Wednesday, April 2, 2014

Data journalism, rich suburbs and a complex NJ district

Some interesting thoughts from Aron Pilhofer of the New York Times on data journalism, a field about which I'd made some preliminary notes recently. Anyone getting into journalism now should heed his warning about what's less likely to be tolerated as much going forward: "Journalism is one of the few professions that not only tolerates general innumeracy but celebrates it. I still hear journalists who are proud of it, even celebrating that they can’t do math, even though programming is about logic."

Also of interest, this piece at FiveThirtyEight: "Rich Suburbs Can’t Save Democrats This November," by David Wasserman, which assembles data suggesting that affluent "Super-ZIPs" are not the reliable Democratic votes they seemed to be in Virginia.

Maybe so, but I see that my own district, NJ-05 represented by Scott Garrett, is high on a list there of "Top 22 GOP House Districts Ranked by Median Household Income, 2012" and the districts being Republican strongholds is explained by Wasserman thus: "Why are they so secure? For the most part, these Super Zip Republicans are pro-business members who de-emphasize wedge social issues and biographically fit their districts well."

There's surely some truth to that in NJ-05 but there's good reason to think it's not the whole story. (Which of course may be true of many data journalism stories; there's a flood of data and anything a journalist culls together will necessarily be selective.) Look at a map showing the very irregular shape of the district. For Garrett, the district having less affluent, rural or semirural, mostly non-commuter areas in the western part of New Jersey is also a key factor in his support. I've stated in the past that Garrett's incendiary role in the debt ceiling standoff might cause him to lose some support in affluent Bergen County--but even if so, that would be less likely to weaken him among the non-Wall Streeters further west. In any case, his reelection prospects are bolstered by more than just "rich suburbs."

As a further cautionary point, I note that you get different numbers and rankings when you look at income figures for the towns in my district of New Jersey. Franklin Lakes, for instance, outranks Wyckoff in income here but not when you plug the towns in here. It all depends where the data come from, and the data are always changing. Good advice to consumers of data journalism is not to be dismissive of the data, but remember it's a snapshot and may not precisely match the questions it is being pulled together (on deadline) to address.

No comments: