Once bitten, twice shy? Market size affects the effectiveness of a leniency program by (de-)activating hysteresis effects

This article presents results of a laboratory experiment testing whether the effectiveness of a moderate leniency program depends on market size. Against theory, the results indicate that a moderate leniency program is not preferable to a policy which only includes fines for detected cartels, either in duopolies or in quadropolies. Surprisingly, the leniency program is not even entirely to be preferred compared to a laissez-faire policy. Yet, the leniency program seems to work better in markets involving four firms, compared to markets with two firms, especially in terms of the prevention of hysteresis effects.