���������������������������������������������������������������
�������������������������
The functions of knowledge synthesis
and unknowns management provide a detailed understanding of the complex problem under consideration and therefore a strong base for developing appropriate policy . Models are also used more directly as instruments of policy support and policy change in several ways . All types of models can be used in various communication roles , such as stimulating new options or explaining a decision . Mathematical and computer models are particularly effective for forecasting and comparing options . These functions all occur in the Use phase and are more familiar as suitable roles for modeling .
Forecasting takes a particular set of assumptions that reflect best knowledge about how the system operates and extrapolates to the future , for example , estimating next year ’ s inflation rate . This provides a description about the likely future situation and is important for planning where decisions must be made in advance . The effectiveness of forecasting depends on whether the model captures enough about the target system so that the model ’ s behavior tracks how the system would behave in similar circumstances .
For more complex situations , forecasting can help to identify potential problems before they develop and motivate further investigation or broaden the range of possible options considered to prevent such problems from occurring . Alternatively , forecasting can help identify opportunities and motivate action to take advantage of those opportunities . More generally , multiple scenarios or hypothetical situations can be modeled to provide ‘ what if ?’ analysis . A scenario is defined by a set of model assumptions and parameters that can reflect not just best knowledge but also possible situations or policy options . Of course , scenarios can be developed without a model , through a thought experiment or discussion . The challenge is to develop scenarios that are plausible , particularly in complex situations with multiple changes or subtle influences that make extrapolation difficult . A model provides a structured approach that captures the relationships between different parts of the system , which ensures that an assumed change to one part of the system has a realistic impact on the other parts . While planning typically focuses on more probable scenarios , exploration of diverse possible futures that are less probable can stimulate new ideas ( Cole , 2001 ; Mahmoud et al ., 2009 ).
With a mathematical or computer model , scenarios can be used to compare policy options . Even where pilot programs or other experimental processes are feasible , using models to compare options avoids the time , expense , practical difficulties , and ethical dilemmas associated with trying out different options in real life . Scenarios can be tested before implementing changes to assess , which are best , enabling a decision maker to understand the expected result of choosing a particular option . It may be that ‘ best ’ is not well defined , with scenarios suggesting that options have advantages against some assessment criteria or only under certain conditions , but disadvantages against others . In such situations , using a model can improve policy advice by focusing consideration on a small set of preferred options and describing the trade-offs involved in the decision .
Scenarios can also be created to reflect different preferred options from different perspectives . This may show that a proposed policy option is not viable or that disagreements are unimportant , as they do not affect which option is better , helping to reduce conflict and move past entrenched
72