The Modal View of Economic Models: Theoretical, Empirical and Methodological Aspects

Economic models are highly complex objects for the philosopher of science. The reason is because these models are subject to a dangerous contradiction. On the one hand, they have been and still are one of the theoretical underpinnings used by economists to explain and predict social phenomena. On the other hand, they are harshly criticized by economists, philosophers and even the general public. In our thesis we argue that the contradiction is spurious. In order to achieve this objective, we shall advocate a specific methodological account of economic models, namely the "modal view". The modal view has been proposed for the first time by the logician and economist Steven Rappaport (1986, 1989, 1998) and has been broadly ignored in the recent literature on economic models. The central point of the modal view is the distinction between the theoretical model and the multiple applications that can be generated from it. Roughly speaking, we construe a theoretical model as a set of assumptions and definitions that describe the behaviour of hypothetical entities. The entities of a theoretical model are hypothetical to the extent that they would exist if the conditions expressed by the model were realized in the real world. Instead, an applied model consists in the interpretation of some variables and general terms of a theoretical model in terms of a real-world situation. In what follows, we shall argue that the distinction between theoretical and applied models -- the core statement of the modal view -- provides a general framework to investigate several issues concerning the nature and role of economic models. The modal view of economic models has been developed in the late 1980s. Thus, although we believe that the central claims of the modal view are still valuable, it needs some methodological refinements. In particular, we maintain that the main shortcoming of Rappaport's version of the modal view is that it does not adequately examine the mutual relationship between applied models and real-world phenomena. Therefore, it cannot account for whether and how economic models can be explanatory and testable. Hence, we shall propose a refined modal view of economic models in order to show that there are several different methods to link theoretical models and the real world -- including robustness analysis, ceteris paribus clauses and applied experimental questions. Economists use all these methods in order to assess their theoretical models and to give them empirical significance. As a matter of fact, the distinction between theoretical and applied models will prove useful to examine the link between theoretical models and the real world. In fact, using the notions of "application" and "applied model" we will show what are the similarities and the differences between explaining via models and testing a model. Following our refined modal view, we shall advocate two main statements. First, we argue that the economist's modelling practice can be divided into two levels: (i) conceptual analysis of possible economies, and (ii) application of the theoretical framework in order to explain phenomena and test the model's theoretical predictions. Second, we show that many workhorse economic models may be epistemically worthwhile, empirically testable and explanatory. In this regard, we maintain that the vast majority of criticisms directed to economic models are misleading because they are grounded on a misunderstanding about the structure and role that models have in economic theory. Let us talk about the structure of the present dissertation. The work is composed by five chapters. In chapter 1 we introduce the notion of "scientific model" and examine its role in science and the philosophy of science. Chapter 2 presents Rappaport's modal view of economic models and focuses on some of its methodological advantages. In chapter 3 we propose our refined modal view of economic models in order to take into account of whether economic models can be said to be explanatory and testable. After that, the thesis develops in two directions, a critical one and an applied one. In chapter 4 we critically discuss the modal view through a comparison with other philosophical accounts of economic models. Chapter 5 applies the modal view to a microeconomic case study, i.e. Hal Varian's model of sales. The reader can choose to read the chapters as they are presented or pick one of the two according to his interests. The conclusion summarizes the main findings, addresses a few open issues and proposes the notion of "levels of abstraction" as a key to capture the gist of economic models