Citation
Three essays on international trade and welfare

Material Information

Title:
Three essays on international trade and welfare
Creator:
Okamura, Makoto, 1956-
Publication Date:
Language:
English
Physical Description:
vii, 63 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
International trade ( jstor )
Labor unionization ( jstor )
Labor unions ( jstor )
Lobbying ( jstor )
Quid pro quo ( jstor )
Relative wages ( jstor )
Skilled labor ( jstor )
Tariffs ( jstor )
Unskilled labor ( jstor )
Wages ( jstor )
Dissertations, Academic -- Economics -- UF ( lcsh )
Economics thesis, Ph. D ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1997.
Bibliography:
Includes bibliographical references (leaves 59-62).
Additional Physical Form:
Also available online.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Makoto Okamura.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
0027601702 ( ALEPH )
37135718 ( OCLC )

Downloads

This item has the following downloads:


Full Text















THREE ESSAYS ON INTERNATIONAL TRADE AND WELFARE


















By


MAKOTO OKANURA


















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA 1997



































Copyright 1997

by

MAKOTO OKAMURA

































I dedicate this dissertation to memory of the late Hiroshi Okamura, my father and Misuzu Okamura, my mother.

























ACKNOWLEDGMENT S




I would like to express deep appreciation and gratitude to Dr. Elias Dinopoulos, chairman of my supervisory committee. Without his invaluable suggestions and encouragement, i could not have completed this dissertation. I also gratefully acknowledge insightful comments by Dr. Richard Romano, Dr. Bin Xu and Dr. James Seale. 1 really appreciate valuable assistance from my friends, Mr. Shinji Yane and Mr. Sang H. Lee.

































iv


















TABLE OF CONTENTS

page

ACKNOWLEDGMENTS........................................................... iv

ABSTRACT.................................................................. v

CHAPTERS

1 INTRODUCTION.......................................................... 1

2 ENDOGENOUS TARIFF FORMATION AND DIRECT FOREIGN INVESTMENT:
GENERALEQUILIBRIUM ANALYSIS 4

2.1 Introduction 4
2.2 The Model 6
2.3 The Equilibrium 9
2.4 The Basic Comparative Statics 12
2.5 Optimal Direct Foreign Investment 15
2.6 Quid Pro Quo Direct Foreign investment 17
2.7 Concluding Remarks 20

3 LEARNING-BY-DOING AND PATTERN OF INTRAINDEJSTRY TRADE 21

3.1 Introduction 21
3.2 The Model 23
3.3 Trade at the initial equilibrium 25
3.4 The Dynamic Behavior of the Economy 26
3.5 The Effects of Government Policies 30
3.6 Concluding Remarks 34

4 TRADE LIBERALIZATION IN AN OPEN ECONOMY WITH A UNIONIZED SECTOR 35

4.1 Introduction 35
4.2 The Model 37
4.3 Uniqueness of Equilibrium 40
4.4 The Effects of Bargaining Parameters 47
4.5 The Effects of Trade Liberalization and Immigration 50
4.6 Concluding Remarks 55

5 CONCLUSIONS 57

LIST OF REFERENCES 59

BIOGRAPHICAL SKETCH 63









v

















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




THREEE ESSAYS ON INTERNATIONAL TRADE AND WELFARE



By



MAKOTO OKAMUPA



May, 1997



Chairman: Elias Dinopoulos
Major Department: Economics


I formulate a two-period, two-sector, specific factor, general equilibrium trade model with endogenous tariff formation. The tariff rate

depends on labor devoted to lobbying in each of the two industries. We examine the possibility of quid pro quo direct foreign investment (DFI). It is shown that more DPI in the exporting sector reduces the level of protection and quid pro quo DFI occurs. Quid pro quo DPI takes place if DPI in the importing sector increases directly the ability of the hostcountry government to resist protection.

I construct a two-country Ricardian dynamic general

equilibrium model with learning-by-doing and transportation cost. The two countries differ only in the distribution of the consumer' s preference. At the initial time, no trade occurs due to transportation costs. There exists a unique time when trade begins. After that time, each country exports the goods for which it has a stronger preference



Vi










than does the other country. A government can completely reverse the pattern of trade by a tax-transfer policy to change the distribution of income. A tariff delays the time when trade begins, while a quota does not affect the timing of trade.

Finally, I formulate a open economy model in a two-by-two

general equilibrium setting. Unskilled workers establish a labor union and bargain with a firm about their employment level and wage, while skilled workers are employed in a competitive market. Both parties are engaged in Nash bargaining. The labor union is assumed to be employment oriented. I establish that trade liberalization decreases the wage on unskilled workers in the unionized sector if the unionized sector is labor unskilled labor intensive. I also show that immigration of unskilled labor decreases the wage of unskilled unionized labor if the unionized sector is unskilled labor intensive.





































Vii













CHAPTER 1
INTRODUCTION





According to traditional theories of international trade,

typically, in the original Hechsher-Ohlin-Samuelson general equilibrium framework, -three assumptions are imposed: First, capital movement from one country to another country is seen essentially as s substitute to a commodity trade. Direct Foreign Investment(DFI) is induced flow from the capital abundant country with low reward to the capital scare country with high reward to seek higher return. Second, technologies of trading countries are assumed to be given. That is, no technical changes occur at these countries. Third, in both product markets and factor markets, perfect competition prevail. In this dissertation, I construct three models to examine and extend these three assumptions.

Chapter 2 of the dissertation analyzes political-economic aspect of DFI. There have been two distinct approaches, which attempts to incorporate various political-economy considerations. The first strand of literature has focused on product markets by modeling resource-using lobbying or by modeling voting behavior. The level of protection is determined through the self-interest maximizing behavior of economic agents. The second strand of literature is associated with "Quid Pro Quo" Direct Foreign Investment (DFI). This strand examines how DFI can be used to defuse the threat of protection (or the actual level of protection) based on political-economy considerations DFI aims at diffusing future protection even if it is associated with current losses. I construct a





I






2




two-period, two-sector, specific factor model with DFT and endogenous lobbying to integrate these two strands. in this model, I examine in which sector, the exporting or importing sector is engaged in quid pro quo DFI to defuse future protection of host country.

Chapter 3 builds a simple dynamic Ricardian general equilibrium model with learning-by-doing effects of technologies. The learning-bydoing effect means that each firm can learn from its experience and improve its technology. Two types of consumers in each of the two countries are considered. Each type has preferences over a particular set of similar products. The distribution of the two types of consumers differs across the two countries. Current productivity of labor, which the only factor of production, increases over time because of learningby-doing considerations. The dynamic nature of the model allows us to determine the duration of autarky based on transportation costs. Transportation costs provide a realistic and natural way to generate an initial period without trade, which is necessary to allow the interaction between national differential tastes and learning-by-doing. I investigates the effects of domestic income transfers across different types of consumers, tariffs and import quotas on the pattern and volume of trade.

Chapter 4 of the dissertation constructs a general equilibrium model of an open economy with unionized sector. The export producing sector is perfectly competitive and utilizes two factors of production, skilled and unskilled labor. The import-competing sector consists of a unionized firm that is protected from foreign competition through an import quota. The domestic monopolist utilizes both skilled and unskilled labor under a constant returns to scale technology. However, unlike the






3



labor demand for skilled workers which is assumed to be competitive, I assume that the unskilled workers in tlie import-competing sector have formed a labor union. The union bargains with the domestic monopolist over the negotiated wage and the employment of unskilled labor. The bargaining process is modeled as an efficient Nash bargaining game that results in simultaneous determination of employment and the wage. I analyze the effects of trade liberalization ( i.e. an increase in the quota ) and immigration on income distribution and domestic prices. I identify conditions on the parameters of the model that establish the robustness of the Stolper- Samuelson mechanism in the presence of sectorspecific labor union. I also characterize conditions for the existence of a unique equilibrium.
















CHAPTER 2
ENDOGENOUS TARIFF FORMATION AND DIRECT FOREIGN INVESTMENT: A GENERAL EQUILIBRIUM ANALYSIS



2.1. Introduction




There have been two distinct approaches to the theory of endogenous protection, which attempts to incorporate various politicaleconomy considerations in the formation of tariffs, Quotas and other instruments of protection. The first strand of literature has focused on

product markets by modeling resource-using lobbying (i. e., Findlay and Wellisz (1982)) or by modeling voting behavior (i. e., Mayer (1984)). In

both cases, the level of protection is determined through the selfinterest maximizing behavior of economic agents. The second strand of literature is associated with "Quid Pro Quo" Direct Foreign Investment (DFI) This strand examines how DFI can be used to defuse the threat of protection (or the actual level of protection) based on political-economy considerations. Bhagwati (1985, 1986, 1987) has identified the new type of DFI which aims at diffusing future protection even if it is associated

with current losses. Bhagwati et al. (1987), Dinopoulos (1989, 1992), Wong (1989) Dinopoulos and Wong (1991), and Zhao (1991) have developed models of Quid Pro Quo OFI.

