Page 80 - Cultures and Organizations

62 DIMENSIONS OF NATIONAL CULTURES

ity survives only where it is matched by obedience. Bernadotte’s problem
was not a lack of leadership on his side; rather, the Swedes had a different
conception of the deference due to a ruler from that of the French—and
Bernadotte was a Frenchman.
Comparative research projects studying leadership values from one
country to another show that the differences observed exist in the minds of
both the leaders and those led, but often the statements obtained from those

who are led are a better reflection of the differences than those obtained
from the leaders. This is because we are all better observers of the leader-
ship behavior of our bosses than we are of ourselves. Besides the questions
on perceived and preferred leadership style of the boss—questions 2 and 3
in the PDI—the IBM surveys also asked managers to rate their own style.
It appeared that self-ratings by managers resembled closely the styles these
managers preferred in their own bosses—but not at all the styles their
subordinates perceived them to have. In fact, the subordinates saw their
managers in just about the same way as the managers saw their bosses. The
moral for managers is: if you want to know how your subordinates see you,
don’t try to look in the mirror—that just produces wishful thinking. Turn
around 180 degrees and face your own boss. 7

Power Distance in Replication Studies

In Chapter 2, Table 2.1, six studies were listed, published between 1990
and 2002, that used the IBM questions or later versions of them with
other cross-national populations. Five of these, covering between fourteen
and twenty-eight countries from the IBM set, produced PDI scores highly
8
significantly correlated with the original IBM scores. The sixth got its

data from consumers who were not selected on the basis of their relation-

ships to power, who were in very different jobs, or, as in the case of stu-
dents and housewives, who did not have paid jobs at all. We investigated
whether the new scores would justify correcting some of the original IBM
scores, and we concluded that the new scores were not consistent enough
9
for this purpose. None of the new populations covered as many countries
or represented such well-matched samples as the original IBM set. Also,
correlations of the original IBM scores with other data, such as consumer
10
purchases, have not become weaker over time. One should remember that
the scores measured differences between country cultures, not cultures in an
absolute sense. The cultures may have shifted, but as long as they shifted