Caseload Management in the Law Courts: Methodology, Experiences and Results of the first Swiss Study of Administrative and Social Insurance Courts

The judiciary in Switzerland – like all state bodies – is under increasing pressure to reform: On the one hand, workloads, the complexities dealt with and procedural requirements are all tending to grow in volume, while at the same time there are scarely any additional resources available to cope with the problem. By prof. dr. iur. Andreas Lienhard and mag. rer. publ. Daniel Kettiger


I. Introduction
The judiciary in Switzerland -like all state bodies -is under increasing pressure to reform: On the one hand, workloads, the complexities dealt with and procedural requirements are all tending to grow in volume, while at the same time there are scarcely any additional resources available to cope with the problem 1 .The outcome is that judicial authorities (conciliation and arbitration authorities, justices of the peace, public prosecutors offices, courts) are forced to increase their efficiency, 2 , and this can ultimately be achieved only through a truly effective system of court management 3 .One element of court management is caseload management.The former Chief Justice of the Federal Supreme Court, Arthur Aeschlimann summed this up as follows: "Each division president will have to work with objectives for his or her division members and staff and keep up to date on individual caseloads."4Some courts in Switzerland are currently addressing questions of caseload management: these include at federal level the Federal Supreme Court (the highest Swiss court) and the Federal Administrative Court (the court that reviews administrative acts taken by the Federal Administration), and at cantonal level in particular the Administrative Court of the Canton of Lucerne.This article intends to show first of all the importance of caseload management in the context of court management, and it also provides information on a research project which the authors were able to carry out on behalf of Administrative Court of the Canton of Lucerne5 -the first project of this kind in Switzerland.
In order to understand the following comments, it is important to know that in Switzerland there is no uniform, nationwide administrative procedure.Each canton enjoys extensive autonomy to regulate its own administrative procedure and the structure of its administrative justice system.There are therefore considerable divergences in the cantons and these affect both the various stages and channels of justice as well as administrative jurisdiction and court organization.Court Management means the administration of the courts, i.e. the "administrative activity that creates and maintains the resources and personnel required for arriving at court judgments and rulings"6 .The essential matters here are the administrative and financial management of the courts 7 -namely those areas that represent the central components of oversight and supervisory control 8 .The former President of the Cantonal Supreme Court of the Canton of Zurich, Rainer Klopfer, described the importance of court management as follows: "A courts, as a major institution providing services, and as the most important supervisory body, needs a professional, efficient administration.This does not happen without management, but this in no way means that the independence of judges is compromised, just the opposite.It produces better working conditions for the judges and means that they can better fulfill their core duty, namely to adjudicate." 9 .

II. Caseload Management as an Element of
At federal level, the requirements in terms of court management can be derived from three provisions in the Federal Constitution.First we should mention the right of the courts to self-government as defined for the Federal Supreme Court in Art.188 para.3 of the Federal Constitution 10 .The right to self-government as an essential component of the institutional independence of the judiciary is simultaneously a right and an obligation for a functioning system of court management 11 .
An essential requirement for court management is also the constitutional requirement that public resources should be used economically.This efficiency requirement at federal level, which is laid down in Art.126 para. 1 of the Federal Constitution 12 , applies not only to administrative authorities but also to the courts 13 .Since courts never receive sufficient resources to meet their notions of quality, the result is an ongoing drive for efficiency.A further guideline is the constitutional requirement in Art.170 of the Federal Constitution, which stipulates that public duties must be carried out effectively.Court management should accordingly ensure that the courts can in fact fulfill their duties.This means guaranteeing the protection of the law, the consistent application of the law, and the development of case law 14 , while all the time observing proper procedural guarantees (Art.29 ff.Federal Constitution), such as prompt and timely adjudication.
In other words: court management is a necessary precondition for guaranteeing proper adjudication.
In recent times, endeavors to improve court management in theory and practice 15 have produced a number of constituents for a system of good court management that may be listed as follows 16  Caseload management here is of key importance for guaranteeing adjudication, both with regard to preventing legal delays or the denial of legal rights (settling cases in good time) as well as with regard to the quality of judgments (settling cases in the correct way in procedural and material terms).

Caseload Management
It goes without saying that not all cases generate the same amount of work.In order to provide reliably for proper case assignment and the allocation of the resources required (and to maintain a good climate in the court and guarantee the presence of "legal harmony" between the court divisions), it is necessary to have some idea of the staff resources (and in particular working hours) that a case in any legal field (or case category) requires on average.
Caseload management is thus closely connected with controlling; it may also be a component of an integrated controlling system 17 .The Ordinance of the Federal Assembly on Judges' Positions in the Federal Supreme Court 18 addresses this concern and sets it down in legal terms as follows: Art. 2 Controlling and Reporting 1 The Federal Supreme Court shall introduce a controlling procedure that will serve Parliament as a basis for supervisory control and for determining the number of judges.
2 In its annual report, it shall comment on trends in the workload and in general terms on the results of controlling.
Today there are still some uncertainties as to the method of caseload management.There is also a notable lack of reliable reference values and documented experiences.On the organization and workload of courts in Switzerland, and in particular the highest administrative and social insurance courts, there is a general lack of surveys and figures that can be compared.The research project outlined below attempted to gain new insights in this regard.
ISSN 2156-7964 recently that caseloads have become a major subject of research again.The Ministry of Justice of Belgium has commissioned a series of studies, for example, 25 , and in Switzerland -as part of different initiatives towards good court management -caseload studies have also been launched. 26is article describes the first caseload study ever made in Switzerland by a University Institute on a scientific background.It aims not to give final answers to methodological questions, derbut rather to give insight to a research project.The study described lead to a lot of discussion in Swiss judiciary and seems to become the starting point of a series of caseload studies within Swiss courts.

III.
The Research Assignment as Point of Departure The Administrative Court in Lucerne reviews the administrative acts of cantonal and communal authorities in the Canton of Lucerne.In administrative matters, it is the highest cantonal court, and at the same time in many cases also the only court.Its judgments may normally be appealed to the Federal Supreme Court.
The administrative court is organized in three divisions.As a result of the growing number of cases in recent years, in particular in the field of invalidity benefit cases, the divisions dealing with administrative law and tax law have increasingly been called on, within the framework of what is termed the "caseload exchange procedure", to take over cases from the social insurance division in order to guarantee their completion within a reasonable period of time.This circumstance made it necessary to introduce a system of caseload management.As a consequence, the Administrative Court of the Canton of Lucerne embarked on the first efforts to construct a caseload management system.After studying the results of these internal steps, in October 2007 the court in plenary session requested its administrative committee to bring in one or more experts in order to investigate and evaluate the workload involved in the settlement of cases from all legal fields on the basis of the current workload and present system of organization.
The external experts finally were given the task of investigating and evaluating the workload involved in the settlement of cases from the major legal areas on the basis of current workload and present court organization as follows: Ultimately, the external experts' essential mission was to investigate how high the average workload was, based on the existing average requirements for the important categories of case and within the limits of evaluation accuracy.It was intended that the investigations would proceed by making comparisons with other Swiss courts.The results expected were as follows: Information for own use:  Basis for objective caseload management;  Basis for the staff allocation in a subsequent merger of the Cantonal Supreme Court and Administrative Court to form a Cantonal Court.Additional information as "added value":  Comparative data for other judicial authorities in Switzerland;  Contribution to knowledge (in substance and method).

Task Analysis and Procedures
In principle, the task set was a business management problem involving the optimization of processes such as is encountered frequently in the private sector in production and service companies.The problem would therefore have to be tackled using the methodology applied to similar optimization processes.In order to obtain the basic data, all court employees and registry workers would be required over a period of at least five to seven months to keep records of the working hours that they devoted to predefined categories of case.At the same time, by conducting interviews with selected court employees, details would be collected on the nature of the work and the processes involved.Such ISSN 2156-7964 analyses have already been carried out in courts, in particular at the beginning of the 1970s in German civil courts of the first and second instance27 .
Procedures such as these are costly, both in terms of time and finance.In addition, the results would not be ready for a good year.The specific situation of the Canton of Lucerne's Administrative Court required us to look for ways of obtaining usable data more quickly and with less expenditure.A modular phased procedure was therefore chosen:  Phase 1: Blanket survey  Phase 2: Validation of the results of Phase 1 by means of interviews  Phase 3: Detailed investigation at the Administrative Court of the Canton of Lucerne This contribution contains only the results of the completed Phase 1.In relation to the case numbers identified, therefore, the values are approximations which are, however, coherent in the overall relationships.

Organization
The administrative court as "the client" was represented on the project by a three person working group: Dr. iur.Andreas Korner (President); Dr. iur.Patrick Müller (Vice-President and from 1 January 2009 President; internal project management); Dr. iur.Heiner Eiholzer (Vice-President).The team of experts comprised the present authors.In order to provide methodological backup and to validate the results, a specialist advisory committee was established comprising the following experts: Prof. Dr. iur.Arnold Marti, Cantonal Supreme Court of the Canton of Schaffhausen; Dr. iur.Hans-Jakob Mosimann, Social Insurance Court of the Canton of Zurich; Federal Supreme Court Judge Rudolf Ursprung.

