From Boston to Los Angeles, “mixed use” development, combining residential and commercial properties, is on the rise. The benefits that have been cited for colocating housing and retail establishments include reduced travel distances, more-pedestrian-friendly neighborhoods, and stronger local character. Recent research suggests another important potential benefit: Retail establishments may play an important role in crime prevention.
In our research, we examined the effect on crime of temporarily shuttering two types of retail businesses: medical marijuana dispensaries (MMDs) and restaurants. Why study dispensaries? In 2010 Los Angeles initiated a mass closing of two-thirds of the dispensaries in the city. The fact that the closings were based on a very arbitrary registration process that took place several years prior allowed us to use the closings as a natural experiment to estimate the causal impact of MMDs on crime.
Surprisingly, we discovered that the closures were associated with a significant increase in crime
the blocks immediately surrounding a closed dispensary, compared with the blocks around dispensaries allowed to remain open. Our results demonstrated that the dispensaries were not the crime magnets that they were often described as, but instead reduced crime in their immediate vicinity. And when breaking down the effect by types of crime, we found that the increases in crime after dispensary closures were driven by the types of crime most plausibly deterred by bystanders: property crime and theft from vehicles.
But our interest was in the effect of retail businesses on crime in general, not just in what happens when MMDs closed. We wondered: Would we observe the same dynamic in another retail context?
We then examined the impact of temporary restaurant closures by the Los Angeles County Department of Public Health for public health code violations. When a health inspector finds a violation that poses an imminent health hazard, the inspector immediately shuts down the restaurant. The restaurant must remain closed until a subsequent follow-up inspection confirms that the situation has been resolved. When we examined crime patterns around these closures, we found essentially the same pattern as we did with MMDs: The area immediately around a closed restaurant experienced an increase in property crime and theft from vehicles, relative to areas around restaurants that were either recently reopened or about to be closed. Furthermore, this increase in crime disappeared as soon as the restaurant reopened.
Given the differences in the nature of these establishments and the reason for and timing of their closures, we were left wondering what common factor might drive the similarity in results. One key factor common to retail establishments, whether MMDs or restaurants, is that they generate foot traffic. And with foot traffic comes informal surveillance.
As Jane Jacobs described in her groundbreaking 1961 work, The Death and Life of Great American Cities, people provide a natural form of incidental surveillance that can increase public safety. This idea, which Jacobs called “eyes upon the street,” has proven enormously influential, and is now a cornerstone of modern urban planning. But while it seems intuitive that criminals may be less likely to commit some forms of dark-alley crimes in front of an audience, the relationship between retail businesses and crime is more complex. Not only are retail customers (and the shops they frequent) potential crime targets, but they may also be perpetrators of crime. For that reason, the impact of retail business on crime is theoretically ambiguous — it could go either way — and until now there’s been virtually no rigorous empirical evidence.
We were able to probe this relationship in the context of restaurant and dispensary closures, offering new insights into the received wisdom. Specifically, to test whether a reduction in eyes upon the street could be the common crime-generating mechanism behind our results, we collected “walk scores” for each establishment in our sample. Walk scores are a measure of an area’s walkability as determined by the number of nearby restaurants, coffee shops, grocery stores, and other features that generate foot traffic. A business with a high walk score is located near many other businesses. Thus the closure of a business in a high-walk-score area should have a very limited impact on local foot traffic. On the other hand, the closure of a business in a low-walk-score area should have a proportionally large impact on total foot traffic. If our results were driven by eyes upon the street, we should find that, all else equal, crime is negatively related to walk scores. This is indeed the pattern we found: The increase in crime associated with business closures was stronger in neighborhoods with less walkability and fewer other businesses around.
A quick, back-of-the-envelope cost calculation using our results and crime costs from a 2010 study suggests that an open retail business provides over $30,000 a year in social benefit just in terms of larcenies prevented — something to keep in mind the next time you are deciding how much to tip at your favorite local restaurant, coffee shop, or bakery. It’s something urban planners should keep in mind as well when zoning neighborhoods. Retail businesses draw customers, and in doing so, lower crime.