I shall approach macroeconomic methodology in this lecture from a different angle. The first four lectures were grounded in problems that arise in a concrete case in applied macroeconomics – Pissarides's model of the labor market. The point was to insure that some issues of practical relevance to practicing macroeconomists were addressed. One might think of this as a bottom-up methodology. It corresponds more or less to the vision articulated by Alexander Rosenberg (recall Lecture 1) in which a methodology is related to the particular state of the scientific endeavor and necessarily changes as the state of scientific knowledge itself changes.

The questions we addressed using this approach were: Is it correct to regard theories (or models) as composites of economic laws that govern the behavior of data? Can models be regarded as idealizations? And, if so, how do idealizations connect to reality? What is the relationship between theories (or models) more closely connected to data at the macro level and those that are regarded as somehow more basic at the micro level? What is the nature of the causal structure in those models, and how can real observations be brought to shed light on it?

These questions are genuine ones that arise for macroeconomists, whether or not they think of themselves as doing methodology. These questions are natural ones and are thought to have a special status as methodological problems only when people see that they generalize beyond the particular contexts in which they are first noticed.