Home > economics community > Rationality and Actions

Rationality and Actions

Via Mark Thoma, I see Dan Little commenting on rationality.   This is somewhat aside from the recent microfoundations dust-up, but it did bring the microfoundations debate back to my mind.

In particular, DL is commenting on the Akerlof and Kranton “Identity Economics” that they have been working on.   The idea is simple: your identity (say as a “police officer” or “school teacher” or “coal miner”) enters directly into your utility function, altering the payoffs for all of your choices.  I’ll just quote him, since its some smart thinking that I won’t easily be able to sum up:

What this comes down to, in my reading, is the idea that one’s “identity” creates a new set of payoffs for some actions, depending on whether the action confirms and enhances one’s identity fulfillment or whether it decreases one’s identity fulfillment..  Crudely, identity-consonance is a plus utility, while identity-dissonance is a minus utility, and actors balance first-order utilities and identity-consonance utilities in their ultimate choice of action. So this construction doesn’t deviate from standard rational choice reasoning much, if at all. Rather, it extends the cost-benefit calculation to include a new category of effect that the agent is hypothesized to value or disvalue–consistency / inconsistency with self concept.

This is a pretty limited conception of how identities work.  A more adequate treatment of identity as a substantive feature of social psychology ought to pay attention to a number of dimensions of practical rationality that are not included in this analysis.  (i) Cognitive frameworks.  Individuals with a specific identity may have distinctive ways of conceptualizing and experiencing the world.  These differences may affect behavior through mechanisms that are quite distinct from calculation of costs and benefits. (ii) Normative motivations. It is possible that people make decisions on the basis of their normative commitments, and that this process is to some degree independent from calculations of costs and benefits.  Moreover, it is possible that different groups have significantly different normative commitments. In this case individuals from different “identities” may behave significantly differently when confronted with apparently similar situations of choice. (iii) Group affinities / identifications.  It is possible that there is a social psychology of “solidarity” that has its own dynamic and behavioral consequences; and that this affective or motivational system has different characteristics in different groups. (iv) Emotional frameworks. It is possible that individuals absorb behaviorally important systems of emotions and feelings through their development within a specific cultural group; and it is possible that differences across groups lead to different patterns of behavior in common scenarios of action and choice.

So I think that Akerlof and Kranton are right to think that the theory of action associated with narrow economic rationality doesn’t do justice to ordinary decision making in a range of important cases.  They are right as well in thinking that the social psychology of identities and normative commitments is relevant to behavior in ways that cannot be pushed aside as “extra-rational.” But I don’t find their solution based on incorporating identity “utilities” into a larger utility function to be an adequate way of incorporating these broader considerations for action into a theory of the rational actor.

This is really smart, and I agree with nearly all of it.   One disagreement I would have is the suggestion (made obliquely, so I could be misinterpreting) that these group motivations are not “rational” in the “narrow economic” sense.   Strictly speaking each one of points (i)-(iv) can be represented in utility form as a maximization problem and are therefore “rational” in a “narrow economic” sense.   What Akerlof and Kranton are doing, IMHO, is identifying a simple model of pro-social behavior–which is to say they are attacking the “selfish” assumption, common in economics while maintaining the rationality assumption–which would be a problematic assumption to drop, if you really insisted on it.

But what DL really reminded me of is another point about utility and microfoundations that I don’t think is really emphasized enough.   It doesn’t aggregate.   Even the Marshallian demand functions, derived from the utility maximization problem, only aggregate in special cases.

This brings up the problem of usefulness, or lack thereof.   If utilities can’t be aggregated, formally, then there can be no representative agent (formally).   If demand derived from individuals using utility theory can’t be aggregated, then even heterogeneous agent models may not capture dynamics.   Which is to say, that the formalism for individual behavior that economists have adopted may not be very useful for describing macroeconomic behavior.   It does no good to say that utility theory encapsulates individual behavior if the problem we wish to solve is aggregate behavior–where utility theory doesn’t apply.

You may still be OK; these problems could be “small”.  That’s not what I want to talk about, though.  Instead, what if you founded your theory, not on desires (as is done in utility theory), but instead on actions?   I’m not sure what a rigorous theory that treats actions as fundamental would look like, maybe it isn’t possible.   Still, it seems reasonable, likely even, that such a theory would have very nice aggregation properties (for example, see this also via Thoma).   It might not be necessary to drop the rationality assumption from economics but it could be beneficial.

Categories: economics community
  1. No comments yet.
  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: