Code for America founder Jennifer Pahlka has spent her career working to bring a more user-centered approach to the way government integrates digital tech. One example? Building a mobile app for SNAP applicants in California that shrank the process down from an hour to seven minutes.

“We have this incredibly structured way of thinking in government where it’s sort of a hierarchy,” says the former U.S. Deputy Chief Technology Officer under President Obama and the co-founder of the U.S. Digital Response. “That’s really hurting us.”

You can hear more from Pahlka and details from her new book, “Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better,” during this lively conversation with Signal’s John Battelle in the video below. Or read through our lightly edited transcript.


John Battelle
Welcome to another Signal Conversation. I’m very excited about this one, a former colleague of mine who’s gone on to great heights, Jennifer Pahlka is joining us, she’s written a new book. She is a former U.S. Deputy Chief Technology Officer under President Obama, a founder of Code for America, and the co-founder of the United States Digital Response. Her book is “Recoding America, Why Government is Failing in the Digital Age and How We Can Do Better.” In this book, she makes the case that government at all levels has limped into the digital age. It’s a problem that affects more than government, any large organization can get in its own way, and we’ll be talking about that. And there’s much to learn from Jennifer’s work. So welcome, Jennifer to the Signal Conversation.

Jennifer Pahlka
It’s a delight to be here, John, thanks so much.

Let’s just start with this book. Just tell me, what prompted you to write it?

I started to see over the years working with and in government, that we’re in really an implementation crisis, and we don’t talk about it. What we talk about, are these partisan fights these fights about values. While I think the fact that we are failing to implement, the things that we do agree on is sort of depressing. It’s also, I think, more hopeful, because it’s probably more solvable than our partisan bickering. There’s actually a lot that people do agree on across whatever the aisle is today, right. We’re in the middle of some strange political realignment. I think in that realignment, there’s actually a lot of agreement that we need to just get together and get government to a place where, though, we’re still going to disagree on some things we can do the things we have agreed on. We have to build the capacity to actually execute on the laws and policies that are passed. I think if people think that way, we can spend a lot less time churning and a lot more time getting stuff done together.

It is strange that we seem to assume that things get done, if we agree on them, and they get passed into law. I think we just throw them over our shoulder and assume they’re getting done. But you started an organization Code for America, which actually attacked the problem, at its root of how government can use digital technology in a different way. Can before we get deeper into the book, can you tell us about Code for America, what its mission is and why you started it?

I think why I started and what I got out of it ended up being a little different in the sense that I like everybody thought that, right? I grew up on Schoolhouse Rock with, “I’m just a bill sitting here on Capitol Hill.” And once the bill gets passed, everybody cheers, and we’re done. When I started Code for America back in 2010, it was there’s something that the tech industry can bring, particularly the web to a world, which I learned from you, John. I got to watch as you were sort of telling the story of this lightweight, user-centered web that was very different from what had come before, and thought, “Oh, my gosh, there’s an incredible value to doing this in government,” that’s actually probably the best and highest use of the principles and values that you and others articulated at that time. But I think what I really learned in doing it was, let’s bring tech people into government was, yes, government has something to learn from tech, tech has a lot to learn from government. But really, it’s not just different programming languages in the cloud. It’s that there’s a fundamentally different way of thinking in the consumer tech world, in that web to a world, which starts from the bottom up instead of the top down and closes the loop. We have this incredibly structured way of thinking in government where it’s sort of a hierarchy. That’s really hurting us. Code for America really started as just a way to bring tech people into government. We worked with city governments. Today, it’s a much larger organization that works more commonly with states, also with the federal government, but they just really show how service delivery can be great, how it can really work for people. I think from that you back into all of the profoundly different ways that those folks are approaching a problem as opposed to the legacy ways of thinking in government.

You’ve had a unique viewpoint in that you with Code for America, you saw, you did a lot of work with local governments. You then went to the federal government and saw inside that sausage factory and how things worked or did not work. You were part of a group of really optimistic people who sort of surged into Washington and said, “We’re going to change this place.” Do you think that change occurred? It’s been almost two administrations since you joined and subsequently moved on. What did you learn? Do you have hope that that kind of change at the big level, the federal level is possible?

Oh, change is absolutely happening. You know, it’s evidenced in part by the fact that there’s all these teams now doing real user-centered work in government that just didn’t exist before. If it happened, it was sort of scraped through the edges, and people were sort of cheating to do it. I went back to DC, before the book launched, but when a number of people had read it. I’d sent it out for people to read, and I was afraid that people were going to react negatively to it. But in fact, people really welcomed it and embraced it, and were excited to get it out in the world. It was then when I started hearing from people,”Your critique of government is correct, we have had this sort of artificial separation between policy and implementation, not enough focus on implementation, but it’s changing.” They were eager to give me all these examples of the ways that delivery, to use another word that really essentially means implementation, or sort of how the public experiences implementation, delivery is becoming a much bigger focus. There is much more tension on it from the parts of our particularly federal government that would never have paid attention 10 years ago.

That’s a great legacy in and of itself. But if we can dive into the book, you identify a number of what you call core concepts throughout the book, and you tell some really entertaining stories about how things didn’t work or got screwed up. I mean, successes as well, of course, but I want to sort of dive into some of the structural problems you noticed and noted, because I think even though the audience here at Signal is mostly corporate, I think they might recognize some some similarities. Large organizations are large organizations. One of them has to do with an idea of a procedure fetish, or this idea that it’s more important to follow the rules and check a box than it is to actually accomplish something. Can you can you unpack that for us?

I encourage folks, of course, you read the book, but if you don’t, Google Nicholas Bagley’s The Procedure Fetish. It’s a short paper that is well worth your time. I think what’s exciting about his work is that he really unpacks what drives this desire to layer procedure upon procedure upon procedure. It’s anxiety about legitimacy. You’re trying to defend that whatever decision you made is legitimate through process. But then we have so much process that really accumulates. It’s not like you there’s a process and then when a new one comes along, the old one goes away, they get layered upon each other, which is how you get this sort of ossification. But you get to the point where it’s so ossified that you’re not getting anything done. Then we really do then have a real crisis of legitimacy, weirdly driven by people’s anxiety about legitimacy. You see that all over government. Of course, people blame the lawyers, but it’s really, I think, the kind of thinking that says, it’s more important that we’re safe in this particular moment than that we pull back and look at, are we actually getting the job done? I can tell, you know, any number of stories from the book about that, but I do think that folks and corporations tell me very often it resonates. We’re doing that too. We’ve got to figure out a way out of this proceduralist trap.

Yeah, and it’s interesting because large companies and oil companies for-profit companies, which of course is not the government’s MO, supposedly, although we suppose sure we could spend a long time talking about the role of money in government. But another one of your key concepts is rethink input. The people that you need to hear from in those valuable input loops. How do we do a good job at whatever were delivered. But the people that you say the people that we need to hear from are not participating. Certainly, that could be said in large corporations as well. Can you unpack this concept of rethinking input?

I’ll tell you sort of one of the ways that it really started to hit home, for me, when I was working with the team at Code for America. We had been redesigning a process for applying for SNAP, Supplemental Nutrition Assistance Program or food stamps. We were doing that because the process in California, if you went online to try to get food stamps in San Francisco County, where we were located, it was a 212-question form, it didn’t work on mobile. So while California is going, “Hey, we want to dramatically increase our participation rates,” California’s a very pro-welfare states, and it had the second lowest participation rate in the country. The only state that was worse was Wyoming. They’re not looking at the fact that, it’s not the policies, it’s not the outreach, it’s the fact that once you go to that form, if you’re not on a mobile phone, if you don’t have an hour to give, if you don’t know how to answer some of those questions, you’re just not going to get through the application process. So we put up this form that you can use, that took about seven minutes on a mobile phone, it was much easier to use.

When a colleague of mine was asked to go, sort of show the form and talk about this practice of user research in front of the consortium that had created that 212 -question form, what he saw was that there was these, representatives from the counties, and they all voted on what features were in this online application. You really had you know, this 20-something, and I call them bureaucrats, but I don’t mean that in an insulting way. They’re their public servants doing their jobs, trying to represent the needs of their county, which each county does a little bit different, and nobody in the room representing the actual users. So when my colleague Jake is up there showing what it’s like for a user to actually try to get through this form, he’s actually bringing a lot of new knowledge to these folks. Because in that case, input is all from the internal stakeholders and meeting their needs. In the language of the book, and my colleagues in the UK, it’s meeting government needs not meeting user needs. You see that all over the place.

You’re right, I think often in corporations, we have that same pattern of getting input from a lot of internal stakeholders, and then fighting over this department needs this and that department needs that, and not looking at the end users. In government, we have an additional problem, which I think is worse than in corporations, which is, we also are required to go to the public through a process called Notice and Comment. There’s various versions of at all levels. So we’re going to make a change to regulations, we put this out, people are able to comment on it, government has to respond to those comments, at least in some way. Of course, that process, which should center the people who are actually going to use the program, or the product, is hugely captured by special interests. To read the Federal Register, you are almost exclusively going to be paid to do that. The average person is not really going to go out there and do that. So we really have to look to what I think of as the discipline of product management. We can borrow from the tech industry, where product management says, “You don’t just let everybody tell you what they want. You really understand the needs of your users and figure out elegant ways of delivering that.” But that means an active approached input, going to the people who are going to use your service and making sure that their needs are centered. Then you can take care of all those internal needs and handle all those stakeholders, but you’re grounded in something that ultimately is driving towards a product that will work for people.

Did you find, just going back to that SNAP example, I mean, it’s pretty extraordinary to think you had a seven minute experience at the end there, and you started with a 212-question form that was impossible to navigate. Was there resistance to the idea of simplifying it because people were somewhat addicted to the complexity?

I think in that case, there wasn’t so much resistance in the sense that we did not ask their permission, we just created this extra form. Then people are either very excited by it, like Leo O’Farrell in San Francisco, who embraced it with open arms and tried to spread it across all the counties, or yes, they were resistant, because they had their internal business processes and this messed with that. So you had this fight between government needs and user needs.

But I think there’s so many ways in which I’ve seen people in government, with all the best intentions, who are really dedicated to what they do, yeah, embrace that complexity, because it is their sense of self worth, essentially. They know all of these internal workings of this very complicated regulation, and that’s what sort of makes them special.If you want to fight against that, you kind of have to recognize that and honor them for it, instead of degrading the value that they believe they’re bringing to the public today.

In companies when you tried to make systemic change, often, what you need to change is what’s being measured, and valued. Did you find that to be the case when working with government as well? Obviously, California very much wanted to increase engagement with in this case, SNAP. Was that a metric that became at the forefront this is our goal, we’re trying to hit this. Is that one way of creating systemic change by changing what you value as metrics of success?

One thousand percent. I think that is what made get CalFresh, this alternate app that we did successful and ultimately embraced by the state. But what’s hard about that is that it’s very hard to align a large number of stakeholders around a central metric. The other thing that can really go wrong in that approach is that people in government are very used to any kind of data or measurement being used against them as sort of a weapon. So I talk in the beginning of the book about my work with unemployment insurance backlog in California during the pandemic, and how the leadership at the Employment Development Department ,which administers UI, Unemployment Insurance, really was very resistant to a lot of the work that we were doing to try to give them some ways of understanding what was going on. Where is your backlog? Is it going up? Or is it going down, and being able to sort of analyze the assignment of the staff that they had brought in. It’s because to them, data measurement, performance metrics, are a grade that they get given, and not a compass that they use, it’s always negative. And it’s always sort of external. I think one of the keys to transformation is if you can get people in sort of those mid-levels to embrace the idea that that measurement is a tool in their hands to get where they want to go instead of something that’s being imposed from the outside, that will always just be something that makes them look bad. That shifts to a compass, rather than a grade can be profound.

That’s a great insight. Have you engaged with large companies in this work? Or maybe the launch of the book has gotten some people calling you. Have you have you found that there might be even more to do in large corporations than you thought perhaps earlier?

I will say that at a lot of talks I had people come up to me and say, “I work in a big corporation, and a lot of these problems are the same or similar.” I think a lot of the principles are relevant. I do think there’s an extra layer of complication in government in the sense that you don’t have a CEO, really. You have very distributed decision making and with Congress in the three branches of government and all the levels of government, it does make change even harder.

I think maybe there’s some comfort in that for everybody in the audience. It’s like okay, It’s it couldn’t be worse, right? Lastly, I just want to because I don’t want to take too much of your time I know how busy you are. If you had advice for this audience, people like those who come up to you at the end of your talks, as to how they might best affect change in large organizations, what would that advice be?

We’ve talked a little bit about the need to center users, and that that would certainly want the one piece of advice. But if I could add on to that, I guess I would say, for people who think from a sort of a top-down perspective, you’re usually going to think in terms of policy and strategy. It’s important to look at the ways in which those high-level frameworks get what I call, eaten by culture. There’s, of course, the old saying, culture eats strategy. In the book, I talk about how culture also eats policy. You have a particular intention. But if it is falling into a culture that essentially perverts it, it reverses the intention, then you have to stop playing with those strategy and policy levers, and start paying attention to how you can shift the culture. You’ve got to take your foot off the policy and strategy break for a little while. The only way to figure out what’s going on in your culture is to spend a lot of time with people you wouldn’t normally spend it with and get down there and see how are the practices that they are using to execute that strategy no longer faithful to the intent. What’s happened there and why have they made those choices? You’ll learn a lot and you’ll end up sort of, I think, strengthen the foundations of your company or organization in a way that’s sort of not extractive. You’re not trying to get a particular output at that point. You’re just trying to strengthen the foundation so that anytime you have a strategy goal or a policy goal, you’ll have more strength to get that that when you get there.

Well, that’s great advice. Jennifer’s book is Recoding. America, Why Government is Failing in the Digital Age, How We Can Do Better. aEzra Klein, who is someone I certainly follow, called it the best policy book he’s ever read. So it’s very high praise. I suggest everyone watching this go get that book, there’s lots of lessons no matter where you work, and what you do. Jennifer, thank you so much for joining the Signal Conversation.

Thanks so much, John. Really love being here.