The Side-Effects of Imprisonment: Harm to the Family

Abstract

Imprisonment affects prisoners’ families. Empirical research tells us that some families benefit from this separation, particularly when they are victims of crime or abuse. But for many, imprisonment has a negative impact on family life, which can be intensified by other pre-existing factors. The main literature on the harm that imprisonment causes family members is empirical, but these empirical findings raise difficult moral questions. How might we approach these side-effects? One view, The Simple View, suggests these side-effects are proportionate so long as the overall harm produced does not outweigh the good punishment aims to achieve. This paper defends The Complex View. Here, proportionality requires us to think about factors beyond the magnitude of harm. Drawing upon existing moral and political philosophy, I aim to show three other factors that may alter how certain harms and benefits figure in proportionality calculations.

1 Introduction

Punishing offenders is only justified when it is proportionate. It can be disproportionate because of the harm that it causes to offenders themselves or others. Perhaps the most important group of non-offenders who are harmed as a result of imprisonment are the families of prisoners.¹ While the effects on the families have received little attention from philosophers, they have also received little support from governments. As Jill Matthews suggests, ‘prisoners’ families…remain out in the cold. No agency has any statutory responsibilities towards them…they are just not on anybody’s agenda.’² The main literature on the harms that imprisonment causes family members is empirical, but these empirical findings raise difficult moral questions.

Punishment has to be proportionate in two different ways. Drawing on the ideas of Jeff McMahan, punishment should be proportionate in (1) the narrow sense, where offenders are not harmed beyond what they are liable to receive given the crimes committed, and (2) in the wide sense, where punishment must be proportional in the sense that the harms inflicted on offenders have to be outweighed by the good that punishment achieves.³ This paper is concerned with the latter, as it is most concerned with people who are not liable for those harms. It examines whether the harms experienced by families make punishment disproportionate and explores how we should address these harms.

A simpler way to approach these harms would be to weigh the negative consequences to families against the good ones that punishment achieves. But by drawing on some important non-utilitarian ideas in moral and political philosophy, we begin to see the limitations of this approach. This paper defends The Complex View on how we should think about and weigh the negative consequences of imprisonment for family members. This view suggests that harm to families should count differently in the proportionality balance. Certain factors mean that in some cases, the harms will count for more, and in others, they will count for less.

In this paper, I propose three factors that may alter how certain harms and benefits figure in proportionality calculations. These factors include (a) option luck and the significance of choice; (b) ordinal proportionality and the importance of preserving equality amongst offenders; and (c) prioritarianism and the need to ensure that we do not make those who are badly off even worse off.

If empirical research confirms that families are being harmed unintentionally as a result of imprisonment, we have important reasons to consider the significance of these harms in our proportionality calculations. In this way, this paper begins to bridge a gap between philosophical literature on the justification of punishment and empirical prison scholarship that highlights the realities of prisoners’ families. If the side-effect harms inflicted upon families count heavily against our justifications for imprisoning offenders, it implies that the state has a responsibility to mitigate the adverse effects of imprisonment on families.

2 Understanding Prison Realities

Empirical research tells us that families are significantly affected by a relative’s imprisonment–experiences are sometimes positive for families, sometimes neutral, but often negative. As the harms caused to family members are unintended, I will refer to them as ‘side-effect harms’.

Some families benefit from imprisonment, particularly where family members are victims of the crime–imprisoning the offender may provide them with vital protection from further harm. Harms that offenders cause to families while they are not in prison might include domestic abuse, but also a wider range of harms. Offenders may bring other wrongdoers into the family home, for example, or encourage police intrusion into the home–their imprisonment would therefore protect the family from further emotional distress. Offenders may also squander resources, and so their imprisonment may improve the financial stability of the family.

Comfort finds that the period of incarceration sometimes eases the tension within family relationships:

For those who grapple directly with the untreated substance addiction or mental illness of a loved one, incarceration can bring a sense of relief – both because of respite from the daily turmoil caused by the person’s difficulties and because of the hope that the inmate may finally receive some degree of care, however partial or inadequate, for troubles that are no longer addressed through conventional forms of social assistance.⁴

Prisons offering structured visitation provide a safe environment where familial bonds can be strengthened or repaired. Studies show several benefits to allowing prisoners to have regular contact with their families and children.⁵ Richard Lippke highlights the longer-term, crime-reductive effects:

If someone outside prison cares about them and is permitted to evince that care in various ways, prisoners seem more likely to believe that they have something to live and hew to the straight and narrow for. This might bolster their efforts to stay out of trouble in prison and improve themselves through education, substance abuse treatment, and acquisition of working skills.⁶

And from a financial point of view, imprisonment may encourage prisoners to develop their skills, increasing their employment prospects, so that when they are released, they can better support their families.⁷

Although imprisonment has some benefits for families, it has an array of negative side-effects. The separation resulting from imprisonment can cause families to suffer decreased psychological and mental well-being.⁸ Families may be neglected or left unsupported, and children have been known to experience both emotional and behavioural disruptions, like depression, anxiety, and aggression.⁹ And the side-effects are not limited to emotional harm. Imprisonment can result in financial costs and a loss of economic opportunities, especially where families normally depend on the prisoner financially. While the work prisoners do in prison can provide an income that they can use to support dependents, captive work is connected to widespread structures of exploitation.¹⁰ Work is usually carried out under poor conditions and legal rules exclude prisoners from protections that other workers have. Any income prisoners earn is limited. In the United Kingdom, ‘a prisoner does not qualify for the national minimum wage in respect of any work which he does in pursuance of prison rules’.¹¹

Families may also feel they have lost control of their private life. Regular contact with the prisoner is not always possible, and even when it is, visits are heavily monitored and controlled. The stigma and negative labelling attached to imprisonment can intensify these harms further, causing families to feel unsupported by the community as a whole.¹²

The experiences of families throughout a relative’s incarceration are not uniform. These side-effects can be short or long term, and dependent upon the type and severity of the sentence. Where many families experience negative side-effects from imprisonment, these harms can often be intensified by other factors, such as gender, race, socio-economic status, low-family income, and lack of parental supervision.¹³

3 The Simple Versus Complex View

These empirical findings raise difficult normative questions about the punishment of offenders. The general issue is that punishment may be disproportionate, in one sense (what Jeff McMahan calls ‘the wide sense’)¹⁴ because the value of punishment is insufficient to justify the side-effect harms it causes. And the effect on family members is a significant side-effect harm of punishment. There are several different variations on how to count the significance of effects on family members in wide proportionality.

Here is a simple view:

The Simple View: If the overall aggregate harm that punishment causes outweighs the good that punishment achieves, the side-effects are disproportionate. If the overall aggregate harm that punishment causes does not outweigh the good that punishment achieves, the side-effects of the punishment are proportionate.

In any philosophical theory aimed at justifying punishment, punishment intends to achieve some good. Any sensible theory of punishment would agree that those goods need to be sufficient to outweigh the side-effect harms. From a consequentialist point of view, the good, i.e. any instrumental benefit of punishment, would need to be weighed against the side-effect harms of the punishment. For the retributivists, the good is the fact that offenders are held responsible in the basic desert sense and receive proportionate punishment. According to Berman, retributivist proportionality formulations are typically built upon two different conceptions. A proportionate sentence may consider internal factors such as the offender’s blameworthiness, culpability, guilt, or desert.¹⁵ Proportionality may also focus on outward factors, such as the gravity or seriousness of the crime committed.¹⁶ Given these factors, it may seem that retributivists would not give much weight to the harm experienced by family members. But even on retributivist grounds, we can still only punish if it is proportionate in the sense that the goods of desert are not outweighed by the drawbacks, including the harm that punishment causes as a side-effect to innocent people.

Whatever the goods of punishment are, they must be sufficiently valuable to outweigh the harm that punishment inflicts, including the harm caused to families. In weighing up the positive impact of punishment against the negatives, The Simple View takes into account some of the important empirical research that looks closely at how imprisonment affects the lives of people in the real world, i.e. the families of prisoners.

The Simple View is a natural way of interpreting the ideas of prominent philosophers of punishment. For example, in Why Punish the Deserving?¹⁷ Douglas Husak argues that desert alone is insufficient to justify the costs of the criminal justice system and claims that consequences also play a key role in deciding whether a particular deserved punishment ought to be inflicted. While it is not clear whether Husak adopts The Simple View, he seems committed to something like The Simple View in this account of the non-desert-based reasons to punish:

Whether an extra-desert component is actually required to justify the punishment of the deserving is a function of three variables:

• (1) the amount of any value in punishing the deserving;

• (2) the extent of the drawbacks of punishment; and

• (3) how (2) is balanced against (1).¹⁸

However, while The Simple View takes into account the drawbacks of punishment, it assumes that all harms that punishment causes weigh the same, and count against punishment in proportion to their magnitude. This has controversial implications. To take one example, it tends to disfavour the punishment of offenders with families, who will be negatively impacted by punishment, when compared with offenders without families. To take another, it treats the magnitude of harm that punishment causes a family as equally significant no matter how badly off the family is. An alternative does not make these assumptions:

The Complex View: Whether the punishment has disproportionate side-effects does not depend only on the overall aggregate harms and benefits it produces. Different factors alter the way in which harms and benefits figure in the overall proportionality calculation.

This view suggests a more sophisticated picture of proportionality and one that may have implications for the significance of harm to family members.

I defend The Complex View. But more importantly, I explore some of the main ways in which harm might bear on proportionality in a way that is not simply a function of its magnitude. Thus, doing this illuminates the significance of harm to families for proportionality.

The paper also raises another important question. When we consider costs to family members, should their significance be determined on a case-by-case basis, or should they matter systemically? The latter option is appealing. If all our acts of punishment are jointly proportionate, this would seem to ensure that each individual act is proportionate. Where offenders are equally culpable, all families and their experiences would be judged equally. Discussions on whether we should have an institution of punishment would then need to consider the expected costs to family members. And if a punishment system can be justified despite these costs, family welfare would play a central role in shaping these institutions. But from the system view, we need not refrain from punishing offenders where our individual punitive acts might seem disproportionate.

However, this approach cannot survive scrutiny, as the following example suggests:

Suppose that if we punish 1000 guilty people with 10 years in prison each, this will prevent 100,000 assaults each of which would result in an innocent person suffering harm to degree h. However, we will cause 2000 innocent people to be harmed to degree h as a side-effect. Degree h is equivalent to suffering 10 years in prison. Punishing all 1000 appears to be proportionate.

Now suppose that we punish 999 guilty people with 10 years in prison. Doing so prevents all 100,000 assaults and causes no harm to others as a side-effect. We are now deciding whether to punish the final guilty person, X. If we punish X, giving them 10 years in prison, no assaults would be prevented and 2000 people would suffer harm to degree h as a side-effect.

The system view implies that it would be permissible to punish X, which can’t be right. We might think that our acts in conjunction are proportionate, but this does not mean that each individual act is proportionate. This would depend entirely on which acts would cause additional harm as a side-effect. On this basis, it seems more plausible to justify each incremental decision.

Therefore, I adopt a more individualistic view and believe that harm to families should be approached on a case-by-case basis.

Here are three ways in which harm might matter in a way that is not just a function of its magnitude: (a) choice and the risk of harm; (b) preserving equality between offenders; and (c) prioritarianism.

I find all of these ideas plausible, though I do not intend a full defence of them. My main aim is to highlight the fact that familiar and pressing issues in moral and political philosophy have not been considered in accounts of proportionality in punishment.

4 In Defence of the Complex View

I will now explore three different ways proportionality depends on more than just the magnitude of the harm.

4.1 Making Choices and Taking Risks

One distinctive way The Complex View departs from The Simple View is the way it considers whether those harmed by punishment chose to risk harm. Do such choices affect considerations of proportionality? If there is a connection between the side-effect harms experienced by imprisonment and the previous choices made by families, it could be argued that the gravity of the harm they experience matters less. This section supports this idea. And while this view can be applied to different types of harm, this discussion will consider economic losses more specifically, where families depended on the prisoner financially and would be left in a worse position following their imprisonment.

A common theme in distributive justice is that choices can determine the significance of inequalities, which may then influence the significance of harm. To understand this idea, we can reflect on two prominent accounts of the significance of choice. The first is ‘luck egalitarianism’, a view most fully developed by Ronald Dworkin.¹⁹ The theory relies on the distinction between brute luck and option luck. Brute bad luck is bad luck that does not arise due to our choices. Option bad luck is bad luck that arises because of how we choose–for example, the risks we voluntarily take. Luck egalitarians claim that it is unjust for some to be worse off than others as a result of their bad brute luck, but it is not unjust, or not as badly unjust, for some to be worse off than others because of their bad option luck.²⁰ Where people are worse off as a result of option luck, they have a weaker claim, or perhaps no claim, to object to the subsequent inequality.²¹ Lippert-Rasmussen illustrates this core concept with an example: ‘If I freely choose to be worse off than others, say, because I want to spend most of my income on protecting a certain endangered species of flower, then the core luck egalitarian claim does not condemn the resulting inequality’.²²

Here is an implication: against the background of equal choices, we could have two people with similar options––but one chooses badly, and the other chooses well. Dworkin’s view is that the resulting inequalities do not matter as much since the harm suffered was avoidable.

A second account explains the significance of choice by appealing to the value of having or making choices. TM Scanlon influentially develops this idea in What We Owe to Each Other.²³ He writes:

By having a choice, I mean having what happens to one depends on how one reacts when presented with certain alternatives under certain conditions. The value of having such a choice lies in the reasons a person has for wanting what happens to him or her to be determined in this way. These reasons depend on the alternatives that are made available to him or her, and on the conditions under which these alternatives are presented.²⁴

Scanlon identifies three reasons why people want opportunities to make choices.²⁵ First, we want to make choices because of their predictive or instrumental value––they allow us to predict our future satisfaction and feature in the reasons people have for wanting certain outcomes.²⁶ Second, choices have representative value––they reflect our tastes, values, and affections.²⁷ And third, they have symbolic value––they represent a form of recognition that we are competent independent agents, capable of making choices.²⁸ For Scanlon, choices are valuable for these reasons regardless of whether we have free will, or whether our choices are determined by other causes.²⁹ But while the opportunity to avoid harm makes a difference, this would not apply if choices were determined by coercion or other unjust kinds of interference. Scanlon’s focus is on whether a person is placed in a sufficiently good position where they have opportunities available that would usually lead people to avoid harm. If they do, then they would have no valid complaint about an unsatisfactory outcome.³⁰ What matters is the value of having the opportunity to choose from alternatives.³¹

Of course, both accounts have received criticism. The aim is not to address these or defend one of the two accounts but to illustrate the common understanding that it often matters that we have a choice to avoid harm. Whether or not these choices matter will depend on the context in which these choices are made.

To begin to see their implications for harm to family members, we should consider different circumstances where imprisoning an offender harms their spouse.

There may be cases where a spouse has voluntarily relied on a partner’s criminal income, for example, income from the sale of drugs. If the partner is caught and sentenced to a period of imprisonment, the family relying on that income will be harmed as a result. While these types of cases raise distinct issues, they are easier to resolve. This is because the family knowingly relied on the ill-gotten gains of criminal activity. They took a risk despite the possibility of the partner being caught and imprisoned. And most importantly, they were never entitled to the proceeds of the crime in the first place. Given this, the family would have a weak, or perhaps no claim against any side-effect harms of imprisonment.

We should examine more challenging cases to illustrate the significance of choice. For example, where family members relied on a partner’s legitimate income, but they were later imprisoned. Consider two pairs of cases:

Choice One: Noah and Lily are married with children. Lily makes the decision to give up her job because Noah’s income is high. Although Noah’s income is legitimate, he does have a criminal background and was previously part of a gang. Noah commits a crime of passion and is sentenced to a period of imprisonment.

Choice Two: Michael and Alice are married with children. Alice makes the decision to give up her job because she has a child with disabilities, and there is inadequate support available. The family becomes fully dependent on Michael’s income. Michael commits a crime of passion and is sentenced to a period of imprisonment.

In both cases, all things are equal in the sense that the spouse and the children financially relied on the offender and will be left worse off following the imprisonment. Once imprisoned, Noah and Michael will no longer be able to support their families. The difference here is that the constraints on choice are varied. In Choice One, Lily could have kept her job, protecting herself from the loss of income. Noah has a criminal background, and so there was a greater risk of him committing a crime and being imprisoned.

Now, some may believe that in both circumstances, the family has no right to object to the imprisonment. It may be argued that Noah and Michael are not entitled to their jobs given that a crime has been committed, suggesting that the right to continued sustenance depends on an offender being entitled to his job. While a crime has indeed been committed, the provision of financial support in both cases was legitimate. Because of this, it would be wrong for the state to undermine Noah and Michael’s role and ability to continue to provide financial support to their family.

However, the harm that Alice suffers would have a greater bearing on the decision to punish Michael, because of the circumstances in which she chose to rely on him. Here is an explanation. Choice One is what we might call a ‘pure case’ where choices are unconstrained––this encompasses circumstances where family members have received financial support from the offender and rely on them being fully aware of the risks involved, and where they have good opportunities not to take those risks. Lily takes a risk to receive a benefit––for her economic comfort or more leisure time, for example. The second case is a ‘constrained case’ where Alice’s choice to rely on Michael was made due to social pressures. While the choice to financially rely on the offender was made, this choice may have less significance due to these factors.

It should be noted that there will be unpredictable cases, like Choice Two, where a partner commits a crime of passion and the option luck analysis would not be applicable. They have not had an opportunity to make a real and informed choice to roll the dice that their partner would commit a crime, and they would subsequently lose their income. However, there will be rare cases where a relationship has formed, and there is a chance that a partner will commit a crime (either they were in a gang or were already committing crimes). These are consistent with the option luck case. Here, the risk of a partner committing a crime and being imprisoned is high, since the non-offending spouse is usually aware of their partner’s behaviour. The Simple View cannot capture the intuitive difference between these cases, for that difference is not concerned with the magnitude of the harm that families suffer, but rather how the risk of harm came about. So this supports The Complex View.

We can now consider how the two philosophical accounts bear on the harms experienced by families. Dworkin attaches value to the choices made and the potential risks associated with those choices. He writes that ‘the possibility of loss was part of the life they chose–that it was the fair price of the possibility of gain.’³² We might say the same thing about Lily in the first of the two cases distinguished above.

