Document Type

Conference Proceeding


Modelling and Simulation Society of Australia and New Zealand


Faculty of Business and Law


School of Law and Justice




This article was originally published as: Syme, G. J., Bennett, D., Macpherson, D.K., & Thomas, J.F. (2011). Guidelines for policy modellers - 30 years on: New tricks or old dogs?. Paper presented at the MODSIM 2011: 19th International Congress on Modelling and Simulation. Perth, Western Australia. Original article available here


Systems thinking and the use of modelling to assist in natural resources decision making started in the early 1960’s (Forrester, 1961). In the early 1970’s a multidisciplinary program was established in CSIRO’s Division of Land Resources Management to, among other issues, research and develop models to assist in the resolution of persistent and multi-faceted environmental issues. The authors of this paper were members of that group. We recently decided to look at what has been learned about these reasonably intractable problems over the more than 30 years period since. If it is true that the half life of modelling papers is around eight years and, apparently, decreasing (Haggett, 2005), then we can expect that the wheel has often been reinvented. So we describe issues as they were expressed up to the early 1980’s and compare and contrast them with concerns after the turn of the century in two contexts - the development of systems thinking in modelling and deciding when and how to model. Three general criticisms of systems analysis were evident then: (1) the models were highly mathematical and therefore opaque to the user, and it was hard even for ’experts’ to explain unexpected findings; (2) welfare, as expressed by economic or utilitarian criteria, was considered to be poorly represented in systems practice (systems thinking was being viewed here as rational, while it was slowly being recognized that apparently irrational behaviour is an important part of the social fabric and could not be ignored (Glazer, 1978)); (3) the systems approach relied on lists derived from planners, but in the absence of “systems thinking” guidelines they were only lists of what planners thought of doing, not comprehensive in any strong sense. These three major drawbacks still exist and in much the same form today. Development of systems thinking is now much more widespread. The applications gap does not appear to be filled, though there have been attempts to broaden the client base by participative model construction to reach a consensus of what is agreeable to most stakeholders. “Irrational behaviour” still threatens systems analysis and comes in several forms. There is lack of acceptance that utilitarian criteria should be used to drive the solution. And the consequences of implementing a model’s solution may lead to unforeseen consequences and behaviours by and to affected parties. Very few papers have been written in recent years on guidelines on when to model and when not to do so. There seems to be no connection with earlier publications. The field does not appear to have progressed. We perceive a world in which those who can, model, and those who can’t still feel little need to try. It is to be hoped that interest in the issues of when and why to model will not wane. It would be a shame, perhaps a source of conflict, if modelling remained yet another marker that separates a ‘them’ from an ‘us’. Some of us have now been much closer to decision-makers. Our observations include: (a) Timing is critical in decision-making. Sometimes there are a few hours for deliberation, for others decades. A modeller has to be close to help with the quick ones. Advice that comes too late will not be relevant. (b) Close contact with decision makers is essential, as is their training. (c) A contract from the client is a desirable/necessary precondition. (d) Trust, and lack of it, will always be important in accepting advice. (e) Committees, especially those charged with doing something and feel that they cannot progress without assistance, are good targets. But committees, like modelling contracts, can be solely methods of delaying decisions. (f) Modelling physical and biological systems is less contentious than using utility to guide advice.