J.S. Mill's Puzzling Position on Prostitution and his Harm Principle

Abstract

J.S. Mill argues against licensing or forced medical examinations of prostitutes even if these would reduce harm, for two reasons: the state should not legitimize immoral conduct; and coercing prostitutes would violate Mill's harm principle as they do not risk causing non-consensual harm to others, their clients do. There is nothing puzzling about Mill opposing coercive restrictions on self-regarding immoral conduct while also opposing state support of that conduct. But why does Mill oppose restrictions on prostitutes’ liberty if those restrictions could prevent harm to third parties? Mill's position is not puzzling once we recognize that his harm principle is not a harm-prevention principle that warrants restrictions on liberty to prevent harm no matter who caused it (as David Lyons famously argued) but instead warrants restrictions on liberty only of individuals who are the morally relevant cause of that harm. Mill's discussion of prostitution shows he prioritizes both individuality and moral progress over harm reduction.

1. Introduction

John Stuart Mill thinks prostitution is immoral. In a letter to Lord Amberley of Feb. 2, 1870 Mill writes that prostitution is ‘second only to rape’ in its ‘evil propensity’ to satisfy sexual desires; it offers not even a ‘temporary gleam of affection and tenderness’ and completely uses a woman as a mere means for a purpose she must find disgusting (CW 17:1693).¹ Because prostitution is immoral, Mill does not think the state should legitimize it by regulating or licensing prostitutes. In 1871 Mill testified against the Contagious Diseases Acts (CW 21:351–71), hereafter referred to as ‘the Acts’. The Acts required suspected prostitutes to be examined and forcibly detained for treatment if found to have a sexually transmitted disease (STD), a primary aim being to protect soldiers who frequented prostitutes (Jose and McLoughlin, 2016, pp. 254–56; Waldron, 2007). One reason Mill objects to the Acts is that like licensing schemes, they encourage immoral conduct by making it safer, and Mill believes that rather than focus on harm-reduction, the state should promote moral progress.² In his testimony against the Acts Mill relies on his moral objection to prostitution also in claiming that police may prevent solicitation in the streets (CW 21:369); presumably solicitation by prostitutes, when done publicly, is the sort of ‘offence against decency’ that in On Liberty he says the state may prohibit (OL, CW 18:295–96). In addition to these two claims – that the state should not license prostitutes, and that it should prevent public solicitation, both of which are motivated by Mill's commitment to moral progress – Mill makes a third claim about prostitution that may seem at odds with these first two. Mill objects to the Acts also because they impose ‘a penalty for being a common prostitute’ (CW 21:352) and he does not think prostitution should be illegal.

It isn't puzzling for Mill to think that prostitution should be legal while also thinking that the state should not morally condone or legitimize it.³ What does seem puzzling is that in his seminal work of political philosophy, On Liberty, Mill supports state restrictions on liberty in order to prevent harm, yet is unwilling to restrict prostitutes to prevent them from harmfully spreading disease. As we'll see, Mill reasons that any harm they cause to their client was consented to, and if their client proceeds to spread an STD to a third party, they and not the prostitute cause that harm. Jeremy Waldron finds Mill's opposition to the Contagious Diseases Acts ‘bewildering’ given Mill's defence of the harm principle. In On Liberty Mill defends individuality (OL ch. 4), which his harm principle promotes by ensuring that individuals are free to engage in self-regarding conduct even if it flouts customs or social norms, so long as their conduct doesn't harm others. According to the harm principle, the only end for which the state may legitimately exercise coercive power is to prevent harm to others. On Waldron's view, given that the Acts aim to curb the spread of STDs and thereby reduce harm, shouldn't Mill support the Acts? (Waldron, 2007, p. 16) I have two objectives: a subsidiary goal is to clear up the puzzlement about Mill's views on prostitution; my larger goal is to show how Mill's position on prostitution casts doubt on a prevalent interpretation of the harm principle and motivates us to seek an alternative understanding of this central feature of Mill's political philosophy. I argue that Mill's position is not puzzling after all, for the goal of Mill's harm principle is not harm reduction.

In section 2 I establish that Mill believes prostitution should be legal, even though he opposes the licensing of prostitutes and believes police should prevent public solicitation. In section 3 I argue that Mill's position rests on a particular interpretation of the harm principle that has been in recent disfavour: that the harm principle warrants the use of coercion only on those who themselves proximately cause non-consensual harm, and not to prevent harm no matter who caused it. This interpretation – that the harm principle is a ‘harm-causation’ principle – was originally laid out by D.G. Brown (Brown, 1972). In section 4 I resurrect and extend Brown's position, defending it against David Lyons’ opposing view that Mill's harm principle is a ‘harm-prevention’ principle, and that Mill would permit the state to coerce an individual to prevent harm to others even if that individual did not proximately cause the harm (Lyons, 1997). In section 5 I show how Mill's claim that the state should not license or legitimize prostitution further supports the position that Mill's primary concern is not harm reduction. Licensing would make prostitution safer, yet Mill opposes licensing, because he thinks the state should promote moral progress. Section 6 then addresses an apparent inconsistency between Mill's view that prostitution should be legal, and his view that the police should prohibit public solicitation. If the harm principle requires that prostitution must be permitted because the prostitute is not the morally relevant cause of non-consensual harm to others, why would Mill restrict public solicitation? While my main purpose is to interpret rather than evaluate Mill, in section 7 I conclude with some evaluative comments about the implications of Mill's defence of a harm-causation as opposed to a harm-reduction principle.

2. Mill's Position on Prostitution

While Mill thinks prostitution is immoral, and for that reason opposes state licensing of prostitutes, he defends a principle of liberty – the harm principle – according to which the prostitute should be free to engage in immoral, self-regarding activity in private. Mill opposes ‘legal moralism’, or the view that the state may legally punish conduct that is regarded as immoral even if that conduct doesn't harm others.⁴ In his testimony on the Acts Mill opposes seduction and bastardy laws, explaining: ‘at present my feeling is against any attempt however much it may be agreeable to one's moral feelings, to restrain illicit intercourse in that way’.⁵ Laws should keep us from harming others, but not force us to be moral.

While consensual sex between a man and a prostitute may not be entirely self-regarding as it can put the man's wife or other intimate partners at risk of receiving an STD, nevertheless Mill doesn't think that prostitutes should be punished for selling their sexual services.⁶ Mill doesn't explicitly say this in On Liberty but there is compelling textual evidence that this is his position. First, Mill says in the ‘Application’ chapter that ‘[f]ornication, for example, must be tolerated’ (CW 18:296). That alone is no proof that Mill thinks prostitution should be legal since one could think that fornication with a prostitute should be treated differently. But he then immediately takes up a puzzle: while fornication must be tolerated, ‘should a person be free to be a pimp?’ (CW 18:296). Pimps are ‘accessories’ to prostitution by facilitating the transaction between prostitute and client, and Mill wonders why we should punish ‘the accessory when the principal [the prostitute] is (and must be) allowed to go free’: why fine and imprison ‘the procurer, but not the fornicator?’ (CW 18:297). Mill was torn by a similar question 12 years later when in his testimony on the Acts he is unable to conclude on the ‘very difficult’ question of whether brothels should be permitted (CW 21: 359–60; cf. 369).⁷ My point is that the question of why we should punish the pimp but not the prostitute is puzzling for Mill only because he assumes that we should not punish the prostitute. This is as close as we get to direct evidence in On Liberty that Mill does not think prostitution should be a crime. In his testimony on the Acts, Mill supports the criminalization of prostitution for girls under 17 – but only because they aren't yet adults and so their liberty can properly be interfered with (CW 21:368) – the implication again being that adult prostitution should not be illegal.⁸

3. The Harm Principle and Mill's Position on Prostitution

Mill defends our liberty to engage in self-regarding activity that could not harm others, but as prostitutes risk spreading harmful STDs, wouldn't Mill have good reason to think prostitution should be illegal? In this section I argue that Mill's harm principle is not a harm-reduction principle. It does not permit the state to coerce me merely if doing so would reduce the amount of harm in the world. It may coerce me only if I am the morally relevant cause of that harm; and Mill does not regard the prostitute as the morally relevant cause of harm when their client spreads an STD to an innocent third party.

In On Liberty Mill introduces the harm principle as holding ‘that the sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any of their members, is self-protection. That the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others’ (CW 18:223). There are several ambiguities as to what the principle means, and I now address two that are important in understanding Mill's position on prostitution. In later sections I address a third ambiguity.

First, what constitutes ‘harm to others’? Some commentators interpret this broadly: if you do something that upsets, offends, or merely displeases me, such as having sex in public in plain view of me or my children, and I don't consent to your activity, you have harmed me and your activity can be regarded as ‘other-regarding’ and subject to possible interference.⁹ On this view, society could have jurisdiction over you if you prevail over me in a competition in business or athletics, or the courting of someone we both love, because that would harm me using this wide conception of harm. But Mill clearly rejects that account of harm. In On Liberty he says the state should not intervene when someone loses out in a competition, because ‘society admits no rights, either legal or moral, in the disappointed competitors’ (CW 18:293). This passage indicates that for Mill, harming others involves more than producing a bad consequence such as offending or displeasing them. It sets back interests they have that are regarded as rights, through an act to which they do not consent.¹⁰ If there is no violation of a right, there is no harm.¹¹ Mill adds that there should be a ‘definite damage, or a definite risk of damage’ or ‘perceptible hurt to [an] assignable individual except himself’ for an action to be placed in the ‘province […] of morality or law’ (CW 18:282). These clarifications help limit the ambiguity of Mill's principle, but Mill still leaves open the question of what rights there are. Does my failing to rescue a drowning person, when I easily could, constitute ‘harming them’ and therefore legitimately expose me to punishment? That depends on whether they had a right to be saved. It isn't always clear on Mill's view what rights society ought to declare. I return to this ambiguity in section 4.

I've already introduced a second ambiguity of the harm principle. It might mean what the words in Mill's introductory statement of it literally say: the state may coerce individuals if doing so will ‘prevent harm’, regardless of whether the person being coerced caused the harm – this is the harm-prevention principle. But there is a competing interpretation: the state may use coercion upon individuals only if those individuals are the morally relevant, ‘proximate’ cause of non-consensual harm to others. According to this ‘harm-causation’ principle, there are two conditions that must be met for the state legitimately to coerce me. First, I must be a proximate cause of harm to others. It is not enough that ‘but for’ my conduct harm to others would not have resulted; my conduct must have a direct connection to the resulting harm, with no intervening voluntary cause of that harm that would nullify my responsibility for it.¹² For example, if I fire a gun at you and miss, but cause you to flee in panic, and as a result you are injured after running in front of a truck that can't avoid hitting you, I am the morally relevant cause of your injury; but if the truck driver could have easily avoided you but purposely struck you, they are an intervening voluntary cause that replaces me as the most proximate, morally-relevant cause of your injury. Second, even if I do directly injure another party, if they consented to my doing so I cannot be said to have harmed them. If I freely consent to fight you in a duel with pistols at close range, knowing your skill as a marksman, and I am injured by your shot, I consented to that risk and so you have not caused non-consensual harm.

Resolving the ambiguity of whether the harm principle is the ‘harm-prevention’ or ‘harm-causation’ principle is essential in addressing prostitution. Waldron, in finding Mill's opposition to the Contagious Diseases Acts ‘bewildering’, assumes that Mill defends the harm-prevention principle.¹³ His Mill is willing to restrict liberty to prevent harm to others, and since the Acts prevent the spread of disease, shouldn't Mill support the Acts?

Mill recognizes that ensuring the health of the community is within the province of government (Acts, CW 21:357). In On Liberty Mill distinguishes the ‘preventive function of government’ from the ‘punitory function’ and says both may be employed to fight crime (CW 18:294). But we should coerce or punish the right party: not the prostitute, but the one who foreseeably and proximately causes harm to a non-consenting, innocent third party. Mill argues this explicitly in his testimony against the Acts. The Royal Commission conducting hearings on the Acts asks Mill, can't the state get involved if the object is to prevent harm to third parties such as wives or other innocent parties who might get the disease from the man? Mill replies that the woman doesn't transmit the disease to these third parties, the man does, and so the man is more properly targeted if that is the Acts’ aim (CW 21:354). Later he makes a similar point: ‘[I]t is only a man who having been infected himself can communicate infection to an innocent person’ (CW 21:362). Mill suggests that the state impose ‘very severe damages’ on the man (CW 21:354–55) and use military discipline to prevent soldiers from engaging in risky behaviour (CW 21:369; cf. 360). But don't coerce the prostitutes.

Waldron is not convinced. Referring to the ‘consented to’ provision of the harm principle – which says that the protected sphere of liberty includes not only conduct ‘which affects only [my]self’ but also conduct which ‘affects others […] with their [c]onsent’ (CW 18:225) – Waldron writes, ‘Certainly, Mill would have had little patience with the objection that the transmission of infection did not count as harm inflicted by the prostitute because the transaction was consented to’ (Waldron, 2007, p. 18). Waldron acknowledges Mill's testimony that the man ‘knowingly places himself in the way of’ the disease and the women have nothing to do with its direct spread to others (Waldron, 2007, p. 28, citing CW 21:354). But according to the harm-prevention principle that Waldron takes Mill to defend, that doesn't matter: the threshold ‘necessary condition’ for coercing me is met merely if I do something that somehow contributes, even indirectly, to the injury of others, including ‘unknowing (and therefore non-consenting) [third] parties’ (Waldron, 2007, p. 18). This is why Waldron must instead turn to other grounds to account for Mill's opposition to the Acts: by applying only to female prostitutes and not their male customers, they impose an unequal burden based on one's sex.¹⁴ Other scholars who explain Mill's opposition to the Acts similarly turn to Mill's commitment to equal treatment of the sexes. For Jim Jose and Kcasey McLoughlin, Mill opposes the Acts because they reflect ‘sexist thinking’: the real aim of the Acts is to enshrine male privilege; and for Clare McGlynn, Mill opposes the Acts because they wrongly target women instead of the male clients who create the demand for prostitution.¹⁵ Mill also raises due process objections. Under the Acts, police could ‘apprehend’ women on suspicion of being a prostitute and bring them to a magistrate, who could confine them for up to 6 months if they refused to be examined (CW 21:351). Mill objects that police discretion can be abused, and that the hearings did not provide for a jury (CW 21:351–53).

While Mill clearly had equal protection and due process objections to the Acts, his opposition is based more essentially on his assessment that the Acts violate the harm-causation principle. It provides a threshold test for when state coercion is permissible, and the prostitute's activities don't reach its bar. The prostitute can be contrasted with the individual who incites a frenzied mob to commit an imminent act of violence and who Mill says ‘may justly incur punishment’ (OL, CW 18:260). The inciter manipulates the mob and might be seen as a proximate cause of the harm the mob, in its ‘frenzy’, proceeds to nonvoluntarily inflict; but prostitutes who spread STDs presumably do not manipulate their clients, and if their client proceeds to give an STD to a third party the client would be an intervening voluntary cause of that harm. In On Liberty Mill does not explicitly defend one interpretation of the harm principle over the other; but his insistence that ‘the prostitute is (and must be) allowed to go free’ (CW 18:297), and his reason, which he provides in his testimony on the Acts – that the prostitute doesn't transmit STDs to third parties, her client does (CW 21:354, 362) – makes sense only according to the harm-causation principle.¹⁶

4. Re-Interpreting the Harm Principle as the Harm-Causation Principle

Does this interpretation of the harm principle stand up in light of other positions Mill takes, or is Mill's position on prostitution an anomaly? Over 50 years ago D.G. Brown noted that if we take Mill's introductory formulation literally as permitting the state to limit one's liberty if doing so would prevent harm, the state could punish me to deter you from causing harm even though I did nothing that risked harming others. So Brown reformulates the principle to say that ‘the liberty of action of the individual ought prima facie to be interfered with if and only if his conduct is harmful to others’ (Brown, 1972, p. 135).

David Lyons, responding to Brown, defends the harm-prevention interpretation instead. He argues that Brown's harm-causation principle fails to account for some positions Mill takes. Lyons focuses on the following passage from On Liberty:

[There are m]any positive acts for the benefit of others, which [one] may rightfully be compelled to perform; such as, to give evidence in a court of justice; to bear his fair share in the common defence, or in any other joint work necessary to the interest of the society of which he enjoys the protection; and to perform certain acts of individual beneficence, such as saving a fellow-creature's life, or interposing to protect the defenceless against ill-usage; things which whenever it is obviously a man's duty to do, he may rightfully be made responsible to society for not doing. A person may cause evil to others not only by his actions but by his inaction, and in either case he is justly accountable to them for the injury. (CW 18:224–25, my emphasis)

Lyons argues that failing to testify at trial, save a fellow creature's life, or pay taxes to contribute to joint undertakings such as the common defence does not cause non-consensual harm, and so if Mill meant to defend Brown's harm-causation principle, coercive interference would not be warranted for these omissions – yet Mill says one may be ‘compelled to perform’ these acts (Lyons, 1997, pp. 116–17). To avoid that inconsistency, Lyons argues that Mill defends not Brown's version of the harm principle but a ‘harm-prevention’ principle that can justify coercive interference in these cases. According to that principle, ‘[h]arm to others can be prevented not just by interfering with acts that can be said to cause, or that threaten to cause, harm to other persons’; merely preventing harm to other persons suffices as a reason for restricting behaviour (Lyons, pp. 124, 118–19). Giving testimony in court can be required as testimony is needed for the criminal justice system to effectively prevent future harm (121); aiding someone who is injured can be required to prevent further harm to others (119), even if the bad Samaritan – who fails to aid – wouldn't be the proximate cause of harm; and we may coerce individuals to pay taxes because cooperation requirements ‘may well provide the only means of preventing or eliminating some significant harms, such as malnutrition and starvation’ (122). Lyons suggests that on Mill's view one might, as a means of harm-prevention, even be forced to contribute to foreign aid efforts for the purpose of preventing war (123).

Lyons’ account fails to explain Mill's position that prostitution must remain legal, and Mill's opposition to the Acts. Though Mill knows prostitutes can spread STDs, and that the Acts could help reduce that risk, he still insists that prostitutes cannot be coerced because they do not proximately cause non-consensual harm to others.

But that is not the only textual evidence favouring Brown's view that the harm principle is a harm-causation principle and not Lyons’ harm-prevention principle. In On Liberty, Mill says that I should be free to publish potentially dangerous opinions, such as that tyrannicide is lawful, or that corn-dealers starve the poor, even though doing so might inspire one of my readers to harm others. But, he continues, if I directly incite a crime, by delivering my opinion to an excited mob assembled before the house of a corn-dealer, or if my encouraging tyrannicide has a ‘probable connexion’ to a wrongful act, my freedom of speech can be restricted.¹⁷ In theory this distinction could be justified using the harm-prevention principle. One could argue that the harm principle permits restrictions in either case – publishing one's opinions, inciting a particular person to act – to prevent harm; but that whether the state should take measures that the harm principle would permit must be decided by the principle of utility, and that principle would support punishment only of the direct inciter, given the tremendous disutility of chilling speech addressed to a general audience.¹⁸ But that is not what Mill argues. Instead, he argues that publications for the general public – even if they could lead to substantial harms, such as tyrannicide, or attacks on merchants – must be permitted because they are not proximate causes of harm. For Mill, we decide whether restrictions on liberty to prevent harm are warranted based not merely on the utility the restrictions would have, but on whether there is a ‘probable connexion’ between the exercise of liberty and the harm: in the case of speech that could lead to harm, between speaker and perpetrator of the harmful act (OL, CW 18:228n).¹⁹ When there is not – as when I publish a tract for a general audience that happens to instigate a reader to commit a crime – the perpetrator's intervening voluntary act absolves me of responsibility for the harm that results.

Mill indicates that I can be coerced only to prevent harm of which I am the morally relevant cause and not, as Lyons holds, to prevent harm regardless of who caused it also when, in laying out his harm principle, he says that we cannot restrict an individual's liberty unless they had a malicious intent: ‘the conduct from which it is desired to deter him, must be calculated to produce evil to someone else’(CW 18:224, my emphasis). Malicious intent to cause harm is necessary but not sufficient to subject one to coercion. Even if the publisher of opinions supporting tyrannicide hoped their publication would incite some reader to commit murder, and that is why they published their views, Mill still would not restrict their liberty to publish without a ‘probable connexion’ between speaker and actor. But Mill says intent is a requirement, and that supports the interpretation of the harm principle as the harm-causation principle. The prostitute is not subject to coercion not only because there is an intervening voluntary cause of any harm to a third party that results from her act, but also because she lacks the intention to injure innocent third parties.

Two other objections to Lyons’ interpretation challenge the evidence he musters to support it. First, the three omissions to which Lyons points as evidence that Mill endorses state coercion for conduct that does not itself cause harm – failing to testify at trial, save a fellow creature's life, or pay taxes – might be construed as proximately causing harm. Second, even if we disagree, Mill's harm principle could still be the harm-causation principle: owing to a further ambiguity in his principle, the interference Mill might support in these cases may fall short of the coercive exercise of power that the principle rules out. I lay out these objections in turn.

Lyons assumes that Mill's support of the use of ‘compulsion and control’ in the three cases can't be accounted for by the harm-causation principle. While there is no direct textual evidence either way regarding whether Mill regards any of these three failures to act as proximately causing harm, a plausible case can be made that they do. My failure to testify in a criminal trial could proximately cause non-consensual harm by letting a dangerous person go free. Lyons may assume that person would be an intervening voluntary cause of any future harm they inflict, just like the prostitute's client who, after receiving an STD from the prostitute, then spreads the disease. Their intervening voluntary act eliminates me as the proximate cause of the resulting harm. But when my failure to testify results in the release of a dangerous suspect, their very release could cause definite damage to assignable individuals who sought justice, or who would suffer anxiety over a looming threat the defendant on trial would pose to them if released. I am the proximate cause of these harms. My failure to pay taxes that help fund the common defence might also be said to proximately cause foreseeable harm to assignable individuals whose interests are setback by now having to shoulder an unfair share of the overall tax burden. While the harm is diffuse – a single individual's evasion of taxes may cause no perceptible damage to anyone in particular – the aggregate effects of non-compliance does constitute perceptible damage; and there is no intervening voluntary cause of that harm to others to nullify the role the tax evader plays as proximate cause. Lyons may assume the harms being prevented when the state coerces me to provide financial support for the common defence are the assaults we would suffer if we had inadequate defences, and he would be right that I am not the proximate cause of those harms, the assaulters are; but the harm Mill could have in mind is the increased tax burden everyone else faces, of which I am a proximate cause.

The hardest case to reconcile with the harm-causation principle may be that of the bad Samaritan, such as the person who doesn't attempt to rescue a drowning swimmer when they easily could. For Mill to think it justified to punish the bad Samaritan using the harm-causation principle, he would first have to think that failing to prevent the drowning itself causes the drowning, even though the drowning would have occurred if the bad Samaritan were nowhere in the vicinity. One might think, in general, that omissions or inaction cannot be the cause of harm. Mill, however, disagrees: ‘a person may cause evil to others not only by his actions but by his inaction’ (OL, CW 18:225).

But Mill would also have to think the omission is a cause of harm. Recall that for Mill, a harm is a setback to interests that are regarded as rights. A swimmer who drowns because of a strong rip current suffers misfortune. But for Mill to justify punishment of the bad Samaritan using the harm-causation principle, he would have to think that in failing to act, they set back interests that are regarded as rights and therefore harmed the swimmer; he would have to think that the drowning person had a moral or legal right not to suffer that misfortune.²⁰

Because of the ambiguity in Mill's harm principle that I discussed in section 3 regarding what constitutes a right the violation of which could be considered a harm, Mill could regard failing to rescue as violating a right, though we can't be sure if he would. In his essay ‘Comte and Positivism’ Mill says that someone who disappoints our expectations of what a moral person would do can properly be blamed: ‘inasmuch as everyone, who avails himself of the advantages of society, leads others to expect from him all such positive good offices and disinterested services as the moral improvement attained by mankind has rendered customary, he deserves moral blame if, without just cause, he disappoints that expectation. Through this principle the domain of moral duty, in an improving society, is always widening’.²¹ Disappointing such expectations might be seen as a breach of promise that sets back interests of others that are regarded as rights, thereby harming them (Berger, 1997, pp. 49–50). In On Liberty Mill says that a breach of contract can be made a ‘subject of legal punishment’ (CW 18:295). Mill could think that a legislature might ‘raise’ a promise or contract by creating a right to be rescued, just as it might create a right that others testify in court cases impacting me, or that I pay only my fair share of taxes and not more.

Yet Mill might be wary of adopting this position. Doing so could set a precedent for legislators to expand the state's authority to restrict individual liberty simply by declaring rights. The state could declare a right not to be offended or displeased. Mill, in defending individuality, forcefully objects to the ‘monstrous principle’ that would establish an expansive social right that others not act to ‘weaken and demoralize society’ (OL, CW 18:288). Mill does say that what rights there are is settled by the principle of utility (Utilitarianism, CW 10:250; cf. OL, CW 18:224), and one might think Mill would trust legislators to reject expansions of rights that threaten individuality using that principle. Yet presumably legislators enacted the Contagious Diseases Acts to promote social utility. To do so, they implicitly asserted a right of innocent third parties not to face a risk of disease, the protection of which right would justify coercing prostitutes. Mill, who opposed the Acts, could doubt that utilitarian-legislators can be trusted to adequately respect individual liberty. He could think we need the harm-causation principle's requirement that to restrict liberty not only must a legislatively-declared right be violated, but the targeted activity (or omission) must proximately cause setbacks to the interest of others that results in ‘definite damage’. Only then would individual liberty be protected against a state that enforces an unduly expansive list of rights, and not be ‘swallowed up’ by utilitarianism. If that is how Mill would resolve the ambiguity in his harm principle of whether there is a right-violation, he may well see the failure to rescue as triggering the harm-causation principle. He explicitly says one can cause evil by their inaction; and he could see my failure to rescue you from drowning as a setback to your interests that causes definite damage – a requirement not met, at least to the same degree, where I merely displease or offend you, or beat you in a competition.

Even if we don't agree Mill could think the three omissions Lyons points to proximately cause harm to others, Mill's harm principle could still be the harm-causation principle. Mill does say that one may be ‘compelled to perform’ the acts in question and be ‘made responsible to society’, and so Lyons has good reason to think Mill allows for punishment even of those who – assuming we don't accept the previous line of argument – don't cause harm. But I now argue that because of a further ambiguity in his harm principle, Mill could claim that the measures the state takes to hold people responsible for their omissions in these three cases falls short of the exercise of coercion Mill reserves only for those who proximately cause harm to others.

In addition to the two ambiguities of Mill's harm principle I discussed in section 3, there is a third: what constitutes an ‘interference with liberty’ or rightful ‘exercise of power’ against one's will that is not warranted unless it prevents harm to others? Legal punishment is the most obvious example, and Mill explicitly refers to it (OL, CW 18:292). But what about fines? Time, manner, or place regulations that merely limit the circumstances under which one may act but do not outright prohibit the activity? Refusing to subsidize or license the activity? What about forms of interference that are undertaken not authoritatively by the state, but by private individuals, such as exhortations, group interventions, or boycotts? Mill isn't entirely clear. When Mill says those who setback the interests of others may be ‘subjected either to social or to legal punishment’ (OL, CW 18:292) he has in mind punishment inflicted not only by state actors but by private individuals, acting either in isolation or in coordination with others. He gives as examples of interference, or an ‘exercise of power,’ ‘compelling’ someone to do their duty (CW 18:224) and ‘compulsory labor’ (CW 18:295), so he has in mind coercive exercises of power – interferences that force one to act in a certain way. That he means to single out ‘coercive’ exercises of power is evident also from passages where he refers to other sorts of interference which he says must be permitted – non-coercive means of persuasion such as exhortations or expressions of contempt. Mill thinks such ‘natural penalties’ are permissible means to morally improve those whose conduct we find distasteful or contemptible (CW 18:282). Not only may they be inflicted in response to self-regarding activity – activity that does not harm others – they may even be a more appropriate form of interference than an exercise of coercive power against someone who does harm others. The harm principle permits or warrants but does not require the use of coercion.

Immediately after giving his examples of omissions for which one can be ‘made responsible’, Mill adds that to be made justly accountable to society for one's inaction ‘requires a much more cautious exercise of compulsion’ than is required to respond to one's actions. ‘To make any one answerable for doing evil to others, is the rule; to make him answerable for not preventing evil, is, comparatively speaking, the exception’. Mill then says that in deciding whether the person failing to act can be held ‘justly accountable’, we need to consider ‘the special expediencies of the case: either because it is a kind of case in which he is on the whole likely to act better, when left to his own discretion […]; or because the attempt to exercise control would produce other evils, greater than those which it would prevent […]’ (CW 18:225).

Here Mill echoes Bentham's argument in Introduction to the Principles of Morals and Legislation that there are ‘cases unmeet for punishment’ where punishment is warranted but for utilitarian reasons is not implemented (Bentham, 1789, ch. 13). But that may not be Mill's main point. In recognizing degrees of responsibility depending on whether one acted or failed to act, Mill may implicitly acknowledge that there are varying degrees to which someone might be said to proximately cause resulting harm. Mill doesn't think a prostitute proximately causes harm to non-consenting third parties and so the prostitute can't be punished or subject to other coercive interference for trading in sex; but there are other cases where there is a less attenuated connection between act or inaction and result. Mill, in referring to a ‘more cautious exercise of compulsion’, may also have in mind how coercion is a scalar property, and that there may be ways of ‘compelling performance’ falling short of punishment. The ambiguity in the terms ‘interference with liberty’ and ‘exercise of power’ in his harm principle provides Mill some leeway so that even if he thought that, like prostitutes, bad Samaritans, tax evaders, or those failing to testify in court did not proximately cause harm, in saying they could be ‘compelled to perform’ he could be referring to means of compelling that fell short of punishment, such as fines, or the exhortations and other natural penalties he allows even for self-regarding conduct that does not proximately cause harm to others. More likely, given that Mill says that a person may ‘cause’ evil even by inaction (CW 18:225), he could think that they proximately cause harm at least to some degree, which could support ways of ‘compelling to perform’ that may even include punishment. In either case, we needn't follow Lyons in rejecting the harm-causation interpretation of Mill's harm principle to explain Mill's willingness to hold these individuals accountable for their omissions.

5. Mill's Objection to Licensing Prostitution

Lyons’ view that Mill is willing to exercise coercion upon an individual to prevent or reduce harm even if that individual is not the morally relevant cause of harm is contradicted by Mill's claim that prostitution must be legal because the prostitute does not proximately cause harm to non-consenting third parties. In this section I present a further objection to Lyons’ reading of the harm principle. As noted in section 1, one of Mill's major objections to the Acts is that by in effect licensing prostitutes, the state legitimizes their conduct.²² Mill's objection is puzzling to those who see his political philosophy as centrally concerned with harm reduction (Waldron, 2007). By turning to his reasons for opposing licensing, we see that Mill is more concerned with promoting moral progress – subject to the constraints imposed by the principle of liberty – than he is with reducing harm.

