Tuesday, July 05, 2011

Review: Judgment in Managerial Decision Making, by Max H. Bazerman

The text of this review crossed over a little with the opinions I shared with the person who recommended it.  I feel sort of weird about that, but hopefully it's changed up enough...

Well, I liked the book, and I didn't expect to, although my skepticism was probably a function of the title and the intended audience of MBA types. I have enjoyed reading about cognitive biases and how they govern our interactions in the past, and Bazerman took me some distance beyond what I already knew or figured out, and he was kind of enough to elaborate. In addition, filling the concepts out with the accepted vocabulary has since proven handy. Some of these biases, like the fallacy of small numbers ("fastest-growing" is a fine indicator of something unimportant), have been long-standing pet peeves of mine, while others, like the Dunning-Kruger effect, where incompetent people are the ones most likely to judge themselves as skillful and correct, were more like inchoate little complaints needing the help of some wise academician to identify and label. Things like "competitive escalation" I readily identified as true (he said half-seriously) because I've seen them in innumerable cartoon plots, while others, like these associative heuristics he warns against, are a little closer to my own areas of psychological susceptibility (in Mr. Downer's case here, intense negative feelings that get weighted too heavily in matters of self-judgment), and were not at all as fun to read, even if I did my best to keep an open mind about them. At least I was comforted that I don't share the thinking patterns of MBA students.

I do want to point out that some biases can be more rational than Bazerman lets on. People judge probabilities poorly, and we know it. While it's fine to assume accurate odds or trusted outcomes for the purposes of a word problem, I think the whole "X will earn 4% interest", or "Y has a 60% chance of failing" thing has, in real life, little chance of being true. The assumption of randomness is often false. The stock or housing market doesn't always grow over meaningful time periods (imagine, for one well-chosen example, that you were foolish enough to start saving in it ten or fifteen years ago) and inflationary discounting omits a lot of factors that actually matter in the cost of living. In real life, people often lie to their own advantage about statistics and odds, or assess them badly, or, of course, weigh them down with the offeror's own assortment of mental fuckups. To his credit, Bazerman gets into that a little in the decision-making chapters, but these remain tragically neat little story problems too, clearly and unrealistically mapped on an x-y axis.

The idea that people suffer from many cognitive biases that reinforce their positive illusions is a powerful one I think, and in my opinion, it goes farther than, say, the entire field of economics in explaining the shape of human society. [I keep entertaining a longer post about this, but every draft I write has come out with too much crackpot in it, and I am furthermore not completely sure where I stand on the larger details.] Maximizing your self-interest, or maximizing mutual interest, are things that go only so far, and there are all sorts of cognitive hiccups that must be served in the meantime. That is, any rational analysis of cognitive bias shouldn't exclude the observation that we tend to believe that we are more rational than we actually are. In the edition I have, Max Bazerman addresses some of these erroneous conceptions of fairness in the appropriate chapters, and does draw some broad pictures of, say, resource conservation on those terms (the usual tragedy of the commons, now with a new twist) and how it suffers. Even the rational strategies he outlines for negotiations (where sometimes there is mutual benefit to ceding power) appear that they could gradually produce inequality, and this does also get a brief and late mention in the short book, some time after I angrily scribbled my note.

I have a few other blurbs that I had noted as more useful or interesting to watch out for. Anyone want some insight to what makes me tick?

  • I thought it relevant that we tend to count gains and losses more in terms of the number of occurrances than as a net value. I often like to laugh at my particular brand of (self-created) bad luck and frustrations, but I do try and realize I'm basically a privileged little dude. He recommends setting up a mental ledger for this sort of thing, and I was happy that it affirmed something I thought was useful, but hadn't ever outlined explicitly.
  • He had a sentence that actions tend to produce more short-term regrets, but failure to act tends to haunt us longer. I am not really sure I agree with that, and I don't think it's a good decision-making heuristic, but I know a little of both sorts of regret, and of course it is indeed wise to weigh the cost of inaction too. (Someone get Neil Peart on the phone, I've got an idea for some song lyrics.)
  • I'd mentioned this in a conversation, but there's a note that depressed people might be less susceptible to positive fallacies. (I suspect we might overvalue negative ones though.) Implicit in the discussion is an observation that a facility for thinking about consequences and counterfactuals is also pretty depressing. Can't disagree with that one.
  • The idea of anchoring to arbitrary reference points is a useful thing to think about, but an annoying one to pretend to quantify. (If I read about the fucking "Overton Window" one more time...) Bazerman writes that in negotions, an offeror's anchoring value is less powerful if you know your alternatives and objectives. I think we can file that one under "cheap advice." On the other hand, remembering that many people have a tendency to reference the status quo is useful, keeping in mind the half dozen caveats I mentioned above.
  • I have been looking for a new job. Bazerman makes a point that people tend to over-emphasize things like prestige, and undervalue the work environment, how happy we will be, when considering these things. Couldn't agree more, and it's a lesson hard-won. (You know, provided you're in a position to be choosey. I wouldn't mention this point to the character in the next book review.)
  • As a final note, I find "regression to the mean" a highly annoying term, mostly for some vocabulary issues with "regression" and "mean" as used in the phrase. Turns out, it's a meaning about as close to the original ideas of regression analysis as you can get. Well, they got better.

  • No comments: