3.

Rational Expectations Solution Technique

Tutorial 14

1. Consider the following model of a closed economy where symbols have their usual
meanings. y, p and m are in natural logarithms. E is the expectations operator and ε is a
random disturbance term with the usual properties.

y t = 20 + 0.25(p t − E t−1 p t ) 14.1

y t = 30 + 0.25(m t − p t ) 14.2
mt = 5 + εt 14.3
(a) Solve for

(i) the full-employment equilibrium value of y,

(ii) the expected price level.

(b) If ε t = −2, what is (i) the price level? (ii) the output level?

2. Consider the following macroeconomic model where m, p and y are the logarithms of
money supply, the price level and output respectively. (14.4) is a money demand function,
(14.5) is a variant of the Phillips curve and (14.6) is the authorities money supply rule. E is
the expectations operator and ε is a ‘white noise’ error term.

mt = pt + yt 14.4
p t = E t−1 p t + 3(y t − 50 ) 14.5
m t = 55 − 0.4(y t−1 − 50 ) + ε t 14.6
(a) Obtain a solution for y t .

(b) Obtain a solution for (i) p t and (ii) its variance around expected price level.

(c) Can the authorities stabilise output in this model?

(d) Solve for p t using Lucas method of undetermined coefficients.

 Solution

1(a) Note that the basic method will suffice for all Rational Expectations models in
which there are expectations (at any date in the past) of current events only. The method
involves three steps:

1. Solve the model, treating expectations as exogenous.

2. Take the expected value of this solution at the date of the expectations, and solve for
the expectations.

3. Substitute the expectations solutions into the solution in 1, and obtain the complete
solution.

Substitute equations (14.1) and (14.3) in (14.2) to get the reduced form:

⇒ 20 + 0.25(p t − E t−1 p t ) = 30 + 0.25(5 + ε t − p t )

0.25ε t + 11.25 = 0.5p t − 0.25E t−1 p t 14.7

In order to get E t−1 p t run the expectations operator along equations 14.1-14.3:

E t−1 y t = 20 + 0.25(E t−1 p t − E t−1 p t ) ≡ 20

E t−1 y t = 30 + 0.25(E t−1 m t − E t−1 p t )
E t−1 m t = 5
Substituting E t−1 m t = 5 and E t−1 y t = 20 in the second expression yields solution for
E t−1 p t .

E t−1 p t = 45 14.8

The full-employment equilibrium value of y ∗ can be found out by looking at equation

(14.1). Note that in equilibrium p t − E t−1 p t = 0. It follows that y ∗ = 20.
 Substituting (14.8) in (14.7) yields:

⇒ 0.25ε t + 11.25 = 0.5p t − 0.25(45 )

p t = 45 + 0.5ε t 14.9

1(b(i)) If ε t = −2, substituting in (14.9) gives the price level:

p t = 45 + 0.5(−2 ) ≡ 44 14.10

1(b(ii)) Substituting (14.10) and (14.8) in (14.1) yields solution for output:

y t = 20 + 0.25(44 − 45 ) ≡ 19.75

2(a) Substitute equations (14.5) and (14.6) in (14.4) to get the reduced form:

55 − 0.4(y t−1 − 50 ) + ε t = E t−1 p t + 3(y t − 50 ) + y t 14.11

In order to get E t−1 p t run the expectations operator along equations 14.4-14.6:

E t−1 m t = E t−1 p t + E t−1 y t

E t−1 p t = E t−1 p t + 3(E t−1 y t − 50 )
E t−1 m t = 55 − 0.4(y t−1 − 50 ) + E t−1 ε t (Note E t−1 ε t = 0)
Substituting E t−1 m t = 55 − 0.4(y t−1 − 50 ) and E t−1 y t = 50 in the first expression yields
solution for E t−1 p t .

E t−1 p t = 25 − 0.4y t−1 14.12

Substituting (14.12) in (14.11) yields solution for output:

 ⇒ 55 − 0.4(y t−1 − 50 ) + ε t = 25 − 0.4y t−1 + 3(y t − 50 ) + y t

y t = 50 + 1 (ε t ) 14.13
4

2 (b(i)) Substituting (14.12) and (14.13) in (14.5) yields solution for the price level:

⇒ p t = 25 − 0.4y t−1 + 3 50 + 1 (ε t ) − 50
4

Substituting for y t−1 (= 50 + 1

4
(ε t−1 )) results in:

p t = 5 + 0.75ε t − 0.1ε t−1

2(b(ii)) Since we have the solution for p t and E t−1 p t it is easy to compute the variance
of the price level.

Definition The variance of X will be denoted by σ 2 whether X is a discrete or a

continuous type of random variable. It is useful to note that
σ 2 = E[X − µ ] 2 = E[X − 2µX + µ 2 ]; and since E is a linear operator,
σ 2 = E(X 2 ) − 2µE(X) + µ 2 = E(X 2 ) − µ 2 .

2
E[p t − E t−1 p t ] 2 = E 3 ε t = 9 σ2
4 16

2(c) The authorities cannot stabilise output in this model as output is invariant to the
money supply rule. This can be proved by multiplying the money supply equation by (0)
instead of (-0.4). The variance of output turns out to be the same ( 161 σ 2 ) irrespective of the
rule in place. This is because the money supply rule is incorporated into agents’
expectations at (t-1) and cannot cause any surprises.

2(d) Lucas method of undetermined coefficients

Substitute equations (14.5) and (14.6) in (14.4) to get the reduced form:

55 − 0.4 (p t−1 − E t−2 p t−1 ) + ε t = p t + 50 + 1 (p t − E t−1 p t )

3 3
 To apply the Lucas method write the solution for the endogenous variables (instead of
being written in terms of the constants and the errors) in terms of the ‘state variables,’ i.e.,
current and past values of the exogenous variables (including the error terms of the model
equations) and past values of the endogenous variables. Thus write the solution for p t (on
which we focus here) as

p t = π 0 + π 1 ε t + π 2 ε t−1 + π 3 p t−1 + π 4 y t−1

By implicaton
E t−1 p t = π 0 + π 2 ε t−1 + π 3 p t−1 + π 4 y t−1
p t−1 = π 0 + π 1 ε t−1 + π 2 ε t−2 + π 3 p t−2 + π 4 y t−2
E t−2 p t−1 = π 0 + π 2 ε t−2 + π 3 p t−2 + π 4 y t−2

Substituting these conjectures in the reduced form equation and collecting terms in ε t ,
ε t−1 ......

(cons tan ts) : 55 = π 0 + 50 ∴ π 0 = 5

(ε t ) : 1 = π 1 + 1 π 1 ∴ π 1 = 3
3 4
(ε t−1 ) : − 0.4 π 1 = π 2 ∴ π 2 = −0.1
3
(p t−1 ) : 0 = π 3
(y t−1 ) : 0 = π 4
7
Substituting the π i s in the conjecture solution for p t yields:
p t = 5 + 0.75ε t − 0.1ε t−1
 Tutorial 15

Consider the following model. Fiscal policy will be held constant and monetary policy
will be the only policy variable affecting demand for output. For expositional purposes the
income velocity of money is also held constant. With these assumptions, the aggregate
demand for output can be written in logs as:

_
m t +v= p t + y t 15.1

The above equation is the equation of exchange in logs. The model is complete with the
introduction of the aggregate supply equation and a money supply rule.

y t = y ∗ + α(p t − E t−1 p t ) 15.2

m t = βy t−1 + ε t 15.3
(a) Given that agents form expectations rationally, use the basic method to obtain a
solution for (i) y t and (ii) p t .

(b) If expectations are backward looking i.e., p et − p et−1 = γ(p t−1 − p et−1 ), where
0 Š 㠊 1, use the basic method to obtain a solution for (i) y t and (ii) p t . Is there any scope
in this model for the policy authorities to influence output through systematic stabilisation
policy?

