Poverty and Crime: Evidence from Rainfall and Trade Shocks in India

Lakshmi Iyer, Petia Topalova


This paper examines the mechanisms underlying the observed relationship between rainfall and crime. Using four decades of district level data from India, the study establishes a robust effect of rainfall shocks on different types of crime, with the strongest effects on violent crimes and property crimes. It then examines to what extent poverty is the main causal pathway between rainfall shocks and crime.

The study identifies trade liberalisation as an additional source of exogenous income shocks for households in rural India, independent of the amount of rainfall. The larger the loss in trade protection a district experienced, the higher is the incidence of these crimes. The similarity in patterns of how criminal behaviour responds to two very different sources of variation in poverty and income suggests that the income channel is the most relevant mechanism behind the observed rainfall-crime relationship.

The study concludes by examining whether policy measures aiming to weaken the rainfall-income relationship also reduce the rainfall-crime relationship. It focuses on two policy measures in India: 1) the nationwide workfare program created by the National Rural Employment Guarantee Act (NREGA), which guarantees a hundred days of minimum-wage employment to every rural household; and 2) the construction of dams.

Key Findings:

  • Trade shocks, previously shown to raise relative poverty, also increased the incidence of violent crimes and property crimes. The relationship between trade shocks and crime is similar to the observed relationship between rainfall shocks and crime. Thus, the study identifies a causal effect of poverty on crime.
  • Violent crimes and property crimes rise during periods of low rainfall and/or higher exposure to foreign competition, while other crime categories such as crimes against women do not show a strong relationship with either of these exogenous income shifters.
  • Trade liberalisation affects the consumption of households at the lower deciles of the income distribution, while rainfall shocks affect consumption over the whole range. Temperature variations have a significant effect on some types of crime, including crimes against women, even though per capita consumption or poverty is not much affected by temperature.
  • Policies such as dam construction or workfare programs do not appear to alleviate consumption in the face of weather shocks to a large enough extent to have an impact on crime.


Iyer, L. and Topalova, P. (2014). Poverty and Crime: Evidence from Rainfall and Trade Shocks in India. Working Paper No. 14-067. Harvard Business School BGIE Unit.