Data used to back Discrimination Claim against US Insurer

The lawsuit is led by Jacqueline Huskey, a black woman living in suburban Illinois, as well as hundreds of other as yet unnamed plaintiffs representing the insurer’s black customers in six midwestern states. The defendant is State Farm, the largest provider in the US of insurance to homeowners.

What makes the case unique is the way in which the lawsuit uses the findings of a nine month study into alleged disparities between the insurer’s treatment of claims from black policyholders, compared with their treatment of claims from white policyholders. This survey provided an evidence base around which the suit took shape and through which it hopes to achieve class status.

There have been allegations in the US about the treatment of claims by black policyholders for some years now. The problem was, as reported in 2020 in the New York Times…

“…that although plenty of Black homeowners had stories to tell about their treatment by insurance companies, the absence of publicly available data made it difficult to evaluate whether individual cases amounted to a pattern of discrimination — largely because insurers keep information about their claims secret.”

Similarity with UK Events

You can see the similarity here with the loyalty penalty and ethnicity penalty reports by Citizens Advice in the UK. Both evidenced their case using data collected and analysed independently of the insurance sector. As with the State Farm lawsuit, micro-outcome data was used to substantiate the allegations being levelled against insurers.

What this means is that while the way in which insurers assess and decide claims and counter fraud issues has until recently been pretty opaque, consumer groups are now willing and able to use data to push back on that opaqueness. This means that judgements about the ethical or legal validity of a claims or counter fraud practice are now no longer the sole preserve of the insurer.

One of the ethical skills I teach is called the Financial Times technique. Is the decision the insurance person is about to take one that they would be happy to read about on the front page of the FT? If not, then they should think again about making it. The message is ‘do not rely on people not knowing’.

Some Detail

The study at the heart of this law suit was carried out by the Center on Race, Inequality and the Law at the New York University School of Law, in partnership with law firm Fairmark Partners. It used YouGov to survey 800 State Farm homeowner policyholders about their claims experience. This is from the New York Times’ December 2022 report on the lawsuit…

“The researchers looked at measurements like the number of interactions that claimants had with State Farm representatives, the length of time it took for a payment to be made after a claim was filed and the amount of extra paperwork that State Farm asked for before agreeing to pay a claim. The findings, cited in the lawsuit, showed that Black homeowners had a significantly harder time by several measures. For most white customers, the process typically took fewer than three interactions before claims were approved. White customers were also one-third more likely to have their claims paid out in less than a month. The study found that Black customers were 20 percent more likely to have to talk to a State Farm representative on at least three separate occasions before having their claims approved. They were also much likelier to have to submit extra paperwork.”

Counter Fraud

The lawsuit didn’t just focus on the insurer. The New York Times 2022 article again…

“State Farm doesn’t handle all of the claims its customers make. Like many of its peers, the insurer relies on specialized technology companies to help process them. The lawsuit used the example of one such company, Duck Creek Technologies, to lay out how the system works. In the case of Duck Creek, once an insurance claim is made, the company uses software from the artificial intelligence firm FRISS, which is based in the Netherlands, to flag claims for potential fraud. According to the lawsuit, FRISS gives each insurance policyholder a “risk score” by running that customer’s information through its computer programs, which analyze the language in the claim narrative as well as the customer’s profile. Each score is based on elements like — as FRISS puts it — “demographic data about the neighborhood, such as the degree of urbanization,” crime statistics and data harvested from social media.”

Some of these data types are highly contentious – urbanisation for example. So if class action status is achieved, we can expect to have them debated vigorously during the court hearings, giving them a public airing like never before. As Duck Creek, FRISS and firms like them are widely used across the US, UK and many other markets, there will be few insurers who do not have to keep a close watch on how this case turns out.

Public Support

You may recall this article from November 2022 about a US survey by the Coalition Against Insurance Fraud, on the use of data to combat insurance fraud. They found that support amongst consumers was strong but also very conditional upon the proper handling of ethical issues. Should this lawsuit gain widespread attention, the questions it raises, and how they are addressed, are very likely to influence public support for the sector’s counter fraud activities.

Let Me Say It Again

I’ve taken the following bullet points from earlier articles I’ve written on the overall issue of discrimination in claims and counter fraud. If your firm is one of the insurers that needs to watch how the Illinois lawsuit progresses, these are some steps that I would recommend your firm taking…

• to revisit with a critical perspective their ethical risk assessments for claims and counter fraud;
• to review how their processes and controls handle those ethical risks;
• to challenge themselves on some of the assumptions built into those processes and controls;
• to particularly challenge themselves on how bias is being managed in both data and analytics;
• to draw the above together in preparation for increasing scrutiny of counter fraud, from inside their firm (e.g. boards) and outside their firm (regulators and consumer groups).

Relevance to the UK

Some of you may be tempted to dismiss the Illinois lawsuit as not relevant to the UK market. I would urge you, in the strongest possible terms, not to dismiss it as irrelevant. Based upon what I’ve been told in recent years, the risk in the UK is pretty similar to that in the US.

This won’t be an easy thing for UK insurers to take on board. There’s a strong and established culture in insurance counter fraud. However, the sector has adopted an opaque approach to tackling insurance fraud. Insurers need to prepare for this to be changed in some way, otherwise there’s a strong possibility that change will be forced upon them.