March 13, 2021   7 mins

Men do not understand what it is like to be a woman — to feel unsafe on the streets. In the aftermath of the death of Sarah Everard, whose body was recently found outside Ashford, that fact has been extremely clear. 

I have walked the streets of London and Liverpool on my own for the past 20 years, at all times of night, rarely giving it a second thought. So reading women’s accounts of their everyday experiences, and talking to the women in my life about theirs, is a glimpse into an alien world. I am left with the terrible, bleak feeling that it will be impossible to ever make women feel adequately safe.

There is a kind of stereotyped masculine response to these female fears. That is: they are irrational. Murder, especially the murder of women, and especially the random murder of women by strangers in the street, is vanishingly rare. You are far more likely to be hit by a car on your walk home. So if you worry about it, goes this response, you are wrong.

This response is not exactly wrong, but it misses the point by a distance. It’s absolutely true that murder is rare in the UK; you are less likely, as a citizen of the UK in 2021, to be murdered than almost any other human in history. The Crime Survey for England and Wales puts the UK’s homicide rate at about 1.1 per 100,000. For comparison, according to the UN, the USA has something like 5 per 100,000; Brazil, about 27.

In the UK, and many other countries, the number of murders per year has been dropping since at least 1990. Going further back, there was a surge in the later years of the 20th century, but on a longer timescale the trend is clear: you were many, many times more likely to be murdered for most of human history than you are now in the UK, or in most other countries.

To carry on in the same stereotyped vein, if you are a woman, you are even less likely to be murdered. Of the 695 people murdered in England and Wales in the year to March 2020, 506 of them were men. And for the specific issue of murders on the street, by strangers — the fear we’re dealing with — the disparity is greater still: 154 men were murdered by strangers, and just 23 women. 

All this is true. But that doesn’t mean that the streets feel safe for women.

For one thing, it’s not just getting murdered that women fear. They fear getting abducted. They fear being raped. They are made by men to feel uncomfortable. Men yell things out of cars at women as they drive past, or follow them as they walk home, shouting out to them. One woman I know finished a half marathon as a bunch of men at the finish line loudly rated the attractiveness of the runners (“would, wouldn’t.”) 

I’m a bit sceptical of surveys like the recent one which found that 97% of young women have been sexually harassed, because the definition of “sexually harassed” varies, and I’d say some of the things they include in their data are borderline. But I bet that most women have some stories like the ones above. 

It’s mercifully true that almost none of those cases will escalate into murder, and few into actual violence. But, equally, I don’t think it’s irrational to be unnerved, in a world where these things happen reasonably regularly, and so to avoid being out alone after dark. 

If men being threatening is the main driver of women feeling unsafe, you might think that we have a clear way to move forward. Simply get men to stop being threatening, and women will feel safe. But I worry that will fail, on two grounds. One, I don’t think it’s possible, and two, even if it were, I don’t think it would work.

The first trouble is that we bump up against the law of large numbers. Imagine that the average woman walks past 200 men in a day. I expect working women in large cities see much more than that, in non-pandemic times, but others who are at home or who live in small communities see far fewer, so perhaps that number’s about right. In a year, that’s 73,000 encounters; in a decade, 730,000. Even if only a tiny fraction of men are dangerous or threatening, women will occasionally encounter them. 

Forget “not all men”: even if it’s “almost zero men”, there are enough to make women feel threatened. And it’s probably not almost zero men; it’s probably some relatively small but non-trivial fraction of them. (Or, even worse, a larger fraction of men, but only some of the time, for instance when drunk.) How many incidents per decade would it take to make you anxious? One? Two? 

There’s more than just this going on though. We’re also battling against human psychology.

In the US and UK, crime has been falling consistently for 30 years. In America, there are something like half as many violent crimes per person per year, and about half as many property crimes, as there were in the early 1990s. In the UK, the crime rate has fallen to something like a quarter of its 1996 peak.

But in both the USA and UK, most people believe that crime is increasing. The percentage who think so has dropped in the UK over the last 10 years, but still, nearly two-thirds of adults think that crime is more common now than it was a few years ago. Almost 80% of US adults think there is more crime now than a year ago. People’s perception of the risk of crime is only very loosely tied to the actual risk of crime in their country. Bringing it down — making people safer — will probably only have a relatively small effect on how safe people feel.

What’s more, while it is true that things like traffic accidents are much more common, humans are much more likely to be scared of dramatic, rare risks than common ones: “They worry more about earthquakes than they do about slipping on the bathroom floor, even though the latter kills far more people than the former,” wrote the security expert Bruce Schneier back in 2006. 

Schneier also points out that personified danger, dangers with identifiable victims, is far scarier than anonymous, statistical risks. 10,000 faceless people dying of diabetes will never get to us in the way that a single named child ill with bone cancer will. If there’s a human agent behind the risk, that’s even worse. And risks that we don’t choose, risks that are beyond our control, are scarier than those we choose. High-profile murders by strangers meet all of these criteria. They are almost designed to scare us. 

And we will always keep hearing about them, once again because of the law of large numbers. There are about 70 million people in the UK. If we get the risk of murder down to an unrealistically low one-in-70-million per year, we will still hear about them once a year, because they will make the news. The news will never dedicate 100 times as much coverage to diabetes deaths as they do to street murders, because that’s just not how the media works or, really, ever could work. A newspaper headline of “100,000 aircraft landed safely yesterday” would not sell many copies.

More than that, women tend to be more risk-averse than men. I want to be really clear about something: this is not “irrational”. There’s no optimal level of risk aversion, no correct amount to be worried. But it is an empirical fact that although men are more likely to be victims of crime and violence, they are less likely to be worried about it. 

That makes sense. Even if I do get jumped in Finsbury Park as I walk home some night, there’s some deluded bit of me that thinks I could fight, or at least run. I would feel like I have some control, and risks we think (rightly or wrongly) that we can control are less scary. Most women do not have that feeling. It is hardly surprising that the world is scarier for women.

Plus, we’re diurnal animals. The night is frightening for us, and has been since before we evolved from our common ancestor with chimps. You don’t have to go all evolutionary psychology to think that being alone on a dark street is going to be scary for humans, whatever the real risks are.

Women also differ widely in their responses to things. There is a great essay by Scott Alexander, called “Different Worlds”, talking about how the same situation will be perceived entirely differently by different people. We all have different levels of threat-perception, and almost all signals we receive from the world are ambiguous. Some people will read some behaviour as dangerous, some as benign.

I read this story, about a woman’s reaction to unwanted attention on a train, and it felt threatening and unpleasant. But perhaps some other women might think it was harmless flirting. And the survey into sexual harassment I mentioned above included “being stared at” as sexual harassment, but I remember some female friends, when they were single, telling each other about that guy at the bar “checking you out”. The line between flirty “checking out” and harassment-level “being stared at” will be different for different women. Which is entirely normal and unavoidable. But it does mean that some women will be much more likely to feel threatened than others. Which will in turn mean that it will be harder to reach a point where all, or enough, women feel safe.

What can be done? Certainly men could be more aware of behaviour that might seem threatening. And doing whatever we can to reduce the level of crime, improving rape conviction rates, reducing workplace harassment and stigmatising the sort of creepy behaviour I talked about above — those things are good in their own right, and probably will have some effect on how safe women feel. I’ve seen women suggesting curfews for men — I’m not sure how seriously — and no doubt that would lower the incidence of attacks, although it’s not really proportional or realistic; it would involve punishing an awful lot of innocent men for the actions of a terrible minority. As would restricting alcohol sales, given that at least half of attackers in violent crime are under the influence.

Fundamentally, we are fighting too many things — especially the law of large numbers, a sensationalist media, and human psychology — for it to be realistic that women will ever feel safe enough. Men will never understand how women feel; but even worse, we may never be able to entirely change it.

Tom Chivers is a science writer. His second book, How to Read Numbers, is out now.