EVALUATION OF LEGISLATION

Proceedings of the Council of Europe’s legal co-operation and assistance activities (2000-2001)

(...)

Evaluation of legislation – contribution to the quality of legislation

Professor Dr. Luzius MADER
Vice-Director of the Federal Office of Justice, Switzerland

1. The development of legisprudence

During the last three decades, legislation has become an important field of scientific interest and practical training. At least in the German speaking countries, Peter Noll was one of the pioneers opening up the way to a new, comprehensive approach to the legislative phenomenon1. This new approach is not limited to certain countries, as efforts to improve the quality of legislation are a general endeavour. It is not limited to the legal sciences or to jurisprudence but concerns a multitude of academic disciplines interested in various aspects of the legislative activity. And it is not limited either to academic reflections and researches but takes into account practical or pragmatic needs and considerations.

Legislative problems are, to some extent, universal problems; they should be dealt with in a multidisciplinary way, and they have a theoretical as well as a practical dimension. The term "legisprudence" that is sometimes used to join together the different elements and to stress the links between different currents and focal points of modern legislative studies or work expresses in a very appropriate way the character of this new, comprehensive approach2: legisprudence aims at furthering the theoretical understanding as well as the technical handling of legislation; it combines elements of science, art and craftsmanship; it is interested both in the content of legislation and in its form.

The following main fields of interest may be distinguished:

- legislative methodology (or material legistic3): deals with the content of legislation; proposes a methodical way of elaborating normative contents and develops practical tools facilitating the different steps or phases of this methodical approach;

- legislative technique (formal legistic in the narrow sense): deals with formal aspects of legislation, with the different types of normative acts, with their formal structure, etc.;

- legislative drafting4: concerns linguistic aspects of legislation, the way of expressing normative contents by normative texts;
- legislative communication: the publication of normative texts or, more generally, the way of communicating normative contents, including a broad range of information about legislation, not only the official publication;

- legislative procedure5: the process of elaborating, enacting and implementing legislation follows procedural rules of various kinds; these rules may influence to some extent the formal and the material or substantial quality of legislation; they may further or hinder the methodical approach of the legists, i.e. of the persons preparing new legislation;

- the management of legislative projects: preparing legislation may be seen as a task for which the principles and techniques of project management are applicable;

- the sociology of legislation: the political process preceding the enactment of legislation, the implementation process and the effects of legislation are an important field for sociological studies and an essential element of legisprudence;

- the theory of legislation: legisprudence includes also theoretical reflections on the role or function of legislation as an instrument of social guidance and control by the State.

Although general legislative problems have a universal character, the more practically oriented dimensions of legisprudence - and, above all, the solutions practised in fields such as legislative technique and drafting, legislative procedure and even management of legislative projects - are closely linked to particular national contexts and have to obey national legal rules. For this reason, the practical training of legists, i.e. persons involved in the preparation of legislation, can hardly be disconnected from a specific institutional and legal framework.

Whereas legislative drafting or, in a broader sense, formal legistic is mainly an affair for lawyers, material legistic is by no means an exclusivity of persons with legal training or a field of activity for which lawyers are necessarily better prepared than persons having another professional background. This is true, in particular, for the different sequences of the methodical approach, including evaluation, for the management of legislative projects and for the sociology of legislation. For dealing with the material or substantial aspects of the legislative work, lawyers - at least lawyers with a traditional legal education - are not better equipped than anyone else. But because of their predominant and privileged position in the drafting stage of the legislative process they have a particular responsibility also for the material aspects of legislation.

2. Evaluation as an essential part of legislative methodology

The evaluation of legislation forms, on the one hand, an important – and, at least in some fields, traditional6 - element of the sociology of legislation. On the other hand, it is also an essential part of legislative methodology. Unlike legislative drafting, it thus concerns the material, substantial aspects of legislation rather than its formal aspects and it is particularly concerned with the relation between normative contents and their consequences in the social reality, in the "real world".

The methodical approach recommended for elaborating normative contents is based on the separation of different analytical steps or sequences. These steps include:
1) the analysis and definition of the problem that is supposed to be solved by legislative action;
2) the determination or clarification of the goals of legislation;
3) the examination of legal instruments or means that can be used to solve the problem and the choice of such instruments (based - among other things - upon a prospective evaluation of their possible effects);
4) the drafting of the normative content;
5) the formal enactment;
6) the implementation;
7) the retrospective evaluation;
8) if necessary or opportune, the adaptation of legislation according to the results of the retrospective evaluation.

