The Realignment: 379 | Jennifer Pahlka: How Government Limped into the Digital Age

Marshall here. Welcome back to The Realignment.

Quick note before I jump into today's episode, this Friday, Saga and I will release our latest

Ask Me Anything episode where we answer and respond to questions and comments from Supercast subscribers.

If you'd like to subscribe and get this or any other exclusive episode, you can go to

Realignment.Supercast.com or click the link in the episode's show notes.

On to today's episode. Speaking with Jennifer Palka, she was the former Deputy Chief Technology

Officer of the United States during the Obama Administration and before that, she was the

founder of Code for America. Think of the City Year program, but for technologists working at

every level of government. She's the author of Recoding America, why our government is failing

in the digital age and how we can do better. As reclined called Jennifer's book, one of the best

policy books he's ever read and given her efforts across the federal and state levels of policy

from helping rescue Obamacare's website in 2013 to helping California fix its buckling

unemployment insurance system during the COVID pandemic in 2020. She knows exactly what she

is talking about. There's obviously a long running debate about the size and scope of

government in the US and what makes this episode particularly relevant is that even if you have

individual thoughts about how we need more government programs or less government programs,

there are still just always going to be specific government mandates and tasks that need to be

accomplished and given that perspective and given the challenges of the 2020s and everything from

infrastructure to global threats and last one events, we need to take Jennifer's critique

of how government succeeds and fails seriously. A huge thank you to the Foundation for American

Innovation for supporting the realignment. Hope you all enjoy the episode.

Jen Paoka, welcome to the realignment. It's great to be here. Thank you.

Yeah, it's great to be with you. So I'll address this in the introduction, but

most listeners will know that I interview folks in the public policy space, but also folks in the

venture capital tech startup founders space. And I want to do my best to combine these two

worlds and these two topics because that honestly speaks to your background and your experiences.

So let's pretend I'm on the spot turning this into a pitch deck or a conversation around

addressing building a startup because I actually think it's a useful framework that I wish people

from the policy world would bring into our conversations more so we could be a little

more mission focused. What would you say the problem from your perspective in the category of

government and digital capacity is? That's a great question. I actually think there's multiple

answers, but I'll give you one of them that I think is super helpful. I think we have let

government sort of accumulate by which I mean like policies and processes and technology

have sort of come down over time in like archaeological layers now and those systems

that have just sort of accumulated over time because somebody did something and then someone

else added on to it and then someone else added on to them often don't work well. And so people

assume that they have been designed to be hostile to their users, whether it's bureaucrats or citizens

who have to interface with these processes and systems. And I think they haven't really been

designed at all. They come off as hostile, but we don't really value the discipline of design,

service design, technology design. We don't say, wait, it's somebody's job really to thoughtfully

go back, look at all the things that have accumulated and say, we're going to throw that

all out and just make something that actually works for today. It's 2023. How would you design this

today? What changes do we need to make in the policy to support how it should work today and

actually do that design? So a lot of people on the left are very upset because they feel like

government is starved by design, that we're intentionally keeping resources from it to

make it work. And that may be true, but I think we're also starving government of design because

it's not considered a core competency. It's sort of no one's job. And you can say that, I'm sorry,

you can see that also in sort of one of the stories in my book that I just start out with

and end with, which is this records clearance problem. So states all over the country have been

decriminalizing marijuana use. And as part of that, they have said, look, if you have a past

marijuana conviction, you shouldn't have a felony on your record that's really holding you back.

It's not even a crime anymore. But the process to get that conviction off is crazy. It's like

finding paperwork from different government agencies, filling out forms that are really confusing,

filing them in the right places, waiting to hear back, showing up in court, filing those papers

again. And it's so hard to do that hardly anybody does it. And it's a perfect example of something

that just hasn't been designed. None of that stuff is actually necessary. If you really think about

it, and this will get to your, I think your tech angle, I think the way a technology person sees

this problem of records is that they're just records in a database, their fields in a database,

and you could just change them. And none of those paperwork processes are necessary. If you really

think about what you're trying to get done here. And of course, that is something that we did at

Code for America. But it's a fundamental redesign rather than just saying, Oh, here's the process,

how can we make this part a little bit better? And as I'm recalling the example of marijuana

expungement in California, weren't there like thousands and thousands of thousands of people

that were hypothetically eligible for it? But in the final accounting, only like 23, 24 people

actually made it through the full process of expungement, because there are all these

pledged together hurdles you have to get across. Is that a way of understanding it?

Not even the 23. So I mean, this is San Francisco County alone, and it's just about a year and a

half after the passage of the law. 23 people had even started the process. And we don't think any

of them had gotten through a year and a half later. And then yes, we found out there was about 9,000

people who were eligible. And ultimately, we were able to work with the DA's office and the courts

to file one petition for expungement with like 9,000 names on it. It's a little less than 9,000

in the end. And then there were some that came through a different way. But yeah, there's no way

that giving government more resources to make all those 9,000 happen would have been the right

thing. Just realizing that all that work was unnecessary was the right thing.

