Tuesday, October 30, 2012

Inside Outside (Where Have I Been?)

Paul Krugman channels Bill James at his best in explaining why "outside" knowledge sometimes trumps "inside" knowledge:

But Martin’s tweet also reveals a broader issue in reporting, which I’ve commented on before, I think (no time to search): the unhealthy cult of the inside scoop.

A lot of political journalism, and even reporting on policy issues, is dominated by the search for the “secret sauce”, as Martin puts it: the insider who knows What’s Really Going On. Background interviews with top officials are regarded as gold, and the desire to get those interviews often induces reporters to spin on demand.

But such inside scoops are rarely — I won’t say never, but rarely — worth a thing. My experience has been that careful analysis of publicly available information almost always trumps the insider approach.
It's very good; read, as they say, the whole thing.

Bill James once talked about what it was like when he became established enough that he started having access to inside information. What he said -- and I think this is exactly correct -- was that the best way to think about inside information was in terms of a temporal lag. It's not that insiders know stuff that outsiders will never learn; it's that they sometimes hear things before they are made public.

Regulars will also remember that one of my all-time favorite things about this is the old McLaughlin Group self-description; the show would give us "inside opinions and forecasts" -- not smart opinions or accurate forecasts, but inside ones.

Now, the thing is that it is valuable for outsiders to hear "inside opinions and forecasts." And inside information, too. To take one of Krugman's examples: knowing what Bush Administration officials thought about Iraq in 2002 was a much better indicator of what was going to happen than actually knowing about, say, Iraq's various weapons programs. Good reporting would have told us both, making clear both what was accurate about the world in general and what the Bush Administration believed.

The trick is balancing the two; the trick is to reject the notion that there's something special or inherently more accurate about what insiders know, even while it's very useful for us to learn what those insiders are thinking. That's not easy. But it's essentially impossible if you mistakenly attribute mystical qualities to insiders, incorrectly treating their insider knowledge as inherently more accurate. I've always thought that the first step out of that mystification is to follow Bill James and realize that there's really nothing special about insider knowledge at all.

15 comments:

  1. "Out of my brain on the 5:15..."

    ReplyDelete
  2. I'd also add that the the inside the beltway crowd has strong economic incentives to dismiss what people like Silver or Sam Wang say. Right now there are not that many people who cover politics at a "major league" level and importantly being seen to being someone "in the know" is not just a way to get stories published, it's also a way to fame and money from things like book deals or jobs outside of the daily grind of real journalism. Richard Ben Cramer in "What It Takes" goes into this a lot and you can see it in cable news all the time; Joe Scarborough might be a blow hard, but he makes a lot of money. Since top tier political reporters generally come out of the same institutions, like elite journalism schools, and have to work for a while to climb the ladder letting someone like Silver, whose comes from a totally different world, arrive on the national scene and start giving his own analysis is a kind of economic threat. If you read Silver why do you need to read David Brooks' column to tell you what to think? Silver represents, to them at least, a threat to their jobs as much as some uppity nerd telling them they are all wrong. Although I'm sure cultural differences play a big part too.

    ReplyDelete
  3. @Jonathan-- So, another way of putting this, why should we care/if David Brooks cuts his hair? (:

    ReplyDelete
  4. Um.....didn't you organize the Political Junkies at IGS?

    ReplyDelete
  5. I read that book Cod a while back, and the part I remembered was the reporting that when Cabot reached the Grand Banks in the late 15th century, he was shocked to discover the ruins of 1,000 Basque ships. When folks have information that is more valuable to them as a secret, it tends to stay that way.

    I don't buy the meme that inside information "wants" to get out; more specifically, that inside info "wants" to get out without some insider benefiting. Example: one of the most incongruous aspects of the American political landscape is elderly support of Republicans, given their reliance on FDR's Great Society.

    Surely the RNC has a fairly detailed view of how that odd alliance holds, no? We all have theories, like culture or inertia or xenophobia or whatever, and all of that might be part of the RNC summary, but surely "theirs" is much more sophisticated than "ours". But who benefits from spilling the beans? What Republican operative, privy to the insider theory of why the old folks stick with those who would stick it to them, is better off for sharing? The answer is pretty clearly no one, which is why the rest of us, like the explorer Cabot, remain in the dark.

    Finally, this also pretty well explains why baseball info leaks more easily than political info. For when Theo Epstein joins the Cubs, there is no presumption, from any sports fan, that he's not utilizing all the sabermetricky stuff he learned while with the BoSox; in fact the Cubs' bad 2012 engenders schadenfreude that maybe the poor bastard Epstein simply forgot to download secrets to his flash drive before he announced his intentions to BoSox management.

    You cannot imagine a figure in the world of politics for whom loyalties are as permeable as what you naturally expect out of Epstein; consequently, we should assume that secrets in the political realm stay secret much longer than those in baseball.

    ReplyDelete
    Replies
    1. @CSH: (1) if as many people know a secret as you're positing, guaranteed there's someone who'd benefit from leaking some part of it. Cultivate allies, undermine opponents, even just trade something to a reporter in exchange for favorable coverage -- idk. In fact -- that's the kind of thing insiders know and we don't know! A lot more detail on some of the mechanisms of transfer of information &c. Which can certainly be helpful in deciding whom to trust when about what. But which can also just be a fog of gossip, a set of very noisy data.

      (2) We know what the RNC is doing to reach out to older people! Because they do it, and then it's public!

      (3) The RNC, really? Not the most unbelievably promising institution to pick as an example of hierarchy, efficiency, and discipline, is it?

      Btw -- it's still raining here. It's been raining straight through, for more than a day, with variations in rain intensity and presence of lightning. It's weirdly creepy. (What's that Ray Bradbury story where the unceasing rain on Mars gradually drives everyone insane?) No real damage though so I'll try to stop complaining every time I post.

      Delete
    2. Having lived in the rainforest, your whine about a couple days' rain and storm is funny ^-^ The driest place I lived as a child got a 100" a year.

      Delete
    3. It's not the amount, it's the sound, or rather some of the sound patterns. (what does Austen say in Persuasion -- "everybody has their taste in noises as well as other matters; and sounds are quite innoxious or most distressing by their sort rather than their quantity") But it's pretty awesome that you've lived in a rainforest!

      Delete
    4. There's also an expectancy factor with rain; if you wake up every morning expecting it will rain, its arguably less depressing when Mother Nature constantly obliges. Not sure if your musical taste runs to postmen-turned-folk-legends, but here's John Prine's "Great Rain" for a lighthearted take - "I was praying for mercy, all he ever brought was you".

      To the other topic: people often say that financial incentives drive these secrets public, but there's something more required; something cultural, follows is a fictionalized account of something that happened to me long ago, a story for which probably half this audience has its own version:

      Suppose you were a grunt working for Disney, and you happened to be working on the due diligence of the LucasFilm acquisition. After you acquire that company, you discover, upon cracking their books, that it is not the wooden dialogue or stunningly inept plot developments (e.g. Jar Jar Binks) that made those movies successful; rather that IL&M had discovered a "golden mean" of cinematic color combination that was not just eye candy but in fact eye heroin.

      Your Disney leadership eagerly commits to carry-forward this newfound technology in episodes 7,8 and 9; the company discloses it on all relevant SEC docs in the way that companies typically do ("use of pixelated techologies from IL&M"). You and your buddies sit in the break room and marvel that if MGM/Universal (or, even, Marvel) knew what those SEC filings meant, why...

      And you talk about how much they would pay you to cough it up. And yet said coughing up of secrets almost never happens. Mainly because it is a terrible idea professionally, partly because there could be legal jeopardy, partly because its just a big pain in the butt.

      Assuming such secret strategies exist (and your point is well-taken about the dubiousness of the assumption), the RNC's scientific view of exactly how to hold the elderly, to which Mary Matalin may be privy, is not the kind of thing she should be whispering in her hubby's ear. He could certainly make it worth her, and their, while! But she just wouldn't do it, you know?

      Cubs GM Epstein, though, we would fully expect to take whatever he learned in Boston and apply it in Chicago. Which, so far, seems to be about nothing.

