Pushing the boundaries of the quarantine model: Philosophical concerns and policy implications

1. Introduction: the quarantine model and its merits

Retributivism is a philosophical theory concerned with the justification of punishment.¹ It is grounded on three fundamental principles.² Firstly, those who engage in morally wrongful acts deserve proportionate punishment. Secondly, proportionate punishment is not only required but also morally commendable. Thirdly, a deliberate punishment of an innocent person or the excessive punishment of an offender are morally impermissible. The legal framework provides the basis for the justification of individual punishments, while broader justifications for the practice of punishment can be found in ethical theories, shared values, or societal goals.

Retributivism is rooted in the concepts of positive and negative desert. The former claims that offenders are morally deserving of punishment for their wrongful actions, while the latter maintains that it is not acceptable to punish individuals who have not committed wrongful acts. In essence, retributivism is a deontological theory that is primarily backward-looking, concerned with assigning moral responsibility and blame for past conduct.

Recently, retributivism has been challenged on philosophical grounds and – as a consequence – has become less socially and politically appealing. In particular, skepticism has emerged concerning the possibility of holding a person fully responsible for the actions she performs. Two types of arguments seek to undermine the existence of free will, and hence of moral accountability for wrongful actions. The first set of arguments is based on neuroscientific experiments, which aim to demonstrate that conscious will is not the real cause of our actions.³ The second is purely philosophical and deals with such metaphysical issues as determinism/indeterminism, manipulation, causa sui, and luck.⁴ Pereboom and Caruso have questioned whether it is ever rational to hold someone morally responsible in the basic desert sense.⁵

In doing so, they have called for a significant reform of the penal and judicial system and have explicitly endorsed a model of incapacitation for dealing with dangerous criminals that challenges retributivism.⁶ This model,⁷ known as the quarantine model, is based on free will skepticism, treats offenders as not responsible for their crimes, and envisages that they should be detained in conditions that are not extremely harsh, while being offered the opportunity for rehabilitation. The main tenet underlying the quarantine model is that the confinement of those who have proven to be dangerous for society is more humane and preferable than alternatives currently used for dealing with criminal behavior.⁸

To justify the preventive detention of dangerous subjects, advocates of the quarantine model typically draw an analogy with the process of quarantining infectious people. Specifically, at least three major similarities are observed between the process of quarantining sick people and the idea of preventive detention of criminals. The first similarity is that both groups of subjects pose a serious threat to those around them and to society in general. As a carrier of the Ebola virus constantly threatens people while they wander around, so does a potentially dangerous criminal, who is continuously on the verge of committing a crime while they are free. Thus, both violent criminals and disease carriers threaten public safety. The second similarity is that, in each case, both the contagious patient and the detained criminal cannot avoid harming others without the adoption of some restrictive measures. That is, the damage caused by carriers of infections and violent criminals, so the argument goes, is unavoidable without the use of some forms of isolation. Lastly, and related to the second similarity discussed above, neither a patient nor a detainee can be held responsible in a basic desert sense for the harm they may cause to others.

This similarity can be substantiated only if we operate under the premise that human agents do not possess free will and, therefore, are not free agents. We generally agree – say the advocates of the model – that it is sometimes permissible to quarantine carriers of infections, and the justification for that is the right to self-protection and the prevention of harm to others.⁹ “For similar reasons, even if a dangerous criminal is not morally responsible for their crimes in the basic desert sense […] it could be as legitimate to preventatively detain them as to quarantine the non-responsible carrier of serious communicable disease.”¹⁰

In Pereboom’s and Caruso’s account, the offender is not the appropriate source of their actions in the sense required for basic desert moral responsibility. In other words, under the quarantine model, offenders are similar to disease carriers, who cannot avoid infecting others with their illness. As individuals cannot choose to become contagious carriers, so subjects are driven towards criminal behaviors not by their own free will, but rather by genetic dispositions, specific environments, and similar factors, which are all outside of their control.¹¹

Advocates of quarantine models believe that, to protect society from the harms of crime, we should not punish offenders; rather, since they do not deserve to suffer for the harms they have caused, we should place criminal offenders in isolation as we would do with those quarantined with infectious diseases. Importantly, the model includes the principle of least infringement, which holds that the least restrictive measures should be taken to protect public health and safety. This ensures that criminal sanctions will be proportionate to the danger posed by an individual, and any sanctions that exceed this upper bound will have to be justified.¹² Finally, prevention (in terms of the alteration of relevant social conditions of crime), rehabilitation, and reintegration should replace punishment as the focus of the criminal justice system, although, if a criminal cannot be rehabilitated, the model requires them to be confined indefinitely.

This proposed approach, however, has raised some criticism. For instance, what criteria should be used to determine who is dangerous and should be subjected to preventive detention? What type of facilities should be used for such a detention, and how long should this detention last?¹³ In addition, since we lack a reliable way of determining who is dangerous, the quarantine models may even encourage certain crimes.¹⁴

In this article, we remain agnostic about the existence of free will and therefore do not include any considerations related to free will skepticism, which is the basis of the quarantine model. We will not even seek to refute the overall validity of the quarantine model. Rather, by taking some of the most important ethical reasons underlying such a model seriously, we would like to highlight some gaps or inconsistencies within it, with the intent to contribute to a constructive debate ultimately oriented towards the defense of individual safety, human rights, and civil liberties.

The article is organized as follows. Section 2 raises two main lines of criticism against the quarantine model, which are concerned with: (2.1) the lack of a reliable way of determining who is dangerous, and (2.2) the possibility that the quarantine model may encourage certain crimes.¹⁵ These two lines of criticism are, in our view, the most relevant ones that the theory should address and are therefore discussed most extensively. Section 3 proposes three more objections to the quarantine model concerning: (3.1) the rights that are potentially suppressed in the quarantine model; (3.2) the role of “genetic justice”; and (3.3) the difficulty to accommodate reasons-responsiveness. These three objections are relatively minor compared to those discussed in section 2; however, we believe they cannot be simply ignored or considered entirely irrelevant. Therefore, they are grouped in a single section and discussed less extensively. Finally, drawing on the lessons learned in the previous two sections, part 4 invites readers to reflect on how the quarantine model can be improved if it is to become a more credible framework for the development of viable policies in criminal justice.

2. Two objections against the quarantine model

2.1. The quarantine model lacks a reliable way of determining who is dangerous

In this subsection, we argue that the incapacitation model fails to properly address the issue of recidivism, because it lacks a reliable way of determining who is dangerous even among people who have already committed a crime.

Pereboom and Caruso believe that people should not be held morally accountable for the actions they perform because they are not free.¹⁶ As a result, they also argue that wrongdoers should not be punished for the crimes they commit; rather, they propose incapacitating them for the minimum time required to rehabilitate them. Incapacitation is justified by appeal to the right of self-defense, and this right in turn justifies the usage of only as much force as is necessary for self-defense, but no more. Rehabilitation is therefore viewed as the best way to ensure that criminals are no longer dangerous. This approach to criminal justice is believed to be preferable to retributivism because it: i. minimizes potential future crimes (reduces serious societal harms by prolonging the quarantine of dangerous offenders); ii. is more viable economically; and iii. offers to wrongdoers a possibility for rehabilitation (is more humane and avoids unnecessary suffering).

To accomplish these declared goals, however, the model needs to be able to assess, with high accuracy, the risk of recidivism. Without a reliable system of assessment, criminals could spend too much or too little time in prison. Deviation in the accuracy of predictions in either direction will make the model ineffective and rather costly for society. If the model’s predictions were harsher than needed, public funds would have to be spent to keep offenders incarcerated and to provide them with a form of rehabilitation they would not really need. Moreover, the offender would be precluded from contributing to the economy and to the lives of friends and family. On the other hand, if the model’s predictions were too lenient, then dangerous individuals wouldn’t spend enough time in confinement and would not be sufficiently reformed (rehabilitated). This in turn would put citizens in danger and society would have to deal with the possible consequences of reoffending (such as supporting victims and initiating a new cycle of incarceration and rehabilitation). Without an accurate assessment of the risk of recidivism, the quarantine model would fail to minimize future crimes, reduce serious societal harms, justify its economic viability, or offer adequate rehabilitation for wrongdoers.

The timely incapacitation of wrongdoers therefore plays a pivotal role in the quarantine model and helps establish the analogy with the epidemiological process of quarantining people. We note, however, that these two processes are not analogous which renders the proposed analogy far-fetched. This is because the level of accuracy underlying decisions within these two fields (criminal justice and epidemiology) is considerably different.¹⁷

More specifically, medical knowledge and the diagnostics tools nowadays available to doctors allow them to make well-informed and extremely effective decisions in epidemiological cases. People who have been in contact with infected individuals must be quarantined but only for the time needed by the virus to develop (this coincides strictly with the so-called incubation period). After the end of the incubation period, the person can be released and we can be confident that she would not spread any disease, hence pose no threat to society. Doctors also know that someone’s quarantine can end even faster if a negative test, or a series of negative tests for the specific virus are produced.

People who develop clinical symptoms or test positive for a disease are immediately isolated and they are kept in isolation until they are no longer contagious (usually when they no longer display clinical symptoms or until a test or a series of tests demonstrate their full recovery). All in all, we can say that there are very effective and reliable mechanisms through which biomedical experts can know, with precision, if any given patient is virus-free; that is, they can know with an extremely high level of probability (close to 100 percent), if she can be released because she does not pose any threat to society.

However, the level of certainty accompanying court decisions to release an offender from incapacitation cannot be said to be in any way close to the one that can be achieved by biomedical experts in epidemiological cases. Sure, there may be some cases when the release should be prevented because of how probably it would lead to a crime, for example in cases involving psychopaths with brain damage.¹⁸ However, for most cases involving wrongdoers, their future dangerousness is not predictable, at least not predictable with the same level of accuracy that can be observed for epidemiological cases. In other words, no test or procedure or textbook in criminology can ever tell us for how long a given person should be incapacitated. Proof of this is the fact that members of parole boards and judges struggle with this problem constantly in their daily practices.¹⁹

Pereboom and Caruso may perhaps object that neural tests for violent tendencies will be developed soon and that such tests will allow us to determine, with high accuracy, whether someone will commit a crime.²⁰ However, we do not have such tests at the moment, we have no guarantee that they will ever be developed, and – most importantly – even assuming that such tests will become available, we do not know yet whether they will be effective and fulfil the promise for which they may be introduced.²¹ Yet, advocates of the quarantine model could rejoin by suggesting the introduction of risk-assessment tools to estimate the probability that a wrongdoer would reoffend. Risk-assessment methods are already widely used around the world.²² Such methods could be adapted to criminal justice and could employ computer algorithms, big data analysis, and advanced statistical calculations to assign a probability of re-offending to a wrongdoer based on the risk factors associated with her profile.

But we should note that there is a very heated debate over the possible adoption of risk-assessment procedures in criminal justice.²³ Criminologists are divided about their usage and many of them see insurmountable ethical difficulties in their application. We also believe that the probabilistic nature of risk assessment methods will inevitably lead to the incarceration of individuals who would not necessarily reoffend. So, the real risk associated with their adoption in society could be that we would end up restricting (at least some) people’s basic liberties for the sake of public safety.

Pereboom could object that doing this, based on statistical generalizations or some sort of scientific algorithm, can be justified by appeal to self-defense. The objection is therefore that doing so may bring overall better consequences than any alternative. However, this amounts to saying that we should justify detention, hence make important decisions in criminal justice, on utilitarian grounds – an approach that Pereboom and Caruso strongly reject elsewhere.²⁴

Thus, if Pereboom and Caruso were to adopt such an approach, they could justify the risk of quarantining individuals who will not re-offend, but this justification would be consequentialist in character, and they have already explicitly denied consequentialism a role in their model.

