On Government Documents Management
Building on my earlier post about listing collaborative options for government or congressional agencies, I’m thinking about useful ways to distinguish between different types of government information, and what that implies about records management.
At the recent IPDI Politics Online panel on radical transparency, Peggy Garvin made a great point about one fundamental distinction that can be made within government information. She suggested that all government information is either collected from regulated entities, or pertains to the operations of government itself. (much more below)
Other fundamental distinctions could be made about government information generally; for example, all government information is either classified or not (attempt to equivocate notwithstanding.)
Reflecting on this point, I thought of the various Creative Commons designations to be asserted over works. These distinctions have been developed as technology made use and reuse of (especially) digital information easier by an order of magnitude, which, in turn, rendered previous regulatory systems inadequate both de jure and in spirit. The same could now be said of both privacy law and many privacy policies, whose phrasing is too vague to apply clearly to new digital circumstances. (Technology that changes rapidly can be a real headache for legislative drafting: for example, the line between terms of art and informal slang is fuzzy at best.)
Back to government information, the same thing that happened to music because of mixtapes, and that happened to video because of youtube, is also happening to government information. Regulatory schemes and records management practices need to be rethought.
What would a forward looking government information management system look like? FDSys, obviously, is the new federal documents delivery system, but I’m wondering about the new policy distinctions that are probably on their way.
For example, some documents have enough public import and enough consistent demand that they should be published online as quickly as possible, in as structured a way as possible. Votes in legislatures would fall under this distinction. Another set of documents would be those things that exist under the domain of the government, and maybe are FOIA-able, but are probably best not public. Personnel information and yearly reviews all online would make for a very complex work environment.
I suspect that people at the GSA or OMB spend their time making exactly this sort of documents management distinction for the federal government, and if anyone has public documents or OMB memos to this effect, I’d love to see them. These management practices, however, are probably based on old legislative mandates like the paperwork reduction act or the OMB A-130 circular.
I’m thinking about something more far-reaching, that could proactively assign distinctions to documents with public access in mind, in addition to other concerns like national security, or commercial or private or sensitive information. Another example, which information sets within the governmental (there are more than a few) are appropriate for APIs? Surely there’s a better way to decide on this than on a case-by-case basis. Which documents should have permanent URLs associated for them? If NIST can dictate that congressional districts will be referred to as "AL – 01" (or whatever), then shouldn’t someone make similar statements about government URL/database structures?
If the US gets a CTO, (or, if the structures in place to carry out these functions are expanded), that person or those people should come into the job realizing that they’re not just coordinating technology, but they’re practical epistemologists, determining in practice what can be known about our national government.
(Full Disclosure: Larry Lessig is on Sunlight’s Advisory board, although I had no contact with him about this post.)