Mill does not think the state should prohibit prostitution, because prostitutes don't proximately cause harm. But to license is not to outright prohibit. Mill thinks individuals should be free to engage in risky self-regarding behaviour without state meddling, but not necessarily at liberty to engage in commerce with each other free from state regulations that could ensure the transactions are safe. In On Liberty Mill says that ‘trade is a social act’ that ‘affects the interest of other persons’ and therefore comes under the jurisdiction of society (CW 18:293). Mill opposes regulations restricting a buyer's ability to purchase goods and services for their self-regarding aims (CW 18:288); but he allows for regulations of sellers. The state can't restrict my liberty to buy poisons for self-regarding purposes, for example, but it can regulate sellers of poison:

To require [of a buyer of poisons] in all cases the certificate of a medical practitioner, would make it sometimes impossible, always expensive, to obtain the article for legitimate uses. The only mode apparent to me, in which difficulties may be thrown in the way of crime committed through this means, without any infringement, worth taking into account, upon the liberty of those who desire the poisonous substance for other purposes, consists in providing what, in the apt language of Bentham, is called ‘pre-appointed evidence’. (CW 18:294)

Mill then explains: sellers can be required to document purchases to deter crimes involving poisons, or to help catch a criminal after the fact (CW 18:295). Because the exchange of sexual services for money is a ‘social act’, then on this reasoning the state should be permitted to regulate its sale as well, without putting up a barrier for the buyer. So why does Mill oppose licensing of prostitutes, which would not create an outright barrier to purchasing sex?

Mill's main objection is that licensing prostitutes will legitimize prostitution. He says this repeatedly.²³ Mill distinguishes ‘attacking evils [such as STD transmission] when they occur, in order to remedy them’, from ‘making arrangements beforehand which will enable the objectionable practices to be carried on without incurring the danger of the evil’ (CW 21:358). He opposes the latter because he does not think the state should ‘enable’ or condone the morally objectionable practice: ‘I do not think that prostitution should be classed and recognized as such by the State’ (CW 21:359); he opposes ‘toleration of that kind of vicious indulgence’ (CW 21:356). By having hospitals devoted to prostitutes, the State would be going out of its way to facilitate prostitution, which would legitimize the practice (CW 21:354).

To be sure, Mill gives apparently prudential reasons for not wanting to legitimize prostitution. If prostitution is made safer it will be encouraged (CW 21:355), increasing the demand for prostitutes and in turn the supply (CW 21:364). If we refuse to condone prostitution, we'd impress on people that it is immoral, and there may be fewer prostitutes on the streets (CW 21:368).

But in wanting to reduce even safe prostitution, Mill shows that his overriding concern is not harm-reduction: it is to discourage immorality, or ‘moral injury’ (CW 21:371). Nor is his main concern, as some have suggested, a feminist opposition to male exploitation of women. Mill objects even to safe prostitution but not because he thinks women are forced into prostitution; in his testimony before the Commission Mill says that women ‘voluntarily’ choose to be prostitutes. He objects, rather, because the life of prostitution that they choose is ‘degrading’ (CW 21:368). Members of the Commission who favoured the Acts because they would reduce sexual disease were clearly irked by Mill's opposition: they asked Mill if he is really fine letting women come out and spread disease right and left, or leaving them to ‘rot and die’ rather than save them with the Acts (CW 21:365, 366). First Mill replies that the question is unfair; anyone suffering a wretched disease can be laid hold of and given proper medical treatment (CW 21:366–67). He then sticks to his objection, concluding his testimony by reiterating that we should not make ‘safer than it would naturally be’ a ‘course which is generally considered worthy of disapprobation’, for if we did, it would not be ‘considered very bad by the law, and possibly may be considered as either not bad at all, or at any rate a necessary evil’ (CW 21:371). Mill opposes the Acts because they would undermine a commitment to moral progress he thinks the state should pursue, a commitment that is grounded in his distinct theory of utilitarianism. While one might think that a utilitarian should support the Acts because they surely would reduce harm, Mill's utilitarianism seeks not harm reduction, but moral improvement. Mill thinks we should seek the ‘higher pleasures’ enjoyed by the ‘cultivated mind’.²⁴ It is not the quantity but the overall quality of lives lived that is to be promoted.

One might object to Mill's position: it is more important to prevent harm than discourage immorality. Mill's position is at odds, for example, with government programs that provide drug addicts with sterile needles and tools to check that the illegal drugs they take are not laced with lethal substances, on the ground that it is more important to reduce harm than morally condemn drug use.²⁵ One might also challenge Mill's assumption that by licensing an activity the government necessarily expresses approval of it. But regardless of whether we agree with Mill, his position on licensing indicates that Mill prioritizes not only individuality but moral progress over harm reduction.

6. Mill on Public Solicitation

Mill opposes criminalizing prostitution based on his defence of individuality: individuals should be free to make and pursue their own choices of how to live so long as they don't proximately cause non-consensual harm to others. While the state must respect individuality by adhering to the harm principle, Mill also thinks the state should promote moral progress. This is why he opposes the licensing of prostitutes even though licensing would reduce harm.

One puzzle remains concerning Mill's position on prostitution. As I noted in section 1, Mill apparently supports public solicitation laws – laws that prohibit prostitutes from advertising their services in public places. In his testimony on the Acts, Mill says the police have a duty to ‘prevent solicitation in the streets’, ‘in order to preserve the order of the streets’ (CW 21:369).²⁶ This might seem to contradict On Liberty's defence of the principle of liberty: if prostitution in private does not proximately cause harm and warrant coercive state interference, why would its solicitation in public? But in a notoriously cryptic passage in On Liberty Mill opens the door to restrictions of normally self-regarding acts when done in public, if they are ‘indecent’. Mill writes:

Again, there are many acts which, being directly injurious only to the agents themselves, ought not to be legally interdicted, but which, if done publicly, are a violation of good manners, and coming thus within the category of offences against others, may rightfully be prohibited. Of this kind are offences against decency […]. (CW 18:295–6)

In addition to solicitation Mill could have in mind acts such as sex in a public place, or offensive displays akin to the displays of swastikas in a neo-Nazi march.²⁷ Targeting offenses against decency sounds like the very legal moralism Mill explicitly disavows in saying that power cannot be exercised against someone's will except to prevent them from non-consensually harming others. If prostitutes don't non-consensually harm others, isn't it inconsistent for Mill to think that public solicitation may be prohibited? (Wolff, 1998, p. 4).

