P-.

, ,'
EuropeanJournalof
POLITICAL
European Journal of Political Economy ECONOMY
ELSEVIER Vol. 14 (1998) 115-137

The political economy of retaliation,

liberalization and trade wars 1
David M. Gould a,*, Graeme L. Woodbridge b,2
a Federal Reserve Bank of Dallas, Dallas, TX 75222, USA
b . . . . .
Austrahan Competttzon and Consumer Commission, Level 35, 360 Elizabeth St., Melbourne,
Vic. 3001, Australia

Received 1 November 1995; revised 1 November 1996; accepted 1 February 1997

Abstract

W e examine the dynamic process behind protection, retaliation and trade wars. Consis-
tent with empirical evidence on the development of trade policies, we model policy
decisions as an outcome of political contests within two trading nations, rather than as an
outcome of a strategic game between two governments. Uncertainty about the incidence and
success of retaliation yields a dynamic political equilibrium in which one country imposes a
tariff that increases gradually over time. Eventually, the cost of the tariff to the other
country's exporting interests induces retaliation. W e show that depending on the character-
istics of the markets in the two countries, retaliation may encourage liberalization or may
cause a trade war. © 1998 Elsevier Science B.V.

JEL classification: C73; D72; F13

Keywords: Political economy; Trade liberalization; Retaliation; Trade policy; Game theory

* Corresponding author. Tel.: + 1-214-9225163; fax: + 1-214-9225194; e-mail:

david.re.gould @dal.frb.org.
l The views expressed in this paper do not necessarily reflect those of the Federal Reserve Bank of
Dallas or the Federal Reserve System, or the Australian Competition and Consumer Commission.
2 Tel.: + 61-3-92901859; fax: + 61-3-96633699; e-mail: [email protected].

0176-2680/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved.

PII SO 1 7 6 - 2 6 8 0 ( 9 7 ) 0 0 0 4 1 - 4
116 D.M, Gould, G.L Woodbridge/ European Journal of Political Economy 14 (1998) 115-137

1. Introduction

Under Section 301 of the Tariff and Trade Act of 1974 and Super 301 of the
Onmibus Trade and Competitiveness Act of 1988, the U.S. government has
demonstrated increased willingness to threaten retaliation against protected foreign
markets. In 1995 alone, the United States, through the use of countervailing duties,
antidumping actions and other sanctions, threatened to retaliate against, Canada,
the European Community, Korea and Japan in an attempt to improve its access to
their markets 3. In some cases, these threats have been successful. Japan buckled
under U.S. pressure and agreed to numerical targets on its auto parts imports. U.S.
threats against Korean exports of semiconductors led the Korean government to
further open its domestic markets to foreign competition. However, retaliation is
not always successful and in some cases has degenerated into a trade war 4.
What generates the decision to retaliate and when does retaliation bring about
liberalization? Traditionally, this question has been addressed in the context of a
strategic game between governments. In the game-theoretic approach to trade
policy, a government retaliates, liberalizes, or follows a mixed strategy to maxi-
mize its own country's welfare 5. However, as Stigler (1971), Peltzman (1976)
and others have argued, governments seldom pursue policies designed to maxi-
mize social welfare. Rather, governments maximize their political support and, in
doing so, implement policies that reflect the interests of the most influential and
vocal self-interest groups. Indeed, although the case for strategic trade policy is
appealing, the United States has rarely acted in this manner (Dixit, 1987). The
decision to retaliate against foreign nations is, arguably, as much an outcome of an
endogenous political process as is the decision to implement tariffs and other trade
policies 6. The decision to retaliate, or the decision to liberalize in response to a
threat of retaliation, is not determined by the exogenous objectives of the
governments but, rather, is determined by the relative influence of competing
interests within two countries 7

3 See Finger(1991) for a descriptionof the retaliatoryactionthe UnitedStates can take to improve
its access to foreign markets.
4 The Smoot-Hawley tariffof 1930 is an extreme example of retaliation and counterrctaliation
between the United States and its trading partners nearly halting world trade.
5 See Richardson (1990) for a broad survey of the recent literatureon strategictrade policy.
6 For a survey of the literaturcon the politicaleconomy of trade protection, see Hillman (1989) or
Quibria (1989).
7 The majority of the literatureconcerning thc politicaleconomy of protection is silent on the causes
of trade liberalizations.Notable exceptions arc Hillman and Moscr (1996) who model trade liberaliza-
tions as agreements between two countries that trade market access, Grossman and Helpman (1995)
who examine the development of trade policies in a two stage game between special interests,voters
and governments and Cassing ct al. (1986), who explain tariff cycles as changes in the political
demands of import-competing interestsand exporting interestsat various stages of the business cycle.
In contrast to this literature,we examine the dynamic political interaction between special interest
groups, which can lead to gradual increases in protection, retaliation,and episodes of liberalizationor
trade wars.
 D.M. Gould, G.L Woodbridge/ European Journal of Political Economy 14 (1998) 115-137 117

The contribution made by this paper is to enhance our understanding of the

dynamic political process behind the decision to retaliate and the response to such
threats of retaliation. We model the process of protectionism, liberalization and
retaliation as an interaction between the foreign political market for protection and
the home country's decision to retaliate against that protection. In contrast to other
recent models of the political economy of trade liberalization, in our structure the
threat of retaliation can play a crucial role in inducing a country to liberalize 8.
The model explains a number of stylized facts in tariff-setting behavior and
retaliation. First, it demonstrates that a government, responding to domestic
pressure groups, will delay retaliation until foreign protection reaches a critical
level. Retaliation will be the politically optimal policy only when the potential
gains to domestic exporters from a foreign liberalization are large. This will occur
when foreign protection is substantial. Second, the model explains the dynamic
pattern of tariffs. Protectionist policies are usually implemented on an industry
basis and tend to increase gradually over time, while trade liberalizations are
usually implemented at discrete points of time and tend to reduce protection at
once. Such liberalizations are often followed by subsequent protectionist pressures
that may partially or completely reverse the original liberalization (Marvel and
Ray, 1983; Ray, 1987).
We demonstrate that as the foreign tariff increases, information is gained about
the tolerance of the home country. This new information alters the political
equilibrium, increasing the equilibrium tariff. Eventually, if protection rises to
such an extent that it induces a threat of retaliation from a nation with a substantial
export market, foreign exporting interests will devote more resources lobbying for
free trade. This increased lobbying effort can generate an episode of liberalization.
If uncertainty about the potential for retaliation remains, protection can subse-
quently rise until another credible threat of retaliation is made and the process is
repeated.
An important finding of the paper is the positive role that large nations can play
in maintaining an open world trading system. Despite the potential for free trade to
maximize a country's aggregate welfare, incentives facing political parties lead
them to design polices to gain the support of pro-protection interest groups.
Without the discipline from trading partners, the political balance within a country
will remain in favor of pro-protection interest groups and policies inhibiting trade
will remain. The model predicts, however, that trading partners, through the threat
of retaliation, can alter the political balance in such a country and encourage
liberalization. Such action is not without its potential costs. Retaliation may have

s Hillman and Moser (1996) model trade liberalization as an exchange of market access which
permits governmentsto mutually increase their political support. Similarly, Grossman and Helpman
(1995) show that cooperative multilateral trade negotiations between governments can reduce the
political cost of acting non-cooperatively.
118 D.M. Gould, G.L Woodbridge/ European Journal of Political Economy 14 (1998) 115-137

the perverse effect of causing a trade war, which will further reduce access to
foreign markets.
Furthermore, because exporters have a great interest in maintaining access to a
major market, large countries, such as the United States, have the greatest power
to encourage other countries to liberalize. This is consistent with the observations
of Lipson (1982), Keohane (1984) and Baldwin (1989) who note that the trend
toward a more liberalized trading environment since World War II may have been
the result of the coercive hegemonic power of the United States. While it appears
from recent experience that the United States is large enough to act as a
hegemonic power in world trade, its relative size in the world market is diminish-
ing. The absence of a single large country may partly explain the emergence of
trading blocs. Large members because of their size, have the power to discipline
smaller countries that attempt to cheat on trade agreements 9. However, in broader
trading agreements, such as the World Trade Organization (WTO), a liberalized
trading environment may be more difficult to maintain if there is no single large
country. This concords with the recent trade dispute over auto parts between the
United States and Japan and the subsequent U.S. threat of retaliation outside the
purview of the WTO.
This paper is organized as follows. In Section 2, we set the background for the
model and describe the industry structure and the behavior of firms. Following
this, in Section 3, we describe the informational assumptions and the timing of the
political process. In Section 4, we describe an electoral contest in the home
country that determines the decision to retaliate, given the level of the foreign
tariff and uncertainty about the success of retaliation. The dynamic political
process that determines the foreign tariff is described in Section 5. We conclude
with some brief remarks in Section 6.

2. Industry structure and the behavior of firms

Consider trade between two countries (home and foreign) and two industries. In
each country there is an exporting firm and import-competing f'mn 10. The home
country exporting firm sends its entire production to the foreign market and
competes in that market with the foreign import-competing firm. Likewise, the
foreign exporting fu-m sends its entire production to the home market and
competes in that market with the home country import-competing f'Lrm. In every

9 Of course, the presenceof a free trade agreementis not a necessarycondition for a country to use a
threat of retaliation to coerce another nation into not raising its trade barriers.
~oThe assumption that there is a single firm in each country's exporting and import-competing
sectors is made for simplification. The model could be generalized to allow for many firms in each
sector. While this generalization would introduce a public-goodcharacter into the lobbying contribu-
tions of each industry (see below), it would not alter the flavor of the results from the model.
 D.M. Gould, G.L Woodbridge / European Journal of Political Economy 14 (1998) 115-137 119

period, firms make production decisions and can observe how the market reacts to
those decisions before the beginning of the next period. Because all market
interactions take place within each period, for simplicity we exclude time sub-
scripts.
The goods produced by the home country exporting firm and the foreign
country import-competing fn'm are imperfect substitutes in consumption 1~. In the
foreign country, the demand for home-country exports and domestically produced
goods are
b
Px=a---X+TxPx (1)
M
and
b
Px* = a - - - X * + 3,xP~ (2)
M
where X ( X * ) is the quantity of the home (foreign)-produced good supplied to
the foreign market, Px ( P x ) is the price of the home (foreign)-produced good in
the foreign market and M is the size of the foreign market. 3'x represents the
substitutability in consumption between the two goods as well as market size
(0 < 3'x < 1). The larger 3'x, the greater will be the fall in foreign demand for the
home-country export good when the price of the foreign-produced good
decreases ~2. The parameters a and b are positive.
The goods produced by the home country import-competing firm and the
foreign country export firm are also imperfect substitutes in consumption and are
modeled symmetrically to the foreign country's demand. In the home country, the
demand for domestically produced goods and foreign exports are Py = a -
( b / N ) Y + TyPy* and Py = a - ( b / N ) Y * + 3'yPy, respectively, where Y (Y*) is
the quantity of the home (foreign)-produced good supplied to the home market, Py
( P y ) is the price of the home (foreign)-produced good in the home market and N
is the size of the home market. ~/y represents the substitutability in consumption
between the two goods as well as market size (0 < Ty < 1).
Assume that there is no possibility for entry into an industry and that all firms
have no costs of production 13. Each firm within an industry selects an output level
to maximize its profit, given the output level of its competitor. The home firm
exporting to the foreign country faces a specific tariff, t*, while the foreign fLrm
exporting to the home country faces a specific tariff, ty.

II This market structure is similar to the one used by Hillman and Ursprung (1988). They analyze the
political choice between tariffs and voluntary export restraints.
12 For a discussion of the form of the utility function underlying these demand functions, see Harris
(1985).
13 This assumption is made for simplicity. The results from the model still hold if one assumes
constant nonzero marginal costs of production.
 120 D.M. Gould, G.L Woodbridge/European Journal of Political Economy 14 (1998) 115-137

n;I.,:-,.,

n;I.,. =-,.o

Hx [ r,,,=-r,,~
n,,l-r,,_-

o 1 ~*~
Fig. 1. Profit functions of the domestic export firm (Hx) and the foreign import-competing firm (H~*)
(Yxt > %o)-

The Nash equilibrium profit function of the h o m e exporting finn is

M [ ( 1 + y x ) a 2 ] (1 - 13" )2
I-I~(flx*)-~ b(l_yx)(2+yx) 2 (3)

and the profit function of the foreign i m p o r t - c o m p e t i n g firm is

M [ ( 1 + 'y~)a2] (1 + ( Tx/3x"/2))2
llx* ( / 3 x ) = b(1 - y x ) ( 2 + yx)2 (4)

where fix* = t ~ / t ~ is the foreign tariff as a proportion of the prohibitive tariff,

-* 14
tx

14 The prohibitive foreign tariff (i.e. the tariff that makes it unprofitable for the home exporting firm
to supply any quantity to the foreign market) is t~" = (2 -- yx)a/2(1 - 7x), while the prohibitive home
tariff is ~r = (2 - yy)a/2(1 - 'yy).
 D.M. Gould, G.L Woodbridge/ European Journal of Political Economy 14 (1998) 115-137 121
In the other industry, the profit function of the home import-competing firm is

