Transcript: The Book I Wish Every Policymaker Would Read

The Book I Wish Every Policymaker Would Read

In "Recoding America," Jennifer Pahlka offers a deeper theory for why the
U.S. government struggles to provide services to its people.

   Tuesday, June 6th, 2023

          [MUSIC PLAYING]

   ezra klein
          In 2009, Jen Pahlka founded Code for America. And the idea
          behind Code for America was to get the best technologists
          working on the public's problems, not just private websites and
          e-commerce platforms and social networks. Even then it was clear
          that public digital infrastructure was going to be really
          important and it was lagging far behind the private sector. And
          Code for America was built to bridge that gap.

          Then, in 2013, Pahlka became Deputy Chief Technology Officer in
          the Obama administration. She was very deeply involved in the
          effort to rescue healthcare.gov and helping create the United
          States Digital Service. Then she went back to Code for America,
          and then, in 2020, was tapped by California Governor Gavin
          Newsom to help untangle the mess of California's unemployment
          insurance program as it buckled under the weight of the Covid
          response.

          I mention all that because Pahlka has worked directly on -- and
          then helped oversee and advise teams working on -- the digital
          delivery of government services really at every level of
          government across much of the U.S. So she's someone who badly
          wants government to do good in the lives of people and has spent
          years now confronting the reasons it so often falls so short of
          its goals.

          One of her big points, one I've come to appreciate a lot more in
          recent years, is that even liberals who care a lot about
          government don't care enough or track closely enough how
          implementation actually happens. Delivery often happens out of
          sight, except for the people who need that delivery.

          In our media -- I mean, we're a part of this. I'm part of this
          problem, too. There's a ton of focus on politics, on elections,
          on big policy questions and fights and theories. But then the
          bill passes and the nitty-gritty of how that policy actually
          shows up in people's lives is left up to someone somewhere. And
          when it doesn't show up in people's lives or even makes people's
          lives worse because of how it is implemented, there's often no
          outcry because there's no attention, and so there are no fixes.

          Now Pahlka has written this book, "Recoding America." And I want
          to say it clearly. This is one of the best policy books I've
          read. It's a book I hope future governments and current
          governments will absorb. We get at a lot of it in this
          conversation, but not enough.

          But one thing she really does is tell clearly the stories and
          tell clearly the lessons she has seen and learned about why
          things go wrong even when the people involved are trying to make
          them go right. So people who care about these issues really
          should pick it up and read it in full. But this conversation is
          a really rich place, I think, to start.

          As always, my email [email protected].

          [MUSIC PLAYING]

          Jen Pahlka, welcome to the show.

   jennifer pahlka
          Thanks for having me.

   ezra klein
          So I want to begin with a quote that feels in a way like the
          organizing thesis of the book, which is, "When systems or
          organizations don't work the way you think they should, it is
          generally not because the people in them are stupid or evil. It
          is because they are operating according to structures and
          incentives that aren't obvious from the outside."

          And I love that, because, for me, it's a bit of a credo in
          reporting on government. But you've been inside government. So
          tell me a bit about how you came to that view.

   jennifer pahlka
          I've seen so many people frustrated with government services,
          and they make a lot of assumptions about how that government
          service got to be the way it is. One example I spent a lot of
          time on was SNAP in California -- Supplemental Nutrition
          Assistance Program -- where the application form was over 212
          questions.

          It took a long time to fill out. And so people naturally assume
          that this is done on purpose, that the people making the form
          must not want people to get SNAP. And, for example, there,
          that's not really what was going on. California is a very
          pro-welfare state. There's a huge number of incentives for local
          communities to sign people up for SNAP. It's great for their
          economy. It's great for getting people out of emergency rooms,
          for instance.

          But there are just so many stakeholders in this that they all
          got to pile on their questions, and you ended up with something
          that was really, really burdensome. And it's hard to see those
          incentives from the outside. That's just one example of them.
          But our public servants are incented to do things that, if you
          look at them rationally, don't make sense.

   ezra klein
          You have a great line in that section, which I'll probably get
          wrong in the paraphrase. But you say something like, people
          often think it was designed this way, when the truth is it
          wasn't designed at all.

   jennifer pahlka
          Yeah. I think that's true of so many things that we think are
          systems. I also think it's true of the policy. Take unemployment
          insurance, for example. We've had layers and layers of changes
          on it over time.

          Those changes come from federal government. They come from state
          government. They come from executive, legislative and judicial
          branches. And so what you end up with, 90 years later almost, is
          something that's just sort of a mess of changes, but not
          something thoughtfully crafted by policymakers in D.C. that is
          now operating in some rational way at the state level.

   ezra klein
          So I thought a lot about whether or not to do this, because
          there is probably a best practice of podcasting where you start
          in the examples that are going to be attainable and people
          already know. And we have a bunch of those -- healthcare.gov and
          California unemployment insurance. But I want to start actually
          with a weedsier one, which is: Tell me about the Enterprise
          Services Bus.

   jennifer pahlka
          [LAUGHS] OK, you want to hear about the E.S.B.

   ezra klein
          Yeah, because I think it gets at this point of systems where
          clearly nobody was trying to do something wrong here -- and yet.

   jennifer pahlka
          Yeah, I think everyone was trying very hard to do it right. So a
          friend of mine, Matthew Weaver, was assigned to go help a team
          at Raytheon who had a contract with the Air Force to do,
          essentially, the next generation of software for satellites that
          enable Global Positioning System, G.P.S. So something that's
          pretty important to all of us, especially people like me who
          need it all the time.

          And when he got there, he found that one of the big things that
          was hanging them up was that they had to get the data from the
          satellites to the ground stations. And there's a very industry
          standard way of doing this. It's the obvious way you would do
          it. But instead, the team had inserted this thing called an
          Enterprise Service Bus, or E.S.B, which technical people will
          know was something that was really popular in the `80s and `90s.
          But it's a big, clunky piece of software.

          And it was kind of creating this Rube Goldberg machine that made
          the data go from the normal, the easy protocol into this crazy
          mess of code and then back out again. And it would time out. You
          couldn't get the data from one place to the other in time.
          They'd spent months on this and they couldn't figure it out.

          And he's like, why are we using this E.S.B. when this other
          protocol would work fine? It's a requirement in the R.F.P. that
          the Air Force had put out.

   ezra klein
          An R.F.P. being a Request For Proposal. So the R.F.P. is how the
          Air Force ends up with Raytheon building this thing.

   jennifer pahlka
          Exactly.

   ezra klein
          And we're going to talk about the contracting in a bit.

   jennifer pahlka
          Yeah, we'll talk about the contracting. So Raytheon can't take
          it out because they say, no, the Air Force is requiring us to do
          this. Why is the Air Force requiring it? Because the Department
          of Defense requires it. Why does the Department of Defense
          require it? Because there's something called the Federal
          Enterprise Architecture that requires it.

          The story of the Federal Enterprise Architecture goes back to
          this Act called Clinger Cohen, that was I think 1996, where
          Congress said we need to get a better handle on digital. We're
          not doing technology well in government. Let's make every agency
          create a plan. And then let's have those plans be coordinated.
          And the Federal Enterprise Architecture was the attempt to
          coordinate those plans.

          So one of the things they put in there was that things need to
          be interoperable. And they offered in there, as an example, this
          very common type of software at the time -- Enterprise Service
          Buses -- as an example of how you might get things to be
          interoperable. They weren't actually a great idea at that time.
          They were already kind of becoming sort of obsolete.
          Essentially, Application Programming Interfaces -- APIs -- that
          run the internet now took them over.

          The problem really wasn't that they were specifying something
          outdated. It's that because it was in this document as a
          suggested way of doing what the Federal Enterprise Architecture
          said should happen, when it got translated down to the
          Department of Defense Enterprise Architecture and the Air Force
          Enterprise Architecture, each layer in that hierarchy
          interpreted it more and more rigidly, until at the Air Force
          Enterprise Architecture level, it's assumed to be required in
          all software forever.

          And so the team is saying, I can't take this out. The law
          requires it. In fact, the law does not require an Enterprise
          Service Bus. But the people who have made all of the documents
          that govern the creation of the software for these satellites
          believe that it is.

   ezra klein
          When I was reading that story, it reminded me of a piece I just
          spent a couple of months working on about what I called
          "everything-bagel liberalism." And that was a piece that was in
          part about affordable housing and part about semiconductor
          contracting. And it was a weird piece of work, to be honest, in
          its structure.

          But when I was reporting on the, basically, request for proposal
          the Biden administration put out for its semiconductor grants,
          one of the things that came up, as I was talking to members of
          the administration about all of the additional layers of
          requirements and asks, and having a child-care center and this
          and that ended up in the bill, I would hear about people say,
          look, we think those things are good. You don't realize, we
          actually fought a bunch of these battles and knocked a bunch of
          stuff out that otherwise would have been there.

          And then I would read the text, and I would see the thing they
          told me was knocked out in the text, but it was a suggestion. It
          was a we're not requiring this, but one good example of how you
          might achieve this goal we are requiring is this. And it was
          really striking to me that people were telling me, on
          background, that they had fought and won this fight to not have
          X in the notice of funding opportunity. And then I'd read the
          notice of funding opportunity and X was there.

          And it was true that it wasn't required, but it seemed very
          likely that the semiconductor firms were going to think it was
          required or at least think -- and this was a plain reading to
          me, too -- that they would have a better chance of winning the
          grant if they put it in there. And so there is this way in which
          even when people seem to think they've created flexibility in
          the actual language, once the lawyers get hold of it and the
          contractors get hold of it, it's really not that flexible.

   jennifer pahlka
          There's so much better safe than sorry thinking going on, and I
          would say with contractors and with people who do the
          compliance. Those bids that those companies give back, they know
          they'll be judged by people who are really just checking boxes.
          Does this meet this? Does this meet this? And they're going to
          check that box.

          I think it's a good example of the disjuncture between the
          thinking of policymakers and the reality of how that policy
          plays out. Another good example of it is FISMA -- it's the
          Federal Information Security Management Act.

   ezra klein
          Yeah, which I only know because I have your quote on this
          sitting here in my next paragraph of notes.

   jennifer pahlka
          I've gotten pretty good at remembering what all the acronyms
          stand for. But there are a lot of them. It can be hard. So FISMA
          is a law that covers cybersecurity. And it has these 300
          controls in it. So 300 different things that you, according to
          law, could do to secure an application.

   ezra klein
          So it's like a menu, not 300 mandates.

   jennifer pahlka
          It's a menu. But in practice, if you ask pretty much anybody who
          does software in government, they have to do all 300. And it's
          because I think it's evolved such that in order for people to
          cover their butts, they have to explain why you didn't do
          numbers 138 through 192 or something. And it's really hard to do
          that. In fact, it's sometimes easier just to do the controls
          than it is to explain them.

          But the people in the middle who are doing the compliance have
          no idea if, for your application, numbers 35 and 95 are the two
          really, really important ones, and you should spend the rest of
          your time on the thing that's most important to security, which
          is testing. All they know is that they will be responsible if
          anything bad happens. And if it was control number 96 that got
          you, it's their butt on the line. So they will insist on all of
          them. And if you go talk to the people who wrote FISMA, they
          will say, we never intended that.

   ezra klein
          Well, this gets at a, I think, a really important point you make
          about that incentive system -- how it is actually designed
          versus how it might look from the outside. You write that public
          servants are trapped between two distinct systems of
          accountability. So what are the two?

   jennifer pahlka
          Well, when things go wrong, we're very upset with public
          servants. I mean, healthcare.gov is a famous example. There were
          like 10 hearings in one month about this. And they were
          televised. And we're upset because the thing doesn't work.
          That's an outcome. But what public servants are actually hired
          for, promoted for, what their careers depend on is fidelity to
          process. It's a compliance framework.

          And so, yeah, if you didn't do number 96 and something goes
          wrong, that's what the administrative state will focus on. The
          people who do oversight focus on whether you checked all the
          boxes not whether it works. And in fact, it's quite true, in
          people's careers, that they are not punished for something not
          working as long as they can say, I did what the procedure
          prescribed.

   ezra klein
          Something this gets at is that we have civil service
          protections. That politicians are mad at the outcome of
          something civil servants did. It might allow them to embarrass
          the agency. They might even cut funding for the agency in the
          future, although often they don't. But they can't fire anybody.

   jennifer pahlka
          No.

   ezra klein
          They can't really create that many consequences for the agency.
          Whereas, failing to follow process, you really can get fired.
          You really can lose your promotion track. There really are
          consequences. And so in a weird way, the people who made the
          policy in the first place, and, in theory, are exposed to the
          voters who are supposed to be the judges of the policy, actually
          do not have power really to hold the civil servants accountable.

          And there are good reasons for that. I mean, there are good
          reasons for civil service protections. But there is also
          consequences to it, which I think, frankly, on the left, are
          pretty underplayed.

   jennifer pahlka
          Yeah, I mean, I heard so many times, when I worked in
          government, you could get fired for that. In fact, you hear
          frequently, you could go to jail for that. It doesn't really
          match the reality. People don't get fired that much. And they
          certainly don't go to jail.

          We used to have a joke about, really, I'm going to go to
          procurement jail? Where exactly is procurement jail? Right? It's
          like such a nerdy piece of law and policy that certain people
          understand, but there's this perception that a small violation
          of it is going to end you in jail. It doesn't really match the
          reality, but the culture has definitely evolved to highlight any
          deviance from process and make it very, very risky.

          It can be risky on just a practical level. You might not get
          demoted. You might not lose your job. But suddenly people are
          not wanting to work with you. [LAUGHS] There can be cultural
          consequences to it and just your ability to get the job done.

   ezra klein
          Tell me about the Paperwork Reduction Act.

   jennifer pahlka
          So this is a great example of a law that sounds great. And it
          sounds like something that's really important to me because it's
          supposed to reduce the paperwork burden on the American public.
          And I think the paperwork burden is a real problem, in part,
          because, as I say in the book, paperwork favors the powerful.
          The more paperwork that's required, people with resources will
          get through it and people without them will not.

          So this law said, essentially, something we all believe. If you
          measure it, you'll be able to manage it. So I always call it the
          comically misnamed Paperwork Reduction Act because it's designed
          to measure the burden of any, say, form that you have to fill
          out. Like your tax form, for instance, the 1040. That has an
          O.M.B. number, which is required by the Paperwork Reduction Act.

          And it means that the Office of Information and Regulatory
          Affairs, known as OIRA, reviewed the time it would take to fill
          out this tax form and stamped its approval. And so it's got this
          measurement on it.

          The problem is that it got interpreted very broadly to say any
          time you speak to a member of the public about a form like this,
          that has to get OIRA approval.

          So in the tech world and consumer product world, we have
          something that's really important to the usability of our
          services called user research. The way we make a form easier,
          like that tax form, is that we watch people filling it out, and
          when they get stuck at a question -- like, what were your last
          year's state taxes -- we go, OK, we've got to make that easier
          on that person and we redesign it.

          So we have this very important discipline called user research.
          That is how we streamline forms and applications and make them a
          lot easier to use. So it involves something like watching you
          fill out that tax form and seeing where you get stuck, seeing
          where you have questions. Asking questions like, does this
          question actually need to be on this form? How do we make it
          easier for the user?

          But when you assume that every single interaction with a member
          of the public needs to go through OIRA, then your user research
          needs to go through an OIRA approval process. And because that
          process can easily take nine months and sometimes longer, you
          essentially can't do user research.

   ezra klein
          But this is weird because -- you point this out in the book --
          it would seem, from a plain reading of the law, that user
          research is exempted.

   jennifer pahlka
          Yes.

   ezra klein
          And yet --

   jennifer pahlka
          And yet, there are examples of this going on every day in
          federal government now. There's also examples of people getting
          through it and not being required to get OIRA approval for user
          research. But essentially, what happens is -- in fact, a friend
          of mine whom you know, Erie Meyer, who's at the Consumer
          Financial Protection Bureau, had this happen to her the other
          day. She was already doing user research. And they said, oh, no,
          we're going to have to send this through OIRA for approval. I
          mean, they were already doing it, which is the crazy thing. And
          essentially, the person in the agency decides that better is
          safe than sorry. It's better to be safe than sorry. And says, I
          understand your interpretation, but it's my butt on the line if
          someone else calls a foul on this, so we're sending it up to
          OIRA for their review.

          Now, OIRA could say, obviously, this is wrong and silly and
          there's no need to get our approval. But they also take a better
          safe than sorry stance, again, because of the incentives to
          them. And the result of Erie's request was that they said this
          will have to go up on the Federal Register, and we will have to
          get comments from the public about whether these people can get
          comments from the public. Oh, and in the meantime, there'll be
          no user research or no comments from the public happening until
          this process is over.

   ezra klein
          I just want to note that everybody in this story is a liberal
          who wants government to work well.

   jennifer pahlka
          Yeah.

   ezra klein
          I just think that's important to say here because it's something
          you say towards the end of your book, that working in government
          is a really valuable thing to do, in part because it will change
          your view on what sometimes makes government work poorly.

          And one of the almost spiritually difficult things about reading
          your book, if you're somebody who does care about government, is
          knowing that this is a lot of people who are doing their best,
          who have, in many cases, much more difficult jobs that require
          much more endurance and creativity to serve the public than
          anybody in the private sector does. And yet, we're getting this
          outcome somehow, right?

          I've known a lot of people at the top of government, the people
          who work in offices in the West Wing. And they're all very, in
          some conceptual way, aware that government is inefficient. They
          don't want it to be. Bill Clinton had Al Gore do reinventing
          government. Barack Obama hired the first C.I.O. to look over the
          government -- Chief Information Officer. Joe Biden is somebody
          who when he was a Senator often talked about government
          inefficiency. Their chiefs of staff are impatient
          management-oriented people.

   jennifer pahlka
          Yeah.

   ezra klein
          And yet, somehow from the top, even though they talk about this
          all the time on the campaign trail -- no Democrat wants to be a
          pro-inefficient government Democrat -- this does not filter down
          somehow. Why?

   jennifer pahlka
          I think you have to look not just at the executive branch in
          administrative agencies, though you shouldn't not look at them.
          I mean, I think it essentially comes down to you come in with a
          passion for cutting the red tape and making this easier. And two
          things happen. You get worn down and you have other priorities.

          It's really, really hard to spend all your time on this. And
          you're not in control of it.

          When I worked in the White House trying to stand up what became
          the United States Digital Service, we were originally going to
          do it in G.S.A.

   ezra klein
          Which is the --

   jennifer pahlka
          General Services Administration. And the head of the GSA is a
          fantastic public servant named Dan Tangherlini. And he really,
          really wanted this unit in his agency. And he was actually
          blocked by a guy below him who kept calling out that better safe
          than sorry on some very obscure interpretation of how the money
          that we were going to use to stand up the unit could be used
          legally.

          I mean, it was wrong. But he could construe the language that
          way. So you have to actually have some empathy for people at the
          top who are fighting as hard as they can to get stuff done and
          are blocked by people not just above them, but around them and
          below them.

   ezra klein
          There's another dimension here that you point out. And it's a
          sociological observation about government that as soon as I read
          it, I realized is completely true. That knowing a lot about
          politics is very high status in government. If you're considered
          a tactical political genius, that's great. Knowing a lot about
          policy, having very big policy ideas, that's also very high
          status. Implementation isn't.

   jennifer pahlka
          No.

   ezra klein
          Why?

   jennifer pahlka
          I think it goes back very deep in our culture. I mean, I think
          that's not just true in government. Let's be honest, right? Big
          ideas people get a lot of status in our society.

   ezra klein
          Thank you for listening to my Ted Talk.

   jennifer pahlka
          Well, yes, big ideas guy. [LAUGHS]

          A friend of mine who worked in the government digital service in
          the U.K. pointed me to this idea that's written about the
          British civil service. That it is literally divided between the
          intellectuals and the mechanicals.

          Now, I think that the digital age has complicated this in a lot
          of ways. If you look at, say, metaphysical Silicon Valley, if
          coding is a mechanical task, it's the way you implement
          something, all those companies were founded by coders.
          Mechanicals in that framework are at the top.

          And lots of interesting -- good and bad -- things have come out
          of that culture, which is really distinct from the culture of
          government where the intellectuals are at the top and have the
          power. And the people who write the code or do the designs that
          implement these programs are really, really, really far down.
          They're at the very bottom.

   ezra klein
          In a way, they're not even at the bottom. They're outside.

   jennifer pahlka
          They're outside. This is true.

   ezra klein
          Tell me about Circular A-76.

   jennifer pahlka
          So a couple of decades ago, the federal government said -- this
          is actually coming out of O.M.B., the Office of Management and
          Budget, but it really derives from a long history of trying to
          streamline government -- said, if someone else can do your job
          in government, we're going to outsource it. It's the reason we
          have things like the concession stands in national parks are run
          by third parties. They're not run by the government. That's all
          A-76.

          And there's a real valid justification for some of this. There's
          a famous case in which, I think during World War II, government
          decided to make its own steel because there wasn't enough. Well,
          by the time they finished that, it was wildly expensive and very
          bad. And it's a commodity. The market can provide us good steel.
          That is the best way to do that. We should not get in the
          business of steel.

          And so on that basis, they roped a lot of other things into
          that, like concession stands in federal parks. There's no reason
          that's inherently governmental, which is an important phrase
          that very few people talk about. Because the distinction between
          something that is inherently governmental, like a government
          person must make this decision -- i.e. most policy decisions --
          and something that is inherently commercial is like, OK, well,
          we'll make the policy decision and then we'll contract it out to
          somebody who just delivers on it.

          And it creates this real separation between policy and
          implementation, which has a lot of complications to it. And I
          think over time has snowballed into some really bad outcomes,
          because we're trying to do policy over here, implementation over
          here -- I have my hands very far apart since your listeners
          can't see -- and so much gets lost in the middle.

   ezra klein
          So that's, in a way, the guiding principle here. You say it's
          gone now really far. I mean, you go quoting someone else as far
          as really saying that civil servants manage; they don't
          implement. So particularly in the digital sector, tell me about
          what the government does and what it contracts out. What
          functions in your experience are done by somebody on a public
          payroll? And what functions are done by somebody who is being
          paid out of public taxes but is on a private payroll?

   jennifer pahlka
          So let me speak to what's still pretty standard but changing,
          like when I was in government in 2013 and 2014, was really,
          really standard. If you're trying to get, say, healthcare.gov
          built, and you're the digital person responsible for it, what
          you will actually do is work with contracting officers, people
          to gather a bunch of requirements, lawyers, and you will have to
          be master of an incredibly complex apparatus of procurement
          rules, and operations, and forms that need to be filled out, and
          processes that need to be followed.

          That actually is a very robust skill set and knowledge base in
          its own in order to hire people who will actually do the digital
          implementation of whatever it is you're doing. So I talk about
          somebody in the book who's, in fact, a fantastic digital leader
          and gets really good outcomes because he is a good digital
          leader. But like 90 percent of his time is spent on procurement
          and legal.

   ezra klein
          So I want to talk about how procurement actually works. Because
          just like it is different to hire and fire in the public sector
          because of civil service rules, it is different to procure, to
          pay an outside vendor to do something in the public sector.
          Because we have a lot of rules meant to clamp down on cronyism,
          on patronage. We're trying to make the process fair, and we make
          it cumbersome.

          One of the things you talk about a bit is the way the system can
          get slowed down by vendor protest. That if you don't win a
          contract, you can actually protest the entire process, holding
          it up. Sort of like if you are upset about people building
          something, you can bring an Environmental Quality Act objection.
          Tell me about these protests. Tell me about the ways in which
          the vendors actually have quite a lot of power in the process.

   jennifer pahlka
          Yeah, these rules were designed to make the process fair. So for
          instance, an agency will announce a winning bid. And if other
          companies who bid on this think that either something genuinely
          went wrong or more likely that they can simply take some
          advantage out of this, they can protest the bid.

          And because these bids are so wildly complex -- I mean, the
          paperwork required to bid on a major government project, whether
          it's technology or anything else, is massive. There's a lot to
          review. And there's very little evidence of actual competency
          that are required in this paperwork. It's all evidence of
          compliance.

          And so it's very easy to find something that's not compliant.
          But the result is easily six months, often year or two, delays
          in getting something done.

          [MUSIC PLAYING]

   ezra klein
          I think, actually, the first time we started talking, I became
          more aware of your work, was in the healthcare.gov debacle. And
          at the core of that debacle was procurement and who got chosen
          to build that. So how did these rules end up affecting who got
          that contract?

   jennifer pahlka
          If you look at the timeline in the A.C.A. -- and I'm going to
          forget the exact times -- but essentially, they only had three
          years. The Centers for Medicare and Medicaid Services, C.M.S.,
          had about three years to get healthcare.gov up and running. They
          needed to launch it by Oct. 1, 2013, if I recall.

          And as I just described, the process of getting a bid together,
          that's incredibly complex. Getting all your requirements,
          getting that contract out to bid, reviewing the bids and then
          awarding it and then, potentially, protests, that's easily three
          years right there. So they had to do something else.

          So CMS actually had some vendors on a very obscure vehicle, we
          call it, called I.D.I.Q. It stands for Indefinite Delivery
          Indefinite Quantity, which is essentially a way of saying I've
          hired you to be around and do what I need. [LAUGHS] Actually, an
          I.D.I.Q. can be a very good thing to have someone just
          on-boarded and ready.

   ezra klein
          They're basically on retainer.

   jennifer pahlka
          They're on retainer. But yeah, for a wide variety of things. And
          so C.G.I. was one of the companies that was on an I.D.I.Q.
          already at C.M.S., so they didn't have to bid certain pieces of
          it out. Now, let me be clear, there were 60 different contracts
          to vendors to build healthcare.gov. Some number -- like 12 of
          them -- went to C.G.I., which is considered the main contractor,
          and others went to others. So it's not like they had the whole
          thing.

          But they really couldn't have done that full process where they
          bid it out, because if anybody had protested, they had no hope
          of making their congressionally mandated deadline.

   ezra klein
          I think it's worth spending a minute here. Because I covered the
          drafting of the Affordable Care Act really closely. I covered
          how they were thinking about building the --

   jennifer pahlka
          And you missed this. [LAUGHS]

   ezra klein
          Not only did I miss this, but I can say with certainty that
          there was a lot of excitement in Congress about we're going to
          create these online digital marketplaces and it's going to be a
          next generation website. You're going to be able to compare
          everything.

          And I'm sure in Congress's mind, having talked to people, the
          way they thought about this is the government would go for this
          unbelievably marquee project and get a really great website
          design player. I mean, Google doesn't do a ton of government
          contracting in this particular way. They do other things. But
          someone like a Google, just whoever is the best. And this would
          be a great project.

          And instead, you end up with a pretty obscure contractor -- and
          to me this is important -- who -- I'm not saying this is their
          only competency -- but across a bunch of these contractors, what
          they have is a competency in contracting.

   jennifer pahlka
          Yeah.

   ezra klein
          They know how to work with the government, which is a huge pain.
          And this feels like another place where the procurement rules
          and the way different regulations, et cetera, have been
          interpreted, it's really not what the legislature was trying to
          do. They wanted a great website.

   jennifer pahlka
          Yeah.

   ezra klein
          What they got was a total disaster. But they didn't then go back
          and change procurement rules either, really, to my knowledge.

   jennifer pahlka
          Well, the legislators would say, we have given agencies -- and
          I'm not sure about this particular case at that particular time
          -- but this happens in the Department of Defense all the time.
          We have given you lots of exceptions that you can use. There's
          something called an Other Transaction Authority, which basically
          gets rid of a lot of those rules and lets you choose who you
          want to choose in a streamlined kind of way.

          But they just don't get used. And that, again, goes back to
          people say better safe than sorry.

          If something goes wrong, someone is going to come back and say,
          what justification did you have for using O.T.A., Other
          Transaction Authority, instead of the more robust way that we
          know will ensure, quote unquote, "fairness"? So it's really that
          culture that comes back to it.

          I mean, it's a reasonable example I think of what I talk about
          in the book is culture eats policy, right? The policy that is
          designed by policymakers wants to get the outcome of flexibility
          for the agencies because they know that the rigidity is hurting
          the outcomes. But the culture just goes back to, I see that
          you've given us that, I'm choosing not to take it because I feel
          like it's going to make me safer.

   ezra klein
          So healthcare.gov was this huge project. It had to be contracted
          from top to bottom. But I also want to talk about the ways in
          which this kind of contracting makes it hard to update systems
          as policy is changing. And I want to use a pretty horrific
          example that you talk about in the book, which is the child
          separation policy at the border.

          So obviously, the actual policy there was terrible. But one
          thing that made it much worse was the rigidity of the system
          being used. Can you go through that?

   jennifer pahlka
          When the Trump administration ordered that kids and parents be
          separated at the border, the systems that tracked families could
          only assign one what's called A number, an Alien number, to a
          family. That's how it was set up and had worked for years.

   ezra klein
          So just horrible. It's called an alien number.

   jennifer pahlka
          It's called an alien number. Though, they call it A number,
          which is also really confusing. So if a family of four comes
          through, they get one A number. Because in the past, they were
          treated as a unit. This horrific and cruel policy comes down.
          They separate them.

          They now give you one A number, your partner one A number, and
          your kids get A numbers separately. And there is no affordance
          for connecting your A number with your children's. They were
          pinning sticky notes with A numbers on infants who had been
          taken from their parents.

          And there were wonderful public servants who were writing down,
          keeping spreadsheets, trying to track the connections between
          parents and kids. It wasn't like there was no effort made. But
          there wasn't a systematic way to do it. There was a lot of
          chaos. The person who's trying to reunite them is not
          necessarily the person who separated them in the first place.

          And so when, thankfully, that order was reversed, there was very
          little capacity within the Customs and Border patrol to fulfill
          the demand to reverse it and reunite these kids because the
          software didn't have any affordance for making this connection.

   ezra klein
          So something you then talk about is it would have been a fairly
          obvious move -- again, without either of us endorsing a terrible
          policy -- but it would've been a fairly obvious move, if you're
          doing that policy, to change the software so that the A numbers
          are tracked together. Why didn't they? What would that have
          taken?

   jennifer pahlka
          So I should caveat by saying, I don't know the specifics of the
          C.B.P. contracts that govern this software. And I've never seen
          this software in action. But I do know that the way this
          generally works in government is if you want an update to your
          software, you take it out to bid. And that takes a really long
          time. And yes, the company that may have originally written it
          will bid on it, but you're not guaranteed to give it to them.
          Maybe that's the right thing. Maybe that's the wrong thing. But
          if you have a new vendor, they have to get up to speed on the
          code base.

          But essentially, the way we think about software in government,
          it's a project. You start it. You finish it. And then there's a
          little bit of money trailing around for you to keep it up and
          running. So when you have a change like this, that's frankly
          incredibly urgent, you don't have people who work on that
          software day to day, week to week that are there ready to change
          it and know how the code base works.

          So in contrast, Google didn't create Search and then lay off 95
          percent of the staff and keep just a skeleton crew around it to
          keep it up. The way things that really work in the real world
          are staffed is on an ongoing basis, not only to make sure that
          changes in some connecting database somewhere can be updated so
          the thing doesn't crash, but also because we live in a dynamic
          world. We can't expect there not to be policy changes that have
          to be incorporated.

          And frankly, we should be constantly looking at the software and
          saying, is it doing what it could be doing as well as it could?
          There's constant improvements in the consumer products that we
          use. So the curve of staffing in a consumer product is it starts
          really slow, and then as the product becomes more and more
          clearly valuable and popular, the staffing goes up and up and
          up. It's just this constant curve up. That's how the software is
          constantly better and more usable and more valuable to us.

          In government, it's this weird, like, there's nothing, then you
          hire all these people, and then there's this couple-year period
          where lots of people are working on it, and then it drops to
          almost zero. So after almost zero, you have no capability to
          change that software. You have acquired software, but you have
          not acquired the capability of doing what that software is
          supposed to do, of performing the function on an ongoing basis
          that that agency is responsible for.

   ezra klein
          I think this brings us back to healthcare.gov, which is a real,
          I think, before and after moment for how digital is viewed
          inside the government, because it is built and broken by this
          contractor or set of contractors. And the rescue effort is a
          much more internal process that is built on now the idea that
          you actually need internal capacity and flexibility, and because
          of the president is focused. There's a lot more ability to move
          quickly and make decisions. So tell me a bit about the
          principles and staffing that the rescue effort is built on.

   jennifer pahlka
          The rescue effort came at a time when we were trying to prove
          the point that you need an internal capacity. So as disastrous
          and horrible as it was, it was actually really useful for me and
          others who were saying, hey, we can't just have procurement
          competence in government. We have to have some digital
          competence in government. And in fact, we need it in the halls
          of power in the White House and in O.M.B.

