Ideas tell a man what his interest are
In my dissertation I'm trying to expand standard public choice analysis (and therefore mainstream Chicago assumptions) to incorporate ideas. Whereas most attempts to deal with ideology will incorporate it alongside material interests, I think that ideas can only be analyzed within a subjectivist framework.
Consider Roy Keane's decision about who to join. It's clear that financial considerations are important, but not all-deciding. He is reward sensitive, but not reward maximising. The chance of winning honours will come into consideration, but so too will personal factors. Some players could never contemplate joining a rival, whereas others respond to such acrimony. These matters are psychological, and will bound and define the decision making process.
Joseph Schumpter refers to a "pre-analytic cognitive act", which is a stage of thinking that preceeds our standard analysis. It is similar to Israel Kirzner's concept of "alertness", which refers to the moment of discovery as perceived by an entrepreneur. By definition it is costless, meaning that there's a stage of human action above and beyond simple maximisation. Whereas most economists reject a term that is essentially uneconomical (after all, economics is about scarcity), it's always important to retain exogenous variables. If we are studying human action, there must be something essentially mysterious.
In short, I want to claim that expectations precede action. There is therefore a theory of the mind that is implicit, and costless. This does not imply that incentives don't matter, and prediction is impossible. However it does prove that ceteris parabus is a key assumption, before we can make a prediction. Although the truthfulness of the theory does not require empirical support, the applicability of it does. Consequently there's a vast role for applied theory to find out the subtlety of time and place that make ideas a more important condition than "mere" interests.