Background
1. In order to provide information about costs and to inform the work
of the Inquiry into civil litigation, in 1995 Lord Woolf asked the Supreme Court Taxing Office (SCTO) to collect comprehensive information about costs from bills submitted for taxation. The results of a preliminary analysis of litigation costs in 673 cases were published as part of the Inquiry's interim report Access to Justice in June 1995. Following further data collection, a fuller analysis of costs and case length in High Court litigation has been carried out on 2,184 cases sampled from those submitted to the SCTO during the period 1990-1995 1
2. The analysis describes total average costs (on the winning party's side only 2) in cases of different type, weight and value, and provides a breakdown of average costs for particular elements in bills, such as counsel fees, expert fees, discovery and documents. Variations in case length have been analysed in relation to factors such as case type, weight, claim value, and the presence of legal aid. In addition, three separate statistical modelling exercises were undertaken, in order more fully to understand the factors that affect the length and cost of litigated cases. Some of the significant findings of the study are summarised here.
Characteristics of sample
3. Case type: The sample comprises 2,184 cases divided roughly equally between 10 case types: medical negligence, personal injury, professional negligence, Official Referees', breach of contract, judicial review, Chancery, Queen's Bench 'other', Commercial, and bankruptcy/Companies Court cases.
4. Claim value: Claim value was based on the amount of money recovered in the action. The highest claim value in the sample was £660 million in a Commercial case, and the lowest was £5 in a breach of contract case. There was a wide range of claim values in the sample both within and between case types. Commercial cases had the highest median 3 claim value at £118,454. Other case types with high median claim values were bankruptcy/Companies Court (£50,269), personal injury (£50,000) and breach of contract (£37,875). The lowest median claim value was among Queen's Bench 'other' cases (£29,000). About one-third of cases in the sample had claim values of £25,000 or less, and one-quarter had claim values of £100,000 or more.
5. Case weight: The weight of cases in the sample was assessed by the taxing office using five categories: category A (heaviest weight) to category E (least heavy). Factors taken into account in the categorisation were: complexity of issues, importance of case to the parties and to other litigants and potential litigants, unusual expedition, number and importance of documents, and quantum. The vast majority of cases (90 per cent) were assessed as being at the light end of the range (C-E). There was, however, much variation between case types in this respect.
6. Legal aid: Some 35 per cent of all winning parties had legal aid, but the proportions varied among different case types. Legal aid was most prevalent among medical negligence (92 per cent), personal injury (59 per cent), judicial review (58 per cent), and professional negligence cases (38 per cent). Among other case types, only a minority of winning parties had legal aid (29 per cent Chancery; 18 per cent Official Referees'; 17 per cent breach of contract; 18 per cent Queen's Bench 'other' and 9 per cent bankruptcy and Companies Court).
7. Stage of proceedings reached: There was considerable variation among cases in the stage that had been reached before proceedings ended. About a quarter of cases got only as far as the issue of proceedings before the case was settled. In just over a quarter of cases the matter was concluded after it had been set down, but before trial. In 17 per cent of cases the parties reached the doors of the court before settling, and about one-third of cases had proceeded all the way to trial.
8. How proceedings were concluded: Nearly half the cases had been concluded on the basis of a consent order, and one-quarter on the basis of a judgment. In a further 9 per cent of cases, a payment into court had been accepted; 4 per cent of cases were concluded on the basis of a consent judgment; 2 per cent of cases had been struck out; 1 per cent of cases had been dismissed; 7 per cent had been concluded following summary judgment; and a further 1 per cent had been withdrawn or discontinued.
9. Successful party: 86 per cent of cases in the sample had been won by plaintiffs, reflecting the kinds of cases that are submitted for taxation. Among medical negligence cases none had been won by the defendant.
