College Of Business and Economics

Download Report

Transcript College Of Business and Economics

The Real Business
Cycle School
Intermediate Macroeconomics
ECON-305 Spring 2013
Professor Dalton
Boise State University
RBC Macroeconomics


Evolved out of New Classical
economics of 1970s
Major proponents




Edward Prescott (Minnesota)
Finn Kydland (Carnegie-Mellon)
Charles Plosser (Rochester)
Robert Barro (Harvard)
From New Classical to RBC


The late 1970s and early 1980s was a
time of New Classical Dominance
By the early 1980s, however, significant
doubts had arisen



“signal-extraction” problem not robust
enough to explain business cycles
Evidence supportive of monetary neutrality
of announced policy weak
Tobin suggests a “way out” that he does not
himself take seriously – random real shocks
From New Classical to RBC


Kydland and Prescott take seriously this
alternative and develop the foundations
of Real Business Cycle Theory – a theory
that accounts for cycles wholly by
changes in real supply variables.
Kydland and Prescott, “Time to Build and
Aggregate Fluctuations,” Econometrica
(November 1982)
Originally viewed as augmenting the New
Classical approach
Central Proposition
Large random fluctuations in technology
produce supply-side shocks to the
production function, generating
fluctuations in aggregate output and
employment as rational individuals
respond to the altered structure of
relative prices by changing labor
(resource) supply and consumption
(investment) decisions.
RBC Macroeconomics

Assault on all previous 20th
century macroeconomics


Booms are not “good” and
recessions are not “bad”
Recessions are not desired by
agents in the economy but they are
nonetheless unavoidable
consequences of changes in
constraints agents face
RBC Macroeconomics

Assault on all previous 20th
century macroeconomics


Agents react optimally to changes
in constraints and the resulting
aggregate fluctuations are efficient
Supply, not demand shocks, are key
to understanding economic
fluctuations
RBC v. New Classical

RBC replaces the impulse mechanism of
New Classical economics


RBC retains the propagation
mechanisms of New Classical economics


“Technology shocks” instead of “monetary
surprise”
Rational expectations and relative prices
New Classical Economics “Mark II”
Reactions of Leading New
Classicalists

Lucas
Exclusion of money in RBC a mistake
 Viewed as addition to NC models



Approves of methodology


Later says “monetary shocks just aren’t that
important”
micro-based models and use of fully-articulated
artificial economies to compare real with
experimental economics
Business cycles are “minor problem” –
shifts focus to Growth economics
Reactions of Leading New
Classicalists

Barro
RBC promising
 Monetary neutrality of New Classical
models a “mistake”
 Shifted focus to Growth economics
 Defends New Classical Achievements

Equilibrium modeling
 Rational expectations
 Dynamic policy-making and evaluation

Growth and Cycles



Development of RBC and shift of
research to Growth represent revival of
interest in “supply-side”
macroeconomics
Continuation of pre-Keynesian lines of
business cycle research
New technology influences both longrun growth as well as producing shortrun displacements (disequilibrium?)
RBC Antecedents

Dennis Robertson


Joseph Schumpeter


emphasized real forces
“Theory of Capitalist Development”
Knut Wicksell

Changes in marginal productivity of capital
(impulse mechanism) cause divergence of
“natural rate of interest” from “bank loan rate”
leading to endogenous monetary creation
(propagation mechanism), distorting time
structure of production and leading to selfreversing boom
Growth and Cycles
For Real Business Cycle
Macroeconomics,
growth and cycles are
inseparably
interrelated
History and RBC

Supply shocks of 1970s



Apparent failure of Demand-side Keynesian
model
Political emphasis of “new supply-side
economics” of Reagan Administration



Two OPEC oil increases
Tax cuts and deregulation
Renewed interest in statistical properties of
economic time-series
Seminal work of Nelson and Plosser
Cycles and Random Walks

Conventional Approach




Imagines economy evolving along a
growth path reflecting underlying trend
Fluctuations about trend due to demand
shocks
Shocks “die out” over time, so economic
time-series are “trend-reverting”
Yt = gt(Y0) + b (Y – YT)t-1 + zt
Cycles and Random Walks

Yt = gt(Y0) + b (Y – YT)t-1 + zt

At time t1, a shock of size z
occurs, but it dies out over
time and the growth path
reverts to the trend

Conventional approach
consistent with the natural
rate hypothesis
(unanticipated changes in
monetary growth produce
temporary deviations from
YN)
Y
time
Cycles and Random Walks
Nelson and Plosser, “Trends and
Random Walks in Macroeconomic
Time Series: Some Evidence and
Implications,” Journal of Monetary
Economics (September 1982)


