The Logic of Appropriateness

14 December 2007 at 7:52 am Nicolai Foss 11 comments

| Nicolai Foss |

To paraphrase Fritz Machlup, the rational-choice model has been a sort of “universal bogey” for many scholars in sociology, psychology, and management. The nature of the alternative(s) has, however, seldom been clarified. Thus, most models of bounded rationality are really variations on the basis RC model.

However, a much-cited attempt to characterize an actual alternative is James March’s notion of the “logic of appropriateness,” which may be characterized thus:

The logic of appropriateness is a perspective that sees human action as driven by rules of appropriate or exemplary behavior, organized into institutions. Rules are followed because they are seen as natural, rightful, expected, and legitimate. Actors seek to fulfill the obligations encapsulated in a role, an identity, a membership in a political community or group, and the ethos, practices and expectations of its institutions. Embedded in a social collectivity, they do what they see as appropriate for themselves in a specific type of situation” (quoted from this paper).

In in a logic of appropriateness, the agent/actor does not begin by identifying alternatives, preferences, etc. as in the rational choice model, but rather asks, “What kind of situation is this? Who or what am I? What is the appropriate thing to do given who I am?”

Obviously, this idea is very closely related to sociological notions of roles and to ideas about concerning the impact of norms, rules, etc. on action. However, the logic of appropriateness seems to equip action with an element of volition which is sometimes missing from accounts of action that stress roles, etc. Perhaps for this reason the notion has become rather influential perhaps particularly in political science (mainly through the joint work of March and Johan P. Olsen), but also in management and heterodox parts of economics. Thus, a number of evolutionary economists, such as Giovanni Dosi, have embraced it, perhaps because it may be seen as part of the badly missing micro-foundations of the routine construct.

In a recent paper, “Appropriateness and Consequences,” Kjell Goldmann, a political scientist at Stockholm University, very ably dissects the concept. Among other things, Goldmann shows that the two logics of action are in actuality strongly overlapping. I strongly recommend the paper for those who — like me — have been attracted to the concept, but have had reservations because of its mushiness. Here is the abstract of Goldmann’s paper:

Skeptical questions may be raised about the neo-institutionalist approach of James G. March and Johan P. Olsen, in which a “logic of expected consequences” is set against a “logic of appropriateness”: (1) it is difficult to determine what kind of constructs the so-called logics are—whether they are to be seen as perspectives, theories, or ideal types; (2) the logics, far from being mutually excluding, overlap very considerably; (3) analytical utility is debatable not only in the case of the “logic of expected consequences” but also when it is a matter of the “logic of appropriateness”; (4) the normative virtue of substituting a “logic of appropriateness” for a “logic of expected consequences” is less obvious than March and Olsen’s readers may be led to think. It is tempting to conclude that March and Olsen’s approach has proven compelling because of its consequences for the scholarly community rather than by virtue of its analytical appropriateness.

(HT to Nils Stiglitz for the pointer.)

Entry filed under: - Foss -, Papers.

Et Tu, Opera? Impact of the Commodore 64

11 Comments Add your own

1. Joe Mahoney | 14 December 2007 at 1:04 pm

In economics I was taught the following:

1. Rationality has to do with means (e.g., actions) being consistent with ends/goals.

2. As positiivists we do not pass judgment on the goals per se. For example if Captain Ahab wants to catch a big white whale, so be it.

3. If a person puts appropriateness and identity in their utility function as part of what that person wants to maximize, who am I as a mere positivist economist to pass judgment on the objective function?

4, My job as an economist is to set up the contrained maximization problem and serve my humble function.

In this light, I have trouble understanding on what basis any positivist social scientist would have anything to say about what is allowable or not allowable in a person’s utility function.

“De Gustibus est non desputadum.” (Of taste, one cannot quarrel).

