Now Reading
Data Offer Clues on San Francisco’s Crime Paradox. Police Chief and Academics Have Some Ideas Too
Sunday, August 07, 2022

Data Offer Clues on San Francisco’s Crime Paradox. Police Chief and Academics Have Some Ideas Too

There is plenty of evidence that many San Francisco residents believe the city has grown less safe in recent years, and especially during the pandemic.

Eight out of ten San Franciscans believe crime has worsened in recent years, according to one poll released by the San Francisco Chamber of Commerce this summer. Viral videos of shoplifters and other crime stories have spread fast on social media, feeding perceptions of a city out of control. Recently, one Marina neighborhood went so far as to hire private security to patrol its streets. 

District Attorney Chesa Boudin is now facing a likely recall election, at least in part because of growing worries about crime.

Boudin and other public officials, including Mayor London Breed and Police Chief Bill Scott, have all acknowledged that many residents feel less safe.

But they have pushed back on the idea that the city is experiencing an unprecedented crime wave. City data shows reported crime was down about 20% during 2020, though this year’s numbers show a mixed picture. In San Francisco, as in much of the country, overall crime rates remain well below their 1990s levels.

We asked Maxim Massenkoff, an economist at the Naval Postgraduate School, to look at the data for clues to the contradiction. His findings suggest at least part of the answer to the crime paradox: because so many fewer people were out and about during the pandemic, the chances of being a victim of a crime increased even as the overall crime rate decreased. 

Massenkoff used anonymized GPS cell phone data from Safegraph to measure foot traffic in each of San Francisco’s roughly 600 census blocks. The data showed that the number of violent crimes rose from 0.9 per 10,000 people to 1.2 per 10,000 people—a jump of 34%. On Muni, where the number of boardings and the number of crime incidents are both reported, violent crimes per rider nearly tripled.

And when it comes to car crime, parking meter data makes it possible to see that on fully metered streets, the chances of having your car stolen rose 26% in 2020 and 52% so far in 2021 from 2019 levels.

Read Massenkoff’s column here.

Mikaela Rabinowitz, director of data, research and analytics at the District Attorney’s office, said the research was flawed. In particular, she objected to the use of relative rate increases instead of absolute numerical increases. “For example, in looking at crime on Muni, the report notes that violent crime increased from .9 incidents per 10,000 people to 1.2 incidents per 10,000 people; but instead of accurately describing that as an absolute rate increase of .3, the report describes the rate at which the rate increased, which was 30%,” she said in a statement to Here/Say.

She also objected to what she said were inaccurate assumptions and small sample sizes, noting that the discussion of car crime was based on “a sample of 1% of motor vehicle thefts across 10% of city streets, a statistically invalid sample from which to infer citywide trends.”

SFPD Chief Bill Scott acknowledges that the police department has a duty to address the public’s fears of crime, “regardless of what the statistics say.

The fear of crime is a real thing that we have to deal with,” he told us during a ridealong interview on National Night Out. I don’t care how those statistics slice up because if people fear that crime is happening, it’s a real problem.”

Scott believes that viral videos are a major contributor to the perception that San Francisco is dangerous, but he also notes that the underreporting of crimes could account for crime statistics not matching up with citizens’ lived reality. 

See Also

See our interview with Chief Scott below.

Video by Jesse Rogala

Academic experts point to some other factors that could contribute to fear of crime. Deserted streets and empty trains and buses can make crime more noticeable, for one. Now we’ve got two people on the street. One of them is stealing handbags. It becomes very obvious,” says sociologist Sarah Lageson, an associate professor at Rutgers University Newark School of Criminal Justice who studies law, privacy, technology and surveillance.

Social media apps including Nextdoor and Citizen amplify skewed perceptions of crime with constant notifications about incidents, as do viral videos of shoplifting and other crimes, Lageson and others say. The generalized anxiety created by the pandemic is also a factor.

There are plenty of questions about the city’s underlying data too. Supervisor Catherine Stefani, a former prosecutor, believes that the way that crime data is analyzed by city departments and communicated to the public needs to change. For instance, she believes a per capita crime rate based on the number of people in the city (rather than the resident population) would more accurately capture the city’s crime rate by accounting for the decline in tourists and commuters during the pandemic. 

She also observes that “a lot of what people perceive as crime isn’t captured in the crime rate.Drug overdoses, for example, aren’t included in the aggregate crime rate, yet those deaths are likely linked to the criminal activity of drug dealers and illicit drug sales. Additionally, crimes like the spate of violent attacks against elderly Asian Americans that horrified San Franciscans and caught national attention earlier this year, might be lumped in with assaults rather than documented as hate crimes. 

In addition to more nuanced data crunching, Stefani believes that city leaders shouldn’t be afraid to have real talk with citizens about crime. “When you tell people in San Francisco, ‘crime is down,’ you completely insult what they’re experiencing,” she says. “There is a lot more we should be doing.” 

Christina Campodonico can be reached at [email protected].
Jonathan Weber can be reached at [email protected].
  • I don’t understand the response of the DA’s office. On Muni, whether it’s the rate of the rate or an absolute value, it is still evidence that Muni riders were more at risk of experiencing a crime compared to previous years. I voted for Chesa but it makes him/his office seem out of touch with reality when they miss the bigger picture (to me, out of re-election concerns). I think it is doing a disservice to the broader progressive DA movement.

    • I suspect that the broader progressive DA movement is in for some hard times ahead. Boudin has overplayed his hand, as city residents (like all people) want to feel safe. Boudin is on the record championing the out-of-town drug dealers in the TL by blaming our broader economic system that drives them to these life choices. Neat idea for the classroom, but not reality based.

      There have been bad people, people who choose criminality, since our species started walking the earth. One cannot justify the actions of every single one of them. Boudin really makes a heartfelt attempt at justifying all their actions, and he projects a seemingly vicious arrogance towards actual victims of crime.

      I’m a progressive. Boudin is a radical with a worldview axe to grind. There’s a difference. SF voters made (yet another) mistake, although by an incredibly thin margin. Not a fan of CA recall processes but I support this one wholeheartedly.

  • SF residents and businesses have been trained not to report crime because it is useless. Actual crime is different from reported crime, even if Chesa would like to bury his gap-toothed head in the sand about it.

  • Interesting ideas: crime is easier when the streets are largely deserted. But let’s not forget Chesa’s role in this who managed to become DA with a desire to avenge the just punishments of his murderous parents.

  • For me and my friends, the increase in crime is real, not simply perception. In the last 18 months, two of my friends friends have been physically attacked, and my partner’s car has been broken into twice. A news crew a half block from my house was robbed of their equipment in broad daylight. The viral videos only add to what I already know to be true. The increase in crime is real and unprecedented compared to any other time in the last 20 years.

  • I’m very glad to see someone analyze the data on a per-person basis instead of a per-city basis. What most of us care about is how likely is an individual to be a victim of a crime.

    A quick calculation suggests that the number of people in San Francisco fell by as much as 40% during the pandemic. So, if crime rates on the per-city basis stay the same or even fall by 20%, this is still a big jump in crime rates per person.

    Before the pandemic, SF had 25.8 million visitors per year. (If each stayed 1–2 days, this is 100,000 people per day.) Also, pre-pandemic, 700,000 people worked in SF; many commuted in from outside SF. This number dropped a lot because all offices and businesses shut down. There is also an estimate that the resident population reduced by 100,000 during lockdown as people left the city, temporarily or permanently.

Leave a Reply

Your email address will not be published.