Most changes in GDP are permanent, with
no tendency for Y to revert to former trend
GDP follows a random walk process with
drift
Cycles and Random Walks

Nelson-Plosser Approach



Value in one period still dependent on
previous value of the variable, but
shocks (z) change output permanently
Rather than g being underlying growth
rate, g is “rate of drift;” b has value of 1
– “unit root” hypothesis
Yt = gt(Y0) + (Y – YT)t-1 + zt
Cycles and Random Walks

Yt = gt(Y0) + (Y – YT)t-1 + zt

At time t1, a shock of size z
occurs and it permanently
changes the growth path of
the economy
Y
time
Implications of Nelson-Plosser




Observed fluctuations are fluctuations in the trend,
not deviations from a trend.
In NC world, permanent changes in GNP growth
cannot occur from monetary shocks since money is
neutral; therefore main forces causing instability
must be real shocks.
If shocks to productivity growth are frequent and
random, path of Y follows a random walk that
resembles the business cycle.
No distinction between trend and cycle, so theory
of growth and fluctuations must be integrated.
Productivity Shocks





Unfavorable changes in the physical
environment that adversely affect
agricultural output
Significant changes in price of energy
War, political upheaval and labor unrest
Government regulations
Changes in the quality and quantity of
capital and labor; new management
techniques; new products; new production
techniques =>Technological change
RBC Models: Common Features
(1) Representative agent models; agents
maximize s.t. constraints
(2) Agents form Ratex; signal-extraction
problem re permanent v. temporary
productivity shocks
(3) Continuous market-clearing
(4) Exogenous productivity shocks are
impulse mechanism for output and
employment fluctuations
RBC Models: Common Features
(5) Propagation mechanisms vary,
include consumption smoothing,
“time-to-build,” and intertemporal
labor substitution
(6) Fluctuations in employment
voluntary; labor and leisure highly
substitutable over time
(7) Money is neutral
(8) No distinction between SR and LR
Changes from New Classicalism



Impulse factor – productivity shocks
replace monetary shocks
Abandon price level/relative price
misperception emphasis
Abandon long run/short run
distinction
RBC: Model Structure

Production Function


Technology Evolution Parameter


0<þ<1
Ut = f(Ct, Let)
Resource Constraints


At+1 = þAt + єt+1
Representative Agent Utility Function


Yt = At F(Kt, Lt)
Ct + It ≤ Yt ; Lt + Let ≤ 1
Capital Stock Accumulation Equation

Kt+1 = (1-∂) Kt + It
Technology Shocks and
Employment


Technology shocks change A, shifting
the production function upward; the
demand for labor curve will also shift
upward.
Increased labor demand will increase
the real wage and employment.
How much? Depends on supply elasticity!
 If labor supply is inelastic?
 If labor supply is elastic?

Technology Shocks and
Employment



Stylized facts of Business Cycles
indicate small procyclical variations in
real wage are associated with large
procyclical variations in employment
Is labor supply highly elastic or
inelastic with respect to real wage?
What does that indicate about the
intertemporal substitution of labor for
leisure?
RBC and Lucas-Rapping ASH

Lucas-Rapping: in making laborsupply decisions, workers consider
future and current C and Le
Substitution and income effects of
changed real wage
 Temporary v. permanent changes in real
wages

RBC and Lucas-Rapping ASH

RBC
 temporary
technology shocks will
lead to temporary changes in real
wages; no income effect and large
supply response
 Permanent technology shocks will
lead to permanent changes in real
wages; large income effect and
small supply response
Labor Supply and Interest Rates



Change in the real interest rate affects
labor supply by altering relative price
of income earned today v. future
Ls = Ls(W/P, r)
Intertemporal price-ratio
(1+r) (W/Pt)
(W/Pt+1)
 Increase in r increases labor supply;
reduction in r reduces labor supply
RBC AD and AS

r
RAS
LM/P

IS (RAD)
Y


An IS-LM model conforming
to Ratex, Continuous Market
Clearing, and Fullinformation Ms
Output and employment due
to real forces; RAS
determined by production
function and labor supply
Tech improvement shifts RAS
to right and LM/P adjusts so
full employment exists
Problem with model: Labor
supply not dependent on r
RBC AD and AS

r
RAS
re
RAD
Ye

Y
If labor supply is
dependent on r, an
increase in r
increases labor
supply and increases
output
The RAS curve is
positively sloped
Characteristics of Model






