By Geetika Dang, Vani S. Kulkarni and Raghav Gaiha

One in three women throughout the world experience physical and/or sexual violence from a partner. In South Asia, the figure rises to nearly 40%.

In India, the incidence of serious crimes against women, including rape, kidnapping and abduction, dowry deaths, and cruelty by husbands and relatives, rose by about a third between 2001 and 2015, according to National Crime Records Bureau reports. Of the 313 crimes committed against women in India each day in 2015, around 30% were instances of rape (including the intent to rape).

Yet, the conviction rate for crimes against women is low, currently at 21%. The high and rising number of crimes committed against women combined with the fact that most of the perpetrators remain free, shows that women in India are very vulnerable to serious violent and sexual crime. However, the vulnerability of women varies enormously across states.

Our recent GDI Working Paper focuses on two related questions:

1. What are the factors associated with huge inter-state variation in serious crimes against women (CAW) in 2015?
2. Why have CAW risen between 2001 and 2015?

As answers to these questions lie in the interplay of the affluence of a state, religion, demographics (including the female/male ratio), employment opportunities for women, their literacy, the rural/urban population ratio, quality of governance in the state, and media exposure, we carried out a detailed analysis that allows us to assess the individual and joint contribution of each factor to crimes against women generally, and rape specifically, over time and across states.

These were our findings:

• Crimes against women are more common in affluent states and less common where there is a high female to male ratio. This explains why states such as Delhi and Haryana perform so poorly: they are both affluent and have low female to male ratios.
• The ratio of females to males is generally a key determinant: the more women there are relative to men, the lower is the incidence of both CAW and rape. Scarcity of women relative to men results in a squeeze in the marriage market and consequent rise in crimes against women.
• Our results on the incidence of CAW and rape show that female literacy either has a significant positive effect (increasing the occurrence of CAW) or a non-significant effect (for instances of rape) while the female labour force participation rate has a negligible negative effect (on both incidence of CAW and rape). This may seem counterintuitive at first, as improving a woman’s educational and employment options typically strengthens her bargaining power and assertions of autonomy, resulting in a decrease in domestic violence. However, for women who currently have a very low level of bargaining power, education and employment may lead to weak or even counterproductive results: evidence suggests, for example, that a woman gaining employment while their spouse is unemployed may cause tension and domestic violence.
• While crimes against women may occur across different locations and cultures, their forms and frequency vary. This proposition is corroborated by our analyses: there are significant differences between rural and urban populations, and between Hindus and Muslims.
• Alcoholism increases the occurrence of rape despite the ban on the sale of liquor in certain states (eg Gujarat, Bihar, Manipur and Nagaland). Yet, the effect of alcoholism on the occurrence of rape is very large, indicating weak enforcement of prohibition.
• Communities that are more exposed to the media are more likely to report crimes and those crimes are more likely to be deterred in the first place. It is difficult to separate the two, so the combined effect is captured in our analysis, though our results show that better reporting effect dominates the crime deterrence effect.
• Women in India are typically unwilling to report incidents of rape or attempted rape. Sometimes this is because they are unsure whether the harm inflicted reached the level of a criminal act, or because they were able to fight the perpetrator off. Many also doubt the integrity of the system to which they have to report, given such a low conviction rate. A credible judicial system with a higher conviction rate would encourage women to report crimes committed against them, but also deter those crimes from occurring in the first place. In our analysis, we found that for both incidence of CAW and rape in India, the deterrence effect of convictions for rape is positive but negligible. This is not surprising as there are long delays in securing a conviction against heavy odds, including primitive medical tests, falsification of evidence and intimidation.
• Governance makes a difference in the long term, especially the enforcement of law and order and legal provisions for protection of women against violence. We found a quadratic relation between the incidence of rape and quality of governance – an inverted U relationship in which the incidence of rape first rises with improvement in governance, and then declines. As the judicial and police systems tend to be abysmal in several states and intimidation of victims is rampant, slight improvements fail to deter perpetrators of such crimes. It is only after a certain threshold of governance is reached, that it begins to be effective.

As affluence in Indian states grows, there is likely to be a slight rise in crimes against women, but as the female to male ratio has already begun to rise, it is in turn likely to dampen the rise in these crimes. The efficiency of the judicial and police systems may be the deciding factor – and unfortunately, going by recent evidence, it may be optimistic to surmise significant improvements in these systems in the near future.

Note: This article gives the views of the author/academic featured and does not represent the views of the Global Development Institute as a whole.