| Hello and happy August! We’ve both been traveling a lot this month, so we have fewer links and musings to share than usual. That means that today’s newsletter is going to be all business: two issues we think could make for some great journalism. This is also our first foray into offering sample open records requests on MuckRock. Let us know what you think about them or if you have questions or ideas for future editions of The Jump Line. If you watch TV or stream anything with ads these days, you’re going to see lots of companies advertising their AI tools. Every time we open our phones, AI is offering to help write our emails or edit our photos. Because these tools are popping up everywhere we turn, we are planning to feature a number of applications of AI in policing and incarceration (there are many). Today we’re starting with ALPRs. Automated license plate readers (ALPRs) are popping up across the country. These cameras are typically stationary, but many are mobile and in some cases even mounted to police cars themselves. ALPRs use object character recognition to track license plates and record the date and location of a vehicle. They can be used for non-governmental reasons (for example, fast food restaurants are also using the technology to “tailor the drive-through experience”), but we’re going to focus on their use in policing. ALPRs were originally used as red light and speed cameras and in toll systems. But police are starting to use them in ways that have worrying privacy and public safety implications. Newer ALPR companies like Flock Safety, which was founded in 2017, track more than license plates. They can also track the make, model, and unique characteristics (e.g. a dented right bumper) of a vehicle, its immediate surroundings, and even its driver and passengers. Because the cameras are so ubiquitous, they function as real-time surveillance of any given vehicle. Because relatively little has been written about this technology, you can probably start with pretty simple questions: what companies have cameras in my community (Flock, Genetec, Rekor, Samsara, Asura, Motorola, Jenoptik, etc…), what do they track, who has access to that information, and how long is it stored? We’ve started some sample ORR requests on MuckRock here, and you should feel free to duplicate or improve upon them! Here's a clearinghouse with all the state statutes around ALPRs. (This is a work in progress, as many states don’t have anything in place yet.) The Electronic Frontier Foundation has a wealth of information that could inform stories, like this database of local police ALPR usage and this similar, California-specific database. Here are a few angles to consider: Map out the location of the cameras in your community. You can request emails between city or county officials and the company, looking out for any discussion about camera location. Are there sensitive locations in your community to be mindful of (healthcare centers, immigration clinics, etc…)? EFF notes that ALPRs have been used outside of gun shows, mosques, and political rallies. And some smaller communities install cameras at the entrances/exits to town, creating an electronically-gated community. Investigate what police are doing with the information. EFF found that police have used ALPRs to locate people without insurance and people with unpaid court debt. Consider cross-referencing the number of cameras in a given community with the percentage of the municipal budget brought in by traffic fines and fees.
Thomas Brewster has been doggedly investigating Flock for Forbes. He found that the company has broken numerous laws in various states as they install their camera systems without the local Department of Transportation's permission. He also found that a network of 309 Flock cameras in Riverside County, California scanned 27.5 million cars in one month. Many of Brewster’s scoops came from open records requests between Flock and the DOT, which is something you could replicate in your own community.
What happens when cameras get something wrong? The same companies that sell ALPRs are starting to offer technologies that bear striking similarities to older systems with a documented history of failure. For example, Flock recently came out with an “audio shot detector” called Raven that the company claims can detect gunshots with 90% accuracy (notably, the underlying data is not readily available on Flock’s website–a great thing to follow up on!). An older but similar program called ShotSpotter has a poor track record, drawing scrutiny and in some cases contract cancellations in multiple cities including Chicago, New York, and Dayton. So-called “hotlists” are another area rife with potential error. These are lists of license plates associated with vehicles or people that the police are looking for. If the ALPR scans a plate on the list, the system sends an alert to the police. Sometimes there’s error within the ALPR, and sometimes there’s human error; either way, these hot lists have resulted in entirely unwarranted, sometimes violent, police stops. For example, Sasha Lenninger at KOAT reported on a Flock ALPR that wrongly identified the car of a 21 year old woman in Española, New Mexico as stolen. The driver and her 12 year old sister were pulled over, handcuffed, and held at gunpoint. They have sued the city. Jessica Porter at Denver 7 reported a similar story out of Aurora, Colorado.
As you might have heard, it’s a national election year, and federal funding is always top of mind in those years, given how deeply politicalized it is. Connecting Medicaid and public safety As we’ve mentioned before, there is lots of research on the connection between Medicaid and public safety. Here are just a few studies showing that Medicaid expansion leads to lower violent crime, fewer arrests for drug crimes, a reduction in overdose deaths, and lower eviction rates. Conversely, dropping people from Medicaid leads to increases in some types of crime like assault and theft. (Alex Burness at Bolts has a great piece on this last study.) There are so many potential directions for great journalism here. We could imagine stories featuring interviews with doctors, nurses, public defenders, and new Medicaid patients that seek to unearth the reasons why Medicaid–or the lack thereof–is connected to crime and incarceration. There is also lots of room for data journalism here looking at local Medicaid uptake rates, crime statistics, and incarceration rates. Princeton’s Eviction Lab tracks evictions across the country. You could compare eviction rates with Medicaid expansion (or lack thereof) on the county and state level.
If you’re looking for high-level information, KFF has some great explainers on how Medicaid financing works.
Applying a public safety lens to one of the policies currently dominating the headlines: free school lunches Anna Pope at Harvest Public Media has a good rundown of a new federal program called SUN Bucks. Here’s the tl;dr: SUN Bucks provides $40 of grocery money per month per child over the summer to some low-income families. These funds are meant to be a backstop for the free and reduced-price lunches many of these kids receive when school is in session. Twelve states–Alabama, Alaska, Florida, Georgia, Idaho, Iowa, Mississippi, Oklahoma, South Carolina, South Dakota, Texas, Utah and Wyoming–opted out of the program. Because this is a brand new program, there are ample opportunities for journalists to follow the impact, both in states that opted in to the program and those that opted out. The impact of free school lunches on children’s hunger and health is well-documented. Does this program have similar effects? What about its impact on other types of safety? Other cash assistance programs have been shown to reduce some types of crime. Does this program do the same? In opt-out states, journalists can also focus on community-based efforts to fill the gaps. Marigo Farr at Grist has an interesting piece on urban orchards that could serve as inspiration.
Don’t forget about environmental programs There’s a ton of evidence that environmental factors like water, greenspace, and vacant lots have a big impact on crime and violence, so it’s important to track federal funding related to those issues. Michael Phillis at the Associated Press reported on four states’ (Washington, Maine, Oregon, and Alaska) refusal of federal funds allocated for removing lead pipes. He learned that states have to make an official request in order to decline the funds. This presents another open record request opportunity–this time for the communication between the agency and the declining states.
| |
