The World Bank and Oxfam India  co-organized a high energy event earlier this week - Joining Forces to End Violence Against Women . It was an intense two days – about 200 participants from diverse backgrounds gathered to listen, to educate each other, to speak up, and to build alliances; in short, to join forces towards the next step. Several of them congratulated the “movement” on progress – on having coopted unlikely allies, on the fact that more men were involved than ever before, and that public outrage against violence is widespread in South Asia. Surely, this will lead to change, is the implicit hope. But long-time warriors like Flavia Agnes, voiced angst and discouragement, as only those who have spent a lifetime of struggle are entitled to. Finally, the anger came from 21-year old Urmila Chaudhary – freed from bondage as a Kamalari  – “where were you all when I was pledged to a family as a maid at the age of six”, she asked a somber audience?
From the impassioned to the banal, I steered the conversation in the opening panel towards data and research as one of the ways to actualize the hope of action. The Demographic and Health Surveys  (DHS) have been seminal in this regard. They fielded their first module on intimate partner violence (IPV) in 1990, and opened the floodgates to studies on the prevalence and determinants of IPV. These studies helped to establish to skeptics and to those who asserted that violence is a private affair, that it is in fact pervasive, and it is related to a range of poor development outcomes and that policy cannot be silent. The DHS modules are carefully designed, investigators diligently trained and modules suitably revised. This is not the norm for a number of other surveys that also try and measure the extent of violence against women.
There are serious issues in collecting reliable data on spousal violence. Not least of these is ethical. Researchers can well go into homes, asking blatant or even guarded questions about the experience of violence, and laying the respondent open to retaliation, regardless of her answer. This may well affect the reliability of the data as well, but that’s less important than the effect on her well-being. I worry that there is a proliferation of surveys that measure violence of different kinds, especially in the aftermath of the heinous Delhi gang rape. I think there is a collective responsibility to ensure that investigation about the prevalence of violence doesn’t endanger the woman in question. A guide to collecting data on domestic violence that Mary Ellsberg and Lori Heise published almost a decade ago is still very relevant and should be essential reading for anyone doing field work in this area (pdf here ).
A second point about collecting data on violence concerns not actual violence, but the threat of it. This is especially important because the threat of violence and lack of safety in public spaces deters women and girls from leaving their homes to seek better lives such as going to school, college, to health centers or to work. Yet, few surveys in non-OECD countries measure the perception of safety and security. The last Urban Audit Perception Survey  (pdf) conducted as part of the Euro Barometer in 2009 asked respondents in 75 European cities whether they felt safe in their city. As expected, there is huge variation across cities, with residents of Bucharest, Sofia, İstanbul, Athens and Naples feeling most unsafe and the Nordic countries reporting greatest safety. The India Human Development Survey  conducted by NCAER and the University of Maryland now has two rounds of surveys with questions on perceived safety.