For one journalism professor’s students, the Bureau of Labor Statistics is the path to enlightenment

W

hen Keith Herndon started out as a reporter covering rural Georgia in the 1980s he quickly learned that the economic data produced by the US Department of Agriculture (USDA) provided valuable insights into the rural economy. But it was as a business reporter—and later as a consultant specializing in strategic research on the Internet economy—that he really became a fan of federal statistics and, in particular, the Bureau of Labor Statistics (BLS). “One of the best things to leverage as an independent consultant is free data,” says Herndon “and frankly, the BLS data is so deep and so rich that most clients didn’t even know that this stuff was available.”

Now, as William S. Morris Chair in News Strategy and Management and Director of the James M. Cox Jr. Institute at the Grady College of Journalism at the University of Georgia, learning to use BLS data is a critical element in many of the courses he teaches. “I decided that the BLS was probably the best entry level set of data for these students because of three things,” he says. “One, it is free. Two, there’s a structure to BLS data. And three, it is very deep.”

For the majority of his students, federal statistics are terra incognita. And while they will have taken introductory courses in economics, as required by the University of Georgia, concepts like employment and unemployment are still, in a sense, theoretical. When they meet the BLS in their reporting classes, they have to contend with the data behind the concepts.

Their first surprise, says Herndon, is that this data is actually accessible: “There is an ‘a-ha’ moment: ‘Oh my gosh, there’s all this data that’s available for free, and if I invest a little bit of time figuring out what some of this terminology means, I’ve got a treasure trove of information.’”

As he puts it, there are two ways you can tell a story about how a community is doing economically: “I can assign a story and say, ‘Okay, I want you to go do an economic profile on Clarke County.’ The standard story might be, they’ll go find a couple of business people. They’ll go find the mayor or then go find an economics professor at the University of Georgia, and they’ll do their interviews, and they’ll have a nice story based around what people said. But there’s no real meat or substance to it. They’ve created a narrative on what a few people said.”

“You can take that same story,” says Herndon, “and show them how to insert some real data that adds that perspective and gives context to what those people said.” Immediately, they can see how much more powerful that story is.

The critical point is that the data isn’t something you tack on after you’ve done the reporting; it needs to be the starting point for any interview. “The students can then go to the Chamber of Commerce, or the county economic development department chairman,” says Herndon, “and instead of asking whether the county is above or below the state average, they can ask: ‘Why—what’s driving that? What segments might have higher unemployment than others?’”

“I try to teach my students that they can use the data to be informed,” says Herndon. “And then rather than getting informed in an interview, they can get enlightened.”

Intestinal fortitude

None of this is to say that there aren’t barriers to journalism students (and the public) using BLS data. At a time when minimalism and discoverability are driving web design, when even the USDA is using narrative pathways to direct visitors, the BLS.gov is more of a maximalist experience. The visitor confronts a lot of visually undifferentiated information up front and is expected to know where they want to go and what they want to know. While this may meet the needs of the BLS’s core audience, it is overwhelming for novices.

There is no getting around this, says Herndon, government datasets are all, to varying degrees, difficult to access. “I try to explain that yes, this is standard. But if you look at the Georgia data, the Alabama data will be the same; and the Florida data will be the same. But getting in there, in the first place, requires some intestinal fortitude. That’s going to be the case if they find themselves reporting on an airline, or an airport, or if they’re doing FAA data, or if they’re looking at Environmental Protection Agency data. Hell, the BLS is way better than anything the EPA’s ever put together!”

Some fixes for a better user experience

Nevertheless, using the BLS’s riches would be made easier by a “much simpler interface,” says Herndon. “I’d have a button that said national data, a button that said state data, and a button that said local and county data.” Even better would be a national map providing top-line data for the nation, and from which you could click on state for the same comparable data and then click on the county and get the same thing, he says. “Then, you could line it all up and it would be apples to apples comparison.”

Another important step—and perhaps easier to implement—would be to pay more attention to the way seasonally adjusted and unadjusted data are randomly mixed on the website without explanation as to what those terms mean (there are seasonal effects on employment, such as students working summer jobs). This confuses students and opens the door to misinterpretation, says Herndon. “You click on one press release that’s been posted on the BLS, and it’s written in seasonally adjusted terms; you can go on the same page and find another set of data that’s not seasonally adjusted—and then if you’re not careful, you’re using apples and oranges, and you’re talking about one county using seasonal data and another county using not seasonally adjusted data.”

Another problem, and one that is not just a problem for journalists, is that while the BLS industry data is rich, it has not always kept up with changes in the economy. Herndon, who earned a PhD for a dissertation on the U.S. newspaper industry’s relationship with online media, points to the category ‘Internet Publishing and Web Search Portals.’ First, it’s too broad; second, Google data can skew the entire category. “I wish that was broken out,” he says.

A bigger challenge would be creating a repository that managed to integrate all the economic data produced by different agencies and government departments in a cohesive way. “The Bureau of Economic Analysis has its own set of data, BLS has its own set of data, the Commerce Department has its own set of data, the US Department of Agriculture tracks things, the Federal Reserve has its data,” says Herndon. “We have all this great data, but these agencies aren’t really working together,” says Herndon. There is one site—STATSAMERICA—that is attempting to do this, but Herndon finds the interface too confusing for teaching.

The bottom line

Herndon doesn’t buy into the idea that data journalism in and of itself is a panacea for the news business’s woes. It’s not so much a special “thing” as a tool that every reporter needs to be able to use in order to ask better, more meaningful questions; and yet, there is something vital in being able to see how your community—your county—is doing relative to the county next door, the rest of the state, and then the nation. There are few issues more important to Americans than employment and wages—and perhaps no other issue that has as big an impact on voting. In connecting data on employment and wages to people, and explaining what it means for them, it just might be the way journalism can better connect with America.