However, although Dworkin does not focus on this, it is much less plausible that a choice-dependent inequality is unproblematic where the circumstances in which the choice was made were constrained, as is true in the second case. As a result of this, we might also think that there is a spectrum of cases from pure brute luck to pure option luck, where the circumstances of choice determine where on the spectrum any case is. Different factors may be relevant to determining what the basis of this spectrum is. This could include the quality of the options, the quantity of options, and the amount of control people have in making choices. In identifying the distinction between option luck and brute luck, we can see how this might connect to the pure and unconstrained cases described, in the context of side-effect harms to families. It could be argued that pure cases are less problematic as they reflect instances of bad option luck. Constrained cases, on the other hand, could be seen as instances of bad brute luck, or at least some way along the spectrum towards bad brute luck––it is these instances that may determine the side-effect harms experienced by families as disproportionate. When Lily decides to rely on Noah, we might think she takes a risk and ought to bear the consequences of that risk, including the subsequent harms that result from Noah’s imprisonment. Therefore, she has a weaker complaint against any side-effect harm following his imprisonment. In the second case, while Alice did indeed choose to rely on Michael, her options were restricted by pre-existing social factors. These constraints on her options meant that she had to make a risky choice for the best interests of her family as a whole.

There is, however, less difficulty when it comes to considering the disadvantages experienced by the children and their complaints against imprisonment. In both circumstances, the children’s objections to the imprisonment can be considered separately from the decisions made by the partners. Their complaint against imprisonment is substantially greater, as they do not take any risks–their association with Noah and Michael is involuntary, and they have no choice but to rely on their parents.

Now, it ought to be mentioned that in some instances, there are solutions to brute luck. If we take an uncontrollable circumstance, foresee the potential risks, and protect against them through insurance, Dworkin believed that brute luck could be somewhat transformed into a kind of option luck. He explains with an example:

Of course, insurance does not erase the distinction. Someone who buys medical insurance and is hit by an unexpected meteorite still suffers brute bad luck, because he is worse off than if he had bought insurance and not needed it. But he has had better option luck than if he had not bought the insurance, because his situation is better in virtue of his not having run the gamble of refusing to insure.³³

However, I have doubts that the availability of insurance would make a difference here, as this solution is difficult to apply in the context of imprisonment. While families have ample opportunities to medically insure against their loved ones, there are no options to insure against partners going to prison.³⁴

The Value of Choice account can also be applied when thinking about families negatively harmed by imprisonment. It would consider the conditions at the time the choice was made. In the first pure case described, Lily was in a sufficiently good position, with ample opportunities available to avoid harm. The choice has representative value, as the family shaped their lives in a way that expressed their interests and values. It also has symbolic value, as it reflects a judgement made by a competent, independent agent. A decision to compensate for emotional harm, for example, could come across as patronising and may not be well received by families. It might question their ability to make choices at the expected standard of competence. Taking into account these factors, the harm they experience would count less, not because they have chosen those burdens, but because they had the opportunity to avoid harm by choosing appropriately.³⁵

Now let us consider the implications of Scanlon’s account for constrained cases. The Value of Choice account would accept that the choices made in Choice Two are valuable. Although Alice’s decision was influenced by unjust social pressures outside of her control, they still have instrumental, representative, and symbolic significance. However, the difference here is that they may not be expected to bear opportunity-accompanying burdens. That is because Alice’s opportunity to avoid harm was constrained. Members of a just society can then be expected to bear costs to ameliorate the negative effects of her choices.

If the choices we make are significant, in the way that Dworkin and Scanlon suggest, this may offer support to those who believe families should have done more to prevent their relatives from committing crimes. They might argue that the spouses should have discouraged their partners from carrying out unlawful acts and that Lily and Alice are equally blameworthy–they may have had ample opportunities to do so and chose not to. However, the ability to prevent them from committing a crime may have been hindered by other constraints, and therefore, it may not be appropriate for the spouses to bear the consequences of that choice.

This reasoning again illustrates the distinctiveness of The Complex View, highlighting one of the challenges that may be encountered when exploring the side-effect harm to families. In pure cases, the consequences of punishment to the family may not be considered as bad. However, we ought to recognise those cases where the family’s welfare depends on the partner facing imprisonment, and the choice to previously depend on them is not based on the acceptance of risk. In these constrained cases, there is less significance attached to choice and the harm they suffer matters more.

Understanding these experiences involves taking into account a variety of factors that will often be difficult to resolve––there may not always be a clear solution. Furthermore, efforts to explore the choices of families should not be seen as an attempt to stigmatise and place blame on them for the harm they experience. However, this section has aimed to show that consideration of financial harm to families involves more than the extent of the harm suffered. It shows how making choices and taking risks may change the way we view harms and benefits in proportionality calculations and the decision to imprison offenders.

