Mathematics for Economics ECON10005

SOLUTIONS
4 Problem set – Single-variable optimisation

1 Stationary Points
Please associate the following keywords to the various points listed below: stationary point,
inflection point, local maximum, global maximum, local minimum, global minimum.

a. x = 0 for the function f (x) = x5 , stationary and inflection point,

b. y = 2nπ + π/2, where n is an integer, for the function f (x) = sin(y), stationary and
local/global maxima,
c. z = (1, 2) for the function f (z1 , z2 ) = (z1 −1)2 +(z2 −2)2 , stationary and local/global
minimum.

2 Monopoly pricing
A firm producing violins is a monopoly in Ireland. The unit production cost is c > 0.
Demand for violins is characterized by the following function:

p(q) = − ln(αq)/β,

which can be interpreted as follows: in order to produce q > 0 violins, the monopoly
would need to set the unit price at p = p(q). (α, β) are positive parameters characterising
demand.
a. Write the profit of the firm, as a function of a given produced quantity q > 0.
The profit can be written as follows:

π = ( p(q) − |{z}
c ) q = (p(q) − c)q
|{z} |{z}
unit price unit cost quantity

b. Rewrite the profit of the firm, as a function of the selling price p > 0.
In order to do so, we need to invert the demand function, i.e., write the quantity as
a function of the price, rather than the opposite:

p = − ln(αq)/β

e−βp = αq
e−βp
q=
α
The profit of the firm is then:
e−βp e−βp
π=( p − |{z}
c ) = (p − c)
|{z} α α
unit price unit cost | {z }
quantity

1
c. The firm is maximising its profits; write the corresponding optimisation program.
The firm maximises:
e−βp
 
max π(p) = (p − c)
p α

d. Find the stationary point.

We look for the stationary point of the function π(p):

e−βp e−βp
π ′ (p) = − (p − c)β =0
α α
1 − (p − c)β = 0
p = c + 1/β

e. Find the profit maximising price. Comment on your result.

To check whether the stationary point is a maximum, we can analyse the sign of
π ′ (p):

π ′ (p) > 0
1 − (p − c)β > 0
e−βp e−βp
− (p − c)β >0
α α
p < c + 1/β
We can conclude that π ′ (p) is positive below the stationary point and negative above,
so p∗ = c + 1/β is the profit maximising price.

[Note that π ′′ (p) is not negative everywhere, so profit is not concave. So we cannot
use this to verify that the stationary point is a maximum.]

Comment: The firm sets up a mark-up 1/β over the unit cost of production c. The
smaller β, the less responsive consumers are to changes in prices, and the higher the
mark-up chosen by the firm.
f. Assume that the government subsidizes violins, i.e., the price faced by consumers is
p − τ where p is the price offered by the firm and τ > 0 is a given subsidy. Solve for
the new optimal price: does it increase, decrease?
The firm maximises: n o
max π(p) = (p − c)e−β(p−τ ) /α
p
We look for the stationary point of the function π(p):
π ′ (p) = e−β(p−τ ) /α − (p − c)βe−β(p−τ ) /α = 0
1 − (p − c)β = 0
p = c + 1/β
The firm does not change its price, compared to the situation without subsidy. It
will however sell more violins and make a higher overall profit.

2
3 Maximising tax revenue
The British government decides to impose a tax t per French flag sold to UK consumers.
As a consequence, when the consumer pays the price P , the supplier receives P − t. The
supply and demand for French flags are thus respectively given by:

P − t = Q + 8 and P = −3Q + 80

a. Write down the market equilibrium condition and solve for the equilibrium quantity
and the equilibrium price.
The market equilibrium is characterised by:

Q + 8 + t = −3Q + 80 ⇒ Q = 18 − t/4 and P = 26 + 3/4t

b. Find the expression for the total tax revenue as a function of the tax rate t.
The total tax revenue is:

T R(t) = t × Q = t(18 − t/4)

c. Find the revenue-maximising tax rate.

We just need to maximise the following program:

max {T R(t) = t(18 − t/4)}

t

′′
It is easy to show that T R (t) = −1/2: the function is concave. As a consequence, we
can apply a theorem seen during the lectures: if we find one stationary point, it will be
the maximum.
The stationary point verifies,
′
T R (t∗ ) = 18 − t∗ /2 = 0

and thus t∗ = 36 is the maximum.

3
4 Lesotho and Laffer
The Lesotho government would like to raise tax revenues in order to finance a large in-
dustrialisation program. Letting t denote the flat tax rate, we assume that the tax base,
Y (t), decreases with t.

a. Write the expression for tax revenues, R(t).

R(t) = t · Y (t)

b. Under which condition are these tax revenues concave in t?

R is concave if R′′ (t) ≤ 0. We take the derivatives:

R′ (t) = Y (t) + t · Y ′ (t)

R′′ (t) = Y ′ (t) + Y ′ (t) + t · Y ′′ (t) = 2Y ′ (t) + t · Y ′′ (t)

So R is concave if 2Y ′ (t) + t · Y ′′ (t) ≤ 0.

c. Under the previous condition, which relationship should be verified by the optimal finite
tax rate, if it exists?

The government maximises

max t · Y (t)
t
If R is concave, the optimal finite tax rate t∗ would be a stationary point and verify the
first-order condition:
R′ (t∗ ) = 0 ⇒ Y (t∗ ) + t · Y ′ (t∗ ) = 0
Bonus Info:
• The FOC can be re-written as:
t · Y ′ (t∗ )
−1 = = εY
Y (t∗ )
This is the elasticity of the tax base w.r.t. t
• The graph of R(t) is called the Laffer curve:

R(t)

t t

4
5 Income and substitution
Consider a worker paid at a given hourly wage w > 0; the labour income is her only
income, and she consumes her whole income. The worker gets (strictly) concave utility
u(c) from consumption c and (strictly) convex disutility v(l) from working l hours. (I.e.,
We assume: u′ (c) > 0, u′′ (c) < 0, v ′ (l) > 0, v ′′ (l) > 0.)

a. Write the worker’s maximisation program.

The worker wishes to maximise u(c) − v(l) where c = w × l by choosing l:

max u(w · l) − v(l)

l

b. Find the optimal hours worked, l∗ , for a given wage rate w.

At the optimum the worker should be indifferent between slightly increasing or decreas-
ing hours:

∂u(w · l) ∂v(l)
=
∂l ∂l
⇒ w · u′ (w · l∗ ) = v ′ (l∗ ) (1)

∗
c. Find dl
dw and interpret the different effects affecting its sign. [Hint: You need to use
implicit differentiation.]
Equation 1 implicitly relates l∗ to w. I.e. we can now understand l∗ as a function of w:

w · u′ (w · l∗ (w)) = v ′ (l∗ (w)) (2)

We differentiate both sides of equation 1 w.r.t w

(using both product rule and chain rule):
 
′ ∗ ′′ ∗ ∗ dl dl
u (w · l ) + w · u (w · l ) · l + w = v ′′ (l∗ )
dw dw

We re-arrange:
dl u′ (wl∗ ) − wl∗ (−u′′ (wl∗ ))
= ′′ ∗
dw v (l ) + w2 (−u′′ (wl∗ ))
Remember we assume: u′ (c) > 0, u′′ (c) < 0, v ′ (l) > 0, v ′′ (l) > 0, so dw
dl
could be
positive or negative, depending on the sign of the numerator u (wl ) − wl (−u′′ (wl∗ ))
′ ∗ ∗

So, the effect of the hourly wage is ambiguous:

• Substitution u′ (wl∗ ): as w increases, the opportunity cost of not working in-

creases. I want to substitute leisure with work. (positive effect)
• Income −wl∗ (−u′′ (wl∗ )): as w increases, I become richer with the same number
of hours worked. With decreasing returns to consumption, I want to work less.
(negative effect)