Mercy towards decarceration: Examining the legal constraints on early release from prison

There are close to seven million people under correctional supervision in the United States, both in prison and in the community. The US criminal justice system is widely regarded as an inherently unmerciful institution by scholars and policymakers but also by people who have spent time in prison and their family members; it is deeply punitive, racist, expansive and damaging in its reach. In this article, we probe the meanings of mercy for the institution of parole.


Introduction
The number of people under state supervision in prison in America equals the populations of major cities like New York and dwarfs the population of cities like Los Angeles, Chicago, Houston, Philadelphia and Phoenix (Kaeble and Glaze, 2016). Taken as an entire 'system', America incarcerates more people per capita than nearly any country in the world (National Research Council, 2014). 1 Since the 1980s, the average prison terms in the United States have only gotten longer, and people serving lengthy sentences are growing as a share of the prison population; in 35 American states, for example, at least 1 in 10 people have been there for a decade or more (Courtney et al., 2017). These numbers are especially high for those serving time for violent crimes (Courtney et al., 2017;Sered, 2019); representing more than half of the prison population, the release of men and women convicted of these crimes pose a significant challenge for decarceration (Sawyer and Wagner, 2020). There are a number of US-based initiatives aimed at cutting the incarceration rates in the country, such as the Macarthur Foundation's Safety and Justice Challenge, aimed at reducing jail populations, or the initiative known as 'cut50', lead by the Dream Corps and aimed at reducing the prison population by 50% over 10 years. However, researchers have demonstrated that even if all individuals convicted of drug, public order and property crimes were released early or sentenced to alternatives to incarceration, the population would still not be reduced by 50%, and they have thus focused on more serious crimes in these initiatives (Goldstein, 2015). Leading scholars and policymakers have argued that unless the US federal government and states take radical steps to address sentencing for violent crimes and parole policies, the country will not reduce its prison size in a way that sufficiently mitigates the harms of incarceration on the populace more generally (Reitz and Rhine, 2020;Sered, 2019).
The global COVID-19 pandemic has brought calls for mass clemency to the forefront of national and international debates on incarceration (Abdur-Rahman, 2020; Grant, 2020;Oliva and Notterman, 2020). Now is the time to consider the role and power of the state to grant people early release from prison as an exercise of mass mercy -recognizing that despite their crimes and their actions, early release from prison is necessary given our current understanding of the harms produced by over incarceration, even if not clearly warranted by a classic justification for punishment. While such mass release raises serious questions about the specific mechanisms for release, our point here is largely elucidate the way that mercy moves us towards decarceration in substantial and meaningful ways.
There is evidence that this kind of recognition is possible. In the face of the global pandemic that has resulted from the novel coronavirus, many state legislatures have moved towards this position, if not nearly as robustly as necessary. Several US states have taken steps to reduce the jail populations, including decreasing bond payments, reducing arrests and state judges have taken advantage of administrative release (Prison Policy Initiative, 2020). And while US states have not released a significant number of people during this global pandemic, state legislatures have created the pathway for such action. The states of Louisiana, New Jersey, Michigan, Kentucky, Virginia and Oregon among others have all taken legislative or executive action to decrease the prison population (Prison Policy Initiative, 2020). A state appeals court in California has ordered prison officials to half the population of San Quentin State Prison after more than 75% of the population contracted COVID-19 and 28 men died (Egelko, 2020).
The California court decision raises questions about how such winnowing of the prison population might be done. But the answer can be found in the legislative policies and executive decisions that have followed COVID-19. For example, legislation in Virginia requires the Virginia Department of Corrections to develop and implement the Inmate Early Release Plan (Virginia Department of Corrections, 2020). Under the plan, a declared emergency authorizes the Director to release prisoners on lower levels of supervision or other forms of community corrections, or any prisoner with less than a year on his sentence if from imprisonment or supervision that is completely independent of the meritocratic principles underlying parole and other measures. This moment opens the door for us to take mercy more seriously as a vehicle for decarceration. In this article, we focus on parole as the site where mercy can be used towards decarceration in the United States. As Berk et al. (1983) have argued, parole reforms have taken on different meanings over time and place; we argue that now is the time for parole reform to be rooted in the principle of mercy.