I guess the let's stick with the archaeology failure mode, because we could get to a couple

of the other ones you hinted at a little later. I'd love to understand how much of this archaeology

problem is related to digital and the internet versus like what's just inherently an issue when

you're dealing with a government that does things, right, in the sense that if it's the 60s, you're

doing a punch cards, and there isn't even an internet, obviously, you'd still have those same

bureaucratic, accumulatory state, federal, local problems coming up. How much of this is a digital

21st century issue versus this is just how government works. So it is just how government works,

and it's gotten worse in some ways because of digital. So let's start with unemployment insurance,

which I dressed in the book and I worked on the unemployment insurance backlog in California

at the beginning of the pandemic, which is the first place I sort of saw this and said,

oh my gosh, these are archaeological layers. So there, for instance, the complexity of the

policy that governs UI became really clear to us when one of the claims processors that we were

working with kept saying, I don't really understand this. I'm new here. I'm the new guy. And we finally

said, okay, how long have you been here? And he said, I've only been here 17 years. The folks who

really know how to process a claim through UI have been here 25 years or more. That's a lot

of complexity. But if you think back, what did this started in, I think, 1939 with the Social

Security Act? So that's a lot of years that we've had changes to unemployment insurance,

law and policy that comes from executive, legislative, and judicial branches across both

federal and state government. That's going to get very complex. And nobody really does the work

to sort of go simplify it and rationalize it. So of course, that's not unusual. That's how these

things come down over time. But then you have the archaeological layers of technology that

represent that. So here was a change that happened in the 70s. Oh, you can see how the systems were

tweaked to accommodate that change. And then it happens over and over again. But in a very

incremental way, that sort of lacks the ability to go back and just design something that would

work for today. But the other part of it is that we have layered on all of that policy and process

onto technology acquisition itself. So people in the tech world often sort of look at the people

who do government procurement with some disdain, because they have all of these constraints on

them and how they can work with vendors to build systems like, say the unemployment insurance system

in a state. But really, those people are incredibly smart. They're like that 25 year veteran of the

unemployment insurance work who has learned an enormous amount of stuff to get anything done

at all. So the problem of having policy and process accumulate has happened in how we

actually buy and build technology. I would argue that you could be a pretty good junior level

programmer a lot faster than you could become a junior level contracts officer, contracting officer,

procurement officer in government. It is actually more complex to follow all those rules than it is

to learn a new programming language. That's a real problem. If we want to sort of keep up,

I mean, our world is moving really quickly. And all these policies and processes make

our acquisition of software development really, really slow and behind the times.

I'm wondering, do you have a theory of the case for it? And we're doing a state, federal, local

here. So it's going to be complicated. But what is your theory of the case, if any, for how change

actually occurs in these spaces, in the sense that once again, if you're coming from a tech

perspective, your instinct is okay, like this thing doesn't work, like let's start a new thing,

or let's replace this thing and what's to be how much we could reform versus how much we have to

put up with, et cetera, et cetera, et cetera. How do you just broadly explain that dynamic to people

who are looking at it from the outside? I mean, I think there's two big ways that things change.

One is crisis. And that is really consistent in our history. I mean, I happened to work in the

White House during the failure of healthcare.gov. At the time, I was trying to stand up what became

the United States Digital Service, and it was going very, very slowly. There was

not a lot of support for it. And we can talk about sort of the cultural reasons why. But

then healthcare.gov came along and created some urgency. And it was the crisis that ended up

making change. I'm not saying that the existence of USDS is by any means sort of the be all end

all of further change. But it was something that I was glad to see happen and I think has had some

positive impact. We pause there real quick because two things. So I'd love for you to explain

what the digital service is. And then two, there's a more of a decent number of our Zoomer,

Gen Z friends who are listening to this podcast, and they are just not going to be aware of how

much of a disaster the healthcare.gov thing was. So can you explain digital service and then like

what was, because it seems like to me like the healthcare.gov failure is just like the peak

story to just illustrate. During the 2010s, like what could go wrong relative to promise?

So healthcare.gov was supposed to launch October 1st, 2013. It did. It had millions of people

trying to use it. In order for the ACA to actually succeed as a policy, millions of people had to

sign up for health insurance on this website. And the first day, it was so slow, only eight people

could get through the site. I mean, most people think no one got through the site, but it's still

an abysmal. You're like in defense of the person in charge. There were eight people.