4.2 Unfairness Between Defendants

Another failing of The Simple View is that it overlooks a potential unfairness between different defendants about how punishment is distributed. The pressure to preserve equality amongst prisoners may influence the way in which harms and benefits figure in the overall proportionality calculation.

Giving special consideration to offenders with families may be unfair to unpopular offenders––i.e. those without families or people who depend on them.

Consider the following case: George and Brad are both equals, in the sense that they have committed the same crime and would be due to receive the same level of punishment. The only difference between them is that George has a wife and children. He and his wife both provide an income and share childcare duties between them. In the event of George’s imprisonment, financial burdens may arise as his wife is now required to take on additional childcare duties––which may mean further costs from making alternative childcare arrangements. Not only this, but the family may also suffer emotional distress due to the separation caused by George’s imprisonment. Brad, on the other hand, is not in a relationship and has no children or anyone who would normally depend on him–no one else would suffer side-effect harms because of his imprisonment.

Now suppose that authorities must choose between assigning one offender a prison sentence and the other offender a term of probation. Of course, George’s imprisonment is likely to result in some additional side-effect harm to his wife and children. But is it unfair to take account of these harms in determining whom to imprison? Here is a more perspicuous way of asking this question: is there greater value in preserving equality between these two offenders, or are we required to interfere with equality when doing so prevents one family from being left worse off?

One approach to answering this question draws on the principle of irrelevant utilities, first articulated by Frances Kamm.³⁶ This is the principle that some goods or bads do not count in our decisions, such as our decisions on who to benefit or harm. She uses the following case to explain this principle:

Sore Throat: A scarce medicine could be given to save the life of either Jim or Joe, but only if it is given to Jim will there be enough left over to cure a sore throat that Nancy would otherwise have for a week.³⁷

Giving the medicine to Jim would maximise utility, as it would also cure Nancy’s sore throat. However, Kamm’s view, and the principle of irrelevant utilities, suggests that it would be wrong to provide Jim with the medicine solely because of this extra utility. Doing this would ‘treat people “merely as a means” since it decides against the person who cannot produce this extra utility on the grounds that he is not a means. It does not give people equal status as “ends in themselves” and, therefore, treats them unfairly.’³⁸ Therefore, in cases where the additional minor harm (such as a sore throat) seems irrelevant, a randomising procedure that would decide whom to save ought to be the automatic and preferred response.³⁹

However, this idea may be considered less plausible if the costs to families following imprisonment are high. In the original Sore Throat, Kamm fairly claims that it would be wrong to decide to save Jim due to the additional good in gaining a sore-throat cure, and we have good reasons to support this. But would the response be the same if the additional harm was grave? According to Victor Tadros, the only reason why we should depart from a randomised procedure is when the duty-grounding force of the harm is sufficiently strong.⁴⁰ He explains that a randomised procedure is preferred when we face two options that are morally close to each other (though there may be circumstances where we must consider two options that are morally distant from each other in other respects).⁴¹ What if, instead of a sore throat, Nancy would suffer greater harm, such as death or grave disability? It is more plausible to depart from a randomised approach in this case–the value of death or grave disability to Nancy would defeat the duty-grounding force of Joe’s chance of survival.⁴² This analysis therefore raises implications for prison sentencing, particularly for cases where the costs to the families are grave.

Reflecting on the alternative version of Sore Throat illustrates important reasons why we might choose to depart from a randomised procedure. What we ought to do in these cases appears to depend on the level of additional harm that would be caused. If we are faced with a situation where we can save either A or B, but in saving B, we also prevent C from getting a sore throat, it seems that we should flip a coin to determine who should be saved. But if we are faced with an alternative case, where we can save either A or B, but in saving B, we also prevent C from losing a leg, saving B seems to be a better option than saving A.

The theory of irrelevant utilities can be applied in the context of punishment and the additional harm against the families. We can return to the choice of imprisoning George or Brad. In this example, we lack adequate prison places to imprison both offenders so whoever we choose not to imprison will be put on a period of probation. By imprisoning George, his family will suffer some emotional harm. Here, it may be thought of as unfair to depart from equality on the basis that George’s family will experience some side-effect harms. While the family will be upset by what has happened and suffer minor emotional harm, the costs to George’s family are relatively small––we lack good reasons to depart from equality. It seems plausible to preserve equality if the side-effect harms to families are expected to be modest, and use a randomised decision procedure, like flipping a coin, to determine whom we should imprison in this instance. It is better that George’s family bear some cost, rather than exposing Brad to a greater chance of being imprisoned. It seems unfair to think that the type of punishment Brad receives would be re-considered if he had a family or other dependents who would be negatively affected by his imprisonment.

However, if imprisoning George causes his family to suffer substantial emotional and financial harm, it seems unfair to approach this case by using a randomised procedure. We ought to interfere with equality and imprison Brad–assigning George the period of probation–since the costs to the family are more significant and the additional goods we would gain are high.

