Just Because it’s Open Source Doesn’t Mean it’s Objective

Today, our guest Aneesh Chopra, the first Chief Technology Officer of the United States, shared with us the ongoing projects that he and others implemented as a part of the Obama Administration’s attempt at a more open government. These include the MyGov initiative, Data.gov, open-source challenges like the FTC’s RoboCall challenge (which I hope someone can solve), among others.

However, there is an important danger associated with open-source government. The problem is that even though the government claims to be open-source, there is no guarantee that this is completely true. Let me be specific: this need not manifest itself as malicious intent on the part of the government. Indeed, the most insidious problem arises when there is every reason to believe that the government is being genuine, providing false allusion to citizens and developers alike that the open data they see on sites such as data.gov is objective and exhaustive. But since it is the government declassifying and providing the data, there is the inherent issue that it is the government’s choice of which data exactly to put out, how to package it, and how to organize it. I call this presentational spin; the term “data” has objective and neutral connotations, but presentation and choice have everything to do with distorting this objectivity. This has a couple of implications. The first is that open government is not truly achieved; we have a situation in which the government is open … because it says it is. The second is that the open data that the government provides is not neutral, and can be used to influence peoples’ perceptions of issues, such as energy and weather reports; the government chooses not to falsify data, but perhaps to organize and publicize it in such a way as to emphasize the key points that would be beneficial to its motives, or to declassify certain data but not others (we can never really know which government data is not open source).

There is no solution to this problem. As long as we rely on government to keep and classify important data sets, we will have to rely on their presentational spin whenever they choose to release data. There is however a practical compromise, which is to institute something of an Open Data Review Board, with power and basis independent of the Executive Agency (perhaps a congressional committee?), which can oversee that presentational spin of data is minimized. This obviously has its flaws, similar to the ones plaguing the notion of data.gov, but it is the best solution to ensure that truth prevails, in the interest of the people.

Leave a Reply