How we measure violence

I’ve noticed lately that the way we talk about prevalence of gender-based violence has changed lately. While we used to talk mostly about incidence of violence, a measure riddled with problems of underreporting and non-response, more scholars, NGOs and thus media outlets are concentrating more on measures of acceptability of violence. The questions “is wife-beating ever justified?” and “when is wife-beating justified” are garnering more attention than ones that seek to pin down the number of times a wife was actually beaten. The extremely high affirmative response rate to these questions (a recent TrustLaw post cites a UN study claiming at least 25% of people think it’s justifiable for a man to beat his wife in 17 of 41 countries surveyed) reinforces the notion that we might be missing a lot with surveys that get at instance.
it of course, does nothing to mitigate problems of reporting in places where the practice is outwardly condemned. In the US, I’d imagine, the statistic isn’t very useful as you’re unlikely to find many people who would assert that domestic violence is justifiable.
Additionally, it seems that, just like with incidence reports, the answers are subject to social norms and prevailing custom. In that sense, though, the question about justifiability may more closely measure the social norms themselves than questions about incidence.