We have been pioneers in the application of Participatory Systems Mapping (PSM) to appraisal and evaluation. Participatory systems mapping is a method in which a group of stakeholders collaboratively develop a simple causal map of an issue during a workshop. Stakeholders produce a map made up of factors, which can represent anything as long as they are expressed as variables (i.e., they can in some sense go up and down) and links, which represent causal relationships. The map is intended to represent what stakeholders believe to be the causal structure of their system. The process of building a map can be hugely valuable to participants, the digitized version of the map can be a useful resource, and additional analyses can be conducted on the map created.
PSM is especially relevant to the formulation of and evaluation of complex policies, i.e., ones in which there are many influential factors, effects are non-linear and/or there are feedbacks and tipping points. In conjunction with our work on PSM as a methodology, we have developed ideas of actionable complexity, i.e., finding policy interventions that are effective despite complexity challenges, and methods for policy development in the face of deep uncertainty.
We have developed software for constructing PSM online in real-time with groups of stakeholders.
A system map can also be used as a precursor to the creation of a Theory of Change. Moreover, formulating a Theory of Change graphically as a system map makes it easier to adapt it as one learns more about the domain and as the external environment alters. The ‘participatory’ element of PSM is well suited to supporting developmental evaluation, as well as the more traditional formative and summative evaluation styles, and PSM can be a valuable ingredient in agile management approaches, where one is wanting to ‘steer’ rather than command and control a policy area.
PSM leads to system maps that can be analysed using network analysis tools, and we have developed a suite of methods that use such tools to identify critical factors, factors that are vulnerable to shocks, factors that are central to the influence network and so on. Our methods respect the fact that in many cases, quantitative data on cause and effect is not available or not feasible to collect. System maps can also be used as a precursor to the construction of agent-based models (ABM), which simulate scenarios using qualitative (and quantitative) rules. We have long standing expertise, dating back thirty years, in constructing ABM, with recent experience of applying ABM to public policy.
These methods, when applied to a whole system, have the ability to identify interactions between policies and to pinpoint unanticipated consequences and secondary impacts that deserve particular attention. They are therefore especially useful when one is considering cross-cutting policies that may impinge on several policy teams or even across organisations.