Tariff De-Escalation

with Successive Oligopoly

Ian Sheldon

Short Course - Lecture 2

Catholic University of Milan

Cremona, Italy

June 11-22, 2012

Tariff Escalation

● Tariff escalation long-recognized issue in trade policy

literature, (Corden, 1966; Ethier, 1977; Anderson, 1998)

● Cadot et al. (2004) report nominal protection escalates

with degree of processing in both industrial and

agricultural goods

● Extent of tariff escalation highlighted as key issue

affecting developing country exports (UNCTAD, 2002;

World Bank, 2003; FAO, 2007)

● Provides rationale for formula approaches to reducing

tariffs, i.e., percentage reduction in higher tariffs

exceeds that for lower tariffs (Francois and Martin,

2003) – essentially in draft WTO modalities (WTO, 2009)

EU Japan US

Cocoa Primary

Processed

0

17

0

43.0

0

13

Coffee Primary

Processed

4

8

0

12

0

2

Sugar Primary

Processed

406

36

397

80

90

55

Fruits Primary

Processed

8

19

9

19

5

11

Oilseeds Primary

Processed

0

29

60

6

18

5

Grains Primary

Processed

30

34

67

153

1

3

Rice Primary

Processed

40

60

702

847

2

5

End-Uruguay Round Bound Tariffs (FAO, 2007)

Basic Result

● In vertically-related market, simultaneous and equal

reduction of upstream and downstream tariffs has non-

equivalent effects on upstream and downstream firms’

profits

● Result due to within (horizontal) stage and between

(vertical) stage impact of tariff cuts, where latter is

made up of pass-through and pass-back effects

● To extent firms are concerned about relative

profitability, outcome provides potential source of

opposition to tariff reductions

● Generates strong argument for tariff de-escalation

Literature

● Relates to literature on cascading contingent protection

where upstream tariffs have spillover effect, increasing

chance of tariffs downstream (Hoekman and Leidy,

1992; Sleuwaegen et al., 1998)

● Feinberg and Kaplan (1997) in analyzing US anti-

dumping cases found levels of protection upstream

have an impact on protection downstream

● Different, however, to literature on optimal tariffs in

vertically-related markets (Spencer and Jones, 1991,

1992; Ishikawa and Spencer, 1999)

● Paper also abstracts from explicit political economy

considerations in order to focus on mechanisms arising

with simultaneous tariff reductions

ux1

ux2Upstream:

Domestic Imports

x1Downstream: x2

Policy

Final Demand

Tariff - ut

Tariff - dtd ut t>

Stage

u

u u u

x x

x x x

1

1 2

=

= +

Technology:

Vertical Market Structure

Equilibrium

● Three-stage game:

(1) Government commits to ut and dt

(2)/(3) Nash equilibria upstream and downstream,

● Downstream revenue functions:

R x x1 1 2( , ) (1)

R x x2 1 2( , ) (2)

● Downstream profit functions:

dπ x x cR x1 111 1 2= ( , ) - (3)

d dπ x x t xcR x2 222 1 2 2= ( , ) - - (4)

Equilibrium

● First-order conditions are:

cR1,1 1 = (5)

2,2 2 = + dtcR (6)

● Nash equilibrium downstream:

1,11 1,12 1 1

2,21 2,22 2 2

=

c +

u

d

dpR R dx

d dtR R dx (7)

● Slopes of reaction functions:

1,121

1

2 1,11

= = - Rdx

rRdx

(8)

2,212

2

1 2,22

= = - Rdx

rRdx

(9)

Substitutes (complements), i,ijR < 0(> 0), ri < 0(> 0)

Figure 2: Strategic Substitutes vs. Strategic Complements*

X2 X2

X1 X1

N N

R2

R1

R2

R1

R2'

N'

R2'

N'

Strategic Substitutes Strategic Complements

* Bulow, Geanakoplos, and Klemperer (1985)

Equilibrium

● Solution found by re-arranging and inverting (7), and

simplifying notation:

1 2 1-1

2 2 1 2

- = Δ

- + d

a b dpdx

b a dc dtdx

1

u

(10)

where: 1,11 2,221 2 = = a aR R

1,12 2,211 2 = = b bR R ,

and for stability, i < 0a , and -1

1 2 1 2Δ = ( - ) > 0a a b b

● From (8) and (9), substitute i i i= -( ) /r b a into (10):

11 2 1 1-1

22 2 1 2

= Δ

+

u

d

dpdx a a r

dc dtdx a ar (11)

Equilibrium

● Upstream firms’ profits are:

1 1 1 2 1 1= ( , ) -u u u u u uπ R x x c x (12)

u u u u u u u uπ R x x c x t x2 2 1 2 2 2 2= ( , ) - - (13)

● Given technology, upstream Nash equilibrium is:

12 11 -1 1

22 12 2

( )Δ

uU U UU

U

uU UUU U

dca a rdx =

dta ardx dc + (14)

where for stability < 0u

ia , -1(Δ ) > 0u ,and also >u

i ia a ,

i.e., perceived marginal revenue steeper upstream

(see Lemma 1)

Incidence of Tariff Reductions

● To identify market access effects, assume initially that

(i) u ddt dt> 0, = 0, and then (ii) u ddt dt= 0, > 0:

■ Pass-through of udt :

