This transcript has been edited for length and clarity.
Brian Kenny: HHS, HMS, ACA, CMS, QSSI. No, this is not some secret code. These are all acronyms of organizations involved in the launch of Healthcare.gov, the healthcare exchange website that went live on October 1, 2013—and promptly crashed within two hours. It's the project that President Obama described as an unmitigated disaster, and one that nearly derailed the most significant accomplishment of his administration to that point.
Today, we'll hear from Professor Len Schlesinger about his case entitled, HealthCare.gov: The Crash and the Fix
Len Schlesinger is an expert in strategy, service quality, customer satisfaction, entrepreneurship, and organizational change. He's a longstanding member of the Harvard Business School faculty, who also served as the president of Babson College in Massachusetts.
Kenny: This is a case everybody will be able to relate to. We all saw this playing out in the headlines in very dramatic fashion. Can you just take us to the beginning of the case? How does the case start out? What's it mean?
Schlesinger : The case starts on October 5, 2013. It's four days after the supposed launch of Healthcare.gov. It becomes obvious to the president and his chief of staff, Dennis McDonough, that this is an absolutely unmitigated failure. Across the entire United States population, [only] 26 people were able to register on the first day. The only thing they had going for them is the Republicans had shut down the government, and that was distracting attention from the Healthcare.gov fiasco, and gave them the weekend to sort through what might come next.
Kenny: Why did you write this case? We're a business school. We typically stay away from politics, but there's are a lot of politics infused in this.
Schlesinger : There's politics as part of the process of trying to understand how a major managerial task got bungled. To that extent, there's politics in any traditional organization as well. It's in bolder relief in the context of the United States government, but it's still everywhere… In this case a graduating student last year was really interested in delving much more deeply into the administrative processes that were associated with the failure of Healthcare.gov. We collaborated on the case.
Kenny: I'm just curious, you have some great exhibits in the case including internal memos. Did you have access to the White House for this?
Schlesinger : We had access everywhere. One of the joys of the Freedom of Information Act is everything that appears to be supposedly secret is now out there. The memos, which might have looked like we got from extensive conversations with the protagonist—and we did have conversations with a number of them—all were essentially gotten in the public domain, the McKinsey reports, the David Cutler memo, which we're going to reference in our conversation, all of them in the public domain.
Kenny: It might be helpful to our listeners if you could just do a quick reminder for what the Affordable Care Act was asking for and how the Healthcare.gov website factors into that.
Schlesinger : Certainly. We're spending a lot of time now still debating Obamacare, which is what the Affordable Care Act actually got renamed and is actually called by the president himself. People by and large don't remember what it was about and what it's supposed to do. There were three essential components associated with the Affordable Care Act:
The first was to make insurance available to every American citizen, regardless of pre-existing conditions. Up until this point, insurance companies were fully able to reject people for existing health care conditions. That ability to reject people went away.
The second piece was that all individuals were expected to sign up for insurance or they were going to pay a tax penalty of some magnitude after the year in which they hadn't signed up.
Third was there would be income-based subsidies provided by the federal government to ensure that there was an affordability question that would be dealt with effectively through subsidies. Make sure it's available, make it required, and make it affordable.
Kenny: How did they set about operationalizing that plan?
Schlesinger : The reality is they spent so much time building the plan, writing the legislation, and fighting through congress to get it passed, they actually never spent any time on operationalizing the plan. In March of 2010, the act finally passed. It passed without a single Republican vote in favor of the legislation. It was at that time that people began to wonder about, "Now that we've got this passed, how in heaven's name we deal with this?" The tension really became apparent in the context of one of the quotes we have from one of the White House conferences (where the president talked about the need to actually move forward on this) that said assuming implementation doesn't assure implementation. That very much is the theme of this case and the theme of the course.
Kenny: I saw that quote. I actually highlighted that myself because I thought that was really illustrative. Describe the exchange concept to us.
Schlesinger : The notion was that we would deliver the health care choices through a portfolio of exchanges, largely at the state level, that each of the individual states would have a portfolio of insurance companies that were licensed to sell insurance in their states, that would offer up packages, and then the federal government subsidies would be on top of that. There was one exchange that already existed. It was in the Commonwealth of Massachusetts, largely as a result of that health care package that was passed under Governor Romney's administration. The notion was that the vast majority of the states in the United States would in fact do this, that there would be a federal exchange, but it would be a minor exchange that would essentially be a backup for the few states that actually would not want to handle this on their own. Again, because of the joys of politics, 36 Republican governors elected to not participate in the process of developing their own exchanges. The magnitude of the task that fell to the federal government was way out of proportion to any planning or any logic that they had attached to the project.
