Oligopolistic Competition
Cournot (2 firms)
- There are
n
=
2
n=2
n=2 firms, outputs
q
1
q_1
q1 and
q
2
q_2
q2
- zero MC and FC, market demand p = a − b Q p=a-bQ p=a−bQ, where Q = q 1 + q 2 Q=q_1+q_2 Q=q1+q2
- firm 1: π 1 = [ a − b ( q 1 + q 2 ) ] q 1 = a q 1 − b q 1 2 − b q 1 q 2 \pi_1=[a-b(q_1+q_2)]q_1=aq_1-bq_1^2-bq_1q_2 π1=[a−b(q1+q2)]q1=aq1−bq12−bq1q2
- 1’s FOC: a − 2 b q 1 − b q 2 = 0 a-2bq_1-bq_2=0 a−2bq1−bq2=0
- 2’s FOC: a − 2 b q 2 − b q 1 = 0 a-2bq_2-bq_1=0 a−2bq2−bq1=0
- Cournot equilibrium: q ∗ = q ∗ = a 3 b ⇒ Q n = 2 C = 2 a 3 b q^*=q^*=\frac{a}{3b}\Rightarrow Q_{n=2}^{C}=\frac{2a}{3b} q∗=q∗=3ba⇒Qn=2C=3b2a
- First best: S M C = S M R ⇒ 0 = a − b Q ⇒ Q F B = a b SMC=SMR\Rightarrow0=a-bQ\Rightarrow Q^{FB}=\frac{a}{b} SMC=SMR⇒0=a−bQ⇒QFB=ba
- Monopoly: π M = ( a − b Q ) Q = a Q − b Q 2 ⇒ Q M = a 2 b \pi_{M}=(a-bQ)Q=aQ-bQ^2\Rightarrow Q^M=\frac{a}{2b} πM=(a−bQ)Q=aQ−bQ2⇒QM=2ba
- Q M < Q n = 2 C < Q F B Q^M<Q_{n=2}^C<Q^{FB} QM<Qn=2C<QFB
Cournot (n films)
- There are
n
n
n firms, outputs
q
1
,
q
2
,
⋯
,
q
n
q_1,q_2,\cdots,q_n
q1,q2,⋯,qn
- symmetric equilibrium: q 1 = q 2 = ⋯ = q n = q e q_1=q_2=\cdots=q_n=q^e q1=q2=⋯=qn=qe
- firm i: π i = [ a − b ( q i + ( n − 1 ) q e ) ] q i \pi_i=[a-b(q_i+(n-1)q^e)]q_i πi=[a−b(qi+(n−1)qe)]qi ⇒ π i = a q i − b q i 2 − ( n − 1 ) b q e q i \Rightarrow\pi_i=aq_i-bq_i^2-(n-1)bq^eq_i ⇒πi=aqi−bqi2−(n−1)bqeqi
- FOC: a − 2 b q i − ( n − 1 ) b q e = 0 a-2bq_i-(n-1)bq^e=0 a−2bqi−(n−1)bqe=0
- in equilibrium, it is optimal for i i i to choose q i = q e q_i=q^e qi=qe
- FOC ⇒ q e = a ( n + 1 ) b ⇒ Q n C = n a ( n + 1 ) b \Rightarrow q^e=\frac{a}{(n+1)b}\Rightarrow Q_n^C=\frac{na}{(n+1)b} ⇒qe=(n+1)ba⇒QnC=(n+1)bna, clearly, Q n C Q_n^C QnC increases in n n n
- when n → ∞ n\to\infty n→∞, Q n C → Q F B = a b Q_n^C\rightarrow Q^{FB}=\frac{a}{b} QnC→QFB=ba
Stackelberg
- Two firms, 1 is the leader, 2 is the follower
- 1 chooses q 1 q_1 q1 first, 2 chooses q 2 q_2 q2 after observing q 1 q_1 q1
- BI: 2’s view, π 2 = [ a − b ( q 1 + q 2 ) ] q 2 = a q 2 − b q 1 q 2 − b q 2 2 \pi_2=[a-b(q_1+q_2)]q_2=aq_2-bq_1q_2-bq_2^2 π2=[a−b(q1+q2)]q2=aq2−bq1q2−bq22
- 2’s FOC: a − b q 1 − 2 b q 2 = 0 ⇒ a-bq_1-2bq_2=0\Rightarrow a−bq1−2bq2=0⇒ 2’s BR: q 2 B R ( q 1 ) = a 2 b − q 1 2 q_2^{BR}(q_1)=\frac{a}{2b}-\frac{q_1}{2} q2BR(q1)=2ba−2q1
- BI: 1’s view: π 1 = [ a − b ( q 1 + q 2 B R ( q 1 ) ) ] q 1 \pi_1=[a-b(q_1+q_2^{BR}(q_1))]q_1 π1=[a−b(q1+q2BR(q1))]q1 ⇒ π 1 = [ a − b ( q 1 + a 2 b − q 1 2 ) ] q 1 \Rightarrow\pi_1=[a-b(q_1+\frac{a}{2b}-\frac{q_1}{2})]q_1 ⇒π1=[a−b(q1+2ba−2q1)]q1 ⇒ π 1 = [ a − b ( a 2 b + q 1 2 ) ] q 1 \Rightarrow\pi_1=[a-b(\frac{a}{2b}+\frac{q_1}{2})]q_1 ⇒π1=[a−b(2ba+2q1)]q1 ⇒ π 1 = a 2 q 1 − b 2 q 1 2 \Rightarrow\pi_1=\frac{a}{2}q_1-\frac{b}{2}q_1^2 ⇒π1=2aq1−2bq12
- 1’s FOC: a 2 − b q 1 = 0 ⇒ q 1 S t = a 2 b , q 2 S t = a 4 b ⇒ Q S t = 3 a 4 b \frac{a}{2}-bq_1=0\Rightarrow q_1^{St}=\frac{a}{2b},\ q_2^{St}=\frac{a}{4b}\Rightarrow Q^{St}=\frac{3a}{4b} 2a−bq1=0⇒q1St=2ba, q2St=4ba⇒QSt=4b3a
- π 1 S t > π 1 C \pi_1^{St}>\pi_1^{C} π1St>π1C and π 2 S t < π 2 C \pi_2^{St}<\pi_2^{C} π2St<π2C (Revealed Profitability)
Stackelberg_cont
- In equilibrium q 1 S t = a 2 b , q 2 S t = a 4 b , p = a − b 3 a 4 b = a 4 q_1^{St} = \frac{a}{2b},\ q_2^{St} = \frac{a}{4b},\ p = a - b \frac{3a}{4b} = \frac{a}{4} q1St=2ba, q2St=4ba, p=a−b4b3a=4a
- Profits:
- π 1 S t = a 4 ( a 2 b ) = a 2 8 b \pi_1^{St} = \frac{a}{4} \left( \frac{a}{2b} \right) = \frac{a^2}{8b} π1St=4a(2ba)=8ba2
- π 2 S t = a 4 ( a 4 b ) = a 2 16 b \pi_2^{St} = \frac{a}{4} \left( \frac{a}{4b} \right) = \frac{a^2}{16b} π2St=4a(4ba)=16ba2
- π i C = ( a − b 2 a 3 b ) a 3 b = a 2 9 b , ∀ i ∈ { 1 , 2 } \pi_i^C = \left( a - b \frac{2a}{3b} \right) \frac{a}{3b} = \frac{a^2}{9b}, \forall i \in \{1, 2\} πiC=(a−b3b2a)3ba=9ba2,∀i∈{1,2}
- We confirm that: π 1 S t > π 1 C , π 1 S t + π 2 S t < 2 π i C , π 2 S t < π 2 C \pi_1^{St} > \pi_1^C,\ \pi_1^{St} + \pi_2^{St} < 2\pi_i^C,\ \pi_2^{St} < \pi_2^C π1St>π1C, π1St+π2St<2πiC, π2St<π2C
Bertrand
- Two firms, price competition,
F
C
=
0
,
M
C
=
c
≥
0
FC = 0, MC = c \geq 0
FC=0,MC=c≥0
-
p
=
a
−
b
Q
p = a - bQ
p=a−bQ, from 1’s perspective:
- q 1 = 0 q_1 = 0 q1=0 if p 1 > p 2 p_1 > p_2 p1>p2
- q 1 = 1 2 b ( a − p ) q_1 = \frac{1}{2b}(a - p) q1=2b1(a−p) if p 1 = p 2 p_1 = p_2 p1=p2
- q 1 = 1 b ( a − p ) q_1 = \frac{1}{b}(a - p) q1=b1(a−p) if p 1 < p 2 p_1 < p_2 p1<p2
- A unique equilibrium: p 1 B = p 2 B = c p_1^B = p_2^B = c p1B=p2B=c
- Logic:
- (1) Rule out min { p 1 , p 2 } < c \min\{p_1, p_2\} < c min{p1,p2}<c
- (2) Rule out p 1 ≠ p 2 p_1 \neq p_2 p1=p2 ( p i > p j > c p_i > p_j > c pi>pj>c and p i > p j = c p_i > p_j = c pi>pj=c)
- (3) Rule out p 1 = p 2 > c p_1 = p_2 > c p1=p2>c
- Since every firm has an incentive to lower its price.