Average costs
10. Costs and case type: A little under one-half of bills in the sample were allowed at £10,000 or less (42 per cent); just over one-quarter of bills (28 per cent) were allowed at between £10,000 and £20,000; 10 per cent of all bills were allowed at between £20,000 and £30,000; and just over one-fifth (21%) of all bills in the sample, on the winning side only, were allowed at more than £30,000. Average costs varied between different case types. Official Referees' and Commercial cases had the highest average costs (median £19,320 and £18,897 respectively) and judicial review cases had the lowest average costs (median £7,642). The highest minimum costs were found among medical negligence cases (minimum £3,759). Only a quarter of medical negligence and Official Referees' bills were for less than £10,000 (Table 1). The case types with the highest percentage of costs over £50,000 were Commercial cases, Official Referees' cases, professional negligence cases and medical negligence cases.
Table 1: Categories of costs by case type
11. Costs and case weight: Average costs tended to be higher among heavier cases. The difference in the weight profile of case types to some extent accounts for observed variation in average costs between case types. For example, medical negligence and Commercial cases had a relatively high proportion of heavy cases, while the vast majority of breach of contract, judicial review and Queen's Bench 'other' cases were assessed as being light (Table 2). Differences in median costs between case types appear somewhat smaller when comparing cases of similar weight (Table 2). However, it is also clear that average costs among the most straightforward cases are relatively stable at around £6,000, irrespective of case type. It should be borne in mind that these represent the costs of only the winning side in the actions.
12. The average level of costs in cases of different weight is quite striking. Among most case types, simply moving up from weight E to weight D doubles average costs, and there were substantial differences in average costs between the heaviest and lightest cases (Table 2).
13. There is an association between claim value and claim weight which is to be expected since quantum was one of the factors taken into account in assessing case weight. Thus the average claim value among cases in category A is much higher than the average claim value among cases in category E. The pattern is not, however, consistent since there were high value cases that were not assessed as very heavy and vice versa.
Table 2: Median costs among case types in relation to case weight
14. Costs and claim value: There was a huge amount of variation in the level of costs charged for cases of similar value. Among the lowest value claims (up to £12,500) median costs in the sample as a whole were £8,318, but costs for claims with value not exceeding £12,500 ranged from £1,798 to £200,515. The highest value claims (more than £250,000) had median costs of £36,951, with costs ranging from £3,001 to £2,135,412.
15. Analysis of the broad relationship between costs and claim value reveals that, as claim value rises, the proportion of costs allowed at less than £10,000 steadily falls. Among the lowest value claims, 60 per cent of cases had costs of less than £10,000; among the highest value claims (over £250,000), only 18 per cent had costs of less than £10,000, and 41 per cent had costs of over £50,000.
16. There is a lack of proportionality between costs and claim value at the lower end of the claim value scale. Among cases with a value of less than £12,500, about one-third (31 per cent) had costs between £10,000 and £20,000, with a further 9 per cent of cases having costs in excess of £20,000. Thus, in 40 per cent of the lowest value claims, the costs on one side alone were close to, or exceeded, the total value of the claim. By way of contrast, in 60 per cent of claims over £250,000, costs represented less than 20 per cent of the value of the claim.
17. There were substantial differences between case types in the average levels of costs for similar value claims (Table 3). For example, among cases with a claim value below £12,500, the median costs figure in medical negligence cases was £10,482, and in Official Referees' cases the figure was £12,245; median costs in Commercial cases in this claim value category were about half, at £6,187, and the median figure for personal injury cases was £7,099.
18. Table 3 also shows the differences between case types in the proportion of cases falling within different claim value brands. Personal injury and Commercial cases had the lowest proportions of cases with values of £12,500 or less (14 per cent and 9 per cent respectively), while all of the other case types had between a fifth and a quarter of cases in the lowest claim value band.
Table 3: Median costs allowed by claim value
19. When costs are expressed as a percentage of the value of the claim 4, median costs among the lowest value claims (under £12,500) consistently represent more than 100 per cent of claim value, and this holds true among all case types (Table 4). The lower the claim value, the higher the percentage of the claim value that costs represent. At the lowest end of the claim value spectrum, the costs of the winning party only substantially exceed the value of the claim. Among claims with a value of between £12,500 and £25,000, costs as a percentage of claim value range from 41 per cent among personal injury cases to 96 per cent among Official Referees' cases. At the highest end of the claim value scale, costs range from around 1 per cent of claim value among Chancery, and bankruptcy/Companies Court cases, to 19 per cent of claim value among Official Referees' cases.
Table 4: Costs as a percentage of claim value (costs/value x 100)
20. Costs and case duration: A simple analysis of costs in relation to case length reveals higher average costs among longer cases. For example, cases lasting less than one year had average costs of £14,745 (median £8,008); cases lasting two to four years had average costs of £28,921 (median £14,758); and cases lasting for more than six years had average total costs of £30,373 (median £15,685).
Case duration data
21. Case type: Duration was measured from the date of first instruction to the date of conclusion. Medical negligence cases lasted longest on average (mean 65 months), followed by personal injury cases (56 months) and professional negligence cases (41 months). The most rapidly concluded cases were judicial review (12 months), Companies Court cases (about 13 months) and Commercial cases without a claim value (16 months) (Table 5).
Table 5: Average duration (in months) from date of instruction to conclusion by case type
22. Claim value: Analysis of case length and claim value revealed no consistent pattern in the sample as a whole.
23. Legal aid: In the sample as a whole, cases where the winning party had legal aid lasted longer, on average, than cases where the winning party did not have legal aid. With the exception of judicial review cases, this difference held true within case type. The mean duration of case from first instruction to conclusion for the whole sample was 26 months among cases without legal aid (median 22 months) as compared with 49 months among cases with legal aid (median 48 months).
24. Means of concluding proceedings: Cases with the longest average duration in the sample as a whole were those that were concluded on the basis of a consent judgment (48 months ), acceptance of money in court (44 months) or a consent order (42 months). The average length of cases ending after judgment was 25 months from instruction to conclusion, while cases concluded on the basis of a summary judgment had an average duration of 19 months. The association between duration and means of concluding proceedings varied between cases of different type.
25. Weight of case: In the sample as a whole, heavier cases appeared to take longer to conclude. The heaviest cases (weight A) took, on average, 48 months, as compared with 29 months among the lightest cases (weight E). The pattern was not consistent within case types.
26. Interlocutory activity: In the sample as a whole, there was an association between the average length of case and the number of interlocutory applications made. In cases where there were no interlocutory applications, the average length of case was 28 months. This figure rose to 50 months in cases where there were seven or more interlocutory applications, and the pattern held true within every case type.
Delay between first instruction and issue of proceedings
27. Case type: Medical negligence and personal injury cases had the longest average delays between instruction and issue of proceedings. The average delay in medical negligence cases was 21 months and in personal injury cases it was 17 months. The shortest average delays were in Commercial cases with no claim value (two months), Companies Court cases with no value (two months), and judicial review cases (three months).
28. Legal aid: Within the sample as a whole, cases with legal aid generally appeared to have a longer period of delay between instruction and issue of proceedings than those without legal aid. The average period of delay among cases with legal aid was 15 months (median 12 months), as compared with an average of seven months (median three months) among cases without legal aid.
Statistical modelling
29. In order to attempt to understand, rather than merely describe data relating to costs in litigation, a series of statistical modelling exercises was undertaken. Such procedures help to isolate, from the wealth of data collected, systematic relationships between costs and other variables, and to disentangle some of the relationships. For example, an apparent association between costs and claim value might be explained by the fact that cases with higher claim values tend to be more complicated, and therefore when cases of similar levels of complexity are compared, differences in claim value actually make little or no difference to costs levels. It is therefore helpful to analyse the independent effect of such factors as value and weight on costs levels, in order to assess which factors are the most important in explaining costs. Using modelling procedures, it is possible, simultaneously, to compare cases of similar type (in terms, for example, of weight, value and case length) and to assess the effect of different factors on costs levels.
