Recoding America: A Conversation with Jennifer Pahlka

Audio of this conversation is available via your favorite podcast service.

In the United States, it’s fair to say that federal, state and local governments have struggled in the era of digitalization. Decades in to that era, there is still a gap between the policy outcomes we seek and what citizens often get when they engage with government agencies and services online. At its worst this gap means people aren’t receiving critical services that sustain their lives; and at the very least it reduces faith in government to be able to solve problems right at the moment when it’s clear the collective challenges we face are going to

Jennifer Pahlka, who served in President Barack Obama’s administration as deputy chief technology officer and founded the nonprofit Code for America, has written a book that asks us to reexamine how government works, and how it should work, in the digital age. It's called Recoding America:
Why Government Is Failing in the Digital Age and How We Can Do Better, and it's the subject of the podcast today.

What follows is a lightly edited transcript of the discussion.

Justin Hendrix:

Just to establish your bona fides in this space, can you just kind of give the listener a little bit of a background about your experience working at the intersection of tech and government?

Jennifer Pahlka:

Yes. I'm not really a technologist. My first job out of college was at a child welfare agency, but I ended up working in tech media and helped create these Web 2.0 conferences when that was a big deal. And from that, got interested in this concept of Gov 2.0 and that's what inspired me to start the nonprofit I started in 2010 called Code for America. Still around doing amazing things, though I stepped down from it just before the pandemic. But a couple years into running Code for America, was asked to come to the White House and be the deputy chief technology officer under Todd Park. And that's where I helped start the United States Digital Service. Went back to Code for America. Also during that time, I served on the Defense Innovation Board, which the wonderful late Ash Carter started. And I now serve on the board of the United States Digital Response, which is a different organization that allows techies to volunteer with government, the Tech Talent Project, and a couple of other great organizations.

Justin Hendrix:

So this book, to some extent, starts almost with, I guess, a change in your life. You were leaving Code for America, it was just prior to the onset of the COVID-19 pandemic, and I had this sort of almost Godfather like sense of just when you were out, they put me back in. You got mixed up in trying to fix set of government systems in the COVID crisis.

Jennifer Pahlka:

Yes, I was trying to take a break and write this book, actually, but around about July of that first year, we had accumulated in California a very large backlog of unpaid unemployment insurance claims, and the governor convened a task force to go in and figure out how we could clear them and ask me to co-chair it. And I just didn't feel like I could say no, given how really dire the situation was for so many people who had no income and were not getting their checks. Yeah, that was a really interesting experience, which I think gave me a way to tell some of the stories and share some of the observations that I think are true of other government projects, but were just very visible there at the EDD.

One of those is that people really think that we have these systems designed to do something, but when you go look at this sort of complex web of fragile systems that don't really work together, you realize it's more sort of layers of technology that have accumulated over time rather than been designed for a particular purpose. And the other big learning there was that the technology is complex and layered, but so is the process and the policy that governs it, in ways that I think the average person really doesn't understand.

Justin Hendrix:

A lot of this book seems to focus on this gulf between expectations and reality. That's a phrase that you used at one point early on. The expectation of course that technology is going to make things better and easier, the expectations that we now have as users of technology that certain things should be possible when I arrive at a website and want to receive a service. I see this book as sort of plumbing that gulf to some extent, trying to figure out where it comes from. And you've just pointed out one place that it seems to come from, this kind of archeological layering of tech, but also policy. Let's talk about those things a little bit in more depth. How do you see those kind of archeological phenomena playing out?

Jennifer Pahlka:

I think that we have this expectation that if a law or policy that we like passes, makes it through the process, gets signed by the president or whoever, that's when we celebrate and that we don't really hear a lot about what happens next until we're in one of these crises. But we have and have had for I think some time a problem with what happens next. And that's where we get the disappointment from our expectations. As with the unemployment insurance crisis, you can have sort of not enough happen. Congress appropriated a lot more money for unemployment insurance during the pandemic, but only a part of it got out to people. Or in the worst cases, you can actually have the opposite of what the policy intended happen. And that can be really, really tragic. Maybe I'll give you a case that's an example of that, not from the EDD, which we can return to, but records clearance.

