Not Inexplicable At All

In Peter Suderman’s commentary at Reason.com on the HHS IG’s critique of the process that lead up to the flawed debut of Healthcare.gov, he characterizes the decision to let the contracts for the website on a “cost-reimbursement” basis as “inexplicable”. It’s not inexplicable at all. You’d need to be crazy to accept the contract on a fixed bid.

The basic problem was the policy team, i.e. the agency bureaucrats and regulators. The rules were being made on the fly. There was never a fixed scope for the work and, consequently, any fixed bid would have needed to be so sky-high that nobody in his right mind would have accepted it.

This just highlights the point I’ve made for years. The federal government procurement process makes it very difficult for it to manage IT projects reasonably and the loosey-goosey way the PPACA was written made it even harder.

6 comments… add one
  • Cmstanley Link

    It would seem that you need people on both sides of the table who understand the scope of the work and what it costs to get it done. My husband has managed very large construction projects, on both the construction side as well as the owner/developer side. You would think that the bare minimum would be for bureaucrats overseeing these projects to have some IT people in their ranks who have enough experience of big projects to provide the oversight.

  • That wouldn’t be enough. The politicians and policy folks will not and probably cannot relinquish control of these projects. Frequently, that means that they’re doomed and it always means they’re enormously expensive. And it definitely means they can’t be executed on a fixed bid.

  • TastyBits Link

    You go back to have a scope change, and the original contract is amended with additional funds. In many cases, you underbid the project knowing there will be scope changes. This is especially true if your company is the one they want anyway.

    They would need to have hired an outside company to write the specification, but this would require knowing exactly what they wanted. Once the specification is written, any change can have multiple impacts, and in IT, the technology can change before the project is complete.

    Most of today’s whiz kids do not understand that many of the things they take for granted as “just working” have been debugged for the past 20, 30, or 40 years. They may not be pretty, but they work.

  • in IT, the technology can change before the project is complete

    I’ve mentioned this before myself and it’s one of the reasons that I suspect that any really large government IT project is doomed to failure.

  • TastyBits Link

    Most people do not know how engineering works. Engineers design things to known specifications/limitations and established codes. The American Society of Mechanical Engineers (ASME) established the Boiler & Pressure Vessel Code because of a boiler blowing up, but for years, it was common for boilers to explode.

    (Apparently, free-market competition does not apply to anything steam related. You see my leftist friends. If you pay attention, you can learn how to smash that right wing asshole’s talking point.)

    Each branch of engineering has its own set of published standards, and these are usually required to be used. In some cases, they are required to be followed by law (all or part). The standards are derived by studying accidents. When something goes wrong, a rule is made to avoid it.

    This works great as long as things do not change much, but when things change, accidents increase. New materials, new building techniques, new anything can cause an accident until all the problems are identified, and while testing can find most, it cannot find all of them. The first or second generation of anything is best left to others to work out the bugs.

    For computer technology, much of it is changing too fast for there to be any standardization.

  • Additionally, in computing so much is tied to individual manufacturers (or even cultures) that people are inclined to confuse Microsoftology (or Appleology, Linuxology, etc.) with computer science.

Leave a Comment