The Fallacy of Beneficial Ignorance


by Bent Flyvbjerg

In this way, through ‘‘mechanisms of self-deception,” the Hiding Hand tricks decision makers into taking risks they would not otherwise have taken, which is good, according to Hirschman, because this teaches them (a) what the risks are, (b) that the risks are manageable, and (c) that it is therefore okay to be less risk averse for future projects (Hirschman, 1967a, p. 34). ‘‘The Hiding Hand is thus essentially a transi- tion mechanism through which decision makers learn to take risks, and the shorter the transition and the faster the learning the better,” explains Hirschman (1967a, p. 28, 34; emphasis in original). Quick learning is desirable because it makes deci- sion makers and planners ‘‘able to differentiate between acceptable and nonacceptable risk,” thus bringing risks down over time with the help of the Hiding Hand. ‘‘You have to do all these fool things before you do the sensible things,” Hirsch- man wrote in his field notes (Bianchi, 2011, p. 18).

In sum, the Hiding Hand does its work through beneficial ignorance—or ‘‘ignorance of ignorance” as Hirschman (1967a, p. 35) put it. Beneficial ignorance hides two things from decision makers. First, the true costs and difficulties of projects remain obscure, which makes planners take up pro- jects they would not have considered, had they known their true complications. Second, the Hiding Hand also conceals planners’ problem-solving ability in dealing with costly and difficult projects, which makes planners able to successfully deal with such projects when the difficulties manifest them- selves, at which time the planners’ concealed abilities become overt and save the day. Overall, the Hiding Hand is therefore a good thing, according to Hirschman and his followers, because it makes projects go ahead and succeed—based on beneficial ignorance—that would not have done so had plan- ners known at the outset the real costs and difficulties, exactly like Willie Brown argues in the introduction above.


It turns out that Hirschman’s ideas about the Hiding Hand are based on an exceedingly small number of observations and biased data. Hirschman studied only 11 development projects financed by the World Bank, spread over four continents (Hirschman, 1967a, pp. 2–3).11 To be fair, large amounts of empirical data were not available at Hirschman’s time, so it might have been impossible to collect data for a large sample, even if he had wanted to do this. Nevertheless, 11 projects is much too limited a dataset to support the wide conclusions drawn, as we will see. Moreover, Hirschman’s small sample size has typically been ignored or papered over in the sec- ondary literature on the Hiding Hand, leaving the impression that the empirical foundations of the principle are solid. This has become conventional wisdom, because, as often happens, the secondary literature has attracted more readers than the primary source. Elster (1983, p. 158), for instance, says that Hirschman did ‘‘numerous case studies” in support of the Hid- ing Hand; Catino (2013, p. 62) that Hirschman based the prin- ciple on ‘‘many projects analyzed.” Other authors have glossed over the problem of sample size by euphemistically talking of ‘‘a number of” (Cracknell, 1984, p. 17), ‘‘a series

of” (Alacevich, 2014, p. 139), or ‘‘various” (Lepenies, 2008, p. 450) projects studied by Hirschman, or they simply reported his results without reference to the sample size at all, like Klitgaard (1997, p. 1963). It is interesting to note that once a theory has become generally accepted and widely popular, apparently even highly regarded scholars do not bother to check and report on its empirical basis, not to speak of testing it. Based on such writings, which are typical of the literature on the Hiding Hand, readers have been led to believe that the empirical foundation of the Hiding Hand is sound. Only by going back and checking the primary source would a reader get an inkling that, in fact, a big question mark hovers over the principle of the Hiding Hand in terms of the most basic issue for any scholarly work, that of validity and reliability. World Bank officials encouraged Hirschman to enlarge his sample size. But Hirschman wanted to be able to personally visit each project and do his own interviews and observations on site, which made a large sample impractical. So he stuck to his 11 projects, despite initial critique, which was silenced by the popular appeal and positive reception of the work.

In addition to issues of sampling, a second problem with Hirschman’s study of the Hiding Hand is biased data collec- tion. Adelman (2013, p. 388) has systematically gone through the ‘‘abundance of field notes” behind the study. He found the notes convey ‘‘a strong impression of basic confusion.” Hirschman apparently ‘‘struggled to make sense of his obser- vations” and ‘‘was having difficulty keeping his endeavor clear,” because so many of the projects seemed idiosyncratic, whereas Hirschman wanted to ignore peculiarities in favor of ‘‘some broad brush strokes” (Adelman, 2013, p. 388, 396). One might speculate that Hirschman’s need for order and broad strokes became particularly strong exactly because of the confused state of his data collection. In any case, the principle of the Hiding Hand eventually became the main stroke.


No Responses Yet to “The Fallacy of Beneficial Ignorance”

  1. Leave a Comment

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: