Despite the unpleasant nature of the subject, there was yet another packed room of student actuaries eager to hear Paul Murray talk about the very topical issue of ‘abuse claims’.
Paul has worked as an actuary in the London Market for 22 years and is currently a director in the Risk Consulting and Software practice at a consultancy firm. In recent years he has been involved in helping to reach commutation agreements between insurers and reinsurers relating to claims arising from abuse suffered.
What claims are covered and how big is the problem?
So what type of claims are we actually referring to when we talk about abuse claims in the context of insurance? These are any claims which relate to Physical and Sexual Abuse (PSA) where the abuser is someone other than family member or friend and an insured organisation (e.g. church, charity, hospital etc.) could potentially be held liable for failing to protect the individual who has been abused.
Studies indicate that around 10-15% of children are abused, and around 15% of these cases are potentially “insured” abuse. However, not all of these will result in a claim, as understandably, some victims will want to put the abuse behind them and move on with their lives. In addition proving liability can be difficult given the substantial time delays between the event and the claim.
Two examples that have been in the media recently are the Jimmy Savile (operation Yewtree) and Magdalene laundries cases. The alleged abuse carried out by Jimmy Savile spanned for over half a century, and to date 450 different people have come forward and 214 criminal offences have been recorded. Between November 2012 and February 2013, the Savile estate, BBC & various hospitals have been sued on behalf of 67 victims.
The Magdalene laundries were a series of asylum-like laundries run by religious orders in Ireland from 1765 to 1996. There were approximately 10,000 women incarcerated between 1922 and 1996. Direct involvement from the Irish state has now been proved and admitted.
The media coverage that abuse claims have generated recently, including these two prominent examples, has helped to increase public awareness and vigilance and together with the introduction of the Sex Offenders Register should help lower the future incidence of abuses.
What are the legal issues?
In the UK, there are a number of routes that a claimant can follow in order to claim for abuse. The two most common are assault or negligence.
Different limitation periods apply to each. For intentional assault there is a 6 year limitation and up until January 2008 this was strictly adhered to. The absurdity of the strict limitation period was highlighted on numerous occasions, in particular in 1995 when a woman who had suffered abuse as a child at the hands of her father was able to sue her mother for knowingly allowing the abuse but not her father as it was beyond the limitation period!
In 2008 the Law Lords decided that the 6 year limitation was sensible in order to allow people to continue in business, without the threat of claims from distant transactions and to give defendants a chance to defend themselves. However, they did decide that the courts can use discretion in extreme circumstances. The high court must now rule, on each individual case, whether a claim after 6 years is appropriate and if a fair trial is still possible. Despite this, in most cases the claim needs to be made by the time the claimant reaches 24, but by 24 most people are unlikely to have realised the extent of the damage caused!
For negligence there is a 3 year limitation period but this occurs from ‘date of knowledge’. This may be the date the claimant realised the damage caused or the date that they realised there were no controls in place. In the case of negligence, the courts can use their discretion.
Another option is to claim for vicarious liability but this is only possible if the abuser is employed to look after children (e.g. a gardener of a children’s home could not be sued through vicarious liability).
What are the insurance issues?
One major issue for insurers is the aggregation of abuse claims for reinsurance which ideally could be aggregated so the insurer will only suffer one deductible (reinsurers on the other hand would ideally not want claims to be aggregated). For this to be the case claims would need to be from the same ‘event’ but the abuse of different victims is likely to be at different times if not different places.
The abuse is also likely to have taken place over a number of years, so which policy pays? This is calculated as per the Accidental Circle Occupational Disease clause (b) which has appeared on EL contracts since the mid-1980s. It states that when a claim is spread across years, the limit and excess are spread in the same proportion.
So how do we estimate future claims?
In order to estimate claims we should first assess our exposure: look in the policy database, are there any churches/charities/children homes etc.? Have there been any claims so far? This is more difficult for reinsurers who will not necessarily know the full extent of their exposure.
When it comes to modelling these claims a ground-up stochastic model is the best option given the low frequency/high severity nature of them claims, assuming sufficient data is available. An insurer will be interested in their net retention, after recoveries, whereas a reinsurer will be concerned with which claims hit the layers they are on.
The significant assumptions needed in the model are:
- Future notifications by year – 2 perfectly plausible scenarios are that 1) claims have peaked and 2) claims will peak in the future, sometime around 30 years after the key exposures,
- Claims distribution - claims can be very large, taking into account loss of future earnings and other damages,
- Period of abuse - usually the abuse takes place over a number of years,
- Notification delay - can be very long, arguably up to 30 years,
- Payment delay - can be anything from 2 years up to 7 years,
- Inflation - paticular care is needed with setting the inflation assumption. For instance, consider that an inflation assumption of 5%, 10% or potentially 15% could all be justifiable, but with 30 years of inflation these would equate to factors of 3, 7 or 19!
In addition, other assumptions such as policy limits, stability clauses and costs in addition should also be incorporated into the model.
Notable awards to date
The largest award in the UK to date related to abuse carried out by the Archbishop of Birmingham, who abused the victim between the ages of 7 and 18. The victim was 35 when the case went to trial and had never been employed and was deemed unlikely to ever be employed full-time. They were awarded £621,444, the majority of which was for loss of future earnings.
In 2011 the NHS paid out £8m to 165 residents of Buldock hospital in Cornwall.
In 2011, Essex County Council paid £1m in damages to 4 siblings for failure to protect them from their father who had a previous child abuse conviction. This case is unusual in that claims are very rarely made for abuse carried out by a family member. However, in this case the Council could be held liable in failing to protect the children as repeated warnings were ignored.
The total paid out in the US, with a population of 314million, is now well over $3billion. In the Republic of Ireland, which has a substantially smaller population of around 4.5million, over $1bn has already been paid out to date! Further evidence, if needed, that the scale and uncertainty surrounding these claims must be taken seriously by any insurers and reinsurers with the exposure on their books.