Well, I will also defend them by saying that if whatever you feel about the ACA as a policy,

if you're evaluating this as an implementation risk or an implementation failure, by the end of

the open enrollment period, you actually had more people signed up for the ACA than they even thought,

when they had, then they had projected even before the site failed. So it really did succeed,

both as, you know, an initial policy effort and as a website. But that, you know, that was sort of

six months of hell for people in government trying to pull this thing back from the branch of

falling over. I mean, it was mostly down, right? Like the site would just be down. You'd go on,

try to log on and you would get this big error message. And you had an incredible group of people,

including my boss at the time, Todd Park, who was the second chief technology officer ever of the

United States, you know, pulled together a group of people who went over to CMS, the Centers for

Medicare and Medicaid Services, and helped them sort of salvage what they had been given by these

vendors who, who I'm sure were doing the best that they could, but did not deliver.

So my boss had actually hired me to start, what I had pitched him was sort of the equivalent in

the US of a unit that existed in the government of the UK. It was called the Government Digital

Service. And it was a bunch of amazing tech people, tech design, product people in the

center of government, you know, right there in the most powerful part of government. And they had,

you know, not only fantastic talent, but like the authority to go do stuff differently. And

they really shaken stuff up over there. And that's sort of what I wanted to do. And it was a, it was

a really long, hard slog. But in the end, not only did we get the United States Digital Service,

which is now I want to say maybe 250, 300 people who work within the White House,

within the Office of Management and Budget, and get sort of loaned out to agencies to help them,

you know, take a different approach to their digital challenges, but also a really fantastic

unit over at GSA called 18F, that's sort of a consultancy within government to do digital work.

And they do a lot of actual procurement consulting, like helping an agency come up

with the right way to, you know, take a project to market and get good bids on it from good vendors,

really having, you know, done some product management up front, so they really know what

they're looking for. You know, something I'm curious about, when you earlier, just a second ago,

said that ultimately, healthcare.gov, despite the six months of hell, ended up resulting in,

like, more setups than you'd thought, therefore, you know, we should have a question of what failure

looks like. I'm wondering if failure in the healthcare.gov category looks like missed

opportunity, in the sense that now that we have the meme of healthcare.gov, we obviously have

the idea that government can't achieve these big things in these specific categories, in the same

way that we have the meme of the moon landing, whenever we do anything, we're going to say,

this is a moonshot, this is this, this, and that, like, that's a positive meme.

Yeah.

It's like healthcare.gov, except from the question of how many people actually got signed up,

is a negative meme. So can you maybe just, like, talk about that? Like, how do you think about the

question of, like, what can govern, like, what are the actual limits of what government can achieve

in these spaces?

Well, let me pivot back to the question you asked before, which is, what is, like, my theory of

change? And I think that really, that's saying what we pay attention to grows, I think it's

helpful to have healthcare.gov as a thing you can say, like, people say, I don't want this to be

another healthcare.gov. That can spur people in the right direction. But we have an apparatus for

oversight in our government that just finds things that don't work. And so we spend a lot of time

focusing on failure. I tell in the books some stories of amazing public servants who are doing

fantastic things with great success. And those stories are never told. So it's helpful to have

the negative memes because we have to push ourselves to do better. But there is no meme of

Yadira Sanchez, the woman I talk about in the book who works at the Centers for Medicare and

Medicaid Services, which is the same agency that originally screwed up healthcare.gov,

and now is making interfaces for doctors that they love. Like, they're really happy to use these

things that they feel better about working with Medicare because of the fantastic work that Yadira's

team does. Like, why doesn't at least half of our oversight apparatus focus on finding people like

Yadira, lifting her up and spreading her examples of what works? Because then public servants would

really have something to aspire to other than, oh, I don't want to have failed, right? It's easy to

say I don't want to have failed. It is harder to say what does success look like? And are we

rewarding people for the kind of behavior that someone like Yadira Sanchez demonstrates?

And Norma, I tend to avoid the biographical questions. But I think in this conversation,

it's actually useful because you are an outsider coming into government. Therefore,

the question of like, how did you end up in that position in the first place? And then how does that

help you think through these problems in a different way? It's actually deeply relevant.

So how about as much as you can or are comfortable, you tell your story in terms of like, what did

you do first, 90s, 2000s? And then how did that lead you to, you know, Code for America and then

the work with the Obama administration? Yeah, I'm kind of like thought of as a tech person who

doesn't do tech and a government person who doesn't really do government. I'm this hybrid,

you know, neither here nor there. But I got involved in tech. I mean, actually, my first job

out of college, I worked at a child welfare agency. So I think I got to see some of the

functionings of things like child welfare services, which, you know, only government's

going to do that, you know, you don't, it's a very difficult field, very, very high risk.

And when the systems don't work well, the consequences can be really devastating.

If you lose, lose kids and things like that. But I ended up in the tech world, I actually

started out in video games, I used to run the game developers conference for many years,

and then was running something called the web 2.0 conference back when web 2.0 was a big thing.

And it was sort of that connection, wait a minute, if, if, if we can do all these amazing

things in web 2.0, these world changing applications that were built in very lightweight

user centered ways, why can't we apply those principles to the development of the technology

that matters the most, right, our government systems and to the things that, frankly, we are all

stakeholders in. We may all feel like stakeholders in Facebook because we've got a lot of data there.