Many states have passed laws that decriminalize marijuana. And in doing that, they have also made provisions for people who have passed marijuana convictions to have them expunged from their records so that they don't have to live under the burden of a felony, which means at the very least, that they pretty much can't get a job. And the problem is that sort of year, two years in for these things, the law has changed, but for the people who have those criminal records, nothing has really changed. They'll still go try to get a job. And when they do the background check, they still see a felony on their record.

And it sounds really like a failure of pretty basic technology, but I think that's a little misleading. Because it's not just getting the databases to talk to each other, it's the ability, and I think frankly, the will in government to pull back and look at the whole process from a design perspective and say, "Wait a minute." What's happening here when someone tries to clear a criminal record is that the person with the record themselves needs to go collect all sorts of government documents from different government offices, try to fill out very confusing forms, file those forms with other different government offices, wait to hear back, get notices they don't understand, show up in court, then hopefully actually get the expungement and then do the whole thing over again where they take the expungement back to those offices and hope those offices actually update their databases.

Really none of this is necessary and I think sometimes, we think the system is designed to make it hard and that is obviously sometimes true, but very often, it's simply not designed at all. We have these policies and processes and tech systems like at the EDD that have simply accreted over time and it's not so much the difference between user-friendly design and what we would call in tech user hostile design, but more kind of the difference between any design at all and just letting it accrue and accrete. Sort of a no design. Of course, no design ends up as being very user hostile, but it's less the effect of intention and more kind of neglect. I do struggle with the word neglect though, because that implies that it's someone's actual job to do this, and very often, it isn't. I mean, people have very specific jobs they're told to do, and designing systems just is very rarely one of them.

So when we think about this gap between our expectations and what we get, different people have different ways of thinking about it. On the left, we often think of it as the fault of the right because they are starving government by design, they're starving government of the resources they need to get the job done, not giving them enough money. We really do spend a lot of money on government operations and it might be helpful to think about how we're starving government of design, not just by design. I mean, I would argue that if you looked at the records clearance process and you took it for granted, you took it for what it is and you said, "We need thousands more people to come help these records get cleared and get people through this system.", that's the wrong response. The right response is to say, "Let us redesign this so that it actually works." And that is actually what we did.

The city of San Francisco and Code for America got together several years ago and designed a process for automatic expungement, which doesn't take that whole process for granted. It says, "Nope, these are just fields in a database." We can ask the computers to find all the records and change all the records. And none of that is necessary. So I think if we're going to close this gap of implementation, this gap of our expectations, we have to stop asking the wrong questions. Our leaders, our policy class, political class is always asking how much money do we need for a particular task and not asking what are the competencies and capacities that we need, not just to achieve this task, but to have those competencies so that everything that we do is easier and we actually get the job done, we get those outcomes that we want, which I think is just another way of saying, "How do we build the kind of state capacity that we need today?" And I just think that is a far more existential question than people are giving it credit for.

Justin Hendrix:

One of the things that I learned reading this book was maybe some of the historical reasons why especially the US federal government might be so bad at doing the sorts of things you're talking about, about the degree to which procurement as a function has been defined in some cases by legislation, handed to particular agencies, and there's this kind of logic in the way that, especially the federal government does business, that seems to almost be at odds with what you're describing.

Jennifer Pahlka:

I think it's a logic and just a way of thinking. So two historical things that happened in the '90s. One was just the trend of outsourcing. So if you think about when technology started to move from, "Oh, it's cool that we have personal computers." to "Oh, there's the internet and it's profoundly changing our society." That kind of started happening in the '90s, and that is also when we went on a huge outsourcing spree, not just tech. I mean, it was sort of this decision that anything that could be outsourced should be. There's this concept of inherently governmental and a definition of that, and that's sort of the only thing you're not supposed to outsource. But I think the issue is that by the time we started having these discussions, we'd already decided to think of technology as a commodity.

There's this story of the government trying to make its own steel. It was a terrible idea. Government, that's a commodity you should just buy on the market. Will always be cheaper and better to buy it. But tech is something that... I mean, I like to say technology and software is something you do. It's not something you buy. You may buy tech tools, but if you're trying to get things done through technology, it has to be a core competency and something you actually do. So we see that in 1996, you had two members of Congress, Clinger and Cohen, who brought some legislation to the table that became the Clinger Cohen Act. And one of the things in that legislation was asking OMB, the Office of Management and Budget, sort of the most powerful part of the White House, to take on a digital strategy. And what's really telling in this is that OMB didn't want it.

