Conventional wisdom holds that 2020 was a paradoxical year for public safety. On the one hand, homicides in the U.S. rose by nearly 30 percent, the largest single-year increase on record. On the other hand, total crimes fell about 15 percent from 2019. These seemingly contradictory data created a choose-your-own-adventure story for politicians and pundits, with those on the right emphasizing the rise in homicides and those on the left emphasizing the overall crime decline.
This understanding of public safety, based on an analysis of changes in the crime rate, is incomplete. In a new paper with Maxim Massenkoff, I show that the crime rate isn’t a useful statistic to understand what happened to public safety in 2020. A more complete analysis shows that Americans indeed faced a greater risk of crime while in public that year.
Beginning in March 2020, violent street crimes—assaults and robberies occurring in public or commercial spaces—declined by approximately 35 percent. These offenses remained 15 percent below 2019 levels throughout the summer, before returning to 2019 levels in the fall. Yet the way people spent their time changed dramatically with the Covid-19 outbreak. To mitigate disease risk and comply with mandatory lockdowns, we spent more time at home. According to the American Time Use Survey, an annual measure that tracks changes in how and where Americans spend their time, people spent 50 percent less time away from home in the early days of the pandemic than they did during the same months in 2019. While Americans eventually began to return to normal activities, even by the end of the year, they spent about 20 percent less time outside than they did in 2019.
How risky was it to spend time in public? Using official crime data from large cities and citizen surveys, as well as cell-phone location data, we studied whether the number of crimes changed in 2020 per hour spent in public. While traditional crime data show discrete drops in offending during the post-pandemic lockdown period, our analysis indicates that people faced a higher risk when they ventured outside. In 2020, the risk of outdoor street crimes initially rose by more than 40 percent and remained consistently between 10 percent and 15 percent higher than it had been in 2019. The overall crime rate might have fallen, but people were more likely to be robbed or assaulted while spending time in public.
This conclusion withstands several objections. First, one might suggest that the reporting of crimes to law enforcement changed in 2020, owing either to fears of Covid-19 exposure or to distrust in police after the murder of George Floyd. But data from the 2020 National Crime Victimization Survey, a nationally representative survey of American residents, supplied little evidence that crime reporting changed appreciably.
Second, what if the mix of people spending time outside in 2020 changed? If people participating in outdoor activities in 2020 were from groups with a higher baseline risk of victimization, the increased crime risk could have been an artifact of that change in who spent the most time in public. Yet we found reason to doubt this possibility, too. Given that race, gender, and age are powerful predictors of crime victimization, we considered whether higher-risk demographic groups were differentially likely to spend time outside their homes in 2020. The answer is no—if anything, the share of time spent outdoors by older people increased in 2020.
Finally, recognizing the importance of place in predicting crime risk, we considered whether the risk of street crime rose within a given neighborhood. If people from some places spent more time outdoors than people from other places—for example, if individuals living in higher-poverty communities were less able to shelter at home—then the change in risk could have been an artifact of between-community differences in victimization risk. But even within a given neighborhood, risk of street crime victimization increased markedly in 2020.
Though the risk of street crime victimization rose by more than 15 percent, risk remained low in absolute terms. In 2019, there were just over 9 such crimes for every 1 million hours spent in public. In 2020, the risk had risen to approximately 11.5 per 1 million hours. While the causes of the increased risk remain mysterious, as we continue to refine our understanding of trends in public safety, it is important that we account for changes in criminal opportunities. The narrative that the murder spike was an exception to an otherwise safer year in 2020 is misleading. We weren’t safer in 2020—we just weren’t measuring crime risk very well.
Photo by KENA BETANCUR/AFP via Getty Images