Scribd
Upload
41 views

Money and in Ation - The Cagan Model: Econ 208

The Cagan model examines whether inflation is self-generating or caused by monetary expansion. It uses three equations: (1) money supply equals money demand, (2) the time derivative of money supply equals expected inflation, and (3) expected inflation is determined by adaptive expectations. If the product of coefficients α and β is less than 1, then inflation is stable and caused by monetary expansion. If it is greater than 1, then inflation is unstable and self-generating. The model can also be specified with exogenous seigniorage rather than money growth.

Uploaded by

Muhammed Arham
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
41 views

Money and in Ation - The Cagan Model: Econ 208

The Cagan model examines whether inflation is self-generating or caused by monetary expansion. It uses three equations: (1) money supply equals money demand, (2) the time derivative of money supply equals expected inflation, and (3) expected inflation is determined by adaptive expectations. If the product of coefficients α and β is less than 1, then inflation is stable and caused by monetary expansion. If it is greater than 1, then inflation is unstable and self-generating. The model can also be specified with exogenous seigniorage rather than money growth.

Uploaded by

Muhammed Arham
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

Money and In‡ation - The Cagan Model

Econ 208

Lecture 11

March 6, 2007

Econ 208 (Lecture 11) Cagan Model March 6, 2007 1/9


The Cagan Model
Cagan’s question: Are in‡ations self-generating or are they caused by
monetary expansion?
The model (m = log M, p = log P, µ = ṁ, π = expected ṗ)

m p = βπ Money supply equals money demand


µ ṗ = βπ̇ Time derivative of …rst equation
π̇ = α (ṗ π) Adaptive expectations
Could be generated by the steady state of a Sidrauski model in the additively
separable case, taking logs:
M/P = L (c, r + π )
m p = log L (c, r + π )
m p = f (π ) βπ
Adaptive expectations equivalent to using weighted average of past in‡ation
rates: Z t
α (t τ )
π=α e ṗ (τ ) d τ
∞
Econ 208 (Lecture 11) Cagan Model March 6, 2007 2/9
The Cagan Model (cont’d)

m p = βπ Money supply equals money demand


µ ṗ = βπ̇ Time derivative of …rst equation
π̇ = α (ṗ π) Adaptive expectations

The model takes as given an arbitrary (almost everywhere di¤erentiable) path


of m (t ) and an initial expectation π 0 . Need to solve for the equilibrium time
paths of p (t ) and π (t ).
Example: Constant money growth: µ (t ) = µ, constant. Then there is a
steady state equilibrium in which

p (t ) = m (t ) + βµ and π (t ) = µ

But is this steady state stable? That is, if we start with π 0 6= µ, does π (t )
converge to µ?
The answer depends on the values of α and β.

Econ 208 (Lecture 11) Cagan Model March 6, 2007 3/9


The Cagan Model (cont’d)

m p = βπ Money supply equals money demand


µ ṗ = βπ̇ Time derivative of …rst equation
π̇ = α (ṗ π) Adaptive expectations

From MS = MD and Adaptive expectations:

π̇ = α (µ + βπ̇ π ) , which can be written as:


α
π̇ = (µ π)
1 αβ
This is a linear di¤erential equation, which has a unique solution given any
initial value π 0 and a unique rest point π = µ.
If αβ < 1 the rest point is stable, so in‡ation is a monetary phenomenon.
If αβ > 1 it is unstable, and π explodes independently of monetary policy, as
does ṗ, so in‡ation is an expectational phenomenon.
Cagan estimated αβ < 1 for the interwar hyperin‡ation cases he studied.

Econ 208 (Lecture 11) Cagan Model March 6, 2007 4/9


Seigniorage

The implicit tax revenue accruing to the issuer of money, equal to Ṁ/P
Suppose we take as exogenous not the rate of monetary expansion but the
required seigniorage δ
This is more realistic in the hyperin‡ation cases, where …scal pressures were
clearly driving monetary expansion

Ṁ/P = δ, which can be expressed as


Ṁ M
. = δ, which together with MS=MD yields
M P
µe βπ = δ, or:
α
µ = δe βπ , and, since we still have π̇ = (µ π)
1 αβ
α
π̇ = δe βπ π
1 αβ

which has two rest points, if any!

Econ 208 (Lecture 11) Cagan Model March 6, 2007 5/9


Seigniorage (cont’d)

Interpretation of multiple equilibrium: When expected in‡ation is high, the


demand for money is low, so tax base for seigniorage is low, so the tax rate
(actual in‡ation) must be high.
Which equilibrium is stable? H or L?
Again, this depends on the product αβ
(a) αβ < 1
L is locally stable, H is unstable
Given π 0 < π H , π (t ) ! π L
Given π 0 > π H , π (t ) ! ∞
(b) αβ > 1
H is locally stable, L is unstable
Given π 0 > π L , π (t ) ! π H
Given π 0 < π L , π (t ) ! ∞

Econ 208 (Lecture 11) Cagan Model March 6, 2007 6/9


Comparative statics with seigniorage

Assume the “stable” case (a): αβ < 1


When δ " then π L " (µ jumps, followed by a further increase in µ and π)
What happens to ṗ = µ + βπ̇?
In the “unstable case (b): αβ > 1 you get wierd comparative statics
Maximal seigniorage: when the two curves are tangent
in this case you get one-sided stability
this is a fragile equilibrium

Econ 208 (Lecture 11) Cagan Model March 6, 2007 7/9


Cagan with Perfect Foresight
The case of exogenous monetary expansion

Now the model is:

m p = βπ Money supply equals money demand


π = ṗ Perfect foresight

m p = βṗ, or
ṗ = (1/β) (p m)

Given a time path for m (t ) this di¤erential equation has a unique solution
for any given initial value p0 .
This means the time path of the price level is indeterminate.
Almost all equilibria are unstable
So perfect foresight seems to compound the possible instability problem.

Econ 208 (Lecture 11) Cagan Model March 6, 2007 8/9


Cagan with Perfect Foresight
The case of exogenous monetary expansion (cont’d)

ṗ = (1/β) (p m)
Example: constant money supply: m (t ) = m0 constant
In this case the equilibrium is
p (t ) = (p0 m0 ) e (1 /β)t + m0
which yields the stable path p (t ) = m0 if the initial price level is p0 = m0
but yields an unstable path for all other initial price levels.
More generally, for any path of m (t ) and any p0 there is one stable
equilibrium, known as the “forward equilibrium”:
Z ∞
p f (t ) = (1/β) e (1 /β)(t s)
m (s ) ds
t
and a continuum of “backward” equilibria:
Z t
p b (t ) = p0 e (1 /β)t (1/β) e (1 /β)(t s)
m (s ) ds,
0
one for each p0
This raises the question of what to do when the equilibrium is indeterminate?
Econ 208 (Lecture 11) Cagan Model March 6, 2007 9/9

You might also like