The head guys there said, "Look, this is operational in nature and inconsistent with the policy role of the institution." Because they were seeing this just as another tool of implementation and therefore the kind of thing like steel that should always be outsourced. I think that they really mistook this shift in our society for just another change in the tools that implementers would do, but they also were operating that way because culturally, in government, policy is the important thing. That's what important people do. There is a hierarchy, they are at the top. Implementers of all kinds are at the bottom, and technology is sort of the bottom layer of implementation. It's like the last thing that someone asks anybody to do. So this very rigid hierarchy and culture is, I think, also part of the reason that we have not wanted to have tech competency in the most important parts of government, even though I think we can now see how truly important that is.

Justin Hendrix:

I found myself kind of wanting to raise up and look above the parapet, which you do a couple of times in the book looking internationally, you name federalism itself as potentially one of the boogeymen here who we have to contend with. Is there something wrong specifically with our system in the US or is this a sort of problem we see in other governments around the world? Is someone doing it much, much better in a way that we should admire?

Jennifer Pahlka:

The interesting thing is that yes, a lot of governments are doing it a lot better. I think they're doing it better for... I think it's surprising who is doing it better and who struggles really look a lot like ours. So I think it's important to distinguish between who has the same problems that the US does and who has different problems. So smaller nations with less legacy complexity of the kind that I talked to, you see at the EDD, their problem is just digitization of processes that are not that crazy. Our problem, the UK's problem, Canada's problem, I think to some extent like Australia and New Zealand, though I think less so, is the complexity itself and the, because simply digitizing what we have really, really doesn't work.

We'll return to the story of the EDD, I think, for another good example there. My colleague was working with claims processors at the EDD and trying to understand where the bottlenecks were, where could we intervene that would clear this backlog. And one of them kept saying, "I'm not sure I know the answer to that question. Let me go check. I'm the new guy. I'm the new guy." And finally, after about 10 times, she said, "Okay, how long have you been here?" And he said, "I've only been here 17 years. The folks who really know how unemployment insurance works have been here for 25 years or more." So for example, that kind of complexity is extremely hard to digitize in a way that's going to make sense to people and is going to scale in a crisis where you suddenly have 10 or 15 times as many applicants as you had.

Justin Hendrix:

You've got, it sounds like, a labyrinth of policies and systems that no one really understands, people even who've been there more than a decade who can't quite grasp the bureaucratic complexity.

Jennifer Pahlka:

That's right. And it's changing under them too. There's all these changes in policy and process because of COVID. And so you find out that this is really the key to clearing the backlog because what the governor has done and the state legislature, is said, "Great, we need to clear this backlog. We're giving you 5,000 new people who can help." Not really realizing that those 5,000 people couldn't help. They didn't have 17 or 25 years of experience, but what they did do is take up all the time of the experienced claims processors. So in effect, well, not in effect, we were able to actually show very concretely that every person that the EDD hired was slowing processing of claims. That's a kind of complexity in your system that is going to need to be tackled independently if you want a system that scales. I think there's many other countries that are dealing with that, and it's a different problem of getting government services to work than the governments who just need to take what they have, which has not yet gotten that complex, and put it online and make it work for people.

Some of the governments that I really want to go visit and see in greater depth because I hear about the great work they're doing are Bangladesh, Indonesia, Peru. India famously has really great payment systems and great identity verification systems. They're probably doing things that would be hard for us to do with our culture. We're not very inclined to give a national identity to everybody, but I do see places where Bangladesh has made it a real point not to just put anything online before they make it simpler and easier for people to use. Ukraine has done a fantastic job of this, and they did this before the war started. They had an app called Diia that would handle benefits for instance and digital identification, and it made it really easy to automatically verify people and then put money in their accounts.

Well, that's great before a war starts, but it's really critical once you're under attack. Being able to find their people, give them benefits, their people being able to prove who they were without their passports once they'd gone over these refugees, all of this stuff was really critical and it let the government focus on job one, which was fighting a war. They are ahead of us and I think we should take notice.

Justin Hendrix:

Somebody might be listening to this and say, "Well, actually, that's the great beauty of our system. That we don't allow centralization of government and we don't allow centralization of these types of services. And that preserves our freedom and liberty." And you're talking about in some cases, I guess, technocracies that are, in some cases, tilting towards authoritarianism or at least towards perhaps more autocratic direction on some level, particularly in India. How do you square that? Is there something that's in our federalist system that somehow preserves government becoming too powerful, but then also, of course, creates this side effect that government seems ineffective?

Jennifer Pahlka:

Well, look, a national digital identity is not the only way to do identity verification. We have a project here in the US called login.gov that works very, very differently from how the Indian system works and very differently from how China identifies who its folks are. And I think it's much more consistent with those values. And because it's designed to be so. Again, I'll return to that word design.

Justin Hendrix:

I like that answer because you're essentially kind of smartly dismissing my question on some level. You're saying that it's possible to kind of do both.

Jennifer Pahlka:

Yeah, but I also think that those who are worried about an intrusive and overbearing government need to recognize how much what we do to make a smaller, less powerful government actually results in really burdensome services. So when we don't allow anybody to make a decision, a design decision, for instance, when we have multiple stakeholders, when we slow down how things work in government, we end up with what I call policy vomit. Sorry for the indelicate term. But for instance, in California, the application for SNAP, food stamps, food assistance, had 212 questions on it before we worked with the government on a redesign. And that's not because somebody wanted 212 questions. It's because we made the power to create that form very diffuse. Lots of different people had the power to say what could go in it, so we don't have one central authority that could simply make something that worked well. And people experience a 212 question with really kind of insulting and strange questions on it as a really overbearing government, but that came out of a desire to keep anybody from accumulating power.

Justin Hendrix:

One of the things that you talk about specifically that I kind of found interesting, I hadn't thought about, but of course it's true, you talk about the ways in which Liberals and Conservatives sort of try to hold government to account in different ways. You talk about Liberals' pension for suing to enforce regulations, so essentially trying to get the government to do the thing that Senate was going to do. And then Conservatives suing to stop the government from doing the things that it wants to do. I can imagine if you are in compliance or you're also developing agencies or agency approaches to software, it must get really confusing.

Jennifer Pahlka:

Not only do liberals sue to stop, they also are really obsessed with more procedures. There's too often this belief that following the procedure is what gives government legitimacy, not actually the outcome. So that's their sort of dysfunction that I think is the mirror image of the right's dysfunction of trying to make government small but accidentally making it big. I also just not sure that the left and right perspectives are the most valuable here when you're talking about this agenda. There's also this divide, not between left and right, but between people who want to build state capacity, improve the ability of government to do what it says it can do, and those who want to destroy it to really dismantle it.

And those boundaries do not really respect the sort of traditional boundaries of left and right. There are plenty of people on the left who are deeply, deeply skeptical of state capacity for good reasons. And surprisingly, I think there's a lot of people on the right today, not people in the Trump camp, obviously, who think that we actually do need to do what we say we're going to do. And they're sick of this sort of sloppy, overbearing government that we have created and they want to get things done, create housing, and implement some of the provisions of the IRA that are so critical.

So maybe a more relevant divide in a certain sense is people who are thinking about state capacity at all, whether they're trying to build it or tear it down, and those who don't, because the people who are thinking about it, whether their solution is to improve it or degrade it, I think are reacting to the same thing, which is that government is simply too slow and ineffective right now. I don't agree with those who want to be able to fire all public servants. I think we should be able to hire more public servants because when you are frustrated in government that things are moving too slowly, what you need are more good new public servants who can help make these systems work well to balance out those who are really, really resistant to change.

And I think that we should always be dealing with this tension between agility and stability. I don't want to get rid of those folks who stand for stability because there's a lot of reasons we actually need stability in our system, but to me, being able to hire more people to create that correct balance between agility and stability is the way we should be building state capacity. And I don't think that that is a left or right agenda item.