As a footnote I add why would maximization of dollars be “scientific” and maximization of approriateness be “unscientifc” and “mushy”? Pehaps the primary issue is that concepts that are more easily measured (like money) drive theory and win the day.
2. spostrel | 14 December 2007 at 9:26 pm

I agree with Joe about the Protean ability of standard econ to accommodate lots of different preference types. The trouble, of course, is that all those free parameters make it harder to operate the positivist machine–you can explain anything with enough potential preferences. Hence the Chicago penchant for assuming homogeneous and unchanging preferences over standard objects (money, leisure, etc.). I think this approach is too restrictive and throws out a lot of systematic behavior, such as engineers’ love for implementing new technology and for being given strict and clear objectives (“just tell us the rules”).

A rule of bounded rationality that said some potential objects of preference simply never get considered could reduce the number of free parameters. So if it’s inapporpriate to think about Mom’s estate in some contexts, such as when deciding whether to call or visit, but appropriate in others, such as when helping her with her taxes, we might get seemingly inconsistent behavior from a standard econ point of view.

Say that the number of calls and visits is not correlated to the size of Mom’s estate but one’s tax advice is. A logic of appropriateness theory might–I stress might–be simpler than trying to cook up a utility function that rationalizes those choices. And it might make more predictions in other areas.

I’ve now exhausted my ability to defend anti-econ theory.
3. Nicolai Foss | 15 December 2007 at 6:09 am

Joe, I am not sure what you are getting at (and why Steve agrees). The logic of appropriateness is as positive as the standard RC model. “Appropriateness” does NOT have a normative dimension. It is “appropriateness” as seen from the point of the actor, NOT the analyst. It is a theory of how decision-makers set up their decision-situation.
It is true that March and Olsen somehow fits the logic of appropriateness into a broader picture of democracy, but that is really a different story.
4. Joe Mahoney | 15 December 2007 at 11:29 am

Nicolai, Your comment is reinforcing to my original point:

Putting appropriateness in the utility function is no more normative than putting money in the utility function.

Steve’s comments add a lot of insight beyond this point.
5. Nicolai Foss | 15 December 2007 at 4:27 pm

Joe, I may be misunderstanding you, but who said that putting appropriateness in the utility function was “normative”? I didn’t. March doesn’t. To provide a specification of a utility function is not “normative.” It seems to me to be entirely descriptive/positive. To pass moral judgment on the contents of the utility funciton is, on the other hand, to engage in normative reasoning. Is that what you mean? If so, we are in agreement.
6. Joe Mahoney | 15 December 2007 at 7:00 pm

I believe we are in agreement, Nicolai.
7. bee | 16 December 2007 at 7:07 pm

I think our friends March and Olsen are borrowing heavily from psychology on this one. If I am reading you correclt it sounds like they are offering a variant of Ajzen and Fishbein’s (1967) theory of Reasoned Action (please see http://www.ciadvertising.org/student_account/spring_02/adv382j/jagurule/Image1.gif)
8. Nils Stieglitz | 17 December 2007 at 6:22 am

A good, thoughtful paper on the logic of appropriateness is David M. Messick (1999): Alternative logics for decision making in social settings. Journal of Economic Behavior & Organization. Vol. 39, pp. 11-28. It presents some empirical evidence and points to three possible components of the logic of appropriateness. His key claim is that the “approach suggests that a judgment or perception of appropriateness precedes and mediates choice and other judgments, perceptions, expectations, and attributions.”

One should also not forget about Akerlof’s attempt to integrate identity into a standard utility-maximizing framework. However, this approach abstracts away the rule-driven nature of (some) decisions.
9. srp | 17 December 2007 at 7:54 pm

And we should also mention Alan Fiske’s argument that humans perceive all social interaction as falling into one of four distinct categories: authority relations, equality matching, communal sharing, or market pricing. Different people and different cultures place the same type of interaction (e.g. visiting Mom or planning Mom’s taxes) into different categories, which can lead to conflict; but in many cases, people in the same culture will agree on what should be handled which way.
10. Richard Makadok | 14 February 2012 at 2:51 pm

And I thought the whale was white… I’ll have to go back and re-read Melville, because Joe is always right.
11. Weighing Anchor | Teachtree Street | 6 July 2013 at 4:36 pm

[…] part of the American South as part of this emotional and professional move. Having a certain “logic of appropriateness” about my joining Teach For America — I felt compelled to become a young activist […]