30. Three separate modelling exercises were undertaken to explore the effects of a number of factors on costs, on case length and on delay between date of instruction and issue of proceedings. The objective of the modelling exercises was to indicate general trends in case duration and costs. There is considerable scope for further work to be done on analysing both case duration and costs levels in civil litigation.
Modelling of delay between instruction and issue
Effect of case weight/complexity on delay
31. The results of the modelling exercises suggest that the weight/complexity of cases significantly affects delay between instruction and issue of proceedings in five case types, but the impact of case weight is inconsistent. In professional negligence cases, Official Referees' cases and breach of contract cases, increasing weight of cases significantly increases the period of delay between instruction and issue of proceedings. Thus, the heavier the case, the longer the delay.
32. In personal injury cases and Queen's Bench 'other' cases, increasing weight appears to lessen the period of delay between instruction and issue; thus, the more straightforward the case, the longer the period of delay.
33. In medical negligence, judicial review and Chancery cases, weight appeared to have no significant effect on the period of delay between instruction and issue of proceedings. There was no evidence that the period of delay in lighter cases was systematically shorter than in heavier cases, holding other factors constant.
Effect of legal aid on delay
34. When the winning party had legal aid, the period of delay between instruction and issue was affected in most case types, with the exception of medical negligence, Official Referees' and judicial review cases. Whenever legal aid had an impact on delay, the effect was always that the period of delay between instruction and issue was independently lengthened (by about 50 per cent among personal injury and professional negligence cases; by about 60 per cent among Chancery cases; by about 70 per cent among breach of contract cases; and by about 75 per cent among Queen's Bench 'other' cases).
35. Among professional negligence, breach of contract and Queen's Bench 'other' cases, the increase may amount to little more than the time taken to apply for legal aid since those case types have relatively low median periods of delay (nine months, three months and four months respectively). Among personal injury cases, on the other hand, an estimated 50 per cent increase in delay in legal aid cases is likely to be in excess of the time taken to apply for legal aid, since the median period of delay in personal injury cases was found to be 15 months.
Effect of date of first instruction on delay: is delay increasing or decreasing?
36. Among all case types, with the exception of judicial review, there was clear evidence that the period of delay between date of first instruction and issue of proceedings has been declining in recent years. The extent of the decline was as much as 20 per cent or 25 per cent per year among some case types. This effect may be a reflection of change in litigation strategy, or the result of falling workloads among litigation lawyers.
Modelling of case duration between instruction and conclusion of claim
Effect of case weight/complexity on case duration
37. The various models suggest that the weight of cases significantly affects the total length of cases among only four of the case types (Official Referees', breach of contract, Chancery and Queen's Bench 'other' cases). Whenever case weight did affect duration, the effect was always to increase the total length of the case.
38. Among medical negligence, personal injury, professional negligence and judicial review cases there was no evidence that heavier cases systematically took longer than lighter cases, holding other factors constant, and the explanation for variations in case length among those case types must lie elsewhere.
Effect of legal aid on case duration
39. Among all case types, with the exception of judicial review, the fact that the winning party had legal aid significantly affected case length, holding other factors constant, and the effect was always that cases were longer when the successful party had legal aid. For example, the independent effect of legal aid on case length was a 20 per cent increase in the length of personal injury cases, a 25 per cent increase in the length of medical negligence cases and an increase of 45 per cent among breach of contract cases. Given the median case lengths found among these case types (Table 5), the period of increase appears to be well beyond the time taken to apply for legal aid.
40. There are a number of possible explanations for the apparent effect of legal aid on case length. First, the administrative time taken in applying and dealing with the Legal Aid Board; second, the possibility that legal aid work is given a low priority, resulting in cases not being rapidly progressed and hearing dates adjourned; third, the possibility that more work is being done on legal aid cases, or that legal aid cases are being run harder and further than might otherwise be the case, because there is no risk of costs for the party with legal aid.