-
p
=
a
−
b
Q
p = a - bQ
p=a−bQ, from 1’s perspective:
Make the invisible hands visible
1 Market Economy
- Two types of firms, low and high, denoted by F1 and F2
- 10 homogeneous firms of each type ⇒ price takers
- F1: c 1 ( q 1 ) = q 1 2 2 ⇒ m c 1 ( q 1 ) = q 1 c_1(q_1) = \frac{q_1^2}{2} \Rightarrow mc_1(q_1) = q_1 c1(q1)=2q12⇒mc1(q1)=q1
- F2: c 2 ( q 2 ) = q 2 2 4 ⇒ m c 2 ( q 2 ) = q 2 2 c_2(q_2) = \frac{q_2^2}{4} \Rightarrow mc_2(q_2) = \frac{q_2}{2} c2(q2)=4q22⇒mc2(q2)=2q2
- For any market price
p
p
p, using
M
C
=
M
R
MC = MR
MC=MR
- F1: p = q 1 ⇒ q 1 = p p = q_1 \Rightarrow q_1 = p p=q1⇒q1=p
- F2: p = q 2 2 ⇒ q 2 = 2 p p = \frac{q_2}{2} \Rightarrow q_2 = 2p p=2q2⇒q2=2p
- Market supply: Q = 10 p + 20 p = 30 p ⇒ p = Q 30 Q = 10p + 20p = 30p \Rightarrow p = \frac{Q}{30} Q=10p+20p=30p⇒p=30Q
- Market equilibrium:
q 1 m = 1 30 Q & q 2 m = 1 15 Q q_1^m = \frac{1}{30}Q\ \ \&\ \ q_2^m = \frac{1}{15}Q q1m=301Q & q2m=151Q ⇒ q 1 m q 2 m = 1 2 \Rightarrow \frac{q_1^m}{q_2^m} = \frac{1}{2} ⇒q2mq1m=21
2 Central-Planned Economy (CPE)
- Central-Planned Economy (CPE)
- Complete information.
- Government (G) maximizes social welfare (SW).
- On the supply side: Given
Q
ˉ
=
Q
\bar{Q} = Q
Qˉ=Q, choose
(
q
1
,
q
2
)
(q_1, q_2)
(q1,q2) optimally
- Q ˉ = 10 ( q 1 + q 2 ) ⇒ q 2 = 1 10 Q ˉ − q 1 \bar{Q} = 10(q_1 + q_2) \Rightarrow q_2 = \frac{1}{10}\bar{Q} - q_1 Qˉ=10(q1+q2)⇒q2=101Qˉ−q1
-
⇒
\Rightarrow
⇒ For G: choosing
(
q
1
,
q
2
)
(q_1, q_2)
(q1,q2) is equivalent to choosing
q
1
q_1
q1
min q 1 T C = 10 ( 1 2 q 1 2 + 1 4 q 2 2 ) = 10 [ 1 2 q 1 2 + 1 4 ( 1 10 Q ˉ − q 1 ) 2 ] \min_{q_1} TC =10 \left(\frac{1}{2}q_1^2 + \frac{1}{4}q_2^2\right) = 10 \left[\frac{1}{2}q_1^2 + \frac{1}{4}\left(\frac{1}{10}\bar{Q} - q_1\right)^2\right] q1minTC=10(21q12+41q22)=10[21q12+41(101Qˉ−q1)2] ⇒ d T C d q 1 = 10 [ q 1 − 1 2 ( 1 10 Q ˉ − q 1 ) ] = 0 \Rightarrow \frac{dTC}{dq_1} = 10 \left[q_1 - \frac{1}{2}\left(\frac{1}{10}\bar{Q} - q_1\right)\right] = 0 ⇒dq1dTC=10[q1−21(101Qˉ−q1)]=0 ⇒ q 1 ∗ = 1 30 Q ˉ & q 2 ∗ = 1 10 Q ˉ − q 1 = 1 15 Q ˉ \Rightarrow q_1^* = \frac{1}{30}\bar{Q}\ \ \&\ \ q_2^* = \frac{1}{10}\bar{Q} - q_1 = \frac{1}{15}\bar{Q} ⇒q1∗=301Qˉ & q2∗=101Qˉ−q1=151Qˉ
- Under 1 & 2, CPE = Market E, Q ˉ \bar{Q} Qˉ is produced efficiently.
3 Minimum Price
- Start with a market equilibrium
- Demand: p = 3 − Q 15 p = 3 - \frac{Q}{15} p=3−15Q; Supply: p = Q 30 p = \frac{Q}{30} p=30Q
- Market equilibrium: p = 1 p = 1 p=1, Q = 30 Q = 30 Q=30 ( q 1 = 1 (q_1 = 1 (q1=1, q 2 = 2 ) q_2 = 2) q2=2)
- G sets a minimum price
p
‾
=
2
\underline{p} = 2
p=2, demand:
Q
d
=
15
Q^d = 15
Qd=15
-
Efficient way to produce Q s = Q d = 15 Q^s = Q^d = 15 Qs=Qd=15 is:
q 1 ∗ = 1 30 Q ˉ = 1 2 & q 2 ∗ = 1 15 Q ˉ = 1 q_1^* = \frac{1}{30}\bar{Q} = \frac{1}{2}\ \ \&\ \ q_2^* = \frac{1}{15}\bar{Q} = 1 q1∗=301Qˉ=21 & q2∗=151Qˉ=1 -
At p = 2 p = 2 p=2, two types’ ideal outputs are
q 1 i d l = 2 & q 2 i d l = 4 q_1^{idl} = 2\ \ \&\ \ q_2^{idl} = 4 q1idl=2 & q2idl=4 -
Thus, each firm has an incentive to overproduce.
