I usually avoid making overt political statements but its not a political statement to say that the healthcare.gov rollout has been a disaster. Nobody denies this. My interest in this subject, as far as this column is concerned, is that media coverage of this has brought the process of software engineering into the public eye in a way that it usually isn’t. Everybody is highly familiar with useable software, and none of us are strangers to getting befuddled by poor software, but the political thrust of healthcare.gov’s failure has made this about more than just the droll humor that usually surrounds bad software. The public is, if only temporarily, overtly interested in how bad software happens, (and, if you are less cynical, how good software happens). Given the stated aims of this column, I would be remiss to ignore this buzzing topic.
It’s a testament to the software industry that most products work as well as they do – just like it’s a testament to the construction industry that even the cheapest homes don’t collapse. Conversely, anyone who has been frustrated at an IRS office or DMV or similar such institution isn’t exactly surprised that a government website turned out not to be as slick as Apple or Google’s latest efforts – admitting this much doesn’t require much cynicism about government. Indeed, the best thing to come out of all of this
is the sensible, bipartisan recognition that the federal government doesn’t have the best policies when it comes to procuring technology solutions. The greatest hurdle is one that our government might not be able to overcome, which is that any technology built in the interest of this nation will be disinclined to employ foreigners in its construction.
For what it’s worth, I’m not so sure that U.S. citizenship or security clearance was a requirement of anyone working on healthcare.gov, but such requirements typically hold for most anything the government would spend money on, (and for sensible reasons). This is a huge disadvantage that most software developers don’t have to deal with. In the software world, if you are limited to only Americans, then you are limited to a talent pool about 1/3 third the size normal corporations get to enjoy. One thing the public is largely unaware of is that the production bottleneck of most software firms is attracting and retaining talent. The biggest fight between Google and Microsoft and Apple is the fight for who can hire the most and best engineers. The U.S. government is hopelessly outclassed in this fight because of its inherent need to favor homegrown talent.
What makes this worse is the cultural homogeneity among American engineers can be a further bane on innovation. Recent research has indicated that more diversity among groups actually improves performance and innovation. Software engineering is generally too competitive to ever give way to preferences for any specific ethnicity or country of origin, and workplaces at tech firms are usually very diverse. Federally funded projects, by contrast, can hit a wall given the pool of people who are available and interested in the projects. If this sounds silly to you, consider that the last time our survival depended on winning a tech race, we leaned on a group of European expatriates. Consider that our dominant Olympic teams often feature athletes originally from other countries. Consider that America often outcompetes everyone, and is one of the few nations that is overtly dedicated to being a cultural melting pot. It’s not a coincidence.
So while many many observations are to be made about healthcare.gov’s implementation – and corporate engineers are not exactly hiding their opinions from the media right now – I feel that the larger story is the one we’ve all been side-stepping for years, which is that the speed of our government is significantly slower than the speed of technological growth, and that needs to change.