But at the end of the day, I'm not sure Facebook cares that much about us. We do, in theory,

have some control over what the government does and is the, the digital services that it develops,

it in theory develops in the public interest. So I got interested in that through this gov 2.0

conference that we put on after we were kind of sick of the web 2.0 stuff. And, you know,

through conversations about, you know, how government builds and buys technology and certainly did

back then in 2009 and contrasting it with what I'd seen in the world of Twitter and Flickr

and Facebook back then when they were babies, right? I decided to start Code for America,

which was at the time, and it's changed a lot since then, essentially a service year program.

So took people out of tech for a year and said, instead of working for a startup or

Google, you're going to work with a city government and help them build an application

that is as easy to use and as useful for it, the people who need it as the kinds of things

that you would build in a consumer internet startup. And that's how I got started in it.

I was sort of from there and showing some great examples, particularly with the city of Boston,

where we had our first, one of our first big successes that I got recruited to the White

House by Todd, who wanted to help me, have me help him run the presidential innovation fellows.

But as I said, I had sort of just gotten to know the folks in the government digital service in

the UK. And so I said, well, be great if we could not just do fellows, but actually have

something that's more than a year that's really got the power to make change. And that's how I

ended up at the White House. Then I went back to Code for America and Defense Innovation Board got

pulled back into DC through that. I've touched this issue from a lot of different places at the

state, federal and local level. And this podcast has spent a lot of time in the defense category,

so let's just hear what that was like, of course, too. Being on the Defense Innovation Board was

one of the most powerful and meaningful experiences. I initially turned it down when I was asked to

join. I kind of felt like, oh, that's not really my thing. I work on social safety net, criminal

justice issues. I didn't feel like I had a personal tie to the military. And then a week or so after

I'd said no, I heard Stanley McChrystal speak, and he was talking about all the same issues that we

see in social safety net, for instance, where government contracts aren't working out, people

don't have the tools they need to do the job. And I thought about the fact that these people in

uniform are doing it to serve the country, and we are not serving them, because these policies

and processes don't result in tools that work for them. And I called the guy back up and said,

Josh Marqueuse was running the Defense Innovation Board then. And I said, I think I'd like to hear

a little bit more if I didn't just knock myself out of consideration. I want to hear a little bit

more. And ended up joining and going all over the world with these folks, going to bases,

listening to people at all levels in the military and hearing their frustrations.

And I really developed such an appreciation for what they go through. And I think it only

strengthened my feeling that we've got to do better for the people in the branches of military

service and in the DoD. And I am now actually quite concerned that if we don't do something

different in digital within the DoD, we are not going to be able to defend our country in the

way that we are going to need in the coming decades. I guess the next question is because

this is so interesting that it was once again, like I said, I don't normally do biography,

but I think your biography is interesting in that it kind of tells the story of both parts

of these industries. We're talking about government and the tech industry. I always think back to

2013, President Obama gives his second inaugural address and he's talking about we're the country

of Google and Facebook. And then obviously your story intersects with that in the sense that

we in government are desperate for the knowledge and the background of people in the tech industry.

So you're doing these programs, people come in, et cetera, et cetera, et cetera. Obviously,

a couple of things happen 2015, 2016, 2017, and the government-tech relationship takes it on a whole

other, just sort of like coding separate from just like the Trump era. But what really happened

in this space? I know Michael Cratsios was there, but just what happened to this like narrative?

And I feel like it was different than if Hillary had won election. I think that we would just think

of this category differently than we did. So from your perspective, what happened?

Well, I think people were really shocked that groups like the USDS and 18F not only survived

the Trump administration, but thrived in it. It was hard because they were staffed by a lot of

people who'd been hired under a different administration. And a lot of them decided to

leave and I respect that choice. It wouldn't have been my choice had I been there. I felt like

you're serving your country first and foremost. And the capacity of the government to deliver

is still important, even if you don't agree with the person who is sitting in the Oval Office.

I get that there's concerns that folks, that their work might be used in ways that are is

inconsistent with their values. But what I would say is there was no shortage of work during those

four years that couldn't be done well and in keeping with the values of most of those people,

or probably with all of those people. There was just a lot of work that still needed to be done.

And in fact, I think there was some good stuff that came out of that. I mean,

there was more of a willingness to say, I keep hearing that this can't change because it's the

way it's always been done. And the way it's always been done is not necessarily the right way to do

it in 2023 or I guess back then 2016 through 2020. So there's a wonderful guy named Matt Leera who

came into the White House with Trump, who I knew from years back, he had been McCarthy's digital

guy. He and Matt Cutts, who were, you know, who's running the USDS at the time, like really made

sure that this work got to continue. And while I think there was a lot of, you know, a lot of

worry about the work, both enough people stayed and the folks in power during those four years

appreciated what I think the long term or maybe they weren't noticing. Maybe Matt just made sure

they didn't notice and sort of went on and did other things. But I think the other thing that

happened during that time or certainly happened sort of since then until now is that when I started

doing this work, I mean, I guess I started in sort of 2009 right after Obama's election,

it was definitely conventional wisdom that nobody from tech wanted to work in government.

