II. Compassion and substance

In order to consider the relationship between compassion and the rule of law, it is useful to distinguish several ways in which compassion could influence legal decision-making. First of all, we might have a statute that explicitly permits the use of compassion as part of the exercise of allowable discretion. A prison parole scheme might permit compassionate release, for example if the prisoner has terminal cancer. Immigration laws might take compassion into account in granting asylum. In these situations, compassion does not pose a special challenge to the rule of law because there is advance notice that it will be employed as a factor, and there are guidelines for its use, and there is a body of precedent. Indeed, it is not even clear whether compassion is the appropriate term for what these statutes take into account. Although those arguing for exceptions may seek to elicit compassion as a way of driving home their arguments, that grant of compassion often has to meet prescribed criteria. For example, federal sentencing guidelines may take into account the fact that the defendant has young children, but the standard there is the best interests of the children,Footnote ⁹ rather than compassion for the prisoner. Asylum is granted where the applicant shows a ‘well-founded fear of persecution’ on certain specific grounds.Footnote ¹⁰ In other words, these criteria are announced in advance, and they require factual support. If compassion is a feeling, it is not really in play here – these are legal arguments like any other legal arguments.

Of course, pre-prescribed factors may be misused or used inconsistently, but that problem can occur with the application of any legal factor. The pardon power and the clemency power present a more difficult case for the rule of law and compassion. They permit the exercise of compassion in a less constrained context, because the discretion to grant a pardon or executive clemency is so broad.Footnote ¹¹ Each case is sui generis. Decisions may be based on compassion, or they may be based on raw politics. And there may be no way of knowing the actual grounds, if the executive is not required to explain them. Yet, because the discretion is so broad, no decisions exceed it. In that kind of legal regime, you cannot score points by pointing out that a particular decision lacks consistency or predictability – because clemency and pardon do not generally require consistency or predictability. Yet, to say these decisions are lawless is not precisely accurate – they are decisions taken in a field of immensely broad discretion.

A more serious rule-of-law problem is posed when compassion is used to make unauthorised exceptions to a rule. Decisions applying the US federal sentencing guidelines illustrate both of these possibilities – determining where in the allowable range a decision falls, and impermissibly exceeding the range. A judge may use compassion to arrive at a sentence at the low end of the allowable range. But the guidelines impose mandatory minimum sentences that have shocked the conscience of many judges – decades in prison for a first time, low-level drug offence, for example.Footnote ¹² Here is where compassion's role becomes complicated. The guidelines were meant to regularise sentencing and avoid some of the glaring differences in sentence caused by the personalities or proclivities of individual judges. Sentencing, it was argued, should not depend on the luck of the judicial draw, or whether the judge is in a punitive mood on the particular day, or whether the defendant's plight triggers something in the judge based on his own experience or those of his family.

The sentencing guidelines eventually came to be viewed by many as a cure that was worse than the disease. A handful of judges felt they simply could not abide by them in good conscience. At least one federal judge resigned his life tenured position.Footnote ¹³ Others just refused to comply in cases where they believed the sentence would be exceptionally unjust. Judge Jack Weinstein, for example, a federal judge from New York well known for his creativity and courage, has defied the guidelines, arguing that judges have a duty not to impose unfair sentences, a duty to expose injustice and a duty to interpret law with humanity.Footnote ¹⁴ Many believe he is quite correct in his assessment of the injustice of the guidelines. What follows from this? If defendants are lucky enough to draw Judge Weinstein, a powerful, iconoclastic, deeply moral judge, they will receive more compassionate sentences (albeit sentences that will soon be reversed by a higher court). The rule-of-law problems here are obvious – lack of notice or predictability, unequal treatment depending on the luck of the draw, arbitrariness.

But there is also the greater problem that, if we permit a few courageous judges to act as a safety valve, we miss the larger problem – the guidelines themselves are problematic. Arguably, the entire scheme reflects a failure to take compassion into proper account. And perhaps that state of affairs will continue – an unfair statute mitigated by individual acts of judicial non-compliance – so long as we let a handful of individual judges make exceptions in some egregious cases. But that cannot be considered a victory for compassion – or for the rule of law. A greater victory would be a sentencing scheme that was premised on compassionFootnote ¹⁵ – for example, a sentencing scheme that recognises that people make mistakes and that one bad choice should not derail a person's life, or a legal regime that creates effective alternatives to incarceration, giving people the skills and resources they need to lead a more productive life.

