Government doesn’t struggle because of a lack of good intentions. It struggles because the system beneath those intentions was designed for a different era.
In this Mission O/S Livestream, Jennifer Pahlka (author of Recoding America and former U.S. Deputy CTO) joins Bryon Kroger to examine the structural gap between policy intent and shipped outcomes. They unpack why industrial-era budgeting and oversight models continue to undermine digital delivery, why “big-bang” project funding leads to endless modernization cycles, and what it means to shift from projects to products inside government.
The conversation covers:
- Why mandates alone don’t create change
- The limits of project-based funding
- Capability-based budgeting as a path forward
- The importance of path-to-production infrastructure
- How test-and-learn operating models unlock real outcomes
If government is serious about delivering in the AI era, the operating model has to change. This discussion explores what that change requires — and where to start.
Watch the full livestream, and then check out Bryon's free course on shipping outcomes in government.
Transcript
Bryon Kroger (00:00):
Hello everyone and welcome to the Mission OS Livestream. I'm your host, Bryon Kroger. And for those that don't know me, I spent 10 years active duty Air Force, the first seven as a targeting officer using really terrible software, which we'll talk about, to conduct really critical missions. And then the next three, fixing that software over in Air Force acquisitions where I co-founded a very successful transformation project called KesselRun. I got out, started RISE eight to help others do the same, and now we've launched Mission OS to share all of our lessons alongside this live stream, where we will extend the conversation with gov tech leaders and explore what it really takes to ship across people, process, and technology inside of government. Now, today is our very first live stream, and I'm incredibly excited to welcome Jennifer Palka. Jennifer Palka is a co-founder and board chair of the Recoding America Fund, an executive in residence at the John Hopkins University School of Government and Policy, a senior fellow at the Nixon Inn Center, and a senior advisor to the Abundance Network.
(01:05)
You also may know she served as Deputy Chief Technology Officer of the United States under President Obama and a member of the Defense Innovation Board under both Obama and Trump. She founded Code for America, one of my favorite conferences in gov tech, which she led for 10 years. Palka is a winner of the Skull Award for social entrepreneurship, was also selected by Wired Magazine as one of the people who have most shaped technology and society in the past 25 years. Ezra Klein called her book, Recoding America, "One of the best policy books I have ever read and the book I wish all policymakers would read, and I couldn't agree more." Jennifer, welcome.
Jennifer Pahlka (01:47):
I'm so glad to be here, Bryon. One part of my bio that you didn't add is that I'm just a big fan of yours and all the work that you're doing to shape government's ability to deliver the goods.
Bryon Kroger (01:59):
For those who maybe haven't read the book or maybe you've evolved your thinking since, how would you summarize why government struggles so much to deliver in the digital aid?
Jennifer Pahlka (02:11):
I feel like my final answer on that just took a really long time for me to get to. Like you kind of look at the problems of software and then you find yourself sort of going upstream over and over again to, okay, but we can't solve this problem at the level it was created and certainly not with the thinking that created it. And ultimately, I've just come to the conclusion that it's as simple as this. We're still operating in an industrial era model that was the right thing as we were coming out of World War II, sort of post New Deal, all of this stuff. And we got very good at that way of working. And that's not how the world works today. Things move much more quickly and the very heavyweight planning processes that characterize that industrial era's expertise really need to give way to much more of a sort of test and learn kind of way of not just doing software, but really of doing everything.
(03:29)
And we're just very, very rule bound at a time when we need to be far more agile and adaptive, especially as the world, the changes that we see in the world around us are speeding up even at this point every day. So it's really just a huge paradigm shift. And we missed the chance really. I mean, we did not really bring government into the internet era. We did some good work, but the net result was essentially slapping websites on the front of industrial era thinking in the backend. And so now that we're in this AI era, we really have to just think about leapfrogging. How do we leapfrog the government into the model that it needs to adopt in order to meet the moment?
