Note: the original hard copy of this report is 31 pages. Would you like to order a hard copy?

To search for keywords use Edit>Find

cover image

Model key performance indicators for NSW Courts  

, 2000 This report represents the results of a project to develop model key performance indicators for New South Wales Courts. The project was a joint undertaking between the New South Wales Attorney General`s Department and the Justice Research Centre. Its aim was to improve the management information used to help the courts do their work effectively.

Ch 1. Introduction


This report presents the results of a project to develop model ‘key performance indicators’ for New South Wales Courts. The project was a joint undertaking between the New South Wales Attorney General’s Department and the Justice Research Centre. Its aim was to improve the management information used to help the Courts do their work effectively.

Despite an interest in management statistics which can be traced back to the earliest introduction of ‘caseflow management’ reforms (largely imported from the United States),1 and the more recent extension of modern public sector management requirements of ‘accountability’ to their specialised sphere,2 Australian Courts have few, really good models of management information systems available to them. There is, in fact, a familiar litany of complaints about the deficiencies of court statistics. In a very recent assessment, the Audit Office of New South Wales3 described ‘the management information systems currently supporting managers in the courts system [as] relatively inflexible, limited in content and not well integrated.’

A complete management information system has to be designed, of course, to address a hierarchy of different needs. The ‘Model KPIs’ were designed with a narrower but nevetheless ambitious (and elusive) purpose in mind. This was to produce, using the fewest statistics possible, the simplest, clearest, most comprehensive picture possible of how well the Courts were performing, in terms relevant to the operational needs of their managers but also addressed to the interests of a wider public. Thus conceived, it is hoped that the Model KPIs can form something like the ‘keystone’ of a fully adequate management information system for Courts.

Basic Principles

The design of the Model KPIs has been influenced by four basic axioms (or, possibly, prejudices of the authors) about Court performance measurement, which should be expanded upon briefly before turning to the details of the Model.

Performance has to be measured against goals fixed by the Courts

The first among these is that, of course, performance indicators should relate to the ‘performance goals’ of a Court.

A few important observations should be made in relation to this point. Not all Courts presently have explicitly stated aims. However, it is clear that all Courts are concerned, broadly speaking, to serve the public by resolving disputes according to law, when required, through processes which are fair, expeditious and cost-effective.4 The Model KPIs focus on timeliness and cost-effectiveness — or what might be described collectively as the ‘case management’ or ‘processing’ goals of the Court system.

This focus is not intended to imply that wider goals of fair process or just outcomes are somehow unimportant. To the contrary, they are givens and, if management jargon can be used without trivialising the point, ‘performance’ in relation to these goals is assured by entrenched structural features of our court system and its procedural regimens. Everyone will immediately see the sense in trying to reduce timeliness and cost to simple performance goals and measures; it is a nonsense to try to do the same with the goals of fair process and just outcomes.

The Courts should set goals for themselves, in measurable terms

Courts should, in fact, adopt precise (that is, measurable) ‘performance standards’ in relation to the general goals of ‘timeliness’ and ‘cost-effectiveness’.

This has a significant benefit for the Courts. Performance standards or benchmarks overcome the frustration and sensitivity of Courts used to dealing with less sophisticated measures of performance. The most familiar example is perhaps the use of ‘time taken’ as a measure of ‘delay’ — this completely fails to take account of the fact that the process must take time and, even, in some cases may need to take a lot of time. A time standard provides a rational, defensible basis for distinguishing between time necessarily taken by the process, and unreasonable delay. Similar observations can be made about cost-effectiveness.

Performance measurement should support management activity

Key performance indicators should encourage Courts to manage to meet their performance goals.

This means, among other things, that performance indicators should, as much as possible, be directed to the Court’s present and future, rather than its past. They should be related to simple operational targets or benchmarks of satisfactory progress. They should also require no, or minimal, interpretation or synthesis; if things are not going well, ideally they will indicate, in general terms, what sort of corrective action is required.

‘Key’ means comprehensive but simple and few

Key performance indicators should be comprehensive but also simple and as few in number as possible.

This essentially repeats the central idea introduced above but it bears repeating. It would only be a little unkind to observe that Australian Court statistics seem to have followed the climatic cycle of drought followed by flood. After years of keeping only the most rudimentary data on ‘ins’ and ‘outs’, and ‘pending’ case numbers, some Courts now produce whole books of almost bewildering tables and charts. When presented in well-designed reports, this detail has an important place in the management of the day-to-day operations of Courts. However, the aim of a key performance report should be to provide the maximum amount of useful information with a minimum amount of data, in an easily absorbed form. The ultimate message should be simple: ‘All is well’ or ‘All is not well.’

A discussion paper presenting an earlier draft of the Model KPIs was circulated in March 1999 to the chief judicial and administrative officers of each of the New South Wales Courts, for their advice. Several very valuable comments were received, and the model has been revised in light of them. In this report we have endeavoured to reflect fully all of the comments received, and to indicate how we have responded to them.



Ch 2. Overview and general observations


The model key performance report is based on four simple measures —

Backlog is related directly to the Court's performance against its case processing time standards. It is the number of pending cases which are taking too long. The backlog measure indicates whether the Court is meeting its time standards.

Overload relates the size of the Court's caseload to its time standards. It is the number of cases on hand in excess of the number the Court can be expected to process within time. The overload measure indicates whether the Court will continue to meet its time standards in the future (assuming existing conditions are maintained).

Clearance Ratio relates the Court's caseload to its capacity. It is the ratio of the Court's new registrations to the number of finalisations over the relevant reporting period. The clearance ratio indicates whether the Court is heading for, keeping out of or getting out of trouble, in terms of its capacity to meet its time standards in the future.

Attendance Index relates to the efficiency and effectiveness of the Court's processes. The attendance index requires the Court to adopt a standard for the maximum number of times it should be necessary for the parties (or their representatives) to attend at the Court before their case is resolved. The measure itself is the number of pending cases in which there has been more than the benchmark number of attendances. The number of 'trips to the Courthouse' is the most easily obtained datum which is highly correlated with the cost of litigation and efficient resource utilisation.

Figures 1 and 2 show two somewhat different mock-ups of a 'monthly report'. Although the Model KPIs are intended to standardise key performance measures and the formats in which they are presented, the performance standards used by the Courts is a matter for them. Figure 1 illustrates the format of a Model KPI report in its most 'generic' form, and shows the extent to which individual Courts can adapt the format to their own circumstances.5

However, the generic format is necessarily 'filled' with blank spaces, and the formulae given later in the text for calculating each of the measures use symbols rather than actual values, which together may make Figure 1 a somewhat enigmatic explanatory aid for readers who prefer their examples a bit more concrete. So we have imagined a Court which deals with about 12,000 civil cases a year, and has adopted the following case processing standards: 90% of its civil cases should be finalised within 12 months of commencement and all cases should be resolved within two years; 90% of cases should be resolved after no more than three attendances at Court, and no case should require more than five attendances to be concluded. Figure 2 shows what the monthly report of this Court might look like. As described, this hypothetical Court resembles the District Court, but only in part, and the concrete values set out in the report are plausible but made up. The liberties taken in the example stop, then, with naming it the 'Intermediate Court'.

After studying the formulae for the KPIs, given later in the text, interested readers may wish to check the calculations in Figure 2. The additional information required to do this is as follows: We imagined that the Intermediate Court had a pending civil caseload at the end of June 1999 of 16,996. The number older than 12 months was 4,052 (including the 1,934 cases older than two years). The number in which there had been 3 or more attendances was 3,425 (including the 1,569 cases in which there had been more than 5). The number of new registrations in June was 1,100 and the number of finalisations was 1,196. The number of finalisations in the preceding 12 months was 14,376. In the preceding 6 months the number of finalisations was 7,192 and in the same 6 month period a year before it was 6,597 cases. The number of new registrations in the preceding 6 months was 6,614 and in the same 6 month period a year before it was 6,589.

It is envisaged, as Figure 1 and Figure 2 imply, that Courts would publish monthly reports. Before turning to a discussion of the particular measures, a few further general observations about the format of the Model KPIs are in order.

Figure 1: Monthly KPI Report — Generic


Key Performance Indicators

[Month Year]

[specify] Court

[specify e.g. Civil or Criminal] Cases


1 Backlog



2 Overload = [ ] cases or [ ]% (should be 0)

3 Clearance Ratio = [ ]% (should be 100%)



4 Attendance Index




Figure 2: Monthly KPI Report — 'Intermediate Court' (fictional)

Key Performance Indicators

June 1999

'Intermediate' Court

Civil Cases


1 Backlog



2 Overload = 1182 cases or 7.5% (should be 0)

3 Clearance Ratio = 109% (should be 100%)



4 Attendance Index


Reporting for different classes of cases


It is also envisaged that a separate report would be produced by each Court for each major class of business for which the Court has established separate performance goals. Court managers will, presumably, wish to monitor performance at much finer levels of operational detail — for example by division or 'list', and by registry — but there is no need to report this detail beyond the precincts of the Court itself. (Indeed, detail of this kind will be meaningless and obfuscating to readers not involved in the Court's day-to-day management.)

