The Flop

On November 14, six weeks after the failed launch of HealthCare.gov, the federal government’s insurance marketplace website and the public face of the endlessly contested Affordable Care Act, President Obama stood behind a podium at the White House and offered a mea culpa that was anything but. True, he did repeat “that’s on me” and “that’s on us” when he wasn’t invoking jocular sports metaphors (“we fumbled the rollout on this health care law”), but he also said this:

I was not informed directly that the website would not be working the way it was supposed to. Had I been informed, I wouldn’t be going out saying, boy, this is going to be great. I’m accused of a lot of things, but I don’t think I’m stupid enough to go around saying, this is going to be like shopping on Amazon or Travelocity a week before the website opens if I thought that it wasn’t going to work.

And so we are left to wonder how it was that the president not only had not been informed (note the passive construction), but how it was that he did not make it a point to be informed, not merely because he is the president and that’s his job, not merely because he has made health care the legislative priority of his administration, not merely because affordable health care is the hook on which he hopes to hang his presidential legacy, and not merely because the Republicans have been trying to thwart him at every turn, but because of all of these. Knowing how sensitive Americans are to the shortcomings of big government should have made Obama especially vigilant to the disasters ahead if his administration botched HealthCare.gov. On November 19, his spokesman, Jay Carney, admitted that the president had in fact been briefed in April about the danger that the website would fail when it was launched in October. His carelessness will be a good place for historians to begin to unwrap the enigma of a president who appears to run the country on a need-to-know basis, ceding what needs to be known to other people.

In this case, other people knew almost from the start that the website was going to be difficult to build. According to The Washington Post, in May 2010 an outside health adviser told Larry Summers, then the director of the National Economic Council, and Peter Orszag, the head of the Office of Management and Budget, that “no one in the administration was ‘up to the task’ of overseeing the construction of an insurance exchange and other intricacies of translating the 2,000-page statute into reality.” Subsequently Summers and Orszag, according to the Post, recommended that the president appoint a health reform “czar” from outside the administration with triple expertise in business, technology, and the health insurance industry.

This last is crucial since the health care reforms championed by the Obama administration have little to do with the delivery of health care and almost everything to do with the sale and provision of health insurance. Under the ACA, no American can be denied coverage and the cost of coverage cannot exceed more than 9.5 percent of their income. But health insurance is regulated state by state, and a federal system that mandates universal coverage—which is to say a federal system that requires everyone to either buy coverage themselves from private companies, or get it through an employer, or obtain it as a government benefit—has to take into account the many rules and regulations that pertain to each state.

This might have been a manageable task if each state were to run its own health insurance exchange website where residents could log on and shop for the best policy available, but only fourteen states chose to do so, since opting out by a state was yet another way for the Republicans to muck things up. For the law to work, then, the federal government had to create a single national website that enabled residents of the remaining thirty-six states to shop for and purchase insurance in their particular state, with the price reflecting the new tax credits provided by the ACA.

“There aren’t a lot of websites out there that have to help people compare their possible insurance options, verify income to find out what kind of tax credits they might get, communicate with those insurance companies so they can purchase, make sure that all of it’s verified,” the president said in that November news conference. He then suggested that such a task was too difficult for the same government that, for instance, has designed the greatest online data-mining surveillance apparatus in the world: “The federal government does a lot of things really well. One of the things it does not do well is information technology procurement.”

Even so, there are many successful computer systems and websites both within the government (the Defense Logistics Agency comes to mind) and outside of it (Amazon, Google, Kayak) whose complex gearing is housed behind a simple interface. Businesses know that they’ve got five seconds of a user’s attention before that potential customer moves on—it’s what drives them to make the navigation of their websites as easy as possible. Since its launch, HealthCare.gov has often exceeded this by over 3,600 seconds due to software glitches, log-in overload, and a user experience that is remarkably cumbersome: it requires people first to set up an account that includes divulging personal information and having it verified by the Department of Homeland Security, the Internal Revenue Service, and the Social Security Administration before they can see how much it will cost to buy insurance. Imagine having to prove your income and residence and employment status before being able to look for—let alone buy—a pair of pants on Amazon.

And now, suddenly, the Republicans are loudly expressing a newfound interest in data security, using it as yet another way to stymie the ACA. This is just one in a long series of feints and jabs that began before the law was passed and has continued with relentless fury for the past three years. Even without landing a fatal blow, the Republicans have kept the Obama administration ducking and ducking. According to The Washington Post, HealthCare.gov “was hampered by the White House’s political sensitivity to Republican hatred of the law—sensitivity so intense that the president’s aides ordered that some work be slowed down or remain secret for fear of feeding the opposition.”

That fear led the administration to house implementation of the law at the Department of Health and Human Services and then, when it was clear that HHS was out of its depth, at the Centers for Medicare and Medicaid Services (CMS), where it would be buried in the bureaucracy, out of Republican crosshairs. And Republican opposition to the law ensured that no money was allocated to fund a project administrator to coordinate the effort, and it also meant that there was no provision in the law to finance the national exchange website.

Clearly, the $400 million or so the administration has scrounged up by moving money from one HHS account to another and from the law’s implementation fund has not been enough to create a stable, working website. This has less to do with the challenge of writing code than with the fractured way that code has been written and the absence of a single overseer to smooth out the fissures. HealthCare.gov is essentially an extensive public works project with no one paying attention to whether the pipes on the first floor are the same diameter as those on subsequent floors or whether the doors have been sized for the jambs and the traffic expected to go through them.

As with many government endeavors these days, work on the website was jobbed out to contractors. And not only that, it was jobbed out to numerous contractors, so that one company, Development Seed, made the front end—including the visible website—and another, CGI Federal, made the back end—the programs for processing data for both users and insurers—and that company then hired fifty-five subcontractors (Oracle, UnitedHealth Group, Booz Allen, and Verizon among them) with no one making sure that each piece fit with the others.

Government procurement and bidding being what it is—which is to say bureaucratized and arcane—meant that at least one of those contractors, CGI, was hardly vetted. If it had been, perhaps its three years of missed deadlines on a diabetes registry website for the Ontario government, which eventually fired the company, might have given someone pause—the secretary of health and human services, Kathleen Sebelius, say. And maybe someone might have noticed that the software platform that CGI was proposing to use was outdated by about a decade.

In her prepared testimony for the House Committee on Energy and Commerce delivered three weeks after the misguided launch, Cheryl Campbell, a senior vice-president at CGI, stated that “no amount of testing within reasonable time limits can adequately replicate a live environment of this nature.” In fact, the site was tested for a scant two weeks. As Andrew Slavitt of Optum, the company that owns the company that did the testing, told that same House committee, it was clear in the weeks before the launch that “more testing was necessary…[and that] pieces of the system we had tested had issues” and that administrators at CMS (and presumably CGI) knew this.

Either no one told the president, the president didn’t ask, or, as Sebelius more or less admitted on November 6 when she was called to testify before Congress, the administration decided to go out with a compromised product to at least get something up and running, come what may. Remarkably, two weeks later, on November 20, in his testimony to Congress, Henry Chao, the chief information officer at the Centers for Medicaid and Medicare, stated that as much as 40 percent of the website’s back end still needed to be built.

“Like every new law, every new product rollout, there are going to be some glitches in the sign-up process along the way that we will fix,” President Obama said in a speech on October 1, launch day. “Consider that just a couple of weeks ago, Apple rolled out a new mobile operating system. And within days, they found a glitch, so they fixed it. I don’t remember anybody suggesting Apple should stop selling iPhones or iPads—or threatening to shut down the company if they didn’t. That’s not how we do things in America. We don’t actively root for failure.” Would it be too much to root for competence?

—November 21, 2013