Estimating Overcharges in Antitrust Cases Using a Reduced-Form Approach: Methods and Issues

Abstract

This paper presents several methods and discusses salient issues pertaining to the use of reduced-form models to estimate overcharges in antitrust matters (e.g., price-fixing) where "but-for" prices may be less than actual prices during the anticompetitive period. In particular, two common types of reduced-form estimations are discussed: the "dummyvariable approach" and the "forecasting approach". Under either methodology, an error correction model is then specified as one way to address technical problems often found in applied time-series analysis-nonstationary data and the existence of short-term and longterm dynamics.

Heterogeneous expectations in asset pricing: Empirical evidence from the S&P500

“…We can observe them indirectly by estimating the fully identi…ed reduced-form system. The reduced form econometric model is As we need to estimate four reduced form coe¢ cients and we have four unknowns, and the structural parameters and are present in both (19) and (21), all coe¢ cients can be retrieved; see, for example, Nieberding (2006) who describes a similar issue surrounding the identi…cation of linear supply and demand functions. Because we are dealing with estimated coe¢ cients with an accompanying standard error, the structural coe¢ cients , d, , and cannot be found analytically and thus have to be retrieved numerically.…”

Competition Law Enforcement in the BRICS and in Developing Countries

“…The ability to measure the overcharge accurately depends upon how reliably and precisely the analysis can distinguish the collusive effect on prices from other influences that are unrelated to the anticompetitive conduct (Nieberding 2006). There are two ways of doing this.…”

“…If the influence of supply and demand factors affects equilibrium price differently in these two periods, then the DV approach may not reliably measure overcharges vis-a-vis an approach that accounts for such a change (Nieberding 2006). There are several important assumptions that are implicitly being made when one approach is selected over the other.…”

Cartel dating

“…Also, the use of explanatory variables that may have been affected by the cartel, such as capacity, should be avoided. (For further discussion on the relative merits of both approaches, see Godek, ; Marshall & Marx, ; McCrary & Rubinfeld, ; Nieberding, ; White, Marshall, & Kennedy, )…”

“…Many recent cartel studies make use of dynamic regression models to quantify cartel overcharges (see Bernheim, ; Davis & Garces, ; Nieberding, ; Roeller & Steen, ; White et al, ). It is natural and advantageous to include lagged prices as additional explanatory variables.…”