Friday 05 July 2019

Which difference matters in insurance?

(Private) insurers used different forms of information to assign insurance premiums to risks that need to be insured. Just think of the questionnaires that we all have to fill out when taking out car, fire or life insurance. These data are then linked to population data relating to the risk to be insured. Insurers say they do this to ensure that the ratio between the insurance premiums and the expected cost is fair. Without this link, offering insurance would not be profitable due to the risk of ‘adverse selection’, which would create an imbalance between the cost of the insured risk and the contribution the insured pay. Moreover, it would not be fair that low risk groups bear the onus for high risk groups. In short, differences in terms of risk must be expressed in fair differences in the premiums that people pay to insurers. In an ideal world, this would create a sense of solidarity within risk groups, between the ‘unfortunate’ and the ‘fortunate’.

Despite this reasoning, insurers will not take all the known differences within risk groups into account. For example, they may not use genetic risks or gender-based risk differences to determine insurance premiums ‘because we do not control’ our gender or our genes. Besides this, insurers must weigh the cost of collecting data against the benefits of creating smaller risk groups. In addition to the solidarity within groups, the anti-discrimination measures also promote solidarity between the groups: the low-risk groups subsidise the high-risk groups. This shows that not every difference (between the risk groups that need to be distinguished) makes a difference (in terms of insurance premiums). Some forms of solidarity transcend a difference, despite the fact that they explicitly recognise this difference.

Our smartphones, for example, contain a growing number of increasingly accurate sensors that can monitor and visualise our lifestyle or health behaviour. The datafication of our daily life has produced more differences: differences between people who walk more or less than 7,901 steps a day, differences in the number of connections on social media, differences between people who always pay their invoices on time and those that are unable to or don’t, differences between people who never forget to switch off the heating and those who forget now and then. Techno-utopian thinkers claim that health insurance and insurance as a whole will become cheaper, more efficient and more qualitative as a result of this new technology. Critics, meanwhile, point out the threats to privacy and the freedom associated with the penetration of all these sensors and the resulting daily and increased surveillance. The claims of these utopian thinkers and the critics are often founded on presumptuous assumptions, presenting these developments as a fait accompli. However, this technological functionality must still be transposed into practice, into specific insurance products, something that will not be done in a vacuum. The main things to determine is which differences we want to apply as a difference in insurance practices.

Crucial questions include: which known difference must give rise to a different treatment? Will we, as a society, use every known difference to make a difference, thus making insurance less affordable for some (risk) groups? Will we be capable of ignoring certain known differences as differences that do not matter? If yes, when is it fair to take a difference into account and when should we invoke solidarity? What will constitute the basis for solidarity in future data-driven societies/insurance markets? Should a difference as a result of bearing a risk carry more weight than a difference as a result of taking risks – and how easy is it to distinguish between the two? We need a societal debate to determine which principles we must apply when transposing this technology into practice to avoid a potential, undesirable fait accompli in the future. How we make a difference makes a difference!

Gert Meyers (CeSo, Catholic University of Leuven) is a postdoctoral researcher in the Life Sciences & Society Lab, studying how the implementation of new technologies will redefine the role of solidarity in insurance. His doctoral dissertation, titled ‘Behaviour-based personalisation in health insurance: a sociology of a not-yet market’ (2018), was awarded the Geneva Association Research Grant, which, in 2018, focused on Big Data in insurance as well as the biennial 2018 Derine Prize for the contribution that his doctoral research made to the debate on justice in society.