In the weeks since the Journal News‘s publication of a controversial interactive map of gun permit holders’ addresses in Westchester and Rockland counties, industry commentators have been raising questions about the ethics and value of its publication. “Should data have a conscience?” David Carr wondered in his January 13 New York Times column:
As a journalist, I am trained as an absolutist in matters of open data. Public records should be just that, public, and as agents of transparency, news media outlets should help cast sunlight on those records. In that context, data is not good or bad, right or wrong; it is information that should be there for the asking, taking or publishing.
But Carr eventually concludes that the treatment of the data by the Journal News didn’t add enough context to make the map journalistically worthwhile—making it, perhaps, “data-driven link bait.”
At Reuters, Jack Shafer had already rejected this notion on January 2, when he queried whether critics will “be mollified if somebody attaches the Journal News data set to a slicker narrative?” Even on its own, Shafer asserts, the map constitutes a “superb data dump.” Meanwhile, Jeff Jarvis suggested that in the world of digital publishing, editorial judgment has become “moot” regardless, as without print space restrictions, “there’s no longer a scarcity of space to control, to edit.”
As a data journalist and developer, I find the sensibility of these assertions troubling. Carr’s opening question, though rhetorical, suggests a disturbingly naive approach in itself to data as a journalistic source, whether the material is a matter of public record or not. All data is, at best, a dependent authority: It is the product of choices made by human beings as to what is worth observing, recording, and disseminating.
As journalists and citizens, it is our job to question those choices; this is the embodiment of the checks and balances our society embraces. Legality is not a proxy for propriety or fairness, as Bill Keller points out:
While we were all distracted by the dance on the fiscal cliff, the 112th Congress in its final days whisked through a renewal of the law that governs eavesdropping by American intelligence agencies on Americans’ phone calls and e-mail traffic.
That collecting or publishing information is legal — or even a matter of public record — does not reduce our individual and professional responsibility as journalists to make appropriate choices about it.
Likewise, Jarvis and Shafer’s contention that making editorial judgments about what to publish is irrelevant because “someone” would have published (or republished) this map seems little more than a convenient bit of fatalism, which implies that we should shape our work to fit our tools, rather than the other way around. This brings to mind the introduction to Douglas Rushkoff’s book, Program or Be Programmed. He wrote:
The websites and social networks where we do our work and play — they are not nature. They have been constructed by people (or at least for people) with real agendas.
Imposing a journalistic agenda on technology is far from impossible, however. In 2011, Simon Rogers of the Guardian‘s DataBlog demonstrated this perfectly when covering the UK riots, in which he mapped the addresses of arrested individuals.
“Two days [after the riots] David Cameron said, ‘These riots have nothing to do with poverty,’ which is an interesting thing to say with no evidence,” Rogers said. “We wanted to see if people were from poor places.”
Though he used the exact same technologies as the Journal News — the Google Maps API with data drawn from Google Fusion Tables — and mapped actual addresses, Rogers intentionally limited access to names and street numbers of the accused.
“We did it not so much to show where these people lived, because we didn’t want to set up a vigilante’s charter, even though you’re over 18 and you’re in court, so your address is public,” Rogers said.
Instead, he strove to make the map useful while taking steps to protect the underlying data. “We didn’t allow exports, which is just a tick box. In Fusion Tables you can edit what goes in the info window. You can set the zoom as well, which is what I did — it’s not rocket science.”
When what he wanted to do was beyond his expertise, he reached out to Google for help. “I’m not a programmer,” Rogers says. “I asked them, ‘How do you limit zoom?’ Google are phenomenally helpful with this stuff. ” Although the DataBlog usually publishes its underlying data for download, Rogers decided against it, due to public sensitivity about the riots. “It was such an emotional thing here,” Rogers said. “People were still quite raw about it.”
Rogers acknowledged that this didn’t mean the data was totally inaccessible. “If you were a developer, you could go into the code and look for the table ID and go in and get the information, but it seems to me that’s as much effort as a FOIA request.”
Though the code on the Journal News site suggests that they had the technical expertise to do so, the paper declined to put similar restrictions on their data or map views. For at least a week following publication, the data remained downloadable from the underlying Fusion Table.
Rogers was circumspect in his assessment of the Journal News piece. “It’s exciting,” he empathized. “It’s a big dataset. But now it’s been three weeks since they’ve published, and it could have been this really beautiful, detailed thing. The map would have been more interesting to me if they had overlaid income or unemployment or whether people have lots of kids. To me, those are more interesting questions.” As Rogers knows from experience, the technology is a barrier only if one lets it be.
Practicing journalism requires constant value judgments: what stories to pursue, what sources to use, what — and how — to publish. The choice that faced the Journal News was not simply whether to map gun permit holders’ addresses, but how. The Guardian’s map illustrates that showing judgment in that task is feasible; the Journal News map, that it is necessary. Data and technology do not have independent agency; they embody the agendas of their designers and creators. If, as journalists, consumers, and citizens, we fail to use these products in a way that meets our own objectives, then our work will almost surely end up serving someone else’s.
UPDATE, January 22: A few days after this piece ran, the Journal News made some changes to its gun permit map, including removing the underlying data as well as limiting zoom and removing satellite view.
Susan McGregor is Assistant Director of the Tow Center for Digital Journalism and Assistant Professor at Columbia School of Journalism.