Envisioning mercy through the eyes of the incarcerated
In the sections below, we briefly discuss two case studies chronicling the efforts of men who have sought an early release from prison. These cases typify the institutional responses to people with violent convictions. People with convictions, particularly violent convictions, are arguably viewed as dangerous and morally offensive, and thus disqualified from due and proper recognition and consideration as moral subjects, which has implication for their treatment as legal subjects (see also McNeill, 2018). The case studies selected below represent the beginning of our efforts to understand the challenges that exist in the contemporary US justice system described above. The studies are intended to highlight the ways that the institution of parole, as currently devised in the United States, derails mercy.
Michael, 2 who was convicted of murder at the age of 16, was denied parole in New York State seven times since he first became eligible, 24 years after he was convicted. Although Michael had received a risk score of 'low' on a re-entry assessment, had participated in higher education, engaged in volunteer programming and remained incident free, 3 he continued to be denied parole for 38 years. Because Michael was not a US citizen, he faced immediate deportation upon release. While he might have challenged his deportation, his counsel had made it clear to the parole board that this was not his intention. The parole board decisions were in part due to the continued pressure by the family of his victim.
Although seen by staff in the prison as a 'model' prisoner, Michael's attempts at gaining release through the parole process were consistently thwarted because of the serious nature of his crime. Like many other similarly situated prisoners, the parole board relied on the nature of his crime to deny release. But such a decision subverts the function of parole. Had the state legislature not wanted people with Michael's convictions to be parole eligible, those crimes would have been precluded from the parole process. In a literal sense then, the parole board's decisions over the span of three decades denied Michael his legal agency. More than ensuring that addressing the impact of his crimes would be a condition of his release, the board sought to guarantee that those crimes would be the block preventing him from obtaining his freedom. And while a subsequent parole board decision may lead to his release, that will not erase the effects of three decades of imprisonment over the course of Michael's young adult life.
Michael's case exemplifies the confused logics of what has been termed the 'punitive turn', after the 1970s, when penal philosophies collided with neoconservative and neoliberal ideologies of control and 'reform': not only was he required to engage in a required regime of programming in prison, he was subjected to a battery of risk assessment instruments, all examples of the 'risk' oriented models that emerged in response to neoliberal shifts towards the responsibilization of people in prison (O'Malley, 2008). Yet, his fate largely rested on the voices -and power -of the family of his victim (a young, White woman, he is a Black man) -who put pressure on the parole board not to release him. The ascendancy of the role of victims in the punitive process represented a shift that reflected existing societal values and hierarchies of victimization (rooted in race, class and gender ideologies), but also which intersected with risk logics (Cox, 2003;O'Malley, 2010;Zedner, 1991). So, even though Michael did everything that was expected of him, his freedom was beyond his control; his attempts at obtaining parole were systematically denied without the parole board needing to make any justification for its claims, his crime became the determining feature of his existence and he was unable to be viewed as a person of dignity and worth beyond that crime.
Michael was not 'begging for mercy' in a way that was untethered from a rational quest for justice, one rooted in principles of equity and fairness. He sought an opportunity to be released and continue an ordinary life, one where he could exist outside of confinement and allow himself to grow without the burdens and demands of incarceration. Although he expressed deep remorse for his crime, that remorse remained unrecognized. Legal scholar Stephanos Bibas (2004) has argued for the role of remorse and humility in the legal process and has pointed to the ways that these are relational and dyadic; yet, in the case of Michael and others, there is no opportunity for reconciliation. The victim's family provided their input, and the parole board largely followed their desires.
