Skip to main content

View Diary: Sebelius resigning as Health Secretary (125 comments)

Comment Preferences

  •  I bet she believed things weren't getting done (2+ / 0-)
    Recommended by:
    commonmass, raptavio

    as she expected > from her chain of command.  The administration didn't want to stir the pot.

    •  I bet she was asked to resign after the (6+ / 0-)

      disastrous rollout, after the signup deadline. She really blew it with the rollout.

      •  No start up system is perfect. If we begin to (19+ / 0-)

        accept everything as failure and assume heads have to come off just because the media is trying to fill its empty columns, we failed in common sense, judgement and leadership.  Less than 4 weeks was nothing compared to getting through to other organizations.  It was bound to get fixed very soon.  This was something new and never tried before.  I do believe her team should have investigated the handling of possible overload issues and the designer should reduced their price.

        •  Agree. Need to look no further than (3+ / 0-)
          Recommended by:
          wader, janemas, Loose Fur

          Medicare Part D.

        •  I think you're being way too charitible. (5+ / 0-)

          Of course no modestly complex system is perfect at release (or ever). And, yes, HealthCare.gov is somewhat complex and there were changing requirements (as there always are).

          But it really isn't a high volume or transactionally complex system by today's standards so that's no excuse. The amount of user data to be managed and tracked is ridiculously small. The transaction load is not very high. One user's actions don't interact with another user's actions -- there is simply no interaction between users. If I sign up for plan A, the system doesn't have to exclude another user from doing so 100 msec later. The database of information necessary to process a user's requests is not changing by the minute. Plan pricing, features, and options are not being rejiggered to optimize profit dynamically in real time.

          Very basic things didn't work reliably that would "just work" if the system had been designed, implemented, and built to scale and subsequently tested at scale. It obviously wasn't.

          Registration, for example, just isn't that difficult a problem even in HealthCare.gov and any competent designer would have gotten it nearly perfect - there are only tens of millions of people who would ever create a user identity. A skilled design and implementation team with would actually have to work hard to make it not work under load -- intentionally going against their instincts and experience regularly just to create a monster.

          Issues like losing, misplacing  people's pending signups or relegating them to a state of permanent purgatory are also just ludicrous.

          If "load" problems lose, misplace, or corrupt data, the problem is not the load, it's the fundamental specification, design and/or implementation.

          It would be more understandable if high loads cause session timeouts and make the system sluggish but still unacceptable unless the load exceeds the requirements (which obviously should have been clearly specified and agreed to by both the government and the contractors - that's the job of the customer to spec). Any spec for a system such as HealthCare.gov whose load may be difficult to anticipate with absolute certainty should have included requirements regarding how the system would degrade when the load exceeded the maximum specified load per server/system.

          It also would have even been understandable if the problems had been things like "The only plan in County X in State Y just withdrew and the screen residents of that county see is nonsensical because no one thought that would happen because of course every citizen will have at least one plan available to them".

          If I had ever even suggested even releasing something of the quality of HealthCare.gov, I would have lost my job every time.

          This was clearly a complete management screwup and Sebelius was responsible for it. Sure, she probably has no idea how to spec a computer based system or how to test it, but, as a manager, she should understand basic project management and hire the right experts. If she is not accountable, who is? Surely a President should be able to delegate a modest task such as this to a cabinet member and expect it to get done without reviewing detailed development milestones and running regular project audits.

          The only reason I can think of that some people seem to think that HealthCare.gov was "really hard" was because it failed so miserably. If it had worked well, I'm pretty sure we wouldn't have heard people gasping in amazement "Wow, no one ever thought that could work, yet it does - that's among the most amazing thing I've ever seen".

        •  she lacks managerial and executive skills (2+ / 0-)
          Recommended by:
          AverageJoe42, annecros

          it's got nothing to do with the task and everything to do with her

      •  They couldn't have filled her spot until Nov. 21st (2+ / 0-)
        Recommended by:
        wader, Heavy Mettle

        (nuclear option day) that's a pretty steep learning curve for a Secretary for a March 31 deadline.  So they waited a little bit because an April 1st sacking would have been trashy.

        She either was pushed or jumped so she wouldn't be.

        But a resignation between enrollment periods makes sense.

        I'll always be...King of Bain...I'll always be...King of Bain

        by AZphilosopher on Thu Apr 10, 2014 at 05:26:59 PM PDT

        [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site