Crime and Policing in San Francisco
By Peter Elden
Shortly after I began researching crime trends and police response times for this article, the topic became personal. Late one night in January, a disturbed and aggressive individual was overturning trash bins and throwing garbage into the street in front of my condo. When I asked him to stop, he started throwing rocks at my window. He needed help, and he needed to be restrained from causing damage to property and hurting people.
I called the police and explained the situation. The dispatcher said that officers would respond as soon as possible but, unbeknownst to me, many officers were at that moment pursuing a violent criminal in a fleeing vehicle. I waited in my living room, hoping the police would show up and keeping an eye out for the rock-thrower. Eventually, I fell asleep on the couch until an officer arrived at 5 AM, five-and-a-half hours after the incident occurred.
The officer apologized for taking so long and told me about the car chase. He explained that SFPD is short-staffed and they have to make tough decisions about which calls can wait. He asked whether I wanted to file a police report. I declined. Filing a report was going to keep an officer off the streets for at least another hour and I was pretty sure that nothing would come of it.
The research I later completed showed me that my story is a common one, both in terms of the criminal incident I experienced as well as the response to it. After several months of exploration and analysis, I identified seven key insights on crime and policing that should be heeded by anyone interested in making our city safer:
Property crime is sky high in San Francisco
Violent crime rates in San Francisco are similar to those in other large U.S. cities
SFPD response times are increasing
SFPD is solving fewer crimes than before
San Franciscans don’t feel safe in their city
SFPD is short hundreds of officers
SFPD is spending less and less time on proactive policing
Let’s dive into these insights one by one.
Note: The data in the following charts comes from a combination of sources, including the FBI Crime Data Explorer website, US Census Bureau, New York City Police Department, Los Angeles Police Department, the San Francisco Police Department, NeighborhoodScout, the San Francisco Standard, and the Board of Supervisors Budget and Legislative Analyst. In the charts, we compare San Francisco to the 17 largest cities in the country as well as five cities that are similar by virtue of size (Charlotte), being similarly dense (New York), having a robust technology industry (Austin), or being located in California (Los Angeles and San Diego). Comparisons are provided through 2021 as the FBI has not released data for 2022 yet. 2021 crime detail data for Los Angeles was not available as of the publication date. The crime rate charts show reported crimes per 100,000 people.
Insight 1: Property crime is sky high in San Francisco
About one property crime occurs per 20 residents of San Francisco per year. This rate is more than double the average of America’s 17 largest cities and five times higher than New York City’s. Our high rate of motor vehicle theft is also notable, given that San Francisco is among the ten American cities with the fewest number of cars per household.
Our actual property crime rate is likely much higher due to underreporting (the likely causes of which are discussed in An Insider’s View of the SFPD and A Shopkeeper’s Experience of Crime, elsewhere in this issue). Former SFPD officer (now citizen journalist) Lou Barberini has shown, for example, that if we follow the naive assumption that incident reports track actual incidents, we’d be forced to conclude that there are only six shoplifting incidents per day in the entire city of San Francisco.
Another reason police department data may undercount property crimes is the Hierarchy Rule, according to which only the most serious crime in a multiple-offense incident is counted. The FBI requires local law enforcement agencies to collect and report incident data according to this rule. Therefore, if a person commits aggravated assault during a motor vehicle theft, only the assault is counted. The impact of the Hierarchy Rule across cities is difficult to determine, but it’s reasonable to assume that it depresses property crime counts more in cities with more violent crimes, especially violent crimes like robbery that typically are accompanied by associated “lesser” infractions.
Note: Burglary involves unlawful entry into a structure, like a home, in order to commit a felony or theft. Larceny is defined as the unlawful taking of another’s property, like a purse or iPhone. Theft of property from a car, or of car parts – rather than the car itself – is categorized under larceny, not motor vehicle theft.
Insight 2: Violent crime rates in San Francisco are similar to those in other large U.S. cities
While property crimes make up 90 percent of the total crimes committed in San Francisco, violent crimes, including murder, robbery, and aggravated assault, are also present in our city.
Robbery is more common in San Francisco than in comparable cities, but we see fewer homicides than elsewhere.
