JAMES MARSHALL, University of Auckland
MICHAEL PETERS, University of Auckland
ROBERT SHAW, New Zealand State Services Commission
In 1984, a project was undertaken by the New Zealand State Services Commission – the administrative decision-making skills project – to produce a model of administrative decision-making and an associated teaching/learning package for use by government officers. This report describes the model of decision-making, its development, its evaluation and implementation within the New Zealand Public Service.
The State Services Commission is one of the three control departments (along with the Treasury and the Ministry of Works and Development) which is responsible for government administration in New Zealand. It has overall responsibility for the management and organization of government departments and the training and welfare of public servants. The project originated with concern over the implementation of the Official Information Act 1982 and the number of departmental decisions taken by officers that were being questioned by the Ombudsman. The Act follows similar statutes passed elsewhere and has been regarded as the most important constitutional innovation in New Zealand since the office of Ombudsman was established in 1962. Part of the Ombudsman’s responsibilities is to investigate and review the decisions of minister, government departments and organizations covered by the Act, and, also, decisions not to make available information as requested under the Official Information Act. The Case Notes issued by the Ombudsman suggested that many of the cases investigated involved decisions in the area of discretion, where departmental rules and established precedents no longer have direct application and where public servants must exercise judgment. Here moral considerations such as respect for justice, equity and cultural differences loom large. There arose a perceived need for improved decision-making in relation to the Official Information Act. For this purpose Robert Shaw was seconded to the Commission in November 1983, and James Marshall and Michael Peters of the University of Auckland were contracted to advise upon the development of the model and its associated teaching/learning package, and to provide a substantive evaluation of the total project (Peters, Marshall and Shaw 1986). The developmental project was undertaken between March and October 1984 and the course was used with public servants in 1985 and 1986.
Over almost two decades, professional administrators and academics have written of the importance of placing ‘ethics’ at the center of practical administration. Loosely called the ‘new age theorists’, some wish to make radical changes in the way government organizations operate, whilst others seek adjustments to the balance of competing concerns (see e.g. Davis 1969; Dvorin & Sims 1972; Frederickson 1981). These concerns have been repeated in New Zealand, e.g. by the chairman of the State Services Commission (Probine 1984 a & b).
If these sources provide a rationale for looking at decision-making procedures in terms of morals/values, it is not necessarily a rationale for adopting any one particular approach to decision-making. Though the expressed concerns are with the area of discretion, with values, with morality, and with making the individual the most important concern of public power, these do not entail any particular approach(es) to decision-making. Should emphasis be placed upon the psychological or sociological aspects of decision-making to achieve objectives consistent with the direction proposed? Or should one follow Davis’ proposals (1969) to extend the rule of law, thereby decreasing the areas of discretion and obtaining justice through uniformity of administration? In short, in practice: how is one to improve decision-making and to place ‘a refreshing emphasis on values’ in administrative decision-making? Furthermore, how is one to define ‘morality’?
Shaw (1982) had provided an initial step towards a philosophical model by considering values as one set of factors among others – for example, the alleged facts and the legislation potentially pertaining – which must be taken into account when a decision is made (or when reviewing a decision already made). If decisions are thought of in this manner then the approach to decision-making is in terms of factors, or logical components, which need to be considered in making decisions (cf. Wilson 1967). This philosophical type of approach would indicate not only which types of factors need to be considered but it could further provide a potential ‘checklist’ to follow when making a decision and, thereby, a systematic means of analyzing and recording the decision.
Second, if the concern is with areas of discretion and discretionary justice, then the meaning of the concepts of ‘discretion’ and ‘discretionary justice’ need to be made clear. Getting clear about these concepts (or getting clear about which concepts are being prescribed) involves, as part of the definitions, another set of logically related concepts and these in turn, it might be expected, would refer to logical components concerned with any exercise of discretionary justice.
