A surplus of compassion could mean failing to truly help someone, or failing to design something that can.
I’m sitting in a conference room with my business analyst. It’s eight forty-five in the morning and I’m two hotel-breakfast-buffet-coffees in for the morning. I’ll follow it up with a third from the machine in the kitchen and my BA will have her usual, a Diet Coke. We’ve got our game faces on and we’re ready for a long day.
One by one, our subject matter experts—a team of social workers—file in and get settled. They all have their own refillable water bottles and coffee mugs. They dress sensibly, but nicely. Some of them bring blankets because the conference room is cold, even with the heat on (it’s February of this year and it’s snowing off and on this week). They have snacks and drinks and printed-out copies of the “Workshop Packets,” which are Powerpoint slides with baseline screenshots on it and lists of business requirements.
I’m in another state (I won’t say which one, for all our sakes) to help with the implementation of their forthcoming social work and child welfare application. The state in question is adopting a system used in another state and modifying it for their own purposes. The workflows of the two states are similar enough that this is a practical decision, but different enough that we need to take it step by step, area by area, and make sure that nothing will be missed. I’m here not only as a scribe to capture what is decided during the session, but as the user experience and design consultant. One of my better skills is summarizing an hour’s worth of discussion succinctly.
There are compromises and concessions that have to be made here. Will the technology have to bend, or will the process? Typically, the process wins, but this session also falls prey to the Observer Effect: by studying the process, the process inevitably changes. This means that, occasionally, the more senior social workers will insist on integrating some of the other state’s processes into the “new” application, knowing it will be more work in the long run, but that it will be worth it.
Documentation is everything, and everything is documented. If the records are not kept in painstaking detail, bad things happen. Kids can be lost in the shuffle. Abusers could slip out of the reach of the law. The Department won’t get paid. The states are required to comply with the Federal guidelines for Title IV-E of the Social Security Act, which makes up about half of any given state’s social welfare budget on average.
The reason: each record is THE record. What is written in the files of an individual will directly affect their life. Every word written, every incident recorded, can have an impact. There are certain records that cannot be deleted, ever. If the user enters the record incorrectly and needs to make a new one, they need to mark the old one as “created in error,” but that record stays in the system. Deleting those records would be ethically troubling at the very least and, in certain cases, illegal.
This system is meant to make the record keeping easy, which should make the worker’s jobs easier. And that matters. Because this is not an easy job.
I was told a while back, a few months into the job, that there is a directive that guides everything a social services department does:
Keep families together, if you can.
The Department doesn’t want to take anyone’s kids away unless it’s absolutely necessary. If parents are suffering from drug dependency or alcoholism, the Department will try and get them help. If the parent shows signs of improvement and a desire to care for their children, the Department will reunite them. Departments treat this as a moral imperative, which I, too, believe it is. There are practical reasons, as well: when a child is removed from their home, the State assumes custody. States don’t want to raise anyone’s child; they want the child to be placed with caring individuals who can see to that child’s needs.
The directive is noble, but it’s also conditional; “…if you can” implies that, sometimes, you can’t. When a parent is proven unfit in a court of law, their parental rights can terminated. How are parents proven unfit? The worker brings all the documentation in the system to court and provides evidence that a family court judge (or in some cases a jury) can consider. Every bit of data recorded about that case becomes the basis of the State’s claims against the parent. Every paragraph, every word, every punctuation mark matters in those moments. Everything entered counts, and everything omitted does not.
The system we’re building isn’t just a tool, it’s a shield. The data is critical to helping struggling families thrive, or to placing children with a family that can care for them adequately. But the data isn’t just “data,” it’s a collection of stories. Stories of a family in crisis. Stories that are hard to hear.
It’s August of last year. I just transferred from the Federal branch of my company to the State Government/Enterprise branch. On my first day, the design group had a stand-up meeting. I sat in and listened as our Design Leads welcomed me to the group and discussed how things were progressing.
Chris, my co-worker, mentioned that he was making headway on a screen known as “The Deceased Foster Child Checklist.”
I shook my head. What? He had to have misspoken.
He hadn’t. There, on the shared conference screen, was a long form.
“The purpose of this form is to record what happened that might have lead to the death of a child in foster care,” Chris explained.
I was completely dumbstruck. How could you boil something like this down to a form? I wondered.
Answer: you have to boil EVERYTHING down to a form. Every ugly scar, every nightmare scenario. It all has to be recorded. “Narratives” are applied liberally throughout the application. These are text areas that can carry upwards of four thousand characters. Sometimes, entries in these fields are mandatory.
I came home that day and my mother-in-law, a now-retired social worker of thirty-eight years, was watching my daughter. I told her how shocking it was, the sight of the “Deceased Foster Child Checklist.”
“You know, in all my time in Social Services, I think I only ever had to use that form once,” she offered.
That simple statement was so heavy, and the way she said it bowled me over. To be honest, it was somewhat comforting: she only had to use it once. But she had to use it once. There was a given day, some time in thirty-eight years, in which she had to put that awful form on her desk and fill it out for the sake of the case.
In her time, they never had a system like the one we’re building now in this other state. They wanted it for years, and for all the reasons we’re building this one: consistency, clarity, accountability. In thirty-eight years, it was discussed multiple times but never implemented. Getting as far as our client has come can be next to impossible in some states. I have to believe that’s because empathy is not commonly part of a government’s approach to budgeting.
Mike Monteiro gave a talk recently about the responsibility of designers to their users. It’s a great talk and there’s one moment that REALLY stuck with me. He was talking about the concept of the “Veil of Ignorance,” in which you imagine that your role is chosen after a system is implemented (e.g., if you create a society in which slavery is legal, you may want to consider what happens if you are cast in the role of “slave”).
Long story short: design it like you might be subjected to it one day. Design it like it might make a difference in your life. It’s all about empathy, a word we throw around a lot and, hopefully, practice at least half as often.
We have arrived at a moment in our society when there are two kinds of empathy: universal empathy, in which we can picture ourselves in the place of someone we may not like or understand; and selective empathy, in which we only choose to feel deeply for those with whom we already align. For more on this divide, please take a moment to read this article “The End of Empathy” by Hanna Rosin. I can easily see how empathy could be a hindrance. Selective empathy, especially, could make a worker start thinking of a case member as a friend, or a family member. It could cause a worker to step outside of some very well-defined lines.
It’s easy to use selective empathy when we discuss “users” because we speak with users. But we’re not just building this system for the people who enter the data. As I began my work on the social services application, I kept trying to remember that the clients – the mother losing her children or the father trying to adopt or the couple working out their issues – are subject to this system. When a message appears letting a social worker know that a client of theirs missed a key court date, someone has to supply the “why.” That’s the client’s duty. They don’t log in, but they are simultaneously a “user” and a source of data. They are the subject matter experts of their own lives.
It matters because whatever forms we create in the application will have to be filled out, and the worker needs to be sure they’re asking the right question, because the answers may change based on how the question is presented.
Many people in the system are in poverty, and where there’s poverty, there is a lack of opportunity that may hinder that client in other ways. If the client is unemployed, how can they afford a car? If they have no car, how can they get out to find another job, or take their children to school? If the kids are not in school, are they being fed? If they’re not fed, they’re going to perform poorly in school anyway, which may require the parent to attend conferences with the teacher, which would require a car and time away from any job the parent might try to get. Each difficulty can create or amplify a second difficulty. Problems in the lives of these people rarely come one at a time. Simply stating, “the client was late to their court-ordered counseling” doesn’t begin to tell the whole story, and it certainly isn’t fair.
Another real-world example that really blew my mind: imagine that you have a client whose children are being taken into protective custody. One of those children is thirteen…and she has a child of her own. That child has to be accounted for, too, even though the thirteen-year-old is technically the parent in that case. I tried to think about what that would be like, to be a young girl who ended up with a baby and now my parent was abusing both of us and we needed to leave home.
I have tried really hard to put myself in that girls shoes, in that moment in her life. And I can’t.
I had to simply admit at a certain point that the idea of being subjected to this system terrified me. For many case members, the presence of social services in their lives is expected. When you are in the system, you are assigned a case worker. In a perfect world, that worker deals with you as little as possible and you and your family move on from whatever it was that afflicted you.
More often than not, though, a social worker is involved in your life from the moment you first enter the system until you age out of it and into your own livelihood. The workers know their case load pretty well, and there are plenty of those who need a consistent presence around them to stay on the path toward recovery and stability. But the relationships aren’t one-to-one; social workers have many clients and a lot of work goes into each one.
As I began to get deeper into the work, I started trying to picture myself in the roles identified in the system.
“I’m a child. I’m eight years old. I have a little brother who is four. We’re being taken from our house to emergency foster care. It’s two o’clock in the morning…”
Fear. Confusion. Wanting my brother to stay close.
“I’m a parent and I’m an alcoholic. Someone has come to my house at ten in the morning and the sheriff is with them. They are coming to take my child away…”
Anger. Depression. Self-Defense. Sickness.
I have a daughter who is coming up on two years old. She and my wife are the light of my life. The thought of anyone taking her somewhere my wife and I couldn’t get to her makes my body lock up with terror.
Putting myself in the shoes of a person whose child is being taken? It’s really and truly unthinkable to me. There’s an extra layer, too: the person taking the child has a reason, and they have the law on their side. How can you not feel ganged up on at that moment? How can you not feel hopeless and sick, or angry enough to do something drastic?
And these things happen. The social workers all stood around in a circle one day during lunch at a workshop and told me about the times someone had threatened to shoot them, or to shoot their children.
“They threatened to come after your kids?!” I said, my eyes widening.
“Oh, yeah, they say all kinds of things about how you need to watch your back and all,” my BA says. She was a social worker for years before joining the private sector to help implement these systems.
“It happens a fair amount, and you always take it seriously, but…” one of them told me, before shrugging. He said he had been directly threatened several times.
“I could never do this job,” I say to him, a refrain I’m becoming famous for at these workshops.
“Aw, you’d be fine if you make it to month three,” one of the lifetime workers says to me.
Three months seems to be the do-or-die time, according to another senior worker. “When we get new hires, I know in the first three months if they’re going to work out, because anyone who hasn’t quit at that point is made for it.”
“I was only supposed to do this job for six months,” one of the younger batch told me. “That was twelve years ago,” she laughed.
“And you’re not allowed to leave, either!” one of her co-workers said to her. They shared a good laugh.
That’s what often surprises me in these workshops: that anyone can face this kind of work each day and still have such a positive attitude. My mother-in-law is like this, too. Her optimism is ever-present, and she carries herself with a kind of assertiveness that reminds me of World War II-era civics: “We can do it,” “Keep calm and carry on,” that kind of thing. The workers can be a little jaded at times, but the job isn’t breaking them. Perhaps in order to be effective at this, one has to have a kind of “happy fortitude?” I like to think I’m a happy person, but I genuinely can’t picture myself doing this work.
“I’m a case worker and I have been called in to secure two children whose parents were involved in a violent altercation. I have to get them to emergency foster care, but the nearest person who can take them is two hours away and they’re both completely losing it because they saw everything happen…”
And that’s about where I stop imagining it. I get too lost in the doom and gloom.
So much for empathy. I can’t put myself in the place of a client. I can’t put myself in the shoes of a worker. So what can I do?
This system we’re building is meant to make that simple by taking the hundreds, sometimes thousands of individuals into account and making the connections between them and their families (or would-be families) clear. It is there to say “Yes” sometimes and “No” sometimes, and more importantly, to explain “Why” and “Why Not”. It is there to help people in need. I had to stop staring too hard at the underlying causes and realize that this application is key to the solution.
I started to re-frame the use cases in my mind:
“I need to get these kids to safety, and they need to stay together.”
“Mrs. So-and-so needs a small subsidy to cover the cost of her Insulin. Can that come from MedicAid? If not, can we get state funds allocated? How can I do that quickly? Her supply has almost run out and she really needs it.”
“There is a foster parent with two children in placement and the parent needs surgery. Who can take the kids while the parent recovers?”
Empathy is essential, make no mistake. Without it, there would be no social services; there would be no safety net of any kind. Parents would kill their children. Older people would starve to death in their homes, with no one to offer them assistance. We like to tell ourselves that churches and/or the community would step in. The mere existence of social services is proof they wouldn’t. The point of social services is to do the hard, ugly work. Yes, at taxpayer expense, but never as big a slice of the pie as they need to be truly effective at scale.
But empathy can also work against you. You can get so focused on the darkness that fuels the work that you forget that the work needs to be done. Worse, an excess of compassion can stop you from doing things by the book, and that would put you in a whole world of trouble. When writing user stories, when thinking about the process, I try to focus on the fundamentals. For the worker, it boils down to “This is unpleasant, but this is the job; let’s get it done quickly and efficiently and try and find a good outcome. And these aren’t just ‘case members,’ they’re people, and people deserve dignity.”
When good social work is being done, it can improve the situation of everyone involved. The system my teammates and I are retrofitting for the state can’t make a great social worker, but it can help social workers do great work.
The end goal:
- The system should work well in areas with poor internet connectivity. This can be especially challenging in remote and rural areas.
- The system should allow the user to enter narrative and be specific, but should also allow the user to flag certain data points without relying on deep reading. For example, if a child in foster care is a flight risk, the user should be able to highlight that in that child’s case file without having to write it out in a narrative, and that file should be flagged so everyone involved knows that the child might run.
- Many social workers have long histories in the department, and many of them are retiring. It should be fairly easy for new trainees to learn the system for themselves without relying on institutional knowledge or hearsay.
- Where possible, there should be ways to gather “big data” about the system. “How many cases went through intake this year with opioid dependency as a contributing factor?” is a question that might give insights to drug dependency and the opioid epidemic’s role in harming children and families. It should be fairly easy to figure that out and to create meaningful analysis, if we do our jobs correctly.
- Then, there’s the basic UX/UI work we can do to streamline and speed up processes that might take too long now. Shortening the workload of a worker and smoothing the path between A and B is essential. A worker who can conduct their business quickly, efficiently and adequately is a worker who can return their focus to the hardest parts of the job.
As I write this, I am told by our BAs and project managers that our implementation is going well. It’s comforting, knowing that the project is running the way it should. It means we will give our stakeholders something useful that they can honestly say they helped us implement and refine. What effect the new system will have will only be truly evident after the initial adoption phase ends and the workers begin integrating the new system into their working lives.
The effect the system has had on me is already evident. I think about these people and the way they conduct their business all the time. I have a new appreciation for it, and a new reverence for the field. I am immune now to the Dunning-Kruger effect as it pertains to social services: I don’t imagine it’s easier than it looks, and I know I couldn’t do it. But what I can do is offer practical guidance and do my work as well as I am able.
I keep thinking about how you’re not supposed to jump in the water to help a struggling swimmer; your best bet is to throw them a lifeline of some kind and help pull them in. I think we’re inclined to think of empathy as being in the water, neck-high with the person in crisis. But real empathy isn’t supposed to bring both of you down. It merely helps you recognize the other person’s situation and determine the best way to help them without getting caught in the current. If we do our jobs correctly, our users will have the length of rope they need to help someone and, God willing, to pull them out.