This week, as programmers in Washington scrambled to fix the glitch-ridden healthcare.gov, the centerpiece of the Obama administration's Affordable Care Act architecture, their counterparts in Silicon Valley viewed the situation with a mixture of rolled eyes and sympathy.
Many Silicon Valley entrepreneurs have experienced a flawed launch of a big product, accompanied by unexpected delays and over-budget implementation. It's the way things go here, and failure is often embraced as a learning experience. Still, around the coffee shops of San Francisco and the tech campuses of the South Bay this week, the topic of the government's health-care push often elicited cynical jokes — one tech worker quipped that a Stanford computer-science class could have built healthcare.gov for a thousandth of the reported $400 million–plus budget — and confusion. How could a project involving a budget of that size and five years of lead time fail so badly? What was so hard about building, as some have called healthcare.gov, a Kayak for health insurance?
Matt Mullenweg, the co-developer of WordPress (the open-source platform that five states are using to run their own health-insurance exchanges), was among the confused. Healthcare.gov, the largest chunk of which was overseen by Canadian IT firm CGI Federal, "gets more money than all the revenue of most Silicon Valley start-ups combined," he said. He added that although "start-ups have failed launches all the time" — citing Twitter as an example of a site that was "unreliable for years" — the scale of the health-care site, and its importance to the Obama administration's policy aims, should have made transparency a priority.
"If the government is going to spend that much money on something, it should be open source," Mullenweg told me in a phone call. "How cool would it have been if the site launched, didn't work, and some passionate coders came in and said, 'Oh, here's a problem, and here, and here?' But the people who want to help aren't able to."
Eric Ries, the author of The Lean Startup and a member of Silicon Valley's cognoscenti, echoed Mullenweg's concerns. Speaking after an event in San Francisco's Soma district that was filled with tech types, Ries told me that the botched healthcare.gov rollout was emblematic of what he called "state-of-the-art incompetence." He explained that the old-school approach to the site's development, which relies on a management system known as the "waterfall model," and whose back-end database appears to run on Oracle software, had all but ensured it would run over-budget, late, and riddled with technical problems:
"You could take any engineer on the street here and ask them, 'I have a friend who works for a private company — don't mention the government — who's thinking about a five-year, $100 million Oracle installation, and they've hired an outsourced contractor to build it for them. It's going to be proprietary, hosted in their own data center, Oracle-based, with waterfall management. What are the odds that it's working on Day One?' And everyone here will tell you: zero percent."
But while Ries blames the specific structure of healthcare.gov for its failures, other programmers, perhaps sympathetic to Silicon Valley's emerging suspicion of government, see it as symptomatic of the government's ham-fisted approach to technology. The site is "only the latest episode in a string of information technology debacles by the federal government," wrote Clay Johnson and Harper Reed, two programmers with political pasts (Johnson was Howard Dean's lead programmer in 2004; Reed was the brain behind Obama's 2012 digital campaign) in a Times op-ed. The pair went on:
This latest failure is frustrating for us to watch ... We must find a fix to the federal procurement process that spares the government’s technology projects from the self-inflicted wounds of signing big contracts whose terms repeatedly and spectacularly go unmet.
The White House has promised that a "tech surge" of the "best and the brightest," including from Silicon Valley, has been brought in to repair the health-care site. But nobody knows, or is telling anyone, who those people actually are. Contractors grilled at a congressional hearing this week wouldn't name names, and several tech executives questioned yesterday said they had no inside information on the makeup of the government's emergency coder team. Verizon has reportedly been tapped to help with the revamp, but the extent of its involvement isn't known. Former budget official Jeffrey Zients is overseeing the rescue, but since he's not a known coder, his role is limited to managing the project.
Most plugged-in Valley insiders suspect, and hope, that Todd Park is involved. Park, the nation's chief technology officer since last year, is a founder of two health-care-focused start-ups — Athenahealth and Castlight Health — and is seen by many here as the only person in earshot of the White House who truly knows his stuff.
"If Todd Park is in charge of this, they're going to be fine," Ries said. "That guy knows what he's doing; he's an entrepreneur by background. And nobody I know has been able to reach him since October 1, so we all assume he's in there fixing things."
Whether the "tech surge" consists of a few more back-end developers or a full-scale revamp of healthcare.gov, no one doubts that the site will be fixed eventually, and the implementation of the Affordable Care Act will continue apace. (In a conference call this morning, Zients predicted that the site would "work smoothly" by the end of November.) But the episode has confirmed the suspicions of many in Silicon Valley that the government's infrastructure, even on key projects, relies too heavily on outdated legacy systems and could be run more efficiently by those inside their own camp.
"I wish there was a better way to help," Mullenweg said of the health-care site's emergency overhaul. "It's just code."