Please note: this discussion is based largely on the 1st edition of Understanding Public Policy. The second edition devotes most of chapter 7 to
Policy Concepts in 1000 Words: the Institutional Analysis and Development Framework (IAD) and Governing the Commons (podcast in S2 500 words series] [see also Policy Concepts in 1000 Words: it’s time for some game theory]
From Policy Concepts in 1000 Words: Rational Choice and the IAD
‘Rational choice theory’ is easy to caricature and dismiss, but difficult to define and describe, because it refers to a very broad and diverse body of work. So, we can identify some broad features but recognise that some studies display them more than others:

Inspiration – the application of ideas and methods from economics to politics.
Approach – models and deductive reasoning. It creates models of the world based on a small number of propositions and a logical examination of their connections.
Assumptions – ‘instrumental rationality’. Individuals fulfil their preferences according to their beliefs regarding the most appropriate means to achieve them. This is an ‘intentional’ explanation of behaviour based on the goals of individuals rather than motivation by ‘habit, tradition, or social appropriateness’*
Aim – to establish how many, or what proportion of, political outcomes one can explain with reference to the choices of individuals under particular conditions.

We can also identify two main types. The first is the abstract work which often involves building models or creating discussions based on openly unrealistic assumptions – for example, people have perfect information and judgement; they can act ‘optimally’ when faced with any situation.
‘Optimally’ is potentially misleading, since it refers to an ability to fulfil their individual preferences, by ranking them in order and being able to fulfil them. It does not necessarily refer to an optimal overall outcome, because things get complicated when many individuals, each seeking to fulfil their preferences, interact. We should also note that ‘rational’ refers to the ability to reason and act on reason (crucially, we do not have to assume that rational beings are selfish beings).
The second type involves more detailed and/ or realistic assumptions regarding the preferences of individuals and how they relate to specific institutional settings. In this case, the aim is to help explain outcomes.
The first type of work is a logical exercise, to help think through problems and often produce ‘paradoxical results’. Famous examples include:
[see the post for more]