Original title: Beyond the Dark Figure: A Causal and Probabilistic Analysis of Crimes Against Women in India
Rudrani Ghosh · September 2026
Police crime data in India can be misleading because people do not report every crime, and district populations vary wildly. To fix this, researchers linked two decades of police data with 2011 Census details. They used causal tools and statistical models to find the real patterns behind violence against women. First, violence does not bunch up in specific regional clusters; it spreads randomly across the nation. Second, broad state policies matter very little, accounting for less than 1% of local crime rate differences. Third, city growth, rather than the gap between male and female literacy, is the main force driving higher reporting. Finally, a major 2013 legal reform caused an immediate jump of 88,879 reported cases per year. This sharp rise reflected changes in administrative reporting rules rather than a sudden wave of new crimes.
Imagine trying to count how many apples fall from trees in a town. You only have reports from people who bother to bring them to the town hall, and some neighborhoods rarely bring apples even if they fall. Also, after a big storm, many more apples are reported simply because people notice the storm, not because more fell. In this paper, the 'apples' are crimes against women, the 'town hall' is police records, and the 'storm' is a new law that changes how many are reported. The researchers built models to separate the true falling rate from the reporting rate.
Policymakers rely on crime data to allocate resources, but if reporting biases distort the numbers, they might target the wrong places. This study shows that urbanization, not actual violence, drives reporting, and that national laws create huge reporting shocks that look like crime waves. Without these corrections, spending and prevention efforts could be misdirected for decades.
The abstract does not explicitly state limitations. The analysis depends on NCRB and Census data, so it may miss unreported crimes and recent demographic changes. Causal claims rely on assumptions of the do-calculus and DAGs. The findings about the 2013 law as a reporting shock are specific to that context.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Beyond the Dark Figure: A Causal and Probabilistic Analysis of Crimes Against Women in India — Rudrani Ghosh — SocArXiv — doi:10.31235/osf.io/4fxsm_v2
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.