The scope of what underwriters need to have on their radars is getting ever wider. The expanding horizon of rating factors is one driver of this, as is the increasing personalisation of the products being offered. This increases the amount of information underwriters have to process, which is a challenge. There’s just an awful lot of things to weigh up and accommodate. One way in which we would typically adjust to this is by a process of filtering. That however can open the door to an ethical problem called the self serving bias.
Research has shown that if two sets of people holding relatively opposing views are shown an ambiguous document, each of those two sets of people will interpret the document in a way that supports their particular viewpoint. Hardly a surprise you may think, but it has wide ranging implications for the decisions we take. For example, what looks like a well designed and clearly worded policy to the underwriter may be confusing and difficult to interpret for the policyholder.
What is happening here is a tendency to ‘see’ what is most supportive and advantageous to our own viewpoint, while playing down the importance of things that are in conflict with that viewpoint. We filter in what we think is significant and filter out what we think is not significant, with significance being of course in the eye of the beholder.
Now this is not something unique to underwriters, nor to insurance people. We all have a tendency to hold self serving biases, and how they form, and how we handle them, are part of our social development. On the one hand, they teach us how to present our case, but on the other hand, such teaching needs to be balanced with the ability to reflect upon the case put forward by others as well. This is one reason why I suggested that listening could perhaps be the most ethical activity of all, in this earlier post.
Two factors make self serving biases a challenge that underwriters should watch out for. Firstly, underwriting is a skilled and technical task, but this in turn tends to build more than a little self belief in the uniqueness of one’s knowledge and in one’s ability to ‘see things as they really are’. And secondly, the product that the underwriter designs and helps deliver is a promise, made tangible by the way in which its wording is framed and laid out. Both are fertile ground then for self serving biases to take root.
Consider this: one window into the impact of self serving bias in underwriting could be through the reports of the Financial Ombudsman Service (FOS). The FOS deals with cases that have already had a complaint to the insurer reviewed and turned down. So when a line of business has an uphold rate of more than say one third, it could point to some level of self serving bias not helping underwriters see the outputs of their work as everyone else sees it.
So what should insurers do? Three approaches come to mind.