Article Text

Download PDFPDF

Prostitution, disability and prohibition
  1. Frej Klem Thomsen
  1. Correspondence to Dr Frej Klem Thomsen; Roskilde University, Building 3.1.3, Universitetsvej 1, Postbox 260, Roskilde 4000, Denmark; fkt{at}


Criminalisation of prostitution, and minority rights for disabled persons, are important contemporary political issues. The article examines their intersection by analysing the conditions and arguments for making a legal exception for disabled persons to a general prohibition against purchasing sexual services. It explores the badness of prostitution, focusing on and discussing the argument that prostitution harms prostitutes, considers forms of regulation and the arguments for and against with emphasis on a liberty-based objection to prohibition, and finally presents and analyses three arguments for a legal exception, based on sexual rights, beneficence, and luck egalitarianism, respectively. It concludes that although the general case for and against criminalisation is complicated there is a good case for a legal exception.

  • Criminal Law
  • Disability
  • Ethics
  • Minorities

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Criminalisation of prostitution has been an issue of heated policy discussion in recent years, because countries that in other respects share political cultures have taken very different stances on the issue, and because it concerns elements almost designed to provoke heated debate: sex, crime, and gender inequality.i

Similarly, the needs and rights of disabled persons has been an issue of intense and ongoing debate, resulting in the implementation of progressive legislation in most European and Anglophone countries, and in the adoption of the United Nations convention of the rights of disabled persons.ii

However, virtually no scholarly attention has been granted to how these two issues intersect, that is, the purchase of sexual services by persons with disabilities, and the justifiability of prohibiting their doing so. Two recent articles in the Journal of Medical Ethics are exceptions to this trend. Jacob M Appel contends that for some persons with disabilities, having sexual relations is either extremely difficult or impossible except by purchasing sexual services, that persons have sexual rights, and that for this group of disabled persons their right justifies an exemption from legal prohibition of prostitution and state support for bearing the costs.3 Meanwhile Ezio Di Nucci counters that sexual needs do not ground a right, and that the sexual needs of the group of disabled persons who face difficulties satisfying their needs are best served by the establishment of volunteer organisations offering them free sexual services.4

Both articles raise important points, but I believe that both understate the complexity of the issue, and that there is, therefore, much more to be said. Providing an answer to the question of whether disabled persons should be exempted from a general prohibition on prostitution is difficult, because it involves assessing at least four issues: first, there is the contested sociological issue of the character and consequences of prostitution; second, there is the complex moral issue of the badness of prostitution; third, there is the criminal justice ethics issue of whether and, if so, how to prohibit prostitution; and finally, there is the minority rights issue of when and why to exempt minorities from broader legislation.

My focus in this article is the case for an exemption from prohibition, but as this depends, in part, on how strong the case for prohibition is, which again depends on the strengths and weaknesses of prohibition compared to the policy alternatives and how bad prostitution is, it is inevitable that we shall touch upon the other issues as well; there is no shortcut around them.

I proceed as follows. In section 2, I sketch an argument that prostitution is bad because it harms prostitutes, and explore and criticise a recent powerful counterargument. In section 3, I turn to the issue of prohibition, sketching the requirements of justifiable legal prohibition of prostitution. Finally, in section 4, I present three versions of an argument for exemption from a general prohibition for disabled persons, one rights-based, one beneficence-based and one luck-egalitarian, respectively. Section 5 summarises and concludes.

How bad is prostitution?

The literature on the ethics of prostitution covers a range of different arguments,iii including arguments that focus on the intrinsic badness of trading sex, objectification, exploitation, gender hierarchy and harm.5–16 I cannot hope to present or discuss all of them here, and so shall sketch only what I take to be the strongest argument: that prostitution harms prostitutes.iv

That prostitutes suffer harm is empirically well documented.v Melissa Farley cites a 1996 study showing that prostitutes suffered from a slew of work-related health problems, including: ‘exhaustion, frequent viral illness, STDs, vaginal infections, back aches, sleeplessness, depression, headaches, stomachaches, and eating disorders.’20 Her own work concludes that 68% of active prostitutes present symptoms of PTSD.19

We can distinguish several different sources of harm. First, the level of violence prostitutes suffer at the hands of customers is appalling. Melissa Farley cites a range of studies showing that roughly 60–80% of prostitutes experience physical assault and/or rape in connection with their work, and one US survey that shows murder accounting for fully 50% of the deaths among active prostitutes.19 ,20

Second, prostitutes are at elevated risk of contracting sexually transmitted diseases, although perhaps less so than one might suspect. Studies from Scotland, Italy and Spain show 0%, 1.6% and 3.4% rates of HIV infection, respectively, among prostitutes who did not use intravenous drugs.23 ,25 Other STDs appear to be more common. It is noteworthy, though, that there are strong indications that the risks are dramatically reduced with access to information and better control of working conditions.