-
4 Inefficiency of Minimum Price
- Efficient way to produce 15 is
(
q
1
∗
=
1
2
,
q
2
∗
=
1
)
(q_1^*=\frac{1}{2},\ q_2^*=1)
(q1∗=21, q2∗=1)
- T C ( q 1 ∗ = 1 2 , q 2 ∗ = 1 ) = 10 [ 1 2 ( 1 2 ) 2 + 1 4 ( 1 ) 2 ] = 30 8 TC(q_1^*=\frac{1}{2},\ q_2^*=1)=10[\frac{1}{2}(\frac{1}{2})^2+\frac{1}{4}(1)^2]=\frac{30}{8} TC(q1∗=21, q2∗=1)=10[21(21)2+41(1)2]=830
- Any other way is less efficient—i.e., more costly
- T C ( q 1 ∗ = 3 4 , q 2 ∗ = 3 4 ) = 10 [ 1 2 ( 3 4 ) 2 + 1 4 ( 3 4 ) 2 ] = 270 64 > 30 8 TC(q_1^*=\frac{3}{4},\ q_2^*=\frac{3}{4})=10[\frac{1}{2}(\frac{3}{4})^2+\frac{1}{4}(\frac{3}{4})^2]=\frac{270}{64}>\frac{30}{8} TC(q1∗=43, q2∗=43)=10[21(43)2+41(43)2]=64270>830
- T C ( q 1 ∗ = 0 , q 2 ∗ = 3 2 ) = 10 [ 1 2 ( 0 ) 2 + 1 4 ( 3 2 ) 2 ] = 45 8 > 30 8 TC(q_1^*=0,\ q_2^*=\frac{3}{2})=10[\frac{1}{2}(0)^2+\frac{1}{4}(\frac{3}{2})^2]=\frac{45}{8}>\frac{30}{8} TC(q1∗=0, q2∗=23)=10[21(0)2+41(23)2]=845>830
- T C ( q 1 ∗ = 1 , q 2 ∗ = 1 2 ) = 10 [ 1 2 ( 1 ) 2 + 1 4 ( 1 2 ) 2 ] = 45 8 > 30 8 TC(q_1^*=1,\ q_2^*=\frac{1}{2})=10[\frac{1}{2}(1)^2+\frac{1}{4}(\frac{1}{2})^2]=\frac{45}{8}>\frac{30}{8} TC(q1∗=1, q2∗=21)=10[21(1)2+41(21)2]=845>830
- T C ( q 1 ∗ = 3 2 , q 2 ∗ = 0 ) = 10 [ 1 2 ( 3 2 ) 2 + 1 4 ( 0 ) 2 ] = 90 8 > 30 8 TC(q_1^*=\frac{3}{2},\ q_2^*=0)=10[\frac{1}{2}(\frac{3}{2})^2+\frac{1}{4}(0)^2]=\frac{90}{8}>\frac{30}{8} TC(q1∗=23, q2∗=0)=10[21(23)2+41(0)2]=890>830
- T C ↑ ⇒ P S ↓ ⇒ D W L ↑ TC\uparrow\ \Rightarrow PS\downarrow\ \Rightarrow DWL\uparrow TC↑ ⇒PS↓ ⇒DWL↑
An Example on Public Good
PG under private contribution
- Two agents 1 and 2, agent
i
i
i’s income
w
i
>
0
w_i > 0
wi>0, consumption on private good
x
i
x_i
xi, public good contribution
g
i
g_i
gi; both prices are normalized to unity.
- u 1 ( x 1 , g 1 ) = x 1 + G = w 1 − g 1 + g 1 + g 2 u_1(x_1, g_1) = x_1 + \sqrt{G} = w_1 - g_1 + \sqrt{g_1 + g_2} u1(x1,g1)=x1+G=w1−g1+g1+g2
- u 2 ( x 2 , g 2 ) = x 2 + G = w 2 − g 2 + g 1 + g 2 u_2(x_2, g_2) = x_2 + \sqrt{G} = w_2 - g_2 + \sqrt{g_1 + g_2} u2(x2,g2)=x2+G=w2−g2+g1+g2
- Let the equilibrium contribution be ( g 1 ∗ , g 2 ∗ ) (g_1^*, g_2^*) (g1∗,g2∗).
- Given g 2 = g 2 ∗ g_2 = g_2^* g2=g2∗, FOC for agent 1: 1 2 g 1 ∗ + g 2 ∗ = 1 \frac{1}{2\sqrt{g_1^* + g_2^*}} = 1 2g1∗+g2∗1=1
- Given g 1 = g 1 ∗ g_1 = g_1^* g1=g1∗, FOC for agent 2: 1 2 g 1 ∗ + g 2 ∗ = 1 \frac{1}{2\sqrt{g_1^* + g_2^*}} = 1 2g1∗+g2∗1=1
- Agent
i
i
i choosing
g
i
=
g
i
∗
g_i = g_i^*
gi=gi∗ is optimal in equilibrium, from above FOCs.
G ∗ = g 1 ∗ + g 2 ∗ = 1 4 G^* = g_1^* + g_2^* = \frac{1}{4} G∗=g1∗+g2∗=41
Efficient PG and a Comparison
- Social Planner’s Choice
- u ( 1 + 2 ) = w 1 + w 2 − g 1 − g 2 + 2 g 1 + g 2 = w 1 + w 2 − G + 2 G u_{(1+2)} = w_1 + w_2 - g_1 - g_2 + 2\sqrt{g_1 + g_2} = w_1 + w_2 - G + 2\sqrt{G} u(1+2)=w1+w2−g1−g2+2g1+g2=w1+w2−G+2G
- FOC implies that
1 = 1 G ⟹ G F B = 1 1 = \frac{1}{\sqrt{G}} \implies G^{FB} = 1 1=G1⟹GFB=1
- A Comparison
- G F B = 1 > G ∗ = 1 4 G^{FB} = 1 > G^* = \frac{1}{4} GFB=1>G∗=41
- Under private contribution, we have multiple equilibria:
- e.g., (1) ( g 1 ∗ = 0 , g 2 ∗ = 1 4 ) (g_1^* = 0, g_2^* = \frac{1}{4}) (g1∗=0,g2∗=41); (2) ( g 1 ∗ = 1 8 , g 2 ∗ = 1 8 ) (g_1^* = \frac{1}{8}, g_2^* = \frac{1}{8}) (g1∗=81,g2∗=81)
- In (1), agent 1 free rides on agent 2; in (2), no obvious free-riding
- However, PG in (2) is also underprovided
More Analysis on Externalities
Coase theorem
The owner is B
- The case in which the owner is B (with quasi-linear preference)
u A = m A + f ( s A ) = y A − p s A + f ( s A ) u_{A}=m_{A}+f\left(s_{A}\right)=y_{A}-p s_{A}+f\left(s_{A}\right) uA=mA+f(sA)=yA−psA+f(sA) d u A d s A = 0 ⇒ f ′ ( s A ) − p = 0 ( 1 ) \frac{d u_{A}}{d s_{A}}=0 \Rightarrow f^{\prime}\left(s_{A}\right)-p=0 \quad (1) dsAduA=0⇒f′(sA)−p=0(1) u B = m B + g ( 1 − s B ) = y B + p s B + g ( 1 − s B ) u_{B}=m_{B}+g\left(1-s_{B}\right)=y_{B}+p s_{B}+g\left(1-s_{B}\right) uB=mB+g(1−sB)=yB+psB+g(1−sB) d u B d s B = 0 ⇒ g ′ ( 1 − s B ) − p = 0 ( 2 ) \frac{d u_{B}}{d s_{B}}=0 \Rightarrow g^{\prime}\left(1-s_{B}\right)-p=0 \quad (2) dsBduB=0⇒g′(1−sB)−p=0(2) -
∃
\exists
∃ a unique
p
p
p and
s
∗
=
s
A
=
s
B
s^{*}=s_{A}=s_{B}
s∗=sA=sB that satisfy (1) and (2)
- (a) f ′ ( s ) f^{\prime}(s) f′(s) decreases in s s s, and g ′ ( 1 − s ) g^{\prime}(1-s) g′(1−s) increases in s s s
- (b) f ′ ( s ) > g ′ ( 1 − s ) f^{\prime}(s) > g^{\prime}(1-s) f′(s)>g′(1−s) if s = 0 s=0 s=0 and f ′ ( s ) < g ′ ( 1 − s ) f^{\prime}(s) < g^{\prime}(1-s) f′(s)<g′(1−s) if s = 1 s=1 s=1
- ⇒ \Rightarrow ⇒ unique cross point of f ′ ( s ) f'(s) f′(s) and g ′ ( 1 − s ) g'(1-s) g′(1−s), for s ∈ [ 0 , 1 ] s\in[0,1] s∈[0,1]
The owner is A
- The case in which A is the owner (with quasi-linear preference)
u A = m A + f ( s A ) = y A + p ~ ( 1 − s A ) + f ( s A ) u_{A} = m_{A} + f(s_{A}) = y_{A} + \tilde{p}(1 - s_{A}) + f(s_{A}) uA=mA+f(sA)=yA+p~(1−sA)+f(sA) d u A d s A = 0 ⇒ f ′ ( s A ) − p ~ = 0 ( 3 ) \frac{du_{A}}{ds_{A}} = 0 \Rightarrow f'(s_{A}) - \tilde{p} = 0 \quad (3) dsAduA=0⇒f′(sA)−p~=0(3) u B = m B + g ( 1 − s B ) = y B − p ~ ( 1 − s B ) + g ( 1 − s B ) u_{B} = m_{B} + g(1 - s_{B}) = y_{B} - \tilde{p}(1 - s_{B}) + g(1 - s_{B}) uB=mB+g(1−sB)=yB−p~(1−sB)+g(1−sB) d u B d s B = 0 ⇒ g ′ ( 1 − s B ) − p ~ = 0 ( 4 ) \frac{du_{B}}{ds_{B}} = 0 \Rightarrow g'(1 - s_{B}) - \tilde{p} = 0 \quad (4) dsBduB=0⇒g′(1−sB)−p~=0(4)-
∃
\exists
∃ a unique
p
~
\tilde{p}
p~ and
s
~
∗
=
s
A
=
s
B
\tilde{s}^* = s_{A} = s_{B}
s~∗=sA=sB that satisfy (3) and (4).