We could just dismiss the passage.²⁸ But that would be a mistake: we shouldn't ignore Mill's commitment to moral progress – it is a central component of his political philosophy. But how can we reconcile it with Mill's defence of the harm principle?

Riley finds no inconsistency, by arguing that public indecencies cause harm, and while I agree that we can reconcile Mill's position on public solicitation and other indecencies with his harm principle, I would take a different route to do so. Some of Riley's examples of indecencies involve threats to public health – public urination, defecation, vomiting, sneezing (Riley, 2015b, p. 275) – and I agree these present no conflict with the harm principle as these actions could foreseeably cause harm. In the case of nuisances most of us would regard as non-harmful, such as public sex, Riley presents what seems to me an unconvincing argument: they cause perceptible damage by crowding out higher priority uses of public places (276) – unconvincing because failure to maximize efficient use of public resources violates no right of an assignable individual not to suffer definite damage or perceptible hurt. Riley suggests that such public indecencies can disappoint ‘legitimate expectations’ that emanate from laws and customs, and thereby deserve ‘moral blame’, and that ‘deliberative majorities’ may reasonably declare them as wrongful (274–75). In that case, leaving aside the harm principle's requirement that there be ‘perceptible hurt’ and ‘definite damage’, the public indecency would setback interests that are regarded as rights, meeting a key criterion for causing harm. But as I noted in section 4, that approach risks swallowing up the harm principle into utilitarianism, as legislators could simply declare rights not to be displeased or offended.

There is another way to resolve the apparent inconsistency. When in his testimony on the Acts Mill agrees that the police have a duty to prevent public solicitation, or in ‘On Liberty’ he says public indecencies can be rightfully prohibited, he doesn't clarify what measures the police may take to preserve the public order. This calls to mind the ambiguity in his harm principle that I introduced in section 4: what constitutes an ‘interference with liberty’ or rightful ‘exercise of power’ that is warranted only to prevent harm to others? Mill's testimony would conflict with his harm-causation principle if he means the state may forcibly detain and punish prostitutes, because prostitutes, on his view, don't proximately cause harm to others. But if the ‘rightful prohibition’ of indecent public activity that Mill allows for in ‘On Liberty’ is a time-manner-place regulation, it could be distinguished from the coercive restriction of liberty that the harm principle rules out, inasmuch as the activity is still permitted in private.²⁹ When in his testimony Mill agrees that police may prevent solicitation ‘in the streets’, he may have in mind a ‘place’ regulation similar to zoning laws that restrict the location of bars and adult entertainment clubs.

7. Conclusion

Mill believes the state may not punish prostitutes because prostitutes do not proximately cause non-consensual harm, and we shouldn't coerce people merely because we think they are acting immorally; yet he opposes state licensing of prostitutes, which could reduce harm, because he does not want to legitimize an immoral practice. While Mill defends his harm principle because it protects individuality, his defence of non-neutral state policies that promote moral progress, such as refusing to license prostitution, can potentially be more of a threat to individuality than Mill allows for. Consider laws that recognize marriages between a man and woman but not between same-sex couples, and that deny important benefits to non-married partners. One might argue that these laws do not restrict liberty in the way a law prohibiting homosexual sex would, because being denied tax benefits, hospital visitation rights, or countless other benefits is not the same as having one's liberty curtailed: liberty is freedom from hindrance and physical restraint, not entitlement to government support.³⁰ But this argument fails to recognize that when the state refuses to recognize a marriage it inflicts dignitary wounds upon, stigmatizes, and demeans same-sex couples, and can injure or harm their children.³¹ Mill's commitment to moral progress not only can risk increasing the amount of harm in the world; it can also sometimes threaten the very individuality Mill wants to protect. The threat the pursuit of moral progress poses to individuality is apparent when we turn to passages in which Mill supports non-coercive interference both by the state and individuals to promote moral progress, interference that is permitted by his harm principle because it falls short of ‘exerting power’ or ‘compulsion’. Such natural penalties – where I shun or voice displeasure or contempt to you because I disapprove of your self-regarding activities – can promote individuality by being a means of exercising our freedom of expression and association.³² But they can also stifle individuality. Mill is well aware of this and sets limits on the exertion of social pressures: we may avoid the offending person but we may not parade our avoidance (OL, CW 18:278); we may privately warn our mutual friends about him, or deny him the ‘perks of affection’;³³ but perhaps not organize boycotts.³⁴ But by leaving his harm principle ambiguous as to where in the range of the scalar property of coercion an exercise of power becomes illegitimate, Mill risks justifying forms of interference that may compromise his commitment to individuality and liberty.³⁵

Mill's discussion of prostitution and the Acts may not leave us with an entirely satisfactory position, but it is significant. It strikingly illustrates how both his concern for individuality and his utilitarian-grounded concern for moral progress prevail over the goal of harm reduction. We miss this significance if we construe the harm principle as a harm-prevention principle.³⁶