Solution

(a) Substituting equations (15.2) and (15.3) in (15.1) yields the reduced form:

_
βy t−1 + ε t +v= p t + y ∗ + α(p t − E t−1 p t ) 15.4

In order to solve for E t−1 p t using the basic method:

1. Solve the model, treating expectations as exogenous.

 2. Take the expected value of this solution at the date of the expectations, and solve for
the expectations.

3. Substitute the expectations solutions into the solution in 1, and obtain the complete
solution.

_
E t−1 m t +v = E t−1 p t + E t−1 y t 15.5
E t−1 y t = y + α(E t−1 p t − E t−1 p t ) ≡ y
∗ ∗
15.6
E t−1 m t = βy t−1 + E t−1 ε t ≡ βy t−1 15.7
_
∴ E t−1 p t = βy t−1 +v −y ∗

Substituting the solution for E t−1 p t in (15.4) yields solution for the price level:

_ _
⇒ βy t−1 + ε t +v= p t + y ∗ + α(p t − βy t−1 −v +y ∗ )
_ _
⇒ βy t−1 (1 + α ) + ε t +v −y ∗ + α v −αy ∗ = (1 + α )p t

_
∴ p t = βy t−1 − y ∗ +v + 1 ε
1+α t

Substituting the solutions for p t and E t−1 p t in (15.2) yields:

_ _
⇒ y t = y ∗ + α βy t−1 − y ∗ +v + 1 ε t − (βy t−1 +v −y ∗ )
1+α

yt = y∗ + 1 ε
1+α t

The solution for y t consist of an expected part (y ∗ ) and an unexpected part (functions
of ε t ). Rational expectations has incorporated anything known at t-1 with implications for y
at time t into the expected part, so that the unexpected part is purely unpredictable.

(b) If expectations are backward looking i.e., p et − p et−1 = γ(p t−1 − p et−1 ), where
0 Š 㠊 1:
 ⇒ p et = γp t−1 − γp et−1 + p et−1
⇒ p et = γp t−1 + (1 − γ )[γp t−2 − γp et−2 + p et−2 ]
⇒ p et = γp t−1 + γ(1 − γ )p t−2 − γ(1 − γ )p et−2 + (1 − γ )p et−2
⇒ p et = γp t−1 + γ(1 − γ )p t−2 + (1 − γ ) 2 p et−2
By continuous forward substitution for p et−2 , p et−3 , ....

∞
p et = E t−1 p t = γ ∑(1 − γ ) i p t−1+i 15.8
i=0

Substituting (15.8) in (15.4) yields:

∞
_
⇒ βy t−1 + ε t +v= p t + y ∗ + α p t − γ ∑(1 − γ ) i p t−1+i
i=0

∞
_
pt = 1 βy t−1 +v −y + αγ ∑ (1 − γ ) i p t−1+i + ε t
∗
1+α i=0

Substituting p t and E t−1 p t in (15.2) yields solution for output:

_
∞
βy t−1 +v −y ∗ + αγ ∑ i=0 (1 − γ ) i p t−1+i
yt = y∗ + α 1
1+α ∞
+ε t − γ ∑ i=0 (1 − γ ) i p t−1+i

∞
After collecting terms in γ ∑ i=0 (1 − γ ) i p t−1+i and resorting to some algebraic
manipulation we get:

∞
_
yt = y∗ + α βy t−1 +v −y ∗ + ε t − γ ∑(1 − γ ) i p t−1+i
1+α i=0

Note that when expectations formation is backward looking (adaptive/error learning)

the monetary policy or feedback rule (β -the feedback parameter) enters the solution for
output i.e., authorities can stabilise output in this model.
 Tutorial 16

1. Consider the following model where symbols have their usual meanings:

mt = pt + yt 16.1
_
m t = m −β(y t−1 − y ∗ ) + ε t 16.2
y t = y ∗ + q p t − 1 (E t−1 p t + E t−2 p t ) 16.3
2
(a) Using the Muth method, obtain a solution for (i) y t and (ii) p t .

(b) Given that agents form expectations rationally, is there any scope in this model for
the policy authorities to influence output through systematic stabilisation policy?

Solution

1(a) Substitute (16.2) and (16.3) in (16.1) yields:

_
m −β(y t−1 − y ∗ ) + ε t = p t + y ∗ + q p t − 1 (E t−1 p t + E t−2 p t ) 16.12
2

It follows from (16.3) that:

y t−1 = y ∗ + q p t−1 − 1 (E t−2 p t−1 + E t−3 p t−1 ) 16.13

2
Substituting (16.13) in (16.12) yields:

_
m −β y ∗ + q p t−1 − 1 (E t−2 p t−1 + E t−3 p t−1 ) − y ∗ + ε t = p t + y ∗ +
2

q p t − 1 (E t−1 p t + E t−2 p t )
2
or
 _
m −β(0.5q[(p t−1 − E t−2 p t−1 ) + 0.5(p t−1 − E t−3 p t−1 ) ]) + ε t = p t + y ∗ +

0.5q[(p t − E t−1 p t ) + (p t − E t−2 p t ) ] 16.14

The Muth Method of Undetermined Coefficients:

In a stochastic linear economic system, such as the model we are currently working
with, a variable can always be written as an infinite moving-average process in a random
error. This is Wold’s decomposition theorem. We thus write a random variable y t as:

∞
−
y t =y + ∑ π i ε t−i
i=0

or
−
y t =y +π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....

−
where y = the mean of the series, π i = constant parameters, and ε = a normally
distributed error with a mean of 0, constant variance σ 2ε and zero covariance. The same is
true for the variable p t in our model. The Wold’s decomposition theorem forms the basis
for the Muth method of undetermined coefficients. Using the Muth method we can write
the solution for

∞
−
p t =p + ∑ π i ε t−i
i=0

Ignoring the constants and expanding results give:

p t = π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....

p t−1 = π 0 ε t−1 + π 1 ε t−2 + π 2 ε t−3 + ....
E t−2 p t−1 = π 1 ε t−2 + π 2 ε t−3 + ....
E t−1 p t = π 1 ε t−1 + π 2 ε t−2 + ....
E t−2 p t = π 2 ε t−2 + ....
E t−3 p t−1 = π 2 ε t−3 + ....

Substituting in (16.14) yields:

 _
⇒ m −0.5βq(π 0 ε t−1 + π 0 ε t−1 + π 1 ε t−2 ) + ε t =

π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + .... + y ∗ + 0.5q[π 0 ε t + π 0 ε t + π 1 ε t−1 ]

or

_
m −βqπ 0 ε t−1 − 0.5βqπ 1 ε t−2 + ε t =

π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + .... + y ∗ + qπ 0 ε t + 0.5qπ 1 ε t−1 16.15

We need to now evaluate π 0 and π 1 etc i.e., the undetermined coefficients. For this we
collect terms in ε t , ε t−1 , ε t−2 ......

ε t : 1 = π 0 + qπ 0 ⇒ 1
(1 + q )
ε t−1 : −βqπ 0 = π 1 + 0.5qπ 1

⇒ −βq 1 = π 1 + 0.5qπ 1
(1 + q )

−βq
∴ π1 =
(1 + q )(1 + 0.5q )

ε t−2 : −0.5βqπ 1 = π 2

−0.5βq(−βq )
∴ π2 =
(1 + q )(1 + 0.5q )

ε t−3 : 0 = π 3
We know from (16.3) that
 y t = y ∗ + q p t − 1 (E t−1 p t + E t−2 p t ) = y ∗ + 0.5q[π 0 ε t + π 0 ε t + π 1 ε t−1 ]
2

q 0.5βq 2
∴ yt = y∗ + εt − ε t−1
1+q (1 + q )(1 + 0.5q )

1(b) Since β the parameter in the money supply rule enters the solution for output,
stabilisation (monetary policy in this case) policy is effective. This illustration shows that
in the case of private agents not being able to respond to new information by changing their
wage-price decisions as quickly as the monetary authorities can change any of their control,
then scope once again emerges for systematic stabilisation policy to have real effects.
 Tutorial 17

1. Consider the following model of a closed economy where symbols have their usual
meanings. y, p and m are in natural logs. E is the expectations operator and ε is a random
disturbance term with the usual properties.

y t = −αr t 17.1
y t = y ∗ + γ(p t − E t−1 p t ) 17.2
y t = y ∗ + γ(p t − E t−1 p t ) 17.3
_
mt = m +ε t 17.4
R t = r t + E t−1 p t+1 − E t−1 p t 17.5
(a) Derive the solution for output using the Muth method of undetermined coefficients
and comment on the response of output/unemployment to interest rate changes?

(b) Derive the solution for the price level using McCallum’s “minimal state variable”
criterion?

Solution

1(a) Substituting equation (17.5) the Fisher equation in (17.1) the IS curve yields:

⇒ y t = −α(R t − E t−1 p t+1 + E t−1 p t )

1
Rt = α −y t + αE t−1 p t+1 − αE t−1 p t 17.6

Substituting equations (17.6) and (17.4) in (17.3) yields:

_
1
⇒ m +ε t = p t + y t − ψ α −y t + αE t−1 p t+1 − αE t−1 p t

Substituting for y t results in the reduced form equation:

 _ ψ ψ
m +ε t = p t + 1 + α [γ((p t − E t−1 p t ) + y ∗ )] − α [αE t−1 p t+1 − αE t−1 p t ] 17.7

Using the Muth method we can write the solution for p t as:

∞
−
p t =p + ∑ π i ε t−i
i=0

Ignoring the constants and expanding results in:

p t = π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....

p t+1 = π 0 ε t+1 + π 1 ε t + π 2 ε t−1 + ....
E t−1 p t+1 = π 2 ε t−1 + π 3 ε t−2 + ....
E t−1 p t = π 1 ε t−1 + π 2 ε t−2 + ....

Substituting all this in (17.7) yields:

_
m +ε t = π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + .... + 17.8

ψ ψ π 2 ε t−1 + π 1 ε t−1 +
1 + α γ [π 0 ε t + y ∗ ] − α α −α
π 3 ε t−2 + .... π 2 ε t−2 + ....

We need to now evaluate π 0 and π 1 i.e., the undetermined coefficients. For this we
collect terms in ε t , ε t−1 , ε t−2 ......

ψ
ε t : 1 = π 0 + 1 + α γπ 0

∴ π0 = 1
ψ
1 + γ (1 + α )
 Note that we do not need ε t−1 , ε t−2 etc for computing the solution for y t . Substitute the
solution for π 0 in (17.2) to get solution for output.

y t = y ∗ + γ(p t − E t−1 p t ) ≡ y ∗ + γ(π 0 ε t )

γ
∴ yt = y∗ + ψ εt
1 + γ (1 + α )

Note that the parameter for real interest rate α enters the solution for output. It implies
that a rise in the interest rate shifts the aggregate supply curve upwards increasing output.
Intuitively, a rise in r increases the attractiveness of working today relative to working
tomorrow. Thus, there is an increase in employment and output. The response of labour
supply to the interest rate is known as the intertemporal substitution in labour supply
(Lucas and Rapping, 1969). Equilibrium business cycle theory (which we analyse later)
uses the intertemporal substitution of labour to explain why employment and output
fluctuates over the business cycle.

1(b) McCallum’s MSV Criterion

The MSV criterion (see McCallum (1989(a))) is designed to yield a single bubble-free
solution by construction. Its definition begins by limiting solutions to those that are linear
functions - analogous to our model - of a minimal set of “state variables,” i.e.,
predetermined or exogenous determinants of current endogenous variables. Thus the
solution values involve variables that only appear in the model’s structural equations - no
“extraneous” state variables are included, unlike the Lucas method discussed earlier.

It follows that the relevant determinants of p t include only ε t and a constant, and we
conjecture that the solution is of the form:

pt = φ0 + φ1εt

To find φ 0 and φ 1 , we note that if our conjecture solution for p t is true, then

E t−1 p t = φ 0

E t−1 p t+1 = φ 0
 Substituting these conjectures in the reduced-form equation (17.7) yields:

_ ψ ψ
m +ε t = φ 0 + φ 1 ε t + 1 + α [γ((φ 1 ε t ) + y ∗ )] − α [αφ 0 − αφ 0 ]

The following conditions must pertain to the φ 7 s:

_ ψ
(cons tan ts ) : m = φ0 + γ 1 + α y∗
ψ
(ε t ) : 1 = φ1 + γ 1 + α φ1

Now these two conditions - obtained by equating coefficients on both sides of the
reduced form are just what is needed, for they permit us to solve for the two unknowns
(“undetermined coefficients”) - the φ 7 s. Thus we have,

_ ψ
∴ φ 0 =m −γ 1 + α y ∗

∴ φ1 = 1
ψ
1 + γ (1 + α )

Therefore, the solution for p t is (substituting φ 7 s in the conjecture solution for the price
level):

_ ψ 1
p t =m −γ 1 + α y ∗ + ψ εt
1 + γ (1 + α )
 Tutorial 18

1. Consider the following model of a closed economy where symbols have their usual
meanings. E is the expectations operator and ε is a random disturbance term with the usual
properties.

y t = −α(R t − E t−1 p t+1 + E t−1 p t ) + βG t − ψT t 18.1

y t = y ∗ + δ(p t − E t−1 p t ) 18.2
_
G t = G +ε t 18.3
T t = λy t 18.4

where β and ψ are government expenditure and tax elasticities respectively.

(a) Compute the solution for y t using the Muth method and show that fiscal policy

affects output? If so why?

(b) Compute the variance of output and explain the role of automatic stabilisers? What
happens if we have a high tax elasticity?

Solution

(a) Substituting equations 18.2-18.4 in (18.1) yields the reduced form:

⇒ y ∗ + δ(p t − E t−1 p t ) = −α(R t − E t−1 p t+1 + E t−1 p t ) +

_
β G +ε t − ψλy t

or

_
y ∗ + δ(p t − E t−1 p t ) = −α(R t − E t−1 p t+1 + E t−1 p t ) + β G +ε t −

ψλ(y ∗ + δ(p t − E t−1 p t ))

Using the Muth method we can write the solution for p t as:
 ∞
−
p t =p + ∑ π i ε t−i
i=0

Ignoring the constants and expanding results in :

p t = π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....

p t+1 = π 0 ε t+1 + π 1 ε t + π 2 ε t−1 + ....
E t−1 p t+1 = π 2 ε t−1 + π 3 ε t−2 + ....
E t−1 p t = π 1 ε t−1 + π 2 ε t−2 + ....
Plugging in these values in the reduced form equation yields:

R t − (π 2 ε t−1 + π 3 ε t−2 + .... ) +

y ∗ + δ(π 0 ε t ) = −α +
(π 1 ε t−1 + π 2 ε t−2 + .... )

_
β G +ε t − ψλ(y ∗ + δ(π 0 ε t ))

We need to now evaluate π 0 and π 1 i.e., the undetermined coefficients. For this we
collect terms in ε t , ε t−1 , ε t−2 ......

ε t : δπ 0 = β − ψλδπ 0

β
∴ π0 =
δ + ψλδ

ε t−1 : 0 = απ 2 − απ 1
ε t−2 : 0 = απ 3 − απ 2

∴ π1 = π2 = π3
 Substituting the value for π 0 in the Phillips curve equation (18.2) yields solution for
output:

⇒ y t = y ∗ + δ(p t − E t−1 p t ) ≡ y ∗ + δ(π 0 ε t )

δβ β
∴ yt = y∗ + εt ≡ y∗ + εt
δ + ψλδ 1 + ψλ

E t−1 y t = y ∗ (since E t−1 ε t = 0)

Note that both the fiscal instruments (β and ψ are government expenditure and tax
elasticities respectively) enter the solution for output. In order to explain why fiscal policy
affects output in this model see solution for E t−1 p t .

E t−1 p t = π 1 ε t−1 + π 2 ε t−2 + ....

