1 Introduction

It is well known that persons with criminal records (PCR)Footnote 1 will often find it more difficult to access housing than otherwise similar people without a criminal record (Leasure & Martin, 2017). Moreover, because of this, PCR experience more, and longer, periods of homelessness than is typical in the general population (Caton et al., 2005, Legal Action Center, 2016, Bowman & Ely, 2020). These difficulties are not fully explained by factors such as lack of education, low income, unemployment, and the mere possession of a criminal record (Petersen, 2016), or by the fact that PCR will often be coping with certain personality disorders, such as borderline or anti-social personality disorder, or have a history of substance abuse. A further, and important, barrier is the existence, in a significant number of countries, of laws, and municipal policies sanctioned by them, that permit the authorities managing public as well as private housing to deny prospective tenants with a criminal record access to housing (Hoskins, 2019; Rubenstein & Mukamal, 2002; van Tongeren, 2022).Footnote 2 Many of the housing providers and authorities exclude PCR from housing without offering alternative accommodation.

In what follows, we shall focus on the policy of excluding PCR without offering alternative housing, which is widely applied in such countries as the USA,Footnote 3 the UK,Footnote 4 The Netherlands (van Tongeren, 2022) and Denmark.Footnote 5 Housing authorities in these countries can, in addition, evict tenants who are engaged in criminal activity [see e.g. van Tongeren (2022), Legal Action Center (2016)]. There is no doubt that difficulty in finding and retaining safe and affordable housing, or being evicted from one’s home, has a tremendous negative effect on the well-being of many PCR. For, relatively few PCR can rely on their own resources, or on family and friends, for housing, and some will therefore end up homeless. Homelessness is devastating to the well-being of PCR (and most other people) and their relatives, as homeless people have an increased risk of early mortality, drug abuse, being the victims of violent attacks, and unemployment (Tweed et al., 2022; Mayer & Reichert, 2020). Moreover, studies show that no, or poor, access to secure and affordable housing increases rates of recidivism in comparison with those among PCR who do have a stable housing situation (Jacobs & Gottlieb, 2020; Legal Action Center, 2016; Fontaine, 2013).

Given that the laws and practices touched upon above have such negative outcomes, not only for PCR, but also for future victims of crime, it is reasonable to expect there to be strong reasons in favor of them. Why deny, or even ensure there is a possibility of denying, ex-offenders access to such an important social good as housing when they have already been punished by the state? Housing is, after all, a social good which, according to the Universal Declaration Human Rights, is also a human right (UDHU, article 25, 1948).

This article critically discusses some answers to this question. It proceeds as follows. In Sect. 2, we present a few examples illustrating the way in which housing authorities, with the law as an aid, can make it extremely difficult for PCR to access housing, sometimes denying access outright. In Sect. 3, we examine two central reasons for denying PCR access to housing—namely, that doing so prevents crime, and that it reduces public fear of crime. These reasons are often canvassed in the literature on PCR and housing, but to our knowledge they have not been discussed, or challenged, in a detailed manner. In Sect. 4, we provide an argument which concludes that the current laws and practices allow wrongful discrimination. Section 5 sums up the paper’s findings.

Before we move on with these analyses, a further comment on the relevance of this article is worth making. The sheer number of people with a criminal record is enormous. In the UK alone approximately 18 million people had a criminal record in 2018, and in the US the number for 2018 was approximately 77 million people (NCSL, 2018). This implies that around 30% of the adult population in these two countries had a criminal record.Footnote 6 Therefore, which policy we have for the housing of PCR has a huge impact on the well-being of not only PCR, but also for their relatives, and, as we shall see in Sect. 3, for society. Especially when we know that these housing policies, as least in the US, restrict access to those with a criminal record of any sort, no matter how minor the crime on the record is (Legal Action Center, 2016).

2 Home Sweet Home: But No PCR in My Neighborhood

A number of public housing authorities and private property owners have outright bans against housing people with a criminal record (Legal Action Center, 2016). In the US some of these bans are federally mandated. In federal law, housing providers must impose permanent bans on PCR who are required to register as sex offenders for life or are convicted of producing methamphetamine on federally assisted property (Crowell, 2017). US public housing authorities also have broad discretion to set their own strict screening criteria, sometimes resulting in them imposing outright bans on the housing of PCR (Legal Action Center, 2016). There are numerous examples of these bans. Let us briefly describe some of these.

In Georgia, an apartment complex makes it noticeably clear that a prospective tenant with a criminal record, no matter what the offence is, or how old it is, shall not be considered for tenancy. The complex has almost the same policy on anyone who has been charged with, or arrested for, an offence, even if the individual has not been convicted of the offence (Legal Action Center, 2016). The wording of the relevant part of the rental criteria is as follows:

Applicants with felony convictions will be denied. Felony arrest with no conviction must be 10 years old and may be denied based on charge. Misdemeanor arrest, and/or conviction less than three years old may not be accepted (Legal Action Center, 2016, 5).

Another example we want to mention is the Housing Authorities of Baltimore City (HABC) which, in its rental criteria, writes: “HABC denies eligibility for admission based on any history of criminal activity on the part of any applicant’s family members, including drug-related criminal activity” (Legal Action Center, 2016, 8). It is worth adding that strict admission criteria like those above are relatively easily enforced by housing authorities in the US, given the existence of publicly available criminal records. A swift search on the internet can reveal whether or not a prospective tenant has a criminal record.Footnote 7 In other countries, such as Denmark and the UK, housing providers “only” have indirect access to this information. To obtain access to criminal records, which are held by the police and the courts, providers must obtain the informed consent of the relevant applicants.

Outside the US, Danish public housing administrators have been permitted by Danish law, since 2018, to deny housing to prospective tenants if any of the individuals who will occupy the dwelling have committed offences “considered to be relevant to the safety of the relevant residential area” (Indenrigs- og Boligministeriet 2018, § 29 stk 2, our translation). In the Netherlands, the Extended Urban Areas Special Measure 2016 similarly makes it lawful for authorities to systematically ban PCR from living in certain residential areas (van Tongeren, 2022). Moreover, in England and Wales, the Homelessness Act 2002 allows housing providers to deny housing to persons deemed unsuitable as tenants as a result of their past unacceptable behavior, of which at least some kind of criminal behavior counts as unacceptable (The National Archives, 2002, Sect. 14).

Clearly, therefore, it is not uncommon for housing authorities to be permitted (or even required) to operate policies denying housing to PCR. What is less clear are the reasons that may be said to support such policies, and whether these reasons are ethically defensible. In the next section, we turn to these two questions.

3 Is the Exclusion of Ex-offenders from Housing Effective in Preventing Crime or Reducing Public Fear of Crime?

It is no surprise to learn that the residents of a neighborhood often attempt to protect themselves from negative developments that may, for example, threaten their safety. When it comes to the housing of PCR, several studies have shown that residents do not wish to have ex-offenders living in their neighborhood, and this is for at least two reasons (Hoskins, 2019; Wynne-Edwards, 2003; van Tongeren, 2022). First, they believe that the fewer ex-offenders who are permitted to live in their neighborhood, the fewer crimes will be committed there. Second, they think that having fewer ex-offenders in their neighborhood will reduce the fear of becoming a victim of a crime in the locality. As van Tongeren highlights, at a more general level it can plausibly be suggested, as well, that the state has a moral duty

… to act when an ex-offender’s return into society results in a (possible) threat to residents’ (feelings of) safety. Although it does not entitle the ex-offender’s (future) neighbours with a fundamental right as such which they can invoke to force the state to act, it does provide public authorities with an intrinsic obligation to take into account citizens’ safety and quality of life when housing an ex-offender in their neighborhood. (van Tongeren, 2022, p. 62).

It is important to understand that we are talking here about two distinct types of reason (though of course these can be related). It is one thing to be exposed to an increased risk of being a victim of crime, but quite another to experience increased fear of being a victim of a crime. These two worries may of course be causally connected, in the sense that the increased risk will also, often, increase the fear. However, there is no necessary connection between the two worries. Studies have shown that fear of crime may increase among the population even as the risk of actually being a victim of crime decreases (e.g. Curiel & Bishop, 2018).

In what follows, we will critically discuss these arguments in support of the current practice of denying prospective tenants with criminal records housing.

3.1 Crime Prevention

We can understand the argument from crime prevention in at least two separate ways. On the one hand, it could be argued that the current use of laws and practices of excluding PCR from housing prevent crime in the specific areas from which PCR are excluded. On the other hand, the suggestion may be that these laws and practices prevent crime for everyone, through the general deterrence of crime everywhere in our society. The latter claim about a general deterrent effect presupposes that the public is informed about, and understands, the law, and how often housing authorities’ access the criminal records of people applying for tenancy.

If these claims about crime prevention are true, we have a justified reason to stick with the current practice. If they are not, these claims may be pure speculation. In our view, several considerations weaken the appeal to crime prevention. First, we have not been able to find evidence for the assertion that the current practice of denying ex-offenders housing is indeed crime preventive (either in reducing crime in a particular neighborhood or in reducing it more generally in society).Footnote 8 In fact, there are studies showing that some of these practices do not reduce crime in generalFootnote 9 and may even promote crime. A meta study by Bratina (2013) shows that such measures are either ineffective or counterproductive, from the viewpoint of crime prevention, when it comes to the practice of not allowing sex offenders to live, for example, within a mile of schools and playgrounds. This study concludes that “…residency restrictions have negative consequences on outcome, whether collateral consequences, re-offending or proximity to targets (children)” (Bratina, 2013, p. 214). The negative consequences are connected with several things, but chiefly the fact that isolating ex-offenders from supportive family and friends, and making it harder for them to get a job or proper housing, make re-offence very likely (Bratina, 2013, p. 209). Another study by Alm and Bäckman (2022) showed that housing eviction tended to increase the risk of criminal conviction in respect of all the crimes that were the target of the study, including, for example, drug offences and violent crimes. The Alm and Bäckman Study concludes that while that are several reasons why people evicted from housing tend to become more engaged in criminal activity (e.g. poverty, lack of employment, mental illness), “… an eviction event [also] appears to have an independent effect on subsequent criminality” (Alm & Bäckman, 2022, p. 628). From these few studies we should not, of course, infer that excluding offenders from housing will not sometimes decrease criminal activity within an area or in general. Our point is more modest. Namely, that when it comes to crime prevention, there is almost no evidence that excluding ex-offenders will decrease crime, while there are several studies showing that this type of exclusion increases rates of crime in general. Although there are some studies showing that gathering a large concentration of people in a neighborhood who are, for example, on low-incomes, or who are PCR, increases crime within that area, we believe this observation does not speak in favor of current laws and practices (Ludwig & Burdick-Will, 2012).Footnote 10 For one way to decrease the concentration of PCR in a neighborhood is to allow PCR access to housing opportunities in as many neighborhoods as possible, rather than following the current laws and practices, which deny them housing in many neighborhoods. Furthermore, it is precisely as a result of the current situation that PCR gather for housing in areas that do not exclude ex-offenders. From a crime preventive point of view, this once again speaks against the current laws and practices.

Second, even if it were true that the current laws and practices prevent crimes in certain neighborhoods, this would not necessarily provide a convincing justification for those laws and practices. For it seems that the overall preventive effect would be greater if we abandoned the current laws and practices. This policy would be “overall preventive” in the sense that the crime preventive effective in certain neighborhoods (which could be negative, although we have argued for the opposite) would be outweighed by general reduction of recidivism in society. In other words, the crime preventive effect will be greater because, as we observed in the introduction, better housing opportunities make it easier for PCR to reduce their risk of recidivism and to cut the risk of people becoming victims of crime. If this is correct, then, by abandoning the present laws and practices we will very likely ensure that, all things considered, we have fewer criminals, fewer crimes, and fewer victims than we would have had under the current laws and practice. As this is indeed the case, the current rules are counterproductive in the effort to reduce crime. In cases where a change to the laws and practices concerning housing along the lines we have proposed would increase criminality within one neighborhood, it seems right to offer the residents within that neighborhood some residual protections. In Sect. 3.2, we describe how such protections (e.g. increased community policing) could also be part of reducing the fear of crime within a neighborhood where people with a criminal record are afforded access to housing.

Third, when it comes to preventing crime within neighborhoods, there are several alternative ways to combat crime other than by the exclusion of people with a criminal record. These include: increased surveillance, citizen patrols, and better lighting. Since several studies have shown that such measures work, we should use them instead of introducing housing criteria whose preventive effect is far from well supported by empirical data (Lap 2019). We should do this because denying people access to housing should be a last resort—one to be adopted only if, and when, we know it will prevent crime.Footnote 11