Model entirely real (M and P have no
impact on Y or L)
No LR-SR distinction
RAS traces out labor market equilibria
r equilibrates goods market
Shifts in RAS lead to variations in Y
and L
Temporary variations in RAD can
cause Y and L variations
Technology Shock: RAD-RAS
Y

Y
Y*
Y=Y
b
b
Y
a
a
b
w2
w1
Y
SL2 (r2)
w

r
SL1 (r1)
r1
DL2
RAS1
a
r2
a

RAS2
b

RAD
DL1
L 1 L2
L

Y1 Y2
Y
Begin in equilibrium.
Labor market clears at
real wage w1 for given
production function Y.
At current r1, RAS and
RAD clear at Y1.
Favorable productivity
shock increases A and
production function
increases to Y*.
RAS increases, driving
down the interest rate.
The lower interest rate
lowers the supply of
labor and favorable
productivity shock
increases labor
demand.
Labor market
equilibrium moves to
b, employment
increases and output
increases at the lower
interest rate.
Expenditure Shock: RAD-RAS
Y

Y
b
Y=Y
Y
b
a
a

Y
SL1 (r1)
w
r
SL2 (r2)
w1
w2
r2
r1
a
RAS
b
a

b
RAD2
DL1
L 1 L2
RAD1
L

Y1 Y2
Y
Begin in equilibrium.
Labor market clears at
real wage w1 for given
production function Y.
At current r1, RAS and
RAD clear at Y1.
Increase in government
purchases shifts RAD
to the right, increasing
the real interest rate.
The higher interest
rate increases the
supply of labor and
reduces the real wage
rate.
Labor market
equilibrium moves to
b, employment
increases and output
increases at the higher
interest rate.
Temporary v. Permanent Shocks



In the previous model, wealth effects were ignored.
If shocks are permanent, wealth effects can’t be
ignored. When permanent shocks occur, induced
changes in the real wage also will led to additional
changes in RAD.
A change in technology will cause RAS and RAD to
move in the same direction.


A positive technology shock that raises RAS raises the real
wage, increases real income and increases RAD.
A change in expenditures will moderate the change
in RAD.

An increase in government purchases that reduces the real
wage reduces real income and decreases RAD.
Temporary v. Permanent Shocks

r
RAS

RAS2
re
r2


RAD RAD2
Ye
Y2
Y
A positive technology shock
increases RAS to RAS2.
If the shock is temporary, the
wealth effect is small and
RAD increases by a small
amount.
The real interest rate falls as
higher output is achieved.
This does not change the
prediction of the model
which ignores wealth effects.
Temporary v. Permanent Shocks

r
RAS

RAS2
re
RAD
Ye
Y2
RAD2
Y


A positive technology shock
increases RAS to RAS2.
If the shock is permanent,
the wealth effect on
expenditures will be large
and increase RAD by roughly
the same amount as the
increase in Y.
Out put increases but the
interest rate remains
approximately the same.
This prediction of the model
is different than that which
ignores wealth effects.
“Testing” RBC Models



Kydland and Prescott were first to show that
RBC models could generate time-series data
that possessed statistical properties similar
to actual US business fluctuations.
RBC theorists generally have not attempted
to provide models capable of econometric
testing.
Instead, RBC theorists have developed the
method of calibration to test their models.
Calibration Method
(1) Construct RBC equilibrium model
(2) Provide specific functional forms
(3) Calibrate the model
- simulate random shocks with
computer generated random
numbers
(4) Trace out key macroeconomic
variables from exercise and compare
with actual time-series
Calibration Method


Such exercises are able to mimic the
actual economy with respect to
important time-series data and
replicate the stylized facts of business
fluctuations
Problem: How to choose between
competing models? No criteria
equivalent to significance testing in
econometrics to answer such a
question.
RBC and Money



Accepted stylized fact: Positive correlation
between money and output.
Generally accepted (Friedman and Schwartz)
by Keynesians, Monetarists and New
Classicalists that changes in monetary
growth cause changes in real output
growth.
In RBC models, money is “super” neutral.
How do RBC models account
for the accepted stylized fact?
RBC and Money


Caveat: Positive correlation between money
and output may indicate that money
responds to output.
But then why does it look like monetary
growth comes before output growth?



