How community and context helped my reporting on death rates in county jails

When I embarked on this project, I knew I was signing up for months of research into my topic of deaths in county jail custody. It was something I had been wanting to look into for a while, and with the California Fellowship, I’d have the opportunity to put my local jail’s numbers in context. 

The idea for a deeper dive into the issue of jail deaths came in the fall of 2019, when the Shasta County Jail had three people die in its custody in a two-week span. I wanted to know how typical something like that was. It turned out that half of the state’s county jail systems, including more populous jurisdictions like San Francisco, San Mateo, Merced, Kings and San Luis Obispo reported fewer than three deaths the entire year. 

My goal was to highlight the rate of jail deaths in Shasta County and other parts of the state with a three-part series. The first part introduced to the topic by highlighting a local family with a loved one who died in custody and topline observations about the Shasta County Jail. 

The second part highlighted the discrepancies between death rates among jails. That was important because a common refrain from jail officials is that many people come in with poor health, which is true. But the point about discrepancies stands, since all jails are housing a vulnerable population and some still have much higher death rates than others. The second part also included insight from experts about the factors that generally lead to higher rates of deaths in jails. 

The third part was about reforms governments can impose to reduce the number of jail deaths, based on an interview with the San Francisco County Sheriff, as his department has seen a substantial improvement over the last decade, and reports into jail deaths from attorneys and experts. 

The topic of jail deaths is especially important in California’s North State. That’s because many counties here have among the highest arrest rates in the state. The result is more people in jail in mostly smaller, rural counties that may not have the resources to take care of their increasing incarcerated populations. When that happens, the jail’s limited resources are stretched thin, making it harder to supervise people in the jail and therefore ensure their safety. 

Here I’ve compiled a few tips for people working on similar projects, and my takeaways from reporting specifically on jails and jail deaths. 

For large projects on specific government-related issues: 

• Start thinking about visuals early on: This was something I wish I had done a better job of at the start of the project. In my case, COVID-19 made taking photos from inside the jail a non-starter. In hindsight I wish I would’ve been more creative, perhaps finding other sources for photos or substituting them with illustrations. 

• Keep your focus narrower than you initially think it should be: In my experience, large projects are going to expand on their own the longer you research them. While the topic of deaths in county jail custody sounds narrow enough, there are endless branches you can pursue: analysis of causes of death, health care spending, legal battles over safety standards, and so on. So be mindful that starting with the narrowest focus possible can be a big help. 

• Make comparisons to contextualize things: There’s a big difference between saying a jail “had three deaths in 2019” and saying it “had three deaths in 2019, which is more than half of the counties in the state.” Don’t assume readers will know that three deaths in a year — or whatever measure you’re reporting — is significant. In certain projects where data isn’t standardized (fortunately mine was, according to the state database I got it from), know that comparisons among jurisdictions might not be apples-to-apples due to idiosyncrasies in the way they report data. If you run into an issue where that’s hurting your ability to make a comparison between jurisdictions, try instead to make a comparison between two different time periods in a certain jurisdiction. 

• Don’t underestimate time for data entry: If you’re using a lot of data in your project, be cautious about planning your project with the assumption that you’ll have your data work done within a given period of time. It could take two or three times as long as you imagine, even if you don’t find irregularities that take extra effort to resolve, both of which happened to me. 

• Search in local government agenda packets: If you’re covering a local issue in a county or city, see if you can search for all references to “jail deaths” or “jail health care” or whatever your topic is in past agenda packets. This will give you an idea of how local public officials are framing the topic. For example, I learned that when the jail comes up in Shasta County Board of Supervisors meetings, it’s more often about the desire for more jail beds, rather than concerns about confinement conditions. That’s helpful for knowing where public officials’ priorities lie on a given issue. 

• Use community engagement: As part of my Fellowship project, I also worked with Center for Heath Journalism engagement editor Danielle Fox on trying to incorporate the people most affected by the issue and their concerns into the reporting. One of the ways we did that was with a Google form that asked people who had been to jail for notes on their experiences. That ended up flagging an important angle for my project — access to medication — that I might not have realized was such a pervasive concern otherwise. 

Reporting on county jails: 

• The most helpful online resources: The Board of State and Community Corrections has a wealth of data collected from county jails on bookings, spending, inspections and more. The California Attorney General’s open data portal has various helpful databases that cover criminal justice issues from deaths in custody, law enforcement staffing, crime rates and others. For individual anecdotes, use PACER (comes with a small charge per search) to search for federal court cases against county jails or sheriff’s offices. This will help you find examples to highlight if people have alleged negligence or mistreatment. 

• Who to talk to: If you can’t find a specific piece of information online and want to call the sheriff’s office, be sure to speak to a records or jail supervisor rather than a first-level records worker. They get a lot of ignorable calls during the day and you might get left aside if they don’t understand the significance of your call. 

• Let consultants do your work: A lot of counties will hire consultants every so often to review jail operations. If they exist, try to get ahold of these reports as they will speed up your reporting process by isolating problem areas. Citing them can also add credibility to your reporting since the consultants have already been vetted by county officials. 

• Talk about taxpayer costs: Inevitably when reporting on criminal justice issues there will be people quick to cast aside any concerns about the welfare of those accused of crimes. While it can be helpful to use moving anecdotes to humanize those folks, another way of getting the attention of someone who might not care about the topic is by showing how much taxpayer money is being spent on an inefficient system. In the case of jail deaths, that money can be sums paid out in wrongful death suits or other court-related penalties. 

• Be fair to the jail: When you’re reporting unflattering things about an agency, make sure to explain the jail’s side of the story. There’s a good chance they didn’t make a decision to have a smaller staff and operate on less funding on their own. They’re facing their own challenges. It’s important that the officials at the center of your story believe the reporting is fair and that their perspective is thoughtfully included, or else it will be written off as biased, leaving little chance for an ongoing policy discussion. For my project, simply talking with officials about the importance of fairness in my reporting built trust because it showed I wasn’t in it for some sensationalist headline.