Kenny: This is where the fun begins.
Schlesinger : Where the fun really begins. We find ourselves having to create these exchanges, having to create it on a much larger scale than had ever been imagined, and, by the way, having no idea of, a) who was going to do it, or b) how it was going to get done.
Kenny: They had the added challenge of needing to attract younger people to the exchange, because they had to balance out the scales of people who might have ill health, people who are aging and maybe having health issues, with younger people who are in better health. This idea of an online sign-up process becomes very, very important in that context.
Schlesinger : We had two things going on in the administration as they were thinking about it. One was the issue of just setting up the technology that would ease enrollment. The second piece of that was, as you're suggesting, getting a portfolio of young, healthy people to sign up. Remember, the second stool of the Affordable Care Act was the mandate to sign up for insurance or risk a tax penalty. To that extent, they created a private organization called Enroll America, which were some of our finest big-data minds, that was going to be organized exclusively with private money to find young people and to convince them of the desirability of being able to sign up for insurance, when in fact they wake up every morning feeling pretty healthy.
Kenny: Let's go back to the alphabet soup that I started the podcast off with. Who was in charge?
Schlesinger : Nobody was in charge. That's the whole story. We started with the process of the administration having to figure out what agency was going to bear responsibility for this. The case highlights a series of debates between the health care policy team and the economic policy team. The health care policy team essentially argues that this was a logical continuation of the process of health care policy development, and that they had lots of talent to be able to do that work. They already controlled the Center for Medicare and Medicaid Support (CMS).
The economic policy team, on the other hand, was arguing that that was completely absurd, that the policy development had no relationship to the implementation task, that in fact there was nobody in the government who really knew how to do this, and that it was a great opportunity for us to contract out to external expertise for a task the likes of which the government had never, ever seen. The president and (Senior Advisor to the President) Valerie Jarrett weighed heavily on these issues, coming down on the side of the health care policy team, and then the fun began.
Kenny: I would imagine that the issues that emerged as they went down this path with multiple players involved, but nobody really taking leadership or ownership of the outcome, is pretty common in large organizations. We've done a podcast about NASA and it sounded similar to me. It reminded me of this. Can you talk a little bit about that?
Schlesinger : It's unfortunate that we have two examples of essentially government, or quasi-government organizations, where we're asking the question of, "Who's on first?" The issues of technical, political, and cultural dimensions of large organizational change occur in all organizations, for-profit and not-for-profit. They are brought to you in much bolder relief in the context of the Healthcare.gov situation.
They're incredibly instructive to all of us who think about the systems and processes of large organizational change, but they include the issues of: Who is in charge? What is the hierarchy for decision-making? What is the logic for decision-making? What are the interim checkpoints to figure out whether we're on target? How do we ensure a high quality information flow? Who calls the question when we're actually not progressing on the plan in a positive direction?
Every single one of those pathologies emerged in the context of Healthcare.gov, and every one of those was botched in the process of bringing this new technology to market.
Kenny: There was fair warning. You talk in the case of that, the McKinsey report.
Schlesinger : There was no shortage of warning here. One was McKinsey, which had gone in and done a deep organizational decision-making analysis, and the case includes some of the documents that were associated with that. Turning Point Solutions, a consulting firm, had assessed the technology and said the technology wasn't ready for primetime. It is alleged that neither of those reports got to the decision-makers. Therein lies a large part of the problem that exists with the culture of any organization where bad news is not welcome. The notion, in the context of Healthcare.gov, was this was such a critical initiative to the Obama administration that the word went out that failure is not an option.
When the word goes out that failure is not an option, anything that smacks of anything less than perfection just gets buried. The reality of it is, the night before the launch, the Chief of Staff, Dennis McDonough, was getting full information that indicated that October 1st was going to be a magnificent day. Certainly it was anything but. You have a sense of the ability to actually withhold information in a large, complex system in ways that are drawn out of a) fear of being the bearer of bad news, and b) a desire to somehow hope that good news emerges when all the signals indicate that that's not going to happen.
