Data Journalism and the Big Picture
The web-o-sphere this week brought forth a collection of opinions on the value of data journalism and the skills that go with it. To wit:
- Tim Berners-Lee, he who invented the World Wide Web, told the Guardian that “journalists need to be data-savvy” and that “data-driven journalism is the future.” The story then goes on to question whether data analysis could ever replace traditional reporting.
- The blog 10,000 Words declared that one of the “5 Myths about digital journalism” is that “journalists must have database development skills” and suggested that most journalists should leave high-level hacking to the experts.
- Another site, FleetStreetBlues, opined that “amidst all this hype, earnestness and spreadsheet-geekery, here’s the truth about so-called ‘data journalism’. It’s still about the story, stupid.”
There’s been a bunch of reaction to these posts, including a few people pointing out a 1986 Time story that sounds similar to the one this week from the Guardian. And therein lies the problem with all three pieces: None of them benefits from a big-picture, historical perspective on data journalism — not where it came from, not how it’s changed and especially not the massive amount of ground the label covers these days.
We used to call it CAR
Back when software came on 5.25-inch floppy disks, or maybe before then, the idea of using a PC to “crunch numbers” was christened “computer-assisted reporting.” These days, we call it data journalism because, along the way, it became obvious the old name was anachronistic. As Phil Meyer once said, we don’t talk about telephone-assisted reporting, do we?
When I got into the game — when Paradox was the desktop database manager of choice — our newsroom had a personal computer designated as the “CAR station.” While others worked on dumb terminals connected to a mainframe, I was surfing the web with Netscape and ringing up Paul Overberg for advice on Census data. I was the newsroom data expert — the guy reporters called when they had a spreadsheet on a disk or an idea to get data from city hall.
In that era — with database-driven web startups like Amazon.com spreading cultural revolution — it was easy to foresee a time when reporters wouldn’t just get the occasional spreadsheet but find themselves inundated with data. Thus was born (at least in my sphere) the drive to evangelize CAR in the newsroom. We taught Excel, we sent people to IRE boot camps, we set up presentations showing the kinds of stories journalists were landing with these skills. The message of CAR was about finding stories and using simple tools to do it: spreadsheets, databases, maps, stats.
Now we call it hacking
Soon enough, though, the craft began to change and so did the talk at IRE CAR conferences — especially in the hands-on classes and demos. In Philadelphia in 2002, the hands-on classes mostly covered Access, Excel, SPSS and, for the adventurous, SQL Server. Just a few years later, in Cleveland and Houston, the offerings included sessions on web scraping, Perl, Python, MySQL and Django.
The growth of the web and the availability of data helped push the change. I also suspect that “CAR specialists” who started down the data journalism road in the 1990s had pretty much exhausted the boundaries of Access and Excel and were, as we should have been, on to new things. Either way, by the time PolitiFact won a Pulitzer, the era of news apps was in full bloom and the concept of programmer-journalist was simply the next natural evolution of data journalism. Hello, Hacks/Hackers.
But the message in the CAR (now data journalism) community remained the same: We use these tools to find and tell stories. We use them like we use a telephone. The story is still the thing.
On the outside looking in
Back to the week’s three disparate-yet-related stories, one of which really raised the ire of the aforementioned Pulitzer winner. Each one misses the point because it’s missing that context:
- Once a pioneer, Sir Berners-Lee is late to the party in declaring data journalism “the future.” That future has passed. The ability to handle data is no longer a skill journalists ought to learn — it’s a basic life skill my kids are learning in middle school. Plus, I cannot think of an instance in the last 15 years where someone in the CAR community suggested that data journalism was a replacement for shoe-leather reporting. That the writer raises the issue tells me he’s reacting to the hype of Berners-Lee’s statement, not assessing the reality of what’s practiced.
- The CAR/data journalism community’s always been heavily geared toward helping people build these skills. The entry-level boot camps at each year’s IRE CAR conference walk people through Excel and Access — great places to start. But I don’t tell people there’s a limit on what they can do. In the same way that the craft as a whole has evolved, journalists who start down this road usually move on to more complex skills. The only limits on people are the ones they place on themselves. Indeed, the only myth that needs busting here is the one that says you have to be Einstein to learn this stuff or have some magical left brain/right brain balance. No, you just have to be persistent.
- “It’s still about the story.” It’s never not been. The panel descriptions for the last 10 years’ worth of IRE CAR conferences speak for themselves.
If anything, this collection of stories should remind us that more than ever, we need organizations such as IRE, Hacks/Hackers and others to not only impart skills but provide the context for why they’re so desperately important.