Effect of date of conclusion on case duration: is case length increasing or decreasing?
41. The date of case conclusion appeared to affect case length among Official Referees', breach of contract, Chancery and Queen's Bench 'other' cases. Among each of these four case types, there was clear evidence that the length of case has been decreasing in recent years. The rate of decrease in overall case duration is about 8 per cent among Official Referees' cases, about 10 per cent among breach of contract and Queen's Bench 'other' cases, and about 13 per cent among Chancery cases. Thus, whenever a change in case duration over time was identified in the models, the change was always in a downward direction. In none of the case types modelled was there any evidence that the overall length of case has been increasing in recent years.
Modelling of costs
Effect of case weight/complexity on costs
42. The weight/complexity of cases significantly affected costs in every case type analysed. The effect of weight on costs was always to increase the level of costs, although the impact of increasing weight on costs varied in different case types.
Effect of case duration on costs
43. The total duration of cases, from date of first instruction to conclusion of case, was found significantly to affect costs in all case types, except medical negligence and judicial review. When duration has an independent impact on costs, the effect of increased case length is always increased costs, although the impact of long case duration varies between case types.
44. Among medical negligence cases and judicial review cases, there was no evidence that case duration had a significant, independent effect on costs, when other factors were held constant.
Effect of claim value on costs
45. The results of the analysis of costs indicates that the value of claims (in terms of the amount recovered) affects costs in personal injury, professional negligence, Official Referees', Chancery and Queen's Bench 'other' cases. In general, the effect is that, holding factors such as weight and duration constant, increasing claim value increases costs, although the pattern is not always consistent. For example, among personal injury and professional negligence cases, doubling claim value independently increases costs by about 15 per cent.
Effect of legal aid on costs
46. The costs models suggest that legal aid has a significant and independent effect on costs in personal injury cases and in judicial review cases. When the winning party has legal aid in personal injury cases, the costs of cases of the same weight, value and length will be about 10 per cent higher than if the winning party does not have legal aid, and this is in addition to the contribution of legal aid to costs in personal injury cases through increased case duration.
Effect of date of conclusion on costs: have costs been increasing or decreasing?
47. In only one of the case types analysed did the date of case conclusion affect costs. In medical negligence cases, there is clear evidence that costs are increasing at the rate of about 7 per cent per year, independent of case weight, case length or claim value.
48. In all other case types analysed, the date of conclusion has no significant impact on costs. This means that costs in all case types, other than medical negligence, have remained relatively stable in recent years, neither increasing nor decreasing. Since costs have not been adjusted to take account of inflation this means that there has probably been a decline in costs in real terms in recent years, in all case types other than medical negligence.
Medical Negligence
(a) Case type characteristics
| Mean | Median | ||
| ____________________________________________________________________________________________________________ | |||
| Costs allowed | £29,380 | £15,531 | |
| Claim value | £134,748 | £33,381 | |
| Delay between instruction and issue | 21 months | 20 months | |
| Duration | 65 months | 61 months | |
| Average total counsel fees | £7,028 | ||
| Average total expert fees | £4,413 | ||
| Average costs discovery + documents | £7,590 | ||
| % of successful parties with legal aid | 92% | ||
| ____________________________________________________________________________________________________________ | |||
(b) Results of statistical modelling
Delay between instruction and issue of proceedings: Case weight and legal aid were not found to have a significant effect on the period of delay between instruction and issue of proceedings in medical negligence cases. The period of delay has declined in recent years. The median period of delay has fallen from about 28 months among cases commenced in 1975, to 22 months among cases commenced in 1980, to 13 months among cases commenced in 1990.
