Business Cycles, Real
Business Cycles, Real
It is accepted knowledge in empirical macroeconomics that aggregate data are typically characterized by trend and cyclical components. A natural by-product of this is a dichotomy of (empirical) macroeconomics into those who concentrate on growth and those whose main concern is with the cyclical aspects of the data. Although this entry focuses exclusively on the cyclical side of the dichotomy, the business cycle and growth processes are not mutually exclusive features of macroeconomics. That is, one cannot draw any meaningful conclusion about one feature of the data without a comprehensive understanding of the other. It is invariably the case that empirical results hinge on the methodology used to isolate the trend and cyclical components of the data.
What drives business cycles is an important issue for economists and policymakers alike. In his 1994 article “Shocks,” John Cochrane reiterated R. E. Lucas’s 1977 assertion that “business cycles are all alike” (p. 296) in that each cycle exhibits co-movements among macroeconomic variables that are so remarkably similar that the cycles are more likely to be driven by a common force and less so by a composite of several shocks. To that end macroeconomic theories have been written suggesting one shock or another that could possibly explain this cyclical phenomenon.
With abundant research devoted to real business cycles (RBC), the search for that one driving force is still ongoing, and from the early-twenty-first-century state of the literature little progress has been made. (Even though the literature does not seem to be converging on one particular shock the search process has resulted in the development of solution techniques that have contributed greatly to macroeconomics and other related literature. For example, researchers have witnessed the birth of dynamic general equilibrium models and the use of dynamic programming to solve such models.) The literature has presented a gamut of candidate shocks responsible for business cycle movements; shocks to individual preferences and tastes (fads), oil price shocks (e.g., OPEC crises in the 1970s and the Gulf War in 1990), monetary policy shocks, government spending and tax shocks, and technology shocks (shocks that shift the production possibility frontier of a nation). All these shocks, with the exception of technology shocks, have been discredited on grounds of either failing to qualitatively match actual business cycle movements or failing to explain a sizeable portion of the forecast error variance in output. That is, they fail to match the data on quantitative grounds. However, since 1999, empirical studies have begun to question the role of technology shocks in the real business cycle. A host of studies finds results that contradict the technology-driven real business cycle hypothesis. However, Finn Kydland and Edward Prescott received the Nobel Prize in economics in 2004 “for their contribution to dynamic macroeconomics: the time consistency of economic policy and the driving forces to business cycles.” It is obvious, therefore, that there is indeed some benefit to real business cycle research.
The remainder of this entry provides a brief history of business cycle research and the transition of the literature over the years. This account of the history and transition of the literature to the present will be deliberately brief and could be considered an injustice without any objections from the author of this entry. However, this allows the entry to devote more time to the recent developments in real business cycle research. In the end this entry provides suggestions as to where real business cycle theory should be heading.
Business cycle research examines the periodic upswings and downswings in macroeconomic activity that is a feature of industrialized nations. The notion that business cycles are driven by real factors—termed the real business cycle hypothesis—has itself experienced periods of high and low activity over the years. Prior to the Great Depression real theories was the cornerstone of macroeconomics. The onset of the Great Depression then turned the tide in favor of monetary theory. In fact, monetary policy, more accurately monetary policy errors, was believed to be the impetus behind the devastation of the 1930s and it was widely believed at the time that corrective monetary policy and/or more active fiscal policy were the most likely candidates to turn around the economy.
Since the late twentieth century, real shocks have been the focus of macroeconomists and since the 1970s and early 1980s real business cycle research can be characterized as a period in which candidate shocks began to be eliminated for one reason or another. For example, shocks to government policy fail to deliver the requisite co-movements among macro variables by predicting a fall in household consumption in time of economic expansion. Changes in capital and labor income taxes occur too infrequently to make them serious candidates for business cycle fluctuations. Oil price shocks, even with volatile energy prices are too small a fraction of value added in production to have any sizeable effect. Monetary policy shocks, even with the addition of some type of economic friction, have small effects. However, research conducted in the 2000s has shown that monetary policy used in tandem with technology shocks can produce responses very similar to those observed in the post–World War II U.S. economy.
Technology shock has been the only survivor of this elimination process and thus the literature has accepted it as the main driving force of business cycles. The argument is that when there is a positive technology shock, peoples’ marginal productivity increases, causing real wages to rise; workers will work more hours and output subsequently rises. The opposite holds when the technology shock is negative. This is a pattern economists have observed for industrialized economies since the beginning of the nineteenth century.
Kydland and Prescott (1982) have convincingly argued that a dynamic stochastic general equilibrium model driven by technology shocks can mimic the main statistical features of U.S. macroeconomic time series when calibrated using means and variances of macroeconomic data (using reasonable parameter values to fit the real world) from the U.S. economy. An oft-quoted statement used in support of technology shocks attributed to Prescott (1986) can be found on page 7 of Rebelo (2005), asserting that such shocks “account for more that half the fluctuations in the postwar period with a best point estimate near 75%.” This conclusion has received its fair share of criticisms but somehow the technology-driven paradigm has addressed any serious criticism hitherto levied against it: the theory survived by adding new features such as labor hoarding, indivisible labor, or capital utilization, which make technology shocks less volatile (thus reducing the probability of technological regress) and enable the theory to better match microeconomic measures of labor supply elasticities.
However, since 1999 a new wave of attacks on the real business cycle hypothesis appears to have delivered the biggest blow yet. These recent studies have found that positive technology shocks, identified using an econometric technique known as structural vector autoregressions, are contractionary on the part of labor input, contrary to business cycle experiences. Contemporary research conducted by Neville Francis and Valerie A. Ramey (2005a) has concluded that the technology-driven real business cycle hypothesis appears dead. The reaction to such a bold statement has been swift and numerous, giving the literature new life. Real business cycle research has reemerged to the forefront of macroeconomics since the mid-1990s.
In 1999 Jordi Galí published an article in the American Economic Review that empirically challenged the notion that all factors of production should respond positively to positive technological innovations. Using the econometric technique, structural vector autoregression, Galí identified as technology that shock which has positive effects on labor productivity in the long run. Such identifying assumption is a feature of most RBC models. Using this technique to identify technology shock leads to a fall in per capita hours and not a rise as predicted by the standard RBC theory and as evidenced from actual business cycles experiences. Therefore, the empirical prediction for the role of technology shocks in driving business cycles contradicts the underlying theory. In this regard the literature seems to have undergone a paradigm shift and the search for a new theory of business cycle, which matches with the empirics, should be under way.
However, proponents of the technology-driven RBC paradigm have not taken this blow lightly and have tried to save one of the hitherto cornerstones of macroeconomics. Their criticisms of Galí’s approach are twofold. First, they state that the manner in which the trend in hours per capita is removed is erroneous. In fact they assume that per capita hours being a bounded series should not be treated as having a long run trend and thus no attempt of removing any such trend should be made. If one agrees with this suggestion one will recover, using said structural vector autoregressions, the result from standard theory that all factor inputs rise after a technological innovation—here is an example of where the method used to remove trend has implication for the cyclical results as stressed in the opening of this entry. The second point they make is that the tool used in the empirical search, structural vector autoregression, is inappropriate and introduces bias into the results.
In their work in 2004 and 2005, Francis and Ramey address the “trend versus no trend” issue in per capita hours for the United States, both for the post–World War II era and the late twentieth century. They suggest that in order to align the theory to the data there are certain demographic features of the data that are absent from the theory, for example, trends in schooling, aging of the population, and the composition of workers into private and public enterprises. Since the real business cycle theory does not address these issues and they factor so prominently in the data, one might consider removing them from the data. Their removal will make the theory and empirics more compatible. The prediction of labor hours after this exercise is not beneficial to supporters of the technology-driven real business cycle hypothesis as per capita hours again respond negatively to a positive technological innovation.
The search for the driving force of business cycles is ongoing. While researchers have been unsuccessful in finding the candidate shock responsible for this phenomenon they have made some strides in macroeconomics along the way. In particular, research efforts have developed tools like dynamic programming that should figure prominently in economics for some time to come. They have also gone beyond static models and have begun to write models in more dynamic settings. Certainly, this is a practice that is good for the profession, though some researchers have taken it to the extreme by being so mathematical in their approach they have neglected the economics.
The driving force(s) of business cycles is an interesting concept and it will definitely continue. Some have suggested that researchers should first refine the existing model to account for the demographics that are absent from the theory. They maintain that it should not be too much of a stretch to incorporate schooling, aging, and employment composition in the commonly used models of the early twenty-first century. Finally, there are perhaps other shocks out there that have not yet been eliminated. For example, in 1997 Jeremy Greenwood, Zvi Hercowitz, and Per Krusell suggested technology shocks that work through investment goods; however, as of 2006 no consensus has been reached on the plausibility of these types of shocks.
SEE ALSO Business Cycles, Empirical Literature; Business Cycles, Political; Policy, Monetary
Barro, R. J., and R. G. King. 1984. Time-Separable Preferences and Intertemporal Substitution Models of Business Cycles. Quarterly Journal of Economics 99: 817–839.
Chriastiano, L., and M. Eichenbaum. 1992. Liquidity Effects and the Monetary Transmission Mechanism. American Economic Review 82: 346–353.
Cooley, T., and G. Hansen. 1989. The Inflation Tax in a Real Business Cycle Model. American Economic Review 79: 733–748.
Cochrane, John. 1994. Shocks. Carnegie-Rochester Conference Series on Public Policy 41: 295–364.
Francis, Neville, and Valerie A. Ramey. 2004. The Source of Historical Economic Fluctuations: An Analysis using Long-Run Restrictions. NBER working paper 10631.
Francis, Neville, and Valerie A. Ramey. 2005a. Is the Technology-Driven Real Business Cycle Hypothesis Dead? Shocks and Aggregate Fluctuations Revisited. Journal of Monetary Economics 52 (8): 1379–1399.
Francis, Neville, and Valerie A. Ramey. 2005b. Measures of Hours Per Capita and Their Implications for the Technology-Hours Debate. NBER working paper 11694.
Galí, Jordi. 1999. Technology, Employment, and the Business Cycle: Do Technology Shocks Explain Aggregate Fluctuations. American Economic Review 89: 249–271.
Greenwood, Jeremy, Zvi Hercowitz, and Per Krusell. 1997. Long-Run Implications of Investment-Specific Technological Change. American Economic Review 87 (3) (June): 342–362.
Kydland, F., and E. Prescott. 1982. Time to Build and Aggregate Fluctuations. Econometrica 50: 1345–1370.
Lucas, R. E., Jr. 1977. Understanding Business Cycles. In Stabilization of the Domestic and International Economy. Vol. 5 of Carnegie-Rochester Conference Series on Public Policy, ed. K. Brunner and A. H. Meltzer. Amsterdam: North Holland Company.
Prescott, Edward. 1986. Theory Ahead of Business-Cycle Measurement. Federal Reserve Bank of Minneapolis Quarterly Review 10: 9–22.
Rebelo, Sergio. 2005. Real Business Cycle Models: Past, Present and Future. NBER Working Paper 11401.