From Variance to Perception: How Observing other Firms led Managers to Overestimate the Performance Benefits of New Product Development



In this paper, we examine variance-induced perception bias: when a particular strategy on average reduces firm profitability but also increases its variance, observers may develop the erroneous perception that the strategy is generally associated with superior performance. We argue we encountered such a bias when we examined new product development in the Chinese pharmaceutical industry in the period 1991-2000 (study 1). Retrospective interviews suggested that industry insiders believe that during this period, firms that engaged in new product development performed markedly better than non-innovators. Our quantitative analysis, by contrast, reveals that this belief is incorrect: in this setting, non-innovators on average significantly outperformed the firms that did engage in it. To explain this discrepancy, using multiplicative heteroscedastic models, we also show that new product development usually not only lowered firm performance but also significantly increased its variance, so that the very top performers in the industry were indeed often innovators. A complementary experimental study (study 2) confirms that people may overestimate the effectiveness of a strategy as a result of its higher variance. We discuss the implications of our findings for research on perception biases in strategic management, and draw attention to the importance of analyzing variance as a dependent variable in its own right.