Often, when a plaintiffs’ counsel seek to certify a class asserting a hard-to-prove financial injury, they will rely on a statistics or economics expert to demonstrate that there has been some kind of “common overcharge” for the product at issue.  This method is extremely common in antitrust class actions, but also shows up increasingly in various kinds of consumer class actions, including product liability class actions (“we would not have paid this much for a product with a defect”) and food-labeling class actions (“we would not have paid so much for Nutella if we’d known it was sugary”).

The “common overcharge” argument is seductive.  It provides a judge a common method of determining a key element of liability, offered by an expert, and backed up by math.  But, as economists have pointed out, it’s not necessarily accurate.  And this critique is not coming from the perspective of nailing down everyone’s damages to the last decimal point.  Instead, economists are concerned about adequately accounting for heterogeneity of demand when they construct hypotheses. That is, they’re trying to make sure that the model they have created actually acts like the buyers it’s supposed to represent.  Doing otherwise would mean either overcharging the defendant (not good for either deterrence or fairness) or under-compensating a class (not good for compensation or fairness).

In their article Turning Daubert on its Head: Efforts to Banish Hypothesis Testing in Antitrust Class Actions, economists Laila HaiderJohn Johnson, and Gregory Leonard take on the question of whether a single statistical technique can really account for a heterogenous group.

It has become standard in modern empirical economics to recognize and, where possible, account for, possible heterogeneity across economic agents(e.g., customers and suppliers) in their responses to changes in economic factors. For example, the well-known “BLP” approach frequently used in demand analysis allows for the possibility that different consumers have different levels of price sensitivity (including possibly zero sensitivity). Thus, for a defendant’s economist to raise the possibility that a regression model may differ across customer-suppliers is consistent with the current state-of-the-art in economics research.

(Footnotes omitted, emphases added.)  (“BLP” refers to “best linear predictor,” the old “fit the line to the scatterplot” method.)

In other words, there are solid economic reasons not to accept the plaintiffs’ expert’s use of a broad regression that downplays differences among class members.  In fact, it is often good Use of individual regressions/hypothesis testing in class actions.

The authors specifically discuss antitrust class actions, but their analysis easily transfers to products liability and consumer class actions as well.  In fact, any time a plaintiff’s economist claims she can come up with a simple averaged overcharge, defense counsel should be on guard.  As econometrics get more sophisticated, it is becoming clear that an “average overcharge” may be just about as useful as “average state law.”