Agent-based simulation of the impacts of alcohol outlet zoning policy on police encounters and violent crimes - a case study of Baltimore city, Maryland

Abstract

Greater alcohol outlet density contributes to higher rates of violent crimes, especially in American cities. Alcohol outlet zoning policy, such as zoning codes to reduce the number of off-premise alcohol outlets (e.g., liquor stores), has been suggested to reduce violent crimes and improve public health. But such an argument was mainly derived from the outlet-density research that explored the impact of increasing alcohol outlet density (AOD) on alcohol-related outcomes. Few studies have examined the effectiveness of limiting AOD in reducing violent crimes. Besides, the policy-affected neighborhoods are often the low-income urban communities of color where the residents rely on the non-alcoholic services of liquor stores (e.g., food, ATMs) and prefer to hang out and gather near these stores. But there is still a lack of quantitative studies revealing whether the alcohol outlet zoning policy could lead to unanticipated negative impacts on these poor communities; if so, which factors are critical in contributing to these negative consequences. The present study built a spatially explicit agent-based modeling (ABM) to simulate the intended and unintended consequences of alcohol outlet zoning policy on low-income urban communities of color. This model allows us to incorporate interconnected neighborhood factors linking alcohol outlets and crimes and test different hypotheses through what-if scenarios. We selected a predominantly Black neighborhood in Baltimore city with a high crime rate and AOD as our study area, where the liquor stores in residential areas were expected to close due to the city’s zoning code rewrite. We simulated the daily-life routine activities of resident agents and their interactions with other resident agents and police agents, which lead to the emergence of violent crimes in the neighborhood. The model was built upon well-known criminological theories and validated with numerous first-hand and second-­hand data, including national and local surveys and government datasets. Results show that the closures of non-conforming liquor stores were not effective in reducing the residents’ alcohol use in the neighborhood, with the reduction of 5% of alcohol drinkers by closing 57% of liquor stores. The zoning policy would increase illicit drug users because the residents with alcohol use disorder would turn to illicit drug use for self-medication, which could trigger more drug-related violent crimes. But the zoning policy would not increase the environmental inequity in accessing food and ATM services because the residents can easily find alternative corner stores with these services within 5-minute walking distance. Although the policy will reduce alcohol consumption in communities, it would unintendedly increase alcohol-related crimes because more drinkers purchase at and loiter around the remaining conforming liquor stores, imposing other residents at a high risk of being victimized. Furthermore, the policy would lead to over-concentrations of purchasers and loiterers at the remaining corner stores, which would attract more racialized police-citizen encounters and police use of force. City officers are suggested to launch companion interventions, such as providing mental health treatments and rehabilitations and monitoring the police activity in the policy-affected neighborhoods, for mitigating the negative impacts of limiting liquor-store density in socially disadvantaged communities.