MT - 05. RBC

Question 1

Lucas (1978)

Answer 1

Aim: examine stochastic behaviour of equilibrium asset price in pure exchange economy, single good.

Conclusion: time separability is a nuisance, we should add recursive, non-additive preferences with sufficient impatience.

Question 2

Barro and King (1984)

Answer 2

Aim: review time separable utility restriction in RBC models

Conclusion: time separability restricts the relative responses of leisure and consumption to changes in relative price and permanent income. Crucial in evaluating response to expectations.

Question 3

King, Plosser and Rebelo (1988)

Answer 3

Aim: analyse business cycles when driven by exogenous technological change at a constant rate

Conclusion: neoclassical view enhances our understanding of business cycles, yet some key shortcomings arise

• Substantial persistence in technology shocks is required, ie. Model lacks endogenous persistence
• Also fails to match the data in terms of amplification, the magnitude of fluctuations predicted by the model is too small.

Question 4

Stock and Watson (1991)

Answer 4

Aim: analyse the post-war US empirical relationships across the business cycle

Linear detrending –> Over pronounced cycles –> HP filter (1981) better than first differencing

Conclusion:
Strongly Pro-cyclical: consumption, investment, inventories, imports, exports, aggregate employment, productivity and capacity utilisation

Strongly counter cyclical: unemployment

Acyclical / debateable: prices and wages
o Link to Kydland and Prescott (1990) – prices are countercyclical

Procyclical: nominal interest rates

Question 5

King and Rebelo (1999)

Answer 5

Revives RBC model post critiques by embedding vairable capital utilisation.

RBC model only matches the data is the persistence parameter on technology shocks is close to 1 (i.e random walk).

Propagation mechanism from investment means a stronger external propagation mechanism than is probably valid is required to match the data: such as the AR parameter equal to 0.979 calibrated by King and Rebelo (1999)

Conclusion: RBC model with variable capital utilisation gets realistic cycles with small, non-negative technology shocks

Question 6

Rotemberg and Woodford (1996)

Answer 6

Persistence of shock often causes wage to become very pro-cyclical, which is not observed in the data.

Question 7

King, Plosser and Rebelo (2002)

Answer 7

Aim: extend the RBC model flexibly to explain three things
• Exogenous steady state growth – labour augmenting form
• Distortionary taxation
• Time varying government spending

Conclusion: Time devoted to work cannot grow as it is bounded; consumption investment and output must grow at the same rate

Question 8

Cooley and Prescott (1995)

Answer 8

Allow for population and productivity growth in RBC.

• Calibration of factor shares doenst move much in US for long period of time and is well identified for calibration.

Question 9

Kydland and Prescott (1982)

Answer 9

Seminal RBC Paper –> rekindled interest in cycles
- Approach of calibration based on micro evidence

The RBC theory of business cycles has two principles:

1. Money is of little importance in business cycles.
2. Business cycles are created by rational agents responding optimally to real (not nominal) shocks - mostly fluctuations in productivity growth

Method for compariong models rests on calibration, simulation and comparison to cycle data.

Question 10

Hall (1978, 1988)

Answer 10

– intertemporal consumption substitution evidence is weak, contemporaneous wealth has the strongest influence

Question 11

Romer (1989)

Answer 11

– inconsistent data pre and post war, great moderation meaning cycles less important and less frequent

Question 12

Hodrik-Prescott (1981)

Answer 12

HP filter

Question 13

Baxter and King (1994)

Answer 13

Bandpass filter

Question 14

Stock and Watson (1991)

Answer 14

– overview of cyclicality and in real postwar data

Question 15

Gali (1999),
Francis and Ramey (2004),
Basu et al (2006)

Answer 15

Negative relation TFP and Hours –> Reject RBC

Question 16

King and Rebelo (1999), Christiano et al (2004)

Answer 16

Correct some of the issues to resurrect the RBC

Question 17

Mankiw, Rotemberg and Summers (1985)

Answer 17

– separability of the utility function is not a good assumption on utility for matching the data

Question 18

Topic Summary (5)

• Methodology
• Predictions
• Assumptions

Answer 18

Main framework to analyse economic fluctuations firmly established DSGE framework: rational expectations, calibration, simulation and evaluation.

Revolution of RBC modelling was in three conceptual leaps
i. Efficiency of business cycles: RBC model fluctuations are fully optimal –> Stabilisation policies may not be desired

ii. Importance of technology shocks: Came from ability of RBC to generate “realistic” fluctuations when shock matched Solow residual (heavily critiqued)
iii. Limited role of monetary factors: full abstraction from monetary sector

• RBC model labour market predictions are a correlation of around 0.98

Strongly procyclical wages –> Booms driven by tech shocks raising MPL
Employment very procyclical –> Supply more when productivity is high

• Dramatically different story in the data –> Real wages almost acyclical, employment varies positively but less than in the model

Employment adjusts on extensive not intensive margin –> Model has Rep. agent and No unemployment
Could be addressed by Hansen’s lotteries trick

• RBC model lacks amplification and endogenous propagation –> Cannot generate hum shaped responses

Question 19

Seminal papers (6)

Answer 19

Kydland and Prescott (1982), Prescott (1986), Long and Plosser (1983)

Mehra and Prescott (1985), Wel (1989), Hansen and Jagannathan (1991)

Question 20

Okun (1962)

Answer 20

Okun’s law relating output and unemployment: 3% fall in GDP vs 1% rise in unemplyment…recently more like 2:1

Question 21

Lucas and Rapping (1969)

Answer 21

Intertemporal substitution in labour, mechnaism is crucial to business cycle flux in employment

• Rise is r –> Saving more attractive so work more today

Question 22

Nelson and Plosser (1972)

Answer 22

Persistence of fluctuations test to conclude fluctuations have a permanent component based on an ADF test.

Question 23

Campbell and Makiw (1987)

Answer 23

Proposed measure of persistence generally exceeds 1, ie. shocks to output generally followed by further output movements in the same direction.

Question 24

RBC Approach Critiques (4)

Answer 24

• Use of technology shocks not representated well by Solow residual measure of ignorance
• Intertemporal substitution in labour is central propagation mechanism, significant willingness to substite labour required –> Not seen in micro data
• Omission of monetary disturbances, if nominal shocks matter for the real economy then the RBC model features are not very useful
• Is calibration a valid approach? Does matching data moments mean the model is good?