The assumption that men cheat in relationships more often than women is received wisdom, a common trope. The viral HurtBae youtube video and this article from BBC News online are two examples among many:

men cheat more than women

Not only do people seem to widely believe this, many also think they can explain it according to received wisdom about the differences between the sexes. Here for example is the Huffington Post:

why men cheat

Both the idea that men cheat more than women and the male and female characteristics which rationalise that idea are so widely held that they cannot have surprised you. The articles quoted above, like nearly all writing on this subject, come with highly suggestive and stereotyped images which confirm the received ideas they illustrate. This is why these stories make for such good clickbait.

Do you Lie?

Like a lot of received ideas, this one is not necessarily accurate nor supported by statistics. The subject of infidelity in particular is one which has deliberately evaded statistical analysis throughout human history. What happens in the bedroom (or the car, the field, the airplane toilet, wherever) stays there, unless the people involved tell you about it – and people who have affairs tend not to shout about them from the rooftops. You can’t count affairs the way you do tonnes of iron ore output or road traffic accidents. There is no independent body which can record all affairs as they happen (not yet anyway, thank God). So any “statistics” will not be based on records, but on surveys, in other words anonymous confessions of infidelity to pollsters. If you use such surveys you never know, for example, whether more men really have affairs than women, or whether women are just less likely to admit to having affairs. It’s not for nothing that novelists, playwrights and poets rely so deftly on the dual meaning of “lie” to mean “lie down” (with a spouse or a lover) and “tell a lie.” This ambiguity is cradled by the incredibly talented, late and bitterly lamented Prince in these lyrics:

Les enfants qui mentent ne vont pas au paradis [children who lie don’t go to heaven]

O.K Merci.


When I lie awake in my boudoir, I think of you dear
Do you think of me, or do you lie, do you lie?

(“Do U Lie” from Parade, 1986)



Two women for every man?

But what bothers me about this particular received idea is this: if you look at the photos above, in the first you have one man on the couch and two hot women competing for his attention, and in the second you have one man in the background with a wife on the phone to his mistress. It is a basic fact that for a man to have an affair he needs (at least) two women (a wife and a mistress) and both photos, accordingly, portray a world in which there are two women for every man! Personally I would love the real world to be like this, particularly if every second woman looks like the dark haired beauty on the couch. Sadly, the World Factbook reports that for 15-64 year olds (the population segment most likely to have an affair) there are actually 1.07 men for every woman in the United States of America. Far from there being two women for ever man, there are slightly more men than women in the USA (roughly 3.4% points of total population per the World Factbook figures). But if there are more men than women and men are more likely to have affairs than women, who on earth are the men having these affairs with??

Whatever the accuracy or inaccuracy of the surveys, there’s a basic problem of arithmetic with this supposed phenomenon (which the BBC’s Fiona Woods alludes to without analysing it): for men to have more affairs than women they have to have affairs with women. Looking at the numbers, if we accept the survey data in the BBC report above that 15% of men and 9% of women have “overlapping relationships,” then 7.75% of the population would be men in overlapping relationships and 4.3% would be women in overlapping relationships, meaning a total of 12.1% of the total population are in “overlapping relationships” (represented in the third bar of the chart below).

percentage unfaithful

The next chart further analyses that 12.1% of the population in “overlapping relationships.” Assuming a total population of 1m, you would have a total of 121,000 unfaithful cheats (12.1% of total population), of which 77,500 unfaithful men (7.75% of total population) and 43,000 unfaithful women (4.3% of total population). But with whom could these 77,500 Cassanovas be having their flings? 43,000 of the 77,500 male cheats (i.e. 4.3/7.75 = 55% of all philandering men) can cheat with their 43,000 female cheat counterparts. But then with whom can the remaining 34,500 male philanderers possibly be creeping?

percentage unfaithful match

Either there are more unfaithful women out there than assumed, or the unfaithful women are raunchy devils, having affairs with almost two men at a time (43,000 unfaithful women would be having affairs with 77,500 unfaithful men!). These basic facts would naturally lead you to be suspicious of the intuitive reasons given for the notion that men have more affairs. The “stronger sexual impulses” cited above would be particularly lame, as the actual numbers only make sense if the unfaithful women are twice as “active” as the unfaithful men! The unfaithful women would be like R&B legend Betty Wright’s clean-up woman, “a woman who, gets all the love we girls leave behind.”

Causal Fallacy

These popular notions follow a well established psychological pattern analysed by behavioral economics, which I will refer to as the “causal fallacy.” The causal fallacy describes the mind’s tendency to see events, or a sequence of events, and attempt to find a cause for those events from a series of readily available, plausible explanations, even though statistics may show that there is nothing in those events to support this causal explanation. In Thinking, Fast and Slow (2013) Daniel Kahneman uses relationships between men and women to illustrate the causal fallacy. He analyses Francis Galton’s insight about regression to the mean with “a proposition that most people find quite interesting” (page 181):

Highly intelligent women tend to marry men who are less intelligent than they are.

Kahneman writes: “You can get a good conversation started at a party by asking for an explanation, and your friends will readily oblige. Even people who have had some exposure to statistics will spontaneously interpret the statement in causal terms. Some may think of highly intelligent women wanting to avoid the competition of equally intelligent men. Or being forced to compromise in their choice of spouse because intelligent men do not want to compete with intelligent women.”

As with the notion that men cheat more than women, the idea that intelligent women tend to marry men who are less intelligent than they are is easily confirmed by examples which readily spring to mind, and it spontaneously encourages any number of explanations based on characteristics associated with women (wanting to avoid competition out of female insecurity, say) or men (wanting to avoid competition out of male pride, say). And it’s as easy to start a conversation with these familiar notions as it is to get a click with an article about male infidelity.

Kahneman’s alternative to this easy, homespun causality is basic arithmetic of the kind I have been trying to apply to the question of affairs. If a woman is in the top 1% of brightest people, he points out, then there is only a 1% chance she will find a husband as intelligent as she is and a a 99% chance she will find one less intelligent (assuming similar distribution of intelligence among the sexes of course). There is nothing in the female or male psyche which explains this trope, it’s just the law of numbers.

Bu you can’t dismiss the popular notion of male infidelity just because it’s commonly held. Many of these rules of thumb are true, and people believe them for a reason. I therefore want to explore the potential explanations, but from a numerical point of view.

Homosexuality as explanation?

One potential rationalisation for heterosexual men having more affairs than heterosexual women may be the differing rates of homosexuality in the genders (of course you will notice that the articles above refer to men and women without specifying their sexual orientation because of their reliance on a familiar stereotype in which heterosexuality is a given). If a greater percentage of men than women are homosexual that would leave fewer potential heterosexual male partners for the heterosexual women. If, in turn, there are fewer heterosexual men than women then that leaves scope for some of them to have relationships with one woman and an affair with another, or even many others. Off camera, in the images above, would be a gay man who leaves a field in which two (straight) women have to compete over one (straight) man.

The idea that all the philanderers and Lotharios out there actually have their gay brothers to thank for their opportunity (“high five”!) is an attractive irony and would probably make for awesome clickbait!

gay family guy

But even though gay men may be more visible, particularly in certain cities, than gay women, the statistics are not straightforward. As with affairs, sexual orientation is a private matter, and can only be measured with surveys. Moreover, there are different ways of defining sexual orientation and identification and – again, as with affairs – the sensitivity of homosexuality as a subject can affect how people respond to surveys on the topic. According to the National Survey of Sexual Attitudes and Lifestyles, only 1.5% of men and 1% of women in the UK identified as homosexual. This minute 0.5% gap, if it is true in the USA, would not even be enough to compensate for the 3.3% greater percentage of men relative to women in the relevant population. Alfred Kinsey’s studies of sexual attitudes in the US in the 1940s however found that a full 13% of men had a high score of homosexual orientation versus only 7% of women (the difference between Kinsey’s and Natsal’s results constituting an eloquent demonstration of the difficulty of analysing this topic with rigorous statistics). If you accept that wider disparity – which is a big assumption – then you get to a roughly equal percentage of heterosexual men and women (52% of the 15-64 year old population are men of which 87% are heterosexual means 52% x 87% = 45% of the 15-64 year old population are heterosexual men; 48% of of the 15-64 year old population are women of which 93% are heterosexual means 48% x 93% = 45% of the 15-64 year old population are heterosexual women; ratio of heterosexual men to heterosexual women = 1:1). Even if you go all the way back to the 1940s and use Kinsey’s questionable numbers you can’t account for the possibility of more men than women having affairs.

Not all men are created equal

The other potential explanation is distribution of relationships. Perhaps some men have relationships with many women, while other men have no relationships at all. The greater number of men having multiple (“overlapping”) relationships would compensate for and be balanced by the men whom women won’t go near to. This would fit strongly with intuitive explanations of the difference between the sexes according to patterns observed in the animal kingdom. Just as the dominant or king lion in a tribe mates with all the females, while the other male lions live a life of frustration, perhaps some men attract, have relationships with and are therefore unfaithful to many women, while other men just get no action (and presumably play video games and watch porn instead).

alpha lion

This would in turn fit certain genetic explanations of sexual preferences. Men, according to many Darwinian inspired theories, because they can reproduce any time, want to spread their seed among as many women as possible to maximise their chances of reproduction. Women, who carry any child they have for nine months, want to maximise the quality of the genes which their child carries, and therefore have kids with the man with the best genes, who is often the same man with whom other women also want to have children – hence the affairs.

I have no idea whether these theories have any foundation whatsoever. Such intuitive, naturalistic analogies are the sort of thing of which I’m naturally skeptical. And again, this notion can only be tested with surveys. In this case, surveys of whether you’re single or not are more likely to be responded to truthfully than surveys of whether you’ve had an affair or not, and so might provide an indirect glimpse of the truth. This relatively recent yougov survey, reported by the brilliant statistics and prediction site provides the following insight:

percentage single

At the bottom we find a full 3% points more of all men had never had a long term relationship than the equivalent figure for women, at the time of the survey. Because these guys are out of the game, you might argue, they leave more opportunity for the philanderers to have multiple relationships. But if the disparity in the percentage who never had a relationship, reported by yougov, is applied to our base population of men and women, then males who have never had a long-term relationship are 5.7% (11% x 52%) of the total population and women who have never had a long-term relationship are 3.9% of the total population (8% x 48%). The difference between the two figures is 1.8 percentage points of population, again less than the 3.4% gap between men and women among the total 15-64 year old population.

Urban myth?

You can, at a push, combine the impact of the Kinsey percentages of homosexuality among men and women and the yougov figures on the percentages who have never had a long-term relationship, to arrive at a percentage of the population who are both heterosexual and have not been single all their life. This is of course very, very dodgy statistics, as you have no idea whether the percentage of homosexuality among people who have never had any long-term relationship is the same as for the overall population, or vice versa whether the percentage of people who have never had any long-term relationship is the same among homosexuals as it is in the overall population. On top of that you are combining different surveys of different population groups carried out at different times with different methodologies. Tsk tsk.

But even if you accept this approach, you get 40% of the 15-64 year old population consisting of heterosexual males who have had a long term relationship versus 41% consisting of heterosexual females who have had a long term relationship. That measly 1% difference does not account for anything like the various survey reports of infidelity or, more importantly, the potency of the commonly held belief that “men are cheats.” It isn’t even larger than the margin of error.

The reason for this (and the reason why these numbers may not be so dodgy after all) is the fact that we are dealing with small variations. The difference between the male and female percentage of the relevant population is small (3.4%); the percentage of homosexuals in the total population is small (10% per Kinsey), as is the difference in prevalence of homosexuality between men and women (6% per Kinsey); the percentage of people who never had a long term relationship in the total population is also small (c 10%), as is the difference in prevalence of people who never had a long term relationship between men and women (3%). With such small numbers it is very hard to arrive at any meaningful difference between men and women which would justify the received wisdom about men cheating.

Assume, just for the sake of argument, that we drastically increase those numbers, say by just under 50%, adding 5% to the percentage of homosexual men (which would mean a full 18% of men, or nearly one in five, were gay) and men who never had a long term relationship (which would imply a massive 16% of men never had a long term relationship). You would then have “only” 35.6% of the population consisting of heterosexual men who had had long term relationships, versus 41% of the population consisting of heterosexual women who had had long term relationships – 5.4% more women than men. Only such extreme, and frankly far fetched assumptions are enough to almost support the survey responses on the disparity in prevalence of infidelity recorded by Natsal (reported by the BBC article above), of a 6% higher percentage of men than women admitting to infidelity!

Power as soap opera

My suspicion is the following. Power is concentrated in the hands of a few people, today as it was in the past. Be it political, financial or cultural. These people are predominantly men. Powerful men tend to be married, for various reasons, many of them many times. These powerful men have plenty of opportunity to have affairs, and when they do it is often with very glamorous women. Such affairs are widely reported, “available” and easy to remember. In this particular, highly visible and reported world, it really is “the men who have the affairs.” Not because their wives don’t have affairs. It’s just that the media and those who record events (gossip columnists, poets, dramatists, novelists, film makers, artists etc.) are typically interested in the men’s affairs because of their power.

These power structures have existed a long time. And even if we have writers like Boccaccio whose Decameron (1351-1353) is full of women merrily cheating on their husbands, most literature features men having the power and the affairs. It is quite possible that these narratives were reinforced for a generation of women who grew up during WWI and WWII, when they would have experienced a real shortage of men, due to the men who were fighting on the front, and afterwards due to the number of men who died in the conflict. Whatever the case, such narratives have, I think, been transposed to infidelity in the population at large for centuries. They have been used to make sense of and give coherence to the affairs people notice, report on and talk about, and this tendency in turn has made people more receptive to such narratives, and so on, in circular fashion, for centuries. The transposition is easy and familiar (it is described as a “cascade effect” by Kahneman). It is because of this that you will find it easier to remember examples of men cheating than vice versa, and that women who have been cheated on will find it easier to talk about the fact, whereas women who cheat will be more reluctant. And remember, we don’t have real statistics to disprove it because we don’t like to ‘fess up about these things. Yet most people love to talk and speculate about affairs. Even the best monogamous relationships can be boring. The transposed narrative of male cheating satisfies our natural desire for explanations that make sense of the world; it comforts us in our identity as a man or a woman, however stereotypical; and it gives us an easy story to make ordinary time pass more quickly.

The maths however make me suspect that reality is very different from this narrative, as this 2000 R&B classic by the (utterly stunning) singer/songwriter Kandi eloquently asserts:

In her immortal words, “I catch a bone, while you’re dogging me.” The statistics in this piece are there to show that there are only so many bones to catch, and if a man is catching one, then so is a woman.

Agree? Disagree? Have an explanation for the belief about male infidelity? Please comment. I always respond.