It already wasn't true back then. Like our very first year at Code for America,

this service year program paid $35,000 for people who could be working at a tech startup.

We still got- For normal corporate America, that's a big cut. That's not just a tech thing.

That's not very much if you could be working. Yeah. It was a stipend, not an actual paycheck.

We still got 525 applications the first year for 20 slots. Like a lot of people wanted to do this.

But even though I think that it was a bit misleading back then, it still is true that

you didn't have the thousands of people wanting to move from tech into government in part because

tech was different then and government was different then. Flash forward to today,

people still say like you're not going to get the top tech talent in government. Now they're

really wrong. We have a lot of people obviously being laid off from the tech firms who are now

considering jobs in government. But even before the layoffs started, we had a lot of

people who were saying, this work matters. It's meaningful. I get that it's going to be really

hard and I'm going to have to put up with a lot of bureaucracy. But this is how I'm going to have

impact in the world is by going into government. And there's like a line out the door now for

government jobs, which brings me to another issue that I hope we get to talk about, which is that

the hiring is very broken. So we're not absorbing all that great talent when we should-

The neuro-wanna anecdote at the start is no doubt relevant to the hiring conversation.

Oh, I think that they've actually undone some of that for federal jobs. Yeah, I don't remember

which ones, but now that's less of a problem. Yeah, unless you're, unless you are, because

there's actually, I think that's probably you're right. That's probably more of a political pointy

issue because all a bunch of younger folks who applied for early Biden jobs ended up getting

caught in that. Okay, let's talk about- This is good. Let's talk about the hiring process then.

So it seems like if we're falling, if we're going back to the starting point around a,

you know, problem solution, et cetera, et cetera, et cetera, it seems that

the obvious solution is like, look, we're America. We have all this talent. We're,

it's the government. The government serves all of us. Like, let's get the best people.

This is very technocratic, but like, it's coding, right? It's tech. It's coding. There's something

analytical. So saying the best people is a little more relevant than it would be in saying

something like foreign policy where it's more value-based. What is the gap between our need

to get the right people into the right jobs and what we're not actually saying?

The hiring just needs some significant attention. I would say reform, but then most people think

I need policy reform, but I think a lot of this can be done without policy reform.

But- Oh, quick pause then. What's the difference? Because I think of those two as the same thing.

What's the difference? Well, so a lot of things need to change where the process that derives

from the policy needs to change. And people will say, oh, I'm sorry, it has to be done this way.

And I'll give you an example in hiring, right? But it has to be done this way. That's what the

law says. But if you actually go back all the way to the law, like the one thing that you

cannot change without, you know, congressional action, the law doesn't say that at all. The thing

that sort of a status quo person, a defender of the status quo is saying is this process has

been around for a long time. We know it's safe for us to use it. We're not going to get fired if we

use it. So it's the thing we're going to do. And actually, I think they often believe that it's the

only way that you can do something consistent with the law. Let me give you a really concrete

example. There's a team at the USDS that started partnering with OPM, Office of Personnel Management,

and the GSA, General Service Administration, back in, I think, 2019, to look at why hiring

is so weird. And by weird, I mean, like, let's say you're trying to hire a web designer, which

is something I actually tried to do when I was there with my friend Marina Nizza. She was a CTO of

the VA, finally got a headcount, tried to hire a web designer because she was doing all of the

websites for the VA on her own. You have to work really hard to get the position

description approved because they just didn't have any position descriptions for web designers.

Like, the VA never had a web designer, I guess, in 2013. So that part is hard in a lot of ways

that I won't go into. Then that job goes up on USA Jobs or whatever, and she's not allowed to touch

it for a while. And the HR people do their stuff. And she gets back a cert, which is a list of names

of people that HR has determined are qualified for the role. None of them are web designers.

It's not that they're bad web designers or they're junior web designers, they're like,

not the high quality, they're just not web designers. And the way that happens, it just,

it seems like, oh, are people just, you know, are people here just don't know how to do their jobs?

Actually, they are doing their job very, very well. But what they are doing is taking a very

large pool of applicants and doing a minimal qualifications process, which basically is

I'm sending all of the applicants to this questionnaire where they get to rate themselves

on various issues, various skill sets. And all the people who rate themselves as, and I kid you

not, the word I think is master level, go to the top. And then they do a resume screen,

and the resume screen looks for exact words. Like, basically, it privileges people who are willing

and know to exactly copy and paste from the job description into their resume.

So no web designer from the tech industry is going to do this.

Quick thing, any listeners who are looking at their summer internships should be taking

notes as we're having this conversation.