It appeared from the consultations that some Courts were quite concerned about significant differences between types of cases in their caseloads, and these Courts may contemplate addressing the differences by establishing different performance standards in relation to them. Ultimately this is a matter for each Court to resolve for itself. However, setting different performance goals for different categories of business does present some difficulty, and it therefore seems advisable to add a few cautionary words specifically on this issue.

It has been a long-standing convention in New South Wales Courts, as in Australian Courts generally, that it is appropriate to apply different performance standards to criminal and civil caseloads. It is then appropriate that Courts having both jurisdictions should produce separate performance reports for each. Otherwise, however, there are important reasons why Courts should be extremely reluctant to differentiate among categories of work and 'segment' their public performance reporting in this way.

The Model KPIs were designed to provide Courts with the means to produce a simple, clear and comprehensive picture of how they are performing. This aim will be diminished by reporting for multiple categories of cases. But more importantly, having different performance standards for different classes of business raises an equity issue. Different performance standards are apt to affect resource priorities, and in this sense they can be discriminatory — greater resources are likely to be given to the class of cases with the most demanding standards than to others. This kind of discrimination may well be justified6 — the important point of principle is that it must be, whenever the Court contemplates applying different standards of performance to different classes of cases.



Choosing performance standards


The illustrated reports provided in Figure 1 and Figure 2 both use examples of performance standards which are two-tiered or 'split'. This is not intended to be a prescription — there is no particular reason why a Court should not choose simply to set a single standard, for either the time or the number of trips to the Court it should take to complete a case, which relates to all cases (i.e. '100% within 18 months' or 'no case should require more than 3 attendances'). However, two-tiered standards (such as the example of '90% within 12 months and all within 2 years') are commonly adopted by Courts and are an extremely sensible way of acknowledging the existence of 'ordinary' and 'exceptional cases. In technical parlance, the standard set for the 'ordinary' group or 'most' cases is described as the 'norm' and the standard which sets the outside limit for 'exceptions' to the norm is described as the 'boundary'.

When a split standard is adopted, the Court will need to report performance against both the norm and boundary performance standards.

Examples can be given of Courts which have set time standards which do not apply to all of the Court's caseload, as in '98% of cases within 18 months'. The effect of this is to create a class of the Court's business (the extreme 2% in this example) which is not subject to any performance standard and this cannot be regarded as good management practice. The view adopted here is that, whether single or tiered, the performance standards adopted must establish a true boundary (i.e. apply to 100% of a Court's business).

A fair amount of concern about setting appropriate performance standards was expressed over the course of the consultations. Some of this concern may reflect undue or misplaced concern about the criticism a Court may face if it fails to meet its own performance standards. Performance standards are goals. They should be something to aspire to. Setting them will involve an assessment of what the Court's processes require, and it is both appropriate and sensible that these assessments should be informed by the Court's experience. Ultimately, however, the Court must make a value judgement about what is reasonable.



Ch 3. Backlog


The backlog indicator is a direct measure of the Court’s performance against its case processing time standards. However it translates this performance into simple data concerning the Court’s pending caseload. This is extremely significant from a management point of view. There is an argument for saying that, if a Court has adopted a time standard, like ‘all cases should be finalised with 18 months of commencement’, then the key performance indicator should be constructed in exactly the same terms as the goal — that is, the report should indicate the percentage of cases finalised in the relevant period within the time standard. The problem with this view is that the key performance indicator is then historical, and tells the Court that it met its time standard. The Backlog figure tells the Court if it is meeting its time standard. Moreover, the message is in simple terms: ‘The Court has n cases on hand which are “past their use-by date”.’ (‘n’ is a standard symbol used in mathematics to signify a variable which is a whole number.) Particular attention can be given to those cases.

Both of the model reports illustrated in Figure 1 and Figure 2 use case processing time standards which are two-tiered or ‘split’ as, in the example of the Intermediate Court (Figure 2), ‘90% of all cases should be finalised within 12 months of commencement, and all cases should be finalised within 2 years.’ As already noted, this is not an essential feature of the model (and individual Courts might choose to adopt standards in a different form), but two-tiered standards like this are common and are an extremely sensible way of acknowledging the existence of ‘ordinary’ and ‘exceptional’ by defining a norm for ‘most’ and establishing a ‘boundary’ around the exceptions. Where a split standard is adopted, then two backlog figures are reported.

In accordance with accepted definition, the Backlog is the number of ‘old’ cases on hand in excess of the number allowed under the time standards. As a consequence, the formula for calculating this number contains a factor which adjusts the number older than the norm by the proportion allowed under the standard. The general formulae for calculating the Backlog figures then are —


The calculation using the more ‘concrete’ example of the Intermediate Court (Figure 2) is as follows — In light of the consultations on the Discussion Paper, it is recommended that the Backlog indicator should also report the proportion of the pending caseload older than the relevant time periods adopted under the standards. Including proportions as well as simple numbers makes the statistics more meaningful and informative. It also makes the time standards adopted by the Court more transparent. It should be noted that the proportion of the caseload older than the normative time standard is not the same as the Backlog(>norm) — only the proportion (if any) greater than the proportion allowed by the standard7 is properly characterised as backlog. This point is, however, made evident by the format of the Report.


Ch 4. Overload


The overload indicator relates the size of the Court’s caseload to its capacity to meet its time standards. A Court may be meeting its time standards (that is, have no backlog) but nevertheless be headed for trouble. The overload indicator tells the Court whether it has more cases, and what number more, than the number it can finalise within its time standards (assuming existing conditions are maintained). In short, it can provide an early warning that a Court may not continue to meet its performance standards, and it indicates how the Court will need to alter its finalisation rate if it is to avoid getting into, or needs to get out of trouble. The message is, again, simple: ‘the Court has n more cases on hand than it can handle within the time it has set for itself.’

The general formula for calculating this number is as follows —

Overload = Ptotal - (T * F), where Ptotal is the number of pending cases, F is the average number of cases the Court finalises each month and T is the maximum average time which can be taken by all cases if the Court is still to meet its time standards.

Essentially, the Overload formula calculates whether the Court has the capacity to turnover its total pending caseload within the range of times defined by its time standards — assuming it continues to perform at its current level. The fundamental measure of the Court’s capacity used in the formula is F, average monthly finalisations. This is, of course, a historical measure — but also one chosen as a future estimate. Although finalisation rates are typically more stable than new registrations, it is nevertheless usual that there are significant month-to-month fluctuations, particularly at some times of the year and in the higher courts. No hard and fast directions can be given about the period over which the average monthly finalisation rate should be calculated, beyond observing that a fixed period should be settled upon which will tend to ‘smooth out’ month-to-month fluctuations and produce a reliable long-term indication of the finalisation rate. As a rule of thumb, it can be suggested that the average could be calculated for the immediately preceding period of months equal to — and certainly no greater than — the normative time standard. This, however, might be safely shortened in a Court with no history of significant month-to-month fluctuations in finalisation rates.

Taking the example of the Intermediate Court (Figure 2) the formula for calculating Overload becomes —

Overload = Ptotal - (13.2 * F), where Ptotal is the number of pending cases and F is the average monthly rate of finalisations for the preceding 12 months (total finalisations in the 12 month period divided by 12).

The apparently mysterious 13.2, the value of T in this example (the maximum average time which can be taken by all cases if the Court is still to meet its time standards), is simply (0.9 * 12) + (0.1 * 24). It is possible to give a complex mathematical derivation of this formula but, for the less mathematically inclined, it can be understood intuitively by explaining that the maximum average time will be taken if the 90% of cases which must be finalised within 12 months are finished in exactly 12 months, and the remaining 10% are finished in exactly 24 months (2 years).

Overload can also be expressed as a percentage of the Court’s capacity to handle its caseload within time standards, by modifying the general formula given above only slightly —

Overload(%) = 100 * [Ptotal /(T * F)] - 100

The result delivers the basic message in altered terms: ‘the Court’s caseload is larger than the number it can handle within its time standards by y%’. (The calculation may yield a negative number, which actually indicates that the Court is ‘underloaded’ or has surplus capacity.) This statistic has been added since the Discussion Paper as a recommended inclusion in the Model Report because it enriches the amount of information and is arguably a more meaningful statistic for readers of the Report who are not very familiar with the circumstances of a particular court.



Ch 5. Clearance ratio


The Clearance Ratio is very similar to the Overload indicator but, instead of comparing the Court’s pending caseload to its capacity, it relates the incoming volume with the Court’s capacity to finalise, or ‘clear’ its cases. It is a slightly more useful monitor for the Court which is staying out of trouble, while overload is primarily a measure of how much trouble a Court may be in.

Most Courts maintain monthly and annual statistics on new registrations and finalisations. The Clearance Ratio is a simple way of analysing the trends in this data and synthesising the essential message: ‘The Court is (or is not) keeping up with its workload.'

