in spending and taxation through a TEL is a tool that more states could adopt to restrain carceral growth and perhaps even help decarcerate, although it is quite a blunt instrument. Subjecting new criminal laws to a TEL poses tradeoffs between more criminal law or other ways to spend a state’s limited money. Maybe that’s good.162 Unlike with some other tools, TELs do not afford legislatures the ability to swallow hard and raise taxes to pay for more criminal law without cutting spending elsewhere. Although if they feel too constrained, some legislatures can override the TEL with a majority or supermajority vote.
TELs limit new spending and force spending-spending tradeoffs. But TELs are a rigid tool that may inhibit expensive but cost-justified changes—a perverse outcome if the ultimate goal is public safety rather than simply constraining spending. Legislators might cut social safety net programs to pay for more mass incarceration, which can further a vicious spiral.163 Cutting programs that provide food assistance or medical care could be deeply counterproductive to public safety.
To the extent that states apply TELs to criminal law, one design feature of a TEL that would be especially important for criminal law is adjusting the limit when financial responsibility shifts to another level of government. Costs of criminal law are already diffused across state, county, and city governments. Thus states can and sometimes do shift financial responsibility over aspects of criminal law to local governments, such as California requiring counties to pay for a portion of the prison beds their cases use.164 States may also devolve costs onto counties by paying for less of their prosecutors’ office budget or offloading costs like indigent defense to the local level.165 States may even devolve costs by pushing more sentences into jails (paid for by the county) rather than prisons (paid for by the state).166 A well-structured TEL to address criminal law (and other areas) would thus reduce the state’s spending and taxation limit if it offloaded some of its financial responsibilities onto local governments.167
B. Budgetary Offset Requirements
Congress navigates an array of procedures and budget rules, including budgetary offset provisions,168 although the applicable statutes and rules change over time. The most well-known of these offsets is “pay-as-you-go” or PAYGO.169 Although the particulars of PAYGO and other offsets are quite detailed, a brief overview will suffice for present purposes. PAYGO was designed to avoid increasing the federal deficit.170 To that end, it required Congress to pay for or offset within the same budget year new expenditures that would be incurred in the five years following the legislation.171 Paying for the legislation means raising taxes, reducing tax subsidies, or reducing spending for existing entitlement programs.172 Other offsets require different pay-for time periods; for some types of legislation, Congress can shift expenditures outside the relevant time window and thus evade the offset.173 Because raising taxes is so politically unpopular, as a practical matter, legislators and lobbyists often look for offsets elsewhere.174 When looking for discretionary spending offsets, cuts to “provision[s] that benefit[] the relatively needy, rather than an expenditure that benefits organized and wealthy economic interests” are politically desirable.175 Although that insight is hardly surprising as a matter of public choice, it is nonetheless important when thinking about how offsets can work in criminal law.176
Unlike the PAYGO federal statute that required offset within the budget year, a Senate pay-as-you-go rule required that each individual bill pay for any expenditure increases.177 The 1986 Tax Reform Act required a subject-specific offset, requiring savings for a tax cut to be found in the tax code.178 Offsets like PAYGO force choices about competing priorities and interests.179 Unlike TELs, offsets afford legislatures the (unpopular) opportunity to raise taxes to offset new spending. For all the criticism PAYGO has received, including criticism about entrenching the status quo and distorting substantive objectives, PAYGO has helped restrain new spending.180 Of course, requiring the legislature to pay for proposed legislation can work only as well as the quality of the information about the anticipated costs of the bill.181
In criminal law, a few states require (or have required) themselves to pay at the time of passage for part of the changes that will later increase incarceration. Virginia requires bills expanding criminal law to identify a source of revenue to pay for that expansion.182 The State’s sentencing commission provides a fiscal impact statement for a carceral criminal bill, and a finance subcommittee must determine whether there is funding in the bill to support that carceral expansion or else the bill dies without reaching the floor.183 But Virginia (troublingly) excludes the capital costs of building new prisons—a potentially huge expense.184 Virginia’s offset law encompasses not only changes to sentencing law but also changes to substantive criminal laws that will likely increase incarceration.185 An early analysis found that Virginia’s incarceration rate grew well below the national average with these offset rules in place,186 and Virginia did not enact any new mandatory minimums for at least the first three years under this procedure.187 The Virginia experience was far from perfect, however. For the first several years of this law’s existence, many of Virginia’s fiscal impact statements simply said that there was insufficient information to make a projection—a problem that Virginia remedied by setting $50,000 as the default when no other estimate could be provided.188 But Virginia has remained at least somewhat committed to fiscal restraint in criminal lawmaking: Its sentencing commission continues to hone its cost estimation models in ways that may prove worth emulating elsewhere.189
Tennessee similarly requires criminal law bills to also include appropriations to cover any increased costs of imprisonment in state prisons that are estimated in the three years following enactment of a new law.190 Perhaps because of Tennessee’s structure requiring pay-fors, as a practical matter, any legislation including criminal laws with an estimated cost above a certain threshold will not reach the floor of the legislature.191 Unlike in Virginia, where the sentencing commission produces the cost estimate, in Tennessee estimates come from the fiscal review committee in the legislature.192 That lack of a formally independent source is less good. Tennessee’s fiscal restraint procedure comes with a huge hole like Virginia’s: It reaches only the variable costs of housing people in state prisons and not any additional capital costs necessary to build new facilities or maintain existing ones.193 Interestingly, in Tennessee, laws that decrease incarceration in state prisons also require a decrease in the appropriation for the department of corrections.194
Until 2022, Colorado required budgetary offsets for laws that would increase costs of imprisonment.195 Any bill that would yield a “net increase in periods of imprisonment in state correctional facilities” required an appropriation in the same bill to pay for that increase for five years after the bill’s enactment.196 In 2022, however, Colorado paused the law for three years,197 and before the three-year pause expired, the legislature repealed the law.198
For a few years in the early 1990s, Nebraska had a similar law.199 Unlike the other states discussed where the legislature had to find a pay-for in the same bill, Nebraska’s legislature required itself to appropriate sufficient funds during the same legislative session to cover the estimated “per diem and medical expenses” of legislation projected to increase the adult or juvenile prison population.200 The legislation also required that the money appropriated for estimated increased prison expenditures sit in a contingency fund to be used only to pay for increased incarceration costs.201 Puzzlingly, the Nebraska Supreme Court held that this statute violated the State’s constitution because it impermissibly sought to bind a future legislature.202
Building on these models, W.C. Bunting proposed that states require themselves to appropriate money for one year in the current fiscal period to offset expected fiscal impacts of proposed changes in sentencing policy.203 The goal is to urge current legislators to face costs that would otherwise be deferred until after someone was convicted and serving a prison sentence under the new law.204
For carceral legislation that proceeds by initiative or referendum directly to the voters, the legislation can and should be paired with a bond measure that will fund the increased costs.205 The two should rise or fall together.
Instead of requiring that states pay financially for increases to the prison population, James Ganas urges states to adopt a sort of PAYGO for sentencing reform such that legislation can yield no net increase in incarceration.206 In other words, Ganas argues that states should require that any legislation that increases criminal sentences in one area of the criminal code be offset by a decrease in sentencing in another area of the code, measured in time of incarceration rather than dollars.207 The idea is to treat criminal sentencing as a closed system and require the legislature to make tradeoffs that do not increase incarceration on net.208 Ganas’s approach parallels the 1986 Tax Reform Act, in a way, because, unlike with PAYGO, the offset must come from within the same subject area.209
Constraining prison capacity would and has worked similarly to mandatory offsets, especially as it allows legislatures to increase incarceration so long as they are willing to spend more money to build more prisons.210 In Minnesota, capping prisons at 95 percent of their capacity meant that either system actors needed to limit their use of the prisons or the State needed to keep paying to fund more prison construction.211 Minnesota did some of both.212 It built more prisons to increase capacity somewhat, but it also substituted violent offenders for property offenders in existing prisons.213 Similarly, North Carolina has reduced the growth of its criminal legal system through sustained attention to cost in interactions between the legislature and the sentencing commission—looking for offsets that don’t require prison growth but permitting those that do.214
Unlike fiscal notes discussed below, mandatory budget offset requirements like PAYGO seek to force legislatures to pay for their cost choices rather than simply forcing them to listen to how much their bill will cost. As such, mandatory offsets could be a potent tool to fiscally constrain the expansion of criminal law even in the face of political pressures toward expansion.
As the examples above illustrate, mandatory offsets take different forms that pose different choices for legislatures. Some offsets require that the new criminal law bill itself include the necessary appropriation, while others allow the legislature to appropriate adequate funds later during the budget year. Some allow legislatures to cut spending unrelated to criminal law to offset new criminal law spending. Some proposals refuse to allow any increase in incarceration even if the legislature were willing to take the politically difficult approach of raising taxes. We offer our thoughts on each tradeoff.
In our view, legislatures ought not be afforded the opportunity to find the money later in the budget year. Rather, as in Virginia and Tennessee, each bill that is likely to increase imprisonment ought to fund that increase. Our goal is to slow the tailwinds from which carceral legislation benefits by promoting greater legislative deliberation. Allowing pay-fors to come later skirts that objective.215
By contrast, we think that legislatures should be able to cut spending unrelated to criminal law to fund additional criminal law spending, although we recognize that this design choice poses real concerns when we combine political power dynamics and what we know about public safety. As discussed above,216 allowing legislatures to cut other spending such as spending on education or social welfare programs to offset increased costs of criminal law could be quite counter-productive if the goal is to promote public safety.217 But again we think structuring the legislative process to facilitate deliberation is all we can accomplish. We can’t save legislators from their own unwise choices.
So too do we think legislatures ought to be able to raise taxes to pay for harsher criminal law.218 Ganas is right that allowing legislatures to spend more money to build more prisons might mean we get more prisons and more incarceration, as in Minnesota. But pitting the desire for fiscal restraint in its most direct form of a tax increase against carceral sentiment achieves our goals.219 Getting more mass incarceration than we’re willing to pay for is deeply troubling.220 But if voters are aware of and willing to bear a greater personal financial burden to be more carceral, then that outcome seems like one that a democratic system that leaves criminal justice policy to the political branches must tolerate. In part our view is driven by the idea that any rule that outright constrains increased incarceration without pressure release valves would not last or would be regularly evaded.221
Capital expenditures for prison construction or repair should not be excluded from budget offsets in the way that Virginia and Tennessee do.222 Rather, states ought not be sending more people to prison than existing capacity affords or sending people to crumbling prisons unless they are willing to pay to build new prisons or refurbish the crumbling ones.223 Ignoring prison capital expenditures when debating carceral legislation allows the legislature to ignore huge costs when purporting to decide whether a new law is more beneficial than costly. When legislatures don’t approve the necessary spending up front for the laws they pass, they allow themselves to balance their budgets later by ignoring necessary prison construction or remodeling,224 except in the extraordinarily unlikely event that a court orders the state to do otherwise.225 By contrast, the Minnesota approach, using caps on imprisonment that yielded some prison construction and some shift to incarcerating people for different sorts of offenses, makes more sense.
That Colorado repealed its budgetary offset law for incarceration and reverted instead to the idea that it simply must figure out how to pay for its current prison population each year suggests that the budgetary offsets mattered. When Colorado paused that law in 2022, the bill’s fiscal note said that it would reduce appropriations to the Department of Corrections by funding only the actual prison population rather than the population change predicted five years in advance.226 If initial estimates had simply overshot the mark, legislatures could reduce appropriations later. But we suspect that a different tradeoff between ex ante five-year estimates and ex post payment for prison costs is afoot. Ex ante estimates involve taking the costs of housing inmates, considering the existing capacity of prisons, and multiplying by an estimated number of new incarcerated people. By contrast, simply finding ways to make the prison budget work each year gives the legislature ways to cut corners. It may simply demand that the prisons do more without more money. Pack more people into the same decrepit building. Feed them for even less. Compel inmates to work for free instead of paying someone to do the work.227 That ex post approach potentially yields overcrowded (and underfunded) prisons.228 In other words, we suspect that legislatures may use the “make the budget work” approach to shift more costs onto incarcerated people that the state ought to bear.
There are also questions about which bills these procedures would reach—only bills that change sentences for existing crimes, or laws that create new crimes too? Although the cost of changes to sentences for existing crimes will be easier to calculate than the cost of laws creating new crimes because sentencing commissions will have access to existing enforcement numbers for the former,229 leaving substantive criminal law aside would be too large of a loophole. Calculating cost would just require some more estimating conventions relative to laws that are already being enforced and would especially benefit from revisiting a few years later based on the actual data.230
Budgetary offsets seem to hold the most promise for bringing fiscal discipline to bear on criminal legislative excess if they are structured to require the legislature to find in the same bill an offset for any new criminal law spending either by cutting other spending or increasing taxes. Legislatures should have to appropriate money for a non-trivial period such as the impending five years (as Colorado did). Criminal laws won’t yield immediate costs, so setting too short of a period allows legislative machinations.
C. Fiscal Notes
The overwhelming majority of states require fiscal notes that estimate what a bill will cost for all types of legislation before the bill can pass;231 in some of those states the requirement applies only to bills exceeding a certain dollar amount.232 A few other states require fiscal notes to estimate the cost of administrative rules or for only specific types of legislation.233 Fiscal note requirements also apply in many states to direct democracy measures that go straight to voters—initiatives or referenda.234 Many states require that the costs to local government be identified in advance (also typically called fiscal notes but sometimes called local notes).235
The goal of fiscal notes in the legislative process is straightforward: estimate the cost of a bill such that legislators can consider whether its benefits exceed its costs.236 In so doing, fiscal notes provide data and increase the salience of cost in legislative deliberation.237
Fiscal notes are often prepared by a legislative agency akin to the Congressional Budget Office.238 And there is good reason to prefer fiscal notes prepared by such neutral entities rather than notes prepared by a legislative member or their staff.239
Although fiscal note requirements are widespread, they are not as effective in practice as we would wish.240 Some fiscal notes do not estimate costs to be imposed onto local governments,241 sometimes writing instead “data not available.”242 Sometimes fiscal notes simply aren’t prepared even though they ought to be by rule or statute.243
At least as a matter of positive law, many states require fiscal notes for sentencing laws or changes to substantive criminal law.244 Some states mention substantive criminal law specifically in their fiscal note statute or rules,245 while some other states’ codes reach any laws that would affect prison expenditures.246 Many other states simply have broadly worded fiscal note requirements that do not mention criminal law or sentencing law, but nor do they exempt such laws from their reach.247 Some of the statutes are usefully explicit about what sorts of costs of criminal law the relevant entity must estimate and, in some especially useful instances, include costs imposed on local governments.248
To drill down on one example, in North Carolina, when a criminal law bill is considered, the legislature must have before it a five-year cost estimate, including estimates of any necessary capital expenditures to account for increases in imprisonment.249 The legislature’s Fiscal Research Division consults with the state sentencing commission to provide these estimates.250 North Carolina adopted the fiscal note requirement in 1993—a time when the legislature was especially concerned about reining in the fiscal effect of its carceral state to allow it to promote other legislative priorities.251 When one prominent Republican legislator tried to repeal the fiscal note requirement because he thought it prompted legislators to overlook the long-term benefits of crime reduction, other prominent Republican legislators convinced him that fiscal notes provided useful fiscal discipline and were far better than when “[w]e just guessed” at the cost.252
Because state criminal legislation spreads costs across various levels of government,253 there is much to recommend fiscal note requirements like Oregon’s that explicitly consider costs foisted onto county or local government and not just costs such as prisons that the State bears directly.254 Costs to local government include costs to law enforcement agencies, local courts, prosecutors’ offices, public defenders, local jails, and parole and probation supervision.255 Although state legislators have stronger self-interested motivations to constrain the state’s spending than they do local governments’ spending, they may nonetheless benefit from seeing the full picture and may face public pressure if county and city budgets suffer as a result of state legislative action. Other states have broad provisions that require the legislature to estimate any costs that any bill would impose on local governments, although these provisions do not explicitly mention criminal laws.256
Even statutes that do not require calculating costs foisted upon local governments can usefully detail the costs that a fiscal note must estimate. Missouri, for instance, requires a prison impact statement explaining how the law is projected to affect prison, probation, and parole populations, as well as the estimate of prison construction and operation costs for up to ten years.257
Colorado requires fiscal notes for proposed changes in substantive criminal law to analyze how the bill would overlap with existing provisions of the State’s criminal code258—a sensible approach when the political economy of criminal law tends to yield deep codes.259 That analysis can help promote deliberation on the value of the new law and take a more realistic look at projected enforcement cost changes.
