Labor Market Collusion through Common Leadership

To join the lecture via Zoom, please register here.

We study an alleged scheme involving more than fifty tech companies’ “no-poach” agreements. These are illegal agreements to suppress labor market competition by ceasing to recruit one an- other’s workers. We study entry into these agreements and their effects on workers. We first document an overlooked potential mechanism for collusion: what we call common leadership, where two competing firms share high-level leaders such as C-suite executives or board directors. Common leadership is prohibited by antitrust law under Clayton Act Section 8, but Section 8 was seldom enforced until recently. Using data on the networks of firm executives and board members from BoardEx combined with a transparent measure of collusion from court documents, we show that pairs of firms are 11 percentage points more likely to enter no-poach agreements after they begin to share common high-level leaders, even after conditioning on the identities of the firms. This supports increasing enforcement of the prohibition against common leadership.

After establishing the link between common leadership and collusion, we turn to the effects of the no-poach agreements on workers. We use microdata on workers’ employment histories from LinkedIn to show suggestive evidence that the agreements depressed labor flows across pairs of colluding firms. The effects are not limited to managerial and technical positions. Workers are unlikely to have been paid a compensating differential in exchange for their reduced labor mobility. First, no-poach agreements are kept secret from workers, removing the opportunity to bargain for a compensating differential. Second, we find negative impacts on promotion rates within the firms. We also find suggestive evidence in support of the Harrington (2004) hypothesis that firms caught colluding will only change their behavior after all legal damages have been determined, rather than immediately upon discovery.