Understanding “reverse” knowledge flows following inventor exit in the semiconductor industry

Organizational learning research suggests that employee exit lowers firm performance by eroding its human and social capital. We have a rather limited understanding of the conditions under which exit from a focal firm, defined as the firm from which exit takes place, may stimulate learning and reverse knowledge flows from the hiring firm. We developed a model of learning-by-exit to address this gap and tested it using a long panel of data (1985–2012) from the semiconductor industry. Our model suggests that the focal firm is likely to benefit more from reverse knowledge flows from the hiring firm when it is less aware of the latter. A focal firm is less aware of the hiring firm when there have been no prior inter-firm interactions between them, and when they are separated by a larger geographic and technological distance. Econometric analysis of our data using zero-inflated Poisson regressions provides empirical support for our model. This research contributes to our understanding of knowledge spillovers by highlighting the criticality of firm heterogeneity in the relationship between employee exit and reverse knowledge flows. • Employee exit constitutes an event that is significant enough to draw a focal firm's attention to another firm. • A focal firm losing an employee may learn by benefiting from reverse knowledge flows. • Learning by a focal firm after employee exit is greater when it is less aware of the hiring firm ex-ante. • Learning is more in absence of interaction between the two firms prior to the exit, i.e., no alliance or employee mobility. • Learning is more when focal and hiring firms are separated by a greater inter-firm technological or a geographic distance.