Many interesting macroeconomic models are either sufficiently complex that they must be solved computationally, or the questions being asked are inherently quantitative and so they should be solved computationally. The first group includes almost any empirically relevant version of the neoclassical growth model. The second group includes such basic questions as business cycle fluctuations: How well does the neoclassical growth model do in producing variation in macroaggregates (like output, consumption, investment and hours worked) that ‘look like’ those seen in the data. These are quantitative questions for which qualitative answers are insufficient. Calibration is an effective tool for imposing discipline on the choice of parameter values that arise in such models, taking what would otherwise be a numerical example into the realm of an empirically relevant exercise with parameters tightly pinned down by either long- run growth facts, or microeconomic observations. As such, calibration is a useful part of the macroeconomist’s toolkit. This chapter is concerned with measurement as it pertains to calibration. Kydland and Prescott (1982) provided the foundations for the calibration procedure; key subsequent developments have been made by Prescott (1986), Cooley and Prescott (1995) and Gomme and Rupert (2007). This chapter builds chiefly on Gomme and Rupert. Like this earlier paper, our goal is to provide a sufficiently careful and detailed description of our procedures for others to be easily able to replicate our work.
You are not authenticated to view the full text of this chapter or article.
Get access to the full article by using one of the access options below.