Third, the psychological and social effects of the stigma associated with prostitution is often highlighted as one of the most detrimental factors, in that it induces shame, strains personal relationships with friends, family and partners, and deprives prostitutes and ex-prostitutes of valuable opportunities through negative stereotyping and discrimination.26

Fourth, prostitutes can experience psychological harms as a result of their selling sexual services. As previously noted, Farley found symptoms of PTSD in roughly two-thirds of interviewed active prostitutes. In Peter de Marneffe's analysis: ‘This stress is created partly by the constant risk of violence and verbal abuse, but also by the fact that this kind of work demands an exhausting kind of emotional pretense. A successful sex worker must typically pretend to enjoy the company of her clients and to be sexually interested in and aroused by them even when they strike her as stupid, offensive, pathetic, or repulsive. As a result, those who do this work come to associate sex with emotional pretense and manipulation, exploitation and abuse, and, partly for this reason, doing this kind of work makes it difficult for sex workers to enjoy sex (with women as well as men) and to form stable, trusting, mutually supportive and respectful intimate relationships, both while they are working and well into the future once they have stopped.’18 And, as de Marneffe notes, the fact that this effect accumulates gradually and imperceptibly, and varies from person to person and with contextual conditions, should not mislead us into thinking it does not Even critics of the harm argument, such as Lars O Ericsson, typically admit that there is a risk of what he calls a ‘disturbed emotional life’.7

One important caveat about this account of harms is that they may, and by all indications do, differ between various forms and practices of selling sex, with, for example, street prostitutes at far greater risk than escort prostitutes. If some practices, such as so-called sexual surrogacy, where a trained therapist is often attached to a clinical centre, are sufficiently benign, then it may ultimately be mistaken to formulate policies directed at the sale of sexual services as a whole. However, while estimating the ways in which kinds and degrees of harm differ is important, it is also empirically difficult, and a task I cannot hope to accomplish here. For the sake of the argument at stake we need say only that at least some forms of prostitution are harmful.vii

A counterargument: intrinsic versus extrinsic harms

In a recent powerful article, Ole Martin Moen27 challenges the harm argument, drawing on a type of distinction that is sometimes labelled extrinsic versus intrinsic harm,5 that is, those effects that are inherent to an action or event, and those that only obtain under particular circumstances. With minor variations, the gist of the argument is that the harms that prostitutes suffer are caused by contingent features of prostitution as it is currently practised and, therefore, not properly attributable to prostitution as such.

Theories of harm are controversial, and I concur with Moen when he suggests ‘it is useful, as far as possible, not to tie one's argument to specific positions in other areas of philosophy.’28 However, broadness of scope must be weighed against clarity, and I believe that a minimal definition of harm will be useful. Let us say that a person is harmed if and only if she suffers a decrease of her total well-being, and that an action causes harm if and only if its occurrence or effects are a prominent and necessary part of an actual and jointly sufficient set of conditions for a person being harmed.viii

Among the sources of harm listed above, stigma and violence are the two most obvious candidates for an argument that they are extrinsic harms. The idea was forcefully put by Ericsson: ‘The paternalist does not ask himself why depressions and neuroses are common among harlots, why they display self-degrading and self-destructive tendencies, why their behavior often is antisocial, and so on. Yet, the answer should be obvious: the principal cause of these psychological and sociological ‘dysfunctions’ is the social anathema attached to their way of life. [...] It is thus the degradation in which the harlot is held, and as a result also often holds herself, that constitutes the greatest danger to her physical and mental health.’7

Similarly, there seems to be a relevant difference between types of work where suffering violence is an integral part of the job, such as professional boxing, and the types of work where it occurs (far too) regularly but does not belong, such as driving a taxi, caring for psychiatric patients, or prostitution.

There are two challenges for the argument, however. The first is that it seems less obvious that increased risk of STDs and the psychological harm that results from selling sex could be labelled extrinsic. Moen attempts to dispatch the second, by pointing out that stigma (or other extrinsic factors) may be responsible for all the psychological harm that prostitutes suffer, and concludes ‘...statistical correlation between prostitution and various psychological problems is not alone sufficient to conclude that prostitution leads to these problems. Since an argument from mere correlation with psychological problems alone fails to establish C, we will need additional arguments to show that prostitution is harmful.’27

However, as Moen recognises, it is ultimately an empirical question what causal role prostitution plays in the psychological harms prostitutes suffer. Settling it may indeed require further study, but it remains perfectly plausible that prostitution itself can cause psychological harm; the presence of another harm-causing factor (ie, stigma) that potentially explains the harm prostitutes suffer does not entail that prostitution does not or cannot cause harm. Particularly since the theory has at least some empirical support. Vanwesenbeeck found that roughly half—but no more than half—the variance in depersonalisation and emotional exhaustion experienced by indoor prostitutes in The Netherlands was explained by external factors including stigma, lack of control and poor working conditions.31

The second challenge is that it is less clear, I think, than Moen assumes, that distinguishing between extrinsic and intrinsic harms makes prostitution harmless. After all, it seems to amount ultimately to no more than that prostitution is only currently harmful because of contingent social conditions, which clearly does not mean that prostitution is not, under these social conditions, harmful. Moen suggests that we apply a comparative test with homosexuality in the early decades of the 20th century. Homosexuality, he argues, was no more harmful then than it is now, although homosexuals suffered harm from other (extrinsic) causes. And we should ask whether prostitution is currently ‘harmful’ in a different sense than homosexuality was then. The problem here, I think, is that it seems perfectly sensible to say that harm may have several causes, and that homosexuality, conservative sexual norms, and discrimination were all causes of the harm homosexuals suffered at that time.ix Consider a different example: given the choice between putting your child into a class at two different schools, one of which is fine, and one of which is socially dysfunctional due to the presence of a large group of bullies, it seems reasonable to say that your choice of the dysfunctional class, and your child's attending this class, causes your child harm. And this remains true even if the harm is (also) caused by the stigma that the bullies heap on their victims, and even if it is perfectly possible, as evinced by the other class, to have a school-class where no students are stigmatised.

The underlying problem may be an implicit moralising of the metaphysics of causing harm. We may be tempted, in cases of harm-doing, to ascribe causation to those factors that we feel involve wrongness or would prefer to change, but this strikes me as a form of reasoning we should resist. Perhaps it is preferable, if possible, to remove stigma and violence in prostitution instead of prohibiting it, but this does not affect what does and does not cause harm. Causality is not a moral concept.

Criminalising prostitution

In the previous section, I have argued that there is at least one plausible argument for the badness of prostitution. In order to fully clarify the badness of prostitution we would need to also assess other plausible arguments, a task the scope of which falls outside this article. However, even a full resolution of whether and, if so, how prostitution is morally bad would not settle the next issue of whether and, if so, how prostitution should be prohibited. It is to this issue that I now turn.

In terminology defined by Peter de Marneffe, we can divide the relevant forms of legislation into prohibition, decriminalisation and regulation.18 The first refers to policies that criminalise prostitution itself, the second to laissez-faire policies that treat prostitution as it does the purchase and sale of services generally,x and the third to policies that permit prostitution but impose particular restrictions, for example, a minimum age, health examinations, prohibiting pimps and brothels, or conversely requiring that prostitution takes place only in a licensed brothel.

Furthermore, we can distinguish in the prohibition between the different parties to the exchange, for example, by letting it remain legal for a prostitute to solicit and accept solicitation of sex for goods, while making it illegal for the customer to either solicit or accept solicitation.xi In this respect, there is relative consensus today in the literature that prohibiting the sale of sex, which imposes sanctions on the prostitute rather than the client, has few benefits.18 ,25 Thus, I shall focus on ‘the Swedish-model’ where it is the purchase of sex that is prohibited, and therefore, the customer rather than the prostitute who risks sanctions.

Supposing that one wanted to justify a policy of prohibition, how might one proceed? Perhaps the most obvious route—although not, of course, the only one—builds on the account of badness we have already explored. Thus, a harm-based argument for prohibition might be formulated as follows:

  1. Prostitution is bad because it causes harm to prostitutes.

  2. We have reason to avoid harm to persons.

  3. Prohibiting prostitution will reduce harm to prostitutes.

    C. Ergo, we have reason to prohibit prostitution.

The first premise of the argument has already been discussed, and I have argued that it is plausible. The second premise is perhaps as uncontroversial a claim as one can make in moral philosophy. Defending it would take us far afield, and is unlikely to persuade those unwilling to grant it. Thus, I shall assume it for present purposes. This leaves two ways of countering the argument. The first is to challenge the third premise; the second is to argue that there are countervailing reasons against prohibition. Let me sketch each of these briefly in turn.

The third premise of the harm-based argument claims that prohibition will reduce harm to prostitutes. The question is, compared to what? This is important since the reason at stake concerns reducing harm to persons, and if a different policy will better achieve this objective, then prohibition will not reduce harm compared to this alternative, and all else being equal, we have reason to implement that policy.

Furthermore, a challenger might say, it is actually the case that when compared with regulation, the choice is one between quantity of harms and severity of harms. That is, prohibition will decrease the market for sexual services, ensuring that fewer prostitutes suffer harm simply by reducing the number of transactions. However, as is widely recognised, prohibition does not abolish prostitution. Partly, at least, this is because prohibition will leave some former prostitutes with the unenviable occupational choice between (1) criminalised prostitution, (2) other dangerous, poorly paid and/or illegal work and (3) unemployment in absolute poverty. A further objection, therefore, contends that for some women who abandon prostitution after prohibition this may leave them equally badly or even worse off,6 ,11 and it will force the remaining prostitution underground, where the risk of each individual transaction will be greater.

On the other hand, Melissa Farley argues, based on experiences from The Netherlands, Germany and New Zealand, that decriminalisation and regulation may not provide many of the benefits that proponents believe it promises, because (1) stigma will continue to attach to prostitutes and colour the reactions of customers, healthcare workers, and police officers, (2) prostitutes may prefer illegal prostitution to registered legal prostitution, irrespective of the benefits attached, in order to preserve anonymity and (3) zoning and social pressures may push legal prostitution into remote, isolated areas increasing the risks prostitutes face.20

Which of the policies produces the least harm will be a difficult question to answer, since it hinges on complicated and controversial empirical assessments of the effects of each policy, and on contested normative principles about how to weigh distributions of harms, but it is not given that prohibition will enjoy an advantage. Even if it does, the smaller the advantage, the weaker the reason the harm-based argument will generate, which is, in turn, important when we consider countervailing reasons.

The second way of countering the harm-based argument is to introduce such countervailing reasons. The most prominent of these is the antipaternalist challenge, that prohibition constitutes an unjustifiable interference in the freedom of consenting adults. In de Marneffe's version, the argument is ‘Discretionary control over one's own sexual activity, over what sexual acts voluntarily to engage in with other willing adults, is central to sexual autonomy, to control over one's sex life, to control over one's body, and so to personal autonomy. [...] It is important that adults have the discretion to make personal choices about the kind of sex they engage in with other adults, even if these choices are unwise. So it is objectionable for the government to prohibit a person from using her own body and sexuality for prostitution.’18

This is sometimes held to be an absolute constraint, so that any form of hard paternalism, that is, paternalism which does not act upon ill-informed or non-autonomous choices, is impermissible.4 ,7 ,14 Although the limits of paternalism is a complex topic, which deserves a fuller treatment than I can give it here, this seems to me clearly too strong. As a rule of thumb we can, I believe, apply what we might call the ‘seatbelt-condition’: a plausible antipaternalist principle must not rule out such intuitively benign forms of legislation as the prohibition against driving without wearing a seatbelt.

How might we account for such a moderate antipaternalist principle? One possibility is to hold that the personal freedom of autonomous individuals has moral value. That is, we might proffer a freedom-based argument against prohibition

  1. Prohibiting prostitution limits freedom.

  2. We have reason to avoid limiting freedom, the strength of such reason correlating directly with the degree to which freedom is limited.

    C. Ergo, we have reason to not prohibit prostitution.xii

The first premise is obviously true. We can account for the second premise by holding with Mill that freedom is instrumentally valuable,xiii in that persons have different preferences and tend to better promote their own good than others, so that giving them limited freedom to pursue their ends tends to maximise the total sum of good in the long run.33 Or we may hold that personal autonomy has moral importance independent of its contribution to well-being, or to value more generally. Either account allows us to say that we have a reason against a prohibition that must be outweighed by reasons in its favour, which explains the seatbelt condition, since a small limitation of freedom, such as that of being required to use a seatbelt, will produce a weak reason that is outweighed by a substantial reason, such as that produced by avoiding the harm of losing lives in traffic accidents.xiv

Let me mention one potential complication before concluding this section. For some may worry that the distinction between prohibitions against purchasing and selling sex is relevant to the liberty-based argument. That is, could there be different reasons against prohibiting one and the other? One suggestion could be that the crucial liberty at stake is the freedom to consent to sex, while the opportunity to have sex is not similarly morally important. And, the argument could claim, prohibiting the sale of sex limits the first, while prohibiting the purchase merely limits the second.xv

While this is an issue that deserves fuller treatment than I can give it here, I believe there is some reason to be skeptical that the distinction is morally significant in this way. I am inclined to accept, at least for the sake of argument, that some elements of sexual liberty are morally more important than others, but I do not think that the distinction between them tracks the division between prohibiting the sale and purchase of sex, respectively. Consider, that certainly both forms of prohibition deny prostitute and client of opportunities for sex, in at least roughly equal measure. But it also seems to me that both prohibitions limit the freedom to consent to sex in much the same ways—after all, the client who is propositioned by a prostitute is no more free to consent to sex with her, under a prohibition against purchasing sex, than the prostitute who is propositioned by a client is, under a prohibition against selling it.

To sum up this section, I hope to have shown that the need to balance contrary reasons means that to settle even the limited part of the issue of the justifiability of prohibiting prostitution at stake here, we will need to compare the strength of the reason we have to prohibit so as to limit harm with the strength of the reason against prohibiting because this limits freedom.xvi I shall not attempt to do so here, because once again, to do so properly we will need to resolve difficult empirical and further normative issues.xvii This no more means that prohibition is unjustifiable than that it is justified, nor does it mean that deciding the issue is unresolvable. It does mean, however, that the case for prohibition is murkier and weaker than its proponents sometimes suggest. And this, in turn, means that the burden on an argument for exemption is correspondingly lighter.

Disability and sexual need

I have suggested in the above that the case for prohibiting prostitution is more complex than one might hope. The answer hinges ultimately, I believe, on empirical issues about which we still have conflicting and insufficient data, as well as on further clarification of the moral principles at stake. However, let us assume for the sake of argument that the balance of reasons favours a general prohibition of prostitution. This finally allows us to turn to the key question of whether we should make an exception for disabled persons.xviii

There are two facts which, together, form the basis of the case for an exception: the first is that many or most persons have a sexuality that generates strong needs for sexual relations; the second that some disabled persons are partially or entirely incapable of satisfying this need except through the purchase of sexual services from a prostitute.

The most obvious cases concern persons with conditions that at once limit their range of prospective partners because they are viewed by many as sexually undesirable and make them physically incapable of having sex with many or all potential partners. In an illustrative example Teela Sanders quotes a prostitute, one of whose regular clients was a man ‘...who couldn't walk and his carer would bring him. You had to lift him out of the wheelchair and into the jacuzzi and he was stiff because he didn't move his arms or legs. He couldn't move, could get an erection but that was about it.’35

But the difficulties need not be caused exclusively by disability in this narrow sense. Sanders quotes a 59-year-old male respondent with restricted movement in his legs, who sought out sexual services because his disability in combination with low self-esteem, which was itself partly a result of his disability, made other sexual relations difficult: ‘Because I'm not very big, I'm only 5 foot 4, I've got small hands, small feet and small something else, I'm not your alpha male and so tie that in with my own insecurities, I guess sex has always been difficult. [...] I have quite a low self-image and I thought hey even a bloke like me can do this [commercial sex]. It was a matter of feeling in inverted commas ‘normal’ because remember I'd been celibate for about 16 years.’35

Defining disability is notoriously difficult,1 ,36–38 including crucially the extent to which the limitations a disabled person faces should properly be seen as the result of a biological condition or as the result of the social setting in which she functions.xix I shall not attempt to settle such issues here. I adopt, instead, what I hope is a suitably circumscribed definition, according to which a person is relevantly disabled, if and only if (1) she has sexual needs and desires to exercise her sexuality and (2) she has an anomalous physical or mental condition that, given her social circumstances, sufficiently limits her possibilities of exercising her sexuality, including fulfilling her sexual needs.

