How many men in Asia admit to rape?

Indian protester against rape

"One in four men across Asia admit to having committed rape." This statistic was widely reported around the world, following the publication of a UN study. Could that possibly be the case?

The numbers look shocking. A quarter of men in Asia admitting to having raped? It seems a huge proportion.

But returning to the original study, it is soon apparent that the stark headlines are not supported by the evidence. The reality is more complicated.

The researchers from the UN Cross-sectional Study on Men and Violence surveyed about 10,000 men in a handful of urban and rural areas in six countries in Asia and the Pacific - half were from Bangladesh and Indonesia, the others from China, Cambodia, Indonesia, Papua New Guinea and Sri Lanka.

So, contrary to many news reports, conclusions cannot be drawn about the behaviour of half a billion men across the continent of Asia.

Nor can conclusions even be drawn about the entire male population of the individual countries surveyed. The authors of the study state clearly that only one sample, in Cambodia, could be described as "nationally representative". The other samples do not represent a full cross-section of the population.

The press release issued to publicise the study was far more cautious than the media reports, saying only that: "Nearly a quarter of men interviewed reported perpetrating rape against a woman or girl."

But putting aside those reports about "Asian men" to one side, it's still alarming that a quarter of men interviewed admit to rape.

It is significant that two of the areas included in the survey have relatively recently suffered periods of conflict - the island of Bougainville in Papua New Guinea and Jayapura in Indonesia.

The numbers reporting rape here are much higher than in the other areas under study and have pulled up the average.

You can see this both in the figures for rape within relationships and rape outside relationships.

About 12% of men admitted to rape outside relationships. This number falls to 6% if you take out the post-conflict areas.

About 24% of men who have ever had a wife or girlfriend admit what's classed in this study as "sexual violence with girlfriends or wives past and present", or "partner rape", but this number falls to 18% if you take out the post-conflict areas.

Again, though, 18% is still a big number - it's nearly a fifth of men.

But another interesting detail is what questions the men were asked.

The first country to be surveyed was Bangladesh, and the numbers of men admitting rape were lower here than anywhere else. Here, the researchers talking to the men about their relations with partners or former partners asked a question about "forced sex".

In the other countries, a second question was also asked, and while a positive answer to this question was treated as an admission of rape, it doesn't explicitly mention force, violence or coercion.

It was phrased like this: "Have you ever had sex with your current or previous wife or girlfriend when you knew she didn't want it but you believed she should agree because she was your wife/partner?"

The number answering yes to this second question was larger than the number answering yes to the first question in all areas surveyed, apart from Sri Lanka.

"The question is basically, 'Did you make your wife have sex when she didn't want to because you thought she should do so?'" says Professor Rachel Jewkes, the lead technical adviser on the study.

"If you were to stand up in a British court and say to a judge, 'I didn't rape her, I made her have sex because I'm her husband - she didn't want it, but that's why I did it,' you'd get convicted of rape, and you'd get sent to jail in Britain. Sex is rape if you don't consent to sex."

But is it clear from the question that we're talking about making someone have sex, or someone not consenting?

Jewkes told the BBC: "It's implicit in the question that that's what had happened."

Felicity Gerry, a British barrister and co-author of the Sexual Offences Handbook blog, agrees with Jewkes.

She says the question is "carefully worded" to reflect the situation where a husband knows his wife is not consenting.

"It is not dealing with reluctant consent," she says. "It is another way of expressing a situation that amounts to marital rape."

Dr Jenevieve Mannell of the LSE's Institute of Social Psychology is more doubtful.

"Let's take a scenario where a husband and wife discuss whether or not to have sex," she says.

"She doesn't want to and he does. In their discussion the man raises the argument that she is his wife and that sexual intercourse is part of her role, and in the end the wife agrees."

In this case it would be correct for the respondent to answer Yes to the question whether he had ever had sex with his wife when she didn't want to, Mannell says.

But she asks: "Is this a case of coerced sexual intercourse or oppressive gender norms?" Her answer: "I would argue the latter."

It was nonetheless an important question to ask, Mannell says. A direct question about coercion would not be answered accurately by men who saw it as the woman's obligation to have sex, whether they wanted to or not.

So this less direct question, in her view, may not exactly have assessed "evidence that coerced sexual intercourse took place". But the conclusion of the report, she says, was absolutely correct: "We need to start paying attention to the ways in which violence against women has become a social norm."

Follow @BBCNewsMagazine on Twitter and on Facebook