The cost of not linking is that I don't believe you

So yesterday as I was lazing about watching Up With Chris Hayes, Ms. Heel-Filcher* came into the room with her ears perked up because of something Hayes had said about education research. Upon rewinding, we found that the cited story, as reported in the Wall Street Journal, was claiming that, while previous generations have all tended to get more schooling than their parents, that trend is reversing.

This is, naturally, relevant to Ms. Heel-Filcher's interests, not least because of the myriad final papers she is writing as we speak. But in order for this to move from the "interesting" to the "useful" column, we need citations.


Offender #1: Chris Hayes. Here's the segment, I think. (Sorry for lack of transcript.) Chris cites the claim to the Wall Street Journal at an unspecified date this week; for the purposes of TV, that's fine, but MSNBC maintains a blog for the show for precisely the purpose of exceeding the news-reporting capability of the medium.

All right, have to consult the googles. First few links does indeed find Offender #2: the front-page WSJ story "Education Slowdown Threatens U.S.", by David Wessel and Stephanie Banchero which indeed makes the claim:

Throughout American history, almost every generation has had substantially more education than that of its parents. That is no longer true. When baby boomers born in 1955 reached age 30, they had about two years more schooling than their parents, according to Harvard University economists Claudia Goldin and Lawrence Katz, who have calculated the average years of schooling for native-born Americans back to 1876. In contrast, when Americans born in 1980 turned 30 in 2010, they averaged about eight months more schooling than their parents.

 All right! We have scholars' names and affiliations! All we need to find the source of the claim!

Or not. Here's Goldin's faculty webpage. Here's Katz's. Look through the titles. Look through them twice, like I did, before realizing that both pages are a little out of date, so if the claim is coming from this year's research, it won't be on there anyway.

And now we're stuck, because none of those papers sounds remotely like something that would support such a claim, except in some side note or data table or wild goose somewhere that I could spend hours chasing.

A word to journalists, both print and TV: You need to link. And more than that: if your report is based on published or soon-to-be-published scholarly work, you need to include not just the scholars' names, but the title of the paper or book or whatever the interested and motivated reader needs to get her hands on the source. If your report is based off unnamed and uncheckable sources, there needs to be a very good reason for that; and this is not that case, you're not shielding sensitive information here, you're just lazily treating a scholarly finding the same way you do a politician's pronouncement -- as a quote to be marshaled in support of a story, and not as an independently checked piece in the large repository of more-or-less dependable, methodologically-vetted knowledge that is one of the products of our modern scholarly apparatus.

So to Wessel and Banchero: I get that the print edition of the Journal isn't formatted to include footnotes to claims. You know what is? The web edition.

And to Hayes: You have a full-time production staff with people whose job it is to police facts and citations. This claim should have never made it onto your show without an annotation to your watchers of where they can go to get the source. And this one's not hard: one phone call to either of the scholars' offices in question would have gotten you this information, especially if your staff times it right, for example by phoning during their faculty office hours.

[*] This will be her title until she chooses a new one.