Last month the Obama administration announced an executive order mandating an open, machine-readable default setting for government datasets, a move lauded by (among many others) open tech writer Alex Howard, whose piece in Slate was headlined “The Best Thing Obama’s Done This Month.” This is indeed an important move, and good policy for all the reasons Howard points out. It is also in tune with the times, as it is at this point a truism to note that the oft-related concepts of “open data” and “open government” are hot topics. Analysts and crusaders declare that transparency and information-sharing constitute the elusive geist of our zeit. Opening government data – that is, releasing it in machine-readable format online –is seen as holding the key to a host of goods, from better-performing governments to more efficacious health-care to wealth-creation and prosperity.
But Obama’s executive order, while a fantastic step forward in designing policies for transparency in government data, is no more than that: one step.
It is appropriate to note here that “open government” and “open data” – related, but not the same thing – are concepts very much in need of unpacking, in more ways than one. Scholars are doing good work to nail down and demystify these ideas. A prime example is the decoupling noted at the beginning of this paragraph. Yeoman work toward that end has done by Harlan Yu and David G. Robinson in what deserves to be a seminal piece in a fairly new field, “The New Ambiguity of ‘Open Government.’” They note that the idea of “open government” has moved away from denoting the provision of politically sensitive information, information enhancing accountability in democratic regimes, and damaging to authoritarian ones. Now “a regime can call itself ‘open’ if it builds the right kind of website – even if it does not become more accountable or transparent.” This is all quite true – these phenomena are known anathema to activists who work tirelessly to counteract them – and this awareness is fast seeping into the literature.
But while acknowledging that the ontological contours of open data and open government, and their intersections, are becoming clearer only as they’re contested and worked out, there are other, normative and policy, axes on which these concepts urgently need to be positioned, and in this lies the implicit critique of Obama’s executive order.
To wit: is it more important that relatively innocuous, but potentially very useful (and monetize-able) information be made widely available? Or should considerations of government transparency and “openness” to the public be treated as paramount? This is part of the argument taken up by Yu and Robinson, and by others, and I argue that these represent related, but separate, considerations, with related – but separate – kinds of data appended to each.
The first stance mentioned above emphasizes the packaging and diffusion of government-collected data sets, with a focus on data that are of immediate use to consumers in their daily lives, e.g. bus schedules made accessible from a smartphone. As with meteorological and GPS information, “health information and consumer finance” datasets can be parsed by analysts and packaged by entrepreneurs into apps and other products that people will want to buy because of their usefulness. This government data, opened, has the potential to spur a wave of innovation in fields as diverse as health care, weather forecasting, and risk analysis (and sometimes all at once!).
The other kind of open government information is meant to increase government accountability to its citizens. Much of this information, as Jim Harper of Cato points up in this excellent piece, is essentially procedural. Who does what in government, and do they comply with laws and norms meant to govern public life? And do they do it in a way that enables informed deliberation and public participation? But it goes well beyond process as well. Government spending data is a powerful tool for the empowerment of citizens to hold their governments to account, deepening democracy and leading to better outcomes for the expenditure of public monies. This is why, for example, the International Budget Partnership works tirelessly around the globe for the release of government spending data (disclaimer: this tireless organization is a former employer of this author).
The answer to the question of relative importance should be easy: both are vital functions. But the temptation is to act as though open data for business use is the same as open data for citizen use and government accountability, or that opening government data for entrepreneurial innovation will drive the same suite of public goods as would opening government data for government accountability. This is patently not the case, and government functionaries, analysts, and activists working at the multiple intersections of (open) (government) data with public life and the business sphere must resist the temptation to conflate the two, or treat them as substitutable goods, or to simply ignore open citizen-user data in favor of open enterprise-user data.
This last tendency, in the form of privileging open data its usefulness to the business community, predominates in much of the public discourse on open data. For example, the venerable patron saint of borderless capitalism and oft-numinous innovation, writing in the New York Times recently proclaimed that “…one area where [the Affordable Care Act is] surprising on the upside. And that is the number of health care information start-ups it’s spurring.” That is to say, a law whose most important features were supposed to be about justice – that is, making health care available to those without – and public finance in the form of “bending the curve” is treated as though it were actually about knowledge industry entrepreneurs creating new products.
Now, knowledge industry entrepreneurs create important benefits, and no mistake. Prosperity and growth are obviously important. So are the benefits provided by innovative health-care startups like Friedman’s example of “Lumeris, which does health care analytics that uses real-time data about every aspect of a patient’s care, to improve medical decision-making, collaboration and cost-saving.” The work of Todd Park (to take only one example) in spurring these kinds of innovations is nothing short of inspiring, and holds tremendous promise for the future. But we should not act as though Park’s work is simply about empowering the private sector to save the public, and make a bundle doing so. Indeed, this devalues the work of Park, and policy figures like him, whose work is about policy interventions to advance social justice. Indeed, big open public data may well open avenues for policy interventions to fix health market anomalies that are very much an issue of social justice, and sustainable government fiscal policy.
It is vital, then, that advocates for open government data not forget notions of justice and citizen accountability when they (we!) talk about openness. This means not just the release of government weather data; it means the release of government spending data. It’s not just about allowing entrepreneurs to add value to data, enriching themselves and the lives of those who use their apps; it’s about allowing citizens to hold their representatives to account for decisions made about their (taxed) money.
I would not devalue the importance of open data with an emphasis on “adaptability,” machine-readability, and open licensing. It’s just that the ability to remix public data, which is a huge deal because it allows software engineers the freedom to turn datasets into bus-tracking apps and health-cost-tracking apps, is a concomitantly big deal because it allows activists to remix government spending data into presentations that inform their fellow citizens about government functioning. These components of open government data are both important, and they have to both be treated as such, rather than ignoring the transparency-for-accountability aspect in favor of open-data-for-entrepreneurs.
To bring this back to Alex Howard, and the executive order that sparked this essay in the first place: a higher standard for the release of government data may well be the best thing President Obama did in May. But it’s June now, and further progress on opening government data – not just for entrepreneurs, but also for citizens; not just for prosperity, but also for justice, accountability, and deepening democracy – is imperative.