2.2. The quarantine model could encourage specific crimes

In this section, building on Smilansky’s criticism, we argue that the quarantine model is not effective in deterring crimes performed by one-time offenders. Smilansky famously argued that the quarantine model cannot provide enough deterrence for one-time offenders, and as a result, it does not offer any resource to incapacitate recidivistic wrongdoers.²⁵ Pereboom remarked that Smilansky’s criticism is only applicable to exceptional cases and would not threaten the spirit of the quarantine model.²⁶ In addition, he conceded that in order to deal with this potential issue his model could allow financial penalties and limited incarceration for the sake of general deterrence.²⁷

Contra Pereboom, we argue that Smilansky’s cases are not as rare as he thinks, and that adding financial penalties and limited incarceration won’t do much to establish deterrence. In addition, we claim that such a move may add a retributivist element into the quarantine model, which would defeat its original humanitarian purpose, while also contributing to making it even more different from the relevant epidemiological process on which it is based. Consider the case of violent revenge as a paradigmatic illustration of our criticism:

In cases of violent revenge, the offender is not necessarily dangerous for society after he has committed his crime. In fact, it may be easily demonstrated that the perpetrator of a violent revenge is no longer dangerous after his desire for revenge has been quenched. In 2004 Vitaly Kaloev (a former Russian architect) killed Peter Nielsen, an air traffic controller, as an act of revenge. Kaloev held Nielsen responsible for the death of his family, which was aboard the Bashkirian Airlines Flight 2937 that collided with DHL Flight 611 over Überlingen in Germany on the 1^st of July 2002. Kaloev was arrested and then convicted of the premeditated killing of Peter Nielsen. He was sentenced to eight years in prison. However, after two years in prison, Kaloev was released. Before this incident, Kaloev had no criminal history and was a reputable architect. After his release, he never committed another crime.²⁸

The analysis of this case study shows that Kaloev had the intention to kill only one person (Peter Nielsen), who – in his opinion – was responsible for the death of his family and of the 60+ passengers who were on board of the airplane that crashed over the skies of Germany in 2002. After his desire for revenge was quenched, Kaloev wished no harm to anyone else and never committed a crime again. Kaloev’s case is a perfect real-life example of how a law-abiding citizen may commit a vicious crime without necessarily becoming a threat for society. Many cases like Kaloev’s case do exist: just think about Gjakmarrija – the Albanian blood feud – that obliges people to commit murder to salvage the honor questioned by an earlier killing.²⁹ Similar practices based on revenge existed in certain parts of Italy, especially in Sardinia, and still exist in areas of India (like Rajasthan), Kosovo, Montenegro, Brazil, Ecuador, Egypt, Pakistan, Ethiopia, Turkey, and Uganda.

We believe that the incapacitation model does not provide an adequate answer for such real-life scenarios. Avengers usually do not pose any significant danger to the public or to society; thus, society does not need to be protected from them, so, according to Pereboom and Caruso, such people should be set free. This is a case where there would be a complete lack of punishment for the perpetrators of vicious crimes if we were to follow the quarantine model. Such offenders could simply commit those crimes (and possibly others) without having to face the consequences usually associated with such actions. Pereboom acknowledges that, for some offenders, the quarantine model does not provide sufficient deterrence.³⁰ He asserts that we should implement small fines and minimum prison sentences to increase the level of general deterrence of the quarantine model. So potential offenders will know that they may face unpleasant consequences for their actions even if their risk of recidivism is negligible. However, such measures would not be an effective deterrent for those who really seek revenge (as in the real-life scenario we described above).

In addition, if Pereboom is ready to implement hybrid measures (such as fines or short-term prison sentences) in his model, it seems to us that he is diverting from its original humanitarian purpose. To save the quarantine model, Pereboom (and perhaps Caruso) may try to draw an analogy with the punishment that is inflicted on those who break quarantine. However, those people are punished not because they have become dangerous, but rather because they intentionally violate an epidemiological protocol.

