“Everybody that I know who doesn’t live in Portland is convinced that we burned it to the ground in 2020,” Aaron Roussell, an associate sociology professor at PSU and an expert on policing, said. “That it’s some sort of nightmarish hellscape that looks like Mad Max, only with rain. It’s not grounded in anything, and people just get away with saying this nonsense.
“I don’t know how to rebut complete nonsense.”
This is Part 1 of a 3-part series examining data regarding Portland crime rates. Read Part 2 here.
The non-Portlanders in Roussell’s life aren’t alone in their misconceptions. A false narrative of widespread protest-related destruction and its similarly widespread effects on businesses occupied hundreds of column inches in publications around the world in 2020.
As the largely localized protests died down, the narrative shifted toward homelessness, drugs and pervasive crime. Public drug use and sales were termed open-air drug markets, and tales of lawless city streets populated the pages of mainstream publications. Alleged criminal offenses — both serious and low-level — became proof of Portland’s decline. Dramatic headlines asked readers if Portland was on its death bed.
The narrative took hold and hasn’t let go. This June, London-based newspaper The Times ran a headline beginning with the entirely incorrect refrain, “This city defunded its police,” referring to Portland. Portland Police Bureau, or PPB, has its largest-ever budget — $295 million — and it only ever saw a brief 4.8% dip in funding consistent with across-the-board pandemic budget cuts in 2020.
The city experienced a tragic spike in gun violence and homicides beginning in 2020 — the annual average of 26 homicides from 2016 through 2019 increased to 81 from 2020 through 2023. Though the frequency of homicides is now declining, Portland was one of at least 12 cities to have a record number of homicides in 2021. A spike in reported vandalism and car thefts also began in 2021.
While the city’s real, exaggerated and fabricated woes continue capturing headlines, a wider breadth of data telling a different story fails to gain as much traction.
Arrests plummeted over the last four years. The monthly average number of calls to the city’s Bureau of Emergency Communications, or BOEC, since 2020 — including emergency and non-emergency calls — is essentially the same as it was in the two years preceding the pandemic. And despite an increase in 911 calls to BOEC over the last four years — which includes calls for police, fire and medical personnel — calls requiring police dispatch dropped precipitously in the same time frame.
Even homicides, shootings and car thefts decreased substantially in the last year, mirroring a national reduction in the same categories.
Only PPB’s “reported crime” data shows a modest increase in the last four years, less than 10%. PPB’s “reported crime” data is not the raw number of reported crimes but rather data “compiled by the specialized crime analysts in our Strategic Services Division,” according to Mike Benner, PPB public information manager. It is the only non-raw data source Street Roots analyzed, and the numbers include unverified reports.
Still, reports of crime dropped in many categories, including thefts and sex offenses, according to PPB’s data, while the majority of categories remained relatively consistent between the two four-year periods. The 9.7% increase in the average total number of reported crimes from 2020-2023 compared to 2016-2019 (5,806 more reports per year, on average), is largely attributable to a substantial increase in reports of vandalism (3,903 more reports per year, on average) and car thefts (2,251 more reports per year, on average).
There were also smaller increases in reported burglaries (1,208 more reports per year, on average) and assaults (975 more reports per year, on average).
According to all available PPB data, Portland is not and never was in the grips of a widespread crime wave.
Nevertheless, some media, Portland police and many city and state officials climbed aboard the bandwagon: Portland is going to hell in a handbasket and, according to some, it needs more police.
In early 2024, Mayor Ted Wheeler and PPB leadership began trying to turn the wheel of public perception in the other direction highlighting decreased crime in 2023. Just days after a January press conference about reduced crime, Wheeler cheered on a secretive city-state-business partnership — organized intentionally to exclude the public and evade government transparency laws — resulting in plans to increase city and state police presence downtown.
Wheeler has since touted recent decreases in shootings, homicides and car thefts, largely crediting Office of Violence Prevention efforts and increased PPB staffing.
When data and academic rigor enter the picture, the apparent criminal hell Portlanders lived in for the last four years appears similar to, if not less criminal, than the four years before that.
Hard and soft numbers
Officials and PPB personnel routinely cite eye-popping statistics — and anecdotal evidence — bolstering the idea that Portland became an incredibly dangerous city early in the pandemic, even when celebrating stated successes. However, the city’s data often seems to contradict what officials present to the public.
Calls regarding shoplifting increased by 88% in 2023, PPB Chief Bob Day said in the January press conference flanked by Wheeler and Mike Myers, Portland’s Community Safety Division director.
However, 2020-2023 shoplifting arrests are down by 67.6% when compared to 2016-2019, despite the bureau’s increased focus — and public messaging — on retail theft. Benner referenced PPB’s frequent missions (stolen vehicle operations, retail theft missions, criminal interdiction missions) when asked why PPB believes crime appears to be declining, but said it is difficult to say with certainty. Dispatch calls regarding larceny (theft) of any kind in 2023 — save for car thefts — reached the lowest number in the last eight years, according to the city’s data.
Benner told Street Roots that Day was on vacation, so the bureau was unable to provide a source for Day’s stated 88% increase, but said Day typically gets his data from the Strategic Services Division.
The Oregon Criminal Justice Commission awarded nearly $475,000 to support PPB’s “retail theft missions” on Aug. 1, which cost “thousands of dollars to run,” according to an Aug. 7 PPB press release.
While officials paint a picture of an ever-growing demand for a police force stretched to its limits, the city’s data shows the annual average number of public calls — emergency and non-emergency — resulting in police dispatch from 2020 through 2023 was 11% below the 2016-2019 annual average. Even the average number of dispatched calls per 1,000 residents decreased over the last four years.
While it’s true PPB routinely had fewer sworn officers since 2020 — between 7% and 20% depending on the month and year — PPB made 44% fewer arrests in the last four years compared to 2016-2019, according to PPB data.
But it’s not just the publicly alleged number of crimes, arrests or dispatched police that don’t line up with the city’s data. It’s also the purportedly exploding call volume BOEC faces as it relates to crime and wait times.
Commissioner Rene Gonzalez, the public safety commissioner until Wheeler seized all city bureaus July 1, said BOEC dispatchers experienced a 35% increase in workload in the first six months of 2023 compared to the first six months of 2018, in a July 24, 2023 social media post, but suggested BOEC wait times were beginning to improve.
“While crime, heavy drug use, and lack of collective public health continue to create roadblocks, much needed staffing increases are finally having an effect at the call center,” Gonzalez said on X, the social media platform formerly known as Twitter.
Gonzalez’s statements include a mixture of fact and fiction facilitated by the conflation of workload with actual call volume. When comparing the first six months of 2018 and 2023, BOEC call taker staffing was 13% lower in 2023, but the average monthly call volume was only 2% higher. So while it’s true calls per call taker increased by 18%, there’s no actual data supporting Gonzalez’s claim that “crime” created any additional “roadblocks” regarding wait times.
Gonzalez made a similar statement the following month to KATU regarding a first six months comparison, but compared 2023 to 2019. In that interview, Gonzalez explicitly said call volume increased by 35%, according to KATU. In reality, call volume was 6% higher in the first six months of 2023 compared to the first six months of 2019.
Gonzalez’s office did not respond to a request for comment on his previous statements, or the expansive disconnect between his messaging and the city’s data.
A careful selection of data points for comparison can also factor into creating narratives that ultimately crumble under scrutiny, according to Melissa Thompson, a criminologist and sociology professor at PSU.
Thompson said, in general, when someone makes an argument comparing small sample sizes, even year-to-year, it raises red flags for her as a criminologist because of how easily someone can manipulate or misinterpret a small sample. She said that’s especially the case when someone appears to arbitrarily compare one short time period to another.
It’s a common problem with how crime statistics are presented, Thompson said. There’s even a term for the practice — cherry picking.
“When you zoom in really narrowly, you’re missing the larger trend, the larger picture, and you can very much cherry-pick the conclusion that you want to come to, and so that’s not considered to be very valid scientifically,” Thompson said. “In terms of coming to a conclusion, you would want a larger sample. A very common thing that politicians will do on all levels is they will cherry pick the year and the comparison that supports the conclusion they want to draw.”
Gonzalez, Wheeler and BOEC Director Bob Cozzie have all made public statements regarding a substantial increase in 911 call volumes, but since BOEC wait times ballooned in 2021, the average total monthly call volume is only 0.6% higher than it was from 2018 through 2020.
911 calls did increase in that time period by about 20%, while non-emergency calls decreased by about 20%. However, reported serious crimes stayed relatively flat, and calls resulting in police dispatch decreased, according to PPB’s data.
Cozzie told Street Roots he did not believe the increase in emergency calls mirroring the decrease in non-emergency calls indicated more people were simply calling 911 for non-emergencies. Often, emergency events result in multiple calls, while non-emergencies result in a single call, he said. Instead, he believes it’s largely a coincidence caused by the bureau implementing automated systems capturing some of the non-emergency reports.
911 wait times also began decreasing last year due to those automated systems and increased staffing, Cozzie said, adding non-emergency calls often take up more time due to being less straightforward in terms of how a call taker handles it.
Cozzie said he didn’t have a clear explanation for why emergency calls are increasing while the number of calls resulting in police dispatch are decreasing, but he did highlight increased medical calls. BOEC’s open data also reflects a slight increase in fire calls over the last four years. Anecdotally, Cozzie said dispatchers are sending police on calls back-to-back without a down period more often than in the past.
What’s in a rate?
Regarding crime, Roussell thinks the common understanding and the conversation around it are deeply flawed, even when including expansive raw data.
“I can’t emphasize this enough that the word ‘crime’ doesn’t describe anything,” Roussell said. “The word crime means something that’s against the law, but the law is broken every single fucking day.
“I think what (people actually) mean is social harm.”
Roussell said the very statistics police and the public rely on the most — per capita crime rates — have little, if any use, in determining the prevalence of harm or officially designated “crime” in a community. Roussell said a number of social harms aren’t included in crime statistics, which are ultimately generated and contextualized by one side of an uneven power dynamic, referring to the connection between evictions and homelessness as one example.
“I (don’t) give a shit about the crime rate because the crime rate is a random political football,” Roussell said. “I do give a shit about people … that are starving for medical care. They’re living outside and slowly going crazy. They weren’t that way when they started, but they are when they finish because that’s what the violence of being forced from your home and onto the street is — all of which is perfectly legal.
“The sheriff is the person who goes and evicts people. So I just can’t get invested in that conversation (about ‘crime’).”
Roussell and Thompson both pointed out crime rates don’t include harm or misconduct committed by police, further diminishing their usefulness in determining harm in a community.
Thompson is also skeptical of how officials and the media represent crime, saying publicized crimes are often the least common types. Noting a declining number of crimes in Portland over the last several decades, Thompson said there’s an inverse relationship between how often crime happens, and how often the media and officials discuss it.
Roussell’s and Thompson’s at times similar arguments that crime can be a shaky concept with statistics easily manipulated to fit the authorities’ desired narrative isn’t a consensus in academia. Nor is the idea that crime, in how law enforcement defines it, hasn’t really increased, despite fewer police officers being dispatched to calls, fewer arrests, and a relatively flat number of calls and reported crimes when comparing 2016-2019 and 2020-2023.
Kris Henning, PSU criminology and criminal justice professor, said looking at crime rates versus crime numbers can reveal different patterns. He said comparing a single city over time can yield one pattern — assuming reporting practices or systems have stayed the same — while comparing multiple cities over time can reveal other patterns. Additionally, if data systems or victim survey questions change over time, the results may show perceivable changes while the lived reality is much the same. Henning also stressed the importance of evaluating the number of police officers per capita and controlling data analysis for population changes in a given city.
Henning said analysts do not know how many crimes are actually happening in Portland because they can only see what is reported, adding that he didn’t see anything confirming Wheeler’s recent statement that “crime is dropping like a rock.”
Per capita crime rates, while still prevalent in law enforcement and media circles, have fallen out of favor in academia over the last several decades, particularly for the purposes of determining short-term trends in small geographical areas.
Beginning in the 1970s, academics began examining the long-held assumption a linear relationship existed between population size and numbers of alleged crimes. Study after study showed a non-linear relationship between population size and numbers of alleged crimes, meaning more people didn’t necessarily mean more crimes, and fewer people didn’t necessarily mean fewer crimes.
Another problem with traditional per capita crime rates emerged as the science advanced. The “per capita” portion of per capita crime statistics fails to capture much of a city’s ambient population, which is the number of people in a city at a given time. This is particularly relevant for cities like Portland, the only large city in a metropolitan area, which typically draw people from outside their city limits each day as social and economic hubs.
Thompson said she still likes to examine per capita crime rates, but typically over the course of several decades and with larger populations, such as a state or nation, as a way to understand macro trends.
For more precise statistics, particularly in smaller areas across shorter time periods, academics now favor calculations factoring in things like population density and ambient population.
Henning called this the “denominator problem,” saying that in an ideal scenario, analysts would know how many people could have been victimized in city X and city Y during a given time period or neighborhoods X and Y. However, determining that figure is challenging. Census data provides an estimate for people living in a location but not the ambient population.
“Suffice it to say that all approaches to this issue come with major assumptions and caveats,” Henning said.
The same can be said for the numerator — the count of crimes that happened. Not all crimes are known to a victim, reported to the police or classified correctly by the police, according to Henning.
“Hence our counts of crime also come with major caveats,” Henning said.
Street Roots is an award-winning weekly investigative publication covering economic, environmental and social inequity. The newspaper is sold in Portland, Oregon, by people experiencing homelessness and/or extreme poverty as means of earning an income with dignity. Street Roots newspaper operates independently of Street Roots advocacy and is a part of the Street Roots organization. Learn more about Street Roots. Support your community newspaper by making a one-time or recurring gift today.
© 2024 Street Roots. All rights reserved. | To request permission to reuse content, email editor@streetroots.org or call 503-228-5657, ext. 40
This article appears in August 21, 2024.