Note that π 0 (which contains β and ψ) does not enter the solution for the expected price
level. In other words the fiscal instruments are not taken into account when agents’ form
expectations at t-1 and hence can cause surprises.

b) The variance of output is given by:

β 2
⇒ σ 2y = E(y t − E t−1 y t ) 2 = E y ∗ + εt − y∗
1 + ψλ

β 2
∴ σ 2y = σ2
1 ψλ
+

Automatic Stabiliser:

By an automatic stabiliser we mean a mechanism in which a variable (say, tax

liabilities) responds to current income levels, and therefore provide an automatic and
immediate adjustment to current disturbances. Automatic stabilisers are designed to reduce
the lags associated with stabilisation policy. The long and variable lags associated with
monetary and fiscal policy certainly make macroeconomic management more difficult.
Automatic stabilisers are policies that help stimulate or depress the economy when
necessary without any deliberate shift in policy stance. In this particular model ψ is the
elasticity of spending to temporary variations in tax liabilities. Note that the solution for
output is not independent of the automatic stabiliser. A high tax elasticity reduces the
variance of output.
 Tutorial 19

1. Consider the following model of a closed economy where symbols have their usual
meanings. Note that there is persistent shock to output in this model. E is the expectations
operator and ε is a random disturbance term with the usual properties.

y t = −α(R t − E t−1 p t+1 + E t−1 p t ) + µ f (y t−1 − y ∗ ) 19.1

y t = y ∗ + β(p t − E t−1 p t ) + γ(y t−1 − y ∗ ) 19.2
m t = p t + y t − cR t + v t 19.3
_
m t = m +µ m (y t−1 − y ∗ ) + u t 19.4
where µ f (y t−1 − y ∗ ) and µ m (y t−1 − y ∗ ) represent fiscal and monetary feedback response
respectively.

1) Using the Muth method, obtain a solution for y t .

2) Is there a scope in this model for policy authorities to influence output?

2. Solve the following model for p t and y t .

m t = p t + y t − α[E t−1 p t+1 − E t−1 p t ] 19.11
y t = y + β(p t − E t−1 p t ) + λ y t−1 − y
∗ ∗
19.12
_
m t = m +ε t 19.13
Solve by (a) Sargent’s forward root method (b) difference equation method and (c)

Muth method.
Solution

1) Note that we can transform (19.2) using the lag operator:

⇒ y t = y ∗ + β(p t − E t−1 p t ) + γ(Ly t − y ∗ )

⇒ y t = y ∗ + β(p t − E t−1 p t ) + γL(y t − y ∗ )
⇒ (1 − γL )(y t − y ∗ ) = β(p t − E t−1 p t )
 β(p t − E t−1 p t )
yt = y∗ + 19.5
(1 − γL )

From equation (19.1) one can get the solution for the nominal interest rate:

1 − y t + αE t−1 p t+1 − αE t−1 p t + µ f (y t−1 − y ∗ )

Rt = α 19.6

Substituting (19.6) in (19.3) and equating the resulting equation with (19.4) gives:

_
⇒ m +µ m (y t−1 − y ∗ ) + u t = p t +

1 − y t + αE t−1 p t+1 −
yt − c α + vt
αE t−1 p t + µ f (y t−1 − y ∗ )
_
c y t − c(E t−1 p t+1 − E t−1 p t ) + v t
⇒ m +µ(y t−1 − y ∗ ) + u t = p t + 1 + α

where µ = [µ f c
α + µ m ].

Substituting the solution for output in (19.5) into the reduced form equation yields:

_ β(p t−1 − E t−2 p t−1 )

⇒m +µ y ∗ + − y∗ + ut = pt +
(1 − γL )

c β(p t − E t−1 p t )
1+ α y∗ + − c(E t−1 p t+1 − E t−1 p t ) + v t
(1 − γL )

or
_
m (1 − γ ) + µβ(p t−1 − E t−2 p t−1 ) + w t (1 − γL ) = p t (1 − γL ) +
1+ α c (1 − γ )y ∗

c β(p t − E t−1 p t ) − c(1 − γL )(E t−1 p t+1 − E t−1 p t )

+ 1+ α
 where w t = u t − v t .

Using the Muth method we can write the solution for p t as:

∞
−
p t =p + ∑ π i ε t−i
i=0

We need to now evaluate π 0 and π 1 i.e., the undetermined coefficients. For this we
collect terms in w t , w t−1 , w t−2 ......

c π0
wt : 1 = π0 + β 1 + α

∴ π0 = 1
1 + β (1 + c
α )

The identities in the other errors are irrelevant for our purpose here as it is clear by
looking at the solution for output. Substituting the solution for π 0 in (19.5) yields solution
for output.

2) After substituting for π 0 in (19.5) we see that π 0 does not depend either on µ m or µ f .
Thus it is clear that systematic monetary policy does not influence the variance of output in
this model inspite of persistence in the aggregate supply curve.

2(a) Sargent’s forward operator method

Take expectations at t-1 to find E t−1 p t from (19.11-13):

_ λβπ 0 ε t−1
m= E t−1 p t (1 + α ) − αE t−1 p t+1 + y ∗ + 19.14
1 − λL
βπ 0 ε t−1
Note that p t − E t−1 p t = π 0 ε t and y t−1 − y ∗ = 1−λL
.

Hence,
_
m= (1 + α ) 1 − α B −1 E t−1 p t + y ∗ + λβπ 0 ε t−1 19.15
1+α 1 − λL
where the operator B is defined by B −1 (Ex t+j 3 Ω t ) = Ex t+j+1 3 Ω t , where Ω t is the
information set at time ‘t’, in other words B instructs to lag the variable while leaving the
date of expectations unchanged.
_
m −y ∗ λβπ 0 ε t−1
∴ E t−1 p t = − 19.16
(1 + α ) 1 − α
1+α
(1 − λL )(1 + α ) 1 − α
1+α
B −1

Note the last term (using the summation of an infinite series) can be expressed as:

λβπ 0 ε t−1 λβπ 0 ε t−1

=
(1 − λL )(1 + α ) 1 − α
B −1 1 + α(1 − λ ) 1 − λL
1+α

Substituting this in (19.16) yields:

_ λβπ 0 ε t−1
E t−1 p t =m −y ∗ − 19.17
(1 − λL )(1 + α(1 − λ ))
It follows that:
_ λβπ 0 ε t
E t−1 p t+1 =m −y ∗ − E t−1
(1 − λL )(1 + α(1 − λ ))
_ λβπ 0
E t−1 p t+1 =m −y ∗ − E t−1 (ε t + λε t−1 + λ 2 ε t−2 + ...... )
1 + α(1 − λ )
_ λ 2 βπ 0 ε t−1
E t−1 p t+1 =m −y ∗ − 19.18
1 + α(1 − λ ) 1 − λL
Note that the reduced form equation of this model can be expressed as:
_
m +ε t = (p t − E t−1 p t ) + E t−1 p t − α(E t−1 p t+1 − E t−1 p t ) + y ∗
λβπ 0 ε t−1
+ β(p t − E t−1 p t ) + 19.19
1 − λL
Substituting (19.17) for E t−1 p t and (19.18) for E t−1 p t+1 in (19.19) and multiplying
throughout by 1 − λL yields:

ε t (1 − π 0 − βπ 0 ) − λε t−1 (1 − π 0 − βπ 0 ) = 0

Using the method of undetermined coefficients yields:

(ε t ) : 1 − π 0 − βπ 0 = 0 ∴ π0 = 1
1+β
(ε t−1 ) : Terms in ε t−1 also yield this.
Thus the solutions for p t and y t can be expressed as:
 1 ε t +m _ λβπ 0 ε t−1
pt = −y ∗ −
1+β (1 − λL )(1 + α(1 − λ ))
_
p t = m −y ∗ + 1 ε t − λ (y t−1 − y ∗ )
1 + β (1 + α(1 − λ ))
β
yt = y∗ + ε + λ(y t−1 − y ∗ )
1+β t

2(b) Difference equation method

Take expectations at t-1 to find E t−1 p t from (19.11-13). Rearrange by multiplying

throughout by 1 − λL to get:
_
(1 − λ )(m −y ∗ ) = −αE t−1 p t+1 + αλE t−2 p t + (1 + α )E t−1 p t
− λ(1 + α )E t−2 p t−1 + λπ 0 ε t−1
Take expectations at t-2:
_
(1 − λ )(m −y ∗ ) = −αE t−2 p t+1 + (αλ + 1 + α )E t−2 p t
− λ(1 + α )E t−2 p t−1
Therefore
− (1 − λ ) _
α (m −y ∗ ) = E t−2 p t+1 − λ + 1 + α
α E t−2 p t

