Technological Advances in Life Insurance Underwriting Processes

Background: Emergent technologies such as big data and predictive analytics present great potential for life insurance underwriting. As technology continues to develop, including wearable technology and non-traditional data sources, the impact is largely unknown with limited exploration to date in this traditional, heavily regulated, insurance activity. The lack of understanding is further compounded by the infancy of the technology, but the potential radical impact of the technology means this is an area which demands examination.
Aim and Objectives: The purpose of this study was therefore to assess how Big Data and AI will impact life insurance underwriting. The objectives of the study were:
· To establish how Big Data and technology can help to overcome the problems that exist with the traditional life insurance underwriting process;
· To explore the role and value of non-traditional forms of data; and
· To identify the ethical and privacy implications of the modernisation of life insurance underwriting. Value of the Study The study is valuable for a number of reasons with both practical and academic implications. Drucker (2002) highlighted that most innovations, especially successful ones, result from a conscious and purposeful search for opportunities. This is typical of the US life insurance companies who have been exploring how to advance the traditional underwriting process by taking advantage of technological developments and opportunities that currently exist. This presents challenges around the use of personal data in the life insurance evaluation process including privacy and ethical considerations. Whilst this research has a focus on the US life insurance industry, the issues presented resonate across the globe with most US based insurers operating on a global business platform. Since technological innovations will impact the life insurance underwriting industry, we need to know more about how these are likely to materialise both for the consumer and industry – to ensure the most accurate industry stakeholder decisions/choices whilst ensuring the value for the consumers and protection of their rights. Technology and Underwriting in the Insurance Industry Underwriting is defined as ‘the insurance function that is responsible for assessing and classifying the degree of risk a proposed insured or a group represents and making a decision concerning coverage of that risk,” (Bhalla, 2012: pp. 160). Typically, insurance companies have relied on age and policy history as a means of assessing risk, but technology has now advanced this. The existence of vast historical data, or Big Data, and new techniques such as machine learning and predictive modelling provides an unprecedented opportunity for the industry to transform the underwriting process (Maier, Carlotto, Sanchez, Balogun, and Merrit, 2019). More factors can now be considered to support the insurer’s decision(s) with some concluding that this will, in fact, make the risk assessment, or underwriting process, more personalised and accurate, (Bhalla, 2012). In life insurance, Big Data can be defined in terms of the ‘five Vs’ (Andrews, 2018) namely Volume, Velocity,Variety, Veracity and Value. For example, Mass Mutual used Big Data from a fifteen-year period to increase assessment accuracy with a 6% reduction in claims in the healthiest pool of applicants, a reduction in time to issue by >25% and an increase in customer acceptance by >30%. This translated into cost savings. The opportunities for technology in underwriting are therefore evident. A key technological development, identified by Kerbeshian and Matson (2018), is accelerated underwriting which is the use of “data and predictive analytics …. to allow the company to offer a price that is competitive with traditional underwritten policies and often more quickly”, (pp. 9). Many new data sources can, and are, being incorporated into the underwriting process but the challenge is determining what the incremental value of each is. The possibility that life insurance companies use genetic data in the underwriting process has given rise to concerns that genetic discrimination could emerge. A growing number of countries have introduced measures to prevent this through voluntary arrangements and legislation such as the UK (Burton, Joly, Knoppers, Feze, Dent, Pashayan, Chowdhury, Foulkes, Hall, Hamet, Kirwan, Macdonald, Simard and Van Hoyweghen, 2014). While insurers are not allowed to discriminate the consumer can widely avail of genetic testing kits from companies. Insurers could now face a problem known as ‘adverse selection’, where policyholders use genetic testing to ascertain their need for insurance and healthy people could dismiss life insurance (Gapper, 2018). Individuals who discover that they have a greater risk of disability or death will disproportionately seek to purchase life insurance which can lead consumers to make choices that may be counter to their rational position, (Schmidt, 2018). The literature generally indicates that life insurance underwriting is entering an exciting, yet challenging, period of change where technological innovation is impacting the cumbersome traditional process. This paper will attempt to explore this in the context of real-world insurance underwriting by exploring the why’s and how’s around diffusion. Methodology This research was conducted using semi-structured interviews with expert witnesses. A sample of management level employees were selected from a variety of business activities in the industry including data science, actuary, machine learning, data analytics and legal. This type of interview allows for a great deal of flexibility and wealth of information, (Collis and Hussey, 2014). The preparation of 12 predetermined questions acted as a guide. However, the researcher can probe answers for further explanation and to build on their responses around innovation and underwriting. Interviews often result in rich detailed data (Saunders, Thornhill and Lewis, 2016). For the purposes of this study, to eliminate the issue of geography and accommodate international time differences and executive schedules, Skype interviews were conducted. In total 8 expert interviews took place in March 2019. The findings were analysed using Thematic Content Analysis. Key Findings A number of key outcomes were identified through this study. It was confirmed that both accelerated underwriting and the use of wearable technology have the potential to bring significant performance improvements to life insurance operations. The study concluded that the key drivers of this diffusion include reducing costs and increasing profits. Interestingly facial analytics is suggested as simply a way to engage customers with no substantial performance improvements to be gained from this to encourage adoption and thus it is no more than an incremental innovation. It was not possible to achieve consensus among the experts on the use of non-traditional data. It can bring performance improvements when combined with accelerated underwriting. However, the limitations particularly with social media use mean that as a standalone piece it may not bring sufficient cost reductions or profit impact and as such this is likely to delay the widespread adoption of this. For this reason, it is deemed to be an incremental innovation. Although academic literature in this area is relatively new and the technologies are in their infancy the findings of this research corroborate existing thinking, that life insurance underwriting is on the cusp of being changed beyond recognition. Importantly, the research concluded that the use of these technologies will greatly affect individual data privacy and have ethical ramifications for life insurance companies.