A quick step back….
Lump sum “serious illness” insurance is approaching a 30-year anniversary in the UK following its introduction in the late 1980’s. The product was based on a successful modest lump sum insurance policy originated in South Africa protecting the insured against cancers – typically ovarian and prostate cancer. The range of cover was significantly expanded to include other diseases, such as heart attacks, stroke, and all form of malignant cancer. The first mass product was marketed as Dread Disease – a term that (fortunately) was amended to “Critical Illness” (CI) insurance. In addition to the specified illnesses there was further cover provided against Total and Permanent Disability – abbreviated to TPD, TDB and PDB. The product sold, successfully, through tied distribution, with Abbey Life perhaps having most early success. However, it took time before this product became a staple recommendation within the independent financial adviser community due to the lack of standardised definitions that left brokers exposed to recommending the wrong product. The central body representing the insurance industry within the UK, the Association of British Insurers (ABI) reacted positively and worked with insurers and reinsurers to produce standard CI definitions. The scene was set for adviser firms to make informed comparisons and sales grew very strongly.
Naturally, one is more likely to claim under a CI policy than a life policy and hence the spotlight moved to claims outcomes. It is possible to fail to meet the definition of being “dead” in the context of an insurance such that the lump sum is not paid – for example, numerous firms would return premiums on suicide in the first 12 months from inception, or the cause or reason for death was excluded within the Terms & Conditions – acts of war, and terrorism were routinely excluded then. However, CI claims produced more disputes. In 2000 the Financial Ombudsman Service (FOS) was launched in the UK and given statutory powers in 2001. Its purpose was to resolve disputes by providing independent and free services to consumers. Providers were charged a few hundred pounds for each dispute FOS reviewed but the cost now is around £500 per complaint for the firm.
The broker’s nudge
Around the time, FOS were starting out a national discount telephone brokerage lobbied the industry to produce claims statistics. There was some reluctance initially, but gradually insurers started to express their claims outcomes showing the percentage of claims paid. All ABI members submitted total payments made under their various products and this information enabled the ABI to demonstrate the benefit of insurance within the UK, and in some way, satisfy regulatory bodies, including the Office of Fair Trading (OFT) that claims were being paid. All this was very positive, and informative – insurers and reinsurers could understand claim patterns, and refine underwriting and claims handling standards as more data emerged.
The publication of claims statistics and media reports also raised the profile of “non-disclosure” onto a (slightly) more public footing. The failure to answer questions accurately, or to fail to reveal material facts, emerged as a significant issue for insurers particularly as underwriting standards were relaxed to encourage new sales. The insurer’s position was clear – if non-disclosure was discovered and the terms of the contract would have been materially different had the underwriter known the fact, the policy was invalidated from inception. All premiums were returned, commissions recovered as the contract repealed as if it had never existed. There was little proportionality, nor any link established between what was not disclosed and the cause of the claim. The media were (and remain) keen to report such stories that invoke a sense of injustice, one example being a breast cancer sufferer whose claim was rejected due to her failure to disclose a long history of mental illness (that would have changed the terms). Whilst each insurer’s claims’ policies would vary, it only took a few similar stories to undermine public trust and reduce confidence from the adviser market at a time when rising acquisition costs reduced the number of tied, and direct sale forces.
The FOS dealt with significant numbers of claims disputes, and where the “agent” was deemed to have assisted in non-disclosure, insurers with tied, or direct, sales force, were (in many cases) obliged to pay the benefits. Conversely, if the “agent” was deemed to be independent the insurer was not obliged to make any payment and any redress to the consumer was a matter between the insured and their agent. Whilst the approach at the time was a clearly a legitimate application of contract law, it was not one that appeared to sit comfortably with many in FOS. The FOS moved the focus onto insurer’s processes or handling of claims and where they found fault would oblige insurers to pay claims. Such claims have always been in the minority, but in the absence of any active publication that shifted the focus to the many successful claims, it is understandable why a large intermediary would seek to create more transparency. Insurers began to report the percentage of claims that were paid and those that failed due to non-disclosure and where the definition had not been met. However, there was insufficient industry collaboration, cooperation and no agreed basis on how and what would be reported. Perhaps it was assumed this was obvious.
