I can’t resist the temptation of saying “I told you so” to the many supporters of open government that I pissed off repeatedly by underlying open government’s challenges and limitations (see here and here for the latest controversial posts in a long series), after reading an article on NextGov about widespread errors on USASpending.gov, the web site that provides transparency about budgets and spending for the US government. The article goes on highlighting problems with Recovery.gov and Data.gov (for the latter questioning that data is not sufficiently high value for potential users).
Everybody knows that data quality is a key issue with open government, and something that is seen as a considerable cost and risk factor. Of course people do expect government to provide perfectly accurate and up-to-date data, and we all know that when that’s not the case, government becomes and easy target for criticism.
But what if we take a different spin to all this and try to use open government to solve data accuracy problems rather than making them bigger. In the same article, the executive director of the SunLight Foundation, points to a web site where they track those mistakes. This constitutes an essential resource to provide feedback to government. Also, with the same data being available in different ways (e.g. through aggregated information stores like USASpending.gov, individual agency web sites, private enterprises, advocacy groups and so forth), there is no reason why one of them should be officially tagged as the single source of truth, as opposed to using them all to build a more accurate (or less inaccurate) version of the truth.
As I have said so many times, open government should focus more on helping government do its job better – including improving data accuracy – than on creating a platform for as-yet-unspecified players to create new services and value.
We should not consider these mistakes as unforgivable sins, but as a natural part of the process where the “crowd” (professionals, citizens, other agencies) collaborate on making data collectively better.
Open government will succeed if we will accept mistakes as part of the process and their correction as a joint and not just a government responsibility.