Most cartel enforcement stories follow a familiar script: competitors agree on price, get caught, the cartel dies, and prices fall back to earth. Except, some cartels never touch price directly: steel producers close blast furnaces, and airlines quietly agree not to add seats. The collusive device is capacity, rather than price.
This week in Publications: “Capacity Collusion and Overcharge Estimation,” in Review of Industrial Organization (Fagart & Boshoff, 2026).
Unlike prices, capacity has memory. Once installed, it survives long after the investment decision that created it. Prof Willem Boshoff (CCLE Co-director) and Dr Thomas Fagart (CCLE Research Associate) argue this changes how cartel damages should be modeled. The authors formalise this with a dynamic model of capacity investment under partial irreversibility, focusing on what it does to prices before, during, and after a cartel operates.
Simulating the equilibrium path, they identify three effects: the ramp-up, the wind-down, and the lingering effect. The lingering effect considers capacity changes during the cartel that can permanently redirect the industry’s long-run equilibrium path. In this way, the post-collusive market may never fully return to where it would otherwise have settled. Depending on demand conditions, this effect may even persist indefinitely.
From a quantitative perspective, the authors show that these dynamics may bias overcharge estimation itself. Both the forecast method and the dummy-variable method, which rely on pre- or post-cartel data to construct a “competitive” benchmark, can be pushed off by as much as 40%.
For enforcement, they find that post-cartel price persistence isn’t automatic evidence of continued tacit collusion. Some of it is simply the mechanical residue of capacity hysteresis, with real consequences for post-cartel monitoring and for how the gravity of an alleged offense gets assessed.
Read the full article here: Fagart_et_al-2026-Review_of_Industrial_Organization