In aggregate litigation the damages model is not just a calculation. It is the gate the case must pass through, and it gets attacked years before any number is due.
Start a conversation with Cournot, the Institute’s damages concierge, already scoped to class & aggregate. Select a subject area to prompt it, or describe the dispute directly.
Aggregate litigation inverts the usual sequence. In a two-party case the damages model is built late and fought at trial. In a class or mass action the model is fought at certification, because the threshold question is whether impact and damages can be established for the class with common evidence rather than claimant-by-claimant proof. That makes the damages expert a gating figure: a model that cannot measure class-wide harm with a common methodology, or that measures harm the liability theory does not assert, can end the case before the merits are reached, and defendants brief exactly that. The methods themselves are the most statistical in damages practice. Overcharge estimation in antitrust runs on regression against benchmark periods and markets. Consumer claims lean on surveys and conjoint analysis to value the thing misrepresented or omitted. Wage and hour and product cases reach for sampling and extrapolation, proving the pattern from a representative subset. Each method carries a shared burden: it must handle the individual variation inside the class honestly, because the recurring attack in every aggregate case is that the average conceals members who lost nothing. This area covers the class-wide modeling problem, the antitrust overcharge machinery, and the sampling and survey tools, with the same boundary as everywhere on this site: what a given claim permits, and what a given court requires, is counsel's terrain.
The gate, the machinery, and the statistics.
Common impact, the fit requirement, and the modeling choices that survive certification scrutiny or end the case.
investigateThe but-for price, the benchmark that estimates it, and the pass-through question that decides who was actually harmed.
investigateProving the whole from a part: statistical sampling, survey and conjoint evidence, and the design standards that decide whether extrapolation survives.
investigateHow the Institute approaches an aggregate damages question.
Because certification turns on whether common questions predominate, and damages methodology is where defendants look for individual questions. The modern practice is rigorous scrutiny at the certification stage: courts examine whether the proposed model can actually measure class-wide impact and damages with common proof, and whether it measures the harm the liability theory alleges rather than some broader loss. An expert retained on the eve of the certification motion, presenting a sketch of a model to be built later, is a recognizable vulnerability, and the opposing expert's job is to show the sketch cannot work: that the data will not support it, that the class contains uninjured members it cannot identify, or that its logic proves individual rather than common harm.
The recurring structural attack on aggregate proof. Most classes are defined by exposure to the conduct, everyone who bought the product, paid the fee, saw the representation, but exposure is not injury, and inside almost any real class sit members whose circumstances mean they lost nothing: buyers who would have paid the same price anyway, employees whose hours never crossed the threshold, purchasers indifferent to the misrepresented feature. The economic questions are what share of the class that describes and whether the model can identify or account for them; the legal question, how many uninjured members a class can tolerate and at what stage, is contested and jurisdiction-dependent, which is exactly why counsel and the economist need to confront the issue together, early, rather than let the defense frame it first.
By an allocation step that is distinct from the aggregate model, and worth planning at the same time. A model can establish the class-wide loss while saying little about each member's share, and the allocation then runs on claims data: transaction records to compute each purchaser's volume, payroll data for each employee's hours, formulas that scale the aggregate to individual circumstances. Whether an aggregate-then-allocate structure is permissible for the claim is counsel's question, and the answer shapes the model itself: a court expecting individualized distribution will want the methodology to support it, and a model built with allocation in mind, preserving the member-level detail rather than collapsing it early, is more robust at certification too.
Usually, yes, because the center of gravity is econometric and statistical rather than accounting. An overcharge regression, a conjoint survey or a sampling plan is built and defended with formal statistical methods, and the expert will be examined on specification, robustness, significance and design at a level of technicality most commercial damages disputes never reach. The strongest teams often pair disciplines: an econometrician or statistician for the machinery, an industry or accounting expert for the institutional facts the model must respect, and, in consumer cases, a survey specialist whose design work meets the standards that field enforces. The architecture question, who relies on whom, deserves deliberate attention, because a chain of experts fails at its weakest link, and certification briefing will look for that link.
Describe the class and the conduct. The Institute will help you see whether a common methodology can carry the case.