Swiss Re’s recent paper ‘Life insurance risk selection: required differentiation or unfair discrimination’ is a worthwhile contribution to an important debate. What intrigued me about the paper however was not so much the arguments and examples it put forward, but the language it adopted in presenting them.
The choice of language matters, for it signals the character of the discussion that the writer wants to have around the questions being raised. What this meta-discussion does is generate the background assumptions and constructs in ways that influence conclusions and actions.
Why does this matter from an ethical perspective? Well, language can be used to form metaphors and metaphors come with all sorts of assumptions attached. When people use metaphors, they embed those assumptions into the discussion. In short, the choice of metaphors helps set the moral dimension to the issues being discussed.
So when the Swiss Re paper uses words like ‘desire’ and ‘commendable’ when talking about society’s interest in tackling unfair discrimination, the impression left is that this is well-meaning, but not necessarily in everyone’s best interests. And when the paper uses words like ‘impose’ and ‘curtail’ to describe regulatory interest in tackling aspects of unfair discrimination, the impression left is that of a sector being unfairly constrained.
Upon such foundations can then be built arguments about change undermining the sector’s ability to deliver long-standing benefits to society. The impression left is of insurance as a well-oiled machine and change as having serious consequences. Yet this is a sector replete with conferences at which insurers talk about the status quo
being dangerous and change being urgent and necessary. And the paper goes on to talk about mutuality as a concept being forced upon the sector and one that is doomed to failure. Yet the sharing of risk has been fundamental to the development of both insurance and reinsurance.
Can insurers expect a regulatory status quo while at the same time developing new forms of underwriting and distribution? It feels like a big call to put to society, especially when trust in the sector is so low. Behind much of the debate on differentiation and discrimination lies big data, being used by insurers to broaden and deepen what the paper refers to as the ‘risk puzzle’. Yet this is less like an engaging, intellectual game and more like the use of information about how each of us lives, works, travels and feels.
The idea of a ‘risk puzzle’ illustrates an interesting characteristic of the paper. Puzzle is a metaphor for the intellectual unravelling of an unknown. And what runs through the paper is the image of the disinterested application of the scientific method to the understanding of risk. This builds towards what seems to be the central but unwritten point behind the paper, that of ‘trust us; we’re the ones who understand risk’. Yet insurers need to recognise that their investigation of risk is much less disinterested than they think.
Should insurers even be expected to investigate risk in an disinterested basis, or write papers in a neutral manner? Of course not: that would be asking too much of them. Take this article – I can’t say that it’s disinterested or neutral, although I have my own opinion about degree. Yet for the sector to truly engage with the public on issues like fairness and discrimination, they need to be better at engaging in debate in a more disinterested manner (and perhaps at times from behind a ‘veil of ignorance’) in order to fully understand the implications their decisions can have.
How might insurers take steps in that direction? Well, they need to look more carefully at how they’re constructing their argumentation around fairness and discrimination. What matters in constructing a persuasive argument is not what premises you believe are best, but which premises your audience is most likely to accept. After all, the whole purpose of insurers engaging in debates about fairness and discrimination is to persuade the public to leave behind premises they’re already familiar with, and move towards conclusions they haven’t yet accepted. The public is more likely to listen to insurers’ arguments when they feel their arguments are being listened to too. Only then will attitudes around trust start to change.