Justin Hendrix:

This book recounts a variety of different scenarios with particular agencies. One that you focus on a little later in the book especially is the Centers for Medicare and Medicaid Services.

Jennifer Pahlka:

CMS.

Justin Hendrix:

CMS. You tell a story of Yadira Sanchez. Who's Yadira and why does she come into this book?

Jennifer Pahlka:

Yeah. When I talk about needing to be able to hire people, I think about Yadira who is a career public servant. She has now been at CMS for about 25 years and she is currently trying to hire product managers. And I want it to be easier for her because she is such a fantastic example of the kind of public servant that we need. So she had been at CMS for quite some time as an IT project manager. Very successful one who did already kind of colored outside the lines. She liked to talk to users about what they actually liked and didn't like in the software, which is sort of not technically the job of a project manager. And when healthcare.gov launches and crashes on its first day, it can only serve eight people, sadly, she of course gets thrown on this project because she's just a fantastic problem solver. And the key thing that happens there is not just that she helps bring healthcare.gov back to life, but she gets exposed to what's known in the software world as agile user-centered development.

I would argue she was always doing it, but she didn't have the framework for it. She told me she never heard the word agile until healthcare.gov happened. You had these sort of catalysts coming from the outside. Much has been made of the, quote, unquote, "rescue team", that my boss Todd Park brought in to help get the site back up and running. But it was really people like Yadira Sanchez, who kind of got... Didn't get as much attention, but they were so critical to that effort. And she comes out of healthcare.gov really wanting to change the agency long-term, not just be able to respond in a crisis, but be able to do it right the next time. Soon thereafter, CMS gets their next legislative mandate. It's called MACRA, the Medicare Access and CHIP Reauthorization Act. And people like Yadira are now determined to do that one right. And she quickly realizes as she knew that it's not going to be just a question of agile development and the trappings of that, like stand ups in the morning and these things that people associate with agile software development. They encounter immediately problems of policy.

So MACRA is a law designed to improve the quality of care that patients get by paying doctors better for better outcomes, not just for doing more procedures, but what Yadira and her team know is that doctors are fed up with CMS, they're fed up with how Medicare asks them to report their quality data already. They find it confusing. They don't know how to use the systems. They feel like when they put their data in, they never know if they've done it right and there's no way to know. And so they're varied that a year later, they're going to find out that they got paid less because they didn't actually format it the way that it was supposed to. They have spent lots of money and lots of staff time on electronic healthcare records. And in this context, this new law has calmed down and the only thing they like worse, they like less than the way they're doing it now is the thought that they're going to have a whole new way of doing it after this law that will be just as confusing and complex.

So Yadira and her team know that the key thing to deliver here is an interface to those doctors that makes sense to them, that doesn't drive them crazy. But one of the first things they are asked to do is just make a website that explains this program and they say, "Great." First choice they're going to have to make is whether the doctor is a sole practitioner or in a group practice. Well, it turns out there are nine different definitions of a group practice. And I think the old CMS would've said, "Okay, well let's build some software and a long explainer so that everyone can understand all those nine definitions." But what this team does today is say, "We can't work with that. We're going to have to resolve some of those definitions. We really need to have one." They don't get to one, but they get to two.

And I think that's pretty good illustration of how hard it is to work with these people who are... Not to work with the people, but to work with policy that's so complex and people who are saying, "It has to be nine. That's how we're compliant with the law." So this sort of obsession with these fine details when some of them really could be resolved. So they get to two, which is going to make the software a lot easier. Another example of what they do is that the law exempts doctors who only take a few Medicare patients. There's a certain threshold. Those people shouldn't have to try, do new electronic healthcare records software, they shouldn't have to train, they shouldn't have to go through the program at all. But the way the regulators are interpreting the law is that everyone will go through it the first year, and if at the end of that first year you can show you didn't have as many Medicare patients as would hit the threshold, then you would be exempted.

But this is going to drive a whole section of doctors absolutely crazy for no reason. So they fight back and they push back and they say, "No, let's just do it based on the previous year's data." And because they fight a whole bunch of fights like that that are about, as the team said, I get it's complicated, but it has to make sense to a person, then they're able to deliver a project that is not only under budget, on time, but the doctors really like. Instead of flooding the call center with complaints, they call saying, "Something must be wrong. This is too easy."

And it's people like Yadira Sanchez who are doing this, who are sort of standing up and saying, "I need a seat at the table, the policy table, not just writing the code, not having just my team write the code that you tell me to write if we're going to earn back the trust of doctors." And I think that's true for every service. We need to earn back the trust of the American people, whether it's doctors or patients or SNAP recipients or people with a criminal record. And it's people like Yadira who can do that, but they have to change their idea of what their job is and it's empower themselves.

Justin Hendrix:

It seems like part of what Yadira seems to create for herself and that you seem to admire is some wiggle room. Just the ability to maybe do things a little bit differently than the law might have suggested that it should be done. Just take a little bit of liberty with delivering on exactly how the policy may have been laid out in words in order to look more at the goal. What was the goal? What did this law intend to do? Is that right? Is that part of the issue here that we've got almost a system that denies learning in a way because it stipulates to specifically what it wants us something to do?

Jennifer Pahlka:

I think that's exactly right. When you hear about people like Yadira. And she did something else after that I talk about in the book. Later on, she's asked to take this data about pharmaceuticals and extract it and give it out to the ecosystem every quarter. And she says, "There's a better way to do that. Let's create an API, an application programming interface. That will actually honor Congress's intent much better and cheaper than these data dumps that are specified." And when she does something like that, many people have said, "Oh, my gosh. She's not following the letter of the law." But she is very, very well following the intent of the law. She is fulfilling the intent of the law in the way that actually makes the people above her happy and makes the public happy with the services we get. I do think... I don't know if wiggle room is exactly the right word.

I would really focus on are we honoring the intent of the law and policy that's passed or are we simply doing what we're told and then saying, "Well, it didn't work. Sorry, it's not my fault. I followed procedure."? I tell the story of a different leader, extreme contrast to Yadira, somebody at the VA who kept telling me, "I don't have any opinions on the requirements of the software." I have spent my career teaching my team not to do what Yadira says, and I ask him why. And he says, "Because then, it's not our fault." And he says, "If they ask us to build a concrete boat, we'll build a concrete boat." But the problem with that concrete boat is that it won't float. And in that case, the boat that wasn't floating was healthcare for veterans.

At that time, there was something like 18 veterans committing suicide every day, many of them because they couldn't get access to mental healthcare. I don't think it's okay to say, "It's not working, but it's not my fault." And I think people like Yadira are saying, "It is my job to deliver for these people. It is my responsibility. And I'm going to be empowered to do that." I will absolutely have an opinion on business requirements because the people who rely on Medicare need me to have that.

Justin Hendrix:

Late in the book, you write, "I often hear that tech changes so fast, it's difficult to keep up, but what government needs to keep up with is not tech, but people." That seems to be basically the message of this book.

Jennifer Pahlka:

It does come down to people in a number of ways. I think when I say we need to keep up with people, we need to recognize that people's needs have changed. The way people interact with governments, with companies, with systems has been in this crazy whirlwind since the beginning of the internet. People have a lower tolerance for paperwork now. They have shorter attention spans. And our country is in greater need. We have greater inequality, we have bigger problems, we have climate change. The needs of the country and the people have changed and we can't just say this is how it's always been done. But it also comes down to people in the sense that people are the solution. People like Yadira, there's no technology or policy that creates a Yadira Sanchez. We are all, I think, responsible for elevating or recognizing and celebrating people like her so that we get more like her. It's people that need government and the people of government that are going to matter here.

Justin Hendrix:

Late in the book, you also talk about the sort of Silicon Valley mindset that sees building government capacity as perhaps not even possible that this nihilism about government's ability to do anything at all, the thought that really, it should be private companies that remake the services and provision of all the things that make it possible to deliver on the state's interests. You point out Marc Andreessen as the sort of epitome of that attitude. Your book is kind of coming out in that context. This is a kind of ideological fight at the moment.

