"When James Tobin, a Nobel Prize winner in economics, was on the US Council of Economic Advisors some years ago, I asked him if he used complicated macroeconometric models in his work for the Council. He responded that he didn’t because he and others could not understand the workings and output of such models and thus did not have much confidence in them. When I asked what he actually did use to analyse problems, he remarked that he used a simple multiplier–accelerator model on the back of an envelope and even though he didn’t have too much confidence in it, he said at least he understood what he was doing."