Stewarts' Response to the Scrutiny Review of the Draft Damages (Jersey) Law 201-

About Stewarts

Stewarts is an international litigation firm specialising in complex high value disputes. Our practice areas include Personal Injury, Clinical Negligence, International Injury and Aviation.

Within these practice areas, the claims we undertake exclusively relate to injuries of the utmost severity or death. Our specialisms are recognised by top-tier rankings in the leading legal directories, Chambers, the Legal 500 and the Times.

Whilst there are many firms of solicitors which do some complex and high-value personal injury litigation, we are one of very few firms in the UK that exclusively specialise in such claims and do not conduct claims relating to non-disabling injuries (aside from the rare scenario when there are secondary claimants involved in the same incident as a claimant with severe injuries). We are well-placed to provide a response to this consultation, as almost all our cases involve issues surrounding periodical payments orders and the inevitable contrast with a lump sum award applying the discount rate. We acted for the Claimants in a number of the leading cases concerning periodical payments[1] and concerning negative discount rates in Bermuda[2].

We take the opportunity to point out that we are a relevant stakeholder for this review on the basis we have and continue to represent a number of Jersey based clients and this review, together with this response, are made in their interests. We work in conjunction with Jersey co-counsel David Benest at BCR Law in representing our Jersey clients.

Julian Chamberlayne, Partner and Head of International Injury acted for the plaintiff, in the case of X v Estate of Y (Deceased) & oths heard in the Royal Court of Jersey, Samedi Division (2013/143), a polish citizen injured in a road traffic accident in Jersey. We were able to obtain the first PPO in Jersey in this case.

In addition, we have two current catastrophic injury cases running in Jersey[3] in which our expertise is required to ensure those clients have the best chance of obtaining the correct level of compensation.

Accordingly, our submission has a particular focus on the impact of the issues raised by this review on those with seriously disabling injuries, with reference to those Jersey cases mentioned above.

We attach the following documents to assist in interpreting our comments below:

• The Discount Rate – a report for the Ministry of Justice prepared by Paul Cox, Richard Cropper, Ian Gunn and John Pollock. Dated 7 October 2015;

• PFP – MoJ call for evidence; and
• Stewarts Discount Rate Calculation example.

Wells v Wells approach and ILGS

It is accepted within the Draft Damages (Jersey) Law ("Draft Law") that Jersey's courts have generally adopted English common law'.

In our view, the English common law approach determined by the House of Lords in Wells v Wells remains the fairest and best way of meeting the full compensation principle.

In his leading speech in the House of Lords, Lord Lloyd of Berwick stated that the judges in the three jointly heard cases of Wells:

"were right to assume for the purpose of their calculations that the plaintiffs would invest their damages in I.L.G.S. for the following reasons:

1. Investment in I.L.G.S. is the most accurate way of calculating the present value of the loss which the Plaintiffs will actually suffer in real terms.
2. Although this will result in a heavier burden on these defendants, and, if the principle is applied across the board, on the insurance industry in general, I can see nothing unjustNo doubt insurance premiums will haveto increase in order to take account of the new lower rate of discount
3. The search for a prudent investment will always depend on the circumstances of the particular investor. Some are able to take a measure of risk, others are notStill less is it his duty to invest in equities and gilts in order to mitigate his loss.
5. How the plaintiff, or the majority of plaintiffs, in fact invest their money is irrelevant. The research carried out by the Law Commission does not suggest that the majority of plaintiffs in fact invest in equities and gilts but rather in a building society or a bank deposit"

The Wells approach was validated by the Judicial Committee of the Privy Council in Simon v Helmot with Lord Hope explaining at paragraph 47 that:

"finding a solution today is not so dependent on surmise and speculation as it was when the issue was being discussed in those cases three decades agoThe investment scene has been radically altered. The breakthrough came with the introduction of ILGS and the appreciation that there was at last a tool that could be used to provide protection against inflation. It is tailor-made for investors who want a safe investment for the long term. In practical terms that is risk free. As nearly as possible, it guarantees the availability of the money to meet costs as and when it is required. It may not be perfectBut it is the best tool that is available."

In the Draft Law, in line with the Ministry of Justice's (MoJ) 2017 consultation paper, a heavy emphasis is placed on investment return and investment risk. However, as Lord Hope's explanation suggests, the impact of inflation is equally as important. One of the reasons for adopting ILGS as a proxy return on which the discount rate should be derived was the difficulty courts otherwise faced in predicting future inflation. In the Court of Appeal case of Helmot, Sumption JA (as he then was), commented, stating at paragraph 38:

"The reason for calculating a lump sum award on the footing that the lump sum will be invested in index-linked gilts is that it achieves an automatic adjustment for future increases in the general level of prices."

The MoJ's 2017 consultation was right to acknowledge that it is essential for the discount rate to be set in a manner that fairly represents the 100% compensation principle. This echoes the MoJ's 2013 consultation on discount rate where it was noted at paragraph 6 of that paper that:

"if the rate is too high, the burden would fall on the victims of the wrongful act, who include some of the most vulnerable members of society, who would be under- compensated, thereby possibly increasing their reliance on the state and, therefore increasing costs to the taxpayer."

Underlying that observation and the points made by Lord Hope in Wells is the simple concept that the cost of meeting the long-term losses and needs of those seriously injured by the wrongful acts of another party may fall to a combination of the following:

1. The injured party;
2. The party who caused the injury or more commonly their insurer; and
3. The State/taxpayers.

It is clearly appropriate for the party who caused the wrong, creating the loss, to be responsible for the financial consequences. Where those losses continue into the future, the resultant financial risk as to how those losses should be met in the long-term should be borne by the wrongdoer. There is no sound argument in principle as to why either the loss or the risk of meeting the long-term needs should fall to the injured party or to the State/taxpayer, unless of course the State is the wrongdoer.

The adoption of a discount rate based on ILGS does not lead to the single Claimant being "over compensated". Rather, Claimants may choose to take investment risks so as to attempt to improve their financial position. To require them to do so, however, as is contemplated by this Draft Law, would make them the only class of investor required to take risk so as to advantage some other party (the wrongdoer/the wrongdoer's insurer). This cannot be considered either fair or just. In reality most seriously injured Claimants base their decisions on investing their damages around meeting their disability needs and very few seek to take investment risks to outperform their needs.

Proposed Statutory Discount rates of +0.5% and +1.8%

We strongly oppose the introduction of statutory discount rates, to be set at +0.5% and +1.8%, because such rates would not satisfy the full compensation principle and the basis for proposing those rates is inherently flawed. The key reasons why we are of this view are:

1. reliance on the model portfolios from GAD report 2017;
2. emphasis on the investment decisions assumed to be made by individual Claimants;
3. allowance for investment management charges and tax;

4. The proportion of claimants who will be under compensated utilising a mixed portfolio approach is inconsistent with a full compensation regime;and
5. The dual rate split: the Ontario Model and earnings related model.

1. Reliance on themodel portfolios from GAD report 2017

The Draft Law relies on the GAD report 2017 which followed the 2017 MoJ consultations; we refer to our 2017 consultation response to the MoJ which has been provided with this response.

The MoJ asked the Government Actuary Department ("GAD") to analyse two model portfolios based on unpublished anecdotal reports of claimant investor behaviour. Portfolio A was considered to reflect typical low-risk' claimant behaviour, whereas Portfolio B was the highest end of the low-risk' spectrum. They were not asked to and did not analyse the lowest end of the spectrum. Therefore, not only are the conclusions on which the MoJ proposes this legislation is passed derived from an unpublished and small evidential base, they also do not reflect the full range of claimant investor behaviours.

Additionally, the low-risk' categorisation of the portfolios modelled by the GAD is disputed on the grounds that both models carry a serious risk for the claimant's ability to achieve full compensation. According to the GAD model, Portfolio A had just 5% invested in ILGS (with 46% in equities or alternative investments) and this figure was only 3% in Portfolio B (with circa 65% in equities or alternative investments). In contrast, the majority of the 2015 MoJ expert panel (the full report has been provided alongside this response) described the low-risk portfolio approach as one in which at least 75% of the portfolio is allocated to ILGS.

In preparing this response we have sought the views of 2 of the authors of that expert report: Richard Cropper and Ian Gunn. They have confirmed that they stand by the view that no account was taken of the fact that any claimant investment behaviour was in light of the need to hedge mortality risks, the impact of having insufficient capital following the purchase of a property (due to the Roberts v. Johnstone approach to future accommodation giving rise to a capital shortfall), earnings (rather than price inflation), taxation and active investment charges.

Personal Financial Planning (member of the MoJ's expert panel) have repeatedly warned that if one is to take account of real-life experience, that must be in respect of both claimants' liabilities to real earnings growth, taxation and investment charges, as well as the return on investing in a mixed portfolio.

By only looking at what claimants invest in and not why they are forced to invest in those assets, the discount rate is bound to be set at too high a rate, as claimant will have to invest with exposure to even more risk to cover real earnings growth, taxation and investment charges.

2. Emphasis onthe investment decisions assumed to be made by individual Claimants

The House of Lords in Wells v Wells, as followed by subsequent courts who have considered the issue of whether ILGS is the best tool by which discount rate could be determined, have renounced the proposition that it was based on how Claimants choose to invest their compensation. The utilisation of ILGS as the benchmark is only ever a pragmatic proxy for the net real rate of return. It created a fair and predictable method by which future losses could be calculated that was not reliant upon expert evidence. It did not require speculation about the composition of any investment portfolio, the returns on the portfolio, the level of charges of investment managers and stockbrokers, the level of taxation nor a prediction of future inflation.

An individual Claimant's decision to take an investment risk is clearly their choice and a matter only for them. If their investments go well and they out-perform the discount rate, then their financial position is enhanced. However, that is as a result of the investment decisions/risks they choose to take after the conclusion of their case, coupled with the good fortune of those investment risks working out in their favour. If a Claimant who makes numerous risk-based investments were to achieve poor returns or actual losses, it would result in a performance below the discount rate and ultimately lead to compensation running out during his or her lifetime. There is nothing in law, currently, which would allow for him or her to return to the Defendant who caused their injury to seek further compensation. Once again, this potential outcome is a consequence of the risk the Claimant chooses to take in relation to any investments. The law should not be constructed in such a way that either forces the Claimant to take those risks, or requires them to do so for the benefit of the party who caused their losses in the first place.

We endorse the following points made by Personal Financial Planning[4]: that in their experience Claimants (1) do not invest in risk assets with the aim of maximising returns in order to generate over compensation that they can spend on wants' rather than needs', and (2) sometimes accept risk in order to facilitate the maintenance of their needs over time, often having also looked at family support to create a saving on care costs, state support and compromising or foregoing needs.

3. Allowance for investment management charges and tax

It is essential to make full allowance for administration, management expenses, investment advice and the costs of purchasing and selling investments (both initial costs and subsequent changes). The need for such adjustments have been consistently acknowledged by GAD and the MoJ, but the level of the adjustment has not yet been determined nor applied to the modelling conducted by GAD.

Our Mr Chamberlayne, in his role as Chairman of the Forum of Complex Injury Solicitors (FOCIS) has been gathering data on this point from UK firms who regularly act as Deputy and/or Professional Trustee for seriously injured Claimants who have received substantial damages awards. We append a summary of the preliminary conclusions of an analysis of that data by Ian Gunn of PFP as recently submitted to the MoJ and GAD. The broad indication from that data is that overall costs, including investment advice, tend to fall in the range 1.5% to 2%. This is materially higher than the 0.5% figure tentatively suggested by GAD in their 2017 report for the MoJ. It is within, but at the upper end, of the range the GAD referred to in their 2018 report to the Scott ish Parliament. Note that this data did not include the impact of tax on the investment return, as that impact varies considerably between Claimants based on their personal circumstances, but GAD accept that a further adjustment for tax is necessary. This may be achieved in crude terms by rounding up rather than down, until someone conducts the research to enable a more informed adjustment to be made.

In this summary Ian Gunn adds:-

"I should point out that the costs set out above are incurred by claimants in managing cash flows and risk, selecting what to buy and what to avoid, when to sell, holding and keeping track of investments, the income they generate, capital gains and capital losses, and all regulatory compliance. They are not performance fees, i.e. payments to an investment manager in the expectation of achieving superior returns compared to a benchmark, so that the investor ends up better off than without active investment management".

It is essential that the adjustment for all aspects of investment management charges and tax is made to ensure the resultant discount rate reflects the likely net return that these Claimants will achieve on a low risk basis. The alternative is that the law would need to be changed to make such costs a recoverable head of damages. However we contend that is a less satisfactory approach as it will inevitably lead to disputes in many claims.

4. The proportion of claimants who will be under compensated utilising a mixed portfolio approach is inconsistent with a full compensation regime

As strong advocates of the Wells v Wells approach, we are of the opinion that the assumed investment risk profile of the Claimant should be assumed to be that of very risk averse or "risk free".

These issues were previously considered and addressed in a 2015 expert report commissioned by the Ministry of Justice. The appointed panel of experts, consisting of Paul Cox, Richard Cropper, Ian Gunn and John Pollock, came to the majority view that the discount rate should continue to be based on the  Wells approach. The panel cautioned that if the approach of a mixed portfolio balancing low risk investments were to be taken (category b), then only up to 25% of that portfolio should be allocated to non- risk-free assets. The modelling also demonstrated that if the approach were adopted, the discount rate would be 0%.

The majority of the panel did, however,  emphasise that having reviewed Dr  Cox's research on the modern portfolio theory, the emerging risk measures appeared to be inconsistent with the requirement for the portfolio to be considered very low risk. This analysis highlights the very significant deviations that could arise even with a low risk or very low risk mixed portfolio. They could cause a significant proportion of Claimants to be undercompensated and hence leave them unable to meet their needs.

In the Bermuda case of Thomson v Thomson and Colonial Insurance Company Limited, similar issues can be described as having arisen. At first instance in the Supreme Court of Bermuda, Chief Justice Kawaley observed at paragraph 38 that the case appeared to be the first occasion in which a common law court has been required to consider, in the future damages discount rate assessment context, the respective merits of an assumed investment of the entire lump sum to be awarded in ILGS as opposed to in a mixed basket of investments. The Defendants in that case had relied on the expert evidence of

a Canadian Actuary, Christopher Gorham, who advocated a mixed portfolio moderated with what he described as a "safe fund". This was similar in concept to the financial economics approach that is advocated by Dr Cox and considered by the other members of the Ministry of Justice's expert panel back in 2015. This mixed portfolio was also broadly comparable to the "low risk" portfolios that feature in the GAD 2017 report to the MoJ. At paragraph 93 of the Thomson Judgment, it was observed that Mr Gorham: "conceded under cross-examination by Mr. Harshaw that on his investment model between 50 and 33% of plaintiffs would not have sufficient funds. He viewed his approach as fair to both claimants and defendants."

Chief Justice Kawaley commented:

"I viewed his approach as a stunning dilution of the prevailing legal policy preference, in the future loss discount rate calculation context, for a hypothetical investment in an instrument likely to generate a risk-free rate of return."

At paragraph 100, Chief Justice Kawaley also made reference to the evidence of the Claimant's Actuary, (former Government Actuary), Christopher Daykin, as follows:

"As Mr. Daykin explained, institutional investors are able to safely invest in a wider range of investment instruments because they are investing on behalf of multiple ultimate investors whose needs to redeem their investments stretch out over multiple lifetimes. Such investors are also able to hedge against short-term risks in ways which are generally impossible for the typical individual personal injury Claimant. I find that there is a fundamental distinction between the investment goals of the hypothetical prudent investor, especially an institutional investor, (who is not investing sums received by way of compensation for tortious injury), and the investment goals of the hypothetical prudent plaintiff. As regards the latter, there are explicit legal policy imperatives which require the courts assessing the appropriate level of lump sum to award to:

1. assume that the least possible risk will be taken when the lump sumis invested by the prudent Claimant with a view to achieving the goal of full compensation;
2. ignore the commercial realities of how lump sums may actually be invested; and
3. utilize tools for the calculation of the discount rate which are sufficiently simple to be conveniently deployed both in the context of assessing damages in and out of court without the need for the expense of expert evidence save in exceptional cases."

The Bermuda Court of Appeal subsequently fully endorsed the Judgment of the Chief Justice.  

At paragraph 23, Bell JA commented further that:

"What Mr. Daykin was saying is essentially that Mr Gorham's theory of sufficiency demonstrated that, using his model, there is approximately a 50% chance of a Claimant receiving a fund sufficient to meet expenses and losses, with the other side of the coin being that 50% of Claimants would not have sufficient assets to do so. Consequently, Mr. Daykin concluded that these figures come nowhere near meeting the principle of full compensation which has been accepted over many years by the courts. What Mr. Daykin said in relation to the 90 to 95% figures was not that these represented over- compensation on the basis of the Chief Justice's ruling, but that if one were to test a model proposed in place of the Wells mechanism (as advocated by Mr Gorham), then

there would have to be a demonstration that the payments were sufficient for the Claimants in at least 90 to 95% of cases in order to come close to providing full compensation."  

The Ministry of Justice's expert report in 2015 does not appear to have directly addressed the question of sufficiency, this being the percentage of Claimants who would, under the alternative portfolio models proposed by Dr Cox, end up with insufficient assets to meet their future needs. However, reading between the lines in relation to the panel of experts' comments concerning the similar concept of deviation, a significant proportion of those Claimants would end up with insufficient funds. This no doubt led the majority of the panel to come to the conclusion that it was an inappropriate way for calculating the discount rate.

In the GAD 2017 report they acknowledged that the need for the MoJ to consider the issue of what level of under compensation was acceptable within a full compensation regime then make any necessary adjustment to the figures derived from their modelling before setting the discount rate. This issue was echoed in the Justice Committee's Report, published 30 November 2017, in particular paragraphs 22 -24.

The sufficiency question was posed by us to Wrigleys Solicitors and Fensham Howes, firms which routinely advise Claimants on long-term investment following the resolution of their litigation. They both indicated that when considering low risk investment portfolios for their seriously injured clients, they tend to strive for a sufficiency ratio of 90 to 95%. In doing so, they take account of the life expectancy risk to try and ensure that Claimant funds would last to the upper end of the range of probable ages. Fensham Howes' approach to investment is based upon years of research which has repeatedly shown that the perfect investment offering both low price volatility and high long term returns above inflation does not exist.

Due to the boom and bust cycle of the economy the uncertainty caused by this lack of stability means that if a Claimant is in need of funds when the market is low, it will adversely affect his remaining award to his/her detriment and could therefore leave them reliant on the state when the money runs out. This volatility, as outlined in the expert report prepared for the Ministry of Justice in 2015, creates a sequencing risk "which occurs where one year of below RPI investment returns is immediately followed by another, which is immediately followed by another etc. Poor investment return sequences combine with portfolio withdrawals in a highly destructive way because more fund units need to be enchased [sic –encashed] to generate the same annual income. The double erosion of capital following a market fall -the market drop and the drawing an equal income at depressed fund value -is what makes sequencing risk potentially destructive One of the lessons of the technology boom and bust followed shortly by the financial crisis was the importance of the order, or sequence, of extreme investment returns. If a sequence of market drops means the capital of a fund is 50 per cent lower than planned, a 100 per cent gain is needed to return the fund to where it should be"[5]

5. The dual rate split: the Ontario Model and earnings related model 

The Draft Law refers to the UK GAD's analysis and proposes Jersey statutory discount rates of +0.5% up to 20 years and +1.8% applicable to all the years where the loss is deemed to be for a period of over 20 years. The proposed rates and the use of a form of the Ontario model, in splitting periods, lacks adequate supporting analysis for either the 20 year period or the proposed rates. We understand that the 1.8% discount rate is a hybrid of Portfolio A and B over 30-50 year period but we strongly believe this is flawed for two reasons. First, claimants are likely to have a shorter investment life than a composite of 30-50 years. The only claimants that would have an investment life of that length of period would be minors. Since the beginning to 2016, only 3% of Stewarts' PI cases involved minors, therefore the assumption regarding investment period is inappropriate. Secondly, as discussed above, the Portfolio A and B contain 46% and 65%, respectively, in equities and alternatives and therefore cannot be considered a sufficiently low risk portfolio appropriate for claimants. Furthermore, if the Draft Law were passed as proposed, it appears that a Claimant with a loss spanning 19 years would receive more compensation that one of a little over 20 years, as evidenced below:

That is clearly perverse.

We contend that the more important issue is there are to be differential discount rates is to provide full compensation for earnings-related heads of future loss, notably loss of earnings and future care and case management. It is also applicable to other earnings- related heads of loss, such as the cost of a Professional Deputy and medical professionals including therapists.

The impact of earnings inflation and the need for an adjustment has already been closely scrutinised and accepted in the context of a periodical payment regime as reflected by the Court of Appeal Judgment in Flora v Wakom (Heathrow) Limited and Thompstone v Tameside. This has resulted in the general practice thereafter being to apply ASHE 6115 indexation to PPOs for care and case management and less frequently, other earnings- related indices for other earnings-related losses.

The application of this adjustment to lump sum compensation has been considered by the Guernsey Court of Appeal (in the highly regarded Judgment) of Sumption JA (as he then was) and the Privy Council. It has also been considered by the Chief Justice of Bermuda, and subsequently endorsed by the Court of Appeal in Bermuda in  Colonial Insurance Company Limited v Thomson (conjoined with Harvey v Warr en).

Lord Hope in the Privy Council Judgment concerning Helmot v Simon quoted favourably from the Judgment of Sumption JA as follows:

"if an adjustment could be made which would serve to compensate the respondent more exactly for his losses there was no legal reason why it should not be made" (paragraph 40) and "having considered the evidence, Sumption JA said that it seemed to him to constitute strong unchallenged evidence of both the existence of a gap between price and earnings inflation in Guernsey of the order of 2%, and the likelihood that over time it would persist" (paragraph 41).

"As for the question whether there should be more than one rate this seemed to him to be  correct  in  principle  in  a  case  where  there  was  a  significant  difference  between elements representing loss of earnings and care costs" (paragraph 42).

Lord Hope's conclusion, strongly supported by Lord Brown at paragraph 75, was that: "75. if this Claimant is compensated for future loss only on the basis or projected RPI inflation, his award (intended to compensate for increased future losses over the period for his predicted further lifespan of 40 years) will almost certainly prove inadequate to meet his long-term future care needs.

76. The real question in the Appeal must therefore be this: Which approach is preferable: one which ignores the established claim showing an increase in real earnings and so produces an award of almost certainly less than will be sufficient to meet the Claimant's likely future care costs, or one which uses the best available evidence
77. The answer must therefore be obvious."  

Lord Brown at paragraph 82 referred to Sarwar v Ali and oths in which Lloyd Jones J accepted what he described (at paragraph 143) as:

"A consensus between the experts that the RPI is not an appropriate index for Periodical Payments in respect of future loss of earningsThe same is true of the future cost of care and case management."

Finally and pithily from the Helmot Judgment, Lord Dyson at paragraph 113 said:

"There can be no justification for holding that, on these admittedly bare and rather crude facts, damages should be assessed using a discount rate based on RPI inflation. Such an assessment would be bound to lead to under-compensation."

In Sarwar v Ali & the MIB, the Defendant did have an Economist, a Mr Cooper. At paragraph 142 of the Judgment, Lloyd Jones J stated:

"there is near unanimity on the part of the experts in relation to certain of the issues. Dr Wass (Economist), Mr Hogg (Accountant), Mr Copper (Economist) and Mr Hall (Accountant) all agree that average earnings generally increase at a faster rate than prices, that on the balance of probabilities average earnings growth is likely to exceed growth in prices in the future and that, on the basis of historical data, linking Periodical Payments to loss of earnings for RPI would be very likely to undercompensate the Claimant."

At paragraph 161, he observes that:

"She (Dr Wass) concludes that all the measures of earnings growth which she examines exceeds the corresponding measures of growth in pricesShe concludes that measures of earnings growth in the care sector have consistently exceeded the growth in aggregate earnings over the period 1998-2005."

The Judge then quoted from the evidence of the Claimant's Accountant, Mr Hogg, which had confirmed that:

"For the whole period 1963-2006 earnings (AEI) increased on average at 1.9% per annum faster than prices (RPI) but over the last 20 years the rate increase has been lower at 1.53% above price inflation as mentioned by RPI."

The Judge noted at paragraph 163 that:

"Mr Cooper and Mr Hall agree that there is a longstanding pattern showing average earnings rising faster than price inflation."

At paragraph 166 Lloyd Jones J dismissed one of the Defendant's counter-arguments, being:

"wholly unpersuaded by Mr Cooper's arguments based on the possibility of improved productivity in the provision of care for the Claimant in the future."

He observed that was not supported in any way by the care experts.

As can be seen, when this issue has been closely analysed by the courts, the evidence in favour of making an adjustment for earnings-related heads of loss was on each occasion considered overwhelming and required to achieve full compensation. The evidence demonstrated that the scale of the adjustment required tended in the long-run to be between 1.5 and 2%. This would mean it is necessary for there to be a differential earning related discount rate that is between 1.5 and 2% lower than the general (prices) discount rate. This would ensure that the full compensation principle is adhered to and avoid manifest unfairness, notably to those Claimants for whom a Periodical Payment for those heads of loss was wholly unavailable, not offered by the Defendant, or not appropriate for a variety of other factors.

We note that there has apparently been a more modest earnings inflation differential in Jersey in recent years. The short timescale for responding to this consultation has meant we have not been able to scrutinise that phenomenon. It conflicts with the economic evidence advanced in all of the above mentioned cases concerning negative discount rates in Guernsey, Bermuda and Jersey[6]. However, rather than having this issue prescribed by primary or secondary legislation, we propose that the Courts be given a broad discretion to set a differential earnings related discount rate, if the evidence supports such a conclusion. This would be in line with the approach to the issue of the indexation of periodical payments for earnings related losses pursuant to the Damages Act.

Reverting to the concept of the different rates for different periods of time, our position is that we do not consider that to be either necessary or appropriate. We observe this concept was carefully considered and dismissed by the Ministry of Justice's panel of experts, with them commenting at their paragraph 6.7 that they were:

"not persuaded that the discount rate needs to vary by term of loss at the current time, if the discount rate was set by reference to ILGS alone, as the real interest rate yield curve is relatively flat over the long durations."

Fensham Howes, Financial Investment Advisers, emphasise that the risks Claimants face vary with investment time horizon, but do not inevitably, let alone predictably, reduce. For investors with short time horizons, volatility is the greater problem and this leads the prudent advisor to recommend portfolios with little or no volatility risk. Such portfolios will by nature contain a high proportion of low risk fixed income securities and will therefore have very limited potential for real growth above inflation. For investors with long investment time horizons, protecting against inflation is usually the greater problem.

The only common law Judgment which we are aware of in which an adjustment to the discount rate was varied by the term of loss is the Hong Kong case of Chan Pak Ting v Chan Chi Kuen et al. The Defendants in their above-mentioned Bermuda case of Thomson v Thomson and Colonial Insurance Company Limited urged the Bermuda Court to take a similar approach. It was a path adopted by the Defendant's Actuary, Christopher Gorham. However, as Chief Justice Kawaley in  Thomson observes at paragraph 34:

"Chan Pak Ting must be understood in the context of its facts. There was a joint expert report and it was common ground that the ILGS measure was inappropriate as it was not even a potentially available instrument vehicle for the hypothetical reasonable Hong Kong Plaintiff. The vehicles taken into account were not selected as a more appropriate reference tool than ILGS but merely as the best available measure. The Hong Kong Court clearly considered the measures available as a "next best" rather than a "preferred" option."

A statutory introduction on Periodical Payment Orders in Jersey

In England and Wales, the Courts are required to consider, in all cases, whether to make a PPO. The Court can order a PPO even if the Claimant prefers a lump sum, where it is deemed to be in the Claimant's best interests (Sarwar v Ali & the MIB). Our experience is that virtually all UK insurers predominantly negotiate on a lump sum basis, often under Part 36 Civil Procedure Rules, carrying costs consequences for Claimants in failing to accept and fail to beat that offer. Many Claimants, who would prefer PPs, at least for future care, are often forced to settle on a lump sum basis as they are never given a PP based offer.

We fully support the introduction of PPO availability in Jersey, however the power to set the indexation applied to any PPOs should lie with the Courts. Leaving the power in secondary legislation, as is suggested in the Draft Law, is far too crude a measure, which risks falling out of step with use of PPOs. It is far more appropriate to grant the Courts wide discretion so that submissions can be made in individual cases insuring that the Claimants are fully compensated.

The Courts having the discretion to bespoke indexation of a PPO was crucial in the case of X v Estate of Y (Deceased) & oths, in which we represented the Plaintiff. If the PPO indexation had been set by secondary legislation, the PPO would have been entirely inappropriate considering that the Plaintiff is a Polish national who had returned to reside in Poland and consequently faced future losses that were not likely to rise in line with the Jersey economy.

Furthermore, if the indexation of PPOs were to be set by secondary legislation, it is not clear which index would be chosen. There appears to be three main options, being:

1. Price indices, which we argue does not adequately reflect the overall economic trend to be suitable as the set benchmark for all heads of loss types;
2. Jersey earnings indices, which is based on a relatively small group and also may have a high earner skew (due to the socio-economics of the jurisdiction), which may be unsuitable as a bench mark for carers earnings; or
3. English indices, which are based on a sufficiently large sample group benefitting from the wealth of data from ONS as set out in the ASHE tables, notably those related to the earnings of carers. Thisis particularly important as the care claimis usually the largest head of future losses for those with serious injuries. If the indexation is inaccurate then over time the PPO will not provide for the cost of the claimants care claim and that phenomenon will compound year on year.

As shown in the above example, the setting of any of the above indices across the board would be inappropriate and would not sufficiently cater for all heads of losses. This issue is far better left to the judges to apply on the facts and evidence in each case, albeit as with the common usage of the ASHE 6115 index in England, it is likely that a few early court decisions will then establish a commonly applied norm.

What we would like to see in relation to the statutory discount rate?

It is our proposal that the annual revision of the discount rate follows a pre-established formula averaging the real yields on ILGS over the preceding 12 months. The new rate would then simply need to be published/announced by the Government Actuaries. A prescribed formula would remove the danger of subjectivity and personal views of the individuals otherwise chosen to set the discount rate, some of whom may be subject to pressures from different lobbying groups.

A broader review should be conducted by a panel of independent experts. The Government Actuary, acting on the advice of an expert panel, in order to set the differential rate for earnings heads of loss and, if necessary, report to the States of Jersey on any issues that warrant a change to the methodology concerning the discount rate.

We consider that the use of the Ontario model, splitting years and applying arbitrary rates to these periods should be avoided.

If the proposed assumed rate of return on a "low risk" portfolio is taken into account when setting the discount rate and a higher discount rate consequently applied, Claimants will be forced to take risk to overcome these issues. It is also unclear how you calculate the net return on alternative assets without basing data on past performance or simulated past performance which is neither reliable nor compliant with the FCA COBs 4.6.7 - 4.6.9.

We invite the States of Jersey to follow the decisions of the Privy Council in Helmot v Simon, the Bermuda Court of Appeal in Colonial Insurance Company Limited v Thomson and the PPO cases of Thompstone v Tameside and Sarwar v Ali & the MIB. This would allow a differential discount rate for earnings-related heads of loss, notably future losses for care and case management and loss of earnings, to reflect the real and very longstanding differential between pay and earnings inflation which has been variously put in the range of 1.5 - 2%.

