Home | Blog

Friday, November 4th, 2011

Is modern insurance under threat?
posted by Ross Dawkins

I recently participated in a discussion on the impact of the ECJ decision to prohibit the use of gender as a pricing factor.  (My slides are here).  This is a topic that I have discussed here before so I will limit myself to my newer thoughts on this, specifically whether there are implications for other pricing factors currently in use — age, disability and such like.

An important caveat on the below is that it is based upon my reading of Attorney General Kokott’s opinion.  The final judgement of the ECJ is notably more circumspect with respect to detail on the reasoning underlying their decision — I am assuming that this does not diverge significantly from that presented by Attorney General Kokott, but at present this remains uncertain.

Attorney General Kokott disliked the idea that there is information about insurance risk in someone’s gender.  This is because gender is a characteristic of the customer that is inherent and immutable but, in Kokott’s view, lacks a direct causation to differences in risk.   Even with longevity her argument would appear to be that this is due to lifestyle choices rather than someone’s sex per se.  Statistical inferences are treated somewhat dismissively.  This is of course rather worrisome in the context of an industry which is in large part built upon statistical inference.

Let’s now consider three cases: age, disability and credit rating.  All are used as rating factors by at least some UK insurers in particular products.  Starting with your credit rating, the use of this to price insurance does not seem to be affected by the ruling.  It’s correlative rather than causative in its effects — but it is not an intrinsic personal feature although restoring it to good health is no doubt harder than impairing it.

Disability is clearly hard to change, subject to medical advances.  However, it’s use in the UK is in a precise way (i.e. it’s not used as a blanket factor to affect pricing) — so that it has causative properties.  However, some judgement may be used beyond statistical evidence: to this extent there seems to be an element of exposure to challenge as a direct consequence of the ECJ decision.

Age is perhaps the most commonly used variable of all.  Mostly, there is causation too: a young driver is also an inexperienced driver so pricing effects in motor insurance do not appear unfair.  Similarly, death is likely to be nearer as you age, so that life insurance pricing should still — surely — reflect that (let’s face, this product would disappear otherwise).  Kokott’s reasoning can be applied with unsettling consequences: if one looks at health insurance then, yes, more ill-health is correlated with getting older — but how you choose (or chose) to live your life also impacts.  However, the Attorney General herself draws the helpful distinction that your age changes such that over a lifetime it should all balance out. 

Overall then it seems that gender should not be the thin end of the wedge that undermines the basis of all insurance delivery.  That does not make the ECJ’s decision a good one (in my economic view) but is does mean that the further fallout should be limited.

 


Blog Archive

Blogroll

Latest Publications