The Affordable Care Act Website Disaster: A Management Lesson from the Manhattan Project

The recent website disaster of the Affordable Care Act (ACA) resonates with those of us who come from engineering backgrounds. Engineers appreciate and quickly learn to anticipate what can go wrong…in any situation. I like to think of it as the “engineering mentality.” The recent ACA website problems seem to validate the scarcity of that mentality in Washington.

Murphy's Law_2

I wish to state up-front that my comments, here, have nothing to do with the inherent merits (or not) of the legislation itself or the politics that surround it.

I am interested specifically in what went wrong with its website implementation – not from a software developer’s technical point of view, but from a program management standpoint.

 What went wrong is abundantly clear: No specific person was designated to coordinate the effort who possessed the necessary independent authority and the technical competence to make the correct and necessary decisions. The task clearly required someone’s full-time attention to the effort because of its scale and importance. Furthermore, the Washington bureaucracy involved in the project failed to sufficiently monitor those upon whom they depended – specifically, the three or four major contractors.

Management and tracking of the website’s progress, even by the Washington folks with a huge vested interest like Health and Human Services (HHS) and the president himself, seems anemic at best. It appears they thought since the project is being handled by a “bunch of professionals,” everything will turn out to be just fine. A “bunch of professionals” is a useless entity for assuring true accountability and effective program coordination.

Capitol Building

Washington seems quite unfamiliar with the cornerstone of the engineering mentality – Murphy’s Law – which states, “If anything can go wrong, it will !”

 The Manhattan Project and the Development of the Atomic Bomb:
The Primary Take-Away Lesson for All Large-Scale Programs

The number-one lesson: Have a focused responsibility for overall program coordination and accountability – like the Manhattan Project during World War Two, which was even more demanding than “rocket science.”

 Let us look back at that program, its management, and its accountability/visibility. As I pondered the stinging review of the website fiasco in Peggy Noonan’s Wall Street Journal column this past week, the focus of my mind’s eye settled on two individuals who were critical to the Manhattan Project and its successful development of the atomic bomb – a really difficult challenge. Those two individuals were the catalyst which made the technical effort at Los Alamos, New Mexico, the success that it was. The project mandate and schedule were of a scale and complexity never before attempted.

The original charter behind the Manhattan Project was to save the world from a potentially nuclear Germany during World War 2 by developing a weapon first. Albert Einstein, pacifist though he was, nevertheless signed that famous letter to Franklin Roosevelt in 1939 which hinted at the possibility that Germany might attempt to develop a nuclear device; his signature on the drafted letter reflected Einstein’s concern about Hitler and served to raise Washington’s awareness of the potential problem. Other than his purely scientific role in 1905 in formulating the basis of nuclear energy, the famous equation e = mc2, Einstein had no further role in nuclear weapons development. The day before Pearl Harbor, Roosevelt approved the Manhattan Project.

In Charge and Accountable!
General Leslie Groves and Robert Oppenheimer

The government’s point-man in-charge was one Leslie Groves, a general from the Army Corps of Engineers. History proved him a fortuitous choice in many ways. His most important and telling decision was one of his first, the appointment of a brilliant Berkeley physics professor to be the technical point-man for the entire scientific/engineering effort – reporting directly to Groves. His name: Robert Oppenheimer, a certified liberal whose close circle included ties to the communist party – not exactly the choice a top-secret program would espouse during an era when the motives of Russia as an ally were highly suspect.

Despite protests from those in charge of government security, Groves had a hunch about Oppenheimer and pushed through the appointment – a most improbable one at the time. Oppenheimer knew physics, no doubt, but he had no management experience. Program management ability and brilliance in nuclear physics are not common bed-fellows, to be sure, but Groves had the management foresight and the people-skills necessary to recognize that Oppenheimer was not to be so easily pigeon-holed by conventional wisdom.

Can One Person in Charge and Accountable Forge
the Success of a Major Program? The Answer is Yes.

Quite likely, there has never been anyone so “in the right place at the right time” as Robert Oppenheimer. Imagine being put in charge of a huge, all-out scientific program with dire national security implications. Imagine having to start from scratch by recruiting numerous Nobel Laureates in Physics and Chemistry and relocating them and their families from idyllic college campuses across the United States (and abroad) to the empty high-desert of Los Alamos, New Mexico. Imagine trying to make a practical “gadget” (as the bomb was called) which depended upon scientific findings only discovered within the previous few years. Imagine the immense pressure of staying ahead of often ego-centric Nobel Laureates and earning their respect while coordinating their efforts over several major technical disciplines. Finally, imagine the pressure of possibly failing – producing a “dud” or perhaps nothing at all while ever-mindful that Germany might get there first and threaten world-dominance.

In the end, Germany did not succeed, and the very first and only test shot of the “gadget” in July of 1945 was a complete success. Japan would experience first-hand the next two detonations, unfortunately. Whatever one’s feelings are about the use of these weapons, they did abruptly end the war while illustrating that mankind should never again use them. The program to develop the bomb was an undeniable management success, and that is the central point to be made, here.

The other pertinent point is this: Many who were at Los Alamos subsequently said after the war that the whole endeavor could not have happened without Oppenheimer at the helm…and they sincerely meant it.