We suggest, then, that instead of making decisions based on speculations about the crime preventive effect of excluding PCR from housing, we should act on the basis of what we know. What we know, as we saw in the Introduction, and in this section, is that affordable and safe housing is an important insurance against re-offending. Therefore, making it illegal for providers to deny PCR housing, and ending the current laws and practices of exclusion, will certainly offer more housing opportunities to ex-offenders than they have now. And since we know that access to safe and affordable housing will mean less crime and fewer victims, we have a strong reason to abandon the currently widespread practice.Footnote 12

3.2 Reducing Fear of Crime

If what was argued in the last section is correct, the view that excluding PCR from housing is permissible because doing so will prevent crime should be rejected. However, as indicated above, it is sometimes argued by policy makers, housing authorities and others, that there is another important reason why PCR can permissibly be excluded from housing. This is that granting such people housing will lead to more fear of crime among the other occupants in the building the PCR will reside in, or in the wider residential area. Studies have documented this worry. Doble and Lindsay, for example, found community members to be extremely fearful of what would happen if ex-offenders were living in their neighborhood. “Safety was by far the neighbor’s greatest concern” (Doble & Lindsay, 2003, p. 2).Footnote 13

Following contemporary fear-of-crime researchers, we shall understand the fear of crime as a negative “emotional response to potential victimization” (Rader, 2017, p. 2). Now, considered in isolation, it is surely an ethically noble aim to attempt to reduce (or at least, not heighten) levels of the fear of crime among residents. For one thing, it is speculated that fear of crime causes, or at least, worsens, mental health issues among certain groups (Rader, 2017, Lorenc et al., 2013).Footnote 14 At a more general level, being afraid (of victimization) is an emotion most of us would surely prefer to avoid—one whose absence would make our lives go better. Hence, it seems that if excluding PCR from housing reduces fear of crime, this provides us with a reason for the exclusion. In our view, however, the fear of crime rationale suffers from at least three shortcomings.

First, those who refer to mitigation of the fear of crime as a rationale for excluding ex-offenders from housing do not seem to recognize that not all forms of fear should be given equal weight in our decision-making. Arguably, fear of crime that is irrational should be considered of little moral importance when deciding whether to exclude PCR. This point has also been made by Zachary Hoskins (2019), who, in discussing the general question of the permissibility of legal restrictions of ex-offenders’ behavior (including access to housing), swiftly rejects the notion that such restrictions can be motivated with reference to the feelings of potential victims. As Hoskins observes “People may be fearful of others for any number of irrational reasons—skin color, religion, and so on—that are not appropriate considerations in deciding whether to impose restrictive policies on certain segments of the population” (Hoskins, 2019, p. 179). What Hoskins highlights here, in our view rightly, is that what residents report, or experience, as fear of crime may well be based on features of other people that should not be given decisive weight in decision-making.

We might add another, partly overlapping, way in which fear of crime can be of a kind that should not be accorded paramount weight in our decision-making. Namely, if the fear persists even in the face of evidence that shows it to be very probably groundless. Suppose, for example, that a resident of a public housing facility learns that a convicted rapist wishes to move into the apartment next door upon being released from prison. Suppose also that she is informed that the event that led to the conviction happened 20 years ago, that the ex-offender has been a model inmate during his imprisonment, that he has expressed sincere regret about his criminal action, and that he—now 81 years old—belongs to a group of offenders who are very unlikely to reoffend upon release.Footnote 15 In this case, it would arguably be irrational for the resident to fear being victimized by the ex-offender, since the evidence shows this fear to be in all probability groundless. However, there may of course be other types of case in which fear of crime on the part of residents is not irrational. We say more about this shortly.

We share the intuition with Hoskins that it seems wrong to exclude the ex-offender based on either of these two types of irrational fear. However, we also believe that the intuition can be supported by theoretical considerations. Specifically, it seems plausible that relative to the welfare diminishment people may experience by rational fear of crime, such diminishment caused by an irrational fear of crime can often be addressed in ways that do considerably less harm than denying housing to the PCR. The irrational fear may, for example, be addressed through therapy, although not all will be able to pay for, and thus benefit from, this. Other ways may include informing the victim of the actual risk of becoming a victim again, or generally increasing neighborhood watch in the relevant neighborhood (see below).Footnote 16 If this is correct, excluding PCR on the basis of irrational fear of crime is morally dubious, because the harm caused to the ex-offender by that policy (and, in the long run, the harm to potential victims of crime caused partly by increased recidivism among those who are homeless) could have been avoided by addressing public fear in other, morally less costly, ways.Footnote 17

