Data Ethics in Insurance

Data ethics is now recognised as important to the success of an insurer’s digital strategy. That’s because it:

  • builds trust amongst clients and consumers
  • reduces risk through awareness and responsiveness;
  • supports growth through more trust and less risk;
  • enhances governance for long term accountability.

That importance was emphasised by a Gartner report in September 2020, when it made data ethics one of five key marketing trends:

“Over the next five to 10 years, we’ll see demand for the ethical treatment of customer data intensify as consumer trust decreases.”

What this means is that if an insurer wants to compete in a digital market for insurance, then it needs to address data ethics, and be seen to do so. And one key audience for this will be the UK regulator, who is to be reporting on data ethics in financial services during 2021.

There are five key themes that insurers need to think about under the heading of ‘data ethics in insurance’. You can explore each of them in more detail below.

Trust in Digital Insurance

Research in 2020 for the Association of British Insurers found that consumers had a double layer of mistrust when it came to insurers and their data. And the implications of this for growth are recognised by insurance executives. Over a quarter are “extremely concerned” that lack of trust will impact their firm’s growth.

This means that insurers’ digital strategies must engage with data ethics. If they do not, those digital strategies will act as a brake rather than an accelerator to the success of their firm. So how well does your digital strategy build trust in your firm?

To measure this, you need to turn to the four components of trustworthiness. These are competency, reliability, honesty and goodwill. And when assessing how well your digital strategy displays these attributes, it’s vital to think of this from the consumer perspective, not the insurers. After all, it’s their trust you seeking, not your own.

And when weighing up what your digital strategy is doing to build trust, you should also look at where it could be building mistrust. If your digital strategy is built around getting closer and closer to the customer, that might just be your main data ethics problem.

Read more about Trust in Digital Insurance

Data Ethics Risks for Insurers

The sector’s embrace of data and algorithms has led to renewed interest in two of fundamental ethical risks faced by insurers: fairness and discrimination. And that interest comes not just from consumer groups, but from Parliamentarians as well. In other words, the political masters of the regulator. Together, those three audiences are capable of mounting campaigns of considerable impact.

And while every firm has their processes for managing enterprise risk, the challenge that insurers face is twofold.  Firstly, they have to understand data ethics risks from an outcomes perspective, for that is how the risks are shaped and realised. And secondly, they have to get used to the evolving and complex nature of data ethics. Data ethics in insurance is a journey, not a destination.

All of this means that insurers need to think carefully about how they go about assessing data ethics risks. Some are big, obvious and local (the ‘tree level’ risks), while others are broad with sweeping implications (the ‘forest level’ risks). Insurers need to challenge themselves on how they see and respond to data ethics risks, before others do so.

Read more about Data Ethics Risks for Insurers

Accountability for Data Ethics

A recent survey found that 28% of UK insurance executives thought that the greatest risk they expect to face will be personal regulatory sanction because of their firm's compliance failings. And without doubt, the chief source of such sanctions will come from what their firm has been doing with data and algorithms.

Let’s be honest though: accountability for data ethics is a challenge. Indeed, the FCA’s lead adviser on data ethics wrote in one of his books about seven distinct accountability challenges presented by data and algorithms. This means insurance executives have to organise their approach to data ethics very carefully. Some will do so through the depth and detail of their assessment, but the more successful ones will address the scope of their perspectives as well. As the recent UK pricing review so clearly illustrated, data and algorithms can move insurers into areas of risk without their apparent awareness.

Effective oversight of digital strategies must therefore be very tuned into the views and concerns of a variety of audiences. And how they tune in is important too, for engaging with such audiences is just as much about listening as it is about talking.

Supervisory Technologies in Insurance

Insurers have focussed consider attention on their digital strategies, yet they have often failed to notice just how much progress the UK regulator has been making with their own use of data and analytics. This was made very evident when in February 2019, Chris Woolard, interim CEO of the UK's Financial Conduct Authority, had this to say during hearings on insurance pricing before the Treasury Committee

“We have the resources and expertise to pick inside those insurance models.”

And what’re more, he spoke those words in relation to one of the key data ethics risks: discrimination.

The regulator’s digital capabilities have been growing since 2015 and their application to ethical issues becoming clearer. Adviser mis-selling of financial products is just one of the interesting developments.

So what do supervisory technologies (as they’re called) mean for insurers? In broad terms, it means the regulator is now able to monitor adherence to key requirements on a real time basis. The new rules on price walking is one obvious example. And in specific terms, they will be able to pick up on emerging problems and call in the relevant senior management function for an explanation. That’s what Chris Woolard meant when he talked about picking inside those insurance models.

The Personalisation of Insurance

Personalisation is the most fundamental change happening in insurance at the moment, and it’s being driven by data and analytics. Its impact will be seismic and the consequences profound. Some see it driving insurance onto a digital superhighway, while others think it will drive insurance off a cliff.

The ethical issue most often raised in relation to personalisation is fairness. Paying for your own claims and not those of others is how personalisation is often promoted to consumers. Yet this argument is premised upon only one dimension of fairness, that of merit. Other dimensions of fairness are however high on the regulator’s agenda, such as fairness of access and fairness of need.

At the moment, regulatory and political attention to personalisation seems to be ‘on hold’. What we are seeing though is growing attention in academic circles to some of the problems that the personalisation of insurance can create.

Should insurers just ‘carry on regardless’ then, as some evidently are? What I can say is that every insurer should be careful around what types of data they personalise premiums and claims upon. Personalising based upon certain types of data would be reputationally disastrous, while other types of data would just be ‘waved through’. Insurers need to carry out a careful ethical assessment on a case by case basis. 

Read more about The Personalisation of Insurance

I help insurance people to understand more about data ethics...

I provide insurers with insight into the implications of emerging trends around data and analytics in insurance. This is independent advice, free of parallel interests. The perspective is that of building trust in a digitised market.

I help insurers avoid the drift into digital group think, with pragmatic challenge on questions that can sometimes be missing from the project. This is about highlighting connections and consequences for people to think about.

Being probably the only person active in the UK insurance market with a postgraduate degree in innovation, my input to client projects is a blend of market knowledge and academic thoroughness. I have an ongoing engagement with academia. 

I’m part of the debate on the future of insurance, and clients draw on that foresight to help shape their own long term plans. This helps them position ‘trust’ as an active part of their digital planning, giving it a clear and structured voice.

Get in touch for an informal chat - my details are on the contact page