The analytical standard
We combine economic theory, econometric analysis and a detailed understanding of the market. Methods are selected for the question at issue, and their assumptions, strengths and limitations are explained.
The analysis examines the position a claimant would have occupied without the infringement. In purchaser claims, overcharge, passing on and the associated volume effects need to be assessed together, with attention to differences across transactions, customers and time.
Estimate the difference between observed prices and those likely to have prevailed without the conduct. The analysis accounts for relevant market developments and considers how any overcharge varies across products, purchasers and periods.
Assess the effect of the conduct on sales volumes and the associated profit or contribution margin. This requires a view of demand responses and the costs that would have changed with those sales.
Examine whether and to what extent an overcharge was reflected in downstream prices. The assessment considers pricing practices, market conditions and the related effects on sales volumes, avoiding double counting across components of loss.
02 / Three families of evidence
Benchmark outcomes against periods or markets unaffected by the conduct. Before-and-after, cross-market and difference-in-differences analyses require careful assessment of comparability and of other factors that could explain the observed changes.
Examine costs, margins and business performance to inform the counterfactual. Financial records need to be interpreted in their commercial context, including the treatment of costs, investment and normal returns.
Model competitive interactions and simulate outcomes without the conduct. The assessment depends on the economic mechanism represented, the information used to estimate the model and the sensitivity of its predictions to alternative assumptions.
Different methods can provide complementary perspectives. We assess their underlying assumptions, compare results and explain material differences, using the level of analytical detail the evidence and the matter require.
We assess how estimated effects vary across the relevant products, purchasers and periods before considering their combined implications. The treatment of interest, valuation dates and any other adjustments follows the applicable framework and the circumstances of the claim.
Where different specifications or methods produce different estimates, we examine the reasons. We explain the basis for our conclusions and, where appropriate, a central estimate or range, together with the sensitivities that matter.
We distinguish uncertainty arising from the data from uncertainty about the model or its assumptions. Sensitivity analysis, robustness checks and alternative specifications show which conclusions are well supported and where judgement remains necessary.
05 / The replication commitment
Our analytical work is supported by documented data preparation, code and a record of modelling choices. Subject to confidentiality arrangements, another economist should be able to trace the reasoning, examine the assumptions and reproduce the principal results.
The framework draws on Quantifying antitrust damages: towards non-binding guidance for courts, the 2009 study prepared by Oxera and its advisers for the European Commission. Method selection is case-specific and subject to applicable national rules of evidence. Study ↗