研究发现，警察更倾向于拘捕有家庭暴力行为的男性，而对同样引发家庭暴力的女性则多数倾向于听之任之。

A recent report about domestic violence could actually confirm that male victims are not taken seriously by police

The debate on domestic violence （DV） statistics took a fascinating turn last week. Recent years have seen numerous surveys which suggest that incidents involving male victims and female perpetrators are more common than had previously been assumed. Men's charities have long argued that the official crime figures significantly under-representthe problem, because so few men report abuse to the authorities, and those that do are likely to be disbelieved, often to the extent of being arrested themselves when police arrive.

Last week, the media reported a study by Professor Marianne Hester, which analysed amatched-pairs sample from the Northumbria police database of DV incidents over a six-year period. Hester's report actually tells us little that we didn't already know.The most eye-catching details, to me at least, are the findings that female perpetrators are more than twice as likely to use a weapon as their male equivalents, and that couples where both parties are prone to aggression have police call-outs about four times as often as those couples with only one regular aggressor. In that light, it is disappointing that the University of Bristol press release, which formed the basis of all the press reports, chose to lead with the revelation that women are three times more likely to be arrested as men when they commit such acts. That hardly fits with claims that police don't take male victims seriously. Can these competing claims be in any way reconciled? I think they can.

Hester's study found that male perpetrators are arrested on average after one in 10 incidents, whereas with a female perpetrator it is one in three. When an incident of suspected DV is reported to Northumbria police, officers will always attend the scene. When they arrive they are expected to identify the "primary aggressor" and, according to ACPO guidelines, if there is any evidence of a crime having been committed or if someone is in danger, they should always make an arrest. Since the priority is to defuse the situation and ensure the victim's safety, they are instructed to make only one arrest where possible, and according to the data, dual arrests are indeed extremely rare. Officers will later enter a short description of the incident in the database.

The statistics here tell us that even though police are required to make an arrest if there is any evidence of a crime having occurred, they usually do not – presumably because there are no grounds to do so. This is not surprising. In practice, officers will often arrive after a situation has pacified, with neither party wishing to make a formal complaint. Or they may find an ongoing, chaotic scene of anger, tears, accusations and counter-accusations, often involving people influenced by drink or drugs, but no actual violence evident and no coherent testimony. If it seems likely that the situation will escalate into violence again when police leave, an arrest will have to be made, usually on grounds of breach of the peace – by far the most common offence recorded in the study. When no arrest is made, it will usually be because the scene is relatively safe and/or ambiguous, with no clear perpetrator, victim or specific evidence of an offence. Nevertheless, for the purposes of the police DV database, officers are still required to identify a person whom they think is the "primary aggressor". That one little judgment may be the key to this mystery.

What we have here is not the ratio of how many men/women are arrested against how often they have committed an offence, but the ratio of how many men/women are arrested against how often a police officer decides that, on the balance of probabilities, they might have been more at fault.

If the men's groups' claims about the police are valid, these results are in fact exactly what we should expect to see. The assumption, in the vast majority of ambiguous cases, has been that the man is to blame. What initially looks like a clear case of discrimination against women may in fact be a product of prejudice against men.

Of course, it could be that the police officers are entirely accurate in all their assessments, and the men are indeed the primary aggressors in the vast majority of those incidents. Or these data could be hiding a rash of violent attacks against men.

Unfortunately, one needs to read Hester's paper very closely to unravel this explanation. Nowhere in the report does she consider the possibility that the data she is using may be corrupted by the underlying biases, assumptions or slapdash paperwork of overworked police officers.

最近一份有关家庭暴力的研究报告几乎可以肯定男性受害者没有受到警方认真对待

上周有关家庭暴力数据的辩论作出了一个华丽的转身。近些年大量调查表明，涉及男性受害者女性施暴者的事件比原先估计的要平常的多。男性团体长期以来认为官方犯罪数据严重低估了这个问题，因为几乎没有男性向有关部门报告受虐，那些报案的男性也可能不被采信，这已经到了当警方到达的时候，通常会拘捕那些男性报案者的地步。

上周媒体报导了一项由玛丽安娜·海丝特（Marianne Hester）教授进行的研究；她分析了来自诺森比亚（Northumbria） 警局数据库时间跨度6年的家庭暴力事件的配对样本。事实上海丝特的报告没有告诉多少我们所不知道的东西。其中最吸引眼球的细节，至少对我而言，是发现了女性施暴者使用武器的倾向是相同条件下男性施暴者的两倍多；双方都有攻击倾向的夫妻报警的频率是只有一个稳定施暴者的大约4倍。如果鉴于上述发现，那么布里斯托大学（University of Bristol） 的这篇学报，以及基于这篇文章的其他报告读来就会令人感到遗憾；上述报告均选择把女性被捕的可能性3倍于犯下同样罪行的男性作为其首要发现。这几乎无法同警方不认真对待男性受害者的主张相吻合。相互竞争的这些主张是否可能以某种方式相调和呢？我认为它们可以。

海丝特的研究发现在10起案件中平均有1名男性施暴者被捕，而女施暴者3起中有1个。当一起疑似家庭暴力事件报告给诺森比亚警局的时候，警官们总会进入案件现场。当到达的时候，他们要确认"主要侵犯者",而其依据英国警察协会（ACPO）手则，如果有犯罪发生的证据或者如果某人处于险境，警方应该做出拘捕。由于首先要缓和局势并保证受害者的安全，警官被要求在必要的的时候只逮捕一人；根据数据显示，逮捕双方极为罕见。警官们将会在以后把事件的简要说明输入数据库。

这里的数据告诉我们，如果有某些犯罪的证据，甚至是在警方被要求采取拘捕行动的时候，他们通常也不会做出行动--大概是因为没有这样做的理由。这并不让人吃惊。在实践中，警官门通常在事态平息后到达，当事双方都不想做一份正式控告。或者警官可能发现混乱的状态仍在继续，那里有愤怒、眼泪、控诉还有反控诉，通常当事人都涉及酒精和毒品，但是没有确实的暴力证据，以及一致的证词。如果看上去警方一旦离开，局面就要再次升级为暴力，此时才必须做出拘捕，通常是以妨害治安的罪名--到目前为止这是研究中最常见犯罪记录。没有做出拘捕的时候，通常是因为局面相对安定，还有/或者也搞不清楚到底形势如何，这造成了没有施暴者、受害者或者是某项犯罪的具体证据。尽管如此，为了警方家庭暴力数据建立的目的，警官们必须确定一个他们认为的"主要侵犯者".这一小小的判断可能就是揭开迷局的钥匙。

我们目前掌握的比率，不是多少被捕的男性或者女性与他们多久犯下罪行之比；而是多少被捕的男性或者女性与多久一个警官作出自己的判断之比，在权衡了各种可能之后，这些数据更可能是错误的。

如果男性团体有关警方的断言是有根据的，上述结果事实上就是我们应该期望看到的。在绝大多数模棱两可的案件中，都假设应该谴责男性。起初貌似歧视女性的案子可能事实上是对男性偏见的产物。

当然，有可能警方所有的评估都是完全精确的，男性真的在绝大多数案件是主要侵犯者。或者这些数据掩盖了一系列针对男性的突发暴力攻击。

不幸的是，你必须非常仔细的阅读的海丝特的论文才能弄清楚这个解释。她完全没有在报告中考虑到会有这样的可能性，正在使用的数据可能会受到某些潜在偏见，那些过劳的警官们的推测，或者草率的文书工作的误导。