Lecture 3: Real Business Cycles

Tiago Cavalcanti1

1 University of Cambridge

E200: Macroeconomics
Cambridge
Michaelmas 2022
Main reading:

Romer, (2011, ch. 5)

Introduction
Definition:

I Burns and Mitchell, “Measuring Business Cycles” (1946,

National Bureau of Economic Research):

“Business cycles are a type of fluctuation found in the aggregate

economic activity of nations that organize their work mainly in
business enterprises: a cycle consists of expansions occurring at
about the same time in many economic activities, followed by
similarly general recessions, contractions, and revivals which
merge into the expansion phase of the next cycle.”

I To identify cycles, Burns and Mitchell assume that they are no

shorter than 6 quarters, and found a maximum length of 32
quarters.
A Business Cycle
U.S. Business Cycles (from the NBER) (in month)
Log of the US Real GDP, 1957q1-2010q2. Source:
International Financial Statistics (IMF).
Cyclical component of the Log of the US Real GDP,
1957q1-2010q2. Source: International Financial Statistics
(IMF).
US unemployment rate, 1957q1-2010q2. Source:
International Financial Statistics (IMF).
10

8
Unemployment rate (US)

3
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Year
Business Cycles Regularities

Lucas’ Definition: (Business Cycles) Recurrent fluctuations of output

about trend and the comovement among other aggregate time series
(consumption, employment, investment).

Business Cycles are characterized by a set of statistics:

I Volatilities of time series (standard deviations);

I Comovements of time series (correlations, serial correlations).

Log of the UK Real GDP, 1957q1-2013q1. Source:
International Financial Statistics (IMF).
Cyclical component of the Log of the UK Real GDP,
1957q1-2013q1. Source: International Financial Statistics
(IMF).
Cyclical component of the Log of the UK Real GDP and
Consumption, 1957q1-2013q1.

Observation: Consumption includes households expenditures on

durable and non-durable goods. Usually, consumption of non-durable
goods is much less volatile
Cyclical component of the Log of the UK Real GDP,
Consumption and Investment, 1957q1-2013q1.
Business cycle facts: Quarterly US data, 1947-1997, Stock
and Watson (1999)
Some stylized facts

1. Consumption (of nondurables and services) fluctuates much less

than output.

2. Investment fluctuates much more than output.

3. The capital stock fluctuates much less than output and is largely
uncorrelated with output.

4. Productivity is slightly procyclical but varies considerably less

than output.

5. Wages vary less than productivity.

Some stylized facts

6. Employment fluctuates almost as much as output and hours of

work, while average weekly hours fluctuate considerably less.

7. Government expenditures are essentially uncorrelated with

output.
Shocks and Propagation Mechanisms:

1. Shocks

I Technology shocks: internet; new energy source;

I Weather shocks and natural disasters.
I Monetary and Fiscal policy shocks;
I Political shocks;
I Expectations shocks.

2. Propagation mechanisms
I Consumption/investment decision;
I Labour decisions;
I External finance premium; net worth; Default
Overview of Business Cycle Theory

1. Keynesian
I Fluctuations due to aggregate demand (monetary and fiscal
shocks, ‘animal spirits’).
I Key ingredient: nominal rigidities.
I Aggregate relations without micro-foundations.

2. Real Business Cycles

I Fluctuations due to aggregate supply (technology shocks).
I Key ingredients: flexible prices, perfect markets.
I General equilibrium with optimising behaviour.
Getting Intuition: Static Labour/Leisure Choice
Consider the following optimisation problem:

c1−σ (1 − l)1−σ
max +γ , σ, γ > 0
c,l 1−σ 1−σ
Subject to c = wl.

c1−σ (1 − l)1−σ
L= +γ + λ[wl − c],
1−σ 1−σ
FOCs:

c−σ − λ = 0,

−γ(1 − l)−σ + wλ = 0,
 σ
c
γ = w.
1−l
Getting Intuition: Static Labour/Leisure Choice

Combining first-order condition with budget constraint implies:

 1
w σ 1
(1 − l) = wl ⇒ l = 1 σ−1 .
γ 1 + γσw σ

1 σ−1
σ−1 σ
dl w γ w σ
l,w = = − σ 1 σ−1 R 0 if σ R 1.
dw l 1 + γσw σ
I When σ → 1, then labor supply is independent of w.

