Abstract
We show how collusion in market entry can be detected from entry choices alone, without price data. In an infinitely repeated entry game, competition and collusion generate opposite responses to persistent cost shocks: an adverse shock reduces joint entry under competition, but under collusion, it can reduce the value of punishment by more than it increases the gain from a one-shot deviation, so that the cartel must allow more entry. Monte Carlo evidence shows adequate power with a few hundred markets. In the Ohio school milk cartel, participation rises with persistent cost increases where cartel activity concentrates.