The analytical or theoretical model underlying the methodical approach to be used for preparing legislation considers the legislative process as a reiterative learning process; a process in which the evaluation of the effects of legislation is one of the fundamental prerequisites - and tools - for learning; a process, too, in which the responsiveness of the legislator to social reality and the social adequacy of legislative action should be guaranteed.

Of course, this methodical approach to legislation has its limits; limits that are due, in particular, to the fact, that the approach is based, essentially, on an instrumental view of legislation: legislation is considered to be an instrument of social guidance and control, as a tool of social engineering, as it is said sometimes. In other words, this approach is grounded on the assumption that legislation is a rational activity that is aimed at realising specific purposes or goals, at achieving specific results in the social reality.

This assumption may be questioned. It is only partly accurate: not all legislation has an instrumental character; or, at least, the instrumental character of legislation is not always predominant; it is, for example, likely to be more important in the field of public law than in the field of private law. Furthermore, in some cases, legislation may be seen more realistically as the expression of political constraints rather than as an attempt to solve a problem by changing the behaviour of its target population. It may sometimes be seen, also, as a more or less fortuitous result of political debates and compromises rather than as a rational effort to bring about social change. It may be seen, finally, as a symbolic act without any real ambition to produce practical results; symbolic in the sense that its main function can be to express the political authorities' awareness of a problem, their willingness to share the preoccupations of the people, the fact that they are concerned and feel the need to show that the situation is under control. In this sense, the instrumental, rational view of legislation underlying the methodical approach is a partial, limited approach. Nevertheless, we can admit that in many cases legislation very clearly and undoubtedly has a predominantly instrumental character or is at least used in an instrumental perspective. For these cases, the methodical approach is reasonable and appropriate.

After these preliminary observations aimed at clarifying the general context it may be useful to define what is meant by "evaluation" in the context of legisprudence7. In fact, the word "evaluation" is used today in so many different ways and senses that it is indispensable to clearly define it8. By "evaluation" we mean the analysis and assessment of the effects of legislation9. This very general definition comprises three elements. First, evaluation has to do with legislation, i.e. with normative acts and - implicitly - with the administrative decisions based on these acts. Formally its object therefore differs from the object of policy analysis; however, at least in legal orders respecting the rule of law, there is no fundamental difference in practice between the evaluation of legislation and the evaluation of public policies10. Second, evaluation is interested in the effects; it examines the changes, respectively the non-changes, of attitudes, behaviours or situations - and the consequences of these changes - that are due (potentially or really, according to the perspective) to legislative action. In other words, evaluation is interested in the causal relations between legislative actions and social reality. Third, evaluation tries to analyse and assess the effects of legislation in a methodical way.

The third element of this definition needs further comments. "In a methodical way" means:
1) in a way that can be generally understood and followed, in particular because it clearly indicates the grounds and foundations, or the sources, of statements or judgements about the - potential or real - causal connections between legislative enactments and the observable social reality;
2) in a way that is as systematic as possible; in other words: a way that takes into account all relevant effects;
3) in a way that is as impartial or objective as possible; in other words: a way that is not interest-driven, that is free from particular interests.

The results of evaluations meeting these three requirements are in most cases put down in writing, in the form of notes, memoranda, reports or studies. Thus, they may be published or at least communicated to interested persons or institutions and critically examined. Defined in this pragmatic sense, evaluations are, on the one hand, clearly much more than purely impressionistic, intuitive political appraisals; on the other hand, they do not necessarily comply with the most severe criteria of a truly scientific analysis. They take into account intuitive assessments of the effects of legislation made by the relevant political actors or target groups, build on practical experiences and especially on the specific knowledge held by the authorities implementing legislation and try to complete this knowledge by methodical efforts including often, but not necessarily, the use of various methods of the social sciences. In other words, evaluations are a pragmatic effort to produce more relevant and more accurate information or knowledge about possible or real causal relations between legislative action and attitudes, behaviours or situations that can be observed in social reality.