We acknowledge that the Draft Law proposes the introduction of 0% floor to the discount rate. While we are not strongly opposed to this, we would like to see this implemented only in relation to price based heads of loss, with the courts being granted discretion to make further adjustments as is necessary, e.g. for cases involving earnings heads of loss, foreign nationals and other such differentials. Implicitly, this would mean departing from the aim of full compensation, which we contend the States of Jersey should not do absent the clearest of evidence and compelling policy reasons.

If notwithstanding all of our submissions the States of Jersey remains minded to pass legislation as contemplated by the Draft Law, it is essential that the rate does not depart from the full compensation principle and does not result in a significant proportion of Claimants going under compensated with the result that their damages run out during their life time. The consequence of this is Claimants would be unable to meet their injury related needs and fall back on the state for support. If this approach is to be taken then we invite the States of Jersey to apply the appended methodology when determining the discount rate(s) to be set. You will see that, using the GAD's highest return on RPI+1%, the result would, on our analysis be, a discount rate of -2.75% for prices related heads of loss and of -4.5% for earnings related heads of loss. If the 0% floor is to be implemented, that would obviously override and so the before mentioned rates would not be considered. However, what this does illustrate is the implausibility of setting a positive discount rate while succeeding in upholding the long established 100% compensation principle.

Julian Chamberlayne

In the event of any queries, please contact Julian Chamberlayne via email at jchamberlayne@stewartslaw.com

Stewarts Law

Solicitors

5 New Street Square

London

EC4A 3BF

For the attention of Julian Chamberlayne

23rd October 2018

Your ref: Our ref: IGG

Dear Julian

MoJ call for evidence: summary of responses by FOCIS member firms

Thus  far,  you  have  received  responses  from  seven  firms  providing  professional deputyship/trustee  services  to  clients  with  personal  injury  awards. The  responses  provide information about the cost of investment, including financial advice, investment management, transaction costs, custodianship and collective fund management costs.

The evidence shows a clear and unsurprising inverse relationship between investment costs and the amount of capital invested. In other words, as a percentage of the capital invested, investment costs decline as the amount of capital invested rises.

Five of the firms provided anonymised data about 98 individual clients, with a total reported value of a little under £120 million. One firm provided aggregate data for 31 clients, with a total value of £30 million, and aggregate data for a further 100 portfolios but without quantifying their value. Therefore, there is data for 129 clients with a total value of around £150 million. The average portfolio in the sample is £1.16m.

The first observation to be made is that, on average, these portfolios are significantly larger than those modelled by GAD for the MoJ. For the latter, GAD was instructed to model portfolios to provide for a loss of £10,000 per annum for 30 years (therefore £300,000 +/- depending on the discount rate applied). Averaging the cost data provided in the sample is therefore bound to understate the real-world cost for such an assumed claimant, and this is an important caveat to the comments overleaf.

Respondents indicated that:

• Annual  fees  for  independent  financial  advice  to  manage  cash  flows  and  overall  risk parameters  were  largely  in  the  range  0.25%  to  0.75%  of  the  capital  invested,  with  the majority in the middle of that range, i.e. around 0.50%. These fees are exempt from VAT.

• Some  respondents  use  individual  discretionary  fund  managers  to  construct  a  tailored portfolio, with or without an IFA. Their reported fees range from 0.85% to 1.0% plus VAT (1.02% to 1.20% including VAT). Additional costs with discretionary fund managers include: custody  fees,  internal  (in-house)  fund  costs,  market  transaction,  brokerage  and  third - party costs. Some of these are paid per transaction and some as a percentage of value. The data is very noisy' and no meaningful average can be calculated for these additional costs, although some allowance for them is necessary. IFA fees are additional, as above.

• Alternatively,  respondents use collective investment fund managers with advice  from an IFA. Fund management fees range from 0.5% to 1.5%. Additional costs with investment funds include custodianship, audit, accountancy and fund administration. The overall cost figure for collective investments tends to fall in the range 1.0% to 1.5%. No VAT is charged on these costs or fees. IFA fees are additional, as above.

• Platform fees are reported in the range 0.1% to 0.3% depending on the value of the portfolio. No VAT is charged on these fees.

• There are outliers above and below the ranges referred to, as would be expected in such a small sample size and with a diverse population of clients.

The broad indication is that overall costs, including advice, tend to fall in the range 1.5% to 2%. I should point out that the costs set out above are incurred by claimants in managing cash flows and risk, selecting what to buy and what to avoid, when to sell, holding and keeping track of investments,  the  income  they  generate,   capital  gains  and  capital  losses,  and  all  regulatory compliance. They are not performance fees, i.e. payments to an investment manager in the expectation of achieving superior returns compared to a benchmark, so that the investor ends up better off than without active investment management. This is often what conventional' investors are seeking, although on average they are unlikely to achieve it. This is a crucial distinction when considering the real world that claimants have to inhabit when exposed to investment risk.

Yours sincerely

Ian Gunn Consultant

2

Adjustments for consideration in setting the discount rate

Step 1:  Estimated annual investment rate of return1 1.00%

Adjustment to account for possiblity that 10% of claimants

Step 2:  will be under compensated2 -1.50% Step 3: Adjustment for investment management fees and tax3 -1.75%

Adjustment to account for differences between UKRPI and

Step 4:

JRPI -0.50% Discount Rate for price related heads of loss -2.75% Step 5:  Adjustment for earnings inflation[1] -1.75% Discount rate for earnings related heads of loss -4.50%

*Please note that even making the above assumptions claimants are still exposed to other risks including: risk that inflation may actually be higher (as only a very limited proportion of the model low risk Portfolios are inflation linked assets); the longevity risk that they may outlive their life expectancy; risk that their needs worsen; the risk that RPI is not representative due the nature of goods required for their specialist needs; and currency risk.

1RPI+1%, chosen as it is the highest return analysed in the GAD report 2017. 2from figure 1 (page 3) of the GAD report 2017. From that graph, this is the

discount rate that would result in roughly 10% under compensation risk with reference to Portfolio A. Therefore adjustment is the difference between +1% (our assumed nvestment return) and -0.5% (the discount rate that ensures only 10% undercompensation risk) = +1.5%

3Based on our initial analysis conducted with PFP (please find appended). Taken the midpoint between 1.5% - 2%.

The Discount Rate

A report for the Ministry of Justice

Prepared by

Paul Cox,

Richard Cropper, Ian Gunn &

John Pollock

7 October 2015

This report has been prepared within the context of guidance provided by the Ministry of Justice on the parameters within which the advice has to be provided. The authors can provide more extensive comment and advice if this is considered helpful.

The content of this report is the considered opinion of the authors but, for the avoidance of doubt, does not necessarily represent the opinion of any professional bodies or working parties where the authors are fellows, members or participants. In particular any input made to this report by Dr Pollock cannot be assumed to be necessarily reflective of the opinions of the Institute and Faculty of Actuaries or the Ogden Working Party.

AUTHORS

Dr Paul Cox is Senior Lecturer of Finance at Birmingham Business School, University of Birmingham and an Investment adviser at The National Employment Savings Trust (NEST). Since 2007 he has provided pension and investment expertise to the Department of Work and Pensions, the Personal Accounts Delivery Authority, and the Department for Business, Innovation and Skills. Paul has been involved in drafting several government consultations on pensions and investments and recently authored reports include Helping consumers and providers manage DC wealth in retirement' (2015) and Metrics and models used to assess company and investment performance' (2014). Prior to academia Paul was a fund manager at Kleinwort Benson and RCM Global Investors. He has been a representative on many regulatory and institutional investment working groups. He's currently a chief examiner of higher level practitioner examinations at a large chartered professional body.

Richard Cropper and Ian Gunn are consultants at Personal Financial Planning Limited, a company of Independent Financial Advisers dedicated to advising recipients of personal injury damages. Richard and Ian provide post-settlement advice on investment planning and management of personal injury damages, personal injury trusts, State Benefits entitlement, statutory provision of care and assistance, taxation and assurance. In addition, they both provide pre-settlement expert opinion on form of an award' issues (periodical payments or lump sum). Richard provided expert financial advice for the claimants in all four of the Thompstone et al  indexation of periodical payments' cases, which have been pivotal in changing the damages landscape in favour of periodical payments in maximum severity.

Dr John Pollock is a Consulting Actuary. He has a background in institutional investment management and within the University sector but has specialised for the last 20 years in providing expert evidence in the UK Courts on a range of matters where actuarial input is required. He has been the representative of the Institute and Faculty of Actuaries on the Ogden Working Party' since the publication of the 4th edition of the Actuarial Tables with explanatory notes for use in Personal Injury and Fatal Accident Cases (the Ogden Tables') in 2000. He was recently involved in drafting the Actuarial Professional Standard X3 : The Actuary as an Expert in Legal Proceedings.

1. The current legal framework and a reworking of the past approach in current economic circumstances.
2. A review of actuarial practice in setting discount rates more generally.
3. What is the appropriate measure of inflation?
4. A financial economics approach to the discount rate.
5. Should discount rates vary by duration of loss?
6. Summary and recommendations.

Appendices

1. Sources of data and evidence used.
2. A financial economics approach to the discount rate – an expanded discussion.
3. Trigger for review.

Set out below is a list of the issues the panel of experts has been asked to consider, cross- referenced to the relevant section(s) of our report.

The current legal framework and a reworking of the past approach in current economic circumstances

1. The function of this report is to assist the Lord Chancellor, and his counterparts in the devolved administrations, in their review of the responses to the first of the two consultation papers issued by the Ministry of Justice in 2012 and 2013. The first was entitled Damages Act 1996: The Discount Rate and how should it be set.  1  The second was entitled Damages Act 1996: The Discount Rate - Review of the Legal Framework. 2
2. The review of these consultation responses, this report and other investigations that are deemed necessary, will assist the Lord Chancellor, and his counterparts in the devolved administrations, in their decision making about at what level the discount rate (or rates) should now be set, how the rate (or rates) should be calculated and what circumstances should trigger future reviews.
3. The Ministry of Justice has provided us with copies of the responses to the first consultation paper. Our brief review of these responses is summarised in Appendix 1 to this Report.
4. In England and Wales the Lord Chancellor has the power to prescribe the discount rate to be used in cases of personal injury and fatal accidents under Section 1 of the Damages Act 1996. More than one discount rate can be set and the Courts retain

1  Ministry of Justice. (2012). Damages Act 1996: The Discount Rate. Available:

https://consult.justice.gov.uk/digital-communications/discount-

rate/supporting_documents/discountratedamagesact1996consultation.pdf. Last accessed 19th May 2015. 2 Ministry of Justice. (2013). Damages Act 1996: The Discount Rate. Available:

https://consult.justice.gov.uk/digital-communications/damages-act-1996-the-discount-rate-review-of-

the/supporting_documents/damagesact1996discountrateconsultation.pdf. Last accessed 19th May 2015.