The theory of irrelevant utilities is mostly applicable in cases where families have been left worse off because the offender cannot be used to help them, like those cases where families suffer considerable financial disadvantage as a direct result of the offenders’ imprisonment. Equality may be preferred in cases where families suffer some emotional harm, as well as in wider contexts, for example, in the allocation of resources to families. Consider two offenders who are both in need of urgent lifesaving medical treatment. One offender has a family who will suffer emotionally should he not receive the treatment. However, the other offender has no family who would be impacted by a denial of treatment. In this instance, the appeal to equality would be the most intuitive response. It would be unfair to deny the unpopular offender medical treatment on the basis that he has no one who would be impacted by his death. This shows that there may be situations where it would be more appropriate to adhere to principles of equality. In more complex cases, there is a wider cause for concern when families are expected to suffer grave harm–it raises a greater need to depart from equality and their suffering may count for more in calculations of proportionality.

The ideas explored in this section give rise to The Complex View. The idea that we ought to depart from equality when the additional harms posed to the family are substantial aims to align with moral instinct. It means that we can acknowledge that families will be impacted by imprisonment, but how we respond to the side-effects will depend on the type and level of harm they are expected to suffer.

4.3 Equality Versus Priority

Imprisonment can make family members worse off. But from what starting point? And does the starting point make a difference? Other factors may already be affecting families, including difficulties surrounding unemployment, mental health problems, abuse or neglect, human capital/skills deficits, or marital difficulties. Families are also affected as individuals, often by broader forms of marginalisation and discrimination, including racial, economic, and gender inequality.⁴³ This is true of many, but not all, families of offenders.

Suppose we have two families who are expecting to suffer financial harm as a result of a relative’s imprisonment. For Family 1, the offender was in a high-paying job and provided a legitimate income to the family. While the family will still be reasonably well off, they will be in a worse position if we were to imprison the offender. However, for Family 2, the offender was unemployed, and the family was already experiencing financial disadvantage. The family was already badly off and imprisoning the offender would work to aggravate their existing circumstances. In this instance, if we only have the resources to imprison one offender and assign the other a period of probation, should the fact that Family 2 is already badly off mean we ought to give them special consideration?

Two prominent accounts in philosophy support a positive answer. Firstly, we can draw on the principle of prioritarianism, which has been both articulated and defended by writers such as Derek Parfit⁴⁴ and Thomas Nagel.⁴⁵ According to prioritarianism, it is more important to benefit those who are worse off and give them priority in the distribution of advantages.⁴⁶ This is not because of any inequality between those who are worse off than others. Rather, Parfit believed that we ought to benefit those who are worse off because they are at a lower absolute level––the process involves considering how badly off someone would be on an individual basis.⁴⁷ While it is important to show equal concern for everyone’s good, we should favour benefitting the worse off when it comes to making a choice about whom to benefit. Nagel explains that there is a noticeable ranking of urgency: ‘The claims on our impartial concern of an individual who is badly off present themselves as having priority over the claims of each individual who is better off: as being ahead in the queue, so to speak.’⁴⁸ He applies these views to the following hypothetical case:

Suppose I must decide between moving to an expensive city where the second child can receive special medical treatment and schooling, but where the family’s standard of living will be lower and the neighbourhood will be unpleasant and dangerous for the first child – or else moving to a pleasant semi-rural suburb where the first child, who has a special interest in sports and nature, can have a free and agreeable life…⁴⁹

Nagel plausibly claims that it is more urgent to benefit the second child, and this is so, even if the benefit to him would be less than the possible benefits gained in accommodating the first child’s needs.⁵⁰ The priority is not dependent upon how great the benefits would be–this would be the utilitarian approach. To conclude that we ought to benefit the second child, we would only need to determine how badly off he is in this case.⁵¹

In applying the priority view in the context of imprisonment, some side-effect harms may count for more when the families of offenders are badly off. The family was already badly off, and so imprisoning the offender would work to exacerbate already existing stressors within the family. Of course, this view does not intend to undermine the harm that Family 1 expects to suffer––both families will be negatively affected by the relative’s imprisonment. Rather, it argues that we ought to give priority to Family 2, as it is more urgent to benefit them over Family 1. While it may be true that the first family will lose more financially as a result of a relative’s imprisonment, the potential benefits to the second family would count for more––‘benefits to the worse off do more to make the outcome better.’⁵² Benefitting the families who are already being impacted by social pressures seems to be the most intuitive response. Improving the lives of those worse off and highlighting the need to benefit them is more urgent than making improvements that will inevitably benefit those who are ultimately already better off.⁵³

When and why should the state in its capacity as a proportional punisher need to reduce the amount of harm to families of offenders? Prioritarianism is an individualistic theory––the strength of our reasons to benefit a person depends on how badly off they already are. If we consider taking away a unit of welfare from a person who is well off and a unit of welfare from someone who is badly off, there are stronger reasons against taking away welfare from a person who is already badly off. This does have distributive implications because, in this way, we are comparing two different people and the reasons for moving goods between them. However, insofar as the criminal justice system affects family members who are already badly off, the basic prioritarian thought is that we should not make them even worse off.