In all the examples above, compassion acts as a factor, or a stopgap, or a deviation, in a regulatory scheme. For the remainder of my remarks, I will not be focusing on compassion as a stopgap, a safety valve, an impetus to make exceptions to general rules. I will not be focusing on compassion as a means of determining where in a discretionary framework a particular decision should fall. Instead, I want to think about whether compassion can help frame the principles and rules themselves, and about whether compassion can help flag principles and rules that need to be revisited.

The United States Supreme Court has actually weighed in on the question of compassion's role in legal reasoning. In DeShaney v. Winnebago County Department of Social Services,Footnote ¹⁶ in 1988, the Court, in a five-to-four decision, came out against compassion as a factor in judicial decision-making. In that case, the question was whether the state of Wisconsin had violated the constitutional rights of Joshua DeShaney, aged four. Joshua's entire short life had been a nightmarish one, in which he was consistently beaten and brutalised by his father. This behaviour had been repeatedly reported to the Wisconsin Department of Social Services, the agency charged by state statute with protecting him. The state had violated its statutory obligation by failing to offer that protection right up until the day that Joshua was beaten so brutally that he spent the rest of his life institutionalised with irreparable brain damage. (Joshua died in November 2015, at the age of thirty-six, in an institution.) The question the case presented was whether the state, in violating its statutory obligation to protect Joshua, had also violated Joshua's federal constitutional right to due process of law. The Court called the facts of the case undeniably tragic, and observed that:

Sympathy, or compassion, is portrayed as a human impulse that must be avoided by judges – an interference with rational deliberation. In this portrayal, we can see some common and deeply rooted assumptions about emotion: that it is impulsive, that it has no helpful cognitive content, that is interferes with logical analysis. Indeed, DeShaney was a famous catalyst in US jurisprudential scholarship, helping to spawn a re-examination of the role of compassion in particular and emotion in general (Brennan, Reference Brennan1988; Minow and Spelman, Reference Minow and Spelman1988). At the time, I wrote that it was an important step in the right direction to debate the role of compassion and other emotions. But I also cautioned against a tendency to uncritically embrace compassion, sympathy and empathy as soft, merciful and therefore a welcome antidote to the hardness of law. I argued then that whether any particular emotion is beneficial to legal analysis depends on the context. Compassion and sympathy can lead us astray just as they can lead us in the right direction. It depends on who is seeking our compassion, and for what legal purpose (Bandes, Reference Bandes1996).

Using the DeShaney case as an example, a judge ought to seek to understand the suffering of Joshua DeShaney, but also the concerns of the Wisconsin Department of Social Services. In the opinion, it was clear that the Court did a very thorough job of understanding the state agency's concerns – the difficult balancing act faced by child welfare agencies that risk liability when they take children away from their biological parents as well as when they leave children with those parents for too long. Is the problem that they did not fully understand the suffering undergone by four-year-old Joshua DeShaney? It would be hard to make that argument. The fact is that it is not that difficult to comprehend the tragedy of a four-year-old child who has lived his entire short life terrorised by his father, and is then consigned to a lifetime of institutionalisation as a result of violence and neglect. In some ways, this is the easiest kind of compassion. The suffering is serious and entirely undeserved. The sufferer is a small, defenceless child. Compassion for that sort of suffering, as the Court observed, is fairly universal. The harder question is who should take responsibility for alleviating it or preventing it.

The famous dissent by Justice BlackmunFootnote ¹⁸ charged the Court with too little sympathy, or too little compassion. It was famous largely because of its emotional register. It contained a phrase – or actually a bit of punctuation – that seemed to violate the protocols of judicial opinion writing: ‘Poor Joshua!’ (exclamation point):

Justice Blackmun's charge is not that the Court failed to understand Joshua's suffering; it is that the Court failed to respond to it correctly. If compassion requires a judge not only to understand, but also to move to alleviate, then it is a problematic emotion for judges – one that raises serious rule-of-law issues, because compassion is not – at least on its own – a reliable indicator of who ought to prevail in a legal dispute.

