Friday, June 18, 2010

Mulling the Vacuousness of Variables

A few days ago, the Hindustan Times (Bombay edition) featured results from a data compilation conducted by the Tata Strategic Management Group (TSMG). More details can be found here-  http://epaper.hindustantimes.com/Default.aspx?Id=4732524D44634E503452454761516175534D304C706661536D6C2B2F4530656F494E7A6F745436567A5A553D


According to the data, Delhi, Bombay, Pune, Bangalore and Chandigarh are the highest rankers in terms of the well-being (quality of life) index. Low scorers include Orissa, Bihar, Jharkhand, Chattisgarh and Arunachal Pradesh. As far as the Female Security Index, a measure of crime against women is concerned, Orissa, Bihar, Tamil Nadu, West Bengal and Kerala fare well. Maharashtra (including Bombay), Delhi, Haryana, MP and Assam fare the worst. Given that the states with the highest well-being indices are also among the poorest in terms of women's security, the TSMG makes the assertion that there is no link between affluence and security. This is apparently critical information that our government could use to great advantage. Applause!


I thought the entire exercise was a textbook study in how data can be mismanaged and misinterpreted.


To begin with, I find it difficult to believe that any city can be considered to offer a superior quality of life when a significant segment of population feels unsafe within it. Women contribute to cities, build them, rear children in them, and work in them. Shouldn’t their sense of safety be built into any construct that claims to approximate the well-being of the city’s citizens at large? Any distinction between quality of life and security is artificial – a theoretical cop-out.


Moreover, the female security index was developed using two variables – the gender ratio, and the crime rate against women. While these variables may provide a starting point with which to assess the environment in which women function, it’s na├»ve to assume that they in any way represent a woman’s lived experience in a city.


Gender ratios tend to differ between regions in a state, and indeed, within communities and income groups within a single region. This study fails to account for that complexity. Moreover, gender ratios have been skewed in most parts of the country for generations – I am not sure whether this study accounts for any historical improvement or deterioration of the ratio – a better indication of how attitudes to girl children and women are evolving.


The rate of reported crimes against women is also extremely problematic as a measure of women’s safety. Women report crimes when they are secure in the knowledge that they will not be harassed by perpetrators or the police. At the very least, they will seek support from family, social organizations or their community before reporting any offence – and how many women in India can truly draw on such support when needed? Added to this are issues of perceptions and awareness regarding crime – surveys have shown that many Indian women find the idea of ‘domestic violence’ laughable – it is not part of their view of the world. No crime rate can account for the invisible injustices and indignities women encounter every day, and which they tolerate as a part of ‘their lot.’


It's also important to note that a woman’s safety is not only about the likelihood of being murdered or raped. It is about the little things as well, the intimidating body language of passers-by in the street, the fear of wearing a ‘provocative’ outfit, the ability to go out at night knowing that one can use public transport to get back home. Which is why I find it difficult to believe that Bihar ranks higher than Bombay on safety for women. In my experience, Bombay is one of the safest cities for women in the country – women are out and about with much more confidence here than in, let’s say, Patna. It’s possible that my experience is limited and non-representative, but my point is that it’s not enough to look at a crime rate in Bihar and say that the state is safe – it is likely that women in Bombay are just more willing to confront and register crimes than those in other parts of the country.


And given that well-being in this study is constitued of factors such education, hygiene, entertainment, communication, healthcare, transportation, home amenities and kitchen facilities, it is of course no suprise that there seems to be little connection between well-being and women's safety. There is no overlap in how these constructs are measured, and I am not sure whether any attempt has been made to tease out links between education statistics (for instance) and gender ratios, or hygiene and women's health. This is an opportunity lost - the study could have benefitted from a more refined analysis of its data, and a pursual of hypotheses beyond the 'obvious.'


The TSMG doesn't claim to be a research body. But publishing results, seeking backing from various worthies, and presenting a report to the government must go hand in hand with some amount of reflection and a responsible use of data. Both of which seem to have been missing here, along with any kind of consideration for depth, complexity, and nuance. What were they thinking? Not very much.

No comments:

 
Creative Commons License
This work by ToruJ is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.