+λ 1+ α
α E t−2 p t−1

or
− (1 − λ ) _
α (m −y ∗ ) = p et+1 − λ + 1 +
α
α pe + λ 1 + α pe
t α t−1

where p et+1 is p t+1 expected at t-2. Note that this is a second-order difference equation.
Thus the characteristic equation can be expressed as:
_
p et+i = A 1 λ i+1 + A 2 1 + α i+1
α +m −y ∗ (i ≥ 1 )
_
where A 1 + A 2 = p et−1 −m −y ∗
_
A1λ + A2 1 +
α
α = p e −m
t −y ∗

If we appeal to stability requirement so that A 2 = 0, we have

_
A 1 = p et−1 − (m −y ∗ ) and
_ _
p et+i = m −y ∗ + (p et−1 − (m −y ∗ ))λ i+1 (i ≥ 0 )

Now note that

_ _
E t−1 p t+1 =m −y ∗ + λ(E t−1 p t − (m −y ∗ ))
 Since
_ _
E t−2 p t =m −y ∗ + λ(E t−2 p t−1 − (m −y ∗ ))

Substituting for E t−1 p t+1 into (19.14) yields:

_ _ _ λβπ 0 ε t−1
m= E t−1 p t (1 + α ) − α{m −y ∗ + λ(E t−1 p t − (m −y ∗ ))} + y ∗ +
1 − λL
or
_ λβπ 0 ε t−1
E t−1 p t =m −y ∗ −
(1 − λL )(1 + α(1 − λ ))
Rest of the solution follows from what we had done with Sargent’s forward operator
method, especially from (19.17) onwards.

2(c) Muth method

This is in practice the easiest for this model.

− ∞
Let p t =p + ∑ i=0 π i ε t−i . Substitute directly in (19.19) after multiplying (19.19) by

1 − λL to get:
_
(1 − λ ) m +ε t − λε t−1
∞ ∞ ∞ ∞
− −
=p + ∑ π i ε t−i − λ p + ∑ π i−1 ε t−i −α ∑ π i+1 ε t−i − ∑ π i ε t−i
i=0 i=1 i=1 i=1
∞ ∞
+ αλ ∑ π i ε t−i − ∑ π i−1 ε t−i + (1 − λ )y ∗ + βπ 0 ε t
i=2 i=2

Using the method of undetermined coefficients yields:

Collecting terms in:

_ _ _ _
(cons tan ts ) : (1 − λ ) m= (1 − λ ) p + (1 − λ )y ∗ ∴ p=m −y ∗
(ε t ) : 1 = π 0 + β π 0 ∴ π0 = 1
1+β
(ε t−1 ) : −λ = π 1 − λπ 0 − α(π 2 − π 1 )
ε t−i , i ≥ 2 : 0 = −απ i+1 + (1 + α + αλ )π i − λ(1 + α )π i−1

or

0 = π i+1 − 1 + α + λ π i + λ 1 + α π i−1
α α
 This is a difference equation in π i , initial values (i = 2 ) . Thus the characteristic
equation can be expressed as:

_
π i = A 1 λ i−1 + A 2 1 + α i−1
α +m −y ∗ (i ≥ 3 ) where
π1 = A1 + A2
π2 = A1λ + A2 1 +
α
α

Appeal to stability requirement to set A 2 = 0, hence

π 1 = A 1 ; π 2 = λπ 1
Substitute π 2 = λπ 1 in (ε t−1 ) : −λ = π 1 − λπ 0 − α(λπ 1 − π 1 ) to get
−β
λ 1+β
π1 =
(1 + α(1 − λ ))
Thus the solutions for p t and y t is:

_ λβ
p t = m −y ∗ + 1 εt − (ε t−1 + λε t−2 + ...... )
1+β (1 + β )(1 + α(1 − λ ))
_
p t = m −y ∗ + 1 ε t − λ (y − y ∗ )
1+β (1 + α(1 − λ )) t−1
β
yt = y∗ + ε + λ(y t−1 − y ∗ )
1+β t
 Tutorial 20

Consider the following model in natural logarithms.

_
m −p t = α 0 + α 1 y t + α 2 R t 20.1
y t = y ∗ + δ(p t − E t−1 p t ) 20.2

R t = (p t−1 − p t−2 ) + γ(y t−1 − y ∗ ) +

β[(p t−1 − p t−2 ) − 2.5 ] + 2.5 + ε t 20.3
where γ = β = 0.5 is the weight placed on the deviation of output and inflation rate
from target. Equation (20.3) is a simplified version of a backward looking Taylor rule (see
Taylor, 1993) which captures the spirit of recent research in macroeconomics.

R t = Nominal interest rate

p t−1 − p t−2 = inflation rate in t-1

y ∗ = potential real GDP

_
m = constant money supply growth

ε t = white noise error

The policy rule in equation (20.3) has the feature that the nominal rate rises if inflation
increases above a target of 2.5% (set by the Chancellor for the MPC in 1997) or if real
GDP rises above potential GDP. If both the inflation rate and the real GDP are on target,
then the nominal rate would equal 5% or 2.5% in real terms.

1. Solve the model using the Muth method. Can authorities stabilise output in this
model if they use a Taylor rule instead of an explicit money supply rule?

2. Would our conclusions change if we had a New Keynesian Phillips curve of the form
y t = y ∗ + 0.5q[(p t − E t−1 p t ) + (p t − E t−2 p t )] instead?
Solution

1) Substituting equations (20.3) and (20.2) in (20.1) yields the reduced form equation:

_
⇒m −p t = α 0 + α 1 (y ∗ + δ(p t − E t−1 p t )) +

α 2 (p t−1 − p t−2 ) + γ(y t−1 − y ∗ ) + β[(p t−1 − p t−2 ) − 2.5 ] + 2.5 + ε t

or
_
m −p t = α 0 + α 1 (y ∗ + δ(p t − E t−1 p t )) +

(p t−1 − p t−2 ) + γ(y ∗ + δ(p t−1 − E t−2 p t−1 ) − y ∗ ) +

α2
β[(p t−1 − p t−2 ) − 2.5 ] + 2.5 + ε t

Using the Muth method we can write the solution for p t as:

∞
−
p t =p + ∑ π i ε t−i
i=0

Ignoring the constants and expanding results in :

p t = π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....

p t−1 = π 0 ε t−1 + π 1 ε t−2 + π 2 ε t−3 + ....
E t−2 p t−1 = π 1 ε t−2 + π 2 ε t−3 + ....
E t−1 p t = π 1 ε t−1 + π 2 ε t−2 + ....
p t−2 = π 0 ε t−2 + π 1 ε t−3 + π 2 ε t−4 + ....
E t−3 p t−1 = π 2 ε t−3 + ....
Plugging in these values in the reduced form equation yields:
 _
m −(π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + .... ) = α 0 + α 1 (y ∗ + δ(π 0 ε t )) +

π 0 ε t−2 + π 1 ε t−3 +
α 2 (1 + β ) π 0 ε t−1 + π 1 ε t−2 + π 2 ε t−3 + .... − +
π 2 ε t−4 + ....

α 2 γδπ 0 ε t−1 − α 2 β(2.5) + α 2 (2.5) + α 2 ε t

We need to now evaluate π 0 and π 1 i.e., the undetermined coefficients. For this we
collect terms in ε t , ε t−1 , ε t−2 ......

ε t : − π 0 = α 1 δπ 0 + α 2

∴ π0 = − α2
1 + α1δ

ε t−1 : −π 1 = α 2 (1 + β )π 0 + α 2 γδπ 0

(α 2 ) 2
∴ π1 = (1 + β + γδ )
1 + α1δ

The identities in the other errors are irrelevant for computing the solution for output.
Substituting the solution for π 0 in (20.2) yields solution for output:

y t = y ∗ + δ (π 0 ε t ) ≡ y ∗ − δ α2 ε
1 + α1δ t

Note that the parameters of the interest rate rule do not enter the solution for output.
Therefore authorities cannot stabilise output −
in this model. However, if one substitutes for
∞
π 0 , π 1 ,... in the price level equation p t =p + ∑ i=0 π i ε t−i it is clear that β and γ
(parameters in the interest rate rule) enter π 1 (see above). It is clear that the interest rate
rule does affect the solution for the price level.
 2) Suppose we had a New Keynesian Phillips curve, our reduced form equation
becomes:

_
m −p t = α 0 + α 1 (y ∗ + 0.5q[(p t − E t−1 p t ) + (p t − E t−2 p t )]) +

(p t−1 − p t−2 ) + γ(0.5q[(p t−1 − E t−2 p t−1 ) + (p t−1 − E t−3 p t−1 )]) +
α2
β[(p t−1 − p t−2 ) − 2.5 ] + 2.5 + ε t

Plugging in these values in the reduced form equation yields:

_ y ∗ + qπ 0 ε t +
m −(π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + .... ) = α 0 + α 1 +
0.5qπ 1 ε t−1

π 0 ε t−2 + π 1 ε t−3 +
α 2 (1 + β ) π 0 ε t−1 + π 1 ε t−2 + π 2 ε t−3 + .... − +
π 2 ε t−4 + ....