The Clearance Ratio is the number of finalisations for a chosen reporting period, divided by the number of receipts or registrations in the same period, multiplied by 100 (to convert the number to a percentage). A figure of 100% indicates that the Court is keeping up with its workload; more means that it is reducing its pending caseload while a figure of less than 100% means the Court is accumulating cases.

The running period over which the new registrations and finalisation figures are calculated is to some extent a matter of choice (as in the case of the Overload calculation), although if it is too short the figures will be influenced by episodic fluctuations which in fact have no significance. The utility of the Clearance Ratio as a ‘staying out of trouble’ monitor can be enhanced (and the information redundancy with the Overload indicator reduced) by choosing a reporting period which is shorter. This has the effect of increasing the measure’s sensitivity to changes which may portend trouble, although as with the calculation of Overload, the trick is to avoid making the measure sensitive to meaningless short-term fluctuations in the Court’s caseload or finalisation rate. As a point of departure, either a quarter or half year are suggested, although this will depend again on the time standards adopted by the Court and the month-to-month variations in its volumes

The Clearance Ratio by itself is not completely unambiguous: If it is less than 100% it means that the Court is getting through less work than is coming in, but this might be because more work is coming in, or the Court is finalising fewer cases, or a combination of both. Detailed trend information on receipts and finalisations will resolve these ambiguities but requires a high level of synthesis from the reader. The model report provides the same data in already synthesised forms: These are the ‘caseload’ factor and the ‘finalisations’ factor which are reported only as ‘up’, ‘down’ or ‘unchanged’.

The caseload factor is calculated by comparing the running total for new cases (of the relevant case type) in the reporting period with the running total for the same period the year before. If the current period is ‘up’ on the same period in the year previous, this means, rather obviously, that the Court’s workload has increased. Similarly, the finalisations factor is calculated by comparing the running total finalisations with the running total for the same period the year before. If the current period is ‘down’ on the previous year, this means that the Court’s productivity has decreased in relative terms. (This, in turn, could be due to a number of factors, such as increasing average trial lengths, fewer settlements or fewer available judges, which can be the subject of more detailed analysis by the Court’s managers.)

In the Discussion Paper it was proposed that the ‘caseload’ and ‘productivity’ (now ‘finalisations’) indices should be indexed to the total number of judges, because this would incorporate an adjustment for resource changes in the report. In the course of the consultations it was pointed out that some Courts make extensive use of part-time appointees and would therefore have practical difficulties in indexing caseloads and finalisations to available judicial resources. One solution was to translate actual usage of part-time appointees into an ‘equivalent full-time’ measure but on reflection it seemed that by far the easiest solution is not to index for available resources at all. This greatly simplifies the calculation, while also ameliorating some sensitivity aroused by the productivity measure. If there is a notable change in the finalisation rate from one year to the next, the role played by available resources is simply one of the possible explanations the Court’s managers will wish to investigate.



Ch 6. Attendance index


It is probably fair to say that performance measurement related to workloads and timeliness is comparatively familiar and ‘comfortable’ territory for most Australian Courts. This is not true of measures of cost and efficiency. Despite the fact that virtually every known Court believes that the cost of litigation should be as low as possible, and the use of Court resources should be as efficient as possible, no known Court monitors the cost-effectiveness of its process. There are perhaps many reasons why this is so, but no doubt one must be that the data problems are practically insurmountable, if what is contemplated are direct measures of cost efficiency.

All Courts, however, do keep diaries of the occasions when the parties or their representatives are required to attend a hearing or other official ‘appointment’ at the Court,8 and this information can be used as an indirect measure of cost efficiency. The number of times the parties (or their representatives) attend at Court is probably the single, most easily captured datum which is highly correlated with public (i.e. Court) and private (i.e. litigant) cost.9 Moreover, well-established and widely accepted case management principles emphasise the goals of achieving early settlement of cases, and of ensuring that the parties and the Court seriously regard each scheduled hearing as an opportunity for disposing of the dispute. The Attendance Index10 is a directly relevant measure of performance in these terms.

The Index is the number of cases in the Court’s inventory in which there has been a given number or more of occasions — including any ‘appointment’ (however styled) which is adjourned or re-scheduled — requiring the parties’ attendance at the Court. The inclusion of hearings scheduled but adjourned adds an extra degree of refinement to the measure, since adjournments are known to be correlated with efficient resource utilisation.11 The message is, once again, a simple one: ‘n of the Court’s pending cases have been to Court more times than they should have, without a resolution.’ Management attention can then be directed to the problem cases.