Avengers should not be preventively detained as they pose no realistic threat to society. It follows that we cannot punish them for breaking a rule – “quarantine” – they should not be following in the first place. Again, if the advocates of the model allow for the usage of sanctions for the sake of public safety or to achieve deterrence, the resulting model would depart from the original humanitarian spirit of quarantine. The model will start looking like a hybrid model, because it will be partly based on the self-defense/deterrence justification of quarantining dangerous offenders, and partly on some other sort of justification (presumably retributivist or consequentialist) for punishing those offenders who are not dangerous but who must be punished for the model to work. However, these sorts of justifications are precisely what advocates of quarantine wanted to reform in the first place.³¹

3. More objections to the quarantine model

Having presented two main new objections to the quarantine model and analyzed a series of possible replies by its proponents, we next sketch out a series of three additional objections that, taken individually or together, may point to further weaknesses of the proposal. These objections, as noted above, are concerned with (3.1) the rights that are potentially suppressed in the quarantine model; (3.2) the role of “genetic justice”; and (3.3) the difficulty to accommodate reasons-responsiveness. Once again, one should view these objections, necessarily expressed in a brief form for reasons of space, as positive and constructive contributions to the debate. They are not necessarily knock-down arguments that should lead to a total restructuring of the model; however, we believe they highlight controversial or philosophically evaluative aspects that indicate how the proposal of incapacitation is still vulnerable to criticisms of some relevance. These objections should therefore invite us to engage in deeper reflections on the ethical, social, and political implications of the quarantine model, which will require us to balance the competing values of public health, individual autonomy, and social justice. Only by engaging in such a reflective process can we hope to arrive at a more nuanced and justifiable understanding of the theory and of its limitations.

3.1. Some rights are denied in the quarantine model

In this subsection we argue that quarantine is at least in principle as harmful as our current penal punishment. One could object to the idea of preemptive incapacitation, which is at the core of the quarantine model, by appealing to an argument that reconceptualizes punishment in terms of the rights taken away from the subject.³² In Pereboom and Caruso’s model, quarantine is not understood as a form of punishment, but it still has the effect of depriving the person placed in isolation of some of their basic rights.³³ Specifically, those placed under quarantine lose some fundamental parts of their freedom, namely the right to movement and the right to association. In fact, those who are placed in quarantine cannot move beyond the perimeter of the place where they are confined, nor can they physically meet non-offenders, since the purpose of isolation is precisely to maximize public health and safety (it is not however excluded that the person placed in quarantine may have protected meetings with family members, as happens in prisons).

Freedom of movement and association, however, are two of the most fundamental rights that individuals can enjoy. Now, confinement deprives the quarantined person of these fundamental rights. Prison also deprives those who are imprisoned of the right to movement and the right to association. In the quarantine model the deprivation of rights is an undesired consequence of quarantine, which takes place with the purpose of protecting society (and this consequence can be problematic). So, the point is this: if the aim is to protect society and not to afflict the confined person who does not deserve to be deprived of two fundamental rights, then quarantine – in practice – does not do much better than retributivism.

The fact that the prison system is more afflictive and less humane than it should be does not change the moral assessment. There are countries, other than the United States, where prisons are humane and quite effective in re-educating wrongdoers.

One could therefore seek less afflictive methods than quarantine. Indeed, Pereboom and Caruso are in favor of forms of medical and technological intervention or surveillance. However, the so-called neuro-interventions, which, for instance, could limit aggressiveness or sexual desire, threaten to violate other rights. Firstly, the right to authenticity is threatened. This is because drugs can change an individual’s personality. Secondly, chemical manipulation of the brain prevents the person from getting access to her self-knowledge (the right of mental integrity), thereby potentially triggering mental problems. Thirdly, there is an element of coercion in the context where drug therapy is offered as an alternative to prolonging confinement terms.³⁴

So, the quarantine model is confronted with the same difficulties as the “mixed” retributivist system currently adopted in most countries. Only by making the quarantine a “little paradise” would Pereboom and Caruso’s model diverge clearly at the humanitarian level from the system they want to criticize. However, if this is the case, they should address the funishment criticism raised by Smilansky.³⁵ In brief, if confinement facilities should be delightful to compensate for the hardship suffered, then they would become so desirable that they would not constitute a crime deterrent. In other words, offenders would have less fear of ending up in such facilities (if caught) and hence might gamble on the idea that even if they were caught, things wouldn’t be too bad. This entails, according to Smilansky, that hard determinism about punishment would be self-defeating, because it would paradoxically cause more crimes and more confined people. Caruso replied to this objection and argued that it does not have enough grip.³⁶ However, his reply can also be questioned. For example, he says that statistics confirm that countries where prison conditions are less harsh (e.g., Norway) have much lower crime and recidivism rates than the United States, where prison conditions are by common consent very harsh. However, this does not seem to entirely negate the criticism in principle, as there may be other causes that temporarily affect crime rates.

3.2. Can we ignore genetic justice?

Another objection to the incapacitation model may point out that quarantined people are those who are more in need of help due to their genetic profile and are instead deprived of some of their basic rights. In fact, having a genetic make-up that turns out to be socially unfavorable (even without endorsing genetic determinism) can be considered a special case of bad luck. It is a special case of bad luck because we are not currently able to change an individual’s genetic make-up, whereas we can intervene in varying degrees on most cases of bad luck.

As we have seen, Pereboom and Caruso state that “the way we are is ultimately the result of factors beyond our control.”³⁷ Now, these factors can be either external to the individual, internal to them, or a combination of both. In the first case, Pereboom and Caruso would completely link free will skepticism to the environment in which the individual happens to be: in other words, they would posit environmental clues we are unaware of as the sole cause of our lack of free will.

However, they do not seem to go so far as to make such a claim, even though they do cite situationism as one of the elements that demonstrate free will skepticism.³⁸ But then, if the way we are is determined, or at least also depends on internal factors linked to our physical constitution (e.g., our genetic inheritance), or if we consider a combination of internal factors and external factors, then we cannot ignore a recent line of criticism that has come to be known as “genetic justice.”³⁹ This is the idea that we have a specific genetic make-up that strongly affects our health and our behavior; some individuals are luckier than others with regards to their genetic make-up and the consequent outcomes; so, the state has the duty to intervene to reduce the inequalities that arise from the genetic make-up for which nobody is at fault.

The basic idea of genetic justice is in line with free will skepticism, even if the authors who support it have not made explicit statements on the point. Each of us has a specific genetic profile that can also induce us to be more or less prone to compliance with moral rules and/or the law. If our genetic profile is somehow the result of a lottery, over which we have no control, then even the effects of our physical endowment will lead to deviations from an ideal of equality that, according to some theories of justice, should be rectified with public intervention, at least in order to reduce major inequalities.⁴⁰

So, if we have poor health prospects and a low IQ, we should rightly be considered disadvantaged, on par with low-income classes or oppressed minorities. Pereboom and Caruso point out that their model is not based on the idea that dangerous people are all ill in the sense of carrying a deficit or a disorder that can be treated. In any case, if the genetic profile that predisposes us or inclines us to a certain antisocial behavior is the result of a natural lottery that is unfavorable to us, it seems to be a double injustice to subject those who have that specific genetic profile to incapacitation and preventive quarantine. Firstly, they suffer from inequalities from birth and would need help in order not to harm other people. Secondly, based on their genetic profile, which strongly affects their behavior, they would suffer confinement and incapacitation.⁴¹

Thus, quarantine seems not to be a very humane method since it would promote (if only unconsciously) the discrimination of certain individuals. The point is that the principle of least infringement suggests that when dealing with individuals who have committed morally deplorable actions, society should impose the least intrusive or restrictive measures necessary to achieve its goals (such as maintaining social order or preventing harm). Incapacitating dangerous individuals is consistent with the principle (once all the preventive efforts have been made) but it may well result in the double injustice we described above.

