Use of big data risks making some people uninsurable

Oliver Ralph at the Financial Times: “More sophisticated use of data could create an “underclass” of people who cannot afford insurance. According to a new report from the Chartered Institute of Insurance, consumers could miss out on some types of cover altogether if insurers deem them too risky.

Big data are one of the insurance industry’s great hopes for the future. Established insurers and a host of start-ups are investing millions in new systems to better understand the information they hold about customers, and to collect more data. They hope that by better analysing the risks that each policyholder faces, they can not only price their products more accurately but also advise customers on how to avoid problems.

However, the CII paper warns that using data in this way threatens the concept of pooling risk on which the industry was founded.

“Data is a double-edged sword,” said David Thomson, director of policy and public affairs at the CII. “The insurance sector needs to be careful about moving away from pooled risk into individual pricing. They need to think about the broader public interest.”

The report says that the concept of pooling risk “underpins the effectiveness of insurance cover”.

It adds: “Some people may be identified as such high risk to insurers that they are priced out of insurance altogether. Big data could, in effect, create groups of ‘uninsurable’ people. While in some cases this may be to do with modifiable behaviour, like driving style, it could easily be due to factors that people can’t control, such as where they live, age, genetic conditions or health problems.”

The issue of genetic data is a particularly contentious one.

In theory, genetic data could be useful to insurers when deciding how to price life or health insurance. Because of the ethical questions this poses, an agreement signed in 2000 between the government and the Association of British Insurers stops the industry from using predictive genetic test results. The agreement runs until 2019, although a review is due this year.

“You could price people out of the market for health products. There’s a danger insurers will not offer health cover to some people. The government would intervene if people are doing social sorting,” said Mr Thomson.

Better use of data in other areas has already forced the government to act. Improved mapping and data analysis have allowed insurers to more accurately assess which homes and businesses run a high risk of flooding. Many people complained that the resulting prices made cover unaffordable for people living in areas at risk….(More)”.