Many fiscal note requirements require calculation only of increased expenditures or incarceration,260 while others require calculation of savings too.261 In our view, legislatures’ aversion to increasing spending may outweigh their desire for cost savings, although both matter. We thus think that fiscal notes estimating savings are also important to give legislators a specific savings number to publicly tout, but estimating savings is less important than fiscal notes estimating increased expenditures.262
Unfortunately, the statutes tell a different tale from the realities on the ground. In many places fiscal notes for new criminal laws or changes to criminal law simply don’t get calculated.263 Looking nationally, one study found that states wrote fiscal notes for only 40 percent of the bills identified as “significant state sentencing and corrections legislation” during their 2009, 2010, and 2011 legislative sessions.264 New Jersey did not produce fiscal notes for any criminal law bills during the study period.265 Although Delaware law requires fiscal notes for all bills, without exempting criminal law,266 one study found that Delaware typically does not write fiscal notes for criminal justice bills.267 Georgia, Illinois, and Mississippi produce fiscal notes when legislators request them, but during the three-year period studied, apparently no one requested a fiscal note for a criminal law bill in any of these three states.268 South Dakota and Vermont rarely produce fiscal notes for any legislation.269
Other states, including Texas and Tennessee, create fiscal notes for criminal laws more frequently.270 And when the estimated costs in the note are high, the estimated costs can indeed kill the bill.271
The lack of adherence to existing statutes in some states is troubling. One option to increase legislatures’ compliance is to dismiss charges if the statute(s) under which a defendant is charged did not have an adequate fiscal note. We recognize, however, that that approach would have significant drawbacks, including facilitating public perception of defendants “getting off” on technicalities, much like the concerns surrounding the Fourth Amendment’s exclusionary rule.272
Narrowing our focus to the few states that actually wrote fiscal notes for criminal laws, a study by Michael Leachman, Inimai Chettiar, and Benjamin Geare found that the fiscal notes looked only at the financial impact for two or fewer years after passage; many simply indicated whether the anticipated budget effect would be positive or negative without calculating its magnitude.273 Short time periods are problematic in criminal law because the state won’t bear costs until someone goes to prison after a felony conviction. And hard numbers are important too.
Sentencing commissions can be quite valuable in generating high-quality cost estimates. So too can legislative analysts produce useful cost estimates if they’re sufficiently independent from the legislators pushing the bill.274 Sentencing commissions have depth of expertise that legislators likely lack to calculate estimated costs of new legislation and to consider ways to adjust the state’s sentencing law to offset new costs275—as North Carolina did to propose a budget-neutral approach to increase sentences for serious violent crime. Legislatures can then better evaluate the overall impact of a criminal law proposal on the state’s budget when they receive the cost data than can the more narrowly focused sentencing commissions, and legislatures can better appreciate the tradeoffs of spending on criminal law versus other important goals such as health or education—as North Carolina’s legislature did. And, of course, legislators are more politically accountable than are sentencing commissioners.
Much like Colorado, states should exclude from their calculation revenue generated from fines and fees when determining the estimated cost of a change to criminal law.276 That is because in addition to the real risk that they will detract from public safety, fines and fees likely will not yield much money, especially when collection costs are netted out.277
Cost estimation for bills is not always easy, nor will it be here.278 But just because the work is hard and somewhat imprecise is not a good reason to avoid hard choices. Calculating the estimates may cost money. But the carceral state is extremely expensive for the public fisc and imposes severe costs on the citizenry beyond the fiscal. So, spending a bit more on legislative process to promote legislative deliberation before reflexively expanding its reach in ways that likely increase human costs seems a feature rather than a bug.279
In urging increased focus on costs, we hope legislators will also better consider the benefits of proposed new criminal laws. Colorado helpfully requires its fiscal notes for changes to substantive criminal laws to analyze whether the proposed crime could already be charged under current law, how the proposed crime compares to similar types of offenses, and the prevalence of the conduct to be criminalized.280 Because many new crime statutes criminalize conduct that is already a crime,281 this approach to evaluating benefits makes sense.282
D. Prohibiting Unfunded Mandates
Whether by state constitution or statute, nineteen states prohibit their legislatures from requiring local or county governments to spend money unless the legislature appropriates some or all of that money.283 These constitutional and statutory prohibitions on unfunded mandates typically reach only laws that “mandate” or “require” local governments to spend money,284 although a few states use slightly gentler language that seems to require state funding for an even broader sweep of actions that implicate local governments.285
Some of these states’ unfunded mandate prohibitions apply only when there is a tight connection between the legislation and expenditure of local funds, although other states’ prohibitions regulate laws that have a looser connection. For an example of a tight nexus, Massachusetts prohibits state legislation from “imposing any direct service or cost obligation” on its local governments without state funding or local approval.286 New Jersey’s unfunded mandate prohibition similarly reaches only “additional direct expenditures required for the implementation of the law.”287 Montana has similar narrowing language about direct expenditures and narrows itself further to reach only local government activities or services “that [are] not expected of local governments in the scope of their usual operations.”288 Most of the unfunded mandate laws are not drawn so narrowly, though, and encompass any “new activity or service or an increase in the level of any activity or service beyond that required by existing law”289 or any law “requiring [local government] to spend funds or to take an action requiring the expenditure of funds.”290 A few provisions like Alabama’s sweep even more broadly, reaching any law “whose purpose or effect is to require a new or increased expenditure of funds.”291
Prohibitions on unfunded mandates don’t make sense to apply to substantive criminal law or changes to sentencing law. Substantive criminal laws are not mandates on local government—they’re tools provided to local governments to use as they see fit subject to their budget constraints. They may end up increasing the cost to local governments, but they don’t mandate such expenditures.292 City and county officials could increase enforcement by adding more policing of a new crime, but so too could they shift resources from one crime to another without increasing net expenditure. County officials face similar choices about prosecution. County jails would incur increased expenses if more people were detained pretrial or sentenced to short sentences, but those increases would follow from local enforcement choices. Moreover, new substantive criminal laws could decrease enforcement costs by making it easier to demonstrate probable cause and increasing charge bargaining leverage. For similar reasons, changes to sentencing laws also do not function as mandates. Local governments can simply decide not to enforce certain laws if they are trying to avoid triggering increased jail expenditures.293
Of the nineteen states that prohibit unfunded mandates (at least without a supermajority vote),294 seven explicitly exempt new criminal laws or changes to criminal laws from that prohibition.295 Most of these carveouts for criminal law reach statutes that create new crimes or that change definitions of existing crimes.296 A few simply list “criminal laws” as an area to which the unfunded mandate rule does not apply.297 The broadest exception includes not only changes to substantive criminal law but also sentencing.298
Although eleven states prohibit unfunded mandates on local government that do not explicitly exempt substantive criminal law, what sparse evidence exists provides reason to think that most of those states too don’t consider criminal laws as mandates.299 Many of these states nonetheless construe their unfunded mandate statutes narrowly.300 Although few unfunded mandate cases arise in the criminal law context, Tennessee upheld a mandatory minimum jail sentencing law against an unfunded mandate challenge.301 In Hawaii and Montana, unfunded mandate provisions do not reach expenditures that are long-standing county responsibilities,302 as criminal law is.303
Moreover, the point about local enforcement discretion rather than mandates seems all the more persuasive in states that limit the reach of their unfunded mandate provisions to laws that impose “direct expenditure[s]” or “direct service or cost obligation[s].”304 Even though one could muster an argument that new criminal laws operate as mandates in some fashion, rather than as new tools for enforcement discretion, it is hard to see how they could satisfy the directness criterion.
In sum, while we tend to be skeptical of treating criminal laws differently than other legislation as a matter of process, unfunded mandate provisions are rightly thought not to apply to criminal laws; criminal laws aren’t mandates. Nonetheless, that these states treat new criminal law as exempt from unfunded mandate prohibitions sheds light on debates surrounding prosecutorial nonenforcement policies as a matter of state positive law.305
E. Sunset Provisions
One of the ways that legislatures restrain themselves is by enacting laws that are meant to be temporary, using so-called “sunset” provisions that set a law to expire if it is not renewed. This technique is not exactly a financial restraint, but it helps support the financial constraints discussed above. When the legislature comes back to debate reauthorization of a law before it sunsets, the legislature can benefit from greater information about the actual costs and benefits of the law during its effective period rather than having to rely solely on its ex ante predictions.306
Sunset provisions have a long history in the United States, including in the Founding era.307 They have “been used extensively by both federal and state legislatures” in a wide variety of substantive areas.308 In the Federalist Papers, Hamilton explained that sunset provisions would encourage future deliberation about the topic and provide another opportunity to thwart unwise policy.309
Although legislation with a sunset provision may seem temporary while legislation without a sunset seems permanent, the real difference lives in the default rule for continuation of the policy.310 “Temporary legislation can be extended and permanent legislation can be repealed.”311 Informational benefits of sunset provisions accrue only if the legislature considers the new evidence and deliberates over it; simply reauthorizing with “little or no deliberation” doesn’t help.312 Especially relevant for considering criminal law, temporary legislation allows legislators to respond to a crisis and strike when the iron is hot but also encourages them to revisit their choices several years on when feelings have cooled.313 Scholars diverge as to whether, as a public choice matter, temporary legislation allows legislators to extract greater rents.314
One scholar proposed a constitutional amendment making all state and federal criminal laws sunset,315 while others proposed building sunset provisions into substantive criminal laws or changes to sentencing law,316 and there are at least a few examples of states doing just that. After the Supreme Court declared California’s sentencing scheme unconstitutional in the wake of Booker and Blakely, California responded with “a ‘stopgap measure’” to expand judicial sentencing discretion that included a sunset provision and would give the state time to evaluate more permanent changes.317 It did so amidst significant debate about how much the shift to advisory sentencing would cost in increased prison expenditures.318 California has since changed its approach to resolving its Blakely problem by requiring either defendant stipulation or proof beyond a reasonable doubt to enhance a sentence rather than increasing discretion.319 California also experimented with an increase in sentencing for vehicle theft—a law that it passed with a sunset provision that expired three years later.320
Although these examples are usefully illustrative, it is striking how rarely we pause to assess the effectiveness of criminal laws after their passage. Instead, we simply hope things are working out as anticipated. It is hard to imagine that approach in any other area of law.
Whether grounded in concerns about undue cost, on which this Article has focused, or on other values like aligning the criminal code with current moral values,321 the idea of sunsetting one or all criminal laws is meant to address the fact that politics typically makes it easier to pass criminal laws than to repeal them.322 For politicians, voting to repeal a criminal law might be mistakenly equated with condoning the behavior it criminalized.323 Letting legislation expire silently by not renewing it is subtler.324 Sunset provisions are meant to prevent keeping bad laws and “severe sentences on cruise control.”325
The desirability of a sunset provision depends on an estimated comparison between current and future political dynamics; that comparison makes sunsets especially desirable in criminal law.326 Some criminal law is passed in the immediate wake of a high-profile tragedy.327 Perhaps most famously, Polly Klaas’s abduction and murder in 1993 (and the murder of Kimber Reynolds the previous year) is typically credited with inspiring California lawmakers to pass an extremely harsh three-strikes law in 1994.328 The perpetrators in both cases were out on parole at the times of the killings.329 After Carlee Russell falsely reported that she saw a toddler wandering alone near a freeway, Alabama police mustered substantial effort trying to find the (non-existent) toddler.330 After public outcry that Ms. Russell was convicted only of two misdemeanors, Alabama added a felony for false reporting.331 North Carolina lawmakers just passed reforms limiting cashless bail and imposing increased conditions on pretrial detention in a bill titled “Iryna’s Law”—named for a young woman who was fatally stabbed in Charlotte about one month before the vote.332 In the immediate aftermath of a tragedy, the political potency of reflexive carceral legislation may be irresistible. But when passions have cooled and the high-profile crime fades further into the distance, the politics may be less heated and may allow for more thoughtful deliberation.333 We think that every law named for a particular victim should sunset. Perhaps the laws will have been successful, or perhaps prominent victims’ legacies will deeply entrench the laws even if they’re bad; but the politics will likely be less emotionally heated years after a tragedy than in its immediate aftermath.
Sunsetting criminal laws doesn’t necessarily cut in a decarceral direction. In fact, repassing decarceral legislation might prove more difficult than passing carceral legislation, such that sunsets make criminal law more carceral. One design possibility would be to sunset carceral criminal legislation but not decarceral criminal legislation to account for the prevailing political winds. Setting the default to require future action to continue laws that will tend to benefit from political tailwinds might make sense even while setting the default differently for laws that will tend not to have such political tailwinds.334 Rather, legislatures can move policy in a decarceral direction through inaction even if they’d need action to move it in a carceral direction. Politics can of course change over time and vary across locations; this analysis depends on the premise that over time and in most parts of the U.S. carceral legislation benefits from a greater political tailwind in general than decarceral legislation. This asymmetry would provide some incentive for a legislature that wishes to avoid future work for itself to favor decarceral over carceral legislation and civil regulation over criminal law.335
Sunsets provide legislatures with a reason to revisit a law after they have better information about how it will work. In criminal law, that information could and should include information about costs—costs imposed on defendants, their loved ones, cities, counties, and on various parts of the state’s budget. In other words, the sunset can encourage repeating the cost-benefit deliberation at a later time with better information.
To avoid criminal sunsets devolving into “ritualistic readoption of the criminal code,” states should have to repass each criminal law separately rather than en masse.336 Separation increases the chances of meaningful deliberation, and it places scarcity of the legislature’s time as a potential constraint on the scope of criminal law.337 So too does it prevent laws that cannot command majority support from being reenacted by attaching them to laws with strong support.338 As a doctrinal matter, the single-subject rule in most state constitutions could be read to require that each criminal law be readopted separately.339 One reason for single-subject rules is to ensure that “every subject considered and passed into law will be considered on its own merits”340—the same basic idea behind requiring each criminal law to be reenacted separately.
Much of this Article is focused on new criminal laws, but sunsets could be applied far more broadly to sweep in the entire existing criminal code of a state (or large portions of it).341 In his thoughtful proposal of a criminal sunset amendment, Richard Myers considers but ultimately rejects the possibility that only malum prohibitum crimes would sunset rather than malum in se crimes.342 Myers rejects this idea because it short circuits the process by which the people’s representatives are supposed to decide which crimes are sufficiently important to maintain.343 Myers seems quite right that it is hard to fathom a politics so dysfunctional that an intentional murder statute might lapse.344 We tend to share Myers’s view that there is no need to limit sunsets to malum prohibitum offenses. That said, rather than let the perfect be the enemy of the good, we could get behind excluding the statutory equivalents of the common law felonies from sunsetting if that made a sunset proposal more politically feasible.345
There is some downside to the idea of sunsetting all carceral criminal laws. Structuring in necessary fluctuation to criminal codes makes it even more difficult than it already is for citizens to know how to conform their conduct to law. But these concerns about citizens not knowing how to conform conduct to law are more pronounced with the creation of new crimes than the expiration of criminal laws.
* * *
While this Section has discussed the potential to use tools of legislative restraint to constrain criminal lawmaking and highlighted a few instances of states using some of these tools in criminal law, most states do not subject their criminal lawmaking to these procedures. Most states require, as a matter of statute or rule, that the legislature estimate the cost of a new criminal law or change to sentencing law, but even those requirements are often ignored in criminal law. The other tools aren’t even required on paper for criminal law in most states.
It may be tempting for any legislator to try to navigate the desire to pass more criminal laws without running afoul of fiscal conservatism by trying to pass ever-more fines and fees—shifting costs of criminal law away from the government and onto those prosecuted. One way to at least mitigate that outcome would be to prevent accounting tricks. On paper, the fines and fees levied against criminal defendants are staggering. But criminal defendants are overwhelmingly poor.346 If legislatures are going to use fines and fees to offset what would otherwise count as increased spending, they should do so based only on the likely amount collected from those fines and fees rather than the amount assessed.
We recognize that high sticker prices on a fiscal note run the risk of slowing the progress or even thwarting decarceral reforms too. That is especially true when legislatures face the tighter constraints of TELs or mandatory offsets. As an example, consider efforts that have succeeded in some states to prohibit involuntary servitude among the incarcerated.347 If those efforts mean that incarcerated persons would be paid at least minimum wage for their work, such efforts would impose a significant financial cost348—estimated nationally at between $8.5 billion and $14.5 billion.349 The benefits—estimated at between $26.8 and $34.7 billion—to children of incarcerated people, victims, and to society more broadly by reducing future crime and incarceration vastly outstrip the costs.350 But we acknowledge that increasing the potency of costs in legislative deliberation makes expensive things harder to pass even when they’re worth passing, as we have seen with these involuntary servitude amendment processes. Indeed, efforts to amend the California Constitution to abolish involuntary servitude in prisons were thwarted in 2022 because of the large sticker price of paying incarcerated persons for their labor.351 In Colorado, organizers supporting a similar measure avoided a similar fate by arguing that the amendment “would have no immediate impact on prison labor.”352 The fiscal note in Colorado mentioned only the prospect of increased court filing costs if the State were sued more often.353
One possible response discussed above regarding sunsets would be to use these fiscal restraint tools only for carceral rather than decarceral reform. The goal is to urge legislators to deliberate rather than legislate reflexively; these fiscal tools serve that goal best when they push back against prevailing political winds rather than sailing with such winds at their backs.354 Asymmetry in rules is not a panacea though; expensive solutions do not enable politicians to thwart “soft on crime” attacks by claiming fiscal restraint.
