When Should Policy Let Talent Move?

The policy subject is the mobility channel that creates innovating spinouts: non-compete enforcement can restrict it, while entry subsidies can encourage particular kinds of new firms. The decision is not simply whether more startups are desirable. It is which objective matters—more entry, higher-quality entrants, stronger incumbent innovation, faster aggregate growth, or greater household welfare—and which costs that objective is willing to accept. The evidence and model support conditional judgments, not a universal winner.

Removing a barrier changes the composition of firms

The empirical starting point is clear but limited. Across U.S. states, stronger non-compete enforceability is associated with fewer inventor spinouts. The relationship remains after accounting for parent patenting, inventor counts, firm age, and state conditions. A non-compete therefore matters at the margin where an R&D employee decides whether to leave and form a new firm.

That finding does not by itself answer the policy question. A state with stricter enforcement may differ in other institutions that affect innovation, labor mobility, finance, or entry. An index of enforceability summarizes a legal environment; it does not tell us whether a particular inventor’s contract was enforced or what the parent and employee would have done without it. Most importantly, the regression measures a relationship with spinout formation, not the final effect on growth or welfare after all firms adjust.

The calibrated model extends the question from the departure margin to the economy. Its baseline represents a setting with non-compete costs. When those costs are removed, spinout entry increases and aggregate growth rises by 0.10 percentage points. Consumption-equivalent welfare—the change in consumption that would leave a household equally well off—rises by 1.44%. In the comparison calibrated to the strictest enforcement environments, the modeled gains reach up to 0.28 percentage points of growth and 2.57% of welfare.

The mechanism is more interesting than the headline numbers. Greater mobility exposes the economy to more departures, so individual incumbents face more risk of losing an R&D manager. Yet it also creates more high-type entrants. Those firms innovate more and survive longer, shifting the composition of the economy toward organizations with greater growth potential. In the model, that composition effect is large enough to outweigh the negative incentive effect on incumbent firms.

This is a case where the firm-level and aggregate perspectives can point in different directions. An incumbent may rationally prefer a restriction that protects its research investment or makes it easier to retain a key employee. The aggregate economy may prefer more mobility if the additional spinouts bring in enough high-quality firms to generate new innovation and competition. Neither perspective is automatically the correct policy objective. They answer different questions about who receives the benefit and who bears the cost.

The model’s counterfactuals show why the distinction matters. If spinout formation is shut down entirely, aggregate growth falls by 0.33 percentage points and welfare by 2.0% relative to the baseline. The economy loses entry and, more importantly, loses a major source of high-type firms.