If you actually want a government job, ask somebody who knows what a government resume

looks like. It's very different from a private sector industry resume and ask somebody to

coach you through the process because you have to do a bunch of stuff that makes no sense at all

to get the job. So the people who are qualified get knocked out in the first round. The people who

make the list are the people who understand how hiring processes work, know to say,

know to answer master level to every piece of the sales assessment and know to copy and paste.

And then they take the people who have done both of those things, they apply veterans preference,

and that's the cert. And the problem is that is not helping veterans because it is trained

hiring managers to assume that a veteran heavy cert list is a list of unqualified people.

The truth is the qualified veterans are getting knocked out too.

So the USDS team worked with OPM and they created a different process that allows for

real assessment, not self assessment, by other, in this case, web designers,

an actual interview process by what they call SMEs, subject matter experts,

and gets the SMEs reviewing resumes, don't look for the perfect matches, they look for

actual relevant expertise and experience. And so you can get these certs that still may have

either all or many veterans on the cert, but they're qualified veterans, and then the people

actually want to hire them. So it's better for everybody, it's better for veterans,

it's better for the hiring manager. But that required a lot of breaking through this

assumption that the way it's happening may not be perfect, but it's the only way that's legal.

The reality is if you go back and look at the section of the US code that defines veterans

preference, and you read it closely, it actually says that veterans preference is to be applied

after a qualified exam. So the way they're doing it now that everyone told them was going to be

illegal is actually legal. And I think that there's an argument that what they have been doing,

which they say is the only legal way to do it, isn't actually quite compliant with the law.

Okay, so this is the perfect fit to the next question, which is whenever

some in this weird position like you were, because I'm a podcaster who podcasts on policy

and tech topics, I'm both like an insider outsider in both in both categories. But whenever

I turn my most DC, it's when I'm arguing with folks, smart folks who are looking at government

and looking at DC, and they're saying some version of Marshall, when you tell me my ideas are crazy,

or they don't get a more polite than this, but I'm just, you know, making it simple.

These are tech folks, they accuse me of falling into the falling into the this is just how things

are done trap. So I was I'm just curious for you at a narrative level, how does one distinguish

between the two of these things? Because I'm just really thinking you don't have to comment on this,

but like the perfect example that listeners will understand is like yesterday, like Jack Dorsey

tweeted out his endorsement of RFK Junior for the Democratic nomination and someone's like, wait,

how would he make it past the Democratic National Committee? And Jack Dorsey said,

I don't think they matter anymore. He's still going to be able to win. And I get how Jack Dorsey

got there from a tech perspective, because he's saying, look, we're in an era of like,

it's not quite this smart, but he's basically saying like, who cares about the big parties,

no one likes big parties anymore, like we're in the era of like dying institutions, what

matters are the people. But it's like, no, but Jack, it's it's the nomination process for that

actual committee. So the committee in that case does actually matter. So what I'm asking you is,

how do we actually sort between the example you just gave, where it's actually like, because

what I love about your example is, I know we went back to 2015, we could find someone who would

insist until they are blue in the face that no, it's legal, but legally, we have to do the

veterans process the way like, but turned out they weren't correct. So how do we distinguish

between these two situations? Well, I mean, the situation you're talking about with Dorsey is

politics. There's process in there. And I'm talking about sort of the delivery side. But I think I

see your point, like, it's really easy if you're in tech and don't understand government, whether

it's the delivery side or the politics side, just be like, I'm sure with I'm sure this is

an overcomable barrier when in fact, like, it isn't, or it's a lot harder than you think from

your naive perspective. Let me rephrase then, because I think I thank you for helping me get

to the better part of the question. I guess this is more in the self help category. The most useful

experience because I was in DC first, then got into tech. And the most useful thing I picked

up from tech was to a certain degree that like naivete, it's also just sort of like we can do

things, expansive thinking. And that thing that I've learned still does conflict with my DC bureaucrats

status quo bias. How do you advise people think through those class, those seemingly clashing

value systems? Well, if it's people from tech, I think, I think you're right, we actually want

to encourage people to think big and ambitious, and not just take the answer. That's the way it's

always been done. But also, yeah, do the hard work. So I think, I think what is sort of off

putting a little bit about Dorsey's response to that is, who does that hard work of actually

figuring, okay, so you want to have it be different, you think it doesn't have to be the way it has

been. That doesn't magically happen. Like the woman who started this project that I was explaining

that worked with OPM, which by the way is the terrible, terrible acronym, it's called SMEQA,

subject matter expertise qualifying assessments. It really needs to be rebranded. But I think she's

somebody who's looking at this going, wait a minute, it shouldn't be this way. You know,

let's not accept the conventional wisdom that we're hearing that, sorry, this is just what we're,

what we're left with after all these constraints that are piled on by Congress and OPM and agencies.

But she did the hard work to actually go look at the law. And I mean, it was really hard work.