[1]flexibility to diverge from the prescribed rate should they be persuaded that the circumstances of the case merit such a departure.

5. Our instructions are that the report content is to be framed with the context of the first consultation exercise. This is the approach to the setting of the discount rate within the prevailing legal framework. We cannot comment with authority on the legal issues involved but summarise for completeness and ease of reference our lay understanding of these main principles below.
6. We are grateful for the guidance provided to us by the Ministry of Justice who have in turn sought advice from Counsel.
7. The decision in 1998 by the House of Lords in the conjoined cases of  Page -v- Sheerness Steel, Wells -v- Wells and Thomas -v- Brighton Health Care Authority  3 detail the background to the current legal framework adopted by the Courts in the UK. The decision of their Lordships was followed in principle by the Lord Chancellor in 2001 when he exercised his power under Section 1 of the Damages Act 1996 and set the discount rate at 2.5%. There has been no variation in the rate since this time. The Lord Chancellor had the power to set more than one rate, but declined to do so. The legal issues considered in Wells have been revisited in a more contemporary financial environment in the Judgment of Sumption JA in the case Helmot -v- Simon heard in the Guernsey Appeal Court, and the subsequent appeal of the decision in this case to the Judicial Committee of the Privy Council.  [2]  5  The principal investment related issues which arose from consideration of these various cases can be summarised as follows.

• The purpose of assessing the amount of a lump sum award as compensation for future financial losses is to place the claimant into, as far as is possible, the same financial position as if he had not been injured. In Wells, Lord Steyn referred to the "100% principle", i.e. that the victim is entitled to be compensated as nearly as possible for all future losses.

• The availability of Index-Linked Government Stock (ILGS) enabled the Court to consider a theoretical framework within which a claimant could invest in a portfolio of ILGS which, if held to redemption in an adequately structured portfolio, would provide for the future losses or costs as they fell due, without risk of erosion through inflation (to the extent the loss in future would have moved from current levels in line with the RPI, the basis of indexation of ILGS), or loss of capital. The availability of ILGS then provided the most accurate way of assessing the value of future losses in real terms (at least relative to RPI inflation).

• The claimant is not an "ordinary investor" in the sense of being able to wait for long-term returns to materialise, or to ride-out volatility in asset prices.

• The claimant may choose to invest in other securities offering the promise of higher rates of return but this is not relevant to the calculation of lump sum damages. Consideration of what is actually done with an award, how it is spent or how it is invested is not to be taken into account.

• A notional annuity' approach is envisaged with income from the lump sum award being drawn annually with an additional on-going diminution of capital, this being exhausted over the expected period of loss (itself a function of life expectancy).

• Simplicity of calculation of the discount rate and relative infrequent adjustment to the rate was seen to be desirable.

• An investment in equities was deemed inherently risky. Expenses to meet a future loss for a claimant could not be deferred until financial markets recovered (if they had fallen). Lord Hope noted in Wells that there could be no question about the availability of the money when the investor requires repayment of capital and there being no question of loss due to inflation'. It was noted in Wells that insurers and pension funds invest in ILGS to meet index-linked liabilities. Such action was deemed prudent'.

8. Our subsequent review of investment alternatives to ILGS in Chapter 4 is framed with these clearly stated constraints in mind. This is the basis of our instructions.
9. Our additional instructions are that neither general improvements in living standards nor qualitative improvements in an element of the claimant's loss over time are relevant to the Lord Chancellor's task of setting the discount rate. We are therefore instructed not to take these matters into account. The consequences of this constraint on our advice and opinions is considered in Chapter 3 where measures of inflation are reviewed.
10. Given the requirement for this review of investment alternatives to be made within the current legal framework it seems appropriate to review how the Lord Chancellor set the rate in 2001 and to consider what discount rate (or rates) would now result from a comparable analysis being undertaken at the present time.
11. In Wells -v- Wells their Lordships considered average gross ILGS yields over one and three year periods and made a pragmatic reduction for taxation, as a claimant investor bears taxation on income payable from ILGS. At the time the Judgment was made the average net annual ILGS yield (in excess of inflation) was around 3% and the conjoined cases were decided based on multipliers calculated at this rate. What this implied in practice was that if the award of damages was calculated using a 3% discount rate, and the claimant then invested in ILGS securing such a rate of return (if he desired to do so), then the "100% principle" would have been met.

12. The  Lord  Chancellor  sourced  ILGS  yields  from  the  Bank  of  England  Debt Management Office for the three-year period to June 2001.  The rationale behind Lord Irvine's  decision  is  reproduced  in  Appendix  A.1  of  the  first  consultation  paper referenced  above.   Additional  insight  is  gained  from  the  comments  of  Baroness Scotland in the House of Lords debate on the matter on 29th November 2001. In brief terms, Lord Irvine used a three-year average of those ILGS yields with over five years to maturity. The resultant figure was 2.46%.6 A 15% reduction for tax was deemed appropriate giving a yield of 2.46% – 0.37% = 2.09%. The discount rate was then set at 2.5%. What can be seen is that the Lord Chancellor followed the precedent set in Wells and set the rate with reference to ILGS, but it is known that he additionally considered returns on other securities to be relevant, together with the interests of claimants and those of defendants more generally. The authors presume this further review, beyond the largely arithmetic exercise of calculating and then referring to average net ILGS yields, was reflected in the rounding upwards' decision that was made in 2001. The authors note that only the interests of claimants are deemed

6  Although in a document entitled "Damages Act 1996: Briefing Note on Index-Linked Gilt Average Yield Calculation (Data Issues)", dated 6th July 2001, written by the UK Debt Management Office (DMO), obtained by PFP under a Freedom of Information Request, the DMO confirms that their calculation of the yield on all index-linked gilts for the three years up to and including 8th June 2001 was 2.61%. This was said to be based on the following assumptions:

• All IGs were used in the calculation, with no adjustment for the near maturity portions of IGs close to redemption, the rump stock, or WI stock;

• Daily data yield observations;

• Simple average;

• The use of the standard market convention of a 3% inflation assumption in the calculation of real yields from prices.

This figure was then adjusted to 2.50% for various technical reasons and corrections.

[1]relevant in Wells, the impact of any decision about the discount rate on defendants was not deemed relevant.

13. The Courts have the power to use a different discount rate, or rates, if they can be persuaded that the features of the case justify a departure from the rate set by the Lord Chancellor. The authors are aware of a number of cases involving non-resident claimants  7  and are also familiar with the reported decisions in a number of cases considering changes in economic circumstances more generally  8 and where periodical payments are not available. 9  The Courts have not yet used a rate other than 2.5% in circumstances where the Damages Act 1996 applies, although they have in jurisdictions where the Act does not apply and when assessing future losses in cases not necessarily involving injured persons. [2]
14. Before reworking the analysis made in 2001 in the context of current economic circumstances it may be helpful to detail some of the basic features of ILGS.

• ILGS offer coupon payments and redemption proceeds that are linked to the Retail Price s Index (RPI). There has been some debate over whether Consumer Price s Index (CPI) linked ILGS should be issued in future but none have been issued to date.

• The substantially greater part of the return from ILGS comes from the return of indexed capital. In recent years stocks have been issued with trivially small coupons (as low as £1/8 p.a. per £100 nominal). A claimant investor would be taxed on income but not on gains in capital.

• The  ILGS  market  is  both  substantial  and  liquid.   There  are  securities  with redemption dates ranging from 2016 to 2068. The crude average future term is c21 years.

• There are two types of security, those with an eight-month lag in indexation and those with a three-month lag. For practical purposes this can be ignored. The actual  level  of  future  inflation  is  essentially  irrelevant;  ILGS  protect  again inflation as measured by the RPI.

15. The graph below shows the three-year moving average of yields on ILGS securities with over five years to maturity (as were considered by the Lord Chancellor in 2001). [3] Figures are publicly available for 0% and 5% inflation assumptions (only relevant for the lag') and the 0% figures have been used below.  The differences are not material, as noted above.

Gross  over  5 year  ILGS yields  -­ 1 and 3 year  averages

3.00

2.50

2.00

1.50

1.00

1 year 0.50

3 year 0.00  

-­0.50

-­1.00

16. Current and historic average real yields are noted in the table below.

17. The graph shows quite clearly that there has been a marked decline in ILGS yields since 2001, when the current 2.5% discount rate was set.  This is an investment phenomenon that has not been confined to the UK.  In the US real yields on Treasury Inflation Protected Securities (TIPS) currently average 0.7% [4] and in the Eurozone real yields on inflation-protected securities are 0.1%. [5] The lower ILGS yields in the UK relative to yields available on inflation protected securities in the US and Europe may be explained, at least in part, by the basis of indexation, being the RPI in the UK and CPI measures in other countries. Inflation measures are discussed in more detail in Chapter 3.
18. This is to be expected. Real yields will be higher when economies are experiencing growth in real terms and demand for credit is high, and will be low in periods of weak economic growth when consumers and businesses focus on saving, not consumption. The pace of the decline in ILGS yields accelerated after the financial crash in 2008. Those wishing to invest conservatively have had to accept lower rates than in the past.

A Report by the International Monetary Fund contains a useful summary of trends in

this regard. [6]

19. Although opinions can differ on the matter, from an economic perspective one might expect the real yields on ILGS to reflect investor opinions on likely real economic growth in the UK (or real yields in other countries to reflect likely real economic growth in those countries). Government debt has to be financed and repaid from taxation revenue, which is a function of economic activity.
20. Anticipated levels of real economic growth have implications for anticipated levels of earnings growth relative to inflation. One would expect real earnings growth to be high when real economic growth is high (which in turn would most likely be associated with higher levels of real yields on ILGS) and real earnings growth to be low when real economic growth is low (which in turn would be most likely associated with lower levels of real yields on ILGS).
21. The issue of the appropriate levels of inflation to consider in this exercise is discussed in Chapter 3. In this discussion we are mindful that our instructions are that neither general improvements in living standards nor qualitative improvements in an element of the claimant's loss over time are relevant to the Lord Chancellor's task of setting the discount rate. The discount rate is, put simply, the return after tax and management expenses relative to inflation', whatever that might be considered to be.
22. A further consideration mentioned by their Lordships in Wells was the fact that a lump sum award based on a discount rate calculated with reference to ILGS would most closely match the outcome for a claimant in respect of a Periodical Payment Order (PPO). At the time of Wells these could only be arranged by consent (this situation still prevails in Scotland, Guernsey, Jersey and the Isle of Man) but since the Courts Act 2003 the number of PPOs established, primarily for care costs, has grown significantly. If there is to be consistency between those cases settled by means of a PPO and those cases settled by means of a lump sum, then the purchase of a portfolio of ILGS might be considered to most closely replicate the expected PPO cash flows (setting aside for the moment the issue of the measure of PPO indexation, which now conventionally departs from RPI).

23. If the current Lord Chancellor were to replicate the analysis made by his predecessor in 2001, then:

• The starting point would be the historic average of over five year ILGS yields.

• This is -0.30% over a 3 year period or -0.73% over a 1 year period.

• A modification then needs to be made for tax on the (indexed) coupon payments. The tax burden will vary from claimant to claimant, and their Lordships in Wells and, subsequently the Lord Chancellor, took a broad brush approach to solving this problem by simply reducing the prevailing gross real yield by 15%, suggesting a reduction to the gross yield at the time of 0.37%. The change in yields since 2001 and steady issuance of more securities with lower coupons complicates matters. The deduction for taxation will be modest. Time constraints have not allowed us to consider the issue of taxation in detail and for practical purposes we would suggest that this issue can be addressed in the rounding aspects of the calculation.  

24. The calculations outlined above suggest that the discount rate in current circumstances, if set using exactly the same methods used by the Lord Chancellor in 2001, would be between -1% and -0.5% depending on the allowance for taxation and the averaging period being used. From an actuarial perspective, the one-year average is considered to be more appropriate, and the panel is in agreement that rounding should be to the nearest half-percentage rate for use in the current Ogden Tables. On this basis, the discount rate would be -1%.

25. A practical consideration which merits comment is that the ILGS market extends at the current time to 2068, a period of 53 further years. At the time of the decision in Wells the ILGS market extended for 31 years to 2030  [7]. This period of 53 years exceeds the life expectancy of many claimants (those in their early to mid-thirties in normal health) but is less than the life expectancy of some young claimants (if they have average life expectancy, although many do not). Life expectancy at birth is now around 90 years and this is an effective upper bound on the duration of losses likely to be considered by the UK Courts. The concept of discount rates trending towards an ultimate forward rate' features in insurance applications as will be discussed in the following chapter. Currently, nominal discount rates used for terms beyond those available in the conventional gilt market are somewhat higher than those at the longest available terms. This might suggest use of a modestly higher discount rate to apply to cash flows in the very longest tail of the spectrum being valued. The issue of different discount rates for different terms of loss is discussed in Chapter 5.
26. A focus on ILGS returns can however hide more subtle features of the problems under consideration. If an award of damages is calculated at, say -0.5%, and a claimant invests in ILGS of approximate terms securing a net yield of -0.5%, then he would be fully compensated, but only to the extent that the heads of claim being valued will increase in future in line with the RPI. Chapter 3 will discuss whether claimant inflation is indeed properly estimated by the RPI or if alternative measures are more appropriate. These would impact on the discount rate assessed solely by an inspection of ILGS yields. Put crudely, if a head of claim was expected to grow faster than RPI in future then use of an ILGS based discount rate would overstate the discount rate, leading to under-compensation, and vice versa. To the extent that low real interest rates may be predictive of weak economic growth, and hence earnings growth, this issue may be of less importance than at times in the past when real interest rates were higher.

A review of actuarial practice in setting discount rates more generally

1. The assessment of the present value of future earnings, care costs, services or pension rights is fundamentally an actuarial exercise. The Courts now accept this and the Actuarial Tables with explanatory notes for use in cases of personal injury or fatal accidents (known as the Ogden Tables') [1] are widely used by the Courts in the UK when making future loss calculations. In England and Wales they are admissible in evidence by virtue of Section 10 of the Civil Evidence Act 1995 and they are universally applied in other legal jurisdictions in the UK. This was not always the case. [2]
2. The essentially actuarial aspects of the matter are, firstly, that the expected duration of

the payments being valued is a function of the projected mortality experience of the claimant (which varies by both age and gender) and, secondly, that the future payments require to be discounted' for accelerated receipt. The future payments are generally predictable' in real, inflation adjusted, terms and the discount rate is chosen accordingly. This means the Courts can conveniently make their calculations in current money' terms, the impact of inflation' being disregarded. Comparable calculations are made by actuaries in a large range of commercial applications, pricing insurance contracts, valuing pension liabilities and in all reserving and solvency assessment exercises. From the defendants' perspective the sum of money paid by them to a claimant (unless settled by a PPO) is a transaction  which

extinguishes the liability they have incurred for restitution of the damages under consideration. By contrast, from the claimants' perspective the sum of money awarded represents a reserve to provide for the future cash flows as they fall due, anticipated to be exhausted over the period of assumed loss.

3. The basic process followed when assessing the present value of any set of future cash flows is to determine the amount and timing of the future payments and then to discount these to a current time, having assigned to each future cash flow an estimate of the probability that it will be paid. As an example, consider a multiplier selected from the Ogden Tables. The future cash flows being valued are £1 p.a. and the probability that each future annual cash flow will be paid is the probability that the claimant will have survived to this future time of anticipated receipt. The multiplier is then the sum of the discounted expected' future cash flows. The impact of the discount rate is crucial. At a 2.5% discount rate for a term certain of 50 years, the value of a payment of £1 p.a. is £28.72 (see Table 28 of the Ogden Tables). Of this total around 1% is contained in the value of the 50th payment (see Table 27 of the Ogden Tables). At a -1% discount rate the value of a payment of £1 p.a. is £64.96 and of this total 2.5% is contained in the value of the 50th payment. Not only does the discount rate influence the size of the award but lower discount rates necessitate more accurate estimates being made of the future survival probabilities.
4. In recent years the actuarial profession has moved towards a more standardised approach to the setting of discount rates. This is not only considered to be best practice' but has to an extent been driven by regulatory requirements, sometimes EU driven. A helpful reference is the discussion paper presented to the Institute & Faculty of Actuaries entitled  Developing a Framework For Use of Discount Rates in

Actuarial Work by Cowling et al (2011). [1] This discussion paper built on the earlier work of Daykin & Patel (2010). [2] The Actuarial Profession subsequently set out in the form of a note from the Discount Rate Steering Committee recommendations on how actuarial practice is to be encouraged in this area. [3] The key points to note from these various papers, which can be considered to represent a statement of current actuarial practice more generally, are as follows:-

• Two broad frameworks of discounting are considered, matching' calculations and budgeting' calculations. The issue of the quantification of the present value of future cash flows in a personal injury context is considered to be a matching exercise. The Institute & Faculty of Actuaries confirm this specifically in their response to the second Ministry of Justice consultation exercise. [4]

• A matching discount rate exercise uses a market consistent' approach to the valuation problem. This involves use of essentially risk free' discount rates, to the extent these are observable in the market or can be estimated. These rates can be considered to be the term dependent rates implicit from the market prices of matching instruments (i.e. those that would provide the cash flows being valued in both amount and timing).

• Market consistent (matching) calculations are appropriate when assessing a present value for transaction purposes or when assessing solvency or reserves at a particular point in time. The Institute & Faculty of Actuaries Discount Rate Steering Committee states that Where objectivity and fairness between parties is paramount, and the assets and liabilities are essentially already contractual commitments then a matching, i.e. market consistent approach is nearly always to be preferable'.  This is clearly a relevant consideration when considering the award of damages in a case of personal injury or fatal accident.

• Crucially, no account is taken of the actual assets which might be held in respect of the underlying cash flow liability. An investor might choose to hold riskier' assets potentially offering higher returns but this would only be possible if sufficient excess reserves were available to allow a departure from a more matched position. The fact that a higher expected return might be anticipated from chosen investments does not justify use in any valuation exercise of a discount rate higher than that which would emerge from a consideration of the term structure of essentially risk free rates.

• Standard actuarial practice would then suggest valuing cash flows known in nominal terms using risk free nominal discount rates, derived from the conventional gilt-edged market or from the swap' market, which can provide comparable forward discount rates for durations exceeding those available from Government securities. Similarly, standard actuarial practice would suggest referring to ILGS yields when valuing cash flows which are estimable in real, inflation adjusted, terms (to the extent that the inflation in the head of claim under consideration is RPI linked). It is this latter scenario that is relevant for the purpose of this report.

5. To put these more general comments in context it may be of interest to consider more specific insurance applications. Insurers need to use a discounting framework to place a value on future liabilities held in respect of their policyholders and to assess

the value of future expenses that they will incur when managing their business. Insurers are moving towards a Solvency II framework, which will be mandatory across the EU from January 2016. This is overseen more generally by the European Insurance and Occupational Pensions Authority (EIOPA) [5] and, within the UK, by the Bank of England Prudential Regulation Authority.

6. The finer details of the Solvency II framework are complex but the principal feature of interest for present purposes is the use of a market consistent basis of liability cash flow valuation (see above for discussion). The regulatory framework substantially prescribes that risk free nominal discount rates be used when valuing future cash flows estimable in nominal terms. These rates vary by both country and duration. Currently the UK risk free nominal discount rates are below 1% at the shortest terms (as these are tied to bank base rates, currently 0.5%) and rise to 2.75% at a term of 90 years [6] (for practical purposes no personal injury cases will involve durations longer than around 90 years, this being the expectation of life at birth). Insurers would then specifically allow for inflation in the cash flows being valued (which is arithmetically closely equivalent to reducing the aforementioned nominal discount rates by the rate of escalation in the assumed cash flows). The Institute & and Faculty of Actuaries Discount Rate Steering Committee supports the matching framework for liability valuation under Solvency II'.
7. As stated above, the actual assets held by the insurer to back the liabilities are not of relevance, they do not impact on the selection of the discount rate, or rates.  An insurer with weak solvency will be constrained to matching assets but an insurer with substantial free assets could mismatch liabilities with a riskier asset selection. This could advantage (or penalise) them, to the benefit (or detriment) of their stakeholders, but the valuation of the underlying liability is quite unaffected. The underlying principle is that risk free returns are the most relevant and should be used as the default position. Returns from portfolios involving riskier assets, if they were to be used - and the directives are that they should not - would need to be risk adjusted, using the risk-neutral approach returning one to the risk free framework.