N[(I + Ty)a2](l + (Ty/3y/2)) 2

//,( &)
b(1 _ Ty)(2 + T,)2 (5)

and the profit function of the foreign exporting firm is

/-/Y'(J~Y) = N[(1 -]='~y)a2](l _~y)2

b ( l _ Ty)(2_F Ty)2 (6)
where fly is the home tariff as a proportion of the prohibitive tariff.
As displayed in Fig. 1, the profit of the foreign import-competing firm is
increasing in the foreign tariff, /3~* and the profit of the home exporting firm is
decreasing in /3~'. Both profit functions are strictly convex. An increase in the
foreign tariff entices foreign consumers to substitute away from the imported good
toward their own domestically produced good. This substitution increases the
output and profit levels of the foreign import-competing firm at the expense of the
home exporting firm. The larger is 3'x, the greater will be the substitution toward
the foreign-produced good and the larger will be the impact of an increased tariff
on profit (i.e. 6 2IIx* /8¢3 x*8y x > 0 and 8 217x/8~ x,67~ < 0).

3. Timing, politics and informational assumptions

Before the model is fully solved, in this section we describe key assumptions
behind market information, the political process in the home and foreign countries
and the timing of events. The primary purpose of this section is not to completely
describe the model, but rather to make clear the model's fundamental underlying
assumptions.

3.1. Market information

Exporting and import competing fLrrns in the home and foreign countries are
assumed to operate with limited information. Outside each country's domestic
market, firms lack some information, either about the market or the political
process, which makes them unable to predict with complete accuracy how
overseas markets react to domestic trade policy decisions.
For example, U.S. firms will not know with complete certainty how the
Japanese will respond to a threat of retaliation. U.S. firms do not know whether a
retaliation threat will cause a liberalization or a trade war with Japan. They have
enough information, however, to form a probability distribution over possible
Japanese reactions, but the outcome is not known with certainty. Likewise,
Japanese firms cannot predict with complete accuracy when the United States will
122 D.M. Gould, G.L. Woodbridge / European Journal of Political Economy 14 (1998) 115-137

.o

•I-.
J\
~y = 1 ~y = o
11 ~ . -t.,

c~

0
~= 1

"O "~ free trade

;I "
.o .o
o
r~
trade war

etc.
Fig. 2. The dynamic structure of the political game.

threaten to retaliate against Japanese protection. Like firms in the United States,
Japanese fLrrns lack at least one piece of information that prevents them from
completely determining the behavior of the other country.
These informational assumptions are obviously an abstraction from reality.
Domestic finns probably never have full information about their own markets nor
do they always have to form a probability distribution over how foreign markets
will behave. However, the model's assumptions try to approximate reality. Firms
tend to know much more about their own domestic markets than they know about
markets abroad.

3.2. Timing and the political process

Fig. 2 summarizes the dynamic structure of the political process 15. At the start
of the first period, the foreign country, through the political process, determines
what its tariffs will be for that period ( fix* ). The political parties announce their
tariff platforms so as to maximize their political support through the lobbying
contributions of export and import competing industries. As in Hillman and

15 We are indebted to an anonymous referee for suggesting Fig. 2 to us.

 D.M. Gould, G.L. Woodbridge/ European Journal of Political Economy 14 (1998)115-137 123

Ursprung (1988), the political parties use these resources to sway voters, who are
imperfectly informed about the candidates' positions. Industries, in making their
lobbying decisions, weigh the increased probability of their favorite party being
elected against the direct cost of the donation ~6. Foreign import-competing fLrms
lobby for high tariffs because they want greater profits in their sector while foreign
exporters lobby for low tariffs because they worry that high tariffs will cause the
home country to retaliate against their exports. In this set-up, we assume that the
foreign country decides to impose tariffs on the home country and then the home
country decides whether to threaten retaliation 17
In the second period, after the foreign country's optimal tariff is determined,
the home country then decides, though the political process, whether to threaten
retaliation against the foreign country or not (that is, to set /3y = 1 or/3y = 0) 18.
Import-competing firms in the home country always lobby for retaliation because
of the potential that this might generate a trade war with high tariffs. Exporters
always lobby against retaliation when foreign tariffs are low, but lobby for
retaliation when foreign tariffs are high. The higher are foreign tariffs, the less
home country exporters will lose if the retaliation causes a trade war, and the more
they will have to gain if it leads to a liberalization in the foreign country.
In the third period, if the home country does not retaliate after the foreign
country determines its politically optimal level of tariffs, this provides information
to foreign country's firms. With this newly learned information they update their
expectations about when the home country will retaliate and increase tariff levels
in the next period. The continues until a threat of retaliation is made by the home
country.
If the home country threatens retaliation, the foreign country can either
liberalize or raise protection and move towards antarky. This depends on the
relative lobbying ability of the export and import-competing firms in the foreign
country. Once a threat of retaliation is made by the home country, the foreign
country knows the tariff level that triggers retaliation and the tariff cycle is over.

16There are several methods used in the trade policy literature to endogenize the political process.
First, there is the reduced-formpolitical support function based on the work of Stigler (1971), Peltzman
(1976) and Hillman (1982). Here a government's political support depends directly on its policies and
indirectly on how its policies effect the rents accruing to special interest groups. Second, there is the
methodologydeveloped by Brock and Magee (1978) and more fully developed in Magee et al. (1989)
that models the lobbying activity of special interest groups and how it effects the probability of
different political parties winning an election. Our modeling strategy follows the second strand of
literature.
17Which country imposes the tariff or the mechanism of protection is not important. The aim is to
model, in a political economy framework, the protectionist behavior of one country under the
possibility of retaliation and the decision of the other country to retaliate in response to the protection.
Is The threat of retaliation is always credible because, in the context of this model, foreigners can
always observe the outcome of the domestic political process and firms have the incentive to reveal
their true preferences in the political arena.
124 D.M. Gould, G.L. Woodbridge /European Journal of Political Economy 14 (1998) 115-137

Only until new firms enter the political process and uncertainty about home
country retaliation returns will the cycle repeat, but this is not determined in the
model.
The dynamics of the model are driven by learning. In each period the political
equilibrium changes because more is learned about the tolerance of the home
country. A key simplifying assumption is that firms and politicians are concerned
about setting tariffs to maximize profits and political support over the next
political cycle and not subsequent terms. This assumption allows for more
complex within-period behavior and accords well with the lobbying behavior in
many developing and developed countries 19