Online Survey as Phase 1 of the Overall Investigation
a. Basics A first phase involved a blanket survey of all the highest level cantonal administrative and social insurance courts in Switzerland 28 , which in fact consisted of two separate surveys: The aim was therefore to achieve a certain methodological pluralism.The information provided by Survey A, although on the one hand representing clear numerical values derived from statistics, could not all the same be used for comparability purposes on account of its origin, while the results from Survey B arose in accordance with clear, uniform and comparable methodological principles, but are limited by the imprecise nature of estimates.
b. Survey A In the survey of the court administrative offices/court registries, the data summarized below was collected (centrally per court or court authority): -number of staff expressed as percentages (of the total annual working hours in full-time employment) of lawyers working on the preparatory briefing procedure and decisions (including court clerks, part-time judges, etc); -Settlements in total and divided according to the 19 categories of case, and further subdivided according to settlements without final judgment (proceedings settlements) and material settlements; -number of contested decisions; -success rate of appeals.
The following parameters may adversely affect comparability between the cantons and were therefore also surveyed: -avenues of appeal (administrative court as first or second instance); ISSN 2156-7964 -cognition of the court (examination of facts of the case, application of the law, appropriateness); -powers with regard to the judgment (to modify or annul); -adjudicator activity (persons acting as judges as their main occupation, or as a secondary occupation29 , clerks of court); -settlement possibilities (single judge, decisions made by correspondence).
The case numbers for 2006 were recorded (settlements in the financial year 2006), so that a survey of the success rate of appeals in relation to the corresponding decisions is made possible.The survey of staff capacities, however, relates to the year 2008 in order to achieve compatibility with Survey B.

c. Survey B
In the case of the employees of the courts listed in Annex 1, the survey consisted of an assessment of the average expenditure of time per case in 19 specified categories of case.
Only those persons actively participating in the court judgments were interviewed.The survey was conducted anonymously, so although the results can be classified according to certain characteristics (e.g.canton, main occupation or secondary occupation), no conclusions can be drawn with regard to specific persons.Accordingly, this also meant that in the case of this part of the survey, no feedback or follow-up was possible that might have made the answers more precise.

d. Categories of Case
In order to reduce complexity and indeed to facilitate the conduct of the survey, case categories were established that related to various subject areas (cf.Annex 2).For this purpose the following criteria were considered to be decisive:  types of case in accordance with the assignment system currently used by the Administrative Court of the Canton of Lucerne as a starting point;  the list of the case categories must be of a size suitable for the survey (around 15 categories of case);  number of cases per type of case must be taken into account (large number of cases = separate category of case);  the categories of case must be balanced;  business reports/number of cases/classifications made by the administrative courts in the cantons of Bern, Lucerne and Zug as a further basis.
The core component of both surveys was accordingly formed from the following 19  e. Procedure The surveys were carried out electronically via the internet using an online tool.This meant that the invitation to participate had also to be sent by e-mail in order to allow people to click directly on the links needed for online questioning.The survey proceeded anonymously and under password protection.
In order to ensure that the materials reached the right destination, all the judicial authorities concerned (or their presiding bodies) were informed beforehand by letter and requested to provide the research team with the name of a contact person.
The courts thereafter received two e-mails:  a first e-mail with a link to Questionnaire A, which was to be dealt with centrally by the court administrative office; and  a second e-mail with a link to Questionnaire B, which was to be forwarded to all lawyers involved in court judgments.
All the highest cantonal administrative and social insurance courts in Switzerland took part in Survey A (cf. list in Annex 1).
A total of 198 persons participated in Survey B. 15 questionnaires were incomplete or otherwise filled out unintelligibly in a way that precluded them from being taken into account.On the basis of the data obtained with regard to the staff capacities at the administrative courts in Switzerland, this corresponds to a return of around 50 per cent, which can be considered to be better than average.

Methodological Comment on Comparability with Other Studies
The point of departure for the court survey was the deployment of personnel per case, effective personnel on the basis of statistics in Survey A, and estimated personnel in Survey B. Both the data derived from statistics as well as the estimates provided by individuals questioned were based on the number of full-time staff expressed as percentages (percentage of the annual working hours of a full time staff member).There is therefore no scaling present here.The upper values are open.The middle value (and also the median 30 ) do not inevitably reflect the estimated medium workload created by the case.
Other methods are based mostly on estimated values: the values they survey partly represent individual assessments by persons from the respective statistical population (court in plenary session, division), made on the basis of a pre-defined scale (scaled response to questionnaire).The scale employed is abstract, i.e. it represents a weighting system sui generis, which may not be linked to any other system of measurement or assessment.In addition, the scale is closed by an upper and lower limit; the lowest possible value (e.g. 1) and the highest possible value (e.g. 5) are prescribed beforehand.It was furthermore decided in advance that the middle value of the workload scale (in the case of a scale from 1 to 5, the value 3) signifies an average load.With such a procedure it is not clear from the outset whether the scaling is linear (i.e.whether the distance or value difference between the values is always the same size) whether it takes a different course 31 30 The median is the dividing line between two halves.In statistics, the median divides a distribution into two halves.In contrast to the arithmetical mean, or average, the median has the advantage of being more robust in relation to outliers (highly anomalous values) and can be used on ordinally scaled variables (after Wikipedia).31 The question of whether a scale, where there is a scaled response, is linear depends on the precise, verbal definition of the values.For example the scale of values "completely exceeded/partly exceeded/satisfactory/partly unsatisfactory/completely unsatisfactory" is not linear, because in our experience the interviewees tick "partly exceeded" for work that is slightly better than satisfactory, but only tick the highest grade in the case of absolutely perfect work.In practice, the same scales are also worded and used differently in ISSN 2156-7964 Due to the various evaluation systems (more precisely: the reference systems and frame of reference of the evaluation systems), it is not possible to recalculate this or that value in order to make them more comparable.The only possibility for achieving what is still an approximate comparability would be to have completely identical categories of case, both with regard to completeness and to demarcation between the categories.