Case duration: The only explanatory variable found to have a significant effect on total case duration among medical negligence claims was whether or not the winning party had legal aid. The effect is that cases in which the winning party had legal aid were about 25 per cent longer than those in which the winning party did not have legal aid. Lighter cases were not found to be systematically shorter than heavier cases, holding other factors constant. The length of medical negligence cases has remained stable in recent years, neither increasing nor decreasing.
Costs: Case weight and year of conclusion were the only variables included in the costs model found to have a significant effect on costs in medical negligence cases. The heaviest medical negligence cases are estimated to cost about 12 times more than the lightest cases, holding other factors constant. The level of costs in medical negligence cases has been increasing by about 7 per cent per year in recent years, holding other factors constant.
Personal Injury
(a) Case type characteristics
| Mean | Median | ||
| ____________________________________________________________________________________________________________ | |||
| Costs allowed | £19,382 | £12,134 | |
| Claim value | £98,434 | £50,000 | |
| Delay between instruction and issue | 17 months | 15 months | |
| Duration | 56 months | 54 months | |
| Average total counsel fees | £3,744 | ||
| Average total expert fees | £2,413 | ||
| Average costs discovery + documents | £5,016 | ||
| % of successful parties with legal aid | 58% | ||
| ____________________________________________________________________________________________________________ | |||
(b) Results of statistical modelling
Delay between instruction and issue of proceedings: Case weight, legal aid and the year of instruction were found to have a significant effect on the period of delay between instruction and issue of proceedings in personal injury cases. The period of delay has been decreasing in recent years, by about 15 per cent per year. The period of delay is longer among lighter cases than among heavier cases. Delay among weight A cases (heaviest) is about 50 per cent shorter than among weight E cases (lightest). If the winning party has legal aid, the period of delay between instruction and issue of proceedings among personal injury cases will be about 50 per cent longer than when the winning party does not have legal aid.
Case duration: The only explanatory variable found to have a significant effect on total case duration among personal injury cases was whether or not the winning party had legal aid. The effect is that cases in which the winning party had legal aid were about 20 per cent longer than those in which the winning party did not have legal aid. Lighter cases were not found to be systematically shorter than heavier cases, holding other factors constant. The length of personal injury cases has neither increased nor decreased in recent years.
Costs: Several of the explanatory variables were found to have a significant effect on costs in personal injury cases. Case weight affects costs, with costs increasing as case weight increases. Weight A cases (heaviest) cost about eight times as much as weight E cases (lightest). The value of claims also independently affects costs in personal injury cases. Doubling the value of a personal injury case will increase costs by about 15 per cent, holding other factors constant. The duration of personal injury cases also independently affects costs. Doubling the length of a personal injury case will increase costs by about 25 per cent, holding other factors constant. Legal aid also has an impact on costs in personal injury cases. In addition to the effect on costs through increased duration, when the winning party has legal aid, costs will be about 10 per cent higher, holding other factors constant. Finally, costs in personal injury cases have been neither increasing nor decreasing in recent years. If inflation is taken into account, this implies that there has been a decrease in costs in real terms.
Professional Negligence
(a) Case type characteristics
| Mean | Median | ||
| ____________________________________________________________________________________________________________ | |||
| Costs allowed | £32,866 | £14,834 | |
| Claim value | £107,849 | £34,000 | |
| Delay between instruction and issue | 12 months | 9 months | |
| Duration | 41 months | 35 months | |
| Average total counsel fees | £5,757 | ||
| Average total expert fees | £6,445 | ||
| Average costs discovery + documents | £11,320 | ||
| % of successful parties with legal aid | 38% | ||
| ____________________________________________________________________________________________________________ | |||
(b) Results of statistical modelling
Delay between instruction and issue of proceedings: The year of instruction, the weight of the case and the presence of legal aid all have a significant effect on the length of delay in professional negligence cases. There is unequivocal evidence that the length of the delay between instruction and issue, in professional negligence cases, has been declining in recent years, by about 10 per cent per year. However, when the winning party has legal aid, the period of delay between instruction and issue will be about 50 per cent longer than when the winning party does not have legal aid. The heaviest cases (weight A) have a period of delay between instruction and issue of proceedings which is about two-and-a-half times that of other cases (B/C/D/E).