It could in fact be argued that according to the idea of genetic justice the inequalities suffered by those who are induced to crime by their physical endowment should be compensated for. In other words, additional harm should not be inflicted on them to protect those who are already more favored in that specific aspect of genetic endowment. Free will deniers could reply as follows: we are sorry for the disadvantaged offenders, we would like not to resort to incapacitation, but we must defend ourselves from unmotivated attacks/crimes. Nobody would disagree with this, and this is precisely what any criminal system does. The quarantine model should not stigmatize individuals affected by genetic deprivation as people to be isolated, or at least it should not treat them in the same way as those who are not genetically deprived and can behave differently (have better control over their behaviors).

If the quarantine model, while not punishing anyone, fails to distinguish between offenders who are affected by genetic injustice and those who are not, it seems to have a problem of justice to resolve. In other words, the quarantine model should not impose the same restrictions on individuals who are genetically deprived and have no control over their behavior as it does on individuals who are not genetically deprived and can make choices. Obviously, genetic luck is a scalar notion, while satisfaction of control condition is a threshold one (at least in the legal domain). So, here we are concerned with those genetically unlucky individuals that fall below the control threshold due to their genetic make-up on which neither they nor anyone else can intervene.

This raises important ethical questions about the fairness and equity of the quarantine model. Ultimately, the aim should be to ensure that the quarantine model promotes both public health and social justice, while respecting the dignity of all individuals, regardless of their genetic make-up.

3.3. The problem the quarantine model has with reasons-responsiveness

This last issue we raise is connected to the framework of public health ethics described by Pereboom and Caruso themselves. Their attempt to retain a grip over reasons-responsiveness and the control condition within a skeptical view of free will seems to be unsuccessful and, therefore, the quarantine model turns out to be a less credible proposal for crime prevention.

Public health ethics not only justifies quarantining carriers of infectious diseases on the grounds that it is necessary to protect public health, it also requires that we take active steps to prevent such outbreaks from occurring in the first place. Quarantine is only needed when the public health system fails in its primary function. […] Taking a public health approach to criminal behavior would allow us to justify the incapacitation of dangerous criminals when needed, but it would also make prevention a primary function of the criminal justice system. If we care about public health and safety, the focus should always be on preventing crime from occurring in the first place by addressing the systematic causes of crime. Prevention is always preferable to incapacitation.⁴²

What kind of prevention can be carried out if individuals behave in the way assumed by free will skepticism? Pereboom distinguishes between different senses of responsibility. For example, Pereboom says, an agent can be held responsible (not in the basic desert sense) if it is legitimate to expect them to be able to answer a question such as: “why did you decide to do this? Do you think it was the right thing to do?” and to critically assess their own actions and what they reveal about their moral character.⁴³

It is therefore necessary to assess how the authors who defend the incapacitation model propose to prevent crimes. Caruso offers some examples that refer to strategies to overcome cognitive distortions that can generate and sustain criminal behavior. He also mentions creating favorable environments to foster understanding of the reasons for acting morally and according to the law.⁴⁴ As Pereboom and Caruso put it: “What unites policies for treatment of criminals on our view is not that they assume that they are psychologically ill and therefore in need of psychiatric treatment. Instead, they all aim to bring about moral change in an offender by non-punitively addressing conditions that underlie criminal behavior.”⁴⁵

However, when considering individuals who commit crimes or are at risk of doing so and who do not possess any underlying illness, the framework suggests a preference for “methods of therapy that engage reasons-responsive abilities.”⁴⁶ Unlike the illness model, which does not directly address the criminal’s capacity to respond to rational arguments but instead bypasses such cognitive faculties, this approach aligns closely with viewpoints asserting that wrongdoers possess the capacity to comprehend punishment and modify their conduct based on feedback from the criminal justice system. This alignment is not primarily due to the nature of punitive measures (such as imprisonment as a punitive tool) but rather stems from the fact that these individuals possess a sensitivity to rational considerations, and have the ability to deliberate and choose their course of action based on these considerations. Consequently, they can be held accountable for their actions based on their capacity to make reasoned decisions. It is no coincidence that Pereboom and Caruso feel the need to clarify this point:

While many compatibilists see some type of attunement to reasons as the key condition for basic desert moral responsibility, we instead view it as the most significant condition for a notion of responsibility that focuses on future protection, future reconciliation, and future moral formation.⁴⁷

It should therefore be acknowledged that the concern of the incapacitation approach is entirely directed towards the future and that in this sense it has a genuinely humanitarian inspiration, but this aspect seems to go against one of the tenets of the quarantine model; namely, the likening of individuals carrying communicable diseases to offenders, who are both incapable of avoiding doing harm to others.

It is significant in this sense that determinist (neuroscientific) models place the cause of a great deal of recidivism in the morphology and functioning of the brain (on which the subject has no influence) rather than in cognitive styles. Consider the experiment carried out by Aharoni and colleagues, where researchers examined 96 detainees on their release with tests related to rapid decisions and inhibition of impulse responses.⁴⁸ During the test, an fMRI measured the activation of the anterior cingulate cortex (ACC), the area of motor control and executive functions. After a four-year observation, individuals who had experienced the lowest ACC activation during rapid decision-making tasks were those who had the highest rate of recidivism, even discounting confounding factors. Those who were in the lower half of activation values of the cingulate cortex totaled 260% more arrests for all types of crimes compared to the others.

Pereboom and Caruso could claim that a human reason-responsiveness system could exist in a determinist world. However, in that scenario a quite different reason would be at stake. This follows from the fact that Caruso adopts “deliberation compatibilism,” according to which “S’s deliberating and being rational is compatible with S’s believing that their actions are causally determined by antecedent conditions beyond their control.”⁴⁹ However, other authors are more inclined to embrace “deliberation incompatibilism,” according to which there is incompatibility between deliberating rationally and free will skepticism. It is true that there is epistemic uncertainty in choosing between different courses of action and one can decide rationally or irrationally, but, for the free will skeptic, the choices are always metaphysically speaking beyond the agent’s control.

In other words, Pereboom and Caruso seem to hold on to the ability of most people to be reasons-responsive at the descriptive level, but that risks associating their view with a form of normative compatibilism. Again, it is difficult to see why their model should be preferred to a classical one, which in principle is committed to preventing people from wrongdoing and rehabilitating them through punishment, so that they can understand the wrongness of their acts, unless they are judged to be mentally insane and not capable to control their behavior.

4. Conclusion

In this article, we considered Pereboom and Caruso’s quarantine model as a radical alternative to theories of justification for criminal punishment, particularly retributivism. While the quarantine model is gaining attention and support, it appears to be afflicted by several problems that still require close philosophical scrutiny. Some of these have been identified by other scholars, some are highlighted by ourselves in this short contribution.

Firstly, the quarantine model lacks a reliable mechanism for determining who is truly dangerous. This raises the risk of unnecessarily harming innocent individuals by subjecting them to isolation. Secondly, the quarantine model may inadvertently encourage certain crimes, particularly revenge and single-motivated homicides, which are considered serious offenses that society should seek to prevent.

Apart from these weaknesses, there are other aspects of the quarantine model that require further scrutiny. Specifically, we identified objections that pertain to: (a) the rights that are potentially curtailed by the quarantine model, (b) the so-called “genetic justice,” and (c) the challenge of accommodating reasons-responsiveness. While we understand that such objections do not represent knock-down arguments, we believe that they point out important issues that need to be addressed, at least if the quarantine model is to be considered as a credible framework for the development of viable policies in criminal justice.

In this context, we believe that what matters is contributing to a constructive debate that highlights both the potential and the limitations of the quarantine model. The goal of such a debate is to advance research on the topic with the intent to develop a theory that is conceptually, morally, and pragmatically sound and feasible, as should be any theory that wishes to impose limitations on individuals’ rights. Naturally, this debate ought to be open-minded, hence not ideologically driven. Ultimately, the aim is to arrive at a sound and ethical theory that can guide policies aimed at safeguarding society while respecting individual safety, human rights, and civil liberties.