Importantly, our point is not that the fear element in this rationale for PCR exclusion can never be directed at the appropriate target: fearing others because they are prone to crime and fearing them because of the color of their skin, etc. are very different matters. Nor is it that such fear can never be supported, or at least not undermined, by the evidence. Rather, our point is that someone who wishes to exclude an ex-offender from housing because this will reduce fear of crime needs to demonstrate that the fear that is being reduced is the relevant kind of fear. It is not sufficient, in other words, to show that fear of crime may, or will, become more prevalent among residents if people with a criminal record are granted housing in the locality: it must be made plausible that the form of fear being anticipated is of a kind that makes it appropriate to base housing restrictions on it. As we see it, it must be shown that the fear is not of a type that would be more effectively addressed by means other than excluding the PCR. No proponents of the fear of crime rationale for restrictive housing have, to the best of our knowledge, provided such a demonstration.

Second, even if it is demonstrated that the fear reported is of the appropriate kind, there may be other measures that can be employed to counteract the heightened fear of crime in the relevant residential area. Here, besides the interventions we have already mentioned above [e.g. increasing surveillance and street lighting (Struyf 2020)], there are methods of policing usually referred to as proactive policing and community policing. These methods include increasing the opportunities for citizens to personally interact with police officers—e.g. by introducing substations and foot patrols in the neighborhood, or through police officers establishing personal contact with at least one adult residing in each household in the community. Officers can also be active participants in mobilizing the community, perhaps through police-sponsored events (Zhao et al. 2002).

Although it should be noted that the sample sizes in review studies are often small, one review of the literature, examining 19 studies of the effect of proactive and community policing on levels of public fear of crime, found that the majority of publications (74%) reported that community policing reduced fear of crime among residents (Zhao et al. 2002). A more recent review found that slightly more than half of the 16 reviewed studies of community policing (56.3%) seemed to confirm reductions in public fear of crime (Crowl, 2017). As noted above, more work seems to be needed to corroborate these results. But if indeed, it turns out that community policing is an effective way to reduce fear of crime, this (and/or other measures) could be deployed alongside allowing PCR access to housing to ensure that there is no net negative effect on the local level of fear of crime.

Third, and finally, even if we suppose that there is no effective way to reduce, or avoid raising, the fear of crime, among residents, that granting those with a criminal record access to housing may result in, it is far from clear that the reduction of fear of crime should be preferred over the prevention of crime. After all, although it is assuredly unpleasant to fear becoming a victim of a crime, surely it is more unpleasant to be a victim of that crime. However, when we allow those with a criminal record to be excluded from housing in order to prevent fear of crime we do, in fact, seem to enter a trade-off: we trade crime prevention for fear prevention. Now, in any given case it is of course possible that the amount of harm obviated by preventing fear of crime is greater than the harm obviated by crime prevention. And if this is indeed the case, then the tradeoff is, from a harm-reduction perspective, at least, the correct one. However, given what is at stake, whether this really is the case is something that must be argued for rather than assumed. We are aware of no studies showing that fear-related harm in the case of PCR exclusion is, or is very likely to be, greater than crime-related harm—either in general or in a specific kind of case.

To briefly summarize, this section has considered whether an anticipated increase in the level of fear of crime that will be experienced by other occupants of a residential area in which a PCR is to be housed provides us with a good reason to exclude the PCR from the housing in question. We have argued that this does not seem to be the case. First, those who disagree with us have not persuasively demonstrated that the fear experienced by occupants is of a kind on which it is morally appropriate to base restrictive decisions. Second, measures that seem to have a positive effect on the level of people’s fear of crime, such as proactive or community policing, could be employed to ensure that there is no net rise in fear of crime when the PCR is granted housing. Third, we have argued that, due to the positive effects that housing seems to have on levels of recidivism, preventing fear of crime by excluding PCR amounts to a trade-off of crime prevention for fear of crime, and that it is not at all clear that this trade-off is morally desirable.

4 A Discrimination-Based Argument Against the Exclusion of PCR from Housing

In the previous sections we have merely pointed to weaknesses in key arguments in favor of the exclusion of PCR from housing—arguments based on the tendency of the exclusion to prevent crime, or the fear of it. In this section, we shall go further and present an argument against excluding PCR from housing which seems to put even more pressure on the idea that this practice is morally defensible. This argument takes as a starting point the fact that housing providers are often legally prohibited from making housing unavailable to people because they belong to certain socially salient groups. For example, in the US the Fair Housing Act, approved in 1968, explicitly prohibits denying housing to someone on the basis of their race or color, sex, religion, national origin, familial status, or disability (Department of Justice, 2022). Similarly, the UK the Equality Act 2010 rules out discrimination against prospective renters based on their gender, disability, race, religion or belief, sexual orientation, gender reassignment, and pregnancy or maternity (The National Archives, 2010, chs 2 and 4). In our view, this is how it morally should be: no one should be made worse off relative to others by being denied access to housing because they belong to one (or more) of these groups.Footnote 18 If they are made worse off in that way, they are entitled to complain that they are the victims of morally wrongful discrimination.

Can PCR make a similar complaint? We suggest they can. Our main arguments for this claim are that some of the allegedly morally relevant differences between those in the socially salient groups mentioned above and PCR are not morally relevant; and that, of the differences that plausibly are morally relevant, none is sufficient to show that housing discrimination against PCR is permissible. If we are correct, the conclusion to be drawn is this: if it is morally wrong to discriminate against people by making housing decisions based on familial status, sex, disability, etc., it is likewise wrong to discriminate against PCR in that way.Footnote 19

What reasons can be provided for the view that, in making housing decisions, it is morally permissible to discriminate against PCR, but not against those in socially salient groups of the kind we have listed?Footnote 20

First, it could be maintained that there is a relevant difference between PCR and those in the socially salient groups which makes discrimination permissible: the former (unlike the latter) pose a greater risk of harm to other tenants because there is a risk that they will reoffend, and discrimination, it could be argued, is permissible against a group if members of that group present a threat of harm to others. This seems, in essence, to be the line taken in the Fair Housing Act, where it is pointed out that the act, while protecting some groups against discrimination, “affords no protections to individuals … who present a direct threat to the persons or property of others” (Department of Justice, 2022).

Now, there is no denying that PCR, as a group, pose a greater risk of harm to others than those in socially salient groups. One may think that this provides sufficient reason to justify housing discrimination against PCR. However, there is at least one challenge to this view. This takes as its point of departure one of the arguments we presented above, in Sect. 3.1, in response to the contention that excluding PCR from housing is permissible because it prevents crime. There we argued that current exclusionary practices are likely to have a counterproductive impact on crime prevention. If this argument is sound, it also has implications for the present discussion. Specifically, even if we believe that we are justified in discriminating against a group if its members pose a significant threat of harm due to the risk of their recidivism, that justification would collapse where it is the case that the discrimination results in an even greater threat of harm due to what is now an even higher risk of recidivism. That is, insofar as what motivates the view that the discrimination of PCR in relation to housing is permissible is the avoidance of harm, the motivation is undermined, or at least challenged, by the observation that the risk of harm to others is likely to increase following such discrimination. At the very least, it needs to be explained why it should be considered morally more important to protect other tenants from the risk of harm by discriminating against PCR, than it is to protect these same tenants (and indeed other citizens) against a higher risk of similar harms by declining to apply discriminatory housing policies.

The second reason why it may be claimed that discrimination against PCR is morally permissible while discrimination against the socially salient groups we mentioned is not, refers, once again, to the increased levels of fear of crime among tenants. That is, it may be argued that in contrast with what happens when the socially salient groups are housed, housing PCR will result in increased levels of fear of crime among tenants, and that this justifies discrimination against them.

In our view, this suggestion can quickly be dismissed. As we argued in previous section, in any given case there may be ways to reduce the anticipated rise in levels of the fear of crime that housing a PCR produces to the extent that no net increase in fear of crime is observed among the residents. Considering this, it is surely morally preferable to pursue these measures rather than engage in harmful discrimination against the PCR. Further, and as was also discussed in the previous section, we should be reluctant to allow fear of something be decisive when we are making a decision. Imagine that people fear those with disabilities, or those with a religion differing from their own. In such cases, we would not regard it as fair if people with disabilities or a different religion lack access to housing in areas where their fellow residents have such fears.

Turning now to the third reason that it may be claimed that discriminating against PCR is permissible even though doing so against other socially salient groups is not. It might be maintained that PCR are less stable, as tenants, than the other groups in the sense that it is more likely that they will fail to pay rent and/or be short-term residents. To motivate this view, it can be pointed out that PCR often struggle to secure employment, often have poor credit history and may need to vacate their housing after a short period because they have reoffended. It is surely true that many PCR struggle with at least one of these issues, plausibly to a greater extent than those in the socially salient groups we have referred to. However, even if PCR are more unstable tenants than members of other groups are, this is arguably not sufficient to show that housing discrimination against them is morally permissible. To see why, recall that being denied housing has significant detrimental effects on the well-being of the PCR. It also has a negative effect on rates of re-offending, which is harmful to both the PCR and their future victims. Consequently, if one were to argue that the fact that PCR are unstable renters makes discrimination against this group permissible, an implicit assumption is that securing a steady source of income for a landlord, or housing provider, is morally weightier than providing PCR with housing that may well help him turn his life around and keep others safe by reducing, or even preventing, re-offending.

Some may argue that privately owned rental companies are indeed entitled to treat their own interest in a steady income as weightier than the interest the PCR has in his own well-being and the interest that we all have in the benefit of crime reduction—that these latter matters are the responsibility of the state, not private companies. However, surely public housing providers who are often owned, or heavily subsidized, by the state cannot claim that a similar weighting of interest is appropriate for them. Usually, public housing is not intended to provide the state with a steady stream of income. It is meant to benefit the most vulnerable people in society by providing them with a home, with all the benefits that this may bring to society at large. Given this, the fact that PCR are likely to be more unstable renters than the members of socially salient groups cannot plausibly be said to make housing discrimination against them by public housing authorities permissible.

Fourth, and finally, it might be argued that discriminating against PCR is permissible because, as van Tongeren (2022) notes in passing, unlike housing discrimination against socially salient groups, house allocation “discrimination based on a criminal record does not imply discrimination on the ground of a ‘specific character’ but ‘only’ qualifies as discrimination on (past) behaviour” (p. 63, note 8). The suggestion seems to be that discrimination against a group is permissible when it is based on the conduct of members of the group but impermissible when it is based on members’ character traits, and that housing discrimination against PCR is of the first of these kinds.

Should we accept this argument? In our view, no. The past conduct of group members seems to be central to at least some of the groups that most of us believe would be wronged by housing discrimination. Take housing discrimination based on familial status (ruled out by the Fair Housing Act) or pregnancy and maternity (ruled out by the Equality Act). In both groups, membership is most often a result of past behavior by the members themselves. Something similar seems to be the case, albeit to a lesser degree, with religion and some kinds of disability—e.g. those resulting from motorcycle accidents where the rider was not wearing a helmet, or from free climbing. Thus, if it were morally permissible to discriminate on the basis of past conduct when it comes to housing, it would follow that it is permissible to discriminate, not only against PCR in this regard, but also against people who are pregnant, or have children, as well as at least some religious and disabled people. Since the latter surely cannot be correct, we should reject the idea that housing discrimination is permissible insofar as it is based on the past behavior of members of a group.

To conclude this section, it seems clear that housing discrimination against the socially salient groups protected by anti-discrimination legislation such as the Fair Housing Act and the Equality Act is usually morally wrong. It is sometimes argued that PCR differ from those in these socially salient groups in relevant respects, and that this makes housing discrimination against the former permissible. In this section we have argued that prominent ways to attempt ground such relevant differences are unpersuasive, and that the relevant differences that can plausibly be called upon fail to introduce a sufficient reason to think that housing discrimination against PCR is morally permissible. If we are correct, housing discrimination against PCR should be treated as morally on a par with housing discrimination against those who are members of the socially salient groups protected by anti-discrimination legislation. That is, housing discrimination against PCR should likewise be considered morally wrong.

5 Conclusion

In this article, we have critically discussed and rejected two prominent arguments in favor of excluding PCR from housing. Specifically, we have contended that arguments for this practice turning on its crime preventive effect, and its role in reducing or preventing fear of crime, are unpersuasive. We then explained why, in our view, excluding PCR from housing amounts to wrongful discrimination against them. Our analysis suggests that PCR ought to be allowed access to housing to the same extent as other people. This conclusion is, of course, tentative and remains open to challenge. Perhaps there are arguments in favor of the exclusionary practice that we have failed to consider in the article. Perhaps we have overlooked ethically relevant differences between PCR and members of socially salient groups which make housing discrimination against the former permissible. However, it seems to us that unless such considerations are offered the current laws and practices should be abandoned as quickly as possible both to protect the well-being of PCR and to protect the general citizenry from becoming the victims of crimes.