Alternatively, the egalitarian view would also take into consideration how badly off a person is. The difference though, is that egalitarianism is a relational view. This means that, when faced with a choice about whom to benefit in a situation, we would need to consider what levels other people are at.⁵⁴ This would require us to ‘invoke interpersonal considerations that are essentially relational, such as the intrinsic badness of inequality or the comparative strength of the claims of different individuals.’⁵⁵ From the egalitarian perspective, the practice of imprisonment and its impact on families is problematic because it is an institution that is known to exacerbate inequalities between families. We can compare families of offenders who are worse off to families who are better off, where relatives tend not to offend, and if they do, they may not get caught and imprisoned. When choosing between the two families, the egalitarian would––like the priority view–choose to benefit Family 2, but not for the same reasons. After carrying out a comparative examination of the two-family circumstances, the egalitarian would conclude that it is bad in itself, or unjust, that some people are worse off than others–and therefore, we should benefit the family who would be worse off. The aim is to reduce the inequalities between the two families.⁵⁶

Despite their differences, both accounts would agree that making those who are badly off worse counts as a reason against punishment. The prioritarian suggests we ought to refrain from making people worse off by considering their absolute position in society. The egalitarian’s reasons pertain to a person’s relative position in society, exploring considerations of fairness and principles of equality. Some may take a pluralist view and believe that matters of priority and equality both have value. And others may suggest that other factors have force, such as considerations of sufficiency or compassion.⁵⁷ While people may not agree on which view is most convincing, many would agree that what matters is not just how much harm is caused as a result of imprisonment, but to whom the harm is allocated, and its distribution in society. Of course, some of the disadvantages experienced by families will be unavoidable. While we ought to anticipate potential harms, there is no guarantee that families will benefit when alternative responses to wrongdoing are sought out. However, The Complex View recognises that in some cases, imprisonment could contribute to existing familial disadvantage and work to ‘make a bad situation worse’. As explained by Condry and Minson:

The harms of imprisonment sustain and reproduce patterns of inequality, have a profound impact upon the social fabric and organisation of society both in the current context and in the future, and affect the way in which we perceive the place of individuals in a democratic society.⁵⁸

Ultimately, the view encourages us to think about the real-world implications of punishment for the family and explains why they ought to be considered in our proportionality calculations. For example, imprisoning a family member may result in the family being plunged further into poverty and desperation. We can also think about the childcare burdens that stem from parental imprisonment. In single-parent families, children may have to live with relatives or family friends whose financial and emotional resources are already stretched to its limits.⁵⁹ Where this is not possible, children may be placed into foster care––a system that often results in further disadvantage.⁶⁰ Therefore, if we are faced with a choice between preventing severe hardship for a family who is very poor and deprived, and preventing less severe hardship for a family who is better off, it seems most plausible to benefit the family who is already badly off, rather than making them even worse off.

5 Conclusion

This paper presents an alternative way to approach imprisonment’s side-effect harms to families. It begins by outlining The Simple View, which suggests that the side-effects ought to be approached by weighing up the good achieved by imprisonment against the harm experienced by family members. That view fails to take into account the complexities that arise when calculating proportionality.

Therefore, this paper defends The Complex View. This view argues that considerations of proportionality should be prepared to contemplate other factors, which may mean that, all things considered, some of the more alarming side-effect harms that seem to have significance may not count as much on the scales of justice.

Drawing on some distinct areas in moral and political philosophy, this discussion reflects on three factors that may influence how we view the harm experienced by families––factors that go beyond the magnitude of the harm.

It begins by examining the value of choice, and whether decisions made by families make a difference–objections to financial harm may have less strength when financial reliance is unconstrained and based on the acceptance of a fully informed risk.

It then explores issues of unfairness, highlighting the position of less-popular offenders and the challenges faced in the effort to preserve principles of equality.

Last, it considers how imprisonment can work to aggravate already-present stressors in the family––justice may require that priority be given to families already experiencing disadvantage.

Proportionality involves more than just weighing up the benefits of imprisonment against its resulting harms. It involves taking into account additional factors, and in doing so, we may find that some side-effect harms may count more than others. And these are just some of the issues that may arise––The Complex View accepts that there may be other challenges that affect proportionality. This view can also be extended beyond the side-effect harm to families, to the prisoners themselves.