Page 85 - Acertaining Economic Damages Calculation

ITT objected to the sampling methodology used by NiMo’s expert, arguing that his small sample sizes
resulted in the smallest errors in the sample having an enormously magnified impact in the final damag-
es calculation. The court, however rejected this argument and denied ITT’s motion to exclude NiMo’s
expert and, instead, ruled that his methodology would be more appropriately challenged through cross-
examination and testimony by ITT’s experts:

The threshold issue here is the propriety of statistical sampling in the context of this rather
unique litigation (the construction of a nuclear power plant)... [T]he court cannot accept ITT's as-
sertion that NiMo's sampling evidence is inherently speculative. It is true, as ITT points out, that
sampling has generally been allowed in factual settings vastly different than the present one... It
is equally true, however, that the evidence before the court, primarily in the form of damage re-
ports and experts' affidavits, undeniably has a "tendency" to show the amount of damages alleg-
edly sustained by NiMo... Estimates based on "sound sampling techniques" will, in some cir-
cumstances, suffice as the best evidence of proving damages. fn 29

This case reflects the court’s acceptance of statistical sampling as an element that may be used to calcu-
late damages. Although it does not provide specific guidance about how large the sample size should be
or how the sampling should be conducted, it does acknowledge that errors or problems in methodology
could be subject to challenge. In those instances in which the practitioner does not possess sufficient
knowledge to perform the sampling methodology, it may be appropriate to discuss with counsel the ben-
efits of bringing in an expert with this knowledge.

Best Evidence and Multiple Regression Analysis

Under certain circumstances, the use of multiple regression analysis may be useful to establish various
elements of a damage calculation with reasonable certainty, and its omission (without some appropriate
alternative) may leave the court with a lack of credible evidence to allow an award of damages. That be-
ing said, like any analytical approach used by a damages expert, it is important that the expert under-
stand how to correctly perform a regression analysis and to be careful to avoid common errors that could
invalidate its results. As discussed in the Reference Manual on Scientific Evidence:

Over the past several decades, the use of multiple regression analysis in court has grown widely.
Regression analysis has been used most frequently in cases of sex and race discrimination, anti-
trust violations, and cases involving class certification... [m]ultiple regression may be useful (1)
in determining whether a particular effect is present; (2) in measuring the magnitude of a particu-
lar effect; and (3) in forecasting what a particular effect would be, "but for" an intervening event.
fn 30

The following cases address the use of multiple regression analyses by experts.

fn 29 Niagara Mohawk Power Corp., 1992 WL 121726, at *36.

fn 30 Rubinfeld, Daniel, I., "Reference Guide on Multiplication", Reference Manual on Scientific Evidence, 3rd Edition. (2011). The
National Academies Press.

© 2020 Association of International Certified Professional Accountants 83