It’s a bit more complicated than that……
As insurers started to self-report, differences emerged, and the media who had actively encouraged the publication of claims outcomes began to press for all insurers to follow suit. As more insurers produced payment ratios, the focus shifted to those at the high and low end. No insurer wished to present a lower payment rate so the claims data would be more closely inspected – often not by claims officials, but by marketing folk keen to remove some “non-claims” from the pool to project a better image of their firm. Indeed, some insurers were irresponsible enough to produce tables showing how a higher percentage of their claims were successful than their competitors. Suffice to say, the whole topic of claims analysis is a bit more complicated than that.
Consider this simple comparison of death claims experience between two insurers, “A” (in business for 10 years) with “B” (positively youthful – only been trading for five years). Assume again that both insurers have identical products, pricing, distribution, underwriting and claims handling policies. Who will appear to have rejected more claims due to non-disclosure? Insurer B. Both A and B had 250 death claims from policies that commenced in the last five years and 10% had non-disclosure that resulted in the claim being rejected. Company A also had 750 claims from policies that commenced 5-10 years ago; none of these policies had non-disclosure discovered at the point of claim. Insurer A therefore received 1,000 claims and rejected 25, and posted an encouragingly positive headline, 97.5% of claims paid. However, Company B received 250 claims and rejected 25. They publish a 90% claims acceptance rate. As human beings, we have finite energy, and the less focussed effort we need to expend the better as we can do more things. Concentration depletes us. It is therefore easier to conclude that Insurer B is not as safe a bet as Insurer A. Insurer A pays more claims. This was a simple example using a simple “life insurance” policy – one event creates a potential payment. More variables come into play when you introduce “terminal illness”, or “suicide exclusion clauses” to the life insurance product. These two factors would further cause Insurer B to look worse that Insurer A.
But imagine how this becomes more complex when one considers Critical Illness policies? Different generations of policies that used to cover a few conditions, then many; some with TPD, (TDB, PTD) automatically included, some optional; some that cover children; some that have partial payments. Then, add into the mix disability insurance, Income Protection (IP) historically referred to as Permanent Health Insurance into the reporting cycle. A product that pays a regular benefit – usually. Whilst it has less optionality that CI, the claims experience and policies portfolios will differ significantly based on the deferred period – some policies have a “zero deferred period”, that will likely generate a small payment for a severe dose of “man flu” others require a six-month absence before disability income is paid. It is completely logical that most illnesses and injuries that give rise to payments with a zero- or very short deferred period will have a higher success rate. Insurers generally won’t ask how often colds occur, or if you have sprained your ankle, but they will ask about conditions that are likely to cause long term absence.
A further issue specific to IP is that the ABI require its member to submit annual IP claims details. The data includes the claims in payment at the start of the calendar year, new claims and claims that were declined. From this data, it produces a “claims success ratio” that takes “in force claims” within its base number. This is unlike lump sum claims where a claim is either paid or it isn’t. An adviser, and perhaps many within the insurance companies themselves, would expect IP claims’ acceptance rate to follow the same logic as lump sum; “new claims admitted” divided by “all new claims decisions”. However, this will produce a lower percentage than the ABI formula. It appears some insurers will publish the ABI success percentage whilst others perhaps use the “new claim” method.
Apples, pears and moving goalposts
However unpalatable, it is impossible to compare the claims’ performance of one insurer against another. Pricing, product design, underwriting and claims policy and distribution management are some of the most significant variables that impact claims experience. Insurers who adopt a more thorough assessment at the point of underwriting, or who carefully monitor their distributors, are likely to encounter less non-disclosure. Firms that have a more thorough claims investigation will likely detect more non-disclosure. Companies that offer the widest scope of cover are also likely to have more customers who submit claims – running the risk their “definition not met” failures will be statistically higher than the insurer that pays out on fewer eventualities.
In the simple example above, imagine that Insurer B improved its product coverage or offered cheaper premiums. The adviser may see this very positively until Insurer A presents the claims outcomes leaving the adviser in a predicament. This is the reality – unwittingly, the production and publication of claims outcomes could negatively influence recommendations for clients.
One significant change has unilaterally improved claims outcomes for customers following deliberation between the FOS, and the ABI with some, but not much, input from insurers and reinsurers. The consultation delivered a new framework for claims handling that came into place in 2008. Insurers were limited in how extensively they could investigate non-disclosure, where non-disclosure was discovered proportionality had to be considered, as did whether the non-disclosure was linked to the cause of a claim. This resulted in more claims being paid as claimants received a reduced amount of benefit where the non-disclosure would have led to increased premiums and it was not deemed “deliberate” – the burden of proving this sitting with the insurer. The removal of the right to ask for full medical records (unless there was evidence of non-disclosure) meant insurers needed to rebalance their new business underwriting, and claims underwriting, policies. This guidance subsequently became legislation within the UK Insurance Act 2015 that covers all forms of insurance. The legislation is a point of much debate as some believe the more consumerist approach disproportionately favours those who fail to answer questions accurately, compared to those who provide all the information and pay extra, or have exclusions because of being thorough and (dare I say) more honest.
A further development that impacted claims outcomes arose after the introduction of a “tele-claims” process. It was generally accepted that if a customer completed and returned a claim form, this was a “new” claim. As the tele-claims process increased in use, a claims handler could assess, or at least triage, a new claim before a “form” was fully submitted. If during this process, it was clear that the claim would not be successful, the claimant was informed and no formal “new claim” notified. This reduced the number of new claims that were reported as “not meeting the definition” and I estimate that companies fully using tele-claims process could remove between 5-15% of claims. Again, a company requiring customers to submit claim forms rather than operate a tele-claims process would appear to reject more claims.
Understanding how UK insurers might treat various claim situations has been the subject of research by Pacific Life Re both in 2014 and 2016 through benchmarking exercises. In 2016, 13 insurers participated. There were differences in reporting ranging from when a claim is recorded (for example, if a claim was submitted for an excluded condition – some included, some did not), how claims were reported when “deferred” (typically for claims required “permanence” or “terminal illness claims” and how lifestyle concerns was investigated from a non-disclosure perspective. These inconsistencies together with the impact of the size, longevity of the insured pool, and underlying customer and distributions demographics made comparisons challenging, to say the least.
Reporting claim success rates in the UK has had some benefits. Research indicates the public believe the success rates of claims are significantly lower than they are based on insurer’s data. Insurers and advisers can now counter this with formal claims information. In addition, insurers now provide useful information on the amount that is paid out, and for what reason. This data, combined with real case studies and customer testament is compelling and positively evidences the value insurance has in society. Furthermore, many insurers now use their claims outcomes to inform their propositions and risk management through careful and informed outcome analysis. With the CI market, unsuccessful claims can inform changes needed to existing definitions, and creates new conditions that can be covered – or partially covered. These are very positive changes.
What is less developed, or perhaps understood, is the impact of distribution on non-disclosure. Sampling has indicated material differences between firms and between models, “face to face”, “digital”, telesales etc. Insurer’s application forms are easier for customers to understand, and importantly copies of completed application forms are routinely sent back to customers. However, advisers have a significant influence on how thoroughly they present and assist their clients in the accurate completion of new proposal forms.
Perhaps from a customer’s viewpoint speed of claim settlement is one of the most important factors. Accepting there will be debates and differences around when the “stopwatch” starts, surely communicating to consumers, and advisers, how quickly they can expect to receive benefits is important? Perhaps the recent publication of Lemonade’s “three second AI Jim claim” will move the focus towards time metrics. This is most relevant in the IP space, where rents, mortgages, and utilities bills need paying.
What have we learnt after almost two decades of publishing claims acceptance rates?
It hasn’t changed the public perception, but ought to have strengthened the belief of those within the industry that the overwhelming majority of claims are paid. It has led to increased claims information being shared, encouraged more insurers to review their claims data and experiences (from a non-actuarial perspective) and increase the focus onto the most important aspect of an insurance policy. It has led many insurers to set tolerances that ensure debate at board, or senior leadership forums where tolerances levels are not met, or significantly out-performed. This is a welcome change.