Conclusion
The politics of criminal law are badly broken. Some of the problems—perceived need for legislative response to a high-profile event and the upside of looking tough on crime—often seem insurmountable. And maybe they are. But we find reason for optimism that the cost of the criminal legal system can (and sometimes does) usefully cross-pressure the politics of criminal law to improve deliberation. Ultimately, our goal is to increase the salience of costs in legislative debates about the criminal legal system and push lawmakers toward deeper consideration of whether more substantive criminal law or harsher sentencing laws are genuinely good ideas worth their costs. We think the right question for carceral criminal law reforms is: Am I willing to raise taxes to pay for this? If legislatures more often sought to answer that question, we think they’d reach better conclusions than when the benefits of carceral criminal law seem enormous and amorphous and its costs diffuse. Criminal law is enormously costly, and its benefits are far less straightforward than some like to think. Some criminal law likely does increase public safety—a hugely important benefit—but criminal law is not public safety. Criminal law is a tool that imposes huge costs on state and local governments, defendants, their loved ones, and their communities. As such, we recognize that legislators weighing the costs of criminal law to state and local governments would still leave many important costs uninternalized. But we trim our sails to focus here only on urging legislatures to consider costs to governments.
A variety of tools of fiscal restraint that legislatures typically use in other areas of law could help constrain carceral criminal lawmaking. The tools range from TELs that cap the overall size of a state’s budget to fiscal notes that simply require costs to be calculated. In the few states that have TELs, shifting costs to local government and purporting to shift costs onto defendants provide unjustified release valves from the pressure of the TEL. Mandatory offsets such as PAYGO allow legislatures to make a spending-spending tradeoff as TELs do but also allow the legislature to raise taxes to pay for carceral criminal law reforms. Although our primary goal here is to identify the tradeoffs between the different policy levers, we prefer mandatory offsets to TELs because we tend to think that if legislators are willing to vote to increase taxes to pay for criminal law reform, then they should be allowed to make that politically difficult choice. Fiscal notes provide a much gentler tool than TELs or mandatory offsets—simply requiring the legislature to see a cost estimate before it enacts a new law. We think there would be benefit from states adhering to their already-existing statutes and rules requiring fiscal notes in criminal law, although we tend to think that the perversities of criminal law politics are so potent that stronger tools like mandatory offsets may work better than fiscal notes. Lastly, requiring criminal laws to sunset provides another useful tool of restraint insofar as the legislature must either let the bill and its costs expire or consider whether the bill has worked.
Legislatures can (and often do) seek to externalize the costs of criminal law onto defendants and their loved ones by imposing fines and fees. We think that practice is unjustifiable; if it is worthwhile for our government to inflict the misery of the criminal legal system onto some of its citizenry then we all ought to pay for that. But even if one were comfortable in concept with fines and fees, there is a massive gulf between the amounts levied and collected; only the portion of revenue that will likely be collected should count for purposes of offsets and other revenue estimation.
At bottom, our goal is to urge legislatures to think harder about the costs and the benefits of criminal law rather than treating criminal law as impervious to cost and reflexively treating criminal law as public safety. Criminal law isn’t public safety, although the two bear some relationship. Tools of fiscal restraint alone are hardly a perfect fix without greater consideration of the goals we’re trying to achieve. Solutions like cutting the social safety net and removing funding for early childhood education all look like they save money on paper. The problem is that over the longer term they cost vastly more in criminal law than the money they save. And they detract from public safety. We resist easy solutions, but we think the best answers live in treating criminal law more like other areas of legislative policymaking and letting fiscal restraint play a bigger role.
- See Darryl K. Brown, Cost-Benefit Analysis in Criminal Law, 92 Calif. L. Rev. 323, 338–58 (2004) (encouraging cost-benefit analysis in criminal law because benefits are far more salient than costs); see also Cheryl Lero Jonson, John E. Eck & Francis T. Cullen, The Small Prison, in The American Prison: Imagining a Different Future 215, 217 (Francis T. Cullen, Cheryl Lero Jonson & Mary K. Stohr eds., 2014) (“For nearly 40 years, state after state has engaged in a correctional orgy in which offenders have been sent to prison with little thought of how this would be financed.”). ↩︎
- See William J. Stuntz, The Pathological Politics of Criminal Law, 100 Mich. L. Rev. 505, 508 (2001) (snidely remarking about whether “more” is a normative theory underpinning criminal lawmaking). We recognize, of course, that crime imposes costs. Our point is that the costs of crime do not always exceed the costs of criminal law enforcement. ↩︎
- Cf., e.g., Tracey L. Meares, Neal Katyal & Dan M. Kahan, Updating the Study of Punishment, 56 Stan. L. Rev. 1171, 1205 (2004) (discussing the “deliberation-forcing” rationale for courts construing criminal prohibitions narrowly); William N. Eskridge, Jr. & Philip P. Frickey, Quasi-Constitutional Law: Clear Statement Rules as Constitutional Lawmaking, 45 Vand. L. Rev. 593, 631 (1992) (discussing how interpretive presumptions in statutory interpretation and clear statement rules can promote legislative deliberation); Joel S. Johnson, Dealing with Dead Crimes, 111 Geo. L.J. 95, 108–10 (2022) (discussing the idea of deliberation-forcing mechanisms in the context of desuetude). ↩︎
- Ala. Code § 13A-11-82.1 (West 2026) (codifying Act 2025-54). ↩︎
- Telephone Interview with Phillip Ensler, State Rep., Ala. House of Reps. (Nov. 21, 2025). ↩︎
- Legislative Servs. Agency, Fiscal Div., Fiscal Note for H.B. 26, 2025 Reg. Sess. (Ala. 2025), https://alison.legislature.state.al.us/files/pdf/SearchableInstruments/2025RS/FiscalNotes/FN-5FZG3M3-1.pdf [https://perma.cc/K8FE-D3W9]. ↩︎
- See, e.g., Ronald F. Wright, Three Strikes Legislation and a Sinking Fund Proposal, 8 Fed. Sent’g Rep. 80, 80 (1995) (arguing that legislatures discount the future prison costs when passing three-strikes laws). ↩︎
- See, e.g., U.S. Dep’t of Just., Civ. Rts. Div. & U.S. Att’y’s Offs., N. & S. Dists. of Miss., Civ. Divs., Investigation of Central Mississippi Correctional Facility, South Mississippi Correctional Institution, and Wilkinson County Correctional Facility 5 (2024) [hereinafter Mississippi DOJ Report], https://www.justice.gov/d9/2024-02/2024.02.26_ms_doc_findings_report_it_508_reviewed_0.pdf [https://perma.cc/TN7R-R7BS]; U.S. Dep’t of Just., Civ. Rts. Div. & U.S. Att’y’s Offs., N., Middle, and S. Dists. of Ala., Investigation of Alabama’s State Prisons for Men 5 (2019) [hereinafter Alabama DOJ Report], https://www.justice.gov/d9/press-releases/attachments/2019/04/03/notice_letter_and_report_aldoc.pdf [https://perma.cc/T9ED-PNAX]. ↩︎
- See, e.g., Elizabeth Garrett, Response, A Fiscal Constitution with Supermajority Voting Rules, 40 Wm. & Mary L. Rev. 471, 474 (1999) (“[P]rocedures requiring congressional entities to produce particular information before floor debate may highlight issues that lawmakers would otherwise overlook.”). ↩︎
- See Kevin R. Reitz & Cecelia M. Klingele, Model Penal Code: Sentencing—Workable Limits on Mass Punishment, 48 Crime & Just. 255, 276 (2019). ↩︎
- See, e.g., Johnson, supra note 3, at 108–10; Meares, Katyal & Kahan, supra note 3, at 1204–09. ↩︎
- See, e.g., Or. Rev. Stat. Ann. § 173.029 (West 2026). ↩︎
- See Wright, supra note 7, at 81 (advocating such an approach). ↩︎
- Colo. Rev. Stat. Ann. § 2-2-703 (West 2025), repealed by, S.B. 25-210, 2025 Colo. Sess. Laws 468, ch. 112. ↩︎
- S.B. 25-210. ↩︎
- Tenn. Code Ann. § 9-4-210(e)(1) (West 2026); Va. Code Ann. § 30-19.1:4(H) (West 2026). ↩︎
- Infra Part II. ↩︎
- See India Thusi, Policing Is Not a Good, 110 Geo. L.J. Online 226, 230–41 (2022). ↩︎
- Stuntz, supra note 2, at 508. ↩︎
- See, e.g., David M. Schizer, Fiscal Policy in an Era of Austerity, 35 Harv. J.L. & Pub. Pol’y 453, 476 (2012); Russell M. Gold, The Price of Criminal Law, 56 Ariz. St. L.J. 841, 863–909 (2024). ↩︎
- Richard A. Bierschbach & Stephanos Bibas, Rationing Criminal Justice, 116 Mich. L. Rev. 187, 220 (2017). ↩︎
- Id. at 209–33; Brown, supra note 1, at 338–58. ↩︎
- Bierschbach & Bibas, supra note 21, at 212; Brown, supra note 1, at 345–49. ↩︎
- See W.C. Bunting, The Regulation of Sentencing Decisions: Why Information Disclosure Is Not Sufficient, and What to Do About It, 70 N.Y.U. Ann. Surv. Am. L. 41, 57–62 (2014) (criticizing fiscal notes as inadequate for the task); David B. Kopel & Trevor Burrus, Reducing the Drug War’s Damage to Government Budgets, 35 Harv. J.L. & Pub. Pol’y 543, 545–50 (2012) (arguing that fiscal notes could help constrain the war on drugs, dedicating only a small section to fiscal notes); Wright, supra note 7, at 80 (arguing that legislatures ought to be required to set aside funds to pay for distant but certain budgetary impacts of legislation that will increase sentences); James W. Ganas, PAYGO for Criminal Sentencing: Political Incentives and Process Reform, 99 N.Y.U. L. Rev. 320, 343 (2024) (arguing that any increase in incarceration that legislatures pass should be offset by a decrease in incarceration for other crimes). ↩︎
- Wright, supra note 7, at 80. ↩︎
- See Rachel E. Barkow & Kathleen M. O’Neill, Delegating Punitive Power: The Political Economy of Sentencing Commission and Guideline Formation, 84 Tex. L. Rev. 1973, 1986, 2011 (2006). ↩︎
- See Gold, supra note 20, at 843–50, 863–64, 884–90 (arguing that we’re getting criminal law on the cheap and that we should be paying more public money than we are even at current enforcement levels). ↩︎
- See generally Hadar Aviram, Cheap on Crime: Recession-Era Politics and the Transformation of American Punishment (2015) (exploring the role that recession has played in American criminal law reform). ↩︎
- E.g., Rachel Elise Barkow, Prisoners of Politics: Breaking the Cycle of Mass Incarceration 5 (2019); Erik Luna, The Overcriminalization Phenomenon, 54 Am. U. L. Rev. 703, 719 (2005); Paul H. Robinson & Michael T. Cahill, The Accelerating Degradation of American Criminal Codes, 56 Hastings L.J. 633, 634 (2005). ↩︎
- Barkow, supra note 29, at 3. See Robinson & Cahill, supra note 29, at 644 (“Many amendments and new offenses are enacted . . . to respond to an especially grim or headlined case . . . .”); Sara Sun Beale, Essay, The Many Faces of Overcriminalization: From Morals and Mattress Tags to Overfederalization, 54 Am. U. L. Rev. 747, 755 (2005) (explaining that federal criminal law “contains what some have called the crime du jour—legislation drafted in response to whatever crime is the focal point in the media—even if that offense is already defined and punished harshly and effectively under state law”); see also Barkow, supra note 29, at 1 (“[C]riminal justice policy in the United States is set largely based on emotions and the gut reactions of laypeople.”). ↩︎
- Barkow, supra note 29, at 6. ↩︎
- Jonathan Franklin & Juliana Kim, Carlee Russell Is Charged with Kidnapping Hoax, NPR (July 28, 2023, at 16:05 ET), https://www.npr.org/2023/07/28/1190774312/carlee-russell-charged-kidnapping-hoax-alabama [https://perma.cc/NWS7-83FB]; Ala. Code § 13A-10-9(b) (West 2026) (codifying Act 2024-345). ↩︎
- Sun Beale, supra note 30, at 773; Sara Sun Beale, Still Tough on Crime? Prospects for Restorative Justice in the United States, 2003 Utah L. Rev. 413, 425–32; see also Christopher H. Achen & Larry M. Bartels, Democracy For Realists: Why Elections Do Not Produce Responsive Government 267–84 (2016) (explaining that voters misunderstand facts in the world and vote largely out of partisan loyalties, social identities, and group attachments); Barkow, supra note 29, at 106 (discussing voters’ poor understanding of criminal justice issues, including their lack of awareness of crime data). ↩︎
- See Barkow, supra note 29, at 5; William J. Stuntz, Plea Bargaining and Criminal Law’s Disappearing Shadow, 117 Harv. L. Rev. 2548, 2560 (2004); Sara Sun Beale, What’s Law Got to Do with It? The Political, Social, Psychological and Other Non-Legal Factors Influencing the Development of (Federal) Criminal Law, 1 Buff. Crim. L. Rev. 23 (1997); Robinson & Cahill, supra note 29. ↩︎
- Stuntz, supra note 2, at 546–49; see Stuntz, supra note 34, at 2558. ↩︎
- Stuntz, supra note 2, at 546–48. ↩︎
- John F. Pfaff, Locked In: The True Causes of Mass Incarceration—and How to Achieve Real Reform 13 (2017); Gold, supra note 20, at 850, 909. ↩︎
- Univ. of N.C. Sch. of L., The Prosecutors & Pols. Project, Crime and Punishment Across America: A 50 State Legislative Analysis 4, 9 (2025) [hereinafter Crime and Punishment Report], https://papers.ssrn.com/abstract=5506800 [https://perma.cc/47H5-J55G]. ↩︎
- Id. at 4, 7; Darryl K. Brown, Democracy and Decriminalization, 86 Tex. L. Rev. 223, 245–49 (2007). ↩︎
- Crime and Punishment Report, supra note 38, at 4; Brown, supra note 39, at 245–49; see also Alice Ristroph, Just Violence, 56 Ariz. L. Rev. 1017, 1052 (2014) (“Substantive criminal law has steadily expanded; more conduct is criminalized each year, and decriminalizations are few in comparison.”). Criminal law bills fail in state legislatures at about the same rates or slightly lower rates than other bills, at least according to two empirical studies of different time periods. Crime and Punishment Report, supra note 38, at 7; Brown, supra note 39, at 245–49. ↩︎
- See, e.g., Alaska Stat. Ann. §§ 17.38.010–.900 (West 2026); Cal. Health & Safety Code § 11362.1 (West 2026); Colo. Const. art. XVIII, § 16; Cannabis Regulation and Tax Act, 410 Ill. Comp. Stat. Ann. 705/1-1 to 999-99 (West 2026); La. Stat. Ann. § 40:966 (West 2026); see also, e.g., Jonson, Eck & Cullen, supra note 1, at 225; Kate Bryan, Nat‘l Conf. of State Legs., Cannabis Overview (June 20, 2024), https://www.ncsl.org/civil-and-criminal-justice/cannabis-overview [https://perma.cc/M67S-R3PF]. Although we focus on state legislatures, even Congress has reduced some federal drug sentences in the past couple of decades. Fair Sentencing Act of 2010, Pub. L. No. 111-220, secs. 2–3, 124 Stat. 2372, 2372 (codified as amended in scattered sections of 21 U.S.C.); First Step Act of 2018, Pub. L. No. 115-391, secs. 401–404, 132 Stat. 5194, 5220–22 (codified as amended in scattered sections of 18 U.S.C. and 21 U.S.C.). But see Fair Sentencing Act of 2010 secs. 4–6 (increasing drug sentences and calling for the Sentencing Commission to increase drug sentences in some instances). ↩︎
- Lisa L. Miller, The Perils of Federalism: Race, Poverty, and the Politics of Crime Control 121–23, 139, 145, 176 (2008); Nirej Sekhon, Public Defenders and Collective Action, 52 Fla. St. U. L. Rev. 277, 286, 289 (2025); Barkow, supra note 29, at 112–19. Decarceral interest groups may have difficulty gaining legislative traction insofar as they’re seen to benefit people who are socially constructed as deviant. See Jonathan J. Pierce et al., Social Construction and Policy Design: A Review of Past Applications, 42 Pol’y Stud. J. 1, 13 (2014); Anne Schneider & Helen Ingram, Social Construction of Target Populations: Implications for Politics and Policy, 87 Am. Pol. Sci. Rev. 334, 335–39 (1993). ↩︎
- Maria Ponomarenko, Some Realism About Criminal Justice Localism, 173 U. Pa. L. Rev. 789, 831–35 (2024). ↩︎
- Brenner M. Fissell, Localizing Public Order, 78 Vand. L. Rev. 1521, 1524, 1531–37 (2025). ↩︎
- Id. at 1534 fig. 2. ↩︎
- Brenner M. Fissell, Against Criminal Law Localism, 81 Md. L. Rev. 1119, 1138–43 (2022); see Ponomarenko, supra note 43, at 838. ↩︎
- See Reitz & Klingele, supra note 10, at 276 (“In the history of state sentencing systems, however, [correctional resource management] has generally been used to slow or stop incarceration growth.” (citing Richard S. Frase, Just Sentencing: Principles and Procedures for a Workable System (2013))). ↩︎
- See Michael Tonry, Sentencing Matters 5, 58–61, 67, 193 (1996); Rachel E. Barkow, Federalism and the Politics of Sentencing, 105 Colum. L. Rev. 1276, 1285–89 (2005); Russell M. Gold, Prosecutors and Their Legislatures, Legislatures and Their Prosecutors, in Oxford Handbook of Prosecutors and Prosecution 327, 337–38 (Ronald F. Wright, Kay L. Levine & Russell M. Gold eds., 2021). ↩︎
- Ronald F. Wright, Portable Minimalism in Sentencing Politics, 2011 Cardozo L. Rev. DeNovo 9, 20–21; Fox Butterfield, With Cash Tight, States Reassess Long Jail Terms, N.Y. Times (Nov. 10, 2003), https://www.nytimes.com/2003/11/10/us/with-cash-tight-states-reassess-long-jail-terms.html [https://perma.cc/G39Q-72MV] (“In the past year, about 25 states have passed laws eliminating some of the lengthy mandatory minimum sentences so popular in the 1980’s and 1990’s, restoring early release for parole and offering treatment instead of incarceration for some drug offenders.”). ↩︎
- Daniel F. Wilhelm & Nicholas R. Turner, Is the Budget Crisis Changing the Way We Look at Sentencing and Incarceration?, 15 Fed. Sent’g Rep. 41, 43 (2002); Jon Wool & Don Stemen, Changing Fortunes or Changing Attitudes? Sentencing and Corrections Reforms in 2003, 16 Fed. Sent’g Rep. 294, 294, 298–99 (2004); see also Aviram, supra note 28, at 56 (explaining how fiscal pressures in criminal law can forge alliances). Local police departments sometimes make meaningful policy changes because of the financial pressures imposed by insurers. John Rappaport, How Private Insurers Regulate Public Police, 130 Harv. L. Rev. 1539, 1573–76, 1588–91 (2017). ↩︎
- See, e.g., Ian Urbina, Citing Cost, States in U.S. Consider Halting Death Penalty, N.Y. Times (Dec. 5, 2009), https://www.nytimes.com/2009/02/25/world/americas/25iht-25death.20417446.html [https://perma.cc/VH9C-LATW]; Carol S. Steiker & Jordan M. Steiker, Costs and Capital Punishment: A New Consideration Transforms an Old Debate, 2010 U. Chi. Legal F. 117, 118–24; Jolie McLaughlin, Note, The Price of Justice: Interest-Convergence, Cost, and the Anti-Death Penalty Movement, 108 Nw. U. L. Rev. 675, 688–704 (2014); Costs, Death Penalty Info. Ctr., https://deathpenaltyinfo.org/policy-issues/policy/costs [https://perma.cc/HZ5B-P2NV]. ↩︎
- See Barkow, supra note 48, at 1287–88; Butterfield, supra note 49. ↩︎
- Barkow, supra note 48, at 1288–89; Wool & Stemen, supra note 50, at 300–01. ↩︎
- See, e.g., Rachel E. Barkow, Administering Crime, 52 UCLA L. Rev. 715, 775–93 (2005); Barkow, supra note 48, at 1286–90. ↩︎
- See Sun Beale, supra note 30, at 774; Sun Beale, supra note 33, at 435–36; Gold, supra note 48, at 337–38; Ronald F. Wright, The Power of Bureaucracy in the Response to Blakely and Booker, 43 Hou. L. Rev. 389, 404 (2006); Rachel E. Barkow, The Political Market for Criminal Justice, 104 Mich. L. Rev. 1713, 1720–23 (2006); Barkow, supra note 48, at 1277, 1300–12. ↩︎
- Barkow, supra note 48, at 1292. ↩︎
- Pew Charitable Trs., Louisiana’s 2017 Criminal Justice Reforms 1–3 (2018), https://www.pew.org/~/media/assets/2018/03/pspp_louisianas_2017_criminal_justice_reforms.pdf [https://perma.cc/QW2T-QBJW]. ↩︎
- Id. at 6. ↩︎
- Id. at 7, 15. ↩︎
- Id. at 11. ↩︎
- Emmett Sanders, An Act of Regression: Louisiana Takes a Giant Step Backward in Parole and Sentencing Reform, Prison Pol’y Initiative (Aug. 21, 2024), https://www.prisonpolicy.org/blog/2024/08/21/louisiana_parole_reform/ [https://perma.cc/MK3T-6P8M]. ↩︎
- Leah Sakala & Samantha Harvell, Urb. Inst., Justice Reinvestment Initiative: Alabama 1 (2019), https://www.urban.org/sites/default/files/2020/03/06/justice_reinvestment_initiative_jri_alabama.pdf [https://perma.cc/U77R-QXZM]; see Brown v. Plata, 563 U.S. 493, 545 (2011). ↩︎
- Sakala & Harvell, supra note 62, at 1; Council of State Gov’ts Just. Ctr., Alabama’s Justice Reinvestment Approach: Reducing Prison Overcrowding and Strengthening Community-Based Supervision 2–3 (2015) [hereinafter Reinvestment Approach], https://csgjusticecenter.org/wpcontent/uploads/2020/01/AlabamasJusticeReinvestmentApproach.pdf [https://perma.cc/UFE3-8C67]. ↩︎
- Reinvestment Approach, supra note 63, at 1. ↩︎
- Sakala & Harvell, supra note 62, at 1–2. ↩︎
- Ala. Appleseed Ctr. for L. & Just., Unsustainable: Alabama’s Increasing Trend of Keeping the Elderly Behind Bars 2–4 (2022), https://alabamaappleseed.org/wp-content/uploads/2022/11/Unsustainable_Alabama-Appleseed.pdf [https://perma.cc/PHS8-CRWR]. ↩︎
- Ralph Chapoco, Analysis Says Alabama Department of Corrections Received $5 Billion in State Funding in Five Years, Ala. Reflector (Sept. 4, 2025), https://alabamareflector.com/2025/09/04/analysis-says-alabama-department-of-corrections-received-5-billion-in-state-funding-in-five-years/ [https://perma.cc/TT3C-TEZM]. ↩︎
- See Ronald F. Wright, Counting the Cost of Sentencing in North Carolina, 1980–2000, 29 Crime & Just. 39, 52–86 (2002). ↩︎
- E.g., Michael Pinard, Collateral Consequences of Criminal Convictions: Confronting Issues of Race and Dignity, 85 N.Y.U. L. Rev. 457, 459 (2010); I. Bennett Capers, Against Prosecutors, 105 Corn. L. Rev. 1561, 1597 & n.196 (2020). ↩︎
- Wright, supra note 68, at 44. ↩︎
- Id. at 40. There were several steps in between, including North Carolina releasing people on parole to alleviate overcrowding and then passing sentencing reform that preserved so much judicial discretion that it did not adequately constrain costs. Id. at 44–52. ↩︎
- Id. at 52–58. ↩︎
- Id. at 54. ↩︎
- Id. at 58. ↩︎
- Id. ↩︎
- Id. at 70–71. ↩︎
- Id. at 71–72. ↩︎
- Id. at 72–73. ↩︎
- Id. at 75–76. ↩︎
- Id. at 86–90. ↩︎
- Id. at 81–82. ↩︎
- See id. at 90. ↩︎
- See id. at 84 (“The commission also revised its projections each year.”). ↩︎
- Id. at 82. ↩︎
- Michael Tonry, Sentencing in America, 1975–2025, 42 Crime & Just. 141, 156 (2013). ↩︎
- Id. ↩︎
- Id. at 157. ↩︎
- Reitz & Klingele, supra note 10, at 275. ↩︎
- Id. at 278. ↩︎
- Michael A. Wolff, Missouri Provides Cost of Sentences and Recidivism Data: What Does Cost Have to Do with Justice?, 24 Fed. Sent’g Rep. 161, 162 (2012). ↩︎
- Ryan W. Scott, How (Not) to Implement Cost as a Sentencing Factor, 24 Fed. Sent’g Rep. 172, 172, 175–76 (2012) (arguing that a more systemic approach to accounting for costs such as through a sentencing commission recalibrating sentencing guidelines would be better than providing cost information to individual judges); Chad Flanders, Cost as a Sentencing Factor: Missouri’s Experiment, 77 Mo. L. Rev. 391, 398 (2012) (criticizing the reform because legislatures, not judges, should account for the role of costs in criminal law); Chad Flanders, Cost and Sentencing: Some Pragmatic and Institutional Doubts, 24 Fed. Sent’g Rep. 164, 167 (2012) (same). ↩︎
- Bobby Allyn, Philadelphia’s New DA Wants Prosecutors to Talk Cost of Incarceration While in Court, NPR (Mar. 31, 2018, at 08:22 ET), https://www.npr.org/2018/03/31/598318897/philadephias-new-da-wants-prosecutors-to-talk-cost-of-incarceration-while-in-cou [https://perma.cc/55L9-GZGQ]; Phila. Dist. Att’y’s Off., Philadelphia DAO New Policies (2018), https://phillyda.org/wp-content/uploads/2022/04/DAO-New-Policies-2.15.2018-UPDATED.pdf [https://perma.cc/S6N8-QR3V]. ↩︎
- Adam M. Gershowitz, An Informational Approach to the Mass Imprisonment Problem, 40 Ariz. St. L.J. 47, 50–51 (2008). ↩︎
- See generally Russell M. Gold, Promoting Democracy in Prosecution, 86 Wash. L. Rev. 69 (2011). ↩︎
- See, e.g., Bierschbach & Bibas, supra note 21, at 195–98; Gold, supra note 20, at 850, 902–04. ↩︎
- See Gold, supra note 20, at 873; Aurélie Ouss, Misaligned Incentives and the Scale of Incarceration in the United States, 191 J. Pub. Econ., art. no. 104285, at 1 (2020). ↩︎
- Eve Brensike Primus, Defense Counsel and Public Defense, in 3 Reforming Criminal Justice 121, 125 (Erik Luna ed., 2017). ↩︎
- Gold, supra note 20, at 888–95. ↩︎
- E.g., Bierschbach & Bibas, supra note 21, at 195–200. ↩︎
- Franklin E. Zimring & Gordon Hawkins, The Scale of Imprisonment 140 (1991) (coining the term). ↩︎
- See also Jonson, Eck & Cullen, supra note 1, at 226 (arguing that counties should have to bear the costs of the prison beds they use). ↩︎
- Adam M. Gershowitz, Consolidating Local Criminal Justice: Should Prosecutors Control the Jails?, 51 Wake Forest L. Rev. 677, 679 (2016); Bierschbach & Bibas, supra note 21, at 198. ↩︎
- Joan Petersilia et al., Stan. Crim. Just. Ctr., Voices from the Field: How California Stakeholders View Public Safety Realignment 27–35 (2014), https://papers.ssrn.com/abstract=2395498 [https://perma.cc/33QN-ZJ9T]; see also Brown v. Plata, 563 U.S. 493, 502 (2011) (ordering California to reduce its prison population to address inadequate medical care). ↩︎
- Magnus Lofstrom & Steven Raphael, Pub. Pol’y Inst. Cal., Impact of Realignment on County Jail Populations 9 (2013); James Austin, Regulating California’s Prison Population: The Use of Sticks and Carrots, 664 Annals Am. Acad. Pol. & Soc. Sci. 84, 95 (2016). ↩︎
- Petersilia et al., supra note 103, at 43 (“Roughly speaking, the legislature split the current cost of state supervision by about 50% with the counties.”); see also W. David Ball, Why State Prisons?, 33 Yale L. & Pol’y Rev. 75, 82–83 (2014) (explaining the distorted incentives created when local governments prosecute but states pay for prisons, while local governments pay for alternatives to incarceration, such as probation). ↩︎
- Lofstrom & Raphael, supra note 104, at 9. ↩︎
- Id. at 19; Austin, supra note 104, at 101; see also Aaron Gottlieb, Pajarita Charles, Branden McLeod, Jean Kjellstrand & Janaé Bonsu, Were California’s Decarceration Efforts Smart? A Quasi-Experimental Examination of Racial, Ethnic, and Gender Disparities, 48 Crim. Just. & Behav. 116, 125–26 (2021) (measuring the effect of several criminal law reforms in California). ↩︎
- We use the term Latinx here to match the source on which we rely. Gottlieb et al., supra note 107, at 127. ↩︎
- Id. at 126–27. ↩︎
- Austin, supra note 104, at 98–99. ↩︎
- S.B. 681, 1996 Cal. Stat. 22, ch. 6; Ouss, supra note 96, at 3–4. ↩︎
- Ouss, supra note 96, at 4–11. ↩︎
- Russell M. Gold, Paying for Pretrial Detention, 98 N.C. L. Rev. 1255, 1285–89 (2020). Fines are imposed after conviction and intended as punishment, while fees are intended to recoup the expense of administering the criminal legal system. Wayne A. Logan & Ronald F. Wright, Mercenary Criminal Justice, 2014 U. Ill. L. Rev. 1175, 1203; Beth A. Colgan, Fines, Fees, and Forfeitures, in 4 Reforming Criminal Justice, supra note 97, at 205, 206–08, https://papers.ssrn.com/abstract=3019435 [https://perma.cc/8HX2-STVY]. For an argument that the public should bear the costs of the criminal legal process rather than trying to shift them onto defendants through fees, see Gold, supra note 20, at 872–74, 896–99. ↩︎
- Ala. Appleseed Ctr. for L. & Just., Under Pressure: How Fines and Fees Hurt People, Undermine Public Safety, and Drive Alabama’s Racial Wealth Divide (2018), http://www.alabamaappleseed.org/wp-content/uploads/2018/10/AA1240-FinesandFees-10-10-FINAL.pdf [https://perma.cc/Q34D-CA83]. ↩︎
- Ala. Code § 13A-12-281 (West 2026). ↩︎
- One of the authors has observed this approach in numerous courtroom observations and has heard it in interviews she has conducted. See also Beth A. Colgan & Jean Galbraith, The Failed Promise of Installment Fines, 172 U. Pa. L. Rev. 989, 999–1048 (2024) (explaining the growth of and problems with charging fines on installment plans). ↩︎
- Sarah Picard, Leah Nelson, Rae Walker, Kasey Eickmeyer & Ellie Wilson, MDRC, Findings from the Jefferson County Equitable Fines and Fees Project 8 (2025). ↩︎
- Id. ↩︎
- Id. at 7; see also Cortney E. Lollar, What Is Criminal Restitution?, 100 Iowa L. Rev. 93, 100–03, 130–48 (2014) (explaining how restitution has metastasized way beyond disgorging unjust enrichment or even compensation for individual victims); Cortney E. Lollar, Punishment Through Restitution, 34 Fed. Sent’g Rep. 98, 99–100 (2022) (same). ↩︎
- See Logan & Wright, supra note 113, at 1178 (“When the tax-paying public is not asked to fund criminal justice, it gets a distorted message about the real costs of enforcement. . . . [D]oing so weakens one of the key moderating influences in public safety politics.”). ↩︎
- See id. at 1200 (explaining that Legal Financial Obligations (LFOs) that generate substantial revenue for a government unit “pose[] a threat to neutrality” in ways that can violate due process). Greater use of fines and fees that yield revenue for prosecutors’ offices reduce the marginal cost of prosecutions by generating an offset—at least to the extent they’re collected. Gold, supra note 20, at 872–73, 896–99. ↩︎
- See, e.g., Logan & Wright, supra note 113, at 1177 (“Today, criminal justice actors increasingly rely on the income from LFOs to fund ordinary system operations and to expand the system’s reach. When this happens, courts and other criminal justice actors become mercenaries, in effect working on commission.”). ↩︎
- Ala. Off. Prosecution Servs., District Attorney’s Legislative Budget Report, September 8, 2022 (2022); see also Logan & Wright, supra note 113, at 1194 (discussing allocation of some fee proceeds to district attorney offices); Gold, supra note 113, at 1287–88 (same). ↩︎
- See Logan & Wright, supra note 113, at 1183 & n.50 (describing contingency-fee payment structures for prosecutors in the United States at earlier times in our history). ↩︎
- Peter Lushing, The Fall and Rise of the Criminal Contingent Fee, 82 J. Crim. L. & Criminology 498, 505 (1991); Baca v. Padilla, 190 P. 730, 731–32 (N.M. 1920); see also Model Rules of Pro. Conduct r. 3.8 cmt. 1 (A.B.A. 1983) (“A prosecutor has the responsibility of a minister of justice and not simply that of an advocate.”). ↩︎
- Cf. Garrett, supra note 9, at 476 (describing procedural protections to make certain types of legislation more difficult to pass as “lying along a spectrum” of strength). ↩︎
- See supra Part II. ↩︎
- Colo. Rev. Stat. Ann. § 2-2-703 (West 2025), repealed by, S.B. 25-210, 2025 Colo. Sess. Laws 468, ch. 112. ↩︎
- See Brown, supra note 1, at 340–49, 351–58 (arguing for cost-benefit analysis at an individual case level to consider whether the benefits of a particular prosecution exceed its costs and how the costs and benefits of deploying the criminal law tool compare to the costs and benefits of other enforcement tools to address the same wrongdoing). ↩︎
- Id.; see also Jeremy Bentham, An Introduction to the Principles of Morals and Legislation 11–16 (J. H. Burns & H. L. A. Hart eds., Athlone Books 1996) (1780) (describing the principle of utility). ↩︎
- See Brown, supra note 1, at 363 (“The internal rules governing legislatures also affect the translation of preferences into legislation.”). ↩︎
- See id. (“[Cost-benefit analysis] does not eliminate political judgments” or “replace politics with neutral rationality.”). ↩︎
- Other goals for the criminal legal system could include preserving or increasing its use as a tool of racial oppression, e.g., Paul Butler, The System Is Working the Way It Is Supposed to: The Limits of Criminal Justice Reform, 104 Geo. L.J. 1419, 1425–27 (2016). But if that’s the diagnosis, then its ailments are incurable. To that end, this analysis tries to take seriously a shared goal of public safety, even though there is much to work out about its meaning. ↩︎
- See, e.g., Allegra M. McLeod, Beyond the Carceral State, 95 Tex. L. Rev. 651, 686 (2017) (book review) (discussing other means of promoting safety than the carceral state); Dorothy E. Roberts, Foreword: Abolition Constitutionalism, 133 Harv. L. Rev. 1, 44 (2019) (“Abolitionists, therefore, are both developing nonpunitive measures to deal with harm and creating new conditions to prevent harm from occurring in the first place, recognizing both as better approaches to ensuring safety and security than relying on police and prisons.”); Barry Friedman, What Is Public Safety?, 102 B.U. L. Rev. 725, 739–46 (2022) (articulating a broader definition of public safety that accounts for the dangers of being unhoused, for instance); Russell M. Gold, Look What You Made Me Do, 82 Wash. & Lee L. Rev. 1377, 1439–60 (2025) (arguing that stifling counternarratives that might challenge the status quo in the criminal legal system values a perception of public safety in ways that stifle hard conversations about how to actually increase safety); Thusi, supra note 18, at 229 (explaining why policing and public safety are not coextensive and that a community can be both overpoliced and underprotected simultaneously). ↩︎
- For purposes of this Article, we do not take a position on the role that each of the purposes of punishment—incapacitation, deterrence, retribution, or perhaps rehabilitation, e.g., Michael Tonry, Purposes and Functions of Sentencing, 34 Crime & Just. 1, 6 (2006)—play in promoting public safety. ↩︎
- See, e.g., Mariame Kaba, Opinion, Yes, We Mean Literally Abolish the Police, N.Y. Times (June 12, 2020), https://www.nytimes.com/2020/06/12/opinion/sunday/floyd-abolish-defund-police.html [https://perma.cc/SXZ3-YCST]. ↩︎
- See, e.g., Thomas Ward Frampton, Essay, The Dangerous Few: Taking Seriously Prison Abolition and Its Skeptics, 135 Harv. L. Rev. 2013, 2046 (2022); Shaila Dewan, Inside America’s Black Box: A Rare Look at the Violence of Incarceration, N.Y. Times (Mar. 30, 2019), https://www.nytimes.com/2019/03/30/us/inside-americas-black-box.html [https://perma.cc/HAP3-DZEP]; Ruth Wilson Gilmore, Golden Gulag: Prisons, Surplus, Crisis, and Opposition in Globalizing California 16 (2007); Angela Y. Davis, Are Prisons Obsolete? 10, 20–21 (2003). In our home state of Alabama, 325 people died in prison in 2023, making a total of 1,045 deaths in Alabama prisons from April 2019 to the end of 2023. Eddie Burkhalter, Record Loss of Life in 2023 Pushes ADOC’s Death Total Over 1,000 Since DOJ Put State on Notice, Ala. Appleseed (Jan. 29, 2024), https://alabamaappleseed.org/author/eddie-burkhalter/record-loss-of-life-in-2023-pushes-adocs-death-total-over-1000-since-doj-put-state-on-notice/ [https://perma.cc/2E97-SSZH]. Some scholars conclude that we ought to abolish prisons or the prison-industrial complex because the criminal legal system undermines public safety more than it promotes it, and public safety can be accomplished in less harmful ways. See, e.g., Allegra M. McLeod, Prison Abolition and Grounded Justice, 62 UCLA L. Rev. 1156, 1172–207 (2015); see also, e.g., Roberts, supra note 134, at 4–8 (arguing for abolition of the prison industrial complex because its roots lie in subjugating Black people after the end of slavery and because caging people is not necessary to solve social problems); Marbre Stahly-Butts & Amna A. Akbar, Reforms for Radicals? An Abolitionist Framework, 68 UCLA L. Rev. 1544, 1553 (2022) (“[A]bolitionists see the system as working ‘in containing, controlling, killing, and disappearing the people that it is meant to.’” (citation omitted)). Our view is that because we vastly overuse the criminal legal system, many of these uses undermine public safety more than they promote it. But we could use it less frequently—when the harms the system inflicts are worth bearing. See, e.g., Christopher Slobogin, Essay, The Minimalist Alternative to Abolitionism: Focusing on the Non-Dangerous Many, 77 Vand. L. Rev. 531, 535–56 (2024); James Austin & Lauren-Brooke Eisen with James Cullen & Jonathan Frank, Brennan Ctr. for Just., How Many Americans Are Unnecessarily Incarcerated? 7 (2016), https://www.brennancenter.org/our-work/research-reports/how-many-americans-are-unnecessarily-incarcerated [https://perma.cc/46E8-LMEH] (calculating that 576,000 people “are incarcerated with little public safety rationale” and that releasing them would save almost $20 billion per year). And we could make our jails and prisons vastly safer. See, e.g., McLeod, supra, at 1180–85 (arguing that some harms of prison are inherent). Our goal is decidedly less ambitious than the project of abolishing prisons or the prison-industrial complex. Benjamin Levin, Essay, Abolitionisms, 10 UCLA Crim. Just. L. Rev. (forthcoming 2026). ↩︎
- See Stephanos Bibas, The Machinery of Criminal Justice xix, 20–26 (2012) (explaining that criminal law changed when we hid some of it behind prison walls). ↩︎
- See Slobogin, supra note 137, at 552–53; see also Karl Menninger, The Crime of Punishment 28 (1968); Anthony C. Thompson, What Happens Behind Locked Doors: The Difficulty of Addressing and Eliminating Rape in Prison, 35 New Eng. J. on Crim. & Civ. Confinement 119, 165 (2009). ↩︎
- See John J. Donohue III, Assessing the Relative Benefits of Incarceration: Overall Changes and the Benefits on the Margin, in Do Prisons Make Us Safer? The Benefits and Costs of the Prison Boom 269, 307–09 (Steven Raphael & Michael A. Stoll eds., 2009). ↩︎
- See Patrick Sharkey, Gerard Torrats-Espinosa & Delaram Takyar, Community and the Crime Decline: The Causal Effect of Local Nonprofits on Violent Crime, 82 Am. Socio. Rev. 1214, 1219–20, 1234 (2017) (finding “strong evidence that establishment of community nonprofits had a substantively meaningful negative effect on murder, violent crime, and property crime” using data from 1990–2013); see also James J. Heckman & Dimitriy V. Masterov, The Productivity Argument for Investing in Young Children, 29 Rev. Agric. Econ. 446, 446–49, 454–59, 481 (2007) (demonstrating that early childhood education programs for disadvantaged children reduce crime); James Heckman, Rodrigo Pinto & Peter Savelyev, Understanding the Mechanisms Through Which an Influential Early Childhood Program Boosted Adult Outcomes, 103 Am. Econ. Rev. 2052, 2075–76, 2079–80 (2013) (finding significant drop in crime traced to participation in particular type of preschool program for students who could not afford it). ↩︎
- See Picard et al., supra note 117, at 25–26 (finding that harsher fines do not necessarily generate more revenue). Driver’s license suspension and incarceration are common penalties for nonpayment of LFOs, and they yield negative net revenue. Peter A. Jones, Ala. Appleseed, Measuring the Impact of Suspended Licenses on State Tax Revenue (2022), https://alabamaappleseed.org/wp-content/uploads/2024/08/DL-Fiscal-Analysis_Jones_01282022.pdf [https://perma.cc/L9NP-W3PB]. Individuals who cannot drive struggle to find work, earn income, and pay associated income and gas taxes. Id. ↩︎
- Aravind Boddupalli & Susan Nembhard, How Do Fines and Fees Affect Families’ Well-Being?, Urb. Inst.: Urb. Wire (Aug. 14, 2024), https://www.urban.org/urban-wire/how-do-fines-and-fees-affec-families-well-being [https://perma.cc/84PM-CRSZ]; Ala. Appleseed Ctr. for L. & Just., supra note 114, at 31–32. ↩︎
- A 2019 report by the Fines and Fees Justice Center “documented at least $27.6 billion in unpaid fines and fees across the nation,” a “gross understatement of the total court debt people are living with because half the states were unwilling or unable to provide information about the amount of outstanding court debt in their state.” Lillian Patil & Tanisha Pierrette, Fines & Fees Just. Ctr., Imposing Instability: How Court Fines and Fees Destabilize Government Budgets and Criminalize Those Who Cannot Pay 8 (2025), https://finesandfeesjusticecenter.org/wp-content/uploads/2025/07/Imposing-Instability_FinalPDF.pdf [https://perma.cc/GC9P-PT27]. ↩︎
- Matthew Menendez, Michael F. Crowley, Lauren-Brooke Eisen & Noah Atchison, Brennan Ctr. for Just., The Steep Costs of Criminal Justice Fees and Fines 5 (2019), https://www.brennancenter.org/our-work/research-reports/steep-costs-criminal-justice-fees-and-fines [https://perma.cc/K3ED-DEVQ]. ↩︎
- What Does It Cost the IRS to Collect Taxes?, USAFacts (Mar. 25, 2026), https://usafacts.org/articles/what-does-it-cost-the-irs-to-collect-taxes/ [https://perma.cc/EDS3-P7YX]. ↩︎
- Beth A. Colgan, The Excessive Fines Clause: Challenging the Modern Debtors’ Prison, 65 UCLA L. Rev. 2, 8 (2018). ↩︎
- Menendez et al., supra note 145, at 5. ↩︎
- Brian Nam-Sonenstein, Research Roundup: Evidence That a Single Day in Jail Causes Immediate and Long-Lasting Harms, Prison Pol’y Initiative (Aug. 6, 2024), https://www.prisonpolicy.org/blog/2024/08/06/short_jail_stays/ [https://perma.cc/8PHM-8KJD]. ↩︎
- Primus, supra note 97, at 121 (“More than 80% of American criminal defendants are indigent.”). ↩︎
- This quote comes from one of the authors’ field work, the findings of which have not been published. ↩︎
- Ala. Appleseed Ctr. for L. & Just., supra note 114, at 31–32. ↩︎
- Michael Ostermann, Nathan W. Link & Jordan M. Hyatt, Reframing the Debate on Legal Financial Obligations and Crime: How Accruing Monetary Sanctions Impacts Recidivism, 62 Criminology 331, 349–50 (2024); see also id. at 336–39 (recounting the literature about why LFOs increase recidivism). ↩︎
- See Schizer, supra note 20, at 477 (recognizing the informational challenges to making fiscally sound policy that legislatures face). ↩︎
- See Robert M.M. Shaffer, Unfunded State Mandates and Local Governments, 64 U. Cin. L. Rev. 1057, 1067 (1996); see also Barkow, supra note 54, at 775–93 (discussing sentencing commissions providing cost estimates for changes in sentencing law); Elizabeth Garrett, Harnessing Politics: The Dynamics of Offset Requirements in the Tax Legislative Process, 65 U. Chi. L. Rev. 501, 560 (1998) (recognizing that regulated entities can provide useful information about costs of implementation because of their on-the-ground experience). ↩︎
- See Barkow, supra note 54, at 775–93; Wright, supra note 68, at 55–80. ↩︎
- What Are Tax and Expenditure Limits?, Tax Pol’y Ctr. (Jan. 2024), https://taxpolicycenter.org/briefing-book/what-are-tax-and-expenditure-limits [https://perma.cc/Q4XG-RPEX]; see, e.g., Colo. Const. art. X, § 20(7)(a); S.C. Const. art. X, § 7(c). ↩︎
- Michael J. New, U.S. State Tax and Expenditure Limitations: A Comparative Political Analysis, 10 St. Pol. & Pol’y Q. 25, 29–30, 33–35 (2010); see also, e.g., Schizer, supra note 20, at 484. ↩︎
- E.g., S.C. Const. art. X, § 7(c); see New, supra note 158, at 30, 35; Schizer, supra note 20, at 484. ↩︎
- New, supra note 158, at 30. ↩︎
- Henning Bohn & Robert P. Inman, Balanced Budget Rules and Public Deficits: Evidence from the U.S. States 18–19 (Nat’l Bureau of Econ. Rsch., Working Paper No. 5533, 1996), http://www.nber.org/papers/w5533.pdf [https://perma.cc/GQ26-A7U5]. ↩︎
- Bierschbach & Bibas, supra note 21, at 220. ↩︎
- One of the important insights from the literature on budgetary offsets discussed below is that it is politically easier to cut from programs that benefit the needy rather than the wealthy. See Garrett, supra note 155, at 519. ↩︎
- See Petersilia et al., supra note 103, at 43. ↩︎
- Normatively there are reasons to like unifying the cost incidence and the decision-making at the local level. See Gold, supra note 20, at 888–95, 899–902. ↩︎
- See, e.g., W. David Ball, Defunding State Prisons, 50 Crim. L. Bull. 1060, 1079–80 (2014); Ball, supra note 105, at 82–84. ↩︎
- See New, supra note 158, at 30 (describing TELs that adjust when states devolve costs to local governments). ↩︎
- Garrett, supra note 155, at 503. ↩︎
- See id. ↩︎
- Id. at 510; Cheryl D. Block, Pathologies at the Intersection of the Budget and Tax Legislative Processes, 43 B.C. L. Rev. 863, 884 (2002); 2 U.S.C. § 902(a). ↩︎