Dwayne Betts recalls a call from Michael 2 just after the decision in his penultimate parole hearing. Denied again, Michael wanted to know whether Betts believed he would one day be released. The question was not borne out of imagining that Betts had some inside information, but out of an honest wondering about whether either could salvage meaning from a process anachronistically fixated on his conduct nearly 40 years in the past. Michael was ultimately released from prison. Between 2005 and 2015, in Roper v. Simmons, Graham v. Florida and Miller v. Alabama, the US Supreme Court decided three cases that found that for crimes committed before their 18th birthday people could not be executed, sentenced to life without parole for non-homicide offenses or sentenced to mandatory life without parole sentences. Each of these decisions was premised in part on the idea that youth is more than a biological fact and might diminish the culpability of a defendant. At the time when Betts worked on Michael's case, the state of New York's parole statute failed to reflect this jurisprudence about age and culpability. Michael's case challenged this, and he was ultimately granted a new parole hearing. While the board did not grant Michael parole at this hearing, he was granted parole at the subsequent hearing. Under current systems of parole, the person who committed the heinous crime may never able be able to fully demonstrate their capacities for rehabilitation -and the state itself can never prove such rehabilitation. In Michael's case, it took the state 38 years to decide, in his case, under duress, that he merited release. But this decision was deeply protracted, painful for the victim's family, for Michael, and arguably for the state itself.
Charles 4 entered custody in 2001 and was ultimately sentenced to 25 years to life in prison at the age of 19 for a homicide. He has spent the last 15 years bouncing between the maximum-security prisons of upstate New York. Designated as a 'gang leader' by senior prison officials, he has spent his entire time on a closed supervision unit. This unit is reserved for individuals who are identified as gang members, members of the mafia or high-profile prisoners, because of the nature of their case. Once identified as a closed supervision prisoner, it is almost impossible to get off this status. Being on this unit has precluded him from participating in the kinds of activities such as college programs, leadership groups, HIV/AIDS peer training and so on, which may have demonstrated his worthiness for rehabilitation. He successfully completed anger management programming, however, and he has been able to obtain conjugal visits with his wife, whom he married. He has not been incident free -yet his tickets have been for minor insurrections, such as stepping over a painted line or accidentally burning a state bench when he and his wife were cooking dinner during their conjugal visit (he was sentenced to time in solitary confinement for that accident).
Charles's early childhood was marred by neglect: his mother became addicted to crack cocaine when he was a young child, and she left him in the care of his father, a local drug dealer who was well known and respected in his community and who did not shield his young son from his lifestyle and transactions. He also physically abused him, and the state finally intervened and placed Charles in foster care when he was a teenager. He had already begun to participate in the family business, and was subsequently arrested and charged with drug sales, at the age of 16, in a county in upstate New York where harsh justice was the norm: he was sentenced to prison time. After he got out, he ascended through the ranks of a local street organization and became known to law enforcement. He was eventually arrested and charged with the murder of another young man.
Under the law, Charles is not eligible for parole for another 9 years. And all evidence points to the fact that he will not be granted parole at his first opportunity to receive it. Thus, he is trying to apply for clemency; yet, all sources indicate that he cannot be presented as a character who can be redeemed. Law firms that provide voluntary clemency assistance will only take individuals under their caseload who, for them, demonstrate 'exemplary' rehabilitation, which includes an engagement in a broad range of rehabilitative programs and active citizenship in facility life. They have passed his case over because he has not had the opportunity to engage in these activities.
But what is Charles' perspective on mercy? Individuals like Charles, who have spent many years in prison, are not naive about their narrow opportunities for freedom. He knows that his status as an identified gang member prohibits him from participating in the activities that might make him seemingly more redemptive, but he is also under no illusion that the activities in which he is able to participate will even be fully recognized as pathways to redemption. He wants to know and understand the concrete and clear paths to his freedom.
There is not adequate space within the system to recognize Charles' youthfulness when he was first arrested. There is no place in the parole system to recognize the personal journey or pathway he has taken away from the difficulties of his childhood and towards a supportive network of individuals, most especially his wife, who has built a home for him, works full time and is deeply engaged in his personal and spiritual growth. There is no room to recognize that he no longer identifies as a gang member, because the system has stamped him in that way; there is no recognition of the myriad relationships and mentorships that he has participated in and received support from men on the inside who have gone on to become leaders in their community, because they are indeed 'criminals' themselves. For men like Charles, who may not be the poster children of an ideal prisoner, is there a chance for freedom? It is rare. Yet, they, and we might ask, why not? Charles seeks mercy and grace but finds he has no clear avenues to obtain them, despite the ostensibly clear pathways laid out for him in the system. He was convicted of a crime and sentenced to time; he has done everything he can to meet the systemic demands on him for redemption, yet he will likely be denied an opportunity to go home at his first parole board meeting.

Mercy, the law and parole
To take mercy seriously does not mean that mercy lives outside the law. In fact, there are both mercy-enhancing laws and mercy-depriving laws that already exist in the American legal systems. Mercy-enhancing laws exist at the local level, for example, the use of nolle prosequi or the dismissal by prosecutors; and, at the state level, where legislative bodies can set a standard of presumptive release from imprisonment. We are invested in the law as a source of possibility rather than simply a mechanism to reinforce carceral logics; this grows out of our research and work with people who have served very lengthy sentences. These individuals hook their vision for liberation to the reimagination of legal structures that have facilitated their incarceration, in part because they have committed an act of violence and know that they have had to do prison time for that act. We thus believe that an individual's guilt can be recognized not as a source for a lifetime of punishment, but as a wellspring, the start of a conversation about mercy.
We begin with a brief overview of the philosophical and legal conceptions of mercy-like practices. Mercy-like practices are characterized here as practices that lead to outcomes that would be considered mercy, as we define it, but are not considered to be mercy by the actors making those decisions. We then focus on extant appraisals of mercy in the philosophical and legal realms, and their relevance and uses for a consideration of early release from parole.

Mercy-like practices and the issue of merit
In the context of existing criminal justice decision-making, mercy is typically defined as a form of compassion or leniency that is given to individuals without or independent of a legal entitlement, right or claim to such compassion, either because they have been sentenced in accordance with the law's requirements or the facts of their case or crime may not be viewed by the public to merit such compassion or leniency. Yet, in a number of places in the criminal justice system context, leniency (or what can sometimes be viewed as mercy) is offered as a way of recognizing an individual's merit or their worth, as opposed to being an act of grace that transcends consideration of merit. We identify those sites below and raise questions about whether they can be considered acts of mercy.
There are various acts that can take place in the criminal justice process, both at the pre-and post-conviction stages, that have been considered to be merciful. Pre-conviction acts include the judicial dismissal of charges, a failure to indict by a grand jury, the filing of lesser charges or a nolle prosequi by a prosecutor or a plea bargain for an alternative to incarceration. Post-conviction opportunities include judicial leniency at sentencing, prosecutorial support of a sentence modification, parole or clemency. These forms of leniency can be exercised by a judge, a grand jury, the prosecutor, parole board or executive (Barkow, 2008;Huq, 2015;Markel, 2004). However, there is debate about whether decisions to reduce prison sentences constitute mercy. Might those decisions be better characterized as efforts to establish proportionality or parsimony and ultimately identify desert in punishment, not to a desire to exercise the restraint that mercy demands?
Thus, some acts might be more appropriately termed leniency rather than mercy. Some acts that fall into a grey area between justice and mercy may include those exercised by the grand jury, which is legally empowered to refuse to indict a person even if there is probable cause; it can substitute non-capital for capital offense or lower charges (Markel, 2004). But, for example, when a grand jury fails to indict because the jurors do not believe a conviction will follow, such a decision should not be considered mercy. Police and prosecutors can also exercise a form of leniency in that they can decide not to charge or prosecute a crime, or to plea bargain, to lessen the potential effects of punishment. But when the motivation behind those decisions is a disbelief in the merit of punishment, given the context, it is not rightly mercy. Nor is plea bargaining, in its many vagaries, rightly characterized as a place of mercy. While plea bargaining may lead to netwidening and over-criminalization (Husak, 2008), the historical use of the prosecutorial power to nolle prosequi and the more recent adoption of alternatives to incarceration as a vehicle for dismissal of criminal charges reflect the potential for plea bargaining to be a staging point for mercy. 5 The power to exercise leniency is enshrined in the law but does not necessarily require justification by the authority figure who exercises it. Yet, the procedural hurdles to clemency and pardons, and the public justifications for granting them, all highlight the merit of the receiver. The decisions are not about mercy, but correcting a past wrong or recognizing merit.

Defining mercy
Enter mercy here. No stranger to the philosopher's ruminations, the concept of mercy has been explored by scholars from a range of disciplines, from the law, to philosophy, to theology. The relationship between mercy and merit has been dealt with extensively, for example. Some scholars have rooted the expression of mercy in a deeply empathic response to and understanding of human flaws -as an act of charity -a way of conveying benevolence or compassion for an individual . And they have said that this response can only be driven by a sound appreciation of one's own flaws (as a jurist). Aristotle and Seneca pointed to the character strengths inherent in beginning with an approach to others that assumes that we are all part of a complex web of interactions that produce flawed behaviour (Nussbaum, 1993: 104). Justice itself can actually be rooted in these flaws -and judges who exercise justice can do so in a way that equitably appreciates the expression of human flaws through their application of mitigation and mercy (Nussbaum, 1993: 94). On philosopher Martha Nussbaum's account, the expression of mercy facilitates justice.
Philosopher Claudia Card argues that when we temper institutional justice with mercy in deciding how to treat the offender, we consider not only facts about his offense but also facts about his character and suffering which may not be revealed simply by looking at his offense. (1972: 191) Hers is a claim that people deserve mercy as an act of justice. Card centres her concept of mercy around an idea of deserving, based on some concept of suffering, she argues that 'mercy is deserved on the basis of what one has endured and the nature of one's moral character on the whole, rather on the basis of individual performances or omissions ' (1972: 198). She relates mercy, then, to a form of compensation or in legal terms: mitigation. But retributivists such as Jeffrie Murphy (1988) have argued that mercy should not be 'deserved'. Jeffrie Murphy (1988: 4) has argued that the exercise of mercy involves both a tempering of justice but also, to some extent, a departure from it. Mercy, Duff (2007) argues, has no place in the criminal justice system, as it is a concept that has moral value and meaning in systems outside of the justice system and complicates the aims of achieving fairness in justice.
Others have argued for a more formal definition of mercy and its uses in the law. They have challenged the uses of mercy within a system that, they argue, should be focused on desert and proportionality in punishment (Duff, 2007;Murphy, 1988;. In seeking out a remedy to the messiness of the term 'mercy', some scholars have advanced the term 'equitable discretion'. They argue that equitable discretion creates space for leniency in the criminal justice context in a way that forces the decision maker to take account of particulars of the individual's life and actions that are relevant for the application of justice, in contrast to mercy, Mary Sigler (2015) has argued is unbounded and allows for a form of leniency that doesn't hold the decision maker accountable. Perhaps, the strongest adherent of 'equitable discretion' instead of 'mercy' is Dan Markel, whose version of equitable discretion is tied to an individual's reason for committing the crime (2004: 1435-1436). Markel argues that mercy should be awarded for 'justice enhancing reasons', or those 'judgments based on articulable standards of desert in relation to culpability and the severity of the offense' (p. 1441). He argues that mercy is the remission of deserved punishment, in part or in whole, to criminal offenders on the basis of characteristics that evoke compassion or sympathy but that are morally unrelated to the offender's competence and ability to choose to engage in criminal conduct. (2004: 1436) For example, on Markel's account, a judge's decision to sentence an elderly person to a shorter period of incarceration than the statutory minimum because the judge feels that the person does not deserve to die in prison, if that reason is disconnected from the person's role in the crime, is a form of mercy, not equitable discretion. He argues that if a decision maker is to consider an individual's social background in sentencing, the elements of that social background must be connected to that individual's role in a crime, not a desire by a judge to be sympathetic or compassionate to that individual because of their traumatic past. In a sense, Markel argues for transparency in mercy.
Markel's argument raises a core and fundamental question: Is there room for truly merciful acts in a justice system? This is particularly significant in the light of where in the judicial process the scholarly scraps around proportionality, parsimony, equitable discretion and mercy have taken place. The focus has been on the use of these mechanisms at the front end of the legal process -the point of sentencing; rarely has the question of mercy after a sentence has been determined been considered in the scholarly literature. Lacey and Pickard (2015: 673) have argued that once a sentence has been handed down, mercy is 'otiose', and the question of forgiveness remains. We agree that 'justice' is allocated at the point of sentencing is considerably different from how it is allocated after one has been convicted. And in this moment of mass incarceration, what happens at the front end, guided and directed by the confluence of legislative proscription and prosecutorial and judicial discretion, is hopelessly muddled. American punishments have become too harsh; the harsh punishments that pervade American life not only have effects on individuals, but they also have population-level effects-on public health, social inequality and the reproduction of racism, classism and sexism.
Decarceration, as a principle, is constantly at odds with state and federal laws that increasingly proscribe incarceration for more conduct. Mercy should be considered a vehicle for decarceration, ensuring that the question of whether someone should be incarcerated is not simply a product of revisiting broad laws, harsh sentences or past conduct in the light of more recent efforts at rehabilitation.
In his recent book on his work as a criminal defence attorney, representing individuals accused of serious and violent crimes, young people and people on death row, Bryan Stevenson argues that 'mercy is most empowering, liberating, and transformative when it is directed at the undeserving. The people who haven't earned it, who haven't even sought it, are the most meaningful recipients of our compassion' (2014: 314). Stevenson says that our commitment to justice is revealed when we direct our compassion towards those who are the most marginalized -this includes the condemned and the incarcerated. Inherent to Stevenson's message is that an individual's past conduct should not be a barrier to receiving mercy. Stevenson's legal work draws on the Judeo-Christian notions of mercy as an expression of God's grace and compassion -an unconditional expression, one that recognizes that all people, not as particular individuals with particular histories, should be subject to such compassion, regardless of their circumstances, or 'unmerited grace'. For Stevenson, this is captured in his oft-repeated statement that 'each of us is more than the worst thing we've ever done ' (2014: 17-18). But for men and women with the death penalty or life in prison, 'mercy' has functioned to move from a death sentence to a near-death sentence by years. For example, in Florida, Kenneth Young was sentenced to life without parole for a home invasion that he committed with a co-defendant at 15 years of age. After the Supreme Court found in Graham v. Florida that life without parole sentences offended the Eighth Amendment of the constitution, Young appeared before a judge to be resentenced. At that point, after 7 years of incarceration, most without hope of release, he had not received a single institutional charge and had gone about the work of rehabilitating himself. In sentencing him to 40 years in prison, the judge told him that he would not give Young credit for the rehabilitation that the Florida Department of Corrections required him to participate in. The sentence was reduced, Young had demonstrated some 'merit' in receiving a reduction, yet he still faces most of the rest of his life in prison.
This cannot rightly be considered mercy. Instead, it represents the theoretical muddling of parsimony, proportionality and deserts that keeps the prison population from decreasing in any meaningful ways. The problem is state and federal sentencing laws do little to impose restraint on the outcomes of guilty verdicts and guilty pleas. In such a system, parsimony, proportionality and desert will always be in tension. In a world where the ceiling is routinely dozens, if not hundreds of years in prison, understanding when the floor is unjust becomes bogged down in arguments about preference and not justice. Discussions of mercy that are untethered from the legitimacy questions related to state and federal laws, state and federal sentencing guidelines and immediate or near immediate release from prison miss the point. Or, as William Stuntz (2012) puts it, revivification of the role of mercy is necessary in a country that exerts its powerful might against those who have been accused of crimes, particularly those condemned to severe sentences. 6

Parole as an unmerciful institution
The legal scholar Rachel Barkow (2008) argues that the rise of the administrative state -which is embodied by institutions like parole boards -has led to the demise of merciful practices, in part because effective mechanisms of accountability for these administrative institutions do not exist. Ironically, the parole system itself was established in part as a measure to relieve prison overcrowding and reduce excessive sentencing, as well as to ensure that people obtained the right to rehabilitation (Messinger et al., 1985;Rotman, 1986). While there are ways to publicly shame parole agencies for the release of people who the public might consider undeserving of an early release, the converse approach does not exist. The link between a person's demonstrated process of 'change' and successful ability to earn parole is tenuous at best.
Mercy in the context of parole matters because the length of prison sentences has grown and opportunities for parole release have been constricted in America since the 1970s, particularly for people convicted of violent offenses (Courtney et al., 2017;King, 2019). In 1974, following from national conversations about 'just deserts' and proportionality, a group of researchers and policymakers came together to propose and adopt guidelines for the then-US Board of Parole, which operated in federal cases, aimed at making the system more fair and proportionate (Bottomley, 1990). The guidelines imposed a scoring system which sought to strike a balance between offense severity and 'parole prognosis ' (Gottfredson et al., 1975: 37). The new guidelines also required that parole board members provide an individual with the reasoning behind a parole denial. In addition to the federal context, these guidelines were adopted in 15 states (Bottomley, 1990: 345). According to Bottomley (1990), the guidelines went some way towards controlling for disparities and moving towards more fair procedures, but they also had several, more potentially harmful consequences. Four of the original fifteen jurisdictions that adopted the guidelines ended up abolishing parole in their jurisdictions in the 1970s; the US Parole Commission itself ended, effectively abolishing parole for federal prisoners in 1984 with the passage of the Comprehensive Crime Control Bill, which established the US Sentencing Commission (Bottomley, 1990: 345). A number of states followed, and over a quarter of the states had abolished parole by the late 1980s, which coincided with an increase in prison sentences across the country (Bottomley, 1990: 346). 7 The reforms to parole have also involved a rise in the use of putatively 'rational' processes like the use of risk assessment instruments such as the COMPAS tool, which is used in a number of states to assess an individual's eligibility for release, in part based on their participation in prisonmandated programs, but also based on risk factors related to their individuals features (level of education e.g.) and their expressions of remorse (in response to set interview questions). Many people find that they complete hours and days of required in-prison coursework to demonstrate their rehabilitation, only to be denied freedom as a result of their poor performance on a risk assessment (see also Shah, 2017). 8 COMPAS has been demonstrated to have racially disproportionate impacts in terms of its identification of people as being at high risk, and thus not eligible for release, and its misalignment of research knowledge and 'data' (Angwin et al., 2016;Dressel and Farid, 2018).
On our account, a person who has worked through these traditional frameworks, but has failed, should still achieve mercy. What parole boards lack are an ability to address this issue and exercise mercy. As a result, in recent years, parole has become more and more out of reach for individuals who are incarcerated, according to the Pew Charitable Trusts (2014). Individuals who have been convicted of crimes face significant barriers to meet the conditions of release from prison, parole and probation (ACLU, 2016;Phelps, 2016;Werth, 2017). In addition to the states that have eliminated parole or severely curtailed it, parole boards are notoriously opaque in their decision-making structures (Barkow, 2008;Schwartztapfel, 2015). More than one in five people in US state prisons maxed out of their prison sentences because of difficulties in obtaining parole, with significant increases in the numbers of people maxing out since the 1990s (Pew Charitable Trusts, 2014). For far too many people, due process in the parole release system is being notified that the parole board has denied your application, irrespective of the system's ostensible commitment to parsimony in punishment.
Across the United States, the current system of post-conviction relief for individuals has welldocumented problems of arbitrariness, an unequitable distribution of outcomes, and a lack of transparency or oversight, and elusive and unclear standards for rehabilitation (Mehta, 2016;Schwartzapfel, 2015). Individuals who appear before parole boards in many states find that crime victims play a disproportionate role in influencing parole board decisions; that the victim and perpetrator of the crime's race, class and gender impacts disproportionately on their outcomes; and, that their personal characteristics, which may have played a role in mitigating their sentencing outcomes (i.e. their young age upon commission of the crime), do not play a role in whether they are granted parole (Huebner and Bynum, 2008;Mehta, 2016;Morgan and Brent, 2005). They may engage in every rehabilitation program available to them, per the guidelines of the prison system and of the state, but find that the parole board makes their decision simply on the basis of their perception of the heinousness of the offense (e.g. see the case of Kathy Boudin, whose extensive involvement in prison programming, support groups, HIV/AIDS prevention initiatives and higher education programming were tabled in favour of the parole board's focal consideration -the crime in which she participated, which involved the death of a police officer (Associated Press, 2001)).
The answer to this dilemma is mercy. Discretion that does not centre on mercy, as evidenced by the parole grant rates across the country, easily becomes a way to justify continued incarceration.
Yet, because each grant of parole is susceptible to public scorn, parole boards have learned to couch their decisions not in mercy, but in what can be learned from a COMPAS report.
Not only is the process of being released from prison fraught with obstacles and indifference, it includes deep encroachments on personal freedom when one is under the supervision of parole or probation offices. The formerly incarcerated person is something of a contemporary trope. People return to their communities from prison and frequently struggle to acclimate to a society because of the nature and consequences of their status and how that complicates their ability to connect to the key institutions that has shaped American life -the prison. But freedom should not be conditioned on a person's prison experience becoming the albatross around their neck. State governments have the power to decide whether freedom comes after completing a prison sentence or being granted an early release because of extraordinary personal rehabilitative efforts, or through mercy -the last of which imagines relief as being possible without merit.

Conclusion
Scholars have begun to ask whether we believe that justice requires a punishment without end (Loader, 2010). An American commitment to punitiveness, degradation and banishment has played a unique role in the country's approaches to punishment (Beckett and Herbert, 2010;Whitman, 2003). We argue that the exercise of mercy must and should challenge the permanent markers of punishment that affix themselves to the individual.
Mercy has its limits and that we must interrogate them. We recognize that post-release supervision and parole can in and of themselves not be merciful institutions, and the use and overuse of parole violations have created a robust funnel back to prisons (Bülow, 2019;Jacobi, et al., 2013;Klingele, 2013). Mercy should not be an act that only ever can exist at the early stages of the system, when a person's crime is linked to his or her life, and his or her ability to have a full life should never be treated conditionally. In the case of individuals like Michael and Charles, whose lives will continue to be enmeshed with the criminal justice system for many years to come, it must be linked to a sense of transparency and fairness, and a recognition of human dignity and worth.
These are particularly salient questions in the context of states that have abolished parole or created punitive sanctions that allow for little or no discretion on the part of parole boards. If there should be a place within the legal structure of the liberal democratic context of the United States for its citizens, even those who are incarcerated, to demand dignity and rights in the context of those laws, where is the place for them to find freedom? Without vigilance, we risk the perpetuation of the idea that only appropriate sinners can live among us.