α 2 γqπ 0 ε t−1 + α 2 γ0.5qπ 1 ε t−2 − α 2 β(2.5) + α 2 (2.5) + α 2 ε t

We need to now evaluate π 0 and π 1 i.e., the undetermined coefficients. For this we
collect terms in ε t , ε t−1 , ε t−2 ......

ε t : − π 0 = α 1 qπ 0 + α 2

∴ π0 = − α2
1 + α1q

ε t−1 : −π 1 = α 1 0.5qπ 1 + α 2 (1 + β )π 0 + α 2 γqπ 0

(α 2 ) 2 (1 + β + γq )
∴ π1 =
(1 + α 1 q )(1 + α 1 0.5q )
 Note that we do not need ε t−2 , ε t−3 etc for computing the solution for y t . Now substitute
the solution for π 0 and π 1 in the New Keynesian Phillips curve to get solution for output.

y t = y ∗ + qπ 0 ε t + 0.5qπ 1 ε t−1

Note that the parameters of the Taylor rule β and γ enter the coefficient π 1 and hence the
solution for output. Thus, it is clear that short-run non-neutrality depends on what sort of
aggregate supply curve is in place and not on the way monetary policy (money
supply/interest rate rule) is conducted.
 Tutorial 21

Consider a version of Cagan (1956) model of hyperinflation. All symbols are

interpreted as being deviations from equilibrium. γ is the policy parameter describing the
direction and degree of intervention. For example, a value of γ ‹ 0 means authorities seek
to offset the higher inflationary pressures by engaging in monetary contraction.

m t − p t = −β[E t p t+1 − p t ] + ε t 21.1

m t = −γp t 21.2
1. Comment on the solution path for p t by using the Muth method?

2. What happens when we impose the terminal condition π 1 = 0 on the solution path
for the price level?

Solution

1) Substituting equation (21.2) in (21.1) yields:

⇒ −γp t − p t = −β[E t p t+1 − p t ] + ε t

⇒ −p t (γ + 1 + β ) − ε t = −β(E t p t+1 )

β(E t p t+1 ) = (1 + β + γ )p t + ε t

Using the Muth method we can write the solution for p t as:

∞
−
p t =p + ∑ π i ε t−i
i=0

Ignoring the constants and expanding results in :

 p t = π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....
p t+1 = π 0 ε t+1 + π 1 ε t + π 2 ε t−1 + ....
E t p t+1 = π 1 ε t + π 2 ε t−1 + ....

Plugging in these values in the reduced form equation yields:

π 0 ε t + π 1 ε t−1 +
β(π 1 ε t + π 2 ε t−1 + .... ) = (1 + β + γ ) + εt
π 2 ε t−2 + ....

We need to now evaluate π 0 and π 1 i.e., the undetermined coefficients. For this we
collect terms in ε t , ε t−1 , ε t−2 ......

ε t : βπ 1 = (1 + β + γ )π 0 + 1

βπ 1 − 1
∴ π0 =
(1 + β + γ )

ε t−1 : βπ 2 = (1 + β + γ )π 1

βπ 2
∴ π1 =
(1 + β + γ )

or
(1 + β + γ )
π2 = π1
β

ε t−2 : βπ 3 = (1 + β + γ )π 2
 βπ 3
∴ π2 =
(1 + β + γ )

or
(1 + β + γ )
π3 = π2
β

Similarly,

(1 + β + γ )
π i+1 = πi i = 1, 2..........
β

Therefore, the solution for p t is:

_
⇒ p t =p +π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....

(1+β+γ)
_ βπ 1 − 1 ε t−1 + β
ε t−2 +
p t =p + εt + π1
(1 + β + γ ) (1+β+γ) 2
ε t−3 + ...
β

where π 1 is undetermined. This model clearly has the saddle path property i.e., there is
an infinity of paths/multiplicity of arbitrary solutions exist. One way to resolve this issue is
to impose a terminal condition π 1 = 0 on the solution path in order to rule out a bubble.
 Tutorial 22

Consider the following model (a modified version of Taylor (1977)) where symbols
have their usual meanings.

y t = −β 1 [R t − (E t−1 p t+1 − E t−1 p t )] + β 2 (m t − p t ) + u 1t 22.1

m t − p t = y t − α 1 R t + α 2 (m t − p t ) + u 2t 22.2
y t = γ 0 + γ 1 (m t − p t ) + u 3t 22.3
mt = m 22.4
1. Determine the solution for p t using the Muth method and comment on the solution?

2. What sort of terminal condition is required to get a stable solution?

Solution

1) Our first objective is to get the reduced form equation. Note that we can get R t from
equation (22.2).

⇒ −α 1 R t = (1 − α 2 )(m t − p t ) − y t − u 2t

R t = α11 [y t − (1 − α 2 )(m t − p t ) + u 2t ] 22.5

Substituting (22.5), (22.3) and (22.4) in (22.1) yields the reduced form equation:

⇒ γ 0 + γ 1 (m − p t ) + u 3t =

1
α1 (γ 0 + γ 1 (m − p t ) + u 3t − (1 − α 2 )(m − p t ) + u 2t ) −
− β1 +
(E t−1 p t+1 − E t−1 p t )

β 2 (m − p t ) + u 1t
Collecting terms in (m − p t ) yields:

β1 β1
⇒ γ 0 + γ 1 + α 1 γ 1 − α 1 (1 − α 2 ) − β 2 (m − p t ) + u 3t =
β1
− α 1 [γ 0 + u 3t + u 2t ] +

β 1 E t−1 p t+1 − β 1 E t−1 p t + u 1t

or
E t−1 p t+1 = E t−1 p t + δ 1 p t + δ 0 + u t

where
(1 − α 2 ) β2 1 1
δ1 = α1 + − γ1 α1 + β1
β1

δ0 = γ0 1 1
α1 + β1 − δ1m

ut = − 1 u 1t + 1 1 1
β1 α 1 u 2t + α1 + β1 u 3t

Using the Muth method we can write the solution for p t as:

∞
−
p t =p + ∑ π i u t−i
i=0

Ignoring the constants and expanding results in :

p t = π 0 u t + π 1 u t−1 + π 2 u t−2 + ....

E t−1 p t = π 1 u t−1 + π 2 u t−2 + π 3 u t−3 + ....
p t+1 = π 0 u t+1 + π 1 u t + π 2 u t−1 + ....
E t−1 p t+1 = π 2 u t−1 + π 3 u t−2 + ....
 Plugging in these values in the reduced form equation yields:

⇒ π 2 u t−1 + π 3 u t−2 + .... = π 1 u t−1 + π 2 u t−2 + π 3 u t−3 + .... +

δ 1 (π 0 u t + π 1 u t−1 + π 2 u t−2 + .... ) + δ 0 + u t

We need to now evaluate π 0 and π 1 i.e., the undetermined coefficients. For this we
collect terms in u t , u t−1 , u t−2 ......

ut : 0 = δ1π0 + 1

∴ π0 = − 1
δ1

u t−1 : π 2 = π 1 + δ 1 π 1

π2
∴ π1 =
(1 + δ 1 )

or
π 2 = (1 + δ 1 )π 1

u t−2 : π 3 = π 2 + δ 1 π 2

π3
∴ π2 =
(1 + δ 1 )

or
π 3 = (1 + δ 1 )π 2
 Similarly,
π i+1 = (1 + δ 1 )π i i = 1, 2..........

