Melberg, Hans O., Culture and Decision Making

_{[Note for bibliographic reference: Melberg, Hans O. (1996), Culture and decision
making: A review of an article by F. Gaenslen, http://www.oocities.org/hmelberg/papers/960904.htm]

Culture and decision making

A review of an article by F. Gaenslen

by Hans O. Melberg

Occasionally one comes across an article that is refreshingly original and full of rich
insights. The article "Culture and Decision Making in China, Japan, Russia, and the
United States" by Fritz Gaenslen is precisely such an article (World Politics
39 (1986), pp. 78-103). In this review I want to discuss his article a bit further - its
weak and strong points. This effort is a part of a larger work in which I try to assess
the quality of explanations based on cultural traits.

One author has written that trying to define culture is like nailing a pudding to a wall.
Although it is impossible to find a perfect definition, we need to have one to get going.
An analogy may illustrate the point: Definitions can be likened to a tool needed to climb
a mountain. Now, there are two problems: One may spend all the available time on
perfecting the tool and hence one never gets to climb the mountain. On the other hand, one
may fail by trying to use a tool that is not good enough. Gaenslen avoids both these
problems. He spends less than one page, right in the beginning, on the problem of defining
culture. Furthermore, his definition is an adequate tool since it captures many aspects of
the many-dimensional concept of culture. Because I believe this is important I shall
elaborate a bit more on this.

According to Gaenslen a culture consists of assumptions about 1. human nature, 2.
causality, 3. the possible, 4. the desirable, 5. the appropriate, 6. the nature of the
physical environment, 7. the relationship of human beings to their fellows (p. 78). The
question is then how good this definition is.

I believe this is a good definition because it focuses attention on what goes on inside
the individual - his beliefs and values - which in turn is a main determinant of his
actions (which in turn create aggregate behavioural patterns in society). Moreover, the
sub-categories captures many of the different types of beliefs that are important when an
individual act (though, not all. For more on the various sub-categories of beliefs see my
article The Information Required for
Perfect Prediction ). Finally, the definition is concrete enough to enable testing of
hypothesis' about culture because one may use questionnaires, experiments and other
methods to examine whether a society is characterized by certain beliefs or values.
"Which other methods?" one may ask, and Gaenslen gives a very original answer.

In short, Gaenslen has analysed 1000 fictional disagreements as described by 272 different
authors in 514 works. The belief is that the description of these fictional conflicts
reveal something about the real culture in these countries. For example, one may
hypothesise that in collectivist cultures subordinates are more likely to yield to the
superiors whenever there is a conflict, while the opposite would be the case in
individualistic cultures. One may then analyse the conflicts in the literature to see
whether the hypothesis is true or not. It may not be entirely original to use literature
to prove points about national character, but the scale in which this is done in
Gaenslen's article makes it much more convincing than isolated literary anecdotes.

There are, of course, many methodological problems both with this specific method and with
cultural research in general. First, there is the problem of overdetermination i.e. that a
phenomena may be explained both by reference to culture and by reference to rational
adjustment to circumstances. Gaenslen is aware of this problem and he presents five
example of behaviour which some authors have explained by reference to culture, while
others have used rational choice explanations of the same phenomena. One example of this
is the abandonment of the piece-rate incentive system in China (in the 1950s). Some
authors explain this by arguing that the Chinese were inspired by an anti-individualistic
culture. Others argue that China abandoned the system because they did not have the
personnel or the ability to make the system work properly. Both explanations are possible
and the problem is to determine which is the true explanation.

A related methodological problem, also discussed by Gaenslen, is where in the chain of
causes one should focus one's attention. The problem is this: One author may explain a
behavioural pattern by reference to the incentives created by the institutional
environment (i.e. a form of rational choice explanation). He may then argue that it is not
culture, but the institutional environment which determine the behaviour of individuals.
Against this another author may argue that the institutional environment itself is
determined by the culture - as Gaenslen writes : "people hold beliefs, attitudes and
values that lead them to create particular institutions (for example electoral
systems)" (p. 81). The question is then how one can determine which of these is most
important.

One method, mentioned by Gaenslen, is to study immigrants. By studying the Chinese in Hong
Kong and the Russians in the USA one may see which cultural traits that are independent of
the institutional environment. Once again this is a rather common recommendation and I
shall not focus on it here.

What is more interesting, however, is Gaenslen's arguments why there is scope for both
rational choice and cultural explanations. His first (implicit) argument is that in some
situations it is theoretically impossible to act rationally. Behaviour in such situations
must be explained with reference to norms and culture. (For more on those situations in
which it is impossible to make a rational decision, see Jon Elster's article "Some
Unresolved Problems in the Theory of Rational Behaviour", Acta Sociologica, 36
(1993), pp. 179-190. One example should suffice: It is theoretically impossible to make
rational investment-decisions because we do not know how much information we should
collect about how much information we should collect before we make the decision. The
belief that we can act rationally in these situations - what Elster calls hyperrationality
- may itself be a cultural trait in the Western world.) Gaenslen's second argument is that
some situations are so complex that for practical reasons humans do not manage to make a
rational decisions. Thus, in those situations one should rely on explanatory strategies
other than rational explanations (such as cultural explanations) . These are strong
arguments that not everything can be explained by rational choice theories and in
consequence there is a justified role for alternative theories.

Gaenslen's main interest is whether some cultures are more individualistic than others. To
test this he creates three hypothesises: First, in individualist cultures the power of
normative arguments (do X because it is your duty) by superiors is weaker than in
collectivist cultures. Statistical analysis of the literature shows a significant
difference between the American and the other cultures in terms of how often the superiors
"win" the conflict when they use normative arguments. Second, in collectivist
cultures the individual is less likely to win a conflict in a public setting. Once again
statistical analysis (of the literature) supports this hypothesis: A public setting makes
an American individual 6.2% worse off (compared to a public setting), while a Russian is
12.3% worse off. Third, collectivists are more likely to let the interpersonal
relationship affect the nature of the argument than individualists. This hypothesis is
also supported by statistical evidence In uneven relationships the Russians, Chinese and
Japanese were more likely to use normative arguments, while American superiors tended to
use the same arguments regardless of the status of the other person. Taken together these
three findings suggest that the American culture is more individualistic than the Chinese,
the Japanese and the Russian.

Well, one may say, this is not very new. We all suspected that Americans were more
individualistic than the Japanese, the Chinese and the Russians. Maybe so, but after
reading Gaenslen's article the old belief is more justified. One might say that whereas we
before had prejudices, we now have well grounded beliefs. In this light the article is not
a waste of time.

Although this is a very good article, it is possible to criticise a few points. First, I
am not entirely convinced that his findings are significant enough to convincingly prove
that the Americans are more individualistic than, for example, the Russians. For example,
the finding that Russians are 12.3% worse off by a public setting, while Americans are
6.2% worse off, is based on respectively 23 and 63 cases. I would then like to ask whether
this difference is statistically significant (and at what level). Furthermore, even if the
difference is significant is does not necessarily mean that a Russian individual is more
willing to submit to the collective than an American. As Gaenslen himself notes in the
beginning of the article, the Russians may agree with superiors in public (because of the
more severe sanctions superiors have in Russia), while sabotage the decision secretly and
in private. Hence, a Russian may be as much an individualist as an American even if he
more often looses in public settings. Lastly, there is one point of criticism which needs
some elaboration: the gradual weakening of confidence as the chain or arguments grows.

Some of Gaenslen's conclusions have gone through so many analytical steps to be proven
that they are not very convincing. Each step is plagued with some uncertainty and taken
together this may become significant if there are many steps. This may sound confusing, so
I'll give one example. One argument which require many steps is Gaenslen's speculation
that "given the slightest hint of goodwill Chinese, Japanese, and Russians will be
more likely [than Americans] to overreact with cooperation" (p. 99) in bargaining
situations. The chain to establish this is as follows: First, Gaenslen uses literature to
prove that the American culture is more individualistic than the others. This step
requires us to assume that literary conflicts reveal the real culture. One also needs to
assume that the method used to analyse the literature is reliable and not random or
biased. Each point is convincingly argued by Gaenslen, but we cannot assign the arguments
full certainty. Next, Gaenslen notes that there is a correlation between collectivist
attitudes and high interpersonal orientations. Once again this link is probably true, but
even so the step decreases the confidence in the ultimate conclusion. Lastly, research has
shown that those who are high in interpersonal orientation are very reactive to the
strategies of the bargaining partner. This conclusion, too, is convincing - but as always
there is some doubt. Taken together there is doubt at each step and added together this
makes me uneasy about accepting the final conclusion. (I should add that Gaenslen mentions
a similar problem in the beginning of his article. As he writes "the smaller the
conceptual distance between cultural variables and what one wishes to explain by them, the
more compelling a cultural explanation is likely to be" (p. 82). Yet, his arguments
about the implications of his results are presented with a tone of certainty, not caution)

However, these are minor points. Overall the article is very stimulating, well argued and
highly original. Not only are there interesting major insights, there are small empirical
results which in themselves makes it worth reading the article. For example, Gaenslen
documents the impact of censorship on the description of conflict in literature by a
statistical analyse of the difference in how conflicts are described in censored and
uncensored literature. Altogether, minor and major points considered, this is an article
which I strongly recommend.

_{[Note for bibliographic reference: Melberg, Hans O. (1996), Culture and decision
making: A review of an article by F. Gaenslen, http://www.oocities.org/hmelberg/papers/960904.htm]}}