Mary S. Morgan

"Of importance for the is that , in both his cycle books, had concentrated his energies on the statistical evidence of the economic interactions involved in the business cycle and the of these relationships rather than on the relation between the economic cycle and the exogenous causal factor. Moore's concern with evidence end statistical explanation compared to that of Jevons, and the matching change in contemporaries' responses, are both indicative of the development of the econometric approach by the early years of the twentieth century. Yet, it was some years before Moore's broad econometric approach to the explanation of economic cycles, involving a large number relationships linking different parts of the economy, was taken up by who produced the first macro econometric models in the late 1930s."

"From the late nineteenth century, economics gradually became a more technocratic, tool-based, science, using mathematics and statistics embedded in various kinds of analytical techniques. ... By the late twentieth century, economics had become heavily dependent on a set of reasoning tools that economists now call 's': small mathematical, statistical, graphical, diagrammatic, and even physical objects that can be manipulated in various different ways. Today, in the twenty-first century, if we go to an economics seminar, or read a learned scientific paper in that field, we find that economists write down some equations or maybe draw a diagram, and use those to develop solutions to their theoretical conundrums or to answer questions about the economic world."

"The joint work with from that research group, ' ..., is now seen as creating a new strand. The extant philosophy of science thought about s in relation to theory: models were ways of capturing the essence of a theory. What we were doing in that little research group – and what we did in the volume Models as Mediators – was to say, if you look at the way science is practised, you see that scientists treat models as autonomous objects on which they develop arguments. They manipulate them, argue with them, extend them. Models are not in a simple relationship between theory and the world, rather they are at angles to both, so you can use them to interrogate both sides. Models as Mediators is 20 years old, and you can definitely see now that the project as a whole changed the conversation in the philosophy of science about models. I don’t mean that everybody was convinced by it, but it created a big enough presence so that, even if you didn’t agree with it, you had to take it into account. This work was part of a wider move that has been happening toward ‘the philosophy of science in practice’."