As beginning economics graduate students soon discover, most of their time will be spent learning about model building. As James Heckman observes, ‘Just as the ancient Hebrews were “the people of the book”, economists are “the people of the model”’ [2000, p. 46]. But, students will also discover that there will be little explicit talk about how best to go about building models. Instead, they are seemingly expected to learn inductively by example. Over the last fifty or sixty years of educating modern economists, most things have not changed much. However, what one is supposed to think constitutes a model and what one is supposed to think a model can or should do has changed dramatically. Originally, the idea was that common principles of economics could be illustrated with physical models, but later it was that they could be represented by mathematical models. Representative models were spearheaded mostly by the work of Jan Tinbergen in the 1930s. In the last two or three decades, the representative notion of models seems to have been lost. Today, models are objects of research on their own, without explicit recognition of any underlying principles that someone might wish to discuss without reference to a specific model. Stated another way, models today are easily viewed as simple instruments to be used to measure the economy and hopefully to learn about the economy in a trial-and-error manner. Unfortunately, by viewing models as mere instruments, students have been cut off from the origins of ideas underlying the models and hence from an opportunity to learn about those ideas.