Of course, such causal relations are very often extremely difficult to analyse. Causality in social reality is a very complex affair. In most cases, it cannot be proven in a strict sense. However, this does not diminish the importance or even the necessity of evaluation. Because - at least if we admit the pertinence of the instrumental perspective on which the methodical approach to law-making is founded - the legislator inevitably admits the existence of such causal relations. It bases its action explicitly or implicitly on assumptions relating to causal relations. The main purpose of evaluation is not to prove that these assumptions are right or wrong. It is to make sure
1) that the causal assumptions underlying legislative action are as explicit and as differentiated as possible when new legislative decisions are prepared;
2) that they are based on the best possible theories and the most reliable information that can be gathered with reasonable efforts;
3) and that their accuracy is examined after the adoption and implementation of legislation.

In other words: evaluation is a pragmatic effort to improve the legislator's assumptions and knowledge about the effects of legislation. It aims at more plausibility in this field, not at certainty or scientific proof.

3. Prospective and retrospective evaluation

In accordance with the analytical model underlying legislative methodology evaluations should be made both before and after the formal enactment of legislation. We have to distinguish, therefore, between two different perspectives or between two different types of evaluations: prospective (ex-ante) and retrospective (ex-post) evaluations.

Prospective evaluations are made before taking formal legislative decisions in order to have a better insight into the possible or potential effects of planned legislation. They may help, in particular, to choose the appropriate instruments for solving a legislative problem. Retrospective evaluations are made after the adoption of legislation, during its implementation or, in some cases, especially if the norms are limited in time, shortly before or after the end of validity of legal norms. Their purpose is to know better what happens after the entry into force of legislation11 and to apprehend the real effects of legislative action.

This distinction is important, in particular because the tools, the methods and techniques used for evaluating the effects differ between the two types of evaluation. Prospective evaluation, the evaluation of the possible effects of draft legislation, relies at least partly on other methods and techniques than retrospective evaluation (see below). The difference should not be exaggerated however, even from this methodological point of view.

As a matter of fact, from the practical point of view of a legist, the two types or perspectives of evaluation are largely complementary. Prospective efforts to assess methodically the possible effects of draft legislation facilitate considerably retrospective evaluation; the more explicit and differentiated prospective evaluation is, the easier it is to get reliable information retrospectively because it furthers the awareness for existing deficits of knowledge and gives an impulse for taking the necessary measures in order to fill up these deficits in time. Maybe legislation can even provide for these measures and guarantee that the necessary data are collected. And on the other hand, good retrospective evaluation helps to create a more solid basis for prospective evaluation, because it produces useful data for comparisons or analogies and favours the development of theoretical insights that permit a more accurate prognosis of the effects. For these reasons, the same attention has to be paid to both types of evaluation and they have to be developed simultaneously.

4. Experimental legislation

A special type to be mentioned in this context is experimental legislation, i.e. legislation enacted for a limited period of time in order to examine whether a particular legislative measure is appropriate to realise certain goals. In other words: experimental legislation is enacted with a prospective purpose, but from the methodological point of view it asks for retrospective evaluation12.

Of course, every legislative decision has in a certain sense an experimental character; it can be seen as being part of a learning process based on trial and error. This is not what is meant by the expression "experimental legislation". Strictly speaking, experimental legislation is a form of legislation presenting some particularities: it should be limited in time; it should clearly state its purpose; it should specify the goals of the planned legislative action and indicate the criteria used to evaluate whether the instruments introduced on a temporary basis are appropriate; it should specify the data that have to be collected and define the responsibilities for collecting these data and for assessing the results, etc.

Relatively simple examples of such experimental legislation can be found in the field of traffic legislation, in particular norms relating to speed limits. A recent and very instructive Swiss example is the legislation enacted some years ago to test the effects of the medical distribution of heroin to persons addicted to hard drugs13. Based on the convincing results of this evaluation, Switzerland has in the meantime decided to introduce definitely the experimental measures. These measures being very controversial politically, it was essential to make sure that the evaluation led to reliable results.

5. Evaluation criteria
Evaluation is concerned with the effects of legislation. They are its main object; an object that presents different aspects, that can be considered from various points of view and assessed according to different criteria. The most frequently mentioned evaluation criteria are the following three14: effectiveness, efficacy and efficiency.

