1 Introduction

Let us begin with a thought experiment:

Imagine a country town that experiences two concurrent horrors: A flooded stream that causes the drowning death of one young woman, and the murder of another young woman by an unknown assailant. Further, imagine that both deaths occurred in the same parklands and that the murderer appears to have been unknown to his victim. In the opinion of the police, the flooded stream will continue to pose a risk to human life, while the offender will continue to pose a risk to women.

Is it morally justifiable for police to provide safety advice to address one or both risks?

Surely, most people would regard warning the public about the flooded stream as not merely justified, but as morally required. After all, the job of police is to keep people safe, and this is best achieved by preventing them from entering harm’s way. However, reactions to homicide prevention advice will be more mixed. Some people will regard the police responsibility to warn against an ongoing threat as self-evident. Yet others will worry that the warning engages in the morally problematic practice of ‘victim blaming’.

1.1 What is Victim Blaming?

Victim blaming is the practice of holding people responsible when bad things happen to them. Psychologically, it is associated with belief in a ‘just world’, particularly with the idea that the world is just to other people, not merely to oneself (Hayes et al. 2013). As Double (2005, 22) observes:

“Persons who are cognitively or emotionally invested in the belief that the world is just are unlikely to think that bad things happen to persons who do not deserve them.”

According to defensive attribution theory, victim blaming fosters feelings of safety by highlighting differences between the victim blamer and the victimised (Pinciotti and Orcutt 2020). A woman, for instance, might comfort herself that she is unlikely to be raped because she avoids dressing provocatively or consuming alcohol in public. Likewise, a man who does not smoke may find reassurance in the knowledge that a man with cancer had a long-term smoking habit.

Victim blaming presents in many contexts, including in attitudes towards the disabled (Shakespeare et al. 2017), the overweight (Adler and Stewart 2009), the low-paid (Jason and Turgeon 2021), and disadvantaged racial groups (Ryan 1972). Nevertheless, it is often regarded as particularly prevalent (and morally problematic) in respect to violence against women, where victim blaming attitudes and comments “shift blame from the perpetrator to the victim” (Bieneck and Krahé 2011, 1786), harming women by encouraging self-blame and self-curtailed liberty (Matthews 2014).

1.2 Police Officers and Victim Blaming

In the course of their duties, it is common for police officers to issue crime prevention and personal safety advice. This advice can come in many forms, including:

  • Brief public statements on personal safety, e.g., social media posts and ‘ten second grabs’ on the evening news.

  • Printed materials, such as crime prevention flyers and brochures.

  • Public safety talks at community events.

Given the public nature of such advice, it is perhaps not surprising that police officers occasionally attract accusations of victim blaming, both from academic and non-academic sources. Sometimes, the advice issued is truly appalling, as in the following comments delivered by a Canadian constable at a safety forum (Pilkington 2011):

“I’ve been told I’m not supposed to say this – however, women should avoid dressing like sluts in order not to be victimised.”

Less egregious, though still ill-considered, advice includes the following public comment from a detective inspector in the Victoria Police:

“I suggest to people, particularly females, they shouldn’t be alone in parks…I’m sorry to say that is the case.” (Calligeros 2015).

This comment followed the stabbing murder of a 17-year-old woman, who had been walking alone through a Melbourne park at 6.50 pm in March 2015, still within daylight hours.

Taken as general advice for those living within the relative safety of a major Australian city, rather than as localised, short-term advice offered in the face of a current threat, this prescription was quite extreme. Indeed, it may have been influenced by the availability heuristic, a cognitive bias associating the perceived likelihood of an event with “the ease with which instances come to mind” (Kahneman 2011, 129). Murders of this kind are rare, but the emotional response had by most people (including investigating police) may give them a salience that lends itself to a disproportionate response.

Nevertheless, not all putative victim blaming comments are as extreme or even insulting as these examples. Consider the following advice issued by police in the Bahamas in May 2022:

“Females are advised to exercise the necessary caution while exercising and to be aware of their surroundings when arriving at home. On arrival at home, a thorough search of the dwelling should be conducted and ensure that all doors and windows are properly secured.” (Bowleg 2022 A; Bowleg 2022B).