Note three things about this definition. First, the notion of relevance at stake here is, of course, relevance to the problem at hand, that is, the prohibition of purchasing sexual services. Second, my use of the feminine in the definition is not intended to indicate that the relevant persons are exclusively or predominantly female; in fact, we can safely assume the opposite, since this is the general situation with respect to persons who purchase sexual services. Third, while the qualification ‘sufficiently’ is necessary to rule out exempting anyone who faces any difficulties whatsoever, it introduces a problem of where exactly to draw the line between persons who are and persons who are not ‘sufficiently limited’ in their ability to exercise their sexuality. While this is a difficult and an important task, I believe it is also primarily a practical one that can be set aside here. My task at present is merely to investigate the reasons that speak for and against drawing a line at all.xx The question, then, is what arguments can be given for a legal exception for this group.

The argument from sexual rights

Much of the discussion in Appel, di Nucci and Sanders focuses on sexual rights. Appel explictly understands the right in question as first and foremost a negative right grounded in autonomy: ‘In short, sexual liberty means the autonomy to make one’s own sexual decisions independent of state or societal interference.’3 Appel takes this to imply that we have reasons to create a legal exemption for the relevantly disabled: ‘If sexual pleasure is a fundamental right, as this author believes, then jurisdictions that prohibit prostitution should carve out narrow exceptions for individuals whose physical or mental disabilities make sexual relationships with non-compensated adults either impossible or highly unlikely.’3

Appel also appeals to a positive right, which would ground duties on behalf of others, centrally the state, to guarantee the right, that is, by providing sexual services, or the funding to purchase sexual services, to those who could not otherwise obtain them. Di Nucci argues that this positive right to sexual pleasure is implausible because, pace Appel, a negative right ‘to choose and practice one's sexual attitudes, orientation, and partners’4 is incompatible with a positive right to fulfilment of sexual needs.xxi However, Di Nucci takes this to further undermine the argument for a legal exemption: ‘...the consequent need to (partially) legalize prostitution [...] depend[s] on his claim about the positive human right to sexual satisfaction.’4 This is a mistake. The legal exception requires not the positive right, but the ‘negative right to sexual pleasure’ that Di Nucci and Appel agree exists, since this is the right that makes it ‘...wrong to prevent disabled people from satisfying their sexual interests’, which is exactly what legal prohibition does. Apart from this, Appel's argument and Di Nucci's critique focuses on the implications of the positive right, and we can set them aside for present purposes.

There remains some ambiguity about the right at stake in the different formulations Appel employs, specifically whether it is meant to be an interest-based or a will-based right. We can charitably resolve this by considering either version. The argument, it seems then, is

  1. Persons have either (1) a negative (will-based) right to sexual liberty, or (2) a negative (interest-based) right to fulfilment of their sexual needs.

  2. We have (decisive) reason to not violate rights.xxii

  3. Prohibition violates the negative sexual right, by limiting persons’ freedom to engage in sexual relations.

    C. Ergo, we have (decisive) reason to create a legal exception from prohibition for relevantly disabled persons.

Note three important points about the argument: first, that what apparently grounds the alleged right is either an interest or autonomy; second, that Appel offers a claim about a right rather than an argument that such a right exists and third, that the right in question is on either account a right shared by disabled and non-disabled persons alike.

These points raise several problems. Taking them in reverse order, what we might call ‘the scope challenge’ concerns the fact that all persons possess the right in question, and that prohibition infringes their rights equally. This makes it hard to see how it could constitute a premise in an argument for a legal exception for disabled persons. Di Nucci rightly stresses that the alternative, a sexual right that only the disabled enjoy, is implausible,4 and nothing in Appel's text seems to me to support it. It seems, if anything, an argument against prohibition tout court.

Second, what we might call ‘the justification challenge’ concerns the fact that we require an argument that a right exists and has particular content. Broadly speaking, there are two ways in which one can argue for the existence of a moral right. On one understanding, a negative moral right to X is what you have if, all things considered, it would be morally wrong to prevent you from enjoying X. On this account, rights give us no reasons for or against something being right or wrong, rather, in John Oberdiek's formulation: ‘...rights represent the conclusion of all of the reasons bearing on the justifiability of a given action.’41 On this interpretation, premise 2 is true by definition, but premise 1 of the argument is itself the conclusion of a different argument, namely the one which I have suggested is too complicated to here assess: the overall justifiability of legally prohibiting the sale of sex. Clearly, on this account, the premise begs the question.

On the alternative, substantial understanding of rights, roughly, you have a negative moral right to X if there is a moral property of the situation that generates a decisive reason against preventing you from enjoying X. But then we require an argument that such a property obtains, and an explanation of how it generates the reason. Rights are a particular type of powerful moral claim, and to merely postulate them is, in Bentham's memorable phrasing, ‘nonsense on stilts’,42 or, as Norman Daniels somewhat more sympathetically put it: ‘Rights are not moral fruits that spring up from bare earth, fully ripened, without cultivation.’43 In short: on either account, a moral right must be the conclusion, not the premise of an argument. While it is of course possible that such an argument could be produced, until we see it we should not take the existence of the right in question for granted.

Third, we might raise what we can call ‘the superfluence challenge’, which concerns the fact that if we do develop an argument that justifies a right, then most or all the interesting normative work will be done by the components of that argument. Indeed, it seems that we have already considered the reasons we have against prohibition based on concern for personal liberty as part of the discussion of criminalisation, and that we did not need to introduce the concept of a right to do so. Adding a right grounded in autonomy at this point will then be either double-counting or restating the same point in less concise form.

In conclusion, it seems that an argument from sexual rights is likely to be, at best, superfluous and, at worst, misleading. Let us look at two arguments that do better.

The argument from beneficence

Possibly the most straightforward argument for allowing prostitution is that it provides benefits to the clients. In a sense this is almost trivially true: since clients engage in a consensual exchange of goods for services, the client presumptively considers the benefit of what she receives greater than the cost of what she pays. Or bluntly: customers pay for sex because it is worth more to them than other things they could have done with that amount of money.

The above applies to all clients, and therefore provides, if anything, an argument against prohibition in general. However, there is reason to think that the group of relevantly disabled persons stand to lose a particularly great benefit, so that even if, as we have assumed arguendo, the benefit that clients, in general, would obtain from prostitution, is outweighed by the reasons that speak in favour of prohibition, the same may not be true for this group.

The argument therefore is

  1. We have reason to promote the well-being of persons (beneficence).

  2. Most persons have strong sexual needs, the welfare function of which is strictly concave, that is, the welfare gains of sex are greater the less fulfilled the need is, and vice versa.

  3. For relevantly disabled persons, prohibiting the purchase of sexual services will preclude any (or almost any) fulfilment of their sexual needs.

  4. The welfare-loss of these persons from prohibition will be much greater than the welfare loss of the average person (from 2 and 3).

    C. Ergo, we have a much stronger reason of beneficence to not prohibit these persons from purchasing sexual services (from 1 and 4).

The natural question the argument raises is how strong an interest persons have in exercising their sexuality, or more technically, how steep the welfare function is, and therefore, how strong the moral claim that can be grounded in beneficence is. As Davidson stresses, sexual needs that go unfulfilled do not result in bodily harm the way lack of food, water, or sleep do.6 However, most of us accept that persons can have strong interests that do not lead to bodily harm when frustrated; consider, for example, the interest in having children and securing their well-being, or the interest in social interactions and friendship. And Sanders emphasises that although the good traded is sexual services, the need fulfilled should not be understood exclusively as a need for sexual pleasure in its most physical and limited sense. The sexual services are valued also for the physical and emotional intimacy, or, if we suppose that the shared intimacy may not always be genuine, then at least the impression of intimacy that they provide. Sanders quotes a 44-year-old ex-builder who uses a wheel chair after a work-related accident: ‘You know it's wonderful to feel flesh on flesh when you haven't felt it for years. And quite apart from anything else [...] I'm enjoying being with them. We have what the Irish would call the craic, and we talk and share things.’35

The argument will be even stronger for those who accept an axiology in which a benefit is morally more important when it accrues to those worse off, for example, telic egalitarians and prioritarians. How much stronger depends on views about the well-being of disabled persons, but some will probably hold that the group is, generally speaking, worse off in a number of ways, since this is the ordinary result of disabilities.xxiii If this is true, and if as noted above we believe that benefits matter more when they accrue to those worse off, then it further strengthens the argument from beneficence.

Other considerations of beneficence may differ as well. Thus, there could be an undercutting argument, directed at the harm-based argument for the badness of prostitution, to the effect that given the circumstances under which the sale of sexual services to the relevantly disabled take place, it is unlikely to cause harm, or at least, to cause as much harm, as prostitution generally does. Such lessened harm could plausibly be supposed to result from the different power relationship between the prostitute and a disabled client.xxiv As Sanders puts it: ‘In a situation where the client does not fulfil the alpha male stereotype and has a different status because of his impairment the power dynamics may be different because of the differences in social status and physical ability.’35 If this is empirically correct, the case for a prohibition is weakened, while the argument from beneficence sketched above strengthens the case against it.

The argument from luck equality

The second argument builds on the idea of unfortunate inequality providing independent moral reasons. The thinking draws on ideas developed in John Rawls’ ‘A Theory of Justice’,47 but the most influential statements are probably Cohen's48 and Arneson's.49 The basic luck-egalitarian idea can be put roughly as follows: it is unjust when a person is worse off than others, if the person who is worse off is not responsible for being worse off.xxv

The principle will have strong intuitive support for many. In one classic illustration, it seems unjust if A is much worse off than B, say with respect to opportunities for and level of welfare, despite A and B making equal efforts at improving their lives, simply because B was fortunate to be born in a wealthy and resourceful family, while A was unfortunate to be born in a poor and dysfunctional family.xxvi

In the context of an exemption for the relevantly disabled from legal prohibition of prostitution, the argument would be:

  1. It is unjust if a person is worse off than others, without being responsible for being worse off (luck-egalitarianism).

  2. We have reason to prevent injustice.xxvii

  3. For relevantly disabled persons, prohibiting the purchase of sexual services will preclude the feasible way of their becoming less worse off than others with respect to fulfilment of sexual needs.

  4. Relevantly disabled persons would not be responsible for a legal prohibition of prostitution; at least some disabled persons are worse off than others with respect to fulfilment of sexual needs through no fault of their own.

    C. Ergo, for at least some relevantly disabled persons, we have strong reason to not prohibit their purchasing sexual services.

At this point, it may be tempting to introduce two of the standard objections raised by relational egalitarians, for does not the argument above distinguish in an unmerciful manner between those who are and those who are not responsible for their disabilities? And would not a policy, which exempted the latter but not the former, require an unacceptable form of intrusion into the privacy of candidates for exemption, so as to determine which of the two groups they belong to?51

I believe that the most obvious reply open to a proponent of the luck-egalitarian argument at this point is that, on balance, a policy which avoids these problems by exempting all relevantly disabled persons is preferable. That is, a luck-egalitarian might concede that abandoning those responsible for their disabilities is in some respect bad, and that trying to distinguish between the two groups would require a form of intrusion that is also in some respect bad, while maintaining that the state of affairs in which those not responsible for being worse off are worse off is unjust. Saddled with the unenviable choice between the imperfect policies of no exemption, which leaves some worse off through no fault of their own, determinate exemptions, which abandons some and requires intrusion, and blanket exemptions, which exempts some for no reason, it seems that the luck-egalitarian could plausibly maintain that blanket exemptions is the least bad policy, and that it is the least bad for, at least partly, the very reason emphasised by luck-egalitarians.xxviii


Over the preceding four sections, I have explored the problem of whether we ought to provide a legal exception for disabled persons to a prohibition against prostitution. I have suggested that there is sparse and conflicting sociological evidence about the conditions and consequences of prostitution but that it seems clear that prostitutes suffer harm, and argued that causing this harm constitutes one morally bad property of prostitution. I have further discussed the forms of legal policies available and the requirements we need to make of an argument for prohibition, arguing that the relative strengths of alternative policies, and the contrary reason generated by the limitation of liberty which prohibition imposes, jointly weaken the argument for prohibition. Finally, I have attempted to define the group at stake and the conditions that provide the basis of the argument for an exception, critically discussed a rights-based argument, and presented two alternative arguments based on beneficence and luck-egalitarianism, respectively.

I have emphasised throughout that the discussion has been narrow, setting aside additional elements of the problem, complications, and difficult assessments. To that extent, it is largely a sketch of the structure of the problem, which leaves important details necessary for an all-things-considered conclusion to be filled out by future work. But I believe it sets out the overall framework and goes some way towards accounting for the most important moral factors at stake. As such, this paper marks neither the beginning nor the end of the discussion. But I hope that it may constitute at least another step on the way.

Although we are left in a kind of ‘not-enough-data-to-compute’ situation, the question remains what we ought to do about an exception in the meantime. Given the limitations acknowledged above, any reply will necessarily be conditional and tentative, but it seems to me that the arguments from beneficence and luck-egalitarianism provide us with two strong reasons, and that the balance of reasons regarding a general prohibition is at best moderately in favour. While we should devote more effort to exploring the problem and reassess this conclusion in the light of further analysis, there is presently, I believe, a relatively good case for a legal exception.


I am grateful for comments and suggestions from Andreas Albertsen, Rune Klingenberg Hansen, Kasper Lippert-Rasmussen, Jesper Ryberg, Fatima Sabir, Theresa Scavenius, Adrian Viens, and two anonymous JME referees. I owe particular thanks for very detailed comments to Thomas Søbirk Petersen.



  • A draft version of this paper was presented at a Roskilde University Moral Philosophy seminar.

  • Contributors FKT is the sole author and contributor to the paper.

  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • i Thus, Sweden, Norway, Iceland and France have all adopted ‘the Swedish-model’ of criminalising the purchase, but not the sale, of sexual services, while The Netherlands, Germany, Austria and New Zealand have all legalised and regulated the sale of sexual services. In most other European and Anglophone countries, including the UK, Spain, Italy, Denmark, Finland, most of Australia and Canada, prostitution is decriminalised and loosely regulated.

  • ii For good overviews of the contemporary debate on the ethics of disabilities cf. Hartley;1 Wasserman et al.2

  • iii The terminology of the prostitution debate is contested, in that some prefer, for example, ‘sex work’, so as to avoid any negative connotations clinging to ‘prostitution’, and others, e.g., ‘survivor’, so as to emphasise what they take to be its essentially violent nature. In the present context, I use ‘prostitution’, ‘prostitute’ and ‘sexual services’, but I intend to convey no particular attitude to, or understanding of, the phenomena through my choice of these terms. Although this is unlikely to satisfy all participants to the debate, I believe that the substance of my arguments should not depend on the particular terms employed, and invite those readers who prefer a different terminology to read, e.g., ‘sex worker’ or ‘survivor’, for ‘prostitute’.

  • iv Note, that I thereby also exclude a number of arguments based on harm to other groups. For example, one frequently posed argument is that prostitution increases the demand for trafficked women, cf. Dempsey (pp.1752–3);17 de Marneffe (pp.37–8);18 another, that clients are at increased risk of STDs, and clients’ families at increased risk of heartache and divorce.

  • v The empirical evidence and the exact nature and causes of this harm are contested, but that many prostitutes suffer greatly seems indisputable. For an indication of how divided the debate is on this issue, cf. Farley et al;19 Farley;20 versus Weitzer;21 ,22. The only metasurvey with which I am familiar is Ross et al;23 although cf. also Vanwesenbeeck.24

  • vi Cf. also de Marneffe, (pp.23–6);18 Davidson, (p.91);6 and Gauthier, (pp.168–73).8

  • vii I am grateful to an anonymous referee for pressing me on this point.