4. T h e electoral contest in the h o m e c o u n t r y a n d the d e c i s i o n to retaliate

Initially, assume that trade between the two countries is free of tariffs, but then
the foreign country decides to impose a tariff on the home country's exports to the
foreign market. How will the home country respond to foreign tariffs on its
exports?
Because a foreign tariff reduces the access of the home exporting fLrrn to the
foreign market, the home country decides whether to retaliate. Both the level of
the foreign tariff and the decision to retaliate are determined by electoral contests
within the two countries 2o. Decisions to lobby for or against a tariff are made
under uncertainty about whether the tariff will incite retaliation and whether the
retaliation will be successful, In this section, we consider the political choice
within the home country to retaliate in response to a foreign tariff. This is followed
by an analysis of the determination of the foreign tariff itself.
The decision by the home country to retaliate is determined by an electoral
contest between two political parties 2t. To distinguish the political parties, assume

t9 The tendency of lobby groups to concentrate on the short-run is not a new observation. Magee et
al. (1989, p. 110) find that "trade policy lobbying in the United States is a short-term phenomenon."
Their evidence is based on U.S. voting behavior that supports the short-run, Ricardo-Viner fixed factor
model over the longer-run, Stopler-Samuelson theorem. They speculate that periodic policy reviews
cause agents to weight the short-run more heavily. In developing countries where the political
environment is much more volatile, agents are perhaps even more likely to concentrate on the
short-ran.
20The only policy instrument available to the politicians is a specific tariff. Issues relating to the
choice of the policy instrument are not considered in this paper.
21The essential feature here is that the decision to retaliate is the outcome of the political process in
the home country and is not generated by a strategic game between the home government and the
foreign government. See Riezman (1982) for an analysis of the use of retaliation in a strategic game.
 D.M. Gould, G.L. Woodbridge/ European Journal of Political Economy 14 (1998) 115-137 125

that one party (the protectionist party) is predisposed toward retaliation while the
other party (the liberal trade party) is predisposed toward a policy of no retaliation.
This means that if the two parties adopt different policies, the protectionist party
will propose retaliation and the liberal trade party will propose a policy of no
retaliation. However, just because the liberal trade party has a predisposition for
no retaliation, this does not mean it will exclude retaliation as a platform. It may
adopt the platform of the protectionist party if that is the policy which will
maximize its political support. The aim of each political party is to maximize its
probability of election, which is determined by the campaign contributions re-
ceived from the home country firms 22. The political parties have a discrete policy
choice between retaliation (R) and no retaliation (NR) given the foreign tariff, /3x .
Retaliation involves setting a prohibitive tariff (/3y = 1), which is removed if the
foreign country liberalizes 23. The decision not to retaliate leaves the home tariff
unchanged (/3y = 0).
As we show below, retaliation can lead to one of two outcomes. Either the
foreign country liberalizes, which re-establishes free trade (i.e. /3~* = 0 and
fly = 0), or the foreign country counter-retaliates which leads to a trade war and
autarky (i.e. /3x = 1 and fly = 1). The outcome of retaliation is unknown. Assume
that the home firms believe the probability of successful retaliation is ot (0 < a <
1). If the home country retaliates, the expected profit of the home import-compet-
ing firm is

E/-/~ = N [ ( 1 + Ty)a2]
. . . . .
L[l+ ( 1 - a ) y y ( l ~
. .
+ Ty (7)
b(1 - ~,y)(2 + ~,y)

and the expected profit of the home exporting firm is

Eli?= ctM[(1 + Tx)a 2]

(s)
b(1 - Tx)(2 + Tx) 2

22The political model used here is designed to capture the essential features of a representative
democracy in which pro and counter forces contest a policy via lobbying expenditures. Electoral
contests may also be decided by the preferences of voters in conjunction with campaign expenditures.
Mayer (1984) and Mayer and Riezman (1987) determine electoral success using majority voting, while
Young and Magee (1986) combine voting behavior and campaign contributions. See HiUman (1989)
for a discussion of models of endogenous protection under representative democracy,
23 For simplicity, the size of the retaliatory tariff is not a policy decision, although it could be
endogenized. Typically, retaliatory tariffs are punitive in nature and are set at levels much higher than
those set in normal periods. We do not address the question of why retaliatory tariffs are typically set
so high, although we speculate that it may be due to the high fixed costs of getting into the political
process and lobbying for a retaliation (i.e. those referred to by Olson (1965, pp. 10-11)). Conse-
quently, when threats of retaliation do occur, they tend to be large once the fixed costs of getting into
the political process are overcome.
126 D.M. Gould, G.L. Woodbridge/ European Journal of PoliticalEconomy 14 (1998) 115-137

If the home country does not retaliate, the profits of the two finns are

H ~ R= N [ ( I + Ty)a 2]
(9)
b(1-yy)(2+yy) 2

and

HxNR = M[(1 + yx)a2] (1 - fix* )2

(10)
b(1 - T~)(2 + Tx) 2

Assuming risk neutrality, the home import-competing firm will prefer a policy of
retaliation independently of the level of fix* (i.e. EII7R > H ~ r~ for all /3~*).
However, the home exporting firm will have a larger expected profit under
retaliation only if /3~* is greater than [ 1 - (a)1/2]. In the case in which the
interests of the firms are opposed, the decision to retaliate will be the outcome of
an electoral contest. Assuming that the parties propose different policies, the
profits of the fLrmS will depend on the outcome of the election and they will
contribute to the campaign of one of the two political parties. The probability that
the liberal trade party wins the election is given by

Zx
0 = ~ (11)
L~ + Ly

where L~ is the campaign contribution of the home exporting firm to the liberal
trade party and Ly is the campaign contribution of the home import-competing
firm to the protectionist party 24
The import-competing firm will select Ly to maximize

EIIy =OIIy~ + ( 1 - O) E I I ~ - Ly. (12)

Setting 6E1-ly/6Ly equal to zero yields the import-competing firm's campaign

contribution reaction function, given L x. Likewise, setting 6 E H J 6 L x equal to
zero yields the exporting firm's campaign contribution reaction function, given
Ly.
Solving for the Nash equilibrium levels of Ly and L x, given these policy

24Foreign interests cannot directly influence the outcome of the contest by contributing to the
campaigns of the home political parties. Hillman and Ursprung (1988) develop a model in which
foreign finns contributeto the politicalparty pre-disposedto a liberal trade policyand hence influence
the choiceof policyinstrument.
 D.M. Gould, G.L. Woodbridge/ European Journal of PoliticalEconomy 14 (1998)115-137 127

reactions, yields Ly(NR, R) and Lx(NR, R), which, in turn, determine the odds of
the protectionist party being successful in a political contest. That is,
1- 0 Ly(NR, R)
S(NR, R) = ~ = (13)
0 Lx(NR, R)

Proposition 1. If a political contest develops in the home country, the home

country will retaliate if and only if the foreign country's tariff, fix*, exceeds some
critical level, bx .

