Herbert Simon Was Right: Rationality is an Accomplishment, Not an Assumption

A friend of mine just posted a link on Facebook to a website for comparison shopping, Priceonomics: The Price Guide for Everything, and he captioned the link: “Marginal utility maximizers rejoice! Perfect information is just a mouse click away!” This reminded me of an Annual Review piece I read earlier this week on Herbert Simon’s work and his influence on political science (Bendon 2003). Simon is known for many thing, but probably is most associated with the program of “bounded rationality” (BR). BR argues that decisionmakers – individuals and organizations – are not rational because rationality is impossible to perfect, and costly to approximate. Gathering and processing information is difficult, and even more difficult when the space of choices available is unknown ahead of time, and the connection between choices and outcomes is, at best, predictably random, and at worst nearly impossible to determine.

In the review piece, Bendon notes that there are two main branches of BR approaches – that pioneered by Simon (and March), and that associated with Kahneman and Tversky (K&T). These two approaches are compatible, but different in focus. K&T focus on universal or nearly universal errors in judgment (prospect theory) or perception (framing effects, where the description of otherwise identical choices can influence decisions). Simon, on the other hand, was more interested in processes of decision-making in light of the impossibility of by-the-book rationality. For example, Simon argued that organizations were often more rational than individuals because organizations can bring together many specialized individuals (and, I would add, tools, cf. Callon and Latour 1981) towards more clearly specified ends. Thus, rationality might be a better approximation of firm behavior than of individual behavior (though, of course, Simon and all manner of organizational theorists have shown how organizations too have systematic problems with decision-making). But the broader point is that Simon, who turned from political science to cognitive science, was interested in general processes of thought, and ways that people and organizations managed to get things done reasonably well in spite of the impossibility of rationality (hence research on things like “heuristics”, and his later work on expertise).

One thing that the advance of information technologies, especially those distributed openly via the web, has done is to somewhat level the playing field. That is, individuals now have access to a wealth of information on myriad topics that would previously have only been available to specialists working in collaboration. In other words, it’s easier to be more rational about more kinds of decisions now than it used to be. On the other hand, firms too have become increasingly rational – think of Walmart’s supply chain management practices, or Target’s creepily perceptive advertising techniques. But overall, I wonder if the balance hasn’t actually tipped in favor of individuals (at least when it comes to the kinds of decisions about which reliable information is now easily available online, and the kind of individual with enough savvy and resources to access those tools). Of course, the availability of that information is itself contingent on social arrangements (e.g. Wikipedia as a problem in maintaining a commons), and potentially polluted (by the nefarious forces of “Search Engine Optimization”, for example). But overall, I wonder if rationality (of a certain, intendedly maximizing sort) is becoming a better model for many decisions than it ever was. As people become more enmeshed in their tools – as the actor-networks of individuals expand and come to include more and better databases and search engines and so on – we become more like organizations in our decision-making. In other words, Simon was right, rationality is a variable accomplishment, and not just an inaccurate description.

About these ads

Get every new post delivered to your Inbox.

Join 151 other followers