The Human Cost of Bad Systems

In the digital age, technology is considered the first and best solution to many of our problems. It is often used as a Band-Aid—something that can be deployed quickly and easily while we search for a longer term fix. But patching a problem with technology has its own issues—it can surface or even exacerbate fault lines that weren’t visible on the surface of the problem itself. At this year’s Summit, a panel of technologists spoke about two particular challenges that highlighted the need to think beyond tech to address a human crisis.

“You cannot use technology to replace the fact that you still have to build relationships.”

Historically, Rhode Island has had a difficult time placing children who are in need of foster care with families. Adam Williams of the Rhode Island Department of Children, Youth and Families and Marina Nitze of New America found themselves in a situation early last year when there was a significant uptick in the number of children who needed to be placed in foster care. They had been approaching the issue from place of data-driven recruitment, but in light of the increased need found that it was not enough. They decided to step back from the tech and dive deep into the entire process around foster parent recruitment, and found that the most helpful thing they could do was not launch a digital tool, but overhaul and streamline the process itself. Most importantly, they learned that they couldn’t meet an increased need by simply inserting technology to expedite the process, even when there was pressure to use technology first—becoming a foster parent is a life-changing experience, and the most important tool for recruitment is building intimate relationships with the families.

“Tech is not necessarily the most fundamental platform… Trust has to be the most fundamental platform.”

The other panelists spoke about their work combating the immigration crisis at the US–Mexico border. Vivian Graubard of New America, Nina Siulc of the Vera Institute, and Kavi Harshavat of GoodCTZN all traveled to Texas last year to lend their skills to helping children affected by the family separation policy. They found themselves in an incredibly fast-moving environment with a lot of uncertainty and distress, and it was immediately clear that this was not a problem that technology could solve. They spoke to shelter staff about their most pressing needs, both technological and otherwise, and found that the answer was not fancy facial recognition software or blockchain, but very simple tech: cell phones for people to call family members they’d been separated from. And along the way, they learned important lessons about building technology in the midst of a shifting crisis with serious human consequences.

When we’re confronted with a challenge, it’s tempting to turn to digital solutions first. And sometimes, it helps—but we have to be deliberate about building the right technology, not tech for tech’s sake. Most importantly, if your goal is to deploy technology to help people, the most useful skills you can bring to the table are human engagement and empathy.

To hear the panel in full, watch the video or read the transcript below.

Transcript

Robyn Scott:

Good morning everyone. I am delighted to be here moderating this conversation, which, in a room of so many smart technologists who’ve done so much to make government better, could seem almost heretical because we’re here to talk about why technology isn’t always the right solution and certainly isn’t always the first solution. Just to get a sense of where the audience is on this issue, can you raise your hand if you have worked on deploying technology to solve a public sector challenge? And can you keep your hand up if that technology hasn’t fully delivered on its promise in the way you expected? Okay, so to help us explore this question, we have fabulous panelists and two situations where the stakes really couldn’t have been higher.

Robyn Scott:

So first of all, we have Adam and Marina talking about foster care in Rhode Island. Now about 450,000 children are in the foster care system across the country and there’s around 3,600 in the Rhode Island system. So that’s about five times the number of people in this room.

Robyn Scott:

Then we will turn to the border crisis where, last year, I believe within the space for months around 2,300 children were separated from their families, and that’s of around 450,000 adults who, over the course of a year, are detained. So we’re talking very large numbers. We have Nina, Viv, and Kavi to talk about that.

Robyn Scott:

Now before we jump into exploring the question directly, just to give a sense of the authority of our panelists to speak to this, I’d like to ask them each briefly to talk about the moment in their career that is most significant that they’re most proud of.

Vivian Graubard:

Thank you Robyn. One moment that I’m really proud of, in 2014 I was a tech advisor to President Obama’s task force to protect students from sexual assault. One day after work we invited six student survivors to the White House to talk about their experiences and aim to understand the ways in which policies were and were not working. One thing that became quickly evident was that none of the policies that we designed in Washington D.C. were working as intended out in the wild, right? All of these measures that we had put in place to protect students were not doing their job. It was an incredibly humbling moment, a reminder that the things that we do in Washington D.C. often do not work as intended and our ultimate recommendations are better when we include the people in them throughout the process who are most impacted by the decisions that we’re making. It was very special, personally, to be able to have these students at the table.

Robyn Scott:

Thank you. Kavi?

Kavi Harshawat:

Yeah. I think for me one of the biggest and most important moments in my career so far has been when we were starting the digital service, actually with Marina, at the Department of Veterans Affairs, and we had been spending months and months trying to get access to a simple … to a team’s API, essentially, to find missing records at the Department of Veterans Affairs. We had literally spent months trying to find laptops and get access to WiFi and fill out the paperwork to get this simple access. Finally after those six months we finally saw data come through this API and we were able to find those records and our whole team was just ecstatic about it, up in arms, super excited about it. That was a pretty big highlight.

Robyn Scott:

Thank you. Nina.

Nina Siulc:

So I learned last summer that my five-year-old was going around telling people that I was going to be responsible for Jeff Sessions losing his job. So I’m pretty proud of that, that that’s what my kid thinks I’ve achieved. The reason she thinks that, I’ve spent the past couple decades doing research on the human impacts of immigration policy, and a few years ago left academia to return to the policy world. I now work on a team that has a simple, bold vision that anybody facing deportation in the United States deserves to go to court with the help of a lawyer. We have built a network of programs that is working towards that goal. At a time when we have people at the highest levels of our government threatening the protections that our immigration laws offer, we’ve been able to use evidence to convince local, state and federal government to invest in deportation defense.

Robyn Scott:

Great, thank you. Adam.

Adam Williams:

Yeah. In fall of 2016 I left my career in Chicago where I was doing marketing for and testing of different public policy, public health programs, in the Chicago area and joined something called Foster America, which, if you’ve never heard of it, is a program that takes talent outside of the world of foster care and child welfare and brings that talent into the field. It was brand new and I had no idea what I was going to get.

Adam Williams:

About two weeks into the training, and that was all we got was two weeks, I must have had a look on my face one day because some of the movers and shakers in the field who had come to kind of help provide the training looked at me like, “You are panicked.” I just said to myself like, “I don’t know, I can’t tell yet if there’s a place for me in this field.” They looked at me and, one gentleman in particular, and said, “If you want to bring your skills to the foster care world, we will have you for as long as you’re willing to contribute them.” I think that at that moment I found a calling that I don’t know that many people have the privilege to get, and so here I am in it for life.

Robyn Scott:

Fantastic. Marina.

Marina Nitze:

For the first four years where I was the Chief Technology Officer of the Department of Veterans Affairs, we were not allowed to use cloud computing because the inspector general said you could not put it in evidence bag. So after four years of lunches and coffees and trust-building and earning political capital, the highlight of my time there was getting a memo from the inspector general saying that we were allowed to use the cloud.

Robyn Scott:

Fantastic. So we have a panel of people who found their calling and have been extremely effective in tackling problems. Adam, I want to turn to you. Can you take us to Rhode Island and the situation facing the foster care system?

Adam Williams:

Yeah, so Rhode Island has historically been one of the worst states in terms of placing children in families when they need foster care. When they don’t get homes, they end up in a group home or a congregate care setting. It’s, in my mind, basically a modern day orphanage, and so I was asked to come there to help them address this issue, and Marina has a specific technology role that she’s playing in this.

Adam Williams:

If you skip forward, that was in 2016, if you skip forward to January of 2018 we had a kind of a mini crisis going on where there was an uptick in the number of children coming into our care because of the opioid crisis and also an uptick coming in because of a situation regarding a principal who failed to report abuse and neglect and then after he was disciplined for that quite harshly, the phone just started ringing on every incident you could possibly imagine that happens in a school, so we were seeing a lot more kids coming into care and needed to take action.

Adam Williams:

We decided at that point to, in an eight-week time period, completely rethink the way that we recruit and train and license our foster families. We set up a convention, not unlike this, and a very small team of about six or seven people managed to bring in 174 families to get trained at once. To give you some context, normally a class would have about 15 families in it. They’d sit in a circle and talk for a number of weeks. We decided to do this in a three-day period where they would come and, boot camp style, get the bulk of what they needed to participate, to complete their licensing. It was also a response, I think, to a demand that we saw from families that the time period was just very long and that there was a barrier from helping that families wanted to do because of, you know, because the format of becoming a foster parent was a little bit cumbersome.

Robyn Scott:

Super. Marina, tell us why were you there working on this problem?

Marina Nitze:

So two Summits ago I was preparing to leave my role at the VA and I didn’t know what I was going to do next, but I knew I wanted it to be in foster care, and so I told everybody that, which is my main advice. Tell everybody what you want to do next. My former boss at the White House, Richard Culatta, had become the Chief Innovation Officer of Rhode Island, and my former colleague from the White House, Sherry Lachman, had gone on to found Foster America. So they invited me to Rhode Island for the day to meet their child welfare team. I met Adam and, I think, totally fell in love with the problem and saw ways that we could use human-centered design and technology to help fix the licensing process and beyond.

Robyn Scott:

So you fell in love with the problem, how did you approach it and what lessons and mistakes did you make along the way?

Marina Nitze:

So the approach was initially around … there was an interest in doing data-driven recruitment of foster families because, as Adam said, there was not enough. But I think when we dug in, and Adam showed me the way and the ropes of Rhode Island, that when you actually crawled through the process, it was so long and cumbersome that a lot of families were dropping out. So we really had to focus on first things first, and not necessarily technology first, but just what were there ways that we could map out a more streamlined process that then we were able to actually try at recruitment weekend and validate?

Adam Williams:

Yeah, I think the weekend really shined a light on, you know, why weren’t these things being done sooner? I mean, so much of this process really could be done if you just like got families in for two hours and knocked out a lot of paperwork. Whereas … and I think that Marina’s work really highlighted some of those things.

Adam Williams:

I would also offer on the flip side that one of the things that we made sure to do when we trained and licensed the 174 foster families is that you cannot use technology to replace the fact that you still have to build relationships with these families. You have to know, you have to sit in their living room and have tea, which we did, even in that eight-week time period. You have to know their dogs’ names and encourage them to get vaccinated. You have to know what their fears are and what their preconceived notions are and you have to address those during the training and none of that can be replaced just because you have kind of an expedited way, or because you’ve inserted technology, you still have to have those intimate relationships because becoming a foster parent is a life-changing experience.

Robyn Scott:

And you were sometimes under pressure to use technology first. Can you talk a little bit about that?

Marina Nitze:

I think a lesson I’ve learned hard in my career by screwing it up a few times is that the last thing we wanted to do was like build a custom Ruby on Rails foster parent licensing tracker, and so as frustrating as it was sometimes to wait and go through procurement processes and official channels, it was really important that the solutions that we aimed for were sustainable beyond me and Adam.

Robyn Scott:

Super. So we’re going to come back to you, but going to go now to the border. Who in this room was involved in some way in the border crisis and trying to solve it? So maybe five or six hands. Nina, can you paint a picture and take us there?

Nina Siulc:

So we at the Vera Institute run and coordinate national networks of lawyers at nonprofit organizations around the country who work with adults and children in detention. At the end of 2017 we started to hear kind of ad hoc reports of children and grown-ups being separated from each other and that began to escalate over the beginning of 2018 and crystallized in what became an official policy the administration announced, just about a year ago now, called Zero Tolerance. Under this policy they were announcing that they were going to start referring for prosecution any adults who entered the country without presenting themselves at an official port of entry. And what that meant was the children those adults came with were rendered what the government calls “unaccompanied.” They were separated from their adults and sent to the custody of the Office of Refugee Resettlement, which runs a network of shelters that detain children around the country.

Nina Siulc:

In the space of about one month, around 2,300 children, we believe, were separated from their grown-ups. And this is a field in which we never have enough resources to serve everybody, and people were left scrambling trying to figure out what was an official policy that was unclear even to those implementing it, that was changing every day. The adults were not told where their children were sent. The shelters that received the children were not told that the children had been separated. Many of them were too young to be able to articulate themselves that they had been taken from their parents. The public has since learned that the many electronic systems the government uses to track people in immigration proceedings were not used to identify the fact that these grown-ups and children had been separated, so there was no official record-keeping that would have identified who was taken from whom.

Robyn Scott:

So the available technology wasn’t even being used?

Nina Siulc:

That’s right.

Robyn Scott:

Okay. So we have the situation, 2,300 kids. Viv, how did you end up there with Kavi?

Vivian Graubard:

Yeah, so I get a phone call on a Saturday morning from Cecilia Muñoz, who spoke here last year, who’s been an immigration advocate her whole life, that there’s a lack of resources. There’s this crisis, we’re all asking ourselves what is the role of technology in helping, you know, to help resources go further and figure out how we can get more children back together with their parents and to find them legal support and legal representation?

Vivian Graubard:

It became very clear, the first time I spoke to Nina, she was shopping at a Home Depot on a Saturday morning and I was like, “What is happening on the border?” And it became very clear that without physically being there, it was going to be super difficult to figure out where a team of technologists could be helpful. And so I emailed a bunch of people and said, “Who is available to go to McAllen, Texas, with me tomorrow morning?” And Kavi canceled a family vacation to hop on a plane to go.

Vivian Graubard:

We arrived, we were meeting with lawyers, we were going to detention centers and we ended up at a shelter called Catholic Charities Respite where they were seeing 200 families per day that were coming from ICE detentions, and helping process them and reunite them with their family in the US, in a shelter the size of a classroom. Adults all had ankle monitors on them. The children had not received any healthcare over the course of two months. They were all wearing the clothing that they were wearing throughout their two-months journey. It was … we had to be there, right, to fully understand the gravity and the level of crisis.

Vivian Graubard:

One thing that became immediately clear was that all of this that was happening was not a problem that technology could solve. We were getting phone calls from all sorts of researchers and consultants from across the country that were like, “How can facial recognition help? I’d love to deploy my new untested DNA testing kit to solve this problem.” We had one funder tell us that they would love to give us money to work on this, but we had to use blockchain. They didn’t know how or why, but like some Microsoft consultant got to them before I did and they were holding us ransom. They were saying like, “We will not fund this unless you can tell us that you’re using blockchain,” and fortunately other people stepped up to help us figure out where we could be the most helpful.

Nina Siulc:

We were having to communicate chatbots won’t be helpful in detention where you can’t take your cell phone.

Vivian Graubard:

Yeah.

Robyn Scott:

But you did need, Nina, you said you kept a database of facial recognition providers, right? To reference.

Nina Siulc:

Yes, I got so many offers of facial recognition technology that we kept a list just in case we ever need that technology in the future. So we’re trying to find a way to put those people to good use.

Robyn Scott:

So, Kavi, you and Viv are in the shelter, incredibly fast-moving environment, a lot of uncertainty, a lot of stress, a lot of distress. Can you talk a little bit about what you did and what you learned?

Kavi Harshawat:

Yeah. One of the things that we saw while we were there is just so many people coming through and kind of like crisis tourism almost. There was just people who were, in gaggles, coming, taking pictures of the scene, reporters visiting officials who would come–

Vivian Graubard:

Celebrities.

Kavi Harshawat:

… take pictures and then like leave, but they wouldn’t actually help. Meanwhile you have, you know, this room full of hundreds of migrants like overflowing into the parking lot. Even the people who were trying to help out who were visiting through, they’re really coming in and, you know, like playing with the kids or maybe handing out water.

Kavi Harshawat:

So we were asking ourselves, “Okay, well what is the actual thing that’s needed, both from a technology perspective and more than that?” From the tech perspective, the woman who runs the shelter, Sister Norma, who’s an amazing person, she basically started the shelter and has been doing this work for years and years … basically came right up to us as soon as we walked through and said, “You guys are from the White House? You guys are techies? I need a TV, I need a bus schedule, I need Chromebooks, I need a better intake form.” And like that was it. She knew exactly what we needed, and aside from that it wasn’t about technology, as Viv mentioned. It’s not like the fancy, flashy blockchains, it was, you know, we literally, the most helpful thing we could do is get cellphones for people to call their family members who they’d been separated from, and buy a bunch of McDonald’s for dinner for everybody in the shelter.

Robyn Scott:

So you guys ended up working together, the three of you, but in fact they didn’t meet until earlier today backstage. Nina, can you talk a little bit about the second project?

Nina Siulc:

So we realized at the Vera Institute that because we run this national network of legal service programs, we actually knew which lawyers had talked to which children everywhere around the country. What we needed was to find the grown-ups so we could put them in touch with the lawyers who were seeing separated children. So we launched a project we called the Immigrant Connection Network. We set up a war room overnight. We got a cellphone, we assigned multilingual staff to operate the war room and then we were put in touch with Viv because New America was looking to put its technology to good use.

Vivian Graubard:

And like let’s talk about failures because this has kept me up very often over the course of the past year. You know, the moment of a crisis, I have decided, I like sort of knew previously, have decided definitely, is not the time to deploy any sort of new technology. There were moments when I was having to pick up the phone to call Nina or someone on her team and say like, “Is this thing that we’re building right?” There’s this ragtag team of volunteers that I’m so grateful for, but already I feel like we’re making bad technical decisions, like not just using Airtable because we should build something custom. I mean … it was very stressful.

Vivian Graubard:

The question that I had to ask myself every time was, if I call Nina right now, am I going to take time away from actual lawyering and legal resources and helping people on the ground to ask them some question about technology that we could build, when email and emailing Excel spreadsheets, like might give everyone in here like all sorts of nightmares, is actually the most productive thing to do right now? So every second of it felt like I have to be very careful about what I’m asking because it is almost irresponsible and reckless to think that my goal of figuring out how technology can be helpful is actually the most important thing at the moment.

Kavi Harshawat:

Yeah, and I think that, to build on that, one of the … even the technology that we were trying to build, it just wasn’t the right time or the right environment to build it, right? Because there was, you know, a constantly changing environment, I think faster changing than we, either of us, had ever seen working for the federal government, where the administration was changing its policies on a nearly weekly basis in response to ACLU’s lawsuits against it and just other circumstances.

Kavi Harshawat:

So the things that you tend to rely on, even if you’re doing all the right things, like you’re taking time to consider security and you’re taking time to consider how you might iterate on this product and build the right thing and gather information about the environment on the ground before just building a product, you try to do all those things, but it just doesn’t work out in that situation because you’re trying to build it in the midst of this emergency situation, this crisis.

Vivian Graubard:

There were more lessons that we learned about the critical infrastructure that needs to be set up prior to a crisis, like the way that emergency planners think about work, then in the moment of a crisis. Unfortunately the immigration space has not not been in a crisis for, you know, a decade now, and so finding that time to talk about information sharing between, you know, 30 organizations across the country is a luxury that we haven’t had in this space for a long time.

Robyn Scott:

So when we look at what works around the world at Apolitical, we often come to the conclusion that trust has to be the fundamental platform. Tech is not necessarily the fundamental platform. Marina, can you, I mean there’s much that resonates between these two stories, can you talk briefly about trust building and what you’ve learned about how to accelerate that building?

Marina Nitze:

Sure. I think it’s really, it’s about the human connection, and as can be imagined too with Sister Norma, it wasn’t about showing up with a fancy form. It was about him cleaning up vomit off the floor. And that was what impressed her enough to build some trust. I think Adam did a really amazing job of this in Rhode Island too, which was not showing up and saying, “I’m the amazing expert, I’m going to fix all of your things,” it was being really humble, helping them refine their excel spreadsheet so that it worked more effectively, and kind of working with people where they are and earning their trust that way.

Robyn Scott:

Super. So just to conclude, I’m going to ask everyone on the panel to share the question that they would advise others to ask themselves in these fragile, high-stress situations, to avoid embarking on the wrong solutions too quickly. Viv.

Vivian Graubard:

Yeah, I mean, am I doing more harm than good? We are all going to make mistakes in this incredibly difficult work. So why are you doing it? And if the answer keeps you up at night, you should reconsider your path.

Robyn Scott:

Kavi.

Kavi Harshawat:

I think considering whether, you know, if there’s an emergency or crisis that you are not thinking about today, that you should be planning for and preparing for in the future, you should think about that.

Nina Siulc:

Yeah, I think we know that high-tech solutions aren’t always the answer to the low-tech immigration field that I work in, but we also have learned that we really need to pause and build the infrastructure so we can take advantage of what technology has to offer before the next manufactured crisis comes. We’re already seeing that children are beginning to be separated from their parents again and we still don’t have that infrastructure in place.

Robyn Scott:

Adam.

Adam Williams:

Yeah, I think the question that I would ask to myself is have I really engaged with the people that I’m trying to serve? And assume nothing. I think too often we think that we know what people need, and rather than like engage with them and have them at the table, so that we’re co-designing things in a way that makes sense for them and addresses their most immediate needs first.

Marina Nitze:

Yeah. And then ask yourself also, is this solution something that the organization can sustain beyond me? Or will they be calling you until you’re, you know, 85 years old to fix bugs? If you answer these questions correctly, and would like to work with us, Foster America is recruiting for our next cohort at foster-america.org.

Robyn Scott:

There you go, your call to action. There’s so much more to say on this question, but we’re at time, please give a big round of applause to our inspiring panel.