Tuesday, October 29, 2013

Healthcare.gov as a Case Study for the Analog Digital Divide

by Kent AitkenRSS / cpsrenewalFacebook / cpsrenewalLinkedIn / Kent Aitkentwitter / kentdaitkengovloop / KentAitken

I find myself thinking about the launch of Healthcare.gov with a bizarre frequency. I think it’s fascinating as a case study and, I believe, emblematic of a broader unaddressed problem.

The Long Story Short

Healthcare.gov launched on October 1, 2013 to serve as a hub for US citizens seeking health insurance in those states that didn’t set up their own systems in response to the Affordable Care Act. It has been fraught with technical problems, rendering the website nearly useless at times. It appears that the cost will end up being $170M-$300M, double or triple original estimates.

David Eaves weighed in on procurement in general on Friday, suggesting that procurement was borked: that embedding core principles such as openness and transparency in government systems is at odds with how we currently develop and procure those systems.

He pointed to Clay Johnson's piece on Healthcare.gov in the New York Times,
Why the Government Never Gets Tech Right. Johnson also lampooned current procurement models, and noted the dismal (~6%) success rates for large IT projects.

Alex Howard's assessment of the failed launch mirrors Johnson’s: the current procurement model rewards companies that are good at bidding, not building. Their US examples find support elsewhere: in the UK, "70% of IT spending between 1997 and 2010 went to just seven (!) companies."

Proffered Solutions

Eaves, Johnson, Howard, and a few others - Mark Headd, CIO of Philadelphia and Merici Vinton, part of the stunning digital success story of CFPB - call for wholesale procurement reform. Vinton goes one further and calls for hiring reform as well, to allow "in house strategy, design, and tech", which is one of the other common threads: Howard and Vinton laud initiatives such as the Presidential Innovation Fellows and entrepreneurs-in-residence, and Headd calls for more "makers and hackers" in government as the path forward.

Johnson's take:
"The president should use the power of the White House to end all large information technology purchases, and instead give his administration’s accomplished technologists the ability to work with agencies to make the right decisions, increase adoption of modern, incremental software development practices, like a popular one called Agile."

Johnson and Howard noted the dissonance between Obama's massively successful data-driven and digitally savvy campaign and this IT flop. Two of the key figures in the campaign's success - Daniel Wagner and Teddy Goff - were in Ottawa last week with Canada 2020, and their most striking message at that event applies here: science is your friend.

Agile methodology that Johnson referenced is broad, but a key component is the scientific method: hypothesizing, testing, and adapting. Howard noted that Healthcare.gov's front end, built with Agile methodology and open source, works well.

Building on that, one of the major themes that these authors’ admonishments drive at is that we shouldn't pretend that we can establish clear requirements for IT at a singular point in a procurement cycle. Flexibility and in-house talent are essential for managing a world in which that assumption, rightfully, breaks down.

In sum, the messaging is this: don't invest in highways with no exits. (And make sure you have someone to tell you when you need to use them.)

The Digital Analog Divide

Healthcare.gov is a cautionary tale for government’s major IT projects, but I think it’s also a case study for more fundamental issues. There are several possible lessons, but principally I think it’s a symptom of an awkward, incomplete transition from the analog world to the digital one.

Many companies
failed to make the transition, and the Fortune 500 list looks very different now because of it. Government faced no such competition: it’s still standing, applying a procurement model that worked for furniture - with 20-40 year lifespans - to technology that becomes economically obsolete at least ten times quicker.

So the questions I'm bandying about, and will return to next week:
  • Do governments have the appropriate foundations for a technology-enabled world?
  • Have governments reached the limits to outsourcing?
  • To what extent are procurement and hiring models influencing governments' effectiveness?

No comments:

Post a Comment