The Attendance Index is the cost measure parallel of the ‘timeliness’ index provided by the Backlog measure. As with measures of timeliness, it is necessary for a Court to adopt some standards for the number of times it should be necessary for the parties (or their representatives) to attend at the Court before their case is resolved. The standards adopted will depend on an appraisal of process requirements, experience, and, most of all, a normative assessment of what constitutes ‘acceptable’ performance.

To illustrate, we can imagine that the Intermediate Court of Figure 2 has a case management protocol for civil cases contemplating that cases can be dealt with after three hearings (an early ‘status conference’, a ‘directions hearing’ at or near the completion of all interlocutory steps, and an arbitration hearing or, alternatively, a trial). This protocol, then, suggests a starting point for fixing a norm — what can be, should be. While, however, the formal process suggests that most cases should be finalised after, at most, three hearings, experience and reason may indicate that some proportion of the Court’s cases should be expected to involve a greater number of attendances than the norm. The performance standard might then be set accordingly, as in ‘90% of cases should require no more than 3 attendances before finalisation (the norm), and 100% should be finalised after (say) 5 hearings (the boundary).’

The Attendance Index then, using this example, is the number of cases in the Court’s pending inventory which have had three or more, and five or more, appointments scheduled or held. (Once again it should be noted that the count includes adjournments, if they have required an attendance, because the Index is concerned with efficiency.)

In the Discussion Paper the example given was based on a single, rather than a ‘two-tiered’ or ‘split’ standard, and it was suggested that the number of cases exceeding the standard should be reported as a percentage of the Court’s pending caseload. It is apparent from the consultations that the circumstances of most Courts will be better provided for by two-tiered standards. It is also recommended that, again like Backlog, the Attendance Index should be reported in terms of both simple numbers and proportions, as this both enriches the information provided and makes the Court’s standards more transparent. The calculations are quite simple —



Ch 7. General results of the consultations


As already noted above, an earlier draft of the Model KPIs was circulated in March 1999 to the chief judicial and administrative officers of each of the New South Wales Courts, for their advice. Several very valuable comments were received, and the improvements made to the model as a consequence have been noted at several points in the preceding discussion.

At a more general level, most of the responses received were supportive. Some concerns were expressed that related fundamentally to issues about performance standards and measurement. These call for a few comments.

While it was acknowledged, for example, that the suggested Backlog and Overload measures were recognised and useful indicators, it was noted their use required the Courts to establish time standards for case processing. At this point several New South Wales Courts do not have these in place. Several responses raised concerns about fixing appropriate time standards.

It does need to be emphasised, albeit in a somewhat uncompromising vein, that modern principles of court administration dictate that the basic case processing time standards adopted by courts should relate to the time taken from initiation (rather than some defined later point in proceedings) to finalisation (including the time taken to deliver judgment where judgment is reserved). The case management schemes administered by the Courts need not necessarily — but nowadays generally do — reflect acceptance by the Courts that they have a responsibility for managing the progress of their cases ‘from go to whoa’.

Some of the concern expressed about appropriate time standards perhaps reflects undue or misplaced concern about the criticism a Court may face if it fails to meet its own performance standards. Performance indicators are about encouraging better management, not assessing blame. Time standards are goals and should be based on informed value judgments about how long it should take to finalise cases. Experience indicating how long it does take may be useful as a reference point, but obviously unsatisfactory performance should not be elevated to benchmark standards for the sake of avoiding ‘a bad report’.

A similar response must be given to concerns expressed about the need to recognise that a Court’s performance may be affected by a whole range of factors that it cannot control. Quite so. The first — and generally only — consequence for a Court that is not meeting its performance standards is that it should investigate and explain. If its performance is being affected by factors beyond its control, it is important that this should be explained. If there is something the Court can do about its predicament, surely it wants to know that too.

The Attendance Index (termed the 'Hearings Index' in the Discussion Paper) was the most novel of the proposed Model KPIs and, it was evident from the consultations, gave rise to the most concern.

Several Courts noted that their information systems did not capture the data required to produce this measure. (Indeed, a number suggested that they might have difficulty accessing all of the data required for other measures on their information systems.) This it must be acknowledged is a serious practical problem, but one which can, however, be addressed by upgrading information systems, assuming a commitment is made to adopting the Attendance Index as a performance measure.

There was some concern about the difficulty of arriving at an appropriate benchmark, and this concern was accentuated by a perception that the particular characteristics of individual or classes of cases result in considerable variability in the number of hearings required. These concerns strongly parallel the concerns expressed about time standards, before these became a familiar feature of court administration, and in many respects the answers to them are the same.