By effectiveness we mean the extent to which the observable attitudes and behaviours of the target population (individuals, enterprises, public officials in charge of the implementation or enforcement of legislation) correspond and are due to the normative model, i.e. to the attitudes and behaviours prescribed by the legislator. Thus, with regard to this criterion, two questions have to be asked: Is the norm respected or implemented? And can the correspondence between the observable degree of respect or implementation be imputed to the norm or not? For some types of norms, in particular for orders, injunctions or prohibitions, the word "compliance" may be used at least for the first question. For others, such as norms concerning the granting of permits or subsidies by state authorities, the word most frequently used is "implementation". From a traditional legal perspective, it is the effectiveness of legislation that is very clearly in the centre of interest.

By efficacy we mean the extent to which legislative action achieves its goal. This criterion is particularly interesting from a political point of view and it shows also how important it is to clearly define the goal or the goals of a legislative decision: if the political authorities, the legislator, renounces defining the goals, it is impossible to assess the efficacy of legislation. This does not necessarily mean that the goals have to be explicitly mentioned in the normative act itself. They may also be expressed in a report accompanying a legislative proposal or formulated during the parliamentary debate. Without such a politically "authorised" definition of the goals, evaluators have to define themselves what they consider to be the relevant goals of a particular piece of legislation.

Effectiveness is a condition, but not necessarily a guarantee for the efficacy of legislation: it contributes to the efficacy only if the causal assumptions implicitly underlying the legislator's choice of the instruments are accurate. On the other hand, the simple fact that the goals of a particular legislation are realised does not necessarily prove the efficacy of this legislation; the realisation of the goals may result from other factors.

By efficiency we mean the relation between the "costs" and the "benefits" of legislative action15. The words "costs" and "benefits" are used in this context in a very broad sense. The first includes not only the direct financial consequences resulting from the compliance with respectively the implementation of legal norms; it takes also into account immaterial elements, such as psychological or emotional inconveniences, and even all negative effects caused by a particular legislation. The second refers mainly to the goals of a particular legislative action: all effects which are compatible with these goals or further their realisation can be considered as benefits. Evaluating the efficiency of legislation means therefore considering, on the one hand, its costs and on the other hand the extent to which its goals are achieved, facilitating in this way the choice between various measures or the judgement on the proportionality of different measures. The assessment of the costs and benefits is made primarily from the point of view of the legislator but, of course, it can take into account other points of view, especially the point of view of the target population of the legal norms. In this sense, we may distinguish also between internal and external costs and benefits.

These three criteria (effectiveness, efficacy and efficiency) highlight some aspects of the effects of legislation; they emphasise aspects that are particularly important in the law-making process; but, of course, they do not elucidate exhaustively the complex causal relations existing in social reality; they do not necessarily comprehend all relevant effects of a particular legislative action, sometimes designated by the word impact.

A practical example may help to illustrate this fact: In most countries, the drivers and even other passengers of cars have to wear safety belts. The legal obligation is supposed to reduce the injuries in case of an accident. The effectiveness of this norm seems to vary considerably. At least in some countries or areas a relatively high proportion of drivers wear the safety belt and they do so, partly at least, because a legal obligation exists (effectiveness). There is evidence, that wearing the safety belt contributes to reduce injuries (efficacy). And the costs of this measure (technical equipment on the cars, implementation activities of public authorities; in a broader sense also the physical and psychological constraints for the drivers and passengers) are not disproportionate compared with the extent to which this measure contributes to reduce injuries (efficiency). Nevertheless, there seems to be some evidence that the measure produces also negative effects, that could lead to more sceptical conclusions. In particular, wearing the safety belt seems to change the drivers’ attitudes or capacities in an adverse manner, resulting in more accidents and particularly in more serious injuries caused to cyclists and pedestrians. Therefore, the evaluation of the overall impact of this measure probably has to admit some nuances.

6. Evaluation tools and methods
Evaluating the effects of legislation means, in a first step, developing sensible assumptions, hypotheses, concerning the - potential or real - causal connections between legal norms and observable attitudes, behaviours and situations and, in a second step, testing the validity of these assumptions by using all the relevant experience, information and knowledge that is available or can be made available in time and with a reasonable effort.

From the legists' point of view, one of the most important questions is, of course, how can evaluations be practically made. Which are the practical tools, the methods and techniques that can be used to evaluate the effects of legislation? The answer to this questions varies according to, among other elements, the particularities of every case, the specific criteria that are to be considered, the perspective (ex ante or ex post) and the degree of plausibility and reliability of the results. Evaluating prospectively the effects of a change in fiscal legislation on private saving does not call for the same techniques as evaluating retrospectively the effectiveness of newly introduced security standards in the field of construction or of the efficacy of new procedural rights in favour of disabled persons with regard to the principle of equal treatment. An evaluation aimed at having a scientific or quasi-scientific character needs an investment that is incomparably higher than an evaluation aimed at completing an intuitive or impressionistic political assessment by a systematic use of practical knowledge and experience available within the responsible administrative service.

The best and most reliable retrospective evaluations use the different qualitative and quantitative methods and techniques familiar in the field of social sciences: interviews, observation, text analysis, synchronic and diachronic statistical comparisons between target populations and populations not exposed to legal change, etc.16 Prospective evaluations rely on practical tests, demonstration programs, modelling, simulations, forecasting, systems analysis, scenario building, etc.17. In other words: for prospective as well as for retrospective evaluation there is a large variety of very sophisticated tools that can be used separately or, better, in combination ("triangulation") in order to get the most reliable and accurate results.

However, we have to be aware, first, that even these sophisticated tools have their methodological limits: they can rarely give absolute certitudes about causal connections, but they undoubtedly sharpen the legists' and the law-makers' sensibility to this crucial aspect of legislative activity and they may, to some extent, reduce uncertainty and contribute, in this way, to improving the substantial quality of legislation. And, second, we have to be aware also of the practical limits that exist in the legislative process: applying these sophisticated tools needs time, money and personal resources that are not always available in practice.

Legists must not necessarily themselves be able to use these sophisticated tools, but they should at least know in which situations it would be possible and helpful to use them. And for practical reasons, the legists often have to choose more modest tools; tools which are perhaps more appropriate to developing causal assumptions than to testing them, but which are nevertheless very helpful. Two such tools merit mention, particularly in this context: the typology of the effects of legislation and the graphic presentation of the causal connections. Both can be used as auxiliary instruments.

The typology of the effects of legislation distinguishes different categories of effects, for example intentional and non-intentional effects, expected and unexpected effects, positive (beneficial) and negative (adverse) effects, direct and indirect effects, immediate and tardy effects, visible and symbolic effects, anticipatory and retro-acting effects, etc. These distinctions are able to sharpen the legists' sensibility and awareness of the complexity and the multiple dimensions of causal relations in social reality. They have a mainly heuristic function.

The graphic presentation of the causal links between a legal norm and the observable social reality is a way of visualising the causal hypothesis. It helps to critically reconsider and examine these hypotheses. Of course, the graphic modelling inevitably reduces the complexity of the potential or real causal connections. In practice, however, it helps considerably to become aware of certain weaknesses and shortcomings of the hypotheses that have been developed or of the analysis of causal connections. The analysis of the system of relevant social and institutional actors, the goals of the particular legislation and the fields or domains of activity concerned by the legislative action are useful starting points for establishing such a graph.

7. The institutionalisation of evaluation

In the last three decades, evaluation has become quite an important element in the legislative practice of many countries. To some extent it has been institutionalised18. As for the form of institutionalisation, we may distinguish between procedural measures on the one hand and organisational measures on the other19.

Procedural measures are for example evaluation clauses (obligation to make prospective and/or retrospective evaluations) or obligations to produce periodic reports.20 A particular form is the so-called "sunset legislation" propagated for some time especially in some States of the USA. Sunset laws are limited in time and they can only be renewed on the basis of a report confirming their necessity and appropriateness.

Organisational measures concern the creation of special organs or services responsible for the evaluation of legislation. Such organs or services may be created within the different ministries (decentralised solution), in one particular ministry (centralised solution) , within the parliamentary services or as autonomous bodies (e.g. audit office or court). They may themselves evaluate the effects of legislation or commission external specialists.

Institutionalising evaluation presents certain advantages: it makes it easier to take into account methodological aspects and requirements already in the stage of preparation of the legislation; it guarantees that the necessary financial and personal resources are available when needed; it favours an optimal synchronisation of the evaluation with the legislative decision-making process and optimises in this way the integration of the results of an evaluation in this process; it may facilitate the collaboration of the administrative bodies responsible for the implementation of legislation and in particular the access to or the availability of relevant data; it may lead to an impartial, more objective approach; and, finally, it gives greater legitimacy and therefore more political weight to the results of evaluations.

