We use a panel dataset of around 3500 rural households from Southeast Asia and investigate evidence on crime victimization. More concretely, we ask (1) to what extent are rural people affected by crime? (2) What factors determine rural crime victimization? And (3) what are the impacts of crime victimization on welfare of rural households? We use the routine activity approach as the theoretical framework and apply different logit models to identify determinants of crime victimization. We find that 5.46% of the rural households have been victimized, mainly by theft, over the last 12 months, some of them even more than once. Living in a rural region with higher levels of inequality is positively correlated with the likelihood of theft victimization. Households with higher levels of crop commercialization are associated with a higher victimization risk, while households with more livestock and being more specialized in specific livestock species are associated with a lower risk. Moreover, past victimization and exposure to weather shocks are positively associated with the likelihood of being affected by crime. With respect to the impacts, we use the heteroscedasticity-based instrumental variable approach to account for endogeneity and find highly significant negative effects of agricultural theft victimization on food consumption and child health outcomes. We conclude that rural crime requires attention although overall incidence is low in rural Thailand and Vietnam. Guardianship should be promoted, especially in times of weather shocks. Finally, reducing inequality helps preventing theft.

我们使用东南亚约 3500 个农村家庭的面板数据集，调查犯罪受害的证据。更具体地说，我们要问（1）农村人口在多大程度上受到犯罪的影响？ （2）农村犯罪受害由哪些因素决定？ （3）犯罪受害对农村家庭福利有何影响？我们使用日常活动方法作为理论框架，并应用不同的逻辑模型来识别犯罪受害的决定因素。我们发现，有5.46%的农户在过去12个月内遭受过盗窃，其中一些甚至不止一次。生活在不平等程度较高的农村地区与盗窃受害的可能性呈正相关。农作物商业化水平较高的家庭受害风险较高，而拥有更多牲畜且更专业化特定牲畜品种的家庭受害风险较低。此外，过去的受害经历和遭受天气冲击与受犯罪影响的可能性呈正相关。关于影响，我们使用基于异方差的工具变量方法来解释内生性，并发现农业盗窃受害对食品消费和儿童健康结果产生非常显着的负面影响。我们的结论是，尽管泰国和越南农村地区的总体犯罪率较低，但农村犯罪仍需引起关注。应加强监护，特别是在天气冲击时期。最后，减少不平等有助于防止盗窃。