Nate Silver’s FiveThirtyEight site launched on Monday during my orientation here at The Post. And while I was reviewing checklists and getting a tour of the building, other people were publishing skeptical reactions that raise larger questions about Silver’s project, and by extension, efforts at Vox Media, First Look Media and the New York Times that share some of Silver’s premises. (Disclosure: I have friends who work at both FiveThirtyEight and Vox.com and had some conversations with Vox about going to work there before I joined The Post.)
“[T]o me these are ‘tweener’ pieces, too superficial for smart and informed readers, yet on topics which are too abstruse for the more casual readers,” wrote economist Tyler Cowen at Marginal Revolution. “You can’t be an effective fox just by letting the data speak for itself — because it never does,” Paul Krugman chimed in at the Times. “You use data to inform your analysis, you let it tell you that your pet hypothesis is wrong, but data are never a substitute for hard thinking.”
And most interesting to me, Kevin Drum at Mother Jones suggested a rubric for figuring out which stories would work for FiveThirtyEight. “My basic take is that Silver’s data-driven approach to journalism works well with subjects that satisfy two criteria: 1. They lend themselves to analysis via number crunching. 2.They are currently underserved by serious number crunchers . . . ’Life’? That’s a pretty broad category, but I suspect it mostly fails #1.” That’s an easy conclusion to reach, particularly if you’re thinking about criticism and meditation on current mores. But I actually think there’s great potential for data journalism about culture and customs, if FiveThirtyEight and its counterparts can figure out what it is.
First, a long caveat. With my own self-interest heartily in mind, I agree with Drum that not everything under the umbrella of “Life” is going to reveal un-mined truths when subject to rigorous data analysis. Emma Pierson’s piece on relationships in Shakespeare, published on the site’s launch day, is a perfect example of how numbers can add up to considerably less than words when it comes to analyzing cultural content. Pierson is trying to say something about the quality of the relationships in Shakespeare based on how often the couples she’s targeting speak to each other. But for some reason, she doesn’t include “Much Ado About Nothing,” a play that is almost entirely about the erotic lure of good conversation between its two leads, Beatrice and Benedick.
Focusing on quantity rather than quality leads her to some slightly odd conclusions. “One might be tempted to blame this on the nature of the plot; of course the lovers have no chance to converse, kept apart as they are by the loathing of their families!” Pierson writes. “But when I analyzed the script of a modern adaptation of ‘Romeo and Juliet’ — ‘West Side Story’ — I found that Tony and Maria interacted more in the script than did any other pair.” The point of “Romeo And Juliet” isn’t just that the couple is divided by their families, though. It’s that they’re exceptionally young teenagers who destroy themselves over what they think is passionate love even though they barely know each other. And in “West Side Story,” the fact that the play ends with Tony dead and Maria alive to accuse his killers is tragic in a different way. Unlike Romeo and Juliet, Tony and Maria knew each other well. Coupling Pierson’s visualizations with a deep dive into the text might have helped explain why some of Shakespeare’s love stories stick with contemporary readings more than others. But on their own, the data doesn’t do much to transcend or explain our gut reactions to the plays.
While I am not persuaded that data analysis is a substitute for criticism, there are an enormous number of subjects that fall under “Life” where data-driven journalism would actually be a profound public service. The entertainment industry holds up as sacred any number of assumptions that deserve a rigorous, numbers-based fisking, among them that female leads cannot carry movies and that international audiences dislike black actors (Will Smith and Denzel Washington are treated like, dare I say it, magical exceptions to an otherwise hard rule). In television, as the Nielsen ratings increasingly fail to capture Americans’ actual viewing patterns, a deep dive into those practices and the TV ad sales business that returns to the surface with viable suggestions for a new measurement that advertisers would trust and outside analysts could deliver would be high-level service journalism. Vida has made headlines in recent years with its count of female book reviewers in prominent literary publications. It would be useful and important to extend that project to look at race, and to explain whose books are getting reviewed and to determine what effects those reviews actually have to bolster the case for a more diverse reviewer corps or a broader selection of books.
Culture is not the only subject included in “Life” that demands journalism that is either done with data or that parses it. Health reporting can sway radically depending on which study has been released most recently, and journalism that attempts to track those whiplashing conclusions could help ease reader panics, and challenge the sense that the public needs to change their behavior every time a new finding comes out about chocolate or bottle-feeding or standing desks. Storystreams, the timeline tool that Vox Media uses to aggregate coverage, could be an extremely useful way to track these shifts — and to keep data-driven publications honest about chasing clicky health stories.
The data publications could also build reader trust by helping educate the public about which studies to trust and why, a project that would meet Krugman’s call to interrogate data rather than simply present it, and it would be a useful way to service the profusion of so-called service journalism. When I saw the headline on Carl Bialik’s piece on how many calories we burn during sex, I initially rolled my eyes, convinced I had already ready every possible iteration of the piece. But the story itself is actually a breezy, fun read about the study that produced those numbers that left me feeling like I had learned something new. If Bialik had a regular column dedicated to parsing buzzy health studies that paired data analysis with sharp reporting on corporate sponsorship and the track records of the academics involved, I’d read that in a heartbeat.
Finally, when it comes to manners and mores, data journalism might be of use in averting moral panics. The Duke porn star may be getting a lot of media coverage, but is she actually a sign of a breakdown in our sexual culture? Or is her experience a testament to the rising costs of college? Do we need to freak out about old legends like rainbow parties or take seriously the rise of the monocle as a new fashion trend?
Data journalism can be useful in telling us both who we are and what we are not. And when it comes to “Life” and however else Vox, First Look and other FiveThirtyEight counterparts define their culture, style and health coverage, the best journalism of this ilk can use data to determine which lies we are telling ourselves, which truths we want to conceal and then marry the numbers to reporting and analysis to tell us why they both matter.