Beyond Transparency

Is more information always a good thing?

A picture is worth a thousand words—but to whom? To the people who see it? Or to those who present it as their carefully orchestrated version of reality? More readily accessible visualization of data does not necessarily translate into greater understanding, nor is more information more wisdom.

Such observations are leading some thinkers to question the virtues of the increased availability and visuality of government information. In “Mapping City Crime and the New Aesthetic of Danger” (April 2009 Journal of Visual Culture), New York University media scholar Aurora Wallace draws attention to the ever-growing use of crime-mapping software applications. Local police departments, county agencies, state governments, and private crime-mapping firms that use police-department data have been providing crime maps of cities, enabling citizens to assess both the efficiency of their local police and the dangers their fellow citizens pose to them. Implied in the availability of the maps is the expectation that people should actively participate in the maintenance of their own personal safety. They can do so by looking at the crime maps of their neighborhoods and searching for citizens with deviant behaviors (such as sex offenders).

There are no common standards for producing these maps. Certain colors signify different crimes in different cities. Simple icons such as shopping carts can mean shoplifting in one city, vagrancy in another. Some crimes are relatively easy to depict—a tiny gun stands for shooting (but not always), a flame for arson—but what about an icon for rape? Crime maps also confusingly represent space: each icon appears in front of a simplified street-map background, but because the geo-coding of crimes relies on street addresses, maps are unable to show crimes in parks, wilderness areas, and even shopping malls that have no street numbers. If the spatial representation of incidents reported to police is problematic, the representation of time is even more troubling. How many days or weeks of crime are overlaid on a single map? The more days represented, the higher the perceived crime rate.

Crime maps, Wallace argues, have many preconceptions built in. They imply that managing crime is simply a matter of combining police work with individual responsibility. But what about employment levels and jobs policy, the quality of public transportation, or the repair and replacement of defective street lights? Why are these factors not made visual? What maps omit is as important as what they include.

The same principle is true, argues distinguished legal scholar Lawrence Lessig in “Against Transparency” (The New Republic, October 21, 2009), even with the most evidently well-meaning “good government” features of the transparency movement that, Lessig acknowledges, he is himself a part of. Lessig is especially concerned by the lack of context in the online efforts to link campaign contributions to legislators’ voting records. If people read, for example, that Representative Smith voted to bail out failing Moneybags Bank, and that Moneybags made a contribution to Smith’s campaign, casual readers will conclude that Moneybags bought Smith’s vote. But, as any Statistics 101 student will tell you, correlation is not causation. The more likely alternative is that in the past Smith voted for government policy that helped banks, which motivated Moneybags to make its contribution in the first place. The transparency movement thus unintentionally promotes civic cynicism because people with limited time and attention will make inferences the data do not justify.

Visual and textual databases are not good or bad in themselves—people are capable of looking for a multiplicity of causes instead of settling prematurely on one simple cause—but because of what Lessig calls the attention-span problem, data are sure to be misread and misused.

Both of these stimulating articles call attention to a problem for which they have no solutions to propose. But both assert, in different ways and from different frameworks, that it is time to see through transparency.

Michael Schudson and Julia Sonnevend write The Research Report for CJR.