Cross-Cultural Planning Fallacy Research

Cross-Cultural Research

The cognitive analysis of the planning fallacy implies that the phenomenon may also be robust across cultures. Japanese culture provides a particularly interesting test case for studying the optimistic prediction bias. Research to date suggests that Japanese adults show little or no self-enhancing biases in self-evaluation (and in fact, sometimes show a self-effacing bias) (Heine & Lehman, 1995). However, the level of “cognitive” overconfidence, as revealed in studies using general knowledge questions, is as great in Japan as in the United States (Lee, Yates, Shinotsuka, Singh, Onglatco, Yen, Gupta, & Bhatnagar, 1995).

We (Buehler, Griffin, Otsubo, Lehman, & Heine, 2000) distributed identical questionnaires (except for the Japanese translation) to participants at two universities: one Japanese and one Canadian. Participants named a school assignment that was due in the next 4 weeks, reported best-guess predictions for that project, described their thoughts while they made their predictions, reported when they typically finished similar assignments in the past, and later reported when they actually finished the target assignment. Both Japanese and Canadian participants showed the prototypical dissociation between their current predictions and their reports of typical completion times. Participants from both universities expected to finish the assignment nearly 2 full days before its deadline, despite reporting that they typically finished similar assignments less than 1 day before deadline. In fact the target assignments for this study were finished, on average, about half a day before the deadline. The resulting optimistic bias in prediction was slightly (although not significantly) greater for the Japanese than for the Canadian students. Furthermore, in a longer-term pilot study of summer-long research essays, the Japanese students were markedly optimistic – predicting that their essays would be finished 10 days before the deadline, but completing the assignments only 1 day beforehand (and recalling that similar tasks had been completed about 1 day before the deadline).

The thought listing measure also showed a very similar pattern across cultures. Consistent with previous findings, most participants (70%) focused on their plans for completing the task, whereas only about one-quarter of the participants (26%) referred to their past experiences (which implied that a more conservative prediction was in order), and very few participants contemplated the potential problems and delays they might encounter. The Canadian and Japanese samples did not differ significantly in their use of any of the thought categories.

Finally, to examine attributional processes, we asked participants to describe a past prediction failure of their own and that of a school friend and then explain why each of these had occurred. Once again, we found a moderate self-serving bias in Canadians’ attributions for failures to complete tasks by the expected time: They reported causes that were more external, transitory, and specific than the causes they reported for a friend’s late completion. For the Japanese students, however, the opposite pattern emerged: They attributed their own prediction failures to causes that were more internal, stable, and global than those of a friend. Thus, for Japanese students, there was no evidence of self-serving attribution biases that might “explain away” past prediction failures; in fact, the Japanese students seemed to take more responsibility for their own past failures than they assigned to others. This difference in attributional response raises the intriguing possibility that Japanese participants may be more influenced by reminders of past failures than Western participants.