Proof If/31 > 1 - (or) 1/2, there is no political contest because both firms favor
retaliation. However, if fx* < 1 - ( a ) 1/2, the interests of the home country fLrms
diverge. In this case, the import-competing fuan prefers a policy of retaliation,
while the exporting fn'm prefers a policy of no retaliation. The policy will be
decided by the outcome of the electoral contest described above. To demonstrate
that the home country will only retaliate if the foreign tariff is higher than some
critical level, we show that there exists a critical tariff, b], above which both
parties advocate retaliation and below which they both advocate no retaliation. We
initially assume that they propose different policies and show that this cannot be
an equilibrium.
If the parties propose different policies, the odds of the protectionist party being
successful in the contest are given by
1- 0 Ly(NR, R) AE/-/y _-_ NB(1 -- c~)Tr(1 + Ty/4)
S(NR, R) =
0 Lx(NR, R) AEIVIx Ma[(1-flx )2-a]
(14)

where A = (1 - yy)(1 + yx)(2 + yy)2, B = (1 - yx)(1 + Ty)(2 + yx)2, A EIIy =

EII R - HyNR, and A E I I x = FixNR - E I I R.
Solving for S(NR, R) = 1 gives the critical tariff, b~*:
1/2
+ (yy/4))
h; = 1 - l ot + ( 1 - ot) NByy(1
MA (15)

If fix* > bx*, the odds of the protectionist party winning the election are greater
than one-to-one; if f x < b~*, the liberal trade party has a greater chance of
electoral success. Because the aim of the political parties is to select a policy to
maximize their political support, an outcome in which the parties propose different
policies cannot be a political equilibrium. They will both propose the same
platform to maximize their political support. For example, if fix* > b~* and the
liberal trade party chooses a platform of no retaliation, the party will have less
than a one-to-one chance of electoral success. However, if the liberal trade party
128 D.M. Gould, G.L Woodbridge / European Journal of Political Economy 14 (1998) 115-137

chooses the same policy as the protectionist party, firms will be indifferent
between the parties and the liberal party will have an equal chance of electoral
success 25. Consequently, the liberal trade party will alter its policy to match that
of the protectionist party (i.e. it will advocate retaliation). Likewise, if fix* < b~*,
the protectionist party will increase its chances in the election by altering its
retaliation policy to one of no retaliation. So, if fix* > bx, the home country will
retaliate against the foreign country, and if /3~* < bx* the home country will not
retaliate 26. []

5. The electoral contest in the foreign country and the dynamic tariff
equilibrium

The level of the foreign tariff is determined by an electoral contest between a

foreign protectionist party and a foreign liberal trade party. The liberal trade party
selects a level of the foreign tariff, /3x~, to maximize its probability of election,
while the protectionist party selects the tariff level, flx~- It is assumed that the
liberal trade party has a predisposition for lower tariffs, so flxo </3~. Similar to
the political contest in the home country, the probability that the liberal trade party
wins the election in the foreign country is 0 * = L*y/[L~ + L~ ], where Ly is the
campaign contribution of the foreign exporting firm to the liberal trade party and
L~ is the campaign contribution of the foreign import-competing firm to the
protectionist party. While the foreign firms are aware that a tariff may generate a
retaliatory response, they are uncertain of the tolerance level of the home
country 27. As shown above, the home country will only retaliate if the foreign
tariff exceeds the critical tariff, b~*.
The foreign firms do not know b~* and assume that it has the following
cumulative probability density:
F(b~)=b~ .2 b~* ~ [ 0 , 1 ] . (16)
Consequently, for a given foreign tariff level, /3~*, the foreign exporting firm
believes that the probability of retaliation is fix*2. If the home country retaliates,
the foreign exporting firm is excluded from the home market and earns zero profit.
In the event retaliation is avoided, the foreign exporting firm faces a zero tariff and

25 In the Hillman and Ursprung (1988) framework, political parties are constrained from duplicating
the policies of their opponents. The protectionist party must always represent higher tariffs and the
liberal party must always represent lower tariffs. The approach taken in this paper allows terms to
switch platforms and, hence, a Hotelling equilibrium develops. See Hofer and Woodruff (1994) and
Hiilman and Ursprung (1994) for more discussion of this topic.
26 The probability that /3x* = b~* lies in a set of measure zero, so this outcome is ignored.
27 The actual imposition of a retaliatory tariff need not occur. All that is required is the credible
announcement that retaliation will occur. Depending on the consequent actions of the foreign country,
the imposition of a retaliatory tariff may not occur.
 D.M. Gould, G.L. Woodbridge/ European Journal of Political Economy 14 (1998) 115-137 129

n~
En~

ry=vy,

~y='fyo

o l ~*,
Fig. 3. Profit and expected functions of the foreign import-competing firm (//** ) and foreign export
firm (E//:) (~,~ > ~,o, ~yo> "/yo).

earns profit of Hy*(0). Consequently, the expected profit of the foreign exporting
firm for a given tariff, /3~* is
v,)](1-
EHy*(~;)=(1-~x2)H,*(O) = b(l_yy)(2+Ty)Z (17)

The potential for retaliation yields a direct relationship between the foreign tariff
and the expected profit of the foreign exporting firm. As depicted in Fig. 3, the
expected profit of this firm is a decreasing concave function of the tariff.
Furthermore, the larger is the substitutability between foreign and home country
products, yy, the more valuable is access to the home market for the foreign
exporting farm and, hence, the larger will be the rate of decline of its expected
profit as the foreign tariff increases (i.e. 8 2Elly • /6~x •83'y < 0). On the other
hand, the profit of the foreign import-competing firm is increasing in the tariff.
Because the profit of foreign import-competing and export firms depend on the
foreign tariff, the firms contribute to the campaigns of the two political parties. For
130 D.M. Gould, G.L. Woodbridge/ European Journal of Political Economy 14 (1998) 115-137

given policy pronouncements of protectionist and liberal trade parties, the cam-
paign contributions of the foreign import-competing firm and foreign exporting
firm are L~([3xo, [3x~)and Lr([3x0, [3xl), respectively.
Define S* ( [3gO, [3~*!) as the odds that the protectionist party wins the election 28.