Expectations of future output growth may lead
to increases in money demand that increase the
quantity of money supplied.
Bank money (demand deposits) is endogenous;
bank money can be produced faster than real
output.
Money supply changes before output but
output changes cause money supply
changes.
RBC and Money

RBC theorists divided into two camps

Kydland and Prescott, “Business Cycles:
Real Facts and the Monetary Myth,” FRB
Minneapolis Quarterly Review (Spring
1990)


Denial of stylized fact that money leads the
cycle
Plosser, “Understanding Real Business
Cycles,” Journal of Economic Perspectives
(Summer 1989)

Role of money remains an “open question”
Measuring Technology Shocks


How does one measure technological
progress?
“Solow residual”

That part of ∆Y that can’t be explained by
∆K or ∆L
Y = A F (K, L)
 Y = A KβL1-β where 0 < β < 1
 ∆Y/Y = ∆A/A + β ∆K/K + (1- β) ∆L/L
 ∆A/A = ∆Y/Y – [β ∆K/K + (1- β) ∆L/L]

Measuring Technology Shocks


Prescott (“Theory Ahead of Business
Cycle Measurement”) suggested
(∆A/A) is a random walk with drift plus
serially uncorrelated error
Plosser (“Understanding Real Business
Cycles”) uses (∆A/A) and (∆Y/Y) to
show that aggregate fluctuations in Y
are mainly due to fluctuations in
technology
Measuring Technology Shocks
The Stylized Facts



RBC literature led to a renewed effort
to discover and measure the stylized
facts of business fluctuations
Forced a re-evaluation of existing
theories in light of the new data
Central controversies over
Real wages
 Price level

Real Wages and Business
Cycles


Are real wages pro-cyclical or counter-cyclical?
Orthodox Keynesianism and Orthodox Monetarism



RBC



Real wages are counter-cyclical
Changes in AD with sticky or lagging wages (due to
adaptive expectations)
Real wages are strongly pro-cyclical
Changes in technology shift production function and
change the demand for labor
Empirics


Real wages are slightly pro-cyclical
Problem: procyclical wages require elastic labor supply to
produce observed variations in employment and output,
but micro data does not support notion of elastic labor
supply
Price Level and Business
Cycles


Is the price level pro-cyclical or counter-cyclical?
Orthodox Keynesianism, Monetarism, New Classical


RBC


Price level and inflation are pro-cyclical
Evidence from entire 1954-89 period is that price level and
inflation are counter-cyclical
Empirics and Impulse Mechanisms



Impulse determines behavior of price level and inflation
Supply-side changes lead to counter-cyclical prices
Demand-side changes lead to pro-cyclical prices
Policy Implications
(1) More robust case against activism
“Costly efforts at stabilization are likely to be
counter-productive. Economic fluctuations
are optimal responses to uncertainty in the
rate of technological progress.”
- Prescott, “Theory Ahead of Business Cycle Measurement”
(2) Fiscal policy is more potent than monetary
policy but should still be avoided.
RBC and AD Management
Aggregate demand management has
been successful in reducing the
volatility of business fluctuations from
demand-side disturbances compared
to earlier periods; technological
disturbances have emerged as a
dominant source of modern business
fluctuations as a consequence.
- Chatterjee, “Real Business Cycles: A Legacy of Countercyclical Policies”
Criticisms of RBC Theory
(1) Evidence concerning labor supply elasticity
weak
(2) Technology shocks are directly
unobservable


Doubts concerning size and frequency to
produce large variations in Y and L
Doubts that technological regression occurs to
produce recessions
(3) Pro-cyclical Solow residual due to other
reasons

Pro-cyclical utilization rates of labor and capital
Criticisms of RBC Theory
(4) Is large unemployment really voluntary?

pro-cyclical movement of vacancy rates and
voluntary quits
(5) Is money neutral in the short-run?

American and British dis-inflations of the
1980s
(6) Persistence (Unit-root hypothesis or lack of
trend-reversion)


AD changes can produce permanent effects if it
induces technological change
Hysteresis effects
Criticisms of RBC Theory
(7) No micro-economic foundation for
technological change and innovation

Plausible models of demand conditions, R&D
expenditures and “learning by doing” effects
(8) Representative agent models sidestep
rather than address aggregation and
coordination problems
(9) Lack of robust empirical testing
RBC: An Assessment

RBC refocused attention on what we actually
know or don’t know about business
fluctuations




Output does not appear to be trend-reverting,
but rather follows a random walk with drift
Re-integration of growth theory and theory
of fluctuations
Furthered cause of building macro-models
with micro-foundations
Renewal of interest concerning role of
supply-side in macroeconomics