- Garrett, supra note 155, at 527 (explaining PAYGO’s window of time as “five fiscal years”). ↩︎
- Id. at 503–04, 510. ↩︎
- See id. at 527. ↩︎
- Id. at 516. ↩︎
- Id. at 519. ↩︎
- See infra Section III.B. ↩︎
- Garrett, supra note 155, at 513; see also Block, supra note 170, at 885 (discussing the internal Senate PAYGO rule). ↩︎
- Garrett, supra note 155, at 511. ↩︎
- Id. at 505. ↩︎
- See id. at 506, 554–55; Schizer, supra note 20, at 483. ↩︎
- Garrett, supra note 155, at 560. ↩︎
- Wilhelm & Turner, supra note 50, at 47. ↩︎
- Id. ↩︎
- Va. Code Ann. § 30-19.1:4(H) (West 2026). ↩︎
- Wilhelm & Turner, supra note 50, at 47; Kim Hunt, Sentencing Commissions as Centers for Policy Analysis and Research: Illustrations from the Budget Process, 20 Law & Pol’y 465, 484 (1998); see Va. Code Ann. § 30-19.1:4(H) (requiring offsets “[f]or each law enacted which results in a net increase in periods of imprisonment in state correctional facilities or a net increase in periods of commitment or the time committed to the custody of the Department of Juvenile Justice”); Va. Code Ann. § 30-19.1:4(C) (requiring fiscal impact statements for “those bills which add new crimes for which imprisonment or commitment is authorized, increase the periods of imprisonment or commitment authorized for existing crimes, impose minimum or mandatory minimum terms of imprisonment or commitment, or modify the law governing release of prisoners or juveniles in such a way that the time served in prison, or the time committed to the custody of the Department of Juvenile Justice, will increase”). ↩︎
- Wilhelm & Turner, supra note 50, at 47. ↩︎
- Hunt, supra note 185, at 486. ↩︎
- Bunting, supra note 24, at 87. ↩︎
- See Jonathan J. Wroblewski, Sentencing Matters in the States, Sent’g Matters Substack (Jan. 12, 2026), https://sentencing.substack.com/p/sentencing-matters-in-the-states [https://perma.cc/9FU8-AGP6]. ↩︎
- Tenn. Code Ann. § 9-4-210(a) (West 2026). ↩︎
- E-mail from Joe Adams, Rsch. Coordinator, Pub. Affs. Rsch. Council of Ala., to authors (Dec. 8, 2025) [hereinafter Adams E-mail] (on file with authors). ↩︎
- Tenn. Code Ann. § 9-4-210(c); see also Fiscal Review Joint Committee, Tenn. Gen. Assemb., https://wapp.capitol.tn.gov/apps/CommitteeInfo/JointFiscalReview.aspx [https://perma.cc/4NJ5-8DDM]. Tennessee does not have a sentencing commission. Robina Inst. of Crim. L. & Crim. Just., Jurisdiction Profile: Tennessee 1 (2018), https://robinainstitute.umn.edu/sites/robinainstitute.umn.edu/files/2022-05/Tennessee%20SGRC%20Profile%202018.pdf [https://perma.cc/NK37-N3RK]. ↩︎
- Tenn. Code Ann. § 9-4-210(e)(1). ↩︎
- Id. § 9-4-210(b). ↩︎
- Colo. Rev. Stat. Ann. § 2-2-703 (West 2025), repealed by, S.B. 25-210, 2025 Colo. Sess. Laws 468, ch. 112. ↩︎
- Id. ↩︎
- H.B. 22-1330, 2022 Colo. Sess. Laws 485, ch. 104. ↩︎
- S.B. 25-210. ↩︎
- Neb. Rev. Stat. Ann. §§ 50-129–130 (West 1998) (repealed 1998). ↩︎
- Id. § 50-129. ↩︎
- Id. § 50-130. ↩︎
- State ex rel. Stenberg v. Moore, 544 N.W.2d 344, 347 (Neb. 1996); Bunting, supra note 24, at 69–73. Applying that rule to this context, where a second legislature could have simply repealed the fiscal responsibility measure, seems strange. ↩︎
- Bunting, supra note 24, at 45. ↩︎
- Id. ↩︎
- Wright, supra note 7, at 82. ↩︎
- Ganas, supra note 24, at 325. ↩︎
- Id. at 343–44. ↩︎
- Id. at 337–38. ↩︎
- Garrett, supra note 155, at 511. ↩︎
- Ganas, supra note 24, at 338–39. ↩︎
- See Tonry, supra note 85, at 156. ↩︎
- Id. at 157. ↩︎
- Id. ↩︎
- Wright, supra note 68, at 39–42; Amanda Bennett, Penal Code: State Tailors Sentences to Cost of Prison Space with Computer’s Help, Wall St. J., Aug. 5, 1994, at A1. ↩︎
- See Garrett, supra note 155, at 502 (“Better ‘mechanical’ rules lead to more appropriate tradeoffs among the competing claims on limited resources.”). ↩︎
- Supra notes 140–41 and accompanying text. ↩︎
- See supra Section III.A. ↩︎
- See, e.g., Garrett, supra note 155, at 515 (discussing the political risk of raising taxes); R. Douglas Arnold, The Logic of Congressional Action 193–94 (1990) (same). ↩︎
- See Schizer, supra note 20, at 483 (explaining that PAYGO “forc[es] Congress to make tough choices” about “cutting something old or raising taxes” to “do something new”). ↩︎
- Gold, supra note 20, at 843–50, 865–76. ↩︎
- We would have greater faith in the political process to make these difficult decisions if legislative districts were more representative than they are and we eliminated felon disenfranchisement and obstacles to voting by people who come from traditionally marginalized communities. Nonetheless, we think that solutions that require the wealthy and powerful to bear more of the costs of criminal legal systems create an interest convergence, of sorts. Id. at 869 & n.177. ↩︎
- We recognize that it likely makes political sense for legislators to avoid counting the most expensive items. ↩︎
- See, e.g., Mississippi DOJ Report, supra note 8, at 5 (describing “[g]ross understaffing, poor supervision, and inadequate investigations [that] create an environment where violent gang activity and dangerous contraband trafficking proliferate”); Alabama DOJ Report, supra note 8, at 5 (describing “prisons that are inadequately supervised, with inappropriate and unsafe housing designations, creating an environment rife with violence, extortion, drugs, and weapons”). ↩︎
- Although Alabama has allocated $1.3 billion to build a new prison that replaces existing facilities, even the $1.3 billion won’t alleviate overcrowding because it doesn’t increase capacity. Melody Schreiber, Alabama Plans to Use Covid Relief Funds to Finance Prison-Building Spree, Guardian (Sept. 29, 2021, at 05:00 ET), https://www.theguardian.com/us-news/2021/sep/29/alabama-covid-relief-funds-build-prisons [https://perma.cc/BLS2-M8XY]; Ralph Chapoco, Alabama Prison Closures, Key to New Construction Plan, Seem Far Off, Ala. Reflector (Feb. 9, 2026), https://alabamareflector.com/2026/02/09/alabama-prison-closings-key-to-new-construction-seem-far-off/ [https://perma.cc/RVD7-UH2W]. ↩︎
- For the extremely unusual instance of a court ordering a prison release, see Brown v. Plata, 563 U.S. 493, 502 (2011). ↩︎
- See Legis. Council Staff, Final Fiscal Note for H.B. 22-1330, 75th Gen. Assemb., 2d Reg. Sess. (Colo. 2022). ↩︎
- The Thirteenth Amendment exempts the criminal legal system from its ban on slavery and involuntary servitude in ways that are troublingly designed to save states money. Adam A. Davidson, No Exceptions: The New Movement to Abolish Slavery and Involuntary Servitude, 92 U. Chi. L. Rev. 2105, 2124, 2128 (2025) [hereinafter Davidson, No Exceptions]; Adam Davidson, Administrative Enslavement, 124 Colum. L. Rev. 633, 680–81 (2024); see also, e.g., N.C. Gen. Stat. Ann. § 148-26 (West 2026) (“Work assignments and employment shall be for the public benefit to reduce the cost of maintaining the inmate population . . . .”); Vt. Stat. Ann. tit. 28, § 751b(a) (West 2026) (mentioning cost savings as an explicit reason for work requirements for incarcerated people). ↩︎
- See Gold, supra note 20, at 879–80 (arguing the legislators ought to spend much more on prisons or incarcerate far fewer people because current arrangements offload costs that the government should bear onto incarcerated people—in the form of human suffering). ↩︎
- See Bunting, supra note 24, at 88 & n.185; Hunt, supra note 185, at 485. ↩︎
- Cf. Bunting, supra note 24, at 88 (arguing that fiscal analysts were “excessive[ly] reluctan[t] . . . to posit reasonable assumptions that would allow for a fiscal impact to be estimated”). A new criminal law might be unenforced, it might replace some enforcement of a previous law, or it might yield increased enforcement. Estimating costs of substantive criminal law is complicated by the fact that broader criminal codes can make policing and prosecution somewhat easier: Demonstrating probable cause is easier when more things are crimes, and securing guilty pleas can also become easier through charge bargaining leverage if criminal law gets broader or deeper. ↩︎
- We count 37 states with such trans-substantive requirements. Ala. Code § 29-5A-45 (West 2026); Alaska Stat. Ann. § 24.08.035 (West 2026); Ariz. Rev. Stat. Ann. § 41-1272 (West 2026); Colo. Rev. Stat. Ann. § 2-2-322 (West 2026); Conn. Gen. Stat. Ann. § 2-24 (West 2026); Del. Code Ann. tit. 29, § 911 (West 2026); S. Rule 3.13, in Fla. S., The Florida Senate Rules and Manual 2024–2026 (2024); Ga. Code Ann. § 28-5-42 (West 2026); J. Rule 18, in Idaho H. & S., Joint House and Senate Rules (2026); 25 Ill. Comp. Stat. Ann. 50/1 (West 2026); S. Rule VI(A)(40), in Ind. S., Standing Orders and Rules of the Senate (2026); J. S. & H.R. Rule 17, in 91st Gen. Assemb., Joint Rules of the Senate and House of Representatives (Iowa 2025); Kan. Stat. Ann. § 75-3715a (West 2026); J. Rule 4, in La. S. & H.R., Joint Rules of the Senate and House of Representatives (2026); Rule 312, in 132d Me. Legis., Joint Rules (2025); Md. Code Ann., State Gov’t § 2-1505 (West 2026); Mass. Gen. Laws Ann. ch. 3, § 38A (West 2026); J. Rule 20, in Miss. Legis., Joint Rules of the Senate and the House (2026); Miss. Code Ann. § 5-1-85 (West 2026); Mo. Ann. Stat. § 23.140 (West 2026); Mont. Code Ann. § 5-4-201 (West 2026); Rule 5 § 7, in 109th Legis., Rules of the Nebraska Unicameral Legislature (2025); Nev. Rev. Stat. Ann. § 218D.430 (West 2026); N.H. Rev. Stat. Ann. § 14:44 (West 2026); N.J. Stat. Ann. § 52:13B-6 (West 2026); J. Rule V(501), in 68th Legis. Assemb., St. of N.D., Senate and House Legislative Manual 2023–2024 (2023); Ohio Rev. Code Ann. § 103.14 (West 2026); Rule 8.7(c), in 60th Okla. Legis., House Rules (2025); Or. Rev. Stat. Ann. § 173.025 (West 2026); Rule 12(p)(1)(ii), in Pa. S., Rules of the Senate of Pennsylvania (2025); 22 R.I. Gen. Laws Ann. § 22-12-1 (West 2026); S.C. Code Ann. § 2-7-72 (West 2026); Tenn. Code Ann. § 3-2-107 (West 2026); Rule 4(33)(b), in 89th Legis., House Rules Manual (Tex. 2025); Utah Code Ann. § 36-12-13 (West 2026); Rule 95(a), in 87th Legis., Rules of the House (W. Va. 2025); Wis. Stat. Ann. § 13.093(2) (West 2026); Wyo. Stat. Ann. § 28-8-105(d) (West 2026); see also Minn. Stat. Ann. § 3.98 (West 2026) (providing for fiscal notes upon request of a legislator); id. § 3.8853 (same); Wash. Rev. Code Ann. § 43.88A.040 (West 2026) (same); N.M. Stat. Ann. § 2-5-4.1 (West 2026) (requiring that the state create a process for assessing fiscal impact of pending legislation). For another synopsis of the requirements for fiscal notes, see Council of State Gov’ts, The Book of the States tbl. 3.18 (2023), https://bookofthestates.org/tables/2023-3-18/ [https://perma.cc/B9L3-D8PL]. ↩︎
- Ga. Code Ann. § 28-5-42 (West 2026) (limited to a “bill having a significant impact on the . . . expenditure level” of any state entity); J. S. & H.R. Rule 17 (Iowa 2025) (“A fiscal note shall be attached to any bill or joint resolution which reasonably could have an annual effect of at least five hundred thousand dollars or a combined total effect within five years after enactment of two million five hundred thousand dollars or more . . . .”); J. Rule 4 (La. 2026) (containing a $100,000 minimum); Mass. Gen. Laws Ann. ch. 3, § 38A ($100,000 minimum); Nev. Rev. Stat. Ann. § 218D.430 ($2,000 minimum unless the law will increase imprisonment in which case no minimum applies); N.H. Rev. Stat. Ann. § 14:44 ($10,000 minimum); J. Rule V(501) (N.D. 2023) ($10,000 minimum). ↩︎
- See, e.g., Ark. Code Ann. § 10-3-702 (West 2026); Cal. Code Regs. tit. 1, § 2002 (West 2026); Mich. Comp. Laws Ann. § 24.245 (West 2026); N.Y. Legis. Law § 50 (McKinney 2026). ↩︎
- See, e.g., Eric Humphrey, Note, Calvinball in Cole County: State ex rel. Fitz-James v. Bailey, 89 Mo. L. Rev. 989, 990 & n.9 (2024) (citing Mo. Ann. Stat. § 116.175 (West 2026)); S.D. Codified Laws § 2-9-30 (West 2026); Wyo. Stat. Ann. § 22-24-309(a) (West 2026). Bond measures similarly tee up the cost of a particular initiative directly to voters. See, e.g., Wright, supra note 68, at 58 (discussing an important bond vote in criminal law reform in North Carolina). ↩︎
- Shaffer, supra note 155, at 1066–68 (“As of 1992, twenty-eight states had adopted fiscal note requirements.”); see Edward A. Zelinsky, Unfunded Mandates, Hidden Taxation, and the Tenth Amendment: On Public Choice, Public Interest, and Public Services, 46 Vand. L. Rev. 1355, 1366–67 (1993). ↩︎
- See Carrington Skinner, To Get Better Legislation, Start with Reliable Data, Nat’l Conf. State Legs., (Oct. 13, 2025), https://www.ncsl.org/state-legislatures-news/details/to-get-better-legislation-start-with-reliable-data [https://perma.cc/K98Z-LL4X]. ↩︎
- See Shaffer, supra note 155, at 1065–66 (discussing these ideas in the context of expenses imposed upon local governments); Philip Joyce, Evaluating the Impact of the Congressional Budget Office at Middle Age, 43 Cong. & Presidency 279, 289–90 (2016) (recounting, as a criticism of the Congressional Budget Office, that it focuses discussion too much on costs to the federal budget rather than on legislation’s benefits). ↩︎
- See Shaffer, supra note 155, at 1066; Janet M. Kelly, Fiscal Noting Reconsidered: The Experience of the States with Mandate Cost Estimation, 6 Pub. Budgeting & Fin. Mgmt. 1, 9 (1994). ↩︎
- See Michael Leachman, Inimai M. Chettiar & Benjamin Geare, Ctr. on Budget & Pol’y Priorities, Improving Budget Analysis of State Criminal Justice Reforms: A Strategy for Better Outcomes and Saving Money 11–12 (2012), https://www.aclu.org/wp-content/uploads/document/improvingbudgetanalysis_20120110.pdf [https://perma.cc/5BZU-3FJV]; Kelly, supra note 238, at 12; Shaffer, supra note 155, at 1067. ↩︎
- See Zelinsky, supra note 235, at 1366–67; Roderick M. Hills, Jr., The Political Economy of Cooperative Federalism: Why State Autonomy Makes Sense and “Dual Sovereignty” Doesn’t, 96 Mich. L. Rev. 813, 936 n.407 (1998). ↩︎
- Kelly, supra note 238, at 2. ↩︎
- See id. at 13 (focusing in particular on notes estimating costs imposed on local governments). ↩︎
- See Skinner, supra note 236. ↩︎
- See also Leachman, Chettiar & Geare, supra note 239, at 7 & n.21 (explaining that as of 2012, fourteen states had special requirements for fiscal notes on criminal justice bills). ↩︎
- Colo. Rev. Stat. Ann. § 2-2-322(2.5) (West 2026); Rule 52, in Ky. Gen. Assemb., Rules of the Senate (2024); N.C. Gen. Stat. Ann. § 120-36.7(d) (West 2026); Or. Rev. Stat. Ann. § 173.029 (West 2026); S.C. Code Ann. § 2-7-74 (West 2026); Tenn. Code Ann. § 9-4-210(c) (West 2026); Va. Code Ann. § 30-19.1:4(C) (West 2026); see also Miss. Code Ann. § 47-5-39(2) (West 2026) (requiring a fiscal note for laws that establish a new criminal offense or change sentencing for an existing offense, if a legislator so requests). ↩︎