L*~( [3~*o, [3x*~) A IIx*

s'(';°'[3x')= Ly(,xo,[3x,) aEny* (18)
where A IIx* = Fix* ([3~) - IIx* ([3;0) and A ElIy* = EI-Iy* ([3x*O) - EMy* ( [3;! ).
The protectionist party will select its proposed tariff, [3x~, to maximize its odds
of electoral success, given the proposed tariff of the liberal trade party, [3~o-
Likewise, the liberal trade party will select its proposed tariff, [3,~, given the
proposed tariff of the protectionist party, flx~, to maximize its chance of political
success. The political equilibrium will occur when neither party finds it in its
interests to alter its tariff policy.
The foreign tariff consistent with the political equilibrium in this model is not
static. The equilibrium tariff changes from one period to the next, depending on
the home country's retaliation response in the previous period. If the previous
tariff was less than the critical tariff, b~, there is no retaliation, and some
information is gained about the home country's tolerance level of the foreign
tariff. If, on the other hand, the home country retaliates, the foreign political
balance changes as the foreign exporting firm now observes with certainty that its
profit will be zero if the present tariff level stays the same or increases.

5.1. Equilibrium in the first period

Proposition 2. In the first period, the foreign political equilibrium will be a

Hotelling's equilibrium, in which the tariffs selected by the foreign political
parties converge to fl; =2MyxA/[nNB-MYx2A], such that 811x*/Sfl* =
IS E H j / 8 [ 3 ; I and 0 < fl; < 1.

Both political parties converge to the same tariff, /3;, because deviating from
this tariff would generate greater political support for the opposing party 29. The
tariff equilibrium is depicted in Fig. 4. The size of the equilibrium tariff depends
on the substitutability of the home and foreign-produced goods in the two markets

2a S * (., .) is definedonly if/3*o #=/3x*l(which, in this model, implies /3xo</3"1. If/3*0 =/3x~, then
A l-Ix* = A EIIy* = 0 and S" is not defined.
29 Unlike the Hiliman and Ursprung (1988) model, an internal equilibrium is possible because the
expectedprofit functionof the foreignexportingfirm, shown in Fig. 3, is concave,whereas the profit
function of the foreign import-competingfirm is convex. The formal proof of Proposition 2 can be
found in AppendixA.
 D.M. Gould, G.L. Woodbridge / European Journal of Political Economy 14 (1998) 115-137 131

n; (o) ~ ~ ~" End" / ,

I ~ ~ ~ ~ 1
I
I I I ',
i
I I I I
I I I I
I I I

I I

~-x I I

t I

Fig. 4. Dynamic tariff equilibrium in the foreign country.

(3'x and 7y), as well as on the market sizes (M and N). The higher the degree of
substitutability between home-country exports and foreign import-competing prod-
ucts, 3'x, the greater will be the gain to the foreign import-competing firm from an
increase in the tariff. This substitutability increases the willingness of the fu'm to
contribute to the protectionist party, resulting in a larger equilibrium tariff. On the
other hand, the larger is the degree of substitutability between home country
import-competing products and foreign exports, 35 , the greater is the expected loss
to the foreign exporting fn'm from an increase in the tariff. As the foreign
exporting firm stands to lose more from retaliation, increases in the tariff are
strongly opposed through larger contributions to the liberal trade party. This acts
to reduce the equilibrium tariff. Market size also plays an important role in
determining the size of the equilibrium tariff. The larger is the foreign market
(M), the more valuable is the tariff to the foreign import-competing firm. This
increases the equilibrium tariff. However, the larger is the home market (N), the
greater is the loss to the foreign exporting firm if it is denied market access
through retaliation. This, in turn, decreases the equilibrium tariff.
132 D.M. Gould, G.L. Woodbridge / European Journal of Political Economy 14 (1998) 115-137

5.2. Equilibrium in the second period

After imposition of the tariff by the foreign country, the home country will
decide to retaliate, or not to retaliate, depending on the relative size of foreign
tariff, /3~', and the home-country critical tariff b~*. In either case, the initial
political equilibrium in the foreign country will no longer be the initial tariff. The
two cases will be analyzed in turn.

5.2.1. Home country does not retaliate ( /3~ < b~ )

The fact that the home country does not retaliate yields information about the
critical tariff. The foreign firms realize that the critical tariff, b~*, lies above the
current tariff, /3x*. This realization alters their belief about the probability of
retaliation. It is now known that b~* ~ [/3x*, 1]. The probability of retaliation
becomes (/3; -/3~* )2/(1 - / 3 ; )2, which is between 0 and 1 f o r / 3 ; ~ [/3~*, 1]
and strictly increasing in /3~*.
The expected profit of the foreign exporting firm becomes

EI-ly*'(/3; ) =/-/r(0) for/3x* </3x* (19)

and

/3. _a.
r-x___ ,-X 21
Z / / 7 * (0~ for/3 x > / 3 ;
en/'(/3;) = IX (1_/~,)2 J - ' r "-"
(20)

The revised expected profit function is displayed in Fig. 4. The equilibrium tariff
is no longer the original tariff, /3;. At/3x, we find 6Hx*/Bg; > I~eIIy*/6/3; I,
indicating that any increase in the tariff will raise the profit of the foreign
import-competing faro by more than it decreases the expected profit of the foreign
exporting frith. Consequently, if the protectionist party increases its proposed
tariff, /3xl,
* above /3-*x , the party will increase its probability of election. As the
protectionist party increases its tariff, the liberal trade party will follow. A new
internal equilibrium (/3~*) will be established at the point 8/-/%*/8/3; =
I En/'/8/3; I. Solving gives/3" = c / 3 ; , where
4NB~x + 2Ma3'x( 1 -/3x* )2
C= (21)
4NB~x -MA3'x2~; (1 -/3~* )2

If /3; is strictly an internal tariff, then C ~ [1, 2]. The political process will
generate a less than proportionate increase in the tariff that asymptotically
approaches 1 (8C/8/3; < O, 82C/60 ]2 > 0, and lima; _. 1C -- 1). This less than
proportionate increase in the tariff will continue so long as there is no retaliation.
 D.M. Gould, G.L Woodbridge/ European Journal of PoliticalEconomy 14 (1998) 115-137 133

5.2.2. Home country retaliates ( ~ >_b x )

If the home country retaliates, the foreign country faces the following trade-off:
fly = 1 if/3~* > 0 or, fly = 0 if/3~* = 0.
Consequently, the foreign exporting firm knows that its profit will be zero if the
tariff does not fall and will be positive if the foreign country liberalizes and
removes its tariff://y* ( fix ) = 0 for all fix* > 0 or IIy* (/3 x ) = Hy* (0) i f / 3 ; = 0.