If this approach seems overly philosophical and abstract the point must be made that the project proposal demanded that any such model be based upon actual decisions made, and reflect the experiences of public servants. This practical basis was ensured by the continual testing and refinement of the model against the experiences of the developmental project team and participants on the experimental courses.
The model of decision-making is a philosophical model then in that it depends upon key concepts, the derivation of logical components from these concepts, and in that it has a key role in questions of justification. The key concepts were derived from Davis (1969) and Rawls (1972). From these definitions it follows that if a decision is made in accordance with discretionary justice then it would have considered these factors: legality, morality, facts of the case, just procedures, possible decisions; to these were added the implications or effects of decisions, and the strategies that might be used to implement a decision. These were to become the key logical components of the model. In addition key questions were developed for each component to ensure that important aspects of any component were covered. Within each component’s key questions other subsidiary and less important questions were developed also.
Whilst essentially a logical model, there is no claim that the model is an algorithm for calculating or reading off the correct decision. The traditional logical problem of deriving an ‘ought’ from an ‘is’, it but one daunting prospect underlying that processes of any argument designed to permit the deduction of a decision, arguments on normative preference would still abound over the adoption of substantive normative and prescriptive premises. The assumptions underlying the model then were that it would not provide an easy formula for decision-making, but would provide rational reasons for decision (Toulmin 1958).
On the other hand the model, employed in Popperian ‘fashion’ (Popper 1963), provides principles of criticisms which when employed in a critical manner can throw out wrong decisions. This should not be seen as overly negative, as it should be remembered that Popper’s notion of refutation or falsification was married to the highly creative aspect of conjecture, in which scientific theories are conjectured and subjected to rigorous attempts at refutation. This aspect is encouraged within the use of the model.
Associated with the model was the teaching/learning package which evolved in practice from the consideration of case studies, in terms of the model, by the developmental project team. A set of unique educational problems was presented to the project team (Marshall & Peters (1986). These were concerned with the moral and ethical issues raised above – especially cultural differences – and the problem of educating senior public servants in an area which, by definition almost, they have considerable experience and expertise. The problems were then to present a course which was based upon the model, was practically orientated, and which could build upon the experience of course participants whilst achieving its educational aims.
The teaching/learning approach which was developed placed considerable emphasis upon the analysis of actual studies, at various points at which actual decisions had been made, in terms of the model and its components. These case studies were often cases in which the participants had, themselves, been involved as administrators. Participants were required to prepare and structure case studies – actual decisions from departmental files – in a format amenable to analysis in terms of the model. Detailed notes were issued to them (at a half-day introductory session) in a manual concerning the definition, selection and presentation of a case study. At this session they also worked through a standard case developed from the Ombudsman’s case notes. This provided them with the opportunity to evaluate their performance with those of the group and, of course, the Ombudsman.
About ten days after the introductory session participants reassembled with their prepared case studies and the three-day workshop session was devoted to working through the case studies, comparing justification, and refining their knowledge of the model under the ‘eyes’ of a trained course direct. The optimum number of participants was six.
Apart from the project team involved in the development of the model and the associated teaching learning package, a total of 50 senior administrators from 17 government departments and private sector concerns, presented an actual case documented from departmental and business files during the trials. A further 90 senior public servants participated in courses promoted by the State Services Commission in 1985; by the end of 1986, with an unprecedented demand, a further 120 people will have participated in the course and have been introduced to the model. At the targeted level of senior management, by the end of 1986, almost 20% of these senior officials will have been introduced to the model through the training courses. The State Services Commission, which is responsible for the training of public servants, views this course as one of its most successful courses to date.
Whilst provision was made for the evaluation of the developmental aspects of the project it is now a matter of some concern that more thought was not given to the implementation, post-course, of skills and knowledge acquired. The main source of information which attests to the appropriateness of the course is the very high 1986 demand for the course from the public sector.
