Assignment2 2023 Solutions
Assignment2 2023 Solutions
Assignment2 2023 Solutions
Solutions
1. (a) If a = 0, consumers see the products as identical and we have a standard Cournot
duopoly. If a = 1, consumers see the products as completely different and each
firm has a monopoly over its own product.
(b) Firm 1 maximises π1 = (1 − q1 − (1 − a)q2 )q1 − ca with respect to q1 and firm 2
maximises π2 = (1 − q2 − (1 − a)q1 )q2 with respect to q2 . Therefore, for each
firm i = 1, 2 (letting j = 1, 2 with j ̸= i), the first-order condition is
∂πi
= 1 − 2qi − (1 − a)qj = 0.
∂qi
(As the second derivative is −2 < 0, the solution to the first-order condition max-
imises the profit.) It follows that the best response of firm i is
1 (1 − a)qj
qi = − .
2 2
As both best responses are symmetric, q1 = q2 in equilibrium; therefore,
(c) From part (b), we know that for any a ∈ [0, 1], firm 1’s profit will be
π1 = p1 q1 − ca = (1 − q1 − (1 − a)q2 )q1 − ca
1 1−a 1
= 1− − − ca
3−a 3−a 3−a
3 − a − 1 − (1 − a)
= − ca
(3 − a)2
1
= − ca.
(3 − a)2
1
Firm 1 chooses a in order to maximise π1 = (3−a)2
− ca. The first derivative of the
profit with respect to a is
1
The second derivative is
∂ 2 π1 ∂2(3 − a)−3 ∂c 6
2
= − = 2(−3)(3 − a)−4 (−1) − 0 = > 0.
∂a ∂a ∂a (3 − a)4
As the second derivative is positive, the profit function is convex. The solution to
the first-order condition minimises the profit and the maximum is given by one of
the two extreme points: a = 0 or a = 1. (To see this, notice that, with a convex
function, you can always increase the profit by either reducing or increasing a;
hence, there must be a corner solution.) It remains to check which of a = 0 or
a = 1 gives the larger profit. With a = 0, firm 1’s profit is π1 (a = 0) = 1/9 and
with a = 1 firm 1’s profit is π1 (a = 1) = 1/4 − c. As
1 1 1 1 9−4 5
≥ −c ⇔ c≥ − = = ,
9 4 4 9 36 36
the optimal level of advertising is 0 if c ≥ 5/36 and 1 if c ≤ 5/36. (Both 0 and 1
are optimal if c = 5/36.)
(d) The advantage of advertising is to increase the profit in the second stage by reducing
competition. This is worth doing as long as the cost of advertising does not exceed
the additional profit.
2. (a) Using backward induction, we first solve the second stage in which, given s, the
two retailers simultaneously choose their quantities. The profit of each retailer Ri
(i = 1, 2) is πRi = (4 − qR1 − qR2 − s)qRi . The second stage is a standard Cournot
duopoly with a = 4, b = 1, and c = s so the equilibrium quantities are
4−s
qR1 = qR2 = (Topic 2, slide 23).
3
We now turn to the first stage, in which, anticipating the equilibrium quantities in
the second stage, the supplier chooses its price. The supplier’s profit is
2(4 − s)s
πS = (qR1 + qR2 )s = .
3
The supplier chooses s in order to maximise that profit. The first-order condition
gives
∂πS 2
= (4 − 2s) = 0 ⇔ 2s = 4 ⇔ s = 2.
∂s 3
Then, the equilibrium quantities are
4−2 2
qR 1 = qR 2 = ⇔ qR 1 = qR 2 =
3 3
(b) Firm I produces at no cost, sells to firm R2 at price s, and sells directly to consumers
at price P . Therefore, firm I’s profit is
Firm R2 buys from firm I at price s and sells to consumers at price P . Therefore,
firm R2 ’s profit is
2
(c) We begin with the second stage, in which both firms take s as given and simulta-
neously choose their quantities. Firm I’s first-order condition is
∂πI
= 4 − qR2 − 2qI = 0.
∂qI
(As the second derivative is −2 < 0, the first-order condition maximises the profit.)
Therefore, firm I’s best response is
1
qI = 2 − qR 2 . (1)
2
Firm R2 ’s first-order condition is
∂πR2
= 4 − qI − s − 2qR2 = 0.
∂qR2
(As the second derivative is −2 < 0, the first-order condition maximises the profit.)
Therefore, firm R2 ’s best response is
1 1
qR2 = 2 − qI − s. (2)
2 2
Plugging (1) into (2) yields
1 1 1 1 1
qR 2 =2− 2 − qR 2 − s = 2 − 1 − s + qR 2 .
2 2 2 2 4
3 2−s
⇔ qR =
4 2 2
4 − 2s
⇔ qR 2 = .
3
Plugging this result into (1) yields
4 − 2s 12 − 4 + 2s 4+s
qI = 2 − = ⇔ qI = .
6 6 3
Note: The above reasoning is provided for completeness; however, the shortest way
to find the quantities is to recognise that this is a Cournot market with differentiated
cost and use the results from Topic 2, slide 37 with parameters a = 4 and b = 1,
and marginal costs 0 and s.
We now turn to the first stage, in which firm I anticipates the equilibrium quantities
and chooses s to maximise its profit. Firm I’s profit is
3
(As the second derivative is −10/9 < 0, the first-order condition maximises the
profit.) It follows that the equilibrium quantities are
4+2 4−4
qI = ⇔ qI = 2 and qR2 = ⇔ qR2 = 0.
3 3
(d) The merger is good for consumers. The total quantity produced increases from 4/3
(in part (a), each of the two retailers sells 2/3) to 2 (in part (b), I sells 2 and R2
sells 0) as a result of the merger. Therefore, the equilibrium price decreases and
the consumer surplus increases.
Note 1: Any price s ≥ 2 would be optimal for firm I as it would induce a corner solution
in which R2 sets its quantity to zero.
Note 2: The intuition for the result is that firm I prefers to sell directly to consumers
as there is no opportunity for firm I to increase its profit by selling to R2 instead. To
see this, notice first that R2 only sets a positive quantity if P is larger than s so I
would have to sell at a lower price to R2 as it does to consumers. Second, selling to R2
has the additional disadvantage of reducing the price at which I can sell to consumers.
Therefore, it is more profitable for I to sell directly to consumers and keep R2 out by
setting s high enough. This result holds for any values of a, b, and c. It might however
change if R2 had access to consumers that I cannot reach.