Oh, by the way: The "tech surge" isn't going to rescue the ObamaCare website

Or, to be more precise, it’s not going to rescue the site anywhere near as quickly as the White House needs to avert disaster. The task is simply too complex, and Healthcare.gov’s problems too various, to get things humming smoothly within the three-to-four-week timeframe HHS is now realistically staring at. Yuval Levin claimed in a post last night that the “tech surge” is actually a waste of time because everyone already working on the site will have to slow down while the “best and brightest” get up to speed on the software and the scope of the challenges they’re dealing with. My layman’s sense after reading several articles about it and chatting with a friend who works in IT is that it’s like bringing in a bunch of professional writers to rework a 2,000-page first draft of a novel written by a sixth-grader. If you don’t like that analogy, a health-industry consultant told Politico, “It’s like you’ve built a single-story house with no basement, and then the family comes in and says, ‘Oh, we made a mistake, we wanted a multi-story house with a basement.’ Then you have to strip everything down and start over.” It can be done, but it ain’t going to be soon. And it ain’t going to be cheap.

Your must-read on the futility of the “surge” comes from VentureBeat.

The government can’t accept free services from companies, so HHS would need to find additional budget to pay for help.

Another challenge is that CGI, Quality Software Solutions, Booz Allen, and the rest of the contractors would need to agree to open up their code. It’s a huge project — the New York Times quotes one specialist who said that 5 million lines of code would need to be rewritten before the website can function properly.

“That’s the problem here. Even if the government recruits talented engineers from consumer tech companies, whoever builds this site has to take care of it,” said Kocher. “I don’t see this ‘best and brightest’ idea working in practice.”

And the problems with the exchange might not be obvious to a smart Silicon Valley programmer from an Apple or Google. This kind of a project is complex and requires deep domain expertise. As John Engates, the chief technology officer for cloud company Rackspace, put it, “I think the downfall of HealthCare.gov is that they need to integrate with numerous legacy databases: IRS, DHS, Social Security, and myriad state agencies. Any one of those running slow could be the bottleneck.”

One expert told VB that the best thing to do would be to hire a small team (with what money?) and rebuild the site from scratch, rolling it out early next year some time. (Isn’t that too late?) Another thinks the Presidential Innovation Fellows should be enlisted since there’s already money appropriated for them, but it’s unclear if the Fellows have the technical expertise to solve this. Yet another thinks the market will solve this problem as private companies clamor to fill the vacuum created by Healthcare.gov, but who knows how long that’ll take — and how they’ll calculate the all-important subsidies lower-income enrollees are now entitled to. No one, though, thinks a “tech surge” is the missing piece of the puzzle in the short-term. “The system is already built; it’s already a disaster; so you’re going to bring in Silicon Valley to do what?” said Netscape founder Marc Andreessen to Politico. “You could get Mark Zuckerberg, Larry Page and Jeff Bezos on a plane, but what are they going to do at this point?” Which raises the question: Does the White House know that the “tech surge” is largely a waste of time, a contrivance they’re using to placate Americans and stall while they figure out what to do next? Or is the depth of technological ignorance here such that they really do think Jeff Bezos and Mark Zuckerberg could whip this thing into shape in a few weeks?

There’s an irony to the tech surge too, says Andreessen:

“The government has accreted over the last 30 years contracting rules and regulations that make it impossible” for Silicon Valley companies to compete in Washington, he said. “So you’ve wired the system basically to prohibit the people who really know what’s going on.”

That’s one degree of self-imposed difficulty. The second degree was the extraordinary secrecy about the building of the site maintained by the White House and HHS because they were terrified of political embarrassment if their opponents found out how badly things were going. Had they shared more info with private insurers sooner, they might have had useful feedback on how to correct the site’s problems before they reached this point. The third degree of self-imposed difficulty was the fateful decision to force people to create an account on the site in order to comparison shop among plans, a choice that was made possibly because the administration was afraid of “rate shock” scaring away consumers if they weren’t informed upfront about their subsidies. Who made that decision? According to Issa’s House Oversight Committee, sure sounds like it was the White House (“CMS” is the department of HHS that’s overseeing the website):

CGI officials told Committee staff that CMS officials and employees constantly mentioned “The White House” when discussing matters with CGI. For example, CMS officials would routinely state: “this is what the White House wants.” Moreover, CGI officials told Committee staff the ability to shop for health insurance without registering for an account – a central design feature of the health insurance exchange – was removed “in late August or early September.”

They spent years building a site that would let you comparison shop without creating an account, just like every other commercial website does, and then at the very last minute they were told to tear that up by the roots and stick all the plan information behind a reg wall. Why? And who’s the dummy who thought a fix like that would be simple? Computerworld has a story today noting that less than seven percent of IT projects that cost $10 million or more are “successful” in the sense that they’re neither over budget (which Healthcare.gov already is), behind schedule, nor below user expectations. Given how formidable it evidently is to build a mega-site like this, why would they have waited until less than a month before launch to reorganize major parts of it?

The final irony here, by the way, will be if the “tech surge” manages to solve the relatively easy problem of how to create a valid user account before it solves the more daunting problem of calculating subsidies or sharing accurate information about the user with insurance companies. That would enable a huge spike in the volume of enrollments before the system is able to process those enrollments correctly; suddenly you’ve got hundreds of thousands of people in the queue instead of tens of thousands right now. In that sense, it’s oddly analogous to delaying the mandate while keeping the rest of the law up and running, which risks an industry death spiral. Either the whole thing runs smoothly or the whole thing should be taken offline until it can run smoothly. Partial solutions are the surest path to the worst-case scenario.