I’m having a conceptual problem with the idea of rationality we’re using in class. First I’d like to separate the two terms, rational expectations and rational agents.

Rational expectations as I defined it in class “having a model or conceptualization of reality that you update with regards to to new information received” As Dr. Greenlaw pointed out this description is vague enough to be useless and not really amenable to modeling. Any action can be rationalized as rational, even more so once you take into account the transactional costs of forming a model so even non-model based action is of itself in fact a calculated decision (the vending machine example). This conceptualization, which I think is more or less status quo of economic thought, means that even the rambling crazy guy is in fact “rational” because he obviously has an implicit model that tells him he will gain more utility by raving about aliens then he would by holding down a job. Now this might be true but I don’t really think it gets us anywhere closer to being able to model society as a conglomerate. We might be able make post hoc claims explicating peoples impetus for action but because most peoples model is implicit unto itself (not formally defined or articulable even to the people themselves) and only becomes (partially) apparent when revealed by action I don’t see how we can

The second issue I take with the idea of rationality is the fixedness of the agent. Say an individuals model is made up of preferences and costs. The individual is me. I really want a cheeseburger. I go to McDonald’s,  make a considered decision about whether the cost of the hamburger is worth it to me and decide to buy the hamburger. In that instant my tastes and preferences are revealed and I make the exchange for the cheeseburger and happiness subsumes me. But in the next instant, that just revealed preference is eliminated ,”I” no longer want another cheeseburger, my stomach is full and the cost of buying and consuming another one exceed the benefit. This phenomenon is normally just labeled diminishing marginal utility but I think that explanation understates the magnitude of what is changing. If I, as an agent and as a human being, am made up of a series of cost benefit models and an action that I takes changes those models, then it seems quite clear to me that the post transaction I is a different person then the pre-transaction “I”.

Now the cheeseburger example is intentionally trite, but I think the idea has farther reaching implications. Say every individual is perfectly economically rational, a pretty easy standard under the loose rules I talked about above. Not to harp on an old example but said individual would be totally justified in consuming large quantities of crack. If you’re going to completely discount the tastes and preferences of your future self then engaging in “self destructive” preferences at the present time isn’t really a cost. The future you isn’t you, its someone else that bears the consequences of your action. This mode of thinking clearly does not lead to a sustainable society. Thus society and institutions within society develop that force us to value the tastes and preferences of our future self, even if that means changing our “natural” preferences. For example education is clearly future weighted. On one level its designed to make our future selves more employable. Generally by high school most of us (especially the ones likely to be reading this) recognizing that it can be valuable for its on sake. But we didn’t start that way. At some point or another during elementary school , most of you decided that you didn’t want to be at school, that you’d rather stay home and play with toys. Forcing us to go to school, overriding the default preference of our younger selves to discount the future and in effect teaching us to value the preferences of our future selves changes lives. And as a socialized agent of society I’d say that our “lives” are better because of it, if we define life as the collection of different agents sharing the same physical shell. The total consumed utility of the individual is higher.

The take away point of that philosophical rambling is that rational agents utility models are A) impossible to determine in advance of revealed preference and B) that fully rational agents can and will make choices that negatively impact the value of their future selves lives in absence of societal or legal restrictions on those choices. I’m curious if anyone has an alternative interpretation or if anyone thinks you can still create a meaning model of the conglomerate choice given these caveats.