For developers and builders of single-family communities and multi-family housing, an increasingly common tactic is for plaintiffs’ counsel to assert a class action, a procedural device that allows one person to sue on behalf of related persons who are not parties to the lawsuit. Plaintiffs’ counsel then relies on statistical extrapolation from limited inspections of a few homes or units to prove liability and damages.  For example, a plaintiff’s expert may inspect three residences in a subdivision of one hundred single family homes, find water intrusion around windows in those three residences, and reach a conclusion that the windows and associated flashing must be removed and replaced in all one hundred homes.

The statistical extrapolation also occurs at each building inspected.  For example, a plaintiff’s expert may perform test cuts at three locations on a building, find water intrusion, and conclude the entire exterior veneer must be replaced.  In addition to making a statistically questionable generalized conclusion from a small sample, plaintiff’s experts also exhibit a selection bias by choosing for their sample those residences, units, or test locations they believe are most likely to exhibit damages.  Among the many problems this tactic creates for developers and builders are the weaknesses of using statistical methods in this manner and the inevitable pressure defendants feel to settle when faced with a large class action claim even if the plaintiff’s claims are very weak.

Courts continue to struggle with the use of statistical proof in class actions, and the United States Supreme Court returned to the issue in its recent opinion captioned Tyson Foods, Inc. v. Bouphakeo, No. 14-1146, 577 U.S. — (March 22, 2016).  While the Bouphakeo decision relates to a class action for violations of federal overtime wages law, it is important for developers and builders because it limits or eliminates one avenue for forcing homeowners to litigate on an individual, rather than a class, basis and emphasizes that developers and builders should dig into the details of a plaintiff’s statistical proof.

In Bouphakeo, employees in a hog slaughtering and processing facility alleged they were not compensated for time spent putting on and taking off protective clothing required for their work and brought wage law claims on behalf of a class of employees.  To prove a violation of the applicable law, the plaintiffs had to show that by including such time, the employees worked more than forty hours in a week and thus were not compensated for some amount of overtime work.  The plaintiffs were unable to present any records of how long it took each employee to put on and take off their protective gear and instead relied on an average calculated by the plaintiffs’ expert based on “representative evidence.”  The average was applied to all employees of the plant, regardless of what department they worked in and the amount and complexity of their protective gear.  The defendant challenged the purported class on the basis that employees in different departments in the plant wore different types of protective gear, which required different amounts of time to put on and take off, and thus it was inappropriate to lump all employees together to resolve all of their claims through a class action.

The court declined to adopt a strict rule against the use of “representative,” i.e., statistical, evidence in class actions, and instead indicated that the use of such evidence is only limited by the evidentiary rules regarding reliability that apply to all evidence.  Therefore, the important point for defendants faced with class action claims is that a defendant must critically examine and highlight any weaknesses in the statistical sampling and calculations the plaintiff’s expert performs.  Defendant developers and builders can do so in two different, non-exclusive ways.  First, a plaintiff’s expert’s opinions can be challenged as unreliable such that they do not meet the standards for expert testimony and cannot be admitted.  Second, defendants can highlight the deficiencies in the statistical proof the plaintiff offers.  One avenue for highlighting deficiencies is for a defendant to retain a statistical expert to analyze the plaintiff’s statistical proof and the testing and methods used to produce that proof.  A second avenue is for a defendant’s liability expert, i.e., the expert retained to inspect the project at issue and evaluate the plaintiff’s alleged defects, to perform inspections and testing in additional locations and if possible, show that the plaintiff’s purported defects do not exist in those locations.