Case duration: The only explanatory variable having a significant effect on case length was the presence of legal aid. When the successful party had legal aid, total duration was almost 60 per cent longer than when the successful party did not have legal aid, holding case weight constant. None of the other explanatory variables appeared to have any significant effect on case duration. There was no convincing evidence that weight or year of conclusion affects duration in professional negligence cases, suggesting that there has been no systematic increase or decrease in case duration in recent years, and that, holding other factors constant, lighter cases are not systematically shorter than heavier cases.
Costs: The main effects on costs are the weight of the case, the duration of the case and the value of the claim. There is a systematic increase in costs with increasing weight: weight A cases cost nearly 14 times as much as weight E cases, holding other factors constant. Duration also affects costs. If the total length of a professional negligence case doubles, costs will increase by about 20 per cent. The value of claims also affects costs; doubling claim value increases costs by about 15 per cent, irrespective of other factors. Although legal aid does not have an independent effect on costs, it has some impact through its effect on duration, which clearly affects costs in professional negligence cases. The year in which the case was concluded had no significant effect on costs. Thus there is no evidence that costs have been increasing in recent years, and given that costs have not been adjusted for inflation, it is possible that the apparent stability of costs levels represents a decline in real terms.
Official Referees' cases
(a) Case type characteristics
| Mean | Median | ||
| ____________________________________________________________________________________________________________ | |||
| Costs allowed | £35,844 | £19,320 | |
| Claim value | £112,633 | £34,764 | |
| Delay between instruction and issue | 8 months | 5 months | |
| Duration | 34 months | 30 months | |
| Average total counsel fees | £5,396 | ||
| Average total expert fees | £9,653 | ||
| Average costs discovery + documents | £9,956 | ||
| % of successful parties with legal aid | 18% | ||
| ____________________________________________________________________________________________________________ | |||
(b) Results of statistical modelling
Delay between instruction and issue of proceedings: The only explanatory variables that have a significant effect on the period of delay are the year of first instruction and the weight of the case. Delay does not appear to be affected by whether or not the winning party had legal aid (18 per cent of winning parties had legal aid). The period of delay in Official Referees' cases has been declining by almost 20 per cent per year. Heavier cases (categories A/B/C) have a period of delay between instruction and issue of proceedings which is about 50 per cent longer than that of lighter cases (categories D/E).
Case duration: The total duration of cases is affected by the weight of cases, the presence of legal aid and the date on which the case was concluded. Increasing weight has quite a substantial effect; the heaviest cases with weight A/B take twice as long as weight E cases; weight C cases take about one-and-a-half times as long as weight E cases; and weight D cases take about 20 per cent longer than weight E cases. Legal aid also affects duration. Cases in which the successful party had legal aid lasted about 90 per cent longer than non legal aid cases. The length of cases has been decreasing by about 8 per cent per year.
Costs: Costs are affected by the weight of the case, the value of the case and the duration of the case. There is also an interaction effect, in that the effect of weight on costs varies according to the value of the claim. Among cases of weight D and E the value of the claim has no impact on costs. Among more weighty cases, however, costs do appear to increase as claim value rises. Doubling the value of a weight B case will increase costs by about 30 per cent, irrespective of the length of the case. Increasing duration also leads to increased costs. Doubling the length of cases increases costs by about 20 per cent. There was no evidence that legal aid had any independent impact on costs, although it has an influence through its effect on case duration, which affects costs. There is no evidence to suggest that costs in Official Referees' cases have been increasing in recent years. Given that costs have not been adjusted for inflation, it is possible that the apparent stability of costs levels represents a decline in real terms.