The removal of cash from gas stations in Montevideo at night had a significant impact on crime, according to the research paper "Cash and Crime," published in May 2019.
The document was prepared by Dr. Nestor Gandelman—academic coordinator of the Bachelor’s Degree in Economics and professor of Principles of Economics at Universidad ORT Uruguay; Dr. Ignacio Munyo—professor of Economics and director of the Grant Thornton Center for Economics at IEEM, the business school of the University of Montevideo—; and Emmanuel Schertz—a graduate of ORT’s Bachelor’s Degree in Economics and a research fellow in the Opportunity Insights program at Harvard University.
The research was presented at various conferences, including the European Economic Association conference, held in Manchester, and the Workshop on Crime organized by the Research Institute for Development, Growth and Economics (RIDGE), EAFIT University, and the Latin American and Caribbean Economic Association (LACEA), which took place in Medellín, Colombia.
Gandelman described the process of drafting the report "Cash and Crime" and outlined its main contributions.
What is the article about?
The article seeks to estimate the effect of eliminating cash on crime. To this end, it examines the policy of removing cash from gas stations at night.
On average, gas stations generate between $10,000 and $15,000 in sales per day. Before this policy took effect on May 15, 2016, approximately 30% of that amount was paid in cash.
The expected outcome was that crime would decrease for two reasons. First, because if there is less to steal at gas stations, fewer people will be tempted to steal it. A second reason has to do with what happens around the gas station. Whoever was trying to rob the gas station would also rob the person driving by, the woman walking by, the teenager on his way to school—so it’s not just the crime at the gas station itself, but potentially what’s happening near the gas station, in the surrounding area.
What research methodology did you use to write the article?
To conduct this study, we used georeferenced data on reported crimes from the Ministry of the Interior and the geographic locations of gas stations (and other businesses) from Google Maps. This allowed us to determine how far from the gas station each crime was committed.
We examined the period one year before and one year after the date this policy was implemented. We used treatment groups and control groups—that is, groups where the policy was implemented and groups where it was not. The removal of cash was limited to gas stations at night, which left the rest of the city as potential control groups. The study technically identified the most appropriate control groups and observed their evolution over time.
For example, one of the control groups included in the study consisted of all supermarkets (154) and pharmacies (159) in Montevideo, as well as other locations in the city where this policy was not implemented.
We use a 50-meter radius around each gas station and checkpoint. That is the area where we measure crime rates.
What are the main conclusions?
By observing the control groups, it was determined that crime at gas stations during the night was kept in check.
According to one control group, the elimination of cash is associated with a 45% reduction in crime compared to what would otherwise have occurred. In the other control group we used, the reduction is around 30%.
If this measure had not been implemented, crime at gas stations would have been 30 to 45 percent higher than it actually was.
It is important not to confuse our findings with overall crime trends. The reduction in crime associated with the elimination of cash refers to the “counterfactual” (the crime rate that would have occurred in the absence of the policy).
At the same time, in conducting the study, we used data on both robberies and domestic violence. Why? Because both are violent crimes, except that one is financially motivated and the other is not. Domestic violence serves as a control group in the research design. Indeed, we observed—unlike what happened with robberies—that the removal of cash had no impact in the areas where this occurred.
Furthermore, we found no evidence of temporal or geographic displacement effects. In other words, the elimination of cash did not cause robberies to shift from nighttime to daytime or from the area near the gas station to a more distant location.
Why is it important that there was no local displacement? If the policy led to a 30–45% reduction in crime at gas stations and there was no temporary or geographic displacement, then overall, this policy was successful in combating crime.