V. Survey Experiences
At the outset it may be stated that interest in the survey was considerable, the returns were high 32 and the feedback was fundamentally positive.
In the case of Questionnaire A various difficulties arose:  Online questioning was obviously a problem for some court registries, and led to numerous feedback questions during the ongoing survey and to later questions from the experts when the values were being validated.The reason for this may well lie in the fact that online surveys such as this are unusual, and that the survey was also carried out in the main vacation period which meant that it was not uncommon for several people to be working, one after the other, on the same questionnaire. The structuring of the questionnaire could not do full justice in every respect to the various models of organisation for administrative adjudication in Switzerland.This meant that it was not until the courts had consulted the experts that they were able to decide which of the values available were supposed to be entered in what fields. The differing statistics from canton to canton made it difficult to obtain uniform results for the whole of Switzerland.In some cantons, certain values could not be provided, either because they could not be taken from existing statistics and business controls or because they could not be reconstructed on that basis.
In addition, the structure of the application of administrative law with regard to the division of responsibilities between cantonal administrative and social insurance courts was in some cases not completely recognised until the survey was running, and this necessitated a further survey of the relevant highest court whose results were missing.In addition there were a few courts which, despite previous assurances that they would collaborate, had to be once again encouraged to participate in the survey.The conduct of the survey thus encountered difficulties and was delayed.
There were only a few problems with Questionnaire B. This may be ascribed in the first place to the low number of queries from the interviewees, as well as to the rather low number of returned questionnaires that could not be used.It may therefore be concluded that the decision to carry out the survey by electronic means was correct.The difficulties that arose could probably have been avoided to some extent if a more detailed instruction sheet for filling out Questionnaire A had been made available to the court registries.

Preliminary Remarks
The results of the survey permit numerous findings on the organization and procedural aspects of the application of administrative law in Switzerland that go beyond the minimum information that was required in response to the questions posed in terms of the research team's remit.For the time-being, however, we will refrain from a systematic and complete evaluation.In what follows, only those aspects will be presented that were specifically important to the assessment of the workload of the Administrative Court of the Canton of Lucerne or that enable a better understanding of the statements on the workload of the Administrative Court of the Canton of Lucerne.A more detailed evaluation of the data is planned in the context of a larger project.
different cultures: the traditional (Swiss) school marking system with grades from 1 to 6 in state schools is mainly used as a linear system of assessment (often in a test there are 12 questions and each question carries half a point, with the result that 8 points are enough to pass), at universities the tendency is more exponential in the satisfactory range.32 See for more detail No. IV. 3. e above.

2.
Organization of the Application of Administrative Law One of the main findings of the investigation is that the landscape of the administrative justice system in Switzerland, including the highest administrative courts in the cantons, is, in relation to its structure and procedural organization, and with regard to certain parameters of trial practice, far more diverse than was assumed at the beginning of the research project.
This diversity has led to the assumption that in a comparison with the Swiss-wide average, a distortion may arise, in particular due to the values returned by the small, rural cantons.For this reason, the comparison is made based on a selected group, consisting of the courts of the cantons of Aargau, Basel Stadt, Bern, Lucerne, St. Gallen and Zurich.Purely in terms of size and structure, the cantons of Vaud and Geneva suggest themselves as comparable cantons.The administrative courts of the Canton of Vaud, however, cannot be used as a comparative model because they have a different method of counting settlements.The Canton of Geneva presents difficulties affecting a comparison in the area of social insurance judgments, because in Geneva there are disproportionate numbers of cases involving social insurance law (including social assistance).
With regard to the evaluation of specific categories of case, the organizational diversity leads to especially difficult repercussions, in that  in some cantons deprivation of liberty for the purpose of providing care (compulsory custodial care orders) and matters relating to guardianship (wards of the court) are dealt with by the civil courts, while in others the administrative courts have jurisdiction;  in particular in the field of road traffic law (Road Traffic Act), the final cantonal instance in some cantons, (often at the same time the only cantonal authority) is not the ordinary highest administrative court but a special administrative court (e.g. in the Canton of Bern, it is a so-called Appeals Commission33 );  in some cantons the decision in tax cases is made by the administrative court as second instance subsequent to a special tax appeals authority;  in the Canton of Zurich (and in some other cantons), the judgment made by the administrative court as court of first instance in tax cases can on application (i.e. in the case of objection against the penalty imposed) be dealt with in a main hearing;  for coercive measures under the law on foreign nationals, the responsibility lies to some extent not with the administrative courts but with the courts responsible for coercive measures that are affiliated to the prosecution authorities (today often designated as detention courts) (e.g. in the Canton of Bern)34 .

