Determinants of Crime in India: A District-Level Analysis

This paper examines the district-level socioeconomic factors of reported crime in India using data from the NCRB and the Census of India. Using a cross-sectional sample of 481 districts across 17 major states, the study treats female-to-male ratio and inequality as main variables, while population density, work participation rate, poverty, literacy and caste composition are included as controls. While OLS estimates are presented as baseline correlations, the econometric contribution is an IV framework addressing endogeneity in gender composition. 2SLS results show that female-to-male ratio remains significantly negatively correlated with murder rate. For theft rate, female-to-male ratio remains negatively associated with reported crime in the baseline OLS specification and in the IV robustness analysis. The study contributes by combining a large Indian district-level dataset with an IV-based empirical framework. Quantile regression results show that gender ratio is negatively correlated with several reported crime types, while inequality is most consistently correlated with theft and total IPC crime, especially at higher quantiles. This study contributes to economic and social measurement by showing how reported crime is correlated with district-level socioeconomic and demographic vulnerability. A supplementary panel analysis of 27 Global South countries during 2014–2023 provides secondary contextual evidence.