Trade Chaney Slides

Download as pdf or txt
Download as pdf or txt
You are on page 1of 17

International Trade

Lecture 4: the extensive margin of trade

Thomas Chaney

Sciences Po

Thomas Chaney (Sciences Po) International Trade 1 / 17


Target of the paper

Explain the observed role of the extensive margin of trade.


Determin in such a model what parameters govern the elasticity of
trade flows with respect to trade barriers.

Thomas Chaney (Sciences Po) International Trade 2 / 17


New assumptions (versus Melitz)

Many asymmetric countries with asymmetric trade barriers.


No free entry: number of firms directly proportional to country size
(can be relaxed).
Pareto distributed productivity shocks (delivers closed form solutions).
1 extra numéraire sector as in Helpman-Krugman pins down relative
wages (can be relaxed).

Thomas Chaney (Sciences Po) International Trade 3 / 17


Set-up

CES/Cobb-Douglas preferences,
0 1 s s 1 (1 µ)
ˆ
µ s 1
U⌘ qo @ q (w ) s d wA
W

N asymmetric countries: sizes (Li )N N


i =1 , trade barriers, ( tij , fij )i,j =1 .

Thomas Chaney (Sciences Po) International Trade 4 / 17


Productivity distribution and entry

Pareto distribution for productivities j 2 [1, +•),


q
P ( j > F) = F
with q > s 1.

Note: Pareto distributions are invariant to truncation,


✓ ◆ q
⇤ F P ( j > F)
P ( j > F| j > j ) = ⇤
=
j j⇤ q

In country n, Ln potential entrants.

Thomas Chaney (Sciences Po) International Trade 5 / 17


Ownership structure

Without free entry, ownership structure matters.


Aggregate world profits (P) equally distributed across workers.
Aggregate income in country i,
P
(1 + ) Li
L

Thomas Chaney (Sciences Po) International Trade 6 / 17


Zero cutoff profit conditions

Sales by firm j from country i in country j,


✓ ◆1 s ✓ ◆
stij P
rij ( j) = ⇥µ 1+ Lj
(s 1) jPj L

Only firms with j jij⇤ from i export to j,

1 ⇣ ⌘s 1
1
1
pij jij⇤ = 0 , jij⇤ = l1 fijs 1
Pjs Lj tij (ZCPi j)

Entry threshold into j depends on prices in j.

Thomas Chaney (Sciences Po) International Trade 7 / 17


Aggregate prices

Prices in j depends on which firms able to enter j,

N ˆ•
Pj1 s = Â Lk pkj ( j)1 s
dG ( j)
k =1 ⇤
jkj

N ˆ• ✓ ◆s 1
s 1 j
= Â Lk s

tkj
dG ( j)
k =1 ⇤
jkj
N ✓ ◆s 1
q s 1 1
= Â Lk q (s 1) s

tkj
⇤s
jkj 1 q

k =1

Thomas Chaney (Sciences Po) International Trade 8 / 17


Aggregate prices

Prices in j depends thresholds of entry into j.


Threshold of entry into j depends on prices in j.
Simple solution for prices,
✓ ◆ 1q
Lj 1
Pj = l 2 ⇥ ⇥ ( Lj ) s 1 ⇥ rj
L
N
(sq1 1)
with rj q
⌘ Â sk ⇥ tkj q ⇥ fkj
k =1

rj is an aggregate measure of j’s remoteness from the rest of the


world (due to both fixed and variable costs).

Thomas Chaney (Sciences Po) International Trade 9 / 17


Equilibrium exports and selection

Firm level exports (r ), thresholds of export ( j̄), and aggregate profits


( P ),
8 ⇣ ⌘ s q 1 ⇣ ⌘s 1
< Lj r
rij ( j) = l 3 ⇥ L ⇥ tijj ⇥ js 1 , if j jij⇤
: 0 , otherwise
✓ ◆ 1q ✓ ◆ 1
L tij
jij⇤ = l4 ⇥ ⇥ ⇥ fijs 1
Lj rj
P = l5 ⇥ L

1 Note: elasticity of rij w.r.t. tij is


(s 1) as in Krugman.
2 Note: aggregate income elasticity of firm level exports < 1.

Thomas Chaney (Sciences Po) International Trade 10 / 17


Gravity equation

Aggregate exports from i to j, Chaney versus Krugman,


8 ⇣ ⌘ q
>
< X Chaney = l Li Lj tij ( q 1)
ij L rj fij s 1
⇣ ⌘ ( s 1)
>
: X Krugman = l̃ Li Lj tij
ij L r̃j

1 Higher elasticity of trade w.r.t. variable trade barriers, q > s 1.


2 Elasticity does not depend on the elasticity of substitution s.
q
3 Elasticity w.r.t. fixed costs, s 1 1 , is negatively related to the
elasticity of substitution s.
4 Both elasticities are higher in sectors where firms’ productivity is less
dispersed (q large).

Thomas Chaney (Sciences Po) International Trade 11 / 17


Gravity equation

Similar structure, Chaney and Krugman,


8 ⇣ ⌘ q
>
< XijChaney = l LiLLj trij ( s q 1 1) sr
f ij = l N i sij r Lj
j Âk =1 k kj
⇣ ⌘ ( s 1)
>
: X Krugman = l̃ Li Lj tij s r̃
ij L r̃j = l̃ N i sij r̃ Lj
Âk =1 k kj
8
< ( q 1)
rij = tij q fij s 1
with
: r̃ = t (s 1)
ij ij

Exports from i depends on trade barriers from i relative to weighted


average of trade barriers from k’s.

Thomas Chaney (Sciences Po) International Trade 12 / 17


Intensive versus extensive margins

Average size of exporters versus number of exporters,


8 h i
< E rij ( j) | j jij⇤ = l0 fij
⇣ ⌘
: N = l Li Lj tij q f ( s 1 1)
q

ij 0
l fij L rj ij

Problem 1: relative exports conditional on export vary a lot,


✓ ◆1 s
rij ( j) tij
=
rik ( j) j jij⇤ ,jik
⇤ tik

Possible fix: t random shocks + Arkolakis (2008).


Problem 2: average exports decrease with distance.
Possible fix: global fixed export costs (Hanson & Xiang 2008).

Thomas Chaney (Sciences Po) International Trade 13 / 17


Intensive versus extensive margins

! ✓ ◆

∂rij ( j) ∂jij⇤
ˆ
⇤ ⇤
dXij = dG ( j) d tij r jij g jij ⇥ d tij
jij⇤ ∂tij ∂tij
! ✓ ◆

∂rij ( j) ∂jij⇤
ˆ

+ dG ( j) dfij r ( j̄ij ) g jij ⇥ dfij
jij⇤ ∂fij ∂fij
| {z } | {z }
Intensive margin Extensive margin

Thomas Chaney (Sciences Po) International Trade 14 / 17


Intensive versus extensive margins

d ln Xij
z⌘ = (s 1) +(q (s 1)) = q
d ln tij | {z } | {z }
Intensive margin Extensive margin
Elasticity Elasticity
✓ ◆
d ln Xij q q
x⌘ = 0
|{z} + 1 = 1
d ln fij s 1 s 1
Intensive margin
Elasticity
| {z }
Extensive margin
Elasticity

∂z ∂x
) = 0 and <0
∂s ∂s

Thomas Chaney (Sciences Po) International Trade 15 / 17


Some empirics (size distribution and gravity)

ln (Distanceij ) -.9 -.8 -.7


(.04)*** (.04)*** (.04)***
dqh
⇥ ln (Distanceij ) -.09 -.09
sh 1
(.002)*** (.003)***
Languageij .3 1.6 -.4
(.1)*** (.2)*** (.02)**
dqh
sh 1 ⇥Languageij -.4 .4
(.05)*** (.05)***
Borderij .8 3.9 1.4
(.02)*** (.3)*** (.3)***
dqh
sh 1 ⇥Borderij -.7 -.3
(.008)*** (.08)*
R2 30% 31% 23% 25% 32%
Number of obs. 65,687 65,687 65,687 65,687 65,687
Thomas Chaney (Sciences Po) International Trade 16 / 17
Some empirics (differentiation and gravity)

ln (Distanceij ) -.8 -1 -.9


(.02)*** (.02)*** (.02)***
ŝh ⇥ ln (Distanceij ) .02 .015
(.001)*** (.001)***
Languageij .4 1.2 .5
(.04)*** (.09)*** (.05)***
ŝh ⇥Languageij -.02 -.02
(.004)*** (.004)***
Borderij .5 2.3 .6
(.08)*** (.1)*** (.09)***
ŝh ⇥Borderij -.04 -.01
(.006)*** (.006)*
R2 39% 40% 33% 35% 41%
Number of obs. 270,607 257,583 257,583 257,583 257,583

Thomas Chaney (Sciences Po) International Trade 17 / 17

You might also like