theories of risk

Early last month Charles Perrow, as one of the leading experts on organizational accidents, had something to say about the Japan nuclear crisis:

Currently our approach to risk is “probabilistic,” and the probability of a tsunami seriously damaging the Fukushima Daiichi plant was extremely small. But we should also consider a worst-case approach to risk: the “possibilistic” approach, as Rutgers University sociologist Lee Clarke calls it in his 2005 book Worst Cases: Terror and Catastrophe in the Popular Imagination. In this approach, things that never happened before are possible. Indeed, they happen all the time.

In what should be considered a classic case of the failure to take a possibilistic approach, consider this statement by Tsuneo Futami, a nuclear engineer who was the director of Fukushima Daiichi in the late 1990s: “We can only work on precedent, and there was no precedent. When I headed the plant, the thought of a tsunami never crossed my mind.”

Futami was not alone in his thinking. Experts throughout the nuclear industry and government regulatory agencies not only failed to predict the likelihood of a giant earthquake and tsunami, but also failed to examine the vulnerabilities of Fukushima Daiichi’s design to a natural disaster of this scale. Instead, they relied on a history of successful operation as an assurance of future safety. As a result, they ignored or underestimated a number of major risks that have since doomed the plant.

Perrow and Clarke hit on something really fundamental, and perhaps obvious, here: the way we assess or theorize risk partially determines how risky we think a certain action is.  In Knightian terms, risk involves decision-making where we “know” ex ante the probability distribution of a certain event happening. But there are lots of ways to determine the probability that a certain event will happen. Not all event likelihoods are normally distributed, and assessing the probability of an event occurring involves knowing something about the causes of that event. Thus, underlying the assessment of risk is a great deal of uncertainty. We just don’t know enough about the stuff we need to know to adequately assess risk all of the time.

It strikes me as interesting that organizational theory has very little to say about theories of risk, e.g., how it is conceptualized and assessed in different organizational settings. Clearly, this is something worth knowing about, and so I may just be missing out on a huge literature out there (maybe in the sociology of knowledge?) that addresses this.  One fertile setting for this study, I imagine, would be the financial market, where we see poor theories of risk were partially responsible for the financial collapse in 2008. A number of the papers in Lounsbury’s and Hirsch’s awesome two-volume set, Markets on Trial, deal with the problems that ratings agencies and in-house risk assessment divisions in assessing the risks of the mortgage securities market. Most of those papers, though, emphasize the complete failure to recognize those risks early enough.  But as Michael Lewis reminds us, not everyone got it wrong. Some financial specialists’ risk models were better than others (or perhaps they just had better access to information than the big banks did). Perhaps one way to get at this would be to do a comparison of the different models employed by these different specialists. What made certain specialist firms more likely to pick up on alternative models of risk? What accounts for the variation in trading strategies? Some financially literate scholar should get on that.

Written by brayden king

May 5, 2011 at 2:15 pm

10 Responses

Subscribe to comments with RSS.

  1. Daniel Beunza and David Stark have been doing just this sort of work in the financial context. See, for example, Models, Reflexivity, and Systemic Risk: A Critique of Behavioral Finance and Seeing Through the Eyes of Others: Dissonance Within and Across Trading Rooms.

    There are also a lot of fascinating historical studies on the emergence of probability and risk as styles of thought (Ian Hacking, Bernstein’s Against the Gods, etc.).


    Dan Hirschman

    May 5, 2011 at 2:36 pm

  2. Dan – great references. I’d forgotten about Bernstein’s book, which I read in grad school and is sitting on a bookshelf in my office somewhere.


    brayden king

    May 5, 2011 at 2:54 pm

  3. Organization Studies has dedicated a special issue to “organizations and risk in late modernity”:

    More in general I think that NN Taleb would also have something to say on our abilities to predict rare events.



    May 5, 2011 at 3:46 pm

  4. We should really have a better sociology of futures. The dominance of finance-mindset (Taleb’s Black Swan) on thinking about risk means that the importance of theories (how people reason about the future) are underappreciated, with greater focus on information.

    Taleb has said that he has no interest in theory, all his work assumes there is no theory of what will happen in the future beyond statistics. In fact, he is hostile to theory — suggesting that any causal explanation is likely to get it wrong and we should just rely on extrapolation from ‘hard data’. I would claim, however, that causal theories can very easily predict outcomes that are unique and new-to-the-world and therefore outside the scope of finance-mindset and statistical analyses of information.

    The distinction between theory-driven and theory-less prediction of the future might not be entirely accurate, however, since the financial model depends on lay theories of analogy — they must decide what portion of all past information we have represent situations analogous to the future event we are predicting. The categorization of essentially idiosyncratic past events in order to create statistical analyses of future risks seems a bit too unproblematic.



    May 6, 2011 at 6:22 am

  5. […] theories of risk « […]


  6. […] nice post from good-twin Brayden. He quotes Chuck Perrow: In what should be considered a classic case of the failure to take a […]


  7. Nice post, Brayden! Mary Kate Stimmler is doing some very interesting work in this area. She actually responds to Henri’s comment, developing theory about how the financial mindset took hold and how it altered the way people reason about the future. Very important stuff!


    Jo-Ellen Pozner

    May 6, 2011 at 4:10 pm

  8. The idea of doing meta-analysis on the success of certain risk models is very interesting. The popular debates on this issue are centered around what distributions we are drawing from for certain types of events. Mandelbrot and Taleb have argued that asset prices follow a Cauchy distribution, which resembles a normal curve superficially, but is an anomaly in that it has no mean or variance. This makes it hard to do the mathematical acrobatics that we see in the finance literature. Finance scholars like Fama have been aware of this argument since the 70s, but have opted to draw from the normal for convenience. Needless to say, it’s a huge industry now. A lot of this stuff is covered in MacKenzie (2006). [Incidentally, Mandelbrot was a student of Cauchy, who apparently thought he was too applied to qualify as a ‘real’ mathematician, but still worked with him, and Fama was a student of Mandelbrot!]

    I see Perrow’s argument as a substantive expression of these debates in more of a phenomenological frame.



    May 6, 2011 at 8:58 pm

  9. Hi, first time commenting here. I’m reading kobrin’s “managing political risk” (1982). it’s not the type of risk Perrow and Clarke focuses on, but it deals with perception and management of risk by MNCs.



    May 7, 2011 at 10:34 pm

  10. The really scary thing is not so much the “misunderestimation” (goofball word that works in this case) of “possibilistic” risk as it is systematic shifting of standards of acceptable risk seen in situations where regulators are too close to their charges. See link below for an alarming article about near-disaster due to faulty DOE/NRC oversight of a plant outside Chicago. It’s this kind of duplicitous fulfillment of oversight duties that is at the root of the meltdowns chronicled in many of the fine papers in the “Markets on Trial” issue of RSO that Brayden mentioned. Don’t get me wrong—I like the ideas put forth by Perrow and Clarke, but with all respect, there’s more to it than than just using the right framework for assessing risk. It’s also important to set up truly oppositional structures that keep people honest—and make it more likely that appropriately conservative standards will be used.


    Mark Kennedy

    May 9, 2011 at 3:35 am

Comments are closed.

%d bloggers like this: