Data systems in non-profits are nearly always built to track, manage, and analyze individual clients/service users. In fact, this is a base assumption of such systems, and the vast majority of the literature on the topic supports this assumption. The problem with tracking service users.. well, there’s a lot. Here’s a few:
- Client privacy and consent in data collection and storage. Consent especially is very rarely built into these models in any meaningful way.
- A focus on clients inevitably looks to clients to change, and therefore to clients as the victim of their situations, or worse, as the problem.
- Related, focusing on individuals means we easily forget to examine the circumstances and environments affecting clients’ lives. And then we spend less time trying to understand / change those.
- Tracking clients makes clients available to others to track clients, meaning our clients become more visible to other systems that might harm them (e.g. law enforcement).
- Gives the dangerous illusion that the data we collect/track about clients is the important data. We need to fill it in, so we focus on it. Now that we have it, it becomes our focus. It’s harder to remember that other things might matter too.
- The data we choose to collect is often based on racist, colonialist, and other systemic biases and oppressions.
We need not look only at service users. Maybe we don’t need to look at them at all! This prompt in this 5 part series takes the microscope off service users and turns it on the surrounding environments.
I worked once with a group of young people living in an area with high rates of gun violence. They were told that gun violence was somehow their fault – don’t join a gang! Don’t solve your problems with violence! Don’t fight! Don’t offend anyone, lest you be shot! And the programs that “kept them off the street” (many of which were fine programs) would keep track of how many hours they spent in the programs (vs. “who knows where else”) and whether the programs kept them alive and improved their grades. But this wasn’t very satisfying to the young people there. They said: gun violence isn’t even our main problem with violence. We don’t stay off the street because of that – we’re afraid of the police!
So, working with my colleague, they designed a study (Participatory Action Research is another of these data work prompts) to understand the ways policing was impacting the neighborhood’s sense of safety. They did surveys on the street and in parks, asking the community using public spaces how they made sense of safety. I joined them for some of these surveys, and the conversations they sparked were almost better than the survey data. When all the surveys were collected, I helped the group explore their data with exploratory data analysis and data visualization (both ways of exploring quantitative data with groups). In the end, they were able to present to various local stakeholders their results – and to complicate discussions of safety, violence and policing.
Studies like this are not focused on program participants at all. They switch from being the object of study to subjects creating the study and collecting/analyzing the data. In fact, they switch the entire focus from addressing the symptoms to addressing the root causes. Graduating high school isn’t hard because you need math tutoring, when the real distraction are questions like: is it safe to walk home? What will a cop do if they see me? And a host of other questions like that.
There’s a long and storied history of this kind of study. Ida B. Wells-Barnett was an investigative journalist at the turn of the 19th century, examining lynching across the southern United States. Her writing, which included stories but also a tallying of numbers (and, in her emphasis, these should not be separated!), helped to reveal to a white public the extent of lynching and the obviously racist and unjust rationales given by perpetrator-murderers. Wells-Barnett didn’t study the victims to learn about what might have caused others to murder them – she studied and documented the murderers in an effort to expose them. Efforts like this to study white supremacist violence and confront it continue today, including the community research work done to take on Stop & Frisk policing in New York City.
The health care world has started calling studies that examine the structural and systemic causes of harm Social Determinants of Health and is taking the study of them seriously. In short, they are creating data about harmful systems, not individual clients. I simply call this “systems analysis”: creating data about oppressive systems that help to undermine, challenge, and transform them.
If one of the major purposes of data work in non-profits is to improve program service delivery, a study of clients’ environments, especially one that can lead to concrete actions within and outside the organization, is likely to create more significant change than collecting and hording tons of data about clients. It can help inform change in how the organization responds to clients. It might suggest needed organizational partners in confronting challenges. It might be used to create dialog with stakeholders or advocate for policy change.
A great example of a project working from this perspective is OpenReferral. This protocol and technology for sharing data about available community resources reframes challenges as about resource shortages, focusing on finding real-time access to the services clients need, without a focus on managing or tracking clients. The protocol focuses on matching available resources to those who need them, recognizing that sharing real-time information across agencies helps everyone do their work better. The human service system is the target of data collection, not clients. Tracking information about service use and availability over time in a manner like this could also help a group of organizations advocate for more resources or policy change.
This isn’t ivory tower science. It is the grounded study of real problems from the perspective that organizations, institutions, and systems are at fault, not individuals, combined with the belief that better understanding problems from that angle will help us do something about them. It leans on “systems thinking” concepts that help us understand systems, subsystems, and their relations to each other in order to understand an ecosystem of root causes, stakeholders, power relationships, symptoms, and possibilities.
Systems Analysis can be done by staff within an organization (or partnered across several). But it can be better if the community can be involved too. And much like Prompt 1 about Reggio Emilia documentation, involving people in the process changes how they relate to your organization and the staff who work there – part of the hidden pedagogy of how we manage data that we too often forget. Systems Analysis gives you the opportunity to create democratic, collaborative relationships with clients and community that give clients agency, self-determination, and possibly the collective belief that people can confront challenges together.
I differentiate this from Participatory Action Research because PAR could be used to study clients or the internal practices of an organization only. Systems Analysis should look outward, recognizing that an organization is trying to solve problems that are originating elsewhere. The only loophole here might be that client’s experiences of these systems could be part of the study, part of the data that informs an analysis of the problem.
Effective Systems Analysis should include possibilities and proposals. What might be done, at the organizational, neighborhood, local political, or even broader levels to address this problem? Awards to projects that can propose local, immediate, and implementable change, as well as the bigger and slower stuff.
One final example of the last bit. I worked for a while as a youth worker in a community center where young people were often harassed by police. If we looked at the data, we’d have a whole bunch of troubled young people, because so many had citations for things like loitering. But if we looked at the patterns of policing, we’d see that police would pick up kids just for walking around the street, cite them, and then have an excuse to stop them in the future. A quick solution: walk with (older, whiter) staff if possible. Take the center’s van, even if for short outings. Longer term, our group organized a local conference on policing and safety. The organization didn’t need to fix the kids at all, but it could do better by them.