- for the same logic in the previous case
- using (1)-(4), we obtain that p ~ = p \tilde{p} = p p~=p and s ~ ∗ = s ∗ \tilde{s}^* = s^* s~∗=s∗
-
∃
\exists
∃ a unique
p
~
\tilde{p}
p~ and
s
~
∗
=
s
A
=
s
B
\tilde{s}^* = s_{A} = s_{B}
s~∗=sA=sB that satisfy (3) and (4).
A comparison of the two cases
- The case in which B is the owner
- u A O B = y A − p s ∗ + f ( s ∗ ) u_{A}^{OB} = y_{A} - ps^{*} + f(s^{*}) uAOB=yA−ps∗+f(s∗)
- u B O B = y B + p s ∗ + g ( 1 − s ∗ ) u_{B}^{OB} = y_{B} + ps^{*} + g(1 - s^{*}) uBOB=yB+ps∗+g(1−s∗)
- The case in which A is the owner
- u A O A = y A − p s ∗ + f ( s ∗ ) + p u_{A}^{OA} = y_{A} - ps^{*} + f(s^{*}) + p uAOA=yA−ps∗+f(s∗)+p
- u B O A = y B + p s ∗ + g ( 1 − s ∗ ) − p u_{B}^{OA} = y_{B} + ps^{*} + g(1 - s^{*}) - p uBOA=yB+ps∗+g(1−s∗)−p
- Obviously, we obtain that
u A O B < u A O A & u B O B > u B O A u_{A}^{OB} < u_{A}^{OA} \quad \& \quad u_{B}^{OB} > u_{B}^{OA} uAOB<uAOA&uBOB>uBOA
Example with a river
No one owns the river (NOR)
- No one owns the river (NOR)
- steel factory:
π
s
N
O
R
=
p
s
s
−
c
s
(
s
,
x
)
\pi_{s}^{NOR} = p_{s}s - c_{s}(s, x)
πsNOR=pss−cs(s,x)
- Where ∂ c s ∂ x ≤ 0 \frac{\partial c_{s}}{\partial x} \leq 0 ∂x∂cs≤0 if x ≤ x ˉ x \leq \bar{x} x≤xˉ
- Its cost decreases with the level of pollution for x ≤ x ˉ x \leq \bar{x} x≤xˉ
- FOC:
p s = ∂ c s ∂ s ( 5 ) p_{s} = \frac{\partial c_{s}}{\partial s} \quad (5) ps=∂s∂cs(5) − ∂ c s ∂ x = 0 ( 6 ) -\frac{\partial c_{s}}{\partial x} = 0 \quad (6) −∂x∂cs=0(6) - (6) implies that x = x ˉ x = \bar{x} x=xˉ
- fish company:
π
f
N
O
R
=
p
f
f
−
c
f
(
f
,
x
)
\pi_{f}^{NOR} = p_{f}f - c_{f}(f, x)
πfNOR=pff−cf(f,x)
- Where ∂ c f ∂ x > 0 \frac{\partial c_{f}}{\partial x} > 0 ∂x∂cf>0 for any x x x
- Its cost increases with the level of pollution
- FOC:
p f = ∂ c f ∂ f ( 7 ) p_{f} = \frac{\partial c_{f}}{\partial f} \quad (7) pf=∂f∂cf(7)
- steel factory:
π
s
N
O
R
=
p
s
s
−
c
s
(
s
,
x
)
\pi_{s}^{NOR} = p_{s}s - c_{s}(s, x)
πsNOR=pss−cs(s,x)
Fish company owns the river (FOR)
- Fish company owns the river (FOR)
- q q q is the unit price of pollution.
- steel: π s F O R = p s s − q x − c s ( s , x ) \pi_{s}^{FOR} = p_{s}s - qx - c_{s}(s, x) πsFOR=pss−qx−cs(s,x)
- fish: π f F O R = p f f + q x − c f ( f , x ) \pi_{f}^{FOR} = p_{f}f + qx - c_{f}(f, x) πfFOR=pff+qx−cf(f,x)
- FOC:
p s = ∂ c s ∂ s ( 8 ) p_{s} = \frac{\partial c_{s}}{\partial s} \quad (8) ps=∂s∂cs(8) q = − ∂ c s ∂ x ( 9 ) q = -\frac{\partial c_{s}}{\partial x} \quad (9) q=−∂x∂cs(9) p f = ∂ c f ∂ f ( 10 ) p_{f} = \frac{\partial c_{f}}{\partial f} \quad (10) pf=∂f∂cf(10) q = ∂ c f ∂ x ( 11 ) q = \frac{\partial c_{f}}{\partial x} \quad (11) q=∂x∂cf(11)
Steel factory owns the river (SOR)
- Steel factory owns the river (SOR)
- q ~ \tilde{q} q~ is the unit price of pollution.
- steel: π s S O R = p s s + q ~ ( x ˉ − x ) − c s ( s , x ) \pi_{s}^{SOR} = p_{s}s + \tilde{q}(\bar{x} - x) - c_{s}(s, x) πsSOR=pss+q~(xˉ−x)−cs(s,x)
- fish: π f S O R = p f f − q ~ ( x ˉ − x ) − c f ( f , x ) \pi_{f}^{SOR} = p_{f}f - \tilde{q}(\bar{x} - x) - c_{f}(f, x) πfSOR=pff−q~(xˉ−x)−cf(f,x)
- FOC:
p s = ∂ c s ∂ s ( 12 ) p_{s} = \frac{\partial c_{s}}{\partial s} \quad (12) ps=∂s∂cs(12) q ~ = − ∂ c s ∂ x ( 13 ) \tilde{q} = -\frac{\partial c_{s}}{\partial x} \quad (13) q~=−∂x∂cs(13) p f = ∂ c f ∂ f ( 14 ) p_{f} = \frac{\partial c_{f}}{\partial f} \quad (14) pf=∂f∂cf(14) q ~ = ∂ c f ∂ x ( 15 ) \tilde{q} = \frac{\partial c_{f}}{\partial x} \quad (15) q~=∂x∂cf(15)
Comparison between FOR and SOR
- The unit price of pollution
q
q
q, the pollution level
x
x
x, the production levels of the two firms
f
f
f and
s
s
s
- ( q , x , f , s ) (q, x, f, s) (q,x,f,s) that solves (8)-(11) must equal ( q ~ , x , f , s ) (\tilde{q}, x, f, s) (q~,x,f,s) that solves (12)-(15).
- However, firm’s profits in the two cases are different.
- let ( q , x , f , s ) (q, x, f, s) (q,x,f,s) be the solution in either case
- π s F O R = p s s − q x − c s ( s , x ) \pi_{s}^{FOR} = p_{s}s - qx - c_{s}(s, x) πsFOR=pss−qx−cs(s,x)
- π s S O R = p s s − q x − c s ( s , x ) + q x ˉ \pi_{s}^{SOR} = p_{s}s - qx - c_{s}(s, x) + q\bar{x} πsSOR=pss−qx−cs(s,x)+qxˉ
- π f F O R = p f f + q x − c f ( f , x ) \pi_{f}^{FOR} = p_{f}f + qx - c_{f}(f, x) πfFOR=pff+qx−cf(f,x)
- π f S O R = p f f + q x − c f ( f , x ) − q x ˉ \pi_{f}^{SOR} = p_{f}f + qx - c_{f}(f, x) - q\bar{x} πfSOR=pff+qx−cf(f,x)−qxˉ
- it is clear to see that
π s F O R < π s S O R & π f F O R > π f S O R \pi_{s}^{FOR} < \pi_{s}^{SOR} \quad \& \quad \pi_{f}^{FOR} > \pi_{f}^{SOR} πsFOR<πsSOR&πfFOR>πfSOR
When two firms merge (MER)
- When the two firms merge
- π s + f M E R ( s , f , x ) = p s s + p f f − c s ( s , x ) − c f ( f , x ) \pi_{s+f}^{MER}(s, f, x) = p_{s}s + p_{f}f - c_{s}(s, x) - c_{f}(f, x) πs+fMER(s,f,x)=pss+pff−cs(s,x)−cf(f,x)
- FOC:
p s = ∂ c s ∂ s ( 16 ) p_{s} = \frac{\partial c_{s}}{\partial s} \quad (16) ps=∂s∂cs(16) p f = ∂ c f ∂ f ( 17 ) p_{f} = \frac{\partial c_{f}}{\partial f} \quad (17) pf=∂f∂cf(17) − ∂ c s ∂ x = ∂ c f ∂ x ( 18 ) -\frac{\partial c_{s}}{\partial x} = \frac{\partial c_{f}}{\partial x} \quad (18) −∂x∂cs=∂x∂cf(18) -
(
x
,
f
,
s
)
(x, f, s)
(x,f,s) that solves (16)-(18) equals
(
x
,
f
,
s
)
(x,f,s)
(x,f,s) that solves (8)-(11)
- which also equals ( x , f , s ) (x,f,s) (x,f,s) that solves (12)-(15)
- NOR and MER Comparison
- NOR: − ∂ c s ∂ x = 0 ⇒ x = x ˉ -\frac{\partial c_{s}}{\partial x} = 0 \Rightarrow x = \bar{x} −∂x∂cs=0⇒x=xˉ
- MER: − ∂ c s ∂ x = ∂ c f ∂ x > 0 ⇒ x = x ∗ < x ˉ -\frac{\partial c_{s}}{\partial x} = \frac{\partial c_{f}}{\partial x} > 0 \Rightarrow x = x^* < \bar{x} −∂x∂cs=∂x∂cf>0⇒x=x∗<xˉ
- Revealed Profitability: π s + f M E R > π s N O R + π f N O R \pi_{s+f}^{MER} > \pi_{s}^{NOR} + \pi_{f}^{NOR} πs+fMER>πsNOR+πfNOR
More on tragedy of the commons
Tragedy of the commons
- When government chooses
c
c
c (total number of the cows)
- total profit π ( c ) = f ( c ) − p c c \pi(c) = f(c) - p_c c π(c)=f(c)−pcc, where f ′ ( c ) > 0 f'(c) > 0 f′(c)>0, f ′ ′ ( c ) < 0 f''(c) < 0 f′′(c)<0
- First Order Condition (FOC): f ′ ( c ) − p c = 0 f'(c) - p_c = 0 f′(c)−pc=0 ⇒ \Rightarrow ⇒ at optimum, c = c ∗ c = c^* c=c∗, where f ′ ( c ∗ ) − p c = 0 f'(c^*) - p_c = 0 f′(c∗)−pc=0
- When each individual chooses freely
- given c c c cows in the meadow, each player i i i owns exactly one cow, ∀ i ∈ { 1 , 2 , . . . , c } \forall i \in \{1, 2, ..., c\} ∀i∈{1,2,...,c}
- for a new player j j j, j ∉ { 1 , 2 , . . . , c } j \notin \{1, 2, ..., c\} j∈/{1,2,...,c}, π j = f ( c + 1 ) c + 1 − p c > 0 \pi_j = \frac{f(c+1)}{c+1} - p_c > 0 πj=c+1f(c+1)−pc>0
- a new player will not choose to enter when c = c ^ c = \hat{c} c=c^, where f ( c ^ ) c ^ − p c = 0 \frac{f(\hat{c})}{\hat{c}} - p_c = 0 c^f(c^)−pc=0
- from the graph, it is clear to see that c ∗ < c ^ c^* < \hat{c} c∗<c^
Further on tragedy of the commons
- Given c c c, π i = f ( c ) c − p c > 0 , ∀ i ∈ { 1 , 2 , … , c } \pi_{i} = \frac{f(c)}{c} - p_{c} > 0, \forall i \in \{1, 2, \ldots, c\} πi=cf(c)−pc>0,∀i∈{1,2,…,c}
- Consider
j
j
j 's choice of adding one more cow
- for player i i i, Δ π i = f ( c + 1 ) c + 1 − f ( c ) c \Delta\pi_{i} = \frac{f(c+1)}{c+1} - \frac{f(c)}{c} Δπi=c+1f(c+1)−cf(c)
- when player
j
j
j cares about every
i
i
i
- social marginal cost (SMC): c Δ π i = c c + 1 f ( c + 1 ) − f ( c ) c\Delta\pi_{i} = \frac{c}{c+1} f(c+1) - f(c) cΔπi=c+1cf(c+1)−f(c)
- social marginal revenue (SMR): π j = f ( c + 1 ) c + 1 − p c \pi_{j} = \frac{f(c+1)}{c+1} - p_{c} πj=c+1f(c+1)−pc
- social marginal profit (SMP): S M P = S M R + S M C = f ( c + 1 ) − f ( c ) − p c SMP = SMR + SMC = f(c+1) - f(c) - p_{c} SMP=SMR+SMC=f(c+1)−f(c)−pc
- when the G (government) considers adding one more cow:
π ( c + 1 ) − π ( c ) = [ f ( c + 1 ) − p c ( c + 1 ) ] − [ f ( c ) − p c c ] = f ( c + 1 ) − f ( c ) − p c = S M P \pi(c+1)-\pi(c)=[f(c+1) - p_{c}(c+1)] - [f(c)-p_{c}c]\\=f(c+1)-f(c)-p_{c}=SMP π(c+1)−π(c)=[f(c+1)−pc(c+1)]−[f(c)−pcc]=f(c+1)−f(c)−pc=SMP - decision of player j j j (who cares about every i i i) is equivalent to the decision of G (government)