An agreement on the provision of flood insurance has so far eluded insurers and the UK Government. The debate has touched on some ethical points that insurers should take note of and I’ll be exploring them in this and the next few posts.
I’m going to begin with a key question that will affect many policyholders: how affordable will cover for flood insurance be after the Statement of Principles ends in 2013? The insurance sector has been sending out reassuring signals that insurance for homes in areas of significant flood risk will continue to be affordable. A couple of months ago, Nick Starling, Director of General Insurances at the ABI, stressed affordability during an interview on BBC Radio 4’s Today programme. Yet how can insurers deliver on affordability while at the same time seeking to apply a ‘true market rate’ to flood risk properties?
The ethical issues raised by affordability lie in a tension between risk pooling and risk pricing. Insurance works by policyholders pooling their risks and sharing the consequences amongst pool members. At the same time, the contribution each policyholder makes to the pool should reflect the risk they bring to the pool.
Clearly, the more an underwriter knows about a particular risk, the more precisely she can determine the appropriate premium for its policyholder to contribute. However, if you extrapolate this ‘more information equals more accurate premium’ argument, it begins to undermine the risk pooling role also at the heart of insurance. If you’re able to determine the flood risk on a property by property basis, then insurance will resemble little more than a savings plan orientated around when the flood will happen.
Talk of a ‘true market rate’ puts more emphasis on risk pricing, while talk of affordability puts more emphasis on risk pooling. Those who emphasise risk pooling would view a property by property flood rate as unfair on those who find themselves in an area of higher flood risk through no fault or design of their own. Those who emphasise risk pricing would view a more regional rate for flood as unfair on those policyholders who face little to no flood risk within that region. Universal rates may have been necessary when more detailed information (and the means to apply it on a case by case basis) was not yet available, but those days are long gone now.
For insurance to be insurance, risks should be pooled, which means that those policyholders at below average risk of a certain peril will cross subsidise those policyholders at above average risk of that same peril. So cross subsidising per se is not unfair, only its more extended forms: an example would be when someone with driving convictions doesn’t pay that much more than someone conviction free.
At the same time, there’s no point in running a risk pool if the circumstances of the risks being pooled can’t be taken into account. Even the coffee house merchant adventurers recognised that.
How then do you balance these two apparently opposing perspectives? Both are in some way both fair and unfair. One way is through more of a ‘whole policy’ lense, for flood is just one peril covered by a household policy. Consider two policyholders. The one at the top of the hill subsidises the one at the bottom of the hill for flood, but is subsidised in return for storm. The one at the bottom of the hill subsidises the one at the top of the hill for storm, but is subsidised in return for flood.
The current premiums of each of those two policyholders reflect their positions on the risk spectrums of flood and storm, but only to a degree. Neither is paying the true risk premium for either risk. It seems rather unfair if, as seems to be the present case, underwriters are able to drill down their risk pricing to a property by property level for a risk like flood, but not for a risk like storm.
This is not to say that underwriters should only drill down to the level of their weakest set of risk data, but rather to say that they shouldn’t give undue emphasis to one aspect of the risk they’re covering and ignore the contribution of other aspects. A household policy involves several streams of cross subsidisation.
So what this adds up to is this:
I’ll explore these points in a more practical way in my next post.