Jennifer Pahlka:

I do say that I think if Marc Andreessen could see the work of a Yadira Sanchez, it might change his mind. Perhaps I'm being a little generous with him, but I do think people see this very frustrating government. I believe in regulations, but I do recognize that some regulations are implemented in a way that's extremely burdensome for companies and for businesses. And that's a common complaint of people like Marc Andreessen. But you can look at the way a Yadira Sanchez would implement regulations and say, "Oh, that's going to get us a good outcome at a low cost to us." And I think it's, you really have to see the work in order to start to believe again. I think we are starting to see the work.

For instance, covidtest.gov, which came out right in the beginning of the Biden administration, is a great example. It's the exact opposite of healthcare.gov. It's the exact opposite of making doctors go through a year with new medical software before they get exempted. It's the exact opposite of a 212 question form for food assistance. I timed myself getting those COVID tests, it took me 11 seconds to order them. They came two days later. That's delivery in a literal sense, and it happened because people in government, not people outside government said, "We are going to make choices here about what's important." That guy at the VA who said, "It's not my job to have an opinion on business requirements.", the way they develop software, these concrete boats, is that they take often years, sometimes more than 10, to pull together every possible requirement. It's not uncommon for tech projects to have up to 10,000 requirements built into them, and then they try to deliver on all of them.

covidtest.gov said, look, there's a lot of things we could put in this, but let's just value speed, simplicity, scale, and accessibility and all the other... That means we're not going to be able to ask a lot of questions, but we're going to deliver the goods, it's going to be available in different languages, it's going to stay up and handle a huge load, and we're going to get those things off. That is the kind of government that we want, and I think we need to start recognizing those examples and saying, "Yeah, hey, Marc Andreessen, government's not so bad. Let's believe in it and support it."

Justin Hendrix:

At the end of the book, you do conclude, you kind of look up, you think about bigger issues, you talk a little bit about the future and at the risk of asking you to kind of repeat yourself on this battle between those who think of government as incompetent and not where the answer lies, and those who perhaps see more opportunity to do things together in a democratic context, I guess a final word on that, this piece of it, this kind of digital layer where we interact with government seems to be very important now to maintaining our sense of trust that government can be effective. I don't know, where do you think we are in 20 or 30 years on this?

Jennifer Pahlka:

Well, let's talk about this moment for a second. When I wrote the book, we did not have the CHIPS and Science Act, the Infrastructure Act, the IRA, the Inflation Reduction Act. Now, we have three of the biggest pieces of legislation ever, and three that are addressing existential issues that we face in this country. So the idea that implementation isn't as important as policy has never had a better time to be challenged. We have to implement. I'll just take the IRA for example, because it is our shot at averting a climate crisis. We really better be paying attention to that in a way that we haven't paid attention to other implementation.

And I am a little worried. In some ways, implementing IRA looks a little bit like healthcare.gov, in that we are asking agencies, let's say the Department of Energy, that don't normally do consumer facing services, to figure out how to do stuff that everyone's going to have to understand. If you're going to get a rebate for a heat pump or get a tax credit for solarizing your house, that's going to have to happen in a way that you you know about and that you trust that it's going to happen. There can't be a ton of friction in the system. All those skills that Yadira Sanchez is brought to the Medicare program, we now need to bring to implementing this important climate law, and we really need to make sure that those competencies and those skills are there. We miss our one shot at climate.

So where we're going to be in 20 years, I think, depends a lot on the choices that we make today because infrastructure, CHIPS, and IRA are so critical. I think it also matters not just because we could miss our shot at averting a climate collapse or keeping our supply chains resilient, but because we are at kind of the end of our rope with trust and faith in government. There is research that shows that people who engage with means tested benefits that are really, really burdensome and sort of insulting, vote at lower rates. We can't really afford that if we're going to avert electing an authoritarian leader. We have to engage the population again, bring them along. That's really a threat to our democracy when our services don't treat people with respect.

Justin Hendrix:

This book is Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better. Out June 13th, 2023.

Jennifer Pahlka:

That's right.

Justin Hendrix:

Author Jennifer Pahlka, we'll look forward to it and hope that perhaps you'll come back on in a couple of years and let us know if there are some bright points and some movement on some of these issues perhaps in response to your ideas.

Jennifer Pahlka:

There are already bright points and I know that there will be more whether people read the book or not. But I hope they read it. Thanks so much, Justin.

Justin Hendrix:

Thank you.