Trying to organize my thoughts and my readings on the application of behavioural economics to regulatory agencies, it seems to me that there are three sources of bounded rationality in the behaviour of regulators, and all of them have shown up in different papers, although no paper has addressed all of them together:
-Intrinsic preferences: real life regulators are neither benevolent social planners nor bureaucrats à la Niskanen, that is, self-interested and budget maximizers. Rather, most of them care about what they do, perhaps because they have career concerns, but also because they have a public sector ethos. Particular details of these intrinsic preferences depend on experience, professional background, nationality or other aspects of personality.
-Non-optmizing behaviour: in his PhD thesis papers, the regulatory economist Paul Joskow tested a conceptual framework, based on the ideas of Herbert Simon, where regulators were not maximizing any objective function, but they made their decisions consistently with a satisficing behaviour. Years later, some papers have assumed that regulators are adaptive agents, or that they behave with "minimal squawk."
-Expert biases: along the lines of a recent paper by Kovacic and Cooper in the Journal of Regulatory Economics, regulators suffer from expert biases such as over-confidence, confirmation or availability. These biases also affect ordinary people, but they have been shown to be particularly acute in regulators and high profile experts.
As a result of these departures from traditional assumptions, positive models of regulatory behaviour, as well as policy or institutional recommendations based on these models, should carefully fill in the details of the consequences of these real-life regulators.