Unobservability, tractability and the battle of assumptions

Publisher Website
Google Scholar
Abstract
Economic models often include unrealistic assumptions. This does not mean, however, that economists lack a concern for the truth of their assumptions. Unrealistic assumptions are frequently imposed because the effects are taken to be negligible or because the problem at hand is intractable without them. Using the Musgrave‐Mäki typology as the point of departure, these claims are defended with respect to theories proposed by Solow, Hall and Roeger concerning productivity growth and the mark‐up. Since they are unobservable, their values need to be inferred from the values of observable variables. Assumptions such as perfect competition and constant returns to scale are used for making this inference or measurement problem tractable. Other assumptions are justified in terms of negligibility. These findings support the fecundity of the (amended) Musgrave‐Mäki typology of assumptions – including the notion of a tractability assumption proposed here. Finding ways of relaxing tractability assumptions turns out to be an important source of progress in economics.