Don’t get me wrong. I’m all for “open government.” But there are problems inherent to real-world democracy that can’t be automatically solved by making the workings of government more visible to the public.
This is especially true about the reporting by government of data collected from large and complex populations of people and organizations. (I wrote about one example last February in Challenges Facing Recovery.gov and the Recovery Accountability and Transparency Board.)
Why doesn’t “transparency” around large complex programs automatically succeed? There are several explanations:
- People make mistakes. Building complex databases from hundreds, thousands, or even millions of different sources is bound to introduce unintended error somewhere along the way. Quality control efforts are essential but must be implemented throughout the process from data collection through final reporting and access. All that quality control takes planning, time, and money. The resulting data sets are bound to contain errors. People can pounce on these errors to cast doubt on the entire effort.
- Source data can be sabotaged. Many open government efforts depend on widespread public participation, even volunteer effort. If someone wants to make a system look bad, source data can be withheld, selectively reported, or even incorrectly reported. Then when reporting time rolls around those same errors can be pointed out as “evidence” that the overall system is untrustworthy. Has this happened? I don’t know. But when you look at the complexity that surrounds some reporting efforts you realize how easy it would be to throw a “monkey wrench” into the works.
- Facts aren’t sufficient. Even if the system gathers data flawlessly from a myriad of local sources, builds a comprehensive database with squeaky-clean reliability, and makes the contents of the database easily accessible in detailed and summary form, those who don’t support the underlying assumptions behind the data will never be convinced of (or admit to) the facts the data are intended to document. That’s just politics as usual.
- Communication can be restricted. Governments of all shapes and sizes are good at figuring out ways to keep facts away from public scrutiny. While we might all agree in principle that certain national security data needs to be kept away from prying eyes, the fact is that government bureaucracies are well-practiced in burying unpleasant, embarrassing, or inconvenient facts, even in the face of public pressure for openness.
Still, I’m a firm believer that democracy functions best when the inner workings of government are visible. But creating and taking advantage of that visibility is not a simple task. Creating and maintaining reliable sources of data requires time, money, and expertise.
More importantly, people need to be able to gain access to and understand what the data are all about. One problem is, as the complexity of that data increases, the expertise required to understand and interpret that data also increases. Consequently the need for such expertise increases the likelihood for politics to flavor how the facts are interpreted.
What’s the solution? More transparency, it’s obvious to me, is one answer; this current Administration is certainly behind this when compared with previous Administrations. But that’s not all. Vivek Kundra lists the following list of complex elements in Changing the Way Washington Works as he introduces the Federal CIO Council and OMB’s Data.gov public dialog effort:
- Focus on Access
- Open Platform
- Disaggregation of Data
- Grow and Improve Through User Feedback
- Program Responsibility
- Rapid Integration
- Embrace, Scale and Drive Best Practices
This is a long list that will require a LOT of work to implement. But it’s a good start.
It isn’t possible to reverse decades of secrecy, compartmentalizing, and data hoarding overnight. But I’m optimistic since, despite the costs of transparency, it’s probably impossible to reverse the flow once the barriers start to fall.
Copyright (c) 2009 by Dennis D. McDonald. Contact Dennis at email@example.com.