IT procurement and HealthCare.gov: Searching for scapegoats?

There are plenty of serious problems with HealthCare.gov, Steve Kelman argues, but the procurement system itself does not appear to be one of them.

contract signing

President Barack Obama has now raised the allegedly backward nature of the federal procurement system as a presidential concern. Speaking to a business audience about the failures of HealthCare.gov, he stated that what the government needed to do is to "blow up how we procure for IT." FCW has just asked on its cover, "Can IT Procurement Be Saved?"

I'm not sure we yet completely know the whole story of the problems with HealthCare.gov, but to me at this point it looks like there were two big problems with the system's development. One was the constantly changing requirements, a bane in federal IT development. In this case, new regulations were regularly coming down the pike as the system was being developed, creating an ever-moving target that receded each time code writers tried to catch up. The second was the failure to assign a systems integrator to be in charge of all of the many moving parts of this very complex system. That reflects a philosophy that began to spread in reaction to problems with giving systems integrators too much power in projects such as the Coast Guard's Deepwater boat development.

Perhaps the biggest problem was a political one. Not unusually for federal IT (and, to be fair much of commercial IT systems development as well), the system was behind schedule.  But this system had a deadline, the Oct. 1 launch of the new healthcare insurance marketplaces. A decision was made -- to my knowledge, we don't exactly know how or by whom -- to go ahead with the launch. How much those making the decision knew about how far from ready it was is, to me at any rate, unclear.

None of this, however, sounds very much like problems with the procurement system.

I am willing to believe that problems with the procurement system itself bear some significant responsibility for the failure of HealthCare.gov. I promise to join with anyone making a good case that the procurement system bears a notable part of the blame in working for good suggestions to fix problems.

But I have not heard many specifics.

For example, I don't think it has been claimed that delays in awarding the task orders to develop the system were of any consequence in the system's unreadiness on Oct. 1.

FCW's recent cover story is generally a balanced piece of work, and it uses the phrase "procurement system" broadly to include poorly developed requirements, technical difficulties merging legacy systems, and a lot of other general features of government IT.

The one element of the procurement system proper the story criticizes, however -- doing the contract using task orders rather than an integrated contract -- doesn't warrant the criticism, I don't think. Task orders are not inconsistent with having an integrator, or even with having only one contractor rather than several. They would be the appropriate contract form to do agile development. (I was surprised to see Trey Hodgkins, whom I respect, quoted as stating that the system should have been developed as a single contract using full and open competition -- a "back to the future" approach that harks back to the 1980s.)

As a general matter, it is probably true that many of the special social requirements the government puts on contractors (which do not exist for companies not doing business with the government) discourage commercial firms from bidding for government work, especially small firms. However, Obama has often supported expansion of such requirements. 

The reputational risk that a mistake on a government contract can cause for a commercial company is also a problem, but that reputational risk is caused by media coverage, and it's hard to see a solution. (Overzealous auditors might increase the risk by turning small mistakes into apparent crimes.) These issues are likely not relevant to HealthCare.gov, however.

Furthermore, the government's past performance system is too kludgy and riddled with excessive "due process" to be the kind of sharp weapon it often is in commercial IT procurement.

But some of the comments, frankly, have an air of looking for scapegoats for the HealthCare.gov failure.

Again, I by no means see myself as somebody with a stake in defending the procurement system status quo. I have constantly promoted innovative procurement methods, such as contests and share-in-savings contracting -- which are both allowed by the current system. If readers want to make some specific suggestions about problems in the procurement system that bear a noticeable responsibility for the HealthCare.gov fiasco, please put them in the "Comment" section of this blog. I'm all ears.