Proposition 3. If the home country retaliates, the foreign political equilibrium

will be one in which the political parties announce a trade war if MAYx(1 + Tx/4)
> NB, or free trade ifMATx(1 + Tx/4) < NB.
The formal proof of the proposition can be found in Appendix A. The
interesting aspect of this proposition is that the home country's threat of retaliation
will not always move the foreign country toward liberalization. It can have the
perverse effect of inciting a trade war that pushes the countries toward autarky.
For retaliation to achieve liberalization, it must be the case that trade is relatively
more valuable to the foreign country exporting firms. That is, it must be the case
that M / N < B / [ A T x ( 1 + 7 J 4 ) ] which depends on the substitutability of the
commodities in the two markets and the relative market sizes. Both the size of the
foreign market (M) and the substitutability of the foreign good for the home good
(Tx) increase the value of a closed market for the foreign import-competing firm.
Consequently, the more valuable is the domestic market to the import-competing
firms, the greater is the likelihood of a trade war. On the other hand, the larger is
the size of the home country's market (N) and the substitution parameter (Tr), the
more valuable is trade for the foreign exporting firm and the greater is the
likelihood of a liberalization.

6. Concluding remarks
The primary contribution of this paper is to show that there is a dynamic
political process behind protection, liberalization and retaliation that is consistent
with stylized facts in tariff-setting behavior. We often observe periods of gradually
increasing protection followed by trade liberalizations, which are, in turn, some-
times partially reversed by further periods of increasing protection (Marvel and
Ray, 1983; Ray, 1987). Without altering the structure of the economy or the
composition of coalitions, we find that protection may yield a retaliatory response
and can generate a dynamic tariff equilibrium and episodes of trade liberalization.
Uncertainty about the reaction of a trading partner to protection generates a
political outcome in which the tariff increases over time at a decreasing rate until
it is reversed by an episode of trade liberalization. If the uncertainty concerning
the incidence and the outcome of retaliation remains-because of changes in the
characteristics of the home and foreign markets, for example-the tariff cycle will
perpetuate. Of course, retaliation can degenerate into a trade war, which is also an
outcome of the model.
134 D.M. Gould, G.L Woodbridge / European Journal of Political Economy 14 (1998) 115-137

Acknowledgements

We thank Michael Cox, Joseph Haslag, Greg Huffman, Carsten Kowalczyk,

Roy Ruffin, Ping Wang, Mark Wynne and two anonymous referees for helpful
comments.

Appendix A

Proof of proposition 2. Assume that the parties announce different tariffs

(13x~ > J3x0).
Is it in the interests of the two parties to increase or decrease their tariffs?
The protectionist party will select its proposed tariff, /3x~, to maximize S*,
while the liberal trade party will select its proposed tariff, fix*O, to minimize S*,
where

Ma[47x + ~/~( flx*l + fix*O)]

S* ( [3xo, [3~1) = 4 NB( flx~* + flxo (A.1)

It is easy to show that 8S * /813~1 *

= 8S * /813xo* = - M A T J [ N B ( [3xl* + [3xO)
* 2
] < O.
Consequently, the protectionist party will increase its probability of election by
decreasing its proposed tariff towards the liberal party's proposed tariff for any
proposed tariff, ¢lx~, which is greater than the liberal party's proposed tariff, /3~*O.
Likewise, the liberal trade party, will find that it is in its own best interests to
increase its proposed tariff, /3x0, toward /3xl for any [3gO less than [3xr As a
result, the tariffs announced by the two parties converge and, hence, the equilib-
rium will be a Hotelling's equilibrium.
If both parties propose the same tariff, ( flx*O=/3x~ =/3x* ) then foreign firms
will be indifferent between the two parties and both parties will have an equal
chance of electoral success. However, proposing the same tariff will only be an
equilibrium if neither party can gain political support by deviating from the
common tariff.
In order for a party to gain political support and increase its probability of
election, it must be able to deviate from the common tariff, ~ * , and improve its
odds of election above one-to-one. We know that if it is in the interests of either
party to deviate, the deviation will be only a small amount (i.e. if S*(/3~',
/3~* + e) > 1 for any positive e, it will be so for E close to zero). However, if any
party alters its original tariff from/3~* to/3x + ~, this will result in a difference in
the profits of the foreign firms under the alternative policies. This difference will
generate campaign contributions. In the case of the protectionist party, the only
way for it to generate more contributions than the liberal trade party is if the
marginal increase in the profit of the import-competing firm is greater than the
marginal fall in the expected profit of the exporting firm from an increase in the
 D.M. Gould,G.L. Woodbridge/ EuropeanJournalof PoliticalEconomy14 (1998) 115-137 135

tariff, 6/-/x*/813 x > [~EHy*/6/3~* I. But this cannot be an equilibrium because the
liberal trade party would then want to change its policy to match that of the
protectionist party. Consequently, a common tariff in which the marginal increase
in profit of the import-competing firm is greater than the marginal decrease in the
profit of the export firm from an increase in the tariff, 6H~*/6flx >
[6EIIy*/~xl, cannot be an internal equilibrium. Similarly, if 6Hx*/6fl~ <
[6EHy*/613x* 1, the liberal trade party will have an incentive to change its
proposed tariff at the margin and, hence, this also cannot be an internal equilib-
rium. Only if 6II~*/6~x = 1 6 E I I y * / 6 ~ l, then neither party has an incentive to
deviate from the common tariff.
Solving for the equilibrium tariff gives

~x = 2 M T ~ A / [ 4 N B - MTx2A] (A.2)

which, to satisfy the condition of being an internal tariff, must be between zero
and one. That is, for /~x to be an internal tariff, it must be the case that
4 NB > 2 MA T~ + MA T~ and 4 NB > MA y~. []

