The cartel raised the price. By how much, for how long, and who ended up bearing it: three questions, each with its own machinery.
Start a conversation with Cournot, the Institute’s damages concierge, already scoped to antitrust overcharge & pass-through. Pick a starting point, or describe the dispute directly.
Antitrust damages center on a single object: the overcharge, the difference between the price paid and the price that competition would have produced. Estimating it means estimating the but-for price, and the field has settled on a small set of approaches. Benchmark methods compare the affected market to itself before and after the conspiracy, or to markets the conduct never reached, similar products, other geographies, and the modern implementations run through regression, controlling for the cost and demand factors that also move prices so that the conduct's effect is not credited with movements the economy caused. Structural approaches model the industry's competitive behavior directly and are reserved for matters that can carry their assumptions. The estimation inherits every discipline covered elsewhere on this site, clean benchmark selection above all, plus one that is distinctly antitrust: the during period is defined by the conspiracy's actual life, which is itself litigated, and prices may take time to fall after collusion ends, contaminating a careless after benchmark. Then comes the question that determines who was harmed: pass-through. A direct purchaser facing an overcharge may absorb it in margin or pass some of it down the chain to its own customers, and the rate at which that happens, estimable from margin behavior and cost shocks, allocates the harm across the distribution chain. Which levels of the chain may recover, and how pass-through evidence may be used, differs sharply across jurisdictions and claim types; that allocation of rights is counsel's terrain, while the economics of who actually bore the overcharge is the expert's.
Every method is an argument about what the competitive price would have been.
The market against its own competitive periods. Turns on when the conspiracy actually began and ended, which is litigated.
Products or geographies the conduct never touched, adjusted for the ways they differ from the affected market.
Cost and demand variables absorbing what the economy did, so the conspiracy variable measures the conspiracy.
The industry’s competitive behavior modeled directly. Powerful, assumption-heavy, and attacked on the assumptions.
How much of the overcharge each level of the chain passed to the next, estimated from margins and cost-shock behavior.
Non-conspirators raising prices under the cartel’s umbrella, and prices staying high after it ends. Real economics, contested recoverability.
How the Institute approaches an overcharge question.
The number, the class, and sometimes the theory of who can sue.
Every overcharge estimate is a comparison to something. If the before period was already collusive, the after period had not yet recovered, or the comparison market differs in ways the controls miss, the estimate inherits the flaw at full size. Interrogate the benchmark before the regression, because the rebuttal will.
The one whose benchmark the facts can defend, and convergence among several beats elegance in any one. Before-during-after is intuitive and uses the market's own history, but requires confidence about when collusion started and stopped, and clean periods long enough to establish competitive behavior. Benchmark markets avoid the timing problem but import a comparability fight. Structural models avoid both and replace them with behavioral assumptions the defense will attack line by line. Strong analyses often run more than one method, because independent approaches converging on similar overcharges are hard to dismiss as artifacts, while a material divergence between methods is a finding in itself, and better discovered by your own team than by the rebuttal.
From how the intermediary's prices respond to cost changes, observed in data. The standard approach studies documented cost shocks, input price changes, tariffs, freight movements, and measures how much of each moved into the intermediary's own prices and how quickly; the same relationship applied to the overcharge estimates what share traveled downstream. Margin analysis complements it: stable percentage margins over the period suggest costs were being passed along, while compressed margins suggest absorption. The economics depends on market structure, competition at the intermediary's level pushes pass-through up; buyer power below pushes it down, and honest estimates engage with that structure rather than asserting a convenient rate. The same analysis serves both sides of the docket, which keeps everyone disciplined.
Because the law splits the question and the economics has to serve both halves. In some regimes direct purchasers recover the full overcharge and pass-through is no defense; in others, and in many state and non-US frameworks, indirect purchasers may recover what reached them, which makes pass-through estimation the core of the case rather than a defensive footnote. The same underlying economics, one overcharge propagating down a chain, gets modeled from different ends depending on who is suing where, and parallel direct and indirect actions over the same conduct can value the same dollar of overcharge differently. Which regime governs a given claim is squarely counsel's question; the expert's obligation is a pass-through analysis rigorous enough to survive whichever frame it lands in.
Transaction-level pricing over a long window, and the fights over its scope begin early. A credible overcharge regression wants the defendants' transaction data, prices, volumes, products, customers, terms, covering the alleged conspiracy and enough clean period on either side to establish the benchmark, plus the cost and demand data the controls require: input prices, capacity, demand indicators. Pass-through adds the intermediaries' pricing and margin records. The practical counsel questions are about coverage and granularity: aggregated or averaged data can conceal exactly the price structure the model needs, and a clean period shortened by discovery limits weakens the benchmark. Data strategy is model strategy in these cases, and it is set during discovery, not at the expert deadline.
Describe the market and the conduct. The Institute will help you see what the overcharge estimation will require and where it will be attacked.