“Modernizing government technology” is a strong contender for the phrase with the greatest gap between how mind-numbingly boring it sound and how important it is in the lives of real people. How government buys and builds technology shapes everything from our experiences down at the DMV to the country’s implementation of sweeping national policies. What’s more, it’s an industry measured in hundreds of billions of dollars.
Sitting at the middle of that is Jen Pahlka (@pahlkadot). In 2009, Pahlka started Code for America, a San Francisco-headquartered non-profit that partners people with tech expertise with state and local governments. In 2014, it would serve as the inspiration for President Obama’s creation of the U.S. Digital Service, trying the same experiment on the federal level. Pahlka would go to Washington herself to serve as a deputy chief technology officer of the United States.
Pahlka returned to California and Code for America. She stepped down as executive director in 2020, and she’s now at work on a book. It comes as there’s considerable interest in D.C. about how the Biden administration will handle government tech at all levels, across the country. (This interview has been edited for length and clarity.)
Scola: I have to say, I was just laughing about your ants-on-a-log tweet. [She posted about needing help understanding a crossword clue: “Raisins : ants :: celery : ___.”]
Pahlka: When I saw a picture, I was like, “Oh, I think I have seen that before.” But I've never eaten that and had no idea it was called that. [Laughing.] I guess I'm the only person in the country.
You said you recently turned in the first draft of your book. What’s it about?
I’m trying to write what I would call an Atul Gawande book; I know that sets a bar I probably cannot meet, but he writes for regular people. How government has adapted to the digital age is relevant for the general public. Either people see it very black and white, or it’s just frustration over something like Healthcare.gov. But it’s a lot more nuanced than that.
I’m trying to help people understand the dynamics of why government has had such a tough time mastering the digital age.
“The policy class sometimes misses that implementation is how you learn to get the policy right.”
So why has government had such a tough time mastering the digital age?
One of the major themes is that, in fact, it’s not really technology.
Even before everything went digital, there was a real divide between the policy class and the implementers. It’s much deeper than, ‘We don’t know how to program.’ It’s a deep-seated divide between what’s thought to be important.
I think the policy class sometimes misses that implementation is how you learn to get the policy right. And the divide between people who think of themselves as ‘policy people’ and people who think of themselves as ‘tech people’ is not serving us.
How should it work instead?
I have a story from a project at [the Centers for Medicare & Medicaid Services] that I don’t totally have clearance for yet, but you have a 150 person team — some of whom are policy, some of whom are implementation — trying to get a launch out the door because a law’s been passed.
Each individual part of the policymaking apparatus is doing the absolute best job it can. I can’t tell you a story about anybody trying to do something wrong. But the people who are thought of as the tech people are going, ‘Hey, I think we need a giant conference room, where we can put huge sticky notes all over the walls and map out each part of the system.’
What ends up happening is creating an enormous burden on the user, because we can’t see the full thing.
But is the deal sometimes that the technology can’t be built well because the public policy itself is badly designed?
So here in California, Prop 47 passed in 2014 to decriminalize various low-level felonies. People got in a room said things like, ‘We’re going to make, you know, theft of less than $950 a misdemeanor.’ Great, brilliant, love it.
Then two years later Prop 64 passes and legalizes marijuana. Code for America figures out a way to get the state to give access to the database, and then run an algorithm that pulls all the records. So then, if you can do that for Prop 64, go do that for Prop 47.
But the problem there is there’s no field in the database that says how much the property theft was for. You’ll find, you know, ‘In 2003 Jen Pahlka stole a camera.’ What model camera? Was it worth $950? You literally can’t automate it.
You actually have to understand the implementation of the law before you write it.
What should have happened in the Prop 64 case instead? The people writing the law would have consulted with someone who would have told them that ahead of time?
Yeah, and that's actually what's happening now.
The Code for America team worked with Michigan, with Cook County, Illinois, and said, ‘It's great that you want to write a law decriminalizing things. Can you please talk to us before you write the law so that it will so that it can be automated? Because if it’s not automated, it will have minimal effect.’
Someone who did digital work at the Treasury Department recently tweeted that third-party tax preparers objected to their redesign of IRS.gov, saying, she said, that they called it “Elizabeth Warren’s dream” because it makes it easy for people to file directly. I don’t know about that story specially, but how often is it a case that there are people who don’t want government tech to work?
I would say that it's not that there are not people who are working against the public interest. Of course there are, and we should call that out.
But to substitute that anger for the real work that needs to happen — which is mostly good people trying to do the right thing and we can't get out of our own way — is a big mistake, because then we don't do the work that we need to do. If we can’t win when nobody’s trying to stop us, then when are we ever going to win when they are?
“I think we as citizens should stop thinking, ‘I have a DoorDash view of the world, and that’s how things are supposed to work.’”
The country went through a patch where the websites meant to help people set up vaccine appointments weren’t working very well. Does that help make your case?
I don’t think it did. I gotta be honest, if in May or June you looked back at the vaccine rollout in California, it was pretty damn good. But the rhetoric about it in April was that it was a disaster.
People expected this, you know, Ticketmaster-style aggregation. If you’re trying to meet people’s expectations, you didn’t meet them. If you’re trying to get people vaccinated, you did just fine. We didn’t have an information problem. We had a supply problem. I think government was appropriately focused on getting supply.
Everyone’s now talking about the problem the state of California knew was going to be the actual real problem, which is that people don’t want the vaccine. And who’s fault is that? That’s a misinformation problem. It’s not a ‘government is incompetent’ problem.
I guess I’d kinda be defensive of my government friends here. I think we as citizens should stop thinking, ‘I have a DoorDash view of the world, and that’s how things are supposed to work,’ and instead think, ‘Is the job getting done?’ Because that’s what’s important.
Let’s switch gears. What happened with the U.S. Digital Service during the Trump administration?
I think it did really well. I know some people will call that a failure. I think it’s great.
Why would some people consider that a failure?
Because people who hate Trump didn’t want there to be any successes under Trump. To say I’m not a Trump fan is drastically understating it. But that doesn’t mean I want our federal government to fail while he’s in office.
Did [the U.S. Digital Service, a.k.a. USDS] have significant stresses? Yes. Did it generally rise to the occasion? Yes. I feel like the leadership deserves an enormous amount of credit. The people who did service over that time are under-appreciated in my view. It wasn’t that they held the fort. They actually made progress.
“I think [Kushner] not doing anything new but supporting something that was already working is a great outcome. I’m not sure I wanted him doing our next-level strategy.”
Like what?
SME-QA [Subject Matter Expert Qualification Assessments] is a USDS project to reform federal hiring, and I think it may be one of the most important things USDS has ever done.
And it happened entirely under the Trump administration, by dedicated public servants who were like, ‘I don’t care who’s in office. We still have a really messed up federal hiring system that makes it impossible for our agency partners to hire competent technical staff.’
And it’s not just tech jobs. It’s all jobs. They’ve been fighting that fight. It barely has anything to do with technology, but it is fundamental to how government works.
Jared Kushner, as part of what he called his “Office of American Innovation,” was put in charge by President Trump of modernizing federal IT. How did he do?
The only thing I genuinely know about it is that they did a good enough job protecting the USDS and letting it continue to do the important work that needed doing.
I think him not doing anything new but supporting something that was already working is a great outcome. I’m not sure I wanted him doing our next-level strategy.
Part of the U.S. Digital Service model is that people who work in the tech industry will pass through Washington for a few years. But things are a lot more tense between D.C. and Silicon Valley than back when it launched. Does that make it more difficult?
It shouldn’t, in the sense that the people of the tech industry are different from the companies of the tech industry.
What is more complicated is there is much more desire to bring in folks who’ve had government experience. In the early days, there was sort of this vibe that I think people are now a little embarrassed by, of, if you were from a big company, it was, ‘Okay, it’s so great you’re here. Let’s put you in charge.’
Now it’s a little bit more, ‘Okay, you’re from a big company. I’m so glad you’re here. But you’re new to government and you need to actually take a back seat and learn all the stuff we’ve learned, because it is different. And, frankly, it’s harder.’
“I don't particularly care that space has been co-opted by billionaires. I do care that Earth has been co-opted by billionaires.”
You served as deputy U.S. CTO. Six months in, Biden has yet to appoint a U.S. CTO. What do you make of that?
I think it’s hard, and they’re trying to get it right. They have a lot of criteria.
The world around somebody who gets elected president of the United States is not generally full of the right people to be U.S. CTO. So that’s going to be a harder one for them to fill. Of course it’s disappointing. But I do think some of it’s understandable.
I think the real solution is what I was talking about — more connectivity and understanding between these two fundamentally different social structures. I don’t mean Biden and Facebook. I mean the tech-oriented, implementation-oriented people and the political class.
What sort of criteria?
I don’t know exactly what they’re looking for, but I would assume that it would not be appropriate for someone to come in without government experience.
This might not sound like a question about government tech, but given the history of space travel in this country, I swear it is. What do make of all the attention being paid to Jeff Bezos going to space?
I don't particularly care that space has been co-opted by billionaires. I do care that Earth has been co-opted by billionaires.
Look for Pahlka’s book in the fall of 2022.
GETTING THINGS DONE
We love productivity tools here at Slow Build. Pahlka shared hers:
“I’m the most unproductive person…but my friend Marina Nitze just put something out [called TaskTackler]. She’s really organized, and I’m sure it’s amazing.
Oh, okay, here’s my only productivity tip. My friend started a seven-minute workout by Zoom, and now there’s one in the morning and one in the afternoon. At 9:30 and 4:30 every day I get on and do a seven-minute workout with a bunch of friends. That’s actually very good for my productivity.”