Tech-savvy governments are adopting new transparency policies for “Open Data,” broad pronouncements intended to create things like a “presumption of openness,” or to declare a city “Open.”

These declarations are fantastic. They reinforce commitments to transparency, raise public expectations, and give cover to sympathetic government officials who already want to do good work.

Unfortunately, these policies also face severe limitations. When faced with the daunting complexities of public access, government directives often punt. They dissemble, hedge, exempt, and accomodate. They leave difficult decisions to task forces, designated administrators, and future plans.

For example, here is part of the San Francisco Open Data ordinance, as it was introduced:

Each City department, board, commission, and agency (“Department”) shall make available all data sets under the Department’s control, provided however, that such disclosure shall be consistent with the rules and standards promulgated by the Committee on Information Technology (“COIT”).

Seems straightforward enough: agencies have to follow the standards that will be created by the technology authority.

Here are the changes that came out of the committee, though:

AMENDED on Page 2, Lines 2-6 by adding ‘make reasonable efforts to’ after ‘shall’; deleting ‘all’ after ‘available’; adding ‘and with applicable law, including laws related to privacy’ after ‘(“COIT”)’; Page 3, Line 2 by deleting ‘all’ after ‘accounting of’; and on Page 3, Line 5 by replacing ‘would’ with ‘could’.

Those changes are comical. The ordinance now mandates that agencies have to try to follow the standards set by an IT oversight body, to release some information based on an audit of some subset of public data.

This is the language of the low-hanging fruit — the kind of aspirational mandate that isn’t really a mandate at all, but more of a statement of goals and principles, lofty rhetoric with a roadmap made up of other road maps, and plans for other plans.

Again, this declaration, and the others like it, aren’t inappropriate. Their effects probably vary based on the context, based on the actual commitment of everyone involved, from government officials to citizens.

If all these declarations do is to win some of the easy fights, then they’re well worth the effort, because those obvious decisions (like open local transit data) have been gotten wrong far too often in the past, and can have significant positive effects when they’re gotten right.

We need to avoid, however, thinking that these top-level political declarations are something they aren’t. Governments have a vast stores of information, and most of it won’t be reached by these pronouncements. There’s a whole world between the initial urge for government to “put all its data online” and the “please try to put some data online better when you remember to” that that urge decays into. At the first sign of trouble, “all” disappears, “must” becomes “should,” and a mandate becomes a suggestion. The San Francisco ordinance demonstrates just how far a vision for transparency can be from the kind of nuance and structure that makes it possible.

That gap between goals and policies — the long haul of transparency advocacy — will only be filled through hard work, and focused questions about how our governments control their information. Why are legislatures so inept at information policy oversight? Why are IT procurement authorities expected to be competent at controlling access to information? What kind of systematic approaches can audit and characterize government information generally? Why are existing audit and indexing requirements routinely ignored? If e-government and information laws become stale and ineffective so quickly, what can replace them? How can stakeholders and the public at large force the issue?

Those are one approach to the answers we’re really after. What isn’t public, but should be? What is now public, but isn’t online? What should change about what is already posted online? And ultimately, how much can this access truly transform our governments?

Information policy is now only a shadow of its future self. The “presumption of openness” is better than what often grows in its absence (cf. Ashcroft), but it’s no replacement for the thousand tough decisions that get made across governments every day.

Open Government Directives and Open Data Laws are great. They’re not a silver bullet, though, and the ways they fall short will be instructive about the kinds of work that lie ahead.