For better OGD compliance, don't force it. Sell it.
BY Matthew Burton | Wednesday, April 14 2010
Many of us are disappointed by last week's Open Government Directive reports, particularly as they concern the release of data. I think it's pretty obvious what happened.
The agencies with the most impressive data plans--HHS, NASA, Department of Education, National Archives, OPM, OMB--are either research- or data-driven. They already collect, analyze or manage data as part of their core mission. Some other agencies that seemingly should thrive in this area--like the Departments of Labor, Treasury and Veterans Affairs--issued lackluster reports. I don't know why. Maybe they aren't as data-driven as they should be. Or maybe they just dropped the ball.
But for the Departments of Defense, State and others, working with structured data is probably not part of their routine business. The art of diplomacy and war are difficult to think of as data sets. So simply telling these organizations, "Show us all your data," probably left them befuddled. What data? They are institutionally unaware of their own knowledge.
On top of not having standardized data sets to distribute, they likely also lack the personnel to write a serious report on it. The OGD was probably handed to people whose official job descriptions ask them to do other things, whose performance appraisals grade them on work completely unrelated to the OGD. This means the OGD's requirements became just another side project that got attention during rare instances of spare time.
But the President asked them to do this for the benefit of the American public. Is that not sufficient authority to work full time on this? Is the cause not great enough?
First, while the President is technically in charge of the Executive Branch, he doesn't exactly control it. Yes, he appoints the Cabinet, which in turn oversees each department. But in practice, these bureaucracies have a mind of their own. Arthur Schlesinger explains this perfectly:
As the functions of the national government multiplied, a bureaucracy developed dominated by vested interest of its own--vested interests in ideas, in procedures, in institutions. The American government really has today four, not three, coordinate branches: the legislative, the judiciary, the executive and the presidency; and an active President would encounter as much resistance within the executive branch as he would from Congress or the Supreme Court. The increase in the size of the bureaucracy created a split between the "political government" and the "permanent government"; and many members of the bureaucracy exuded the feeling that Presidents come and Presidents go, but they (bureaucrats) go on forever. The problem of moving forward was in great part the problem of making the permanent government responsive to the policies of the political government...The result is that the bureaucracy has an infinite capacity to dilute, delay, obstruct, resist and sabotage presidential purposes. (Journals, Arthur Schlesinger, page 164)
Competing with the President's goals--and the secretaries, deputy secretaries and undersecretaries he appoints to achieve them--is a decades-old organizational culture that dictates how work is done. The department knows that it will be around long after President Obama is gone. While the OGD is one of his administration's core projects, in the context of the department's entire history, the OGD is just a blip. To desk workers, it was likely just another mass email that got a moment's glance before being trashed; there's actual work to be done.
Which leads to the second issue: no, public transparency is not a great enough cause to overcome bureaucratic inertia, because the bureaucracy simply doesn't care about that sort of thing. Cabinet members might, as might other people throughout the government's ranks. But transparency of operations is not a department's reason for being. Transparency is just an ideal companion to the successful performance of the department's real mission. And while that mission is intended to serve the public interest, at the ground level, the public rarely enters the picture. In practice, every government employee is charged with making some other employee happy. A diplomat answers to the ambassador or consulate general. An intelligence analyst works for the policymaker and the warfighter. A systems administrator makes sure the network stays up. If these people perform their jobs transparently and the data is distributed publicly, that's fantastic. But it doesn't make their customers any happier or their performance appraisals any more glowing. It only means they did it transparently. Transforming the bureaucracy into one that intrinsically cares about public transparency will take decades, not months.
How, then, can departments and agencies be convinced to share their data? First, they have to start collecting it, and the only way they will start is if they can be shown how it will further their mission, not just the President's. They have to want to do it.
While data isn't intrinsic to, say, diplomacy, any business process can be improved through data collection and analysis. That is how the OGD should pitch data furnishing to agencies: as a mechanism for streamlining internal business. By collecting data about their own operations and results, they can make better policy and budgetary decisions, which will lead to better performance and happier customers. So instead of asking agencies to publish data in order to "increase accountability, promote informed participation by the public, and create economic opportunity," ask them to catalog their business processes: how many visas were applied for last year? Which consulates had the most applicants per officer? How many visas were granted? Revoked? How many missing passport claims were filed at each consulate last month? This month? Year over year? For the military, this kind of business analysis would do wonders for anticipating threats. While many of it would not be releasable, it would still create the requisite culture and infrastructure for data collection, paving the way for public data distribution.
Government data is a public asset. But it is also a bureaucratic one. Framing it as the latter is much more likely to pique the bureaucracy's interest and earn its cooperation. It would make not just for a more transparent government, but for a more optimized one as well.