Judges or jurors may feel compassion for all sorts of reasons – not all of them admirable. The litigant may be weak or flawed or prejudiced in ways that resonate with a particular fact-finder – like the Texas judge who declined to sentence a young man for a vicious hate crime against a gay man, because the defendant imagined his own son in the role of the attacker and noted that he would want his son, or any virile young man, to act just as this defendant had acted.Footnote ²⁰ We might feel compassion most readily towards those we understand – those with whom we identify. Consider a prosecutor with nearly unfettered discretion about whom to charge, or what charges to bring. Compassion might make her a wiser prosecutor, or she might find her compassion leading her to give a break to those she instinctively understands, sympathises with, identifies with. In the US and elsewhere, this kind of compassion may be selectively doled out based on racial bias and class bias, unconscious or otherwise.

Compassion cannot resolve competing claims. We must determine whether it is a compassion that the law is willing to condone or encourage. In addition, compassion may privilege certain types of litigants for the wrong reasons – the tenant over the landlord, the attractive and likable individual over the remote corporation. Perhaps the landlord or the corporation should lose in the particular case, but not unless we can articulate reasons that go beyond compassion for the plight of the little guy.

But what I have said thus far does not convey the full import of the argument that Justice Blackmun was advancing in dissent, or of the powerful dissent of Justice BrennanFootnote ²¹ in the DeShaney case. Justice Brennan's dissent has proved more influential, perhaps because it is written in a more legally familiar style. And it may be that this is more than simply a stylistic or semantic difference; it may be that it reveals an important limitation of compassion as a tool for reaching substantive results.

One of the most confounding questions about compassion is what actions it leads to. The DeShaney majority did not believe it was ignoring the plight of Joshua DeShaney. It simply believed that his problem did not require a governmental solution. The question is not whether a private compassionate response is called for in the individual case; it is whether there is an injustice by the government that needs to be righted.

And so Justice Brennan's argument was not so different from Justice Blackmun's – except that it was not styled as an argument about what we ought to feel for Joshua, an individual. It was an argument about governmental obligation. The majority in DeShaney says, in essence: the government did nothing wrong here. It did not act at all. This was a private infliction of harm by a father on his son. The Court said that governmental social services are a choice, not an obligation. It would have been beneficent of the state of Wisconsin to step in, and yes, maybe Wisconsin violated its own statutes, but Joshua was not deprived of any constitutionally protected liberty (Bandes, Reference Bandes1990). Justice Brennan responded that the government promised protection, undertook protection, affirmatively placed Joshua in his father's home when his parents divorced, effectively made itself the sole guarantor of Joshua's safety, and so it did act to deprive him of his liberty.Footnote ²² One can feel compassion for a plight that one has had no hand in creating. That is a feeling that speaks well for the compassionate soul. But when one has helped create the plight, the feeling ought to be something more than compassion, a merciful but optional act – it ought to be a duty, a legally enforceable obligation.

The difference in tone and in framing can be illustrated by considering the iconic US case of rights-recognition: Brown v. Board of Education.Footnote ²³ As I will argue shortly, compassion likely played an important role in revealing the true nature of the injustice to Black children educated in segregated, all-Black schools. But it sounds strange and even offensive to our modern-day sensibilities to say that segregated schools were outlawed out of compassion for Black children. We do not rest on compassion as a reason to enforce the guarantees of equal protection of law and due process of law. That formulation sounds uncomfortably in charity, condescension and pity (Garber, Reference Garber and Berlant2004). ‘We’ do not bestow equal and integrated schools on ‘them’ out of the kindness of our hearts, as a charitable act that we perform despite our lack of culpability for the situation. The question to be answered is whether there is a right that has been violated, whether an unjust caste system has been perpetuated under colour of law, whether there should be a ‘we’ and a ‘they’ at all.Footnote ²⁴ Once we begin talking in terms of an obligation to fix an unjust caste system, the term ‘compassion’ begins to sound out of place.

The problem with ‘compassion’ is not only the way it seems to let the giver off the hook for inaction. It is not only the way it seems to bathe the giver in a warm glow to reward her act of charity (Berlant, Reference Berlant and Berlant2004; Spelman, Reference Spelman1998). The term is also problematic for the way it portrays the recipient – as a passive sufferer who is lucky to be on the receiving end of good will. With racial equality, as with gender equality and gay rights, nothing has ever been given out of the kindness of anyone's heart. Every recognition of rights has been hard-fought. I do not think it is usually compassion that impels the correction of injustice. A more relevant emotion is moral outrage by and on behalf of the victims of injustice (Bandes, Reference Bandes2011).