Kenny: It also in hindsight paints a very sympathetic picture for some of the people who were held accountable in the aftermath of this if, in fact, they weren't being given all the information that they needed to make the right decision.
Schlesinger : The vast majority of the people who were involved in the process didn't understand the task. The other fundamental reality we had was that this was going to be a complex end-to-end service management project of building an ecommerce solution. The government had never done any work like that ever in its history, yet, because of its pathological history in procurement, went out to the usual suspects and hired a collection of Beltway contractors who had done big systems, but had no idea how to do this kind of work.
Kenny: There is a happy ending, so to speak. The cavalry comes in. Can you talk a little bit about what turned this around?
Schlesinger : The happy ending doesn't come until the B case. The A case ends in a pretty sorry manner with the world basically coming to an end, nobody able to sign up, embarrassment all over the place, finger-pointing all over the place, and the president and the chief of staff wondering what to do. They went back to a solution that could have been obvious to them early in the process, which was to get Todd Park, the CTO for the United States, and put this in his hands with another fix-it guy, Jeff Zients, and essentially say, "You have no financial constraints, and you have no procurement and contracting constraints. Just get this done."
The two of them put together a tech surge which brought in the finest minds from the industry, from Google, from Amazon, from Facebook. In the context of working literally day and night for six weeks, fixed four hundred bugs, took the reliability up from 43 to 95 percent, and dropped the error rate from 6 to .5 percent. Over a three-month timeframe, enrollments went from 26,000 in the first month to 975,000 by December. It actually wasn't that hard a fix. They came in. They brought the right leadership in. They quickly assessed the system as to whether it needed to be fixed or replaced. They recognized that there were a countless number of people who were working for the existing contractors, who really wanted to fix things, (but) who were working for managers who were interested in protecting their contracts and their own jobs.Ultimately, as opposed to the calcified information flow that we saw in the A case, they went to daily stand-ups with three rules. One was, "We're here to solve problems without blame." Two was, "It's knowledge not rank that drives participation," and, "We're going to triage the immediate risks while organizing the remaining issues to address them.”
The work that happened with the tech surge evolved into Marketplace Light, another collection of young volunteers from great tech firms who essentially rewrote the code for Healthcare.gov, and did extraordinary work there. That devolved into what's now called the US Digital Service, or 18F, which are two organizations that are explicitly focused on not only addressing the technological deficiencies that exist across the United States government, but are also interested in bureaucracy hacking. They've gotten materially more sophisticated about the issues of managing the political and cultural issues associated with intervening in the US government than the government itself had at the start of health care.
Kenny: It's interesting because what we might be seeing here is the evolution of the government moving into the digital age at long last and connecting with this younger generation that's emerging.
Schlesinger : Absolutely, no question about it.
Kenny: You've discussed this in class, I assume?
Schlesinger : Yes, it's a great teaching case. Largely everybody has a very superficial idea of Healthcare.gov and the disaster, and then they go through peeling the onion and looking at all of the fundamental dimensions of failure at the technical level, at the political level, at the cultural level, at the organizational level, and see how easy it is to have a few critical decisions devolve into a process that very quickly is not only out of control, but is out of control and not understood—that actually avoids exposure, as we saw in the context of the McKinsey data and the Turning Point data not seeing the light of day for senior officials. It allows us to easily dump all over all of the principles that are involved in the case. You sit there, and you say, "This is the President of the United States who has access to the finest minds and the finest talent available to solve any problem.” This is the most important agenda item on his plate, and the failure of Healthcare.gov is a simple reaction to the task, as I said earlier, of assuming rather than assuring implementation.
There is a managerial process associated with ensuring implementation that emerges in bold relief in the fixes in Healthcare.gov that are applicable to any managerial setting, where anybody is trying to get something that they care about done, and recognizes that there is generally an inattention to the implementation issues once the policies have been passed.
Kenny: We're in a political year. There's a presidential election coming up. Many of the Republicans have said that they would undo Obamacare. Can you undo this now that it is already built? How do you undo it?
Schlesinger : Undoing Obamacare, which is now by and large portrayed by the vast majority of opponents as simply un-writing the law, has an implementation challenge that wildly exceeds the challenge of actually implementing the law. The good news is that the vast majority of the discussions in the political campaigns are at a level of abstraction where the rubber meets the air. The bad news is if one of them actually wins and tries to do this, we're going to have a lot more cases to follow on the administrative challenges of pulling it off.