The Ethics of Insurance Underwriting
Underwriting has always adapted to new risks and wider changes in society. That’s important for the public, for it allows people to go about their lives and business with greater confidence and security. As a result of this, insurance has become embedded into many key life stages - buying a house and planning for retirement being but two examples.
Has that success lulled the sector into a false sense of security though? As digital innovations move underwriting into a new, and perhaps a more radical, period of change, are insurers being too un-critical about the implications that some of those changes? As one data protection conference discussed, is insurance heading for the superhighway, or the cliff?
For ten years, I’ve followed the implications that digital innovation could have for the ethics of insurance underwriting. A post-graduate degree in innovation and risk, plus many years in the market, give me a critical yet pragmatic perspective.
We are seeing the ethical implications of underwriting innovation at various stages of development. I’ve organised them into these four stages…
The Ethics of Insurance Pricing
It is in relation to pricing that the first significant challenge emerged around the ethics of insurance underwriting, at least in the UK. What we saw was civil society challenging the sector on the fairness of its key pricing model, drawing on its own evidence base to make its case.
What defined these events more than anything was how change was forced on both the regulator and the sector. In 2016, the regulator said it had few concerns about retail pricing practices, despite its own research producing some startling findings. Yet in 2018, it had to rapidly change tack in the face of a super-complaint, And in 2020, it announced that the main pricing model used in UK retail insurance was to be banned.
Never did the words of the economic sociologist José Ossandón ring more true. In 2015, he wrote that insurance pricing…
“…is not simply a matter of supply and demand, but rather the product of a wider range of actors, including regulators, lawyers, policymakers, members of parliament, consumer associations and representatives of the industry.”
There is every prospect that this ‘landscape of actors’ will continue to push the sector to justify its underwriting practices. And the next stage of such challenge is becoming clear.
Read more about The Ethics of Insurance Pricing
Fairness and Discrimination in Insurance Underwriting
The next stage of that challenge is taking shape around two ethical issues – fairness and discrimination. Several years ago, the then CEO of the UK regulator said “…that for leaders today – both in business and regulation – the dominant theme for 21st century financial services is fast turning out to be a complicated question of fairness.”
He was right – fairness is now a dominant theme for insurers, and in more ways than insurers are used to handling. This means insurers are having to rapidly evolve their existing view of fairness (based around fairness of merit) and incorporate other dimensions of fairness that also matter.
What this does is tie together key regulatory initiatives like vulnerability, access to insurance and fair outcomes for consumers. As a result, the overall fairness agenda becomes a lot more complicated, especially for firms who are at the same time pushing forward on ambitious digital strategies.
For those digital strategies to success, they need to be reconfigured around that new, more complicated view of fairness. If not, they are likely to hit a regulatory brick wall. This is because much of the complicatedness of fairness comes from how firms are using data and analytics.
Alongside fairness, but at the same time more reputationally damaging, is the question of discrimination. This is not a distant issue either. Questions have recently been raised in Parliamentary committee and in response, we are seeing regulators coordinate their powers and capabilities.
We know of course that the many good people in insurance do not want to see discrimination feature in the outcomes generated by their firm. And that is important, but not enough. I for one have been reliably informed that discrimination has crept into some decision making within parts of the sector. And clearly, if regulators are being told by Parliamentary committee to address this issue, then insurers of all shapes and sizes need to respond.
So what can insurers do? They need to share the steps their firm has been taking to eradicate discriminatory outcomes, and the evidence for how well this is being delivered. Many have been doing this on an individual consumer basis, but it looks like few have been tackling this at a group level. When I undertook a short survey on this in 2015, it was clear that few insurers were thinking of diversity and inclusion in relation to customers overall. Some civil society groups have since told me that the insurance sector are not fully grasping this, and need to do more.
Read more about Fairness and Discrimination in Insurance Underwriting
Date Ethics in Insurance Underwriting
It is clear from research published in 2020 and 2021 that consumers have concerns about how underwriters could use data and analytics to shape and price their policies. One report from the trade body for UK insurers saw the market’s response to such concerns as vital for overcoming what they called the ‘double lens of mistrust’ that the sector faces.
So the data ethics of insurance underwriting is a big issue. Clearly, for all the digital innovation in underwriting to succeed, consumers need to be convinced that it is being done in the interests of both the public and insurers. The sector needs to bring consumers with them on this.
And lots of insurers understand this. What matters then is how insurers respond to the ethical issues associated with data and algorithms. In doing so, they will face a series of ethical dilemmas, along the lines of ‘we can do this, but should we do it?’ Such dilemmas needs to be handled with knowledge, skill and some challenge too.
What becomes evident from such an exercise is just how often the underlying issues involve conflicts of interest. So, for example, I could do this with this data, and that is in my interests, but should I, which is a question often more in the consumer’s interests.
I often hear insurance people argue that for consumers in the round, such and such a use of algorithms would be in their interests. What they’re missing is the individual consumer perspective, where the argument often doesn’t stacks up. Remember that UK regulations stipulate that the perspective to adopt is that of the individual consumer.
Insurers often use a data ethics framework to structure their approach to data ethics. This is important, for it allows firms to bring the full scope of issues onto their risk radar and build in suitable checks and balances.
Is this perhaps a bit ‘over the top’? Well, I don’t think so, for while pricing has experienced a significant challenge through the super-complaint, what I am seeing points to more serious, and more fundamental, challenges ahead. Insurers need to organise their preparations for this.
Read more about Data Ethics in Insurance Underwriting
The Ethics of Personalisation and Behavioural Fairness
So what could bring about those more serious, more fundamental, challenges? How they will present themselves is difficult to judge at the moment, but they will undoubtedly have their origins in two key themes influencing digital insurance. These are personalisation and behavioural fairness.
Both personalisation and behavioural fairness seem to make sense on one level, but also seem to be self-destructive on another level. They are raising questions about the future of insurance. As one panel at Europe’s conference for policymakers in data protection debated in 2017, ‘is big data driving insurance onto the superhighway, or off a cliff?’
What I am seeing is a growing number of influential academics raising questions like this and looking at how governments and regulators are responding. At the moment, regulators are keeping quiet, doing little. The problem is that other ‘actors’ are asking questions, and in the absence of a regulatory response, could force them to respond, as happening in pricing.
Is it worthwhile then for insurers to pay attention to the ethical issues associated with personalisation and behavioural fairness? Some already are, for they see the implications as serious. And by serious, I mean seriously disruptive to their digital strategies.
These are insurers looking beyond the current digital shift, to what comes after, and more importantly, what matters then. What they’ve found is that trust is what consumers want beyond all these digital developments. And so the question these insurers are addressing is whether personalisation, behavioural fairness and all the ramifications that flow from them, will build trust or reduce trust.
What we have then is a future for insurance that is far from certain, with personalisation and behavioural fairness influencing much of that uncertainty. Insurers should prepare for the debate that is taking shape around the ethics of insurance underwriting, in order to understand and contribute to it.
Read more about The Ethics of Personalisation and Behavioural Fairness