Program transcript:
Grant Reeher: Welcome to the Campbell Conversations, I'm Grant Reeher. Last week we heard from Nina Jankowicz about disinformation and the information war. My guest today also writes about the intersection of technology and government. Jennifer Pahlka was the deputy chief technology officer in the Obama administration and is the founder of Code for America. She's also the author of, “Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better”. Ms. Pahlka will be on Syracuse University campus to give a public talk on October 29th titled, “Governing in the Age of AI: Lessons from Recoding America”. Ms. Pahlka, welcome to the program.
Jennifer Pahlka: I'm delighted to be here. Thank you.
GR: This is a book about government under-performing, even failing in some ways, and also about governing and the digital age. Obviously, there have been deep complaints about government failures well before the digital age. And so my first question is, is government failing in substantially new ways now?
JP: I think it's failing in ways that the people really feel more, and that has sort of come to ahead in terms of a crisis. You saw leading up to the election last year, over the summer there was a poll that said that 70% of people feel the system needs either major reform or to be blown up altogether. And I think that does speak to just, you know, a real frustration with the public. And you can see it in ways, everything from, you know, I had an experience in my hometown in Oakland, California, where I had a home invasion, and I called the 911 center and they didn't pick up for quite a long time. In fact, they dropped the call three times.
GR: Wow.
JP: And then they said, Oakland Police Department will be there as soon as they possibly can and it took them two whole days to show up.
GR: Oh my goodness.
JP: And so that's kind of thing where you expect, right, that when you, at the very minimum, you expect that when you call 911, someone will answer the phone and if there's someone in your home that the police will show up. And that's not something that people feel that they can count on these days. And all the way to that very personal experience at the local level, to the experience that I think people had during the Obama administration, where at a very, at a national level we set some big priorities like, you know, moving to green energy infrastructure or, you know, building more roads and highways or, you know, reshoring chips to manufacturing, or connecting homes to more broadband. And then we spent a lot of money and made a lot of announcements, people didn't see those things actually happening. So you start to get this sense that government can set a goal but not actually achieve that goal, and that can really erode trust and faith in government.
GR: Well, it seems to me that many liberals will argue that government appears to fail because it's under supported and underfunded. And then conservatives will often argue that government fails because in some fundamental way it cannot do or do well what we want it to do, what we hope it will do. What are the root causes of government's current failures or major shortcomings in your view?
JP: Yeah, I think it's time that we sort of looked past the partisan swipes at why. And, in fact, I think there's increasing bipartisan support for the acknowledgment that fundamentally, we moved into an internet age and we left the operating model of government back in an industrial era. I shouldn't say that we haven't updated it to some degree, but it's kind of like we slapped a website on the front end of it, but the guts of it really still are how things operated as we sort of came out of World War Two and moved into an industrial era. It's a very, you know, in the framework of software development, we call this like waterfall development. It's very stepwise, it doesn't have feedback loops in it that, that, you know, you see in the society around us and the things that work well. So, you know, at its fundamental diagnosis, I would say it's time not only that we addressed this operating model of government that's been left behind, but also realize that while we sort of failed to update it for the internet era, we are now entering an AI era. And so we can't really fight the last day’s battles, which I'm prone to do, right? I fought those battles and so they’re still sort of bothering me, but I need to let go of them and say it's time to ours to figure out how to drag government into the AI era, not, you know, not the era where we, we sort of missed the ball for a while. I could talk a little bit about what that operating model looks like, but I think that's the fundamental problem. It's not just that the policies aren't getting what, you know, they intend, but that those policies sit, they rely on this operating model to be successful.
GR: You're listening to the Campbell Conversations on WRVO Public Media. I'm Grant Reeher, and I'm speaking with governance and internet expert Jennifer Pahlka. She's the author of, “Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better”. So, I want to pick up on just that last set of points you were making. And obviously it sounds to me then that what we need is a real, a core rethink, first of all, and a core restructuring. What would be some of the most important elements of what we ought to be doing then that we're not doing, in your view?
JP: I spent the last 15 years trying to work on the digital side of government. And what I realized is that if you want to move government forward, you have to pick up all of the elements of the operating model. And the way I break those down is the following: If you want government that can achieve its policy goals, you have to have the right people, which implicates civil service reform, they have to be focused on the right work, which implicates procedural reform, right? Today they're sort of drowning in paperwork and procedure that doesn't necessarily accrue to the outcomes people expect. Third, you need purpose fit systems for them to work with, and that is the world that I've been in. But trying to get the right digital systems for government, we have the wrong processes and procedures that result in systems that aren't really fit to purpose. And then fourth, and maybe this is the most foundational, we need to be able to operate in test and learn frameworks. Again, not just this idea that we create a plan and then we operate to that plan, even as we can see that it is not getting the outcome that we intended, or not getting it fast enough, or getting a perverse outcome because, you know, we haven't thought this through. When we continue to follow that plan, even as it's not working, because we're locked into this particular framework. But there's a whole bunch of ways the government needs to relearn how it operates that does learn as it goes and has these very much tighter feedback loops where we're able to adjust along the way to get to where we need to go.
GR: That all makes perfect sense. And the last point, which you said may be the most foundational, the first thing that popped in my head when you said that was how difficult that would be to pull off in the media age that we live in, because things get scrutinized so quickly and so critically that the idea that, I mean, it sounds great, you know, we're going to test and learn and we're going to adjust and we're going to do these things quickly. But if something doesn't work right, the media is all over it as this is a government failure, this didn't work the way it was intended. I'm thinking of the rollout of Obamacare, which you're very familiar with. I'm thinking of some of the things having to do with COVID, both during the Trump administration and the Biden administration. So how do you create the space where people will tolerate, I guess, or the media will tolerate that kind of smarter approach?
JP: It's a really good point, and I think we need to bring both the public and the media along on this shift to a new operating model and I’ll give you a really good example. Speaking of very public debacles, we more recently had the FAFSA, it’s the federal student aid form. And Congress ordered a modernization of that form, essentially with the Better FAFSA Act, I want to say back in 2020. But unfortunately, the way we, quote unquote, “modernized” today, it really doesn't work. And so the, you know, what was supposed to be a better new system really failed students. Many of them couldn't get through it and there were millions affected, especially, unfortunately, those who most need the support of the federal government to go to college. The Department of Education did bring in a fantastic team to get it back on track. Some folks from the College Board volunteered their time, and then they sent some people over from the United States Digital Service. And they brought a new way of working that really is consistent with this operating model that we need government to move to. We call it the product model. There's a lot of nerdy stuff that goes into that that your listeners don't want to hear about. But it is what we were talking about, it's being able to adapt as you go. It's having the right internal technical leadership to do it right, some flexibility and really being centered on what users really need. Well, they got the FAFSA very much back on track. If you look at it now, it has very high satisfaction rates, very low error rates. But the Government Accountability Office still wrote them up, with a very, you know, critical report. And the reason the report is critical is that they are holding the Department of Education accountable to the old way of doing things. They didn't realize that the successes have come from this shift to this new product model way of working that's been incredibly successful. And the good news, the other good news, though, is that what typically happens in government when you get a bad GAO report, is everybody sort of gets defensive and everyone gets told to return to the old ways of working because you don't want to get criticized. But in this case, the Department of Ed wrote a letter back saying, here is what you should hold us accountable to. These ways of working are better, they're getting better outcomes, and please, the next time you, you know, assess our success or failure, assess us according to these kinds of criteria, because that is what is actually helping kids apply for financial aid.
GR: That's a good example, I'm sure it's something that a lot of our listeners have lived, so it will resonate. You're listening to the Campbell Conversations on WRVO Public Media. I'm Grant Reeher and I'm talking with Jennifer Pahlka, the governance and internet expert has served in the Obama administration and is also the author of, “Recoding America: Why Government Is Failing in the Digital Age and How We Can Do Better”, and we've been discussing her book and some of her more recent findings and research as well. So I'm going to get a little wonky myself. You were saying you didn't want to get too wonky before, but as I was reading through your book, I got the clear impression that one of your arguments that we haven't really talked about as much, although indirectly, is that the details of policy delivery and implementation have been undervalued and understudied, but that they're terribly important. Do I have that right as take away from your writing?
JP: Yes, I think that's correct. I think that, you know, I came to government from the tech industry where the people who actually, you know, do the work, they create the website, they, you know, deal with the customers are kind of at the top of a social hierarchy, right, if you think about it. Those companies are all run by founders who, you know, did the first version of their product. And in government it's the exact opposite. The people who actually make the interfaces to the public are very, very low on a very, very tall totem pole. And I think that creates real problems because we give respect and status to those who think the big policy thoughts, we don't connect them to the people who are actually going to make that stuff real. And they're so far apart in the sort of social hierarchy, then that's part of how we get really bad interfaces to government.
GR: Yeah, exactly. And the other thing that I thought of was my own field of political science, and that it seems like there's a similarity there too, where it's exactly as you just said, that, you know, there's a value priority put on sort of, for lack of a better word, higher altitude sort of theoretical discussions, less emphasis on the actual nuts and bolts of delivering things. And there's kind of a dearth of interest and attention in the academic field as well. Is that your sense?
JP: Absolutely. There's so much that I would like the academic world to study, that really gets very little attention because it's not about the, you know, the nuts and bolts of government, but in fact, that is where we're failing people. It's not just that we want there to be a closer connection between policymakers and the folks that are doing, you know, the actual work in government. But as I said earlier, these test and learn frameworks, it means that we need a feedback loop. People who make policy and law today really need to understand what implementers do in order to get better at what they're doing. When I spend time on the Hill with congressional staffers, they will secretly or quietly admit that they know that the law and policy that they write is not really implementable. They know that it doesn't get the outcomes that they intend or very often doesn't, and that what they need to learn is sort of more of the full cycle, so that as the people in the agencies that are implementing these laws are learning what, you know, what users want or, you know, what actually happens when you go to implement these things, that they can bring those learnings back up the chain, and we just don't have those feedback loops in place right now. But it's not just about a better communication, it is really about the learning, being, you know, multi-directional, and our policy and lawmakers having the humility to learn from the actual execution.
GR: What you just described for that feedback loop sounds like the right thing to do. But we are currently, you and I are speaking in a moment where the federal government and civil servants, I think it's fair to say, are under an assault where at least they feel like they're under an assault. And how do we create the political space or the trust space for there to be that kind of sharing of information and mutual respect from the lawmaker and the policy designer to the implementer and back again in order that we get this more right?
JP: Well, our founders created a system of federalism, and I think that's a big advantage right now, because when things are not possible in DC, they are possible in the States, and we can experiment with them. I think actually there is more room for maneuvering and learning and talking about new things right now than we're going to, you know, admit publicly because their headlines are quite dramatic, but there is still stuff going in Congress where, where people want to figure out what is next. But leaving that aside, you've got, you know, 50 states or 53 if you count the territories and something like 30,000 local jurisdictions that can experiment with these new ways of working and show what it looks like, and then we can adopt those back into the federal level when, knock on wood, we have a more functional Congress in the future.
GR: Well, one of the things I've noticed as a political observer is that the partisan divide and some of the dysfunctionality of hyper partisanship has now kind of infiltrated to the state and even the local levels. I'm wondering if that is going to compromise that kind of dynamic that you're hoping for from the, you know, 50 laboratories of democracy. I forget which Supreme Court justice used that phrase, it might have been Oliver Wendell Holmes. But anyway, you know, I'm wondering whether the window for that is closing.
JP: Well, I think that it's a tough political environment, for sure. And to the degree that culture war stuff can really distract from the very serious business of governing. Yes, we do have a problem, but, you know, they also, to quote another sort of, tried and true government line, there's no Democratic or Republican way to pick up the garbage, right?
GR: Right.
JP: I mean, these things still matter in communities even when there is political strife. And that is a lot of what we're talking about here. People need to live together, they need to, you know, have roads and schools and garbage pickup. All these things actually really do rest on that operating model of government and are the places where we don't disagree that much. And I think in some ways, by raising the visibility and salience and, and relevance of this core operating model, we actually give people something to talk about where there's just a lot more common ground and a lot more sense of, this is just common sense. This is just, you know, this is part of how we're all going to live together in this community and how things work well. So I think in some ways it's really an antidote to the divisive headlines.
GR: Yeah, that's a good point. If you've just joined us, you're listening to the Campbell Conversations on WRVO Public Media. I'm Grant Reeher and my guest is the governance and internet expert Jennifer Pahlka. I wanted to ask you something that is also in your book that I think would be of particular resonance with people who are living in this part of upstate New York, especially the Rochester area. You write of something that you call the ‘Kodak curse’. What is the Kodak curse, first of all?
JP: The Kodak curse derives from, I guess it was back in the 80’s, when
these companies, starting with Kodak, decided to outsource all of their I.T. And, we've done that in government, right? We've said, look, this isn't something government does, well, let's just buy this from vendors. Well, in a sense, there are many things that can be outsourced well. But when you get to the point where all of the digital competence and understanding of the digital world is outside your company or outside the institution of government, you really can make some pretty bad strategic blunders. And Kodak did this when they, you know, they had a lead on everyone else, they were the first people to have a digital camera. But they lost that lead. And, you know, it really hurt the company because it, you know, sure, you want someone to fix your printer or get you on the internet, I can see why people want to outsource that. But if nobody understands the value of digital expertise in your company, you're going to miss a lot of strategic decisions that would really, you know, put your company forward. Our government has very much treated everything digital as just something you buy. In reality, digital is something you do, right? None of the companies that that run on technology today outsource it entirely. You don't outsource your product. If you think about something like unemployment insurance, yes, people in government think of the politics of unemployment insurance and the policies as the product. And in fact, what people experience is the website is the product And you really have to have control of that and know that it works and know that it's good and that it's going to meet people's expectations. You know, at the end of the day, you're going to be a legitimate government that people trust and want to have work.
GR: This applies so much to the university as well. I'm thinking of the experiences we've had with our website, and I'm also thinking of the experiences with online teaching and buying the packages from elsewhere where they don't actually fit. I wanted to give you some space here at the end, to end on a positive note and to ask, are there, you mentioned before the different laboratories of democracy and this is one of the benefits that we can see. Some of these things that you're proposing might start to be done at the state level or the local level. Are there particular states in the United States, or are there other countries that have caught your eye as leading the way on this issue that the United States national government, federal government could be looking at, to get some important lessons from?
JP: You know, there are bright spots everywhere. There's no place that is such, you know, a perfect model for the United States because every place is different. So at the state level, you know, I'm inspired by Governor Shapiro's leadership around the I-95 rebuild, for instance, where he just said, we have to do this quickly and pulled out all the stops. And I think people really saw a way government can really deliver when it has to. And I think all leaders should take a page from his book on that front. Overseas, I've always been, gotten a lot of inspiration from my colleagues in the UK who started the Government Digital Service before the United States Digital Service came into being. In fact, we modeled the USDS on the GDS and they continue to do great work. And I think they're much more of a peer nation, whereas a country like Estonia, which is very often cited as the leader and truly is incredible in its sort of digital forward in its ability to serve its citizens with incredible ease and convenience and confidence, right? They just don't have the policy legacy, the sort of accumulation of decades of policy debt that we have, which makes it a lot easier for them. But I still think we should aspire to their kind of work. They, for instance, are going to be, I think quite forward leaning on the adoption of AI. And it will just be harder for us, for a number of reasons that we really need to grapple with. Because if you think about this AI world, I mean, think about how fast vectors for fraud and abuse in programs are going to evolve in a world where lots of people have access to AI. If government doesn't avail itself of those tools, we're at an enormous disadvantage. Think about how fast we need regulation to respond. I don't just mean regulation of AI, I mean any kind of regulation. When we move into a world where people can do things very quickly with AI, government will need to respond just as quickly. And we're going to have to start clearing out some of the clutter that keeps us from being able to move into that world. All the time recognizing that AI has, you know, a bunch of challenges that we need to be very careful with it, that we should not be adopting any technology without lots of systems in place for, you know, constantly testing it to make sure it's doing what we intend it to do and not causing real harm to people. But we have to speed that process up enormously. And I think we can look to places like Estonia for inspiration while recognizing that they are, you know, they don't have some of the challenges that we have in terms of really 250 years of always adding to our laws, policies and regulations and never subtracting. This is a real problem that we need to deal with. The good news is that AI also can help us deal with that problem of policy accumulation and cruft. LOM’s are fantastic at this, and I'm really excited to see a lot of people use those tools to do this sort of regulatory and policy simplification that is really long overdue.
GR: Well, I'm happy to leave it there because, you know, this has been such a challenging environment in recent years and you're finding rays of light. And so, you've left me more optimistic than when I started the conversation, so I appreciate that.
JP: I’m glad for that. I’ll just say since you mentioned the federal workforce and the disruptions. I mean, if we may leave on a note that is sobering but also positive, you know, I spent a long time sort of, frustrated with governments sclerosis and it's and it's slow pace. Things are now moving rather quickly, and you know the phrase, you can't make an omelet if you don't break some eggs? I think the present side of this particular moment is that the eggs have been broken, so it's really up to us to now just go make that new omelet.
GR: What we make with it, that's a good analogy, yeah. Jennifer Pahlka is giving a public talk on the Syracuse University campus on October 29th. It's titled, “Governing in the Age of AI: Lessons from Recoding America”, and more information on that talk can be found at the Maxwell School of Citizenship and Public Affairs website. Ms. Pahlka, thanks so much for taking the time to talk to me. I learned a lot.
JP: Thank you, it was a joy.
GR: You've been listening to the Campbell Conversations on WRVO Public Media, conversations in the public interest.
