(UPDATE: OpenColorado is an independent nonprofit, and has no official relationship with Code for America. Apologies for the mistake.)
This past Friday, I headed up to Parker to attend CityCamp Colorado 2013: Change The Game. CityCamp is the annual conference of OpenColorado.org, a nonprofit organization dedicated to “support[ing] a transformation that will lead to a simple, beautiful, and easy-to-use government”. Now, I’m aware that the word limited did not appear in that vision statement. But I believe the pursuit of transparent and accessible government data is one that encourages citizen engagement, and thus at least has the potential to diffuse the policy analysis and implementation process from concentration in the hands of a professional bureaucracy. With properly informed citizens, that is a good thing. However, ethical qualifications to the collection and usage of large government data sets are valid concerns. I was at times a bit worried that I was the only one so concerned.
Another point on the relevance of this discussion: what is true in governance is equally true in advocacy. I was disappointed at my being the only “campaign-side” professional in attendance among many municipal and nonprofit representatives. It seems clear that if government customers’ data is freely and openly released, Conservative advocates gain capabilities. I hope to have some company next year.
Here are a few discussion points and questions to consider.
Data isn’t going away
On the Internet, genies don’t go back into bottles. We have clicked Agree on one too many software licensing agreements, so don’t be naïve. Computers and smartphones didn’t streamline our lives for no reason. Our technology is both useful and free-as-in-beer because it creates value “away from the ball” by revealing our personal needs, wants, and habits. Our present lives are complementary goods to the real money from predicting future opportunities. As a capitalist, I can’t pass judgment on this business model, so long as an opt-out exists. (Does it? Keep reading.)
Data architecture and design are non-partisan; application design and strategy are not
Much of the issue will remain apolitical; there isn’t a Red or Blue way of setting up or analyzing a database. We need not fear the establishment of best practices that are terribly removed from what is done in the private sector, which means that it will remain easy (and therefore transparent) to understand government application design, as long as the designs are made public. But here’s the rub. Anyone worried over private groups’ exercise of influence strategies derived from behavioral data should be downright alarmed by the same actions in a government context. I heard a certain application developer from a certain municipality say that government applications ought to have sticky enough user engagement that they could ‘reward good behavior…and punish bad behavior’. I threw up in my mouth a little (and offline conversations fortunately revealed that I wasn’t the only one). I am glad that the consensus over the course of the day seemed to be that OpenColorado’s direction will turn towards creating dialogue between stakeholders and developers in different organizations and municipalities; this should provide a moderating influence where legal and ethical use is concerned. But the caveat and its teaching moment are not lost: those who handle government data are government officials, and they have both opinions on government and purchase to enact those opinions.
Accountability of data and people, to The People
Perhaps we should use an example less prone to sensationalization. We discussed mobile crowdsourcing to improve, say, public safety. But things ground to a halt with an intractable question. Does everyone in your city have a smartphone?
How about a PC? (Uh-oh.)
Residential broadband service? It went downhill from there…
How will you reach those people?
Doesn’t their data count?
They’re the main reason to improve these services!
Alright, but who owns the data of those without their own devices?
How do we capture any sort of experimental design when they’re off the grid to begin with?
Astride the famous three V’s (volume, velocity, and variety) of Big Data, the old spectre of a “digital divide” returns with a vengeance more fearful than mere inability to use Pinterest: a “data disenfranchisement” that starts at infrastructure and penetration and winds up in outright structural violence as caused by insurmountable knowledge and access deficits. A problem this big naturally attracts with equal ease solutions that are either statist or nonexistent. But who will “develop the developers” so that they build (and ask for resources to build) good things? Make sure you know.
What to watch for
Government’s ethical duties always include pedagogical integrity when introducing new technology to public services. Government faces additional challenges with data strategy, given its arcane technicality and unprecedented scope of applications—and they should be expected to overcome those challenges in good faith. However, citizens, too, have a duty to participate in the decisions being made on their behalf. If an uncomfortable amount of data resides where government can see it, I suggest that it was not an initiative of any exceptional motivation. Given the advance of computing, government’s participation in “Big Data” really means only that the process is easy enough to survive government inefficiencies and still provide value. If government can collect the informatics of our lives, then so too can private parties. Hence, if ownership of large, detailed behavioral data sets is unavoidable, our best response is to “dilute” that ownership with open data. Legal and ethical limits, likewise, are best protected by the attendant transparency and citizen participation of open data principles. And in that case, we can take heart; transparency and civic participation, at least, are nothing new as bulwarks against encroachment or exploitation of the informational and cultural commons.