She didn't just look at the law, tell somebody, and it was done. I mean, it was years. And there's

still many years of this to go to have it actually scale across government. But you've got to get

in the weeds and say, I can't just declare that something should be different. I've got to go

show you and then show you again, and then show 7,000 other people again, over and over again,

that this could be different. And not just like that it, like we could read the law that way,

but I've gotten now 50 people hired. They think the first thing they did was a,

they hired a bunch of data scientists under this SMEQA program, which government desperately needed.

And they could hire them in bulk instead of like each job description, getting one person,

they could put out a general job description for data scientists. And then if they got 100

qualified applications, they could actually give them to 100 different or different positions

in agencies that needed them, so far more. You actually have to show it working.

Like you can't just say like theoretically, you've got to do the hard work to get somebody in

government to agree to try it out with you as a pilot and say, nope, this really works. But we're

talking about a lot of elbow grease here. You know, your answer actually got me to the core

of my sympathy to the Jack Dorsey, like outside her politics was better because it seems like a

reality in tech is that the idea of inevitability matters to a certain degree. So for example,

aside from like the actual companies that were built and the work of, you know, Craig Newark

and my Craigslist, the second the internet was created, eventually the monopoly business model

of like a local newspaper was going to die. So you can and you so once again, you could

sit up in 1991 and say like, Oh, wow, it might not be tomorrow, it might not be next week,

but someday you're not going to be able to make $500 for someone putting a bike on it. That's not

how the math worked. This is before my time, but you weren't you weren't going to be able to build

an entire newspaper off of a geographic monopoly because digital has changed that government

doesn't have to respond to any of what is inevitable in government from that sense in the

sense that like, what did what does the digital era change about the way government works or is it

all just able to be blocked by laws by the wrong mentalities like those different categories?

I mean, I argue that what the digital era has done is less like, okay, now there's technology,

you know, technology isn't adopted well in government and more it's people people's

expectations changed and government is going to be forced in some way to meet them. And I think

we are like COVID test.gov was the absolute opposite of healthcare.gov. It was easy to use

really simple worked great. They did it in three weeks or six weeks or something was crazy.

But I will I do want to argue with you about this idea of inevitable like,

I think almost nothing is inevitable. And we make choices every day that add up to what then seems

inevitable. So it's not inevitable that government won't change. And I think there are just people

working to change it. And that's why it's changing. It's as simple as that. I don't think it was

inevitable that the internet became what it was. Like it was started with such incredible idealism.

And then it happened to sort of very early on adopt an advertising driven business model.

That was a choice people made, it took off, it could have gone differently.

I just think that if we think everything's inevitable, we might as well just go home and

like watch succession for the rest of our lives. And like, there's more hard work to be done,

which will push things in the way in the direction that we want them to go,

especially in government where, yeah, it's the thing that we have some control over.

I don't we don't have any control over Facebook or Meta or wherever it's called.

I think, yeah, that's another 2013 versus 2023 thing. Okay, so in these last few questions,

here's just a big question to ask if I was recording this episode in 2014, 2015, when like tech is like

at the top of like these are America's best industries, to your point about like customer

expectations, like I would say cliche about, you know how you take your phone and press Uber,

and you get an Uber, like that's how X should work. Like that's just that was the 2015 cliche

martial era. I guess the question for you is now that we're after the like, you know, the 2010s tech

boom, some of the luster and separate this is separate from like the Trump era political takes

just that now that like the zero interest rate phenomenon luster has kind of fallen off. What

is the gap between what we should learn from tech in government? And basically what basically

asking you is, yeah, what were there any lessons you could have taken in 2015 that actually were

not lessons and we're actually just various, very specific realities of the era?

I think I understand your question. Maybe we'll go back at if this isn't the answer that that fits,

but like, if you look at tech today, so so in 2015, you had people saying, you know,

these are the crazy things you can do, right? You can summon a car with your phone. It's amazing.

We were all very happy about it. Behind that is a practice, a discipline that the consumer

internet industry got really good at called user research. That stuff doesn't just happen.

It's not that people are just geniuses and they know it. I mean, there's obviously big insights

like yes, maybe you could but like the reason those things are easy is because the tech industry

used user research to figure out how to make them easy for people. And that is a thing you

learn to do. It's not that hard, but like people get very good at it and it's a discipline like

programming or design or anything else. And that can be used in government. And I think that's

something we were saying back in 2015. But I think what we didn't catch is that user research can

also be used to create products that we end up having enormous regret over, right? Like the whole

way that the tech platform started giving people what they wanted, but not what they needed, right?