Proof of proposition 3. As before, the protectionist party and the liberal trade
party will announce their proposed tariffs /3xl and /3gO, respectively, to maximize
(minimize) S*, where

n ; ( /V,) - n ; ( anx*
an," (A.3)
Suppose that the parties announce different tariffs (/3x*l > flxo). Is it in the
interests of the two parties to increase or decrease their announced tariffs?
If both parties announce different, but greater than zero, tariff policies (/3x~ >
/3xo > 0), the profit of the foreign exporting firm will be zero in either case
because of the prohibitive retaliatory tariffs (i.e. lIy*(~x"o) = IIy*(~*~)=0).
Hence, the liberal trade party will not elicit any campaign contributions. But
because the protectionist party's proposed tariff is positive and greater than the
liberal trade party's tariff (/3~ >/3~'0), the protectionist party will attract all the
campaign contributions from the foreign import-competing firm. Consequently, for
any policies announced by the parties which are both greater than zero (/3~"l ~/3~"0
> 0), the protectionist party will always win the election.
The only way in which the liberal trade party can elicit campaign contributions
is to set its proposed tariff equal to zero (/3~*0 = 0), which yield positive profit for
the export firm. This generates the result that if the protectionist party proposes a
positive tariff, the liberal trade party will only propose a zero tariff and opt for
complete free trade (i.e. /3x*0 = 0 for any /3xl).
However, is it the case that protectionist party will propose a trade war if the
liberal party opts for free trade? If the liberal trade party sets its proposed tariff
136 D.M. Gould, G.L Woodbridge /European Journal of Political Economy 14 (1998) 115-137

equal to zero (/3x~ = 0), then the odds of the protectionist party winning the
election are
- n ; (0)
S* (0, /3x] ) = /-/y* (0) (A.4)

By maximizing S* (and, hence, its chance of election) the protectionist trade party
will opt for a trade war and set its tariff to the prohibitive one (/3x] = 1).
Consequently, if the liberal trade party opts for free trade, the protectionist party
will propose a trade war.
If the protectionist party proposes a trade war and the liberal trade party opts
for free trade, then odds that the protectionist party will win the election are:
//~* (1) - n Z (0) MA'yx(1 + ( T / / 4 ) )
S*(0, 1) = -- (A.5)
/-/," (0) --//y* (1) NB
A political equilibrium will only be generated if neither party has an incentive to
change its policy. The equilibrium, however, will only occur if both parties
propose the same policy and each has a 50% chance of winning the election. There
are only two possible outcomes; the liberal trade party changes its proposed policy
and opts for a trade war, or the protectionist party proposes free trade.
If S* (0, 1) > 1, the liberal trade party has less than 50% chance of electoral
success. It will alter its policy to match that of the protectionist party and opt for a
trade war (i.e. /3x~ = 1). On the other hand, if S * (0, 1) < 1, the protectionist party
will increase its chance in the election by changing its original policy and propose
free trade (i.e. /3x* = 0). The two possible equilibrium tariffs are 30
/3xl =/3x"0 = 1 if MA%,(1 + ( y J 4 ) ) > NB (Trade war).
/3x] =/3x~ = 0 if MA'yx(1 + ( y x / a ) ) <NB (Liberalization).
[]

References
Baldwin, R.E., 1989. Trade Policy in a Changing World Economy. University of Chicago Press,
Chicago.
Brock, W.A., Magee, S.P., 1978. The economics of special interest politics: The case of the tariff.
American Economic Review Papers and Proc. 68, 246-250.
Cassing, J.H., McKeown, T.J., Ochs, J., 1986. The political economy of the tariff cycle. American
Political Science Review 80, 843-862.
Dixit, A., 1987. How should the United States respond to other countries' trade policies? In: Stem,
R.M. (Ed.), U.S. Trade Policies in a Changing World Economy. MIT Press, Cambridge, MA.

30 As the probability that S * (0, 1) ~ 1 is essentially zero, that outcome is ignored.

 D.M. Gould, G.L Woodbridge/ European Journal of Political Economy 14 (1998) 115-137 137

Finger, J.M., 1991. The GATT as an international discipline over trade restrictions: A public choice
approach. In: Vaubel, R., Willett, T.D. (Eds.), The Political Economy of International Organiza-
tions. Westview Press, Boulder, pp. 125-141.
Grossman, G., Helpman, E., 1995. Trade wars and trade talks. Journal of Political Economy 103,
675-708.
Harris, R., 1985. Why voluntary export restraints are voluntary. Canadian Journal of Economics 18,
799-809.
Hillman, A.L., 1982. Declining industries and political-support protectionist motives. American
Economic Review 72, 1180-1187.
Hillman, A.L., 1989. The Political Economy of Protection. Harwood Academic Publishers, Chur.
Hillman, A.L., Moser, P., 1996. Trade liberalization as politically optimal exchange of market access.
In: Canzoneri, M., Ethier, W., Grilli, V. (Eds.), The New Transatlantic Economy. Cambridge
University Press, Cambridge, pp. 295-312.
Hillman, A.L., Ursprung, H., 1988. Domestic politics, foreign interests and international trade policy.
American Economic Review 78, 729-745.
HiUman, A.L., Ursprung, H., 1994. Domestic politics, foreign interests and international trade policy:
Reply. American Economic Review 84, 1476-1478.
Hofer, T., Woodruff, C., 1994. Domestic politics, foreign interests and international trade policy:
Reply. American Economic Review 84, 1474-1475.
Keohane, R.O., 1984. After Hegemony: Cooperation and Discord in the World Political Economy.
Princeton University Press, Princeton.
Lipson, C., 1982. The transformation of trade: The sources and effects of regime change. International
Organization 36, 417-455.
Magee, S.P., Brock, W.A., Young, L., 1989. Black Hole Tariffs and Endogenous Policy Theory.
Cambridge University Press, Cambridge.
Marvel, H.P., Ray, E.J., 1983. The Kennedy round: Evidence of the regulation of international trade in
the United States. American Economic Review 75, 190-197.
Mayer, W., 1984. Endogenous tariff formation. American Economic Review 74, 955-967.
Mayer, W., Riezman, R., 1987. Endogenous choice of policy instruments. Journal of International
Economics 23, 377-381.
Olson, M., 1965. The Logic of Collective Action: Public Goods and the Theory of Groups. Harvard
University Press, Cambridge, MA.
Peltzman, S., 1976. Toward a more general theory of regulation. Journal of Law and Economics 19,
211-240.
Quibria, M.A., 1989. Neoclassical political economy: An application to trade policies. Journal of
Economic Surveys 3, 107-137.
Ray, E.J., 1987. The impact of special interests on preferential tariff concessions in the United States.
Review of Economics and Statistics 66, 452-458.
Richardson, J.D., 1990. The political economy of strategic trade policy. International Organization 44,
107-135.
Riezman, R., 1982. Tariff retaliation from a strategic viewpoint. Southern Economic Journal 48,
583-593.
Stigler, G., 1971. The theory of economic regulation. Bell Journal of Economics 2, 2-21.
Young, L., Magee, S.P., 1986. Endogenous protection, factor returns and resource allocation. Review
of Economic Studies LIII, 407-419.