Public funding of impecunious litigants is being radically reformed by the government under the Access to Justice Act 1999. Legal aid is to be deconstructed. Personal injury claims will be excluded from public funding from April 2000 and must be funded under conditional fee agreements backed by insurance policies. Public funding will be provided under a Funding Code with flexible criteria but ones which are more strict than under legal aid.
A draft Funding Code was published for consultation in January 1999. A revised draft was issued in October 1999. Those who need to know about the Code will need to read it in detail. The purpose of this paper is not to summarise the Code but to summarise some of the recent research carried out by the Legal Aid Board's Research Unit on which decisions under the Code were based.1 The research covered a cohort of concluded cases which had been funded by legal aid and contains some highly interesting facts and conclusions on.
- solicitors' initial assessments of success and of damages
- actual success rates
- particular problems of clinical negligence cases
- the impact of the Funding Code on not funding cases previously supplied by legal aid, and conseqential savings in public expenditure.
Assessments by solicitors of likelihood of success of claimants
Likelihood of success bands Predictions of success by % of cohort
(n = 942)>80%25.960-80%45.350-60%10.7<50%0impossible to say17.82
The study found that solicitors find cost prediction difficult. Just under 50% of cases were predicted to involve costs to settlement of between £1,500 to £2,500. These estimates were relatively accurate, since it was shown that costs estimates remained unaltered in 69% of cases where amendment applications were subsequently made.
Predictions of damages
The average predicted value of damages in the sample was £16,383, which was significantly higher than the median, which was in the range of £5,000 to £7,500. This was because of a small percentage of higher value cases. The overall level of damages predicted was low (one third less than £5,000, over half less than £10,000).
Damages=£100,000unpredicted or not applicablePercentage5.17.7126.96.36.199188.8.131.52.324.9
All solicitors claimed that their experience of previous cases and outcomes was paramount in predicting outcome and damages since, at the initial stage of applying for legal aid, they would not have reports from appropriate experts nor knowledge of any defence evidence. Cases differed according to their type in relation to predicting damage levels: it was noted that it is relatively easy to quantify personal injury damages because of the existence of a tariff for particular injuries, but difficult to quantify housing cases and initially impossible to quantify clinical negligence cases.
Damages to costs ratios
The mean damages to cost ratio was 6.4:1, although as the bulk of legally aided cases concern relatively low value claims, the median was significantly lower than the mean at 2.5:1. In only 4% of all cases in the sample did predicted costs exceed predicted damages but the researchers suggested that the number nationally may be significantly higher.
Overall, 45% of cases which had been funded by legal aid would have failed to meet the criteria proposed under the draft Funding Code. The effects would vary between the merits bands, with 82% of the >80% success band remaining funded (if predicted damages had to exceed costs), although the 60-80% band would have been quite severely affected (only 42% funded on a proposed ratio of damages exceeding costs by 3:1) as opposed to the 50-60% band (45% funded if damages exceeded costs by 4:1). All cases with predicted values of £25,000 or more met the proposed requirements of the Funding Code consultation paper, only around 40% of predicted cases of less than £5,000 did so. However, the low value cases overall scored much more favourably in terms of predicted likelihood of success and were also far more likely to be successful in fact - with the exception of personal injury cases.
On the basis of the figures, the Board amended its original proposals in relation to the damages to cost ratios to be applied to different success bands under the Funding Code, so as to apply a 2-fold rather than 3-fold ratio to the 60-80% band, as summarised below. This would permit the funding of more claims, but would still exclude a significant number of cases which have been funded under legal aid.
Prospects of successOriginal proposalRevised proposal Ratio of damages (D) to costs (C)Percentage of legal aid cases which would be fundedRatio of damages (D) to costs (C)Percentage of legal aid cases which would be funded80+%D>C82%D>C82%60-80%D>3C42%D>2C65%50-60%D>4C45%D>4C45%
Actual success rates
The outcome of the cases revealed variations in success rates according to the type category of a case. In relation to those cases involving money, success rates were
clinical negligencec30.6%personal injury71.2%contract36.5%debt21.1%probate42.9%housing disrepair47.3%
The significant points to note here are the comparatively high success rate in personal injury cases but poor rate in clinical negligence cases. (The latter should be compared with the high success rate in clinical negligence cases reported during the past 2 years by those CFA insurers who insist on independent scrutiny of claims before agreeing cover).
The actual results by merit band were:
merit band% successful>80%80%60-80%53%50-60%48%<50%0%impossible to say36%
The researchers concluded that these results show that the predictions of success seemed to be on the high side. Only the results of the >80% band even fell within the predicted range. Less than half of the 50-50% band saw damages awarded to the claimant, and only just over half of the claims in the 60-80% band did so. The researchers concluded that "The results are consistent with the expectation that solicitors will err in their own favour where there is doubt as to precise figures". Two-thirds of the cases in the "impossible to say" category lost.
Costs were fully recovered in 52% and partially recovered in 5% of claimant cases. The average cost of sample claimant cases was £4,758 although there was great variation between individual cases (range £51 to £84,500) and the median was much lower at £2,450. The bulk of cases cost less than £5,000, with around half costing less than £2,500. The mean cost to the fund per sample claimant case was £1,441.
