This paper reviews the role of micro non-convexities in the study of business cycles. One important non-convexity arises because an individual can work only one workweek length in a given week. The implication of this non-convexity is that the aggregate intertemporal elasticity of labor supply is large and the principal margin of adjustment is in the number employed—not in the hours per person employed—as observed. The paper also reviews a business cycle model with an occasionally binding capacity constraint. This model better mimics business cycle fluctuations than the standard real business cycle model. Aggregation in the presence of micro non-convexities is key in the model.

During the downturn of 2008–2009, output and hours fell significantly while labor productivity rose. These facts have led many to conclude that there is a significant deviation between observations and current macrotheories that assume business cycles are driven, at least in part, by fluctuations in total factor productivities of firms. We show that once investment in intangible capital is included in the analysis, there is no inconsistency. Measured labor productivity rises if the fall in output is underestimated; this occurs when there are large unmeasured intangible investments. Microevidence suggests that these investments are large and cyclically important.

Recent developments in business cycle theory are reviewed. The principal finding is that the growth model, which was developed to account for the secular patterns in important economic aggregates, displays the business cycle phenomena once it incorporates the observed randomness in the rate of technological advance. The amplitudes and serial correlation properties of fluctuations in output and employment that the growth model predicts match those historically experienced in the United States. Further, the model continues to display the growth facts it was developed to explain.

The founding fathers of the Econometric Society defined econometrics to be quantitative economic theory. A vision of theirs was the use of econometrics to provide quantitative answers to business cycle questions. The realization of this dream required a number of advances in pure theory—in particular, the development of modern general equilibrium theory. The econometric problem is how to use these tools along with measurement to answer business cycles questions. In this essay, we review this econometric development and contrast it with the econometric approach that preceded it.

The neoclassical growth model studied in Kydland and Prescott [1982] is modified to permit the capital utilization rate to vary. The effect of this modification is to increase the amplitude of the aggregate fluctuations predicted by theory as the equilibrium response to technological shocks. If following Solow [1957], the changes in output not accounted for by changes in the labor and tangible capital inputs are interpreted as being the technology shocks, the statistical properties of the fluctuations in the post-war United States economy are close in magintude and nature to those predicted by theory.

The neoclassical growth model studied in Kydland and Prescott [1982] is modified to permit the capital utilization rate to vary. The effect of this modification is to increase the amplitude of the aggregate fluctuations predicted by theory as the equilibrium response to technological shocks. If following Solow [1957], the changes in output not accounted for by changes in the labor and tangible capital inputs are interpreted as being the technology shocks, the statistical properties of the fluctuations in the post-war United States economy are close in magintude and nature to those predicted by theory.

This essay reviews the development of neoclassical growth theory, a unified theory of aggregate economic phenomena that was first used to study business cycles and aggregate labor supply. Subsequently, the theory has been used to understand asset pricing, growth miracles and disasters, monetary economics, capital accounts, aggregate public finance, economic development, and foreign direct investment.

The focus of this essay is on real business cycle (RBC) methodology. Those who employ the discipline behind the methodology to address various quantitative questions come up with essentially the same answer—evidence that the theory has a life of its own, directing researchers to essentially the same conclusions when they apply its discipline. Deviations from the theory sometimes arise and remain open for a considerable period before they are resolved by better measurement and extensions of the theory. Elements of the discipline include selecting a model economy or sometimes a set of model economies. The model used to address a specific question or issue must have a consistent set of national accounts with all the accounting identities holding. In addition, the model assumptions must be consistent across applications and be consistent with micro as well as aggregate observations. Reality is complex, and any model economy used is necessarily an abstraction and therefore false. This does not mean, however, that model economies are not useful in drawing scientific inference.

The vast number of contributions made by many researchers who have used this methodology precludes reviewing them all in this essay. Instead, the contributions reviewed here are ones that illustrate methodological points or extend the applicability of neoclassical growth theory. Of particular interest will be important developments subsequent to the Cooley (1995) volume, Frontiers of Business Cycle Research. The interaction between theory and measurement is emphasized because this is the way in which hard quantitative sciences progress.

An economic experiment consists of the act of placing people in an environment desired by the experimenter, who then records the time paths of their economic behavior. Performing experiments that use actual people at the level of national economies is obviously not practical, but constructing a model economy and computing the economic behavior of the model’s people is. We refer to such experiments as computational experiments because the economic behavior of the model’s people is computed. In this essay, we specify the steps in designing a computational experiment to address some well posed quantitative question. We emphasize that the computational experiment is an econometric tool used in the task of deriving the quantitative implications of theory.

Prior to the mid-1980s, labor productivity growth was a useful barometer of the U.S. economy’s performance: it was low when the economy was depressed and high when it was booming. Since then, labor productivity has become significantly less procyclical. In the recent downturn of 2008–2009, labor productivity actually rose as GDP plummeted. These facts have motivated the development of new business cycle theories because the conventional view is that they are inconsistent with existing business cycle theory. In this paper, we analyze recent events with existing theory and find that the labor productivity puzzle is much less of a puzzle than previously thought. In light of these findings, we argue that policy agendas arising from new untested theories should be disregarded.