Law enforcement patrol operations are based to a large degree on habit. Criminals tend to operate in zones where they feel comfortable, where the rewards outweigh the risks. The well-established criminological theory of repeat and near-repeat behavior confirms this across different crime types, from burglaries to vehicle crimes to violent crime.
Officers tend to fall into patrol habits as well, whether intentionally or unintentionally. Intentional habits include patrolling areas that are perceived to be crime hotspots based on past experience or the analysis of crime data. These habits can also include officer judgment based on his or her knowledge and experience of the patrol area. Unintentional habits are harder to self-identify. Humans fall into patterns of behavior throughout their daily lives: the route you take to drive to work, where you get coffee or gas, how and when you exercise.
The challenge for law enforcement is that someone intending to engage in criminal activity can recognize these patterns and act accordingly. As Sheldon F. Greenberg notes in his book “Frontline Policing in the 21st Century” (2017):
“Falling into repeated patterns and practices has always been among a patrol officer’s greatest weaknesses and threats. Repeated patterns are easily spotted by others and can be used to their advantage.”
This was one of the founding principles of PredPol: using large datasets of historical crime data allows officers to detect and get ahead of future patterns of criminal behavior. They can proactively patrol those areas and prevent crimes before they occur.
How much does the current outbreak of Covid-19 change all this?
Just one week into shelter in place provisions, we are already starting to see changes in long-term historical crime patterns (note: this is still a small time sample of data). In PredPol cities that are subject to shelter-in-place orders, we are seeing – as you would expect – a decline in residential burglaries, as more people stay at home. Commercial burglaries, on the other hand, appear to be increasing. Traffic collisions are way down, since fewer people are driving. Vehicle crime rates appear to be declining somewhat, but the places and times where they are occurring seem to be moving away from traditional hotspots like shopping center parking lots, and the volume appears to be shifting away from daytime to evening and nighttime.
This is where PredPol comes in really handy. The Covid-19 virus will change many things in our world before we have found a vaccine or cure for it. It may change how and where we work and go to school. It may change how some services, such as healthcare, are delivered. But it will also change the patterns of criminal behavior. Typical times of different crime types will change. Locations will change. The mix of crime types will change. It will be impossible without advanced analytics, it will be almost impossible to pick up those changes in real time. Using PredPol’s machine learning algorithm, however, departments will be able to recognize and react to those new patterns, and change their patrol patterns as well.