Solicitors' cost predictions were accurate only around 40% of the time:
£ <1,5001,500 - 2,500 >2,500All% of cases
The researchers commented "It may well be again this is an example of predictions being given in a manner advantageous to those giving them".
Clinical negligence cases
Evidence was available of the investigation costs in clinical negligence cases. Around one half of clinical negligence cases were discontinued in 1996-7 after initial investigations. The mean cost of these cases was slightly over £2,500, although this cannot be taken as representative in view of the subsequent restriction in handling such cases to specialist firms, and does not include the investigative costs of cases which proceed beyond the investigative stage. In the sub-set of 1996-7 cases conducted by AVMA and Law Society panel, who may be expected generally to conduct more difficult cases, the mean cost of cases discontinued was over £4,000. Fewer cases handled by panel firms were discontinued, had substantially higher success rates and obtained higher damages (£48,653 versus £29,639). The researchers estimated that current investigation costs in clinical negligence cases are mean £5,000 but typically £3,000, made up of around 65% profit costs and 35% disbursement costs, of which just 4% represents counsels' fees.
The average damages obtained was £17,817, which was slightly higher than the average of predicted damages (£14,189). The median (£4,500) was the same in both. There was a similar distribution pattern across the damages bands correlation between predicted and actual damages.
Damages obtained in sample clinical negligence cases were significantly higher (average £25,164, median £6,500) than the general figure. Although the majority of these cases involved damages of over £5,000, only a small percentage involved damages above £25,000. Just over 25% of successful cases involved damages under £5,000 but only 18% of cases involved damages predictions of the same level. Over 25% of cases involving damages estimates of over £7,500 were successful, but only one-tenth were successful where estimates were lower.
The mean damages to costs ratio of cases resulting in damages was 3.6:1 (i.e. lower than predictions) and median was 1.52:1. The researchers concluded that ratios predicted by solicitors were optimistic. 55% of the cases would fail to meet the Funding Code consultation paper cost-benefit criteria on the basis of actual outcome damages.
Savings in government expenditure from changes under the Access to Justice Act 1999
The Access to Justice Act 1999 excludes in Schedule 2 certain types of cases from public funding. The Board estimates that this would directly save around £41 million (less £5 million negated by loss of retained contributions and receipts from the statutory charge), of which £36 million relates to personal injury negligence cases (not clinical negligence), which will have to be funded under a CFA.
Costs in personal injury cases
The Board also analysed 28,523 personal injury case records (excluding clinical negligence) closed over the 6 months to August 1999. Details included:
gross claimant expenditure£100 million, of which 4/5;
was recovered from
defendantssuccess rate 66%average gross case cost£3,600average net case cost£638median gross case cost£2,089mean cost of successful cases after cost recovery£2,946median cost of successful cases after cost recovery£1,702
The 10% of cases with the highest gross costs accounted for 43% of total gross costs, averaging £12,538. However, the general position was less skewed when net costs were taken into account: the difference between most and least expensive 10% of cases was 45 fold in respect of gross costs but just 9 fold in respect of net costs, demonstrating that a far greater proportion of costs are recovered in the most expensive cases (83% in the most expensive 10% of cases but just 14% in respect of the least expensive 10% of cases).
Thus, the great majority of cases involve relatively little gross or net expense. Relatively few cases cost over £10,000, but those that do account for over one-quarter of total gross costs are one-fifth of total net costs. Almost three-quarters of legally aided personal injury cases involve gross costs of less than £3,000, and over half less than £2,000, but only 20% involve gross costs less than £2,000. Just 0.2% of cases involve gross costs of £50,000 or more, but these account for 5.3% of total gross costs and 4.4% of net costs. These expensive cases are unrivalled in money cost benefit terms because of their high damages to costs ratio and very high gross costs to net costs ratio. But they consume massively disproportionate expense compared to low cost cases.
The introduction of requirements to predict the likelihood of success of cases and that cases must satisfy a damages to cost ratio which varies according to predicted success bands will clearly reduce the number of cases that will be given public funding under the Funding Code as compared with the previous legal aid arrangements. It is estimated that this is likely to save the government at least £36 million a year.
The vast majority of cases which receive public funding involve low damages (one third predicted to be less than £5,000 and over half less than £10,000) and costs (the bulk costing less than £5,000 and around half costing less than £2,500, with a mean net cost of £1,441). A very small number of very high cost cases account for a large amount of costs (e.g. 10% of cases account for 43% of gross costs), although the net cost position is significantly better than the gross cost position and a greater proportion of costs are recovered in the most expensive cases.
Success rates vary according to the type category of a case. Only about half of the cases with predicted chances of success less than 80% actually succeeded. Only a third of the cases succeeded where the initial chance of success was considered as being impossible to predict: this category contained a very high number of clinical negligence cases.
1Testing the Code - Legal Aid Board 1999.
2The "impossible to say" category contained a very high number (63%) of clinical negligence cases compared with other categories: personal injury (11%) contract (14%), debt (0%), probate (15%) and housing disrepair (12%). The study noted interview evidence that some solicitors might use this category for "I don't think the client has a hope in hell, but I haven't got the bottle to tell the client".