Comparing actual exposure to violent crime across cities is challenging due to a number of confounding factors beyond simple differences in population. For example, Charlotte’s violent crime rate in recent years has been nearly twice that of San Francisco, but Charlotte measures over 300 square miles, while San Francisco is about one-sixth the size. San Francisco’s relative density means that the proportion of its inhabitants who are regularly exposed to both property and violent crime is higher than in cities where crime is localized in “bad parts of town.” Greater reliance on public transportation among the populace also contributes to more exposure as compared to cities where people spend most of their time in private indoor spaces – at home, in a car, or in an office.
Insight 3: SFPD response times are increasing
The SFPD classifies calls for service into three categories:
Priority A calls involve a present or imminent danger to life, major property damage, or suspects that may still be present or nearby. Examples include shootings, fights involving weapons, or in-progress burglaries.
Priority B calls involve potential for damage to property, a suspect who may be in the area, or a crime that has just occurred. Examples include burglaries without a perpetrator on-scene or verbal altercations.
Priority C calls involve everything else.
As the above chart shows, SFPD response times increased between 2020 and 2022:
Priority A: from 7 to 9 minutes
Priority B: from 21 to 31 minutes
Priority C: from 55 to 71 minutes
Insight 4: SFPD is solving fewer crimes than before
Law enforcement agencies consider crimes solved when they are “cleared” by arrests. For this reason, clearance rates (the ratio of arrests to incidents) are a common indicator of police effectiveness. Clearance rates for homicides in San Francisco have been increasing and are currently at 75 percent, well above the national average of 61 percent, but clearance rates for all other crimes in San Francisco are on the decline.
In 2022, clearance rates were as follows:
Assault: 35 percent
Robbery: 20 percent
Rape: 13 percent
Burglary: 10 percent
Motor vehicle theft: 7 percent
Larceny: 2 percent
Insight 5: San Franciscans don’t feel safe in their city
An October 2022 poll conducted by the San Francisco Standard revealed that 64 percent of San Franciscans say they feel less safe than they did one year ago. Forty-five percent of those polled say they have family or friends who have been victimized by crime in San Francisco.
In addition, a San Francisco Chamber of Commerce poll conducted last year found that 83 percent of San Francisco voters believe that crime has gotten worse. The poll also found that, among voters who say they are likely to leave San Francisco in the next few years, 25 percent say that crime is the primary reason they want to leave the city – up from eight percent in 2020.
Insight 6: SFPD is short hundreds of officers
A major reason for longer response times is that SFPD has lost hundreds of officers in recent years. Specifically, between 2019 and 2022, the department lost 331 officers – over 15 percent of the sworn workforce – and the losses are projected to continue well into 2023.
Note: The difference between “sworn” and “full-duty sworn” represents officers out on disability leave, military leave, or family-medical leave.
In 2021, the city hired a consulting firm to recommend a police staffing level based on an analysis of police workload. The firm concluded that SFPD should have 2,182 sworn officers available for duty (not including patrol duty at the airport) to meet the demands being placed on it.
According to numbers provided by SFPD in April 2023, the actual number of sworn officers assigned to the city is 1,761 – 20 percent short of the recommended level. Making matters worse, 261 of these officers were on some kind of leave, including disability, military service, or family medical leave. Thus, the actual number of officers available to serve the city is 1,500, which is 681 officers short and 31 percent below the recommended level.
SFPD faces multiple serious staffing problems that have built up over the past five years:
Fewer people have been applying to the SFPD, and both the number of police academy recruits and the number of graduates have declined dramatically.
More officers have made lateral transfers out of the department than normal in previous years, in part because of SFPD’s stringent vaccination requirement.
More officers retired in the past few years as compared to previous years – a trend that is expected to continue. According to the SFPD, approximately 384 officers are eligible to retire as of January 2023.
7. SFPD is spending less and less time on proactive policing
Finally, lost in all the other statistics is a rarely acknowledged fact: In recent years, SFPD is spending less time on what are called “proactive” policing activities, which involve officers going out and taking initiative to find criminal activity, rather than just waiting for calls. Examples range from issuing a ticket to someone who failed to stop at a stop sign to detaining someone wandering through traffic and behaving threateningly.
The number of self-initiated activities declined from nearly 350,000 in 2019 to about 150,000 in 2022, which represents a 56% reduction.