  • viii I draw here on views of the complications and how to solve them, presented by Holtug29 and Søbirk Petersen.30

  • ix Moen's only other apposite remark, as far as I can tell, is his somewhat peculiar constraint that persons are not harmed, or if they are, are harmed by their attitudes, if their attitudes are contributing factors; his example being taboos on pork consumption. In the homosexuality case, maintaining this analogy seems to suggest that homosexuals were harmed by neither their homosexuality nor the stigma which society attached to them, but by their (reasonable) desire to be treated with respect. This does not seem to me to help his argument.

  • x That is, it may impose restrictions on prostitution, but only restrictions that apply generally and to all types of service provisions, such as requirements on working conditions, opening hours, zoning, etc.

  • xi Note, that de Marneffe labels prohibiting the purchase, but not the sale, of sex ‘abolitionism’, and distinguishes this from ‘prohibitionist’ policies. This strikes me as an unnecessary terminological complication.

  • xii De Marneffe's antipaternalistic principle initially appears stronger than this. His first formulation holds that a person's benefiting from a limitation of her freedom cannot justify this limitation, unless (1) she lacks mental maturity, (2) she is not opposed to the limitation, (3) the liberty she loses is unimportant, or (4) she desires to do what the policy prohibits because she is relevantly ignorant or mentally unstable, cf. de Marneffe, (p.67). However, in the context of discussing paternalism, it becomes clear that de Marneffe believes that assessing condition (3) involves the type of weighing of competing reasons that I have here sketched, i.e., a liberty is unimportant if its value to the individual is clearly outweighed by her other interests.

  • xiii The obvious limit to freedom being, as Mill emphasised, the freedom to harm others, since this freedom would inevitably lead many to pursue their own good at the expense of others, and thus likely decrease aggregate well-being. This is not, of course, the only reading of Mill's position, or perhaps even the most common one, however, for a persuasive argument that it is the most plausible reading; cf. Crisp, chapter 8.32

  • xiv Obviously, mutatis mutandis, the opposite holds true: a small gain in harm-avoidance will be outweighed by a substantial loss of freedom.

  • xv I am grateful to an anonymous referee for pressing me on this point.

  • xvi To completely settle the issue, we would need, as I have indicated before, to consider also any other reasons that bear on the justifiability of prohibition, such as any other badness attached to prostitution and the human and economic costs of criminalisation.

  • xvii For more extensive discussion of the issue, cf. Dempsey;17 Law;25 Chamallas;34 and particularly de Marneffe, chapter 3.18

  • xviii Note that much like the terminology of the prostitution debate, the terminology of the disability debate is fraught with conflict. As per my comments in that context above, I intend to convey no particular attitude or understanding through my choice of the terms ‘disability’ and ‘disabled persons’. And, although as before, this is unlikely to satisfy all participants to the debate, I believe that the substance of my arguments should not depend on the particular terms employed, and invite those readers who prefer a different terminology to read, e.g., ‘impairment’ for ‘disability’ or ‘person with disabilities’, for ‘disabled person’ where necessary.

  • xix To give a crude example, does a person with limited mobility in her legs face limitations because of her condition, or because of her inclusion in a social setting that implicitly expects persons to be capable of walking, and structures the physical and social environment accordingly?

  • xx The line between concerns of implementation and reasons that count against the policy itself can blur, of course. Thus, if it turns out that any feasible implementation will be misused to grant exceptions to far too broad a segment of the public, effectively undermining prohibition, or that a sufficiently careful procedure to grant an individual exception will be enormously costly, then naturally these concerns count against a policy of exception. As they depend upon, and must be evaluated in the context of various concrete candidate policies, however, I set them aside here.

  • xxi For similar arguments, cf. Davidson, (p.90);6 Sanders, (pp.450–2).35

  • xxii Appel does not speak of rights as providing decisive reasons, but it is commonly held, following Joseph Raz's analysis, that this is one of their distinctive features, cf. Raz.39 In Dworkin's likewise classical formulation, rights are characteristic in that they trump other considerations, cf. Dworkin.40

  • xxiii Cf. Savulescu and Kahane;44 Kahane and Savulescu.45 The idea that disabilities make a person generally worse off is not, however, uncontroversial, cf. Goering.46

  • xxiv I am grateful to an anonymous referee for pressing me on this point.

  • xxv A number of variations and complications exist, concerning what the appropriate currency for measuring being well or poorly off is (i.e., well-being, goods, opportunities, or capabilities), whether the conditional should be exclusive (i.e., if and only if), what to think about cases of being better off through no responsibility of one's own, whether and if so how the results of voluntarily adopted risks should count (option-luck vs all-luck), and how to define being responsible in the relevant sense. For an excellent overview of the discussion and literature cf. Knight.50

  • xxvi Indeed, the most important contemporary criticism comes from relational egalitarians, who argue that in at least some respects, luck-egalitarianism is insufficiently egalitarian, because of how it allows responsibility to heartlessly justify disparities, cf. Anderson;51 for critical responses, cf. Arneson;52 Lippert–Rasmussen.53 ,54

  • xxvii Many luck-egalitarians emphasise that what is at stake here is justice, and take this to produce different or stronger claims than, e.g., beneficence. How much stronger, and why, is a matter of contention.

  • xxviii To deny this, a luck-egalitarian would need to believe that the scope of moral reasons, or perhaps more narrowly justice, is exhausted by concern for those worse off without being responsible for being worse off. But I am uncertain whether this is a position any luck-egalitarians actually hold, and in any case it seems to me a position luck-egalitarians should avoid adopting.

Linked Articles

  • The concise argument
    Russell Powell

Other content recommended for you