      Delete
    5. The Bill James thing wasn't about secret techniques, although it would almost certainly apply to that, too; it was about stuff like who was on drugs, or who was a jerk.

      But obviously I fully agree with the classicist here.

      Both of your examples here are also really good examples of Krugman's original point. Outsiders will have an explanation for both of these things (age-cohort partisan loyalty, Star Wars success), and it's extremely likely that the outsider explanation may be more accurate than the insider explanation. Insiders often don't really have the best vantage point to see these things from, and also often don't have good incentives to see them, anyway.

      Delete
    6. What follows is perhaps tenuous given its the RNC we're discussing, but I don't think its right to say that it is "extremely likely" that outsiders know better than the RNC why the elderly stick with their Republican frenemies. If the RNC is remotely competent (a big if!), indeed it is extremely unlikely that we in the hoi polloi know better than The Machine what drives that relationship.

      We all have theories, no doubt. Our theories are overwhelmingly the result of anecdote; people we know or about which we have read. Maybe we also read some pundits' opinions, which broadens the scope of the anecdote. No doubt that our anecdotes contain more than a little truth!

      At the RNC, considering how their curious alliance with the elderly is so critical to their electoral fortune, it stands to reason their view is built on more than anecdote. (Okay, maybe not the "actual" RNC, but certainly the "hypothetical" RNC). We might expect that a group like the RNC has multiple binders in filing cabinets containing a deep understanding of that relationship, some of which overlaps with our own anecdotal views, and some of which doesn't.

      The DNC is, sadly, not going to pay much for our anecdotal opinions about what drives that relationship. Assuming those binders exist, they are potentially worth silver and gold to the DNC.

      Again, though, its not enough just to have access to private, valuable information - there has to be a reasonable expectation of profiting from its dispersal, which expectation certainly exists in sports, but less so in politics.

      Delete
    7. No, I totally disagree with this. The "we" I'm talking about here are academics (or Nate Silver types) who can apply plenty of data to the question, including long-term voting patterns and good survey data.

      One of the things you learn when you approach it from that outside perspective is that groups remain in place, long-term, for the most part. So any kind of campaign technique is very unlikely to make a huge difference. Could be some, but not a big one. And yet it wouldn't surprise me at all if the people whose job it is to come up with techniques massively overstate their importance.

      Delete
    8. I concede - I suppose there's some chance the RNC's binders would have "better" info than the academic/Nate Silver type's binders, if only because of the possibility of different info making it to a sympathetic audience (e.g. "I hate folks of other races" might be something you'd cough up to a similar/sympathetic researcher but not an anonymous pollster).

      Then again, different info can also be worse, and either way, there's nothing particularly stopping an academic from assembling her own binders that more or less address the issue in about as much detail as a (presumably highly effective) RNC.

      I was wrong. (My goal is not to write that sentence again anytime soon).

      Delete
  6. JB,

    That raises an interesting point. When you talk of "groups" do you mean fixed categories of the electorate (i.e. over-65, under-25, industrial workers, government employees,etc) or do you mean cohorts? If the first, that would seem to imply that people's political behavior changes as they move from one category to another, as all social and economic categories, with the partial exception of race, are malleable and people are moving among them constantly. If, on the the other hand, you mean cohort, that means that a given group of people, usually I guess defined by age, develops a particular set of political behaviors that then remains constant as they move through various categories (particularly age categories).

    In other words, do old people support the GOP because they are old people and that is what old people tend to do, or do they support the GOP because the current crop of 65 to 80 year olds came to political adulthood in an era when the reaction to the New Deal, and then the Great Society, really set in? If the first then one can expect that as society ages the GOP will go stronger. You hear this a lot from conservatives who dismiss the popularity of President Obama among younger voters by saying "they will get conservative when they grow older, people always do." If the second, one can expect that as new age cohorts enter the "elderly" category the political attitudes of that group will change. Thus many Democrats who, with a certain lack of tact, predict a long-term victory funeral by funeral.

    My suspicion is that it is actually some of both, but with the cohort explanation being the stronger. But it wasn't clear what you meant in your comment. Apologies for straying off the point.

    ReplyDelete

Who links to my website?