This advice was issued following the sexual assault of a female jogger by a stranger in the area where the announcement was made, and while the offender was still at large. Despite the criticism, there does seem to be a difference between police warning potential victims about an actual current threat (an at-large offender who has already committed a ‘stranger’ sexual offence) and a non-specific potential threat.

1.3 Are Victims Being Blamed?

Of course, there are legitimate questions over whether some putative victim blaming comments actually blame victims at all. Consider the following controversial social media post from Nottinghamshire Police:

“Taking a risk when it comes to walking alone at night is not one of those things we should be doing. Women who walk alone especially at night are at risk of harassment, or even physical assault.” (Halliday 2019).

Does this message imply that a woman who does walk alone at night is morally blameworthy if she is harassed or assaulted?

As Curchin (2019, 2–3) rightly observes, “everyday people (by which I mean people who are not moral philosophers) are capable of comprehending the distinction between causal contribution and moral responsibility.” They are also, presumably, capable of distinguishing between the notions of ‘should do X’ and ‘blameworthy if one does not do X’. Indeed, if causal contribution and imperative-framed advice do imply moral responsibility, then the same victim blaming conclusion must be drawn about the following common advice:

  • Car owners should not leave valuables on display or car doors unlocked.

  • Holidaying homeowners should have neighbours collect their mail and take in their garbage bins.

  • The elderly should carefully scrutinise the identification of unsolicited tradespeople.

  • Children should be wary of strangers.

  • Everyone should practice online security by not sharing passwords and bank details.

In contrast to women’s safety advice, these recommendations are usually accepted without complaint.

Before proceeding further, a couple of caveats must be issued.

Firstly, the question explored in this paper is whether prevention advice directed at potential victims has a morally justifiable place in the police response to violence against women, not whether it should be the sole response to such violence (or even the sole preventative initiative). It is assumed that the police response should extend far beyond issuing advice. Secondly, this paper is not concerned with comments that are blatantly sexist (such as those describing women as ‘sluts’) or that presuppose that women should occupy traditional social roles. Likewise, it is not concerned with advice that is impractical, extreme or a disproportionate response to an actual threat (such as broad-brush comments advising women to avoid parks). It will be taken for granted that comments such as these have no place in policing.

Instead, the paper is interested in what I will term ‘sound advice’, by which I mean advice that is proportionate, easy to follow, empirically justified, and objectively likely to reduce harm. True, empirical justification and objectivity may be context dependent. Advising women to take a taxi home at night, for example, can be jarring when they have also been advised to be on their guard against sexually predatory taxi and Uber drivers (or people impersonating them). This risks undermining the advice, but it also suggests (i) that police should ensure any advice is current and context specific, and (ii) that they should (in a broad sense) let the public into their confidence.

Regardless, this definition of ‘sound advice’ is very much a working definition, not a definitive set of necessary or sufficient conditions. I hope it suffices, however, to convey the kinds of advice I do (and do not) have in mind. At the end of the article, I will make some recommendation as to the form this advice should ideally take.

2 Walking in the Park: Agent vs. Non-Agent Risks

As we have seen, women’s safety advice bears many similarities with less controversial crime prevention advice, including advice addressing property crime. This may tempt us to reason along the following lines (I will call this the relevant similarity argument):

  1. 1.

    Personal safety advice for women is relevantly similar to other crime prevention advice.

  2. 2.

    Other crime prevention advice is morally justified.

Therefore,

  1. 3.

    Personal safety advice for women is justified.

However, while other crime prevention advice usually does not attract accusations of victim blaming, there are occasional exceptions. An internet search, for instance, identifies victim blaming concerns relating to identity theft (McGarvey 2021), bicycle theft (Hanna 2021) and cyclist safety (Giddings 2015).

Indeed, we might think that we should be concerned about all crime prevention advice. After all, research suggests that robbery victims can be the subject of victim blaming attitudes too (albeit to a lesser extent than rape victims; Bieneck and Krahé 2011), and property crime victims may experience greater feelings of guilt than victims of violent crime (Smale and Spickenheuer 1979). Furthermore, blame attribution can “impact significantly on the well-being of online fraud victims through their inability to disclose to family and friends and seek support as well as through the isolation of victims from support networks” (Cross 2015, 201).

There is even the danger that general crime prevention advice may exacerbate existing vulnerabilities, a concern sometimes raised in respect to the blaming of women victims (Matthews 2014). Parking in well-lit areas is easier in some neighbourhoods than others, for instance, while buying a home alarm or CCTV system is easier for the rich than the poor. In short, many of the negatives associated with gendered violence prevention advice also seem to be associated with more run-of the-mill advice. This suggests that the conclusion of the relevant similarity argument may be wrong precisely because premise two is false: Other crime prevention advice may not be morally justified. Consequently, we will now compare women’s safety advice with advice designed to protect against non-human threats.

2.1 Floodwaters

In many parts of the world, driving over flooded roads and bridges is a danger that is regularly warned against during severe weather. In early 2021, for example, the following comments were made by an Australian politician with ministerial responsibility for emergency services, following the drowning of a man:

“The human cost of these floods has been brought into sharp focus in the most tragic of circumstances and I urge communities to continue to be cautious in the face of continued and imminent threats to life.” (ABC News 2021).

These comments appear to be uncontroversial, and, to the present author’s knowledge, they did not result in any controversy. Importantly, the comments did two things: They acknowledged the tragic circumstances with its “human cost”, and they urged caution “in the face of continued and imminent threats”.

Nevertheless, we would expect more controversy if, instead of commenting on a flooding death, the minister had made the following comments in response to the murder of a woman: “The murder of Miss Smith is tragic, and I urge women in the community to be cautious in the face of continued and imminent threats.” The thought experiment with which this paper opened essentially combines the drowning and homicide scenarios and places them in the same parklands. However, are the two cases relevantly similar? If so, then an argument could be presented in the following form (I will call this the second relevant similarity argument):

  1. 1.

    If actions X and Y are relevantly similar, then action Y is morally permissible if and only if action X is morally permissible.

  2. 2.

    Providing natural disaster prevention advice (X) is relevantly similar to providing homicide prevention advice (Y).

Therefore,

  1. 3.

    Providing homicide prevention advice is morally permissible if and only if providing natural disaster prevention advice is morally permissible.

  2. 4.

    Providing natural disaster prevention advice is morally permissible.

Therefore,

  1. 5.

    Providing homicide prevention advice is morally permissible.

If we accept that it is morally justifiable for police to provide warnings in the case of natural disasters, then the key issue is whether the two cases are relevantly similar.

The following similarities can be readily identified:

  • Both involve current and localised threats.

  • Both involve threats to human life and have recently claimed a human life.

Yet there are also the following dissimilarities:

  • In one case, the warning applies to all people, in the other it is relevant only to a subgroup (women), who have a history of experiencing sexism, inequality, and internalised blame (Matthews 2014).

  • While the ultimate danger in both cases (death) is the same, the cause is very different. In one case it is a natural phenomenon, while in the other it is a moral agent.

Let us consider both dissimilarities in more detail.

2.2 Inequality and Self-Blame

Firstly, does the fact that one of the threats disproportionately harms half the population constitute a relevant difference? Effectively, this is the same as asking whether water safety advice would be unjustifiable if, say, drowning disproportionately affected the female part of the population. Let us consider this. Imagine a possible world that is broadly identical to our own except for biological differences that make women substantially more prone to drowning than men. Intuitively, these differences would not compel us to withhold water safety advice, even if additional measures (such as rescue patrols around waterways) were warranted.

Secondly, does the fact that advice regarding murder, unlike flooding, exacerbates existing inequality by impacting on a broadly disadvantaged group constitute a relevant difference? Probably not. After all, safety advice regarding acts of nature often disproportionately affects disadvantaged groups. Not walking through a park at night, for example, even if the only danger is a swollen stream, may disproportionately affect people who work nights and do not own cars. Likewise, advice to ‘avoid nonessential work’ during coronavirus outbreaks may be easier to follow for the rich than the poor. Consequently, there is the danger that any personal safety advice may exacerbate existing inequalities.

Thirdly, do feelings of self-blame prompted by crime prevention advice (regardless of the intention of the messaging) constitute a relevant difference? Surely not. Based on media reports, people rescued from the ocean often feel intense embarrassment, yet this would not justify withholding safety advice. Souter (2017), for example, described the reaction of her friend to nearly drowning:

“… she’s also feeling embarrassed. A hot line of shame playing out above the rising panic. Lifesavers will later tell me this reaction is common, and it’s one of the reasons some people – especially teenage boys – wait too long to raise an arm for help. They feel stupid for getting into this situation; they’d like to get out of it without anyone knowing” (Souter 2017).

Consequently, we should be very reluctant to conclude that feelings of personal responsibility or shame enable us to differentiate between crime prevention advice and other forms of personal safety advice.

2.3 Agency

A moral agent is responsible for its actions, while non-agents are not responsible for theirs. While there may be non-human moral agents (Clement 2013), and even non-biological ones (Coeckelbergh 2009; Fossa 2018), agency is normally seen to require an element of intentionality or rationality (Clement 2013). Consequently, our homicide offender is a moral agent, while our flooded stream is merely an intention-less, impersonal occurrence – a non-agent.

According to McPherson (1984, 173), “The notion of the moral agent makes no sense in total isolation from that of the [moral] patient.” He was referring to the fact that agency is only significant if there is someone for the agent to harm or fail to meet its obligations towards. Importantly, McPherson’s point also works in reverse: In certain important respects, moral patients can only be harmed by moral agents. No storm, earthquake or volcanic eruption is powerful enough to violate my rights, yet the most ordinary of people can violate them with ease.

Indeed, we might think that a rights violation constitutes a harm over-and-above the more direct harms caused by an agent (Douglas 2021, 147). It certainly seems worse, for example, to be pushed off a cliff than to lose one’s balance and fall, even if the resultant physical and mental suffering is the same. We might also feel that a killing committed by an ordinary rational adult is objectively worse than an otherwise indistinguishable killing performed by an automaton. If so, then while the drowning and the homicide in our parkland scenarios cause the same mortal harm, the homicide would be worse because it also involves a rights violation.

But what is the significance of this for whether police are justified in providing advice?

The most straightforward conclusion is the following:

If death at the hands of a non-agent is sufficiently harmful to justify the provision of advice X, then the provision of advice X should likewise be justified to prevent an objectively worse situation (death plus a rights violation).

This approach is consistent with the priorities of modern policing. While it is certainly the case that police are expected to save people from accidents and natural disasters, their most emblematic function is to deal with criminal offences, which generally violate the rights of citizens.

It could, of course, be argued that an offender’s moral agency is relevant because it implies that the agent could have done otherwiseFootnote 1, which is not something true of a flooded stream. Is this aspect of agency a relevant difference between the two scenarios? I think not. As with the additional harm of a rights violation, if this property of could-have-done-otherwiseness makes the situation objectively morally worse, then it presumably provides an additional reason for preventing it.

Nevertheless, it may be responded that while the could-have-done-otherwise element of the agent’s action does not lessen our obligation to prevent it, it does open up alternative methods of prevention. It might be thought, for example, that campaigns focusing on would-be offenders may be more suitable. Yet this seems unsatisfactory for two reasons. Firstly, there are good reasons for doubting whether alternative interventions of this kind would be successful, or even successful enough to make potential victims materially safer. Even intensive sex offender rehabilitation programmes have at best mixed success (Dennis et al. 2012; Lösel et al. 2020; Ho 2015), so implied solutions in slogans such as “Stop Telling Me – Don’t Get Raped. Tell Men – Don’t Rape” (Orbach 2012, 209) may be overly optimistic. Regardless, many offender-directed interventions are engaged in by police (I provide a brief sketch of these in Sect. 2.4, below). Secondly, alternative interventions could also have been used to prevent the stream from flooding, including building up the riverbank, branching the stream or building a dam. Consequently, it is not clear that this is a difference (let alone a relevant one) at all.

2.4 Expressivist Objection

But perhaps the problem is that issuing homicide prevention advice effectively ignores the agency of the offender. After all, we do not hold floods blameworthy for drownings, so perhaps issuing homicide prevention advice implicitly expresses (or could reasonably be interpreted as expressing) the judgement that a murderer likewise bears little or no blameworthiness. Indeed, expressing attitudes that diminish the culpability of offenders may even make advice givers somewhat complicit in the offending.

Expressivist objections of this kind are common in areas such as reproductive ethics, where, for example, it is sometimes argued that rejecting an embryo based on genetic markers for disability is wrong because it “expresses a negative valuation” of people living with disabilities (Hofmann 2017, 505). Plausibly, any harms caused by this expression must be balanced against other values such as reproductive autonomy (Edwards 2004), yet it could also be contended that the obligation not to express this harmful view trumps all other obligations. In our case, we might think that our obligation to avoid expressing equivalence between a murderer and a non-blameworthy flood trumps our pro tanto duty to provide prevention advice.

