Although an academic treatment, “The Politics of Modelling” is a must-read for those whose lives are affected by modeling results — that’s all of us. The book states that its ambition is “to contribute to a reciprocal understanding between models and society.” And it serves as an aid to better comprehend the assumptions and limitations of models, modeling and modelers.
“The Politics of Modelling” is presented in three main sections: Meeting Models, The Rules and The Rules in Practice. But the book’s foreword on “Mathematical modeling as a critical cultural enterprise” and preface on “The sciences of modeling through” are not to be missed for their spot-on observations of modeling and splendid introductions to the book.
Overall, the book counsels that modelers “should not project more certainty than their models deserve; and politicians must not be allowed to offload accountability to models of their choosing.”
Expanding on this good advice, the book notes that challenges with the development and application of models include policymaking that “refers to values, not only facts and predictions. This shifts the argument substantially: even if we can imagine having perfectly predictive models with little margin for uncertainty, still the problem of finding agreement on values in policy-making would persist. Typically, a decision-maker has to find a balance between evidence and values, though the temptation of politicians is to delegate decisions to scientists, which is the easiest way to avoid delving into the traps of value systems. … In practice, it is impossible to rely exclusively on science, and all policy decisions are inevitably a mixture of evidence and values.”
Public trust in science and many of the tools of science, like modeling used in policy decisions, is at a historic low, so gaining trust will take some effort. Explaining the assumptions and acknowledging the limitations of models should help recover a good measure of public trust. And including more of the public in the model-building process can be beneficial.
“The Politics of Modelling” stresses wide-ranging participation that includes a variety of experts and laypeople. “Real, working dialogs between the community of experts and the ‘extended peer community’ are essential in improving [the] awareness on both sides.” Including numerous participants aids in transparency and increases perspective on effective model preparation and proper application of results. Without proper oversight, “mathematical modeling may more easily advance pretenses of neutrality.”
Educating the public on the finer points of modeling can be challenging. Regardless, modelers should be able to explain in straightforward language the thinking that went into model construction, how the model makes sense of the world, how the knowns and unknowns in the model were handled, how the model has been or will be validated (to the extent possible), and how the results should be interpreted. “The Politics of Modelling” is a tremendous aid in the preparation of successful public communication on this complex topic.
This piece originally appeared at WashingtonTimes.com and has been republished here with permission.
Leave a Reply