I Why? When individuals have no wealth and with logarithmic

utility, income and substitution effects cancel out.
Getting Intuition: Intertemporal labour/leisure choice

Consider the following optimisation problem:

max {ln(c1 ) + b ln(1 − l1 ) + β[ln(c2 ) + b ln(1 − l2 )]},

c1 ,c2 ,l1 ,l2

Subject to
c2 w2 l2
c1 + = w1 l1 + .
1+r 1+r

L = ln(c1 ) + b ln(1 − l1 ) + β[ln(c2 ) + b ln(1 − l2 )] + ...

w2 l2 c2
+λ[w1 l1 + − c1 − ].
1+r 1+r
Getting Intuition: Static Labour/Leisure Choice
FOCs:

b b w2
= w1 λ and β = λ
1 − l1 1 − l2 1+r

Therefore:

1 − l2 w1
= β(1 + r) .
1 − l1 w2

I Implications: w1
w2 ↑ ⇒ work more today relative to tomorrow;

I Implications: r ↑ ⇒ work more today relative to tomorrow.

Point: Shocks can be amplified by households decision to work and
to invest.
Baseline Real-Business-Cycle model

Assumptions:

I large number of identical, price-taking firms;

I large number of identical, price-taking infinitely-lived

households;
I inputs in production are: capital (K), labour (L), and
“technology” (A);
I output is divided among consumption (C), investment (I), and
government purchases (G);
I a fraction δ of capital depreciates each period.
The Household Sector

The representative household maximises the expected value of

∞
X
Ut = Et β j u(ct+j , 1 − lt+j )Nt+j , β ∈ (0, 1)
j=0

Population grows exogenously at rate n:

Nt+1 = (1 + n)Nt .

Time endowment is normalised to 1.

Households own the capital and rent them to firms.

Household optimisation

Assume households own capital. What is their budget constraint?

Nt ct + It = wt lt Nt + rtK Kt , and Kt+1 = (1 − δ)Kt + It

Nt ct + Kt+1 = wt lt Nt + (1 + rtK − δ)Kt .

Divide both sides by Nt and let (1 + rt ) = (1 + rtK − δ) (rt = rtK − δ)
Kt+1 Kt
ct + = wt lt + (1 + rt ) .
Nt Nt

Kt
Define kt = Nt , and recall that Nt+1 = (1 + n)Nt . Then:

ct + (1 + n)kt+1 = wt lt + (1 + rt )kt .
Household optimization under uncertainty II

∞
X
max U = Et β j u(wt+j lt+j +(1+rt+j )kt+j −(1+n)kt+j+1 , 1−lt+j )Nt+j
kt+1+j ,lt+j
j=0

FOC w.r.t. kt+1 ?

−(1 + n)uc,t Nt + βEt (1 + rt+1 )uc,t+1 Nt+1 = 0,

Terminal condition (TVC): limJ→∞ β J (1 + n)uc,t+J Nt+J Kt+1+J = 0.

Consider the following one-period utility function:

ut = ln ct + b ln(1 − lt ), b>0

 
1 1
= βEt (1 + rt+1 )
ct ct+1
Consumption and labour supply

FOC w.r.t. lt ?
wt uc,t − ul,t = 0

wt b
=
ct 1 − lt

bct
= wt
1 − lt
Production Technology:

Cobb-Douglas production technology:

Yt = At (Ktd )α (Lt )1−α , 0 < α < 1.

ln(At ) = Ā + gt + Ãt , Ãt = ρA Ãt−1 + Ãt , ρA ∈ (0, 1).

Labour and capital are paid their marginal products. Real wage and
real interest rate in period t are:

wt = (1 − α)At (Ktd )α (Lt )−α

 1−α
K Lt
r = αAt
Ktd
Equilibrium I

I Capital Market: Ktd = Kt .

I Labour Market: Lt = Nt lt .

I Goods Market:

Nt ct + Kt+1 = At (Kt )α (Nt lt )1−α + (1 − δ)Kt .

Equilibrium II
System of equations describing the equilibrium:

Nt ct + Kt+1 = At (Kt )α (Nt lt )1−α + (1 − δ)Kt ,

" #
Nt lt 1−α
 
1 1
= βEt (1 + αAt − δ) ,
ct Kt ct+1

bct
= (1 − α)At (Kt )α (Nt lt )−α ,
1 − lt

Nt+1 = (1 + n)Nt ,

ln(At ) = Ā + gt + Ãt , Ãt = ρA Ãt−1 + Ãt , ρA ∈ (0, 1).