Like we're gonna feed you stuff that sure, you want to keep consuming it. Really good at knowing

what people wanted. I think taking that concept and saying we're gonna use it in government to

give people what they need. Like they need food stamps. They need to get their criminal record

cleared. Like we need national security. So we have to have things that work for the users inside,

you know, the Armenian Air Force. Like it's the use of techniques from the tech world in government,

but in a framework that's more positive and more in the public interest and I hope in the end more

powerful. It's been unfortunately very powerful the way the tech platforms have used this,

but it will also be, it is also powerful in government because it really helps,

it helps, it helps heal that lack of trust between government and the people, which is a,

when we do that is a virtuous cycle. And it all does come down to can we make things that people

like using? So for the last two big questions, so the example we gave at the start is about, you

know, expunging marijuana records. I'm going to take a quick guess and assume that is not going

to apply to 99.9% of the audience. So I think from a good takeaway perspective, like what are just

immediate and once again, that's not diminishing like the how pressing issue that was, but like I

think that was just a great example to start the book off with of here's an example of how like

the highfalutin stuff we're discussing very tangibly could make a difference in your lives

and how like weird the system is. What are some other examples that would come to mind of like,

maybe they haven't been solved yet, but just like obvious like next step areas, state, local,

federal, the folks who focus on. That I want people to do or what they might get a benefit from?

Yeah, either or whichever one interests you more is a question.

Well, I'll take the opportunity then to say I think more people should go work in government. I

did just say that there's a bit of a backlog in terms of getting people through the door because

of the really crazy hiring process, but I still think that there it doesn't have to be tech people

either. I mean, it's working in government even if it's just for a couple of years is an incredibly

rich and meaningful experience. So number one, do it your life will never be the same in mostly

good ways. So you'll also be frustrated a lot. I think people have to hold their elected leaders

accountable in different ways. We're in a delivery era. It doesn't really matter what the person

you're voting for says if they can't deliver on it. It was very exciting that we could say that these

marijuana was decriminalized and people with these records would no longer have them. But like

if a year or two later, it's exactly the same for all those people. We haven't delivered and

in the end of the day that our elected leaders are going to say, but it's the administrative

agencies that have to do that. But they have enormous control about over the administrative

agencies that they and they need to change their approach to them. Number one thing they could do

is bring implementers to the table when they're writing law and policy. Like actually invite

the people who are going to have to implement what you're the law policy that you're considering.

Get them there and have them speak to how to write an implementable law. We don't ask our

candidates, our elected leaders to operate in these new ways. There's a lot more ways I would

like them to operate. And ultimately, if that's how accountability works in our system, we're

going to have to start doing that. Man, I love, and this is how we'll close, I love the delivery

era framework. I'm going to steal that. I guess the question then to close the loop on tech versus

government. We're talking about delivery in the case of companies venture capital. The reward

system is actually very straightforward. You get an exit, you get a great job that pays millions,

millions of dollars. When you're talking about the marijuana example specifically,

the elected politicians in California, this isn't why they won their office, but like,

I'm a Democrat, I'm in California, we're like with the people on marijuana policy, like that at

a broad level contributes to holding power in the state. So there's a, there's a, there's a

benefit you got there, but there was a gap between delivering the restitution and the actual

implementation. How could we solve the gap? As in, there wasn't a politician who was punished for

not following up. So, and, you know, there wasn't anyone who said like, wait a second,

like you said you were going to get all these people with their records exposed so they get

jobs. You didn't actually succeed at that. Therefore, I'm going to primary you or you're

going to lose the general election, especially because you're doing a felon. So they couldn't

even vote in the first place. How does fixing the delivery side of things like work when you're

doing with government? Well, ultimately, I think it's on us. We've got a, as I said, hold our

elected leaders accountable to sort of different outcomes and also think about how we hold our

bureaucrats accountable. I mean, we forget that they are in this accountability trap of supposed to

be fidelity, you know, excuse me, they're supposed to show fidelity to these processes

that don't really work, but then we also hold them accountable to outcomes. So I think a little

empathy for them and sort of changing the circumstances under which they work is really

important. But I think, you know, I guess one thing that comes to mind there is we talked about

California, a very democratic state, right? I think that we have to start thinking about

less about, are we on the left or on the right? And more about, are we building state capacity

or are we destroying it? I completely understand why people want to. They are frustrated by the

same things that frustrated me when I was in government and continue to frustrate me now that

I work still with government but from the outside. But ultimately building it is the way that we have

a less intrusive government that is less burdensome and more effective and gets us the kind of

government we want. And I, you know, I just don't know that the left and right framework is all

that helpful in thinking about where to put your energy. Very well said. That is an excellent

place to end the episode. Jen Palka, thank you for joining me on The Realignment. The book is

Recoding America, Why Government is Failing in the Digital Age and How We Can Do Better. It

is available now in our bookshop. Thank you for joining me on the show. Thanks so much, Marshall. This is fun.

Hope you enjoyed this episode. If you learned something like this sort of mission

or want to access our subscriber exclusive Q&A, Lotus episodes and more,

go to realignment.supercast.com and subscribe to our $5 a month, $50 a year, or $500 for a

lifetime membership. Thanks. See you all next time.

Machine-generated transcript that may contain inaccuracies.