This relatively young Berkeley Professor of physics, although immersed among Nobel Laureates who were some of the world’s most eminent theoretical scientists, reputedly had the quickest mind of them all. I recall listening to the venerable physics laureate, Hans Bethe, recount, after working under Oppenheimer at Los Alamos, that he had never met anyone quite so quick-of-mind, so able to quickly assimilate complex ideas and information as Oppenheimer. Most who were there agreed that he was deeply involved in the scientific and engineering work that was conducted during the four year effort, much of it at a detailed technical level, both as overseer and contributor.

Incredibly, this brilliant physicist, a scholar who spoke multiple languages and reveled in 16th century French poetry, morphed into an equally capable administrator, able to orchestrate the entire technical effort. Despite Oppenheimer’s various faults, he turned out to be a superb but initially most improbable choice by General Groves to lead the technical effort. By way of comparison:

Who was the Point-Person Accountable for Coordinating
the ACA Website Development? Apparently No
One with the Necessary Authority and/or Capability.

Despite the relative complexity of the ACA website, it is not rocket science, and it certainly is not nuclear weapons technology as was the Manhattan Project. Neither did it require such a rare duo such as General Groves and Robert Oppenheimer to insure a successful, timely result.

What it did require was a software development guru/manager who, like Oppenheimer in the world of physics, had been there before and understood the work at eye-level – one with a complete understanding of how to manage individual contractors while melding their contributions seamlessly into a working site. An effective coordinator must possess the widespread authority necessary to make large and inter-related program decisions, largely free from bureaucratic gridlock. The list of qualified candidates almost assuredly could not have included anyone in the Washington ranks.

When Groves had a technical question or issue, he went to Oppenheimer. Oppenheimer rarely had to go hunting for the answer among the Nobel Laureates; he knew who was doing what, he understood the details, and he could articulately explain it all to a technical layperson like Groves who then effectively interfaced with Washington.

Not Every Race-Horse Can Be Secretariat, but At Least
Find A Horse Who Can Cover the Ground!

Someone like Oppenheimer is a rare breed, like the great race-horse Secretariat, so I cite him as the quintessential, extreme example of how the right person in management can directly induce success, even in an undertaking of enormous proportions like the Manhattan Project.

Surely someone with solid personal credentials and a proven track-record in website development could have been separately contracted by the administration to be their independent point-person and chartered to coordinate the various vendors who each claim to have produced modules which work just fine …therefore  the problem is someone else’s responsibility! A person with the required competence imbued with the necessary authority to make decisions could have assumed total responsibility and prevented many of the mistakes which now appear evident.

Embarrassing, and not acceptable at any level! Washington should have known better than to depend on its bureaucrats to indirectly “manage” a project of this complexity! The message to them? Review the history of successes like the Manhattan Project and NASA’s space program, and re-learn how to manage government programs. It is not nuclear physics or even rocket science! It is actually just common-sense.

Postscript: I highly recommend the following DVD documentary on Robert Oppenheimer and the Manhattan Project, The Day After Trinity. It presents the history of the Manhattan Project, one of mankind’s most remarkable and controversial achievements, while spotlighting the fascinating persona of the main protagonist, Robert Oppenheimer. My sole personal motive for citing this documentary, here, is to share with you my opinion that it is the most interesting and thought-provoking story in my extensive documentary collection. Those who relate the story in the film were participants – there when it all happened.

Day After Trinity

4 thoughts on “The Affordable Care Act Website Disaster: A Management Lesson from the Manhattan Project

  1. After we talked about this the other night, I had a few more thoughts from my years as a software programmer and as a frequent consumer of web services:

    1) While web service development is “not rocket science,” I’m not sure that it is easy to compare the two problem spaces. Many web companies live in fear of a huge surge of traffic. They would much rather ramp up through a steady increase. Many have gone down when a new wave of traffic crests. It’s a little hard to think of good parallels in other areas of science.

    2) I’ve had to simulate user data and server traffic, and no simulation matches real world conditions. Think of the problem: if you expect 100,000 simultaneous users, how do you test that load? You have to set up machines to simulate traffic. How many machines can you reasonably bring to the problem? 1000? If so, each machine has to be tasked to simulate 100 users at the same time. No matter what you do, that simulation will not match 100,000 users on separate machines. Now, what if the number of users spikes to 1,000,000. I’m just trying to make clear the complexity.

    3) Finally, I’m trying to think whether this is an issue of government incompetence, or an issue of the problem space that goes beyond a public vs. private sector issue. Clearly anecdotal, but the private sector example that came to mind was Google’s problems selling Google I/O tickets. I think it is fair to see Google as a company that knows a thing or two about handling massive web traffic. Yet, when it comes to selling tickets for their annual I/O conference, people experience long delays, timeouts, and failures that force them out of the process before the end. All this to sell ~6,000 tickets over the space of an hour or two (i.e., they don’t need to sustain this over a long period of time). Google may have learned from this problem, but it has persisted over multiple years. It has also plagued the rollout of some of Google Nexus devices.

    I actually think that what the government should have done was to release the service on October 1 as an alpha or a beta. They should have come with a warning that it would not reach 1.0 status until it had been fully tested under load for 2-3 months. After all, Google kept Gmail in beta for five years. Set expectations low and exceed them instead of setting them high and falling far below.

    • Thanks for the website/software perspective on this, Scott. Sounds like load-simulation of heavy traffic is not a straight-forward matter, but it is a problem that is well-recognized by competent designers, hence the prudent, gradual approach to website introduction. It seems that those involved with the ACA site either did not appreciate this or chose to ignore it; I’m not sure which of the two is worse!

  2. Hard not to be political here. This disastrous failure is a worrisome portent of the future mess of health care in government hands.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s