Models in Economics: Finding Simplicity in a Complex World
Models dominate economics. As Morgan put it: “these manipulable objects are the practical starting point in economic research work: they are used for theorizing, providing hypotheses and designing laboratory experiments [...] and [e]conomics teaching is similarly bounded: students learn by working through a set of models” (Morgan 2012). However, economists historically seem to prefer a specific type of models, namely highly idealized and simple models. For this reason, the practice of modelling is consistently among the most discussed topics in philosophy of economics. Topics range from the (alleged?) explanation paradox in economic models (Reiss 2012) and debates over the knowledge generated from minimal or toy models (Grüne-Yanoff 2009, Fumagalli 2016, Reutlinger et al 2017 and Ngyuen 2019) to pragmatic approaches of how to understand the generation of knowledge and reasoning with economic knowledge (Morgan and Knuuttila 2012).
During the conference, we will challenge ourselves to look into the future and find out where mod- elling in economics is heading – similar to recent debate in the philosophy of economics (Special Issue “Economic Methodology and Philosophy of Economics twenty years since the Millennium”, Journal of Economic Methodology 2021). We dare to ask: how could simulated, complex, data- driven models change the future of modelling in economics? This is a provocative and inflated question – rather than expecting a clear-cut answer, we use this question to guide us by dissecting its different elements. The question connects to several central issues within the philosophy of science, and more specifically the philosophy of models and modelling practices.
The first central and contested issue is what constitutes a good model, especially in economics. Which criteria should be applied to determine the quality of a model: precision, truth or something else entirely? Should a model be realistic, descriptive or only instrumentally useful? Should it have strong predictive powers? Is a good model one that faithfully represents reality, or is high explanatory power enough? These questions are central to discourses in the philosophy of science (and economics) and many positions can be sketched out.
The second central issue is what types of models are used in economics. This question calls for a typology of common models, like dynamic stochastic general equilibrium models, market for
lemons models and agent-based or simulated models – what characterizes these models from an epistemic and methodological point of view? And how do they relate? Such a typology of models in economics begs the question whether there are types of models which currently only exist at the margins of economic theory and whether these models will or should be of higher importance in the future.
A third issue the above question hints at is what the aim of economics and its models is and how the choice of models impacts the real world, for example through policy. Is a specific type of model better suited to develop inclusive policy? Do models contain implicit biases? What approach to modelling in economics is most sensitive, for example, to disadvantaged social groups or the environment? How do models affect policy and decision-makers?
Finally, what would more efficient policy-making need from future economists and how can philosophers contribute?
By investigating these topics, we hope to elucidate not only the provocative question but also where economics, and therefore the philosophy of economics might be headed in the future.
Join us!
Establishing a dialogue between philosophy and economics, students and staff of the P&E program at the University of Vienna will host a conference that will tackle these issues. From Thursday, September 28th, to Friday, September 29th, Vienna will be the site where experts, academics, and students will contribute to this dialogue.