8. The insurance industry has now settled a large number of personal injury claims with the use of PPOs. The Institute & Faculty of Actuaries has a considerable interest in the PPO market as it represents a large, and growing, part of general insurer liabilities. The insurer faces similar problems as do the Courts, and claimants, in these matters. The duration over which PPO payments will be paid is unknown as it is a function of life expectancy (often in the disabled population) and the level of the future payments is a function of the pre-specified rate of increase in the annual payments, as defined in the Order. Periodical payments are now most commonly linked to statistics extracted from the Annual Survey of Hours & Earnings (ASHE) where details can be found on, for example, carer earnings in any year. Since most PPOs relate to compensation for the anticipated expenses of future care for a disabled claimant it has become standard practice to index payments by reference to ASHE 6115 (the 2000 SOC for care workers', which is created by combining SOC 2010 codes 6145 care workers and home carer' and 6146 senior care workers') at a given percentile of the earnings distribution.

9. The actuarial profession, through the PPO Working Party, has surveyed UK insurers with PPO liabilities. The resulting report and associated presentations also contain information on the level of the real discount rate used by insurers when valuing PPO liabilities. [7] This is clearly of interest; it is precisely the issue under consideration in this  report.  Statistics  on  the  discount  rates  used  by  insurers  when  valuing  PPO liabilities  were  gathered  at  the  end  of  2013  and  the  information  collected  is summarised in the graph below. The discount rate, net of assumed inflation in the payments under consideration - generally ASHE 6115, ranged from -1.5% to +1%, averaging close to 0%. It is notable, but not especially surprising, that at the end of 2013 the average yield on over five year ILGS was also 0%.

10. The extent of the range of insurer real discount rate assumptions is perhaps larger than one  might  expect  given  that  insurers  would  be  expected  to  be  using  a  market consistent basis of valuation. This is most likely explained by different assumptions being made about inflation in ASHE 6115 relative to RPI, rather than any variation in

the underlying nominal discount rates for any given future term, although specific portfolio experience may also be a consideration.

11. To summarise, from an actuarial perspective, the valuation of personal injury compensation falls into the matching framework of discounting practice. This is a market consistent approach and in these circumstances the nature of the assets held in respect of the liability is not considered relevant. Risk free term dependant discount rates should generally be used irrespective of the underlying assets held in respect of the liability. When valuing compensation payments fixed in nominal terms this would support use of nominal discount rates measured by market yields on conventional gilts or longer term swaps. When considering inflation linked compensation the corresponding investment class providing market related risk free real rates of return would be ILGS – or the aforementioned nominal term structure of risk free nominal discount rates deflated by an assumed level of future inflation. It is notable that essentially the same principles were adopted in  Wells -v- Wells, their Lordships anticipating subsequent developments in actuarial practice. The risk free rate was used and the actual assets subsequently invested in by the claimant were not considered relevant.

What is the appropriate measure of inflation?

1. Future inflation must be taken into account in an award of damages for future loss or expense.
2. A conventional lump sum award is arrived at by multiplying the annual sum (the multiplicand) and the multiplier. No adjustment for inflation may be made to the multiplicand. The only option is to use a discount rate that is explicitly net of inflation, in calculating the multiplier.
3. With periodical payments there is neither a multiplicand nor a multiplier, just a defined income stream: inflation is allowed for by linkage of the income to a measure of inflation.
4. Intuitively, it might be expected that the same adjustments for inflation would be made irrespective of the form of an award for future losses.  However, following advice obtained from the Ministry of Justice at our request, it is our understanding that this is not the case.
5. A Note from the Ministry of Justice dated 5th May 2015 sets out the following:

1. This note is prepared in response to the following request from the panel:

"Further to our ongoing discussions the panel members agree that we require legal opinion on which we can rely with regard to whether the principles adopted in Wells and established in Cookson -v- Knowles require an adjustment be made to the discount rate to take account of:

1. general improvements in living standards, so that a claimant's award can keep pace with such general improvements; and/or
2. qualitative improvements in an element of the claimant's loss over time, so that a claimant can purchase better quality goods and services as they become available.

This assumes that no adjustment is made to the multiplicand to take account of either 1. or 2. above, in line with our understanding of Cookson -v- Knowles."

2. The response to this request should be taken by the panel as an addendum to the instructions for their task, by way of clarifying the scope of the exercise of setting the discount rate and therefore the task of advising the Lord Chancellor in setting it.

General improvements in living standards, and qualitative improvements in an element of the claimant's loss over time.

3. Neither of these matters are relevant to the Lord Chancellor's task of setting an appropriate discount rate, and neither should therefore be taken into account by the panel in preparing their advice.
4. The rate to be set by the Lord Chancellor must be set by him or her with reference to the return to be expected from the investment of a sum awarded as damages for future pecuniary loss in an action for personal injury.  The discount rate relates to the real rate of return[1] that the market will provide to the claimant as a reward for forgoing the use of the capital he or she invests. As indicated in the powerpoint presentation made on the 13 March 2015 and attached once again for ease of reference, the claimant is of course taken to be a very risk averse investor and the investment opportunities considered by the panel as potential bases for the rate (or, possibly, rates) are each to be judged with reference to the qualities indicated in that presentation.
5. The court will identify and quantify the claimant's loss by reference to the multiplicand/ multiplier method (the "notional annuity" approach). As the panel will be aware, this involves identifying the claimant's annual future loss at values current at the date of trial, this being the "multiplicand". The court then identifies the expected period of the relevant future loss (e.g. the claimant's life expectancy). That figure is of course not used to multiply the multiplicand because it would not take account of (a) accelerated receipt (getting all your future loss right now as a lump sum rather than in increments as you would have done in fact) nor (b) inflation. So the court identifies an appropriate "multiplier" taking into account a number of years' purchase and the discount rate. The relevance of the discount rate is explained by the Court of Appeal in Cooke's case[2] as follows:

- multiplicands can only be "the figure proved as representing the loss at current prices at the date of trial. Inflation and acceleration are built into the multiplier, and the mechanism for doing that requires that a rate of interest be arrived at as the notional return to be earned on the lump sum over the period in question. This rate of interest is what is known as the discount rate".

6. The setting of the discount rate exercise is therefore a limited exercise. The discount rate applies after the basis of the loss – the annual costs of it – have been identified by the court according to the applicable legal principles. Those principles, and the extent to which they do or don't take account of standards of living, available services and so forth, and likely changes thereto,

are a matter for the common law and thus the courts until such time as Parliament wishes to legislate. The discount rate relates only to the primary question identified in section 1 of the Damages Act 1996, as reflected in §4 above.

6. It is clear from the above that the aim, in respect of the appropriate measure of inflation to be considered in the context of setting the discount rate, is not necessarily to create parity with a PPO, even if this may result in differing outcomes for cases where PPOs are not available or suitable, say cases in Scotland or where there is a substantial degree of contributory negligence. The application of earnings measures to uprate PPOs embraces real earnings growth and pay drift, thereby allowing for both general improvements in living standards and qualitative improvements (specifically in care) over time. As Brooke LJ said in Flora:

This brief summary of the recent history of the discount rate used for the purpose of calculating lump sum awards for future pecuniary loss is sufficient to show that an award of a lump sum is entirely different in character from an award of periodical payments as a mechanism for compensating for such loss. When setting the appropriate discount rate in the context of a lump sum award the House of Lords or the Lord Chancellor had to guess the future and to hope that prudent investment policy would enable a seriously injured claimant to benefit fully from the award for the whole of the period for which it was designed to provide him/her with appropriate compensation.

A periodical payments order is quite different. This risk is taken away from the claimant. The award will provide him or her year by year with appropriate compensation, and the use of an appropriate index will protect him/her from the effects of future inflation. If he or she dies early the defendants will benefit because payments will then cease. It is unnecessary in the context of this statutory scheme to make the kind of guesses that were needed in the context of setting a discount rate. The fact that these two quite different mechanisms now sit side by side in the same Act of Parliament does not in my judgment mean  that  the  problems  that  infected  the  operation  of  the  one  should  be allowed to infect the operation of the other. There is nothing in the statute to indicate that in implementing s 2 of the 1996 Act (as substituted) Parliament intended the courts to depart from what Lord Steyn described in Wells v Wells at pp 382H-383 B as the "100% principle", namely that a victim of a tort was entitled to be compensated as nearly as possible in full for all pecuniary losses (see also paras 18-19 above).

7. Consequently, it is on the basis of our specific instructions that the issue of adjusting for inflation is considered.

8. The costs of the goods and services that the claimant will require over life will alter over time. This is referred to as inflation.
9. The House of Lords in  Wells  often refers to inflation, but does not define what elements of inflation should or should not be taken into account. For example:

• Is the inflation that is to be taken into account that specifically applicable to each element of the claim?

• Should the measure of inflation embrace or discount qualitative change over time?

• Should the measure of inflation embrace or discount general improvements in living standards over time?

10. Therefore, whilst the RPI and CPI baskets are updated to take into account changes in purchasing habits, which would capture general improvements in living standards over time, for reasons explained below, these measures of consumer prices in the UK attempt to control the price impact of changes in the quality of products over time.
11. This is explained in the Johnson Review [1] by way of the following example:

This is illustrated for the case of vacuum cleaners in Figure 12.1. Where there is a gap between the two average prices in January there has been a change in the products that are being collected. As the change in price is due to a basket change, rather than an actual price change, the difference between the two baskets is linked out'. Figure 12.1 uses chain-linking at the item level to show a simplified example of the effect that chain-linking can have. As explained in Chapter 2, in the production of the CPI and CPIH chain-linking takes place at the more aggregated class level.  

The top line in the chart shows what happened to the actual arithmetic mean price of vacuum cleaners collected between 1996 and 2013. The bottom line can be thought of as removing the effect of changes between years to the set of vacuum cleaners for which prices are being collected. This is as a result of linking the years together to remove the gaps. The latter lies well below the former, this reflects a shift towards collecting prices for more expensive vacuum cleaners over time.  

There is, for example, a significant divergence in January 2004, where the average price for vacuum cleaners rises from £136.66 in the old basket to £172.39 for the new basket. This increase in price is due to different vacuum cleaners being priced; this could be due to a change in specification of the product, location or outlet. Figure 12.1 also shows there are occasions, such as in January 2013, where there is little or no difference in the average price between the old and new baskets.

The process of chain-linking removes any price change caused by the products collected  changing,  as  opposed  to  the  prices  of  the  products  themselves changing. In other words, it shows the growth in price of the item as if there had been no change in the selection of products that make up that item. In terms of Figure 12.1, the difference in the two lines shows the increasing prices of the new products entering the basket, which can be seen as the quality of the products improving, over the 17 year period.

12. Whilst the chain-linking in the CPI is undertaken in a slightly different manner and at different times, giving rise to double chain-linking, the effect is the same.
13. The Johnson Review concludes:

Even  though  chain-linking  is  not  referred  to  as  a  quality  adjustment,  in practice  this  is  how  much  of  the  change  in  quality  in  items  collected  is accounted for in consumer price statistics. For example, in a new collection year, bread could be priced from a cheaper shop than in the year before. This reduces the average price of the products being collected, but there has been no change in the price of the products. The impact of this price change does not enter the price index, because of chain-linking.  

Another example could be that in the future, 3D TVs might account for enough spending to be deemed representative of what households are buying. ONS could then broaden the description of TVs to include 3D TVs. This change will likely cause the average price for TVs to change, and this change will again be removed by chain-linking.  

Chain-linking just aligns the price growth of different selections of products. It does not imply positive or negative quality change. It is therefore an implicit quality adjustment and does not control for any of the features, in terms of specification or location, of an item when there is a change in average price between years.  

In practice, this means that goods and services whose quality is improving over time (such as technology goods) may show low or even negative rates of inflation, even though the typical price of a product of that type could be similar to what it was several years ago. It implicitly reflects quality change; the new products are more capable than the old ones.

14. It is also important to consider the impact of a product or service becoming unavailable during the year, resulting in the need to select a replacement. This is a complex issue, the detail of which falls outside of the scope of this report, on which the ONS provides guidance as to how comparable replacements are to be selected. In the event that a new product is not considered to be suitably comparable with the previous product, the ONS has a range of tools that it can apply to ensure consistent comparisons over time.
15. The Johnson Review sets out the following to illustrate the impact of the approach over time:

Figure 12.5 is identical to 12.1 but with the addition of a price index series for vacuum cleaners. This index is created using the Dutot formula (explained in Chapter 10). As in Figure 12.1, the difference between the top two lines shows the impact on average price of the quality of vacuum cleaners being collected changing between years.

The price index controls for the effects of non-comparable replacements on average price, where as the middle line, chain-linked growth in price within years, does not. This means that if a product of higher price (and implicitly higher quality) is used as a replacement for another product it is treated as a price rise in the middle line, but is seen as a quality change in the index. The gap between the two lines can therefore be seen as the impact on average price of the quality of vacuum cleaners being collected changing within years.

Figure 12.5 shows a simplified example of the effect that discontinuation and replacement  of  products  has  on  an  index.  Before  2004,  the  index  falls compared with the middle line. This means the non-comparable replacement products  are  on  average  higher  in  price  and  quality  than  the  original products. Between 2004 and 2008 however, the index is increasing at a faster rate than the middle line.  

This implies that the replacement products are of a lower price over this period, and are implicitly of lower quality. However, for items with a short product life the choice of which products are comparable is very important.

16. In  terms  of  quality  change  over  the  long  term,  the   Johnson  Review  states  the following:

Over the long run, treating quality change appropriately poses additional challenges.  Consider  lighting,  an  example  discussed  by  Nordhaus  (1998). Over the course of hundreds of years, society has moved from candlelight to gas lamps to electric light bulbs. Each change in technology makes light cheaper. Over the same period, the cost of wax, then gas, and now electricity has been rising.  

An inflation measure that puts a price on the means of making light (which is what the CPI does) would show a rise in prices. An inflation measure that puts a  price  on  the  cost  of  light  itself  would  show  a  decline,  because  of  the technological improvements over the same period. Astin (1998) makes a similar point about changes to the cost of listening to music over a period of 60 years. Indeed, Astin argues that the difficulty in comparing significant shifts in technology is such that consumer price indices are not ideal for making long-run comparisons over time.  

This is an important issue but may simply reflect the limits of what a consumer price index can do. But it is something that people who use consumer price indices for very long term contracts, such as pension indexation, should perhaps consider.  

17. It is clear to the panel that neither of the primary consumer price indices available in the UK, the RPI and the CPI, is perfect, particularly over the long run. Specifically, the change in the value of the respective assumed baskets of goods and services is not a good match for the basket of goods and services that make up a personal injury claim and it is reasonable to assume that a claimant does not have the ability to substitute goods over time when compared with the average consumer (which accounts for much of the difference between the two measures). That said, it would be both impossible and impractical to have an adjusted discount rate for each element of the claim for the following reasons:

• Many elements of the claim will not have a suitable measure of their past cost, having removed qualitative improvements and/or the impact of general improvements in living standards; making it impossible to estimate how far out historical price changes for one element of the claim is outwith the RPI or CPI; and

• Even where there is such a historical measure of past differential, it is unlikely to be certain that any previous average of the differential will hold good annually over the claimant's future lifetime; and

• Even if such crystal ball gazing were considered appropriate, this would give rise to the application of so many discount rates in each and every case as to cause heightened conflict and complexity.

18. Consequently, the panel understands that the inflation adjustment to the discount rate can remove speculation by reference to the yield on ILGS, but cannot remove the imprecision of applying a general measure of inflation, as distinct from the more precise path that periodical payments have followed.
19. Both the CPI and the RPI meet the requirements of our brief with regards to the exclusion of the impact of general improvements in living standards and qualitative improvements, insofar as steps are taken to remove any such impact.
20. Clearly, ILGS are linked to changes in the Retail Price s Index (all items); the RPI. The RPI is acknowledged by the Office for National Statistics as being flawed; in March 2013 the RPI was downgraded from being a "National Statistic", and might now fail the test of a reliable measure established by Mackay J in RH[2], for the following reasons:

1.2.1  The Statistics Authority judges that the RPI, including the sub-indices and variants listed in section 1.1.2, does not comply with Principle 4 and specifically with Principle 4, Practice 5 of the Code. This view is based primarily on:

1. the finding that the methods used to produce the RPI are not consistent with internationally recognised best practices (para 3.4); and
2. the decision to freeze the methods used to produce the RPI, and only to contemplate routine' changes (para 3.5).

21. As a result, many pension schemes and State benefits have already switched to the government's preferred measure of inflation, the CPI, and the future issue of CPI- linked  gilts  would  not  be  surprising.   Furthermore,  guaranteed  income  payments under the no-fault statutory compensation scheme for members of the armed forces injured in service are also linked to the CPI.
22. There is also evidence of an increasing wedge' between CPI and RPI. Historically, CPI has been around 0.7% pa below RPI. However, in its paper of November 2011 entitled  "The long-run difference between RPI and CPI inflation"  the Office for Budget Responsibility (OBR) states that it expects the wedge' to rise to between 1.3% and 1.5% pa. Importantly though, the OBR set out the following in their March 2015 "Economic and fiscal outlook":

23. It is clear from the above that the largest contributor to the difference between the two inflation measures is the formula effect'.

24. According to the OBR, the geometric mean is "better suited to accounting for the effect of substitution between goods and services when relative prices change. The [geometric mean] of a given set of non-identical positive numbers is lower than the [arithmetic mean] of the same set of numbers."
25. The CPI is held to be the more accurate measure of price inflation in the UK economy and, therefore, preferred by the government. Indeed the Johnson Review concluded:

The Authority and ONS should make it clear to users that the RPI is not a credible measure of consumer price change. The RPI should not be used for new contracts. Taxes, benefits and regulated prices should not be linked to the RPI.

The RPI should also not be used as the measure of inflation when comparing living standards over time, at least for recent years where better consumer price indices are available. The issuance of index-linked gilts is more complex, but government should move away from selling gilts linked to the RPI, subject to consultation and assurance about the demand for CPI or CPIH linked gilts.

As the headline RPI is not a robust measure of inflation, it makes little sense to continue to produce analytical indices, such as the Pensioner Price Indices and the Tax and Price Index, which are based on the RPI. Where there is demand for analyses of this type to continue, these can be replicated on a CPIH basis (where they are not already), or within the scope of the subgroups analysis set out in recommendation 2.

26. Unlike linking a new PPO to the RPI, relying on the RPI for the basis of the discount rate is not, in our eyes, a new contract, but a reliance on existing contracts for RPI- linked ILGS. That said, on the basis of the above it is clear that on current long-term forecasts the RPI is positively biased, as a measure of price inflation compared to the CPI, by around 1% per annum.
27. Actuaries make assumptions about RPI and CPI levels when valuing pension scheme liabilities and a recent survey of IAS19 Assumptions suggested that RPI might be expected to outpace CPI by 0.95% p.a. [1]

28. As a result, it is appropriate to address whether the RPI-based yields considered in Chapter 1 should be upwardly adjusted to take this wedge into account. We consider that there is some difficulty in doing so, as we are of the view that claimants are unlikely to be able to substitute needs in the same way as the average consumer (with the assumption in respect of substitution accounting for the majority of the wedge) and, particularly, given the contents of Table A above. Whilst these are stated as being long-term' estimates, it is our understanding that they would not be considered robust over a 40 to 60-year duration; which is the duration of loss in many personal injury cases.
29. To illustrate the difficulty, if the discount rate had been increased by 1.4% following publication of the OBR estimate of the long-run wedge in 2011, the fact that the wedge has been 0.7% means that significant under-compensation would have occurred.
30. That said, we refer back to our primary instructions which state that:

The discount rate relates to the real rate of return[2] that the market will provide to the claimant as a reward for forgoing the use of the capital he or she invests.

31. In this regard we note that the Bank of England is the market maker in respect of RPI- linked gilts and continues to issue them through the Debt Management Office (DMO). However, market yield on any gilt can only be derived from observation of actual trades (as distinct from the yield at issue), and there is no market in CPI-linked gilts. In other words, there is no market rate to refer to in making an adjustment for the wedge.
32. The Johnson Review states the following with regards to the issuance of RPI-linked gilts:

The RPI is still used in large numbers of commercial contracts, including in £470 billion worth of index linked gilts. The sale of these gilts is used by the government to fund part of the national budget deficit. Some of these gilts will not be redeemed until 2068 and in future new gilts may be issued with even longer terms. And the continued existence of long-term liabilities linked to the RPI continues to create demand for assets that provide returns also linked to the RPI, regardless of the statistical issues.

So despite its statistical inadequacy, the RPI cannot simply be discontinued. Indeed, cognisant of these needs, ONS also noted that there is significant value to users in maintaining the continuity of the existing RPI's long time series without major change' (ONS, 2013). It was with this in mind that the National Statistician recommended that the formulae used at the elementary aggregate level in the RPI should remain unchanged.  

The UK Statistics Authority and ONS have important roles to play in communicating the primacy of CPIH, explaining what it is measuring and how it should be used. The Authority and ONS should also be very clear in explaining that the RPI is not a credible measure of consumer price change. They should make it clear that the RPI is not fit for purpose and should not be used except where existing legal contracts – for example index-linked gilts – demand it. If a way can be found to discontinue the production of the RPI while maintaining the integrity of these legal contracts then it should be pursued. In the course of the review we have not found a satisfactory solution to this conundrum.