The Obama administration, as of December 8th, 2009, is requiring every government agency to release three high-value data sets to the public within 45 days.

The Sunlight Foundation reports:

In a livestreamed chat, blogpost, and document online, the White House has announced a bold new plan for openness in the Executive Branch. The result of months of interior conversations, and three public components, the new policy introduces sweeping goals and initiatives aimed at bringing citizens closer to their government, through technology, information, and public interaction.

I’ve been impressed by the Obama’s emphasis on opening the door to the raw numbers of government. Data.gov was the first great leap forward. And now this directive to institutionalize his spoken priorities looks promising.

However, I do have one lingering question: how do we ensure accountability when we are allowing the agencies themselves to choose these “high value” data sets? We ought to ask the media, or the general public, to vote for or at least suggest data sets that will be of high value — and let the aggregation of those (filtered) suggestions form the strict directive that agencies must follow.

Allowing the departments themselves to discriminate among what data they choose to release (out of the thousands of data sets within the scope of their work) will almost guarantee that they will release inconsequential data sets so as not to make their department “look bad”. Unless agency leadership carries an ethos of constant improvement and critical self-evaluation, all we will get is the gristle.

If we are not ruthless, we could find ourselves with a dangerous new level of Washington “doublespeak,” where the White House hails a new era of unprecedented transparency, but citizens are only let in on 10% of the data that is actually available from the billions of dollars and many decades of agency operations. Erstwhile we are steered clear from the federal bureaucracy’s most egregious inefficiencies and problems.