From Books and Records to Bits and Bytes
Class actions — legal matters brought by a group of people who are “similarly situated" — can be costly endeavors. Workplace class actions alone cost companies $2.72 billion in 2017, according to Seyfarth Shaw.
When trying to address issues that can affect thousands, if not millions, of potential class members, companies and their counsel may struggle with how to manage the case and put forward the most effective evidence. Technology tools, including data analytics as well as advanced modeling and statistical techniques, are helping companies quickly parse fact from fiction and make the best decisions in the face of a class action. Here are four questions that can help inform and develop case strategy.
1. Is This Truly a Class-Wide Matter?
Even if you hear several convincing stories from plaintiffs about alleged patterns of behavior, the facts that emerge after you dig into the data are often very different.
This is where data analytics come in. “We take all of the relevant data and compare the anecdotal to the actual,” says Mary Beth Edwards, managing director and practice leader of commercial and healthcare disputes and forensics with Navigant.
The data helps identify exactly how often — if at all — the alleged bad act occurred in relation to people who have attributes of the identified class. This is then used to assess whether the affected parties meet the required definition of a class.
Angela Sabbe, director of disputes and economics with Navigant, notes: “The first thing we may do is analyze the data to examine two issues: Are the claims supported by what we observe in the data? And, are they representative of the rest of the defined class?”
Using the data, “We can quickly show whether it’s a pattern of practice affecting a class, or rather an individualized issue,” says Edwards.
“Class certification can involve analyzing millions of data points to evaluate class typicality or commonality,” says Sabbe. For example, “We can programmatically map latitude/longitude/GPS points, plot routes, and calculate distances. This is helpful for class actions involving routes/distances, or for class actions involving environmental events in which we want to assess proximity to the event,” Sabbe notes.
“Some matters, such as those involving allegations of fraud or misrepresentation, may benefit from systematic analysis of transcripts of telephone calls or other communications to assess whether communications to individuals were consistent,” says Debra Aron, managing director in Navigant’s economics practice. “In addition, in some matters, sophisticated data analysis can contribute to a determination of whether class members could be identified from the available information without individual inquiry,” she notes.
Whether it’s a certified class or simply a number of individually aggrieved parties, companies typically want to quickly understand their potential financial exposure and risks in a matter.
“Using complex data-modeling methods, it may be possible in some cases to quickly assess and quantify exposure,” says Sonya Kwon, managing director in disputes and economics with Navigant. “For some clients, this is important to know as soon as possible so they can identify sensible strategies.”
The better informed companies are about their exposure, the better they can make decisions going forward. “You don’t want to turn down a reasonable settlement offer only to end up paying vastly more later because you didn’t use the tools at your disposal to do your homework,” says Kwon.
“Econometrics can be helpful in quantifying impact or calculating damages, particularly when combined with proper economic modeling of the market,” notes Aron. “For instance, if adequate data exist, econometrics can be helpful in estimating the price that would have prevailed in a relevant marketplace if the alleged bad behavior had not taken place,” she explains.
It’s easy to misuse or misinterpret data if it is not put into the context of the relevant industry. Combining advanced analytics, quantum expertise, and industry insights enables a sophisticated analysis of the company’s data and can avoid errors or misunderstandings of the data.
“The model is only as good as the expert’s knowledge of the industry. If you aren’t modeling the right things, it won’t be descriptive of the process you’re trying to quantify,” says Aron.
For example, in Telephone Consumer Protection Act (TCPA) matters, machine-learning algorithms can be used to compare similar but different names from telephone subscriber data, credit reporting agencies, and other sources to determine the similarity or dissimilarity of names across sources. However, using these analytical tools effectively in a TCPA matter requires understanding of phone number porting, as well as why identifying the same individual with nonidentical names is important to ascertaining members of a class.
The data model can also be affected by the regulatory environment, so it’s important to stay up to date on changing rules. “Laws are evolving, which affects what is and is not relevant to your model,” says Sabbe. “For example, in labor and employment, if you are evaluating pay discrimination, in certain jurisdictions, job titles or previous salaries should not be used as explanatory variables, since certain states and cities prohibit using job titles and previous salaries as justification for pay differences,” she explains.
Data analytics can help companies focus on information at the heart of the matter — transactions relevant to the business practice at issue — and separate unrelated information.
“Data analytics enables us to assess the variability in the data, which we then use to help clients understand what’s really at issue in the case,” Edwards says. “Based on this insight, our clients can focus their discovery efforts to gather evidence and testimony to support or refute claims.”
“For virtually all companies, the amount of big data that needs to be analyzed in class actions is growing rapidly,” Kwon says. “We use modern and defensible data and text analytics to quickly distill information from disparate structured and unstructured data sources in order to narrow down to what’s really at issue.”
Data analytics, advanced modeling, and other statistical techniques are essential tools in helping companies understand the data, identify relevant facts and evaluate claims in a class action. The potential stakes in the face of a class action are far too high for companies to overlook or underutilize these valuable tools.
Interested in more insights on key issues from our team? Read the first article in our series: 3 Things You Should Know to Effectively Use Financial Forensics. Additional articles in our series, “From Books and Records to Bits and Bytes,” are coming soon, including: