The impact of tech on insurance discrimination

Three types of fish

British TV viewers will be familiar with the tongue-in-cheek adverts for “Sheila’s Wheels” which offered preferential pricing for car insurance for women. At the end of 2012, the European Court of Justice banned the price advantage based on sex.

In response, a new insurance company launched called ‘drive like a girl’. They provide insurance for any gender. In case of doubt, their website states that ‘drive like a girl’ refers to a safer, smoother driving style. To qualify drivers have to install black box telematics to monitor their driving.

Immediately after the 2012 rule change, the UK average gender price divide for car insurance was £27 according to comparison site Two years later men were paying £51 more, and in another two years, the gap increased to £101.

What happened?

Some male-dominated occupations have a poor claims experience and men tend to drive larger and more costly vehicles. So by using more granular data, the price gap widened even further.

The price disparity could be considered as indirect discrimination, which is covered by legislation in some countries, including the UK. Angela Simpson, employment partner at DWF Law, explained to LedgerInsights that an insurer would not face a legal liability under the UK’s Equality Act. That’s because there’s a sound business rationale, predominantly in the way of financial justification for insurers to charge more for high-risk categories like larger or faster cars.

The impact of tech

Last Month Daniel Schreiber of Lemonade wrote a blog post outlining why newer tech companies like Lemonade can offer better pricing. The reason is they have exponentially more data.

Increasingly the combination of AI, blockchain, and IoT will allow more granular distinctions. That allows for better risk assessment and pricing. But more differences means more significant price gaps between the higher and lower risk customer. The concern is that those in most need of insurance, won’t be able to afford it.

Jeremy Irving, insurance regulatory partner at DWF Law, commented about the UK’s Financial Conduct Authority’s position: “It comes back to the concept of the potentially vulnerable customer and the risk of social exclusion from financial services and in effect creating an insurance underclass.”


Obamacare introduced the ‘individual mandate’, which is about to be repealed. The mandate means there’s a requirement for most Americans to take out basic health insurance. If you don’t, there’s a tax penalty.

The purpose of the mandate is to create a larger pool of people to carry the cost burden. Otherwise healthy people might not bother with insurance. This phenomenon where people who need insurance buy, and the low risk don’t, is known as adverse selection. If the healthy don’t insure, the average cost of healthcare will be higher.

Why can’t you charge everyone the same?

What if insurers were forced to charge everyone the same, say for health insurance? Then the healthy would view the price as too expensive, and the unhealthy would see it as cheap. That would attract more unhealthy people and fewer healthy people. So the insured pool of people would be balanced more towards the sick. This would force the average price even higher, towards the real rate for covering unhealthy people. That would force out even more of the healthy. Plus in most insurance scenarios there are more low-risk people in the population compared to high-risk.

If you consider the endgame, the result is few healthy people have insurance, so less of the population has coverage. And the insurance cost for the unhealthy becomes even higher than it would have been if price discrimination were allowed. It’s a lose-lose for the healthy and the sick.

The conclusion is price discrimination in insurance is a good thing.

To recap. Social public policy arguments lean towards forcing everyone to have coverage. Insurers argue that differential pricing is the way to get the market to deliver the right price.

The math doesn’t lie

UK actuary, Guy Thomas, has done the math. He believes that neither of these is right.

He acknowledges that in some cases price differentiation is essential. A good example is age and life insurance. It makes no sense to give the same pricing to a 30-year-old and an 85-year-old taking out new life insurance.

However, in many cases, when there’s some adverse selection, or some of the low-risk are pushed out by higher prices, the outcome is better for the group as a whole.

What is meant by ‘better’? Thomas’ view is different to the status quo because of a simple switch in perspective: to focus on LOSS coverage rather than just coverage. For example, the highest risk might have four times the losses of the lowest risk.

He told Ledger Insights: “If one insurer (say Lemonade) has more granular information than other insurers, that’s better for Lemonade and worse for the other insurers.”

“But all insurers having more granular information than in the past is not necessarily going to be better for society as a whole. In fact, after a certain point – a certain degree of granularity – it’s going to be clearly worse for society as a whole.”

“This is not an ethical argument or a matter of opinion. It’s a matter of simple arithmetic.” The International Actuarial Association awarded a prize for his paper on the topic.


The result of more granular data available to insurers is likely to be more of the neediest being refused coverage or priced out of the market. That will sometimes mean governments will step in. So the outcome will either be marginally higher taxes or limits on insurers ability to discriminate. Or the extension of legislated insurance such as Obamacare’s individual mandate.