This paper describes the technical aspects of a method of planning and budgeting known as Resources Now (RN). It is written to accompany a description of the system and its applications (Algie and Foster, 1984), and the related computer program (Algie and Com1 UK, 2009).

It should be noted that effective practical use of RN does not require any mathematical knowledge or ability, nor even any accounting know-how: only knowledge of the work budgeted for. However, mathematical knowledge is required to fully understand the technical methodology of the system. The rest of this introduction summarises the basic technical points of the system as a prelude to the fuller explanation in six further sections.

THE BASIC POINTS

RN can be applied heuristically (e.g. using only one step) or fully systematically, or at some degree of sophistication between these two points. It has been applied entirely incrementally (e.g. to specific one-off decisions among the shopping list of financial cuts options), semi¬incrementally (e.g. within an overall incremental budgeting procedure), or comprehensively (e.g. within a zero-budgeting procedure.

It can be applied at several decision levels from macro policy decisions (e.g. on overall organisational expansion)to micro decisions of individual managers and staff (e.g. expenditure decisions under a given budget head). It has been applied to every type of budget, to every kind of plan, and to numerous financial management and control decisions. It can be used as a decision-aid, a (self-instructional) training tool, an investigation or empirical research instrument, or a working system of financial management and control.

RN was developed from analysis of certain converging developments in decision theory (Raiffa, 1968), resource management (Bender, 1963), psychophysics (Whitla, 1968), financial theory (Haley and Schall, 1979), cognitive psychology (Horton and Finnegan, 1973), and epistemology (Armstrong, 1978; Polanyi, 1958) by means of field tests, workshops and decision laboratory experiments, using advanced mathematical and computer techniques. But it may be used quickly and efficiently by managers who have no computer, mathematical or accountancy knowledge, nor any special expertise other than their experience and knowledge of budgeted work. Paradoxically, to fully meet the requirements of simplicity and user-friendliness, quite sophisticated decision-making methods and computer programs based on advanced mathematical modelling and validation procedures are required.

Practising managers can treat all these technicalities as a black box: the mathematical models contain no mysteries for advanced mathematicians, the implicit financial procedures are recognisable by advanced financial planners and managers, and the computer programming protocols familiar to professional programmers (Franks, 1979). Generally, managers only require to know that validated methods are used to elicit the results. They are usually able to judge for themselves that the methods are adequate in that the results adequately incorporate their requirements and intuitive decisions at critical points. Further, many of the results may be tested in terms of how far they accurately predict measurable real-life outcomes. Each step has built-in reliable checks with which users may verify for themselves their results at each stage. However, it is relevant to note that RN is often used as the managers' introduction to the computer age, as an aid to their understanding of financial planning issues, and as a tool for developing their planning and budgetary decision skills.

In outline, the technicalities are as follows. RN constitutes an 'artifical intuition' and 'expert self-support' system based on the Priority Systems approach (Algie and Foster, 1984), and associated mathematical and psychophysical modelling theories. Decision-makers can interact with the RN computer program as they would with expert consultants. The program embodies decision criteria and rules, and applies them in the same way as expert human decision-makers in undertaking financial planning and budgetary decision-making. The program integrates 'qualitative' decisions and estimates of probabilities with whatever hard qualitative data is available. The users' decisions control the program at all points.

Using a range of tailored mathematical techniques, RN evolves 'inferences' from the users' decisions. It incorporates their interpretations of real-life situations that they mB handling, and progressively cumulates their experience for them. The system also provides a 'simulation model' of reality. It pictures and interprets the changing real-life situation for users in their own terms, and -predicts how the situation is likely to evolve. Since RN combines an 'expert support system' with a 'simulation model', decision-makers can have specified for them the probable implications and effects of their decisions on reality, and vice versa.

The Budget Framework and policies are elicited and classified using linguistic variable scaling, repertory grid and cluster analysis techniques. 'Fuzzy scaling' mathematics is used for the calculations (which include correlation measures). Priority and criteria weightings under Budget Priorities and policies are elicited by the judgement analysis method, in conjunction with magnitude estimation and other techniques if required. The judgement analysis method involves a paired comparison exercise constructed on the basis of a validated sampling plan and a standard validated scale, as qualified by self­anchoring stimulus and response scales. The geometric mean technique is used to calculate priority and criteria weightings from this exercise. The relative weightings form a ratio scale called the 'priority scale'. A trace index with a validated test statistic is used in conjunction with a comparison algebra and Spearman and Kendall techniques to obtain measures of consistency, coherence, 'comprehen­sibility' and concordance.

Financial allocations are derived from priority and criteria weights (and vice versa) by means of quadratic and dynamic programming techniques. Feedback from real- life data and/or from users' revised decisions is incorporated, by means of an iterative simulation model combined with quadratic and dynamic programming techniques. This also allows any pre-stipulated constraints to be incorporated at any point, so that users may incorporate their own rules for the program within the bounds of logic and empirically­determined data from real-life. The progressive 'hardening' of initially 'soft' data through this process is achieved by means of a set of 'fuzzy set' mathematical techniques.

A set of additional advanced procedures can be used as part of the method to successively incorporate additional variables as required to successively evolve more 'high precision' results, and to successively incorporate even more rigorous reliability checks. A summary of the methods, techniques and measures used is given in Table 1.