Publishing the “percentage paid” metric has caused some insurers to irresponsibly promote their firms as better than others, others to manipulate data in the quest for a “higher” number leading to significant inconsistencies. It has become another “comparator” for research firms, comparison services and even in tenders for distribution arrangements. This, in my opinion, displays a level of ignorance that borders on negligence.
The assessment of medical claims is subjective, and with imperfect (sometimes contradictory) evidence that leads to individual judgements. There are many thousands of assessors making decisions every day on individual claims. A minority of decisions and assessments that will be poor, some completely wrong. There may be some insurers who culturally adopt a different level of tolerance, or who have poor governance and controls within their claims areas that will exacerbate claims problems. Insurers have a duty to ensure this vital function is undertaken appropriately, fairly and with an appropriate level of care. Insurers should measure (and audit) their claims’ functions against internal processes, regulatory/legislative compliance requirements, fraud controls, customer experience, as well as technical competence, timeliness and what evidence was requested for claim types. The governance around what is included and excluded within annual claims reporting should be understood and independently verified (ideally not by a marketing department to avoid any conflict of interest). It is commonplace within the UK for insurers to also be audited by reinsurers who carry the bulk of the financial risk who wish to ensure claims are being authorised as they would expect. Both require expert skills, knowledge and competence and both also include an element of self-interest. A truly independent audit would provide the Board with the most accurate and legitimate appraisal of how their firm executes its responsibilities and obligations to its customers.
There is no realistic, nor accurate, way that one insurer’s claims delivery can be compared against another, unless of course, the two companies are pretty much identical in every way. None are. Insurers need to understand what measures and outcomes are “right” for them and their customers, considering their own uniqueness. Measuring, monitoring and continually improving their own propositions and capabilities should be their focus. It may not help advisers to see insurers produce markedly different claims outcomes but isn’t that what we ought to expect in a competitive market with new entrants and new and different coverages for customers for delivered? Or does the production of different payments rates simply undermine confidence as we are too busy to think hard?
There are areas in which other regions could learn from the UK experience to avoid much of the noise around claims and claims outcomes. Such lessons should also help avoid advisers, or customers, making incorrect assumptions and consequently making poor decisions. Education is vital. At the time of writing, there are some UK commentators from the broker community who genuinely believe that they and the media have helped deliver improved outcomes to customers. Building trust from distributors, those within the insurers themselves and most importantly the public is crucial. Insurance, particularly against death and serious illness, plays a vital part in society and helps millions – not just those directly protected but their wider families and communities. I often think about the ideal world, but in an ideal world, no-one would need insurance as anything less than ideal simply wouldn’t happen.
Insurers changes lives for the better. At those vital times when our customers and their families or businesses are vulnerable, and in need of our help, we are there. This doesn’t make news – it should – but it is up to those in the industry to make it what people talk about. Insurers compete on price, product coverage, commissions, and service. That’s an old model. There are millions of uninsured lives and in our globally connected world, the insurers that will win will be those whose customers do their marketing, and share their experiences within their community.
If this potted history of the UK’s individual protection claims experience has piqued your interest, or my input could help your firm, distributors, customers or the market in general, please contact me, I am here to help those who are keen to do the right (often not the “easy”) thing.
Worked within the insurance industry since 1983 initially as a claims assessor working for Abbey Life, dealing with life and disability claims until moving into technical life and health underwriting reaching senior underwriter status. After a short sabbatical in Perth, Australia in 1990, Chris returned to underwriting, working as an underwriting manager for HSBC, before taking on the role of Chief Underwriter for Skandia Life in 1994 with responsibility for claims and operational teams with Skandia for 13 years. In 2007, he joined Bupa as Chief Underwriter and Head of Customer Services and after the merger of the protection businesses of Bupa, Friends Provident and AXA headed up the underwriting, claims and operations team for Friends Life until leaving in 2015 after Aviva’s acquisition to set up his own specialist consultancy. Chris has played a leading role in proposition development around the underwriting, claims and product coverage as well as representing insurers through ABI committees. An accomplished and seasoned presenter who lead the move to bring underwriting and claims topics out of “head office” and into regional adviser training seminars to educate and assist brokers in building their own knowledge and helping their clients.
Comment and discuss this article with other industry professionals in the ALUCA Linked-In group