Wednesday, August 22, 2007

Expertise and journalism

Friends know that I've long held that the biggest problem with reporters is that most of them are not experts in anything in particular, which makes them gullible, easily manipulated, and prone to mangling the facts. Recently, there have been several worthwhile posts in this vein from Brad DeLong, Matt Yglesias, and (less directly) Ezra Klein.

Here's an excerpt from an actual email that I wrote two years ago, to an acquaintance who works at a major national magazine, discussing Michiko Kakutani's review of Bill Clinton's autobiography, which Brad DeLong discussed some time ago:

Moving along, if the Times prints book reviews in the daily paper, but never solicits such reviews from people who actually know something about the material in question, then that is a structural problem with the Times. Likewise if the Times rushes articles into print too soon. Perhaps I'm naive and don't know why the newspaper business is set up as it is. Perhaps there are good reasons, and not just ingrained shibboleths, causing the Times to use staff writers exclusively for the daily paper (outside of Op-Ed).

More broadly, I think the fact that journalists are mostly smart generalists is a huge weakness of the press. Save for a few geniuses, generalists are basically obsolete in every other field. Journalism should join modernity and take specialization of labor seriously, with all the attendant educational requirements. I can't count the number of times I've winced at some clueless description of computing technology in the general press. DeLong's frequent outrage at the economic ignorance (and general innumeracy) of the press shows that economics coverage is, if anything, much worse. Kakutani should not be reviewing history; she should be confined to literary fiction, where all the training one needs is to have read a great many books.

Newspapers hire full-time reporters who are not expert in any field because that has historically been an efficient way to produce news articles on deadline. At a high level, a major newspaper hires a hundred people who each write a few articles a week, instead of hiring ten thousand people who each write a few articles every hundred weeks. People assume this is simply the natural order of things, but it's not obviously the better choice: ten thousand experts can give you more substantive coverage of far more subject areas than one hundred generalists. It's true that one percent of a journalist's annual salary is a relatively small sum to pay a freelance expert for a couple of articles a year, but many experts care enough about communicating their subject matter to the public that they would probably work at cut rates (or, as blogs demonstrate, for no pay at all).

So why should newspapers be organized as small bands of generalists, instead of large federations of experts? I think the answer is coordination costs: you would need to maintain the world's biggest rolodex to keep up with all these experts, and aggregating the irregular work output of ten thousand people requires much more sophisticated planning than aggregating the regular output of one hundred people.

But coordination costs are an artifact of available technology. Mechanisms exist today that coordinate the output of many more than ten thousand voices. It's true that no existing system can serve as a drop-in replacement for the modern news reporting organization. However, the field of software-assisted content aggregation is young, and I believe that someday someone will figure out how to exploit the world's existing network of experts directly for reporting on complex issues, rather than relying on secondhand summaries of expert knowledge.

On a slightly related note, it's always laughable to me when I read about junk like this (and more! via Atrios; and more!), where some journalist sniffs contemptuously at the unwashed blogging rabble, and then gets promptly pwned (sorry, that truly is the mot juste).

What's especially rich to me is that, unless I misread Michael Skube's biography, Skube hasn't done hard investigative or beat reporting (at least, not full-time) for about two decades. His biography describes his freelance and city paper reporting from 1975 to 1982, but from 1982 onward his roles are described variously as an editorialist, book editor, columnist, and critic. In other words, it appears that he's spent the better part of his adult life doing exactly what he accuses bloggers of doing: sitting on his ass in a chair, reading what other people write, and commenting on it.


  1. As a whole, in comparison to journalists, bloggers write like confused favorites at the local spelling bee. The shift you suggest would require thousands more man hours to edit and train these specialists.

  2. Ha. I can immediately think of at least three ways you're wrong:

    1. You vastly overestimate the writing ability of journalists. Any profession that elevates Richard Cohen to one of its elite positions has no business crowing about its collective writing chops.

    2. You vastly overestimate the value of well-honed prose versus correct, precise, and insightful thinking.

    3. You say "as a whole". From this, I infer that you're comparing the set of all bloggers to the set of all all journalists. This is a common fallacy when discussing the Internet. I actually have a post in my drafts folder on this subject, but briefly: for any interesting subset of Internet content, quality and attention are both highly unequally distributed. The top of the distribution gets a huge fraction of attention, and the long tail gets relatively little. The "average" blogger barely matters at all to the "average" reader. Given good filtering mechanisms, only the top 0.000001% of bloggers matter to the average reader.

    And the cream of the crop of bloggers produce more valuable writing than all but a handful of journalists on almost any subject you care to name.

    Journalists have a comparative advantage over bloggers in two areas only: (1) local beat reporting and (2) hard, in-depth investigative reporting. Everything else in the newspaper, including everything Michael Skube has done for the past two and a half decades, can be done better by non-journalists. And will be.