          And I think one of the reasons it worked is that they could see,
          through the rescue effort, that we could bring in people whose
          competence wasn't how do we contract with company, but how does
          the software actually work and what's wrong with it. So the
          people who came in were software engineers. Many of them happen
          to have this title at Google and other companies called Site
          Reliability Engineers, S.R.E.s. You've never heard of an S.R.E.
          in government. There literally never was one. That was always
          the responsibility of an outside company.

          So you had people who could actually look at the code. And in
          fact, we found that oftentimes the people at the vendor who were
          supposed to have done the coding didn't really even know how to
          code. So you had all these people who knew how to work with
          government, but didn't actually know how to make the site work.
          And bringing in folks who could actually get in there in the
          weeds was the thing that changed it.

          The process up to that point had relied a lot on planning.
          There's this real focus on planning in government. For lots of
          jobs, if you want to get promoted, you have to have this thing
          called a P.M.P., a Project Management Professional. And it's all
          about upfront planning. But it's really hard to plan in a
          situation like healthcare.gov, where they are still finalizing
          the regulations and rules that govern it right up to the launch
          of the site.

          So the skill that was needed was not this very traditional, plan
          everything, then execute to it. You needed people who were
          basically good at what's going on with users, is it working,
          just go find problems and solve problems at a very detailed,
          technical level, instead of looking at project plans that were
          not actually connected to the experience that users have.

   ezra klein
          Well, one thing you say about the healthcare.gov experience is
          that there were a lot of reasons it went wrong. And there are
          G.A.O. reports on it and --

   jennifer pahlka
          Very good ones.

   ezra klein
          I wrote a piece for "Businessweek" about it back at that time.
          And I mean, there was a huge amount of attention on this. But
          you say if you could boil it down to one, it's that there was a
          lot of process management, program management, but not a lot of
          product management, not a lot of people willing to say, no,
          we're not going to do that, empowered to say, no, we're not
          going to do that. Can you talk about that distinction?

   jennifer pahlka
          So project management is the art of getting things done. It is
          hugely important. I value good project managers very highly and
          everyone should. Product management is the art of deciding what
          to do in the first place. And if we don't do that, then the
          project managers just are overwhelmed with a undifferentiated,
          unprioritized mess of requirements that they have to fulfill.

          And yes, healthcare.gov, there was a lot written in the law
          about what needed to happen. It was probably overspecified. But
          there was also not a willingness to even imagine that you might
          do less. In fact, people were told you can't do less. You can't
          not serve every single edge case from day one, for instance.
          That's illegal. There's reasons why they think it's illegal. I
          do not think it's illegal, and I think we've proven since then
          that it's not.

          But that willingness to say it will launch serving the most
          common set of users, because that way we can handle the people
          who can just use a website to get their health insurance. And
          then the people who have crazy exceptions or rarer exceptions
          are going to go through the call center, they're going to go
          through those in-person service centers that they set up around
          the country, which makes sense. Those people actually probably
          need a person walking them through it because of their
          exceptions.

          It doesn't mean that the website couldn't eventually serve all
          of those people. It means that to get a site that launches and
          works on day one, you constrain what you're trying to do. Get
          those first people through the door. Test the site that way. And
          then you add the functionality later. That's good product
          management. And until recently, it literally hasn't existed as a
          discipline within government.

   ezra klein
          There's both product management as a discipline, and I think
          there's also product management or, in this case, I would call
          prioritization, as a political value, as something the
          leadership values. And you tell a story in there of a conference
          call while the site is in crisis with Marilyn Tavenner, who is
          head of C.M.S., the Centers for Medicare and Medicaid Services.
          Could you tell that?

   jennifer pahlka
          Mina Hsiang was somebody who joined the what we'll call rescue
          effort, though I do try to stay away from that term because
          there are so many great public servants who were part of the
          rescue effort and they often feel unseen by that. But Mina
          Hsiang came in to help.

          And one of her responsibilities was to staff a daily call with
          the administrator, Marilyn Tavenner, and others, including the
          main contractors. And the site was in total disarray. I mean, I
          don't remember exactly how long this was in. But this was before
          we were headed towards, OK, it's starting to work again.

          And one of the things that she brought up on this call was that
          in the original spec, there was going to be a separate site in
          Spanish. Makes sense. I think government services should
          absolutely be in as many languages as we can to serve people as
          well as we can. But at that point, even the English language
          site was very far from being able to work.

          So Tavenner brings this up on the call and says, where's my
          Spanish site? Now, there's all these vendors on the call, so
          Mina knows that they're going to try to placate her and say
          we're going to get to that and then divert resources to the
          Spanish site. So she gets ahead of them and just jumps in and
          says, I'm sorry, Administrator, but you can either have one site
          that works or two that don't. And Tavenner hangs up on her. She
          doesn't want to hear it.

          And God love Mina, she just kept persisting. But her job at that
          point was to keep the team's eye on the ball, which is we have
          to get this site to the point where it's operational and then we
          can use those resources to make a Spanish version. But it would
          be pointless to make a Spanish version of something that didn't
          work. [LAUGHS]

   ezra klein
          One thing that is a worrying dimension of that story is that --
          and you see this in government a lot -- that when things are in
          crisis, people realize they have to bend rules. They have to
          prioritize. They can't do everything at once. They have to make
          the first thing work first.

          But a lot of things aren't in crisis or -- and I think this is
          actually more often the case -- they are, but they're not in
          crisis for the right people, which is to say I think there's a
          lot of attention to delivery and implementation when the people
          being delivered to and implemented upon are middle class, are
          upper middle class, are rich. So healthcare.gov was affecting so
          many people simultaneously.

          If you think about the unemployment insurance failures during
          the expanded unemployment insurance during Covid, that was
          affecting a lot of middle class people who had political power
          all at once. There are obviously a lot of programs where they
          work very poorly for poor people in an ongoing way, but they're
          not a crisis.

          And so the ability to say to the administrator or to the
          secretary or to the president or whomever, if they even care
          about this, hey, we're just not going to do that because this
          crisis is so big, we have to do the first thing first, isn't
          there. And so I worry a bit about that. I'm curious to hear your
          reflections on how things work in what you might call
          implementation peacetime versus implementation wartime.

   jennifer pahlka
          Yeah, we don't get as much done in peacetime. And I think one of
          the tactics of people who I respect in this community is to make
          it a crisis, to make the crisis visible. I'd argue that it was a
          crisis that very few people were getting enrolled in SNAP. But
          it was not hitting the headlines for sure for all the reasons
          that you mentioned. And some great people at Code for America
          made that a lot more visible and change got to happen.

          I talk in the book about what was certainly a crisis at the
          Department of Veterans Affairs where veterans couldn't sign up
          for health care, but it was completely invisible to the people
          inside the building because the form that was broken, if you
          tried to use it from any computer at a library or at your home,
          wasn't set up with the same combination of Internet Explorer and
          Adobe Reader that the computers inside the building were set up
          with. And so the form worked fine for them.

          And the tactic that the team used to finally get this fixed was
          to record a veteran trying to use the service and show the
          people in power how humiliating and frustrating the experience
          was for veterans. But until they were able to bring this veteran
          named Dominick's experience to life inside the building at the
          Department of Veterans Affairs, they were literally told there
          was no work to be done because the requirements in the contract
          had all been fulfilled and all the boxes had been checked.

          So you've got to sometimes make the crisis visible when it just
          won't be because the people who it affects don't have a voice.

          [MUSIC PLAYING]

   ezra klein
          One thing that comes out of the Veterans administration story
          you tell is the way that what a crisis is to the people using
          government services and what a crisis is to the people making
          them can be very different. And I would say throughout the book,
          you're very sympathetic, even to people you're at cross-purposes
          with. You're trying to get something done and they're standing
          in your way. And you tend to see where they're coming from.

          But in this particular story, in the veteran's story, you talk
          about somebody you call Kevin and the attitude that he and his
          team have developed towards development, towards keeping
          themselves out of hot water. Tell me about Kevin.

   jennifer pahlka
          Kevin was also at the Department of Veterans Affairs, but this
          is actually a separate interaction. I had started working with a
          small team to understand problems with the veterans benefit
          management system, which is the internal way that you would
          process people's requests for benefits. We had been told that
          there was terrible latency in the system. Essentially, you can
          click on a link to start processing something and you'd wait for
          a really long time before the page would load. It's very
          frustrating.

          We met with the guy the first day and he said, I'm so glad the
          White House decided to send people to verify that everything is
          OK because latency has been solved. Now, we found out later he
          solved latency by defining it as over two minutes. So if you
          clicked and had to wait two minutes or one minute and 59 seconds
          for that page to load, you were not to report it as latency. So
          that was the solution. [LAUGHS]

          That was my first clue that this man I called Kevin was more
          interested in how things looked than how they actually operated.
          But as I asked him questions about this system -- why had they
          decided to do X instead of Y, what role did the claims
          processors have in this particular part of it -- he kept saying,
          that's not my call, I don't know that. And I asked him why.

          And he said, look, I've spent my entire career training my
          people not to have an opinion on the business requirements. If
          they ask us to build a concrete boat, we'll build a concrete
          boat. And I said, good lord, why? And he said, because that way
          it's not our fault.

          And honestly, that moment for me -- I was seven months into
          working in government then. I'd worked at Code for America
          before. I'd worked with government. I was not new to the
          dysfunctions of government, and I shouldn't have been shocked.
          But I really did take it as a gut punch.

          I didn't blame him, in a certain sense, because I knew what he
          meant about the way we structure responsibility in government.
          But at the time, I think 18 veterans a day were committing
          suicide. Many of them were still waiting for their benefits. And
          it just seemed like such a callous thing to say when, if he
          would push back a little and say this concrete boat is never
          going to float, we could have gotten more veterans their
          benefits.

   ezra klein
          When I finished the book, I had this note to myself that this
          book is about personnel, capacity and autonomy -- that it's
          about who works in government, what kinds of things government
          does versus doesn't do. And then the autonomy, the agency, the
          power we give these people to do the jobs we ask them to do.

   jennifer pahlka
          That's your TED Talk.

   ezra klein
          Yeah, there you go. My TED Talk, a summary of Jen Pahlka's book.
          But one thing that I have always worried about being close to a
          lot of people who work in agencies, and I think that you see in
          that story, is when you make it really hard and annoying to work
          in government, it's not that it pays so little -- you have a
          good section on this. It's not that government pays so little.

          But it is often very frustrating work. You have to deal with all
          these weird procedures and the lawyers tell you you can't do it.
          I mean, working on code in government -- and this is true in a
          lot of your stories, including the stories of your heroes --
          often requires an almost superhuman level of persistence. You're
          trying to get around a lot of very annoying things. And
          sometimes you get a hero out of that.

          But it also, over time, drives out a lot of good people who just
          don't want their days to be annoying. And it also selects for
          people who can make the internal compromise -- in that story,
          Kevin made -- where you sort of shut off the part of you that
          says that doesn't make sense and I'm going to argue -- you lose
          enough fights like that. The lawyers tell you no enough times,
          you get yelled at by your superiors enough times, and either you
          leave or you say, I'm going to stop fighting.

          And that's where things can go, I think, really badly. That it's
          not just that we need good people to work in government. You
          hear that all the time. But we need government to be a good
          place for people to work. And for a lot of creative, impatient
          people, particularly those at mid-levels, not those who are
          appointed to run an agency or run an office in the White House
          or something, it gets very frustrating.

          I've seen a lot of really good people go in, and particularly
          when they're in older agencies that already know how they do
          things, I've watched them go right back out. And I don't think
          they weren't committed. I just think people -- it's hard to have
          a really irritating job.

   jennifer pahlka
          I really think that we're so focused on getting these outcomes
          from a policy sense. And to solve the problem you're talking
          about, we've got to really focus on just basically the overall
          health of the civil service. Yes, it's frustrating to fight
          these battles. I think there's increasingly, certainly in my
          experience, a community of people who support each other around
          it that really makes it tolerable.

          When people stay it's because they've got someone's shoulder to
          cry on when they need it and people to tell them keep going. I
          have found that incredibly valuable and incredibly meaningful to
          me to be part of that. But for instance, one challenge they face
          is they also really can't hire other people quickly. And there's
          a lot that we --

   ezra klein
          Or fire people.

   jennifer pahlka
          Or fire people. We could do so much -- and by we, I mean more
          our elected leaders and how we hold them accountable -- to
          making that environment less destructive, less wearing you down,
          and giving public servants like those who are high profile, who
          are just persistent and heroic, the tools they need to succeed.

          But we just don't think of the tools of legislators or the
          legislative branch as being good at that, right? They have
          money, they have rules, and they have oversight. And those
          things aren't working very well. But if they targeted them at
          the health of the civil service and really focused on it, I
          actually think we could make that a place that doesn't drive
          people away as much.

   ezra klein
          Well, tell me what that would look like.

   jennifer pahlka
          So for example, there are bottlenecks in the system that are
          just vestigial. Like O.P.M. requires that there be industrial
          organizational psychologists that approve a bunch of things, the
          same way that we have OIRA as a bottleneck to getting user
          research out the door. Remove it. But nobody's on that. Nobody's
          looking at these little things like that that, one by one, could
          actually speed hiring and make it easier for people. But there's
          probably 7,000 little things like that.

          And one of my complaints, since you started on the left, is that
          the right actually has a lot of plans around this that are
          focused on firing people. The left really doesn't pay much
          attention to this. They're not obsessed with the details of
          making the civil service have greater autonomy and freedom and
          hiring the right people, and yes, probably firing the right
          people.

          I actually want the right and the left to say this is actually
          more important than this, that, or the other policy outcome that
          we're trying to get because it will enable all of those policy
          outcomes. But it's real work that somebody has to fund and
          somebody has to do. I know probably half a dozen things that
          would go in that, but somebody has to go find the other 36.

   ezra klein
          My impression of this area of politics is that it's become, as
          many things do, very dysfunctional through a kind of crude
          polarization, which is to say that civil service reform, by
          which I mean making it easier to hire and fire, giving the civil
          service more autonomy, but also making sure there's more
          accountability, including accountability from the political
          system itself, because that is often -- parts of that have often
          been a feature of the right wing attacks. And often they are
          attacks that also want to outsource a lot of functions of a
          civil service and starve it of money and so on. They've been
          coded as right. So the way to defend and show you care about
          government is to protect it from these attacks. But if you then
          cover government or you actually know people who work in it,
          even people in the civil service don't like how this works. I
          mean, the difficulty of --

   jennifer pahlka
          They hate it.

   ezra klein
          -- listing a job to get anyone in -- we're so worried about
          patronage. We're so worried about corruption that we've made the
          whole thing work really poorly. And we drive people out because
          they just cannot stand it over time. And many of the people who
          can, we're either asking, again, superhuman work out of them, or
          we're accepting people who maybe are not the people we wish were
          in charge of some of these things. And that's not to take away
          from a lot of people doing hard work.

   jennifer pahlka
          Yeah.

   ezra klein
          But it's also a kind of fallacy if you've ever worked in any
          kind of organization to suggest that everybody in an
          organization is the right person at the right place. I think the
          left has a lot of trouble with this kind of statement, but it is
          important to be able to fire the right people to make
          organizations work. It's not pleasant. It's not what anybody
          should want to do. But to not be able to do it is a way to
          guarantee organizational dysfunction.

   jennifer pahlka
          I agree with that. I think the order of operations, though,
          should be first we make it easier to hire.

   ezra klein
          Yes, no, totally. I think the reason this all gets stopped on
          the left is they're so afraid of the firing conversation. That's
          why I bring this up.

   jennifer pahlka
          They are, and there's some good reasons for that, right? So if
          Schedule F comes back in a second Trump administration or a
          DeSantis --

   ezra klein
          Can you say what Schedule F is?

   jennifer pahlka
          Yeah, so Schedule F was something that Trump instituted, I think
          in the last couple of weeks of his presidency, that would allow
          him to arbitrarily reassign someone from a civil service
          position into a political position. What's the difference?
          Obviously, you can fire a political position, a political
          appointee with very little paperwork. It's quite easy. Whereas,
          a civil service person, you're supposed to not be able to fire
          unless there's -- you go through -- jump through a whole lot of
          hoops.

          I guess what I would say first is both people on the left and
          the right who work in government -- and, of course, civil
          servants are meant to be nonpolitical -- are frustrated with the
          same thing. And I think that polarization is really sad because
          we're actually responding to the same frustration of this
          demosclerosis, things moving too slowly.

          But on a very practical basis, people I see working in
          government are now trying to solve problems. Yes, they have
          these obstructionists around them saying, sorry, this is how
          it's always been done. The E.S.B. has to stay in the software.
          We're going to build a concrete boat. But the first way to
          combat that is to get other people in to balance them out.

          There's a concept always in government and much theory about
          government, which is you're trying to balance between agility
          and stability. And bringing in more people with an agility
          mindset is a good way to balance that, whereas just firing
          everybody is a great way to potentially, in another Trump
          administration, enable authoritarianism.

          So there's a reason they're upset about it. But I think that if
          the left doesn't start acknowledging that we have the same
          frustrations with the civil service and the administrative state
          as the right, we're never going to get anywhere. And I actually
          spend a lot of my time with people on the right with whom I
          agree a lot more than I thought I would.

   ezra klein
          I want to disentangle some thoughts here, at least for me, which
          is, one, I think people go too quickly to the idea that if you
          created a lot more autonomy and flexibility within the civil
          service, made it easier to hire, easier to fire, that has to
          mean it's easier for the political system to hire and fire,
          which I do think is something we should be at least a bit
          careful with.

          Whereas, I'm talking really about the ability of the civil
          service to hire and fire internally unto itself, which is a
          different question already because they're just different rules.
          That's why in Schedule F, Trump is trying to move people from
          civil to political. But you could change the rules for what the
          civil servants can do internally and still keep -- I think maybe
          you do want less wall than you currently have, but a pretty good
          wall between politics and the civil service.

          And then the other thing is I really am skeptical that the right
          and the left have the same concerns here. And there are places
          where procedures might converge. There are places where some of
          the frustrations might converge. Our friends didn't see this a
          lot in a lot of the work I'm doing around permitting reform and
          supply side issues. That often something that is important to me
          now is deregulating the government. And --

   jennifer pahlka
          We're not supposed to call it deregulating.

   ezra klein
          I know.

   jennifer pahlka
          But that is what it is.

   ezra klein
          But that is what it is. I want the government to deregulate
          itself, oftentimes. And deregulation is coded as a thing you do
          to make the government less powerful on the right. But
          oftentimes, I think regulation is one reason the government is
          not able to achieve its ends. And so it kind of sounds to some
          people like I am saying the same thing that the right is saying,
          but I'm not. I actually want a more active and capable
          government.

          And there are things you might want to do from a deregulatory
          standpoint that would get you there. There are other things you
          would want to do from a deregulatory standpoint that would not.
          So many people on the right do not want the E.P.A. to be able to
          regulate carbon. That is a kind of deregulatory decision, but it
          isn't the one I would make. I would like it to be easier when
          the government wants to build clean energy for them to get the
          permits done.

   jennifer pahlka
          Yeah.

   ezra klein
          That is a deregulatory decision that makes the government
          stronger. So I just think that's also true sometimes within the
          civil service debates, that I would like to see a more capable,
          stronger -- I'd like to see more people employed, in many cases,
          by the government. We've outsourced a lot of functions, which is
          not oftentimes what the right means when they say they want to
          do civil service reform. They often don't mean they want to
          bring a lot of these functions back into government purview.

   jennifer pahlka
          Let me say three quick things to that because it's a really
          important point. And it does get gnarly in there.

   ezra klein
          Yeah, totally.

   jennifer pahlka
          One is, yes, but I also think change happens when people with
          different views do have something in common even if they're
          going to use it for different ends, right? And I think we
          probably have some fertile soil in there to get change done,
          even if people on the right want that change for a different
          reason. Obviously, we need to be careful about that. I agree we
          should definitely be regulating pollution.

   ezra klein
          Oh, it's sort of the last thing I'm going to say on that, which
          is that there's the right and there's Trump.

   jennifer pahlka
          Yes, exactly. Well, that was what I was going to say.

   ezra klein
          And there are people on the right you really can have this
          conversation with. It's just not Trump.

   jennifer pahlka
          So that was my third point. My second point was it isn't
          regulation or not regulation. It's how we do it. And when we
          talk about competencies and capacities in government, I think
          one of the main points I'm trying to make is that has to be a
          core competency in government. And we don't even recognize it.

   ezra klein
          I really take the big point of the book as saying there is
          politics, there's policy, and there's delivery. And right now
          delivery is so subordinate to the people who run politics and
          the people who run policy, that in addition to delivery being
          done poorly, the people delivering often don't think they have
          any power.

          They don't have power to interpret the language they're given.
          They don't have power to say no to the lawyers. There's nobody
          empowered over a process to be flexible as it is ongoing. They
          often don't have the power to change contracts quickly.

          And that shift of power from policy or politics to delivery,
          that feels to me like the great revolution you're trying to kick
          off here. And I'm curious if you had two or three places you
          could start kicking it off -- aside from everybody reading your
          book, and they should read your book -- what would you do?
          Because, yeah, culture can need policy, but policy can also
          create culture.

   jennifer pahlka
          I think there's a lot of things that can happen that fall under
          the banner of where we pay attention and where our elected
          leaders pay attention.

          So one really practical thing -- and I have an example of it at
          the end of the book -- is that when we're writing law and
          policy, who's at the table?

          And I take very much your point about everything-bagel
          liberalism where you already have too many people at the table
          and that sounds like just adding someone else on. But if we want
          to make law and policy that are implementable, then we have to
          have implementers at the table when that law and policy is being
          written. And those people are likely to be and hopefully are
          much closer to the users of the system. And they're going to
          provide insights that make your policy better. But they're also
          going to start to shift that power balance that you talked
          about.

          So an example I talk about in the book is clearing of criminal
          records where the way a particular law was written made it kind
          of impossible to implement at scale. But in the aftermath of
          that, you've got people with technology skills who spend time
          with the people who have these convictions and understand the
          constraints in their lives and how they're going to interact
          with it at the table when decriminalization laws are being
          written, those decriminalization laws are usually the ones that
          contain records expungement provisions.

          That's a huge difference. People can barely believe that this is
          happening -- that you've got coders and designers and user
          researchers at the table. This is at the state level when laws
          are being written. So that's one thing.

          There's a saying -- what we pay attention to grows. So our
          elected leaders have this oversight responsibility as part of
          their toolkit. And we also have the whole apparatus of oversight
          in the executive branch, offices -- of Inspector General, the
          Government Accountability Office, G.A.O. And what do they focus
          on? Failures.

          What if they lifted up the public servants who were having
          success by empowering themselves to have that flexibility and
          made half of their work or more shining a light on them and
          saying, this is what we want public servants to be like and
          we're going to reward them? I think it would make a huge change
          in the culture.

          And I do have several policy prescriptions in the book about
          things like how money is allocated that I think are important to
          pursue. But in the end, I keep coming back to those policies
          fail if the culture doesn't change. So the ways we pay
          attention, what we talk about, who we elevate, I think those are
          all reasonable levers and good places to start.

   ezra klein
          So I think that's a good place to end, too. So what are three
          books that have influenced you that you would recommend to the
          audience?

   jennifer pahlka
          The first is a book from 1973 with the longest subtitle I have
          ever seen. It is called "Implementation: How Great Expectations
          in Washington are Dashed in Oakland; Or, Why It's Amazing that
          Federal Programs Work at All, This Being a Saga of the Economic
          Development Administration as Told to Two Sympathetic Observers
          Who Seek to Build Morals on a Foundation of Ruined Hopes." What
          do you think of that?

   ezra klein
          I love that because that's exactly how pamphlets in early
          American history and 17th and 18th century Europe sound. And I
          love those title formats.

   jennifer pahlka
          Well, you're going to love this book. It covers a project that
          started in, I think, 1967, where politicians and bureaucrats
          went and made a big announcement about investment that was going
          to happen in Oakland, which is my hometown and somewhere I think
          you have lived. And everybody celebrated. And then goes into
          excruciating detail, in the same way some of my book goes into
          excruciating detail, about why several years later essentially
          nothing had happened.

          And they really conclude that this separation between policy and
          implementation is the only reasonable explanation for the
          failure. And they pull up all the other obvious reasons you
          would think that it failed and really debunk them. And it's just
          -- it was sort of like they were reading my mail back in 1973.
          So I really loved that.

          The next one is "Radical Help," by Hilary Cottam, who is a Brit,
          looking at the social safety net, the work force training
          system, health care system. And she takes that concept I have in
          the book of user needs over government needs and just pushes
          them so much further in a way that is really inspiring and I
          think very critical of all of these government systems that we
          have today.

          I mean, it's in Britain, but it's very similar to the U.S. And I
          just think she moves the Overton window about what it would mean
          to truly meet user needs in government-funded systems. And I
          hope everybody who works in social services and health care
          would read it.

          The last one is a bit controversial, and it's not one I'm
          recommending in its content. But I do think people of all
          political stripes should read --

   ezra klein
          Retreats are not endorsements here.

   jennifer pahlka
          Exactly. Definitely not an endorsement. But chapter 3 of this
          new report called, "Mandate for Leadership: The Conservative
          Promise," is essentially the blueprint for what the Heritage
          Foundation thinks the next Republican president should do in all
          areas. But chapter 3 covers civil service rules. It's kind of
          hidden in there, but it does say that they would bring back
          Schedule F, which we can debate, but I think is a bad idea.

          But it's ultimately the way I want to call attention to -- they
          are paying a lot of attention to how to reform the civil
          service. And I would like there to be counterproposals on the
          table since theirs is so thorough and detailed.

   ezra klein
          Jen Pahlka, your book is called "Recoding America." I really,
          really recommend it strongly to people. They really should read
          it. We did not cover a lot that is important in it. Thank you
          very much.

   jennifer pahlka
          Thank you, Ezra. This was great.

          [MUSIC PLAYING]

   ezra klein
          This episode was produced by Emefa Agawu. Fact-checking by
          Michelle Harris and Kate Sinclair. Mixing by Jeff Geld. The
          show's production team is Annie Galvin, Jeff Geld, Roge Karma
          and Kristin Lin. Original music by Isaac Jones. Audience
          strategy by Shannon Busta. The executive producer of New York
          Times Opinion Audio is Annie-Rose Strasser. And special thanks
          to Sonia Herrero and Kristina Samulewski. [MUSIC PLAYING]

   [7]The Ezra Klein Show logo The Ezra Klein Show

The Book I Wish Every Policymaker Would Read

In "Recoding America," Jennifer Pahlka offers a deeper theory for why the
U.S. government struggles to provide services to its people.

   (BUTTON)
   Transcript
   (BUTTON)

   transcript
   (BUTTON)
   Back to The Ezra Klein Show
   (BUTTON)
   0:00/1:15:30
   -0:00

   transcript

The Book I Wish Every Policymaker Would Read

In "Recoding America," Jennifer Pahlka offers a deeper theory for why the
U.S. government struggles to provide services to its people.

   Tuesday, June 6th, 2023

          [MUSIC PLAYING]

   ezra klein
          In 2009, Jen Pahlka founded Code for America. And the idea
          behind Code for America was to get the best technologists
          working on the public's problems, not just private websites and
          e-commerce platforms and social networks. Even then it was clear
          that public digital infrastructure was going to be really
          important and it was lagging far behind the private sector. And
          Code for America was built to bridge that gap.

          Then, in 2013, Pahlka became Deputy Chief Technology Officer in
          the Obama administration. She was very deeply involved in the
          effort to rescue healthcare.gov and helping create the United
          States Digital Service. Then she went back to Code for America,
          and then, in 2020, was tapped by California Governor Gavin
          Newsom to help untangle the mess of California's unemployment
          insurance program as it buckled under the weight of the Covid
          response.

          I mention all that because Pahlka has worked directly on -- and
          then helped oversee and advise teams working on -- the digital
          delivery of government services really at every level of
          government across much of the U.S. So she's someone who badly
          wants government to do good in the lives of people and has spent
          years now confronting the reasons it so often falls so short of
          its goals.

          One of her big points, one I've come to appreciate a lot more in
          recent years, is that even liberals who care a lot about
          government don't care enough or track closely enough how
          implementation actually happens. Delivery often happens out of
          sight, except for the people who need that delivery.

          In our media -- I mean, we're a part of this. I'm part of this
          problem, too. There's a ton of focus on politics, on elections,
          on big policy questions and fights and theories. But then the
          bill passes and the nitty-gritty of how that policy actually
          shows up in people's lives is left up to someone somewhere. And
          when it doesn't show up in people's lives or even makes people's
          lives worse because of how it is implemented, there's often no
          outcry because there's no attention, and so there are no fixes.

          Now Pahlka has written this book, "Recoding America." And I want
          to say it clearly. This is one of the best policy books I've
          read. It's a book I hope future governments and current
          governments will absorb. We get at a lot of it in this
          conversation, but not enough.

          But one thing she really does is tell clearly the stories and
          tell clearly the lessons she has seen and learned about why
          things go wrong even when the people involved are trying to make
          them go right. So people who care about these issues really
          should pick it up and read it in full. But this conversation is
          a really rich place, I think, to start.

          As always, my email [email protected].

          [MUSIC PLAYING]

          Jen Pahlka, welcome to the show.

   jennifer pahlka
          Thanks for having me.

   ezra klein
          So I want to begin with a quote that feels in a way like the
          organizing thesis of the book, which is, "When systems or
          organizations don't work the way you think they should, it is
          generally not because the people in them are stupid or evil. It
          is because they are operating according to structures and
          incentives that aren't obvious from the outside."

          And I love that, because, for me, it's a bit of a credo in
          reporting on government. But you've been inside government. So
          tell me a bit about how you came to that view.

   jennifer pahlka
          I've seen so many people frustrated with government services,
          and they make a lot of assumptions about how that government
          service got to be the way it is. One example I spent a lot of
          time on was SNAP in California -- Supplemental Nutrition
          Assistance Program -- where the application form was over 212
          questions.

          It took a long time to fill out. And so people naturally assume
          that this is done on purpose, that the people making the form
          must not want people to get SNAP. And, for example, there,
          that's not really what was going on. California is a very
          pro-welfare state. There's a huge number of incentives for local
          communities to sign people up for SNAP. It's great for their
          economy. It's great for getting people out of emergency rooms,
          for instance.

          But there are just so many stakeholders in this that they all
          got to pile on their questions, and you ended up with something
          that was really, really burdensome. And it's hard to see those
          incentives from the outside. That's just one example of them.
          But our public servants are incented to do things that, if you
          look at them rationally, don't make sense.

   ezra klein
          You have a great line in that section, which I'll probably get
          wrong in the paraphrase. But you say something like, people
          often think it was designed this way, when the truth is it
          wasn't designed at all.

   jennifer pahlka
          Yeah. I think that's true of so many things that we think are
          systems. I also think it's true of the policy. Take unemployment
          insurance, for example. We've had layers and layers of changes
          on it over time.

          Those changes come from federal government. They come from state
          government. They come from executive, legislative and judicial
          branches. And so what you end up with, 90 years later almost, is
          something that's just sort of a mess of changes, but not
          something thoughtfully crafted by policymakers in D.C. that is
          now operating in some rational way at the state level.

   ezra klein
          So I thought a lot about whether or not to do this, because
          there is probably a best practice of podcasting where you start
          in the examples that are going to be attainable and people
          already know. And we have a bunch of those -- healthcare.gov and
          California unemployment insurance. But I want to start actually
          with a weedsier one, which is: Tell me about the Enterprise
          Services Bus.

   jennifer pahlka
          [LAUGHS] OK, you want to hear about the E.S.B.

   ezra klein
          Yeah, because I think it gets at this point of systems where
          clearly nobody was trying to do something wrong here -- and yet.

   jennifer pahlka
          Yeah, I think everyone was trying very hard to do it right. So a
          friend of mine, Matthew Weaver, was assigned to go help a team
          at Raytheon who had a contract with the Air Force to do,
          essentially, the next generation of software for satellites that
          enable Global Positioning System, G.P.S. So something that's
          pretty important to all of us, especially people like me who
          need it all the time.

          And when he got there, he found that one of the big things that
          was hanging them up was that they had to get the data from the
          satellites to the ground stations. And there's a very industry
          standard way of doing this. It's the obvious way you would do
          it. But instead, the team had inserted this thing called an
          Enterprise Service Bus, or E.S.B, which technical people will
          know was something that was really popular in the `80s and `90s.
          But it's a big, clunky piece of software.

          And it was kind of creating this Rube Goldberg machine that made
          the data go from the normal, the easy protocol into this crazy
          mess of code and then back out again. And it would time out. You
          couldn't get the data from one place to the other in time.
          They'd spent months on this and they couldn't figure it out.

          And he's like, why are we using this E.S.B. when this other
          protocol would work fine? It's a requirement in the R.F.P. that
          the Air Force had put out.

   ezra klein
          An R.F.P. being a Request For Proposal. So the R.F.P. is how the
          Air Force ends up with Raytheon building this thing.

   jennifer pahlka
          Exactly.

   ezra klein
          And we're going to talk about the contracting in a bit.

   jennifer pahlka
          Yeah, we'll talk about the contracting. So Raytheon can't take
          it out because they say, no, the Air Force is requiring us to do
          this. Why is the Air Force requiring it? Because the Department
          of Defense requires it. Why does the Department of Defense
          require it? Because there's something called the Federal
          Enterprise Architecture that requires it.

          The story of the Federal Enterprise Architecture goes back to
          this Act called Clinger Cohen, that was I think 1996, where
          Congress said we need to get a better handle on digital. We're
          not doing technology well in government. Let's make every agency
          create a plan. And then let's have those plans be coordinated.
          And the Federal Enterprise Architecture was the attempt to
          coordinate those plans.

          So one of the things they put in there was that things need to
          be interoperable. And they offered in there, as an example, this
          very common type of software at the time -- Enterprise Service
          Buses -- as an example of how you might get things to be
          interoperable. They weren't actually a great idea at that time.
          They were already kind of becoming sort of obsolete.
          Essentially, Application Programming Interfaces -- APIs -- that
          run the internet now took them over.

          The problem really wasn't that they were specifying something
          outdated. It's that because it was in this document as a
          suggested way of doing what the Federal Enterprise Architecture
          said should happen, when it got translated down to the
          Department of Defense Enterprise Architecture and the Air Force
          Enterprise Architecture, each layer in that hierarchy
          interpreted it more and more rigidly, until at the Air Force
          Enterprise Architecture level, it's assumed to be required in
          all software forever.

          And so the team is saying, I can't take this out. The law
          requires it. In fact, the law does not require an Enterprise
          Service Bus. But the people who have made all of the documents
          that govern the creation of the software for these satellites
          believe that it is.

   ezra klein
          When I was reading that story, it reminded me of a piece I just
          spent a couple of months working on about what I called
          "everything-bagel liberalism." And that was a piece that was in
          part about affordable housing and part about semiconductor
          contracting. And it was a weird piece of work, to be honest, in
          its structure.

          But when I was reporting on the, basically, request for proposal
          the Biden administration put out for its semiconductor grants,
          one of the things that came up, as I was talking to members of
          the administration about all of the additional layers of
          requirements and asks, and having a child-care center and this
          and that ended up in the bill, I would hear about people say,
          look, we think those things are good. You don't realize, we
          actually fought a bunch of these battles and knocked a bunch of
          stuff out that otherwise would have been there.

          And then I would read the text, and I would see the thing they
          told me was knocked out in the text, but it was a suggestion. It
          was a we're not requiring this, but one good example of how you
          might achieve this goal we are requiring is this. And it was
          really striking to me that people were telling me, on
          background, that they had fought and won this fight to not have
          X in the notice of funding opportunity. And then I'd read the
          notice of funding opportunity and X was there.

          And it was true that it wasn't required, but it seemed very
          likely that the semiconductor firms were going to think it was
          required or at least think -- and this was a plain reading to
          me, too -- that they would have a better chance of winning the
          grant if they put it in there. And so there is this way in which
          even when people seem to think they've created flexibility in
          the actual language, once the lawyers get hold of it and the
          contractors get hold of it, it's really not that flexible.

   jennifer pahlka
          There's so much better safe than sorry thinking going on, and I
          would say with contractors and with people who do the
          compliance. Those bids that those companies give back, they know
          they'll be judged by people who are really just checking boxes.
          Does this meet this? Does this meet this? And they're going to
          check that box.

          I think it's a good example of the disjuncture between the
          thinking of policymakers and the reality of how that policy
          plays out. Another good example of it is FISMA -- it's the
          Federal Information Security Management Act.

   ezra klein
          Yeah, which I only know because I have your quote on this
          sitting here in my next paragraph of notes.

   jennifer pahlka
          I've gotten pretty good at remembering what all the acronyms
          stand for. But there are a lot of them. It can be hard. So FISMA
          is a law that covers cybersecurity. And it has these 300
          controls in it. So 300 different things that you, according to
          law, could do to secure an application.

   ezra klein
          So it's like a menu, not 300 mandates.

   jennifer pahlka
          It's a menu. But in practice, if you ask pretty much anybody who
          does software in government, they have to do all 300. And it's
          because I think it's evolved such that in order for people to
          cover their butts, they have to explain why you didn't do
          numbers 138 through 192 or something. And it's really hard to do
          that. In fact, it's sometimes easier just to do the controls
          than it is to explain them.

          But the people in the middle who are doing the compliance have
          no idea if, for your application, numbers 35 and 95 are the two
          really, really important ones, and you should spend the rest of
          your time on the thing that's most important to security, which
          is testing. All they know is that they will be responsible if
          anything bad happens. And if it was control number 96 that got
          you, it's their butt on the line. So they will insist on all of
          them. And if you go talk to the people who wrote FISMA, they
          will say, we never intended that.

   ezra klein
          Well, this gets at a, I think, a really important point you make
          about that incentive system -- how it is actually designed
          versus how it might look from the outside. You write that public
          servants are trapped between two distinct systems of
          accountability. So what are the two?

   jennifer pahlka
          Well, when things go wrong, we're very upset with public
          servants. I mean, healthcare.gov is a famous example. There were
          like 10 hearings in one month about this. And they were
          televised. And we're upset because the thing doesn't work.
          That's an outcome. But what public servants are actually hired
          for, promoted for, what their careers depend on is fidelity to
          process. It's a compliance framework.

          And so, yeah, if you didn't do number 96 and something goes
          wrong, that's what the administrative state will focus on. The
          people who do oversight focus on whether you checked all the
          boxes not whether it works. And in fact, it's quite true, in
          people's careers, that they are not punished for something not
          working as long as they can say, I did what the procedure
          prescribed.

   ezra klein
          Something this gets at is that we have civil service
          protections. That politicians are mad at the outcome of
          something civil servants did. It might allow them to embarrass
          the agency. They might even cut funding for the agency in the
          future, although often they don't. But they can't fire anybody.

   jennifer pahlka
          No.

   ezra klein
          They can't really create that many consequences for the agency.
          Whereas, failing to follow process, you really can get fired.
          You really can lose your promotion track. There really are
          consequences. And so in a weird way, the people who made the
          policy in the first place, and, in theory, are exposed to the
          voters who are supposed to be the judges of the policy, actually
          do not have power really to hold the civil servants accountable.

          And there are good reasons for that. I mean, there are good
          reasons for civil service protections. But there is also
          consequences to it, which I think, frankly, on the left, are
          pretty underplayed.

   jennifer pahlka
          Yeah, I mean, I heard so many times, when I worked in
          government, you could get fired for that. In fact, you hear
          frequently, you could go to jail for that. It doesn't really
          match the reality. People don't get fired that much. And they
          certainly don't go to jail.

          We used to have a joke about, really, I'm going to go to
          procurement jail? Where exactly is procurement jail? Right? It's
          like such a nerdy piece of law and policy that certain people
          understand, but there's this perception that a small violation
          of it is going to end you in jail. It doesn't really match the
          reality, but the culture has definitely evolved to highlight any
          deviance from process and make it very, very risky.

          It can be risky on just a practical level. You might not get
          demoted. You might not lose your job. But suddenly people are
          not wanting to work with you. [LAUGHS] There can be cultural
          consequences to it and just your ability to get the job done.

   ezra klein
          Tell me about the Paperwork Reduction Act.

   jennifer pahlka
          So this is a great example of a law that sounds great. And it
          sounds like something that's really important to me because it's
          supposed to reduce the paperwork burden on the American public.
          And I think the paperwork burden is a real problem, in part,
          because, as I say in the book, paperwork favors the powerful.
          The more paperwork that's required, people with resources will
          get through it and people without them will not.

          So this law said, essentially, something we all believe. If you
          measure it, you'll be able to manage it. So I always call it the
          comically misnamed Paperwork Reduction Act because it's designed
          to measure the burden of any, say, form that you have to fill
          out. Like your tax form, for instance, the 1040. That has an
          O.M.B. number, which is required by the Paperwork Reduction Act.

          And it means that the Office of Information and Regulatory
          Affairs, known as OIRA, reviewed the time it would take to fill
          out this tax form and stamped its approval. And so it's got this
          measurement on it.

          The problem is that it got interpreted very broadly to say any
          time you speak to a member of the public about a form like this,
          that has to get OIRA approval.

          So in the tech world and consumer product world, we have
          something that's really important to the usability of our
          services called user research. The way we make a form easier,
          like that tax form, is that we watch people filling it out, and
          when they get stuck at a question -- like, what were your last
          year's state taxes -- we go, OK, we've got to make that easier
          on that person and we redesign it.

          So we have this very important discipline called user research.
          That is how we streamline forms and applications and make them a
          lot easier to use. So it involves something like watching you
          fill out that tax form and seeing where you get stuck, seeing
          where you have questions. Asking questions like, does this
          question actually need to be on this form? How do we make it
          easier for the user?

          But when you assume that every single interaction with a member
          of the public needs to go through OIRA, then your user research
          needs to go through an OIRA approval process. And because that
          process can easily take nine months and sometimes longer, you
          essentially can't do user research.

   ezra klein
          But this is weird because -- you point this out in the book --
          it would seem, from a plain reading of the law, that user
          research is exempted.

   jennifer pahlka
          Yes.

   ezra klein
          And yet --

   jennifer pahlka
          And yet, there are examples of this going on every day in
          federal government now. There's also examples of people getting
          through it and not being required to get OIRA approval for user
          research. But essentially, what happens is -- in fact, a friend
          of mine whom you know, Erie Meyer, who's at the Consumer
          Financial Protection Bureau, had this happen to her the other
          day. She was already doing user research. And they said, oh, no,
          we're going to have to send this through OIRA for approval. I
          mean, they were already doing it, which is the crazy thing. And
          essentially, the person in the agency decides that better is
          safe than sorry. It's better to be safe than sorry. And says, I
          understand your interpretation, but it's my butt on the line if
          someone else calls a foul on this, so we're sending it up to
          OIRA for their review.

          Now, OIRA could say, obviously, this is wrong and silly and
          there's no need to get our approval. But they also take a better
          safe than sorry stance, again, because of the incentives to
          them. And the result of Erie's request was that they said this
          will have to go up on the Federal Register, and we will have to
          get comments from the public about whether these people can get
          comments from the public. Oh, and in the meantime, there'll be
          no user research or no comments from the public happening until
          this process is over.

   ezra klein
          I just want to note that everybody in this story is a liberal
          who wants government to work well.

   jennifer pahlka
          Yeah.

   ezra klein
          I just think that's important to say here because it's something
          you say towards the end of your book, that working in government
          is a really valuable thing to do, in part because it will change
          your view on what sometimes makes government work poorly.

          And one of the almost spiritually difficult things about reading
          your book, if you're somebody who does care about government, is
          knowing that this is a lot of people who are doing their best,
          who have, in many cases, much more difficult jobs that require
          much more endurance and creativity to serve the public than
          anybody in the private sector does. And yet, we're getting this
          outcome somehow, right?

          I've known a lot of people at the top of government, the people
          who work in offices in the West Wing. And they're all very, in
          some conceptual way, aware that government is inefficient. They
          don't want it to be. Bill Clinton had Al Gore do reinventing
          government. Barack Obama hired the first C.I.O. to look over the
          government -- Chief Information Officer. Joe Biden is somebody
          who when he was a Senator often talked about government
          inefficiency. Their chiefs of staff are impatient
          management-oriented people.

   jennifer pahlka
          Yeah.

   ezra klein
          And yet, somehow from the top, even though they talk about this
          all the time on the campaign trail -- no Democrat wants to be a
          pro-inefficient government Democrat -- this does not filter down
          somehow. Why?

   jennifer pahlka
          I think you have to look not just at the executive branch in
          administrative agencies, though you shouldn't not look at them.
          I mean, I think it essentially comes down to you come in with a
          passion for cutting the red tape and making this easier. And two
          things happen. You get worn down and you have other priorities.

          It's really, really hard to spend all your time on this. And
          you're not in control of it.

          When I worked in the White House trying to stand up what became
          the United States Digital Service, we were originally going to
          do it in G.S.A.

   ezra klein
          Which is the --

   jennifer pahlka
          General Services Administration. And the head of the GSA is a
          fantastic public servant named Dan Tangherlini. And he really,
          really wanted this unit in his agency. And he was actually
          blocked by a guy below him who kept calling out that better safe
          than sorry on some very obscure interpretation of how the money
          that we were going to use to stand up the unit could be used
          legally.

          I mean, it was wrong. But he could construe the language that
          way. So you have to actually have some empathy for people at the
          top who are fighting as hard as they can to get stuff done and
          are blocked by people not just above them, but around them and
          below them.

   ezra klein
          There's another dimension here that you point out. And it's a
          sociological observation about government that as soon as I read
          it, I realized is completely true. That knowing a lot about
          politics is very high status in government. If you're considered
          a tactical political genius, that's great. Knowing a lot about
          policy, having very big policy ideas, that's also very high
          status. Implementation isn't.

   jennifer pahlka
          No.

   ezra klein
          Why?

   jennifer pahlka
          I think it goes back very deep in our culture. I mean, I think
          that's not just true in government. Let's be honest, right? Big
          ideas people get a lot of status in our society.

   ezra klein
          Thank you for listening to my Ted Talk.

   jennifer pahlka
          Well, yes, big ideas guy. [LAUGHS]

          A friend of mine who worked in the government digital service in
          the U.K. pointed me to this idea that's written about the
          British civil service. That it is literally divided between the
          intellectuals and the mechanicals.

          Now, I think that the digital age has complicated this in a lot
          of ways. If you look at, say, metaphysical Silicon Valley, if
          coding is a mechanical task, it's the way you implement
          something, all those companies were founded by coders.
          Mechanicals in that framework are at the top.

          And lots of interesting -- good and bad -- things have come out
          of that culture, which is really distinct from the culture of
          government where the intellectuals are at the top and have the
          power. And the people who write the code or do the designs that
          implement these programs are really, really, really far down.
          They're at the very bottom.

   ezra klein
          In a way, they're not even at the bottom. They're outside.

   jennifer pahlka
          They're outside. This is true.

   ezra klein
          Tell me about Circular A-76.

   jennifer pahlka
          So a couple of decades ago, the federal government said -- this
          is actually coming out of O.M.B., the Office of Management and
          Budget, but it really derives from a long history of trying to
          streamline government -- said, if someone else can do your job
          in government, we're going to outsource it. It's the reason we
          have things like the concession stands in national parks are run
          by third parties. They're not run by the government. That's all
          A-76.

          And there's a real valid justification for some of this. There's
          a famous case in which, I think during World War II, government
          decided to make its own steel because there wasn't enough. Well,
          by the time they finished that, it was wildly expensive and very
          bad. And it's a commodity. The market can provide us good steel.
          That is the best way to do that. We should not get in the
          business of steel.