Therefore, the solution for p t is:

_
⇒ p t =p +π 0 u t + π 1 u t−1 + π 2 u t−2 + ....

∞
pt = − δ0 − 1 u t + π 1 ∑(1 + δ 1 ) i u t−i−1
δ1 δ1 i=0

where π 1 remains undetermined.

2) This model therefore has the saddlepath property i.e., there is an infinity of paths all
but one unstable. This immediately raises the question of which solution to choose. Taylor
resolves this by proposing the criteria that variance of the price level be minimised. This
implies π 1 = 0, so that the solution for p t reduces to:

pt = − δ0 + ut
δ1

Here, the terminal condition (π 1 = 0 ) will select the most stable solution by, in effect,
ruling out the root with the largest modulus.
 Tutorial 23

1. Consider the following model (with future expectations) where symbols have their
usual meanings.

m t = p t + y t − α[E t−1 p t+1 − E t−1 p t ] 23.1

y t = y ∗ + β(p t − E t−1 p t ) 23.2
_
m t = m +ε t 23.3

Determine the solution for p t and y t using the method of forward substitution and
comment on the solution?

2. Consider the following model where symbols have their usual meanings.
mt = pt + yt 23.11
y t = y ∗ + β{p t − 0.5(E t−1 p t + E t−2 p t )} + λ(y t−1 − y ∗ ) 23.12
_
m t = m +µ(y t−1 − y ∗ ) + ε t 23.13

Determine the solution for p t and y t using the Muth method of undetermined
coefficients?

Solution

1) Substituting (23.3) and (23.2) in (23.1) yields the reduced form equation:

_
m +ε t = p t + y ∗ + β(p t − E t−1 p t ) − α[E t−1 p t+1 − E t−1 p t ]

In order to solve for E t−1 p t using the basic method:

1. Solve the model, treating expectations as exogenous.

2. Take the expected value of this solution at the date of the expectations, and solve for
the expectations.

3. Substitute the expectations solutions into the solution in 1, and obtain the complete
solution.
E t−1 m t = E t−1 p t + E t−1 y t − α[E t−1 p t+1 − E t−1 p t ] 23.4
E t−1 y t = y + β(E t−1 p t − E t−1 p t ) ≡ y
∗ ∗
23.5
_ _
E t−1 m t = m +E t−1 ε t ≡m 23.6

Substituting (23.5) and (23.6) in (23.4) yields:

_
m −y ∗ α
E t−1 p t = + E t−1 p t+1 23.7
1+α 1+α
Note that this is not the solution for E t−1 p t because E t−1 p t+1 is not solved out; we have
shifted the problem into the future. To solve for E t−1 p t+1 we lead the model by one period:
_
m −y ∗ α
E t p t+1 = + E t p t+2 23.8
1+α 1+α

Leading it further by one period yields:

_
m −y ∗ α
E t+1 p t+2 = + E t+1 p t+3 23.9
1+α 1+α

−−−−−−−−−−−−−−−−−−−
−−−−−−−−−−−−−−−−−−−

Taking expectations of (23.8) and (23.9) at time t-1 yields:

_
m −y ∗ α
E t−1 E t p t+1 = + E t−1 E t p t+2 23.10
1_ + α 1+α
m −y ∗ α
E t−1 E t+1 p t+2 = + E t−1 E t+1 p t+3 23.11
1+α 1+α

At this point we can invoke the Law of Iterated Expectations i.e.,

 E t−1 E t p t+1 = E t−1 p t+1
E t−1 E t+1 p t+2 = E t−1 p t+2

Hence, we can express (23.10) as follows:

_
m −y ∗ α
E t−1 p t+1 = + E t−1 p t+2
1+α 1+α

Substituting successively (forward) for E t−1 p t+2 , E t−1 p t+3 and so on results in:

_ _
m −y ∗ α m −y ∗ α
⇒ E t−1 p t+1 = + + E t−1 p t+3
1+α 1+α 1+α 1+α
_ _
m −y ∗ α m −y ∗ α 2
⇒ E t−1 p t+1 = + +
1+α 1+α 1+α 1+α
_
m −y ∗ α 3
× + E t−1 p t+4
1+α 1+α

or
N−1
i _
α α
∑
N
E t−1 p t+1 = 1 (m −y ∗ ) + E t−1 p t+N+1 23.12
1+α i=0
1+α 1+α

As N → ∞ and applying the stability condition to E t−1 p t+i ,

so that E t−1 p t+N+1 → 0 as N → ∞.

Thus we can write (23.12) as:

∞
i _
α
⇒ E t−1 p t+1 = 1
1+α ∑ 1+α
(m −y ∗ )
i=0

_ α α 2
m −y ∗ 1+ + +
E t−1 p t+1 = 1+α 1+α
1+α α 3
+ .....
1+α
 (Summation of an infinite series)

_
m −y ∗ 1 _
E t−1 p t+1 = α ≡m −y ∗
1+α 1− 1+α

By the same argument but starting from (23.7) we get:

_ _
m −y ∗ α m −y ∗ α
⇒ E t−1 p t = + + E t−1 p t+2
1+α 1+α 1+α 1+α
_ _ _
m −y ∗ α m −y ∗ α 2 m −y ∗
⇒ E t−1 p t = + +
1+α 1+α 1+α 1+α 1+α
+ α 3
E t−1 p t+3
1+α

or
_ N−1
m −y ∗ α α
∑
i N
E t−1 p t = + E t−1 p t+N+1
1+α i=0
1+α 1+α

_
∴ E t−1 p t =m −y ∗

Note that the paths for events can be_ unstable. Our model here implies that all paths for
prices, except that for which E t−1 p t =m −y ∗ , explode monotonically. Prices would be
propelled into either ever-deepening hyperdeflation or ever-accelerating hyperinflation,
even though money supply is held rigid. So we need to impose a terminal condition for
selecting a unique stable path. In this model we use the method of forward substitution in
order to ensure convergence. Substituting E t−1 p t and E t−1 p t+1 in the reduced form equation
yields solution for the price level:

_ _ _ _
⇒m +ε t = p t + y ∗ + β(p t − (m −y ∗ )) − α[(m −y ∗ ) − (m −y ∗ )]

or
_
p t =m −y ∗ + 1 εt
1+β
 Substituting the solution for p t and E t−1 p t in the Phillips curve equation (23.2) yields
solution for output.

β
yt = y∗ + εt
1+β

2) Note that (23.12) can be expressed as

β(π 0 ε t + 0.5π 1 ε t−1 )
yt − y∗ =
1 − λL
The reduced form equation can be expressed as:

∞
_ µβ(π 0 ε t−1 + 0.5π 1 ε t−2 ) _ β(π 0 ε t + 0.5π 1 ε t−1 )
m+
1 − λL
+ εt = ∑ π i ε t−i +p +y ∗ + 1 − λL
i=0

Multiplying throughout by 1 − λL yields:

_
(1 − λ )(m −y ∗ ) + µβ(π 0 ε t−1 + 0.5π 1 ε t−2 ) + ε t − λε t−1
_ ∞ ∞
= (1 − λ ) p + ∑ π i ε t−i − λ ∑ π i−1 ε t−i + β(π 0 ε t + 0.5π 1 ε t−1 )
i=0 i=1

Collecting terms in:

_ _ _ _
(cons tan ts ) : (1 − λ )(m −y ∗ ) = (1 − λ ) p ∴ p=m −y ∗
(ε t ) : 1 = π 0 + β π 0 ∴ π0 = 1
1+β
(ε t−1 ) : µβπ 0 − λ = π 1 − λπ 0 + β(0.5π 1 )
µβ + λ(1 − {1 + β} )
∴ π1 =
(1 + 0.5β )(1 + β )
(ε t−2 ) : 0.5µβπ 1 = π 2 − λπ 1 ∴ π 2 = (0.5µβ + λ )π 1
ε t−i , i ≥ 3 : 0 = π i − λπ i−1

Therefore the solution for p t and y t can be expressed as:

_
p t =m −y ∗ + 1 ε t + µβ + λ(1 − {1 + β} ) ε t−1 + (0.5µβ + λ )π 1 ε t−2
1+β (1 + 0.5β )(1 + β ) 1 − λL
 β µβ+λ(1−{1+β})
1+β
ε t + 0.5β (1+0.5β)(1+β)
ε t−1
yt = y +
∗
1 − λL
 Tutorial 24

Lucas Critique of Econometric policy evaluation:

Consider the following model where symbols have their usual meaning.

y t = β[p t − 0.5(E t−1 p t + E t−2 p t )] 24.1

mt = pt + yt 24.2
m t = µy t−1 + ε t 24.3

1. What is the solution for y t in terms of current and past ε t ?

2. What is optimal µ?

3. What is the solution for y t in terms of current and lagged m t and y t ?

4. If this was treated as the basis for calculating optimal m t , what would this be?

5. If this money supply (m t ) rule was implemented, what would then be the solution for
output in terms of current and lagged ε t ?