1 1,1 1 2 1,1= ( + ) =u u u u u udp dt p dx dx p D

where 1 1,1/ < 0u u udp dx = p , and -1

1 1= {(Δ ) [ (1+ )]} < 0u u uD a r

Likely that 1,1 < 1up D , i.e., under-shifting of reduction in

upstream tariff (linear or weakly convex demand curve generates this result, Fullerton and Metcalf, 2002)

QT

MC

MC + T

D

MR

PT

Q

P

Figure 2: Pass-Through of Tariff Reduction

P

Q

Incidence of Tariff Reductions

■ Pass-back of ddt :

u u uu1

d u u d

dp dp d x xa r p

dt d x + x dt

u-11 1 2

1 1 1,1

1 2

( + )= = Δ (1+ )

( )

(a) -1

1 1 1,1Δ (1+ ) > 0 if < 0u

ia r p r - substitutes

(b) -1

1 1 1,1Δ (1+ ) < 0 if > 0u

ia r p r - complements

● Pass-through and pass-back effects not equivalent:

u u u u up a r a r p-1 -1

1,1 1 1 1 1 1,1(Δ ) [ (1+ )] Δ (1+ )

(see Lemma 2)

Tariff Reductions and Market Access

● Effect of lowering ut on market access:

-121= (Δ ) < 0

uu u

u

dxa

dt

(16)

■ Imports of intermediate good increase

-1 -1 u u u2 2 12 2 1,1 1 1

1

= = (Δ ) [(Δ ) (a (1+r ))]u

u

u u u

dx dx dpa r p

dt dp dt

(17)

u

dx > r

dt2

20 if < 0 or u

dx < r

dt2

20 if > 0

■ Imports of final good fall (increase) depending on

whether final goods are substitutes (complements)

Tariff Reductions and Market Access

● Effect of lowering dt on market access:

d

dxa a r r p

dt

-1 -1 u21 2 1 2 1,1= Δ [1+ Δ (1+ )] < 0

(18)

■ Imports of final good increase

-1 -12

1 1 2 1,1= (Δ ) [1+ Δ (1+ )]u

u

d

dxs a r a p

dt (19)

1 1 2= ( + )u udx d x x , so 2 1 1 1( / ) =1- ( / ) =u udx dx dx dx s

u

d

dx > r

dt2

10 if < 0 or u

d

dx < r

dt2

10 if > 0

■ Imports of intermediate good fall (increase) if final

goods are substitutes (complements)

Tariff Reductions and Market Access

● Net effect on market access of lowering ut and dt :

u uu u u

u d

dx dxa s a r a p

dt dt

-1 -1 -12 21 1 1 2 1,1+ = (Δ ) + (Δ ) [1+ Δ (1+ )] < 0

(20)

■ Imports of intermediate good increase, partly offset

by decline in derived demand downstream

u u u u

u d

u

dx dxa r p a r

dt dt

a a r r p

-1 -12 22 2 1,1 1 1

-1 -1

1 2 1 2 1,1

+ = (Δ ) (Δ ) [ (1+ )]

+ Δ [1+ Δ (1+ )] < 0

(21)

■ Imports of final good increase, as long as vertical

effect of upstream tariff reduction is not too great

Tariff Reductions and Market Access

● Which stage is most affected by change in access?

u d

u u u u u

u u u u

dt dt

a r p a r a a r r pdx

dx a s a r a p

-1 -1 -1

2 2 1,1 1 1 1 2 1 2 1,12

-1 -1 -1

2 1 1 1 2 1,1+

Δ (Δ ) [ (1+ )]+ [1+ Δ (1+ )]= <1

(Δ ) + Δ (1+ (1+ )Δ )

(22)

■ Final good imports likely to increase by less than

increase in imports of intermediate good (see Proposition 1)

● Result rationalizes why some firms may take a

different stance on trade liberalization, reinforcing need for formula reductions in tariffs

Tariff Changes and Profits

● By how much would dt have to change, given unit

reduction in ut , in order to keep change in domestic

firms’ profits equal between stages?

● Tariff rule is to find ˆddt such that:

ˆ

1 1

1 1

+

=

+

d uu

u u

d

d u

d d

dπ dπdt

dt dtdt

dπ dπ

dt dt

(23)

1 1 1 1> 0, > 0, < 0, > 0d u d u

d d u u

dπ dπ dπ dπ

dt dt dt dt

Tariff Changes and Profits

● (i) If ˆ  d udt / dt de -escalation>1, implies tariff

(ii) If ˆ  d udt / dt escalation0 < <1, implies tariff

■ Result (i) means percentage reduction in downstream

tariff should exceed that for upstream tariff

■ Result (ii) means percentage reduction in downstream

tariff should be less than that for upstream tariff

● When vertical effects coupled with horizontal effects,

effects of simultaneous tariff reductions may not have an equal effect on profits of firms located at upstream and downstream stages

Policy Implications

● Equal reduction in tariffs in vertically-related market

may result in greater impact on upstream (downstream)

firm(s) compared to downstream (upstream) firm(s)

● To extent vested interests oppose trade liberalization,

lobbying likely to come from upstream (downstream) –

not just because profits fall, but as profits fall by more

than downstream (upstream)

● Important justification for formula approaches to tariff

reduction – not just simpler negotiations, but also

formal basis in mechanisms arising in vertically-related

markets

● Potentially beneficial to developing country exporters