3.
Procedural Law: General Conditions For all the considerations given below, the following general procedural conditions that apply throughout Switzerland should be taken into account 35 . In social insurance law, the legal channels are specified by federal law36 .There can thus be no cantonal special features in this regard.The procedure differs, however, according to the legal field concerned: in the case of invalidity insurance, the competent authority must first issue a preliminary decision and then a ruling; this latter may be challenged directly by an appeal to the cantonal social insurance court.In cases relating to the Occupational Pensions Act, claims proceedings must be taken to the cantonal social insurance court (original administrative court).In all other social security fields, the insurance fund is required to make a ruling, which may be contested.The decision on the objection is subject to an appeal to the cantonal social insurance court.
ISSN 2156-7964  In the area of tax law, due to the Swiss-wide harmonization 37 of the appeals procedure to a cantonal court, there must always be an objection procedure before the appeal.In all cantons, therefore, appeals relating to direct taxes are appeals against decisions on objections.
A singularity of the Lucerne procedure in administrative disputes is that the administrative court normally acts directly as first appeal instance and simultaneously adjudicates as the only court of first instance, and this means therefore that there is neither a prior objection procedure involving the authority issuing a ruling nor a judgment by an internal administrative judicial authority.The survey reveals that in the case categories of administrative justice, the results across Switzerland tend on comparison to be quite diverse.Notably in the case category of planning, construction and environmental law, the judgments in certain cases are made in other cantons by the highest courts as the first instance:  as the only instance (Solothurn 38 , Vaud) 39 ;  depending on the subcategory of the cases either functioning as first and only instance or as the second appeal instance (Bern, Glarus, Graubünden);  as first and only authority after an internal objection procedure (Bern, Fribourg, Geneva, Neuchatel, Zug).
It may be assumed that cases dealt with by the court of first instance are more time-consuming in the early briefing phase, since there are often questions of fact (incompletely established or contested facts of the case) that have to be cleared up.With regard to cognition, the following may be stated: normally the administrative courts examine only the establishment of the facts and the application of the law.An exception is constituted by the test of proportionality, mostly in cases where administrative courts adjudicate as first appeal instance.This is currently the normal practice in the Canton of Lucerne.
In contrast to our original assumption, the highest cantonal administrative and social insurance courts may modify or annul judgments in almost all cases.The highest administrative courts annul judgments in most cantons only in cases where they (alternatively) do not have available to them the same cognition as the lower court, or when the lower court has erroneously not applied a legal remedy or where there still remains a need for considerable preliminary briefing 40 .With regard to the competence to make a decision therefore, we must proceed on the assumption that these courts are entirely comparable.
The overwhelming majority of the highest administrative and social insurance courts recognise not only the judgerapporteur activity of persons employed as judges as their main occupation 41 -like those in the Administrative Court of the Canton of Lucerne -but also the rapporteur activity of court registrars and clerks of court.In contrast, the administrative activity of part-time judges tends to occur more infrequently.In addition, most courts have the possibility of settlement by a single judge without a decision on the facts.On the other hand, the Swiss administrative justice system, according to the results of the survey of the courts (Survey A), apparently does not recognise the general material settlement by an individual judge.Some cantons recognize the single judge material settlement even in cases with a known amount in dispute, provided this latter is relatively low 42 .Overall there prevails a high degree of comparability with regard to the possible deployment of persons involved in court judgments as rapporteurs a well as of single judges in the case of the highest administrative and social insurance courts.

Whereas the Administrative Court of the Canton of Lucerne recognizes the possibility of decisions by correspondence in all categories of case irrespective of the type of decision, a very nuanced picture emerges for the rest of the cantons.
There is as yet no detailed assessment.
ISSN 2156-7964 Overall it may be noted on the basis of this first overview of the Swiss system of administrative procedure that only one general procedural condition could be decisive with regard to the (un-)comparability of the Administrative Court of the Canton of Lucerne with the highest administrative and social insurance courts of the other cantons, namely the question of whether the court as first appeal instance directly judges or does not judge the rulings of administrative authorities.All other differences in the procedural general conditions are of no relevance to a comparative study.

4.
Staff Aspects In Switzerland, around 415 full-time positions are occupied by judges and court registrars at the highest cantonal administrative and social insurance courts.A total of 345 full-time equivalent positions are deployed directly for adjudication in the highest administrative and social insurance courts of the cantons.This leads to the conclusion that -as an overall Swiss average -of the staff resources at the level of judges, court registrars and clerks of court around 80 per cent are devoted directly to the work of adjudication, and around 20 per cent are used for other functions (management and coordination functions; administrative duties, including statistics; representation; training and education; etc.).

Statistics, Number of Cases
The quantitative data may be summarized as follows 43 :  In Switzerland there are each year around 21,800 cases brought before the highest cantonal administrative and social insurance courts. The average time devoted to each case across all categories of case amounts in these courts to a total of around 38 to 42 hours 44 . The mean, arithmetical workload value across the entire country per case on the basis of a comparison of all cases (21,793) and the total resources effectively available for court judgments (34,439 full-time staff 45 ): 1.89% of the annual working hours of a full-time member of staff per case. mathematical median across the caseload figures for the individual courts surveyed in relation to the entire country: 2.03% of the annual working hours of a full-time member of staff per case. mathematical median across the caseload figures surveyed for the individual courts in relation to the comparison group ( the cantons of Aargau, Basel-Stadt, Bern, Lucerne, St. Gallen and Zurich): 2.07% of the annual working hours of a full-time member of staff per case.
The number of cases were surveyed for 2006 (settlements in the financial year 2006), in order to enable data to be collected on the success rate of appeals.This means that for the courts of all cantons, the changes brought about by the entry into force of the Federal Act of 17 June 2005 on the Federal Supreme Court (Federal Supreme Court Act, FCSA) 46 , have not yet been taken into account.Due to the guarantee of legal recourse, the burden of the highest administrative courts will probably increase, since administrative courts are now required to adjudicate in cases that at present are judged in the final cantonal instance by an administrative authority.
The approximate calculation of the average workload per case contains a methodological lack of precision in that the number of cases relates to the year 2006, while the staff capacities relate to the year 2008.

6.
Success Rate of Appeals Comparisons of efficiency between courts are only admissible if the success rate of appeals is factored in: the success rate of appeals, i.e. the relationship between the number of cases contested and appealed to the Federal Supreme Court to the number of cases in which the Federal Supreme Court overturns the corresponding judgment of the highest cantonal 43 In relation to the limited applicability of these general figures, reference is made to No. VII. 1.And VIII. 2. 44 Assuming an average workload per case of 1.9 to 2.0 % of the annual working hours of a full-time member of staff and net annual working hours (i.e.under deduction of Saturdays, Sundays, public holidays and vacation) of between 1'900 and 2'100 hours.45 Not included are the resources of the Social Insurance Court of the Canton of Geneva, which could not be recorded on the basis of the court statistics.In the event of an average difference between the nominal values and an effective use of resources of around 20 per cent, this would result in an additional 800 to 900 per cent of the time worked by a full-time member of staff.46 SR 173.110.
ISSN 2156-7964 authority, may shed some light on the quality of the work of the cantonal court.In addition a high share of annulled judgments, i.e. judgments that have been overturned and referred back to the lower court for reconsideration, can create an additional workload for the courts without this being reflected in the surveyed number of cases or in the estimates of expenditure of time.
For the time-being we have refrained from a systematic evaluation of the values surveyed with regard to success rate of appeals.A preliminary summary consideration reveals that the success rate of appeals in social security cases is generally lower than for other categories of case 47 : -Cases under administrative law: cantonal court with highest value 12.9%; numerous courts with values between 5.0 and 7.5%; -Tax law cases: cantonal court with highest value 5.4%; numerous courts with values 0.0%; -Social insurance law cases: cantonal court with highest value 42.6%; numerous courts with values around 20%. -7.
Cantonal Differences hamper Comparability The small, rural cantons (in particular Appenzell-Innerrhoden, Appenzell-Ausserrhoden, Nidwalden, Obwalten, Uri) have administrative courts that make the most extensive use of part-time judges.In addition, the number of cases in each of the categories is essentially different from that in the more urban areas or in cantons with both urban and rural or suburban regions.For this reason as well, the comparison in the study undertaken for the Lucerne Administrative Court compares this court with the whole of a selected group, comprising the courts of the cantons of Aargau, Basel-Stadt, Bern, Lucerne, St. Gallen and Zurich.
In various cantons (in particular Basel-Stadt, Schaffhausen, Solothurn), all the administrative and social insurance judgments at the higher level are issued by the Cantonal Supreme Court, as are higher level court judgments in civil and criminal cases, such that the judges basically all have experience of cases from the entire spectrum of court judgments.In view of the small workloads per category of case, this leads to a situation in which it becomes scarcely possible to estimate the workloads involved properly.

Introduction: Basic Comparability Issues
The research remit was to examine and evaluate the workload involved in the settlement of cases from the essential legal fields on the basis of present workload and current court organization.The objective was to create a foundation for a future system of caseload management.On the basis of the surveys of the highest administrative and social insurance courts in Switzerland (list cf.Annex 1), it is now possible to compare and appraise the workload in the individual categories of case.
Comparisons of performance between individual courts, however, would presuppose that the values were produced in the same way as at the Administrative Court of the Canton of Lucerne 48 .A benchmark secured in business management terms would also have to be employed by all participating courts on the basis of a similar scrutiny of time and performance stretching over a lengthy period.
As already indicated, however, it is perfectly possible to make comparisons between the activities of courts that have been properly grouped.This is especially valid for a potential comparison of the Administrative Court of the Canton of Lucerne with the administrative courts of the cantons of Aargau, Basel-Stadt, Bern, Lucerne, St. Gallen and Zurich (selection/comparison cantons).The average values from these cantons taken together can therefore be employed as a basis for the future caseload management of the Administrative Court of the Canton of Lucerne or of another of the comparison cantons.
47 Percentage of cases that were quashed by the higher court (Federal Supreme Court, Federal Administrative Court) when compared with the total of cases for each cantonal court; cantonal judgments from 2006.48 The corresponding data was treated as confidential in accordance with instructions.
ISSN 2156-7964 Proceeding from the surveyed statistical values of case numbers and from the data on staff resources collected from the court registries, the value for the workload per case (expressed as a percentage of the annual working hours of a full-time member of staff) can be obtained for all the courts.By comparing, the following three values can be generated and analysed:  Average workload per case across the whole of Switzerland on the basis of the comparison of all cases (21,793) and the entire available resources effectively devoted to adjudication (34,439 full-time staff): 1.89% of the annual working hours of a full-time member of staff per case;  Median across the caseload figures collected for the individual courts in relation to the whole of Switzerland: 2.03% of the annual working hours of a full-time member of staff per case;  Median across the caseload figures collected for the individual courts in relation to the comparison group (the cantons of Aargau, Basel-Stadt, Bern, Lucerne, St. Gallen and Zurich): 2.07% of the annual working hours of a full-time member of staff per case.
On the basis of Survey A, it may be concluded that in Swiss administrative courts across all case categories, an average case from the average of the comparison cantons requires approximately 2.00% of the annual working hours of a full-time member of staff.
The burden per case in hours is obtained by multiplying the percentage of the annual working hours of a full-time member of staff per case by 19. 53 The ratio of lawyer per number of settlements in the year (and therefore the number of cases per lawyer in case category per annum) may in principle always be derived from the percentage of the annual working hours of a full-time member of staff per case, provided the number of hours worked per year is known.This value does not permit caseload management to be organized, however, and in the opinion of the authors it is not in other respects particularly meaningful.

Findings for Future Caseload Management Systems
a. Recommended variant: management according to the percentage of full-time staff per case The most advantageous method of caseload management is the one that proceeds via the percentage of the annual working hours of a full-time member of staff per case.As a point of departure, the benefits for caseload management of using the percentage of the annual working hours of a full-time member of staff include the following:  The system could readily be adapted to personnel management (number of positions, upper limit on positions, etc.), because personnel management likewise normally uses the percentage of the annual working hours of a full-time member of staff as a basis. Such a system, for example, also enables an adequately accurate personnel need to be estimated directly from the number of unsettled cases. It also makes possible a proper allocation of cases on the basis of the level of employment -but this would require the taking into account of staff resources (expressed as a percentage of full-time employment) deployed on or withheld on account of matters other than court judgments. In the case of known annual working hours, it is always possible to recalculate in terms of working hours. It facilitates the simple calculation of how many cases of a certain case category should be completed annually by a lawyer. In the case of the same or similar annual working hours and also of the same of similar general procedural conditions, a benchmark with other courts becomes feasible.
53 The annual regular working hours for a full-time employee in Canton Lucerne amounts to around 2'050 hours (taking account of days that are not worked, but including vacation).From this must be deducted vacation of 20 days (if 8.4 hours are worked every day, this is equivalent to 168 hours).Assuming no absences due to illness, this results in around 1'882 hours of work for a full-time employee.Accordingly 1% of the working hours amounts to around 18.8 hours.See also the Staff Manual for the Canton of Lucerne.In most other cantons the working hours are similar.
ISSN 2156-7964 On account of the imprecise nature of the results that depend on estimates, and also for the sake of the practicability of the future system of caseload management, it is, however, recommended that rough scale of approximate values be used, subdivided into a total of three to four categories.

b. A Further Variant: System of Points
The basic idea of this system is that although on the one hand the investigated bandwidth of the percentage of the annual working hours of a full-time member of staff per case is quite large, the relationship between the categories of the workloads established on the other hand is relatively constant and may be regarded as a reasonably secure value.The idea of the points system consists in proceeding from the workload difference between the individual categories of case and assigning the categories to a system of points with a scale of three, four or five points.Caseload management is then carried out based on these point numbers for the assigned cases.

1.
General findings The research work presented above commissioned by the Administrative Court of the Canton of Lucerne has demonstrated that it is possible to construct caseload management systems for courts on the basis of estimated values. 54he study made possible for the first time Swiss-wide findings on the workload of the highest cantonal administrative and social insurance courts.At the same time, however, the limits on surveys based on estimates were also revealed (cf.Sections 2 and 3 below).
Even the most reliable method, moreover, should not be allowed to divert attention from the fact that we are basically dealing with a mere quantitative record.This means first of all that the values collected are in need of elucidation, especially in the case of comparisons between several courts -and above all with regard to what we are looking at (test of legality or appropriateness) and in respect of the lower court (judicial authority or ruling authority).On the other hand, the values can only possess a qualitative informative validity if they can be seen in relation to the frequency and success rate of appeals.Or in other words: there is no art involved in spending little time on a judgment.Whether it is a "good" judgment or not is another question 55 .
In the area of administrative theory and research on courts, the study shows that in Switzerland little or nothing exists by way of an overview of the structure and workflow management of the administrative justice system.The modus operandi of administrative adjudication has hitherto hardly been investigated; in particular there is a lack of key management figures and uniform statistics that would permit comparisons between the cantons.

2.
Questions still open The evaluation of the questionnaires raises, among other issues, the question of why there is a considerable difference between the estimates of the workload per case given in response to the questionnaires by those involved in adjudication (Survey B; Results see Sec.VII. 2) and the workload per case obtained from information provided by the court registries (Survey A; Results cf.Sec.VII. 3).Basically there are two possible explanations for this difference:  The persons involved in adjudication give an estimate for their workload per category of case that is too low, which would accord with standard experience in the case of this type of survey.This kind of erroneous estimate may, 54 In the USA you find also caseload studies based on estimated values, but there usually the estimation took place in the iterating process of the Delphi Method, cf.Information Brief on Weighted Caseload Methods of Assessing Judicial Workload and Certifying the need for Additional Judges, The Florida Legislature, Report No. 97-67, p. 55 Examples of indicators of good judgments that may be cited include: factual accuracy, length of proceedings, intelligibility, comprehensibility, development of the law, environment, costs, frequency of appeals, success rate of appeals, cf.Andreas Kämpfer, Was ist gute Justiz?-Qualitätskriterien für die Richtertätigkeit, in: Daniel Kettiger (Ed.), WOV in der Justiz, SGVV-Schriftenreihe Band 44, Bern 2003, p. 129 ff.; Aeschlimann (Note 4), p. 413 also states: "It is not the task of the supreme court to overturn as many cases as possible".
ISSN 2156-7964 however, have other causes (unfamiliarity with making estimates in percentages of full-time working hours, excessive modesty, in self-assessment, etc.). The court administrative offices place too high an estimate on the level of productivity obtained from the nominal full time positions.It could well be possible that individual persons involved in adjudication devote a higher proportion of their working hours to continuing professional education, studying legal texts and instructing new staff or trainees than is generally assumed.In addition, it could be the case that absences due to vacation and illness have not been deducted from the nominal percentage level of employment, which would also push up the workload per case.
It is unlikely that this question can be answered with the required reliability by means of additional interviews (Phase 2).In contrast, a complete survey of working hours (Phase 3) would provide the required clarity. 56

Possible Follow-up Projects
The figures for case workloads for the Administrative Court of the Canton of Lucerne are -as in keeping with the informative value of Phase 1 -flawed due to inaccuracies, and this could have a cumulative effect over several years on caseload management.As was already suspected when the work was begun 57 , sufficiently precise values can only be obtained using traditional business management time recording.In order to obtain this basic data, it would therefore be necessary to record the work of all legal and administrative staff on predefined categories of case, itemizing the time taken on each case, over a period of five to seven months (Phase 3).

Final Remarks
The overall study served first and foremost to develop a caseload management system at the Administrative Court of the Canton of Lucerne.The survey carried out involved setting foot on virgin territory, in particular because the parameters of the general procedural conditions, not to mention the allocation of personnel and the case figures of courts of this type had never before been investigated in Switzerland.Those results shown in the report and in the evaluation that have not yet been made available to the public 58 must not therefore be used indiscriminately to provide benchmarks, other comparisons of efficiency or as a basis for reorganization measures.
Furthermore, it must be stressed that the study, in accordance with the assignment, primarily tackled quantitative aspects.The only qualitative aspect explicitly investigated is the success rate of appeals.The results of the quantitative surveys reflect the average of the qualitative performances of all the participant courts, due to the representativeness of the values obtained.On the question of whether this average corresponds to the current qualitative requirements for good justice system, no comments can be made.

a.
Court administrative offices/Court registries (Survey A): collection of various key figures, unique per court or court authority.b.Court employees (Survey B): estimates of average expenditure of time per case in ten specified case categories.

Annex 1 :
Courts that participated in the Online Survey AG: Cantonal Supreme Court (Part of Court: Insurance Court) AG: Cantonal Supreme Court (Part of Court: Administrative court) AI: Cantonal Court AR: Administrative Court BE: Administrative Court BL: Cantonal Court, Canton of Basel-Landschaft BS: Court of Appeal, Basel-Stadt BS: Social Insurance Court FR: Cantonal Supreme Court GE: Administrative Court GE: Cantonal Supreme Court for Social Insurance GL: Administrative Court 56 On the various phases, see No. IV. 1 above.57 See No. V. 1. 58 See also Note 48.
State pension scheme (AHV)/Compensation scheme for earnings lost while on military service (EO) 19.Unemployment insurance/Other social insurance law cases