A time standard like ‘90% of cases within 12 months and 100% within two years’ effectively defines an acceptable bracket around the ‘tail’ of the distribution of finalisation times for all cases. In principle, exactly the same can be done for the number of attendances at court involved in cases before they are finalised. The fact that many cases might settle after only one hearing, or even none, while others might involve 10 including a trial, should not be seen as an insuperable barrier to setting standards.

The position is no different in relation to finalisation times — some cases may resolve themselves in only a few days or weeks, and some may take years. Just as it is possible to set a norm for most cases (e.g. ‘12 months for 90%’) and a boundary on what can be considered acceptable even for the exceptional cases (e.g. ‘and in no event more than two years’) it is surely possible to choose the number of ‘trips to the Courthouse’ within which most cases ought to be resolved, and an ‘outside’ standard beyond which, even allowing for the ‘odd’ cases, one would say “That’s too many!”.

As with time standards, the actual standards set should be informed by an assessment of what the Court’s process requires, a critical evaluation of experience and, ultimately, a value judgement about how many times it should be necessary for reasonably earnest claimants to attend the Court to get a resolution of their disputes.

Finally, a few of the comments received appeared to accept the idea of an Attendance Index but objected to the suggestion that hearings which required the parties to attend, but were adjourned, should be included in the count. This objection is based on a view that, as adjournments are ‘caused by the parties and not by the Court’ they are not ‘an indicator of the Court’s performance’. (The same objection was not raised in relation to hearings which are re-scheduled when cases are ‘not reached’ by the Court.)

Of course there are many circumstances — some beyond the control of the Court, and indeed the parties — where an adjournment should be granted in the interests of justice, and it would be improper to refuse. Concern, however, over monitoring the frequency of adjournments may be unduly defensive. Nor does it accord with ‘best practice’ in modern court management.

The same point about ‘not our fault’ could be made (and was made) about delay and time standards. The answer is that performance measurement is not about blame. It does, however, involve acceptance by the Court of some management responsibility for the efficiency of its processes. And it is, in fact, well-accepted that the adjournment rate is something which the Court can ‘control’ through efficient management — Courts which experience high adjournment rates typically have more general problems of listing integrity and efficient resource utilisation.

Adjournments will be necessary and will occur. But it is appropriate to establish some sort of benchmark for an acceptable level of occurrence. This can be incorporated into the standard established by the Attendance Index.

It may well be that some of the concern about the Attendance Index should be attributed to its novelty and unfamiliarity. In that event, a very good suggestion received for handling the data collection problem in the short-term (pending IT system improvements) may also address this concern. This suggestion is that Courts could report monthly on the frequency distribution of the number of ‘appointments’ in cases finalised during the month. This data would be comparatively easy to collect manually, although in ‘high volume’ jurisdictions it would perhaps be necessary to base the statistics on a sample rather than the whole population of finalised cases. The additional advantage of this approach is that it does not require reporting against a ‘performance standard’ and it would allow the Courts to develop a ‘feel’ for what is actually going on in their cases, before resolving on an appropriate standard.



The Justice Research Centre


The Justice Research Centre supports the development of legal policy and effective justice system reform with high quality research. The impetus for establishing the Centre was that in the past much law reform debate was based on impressions and anecdotes rather than systematic and objective research data and analysis.

The Centre is recognised nationally and internationally for influential research which is characterised by a sophisticated policy perspective, combined with an empirical approach to issues and rigorous standards of quality, objectivity and independence.

The research program

Our research examines all aspects of the justice system including the policies that shape it, the behaviour of the people who participate in it, the operation of its institutions and processes, and its social and economic effects on the community.

The Justice Research Centre's research program is broadly focussed, across four major areas —


The Centre's research projects are addressed to significant — often controversial — justice issues, where our results make a substantial contribution to policy development and law reform.

How we operate

The Justice Research Centre is an independent public interest research organisation. Its research is influential because of its quality and objectivity.

Core funding is received from the Law Foundation of New South Wales, but the Centre is increasingly dependent on support from other sources. The Centre actively seeks financial support from justice system 'stakeholders' to whom our work has value because it improves policy development and law reform outcomes. The Centre has received grants and donations from the Federal Government and State Governments and from private sector organisations

The Centre's research projects are chosen in consultation with its Advisory Board, which consists of distinguished representatives of the judiciary, the legal profession, government, industry, and the research community. In selecting appropriate projects, great emphasis is placed on the likelihood that the results of research will have significant practical implications for important policy issues.

We welcome project ideas from others. However, the Centre insists on autonomy over the scope and design of projects and complete discretion over publication of results.

