The Departure That Changes Two Firms

The subject now is the inventor’s departure itself: the moment when an employee leaves an incumbent innovator to help form a spinout. The key question is what changes on both sides of that move—whether the new firm gains a durable opportunity while the parent temporarily loses innovation capacity. Understanding this coupled lifecycle prevents a simple “knowledge escaped” story from hiding the costs and institutional choices that make mobility consequential.

One departure, two trajectories

A spinout event is easy to describe as a transfer of talent from an old firm to a new one. Economically, it is a state change for two organizations. The new venture begins with an employee who has experience, technical knowledge, and perhaps an opportunity generated near the parent’s technological frontier. The parent remains in operation, but the person who coordinated research or carried valuable context is no longer available under the same employment relationship.

To measure the parent-side effect, Baslandze and Vardishvili compare two kinds of inventor mobility. In one group, an inventor leaves a patenting parent and a spinout forms. In the other, an inventor leaves for an already existing incumbent firm. Both events involve mobility, so the second group is a more informative comparison than parents whose employees never leave. The question is not whether any departure is disruptive; it is whether a departure that creates a new firm has an additional effect on the parent.

The researchers use a dynamic event-study design around the final year in which the departing inventor patents with the parent. They restrict the analysis to firms experiencing a single mobility event during the sample period and follow outcomes over a ten-year window. The resulting sample contains 43,034 firm-year observations across 4,393 firms; 34% experience a departure that leads to a spinout, while the rest provide the incumbent-move comparison. Firm and year effects, along with time-varying controls for technology class, age, and state, absorb persistent differences and common movements.

The design’s most important diagnostic comes before the departure. The parent firms in the two mobility groups do not show divergent pre-trends in the reported outcomes. That does not prove that the groups would have remained identical under every possible future, but it makes the comparison more credible: the post-event difference is less likely to be a continuation of an already visible decline in the parents that later produce spinouts.

After a spinout departure, the parent’s innovation rate falls by roughly 8% in the following year and remains about 5% below its comparison path for several subsequent years. It returns to baseline at approximately the fifth year. The probability that the parent improves its technological lead shows a similar pattern, dropping by about 15% during the first few years before recovering. The immediate effect is therefore not only fewer research outputs; it is a reduced chance of moving ahead relative to the technological frontier.

Several mechanisms could produce this pattern.