Recoding America
I talked this week to a military veteran who was frustrated by the level of service she receives for a health condition.
How are we going to solve our problems and make government work better for us? One of the foremost experts on this subject is Jennifer Pahlka, the founder of Code for America and former deputy CTO of the US. Jen wrote an excellent new book: “Recoding America: Why Government is Failing in the Digital Age and How We Can Do Better.” I interview her on the podcast this week. If you want to understand why many of us find government to be so ridiculous and frustrating, you should read her book.
Jen has spent over a decade trying to speed up our government, both among cities and states and in DC. “For a long time I thought of our government’s lack of digital competence as an unhappy mishap, a mistake to be corrected . . . I eventually realized that many others inside the Beltway regarded it more as the natural order of things – inevitable and immutable.”
Jen shares the story of improving the intake form for veterans to apply to the Veterans Administration (VA) for healthcare benefits. The catalyst was a video of a military veteran, Dominic, trying and failing to use the webform which only worked on Internet Explorer and Adobe Reader. A senior official upon seeing the video said, “This is unacceptable, let’s update the form so it’s more usable.” They replaced the form, and applications for services shot up to 10 times the previous levels. People within the VA started saying, “Bring back the old form!” because they weren’t used to getting such high volumes. Happily, the leadership stuck with the new form.
In another story, a VA web page struggled with latency in loading – it took up to two minutes to load a document or new web page. So they defined latency to mean anything less than two minutes instead of the more standard few seconds; they solved the problem simply by changing the definition.
Jennifer catalogues a number of reasons why government struggles with technology. First, all of the energy and focus is on crafting or arguing about policy; actually delivering the policy to citizens is considered implementation and not deserving of attention. “We want to talk policy and leave implementation to the mechanicals. That didn’t work out well for the White House [rolling out healthcare.gov] and it’s not going to work out for the American public.” Meanwhile, the website, whether it works or not, becomes the policy to Americans. “No one reads the legal code, they just go to the website to apply for benefits.”
Second, there are layers of policy that get added, but never subtracted. “It takes 25 years for someone to become expert in processing unemployment benefits in California because they have added rules and workarounds and never taken anything away. Someone after 17 years said that they were only partially expert. It’s like archaeology how some of these systems work; it’s layers and layers of legal code causing complexity, before you even get to the software code.”
Third, the policies tend to err on the side of completeness rather than user-friendliness. Jen cites a foodstamp questionnaire of 212 questions that included questions like “Do you own a burial plot?” When Jen explored the rationale behind that question’s inclusion, the drafter said, “Congress asked for a list of assets, and a burial plot is an asset.” It’s safer to follow the letter of the law rather than to use common sense.
Fourth, career civil servants and bureaucrats operate in a culture that makes sticking to process safer for their career – if they deviate from the rules, they could be punished. If they stick to them they’re fine, whether or not the service works or ‘makes sense to people.’ “They’re in a job for twenty years and see taking criticism as simply part of the job.” Jen describes many well-intended people in government that feel constrained by a thicket of rules and bad incentives.
Fifth, tech products are designed using a waterfall development process – you figure out what it’s supposed to do from the policy down and then can’t go backwards/upwards to make changes. Everything flows down from policy. “This results in building something over ten or more years that no longer makes sense at very high expense,” Jen observes. Most companies now use something called agile software development where you get feedback from users and iterate and improve. Jen’s colleague coined a new law – “Byrne’s Law” - that most government tech products could be developed at 85% efficacy at 10% of the current cost; unfortunately no one has the authority to determine which 15% you could leave by the wayside.
Sixth, there are major actors that prefer the current mode because they can bill the government enormous amounts of money with little accountability. A VP at Oracle wrote that “Government’s expertise should be procurement, not technology itself,” in part because Oracle has become expert in government procurement practices and delivering an effective technology product becomes secondary. Jen writes, “When I told her I thought the [new $600 million IT project] would likely fail, she replied, ‘Do you think I don’t know that? The last seven IT projects in this state have all failed.’”
Seventh, there is a cultural attitude that government is bad at tech and techies should head to private industry to solve the real problems, make money and develop their careers. Meanwhile, Jen helped form the US Digital Service, which employs hundreds of coders and engineers to help modernize government. “Victories aren’t always easy to come by, but when they do they are immensely satisfying because you know your work impacts millions of people in a very important way.”
Jen’s book is a compelling insider account of why government often doesn’t work as well as it should. Unlike others, she doesn’t just throw up her hands – she is grinding away to change things. “It’s easy to complain about government but more satisfying to help fix it.” She also thinks that the mission is critical as Americans are losing faith in our own ability to solve problems.
“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,” Jen writes. “The bewildering assumption is that more of what came before will get us different results.”
Truer words are hard to find.
For Jen’s book click here and for my podcast interview of her click here. To join Forward to help improve our government’s incentives to deliver for us, click here. For my upcoming book, “The Last Election” about how the next election could turn out, click here – it comes out in two weeks!
