How Get Good Managers : An Econometric Case Study of a Large Japanese Auto Dealership

M53

M54

Managers could have substantial influences over their subordinates’ productivity. Using the personnel and transaction data from one of the largest auto dealership in Japan, this paper discusses how large the impact of having a good manager is, how managers are assigned to workplaces, which managers are more likely to achieve high productivity, and how much managers can improve their performance through learning by doing. Our primary findings are: (1) replacing “bad” managers at one standard deviation below the average with “good” managers at one standard deviation above the average improves the branch profit by 14%; (2) new managers are assigned to small branches and laterally moved to larger branches as they accumulate experience. In addition, more experienced managers are more likely to be assigned to branches with declining profits; (3) young managers and those with broader experience tend to perform better than old ones and those with narrow experience only in new car sales; (4) the effects of learning-by-doing are small, thus, selecting good managers is more important than training managers.