Damages

• Pages: 167-178

167

CHAPTER VII

DAMAGES

A. Introduction

Conducting a scientific experiment that will yield reliable results

requires patience and attention to detail. Failing to follow the proper

technique or employing inappropriate assumptions could render the

results of the experiment worthless. An econometric damages study is

just such a scientific experiment, with the risk from not following proper

procedures that awards that may be too high or too low by millions, or

even hundreds of millions, of dollars.

An econometric experiment must incorporate a number of

fundamental elements for the practitioner to arrive at a valid conclusion.

These elements relate to the design of the experiment, the structure of the

economic model, the data and econometric techniques used to estimate

the model, the statistical tests applied to identify significant results, and

the economic interpretation of those results. Unfortunately, some

damages studies fail to address all of these elements properly.

Attorneys obviously have an incentive to ensure that their economic

experts do not misuse econometrics in the preparation of damages

estimates. Yet, for those unfamiliar with the theory and practice of

econometrics, there is a substantial leap from understanding a

fundamental concept to being able to identify fatal shortcomings in an

actual analysis. An examination of a few damages studies that failed to

make the grade may provide a more concrete understanding of the

significance and role of these elements. The following sections discuss

the econometric work in two cases: Conwood Co. v. U.S. Tobacco Co.1

and In re Industrial Silicon Antitrust Litigation.2 The econometric

analysis in each of these cases arguably suffered from at least one

1. 290 F.3d 768 (6th Cir. 2002), cert. denied, 537 U.S. 1148 (2003).

2. 1998-2 Trade Cas. (CCH) ¶ 72,348 (W.D. Pa. 1998).