III. Compassion and understanding

I now want to turn to the important contribution that compassion makes to the recognition of legal rights – and that is the way in which compassion illuminates suffering. At this point, the distinction between empathy and compassion, and the ambiguities in both concepts, become especially thorny. Empathy helps judges understand what is at stake for litigants. Although empathy is often assumed to flow only from the strong to the weak, I have long argued that this is a mistaken assumption (Bandes, Reference Bandes1996). A judge may feel empathy for the Department of Social Services just as it feels empathy for Joshua DeShaney; for the school administration just as it does for the strip-searched student; for the Board of Education just as it does for Ms Brown. In part, this is because empathy has a strong cognitive component – it calls for understanding the goals and intentions of others, and that does not require feeling. But even the affective component – the capacity to feel ‘with’ another – does not render empathy something that flows only towards the powerless or the disenfranchised. So a judge might have empathy for a corporate or governmental litigant and, moreover, a judge might have empathy for all the litigants before him. If empathy does not require him to act on behalf of any particular litigant, this is not problematic. The problems arise from selective empathy and from empathic inaccuracy (Bandes, Reference Bandes2009). For judges, these are acute problems, because judges are encouraged to believe in their own omniscience. It is a real occupational hazard – the lack of reminders that their perspective is partial; that they have blind spots and prejudices.

Compassion raises similar concerns, especially if it not only involves taking the plight of others to heart, but also involves a desire to alleviate that plight. Empathy may be inaccurate – an unsuccessful attempt to infer the thoughts and desires of the target (Ickes, Reference Ickes, Decety and Ickes2011). Likewise, compassion can rest on inaccurate perceptions. Indeed, the concerns about inaccuracy are even more acute for compassion than for empathy. The essence of empathy is the effort to accurately understand another's perspective. Compassion is not about feeling with the object of one's compassion; it is about feeling for that person. It might be possible to feel for another's suffering without understanding what the sufferer wants or needs (Blum, Reference Blum and Rorty1980, p. 516). One can feel compassion based on mistaken assumptions about the needs or desires of the recipient; one can be impelled to act to help someone who does not want help, or that kind of help (Carr, Reference Carr1999). Compassion can involve a misinterpretation of the pain of others. The capacity to understand others is educable. As in the example of the Safford case above, it can be educated by exposure to the viewpoints and experiences of others from varying backgrounds.Footnote ²⁵ But one essential precondition to this sort of education is self-awareness. It requires recognition that one's own viewpoint is necessarily partial – that the viewpoints of others need to be considered to gain a fuller picture.

More than that, it requires recognition that we see one another through our own lenses, and that our own blind spots and misconceptions may cloud our ability to see others clearly. Because of this, an essential component of true compassion must be humility. As Murphy (Reference Murphy, Worthington, Davis and Hook2017) recently wrote, there are several strands to the concept of humility. One is this ability for self-criticism – a willingness to examine one's own faults and prejudices to the extent they may interfere with the ability to exercise compassion for others. Another is a tendency to underestimate the role that luck and undeserved good fortune have played in one's own success, thus creating a sense of smugness and superiority towards others.

The philosopher Richard Peters, in his lectures on reason and compassion (Peters, Reference Peters1973), made a compelling argument that compassion is one of a constellation of passions that are essential to reason. The reasonable person, he argued, possesses ‘the feeling of humility which is necessary to the wholehearted acceptance of the possibility that one may be in error’ as well as ‘the respect due to another who may have a point of view worth considering’ (Peters, Reference Peters1973, p. 79).

Consider this well-known example of a failure of empathy – which is also a lack of humility. Justice Kennedy in the Carhart v. Gonzales caseFootnote ²⁶ (often known as the partial birth abortion case) considers the emotional effects of abortion on the woman who has chosen the abortion – and of late-term abortion in particular. He makes an assumption that he admits is unsupported by any evidence, but that to him seems unexceptional, or a matter of common sense:

As it happens, there is substantial evidence contradicting the idea that grief and regret are a common reaction to elective abortion; and yet Justice Kennedy relies on his own belief about what women must feel in order to protect them from making a choice that would lead to those feelings (Guthrie, Reference Guthrie2008).

The compassionate person, to avoid this trap, must be cautious and open and curious about how others feel, what others need and what others need from us. And this knowledge is tied to an understanding of our own fallibility and our own vulnerability. Ultimately, it may lead not just to a reorientation of the individual relationship, but to a more expansive notion of shared humanity and responsibility.