Nevertheless, it is far from obvious that personal safety advice does express this negative view, or, if it does, that avoiding it must trump other considerations. After all, there are many other similarities in the police response to floods and homicides (including patrols of danger areas, prompt response to emergency calls, and first aid for survivors) that are not thought to express moral equivalence. Furthermore, there are also pertinent dissimilarities: Floods are never arrested, charged, and sentenced to lengthy periods of imprisonment. Surely, any purported implicit messages identified in prevention advice should be interpreted in light of explicit offender blaming actions such as these.

Indeed, any expressivist interpretation of police issued personal safety advice should be made considering all preventative actions and messaging employed by police, not merely those directed at potential victims. This includes explicit messaging aimed at actual or potential offenders (such as sexual consent campaigns and public calls for at-large offenders to hand themselves in), various programmes in which police work with at-risk youths to set them up for respectful and law-abiding lives, and high visibility anti-‘violence against women’ initiatives, such as the longstanding police participation in White Ribbon Day events (9NEWS 2013).

2.5 Widespread Police Sexism?

None of this denies that there have been many prominent cases of police officers making outrageous and unjustified claims about women (as quoted in this paper). Some officers have even been convicted of horrendous crimes of violence against women, including the meticulously planned kidnapping, rape, and murder of a young woman in the United Kingdom in 2021 (Morton 2021). Even the current focus of many police agencies on inclusivity and diversity may be seen as an acknowledgement of historical or even current deficiencies (McLeod 2018). Consequently, it is fair to question whether these cases of sexism and violence are the result of bad individuals or bad culture (or, perhaps, some combination). Instances of police sexism and criminality certainly muddy the water, and perhaps it is understandable that safety advice is sometimes interpreted through the lens of these cases, rather than through the lens of the diverse range of women’s safety initiatives promoted by modern police agencies.

Ultimately, the truth regarding police sexism is probably nuanced, with sexism varying across time and place, both within and between police agencies. However, rather than weigh up the evidence for police sexism, let us assume for the time being that it is widespread. Does this constitute a relevant difference between the murder and drowning scenarios?

I argue that even if we assume that police forces are bastions for sexism, it still does not follow that police officers should avoid issuing personal safety advice for crimes that disproportionately affect women, provided that the advice meets our criteria for being sound. Of course, even sound advice may be issued for sexist reasons. A sexist police officer, for instance, may issue sound advice to women to obscure the fact he is otherwise ignoring violence against them. Alternatively, he may see in sound advice confirmation of his own sexist worldview and share it for that reason. An example would be a police officer providing women with genuinely reasonable and timely advice about drink spiking in local bars in the (unstated) hope that they will decide to stay home and cook dinner for their husbands.

Consequently, it is possible to distinguish between several scenarios:

  1. 1.

    Police officers issuing sexist advice for sexist reasonsFootnote 2.

  2. 2.

    Police officers issuing sound advice for sexist reasons.

  3. 3.

    Police officers issuing sound advice for non-sexist reasons.

We are interested in 2 and 3. The challenge, however, is that for a casual observer (a category that captures most recipients of police advice) it can be difficult to distinguish between the two.

Consider the following:

Imagine two (almost identical) situations. In both cases, there is a genuine threat to women in a particular neighbourhood, and in both cases a local police officer issues advice X, which meets this article’s criteria for being sound. The difference is that in one case the police officer is motivated by a genuine desire to prevent further offending against women, while in the other case he is genuinely sexist and only issues the advice because (i) he feels it is the minimum he can get away with doing, or (ii) it is also consistent with his sexist worldview.

I will take it for granted that we would form very different assessments of the police officer’s character in these scenarios. But does this character assessment affect the objective rightness or wrongness of issuing the advice? True, as Markovits (2010, 203) argues:

“When we do the right thing because it happens to suit us, or happens to be in our interest, our action has no moral worth. This is intuitive. Morally worthy actions must be performed for the right (motivating) reasons.”

Nevertheless, they are still right actions. Markovits also notes that “This is not to say that only virtuous people can perform worthy actions – it is possible to act, in this sense, out of character” (203).

Indeed, our whole criminal justice system is designed to get people to act out of character. We clearly would not esteem a man who only refrains from murder because it will result in his arrest and imprisonment. Nevertheless, it is undeniably good that this immoral man mirrors the actions of his moral compatriots, just as it is good that an immoral police officer mirrors his moral colleagues. Our sexist police officer has the wrong motivation, and his belief that he has fulfilled his minimum obligations may be mistaken, yet his actions are still, objectively, right.

As the dissuaded murderer example highlights, right acts performed for wrong reasons can appear in many scenarios. Just as one police officer may pursue a rapist to serve justice and prevent future rapes, another may only do so to preserve his own career. It would even be mistaken to assume that police sexism can only factor in decisions to issue advice. Instead, sound advice may, conceivably, be withheld for sexist reasons. For instance, sexist police could reason (consciously or subconsciously) that “women are unimportant, so let’s not waste any effort on keeping them safe”, or “some women deserve to be raped, so why stop them from getting what they deserve?”

Consequently, we have two additional scenarios:

  1. 4.

    Police officers withholding sound advice for sexist reasons.

  2. 5.

    Police officers withholding sound advice for non-sexist reasons.

An example of 5 may be police officers withholding advice they know to be sound to avoid being labelled victim blamers, or simply because they (rightly or wrongly, as the case may be) choose to prioritise other ways of keeping women safe.

This leads us to a major challenge for the assumption that police sexism counts against the justifiability of prevention advice: If the absence of personal safety advice may be as indicative of sexism as its presence, then the presence or absence of such advice ceases to be a useful marker. True, additional investigation (or access to the minds of police officers) may reveal the motivation in particular cases, but it is not clear that even this insight would make a material difference to whether the advice should be issued. Any good action performed by police officers may be done for the wrong reason, and it would be very brave of us to allow police officers to do the right thing only when we are convinced by the rightness of their motives. Indeed, while it is undeniable that any remaining sexism should be expunged from policing, delivering sound personal safety advice may be the right thing to do even when performed by sexist police officers.

2.6 Widespread Societal Sexism?

However, perhaps the problem is not police sexism, as such, but societal sexism. After all, police are not acting in a vacuum, and the long history of state-sponsored gender discrimination might be the relevant difference between advice addressing gendered violence and other safety advice. Indeed, discrimination is not simply a matter of history. Despite all the improvements of recent decades, many current legal systems (for example) are perceived to be weighted against women, especially those who are victims of sexual violence.

In this context, personal safety advice may be thought to exacerbate or entrench existing discrimination, perhaps because the content of the advice is discriminatory, either directly (by recommending that women adopt different practices than men), or indirectly. The latter may occur when advice disproportionately impacts women, even though they are not discriminated against qua women (Lippert-Rasmussen 2006, p. 170). For example, ostensibly gender-neutral advice (such as “people should not walk alone in parks”) may be indirectly discriminatory if (i) it is more likely to be applicable to women than men (for example, if they are objectively at greater risk than men), (ii) it has a greater impact on women than men (for example, if they are more likely to withdraw from social life), or (iii) women are more likely to be blamed for not following it than men (for example, if male victims are viewed as unlucky while female victims are criticised). In these cases, women are disadvantaged even though the advice itself does not single them out.

Effectively, this advice may limit a woman’s ‘deliberative freedom’, i.e., her freedom to have her life-choices “insulated” from morally irrelevant considerations such as her gender (Moreau 2010, p. 147). Consequently, we might think that the role of police officers is not to train women to operate within the constraints of an unjust and discriminatory society, but to insulate them against the need to do so, perhaps by providing a service that ensures women can safely make the same choices as men.

Of course, as we will see below, it may be that safety advice actually enables women to make choices by providing them with options that might otherwise have appeared closed to them. Regardless, while the ideal situation would clearly be to provide this insulation (or even to eliminate the underlying injustice), achieving this in the short-term seems highly unlikely, even with no deficit of good will. Consequently, we are left with a choice between providing personal safety advice (among a suite of police actions) and allowing preventable harms to occur. In this case, it is surely preferable to provide the advice, provided, of course, that it meets our criteria for soundness.

2.7 A Misguided Response

However, it might be thought that even if safety advice is designed to rectify an underlying injustice, it is fundamentally misguided in the way it goes about it, as the following empirical facts about violence in England and Wales imply (Office for National Statistics 2019):

  1. 1.

    Men are typically more likely to experience violent crime and to be victimised by strangers than women. Consequently, it might be considered unjustified to direct most advice at women.

  2. 2.

    Women are much more likely to be harmed by someone known to them than by a stranger. Perhaps, in focusing on stranger attacks, police are (i) increasing unjustified fear about stranger attacks at the same time that they are (ii) avoiding the most likely threat.

In response to the first concern:

It certainly appears that women are subjected to more unsolicited advice from police than men, although both the perception and the reality may be influenced by media coverage. If the media do not provide coverage of an attack on a man, then that coverage cannot be accompanied by advice from the police. Likewise, police advice to women frequently attracts accusations of victim blaming, which gives the original advice increased airtime. These factors may create the perception that police are more concerned about public stranger violence against women than men.

Of course, it is also possible that police are more motivated to prevent offending against women than men. If so, perhaps this is an acknowledgement of the longstanding discrimination and sexism faced by women, rather than merely an unthinking reflection of it. Regardless, there may be merit in police increasing their preventative efforts against crimes with male victims.

In response to the second concern:

Firstly, despite the relative frequencies of stranger and non-stranger violence, there is still far too much violence committed against women by strangers, which surely justifies it being addressed. Secondly, while women generally may be at greater risk from someone known to them than from a stranger, often the controversial advice is issued in response to an active at-large offender. The immediacy of this threat would seem to justify prioritising it. Thirdly, addressing stranger violence and known offender violence is hardly a zero-sum proposition, i.e., a small number of police are capable of warning against stranger violence at the same time that their colleagues are taking action against domestic violence offenders. Fourthly, even if police provide no advice, cases of violence against women in public places (especially offences such as homicide and rape) still attract substantial public interest and media coverage. Indeed, while some may fear that safety advice will result in women withdrawing from social life, given the link between fear of crime and excessive caution, media reports of serious crimes that are not accompanied by advice from police may do more to prompt women to withdraw from social life than media coverage with the advice. I will expand on this below.

3 Choice Denying vs. Choice Enabling

3.1 Virtue and Flourishing

Regardless of the motive behind the issuing of personal safety advice, it might be worried that such advice effectively limits women’s choices and, thereby, their freedom. It is easy to see why this could be. ‘Don’t go to the park alone’, for example, may result in a woman not going to the park at all. Likewise, ‘don’t walk alone at night’ may effectively prevent a woman from going out. In short, successfully preventing women from coming to harm in public places may come at the cost of confining them to the home (a place that we know is more dangerous for some women than public places) and preventing them from living fulfilling lives.

However, while this worry seems credible, I challenge the assumption that such advice is necessarily choice-restricting. Indeed, it has the potential to be choice-enabling. An Aristotelian account of flourishing demonstrates how.

According to Aristotle, individual flourishing is achieved by avoiding both “excess and deficiency” (Aristotle 2019, 1104a13). As he wrote:

“The same is true, then, of temperance, bravery, and the other virtues. For if, for instance, someone avoids and is afraid of everything, standing firm against nothing, he becomes cowardly; if he is afraid of nothing at all and goes to face everything, he becomes rash… Temperance and bravery, then, are ruined by excess and deficiency, but preserved by the mean.” (1104a20).

An ideal person in respect to personal safety, then, must be one who exercises practical wisdom and operates at the mean between (i) throwing caution to the wind and (ii) being overly cautious to the point of eschewing the goods of communal life. After all, excessive caution is surely as undesirable as insufficient caution, which serves as a useful reminder that while personal safety advice risks holding some people back, it also has the potential to lift others up.

But what, precisely, might it lift them up from?

It is important to remember that while fear of victim blaming may restrict the behaviour of some women, fear of crime itself certainly does (Keane 1998). Indeed, these fears can be debilitating. As Dobbs et al. (2009, pp. 107–108) note:

“The elevated fear of crime that females experience as a result of perceived vulnerability and/or the fear of rape can lead them to engage in either avoidance or protective behaviors. When engaging in avoidance behaviors, a female may choose not to walk alone after dark, avoid dense urban areas and the establishments located in them… or remain inside her home with the doors locked, essentially becoming a prisoner in her own home.”