Solution

1) Substituting equations (24.1) and (24.3) in (24.2) yields the reduced form equation:

p t − 0.5
µβ(p t−1 − 0.5(E t−2 p t−1 + E t−3 p t−1 )) + ε t = p t + β 24.4
(E t−1 p t + E t−2 p t )

Using the Muth method we can write the solution for p t as:
 ∞
−
p t =p + ∑ π i ε t−i
i=0

Ignoring the constants and expanding results in :

p t = π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....

E t−1 p t = π 1 ε t−1 + π 2 ε t−2 + ....
E t−2 p t = π 2 ε t−2 + π 3 ε t−3 + ....
p t−1 = π 0 ε t−1 + π 1 ε t−2 + π 2 ε t−3 + ....
E t−2 p t−1 = π 1 ε t−2 + π 2 ε t−3 + ....
E t−3 p t−1 = π 2 ε t−3 + π 3 ε t−4 + ....

Plugging in these values in the reduced form equation and collecting terms in ε t , ε t−1 ,

ε t−2 ......

ε t : 1 = π 0 + βπ 0

∴ π0 = 1
1+β

ε t−1 : µβπ 0 = π 1 + β[π 1 − 0.5π 1 ]

µβπ 0 µβ
∴ π1 = ≡
1 + 0.5β (1 + β )(1 + 0.5β )

ε t−2 : µβ(π 1 − 0.5π 1 ) = π 2 + β[π 2 − 0.5(π 2 + π 2 )]

∴ π 2 = 1 µβπ 1
2
 π3 = π4 = π5 = 0

The identities in the other errors are irrelevant for computing the solution for output.
Substituting the solution for π 0 and π 1 in the New Keynesian Phillips curve yields solution
for output in terms of current and past ε t :

⇒ y t = β[p t − 0.5(E t−1 p t + E t−2 p t )] ≡ βπ 0 ε t + 0.5βπ 1 ε t−1

1 µβ
yt = β ε t + 0.5β ε t−1 24.5
1+β (1 + β )(1 + 0.5β )

2) In order to compute the optimal value of µ take variance of y t :

2
β 1
1+β
ε t + 0.5β
Var(y t ) = E(y t − E t−1 y t ) 2 = E µβ
× (1+β)(1+0.5β)
ε t−1

where the cross products (covariance matrix) are zero as we assume no serial
correlation and no heteroscedasticity. Taking the expectations operator inside yields:

β 2 2
0.5µβ 2
Var(y t ) = σ 2
+σ 2
1+β (1 + β )(1 + 0.5β )

If we set the value of µ = 0 (optimal policy response) then the variance of output
reduces to:

β 2
Var(y t ) = σ 2 = β2σ2π0
1+β

3) Note that from equation (24.5) we know that:

 β 0.5µβ 2
yt = εt + ε t−1
1+β (1 + β )(1 + 0.5β )

Similarly from the money supply rule equation (24.3) we know that:

ε t = m t − µy t−1 24.6
ε t−1 = m t−1 − µy t−2 24.7

Substituting for ε t and ε t−1 in equation (24.5) yields:

β 0.5µβ 2
yt = (m t − µy t−1 ) + (m t−1 − µy t−2 )
1+β (1 + β )(1 + 0.5β )

Thus the solution for y t in terms of current and lagged m t and y t is:

y t = σ 1 m t + σ 2 m t−1 + σ 3 y t−1 + σ 4 y t−2 24.8

where
β
σ1 =
1+β
0.5µβ 2
σ2 =
(1 + β )(1 + 0.5β )
−µβ
σ3 = = −µσ 1
1+β
−0.5µ 2 β 2
σ4 = = −µσ 2
(1 + β )(1 + 0.5β )

4) From equation (24.8) we can calculate optimal money supply rule by setting y t = 0:

−(σ 2 m t−1 + σ 3 y t−1 + σ 4 y t−2 )

mt = σ1 24.9
 According to our relationship in equation (24.5), this feedback rule for money supply
ought to deliver zero fluctuations in output. Thus our expression for money supply from
(24.9) after substituting for σ 7i s give:

−0.5µβ 0.5µ 2 β
mt = m t−1 + µy t−1 + y t−2 + ε t
(1 + 0.5β ) (1 + 0.5β )

Using the lag operartor L we can simplify the above expression as follows:

m t = µy t−1 + εt 24.10
0.5µβ
1+ 1+0.5β
L

This money supply rule ought to deliver zero fluctuations in output.

5) Substituting equations (24.1) and (24.10) in (24.2) yields the reduced form solution
for output (with a new money supply rule) in terms of current and lagged errors. Thus the
reduced form equation can be written as:

µβ(p t−1 − 0.5(E t−2 p t−1 + E t−3 p t−1 )) +

p t−2 −
µβa + εt =
0.5(E t−3 p t−2 + E t−4 p t−2 )

p t + β[p t − 0.5(E t−1 p t + E t−2 p t )] + ap t−1 +

p t−1
aβ
−0.5(E t−2 p t−1 + E t−3 p t−1 )

where
0.5µβ
a=
1 + 0.5β
 Using the Muth method we can write the solution for p t as:

∞
−
p t =p + ∑ π i ε t−i
i=0

Ignoring the constants and expanding results in :

p t = π 0 ε t + π 1 ε t−1 + π 2 ε t−2 + ....

E t−1 p t = π 1 ε t−1 + π 2 ε t−2 + ....
E t−2 p t = π 2 ε t−2 + π 3 ε t−3 + ....
p t−1 = π 0 ε t−1 + π 1 ε t−2 + π 2 ε t−3 + ....
E t−2 p t−1 = π 1 ε t−2 + π 2 ε t−3 + ....
E t−3 p t−1 = π 2 ε t−3 + π 3 ε t−4 + ....
p t−2 = π 0 ε t−2 + π 1 ε t−3 + π 2 ε t−4 + ....
E t−3 p t−2 = π 1 ε t−3 + π 2 ε t−4 + ....
E t−4 p t−2 = π 2 ε t−4 + π 3 ε t−5 + ....

Plugging in these values in the reduced form equation and collecting terms in ε t , ε t−1 ,

ε t−2 ......

ε t : 1 = π 0 + βπ 0

∴ π0 = 1
1+β

ε t−1 : µβπ 0 = π 1 + 0.5βπ 1 + aπ 0 + aβπ 0

µβ − a − aβ
∴ π1 =
(1 + β )(1 + 0.5β )

The identities in the other errors are irrelevant for computing the solution for output.
Substituting the solution for a, π 0 and π 1 in the New Keynesian Phillips curve yields
solution for output in terms of current and past ε t :

−1
y t = β (1 + β ) −1 ε t + 0.25µβ (1 + β )(1 + 0.5β ) 2 ε t−1

εt
Note that the feedback rule for money supply m t = µy t−1 + 0.5µβ
ought to have
1+ 1+0.5β
L

delivered zero fluctuations in output. However, it has delivered fluctuations nearly as large
as the feedback rule. It was Lucas (1976) who first pointed out that if expectations are
formed rationally, then unless the estimated model equations are genuinely structural, the
implications drawn from such models may be seriously flawed. Where we went wrong was
that we used the reduced form to calculate the optimal rule. But that reduced form itself
depends on the monetary rule because the rule influences (rational) expectations and hence
behaviour of agents.