An illustration of two bar graphs, one right side up and one upside down, mirroring one another

Weekend Reads | Crime Data Report Gives Shallow Analysis

by Kevin Schofield


This weekend’s read is the “Crime in Washington 2021 Annual Report,” published annually by the Washington Association of Sheriffs and Police Chiefs. The report collects and aggregates data from 206 law enforcement jurisdictions in the state of Washington, covering lawbreaking offenses, arrests, and the number of commissioned and civilian employees at each agency. The State is required to collect this data every year and submit it to the FBI for its own annual nationwide report on crime and law enforcement statistics.

Government agencies are good at generating this kind of report: hundreds of pages of data in tables, with some pretty graphics to highlight some particular statistics. Publishing the data is an important public service in the name of transparency, especially when it pulls together data from multiple disparate sources: It would take an enormous amount of time and energy for you or I to try to recreate it by contacting more than 200 law enforcement agencies and asking for the data — then trying to “clean” it so that it’s consistent across the jurisdictions it covers. But more often than not, the charts and graphs are less than useful and more closely resemble “data visualization porn”; they give you a visceral sense that you’re seeing something interesting, but they rarely provide anything beyond a shallow view of the patterns in the data (if that). This particular report is jam-packed with info-vis porn: pie charts, bar charts, and tables of percentage breakdowns that look useful but tell us little. Many of them miss the mark because they don’t answer the most important question when working with statistics: compared to what? For example, on Page 56, we are told that 16.7% of arrests were for DUI, but the report gives us no context to interpret it; is that a high number or a low number? Did it increase in 2021, or decrease?

The entirety of Page 11 is devoted to a “crime clock” (complete with a pocket watch as a background image) that takes 30 different crimes and calculates how often they occur, again without providing any frame of reference as to whether we should think the numbers are high or low. But there is a special place in hell reserved for the person who created this bar chart (on the bottom of Page 15); visually, it appears that violent crime has more than doubled, while in truth it increased by only 12%. 

A misleading chart shows the "Violent Crime Two-Year Trend," making it seem as though there was a 50% increase when in fact it was 12%
“Violent Crime Two-Year Trend” chart from the “Crime in Washington 2021 Annual Report.”
A comparison chart by the Emerald team shows a chart that conveys the crime increase more accurately
Chart created by the Emerald using the numbers from the “Violent Crime Two-Year Trend” chart above.

For the most part, the report sends a pretty strong message that it wants us to look at the data but not think too hard about it. In fact, in the introductory remarks, it tells us explicitly that we should not compare the data in this report to the FBI’s annual crime report, nor should we compare and contrast the individual agencies listed or create “rankings,” since each jurisdiction has its own context with differing demographics, economic base, infrastructure, and proximity to other facilities. That is, in a word, nonsense. The entire point of mandating the collection and publication of the data is to make it possible to compare jurisdictions’ approaches to crime and law enforcement. The context for each agency is indeed important, and we need to be prepared to dig in to understand why things may be different in a particular location; but at the same time, the data on other agencies is a large part of the context in which we interpret what any single law enforcement agency is doing.

The one exception in this report where the authors did well in providing some context is the statistics on each agency’s employment of commissioned officers and civilians. For one, it splits out sheriff’s offices from city and town police departments; interestingly, we discover that while King County has the largest population of any county in the state, both Pierce County and Snohomish County actually have more people living in unincorporated areas — and both have more commissioned officers than the King County Sheriff. They also contextualize the number of employees in an agency by presenting it as a ratio to the total population served by that agency. This is a case where the “law of large numbers” applies: You need a certain minimum number of officers to cover shifts, and in smaller departments, one or two extra officers can swing the ratio significantly. For example, Winthrop’s police department has the highest ratio of any department in the state: 5.71 commissioned officers per 1,000 people — it has only three officers, serving a town of 525 people. Overall, we see a wider variation among the state’s small cities and towns, and high consistency among the largest cities (with the exception of Bellevue) and among sheriff’s offices. The report also gives us a longer historical view on employment levels, noting a multiyear dropoff among police departments (though once again manipulating the chart to mislead us into thinking the change is greater).

A graph shows Police Departments Average Commissioned Rate per 1,000 Population
“Police Departments Average Commissioned Rate per 1,000 Population” graph from the “Crime in Washington 2021 Annual Report.”

Even though the bulk of the report’s “just the facts, ma’am” approach deprives us of any clear understanding of trends or patterns, there’s a lot we can learn by digging in a bit. Some of the tidbits I found while perusing the data:

  • While the report breaks out arrest counts for 42 specific offenses (plus a catchall “all other offenses” category), half of all arrests resulted from only four: simple assault, violating a no-contact or protection order, larceny or theft, and DUI. 
  • That said, there is wide variation across departments on the offenses that occurred in each jurisdiction. This is true even for small towns next to each other, such as Burien and Tukwila. But also for bigger cities: For example, Seattle and Bellevue are seeing some significant differences in the types of crimes occurring within their borders.
  • Only 5.4% of offenses involved drug or alcohol use during the commission of the crime — and much of those were drug-related offenses, such as possession of illegal drugs or paraphernalia.
  • While opiates are dominating the headlines, they represent a relatively small percentage of drug offenses in King County. Stimulants (such as methamphetamines and cocaine) and heroin make up the vast majority of drug offenses in most jurisdictions.
  • Motor vehicle theft is up everywhere in King County, but especially on the Eastside.
  • The statewide aggregate arrest data shows clear, ongoing racial disparities. Less so for white people: Statewide, about 78% of the population is white, and in most categories of arrests, the percentage of white people is near that figure (excepting extortion, bribery, and liquor law violations). But for Black persons, only 4.3% of the state population, the disparity continues to be large: 33% of arrests for robbery; 22.5% of prostitution arrests; 21% of aggravated assaults; 20.9% of arrests for intimidation; and 15.9% of weapons law violations.
  • Fraud offenses dropped off dramatically in 2021. This is discussed in the report: Apparently, unemployment fraud skyrocketed in 2020 as the pandemic hit and aid programs were rolled out. Those numbers have now come back down to earth.

Reports like this — strong on data, weak on analysis — make a good case for learning how to use the basics of a spreadsheet app: to be able to scrape a table of data out of a report, throw it into a spreadsheet, and run the numbers yourself. While we’re often beholden to our government to collect and publish basic data, we can’t let it tell us how to think about that data or dictate the questions we can ask. Personally, I hope that the government employees generating reports like this one are separately doing their own analysis that goes well beyond the inch-deep pablum they serve up to us — though if that’s true, I’m forced to wonder why they are choosing not to share it with us.

Crime in Washington 2021 Annual Report


Kevin Schofield is a freelance writer and publishes Seattle Paper Trail. Previously he worked for Microsoft, published Seattle City Council Insight, co-hosted the “Seattle News, Views and Brews” podcast, and raised two daughters as a single dad. He serves on the Board of Directors of Woodland Park Zoo, where he also volunteers.

📸 Featured Image: Graphic by Emerald team.

Before you move on to the next story …
Please consider that the article you just read was made possible by the generous financial support of donors and sponsors. The Emerald is a BIPOC-led nonprofit news outlet with the mission of offering a wider lens of our region’s most diverse, least affluent, and woefully under-reported communities. Please consider making a one-time gift or, better yet, joining our Rainmaker Family by becoming a monthly donor. Your support will help provide fair pay for our journalists and enable them to continue writing the important stories that offer relevant news, information, and analysis. Support the Emerald!