- J. Rule 4, in La. S. & H.R., Joint Rules of the Senate and House of Representatives (2026); Mo. Ann. Stat. § 217.022 (West 2026); Nev. Rev. Stat. Ann. § 218D.430(1)(b) (West 2026); N.H. Rev. Stat. Ann. § 14:46(VI) (West 2026); Tenn. Code Ann. § 9-4-210; see also S.D. Codified Laws § 2-9-30 (West 2026) (encompassing proposed initiatives and constitutional amendments but not ordinary legislation). ↩︎
- Alaska Stat. Ann. § 24.08.035 (West 2026); Ariz. Rev. Stat. Ann. § 41-1272 (West 2026); Conn. Gen. Stat. Ann. § 2-24 (West 2026); Del. Code Ann. tit. 29, § 911 (West 2026); Ga. Code Ann. § 28-5-42 (West 2026); J. Rule 18, in Idaho H. & S., Joint House and Senate Rules (2026); 25 Ill. Comp. Stat. Ann. 50/1, 50/4 (West 2026); S. Rule VI(A)(40), in Ind. S., Standing Orders and Rules of the Senate (2026); J. S. & H.R. Rule 17, in 91st Gen. Assemb., Joint Rules of the Senate and House of Representatives (Iowa 2025); Kan. Stat. Ann. § 75-3715(a) (West 2026); Rule 312, in 132d Me. Legis., Joint Rules (2025); Md. Code Ann., State Gov’t § 2-1505 (West 2026); Mass. Gen. Laws Ann. ch. 3, § 38A (West 2026); Minn. Stat. Ann. § 3.98 (West 2026); Mont. Code Ann. § 5-4-201 (West 2026); N.J. Stat. Ann. § 52:13B-6 (West 2026); J. Rule V(501), in 68th Legis. Assemb., St. of N.D., Senate and House Legislative Manual 2023–2024 (2023); Ohio Rev. Code Ann. § 103.14 (West 2026); Rule 8.7(c), in 60th Okla. Legis., House Rules (2025); Rule 12(p)(1)(ii), in Pa. S., Rules of the Senate of Pennsylvania (2025); 22 R.I. Gen. Laws Ann. § 22-12-1 (West 2026); Tex. Gov’t Code Ann. §§ 314.001–.002 (West 2026); Utah Code Ann. § 36-12-13 (West 2026); H. Rule 95(a), in 87th Legis., Rules of the House (W.Va. 2025); Wis. Stat. Ann. § 13.093(2) (West 2026); Wyo. Stat. Ann. § 28-8-105(d) (West 2026). ↩︎
- See, e.g., N.C. Gen. Stat. Ann. §§ 164-42(d), 164-43(h) (West 2026); id. § 120-36.7(d); Or. Rev. Stat. Ann. § 173.029; Va. Code Ann. § 30-19.1:4. While somewhat less detailed than others, New Hampshire’s very brief fiscal note provision does require calculating the costs of bills upon prosecution, incarceration, probation, and parole. N.H. Rev. Stat. Ann. § 14:46(VI). ↩︎
- N.C. Gen. Stat. Ann. § 120-36.7(d). ↩︎
- Id. ↩︎
- See Wright, supra note 68, at 81. ↩︎
- Id. at 83; Bennett, supra note 214. ↩︎
- Supra Part II. ↩︎
- See Leachman, Chettiar & Geare, supra note 239, at 21. ↩︎
- See Or. Rev. Stat. Ann. § 173.029 (West 2026); J. Rule 4, in La. S. & H.R., Joint Rules of the Senate and House of Representatives (2026) (so providing, albeit subject to exception if a committee decides not to calculate the fiscal note); N.H. Rev. Stat. Ann. § 14:46(VI) (West 2026); see also Rule 52, in Ky. Gen. Assemb., Rules of the Senate (2026) (requiring fiscal notes for bills that “affect state or local corrections services in a significant manner” (emphasis added)). ↩︎
- See, e.g., Ariz. Rev. Stat. Ann. § 41-1272 (West 2026); Colo. Rev. Stat. Ann. § 29-1-304.9 (West 2026); J. Rule 18, in Idaho H. & S., Joint House and Senate Rules (2026); Kan. Stat. Ann. § 75-3715a (West 2026); Ky. Rev. Stat. Ann. § 6.955(1) (West 2026); Md. Code Ann., State Gov’t § 2-1505(e)(1) (West 2026); Mo. Ann. Stat. § 23.140(2)(4) (West 2026); N.C. Gen. Stat. Ann. § 120-30.47 (West 2026); S.C. Code Ann. § 2-7-76 (West 2026); Tenn. Code Ann. § 9-4-210 (West 2026); Va. Code Ann. § 30-19.03 (West 2026); see also N.C. Gen. Stat. Ann. § 164-42(d) (West 2026) (requiring the state’s sentencing commission to consider the effect of the sentencing guidelines structure on local jails). ↩︎
- Mo. Ann. Stat. § 217.022 (West 2026). ↩︎
- Colo. Rev. Stat. Ann. § 2-2-322 (West 2026). ↩︎
- Stuntz, supra note 2, at 512–19. ↩︎
- See Conn. Gen. Stat. Ann. § 2-24 (West 2026); Minn. Stat. Ann. § 3.98 (West 2026); Miss. Code Ann. § 5-1-85 (West 2026); Nev. Rev. Stat. Ann. § 218D.430(1)(b) (West 2026); N.C. Gen. Stat. Ann. § 120-36.7(d) (West 2026); Or. Rev. Stat. Ann. § 173.029 (West 2026); S.C. Code Ann. § 2-7-74(B) (West 2026); Va. Code Ann. § 30-19.1:4 (West 2026); see also S. Rule 12(p)(1)(ii), in Pa. S., Rules of the Senate of Pennsylvania (2025); Iowa Code Ann. § 25B.5 (West 2026) (requiring fiscal notes for any bills that contain a state mandate). ↩︎
- Alaska Stat. Ann. § 24.08.035(a) (West 2026); Colo. Rev. Stat. Ann. § 2-2-322(2.5); 25 Ill. Comp. Stat. Ann. 50/1(1) (West 2026); Kan. Stat. Ann. § 75-3715a; Rule 52, in Ky. Gen. Assemb., Rules of the Senate (2026); Mo. Ann. Stat. § 217.022; Mont. Code Ann. § 5-4-201 (West 2026); N.H. Rev. Stat. Ann. § 14:46(VI); N.J. Stat. Ann. § 52:13B-6 (West 2026); J. Rule V(501), in 68th Legis. Assemb., St. of N.D., Senate and House Legislative Manual 2023–2024 (2023); Ohio Rev. Code Ann. § 103.14 (West 2026); H. Rule 8.7(c), in 60th Okla. Legis., House Rules (2025); 22 R.I. Gen. Laws Ann. § 22-12-1 (West 2026); Tex. Gov’t Code Ann. § 314.001 (West 2026); Utah Code Ann. § 36-12-13(2)(c) (West 2026); H. Rule 95(a), in 87th Legis., Rules of the House (W. Va. 2025); Wis. Stat. Ann. § 13.093(2)(a) (West 2026); Wyo. Stat. Ann. § 28-8-105(d) (West 2026); see also Ariz. Rev. Stat. Ann. § 41-1272 (not mentioning increases or decreases in expenditures); Ga. Code Ann. § 28-5-42 (West 2026) (addressing “[a]ny bill having a significant impact on the anticipated revenue or expenditure level”). ↩︎
- But see Leachman, Chettiar & Geare, supra note 239, at 7 (arguing that calculating savings is equally important to calculating increased expenditures). ↩︎
- See id. at 9–10. ↩︎
- Id. at 2 & n.2. ↩︎
- Id. at 10. ↩︎
- Del. Code Ann. tit. 29, § 911(a)(2)(b) (West 2026). ↩︎
- Leachman, Chettiar & Geare, supra note 239, at 2, 10. ↩︎
- Id. at 10. ↩︎
- Id. ↩︎
- Zoom Interview with Joe Adams, Rsch. Coordinator, Pub. Affs. Rsch. Council of Ala. (Dec. 10, 2025). ↩︎
- Adams E-mail, supra note 191; Christy Hoppe, Pardon My Fiscal Note, Your Bill Is Being Killed, Dall. Morning News, Mar. 9, 2001, at 33A (“An expensive fiscal note is the best, fastest and easiest way to kill a bill.”). ↩︎
- See, e.g., Akhil Reed Amar, Fourth Amendment First Principles, 107 Harv. L. Rev. 757, 799 (1994). ↩︎
- Leachman, Chettiar & Geare, supra note 239, at 2, 16. ↩︎
- See id. at 11–12 (arguing that independent analysis in state legislative branches are best positioned to create trusted, non-partisan fiscal notes). ↩︎
- Barkow, supra note 48, at 1289 n.63; Barkow, supra note 54, at 812–14; see also Leachman, Chettiar & Geare, supra note 239, at 3 (indicating that best practices for fiscal notes require that they be “[p]roduced by a source that is trusted, non-partisan, and adequately resourced”). Although we are quite sanguine about the role that sentencing commissions can play in facilitating more fiscally constrained criminal law, they cannot play that role effectively when the legislature is not interested in constraining its costs. In this way, the overwhelming majority of states that have to balance their budgets are quite different from the federal government. See Reitz & Klingele, supra note 10, at 276. ↩︎
- Erin Reynolds, Aaron Carpenter, John Armstrong & Clayton Mayfield, Colo. Legis. Council Staff, Fiscal Notes Considerations for Bills with Criminal Justice System Impacts 6–7 (2024), https://leg.colorado.gov/publications/fiscal-notes-considerations-criminal-justice-bills [https://perma.cc/V33W-BVBW]. ↩︎
- See supra Part II. ↩︎
- See also Hunt, supra note 185, at 485 (arguing that fiscal impact statements for proposed “new crime[s] are rarely doable”); Bunting, supra note 24, at 88 & n.185 (recognizing a similar challenge). Calculating the costs of a bill that has already been in effect when it comes up for renewal is an easier task. See infra Section III.E (discussing sunset provisions as a way to force legislative debate after a law has been in effect). ↩︎
- Cf. Leachman, Chettiar & Geare, supra note 239, at 7–9 (arguing that fiscal notes can help restrain criminal law legislative spending). ↩︎
- Colo. Rev. Stat. Ann. § 2-2-322(2.5)(a)–(d) (West 2026). ↩︎
- Stuntz, supra note 2, at 512–19. Deep criminal codes with potential penalties that even legislators may think are too harsh can, as designed, facilitate quicker guilty pleas that require less time from prosecutors on each case. Russell M. Gold, Carissa Byrne Hessick & F. Andrew Hessick, Civilizing Criminal Settlements, 97 B.U. L. Rev. 1607, 1617–20 (2017); Gerard E. Lynch, Comment, Screening Versus Plea Bargaining: Exactly What Are We Trading Off?, 55 Stan. L. Rev. 1399, 1401–02 (2003). Although that may seem like a cost-saving measure, freeing up prosecutors’ time will allow them to charge more cases and increase prison costs. ↩︎
- Colorado fiscal notes are also required to assess the likely effects on groups defined by race and gender—in considering both who the offenders will be and who the affected victims will be. Colo. Rev. Stat. Ann. § 2-2-322(2.5)(e). ↩︎
- Ala. Const. art. IV, § 111.05; Cal. Const. art. 13B, § 6; Colo. Rev. Stat. Ann. § 29-1-304.5 (West 2026); Fla. Const. art. VII, § 18; Haw. Const. art. VIII, § 5; Iowa Code Ann. § 25B.2 (West 2026); La. Const. art. 6, § 14; Me. Const. art. IX, § 21; Mass. Gen. Laws Ann. ch. 29, § 27C (West 2026); Mich. Const. art. 9, § 29; Mo. Const. art. X, § 21; Mont. Code Ann. § 1-2-112 (West 2026); N.H. Rev. Stat. Ann. § 541-A:25 (West 2026); N.J. Const. art. VIII, § 2, ¶ 5; Or. Const. art. XI, § 15; S.C. Code Ann. § 4-9-55 (West 2026); S.D. Codified Laws § 6-15-1 (West 2026); Tenn. Const. art. II, § 24; Wash. Rev. Code Ann. § 43.135.060 (West 2026); see also Shaffer, supra note 155, at 1057 (explaining the prevalence of unfunded mandates, although counting seventeen states that require state reimbursement of localities). Most of the statutes require the state to cover the cost completely, but some don’t. Iowa allows mandates that are partially funded by the state so long as the statute itself specifies the portion of the cost the state will pay annually. Iowa Code Ann. § 25B.2(3)(a). Maine allows mandates so long as they’re funded at 90 percent or greater by the State. Me. Const. art. IX, § 21. Tennessee requires only “that the state share in the cost” and that the legislation so specify. Tenn. Const. art. II, § 24. Rhode Island is not included in this count because, while it requires the State to reimburse local governments, it neither styles its law as a mandate nor specifies the remedy if the State fails to reimburse local governments. 45 R.I. Gen. Laws Ann. § 45-13-9 (West 2026). Neither was Wisconsin included, although it allows localities to obtain waivers from unfunded mandates, subject to exceptions for legislation related to health or safety. Wis. Stat. Ann. § 66.0143 (West 2026). Because Minnesota’s legislature can choose whether a law invokes its unfunded mandate requirements, it too was excluded from the count. Minn. Stat. Ann. § 3.989 (West 2026). ↩︎
- Ala. Const. art. IV, § 111.05(a) (“require a new or increased expenditure”); Cal. Const. art. 13B, § 6(a) (“mandates a new program or higher level of service”); Colo. Rev. Stat. Ann. § 29-1-304.5(1) (“No new state mandate”); Fla. Const. art. VII, § 18(a) (“requiring such county or municipality to spend funds”); Haw. Const. art. VIII, § 5 (“shall be mandated”); Iowa Code Ann. § 25B.2(3) (“requires a political subdivision”); La. Const. art. 6, § 14(A)(1) (“requiring increased expenditures”); Me. Const. art. IX, § 21 (“require a local unit of government”); Mich. Const. art. 9, § 29 (“shall not be required”); Mo. Const. art. X, § 21(1) (“shall not be required”); Mont. Code Ann. § 1-2-112(1) (“requires a local government unit”); N.H. Rev. Stat. Ann. § 541-A:25(I) (“shall not mandate or assign any new, expanded, or modified programs or responsibilities to any political subdivision in such a way as to necessitate further expenditures by the political subdivision”); N.J. Const. art. VIII, § 2, ¶ 5(a) (making not mandatory “an unfunded mandate” that “does not authorize resources, other than the property tax, to offset the additional direct expenditures required for the implementation of the law”); Or. Const. art. XI, § 15(1) (“requires any local government”); S.C. Code Ann. § 4-9-55(A) (“requiring it to spend funds”); S.D. Codified Laws § 6-15-1 (“mandates any county, municipality, or school district”); Tenn. Const. art. II, § 24 (“increased expenditure requirements on cities or counties”). ↩︎
- Mass. Gen. Laws Ann. ch. 29, § 27C(a) (“imposing any direct service or cost obligation”); Nev. Rev. Stat. Ann. § 218D.270 (West 2026) (“requiring one or more local governments”); Wash. Rev. Code Ann. § 43.135.060(1) (“impose responsibility”). A public choice story of unfunded mandates describes them as “an attractive device by which legislators advancing their own political interests opportunistically dispense public largesse to importuning constituencies while deflecting to officeholders at lower levels of government the political costs of taxing to pay for that largesse.” Edward A. Zelinsky, The Unsolved Problem of the Unfunded Mandate, 23 Ohio N.U. L. Rev. 741, 742 (1997). ↩︎
- Mass. Gen. Laws Ann. ch. 29, § 27C(a). ↩︎
- N.J. Const. art. VIII, § 2, ¶ 5(a). ↩︎
- Mont. Code Ann. § 1-2-112(1). ↩︎
- Mich. Const. art. 9, § 29; Mo. Const. art. X, § 21(1). ↩︎
- S.C. Code Ann. § 4-9-55(A) (West 2026). ↩︎
- Ala. Const. art. IV, § 111.05(a); see also Fla. Const. art. VII, § 18(a) (covering any law “requiring such county or municipality to spend funds or to take an action requiring the expenditure of funds”); La. Const. art. 6, § 14(A)(1) (reaching any “law or state executive order, rule, or regulation requiring increased expenditures for any purpose”). ↩︎
- Stuntz, supra note 34, at 2549 (invoking the “menu” metaphor for prosecutors’ charging decisions); see also Colo. Rev. Stat. Ann. § 29-1-304.5(2)(e) (West 2026) (exempting “[a]ny new state mandate or any increase in the level of service for an existing state mandate beyond the existing level of service which is undertaken at the option of a local government” (emphasis added)). ↩︎
- Robert E. Scott & William J. Stuntz, Plea Bargaining as Contract, 101 Yale L.J. 1909, 1962 (1992) (discussing breadth of prosecutorial discretion). Local governments will likely be less motivated to avoid increased prison expenditures because those fall back onto the state. See Ouss, supra note 96, at 1. ↩︎
- Several states allow legislatures to enact unfunded mandates only if they pass by a supermajority vote. Ala. Const. art. IV, § 111.05(b)(5); Fla. Const. art. VII, § 18(a); La. Const. art. 6, § 14(A)(2)(f); Me. Const. art. IX, § 21; N.J. Const. art. VIII, § 2, ¶ 5(c)(6); Or. Const. art. XI, § 15(7)(a); S.C. Code Ann. § 4-9-55(A). ↩︎
