
Photo by Manuel Balce Ceneta, Associated Press
Tim O'Reilly has offered this reply, via e-mail, to my critique of his government-as-platform idea:
You make a thought-provoking critique, but I don't think you really
understand what I mean by "government as platform."
A good platform is more than a foundation. It's also a set of rules. A
good platform provides security, predictable (but not straitjacketed)
user experience, garbage collection :-), protection against bad
actors, and recovery from catastrophic failure.
I'm not calling for government to be a "mere foundation," but an
active partner whose goal is to provide fundamental services while
sparking additional services from the private sector.
I find the iPhone app store versus the traditional cell phone
experience a great analogy. Apple has figured out how to harness the
developer community to build way more services (apps) than they
themselves could or would ever provide, while maintaining control over
user experience, still staying responsible for usability, security
etc. Sometimes Apple is too controlling, even. Being a platform
doesn't mean abdicating responsibility; it means thinking differently
about how to use it most effectively.
I don't think it's easy to see all the implications of the analogy,
but it's a productive one to pursue.
I'd like to thank him for taking the time to reply and further detail his ideas. Whatever criticisms or concerns I may have, Government 2.0 is opening up a vital discussion that's been too long ignored by both politicians and technologists alike. The more minds — from whatever background — that we can get thinking about what to do about our governmental systems breaking down, the better.
He also makes a good point. In retrospect I read too much into his remarks about returning government to "a lean platform for collective action." Both from the explanation above and from further reading of O'Reilly's thoughts on civic action, it's clear that his ideas do not favor as drastic a scaling back as I implied. For that misunderstanding, I owe both he and my readers a sincere apology.
The iPhone analogy is apt, as it implies a striving for balance between individual innovation and a larger structure to check its excesses.
Nonetheless, I'm still concerned about Government 2.0 not accounting for the venality of politics (possibly my own bias speaking here). I spent part of today listening to talks by O'Reilly and others, featured on the Gov. 2.0 homepage. During those, I got some of the same feeling as when I read the original article: rapidly alternating between cheering (hackathons on legislator records? Yes!) and wincing.
There's more to the efforts to improve governmental information (and making that info widely available) than I'd originally thought. It's genuinely exciting stuff, and if successful, would be a godsend for everyone from embattled hard news journalists like myself to concerned citizens such as you, dear reader.
Holding open contests like Apps for America is an excellent idea; something that finally begins to break up the isolation that has hampered governmental services for so long.
A good example of my concerns, though, is Gunnar Hallekson's talk on applying open source principles to government. There's great stuff in there, like his advice that government too, is made up of plenty of smart people and it's thus better to work with them than simply assume they'll screw everything up.
However, he also posits that government programs would be less expensive and more successful if they were less afraid to fail. This is true. Yet no politician is going to want to explain "well we threw $1 million at designing this and it failed utterly, but it's part of a process" to their constituents come election time, and any opponent worth their salt is going to bludgeon them badly with it. Not fairly, no, but dealing with unfair attacks is an essential part of politics.
It's worth keeping in mind that this is because politics is what arises not when we agree (that a needed road repair should happen, to use O'Reilly's example) but when we don't. That's where I think the iPhone analogy has its limits: people don't see eye to eye on which ills should be solved, where roads should be built and through which neighborhoods.
Unlike competing Apps, in those cases it's impossible to pursue both options and see which one is more effective. Laws ban or restrict things or they don't. Projects get built or they don't. Those conflicts are the heart of politics and won't disappear, even if government tries to spark innovation in those areas by active partnerships with private entities.
Solving any social ill is going to require not just getting the "right people" at the table, it's going to mean having to deal with those who may be wrong, but simply have power, and naturally don't think they're wrong at all. If necessary, it's going to require building coalitions and clout to overcome them.
One thing I hope changes is the perception that this, due to the Obama administration inviting more open sourcers to the table, is the big opening before electoral politics takes over. It shouldn't be. It can't be.
Truly changing government for the better will be a long struggle and to be successful, it will have to eventually become the new political consensus, not the hallmark of any single administration. Part of that will be learning to deal with electoral politics directly as an inevitable part of making those changes stick.
Perhaps the largest concern of mine arose in remarks about using data to design better prisons. By itself, there's nothing wrong with this, but it reveals a blind spot: the question of why we have so many prisons reveals far more pressing issues.
In some ways this wave of technologists reminds one of the technocrats who flooded into government in the Kennedy and Johnson eras, eager to apply modern statistics to solve a variety of problems. That worked spectacularly in some areas, but it also proved a test case in hubris, as well as the limits of data and raw intellect to solve problems for which there are no clear metrics.
That doesn't have to happen this time, but it should serve as a cautionary tale going forward. It's good advice for all of us to be aware of unintended consequences and the limits of our own perspectives.
P.S.- O'Reilly had originally tried to send in his remarks as a comment to my previous post, but was denied by Typepad. I'm trying to get this issue worked out now. If anyone else is having similar troubles, let me know via e-mail.