          And so on that basis, they roped a lot of other things into
          that, like concession stands in federal parks. There's no reason
          that's inherently governmental, which is an important phrase
          that very few people talk about. Because the distinction between
          something that is inherently governmental, like a government
          person must make this decision -- i.e. most policy decisions --
          and something that is inherently commercial is like, OK, well,
          we'll make the policy decision and then we'll contract it out to
          somebody who just delivers on it.

          And it creates this real separation between policy and
          implementation, which has a lot of complications to it. And I
          think over time has snowballed into some really bad outcomes,
          because we're trying to do policy over here, implementation over
          here -- I have my hands very far apart since your listeners
          can't see -- and so much gets lost in the middle.

   ezra klein
          So that's, in a way, the guiding principle here. You say it's
          gone now really far. I mean, you go quoting someone else as far
          as really saying that civil servants manage; they don't
          implement. So particularly in the digital sector, tell me about
          what the government does and what it contracts out. What
          functions in your experience are done by somebody on a public
          payroll? And what functions are done by somebody who is being
          paid out of public taxes but is on a private payroll?

   jennifer pahlka
          So let me speak to what's still pretty standard but changing,
          like when I was in government in 2013 and 2014, was really,
          really standard. If you're trying to get, say, healthcare.gov
          built, and you're the digital person responsible for it, what
          you will actually do is work with contracting officers, people
          to gather a bunch of requirements, lawyers, and you will have to
          be master of an incredibly complex apparatus of procurement
          rules, and operations, and forms that need to be filled out, and
          processes that need to be followed.

          That actually is a very robust skill set and knowledge base in
          its own in order to hire people who will actually do the digital
          implementation of whatever it is you're doing. So I talk about
          somebody in the book who's, in fact, a fantastic digital leader
          and gets really good outcomes because he is a good digital
          leader. But like 90 percent of his time is spent on procurement
          and legal.

   ezra klein
          So I want to talk about how procurement actually works. Because
          just like it is different to hire and fire in the public sector
          because of civil service rules, it is different to procure, to
          pay an outside vendor to do something in the public sector.
          Because we have a lot of rules meant to clamp down on cronyism,
          on patronage. We're trying to make the process fair, and we make
          it cumbersome.

          One of the things you talk about a bit is the way the system can
          get slowed down by vendor protest. That if you don't win a
          contract, you can actually protest the entire process, holding
          it up. Sort of like if you are upset about people building
          something, you can bring an Environmental Quality Act objection.
          Tell me about these protests. Tell me about the ways in which
          the vendors actually have quite a lot of power in the process.

   jennifer pahlka
          Yeah, these rules were designed to make the process fair. So for
          instance, an agency will announce a winning bid. And if other
          companies who bid on this think that either something genuinely
          went wrong or more likely that they can simply take some
          advantage out of this, they can protest the bid.

          And because these bids are so wildly complex -- I mean, the
          paperwork required to bid on a major government project, whether
          it's technology or anything else, is massive. There's a lot to
          review. And there's very little evidence of actual competency
          that are required in this paperwork. It's all evidence of
          compliance.

          And so it's very easy to find something that's not compliant.
          But the result is easily six months, often year or two, delays
          in getting something done.

          [MUSIC PLAYING]

   ezra klein
          I think, actually, the first time we started talking, I became
          more aware of your work, was in the healthcare.gov debacle. And
          at the core of that debacle was procurement and who got chosen
          to build that. So how did these rules end up affecting who got
          that contract?

   jennifer pahlka
          If you look at the timeline in the A.C.A. -- and I'm going to
          forget the exact times -- but essentially, they only had three
          years. The Centers for Medicare and Medicaid Services, C.M.S.,
          had about three years to get healthcare.gov up and running. They
          needed to launch it by Oct. 1, 2013, if I recall.

          And as I just described, the process of getting a bid together,
          that's incredibly complex. Getting all your requirements,
          getting that contract out to bid, reviewing the bids and then
          awarding it and then, potentially, protests, that's easily three
          years right there. So they had to do something else.

          So CMS actually had some vendors on a very obscure vehicle, we
          call it, called I.D.I.Q. It stands for Indefinite Delivery
          Indefinite Quantity, which is essentially a way of saying I've
          hired you to be around and do what I need. [LAUGHS] Actually, an
          I.D.I.Q. can be a very good thing to have someone just
          on-boarded and ready.

   ezra klein
          They're basically on retainer.

   jennifer pahlka
          They're on retainer. But yeah, for a wide variety of things. And
          so C.G.I. was one of the companies that was on an I.D.I.Q.
          already at C.M.S., so they didn't have to bid certain pieces of
          it out. Now, let me be clear, there were 60 different contracts
          to vendors to build healthcare.gov. Some number -- like 12 of
          them -- went to C.G.I., which is considered the main contractor,
          and others went to others. So it's not like they had the whole
          thing.

          But they really couldn't have done that full process where they
          bid it out, because if anybody had protested, they had no hope
          of making their congressionally mandated deadline.

   ezra klein
          I think it's worth spending a minute here. Because I covered the
          drafting of the Affordable Care Act really closely. I covered
          how they were thinking about building the --

   jennifer pahlka
          And you missed this. [LAUGHS]

   ezra klein
          Not only did I miss this, but I can say with certainty that
          there was a lot of excitement in Congress about we're going to
          create these online digital marketplaces and it's going to be a
          next generation website. You're going to be able to compare
          everything.

          And I'm sure in Congress's mind, having talked to people, the
          way they thought about this is the government would go for this
          unbelievably marquee project and get a really great website
          design player. I mean, Google doesn't do a ton of government
          contracting in this particular way. They do other things. But
          someone like a Google, just whoever is the best. And this would
          be a great project.

          And instead, you end up with a pretty obscure contractor -- and
          to me this is important -- who -- I'm not saying this is their
          only competency -- but across a bunch of these contractors, what
          they have is a competency in contracting.

   jennifer pahlka
          Yeah.

   ezra klein
          They know how to work with the government, which is a huge pain.
          And this feels like another place where the procurement rules
          and the way different regulations, et cetera, have been
          interpreted, it's really not what the legislature was trying to
          do. They wanted a great website.

   jennifer pahlka
          Yeah.

   ezra klein
          What they got was a total disaster. But they didn't then go back
          and change procurement rules either, really, to my knowledge.

   jennifer pahlka
          Well, the legislators would say, we have given agencies -- and
          I'm not sure about this particular case at that particular time
          -- but this happens in the Department of Defense all the time.
          We have given you lots of exceptions that you can use. There's
          something called an Other Transaction Authority, which basically
          gets rid of a lot of those rules and lets you choose who you
          want to choose in a streamlined kind of way.

          But they just don't get used. And that, again, goes back to
          people say better safe than sorry.

          If something goes wrong, someone is going to come back and say,
          what justification did you have for using O.T.A., Other
          Transaction Authority, instead of the more robust way that we
          know will ensure, quote unquote, "fairness"? So it's really that
          culture that comes back to it.

          I mean, it's a reasonable example I think of what I talk about
          in the book is culture eats policy, right? The policy that is
          designed by policymakers wants to get the outcome of flexibility
          for the agencies because they know that the rigidity is hurting
          the outcomes. But the culture just goes back to, I see that
          you've given us that, I'm choosing not to take it because I feel
          like it's going to make me safer.

   ezra klein
          So healthcare.gov was this huge project. It had to be contracted
          from top to bottom. But I also want to talk about the ways in
          which this kind of contracting makes it hard to update systems
          as policy is changing. And I want to use a pretty horrific
          example that you talk about in the book, which is the child
          separation policy at the border.

          So obviously, the actual policy there was terrible. But one
          thing that made it much worse was the rigidity of the system
          being used. Can you go through that?

   jennifer pahlka
          When the Trump administration ordered that kids and parents be
          separated at the border, the systems that tracked families could
          only assign one what's called A number, an Alien number, to a
          family. That's how it was set up and had worked for years.

   ezra klein
          So just horrible. It's called an alien number.

   jennifer pahlka
          It's called an alien number. Though, they call it A number,
          which is also really confusing. So if a family of four comes
          through, they get one A number. Because in the past, they were
          treated as a unit. This horrific and cruel policy comes down.
          They separate them.

          They now give you one A number, your partner one A number, and
          your kids get A numbers separately. And there is no affordance
          for connecting your A number with your children's. They were
          pinning sticky notes with A numbers on infants who had been
          taken from their parents.

          And there were wonderful public servants who were writing down,
          keeping spreadsheets, trying to track the connections between
          parents and kids. It wasn't like there was no effort made. But
          there wasn't a systematic way to do it. There was a lot of
          chaos. The person who's trying to reunite them is not
          necessarily the person who separated them in the first place.

          And so when, thankfully, that order was reversed, there was very
          little capacity within the Customs and Border patrol to fulfill
          the demand to reverse it and reunite these kids because the
          software didn't have any affordance for making this connection.

   ezra klein
          So something you then talk about is it would have been a fairly
          obvious move -- again, without either of us endorsing a terrible
          policy -- but it would've been a fairly obvious move, if you're
          doing that policy, to change the software so that the A numbers
          are tracked together. Why didn't they? What would that have
          taken?

   jennifer pahlka
          So I should caveat by saying, I don't know the specifics of the
          C.B.P. contracts that govern this software. And I've never seen
          this software in action. But I do know that the way this
          generally works in government is if you want an update to your
          software, you take it out to bid. And that takes a really long
          time. And yes, the company that may have originally written it
          will bid on it, but you're not guaranteed to give it to them.
          Maybe that's the right thing. Maybe that's the wrong thing. But
          if you have a new vendor, they have to get up to speed on the
          code base.

          But essentially, the way we think about software in government,
          it's a project. You start it. You finish it. And then there's a
          little bit of money trailing around for you to keep it up and
          running. So when you have a change like this, that's frankly
          incredibly urgent, you don't have people who work on that
          software day to day, week to week that are there ready to change
          it and know how the code base works.

          So in contrast, Google didn't create Search and then lay off 95
          percent of the staff and keep just a skeleton crew around it to
          keep it up. The way things that really work in the real world
          are staffed is on an ongoing basis, not only to make sure that
          changes in some connecting database somewhere can be updated so
          the thing doesn't crash, but also because we live in a dynamic
          world. We can't expect there not to be policy changes that have
          to be incorporated.

          And frankly, we should be constantly looking at the software and
          saying, is it doing what it could be doing as well as it could?
          There's constant improvements in the consumer products that we
          use. So the curve of staffing in a consumer product is it starts
          really slow, and then as the product becomes more and more
          clearly valuable and popular, the staffing goes up and up and
          up. It's just this constant curve up. That's how the software is
          constantly better and more usable and more valuable to us.

          In government, it's this weird, like, there's nothing, then you
          hire all these people, and then there's this couple-year period
          where lots of people are working on it, and then it drops to
          almost zero. So after almost zero, you have no capability to
          change that software. You have acquired software, but you have
          not acquired the capability of doing what that software is
          supposed to do, of performing the function on an ongoing basis
          that that agency is responsible for.

   ezra klein
          I think this brings us back to healthcare.gov, which is a real,
          I think, before and after moment for how digital is viewed
          inside the government, because it is built and broken by this
          contractor or set of contractors. And the rescue effort is a
          much more internal process that is built on now the idea that
          you actually need internal capacity and flexibility, and because
          of the president is focused. There's a lot more ability to move
          quickly and make decisions. So tell me a bit about the
          principles and staffing that the rescue effort is built on.

   jennifer pahlka
          The rescue effort came at a time when we were trying to prove
          the point that you need an internal capacity. So as disastrous
          and horrible as it was, it was actually really useful for me and
          others who were saying, hey, we can't just have procurement
          competence in government. We have to have some digital
          competence in government. And in fact, we need it in the halls
          of power in the White House and in O.M.B.

          And I think one of the reasons it worked is that they could see,
          through the rescue effort, that we could bring in people whose
          competence wasn't how do we contract with company, but how does
          the software actually work and what's wrong with it. So the
          people who came in were software engineers. Many of them happen
          to have this title at Google and other companies called Site
          Reliability Engineers, S.R.E.s. You've never heard of an S.R.E.
          in government. There literally never was one. That was always
          the responsibility of an outside company.

          So you had people who could actually look at the code. And in
          fact, we found that oftentimes the people at the vendor who were
          supposed to have done the coding didn't really even know how to
          code. So you had all these people who knew how to work with
          government, but didn't actually know how to make the site work.
          And bringing in folks who could actually get in there in the
          weeds was the thing that changed it.

          The process up to that point had relied a lot on planning.
          There's this real focus on planning in government. For lots of
          jobs, if you want to get promoted, you have to have this thing
          called a P.M.P., a Project Management Professional. And it's all
          about upfront planning. But it's really hard to plan in a
          situation like healthcare.gov, where they are still finalizing
          the regulations and rules that govern it right up to the launch
          of the site.

          So the skill that was needed was not this very traditional, plan
          everything, then execute to it. You needed people who were
          basically good at what's going on with users, is it working,
          just go find problems and solve problems at a very detailed,
          technical level, instead of looking at project plans that were
          not actually connected to the experience that users have.

   ezra klein
          Well, one thing you say about the healthcare.gov experience is
          that there were a lot of reasons it went wrong. And there are
          G.A.O. reports on it and --

   jennifer pahlka
          Very good ones.

   ezra klein
          I wrote a piece for "Businessweek" about it back at that time.
          And I mean, there was a huge amount of attention on this. But
          you say if you could boil it down to one, it's that there was a
          lot of process management, program management, but not a lot of
          product management, not a lot of people willing to say, no,
          we're not going to do that, empowered to say, no, we're not
          going to do that. Can you talk about that distinction?

   jennifer pahlka
          So project management is the art of getting things done. It is
          hugely important. I value good project managers very highly and
          everyone should. Product management is the art of deciding what
          to do in the first place. And if we don't do that, then the
          project managers just are overwhelmed with a undifferentiated,
          unprioritized mess of requirements that they have to fulfill.

          And yes, healthcare.gov, there was a lot written in the law
          about what needed to happen. It was probably overspecified. But
          there was also not a willingness to even imagine that you might
          do less. In fact, people were told you can't do less. You can't
          not serve every single edge case from day one, for instance.
          That's illegal. There's reasons why they think it's illegal. I
          do not think it's illegal, and I think we've proven since then
          that it's not.

          But that willingness to say it will launch serving the most
          common set of users, because that way we can handle the people
          who can just use a website to get their health insurance. And
          then the people who have crazy exceptions or rarer exceptions
          are going to go through the call center, they're going to go
          through those in-person service centers that they set up around
          the country, which makes sense. Those people actually probably
          need a person walking them through it because of their
          exceptions.

          It doesn't mean that the website couldn't eventually serve all
          of those people. It means that to get a site that launches and
          works on day one, you constrain what you're trying to do. Get
          those first people through the door. Test the site that way. And
          then you add the functionality later. That's good product
          management. And until recently, it literally hasn't existed as a
          discipline within government.

   ezra klein
          There's both product management as a discipline, and I think
          there's also product management or, in this case, I would call
          prioritization, as a political value, as something the
          leadership values. And you tell a story in there of a conference
          call while the site is in crisis with Marilyn Tavenner, who is
          head of C.M.S., the Centers for Medicare and Medicaid Services.
          Could you tell that?

   jennifer pahlka
          Mina Hsiang was somebody who joined the what we'll call rescue
          effort, though I do try to stay away from that term because
          there are so many great public servants who were part of the
          rescue effort and they often feel unseen by that. But Mina
          Hsiang came in to help.

          And one of her responsibilities was to staff a daily call with
          the administrator, Marilyn Tavenner, and others, including the
          main contractors. And the site was in total disarray. I mean, I
          don't remember exactly how long this was in. But this was before
          we were headed towards, OK, it's starting to work again.

          And one of the things that she brought up on this call was that
          in the original spec, there was going to be a separate site in
          Spanish. Makes sense. I think government services should
          absolutely be in as many languages as we can to serve people as
          well as we can. But at that point, even the English language
          site was very far from being able to work.

          So Tavenner brings this up on the call and says, where's my
          Spanish site? Now, there's all these vendors on the call, so
          Mina knows that they're going to try to placate her and say
          we're going to get to that and then divert resources to the
          Spanish site. So she gets ahead of them and just jumps in and
          says, I'm sorry, Administrator, but you can either have one site
          that works or two that don't. And Tavenner hangs up on her. She
          doesn't want to hear it.

          And God love Mina, she just kept persisting. But her job at that
          point was to keep the team's eye on the ball, which is we have
          to get this site to the point where it's operational and then we
          can use those resources to make a Spanish version. But it would
          be pointless to make a Spanish version of something that didn't
          work. [LAUGHS]

   ezra klein
          One thing that is a worrying dimension of that story is that --
          and you see this in government a lot -- that when things are in
          crisis, people realize they have to bend rules. They have to
          prioritize. They can't do everything at once. They have to make
          the first thing work first.

          But a lot of things aren't in crisis or -- and I think this is
          actually more often the case -- they are, but they're not in
          crisis for the right people, which is to say I think there's a
          lot of attention to delivery and implementation when the people
          being delivered to and implemented upon are middle class, are
          upper middle class, are rich. So healthcare.gov was affecting so
          many people simultaneously.

          If you think about the unemployment insurance failures during
          the expanded unemployment insurance during Covid, that was
          affecting a lot of middle class people who had political power
          all at once. There are obviously a lot of programs where they
          work very poorly for poor people in an ongoing way, but they're
          not a crisis.

          And so the ability to say to the administrator or to the
          secretary or to the president or whomever, if they even care
          about this, hey, we're just not going to do that because this
          crisis is so big, we have to do the first thing first, isn't
          there. And so I worry a bit about that. I'm curious to hear your
          reflections on how things work in what you might call
          implementation peacetime versus implementation wartime.

   jennifer pahlka
          Yeah, we don't get as much done in peacetime. And I think one of
          the tactics of people who I respect in this community is to make
          it a crisis, to make the crisis visible. I'd argue that it was a
          crisis that very few people were getting enrolled in SNAP. But
          it was not hitting the headlines for sure for all the reasons
          that you mentioned. And some great people at Code for America
          made that a lot more visible and change got to happen.

          I talk in the book about what was certainly a crisis at the
          Department of Veterans Affairs where veterans couldn't sign up
          for health care, but it was completely invisible to the people
          inside the building because the form that was broken, if you
          tried to use it from any computer at a library or at your home,
          wasn't set up with the same combination of Internet Explorer and
          Adobe Reader that the computers inside the building were set up
          with. And so the form worked fine for them.

          And the tactic that the team used to finally get this fixed was
          to record a veteran trying to use the service and show the
          people in power how humiliating and frustrating the experience
          was for veterans. But until they were able to bring this veteran
          named Dominick's experience to life inside the building at the
          Department of Veterans Affairs, they were literally told there
          was no work to be done because the requirements in the contract
          had all been fulfilled and all the boxes had been checked.

          So you've got to sometimes make the crisis visible when it just
          won't be because the people who it affects don't have a voice.

          [MUSIC PLAYING]

   ezra klein
          One thing that comes out of the Veterans administration story
          you tell is the way that what a crisis is to the people using
          government services and what a crisis is to the people making
          them can be very different. And I would say throughout the book,
          you're very sympathetic, even to people you're at cross-purposes
          with. You're trying to get something done and they're standing
          in your way. And you tend to see where they're coming from.

          But in this particular story, in the veteran's story, you talk
          about somebody you call Kevin and the attitude that he and his
          team have developed towards development, towards keeping
          themselves out of hot water. Tell me about Kevin.

   jennifer pahlka
          Kevin was also at the Department of Veterans Affairs, but this
          is actually a separate interaction. I had started working with a
          small team to understand problems with the veterans benefit
          management system, which is the internal way that you would
          process people's requests for benefits. We had been told that
          there was terrible latency in the system. Essentially, you can
          click on a link to start processing something and you'd wait for
          a really long time before the page would load. It's very
          frustrating.

          We met with the guy the first day and he said, I'm so glad the
          White House decided to send people to verify that everything is
          OK because latency has been solved. Now, we found out later he
          solved latency by defining it as over two minutes. So if you
          clicked and had to wait two minutes or one minute and 59 seconds
          for that page to load, you were not to report it as latency. So
          that was the solution. [LAUGHS]

          That was my first clue that this man I called Kevin was more
          interested in how things looked than how they actually operated.
          But as I asked him questions about this system -- why had they
          decided to do X instead of Y, what role did the claims
          processors have in this particular part of it -- he kept saying,
          that's not my call, I don't know that. And I asked him why.

          And he said, look, I've spent my entire career training my
          people not to have an opinion on the business requirements. If
          they ask us to build a concrete boat, we'll build a concrete
          boat. And I said, good lord, why? And he said, because that way
          it's not our fault.

          And honestly, that moment for me -- I was seven months into
          working in government then. I'd worked at Code for America
          before. I'd worked with government. I was not new to the
          dysfunctions of government, and I shouldn't have been shocked.
          But I really did take it as a gut punch.

          I didn't blame him, in a certain sense, because I knew what he
          meant about the way we structure responsibility in government.
          But at the time, I think 18 veterans a day were committing
          suicide. Many of them were still waiting for their benefits. And
          it just seemed like such a callous thing to say when, if he
          would push back a little and say this concrete boat is never
          going to float, we could have gotten more veterans their
          benefits.

   ezra klein
          When I finished the book, I had this note to myself that this
          book is about personnel, capacity and autonomy -- that it's
          about who works in government, what kinds of things government
          does versus doesn't do. And then the autonomy, the agency, the
          power we give these people to do the jobs we ask them to do.

   jennifer pahlka
          That's your TED Talk.

   ezra klein
          Yeah, there you go. My TED Talk, a summary of Jen Pahlka's book.
          But one thing that I have always worried about being close to a
          lot of people who work in agencies, and I think that you see in
          that story, is when you make it really hard and annoying to work
          in government, it's not that it pays so little -- you have a
          good section on this. It's not that government pays so little.

          But it is often very frustrating work. You have to deal with all
          these weird procedures and the lawyers tell you you can't do it.
          I mean, working on code in government -- and this is true in a
          lot of your stories, including the stories of your heroes --
          often requires an almost superhuman level of persistence. You're
          trying to get around a lot of very annoying things. And
          sometimes you get a hero out of that.

          But it also, over time, drives out a lot of good people who just
          don't want their days to be annoying. And it also selects for
          people who can make the internal compromise -- in that story,
          Kevin made -- where you sort of shut off the part of you that
          says that doesn't make sense and I'm going to argue -- you lose
          enough fights like that. The lawyers tell you no enough times,
          you get yelled at by your superiors enough times, and either you
          leave or you say, I'm going to stop fighting.

          And that's where things can go, I think, really badly. That it's
          not just that we need good people to work in government. You
          hear that all the time. But we need government to be a good
          place for people to work. And for a lot of creative, impatient
          people, particularly those at mid-levels, not those who are
          appointed to run an agency or run an office in the White House
          or something, it gets very frustrating.

          I've seen a lot of really good people go in, and particularly
          when they're in older agencies that already know how they do
          things, I've watched them go right back out. And I don't think
          they weren't committed. I just think people -- it's hard to have
          a really irritating job.

   jennifer pahlka
          I really think that we're so focused on getting these outcomes
          from a policy sense. And to solve the problem you're talking
          about, we've got to really focus on just basically the overall
          health of the civil service. Yes, it's frustrating to fight
          these battles. I think there's increasingly, certainly in my
          experience, a community of people who support each other around
          it that really makes it tolerable.

          When people stay it's because they've got someone's shoulder to
          cry on when they need it and people to tell them keep going. I
          have found that incredibly valuable and incredibly meaningful to
          me to be part of that. But for instance, one challenge they face
          is they also really can't hire other people quickly. And there's
          a lot that we --

   ezra klein
          Or fire people.

   jennifer pahlka
          Or fire people. We could do so much -- and by we, I mean more
          our elected leaders and how we hold them accountable -- to
          making that environment less destructive, less wearing you down,
          and giving public servants like those who are high profile, who
          are just persistent and heroic, the tools they need to succeed.

          But we just don't think of the tools of legislators or the
          legislative branch as being good at that, right? They have
          money, they have rules, and they have oversight. And those
          things aren't working very well. But if they targeted them at
          the health of the civil service and really focused on it, I
          actually think we could make that a place that doesn't drive
          people away as much.

   ezra klein
          Well, tell me what that would look like.

   jennifer pahlka
          So for example, there are bottlenecks in the system that are
          just vestigial. Like O.P.M. requires that there be industrial
          organizational psychologists that approve a bunch of things, the
          same way that we have OIRA as a bottleneck to getting user
          research out the door. Remove it. But nobody's on that. Nobody's
          looking at these little things like that that, one by one, could
          actually speed hiring and make it easier for people. But there's
          probably 7,000 little things like that.

          And one of my complaints, since you started on the left, is that
          the right actually has a lot of plans around this that are
          focused on firing people. The left really doesn't pay much
          attention to this. They're not obsessed with the details of
          making the civil service have greater autonomy and freedom and
          hiring the right people, and yes, probably firing the right
          people.

          I actually want the right and the left to say this is actually
          more important than this, that, or the other policy outcome that
          we're trying to get because it will enable all of those policy
          outcomes. But it's real work that somebody has to fund and
          somebody has to do. I know probably half a dozen things that
          would go in that, but somebody has to go find the other 36.

   ezra klein
          My impression of this area of politics is that it's become, as
          many things do, very dysfunctional through a kind of crude
          polarization, which is to say that civil service reform, by
          which I mean making it easier to hire and fire, giving the civil
          service more autonomy, but also making sure there's more
          accountability, including accountability from the political
          system itself, because that is often -- parts of that have often
          been a feature of the right wing attacks. And often they are
          attacks that also want to outsource a lot of functions of a
          civil service and starve it of money and so on. They've been
          coded as right. So the way to defend and show you care about
          government is to protect it from these attacks. But if you then
          cover government or you actually know people who work in it,
          even people in the civil service don't like how this works. I
          mean, the difficulty of --

   jennifer pahlka
          They hate it.

   ezra klein
          -- listing a job to get anyone in -- we're so worried about
          patronage. We're so worried about corruption that we've made the
          whole thing work really poorly. And we drive people out because
          they just cannot stand it over time. And many of the people who
          can, we're either asking, again, superhuman work out of them, or
          we're accepting people who maybe are not the people we wish were
          in charge of some of these things. And that's not to take away
          from a lot of people doing hard work.

   jennifer pahlka
          Yeah.

   ezra klein
          But it's also a kind of fallacy if you've ever worked in any
          kind of organization to suggest that everybody in an
          organization is the right person at the right place. I think the
          left has a lot of trouble with this kind of statement, but it is
          important to be able to fire the right people to make
          organizations work. It's not pleasant. It's not what anybody
          should want to do. But to not be able to do it is a way to
          guarantee organizational dysfunction.

   jennifer pahlka
          I agree with that. I think the order of operations, though,
          should be first we make it easier to hire.

   ezra klein
          Yes, no, totally. I think the reason this all gets stopped on
          the left is they're so afraid of the firing conversation. That's
          why I bring this up.

   jennifer pahlka
          They are, and there's some good reasons for that, right? So if
          Schedule F comes back in a second Trump administration or a
          DeSantis --

   ezra klein
          Can you say what Schedule F is?

   jennifer pahlka
          Yeah, so Schedule F was something that Trump instituted, I think
          in the last couple of weeks of his presidency, that would allow
          him to arbitrarily reassign someone from a civil service
          position into a political position. What's the difference?
          Obviously, you can fire a political position, a political
          appointee with very little paperwork. It's quite easy. Whereas,
          a civil service person, you're supposed to not be able to fire
          unless there's -- you go through -- jump through a whole lot of
          hoops.

          I guess what I would say first is both people on the left and
          the right who work in government -- and, of course, civil
          servants are meant to be nonpolitical -- are frustrated with the
          same thing. And I think that polarization is really sad because
          we're actually responding to the same frustration of this
          demosclerosis, things moving too slowly.

          But on a very practical basis, people I see working in
          government are now trying to solve problems. Yes, they have
          these obstructionists around them saying, sorry, this is how
          it's always been done. The E.S.B. has to stay in the software.
          We're going to build a concrete boat. But the first way to
          combat that is to get other people in to balance them out.

          There's a concept always in government and much theory about
          government, which is you're trying to balance between agility
          and stability. And bringing in more people with an agility
          mindset is a good way to balance that, whereas just firing
          everybody is a great way to potentially, in another Trump
          administration, enable authoritarianism.

          So there's a reason they're upset about it. But I think that if
          the left doesn't start acknowledging that we have the same
          frustrations with the civil service and the administrative state
          as the right, we're never going to get anywhere. And I actually
          spend a lot of my time with people on the right with whom I
          agree a lot more than I thought I would.

   ezra klein
          I want to disentangle some thoughts here, at least for me, which
          is, one, I think people go too quickly to the idea that if you
          created a lot more autonomy and flexibility within the civil
          service, made it easier to hire, easier to fire, that has to
          mean it's easier for the political system to hire and fire,
          which I do think is something we should be at least a bit
          careful with.

          Whereas, I'm talking really about the ability of the civil
          service to hire and fire internally unto itself, which is a
          different question already because they're just different rules.
          That's why in Schedule F, Trump is trying to move people from
          civil to political. But you could change the rules for what the
          civil servants can do internally and still keep -- I think maybe
          you do want less wall than you currently have, but a pretty good
          wall between politics and the civil service.

          And then the other thing is I really am skeptical that the right
          and the left have the same concerns here. And there are places
          where procedures might converge. There are places where some of
          the frustrations might converge. Our friends didn't see this a
          lot in a lot of the work I'm doing around permitting reform and
          supply side issues. That often something that is important to me
          now is deregulating the government. And --

   jennifer pahlka
          We're not supposed to call it deregulating.

   ezra klein
          I know.

   jennifer pahlka
          But that is what it is.

   ezra klein
          But that is what it is. I want the government to deregulate
          itself, oftentimes. And deregulation is coded as a thing you do
          to make the government less powerful on the right. But
          oftentimes, I think regulation is one reason the government is
          not able to achieve its ends. And so it kind of sounds to some
          people like I am saying the same thing that the right is saying,
          but I'm not. I actually want a more active and capable
          government.

          And there are things you might want to do from a deregulatory
          standpoint that would get you there. There are other things you
          would want to do from a deregulatory standpoint that would not.
          So many people on the right do not want the E.P.A. to be able to
          regulate carbon. That is a kind of deregulatory decision, but it
          isn't the one I would make. I would like it to be easier when
          the government wants to build clean energy for them to get the
          permits done.

   jennifer pahlka
          Yeah.

   ezra klein
          That is a deregulatory decision that makes the government
          stronger. So I just think that's also true sometimes within the
          civil service debates, that I would like to see a more capable,
          stronger -- I'd like to see more people employed, in many cases,
          by the government. We've outsourced a lot of functions, which is
          not oftentimes what the right means when they say they want to
          do civil service reform. They often don't mean they want to
          bring a lot of these functions back into government purview.

   jennifer pahlka
          Let me say three quick things to that because it's a really
          important point. And it does get gnarly in there.

   ezra klein
          Yeah, totally.

   jennifer pahlka
          One is, yes, but I also think change happens when people with
          different views do have something in common even if they're
          going to use it for different ends, right? And I think we
          probably have some fertile soil in there to get change done,
          even if people on the right want that change for a different
          reason. Obviously, we need to be careful about that. I agree we
          should definitely be regulating pollution.

   ezra klein
          Oh, it's sort of the last thing I'm going to say on that, which
          is that there's the right and there's Trump.

   jennifer pahlka
          Yes, exactly. Well, that was what I was going to say.

   ezra klein
          And there are people on the right you really can have this
          conversation with. It's just not Trump.

   jennifer pahlka
          So that was my third point. My second point was it isn't
          regulation or not regulation. It's how we do it. And when we
          talk about competencies and capacities in government, I think
          one of the main points I'm trying to make is that has to be a
          core competency in government. And we don't even recognize it.

   ezra klein
          I really take the big point of the book as saying there is
          politics, there's policy, and there's delivery. And right now
          delivery is so subordinate to the people who run politics and
          the people who run policy, that in addition to delivery being
          done poorly, the people delivering often don't think they have
          any power.

          They don't have power to interpret the language they're given.
          They don't have power to say no to the lawyers. There's nobody
          empowered over a process to be flexible as it is ongoing. They
          often don't have the power to change contracts quickly.

          And that shift of power from policy or politics to delivery,
          that feels to me like the great revolution you're trying to kick
          off here. And I'm curious if you had two or three places you
          could start kicking it off -- aside from everybody reading your
          book, and they should read your book -- what would you do?
          Because, yeah, culture can need policy, but policy can also
          create culture.

   jennifer pahlka
          I think there's a lot of things that can happen that fall under
          the banner of where we pay attention and where our elected
          leaders pay attention.

          So one really practical thing -- and I have an example of it at
          the end of the book -- is that when we're writing law and
          policy, who's at the table?

          And I take very much your point about everything-bagel
          liberalism where you already have too many people at the table
          and that sounds like just adding someone else on. But if we want
          to make law and policy that are implementable, then we have to
          have implementers at the table when that law and policy is being
          written. And those people are likely to be and hopefully are
          much closer to the users of the system. And they're going to
          provide insights that make your policy better. But they're also
          going to start to shift that power balance that you talked
          about.

          So an example I talk about in the book is clearing of criminal
          records where the way a particular law was written made it kind
          of impossible to implement at scale. But in the aftermath of
          that, you've got people with technology skills who spend time
          with the people who have these convictions and understand the
          constraints in their lives and how they're going to interact
          with it at the table when decriminalization laws are being
          written, those decriminalization laws are usually the ones that
          contain records expungement provisions.

          That's a huge difference. People can barely believe that this is
          happening -- that you've got coders and designers and user
          researchers at the table. This is at the state level when laws
          are being written. So that's one thing.

          There's a saying -- what we pay attention to grows. So our
          elected leaders have this oversight responsibility as part of
          their toolkit. And we also have the whole apparatus of oversight
          in the executive branch, offices -- of Inspector General, the
          Government Accountability Office, G.A.O. And what do they focus
          on? Failures.

          What if they lifted up the public servants who were having
          success by empowering themselves to have that flexibility and
          made half of their work or more shining a light on them and
          saying, this is what we want public servants to be like and
          we're going to reward them? I think it would make a huge change
          in the culture.

          And I do have several policy prescriptions in the book about
          things like how money is allocated that I think are important to
          pursue. But in the end, I keep coming back to those policies
          fail if the culture doesn't change. So the ways we pay
          attention, what we talk about, who we elevate, I think those are
          all reasonable levers and good places to start.

   ezra klein
          So I think that's a good place to end, too. So what are three
          books that have influenced you that you would recommend to the
          audience?

   jennifer pahlka
          The first is a book from 1973 with the longest subtitle I have
          ever seen. It is called "Implementation: How Great Expectations
          in Washington are Dashed in Oakland; Or, Why It's Amazing that
          Federal Programs Work at All, This Being a Saga of the Economic
          Development Administration as Told to Two Sympathetic Observers
          Who Seek to Build Morals on a Foundation of Ruined Hopes." What
          do you think of that?

   ezra klein
          I love that because that's exactly how pamphlets in early
          American history and 17th and 18th century Europe sound. And I
          love those title formats.

   jennifer pahlka
          Well, you're going to love this book. It covers a project that
          started in, I think, 1967, where politicians and bureaucrats
          went and made a big announcement about investment that was going
          to happen in Oakland, which is my hometown and somewhere I think
          you have lived. And everybody celebrated. And then goes into
          excruciating detail, in the same way some of my book goes into
          excruciating detail, about why several years later essentially
          nothing had happened.

          And they really conclude that this separation between policy and
          implementation is the only reasonable explanation for the
          failure. And they pull up all the other obvious reasons you
          would think that it failed and really debunk them. And it's just
          -- it was sort of like they were reading my mail back in 1973.
          So I really loved that.

          The next one is "Radical Help," by Hilary Cottam, who is a Brit,
          looking at the social safety net, the work force training
          system, health care system. And she takes that concept I have in
          the book of user needs over government needs and just pushes
          them so much further in a way that is really inspiring and I
          think very critical of all of these government systems that we
          have today.

          I mean, it's in Britain, but it's very similar to the U.S. And I
          just think she moves the Overton window about what it would mean
          to truly meet user needs in government-funded systems. And I
          hope everybody who works in social services and health care
          would read it.

          The last one is a bit controversial, and it's not one I'm
          recommending in its content. But I do think people of all
          political stripes should read --

   ezra klein
          Retreats are not endorsements here.

   jennifer pahlka
          Exactly. Definitely not an endorsement. But chapter 3 of this
          new report called, "Mandate for Leadership: The Conservative
          Promise," is essentially the blueprint for what the Heritage
          Foundation thinks the next Republican president should do in all
          areas. But chapter 3 covers civil service rules. It's kind of
          hidden in there, but it does say that they would bring back
          Schedule F, which we can debate, but I think is a bad idea.

          But it's ultimately the way I want to call attention to -- they
          are paying a lot of attention to how to reform the civil
          service. And I would like there to be counterproposals on the
          table since theirs is so thorough and detailed.

   ezra klein
          Jen Pahlka, your book is called "Recoding America." I really,
          really recommend it strongly to people. They really should read
          it. We did not cover a lot that is important in it. Thank you
          very much.

   jennifer pahlka
          Thank you, Ezra. This was great.

          [MUSIC PLAYING]

   ezra klein
          This episode was produced by Emefa Agawu. Fact-checking by
          Michelle Harris and Kate Sinclair. Mixing by Jeff Geld. The
          show's production team is Annie Galvin, Jeff Geld, Roge Karma
          and Kristin Lin. Original music by Isaac Jones. Audience
          strategy by Shannon Busta. The executive producer of New York
          Times Opinion Audio is Annie-Rose Strasser. And special thanks
          to Sonia Herrero and Kristina Samulewski. [MUSIC PLAYING]

   (BUTTON) Previous
   More episodes ofThe Ezra Klein Show
   June 9, 2023   o   1:10:19What Communes and Other Radical Experiments
   in Living Together Reveal
   June 6, 2023   o   1:15:30The Book I Wish Every Policymaker Would Read
   June 2, 2023   o   1:32:05Fareed Zakaria on Where Russia's War in
   Ukraine Stands -- and Much More
   May 26, 2023   o   1:13:45If You're Reading This, You're Probably
   `WEIRD'
   May 23, 2023   o   1:09:42Our Obsession With Wellness Is Hurting Teens
   -- and Adults
   May 19, 2023   o   1:20:53Why Are Teens in Crisis? Here's What the
   Evidence Says.
   May 16, 2023   o   56:44A Libertarian and I Debate the Debt Ceiling
   May 2, 2023   o   1:08:08The Culture Creating A.I. Is Weird. Here's Why
   That Matters.
   April 28, 2023   o   1:18:41Democrats, Pay Attention to What's
   Happening in California
   April 21, 2023   o   1:17:05Matthew Desmond on America's Addiction to
   Poverty
   April 18, 2023   o   1:14:35The `Quiet Catastrophe' Brewing in Our
   Social Lives
   April 14, 2023   o   1:04:27This Philosopher Wants Liberals to Take
   Political Power Seriously
   [8]See All Episodes ofThe Ezra Klein Show
   (BUTTON) Next

   June 6, 2023
     * (BUTTON) Give this article
     * (BUTTON)
     * (BUTTON)

   Produced by `The Ezra Klein Show'

   The New York Times Audio app includes podcasts, narrated articles from
   the newsroom and other publishers, and exclusive new shows -- including
   this one -- which we're making available to readers for a limited time.
   [9]Download the audio app here.

   My pitch for this episode is simple: Jennifer Pahlka has written one of
   the best policy books I've ever read.

   Pahlka served as deputy chief technology officer in the Obama White
   House, and she's the founder and a former executive director of Code
   for America, a nonprofit that works to enhance government digital
   services. Over the course of her career, Pahlka has become obsessed
   with an area of policy that is too often ignored by policymakers:
   implementation. She was part of the effort to rescue HealthCare.gov in
   2013 and was tapped by Gov. Gavin Newsom in 2020 to help fix
   California's unemployment insurance system as it buckled under the
   weight of the Covid response.

   [You can listen to this episode of "The Ezra Klein Show" on [10]Apple,
   [11]Spotify, [12]Amazon Music, [13]Google or [14]wherever you get your
   podcasts.]

   It has become a common refrain that the U.S. government is often
   terrible at delivering even basic services. But Pahlka's new book --
   "[15]Recoding America: Why Government Is Failing in the Digital Age and
   How We Can Do Better" -- puts forward a deeper theory of why government
   services are so awful, how policy implementation so often goes awry and
   what it would take to fix those systems so that government could better
   live up to its promises. It's an argument that anyone who cares about
   government in the 21st century needs to take seriously.

   (A full transcript of the episode is available [16]here.)

   You can listen to our whole conversation by following "The Ezra Klein
   Show" on [17]Apple, [18]Spotify, [19]Google or [20]wherever you get
   your podcasts. View a list of book recommendations from our guests
   [21]here.

   Image
   [30eks-pahlka-image-articleLarge.jpg?quality=75&auto=webp&disab
   le=upscale]
   Credit...Andrea Dunlap

   This episode of "The Ezra Klein Show" was produced by Emefa Agawu.
   Fact-checking by Michelle Harris, Mary Marge Locker and Kate Sinclair.
   Mixing by Jeff Geld. Our production team is Annie Galvin, Jeff Geld,
   Roge Karma and Kristin Lin. Original music by Isaac Jones. Audience
   strategy by Shannon Busta. The executive producer of New York Times
   Opinion Audio is Annie-Rose Strasser. And special thanks to Isaac Jones
   and Kristina Samulewski.

     * (BUTTON) Give this article
     * (BUTTON)
     * (BUTTON)

   Advertisement
   [22]Continue reading the main story

Site Index

Site Information Navigation

     * [23]� 2023 The New York Times Company

     * [24]NYTCo
     * [25]Contact Us
     * [26]Accessibility
     * [27]Work with us
     * [28]Advertise
     * [29]T Brand Studio
     * [30]Your Ad Choices
     * [31]Privacy Policy
     * [32]Terms of Service
     * [33]Terms of Sale
     * [34]Site Map
     * [35]Canada
     * [36]International
     * [37]Help
     * [38]Subscriptions

   IFRAME:
   [39]https://www.googletagmanager.com/ns.html?id=GTM-P528B3&gtm_auth=tfA
   zqo1rYDLgYhmTnSjPqw&gtm_preview=env-130&gtm_cookies_win=x

References

   Visible links:
   1. nyt://article/4ac5f147-4318-51d3-9fdd-a7057b6b88e3
   2. https://www.nytimes.com/svc/oembed/json/?url=https%3A%2F%2Fwww.nytimes.com%2F2023%2F06%2F06%2Fopinion%2Fezra-klein-podcast-jennifer-pahlka.html
   3. https://www.nytimes.com/2023/06/06/opinion/ezra-klein-podcast-jennifer-pahlka.html?showTranscript=1#site-content
   4. https://www.nytimes.com/section/todayspaper
   5. https://www.nytimes.com/section/opinion
   6. https://www.nytimes.com/2023/06/06/opinion/ezra-klein-podcast-jennifer-pahlka.html?showTranscript=1#after-top
   7. https://www.nytimes.com/column/ezra-klein-podcast
   8. https://www.nytimes.com/column/ezra-klein-podcast
   9. https://apps.apple.com/us/app/nyt-audio/id1549293936
  10. https://podcasts.apple.com/us/podcast/the-ezra-klein-show/id1548604447
  11. https://open.spotify.com/show/3oB5noYIwEB2dMAREj2F7S
  12. https://music.amazon.com/podcasts/c4a3b1da-5433-49e6-8c14-0e1da53be78c/the-ezra-klein-show
  13. https://www.google.com/podcasts?feed=aHR0cHM6Ly9mZWVkcy5zaW1wbGVjYXN0LmNvbS84MkZJMzVQeA%3D%3D
  14. https://www.nytimes.com/2021/01/19/opinion/how-to-listen-ezra-klein-show-nyt.html?action=click&module=RelatedLinks&pgtype=Article
  15. https://us.macmillan.com/books/9781250266774/recodingamerica
  16. https://www.nytimes.com/2023/06/06/podcasts/transcript-ezra-klein-interviews-jennifer-pahlka.html
  17. https://podcasts.apple.com/us/podcast/the-ezra-klein-show/id1548604447
  18. https://open.spotify.com/show/3oB5noYIwEB2dMAREj2F7S
  19. https://www.google.com/podcasts?feed=aHR0cHM6Ly9mZWVkcy5zaW1wbGVjYXN0LmNvbS84MkZJMzVQeA%3D%3D
  20. https://www.nytimes.com/2021/01/19/opinion/how-to-listen-ezra-klein-show-nyt.html?action=click&module=RelatedLinks&pgtype=Article
  21. https://www.nytimes.com/article/ezra-klein-show-book-recs.html
  22. https://www.nytimes.com/2023/06/06/opinion/ezra-klein-podcast-jennifer-pahlka.html?showTranscript=1#after-bottom
  23. https://help.nytimes.com/hc/en-us/articles/115014792127-Copyright-notice
  24. https://www.nytco.com/
  25. https://help.nytimes.com/hc/en-us/articles/115015385887-Contact-Us
  26. https://help.nytimes.com/hc/en-us/articles/115015727108-Accessibility
  27. https://www.nytco.com/careers/
  28. https://nytmediakit.com/
  29. https://www.tbrandstudio.com/
  30. https://www.nytimes.com/privacy/cookie-policy#how-do-i-manage-trackers
  31. https://www.nytimes.com/privacy/privacy-policy
  32. https://help.nytimes.com/hc/en-us/articles/115014893428-Terms-of-service
  33. https://help.nytimes.com/hc/en-us/articles/115014893968-Terms-of-sale
  34. https://www.nytimes.com/sitemap/
  35. https://www.nytimes.com/ca/?action=click&region=Footer&pgtype=Homepage
  36. https://www.nytimes.com/international/?action=click&region=Footer&pgtype=Homepage
  37. https://help.nytimes.com/hc/en-us
  38. https://www.nytimes.com/subscription?campaignId=37WXW
  39. https://www.googletagmanager.com/ns.html?id=GTM-P528B3&gtm_auth=tfAzqo1rYDLgYhmTnSjPqw&gtm_preview=env-130&gtm_cookies_win=x

   Hidden links:
  41. https://www.nytimes.com/
  42. https://www.nytimes.com/
  43. https://www.nytimes.com/
  44. https://www.nytimes.com/2023/06/09/opinion/ezra-klein-kristen-ghodsee.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  45. https://www.nytimes.com/2023/06/06/opinion/ezra-klein-podcast-jennifer-pahlka.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  46. https://www.nytimes.com/2023/06/02/opinion/ezra-klein-podcast-fareed-zakaria.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  47. https://www.nytimes.com/2023/05/26/opinion/ezra-klein-podcast-joseph-henrich.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  48. https://www.nytimes.com/2023/05/23/opinion/ezra-klein-podcast-lisa-damour.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  49. https://www.nytimes.com/2023/05/19/opinion/ezra-klein-podcast-jean-twenge.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  50. https://www.nytimes.com/2023/05/16/opinion/ezra-klein-podcast-veronique-derugy.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  51. https://www.nytimes.com/2023/05/02/opinion/ezra-klein-podcast-erik-davis.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  52. https://www.nytimes.com/2023/04/28/opinion/ezra-klein-podcast-scott-wiener.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  53. https://www.nytimes.com/2023/04/21/opinion/ezra-klein-podcast-matthew-desmond.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  54. https://www.nytimes.com/2023/04/18/opinion/ezra-klein-podcast-sheila-liming.html?action=click&module=audio-series-bar&region=header&pgtype=Article
  55. https://www.nytimes.com/2023/04/14/opinion/ezra-klein-podcast-danielle-allen.html?action=click&module=audio-series-bar&region=header&pgtype=Article