- Ala. Const. art. IV, § 111.05(b)(2); Cal. Const. art. 13B, § 6(a)(2); Fla. Const. art. VII, § 18(d); La. Const. art. 6, § 14(A)(2)(b); Or. Const. art. XI, § 15(7)(b); S.C. Code Ann. §§ 4-9-55(C)(3), (7); S.D. Codified Laws § 6-15-2(5) (West 2026). Illinois’s law on unfunded mandates is narrower—reaching only certain categories of legislation that do not include criminal law. See 30 Ill. Comp. Stat. Ann. 805/3, 6 (West 2026). ↩︎
- Ala. Const. art. IV, § 111.05(b)(2); Cal. Const. art. 13B, § 6(a)(2); La. Const. art. 6, § 14(A)(2)(b); S.D. Codified Laws § 6-15-2(5). ↩︎
- Fla. Const. art. VII, § 18(d); S.C. Code Ann. §§ 4-9-55(C)(3), (7) (listing “criminal laws” and “laws creating, modifying, or repealing noncriminal infractions”). ↩︎
- Or. Const. art. XI, § 15(7)(b). ↩︎
- See Shaffer, supra note 155, at 1062 (recognizing that many statutes are not mandates because they provide local government with authority but not an obligation). ↩︎
- See City of Worcester v. Governor, 625 N.E.2d 1337, 1338–40 (Mass. 1994) (holding that none of the challenged laws relating to education were unfunded mandates because they did not impose “a direct service or cost obligation” or were “incidental local administrative expenses”); Kennedy v. Commonwealth, 92 N.E.3d 1225, 1231–32 (Mass. App. Ct. 2018) (rejecting a claim that laws imposing requirements on cities that withdraw from school districts constitute local mandates because any costs imposed are indirect and speculative); Mich. Comp. Laws Ann. § 21.234(5) (defining “state requirement” as “a state law which requires a new activity or service or an increased level of activity or service beyond that required of a local unit of government by an existing law”); Jud. Att’ys Ass’n v. State, 597 N.W.2d 113, 116 (Mich. 1999) (explaining that the unfunded mandate provision applies only when “when the state mandates a program” (emphasis added)); City of Concord v. State, 53 A.3d 576, 584 (N.H. 2012) (unfunded mandate prohibition applies only when a law “unequivocally require[s]” local government to spend money); City of Seattle v. State, 666 P.2d 359, 362–64 (Wash. 1983) (holding that requiring electronic recording equipment in courtrooms was “neither part of a new program nor an increased level of services under an existing program” but simply a change to the internal workings of the municipal courts); cf. Gessler v. Doty, 272 P.3d 1131, 1131–33, 1135 (Colo. App. 2012) (holding that a statute requiring local government to bear costs of election administration prevailed over the conflicting unfunded mandate provision); Romer v. Bd. of Cnty. Comm’rs, 897 P.2d 779, 780–82 (Colo. 1995) (finding that local government’s contribution to social services programs is not a subsidy from which the county can exempt itself under a somewhat similar statute); Breitenfeld v. Sch. Dist. of Clayton, 399 S.W.3d 816, 826–32 (Mo. 2013) (holding that law allowing students to enroll outside their home school district did not impose an unfunded mandate because school districts already had the obligation to provide a free public education to all students who attend); Prosecution by the Dekalb Cnty. Prosecuting Att’y, No. 108-90 (Mo. Att’y Gen. July 26, 1990) (explaining that putting a new prison into a county does not impose an unfunded mandate on the prosecutors’ office to prosecute crimes committed in that prison insofar as it is not a new requirement but rather reflects an increase in crime). ↩︎
- Morris v. Snodgrass, 886 S.W.2d 761, 762–63 (Tenn. Ct. App. 1994). ↩︎
- Maint. of Certain Pub. Sts. and Highways, No. 86-15 (Haw. Att’y Gen. June 10, 1986); see Mont. Code Ann. § 1-2-112(1) (West 2026) (reaching only expenditures “that [are] not expected of local governments in the scope of their usual operations”). ↩︎
- McGuire v. County of Hawai’i, 567 P.3d 180, 185–89 (Haw. 2025). There is not sufficient case law nor an attorney general opinion to determine how New Jersey’s Council on Local Mandates construes its unfunded mandate provision. See N.J. Stat. Ann. § 52:13H-18 (West 2026); see also Deptford Twp., No. COLM-0003-15 (N.J. Council on Loc. Mandates Apr. 20, 2016), https://www.nj.gov/localmandates/decisions/Deptford-colm-0003-15.shtml [https://perma.cc/M3PR-9SCG] (finding that a law requiring municipalities to equip police vehicles with video recording systems was not adequately funded); N.J. Ass’n of Cntys., No. COLM-0004-16 (N.J. Council on Loc. Mandates Apr. 26, 2017), https://www.nj.gov/localmandates/decisions/NJAC-COLM-0004-16.shtml [https://perma.cc/LYQ5-32ZJ] (finding that a criminal procedure law was exempt from the unfunded mandate restriction because it effectuated a state constitutional amendment). ↩︎
- Mass. Gen. Laws Ann. ch. 29, § 27C(a) (reaching “direct service or cost obligation”); Mont. Code Ann. § 1-2-112(1) (reaching “direct expenditure of additional funds” and that local government perform an activity “that is not expected of local governments in the scope of their usual operations”); N.J. Const. art. VIII, § 2, ¶ 5(a) (reaching “direct expenditures required” by new legislation). ↩︎
- That states exempt substantive criminal law from their bans on unfunded mandates undermines the force of criticisms about prosecutorial nonenforcement policies as violating separation of powers. See, e.g., Ayala v. Scott, 224 So. 3d 755, 758–60 (Fla. 2017). Those states’ laws don’t view new substantive criminal laws as requiring anything from local governments. ↩︎
- Jacob E. Gersen, Temporary Legislation, 74 U. Chi. L. Rev. 247, 266–72 (2007). ↩︎
- Id. at 249–55. ↩︎
- Id. at 255–57. ↩︎
- Id. at 251 (explaining and quoting The Federalist No. 26, at 164, 168 (Alexander Hamilton) (Jacob E. Cooke ed., 1961)). ↩︎
- Id. at 261; Rebecca M. Kysar, Lasting Legislation, 159 U. Pa. L. Rev. 1007, 1041 (2011). ↩︎
- Gersen, supra note 306, at 261. ↩︎
- Id. at 275; see also Kysar, supra note 310, at 1013 (“[E]ven if accurate and complete cost information regarding temporary legislation is available, there is no guarantee that the estimate will inform Congress’s decisionmaking.”); Zachary J. Gubler, Experimental Rules, 55 B.C. L. Rev. 129, 130 (2014) (explaining that effective policy experimentation requires “the lawmaking process [to] have the capacity for ‘learning’”). ↩︎
- See Gersen, supra note 306, at 270. One scholar argues that Congress should prefer lasting legislation over temporary legislation—a preference that should be strongest for tax cuts and weakest for legislation that responds to a crisis. Kysar, supra note 310, at 1066–67. Some of that critique turns on the particulars of congressional budget rules and cost estimation. Id. at 1021–41. ↩︎
- Compare Kysar, supra note 310, at 1051–56, with Gersen, supra note 306, at 285–86. ↩︎
- Richard E. Myers II, Responding to the Time-Based Failures of the Criminal Law Through a Criminal Sunset Amendment, 49 B.C. L. Rev. 1327, 1356–68 (2008) [hereinafter Myers, Responding to Time-Based Failures]; Richard E. Myers II, Complex Times Don’t Call for Complex Crimes, 89 N.C. L. Rev. 1849, 1874 (2011) [hereinafter Myers, Complex Times]. ↩︎
- See, e.g., Melissa J. Mitchell, Comment, Cleaning Out the Closet: Using Sunset Provisions to Clean Up Cluttered Criminal Codes, 54 Emory L.J. 1671, 1696–99 (2005); Ryan S. Boyce, Recognizing and Remediating Overcriminalization in North Carolina: Solutions for a Modern Criminal Code, 15 Wake Forest J.L. & Pol’y 1, 22 (2024); Dawinder S. Sidhu & Kelsey Robinson, Child Pornography and Criminal Justice Reform, 43 Cardozo L. Rev. 2157, 2193–94 (2022) (proposing sunsets for statutory penalty commands in federal sentencing). ↩︎
- Stephanie Watson, Fixing California Sentencing Law—The Problem with Piecemeal Reform, 39 McGeorge L. Rev. 585, 595–96 (2008). ↩︎
- Id. at 599. ↩︎
- Cal. Penal Code § 1170(b)(1)–(2) (West 2026). Compare id. (requiring stipulation or proof beyond a reasonable doubt to enhance sentence), with Cal. Penal Code § 1170(b) (West 2021) (providing widespread sentencing discretion). ↩︎
- Peter Westen, Lex Mitior: Converse of Ex Post Facto and Window into Criminal Desert, 18 New Crim. L. Rev. 167, 186 (2015); People v. Pedro T. (In re Pedro T.), 884 P.2d 1022, 1023 (Cal. 1994). ↩︎
- See Myers, Responding to Time-Based Failures, supra note 315, at 1335 (arguing that legislative biases toward inaction make it difficult to “maintain widespread public faith that the criminal law accurately represents current public sentiment”). ↩︎
- Id. at 1329, 1335; Johnson, supra note 3, at 108–10. ↩︎
- Myers, Responding to Time-Based Failures, supra note 315, at 1337. ↩︎
- See Mitchell, supra note 316, at 1699–700 (explaining that “sunset provisions . . . avoid[] the political repercussions that may stem from a single legislator standing up and actively voicing his or her support for the removal of” an outdated criminal law). ↩︎
- Sidhu & Robinson, supra note 316, at 2194. ↩︎
- Gersen, supra note 306, at 248. ↩︎
- See Barkow, supra note 29, at 3; Andrea R. Barter, The Practical Realities of Melanie’s Law, Mass. Bar Ass’n (Jan. 2006), https://www.massbar.org/publications/ejournal/ejournal-article/lawyers-journal-2006-january/the-practical-realities-of-melanie-s-law- [https://perma.cc/5DJY-RT6F]. ↩︎
- Christopher Goffard, A 12-Year–Old Girl’s Murder Shook the Country, Inspiring Far–Reaching Laws, L.A. Times (Oct. 30, 2024, at 03:00 PT), https://www.latimes.com/california/story/2024-10-30/polly-klaas-murder-shook-the-country-inspiring-far-reaching-laws [https://perma.cc/5BHF-QFYN]. ↩︎
- Id. ↩︎
- Franklin & Kim, supra note 32. ↩︎
- Id.; Ala. Code § 13A-10-9(b) (West 2026) (codifying Act 2024-345). ↩︎
- Tradesha Woodard & Nathaniel Puente, NC Lawmakers Pass ‘Iryna’s Law,’ Sending Legislation to Gov. Stein’s Desk, WCNC Charlotte (Sept. 24, 2025, at 07:38 ET), https://www.wcnc.com/article/news/crime/nc-lawmakers-introduce-irynas-law-deadly-charlotte-stabbing/275-1fc96400-cad1-4662-8c9b-e066cd61e369 [https://perma.cc/V79X-FQSC]; Hank Lee, ‘Iryna’s Law’: Stein Signs Criminal Justice Bill, Potentially Restarting Death Penalty After Charlotte Stabbing, WCNC Charlotte (Oct. 3, 2025, at 15:06 ET), https://www.wcnc.com/article/news/crime/irynas-law-signed-josh-stein-criminal-justice-reform-death-penalty/275-ecf416ac-6c93-4762-bd2c-2e11dd7d615a [https://perma.cc/E2TB-WWN5]; N.C. Gen. Stat. Ann. § 15A-533 (West 2026) (codifying H.B. 307). ↩︎
- See Myers, Responding to Time-Based Failures, supra note 315, at 1378–79 (arguing that sunset laws help legislatures address “criminal laws that were passed in exigent circumstances” by forcing them to reconsider once “passions have cooled”); Michael M. O’Hear, Perpetual Panic, 21 Fed. Sent’g Rep. 69, 75 (2008) (suggesting sunset provisions as a way to “minimize the[] pernicious long-term effects [of panics] on criminal law”). ↩︎
- Cf. Anna Roberts, Asymmetry as Fairness: Reversing a Peremptory Trend, 92 Wash. U. L. Rev. 1503, 1513–49 (2015) (arguing for asymmetrical treatment of peremptory challenges, including to promote fairness). ↩︎
- See Myers, Complex Times, supra note 315, at 1874 (making the point about civil versus criminal regulation). ↩︎
- Myers, Responding to Time-Based Failures, supra note 315, at 1363. ↩︎
- Id. at 1367. ↩︎
- See Robert D. Cooter & Michael D. Gilbert, A Theory of Direct Democracy and the Single Subject Rule, 110 Colum. L. Rev. 687, 707 (2010) (discussing “riding” in connection with the single-subject rule in most state constitutions). ↩︎
- See Justin W. Evans & Mark C. Bannister, The Meaning and Purposes of State Constitutional Single Subject Rules: A Survey of States and the Indiana Example, 49 Valparaiso U. L. Rev. 87, 88 (2014) (“Forty-one of the fifty state constitutions, or eighty-two percent, contain a general single subject rule.”); id. at 146–47 (summarizing the argument that Indiana’s framers intended for that state’s single-subject rule to be subject to robust judicial enforcement). The single-subject rule has been inconsistently applied. Michael D. Gilbert, Single Subject Rules and the Legislative Process, 67 U. Pitt. L. Rev. 803, 807 (2006) (recounting such criticism); see also Cooter & Gilbert, supra note 338, at 710 (grounding application difficulties in the idea that whether a law violates the single-subject rule “is purely a question of the level of abstraction at which judges believe they should frame the subject”). In one prominent recent example, Arizona’s Supreme Court unanimously struck down a statute for violating the single-subject rule that, among many other things such as gambling and management of a state museum, prevented localities from adopting COVID-19 mitigation measures. Ariz. Sch. Bds. Ass’n v. State, 501 P.3d 731, 740 (Ariz. 2022). ↩︎
- Evans & Bannister, supra note 339, at 149; see also Cooter & Gilbert, supra note 338, at 709 (identifying the purposes of single-subject rules: “preventing logrolling, preventing riding, and improving transparency”). ↩︎
- See Boyce, supra note 316, at 22 (proposing two-year sunsets for new criminal laws and for municipal ordinance crimes). ↩︎
- Myers, Responding to Time-Based Failures, supra note 315, at 1369–71. ↩︎
- Id. at 1369–70. ↩︎
- Id. at 1371. ↩︎
- See Mitchell, supra note 316, at 1697 (arguing that “core” crimes “derived from the common law of England” should not be sunset); see also Myers, Responding to Time-Based Failures, supra note 315, at 1380 (recognizing that measures short of his full proposal could still be helpful if adopted). ↩︎
- Primus, supra note 97, at 121 (“More than 80% of American criminal defendants are indigent.”). ↩︎
- Davidson, No Exceptions, supra note 227, at 2138–45. ↩︎
- A proposed bill in New York would do just that. Id. at 2113, 2158–60. ↩︎
- Edgeworth Econs., A Cost-Benefit Analysis: The Impact of Ending Slavery and Involuntary Servitude as Criminal Punishment and Paying Incarcerated Workers Fair Wages 9–22 (2024), https://static1.squarespace.com/static/58e127cb1b10e31ed45b20f4/t/65b974d55c35ee77d9a1dcea/1706652885542/2024+-+CBA+of+Ending+Prison+Slavery+Report.pdf [https://perma.cc/RSL3-WHYH]. ↩︎
- Id. at 3, 23–42. ↩︎
- Davidson, No Exceptions, supra note 227, at 2153. ↩︎
- Id. at 2166 (quoting Andrew Ross, Tommaso Bardelli & Aiyuba Thomas, Abolition Labor: The Fight to End Prison Slavery 44 (2024)). ↩︎
- Id. at 2142 (quoting Legis. Council of the Colo. Gen. Assemb., 2018 State Ballot Information Booklet 39–40 (2018) https://content.leg.colorado.gov/sites/default/files/2018_english_final_for_internet_1.pdf [https://perma.cc/9KRY-7AJZ]). ↩︎
- Sometimes asymmetry is a sensible design. See Roberts, supra note 334, at 1533–49 (arguing that peremptory challenge law for jury selection ought not be symmetrical between defendants and the government). ↩︎