Advisory Board

The Hon Justice Ronald Sackville
Federal Court of Australia

The Hon Justice Margaret Beazely
New South Wales Court of Appeal

David Bowen
General Manager, Motor Accidents Authority of NSW

The Hon Justice Sally Brown
Family Court of Australia

Dr Janet Chan
School of Social Science & Policy, University of New South Wales

Professor Hilary Charlesworth
Director, Centre for International and Public Law, Australian National University

David Fairlie
Partner, Mallesons Stephen Jaques

Professor Marcia Neave AO
Faculty of Law, Monash University

Clare Petre
Energy Industry Ombudsman NSW

Bret Walker SC
Governor, Law Foundation of NSW and Past President, Law Council of Australia

Dr Don Weatherburn PSM
Director, NSW Bureau of Crime Statistics and Research

Staff

Director

Professor Ted Wright BSc LLB MA
Barrister & Solicitor

Research

April Chrzanowski BAppSc LLB (Hons)

Naomi Doncaster BA MSc

Ann Eyland MSc PhD

Ann Genovese BA (Hons) LLB PhD

Shelley Hampton BA (Hons) LLB

Maria Karras BSc (Hons)

Angela Melville BA BSocSc (Hons)

Carolyn Morris

Natalina Nheu BSc (Hons) LLB

Caroline Sage BA (Hons) LLB (Hons)

Administration

Sue Nicol

Research Associates

Dr Janel Chan
Associate Professor and Head School of Social Science & Policy, University of New South Wales

Professor Michael Chesterman
Faculty of Law, University of New South Wales

Professor John Dewar
Dean, Faculty of Law, Griffith University

Ms Beth Gaze
Faculty of Law, Monash University

Professor Rosemary Hunter
Director, Socio Legal Research Centre, Griffith University

Professor Stephen Parker
Dean, Faculty of Law, Monash University



Publication details


The Authors

Laurie Glanfield is the Director-General of the New South Wales Attorney-General's Department.

Ted Wright is the Director of the Justice Research Centre, Law Foundation of New South Wales, and Belle Wiese Professor of Legal Ethics, University of Newcastle.

ISBN 0 909136 70 X

© NSW Attorney General's Department in right of the Crown February 2000

Published by the Law Foundation of New South Wales.

Any opinions expressed in this publication are the author’s and do not necessarily reflect the views of the Law Foundation’s Board of Governors, the Justice Research Centre’s Advisory Board, or the NSW Attorney-General’s Department.

This publication is copyright. It may be reproduced in part or in whole for educational purposes as long as proper credit is given to the Justice Research Centre, the copyright holder and the authors.




Ch 1. Introduction
2  cf. Steering Committee for the Review of Commonwealth / State Service Provision Report on Government Services 1998, vol 1. See also R Mohr, H Gamble, T Wright and B Condie, Performance Measurement for Australian Courts (1997) 6 Journal of Judicial Administration 156.
3  The Audit Office of NSW, Performance Audit Report: Management of Court Waiting Times (July 1999) at p. 58.

Ch 2. Overview and general observations
5  Even in this endeavour, Figure 1 implies some constraints which are not intended. It assumes, for example, that a Court will have two-tiered or split performance standards for both case processing time and number of hearings. This is not a prescription. See further Choosing performance standards below.
6  Good examples which come to mind are the specialist commercial lists of some Superior Courts in Australia and overseas, and those Courts which set shorter time standards for criminal cases in which the accused are on remand, as opposed to having bail.

Ch 3. Backlog
7  the figure represented by x in the generic example illustrated in Figure 1, or set at 10% in the Intermediate Court example given in Figure 2.

Ch 4. Overload

Ch 5. Clearance ratio

Ch 6. Attendance index
8  Although this information is not always maintained in a form which makes it easy to access or use for present purposesbut this is another issue.
9  cf T Wright, A Eyland and J Cox, Claiming under the Motor Accidents Scheme (Justice Research Centre 1998).
10  In the discussion paper this was called the hearings index but following the consultations it was decided that the normal connotations of hearing were too narrow and potentially confusing, for reasons which may now be apparent in the text which follows.
11  M Solomon and D Somerlot, Caseflow Management in the Trial Court: Now and for the Future (American Bar Association, 1987) and E W Wright, Victorias approach to reducing criminal case delays: Specific initiatives in Papers Presented at the Eighth Annual AIJA Seminar (Australian Institute of Judicial Administration, 1989) at pp. sssssss2551.
12  equal to because the measure relates to pending cases and the standard is the number of hearings within which cases should be finalised.

Ch 7. General results of the consultations

The Justice Research Centre

Publication details