Bryon Kroger (04:22):
That's a really great summary. I'm curious, I know you have a model of a way that you think about solving these problems. Maybe let's start there. Introduce how you think about addressing these problems.
Jennifer Pahlka (04:37):
So as you know, I sort of came from the digital space and now I'm sort of looking at it more broadly. And the way that I pitch it to folks who don't care about software or don't care about the operations of government is that what people do care about is the outcome. In your world, it's we need to deter our adversaries and have a safe nation. And of course we care about that, but we keep pushing on the better policies and better weapons systems or whatever that we think is going to get us that. And forgetting that nobody is working on the operating model that sits below that. So just like you have sort of a hierarchy of needs, Maslow's hierarchy of needs says you can't have self-fulfillment and self-actualization if you're not fed and clothed and housed. The operating model is that basics, the feeding, clothing and housing of government.
(05:37)
We pulled into four parts. If you want government that can actually achieve its policy goals, you have to have the right people. So we need to do civil service reform. They need to be focused on the right things, the meaningful outcomes, not pushing paper that no one actually cares about. So we have to do significant procedural reform, and that takes many different forms. Third, they need purpose-fit systems. This is part of your job here. So we have to reform not just agile approaches, but really the environment in which we build and by technology. The way technology, you'll use the word projects here with a giant asterisks, are funded the way they're conceived of the way they are overseen. They're all in an outdated model. And then lastly though, this really undercuts all of them. We have to be able to operate in test and learn frameworks.
(06:36)
And that'll also take a lot of different modes. But really one way is that while we are trying to get test and learn within agencies, we also actually need test and learn that bridges between the executive and legislative branch, such that when our lawmakers set a goal and say, "Go do this, " they are actually looping back with those who are executing on that intention and saying, "What is working? What is not working? How do we fix what's not working? What can we edit? What can we change? What can we remove so that we can actually get to that goal?" And that is, again, a huge, huge change in just thinking and acting that we have to educate not just executive branch leadership, but our leaders on the Hill and in states and state houses as well. And I think we're not going to have to just talk about, we're going to have to show what that means.
Bryon Kroger (07:38):
Yeah, let's go there. I'm curious your thoughts. I mean, reform is much needed, but the lead time on it is quite long, as you know. What do you recommend to people right now who are in the thick of it? Where can they start pushing the needle within their sphere of influence on those four areas?
Jennifer Pahlka (07:59):
Yeah, that's a great question. So I talk about the approach we're taking at Recruiting America as a sort of a full stack approach, but I'm not talking about a tech stack. I'm talking about the stack from statute to regulation to guidance and policy and guidance and memos, all the way down to just the actual behavior, the practice on the ground. And we do have a C4, we are going to be looking for structural reform and asking for that of our leaders, but we firmly believe that just pushing on that layer will get us a lot more of the same, which is well-intentioned laws that don't actually cascade down properly, but also aren't informed by those lower layers. So everybody who sees the need to move government into the AI era and to be able to deliver for people has a role. If they have any role in what's actually going on in a department trying to deliver on a mission, they are seeing the ways that what we are doing on the ground isn't actually consistent with what we think we want out of this.
(09:16)
And they can call those things out. The whole statutory and regulatory agenda needs to be deeply, deeply informed by the lived experience of people on the ground. A good example of this is like we've had, for instance, the Chance to Compete Act, which is supposed to really change how agencies hire. They're supposed to use assessments in their hiring so that we get the right people in the seats, but we're not actually seeing that happening in the agencies. There's all these barriers to it. And I think if I could get members of Congress to really understand one thing, it's that just because you have ordered something to happen does not mean it will happen.
(09:57)
And you really have to get curious about why it's not happening instead of say, okay, our response to this perceived non-responsiveness is to create another mandate and just order them louder or create more consequences because that's actually making the problem worse instead of freeing the department up to do what they know they want to do. And I think part of that is you have to believe in public servants. I mean, you have to believe that they actually want that right outcome for the most part. We can't talk about people in such a broad way, but for the most part they do and they need to be unlocked from all of this procedural cruft so that they can actually pursue the outcome. And anybody who feels constrained by the decades and decades of mandates and constraints that have been dropped on the department that they're working in can decide to have a voice in saying, "This is what you can remove to get us to a position where we can do what you want.
Bryon Kroger (11:02):
I love that. You mentioned Congress, policymakers, or even the reformers themselves should get curious about why people didn't follow the mandate versus mandating louder. One thing I've seen too though, I mean, it would be great if they got curious and went and asked, but I also find that a lot of the people out there who are doing things and have found successful patterns, but can tell you all of the pain points that will stop the next person from doing it are often the last people to get involved in reform. Is this something you've struggled with trying to get these doers to participate in these changes and inform them better? And if you've been successful, how do you get them to do that?
Jennifer Pahlka (11:42):
They're busy. The people who have been the most successful are the ones who are given often more work, and it's very hard to free them up for it. I think for anyone, really, to me, the barrier is that they have to believe that change is possible. And so on both sides, we have to show change, which will create momentum. And as eventually, I hope drive a real flywheel. There's just huge cynicism
(12:15)
And lack of trust, I mean, particularly between the branches and changing that is not going to be easy. So if you think about when you go talk to somebody on the hill and say, "Can you please go talk to people in the agency and find this thing out, get curious?" They will say to you, and they're completely right, that when they go to talk to agencies, what they get is really stonewalled. The agencies are so afraid of that communication that they sort of lock things down and they sort of interpret any inquiry as a, "We're going to audit you, " essentially. And so innovation also stops. And until people believe that that inquiry is a genuine inquiry in order to consider another way of approaching this problem, hopefully removing a barrier that agency has, then that's going to continue. But both sides have to start to have a little bit of faith.
Bryon Kroger (13:17):
Yeah, that's definitely true. And I think there are challenges. I have one story that I don't recommend people do this, but when we were struggling, the Kesselrun story sounds like it was just made to be and everybody embraced it with open arms, but it was the worst fight I've ever had in my life. Still to this day, scaling a business, everything, that was the worst I've ever had it in terms of I felt like the whole world was out to get me and I was going to war every day. And we had to start advocating for ourself on the Hill. And so I was meeting privately with professional staff members, which is a big no-no. And at one point I got found out and leadership blocked, had a milisus show up and blocked me from speaking to a professional staff member. And I relay that story of just what I learned in that moment, I ended up after that, I was like, "We have to keep doing this still.
(14:08)
We have to figure out how to advocate." But I found avenues for advocacy. I think a lot of people don't know those exist. The one for me that was super helpful was the Defense Innovation Board, which you know well, but are there other avenues where people can contribute without ... Because again, what I was doing actually, I would say is not the right method either and can lead to all kinds of problems. And so what are good outlets that people can use?
Jennifer Pahlka (14:30):
Well, first of all, let me say that there was a saying at USDS that if there wasn't an assassination tempt on that office this week, it was only because we didn't know about it. We felt very much the same way. And actually people after I was there felt very much the same way, that it continued really for years. And yes, it's funny because people look at it and they go, "Oh, you stood up an office. Good for you. " It was not that easy. Yeah. So I mean, I think that there's been a real gap in the ecosystem that we're trying to fill. So I have certainly done this thing where I take someone from an agency or department and I invite them to drinks and I just happen to invite someone from the hill to drinks too. And apparently there's nothing illegal about that, right?
(15:25)
If you get invited to an event and there's someone there, so that is also probably not scalable and not the right way to do it, but sometimes people just have to talk.
(15:38)
I think in some sense, what we are trying to build through Recoding America is that advocacy, because if you look at, it exists elsewhere. When The Hill is trying to decide what to do, sort of legacy industry shows up and they give them a lot of advice. Who is there representing this new way? Who is there representing the public interest, the transformation that needs to take place? If your business model depends on the way it works now, you are not going to be that advocate. And the Civic Tech Movement broadly has not had any advocacy. It's been very, very careful not to be in an advocacy mode, and we need to change that. And that's part of what we're hoping to do. There needs to be people who know how this stuff works and who can speak for really the public interest on the Hill and create those bridges and say, "It's actually our job to go tell the story that you were blocked from telling." And actually, I think that staffers in particular and some members are so hungry for this.
(16:59)
They really want to know the truth. They really know they're not hearing the truth through official channels, and it's very frustrating to them. And they know that then they're limited to things that are probably just going to make things worse, even though they look like the right thing to do.
Bryon Kroger (17:14):
Yeah. Yeah. It's been eye-opening being on the other side of this, seeing how much industry is meeting with all of these same people that I needed to meet with to advocate and have heard stories since about industry lobbyists who didn't like the way we were doing things at Kesselron didn't want to see it take off, who were lobbying against it. But my Pentagon, my leadership chain, I don't think was advocating for it. Nobody was. Even the people that wanted to see it succeed, you don't realize how much it's being advocated against and how much that impacts our ability to execute. So I just think it's really interesting. That's some good advice. Maybe I want to pivot just a little bit to-
Jennifer Pahlka (17:56):
Let me say one quick thing, which is that we are a C3 and a C4, and it is through the C4 that we do this work, just to be clear.
Bryon Kroger (18:04):
Good, good, good. So when it comes to some of the recommendations that you are giving to the government, you also wrote a paper recently or a blog post turned into a paper about product operating model.
Jennifer Pahlka (18:18):
Yeah.
Bryon Kroger (18:20):
Maybe go over the principles, and then I'm curious where you think we need to start. I always like the order, it feels difficult to get the government to do it all at once, but also there's also this feeling if you don't get them to do it all, it doesn't get anywhere,
Bryon Kroger (18:37):
Order matters. So do you see sequence mattering or do you think we just need to go all in and rip the bandaid off?
Jennifer Pahlka (18:44):
Yeah, no, we do need to sequence it. I think let me answer that question then I'll back into what it is. It's just like diagnosing any other problem. It's like, where's the bottleneck? Where's the biggest bottleneck? Because removing a smaller bottleneck isn't going to help if you haven't hit the key one. You get very incremental changes and you're still constrained by that big bottleneck. So we think that there's been a lot more creation of the teams that can do work in a product model in government. So they exist now,
(19:17)
And that used to be the constraint and no longer is and in part due to your advocacy. But when projects are still funded in a sort of big bang model and then all the oversight assumes a big bang model, that to me seems like the big constraint. So one of the places we're going with this, and this is some more excellent papers published by Niskanen Center by a fantastic woman named Solitaire Carroll, who was at the VA, articulate the follow-on piece to product model, which is we call it capability-based budgeting. So how do you actually give an agency money in such a way that will flow into a product model delivery instead of trying to have to translate it from, okay, I'm getting this big chunk that assumes a sort of big bang project and let me sort of translate it into the way that we would need to spend it in order to build the capabilities to do this thing.
(20:21)
So that was out of order in terms of explaining it, but I think your audience is probably somewhat familiar, but the product model really says you're going to do this over a period of time. You're going to have the right in- house talent and you're going to align the correct metrics, et cetera. And I'll get the story because we're talking about The Hill about how that came about. So I'd been asked, this was a couple of years ago now, I guess, almost two years ago, to go talk to a caucus on the Hill, a Senate caucus, and they wanted to give money to an agency and their ask to me was, "We've given them money before, they've spent it and now they want more money. How do we give them one more chunk of money and then never give them any more money again for this project?" And I was like, "And you want me to answer that question?" And they're like, "Yeah, help us because we're sick of funding them for this stuff that they should have finished." And I sat in my ... An hour before this meeting, I sat and made this set of slides that are kind of a fictional representation of what they were asking for, which is really like, I want to give an agency a big chunk of money.