8. Final remarks

Evaluation of legislation, be it prospective or retrospective, is an indispensable element of any methodical, rational approach to law making. These methodical approaches are all based on an essentially instrumental view of legislation. Evaluation tries to contribute - and, to some extent at least, practically succeeds in contributing - to the material quality of legislation as an instrument of social change. In this sense, evaluation is a technocratic tool. As such, it can and should be used in a reasonable way, i.e. pragmatically and selectively, with the necessary awareness of its methodical and practical limits. Evaluation is not a panacea. It would be wrong, therefore, to practise it generally, as an imperative part of legistic routine. As it has been stated by the French sociologist of law more than twenty years ago, the rationalisation of legislative choices becomes unreasonable when it exceeds certain limits of cost and time; beyond a certain threshold, the arbitrariness of the legislator becomes reasonable21.

Reasonably practised, however, evaluation is more than a technocratic tool. From a legal or juridical point of view, the improvement of our knowledge of the effects of legislative action may give a new sense and a new importance to general legal or constitutional principles such as the principle of equality before the law, protection against arbitrariness and, in particular, the principle of proportionality. In this sense, it contributes also to the evolution of traditional dogmatic concepts and to the development of legal theory.

Last but not least, we should not underestimate the political dimension of evaluations. On the one hand, they strengthen the legislator's responsibility for the results of its decisions and improve, in this way, the democratic functioning of the political institutions. And, on the other hand, they create the foundations for a critical - not only technocratic - approach to legislation. As a matter of fact, elucidating the gap between the legislative goals or intentions and the results really achieved may give impulses for the necessary adaptation of legal norms; but it may also procure arguments for a critical examination of the political institutions. It would be wrong, therefore, to consider evaluation just as a tool of social engineering. Its critical function is as important as its technocratic function.

(...)


1 Peter Noll, Gesetzgebungslehre, 1973.

2 See e.g. W. Voermans / Philip Eijlander, Working out the fundamentals for a border crossing post-instrumental doctrine on legis-prudence, in: Charles-Albert Morand (ed.), Légistique formelle et materielle - Formal and Material Legistic, 1999, pp. 47ss; see also Luzius Mader, Legislative Procedure and the Quality of Legislation, in: Ulrich Karpen / Paul Delnoy (eds.), Contributions to the Methodology of the Creation of Written Law, 1996; p. 62; the term "legisprudence" has first been used - in French - by Luc Wintgens: see Luc Wintgens, Création d'une banque de données en "légisprudence", in: Gesetzgebung heute 1992, vol. 3, p.91.

3 For the distinction between formal and material legistics see Charles-Albert Morand, Eléments de légistique formelle et matérielle, in: Charles- Albert Morand (ed.), Fn 2, pp. 17ss.

4 "Drafting" is used here in a narrow sense. In countries belonging to the common law system, legislative drafting generally includes linguistic aspects, the writing of normative texts, as well as other aspects of formal legistic; for the draftsman's functions and the drafting process see David R. Miers / Alan C. Page, Legislation, 1990 (second edition), in particular pp. 51ss. Unlike in Great Britain and in countries influenced by the British tradition in this field, in continental European Countries legislation is usually drafted not by special drafting services but by the administrative units responsible for the content of policies or legislations.

5 Legislative procedure and legislative process are closely interrelated. See in this context also Michael Zander, The law-making process, 1994 (fourth edition).

6 This is particularly true for the field of criminology and, to a lesser degree, also for various fields of private law. In the field of public law, however, the main impulse for the development of evaluation came from the political sciences, not from the sociology of law.

7 See in particular Charles-Albert Morand / Jean-Louis Bergel (ed.), Evaluation législative et lois expérimentales, 1993; Faculté de droit et de science politique d'Aix-Marseille (ed.), L'évaluation législative, in: Cahiers de méthodologie juridique No 9 / Revue de la recherche juridique - Droit prospectif 1994/4; Charles-Albert Morand, Formes et fonctions de l'évaluation législative, in LeGes 1999/2, pp. 79ss; Luzius Mader, L'évaluation législative - Pour une analyse empirique des effets de la législation, 1985.

8 Terminological diversity is, of course, also due to the fact that evaluative approaches have been developed in different academic disciplines and in different national contexts.

9 For this definition and the following comments see for example also the report prepared by a working party set up by the Swiss Ministry of Justice: Département fédéral de justice et police, Mieux connaître les effets de l'action étatique: problèmes, possibilités, propositions (Rapport final du groupe de travail "Evaluation législative"), 1991, Berne.

10 For the evaluation of public policies see e.g. Werner Bussmann / Ulrich Klöti / Peter Knoepfel (ed.), Einführung in die Politikevaluation, 1997; see also Katia Horber-Papazian, Evaluation des politiques publiques en Suisse, 1990.

11 See the subtitle of a classic in the field of implementation and evaluation studies: Eugen Bardach, The Implementation Game - What Happens After a Bill Becomes a Law, 1977.

12 For some recent examples of experimental legislation in Switzerland see Luzius Mader, La législation expérimentale en Suisse, in Charles-Albert Morand / Jean-Louis Bergel, ed. (Fn. 7), pp. 221ss.

13 See Ordonnance du 21 octobre 1992 sur l'évaluation de projets visant à prévenir la toxicomanie et à améliorer les conditions de vie des toxicomanes, Recueil systématique des lois fédérales 812.121.5.

14 The terminology may vary considerably; it is important, however, to clearly distinguish the different concepts. Political scientists interested in the evaluation of public policies distinguish in this context often between "output", "impact" and "outcome"; this not only in Anglo-Saxon but also in French or German literature, "output" meaning in particular the implementation activities induced by a public policy, "impact" standing for behavioural changes of the target population of a public policy and "outcome" for the consequences of these changes. With regard to the goals of the public policy see, for example, Werner Bussmann et al., ed. (Fn 10), pp. 72s. However, these three concepts do not fully coincide with the concepts "effectiveness", "efficacy" and "efficiency".

15 See in this context the papers presented at the conference "Cost-Benefit Analysis: Legal, Economic and Philosophical Perspectives", held at the University of Chicago Law School on September 17-18, 1999; the papers are published in the Journal of Legal Studies, vol. XXIX (June 2000); see in particular Richard A. Posner, Cost-Benefit Analysis: Definition, Justification, and Comment on Conference Papers, in The Journal of Legal Studies, vol. XXIX (June 2000), pp. 1153ss. See also Bruno Binder et al, Berechnung und Abschätzung der Folgekosten von Gesetzen in Österreich, 1999; this study concerns in particular the prospective evaluation of the efficiency of legislation in Austria, but it also takes into account experiences and practices in other European countries as well as in the European Union.

16 For a practical example see Reinhard Giese und Peter Runde, Wirkungsmodell für die empirische Bestimmung von Gesetzeswirkungen - Modellansatz und Anwendung im Rahmen einer Wirkungsanalyse zur Pflegeversicherung, Zeitschrift für Rechtssoziologie 199/1, pp. 14ss.

17 For a good overview see United States General Accounting Office, Prospective Evaluation Methods: The Prospective Evaluation Synthesis, 1990; see also Werner Bussmann, Die Methodik der prospektiven Gesetzesevaluation, in LeGes 1997/3, pp. 109ss; for a practical example see Carl Böhret/ Götz Konzendorf, Rechtsoptimierung mittels Gesetzesfolgenabschätzung: Waldgesetz Rheinland-Pfalz, 1998.

18 The new Federal Constitution of the Swiss Confederation of 18 April 1999, provides in art. 170 that "The Federal Parliament shall ensure that the efficacy of measures taken by the Confederation is evaluated". In spite of its rather narrow wording, this provision is interpreted as a general obligation to ensure the evaluation of the effects of legislation; it does not mean, however, that Parliament itself has to evaluate the effects of legislation. For a general overview concerning the present situation and the future of evaluation in Switzerland see Thomas Widmer et al., Stand und Aussichten der Evaluation beim Bund, 2001.

19 See for example Ingo Unkelbach, Die Institutionalisierung der Gesetzesfolgenabschätzung auf Landesebene, 1998; this author examines the different possibilities of institutionalising evaluation in the German Länder.

20 In Switzerland, today nearly 50 Federal Statutes and ordinances enacted by the government contain such evaluation clauses, one of them, art. 43 of the Federal Statute on the procedure of the Federal Parliament, concerning specifically prospective evaluation.

21 Jean Carbonnier, Sociologie juridique, 1978, p. 412.