If fear of rape, for example, risks making a woman a ‘prisoner in her own home’, then knowing practical steps for reducing the risk may empower her to leave her home. Effectively, such advice may help women be safer, feel safer and reap the benefits of this safety. As Keane (1998, p. 71) notes, after examining the fear and behaviour of Canadian women:

“… it appears that increasing feelings of safety should lead to an increase in the number of options of life experiences available. Simply put, increasing feelings of safety will increase women’s lifestyle choices. The consequence – freedom of movement – may be a liberating experience for many.”

This does not mean that personal safety advice is the only way of reducing fear, and some might be inclined to prioritise reducing crime rates via, say, increased police presence or by a change in offender behaviour. Nevertheless, it is mistaken to assume that fear of crime accurately reflects the objective risk of crime (Warr 2000), so we cannot conclude that simply reducing crime rates will make women feel comfortable exercising their freedoms.

3.2 Instrumental vs. Non-Instrumental Advice

Personal safety messaging can come in many forms. Sometimes it can be very blunt and prescriptive (e.g., ‘don’t walk alone in parks’), while at other times it can be more nuanced or advisory in nature. If we are interested in promoting flourishing rather than constraining liberty, we need to determine how police can formulate and deliver advice that respects the right of women to set their own direction in life.

A partial solution may be to make a clear distinction between advice that relates to instrumental activities that women engage in and advice that does not.

Consider the following two pieces of advice:

  1. A.

    Women should avoid walking through dark parks at night in order to get home safely.

  2. B.

    Women should avoid being promiscuous in order to stay safe.

Both of these sentences have a similar structure and indeed could be rephrased as IF-THEN statements:

A1. If a woman wants to arrive home safely, then she should avoid walking through dark parks at night.

B1. If a woman wants to stay safe, then she should avoid being promiscuous.

Nevertheless, there is an important difference between the claims: While walking through a park is done for the purpose of getting home (A1), engaging in sexual activity is not done for the purpose of safety (A2). Rather, engaging in sexual activity and being safe are two distinct goals that a given woman might have.

Consider that A1 is very different to the following:

A2. If a woman wants to stay safe, then she should avoid engaging in recreation in parks.

In the case of A2, there is a clash of non-instrumental goals, because recreation is not engaged in for safety reasons. If a woman’s non-instrumental goal is to engage in recreation in parks, then acting on this advice would clearly limit her freedom.

I suggest that in delivering personal safety advice, police should favour advice that focusses on instrumental activities in the interests of preserving women’s non-instrumental goals. Many pieces of advice do this already. Consider, for example, advice to women on how to avoid drink spiking (not leaving drinks unattended in bars, etc.), or the recommendation that they catch a taxi home when they have been out drinkingFootnote 3. Stumbling home drunk is hardly a non-instrumental goal, so following this advice should not impact on a woman’s liberty in any meaningful sense. Instead, the advice helps women achieve their goal of having a fun night out.

Of course, favouring instrumental advice need not imply that police should never give advice about non-instrumental behaviours. Such advice may be warranted in more extreme circumstances. For example, while blanket advice about avoiding parks is not ordinarily justifiable, it may be justified when an at-large offender (as in the case of our thought experiment) is actively targeting women in parks in a locality. Nevertheless, the threshold for issuing advice that recommends women abandon or alter a non-instrumental goal would be higher because it comes at a greater cost to women.

4 Conclusion

Are police officers justified in warning the public about the risk posed by a flooded stream? If so, are they also justified in warning women about the threat posed by an at-large murderer? It may be tempting to think that there are relevant differences between personal safety advice addressing gendered violence and personal safety advice addressing threats from impersonal sources, however this paper has argued that these differences disappear under close examination.

Perhaps the strongest concern relates to the long history of state-sponsored discrimination suffered by women. Nevertheless, it seems credible that, rather than reinforcing discrimination, sound safety advice may support women in eschewing excessive caution in favour of a form of prudence that fosters individual flourishing, especially if the advice relates to women’s instrumental activities rather than their non-instrumental goals. If so, then sound safety advice may be better conceived as ‘freedom enabling’ than ‘victim blaming’.

Finally, if my responses to these concerns are found unpersuasive, this still leaves us with the disconcerting position that a police officer may justly warn a female park user that she risks death by drowning but not that she risks death by homicide. This is a difficult bullet to bite.