Initial conditions: N0 , A0 , K0 and terminal condition
1
limJ→∞ β J (1 + n) ct+J Nt+J Kt+1+J = 0.
Special case: Pencil and paper - Brock and Mirman (1972,
JET)

1. Full depreciation of capital: δ = 1;

2. One-period utility function: ut = ln ct ⇒ Households do not
value leisure, so lt = 1 for all t.
3. No population growth: n = 0
Equilibrium conditions:
I Resource constraint:

Ct + It = Yt ⇒ Ct + Kt+1 = At Ktα .

I Euler equation:
   
1 1 1 α−1 1
= βEt (1 + rt+1 ) ⇒ = βEt αAt+1 Kt+1 .
ct ct+1 ct ct+1
Solving the model

We will guess that the solution of this system satisfies a simple rule:
ct = γYt .

Verify: Kt+1 = Yt − Ct ⇒ Kt+1 = (1 − γ)Yt . Moreover,

 
1 Yt+1 1
= βEt α ,
γYt Kt+1 γYt+1
 
Yt+1 γYt
1 = βEt α ,
(1 − γ)Yt γYt+1
 
1
1 = βEt α ,
(1 − γ)

(1 − γ) = βα ⇒ γ = (1 − αβ), verifying the guess.

Solution

ct = (1 − αβ)Yt , ⇒ ct = (1 − αβ)At Ktα ,

Kt+1 = αβYt , ⇒ Kt+1 = αβAt Ktα .

Taking ln of both sides and defining kt = ln(Kt ) and at = ln(At ):

kt+1 = ln(αβ) + at + αkt .

Output:

α
Yt+1 = At+1 Kt+1 = At+1 [αβYt ]α .
Taking ln of both sides and defining yt = ln(Yt ):

yt+1 = at+1 + α ln(αβ) + αyt .

ā+α ln(αβ)
Steady-state: ȳ = 1−α .
White noise process

Suppose that: at+1 = ā + t+1 , t ∼ iidH(0, σ2 ).

ā+α ln(αβ)
Then: ȳ = 1−α . Therefore:

yt+1 = ā + t+1 + α ln(αβ) + αyt ⇒ yt+1 − ȳ = t+1 + α(yt − ȳ).

How does output respond to a shock t+1 = 1, s = 0, s 6= t + 1?

yt+1 = ȳ + 1,
yt+2 = ȳ + α,
yt+3 = ȳ + α2 ,
yt+4 = ȳ + α3 .
yt+s = ȳ + αs−1 .
Dynamics of output with a white noise shock
Random walk Process
Suppose that: at+1 = at + t+1 , t ∼ iidH(0, σ2 ), and a0 = 0.
α ln(αβ)
Then: ȳ = 1−α . Therefore:

yt+1 = at+1 + α ln(αβ) + αyt ⇒ yt+1 − ȳ = at+1 + α(yt − ȳ).

How does output respond to a shock t+1 = 1, s = 0, s 6= t + 1?

yt+1 = ȳ + 1,
yt+2 = ȳ + 1 + α,
yt+3 = ȳ + 1 + α + α2 ,
yt+4 = ȳ + 1 + α + α2 + α3 .
1 − αs
yt+s = ȳ + .
1−α
Dynamics of output with a random walk shock
Dynamics of the full model after a technology shock I

1.2
Technology
Capital
1 Labour

0.8

0.6

0.4

0.2

−0.2
0 5 10 15 20 25 30 35 40
Dynamics of the full model after a technology shock II

1
Consumption
0.9 Output

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1
0 5 10 15 20 25 30 35 40
Dynamics of the full model after a technology shock III

0.6
Wages
Interest rate
0.5

0.4

0.3

0.2

0.1

−0.1
0 5 10 15 20 25 30 35 40
I The people who first advocated the approach (Kydland and
Prescott, 1982) got the Nobel prize in 2004. See their Nobel
lecture: http://www.nobelprize.org/nobel_
prizes/economic-sciences/laureates/2004/
advanced-economicsciences2004.pdf

I Growth and business cycle as unified phenomenon. Recessions

as negative productivity shocks?

I Only real shocks. No role for money in Walrasian setup without

some form of market incompleteness, or price rigidity.
I Look for plausible sources of shocks (taxes, changes in union
power and/or regulation) which can distort labor/leisure choices,
investment choices, and productivity. See Business Cycle
Accounting by Chari, Kehoe and McGrattan:
http://www.econ.umn.edu/~pkehoe/papers/
CKMeconometrica2007.pdf
Next time:

Micro-foundations of investment