(21:41)
And so you would have, if you're going over time, the sort of big spike, and then the idea is that they want it to go down into O&M. And then it's like, great, all you need now is a little bit of O&M and that's not too expensive and we're worried about that. But instead of the way you do a product in the private sector is you start with a very small team doing a discovery sprint, they figure out what actually needs to be done. You add people on, you add people on. I mean, some of the biggest products that we use in the world today started with five or seven people teams, and they actually launched incredibly ... Google Maps. I didn't put the story in the book, but I'd researched for the book. It was seven people when it launched and it was quickly serving millions of people.
(22:29)
And then you add over time because then you were adding to something that you know works and it adds getting real world feedback. And so it's much more of this steady state upwards, but to something that actually works. And in the project model where you're giving them this big chunk of money, you think you're going into O&M, but what you're really going into is, it's unclear how long before you have to do a modernization or it didn't work or you're starting over. And so the thing that you think is like one big chunk of money is actually kind of an endless series of chunks of money because you have funded them in the wrong model. And then the reason I think it really worked for these Senate staffers was that if you take the longer view and see that you have to give them a big chunk here and then another big chunk here and then another big chunk here, you've actually spent far more money in the project model over the course of 10, 20 years than you would if you had just funded a product team that had product vision and could do this in an incremental way and they had the right ... And this is not about insourcing everything, but it is about holding product vision in such a way that you can test and learn your way to a product that will actually get the job done, that will actually serve the need instead of, here are the 5,000 requirements.
(23:51)
We have checked all the requirements, but it doesn't actually meet the need. So I think that there's a real willingness to accept this reality. And you're absolutely correct, it's like, what do you do with that? What is the next step? Well, you advocate for capability-based budgeting, you advocate with oversight bodies to understand that a project that's functioning in the product model should be held accountable to different metrics.
Bryon Kroger (24:21):
Okay. I love that. We talk a lot about not only do you have a better chance at success, but when things are problematic, when they do fail and they will, especially if you're doing more R&D forward-looking things, they're also easier to cancel and reallocate resources. So
Jennifer Pahlka (24:40):
One of the questions- We didn't show that line, but you can absolutely see the discovery sprint figures out there's a real problem and we don't actually go forward with the project. And that should be seen as a huge win instead of a failure because we didn't spend $200 million. We figured out what the problem was with half a million or a million.
Bryon Kroger (25:02):
Yeah. One of the challenges there that I would just plug that ... I know you posted about colorless money the other day, and I think that's a big one, gets you out of the O&M R&D dilemma, but then reprogramming, even though I just said you can move the resources somewhere else, a lot of times you can't because of reprogramming thresholds. And we just have to rethink reprogramming, at least within portfolios. Everybody's talking about portfolios now, but we still don't have portfolio-based budgets where you can freely move money around. And that one is ... Normally, I don't blame Congress. I say, Congress has given us the tools we need, but in this case, we don't have a lot of flexibility, at least in department of war.
Jennifer Pahlka (25:42):
I would love it if you guys would feed us great examples of that, that are stories that we can help get told.
Bryon Kroger (25:51):
Absolutely. Scott asked, and I think you answered this probably at the policy level, what the most important thing or impactful thing we can do to start driving more outcomes and changing behaviors in government. And it sounds like at least in moving to this product model, the most important thing we need to get right now in sequence is fixing how we budget. Is that accurate? Is that your top one?
Jennifer Pahlka (26:15):
Yeah, I think so. I mean, I think that's the thing we're going to work on right now with a sort of fast follow on the oversight side. I think that the single biggest thing though is to believe that change is possible. And then if you believe that, then when you see that something could be better, you will take the time to advocate for it, find people who are willing to listen, reach out to us. Don't be afraid to tell the story. It's funny you say that. Be afraid to tell stories. We all get in trouble for them, but at the end of the day, we don't
Bryon Kroger (26:52):
Tell them. I got in a lot of trouble. I will say, it's funny, Mission OS, I mean, we're giving just tons of playbooks and tactics. Somebody asked me, "What's the most important ingredient?" And I said it was courage. So I think we were right on the same page there. In fact, you've used the term cynicism a lot. And I always quote Mattis. Somebody might've said this beforehand, but I heard him say, "Cynicism is cowardice and it's always really stuck with me.
Bryon Kroger (27:18):
And I think nothing could be more important. But if you are a delivery team, or let's just say you're a program owner, you have a budget, you're trying to execute right now, is there one thing you would recommend to help them get to outcomes within the current constraints?
Jennifer Pahlka (27:35):
Well, I would love to hear your answer to that because I think you've given a lot of thought to this and you're providing a real public good by talking about your approach.
Bryon Kroger (27:48):
Yeah. I think the thing that gets overlooked the most is it seems silly, but it is just the path to production. There was this great paper wrote a long time ago. I came across it in the book, Lean Enterprise, which is from your husband's publishing house. Yes, it is. And it's called Avoiding the Alignment Trap in IT. And they did a huge study and found that the predictor was always whether or not teams first established a well-oiled IT.That's how they framed it. But if you look into it, it's essentially your ability to get code from developer laptop to a production environment was way more important than alignment than having the right outcomes or goals. It was like, can you ship code first and foremost because otherwise you get stuck in an alignment trap where everybody's arguing in boardrooms about what the outcome should be and what they think will produce those outcomes and you can never test them.
(28:53)
And so I tell people, if you can't solve that problem, you're just screwed. And so many people start their programs, they want to build the apps. And I get it, that's where I want to live too. I don't want to live in the platform space. I don't want to do ATOs anymore. I wish I could just build apps, but the reality is the path to prod is everything. And so that's where I would tell everyone, you've got to focus there first.
Jennifer Pahlka (29:17):
At Code for America, our sort of corollary to that would be we would be working with state agencies on their SNAP applications and they'd often have multiple stakeholders, especially in California where SNAP is administered by the 52 different counties and they have these two consortia and the consortia is trying to manage all the different stakeholders in all the different counties. And they would be, well, what would work, what would work? And we would always just say, there's an answer to that question, what is going to actually get us the outcomes we want? And it is knowable through user testing. We can fight about what we all think, but the person who knows This is the user. And I think that's what you're getting at. The quicker you get software that people can actually use, the quicker you find out what works. And we somehow decide to push that all the way to the end of a very expensive endeavor when it needs to be much, much sooner.
(30:15)
As soon as you can possibly do it.
Bryon Kroger (30:18):
Absolutely. I think the thing I'll add there is this magic thing happens once you do that too, is now that users have touched it, it's really hard for the bureaucracy to take it away.
Jennifer Pahlka (30:28):
That's right.
Bryon Kroger (30:29):
They'll try, but it is very hard at that point. And so it's great. Nothing can stop an idea whose time has come.
Jennifer Pahlka (30:36):
Exactly.
Bryon Kroger (30:37):
Well, that feels like a great place to end. Before we go, is one, there anything you want to leave with the audience? And two, where can people or where should people follow along and learn more about what you're doing?
Jennifer Pahlka (30:50):
Yeah, I just think it's a really interesting time. It's a time of disruption that's deeply uncomfortable and a time of political division. And yet this is an area where there's just enormous agreement across many, many communities. And so even though you and I have both been doing this for longer than we want to admit, this is a particular moment where I think we should all really be coming together around this agenda. So that's one of the reasons I'm excited about your work and glad to sort of be here with you and talk to the folks that follow RISE eight. I haven't been putting anything on my Substack very much lately. I put something out today that was written by my co-author, but please do follow us on eatingpolicy.com where I at least last year was writing quite a bit. And then yeah, just thanks so much for having me and thanks for all the great work that you do.
Bryon Kroger (31:59):
Thank you. Appreciate it, Jen. Have a good one, everyone.
