Real Business Cycles (RBC) theory views cycles as arising in frictionless perfectly competitive economies with generally complete markets subject to real shocks (random changes in technology or productivity), it makes the argument that cycles are consistent with competitive general equilibrium environments in which all agents are rational maximizers.

Contrary to what Keynesian, Monetarist, and new classical economicsts believed, RBC theorists, starting with Nelson and Plosser in 1982, found that the hypothesis that GDP growth follows a random walk cannot be rejected. They argued that most of the changes in GDP were permanent, and that output growth would not revert to an underlying trend following a shock.

"In this case, the observed fluctuations in GNP are fluctuations in the natural (trend) rate of output, not deviations of output from a smooth determinist trend." (Snowdon)

Typically RBC models have the following features:

  • They use a representative agent framework, thereby avoiding aggregation problems.
  • Firms and households have explicit objective (utility) functions that they maximize subject to budget and technology constraints.
  • Cycles are created by exogenous productivity shocks (impulse mechanism), which are amplified by propagation mechanisms such as intertemporal substitution, consumption smoothing, investment lag, or inventory building. Kydland and Prescott’s time-to-build model, for example, assumes that it takes 4 quarters to build capital. They furthermore employ a fatigue effect, which incorporates into the model that more labor in t = 1 will lead to higher preference for leisure in t = 2.
  • Their basic assumptions are rational expectations, perfect (competitive) markets, and perfect information.


In the history of economic thought, a process of elimination led to the ascendance of RBC theory in the literatue on business cycles. Essentially, the success of the Rational Expectations hypothesis -- or, more broadly stated, the idea that economic agents do not make systematic mistakes -- was severely damaging to other business cycle theories.

RBC theorists argued that any models attempting to explain business cycles must account for three stylized facts:

  • Persistence: Cycles must not be instantaneous, but must last for some period of time.
  • Comovement: Different sectors or industries within the economy must experience the effects of the cycle concurrently.
  • Changes in the labor supply or employment.

RBC models were generally successful in accounting for persistence and comovement, but less successful in offering convincing explanations for fluctuations in employmen

Anomalies and Modifications[]

Employment Variability puzzle[]

Employment (total hours worked) empirically correlates highly with output and is strongly procyclical, however wages are only slightly procyclical. Shocks that are caused by changes in productivity alter the marginal product of labor, which results in a shift in labor demand. The labor supply curve should therefore be upward sloping. With low wage elasticity of labor (i.e. labor supply curve vertical) most of the shock to productivity would be born by wages and not employment. For RBC theory to work the labor supply curve would therefore have to be almost flat reflecting a high degree of elasticity of labor with respect to real wages.

Critics mention that there is no empirical evidence for high variability in employment (high elasticity). John Kennan (1988) however argues that life cycle models, which are used to estimate wage elasticity of labor, are not fit to capture responses to temporary wage changes within periods of less than a year. And that if short term responses to wage changes were to be measured a higher employment variability could be observed.

Garry Hansen (1985) employs a constraint to generate unemployment in the model, which forces workers to work for a specified amount of time or not at all to. In this case the marginal utility from leisure is fixed and the marginal benefit from working cannot be brought into equality every period. Agents will work as much as possible in times of higher wages (non-convexity in labor supply). This generates a rise in the variability of hours worked.

Productivity Puzzle[]

In RBC models employment and productivity real good are highly correlated because productivity drives output. However empirically employment is countercyclical. <<How can employment be empirically countercyclical here but empirically procyclical above?>>

Accounting for the following factors would help reduce the correlation beween productivity and employment in the model:

  • Shifts in labor supply curve due to nominal wage stickiness
  • Home production
  • Labor hoarding
  • Taste shocks
  • Shocks to government spending (rise in government spending financed through taxes results in a negative wealth effect that shifts labor supply)


Technology is Cobb Douglas in most RBC models, which implies that labors share of income is constant over the cycle. Empirically however it moves counter cyclical.


If cycles are caused by shocks to productivity the correlation between output and productivity should be high. However empirically the correlation between output and productivity is rather low.


There is no unemployment in RBC models, onmporal substitution between labor and leisure because of representative agent framework.

  • Labor market search (Dale Mortenson 1990, David Andolfatto 1994)
  • Risk-sharing contracts (two generations), nominal wage stickiness or minimum wage



Research by Friedman and Schwartz (1963) had supported the accepted stylized fact that money and output exhibit strong evidence of causality running from money to output. Barro (1993) made the argument however that the observed correlation could be pointing to a reverse causation between money and output. This argument was used as evidence for the endogeneity of money and RBC models were developed that endogenised money (King and Plosser 1984).


  • Government expenditure follows an exogenous stochastic process
  • Change in government expenditure don’t seem to contribute much output volatility in some models they may even reduce variability because government expenditure is negatively correlated with the Solow residual (Mary Finn, forthcoming) link.

Open Economy[]

  • RBC models can account for the following: Balance of trade moves countercyclically and trade balance is positively correlated with the terms of trade – the relative price of exports to imports (Backus, Kehoe, Kydland 1994)
  • Savings and investment are positively correlated (Finn 1990)

Imperfect Competition[]

Introducing imperfect competition lowers the ability of productivity shocks to explain output movements is dramatically reduced. Can even reduce employment because the wealth effect of a productivity shock is greater and offsets the substitution effect to a greater extend (Rotemberg and Woodford 1994).


  1. There is no evidence for productivity shocks
    1. Long and Plosser 1983: Multi sector approach – productivity shocks affect individual sectors. Problem: to get similar variation in overall output productivity shocks to individual sectors need to be even larger than shock in simple model.
    2. Interpret shocks as shocks to marginal efficiency of capital – more Keynesian (animal spirits) – which would affect new capital more then existing capital.
    3. Changes in energy prices have historically preceded cycles, but energy prices are fundamentally different than productivity shocks in that they only represent a move along the production function instead of a shift in the function
    4. Changes in terms of trade as shocks – Mendoza 1992- can account for even more fluctuation than productivity shocks empirically
  2. Not subject to econometric testing, no measure of performance
  3. Cannot account for periodicity of cycles – generally weak propagation mechanism
    1. Kydland and Prescott fatigue effect and time-to-build model does not account for enough of a propagation mechanism
    2. Cogley and Nason 1993 examine various models for their ability to generate sufficient output movements from shocks and only home production and labor hoarding can generate serial correlation in output endogenously (for labor hoarding this effect is actually negative).
  4. Cannot account for recessions for this would require negative productivity shocks
    1. Regulatory changes can depress innovation and cause negative cycle. Beauracracy (Hansen and Prescott 1993) “It would not be surprising then, that changes in the legal and regulatory system within a country often induce negative as well as positive changes in technology.”
    2. Louis Corriveau (1994): race to innovate – ex ante allocation of resources to either production or innovation; large allocation to innovation that fails could be negative shock.
  5. Representative agent framework limits the ability to apply RBC to welfare and policy issues
    1. Alan Kirman (1992): Strong attack on representative agent framework; introducing a small amount of agent heterogeneity can have destructive consequences

RBC theorists “Deny either that (involuntary) unemployment exists or that money matters. (the fact that monetary policy is ineffective is of little moment, since in any case the economy is, in this view, efficient, with resources being fully used.) This school of thought focuses on the second problem, that of economic volatility, and proposes exogenous technology shocks as the source of that volatility. The most telling criticisms of this view is the difficulty it has explaining the large negative shocks that mark recession: was there a loss in technological competence?” (Greenwald and Stiglitz 1993 p. 40)


  1. Micro foundations: individual decisions at the micro level lead to output movements at macro level through aggregation (Lucas, 1987)
  2. Explains cycles as real phenomena – money doesn’t matter

Empirical Evidence[]

- Nelson and Plosser (1982) – output data is not stationary with a trend but rather nonstationary and follows a unit root this can be explained by the predominance of real shocks in the economy. - Real wages are procyclical if the cycle is dominated by supply disturbances. They are countercyclical if the cycle is dominated by demand disturbances. No evidence for RBC. - No systematic relationship between prices and output determinable.

More Research by:[]

  • George Akerlof
  • Janet Yellen
  • Olivier Blanchard
  • Gregory Mankiw
  • Edmund Phelps
  • David Romer
  • Joseph Stiglitz
  • Ben Bernake


Government intervention might be worthwhile if the equilibrium situation is not Pareto-optimal to begin with, but the existence of cycles alone doesn’t justify government intervention.

Must be supplemented by other shock sources to be able to explain larger and more persistent fluctuations, so how important are productivity shocks?


  1. “Real Business Cycles” – George Stadler, Journal of Economic Literature, December 1994
  2. Snowdon, Brian, "A Modern Guide to Macroeconomics", 1994.
  3. Lecture by Tyler Cowen, fall 2005.
  • Greenwald, Bruce and Joseph Stiglitz: “New and Old Keynesians” ‘’The Journal of Economic Perspectives’’ Vol. 7, No. 1. (Winter, 1993), pp. 23-44