Let me illustrate the power of compassion in this regard with an example from a sentencing case. I will then move to my main point about the appropriate role of compassion – its role in alerting us to the limits of current legal frameworks and to the need for legal reform.

3.1 Judge Alex Kozinski's story

Alex Kozinski is a federal judge in California – another courageous iconoclast, but one very different from Judge Weinstein, whom I mentioned earlier. Most notably, he is nobody's idea of a bleeding-heart liberal. He supports the death penalty and has little compunction about handing down long prison sentences. He tells a moving story about the time he was asked to sentence a young woman named Catherine Ponce – a woman with no criminal record who stupidly agreed to arrange a drug deal for a man who turned out to be an undercover drug-enforcement agent (Kozinski, Reference Kozinski1997). No mandatory sentencing guidelines were in effect at the time, and so the field was wide open – she could be sentenced to probation, life in prison or anything in between. Judge Kozinski recounts that, as he was pondering her sentence, a seemingly unrelated thought crept into his mind:

As Judge Kozinski pondered Catherine Ponce's sentence, it hit him viscerally that Ponce was

As Judge Kozinski himself observed, this story is not without its difficulties. He did not have to violate any guidelines because none were in place at the time – that would have presented a harder issue. But still, he worried about basing a decision on the happenstance of personal experience. And the worry itself is an important part of the story. Personal experience, identification, compassion that flows for all sorts of reasons, articulated or unarticulated, will always influence decision-making. To the extent we can articulate our compassionate impulses, we can also examine and evaluate them and determine whether they are relevant. In this story, Judge Kozinski captures the most powerful aspect of compassion – the understanding of shared human fallibility. He was right to be concerned about the vagaries and fortuities of his reasoning process in this case. The happenstance of the recent incident involving his own toddler; whatever it was about Catherine Ponce that triggered his visceral identification; these do not seem like firm foundations for a sentencing jurisprudence. What makes this story powerful is his self-awareness. It plays two roles. It instructs him to err on the side of giving second chances to others, because we all make mistakes. It also reminds him that he may be making a mistake in this very case, and that he needs to be hyper-vigilant about what factors affect his sentencing decisions.

This is my argument about the role of compassion in law. It cannot resolve contests between competing viewpoints, but it does offer a way of taking the viewpoint of others, with understanding and humility. It may even offer a way of understanding the values and tensions that underlie constitutional interpretation.

How do judges reflect upon the values and ideals of documents like the US Constitution or other founding texts that require interpretation in light of changing conditions and changing values? How do they seek to understand what those ideals would require in the practical world they confront? How do judges give content to spacious indeterminate language like ‘equal protection’ and ‘due process’, and how do they resolve the inevitable tensions between principles like liberty and equality, privacy and security? People do not begin with abstract principles. They begin with moral intuitions that grow from their own experience and take shape within their own social worlds (Johnson, Reference Johnson1994), and those intuitions prove pretty unshakeable under most circumstances. Most of us are not very good at re-examining our moral intuitions under our own steam. For people to re-examine their moral intuitions, or to try to expand their moral universe, usually requires interchange with others who hold different perspectives (Bloom, Reference Bloom2004). At the very least, it requires the openness and humility to admit that other perspectives are worthy of respect.

This possibility takes us to a deeper problem with the DeShaney opinion, which is also a deeper point about compassion. Because the majority opinion assumes that the government is not required to do anything to protect its citizens, only to refrain from doing them harm, its reasoning turns on this question of who actually created the situation leading to Joshua's injuries. This is a debate about the meaning of due process of law. If, indeed, we have only the right to be left alone – if government's only role is to refrain from directly harming us – then any protection we receive from government might be portrayed as a gift, a compassionate exercise. But there is an alternative view, that due process creates affirmative duties flowing from the government to the people – not just the freedom to be left alone, but the freedom to survive and thrive: the right to be protected from private violence as well as government violence. Justice Brennan's conception of governmental responsibility under the US Constitution is very different from that of the DeShaney majority. In the watershed Supreme Court case of Goldberg v. Kelly,Footnote ²⁸ involving the due process protections due to welfare recipients, Justice Brennan reflected on the purpose of government, observing that ‘[f]rom its founding the Nation's basic commitment has been to foster the dignity and well-being of all persons within its borders’.Footnote ²⁹ Therefore, he reasoned, public assistance ‘is not mere charity, but a means to (in the words of the Preamble to the U.S. Constitution) “promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity”.’Footnote ³⁰