Public servants who followed the full teaching/learning package as a course in decision-making and the trial of the model and teaching/learning package were, on completion of the course, generally positive and favourable towards the model. Many recognized it as representative of their practice, formalizing their ‘unconscious’ procedures. They saw it as a valuable systematic approach to decision-making (see Peters, Marshall and Shaw 1986). In the evaluation process there were opportunities for participants, course directors, project team members and evaluators to appraise the model and its practical efficacy. It should be noted that the project was developmental, with successive stages incorporating earlier comments and criticisms as refinements. The end product thus represents the collective experience of a large cross-section of administrators, both public and private, and academics, working together on a number of practical case studies within the developing framework of the model. Several were sufficiently impressed with the model to wish to adopt it as a basis for training within their own departments. It was generally accepted by participants that the model will assist decision-making in four main ways:
The course was designed to address problematic decisions which in fact had passed through various levels of administration, in many cases ending with the Ombudsman. In most cases the case study was chosen because historically it had presented a complexity for the department, or the official concerned. Sometimes participants chose cases with which they were involved and used the course to identify, clarify, and choose options and strategies based upon the collective wisdom of the group, to deal with their existing responsibilities.
The orientation towards practice maintained interest and motivation, because participants were required, as a part of a learning community (a theoretical concept developed as a consequence of the project – (Marshall and Peters 1985), to adopt a critical stance to every decision, irrespective of the actual status of officials involved. In practice this resulted in participants being openly critical of decisions made by the ombudsman, departmental heads and themselves – a reflection of the underlying Popperian critical stance adopted by the model. The importance of the case study is highlighted by the decision of the State Services Commission to finance a second stage of the research programme focusing on these case studies as a source of privileged data on administrative decision-making in the public service. This research will be undertaken by Michael Peters and will involve not only an analysis of ‘what went wrong’, but also how the use of the model might have obviated problems, difficulties and acknowledged errors before they surfaced.
Finally it should be remembered that an aim of the project was to get public servants ‘to place a refreshing emphasis on values’. Thus whilst the project was collaborative and developmental there was an underlying educative role vis-à-vis the place of moral and cultural factors in decision-making. With regard to these factors the summative evaluation of the project (on the trials) detected changes of attitudes, between pre-course and post-course tests, towards these factors. Given the short duration of the courses, the average age of participants (mid 40’s) and their average length of service (22 years), the small but significant change in attitudes to moral factors detected by the evaluators must be seen as significant in relation to these aims. With more careful follow-up and implementation strategies the model has considerable potential not only for improving decision-making but also for achieving this general educative aim.
Davis, Kenneth Culp. 1969 Discretionary justice: a preliminary enquiry. Baton Rouge: Louisana State University.
Dvorin, E.P. & R.H. Simons. 1972. From a moral to human bureaucracy. New York: Harper Row.
Frederickson, H. George 1981. New public administration. Birmingham: University of Alabama Press.
Marshall, James and Michael Peters. 1985. Evaluation and education: the ideal learning community, Policy Sciences 18. 263-288.
---. 1986. Evaluation and education: practical problems and theoretical prospects, New Zealand Journal of Educational Studies 21.
Peters, Michael, James D. Marshall & Robert S. Shaw. 1986. The evaluation of the development and trials of a decision-making model, Evaluation Review 10, 5-26.
Popper, Sir Karl R. 1963. Conjectures and refutations. London: Routledge and Kegan Paul.
Probine, M.C. 1984a. Accountability in the public service. Paper presented to the Civil Service Institute, Wellington, 21 June 1984.
---.1984b. Efficiency in the public service. Address to the Financial Management Course, Victoria University of Wellington, 22 November 1984.
Rawls, John. 1972. A theory of justice. Oxford: Oxford University Press.
Shaw, Robert S. 1982. Administrative discretionary justice. Paper read at New Zealand Association for Research in Education’s Annual Conference.
Toulmin, Stephen 1958. The uses of argument. Cambridge: Cambridge University Press.
Wilson, John Barry Sugerman & Noman Williams, 1967. Moral education Harmondsworth: Pelican.