These two distinct approaches to endogenous protection aim at understanding the instruments and channels which determine the level of protection, but have ignored each other's implications. For example, although Findlay and Wellisz (1982) have emphasized the resources which are utilized in the lobbying process within the context of general equilibrium, Bhagwati et al. (1987) do not model the resource cost of



4






5


lobbying. In addition, by focusing on product markets, the first strand the literature has abstracted from factor mobility influences on the

of protection. Moreover, since the Quid Pro Quo DFI literature examines the potential for defusing (or reducing ) existing protection, r has opposite implications to the resource using lobbying literature which studies the conditions under which protection emerges. Therefore,

there is some scope for integrating the two approaches in order to analyze how endogenous protection based on resource using lobbying is affected by DFI.

Quid Pro Quo DFI corresponds to a novel form of foreign lobbying which utilizes resources in the present (by incurring a welfare loss for the foreign country) in order to reduce protection in the host country in the future (with the associated benefits). Therefore, it is possible to build a model of endogenous protection with domestic lobbying which uses resources and foreign lobbying which causes DFI in a quid pro quo fashion. I construct a two-period, two-country, specific-factor general equilibrium trade model with direct foreign investment.

Following Findlay and Wellisz (1982), a tariff formation function is added in the second period. Free trade prevails in the first period. The two-period, two-country framework is adopted from Bhagwati et al (1987).

By adopting the two-sector specific-factor model, I can examine in which sector, the exporting or the importing, does quid pro quo DFI take place. This paper shows that the level of DFI is substitutive (complementary) to the lobbying activity and protection when DFI occurs in the exporting (importing) sector. This reduces the level of

protection, which encourages the foreign government to engage quid pro quo DFI. On the other hand, the difference between lobbying activities expands as the result of DFI in the importing sector, which leads to a higher tariff rate. However, if DFI increases the host-country government






6


resistance to lobbying and this effect overcomes the effect of lobbying activity, the tariff rate is reduced by DFI and there is scope for quid pro quo DFI. Another channel by which DFI influences protection is also

shown: the amount of labor devoted to productive activities may increase as the result of DFI. The foreign country exports more and obtains a higher level of welfare. The foreign government chooses DFI taking this
"productive labor expansion effect" into account. This effect emerges in the general equilibrium framework of the model in which the tariff is determined through resource-using lobbying activities.

Section 2.2 formulates the model. Section 2.3 characterizes the equilibrium. Section 2.4 develops basic comparative statics. Section 2.5

calculates the optimal DFI levels. Section 2.6 analyzes the possibility of quid pro quo DFI and section 2.7 offers the concluding remarks of the chapter.



2.2. The Model





I formulate a two-country, specify ic-f actor, general-equilibrium model with endogenous tariff formation using building blocks from Findlay and Wellisz (1982). There are two countries, foreign and domestic, in the world. A manufactured good and an agricultural good are produced under perfect competition and traded in both countries.

First I describe the host country. The manufactured (agricultural) good is produced by a specific factor, capital (land), and a mobile factor, labor. I denote the manufactured (agricultural) good by X (Y) The production functions are



X = F (K, Lx) (1)

Y = G (T, LY) (2)






7


K = Kd + KT = Td + T.(3)



The host country goverment imposes a tariff (subsidy) on each industry. The domestic price of each good is



q= (1 + t ) p, (4)

qy =(1 +~ s )Py (5)



where q,, (qy) is the domestic price of X (Y) and p~. (py) is the world price of X (Y) The tariff (subsidy) rate is determined endogenously through lobbying by each industry. Each industry hires lobbyists who put political pressure to the government to obtain protection. The lobbying

activity utilizes only the following endogenous protection function which determines the size of the tariff as a function of lobbying



t =t ( L, La) s= s(L-Lm ), (6)



where Lm (La,) is a lobbying input by manufactured (agricultural) sector. We assume that the shape of the tariff (subsidy) function, is as follows:



t(o) = 0, t, > 0, t" < 0,

s(o) = 0, s' > 0, s" < 0. (7)



This implies that for a given rival's lobbying behavior, the marginal product of the lobbying by each industry is positive but decreasing. These equations mean that when the X industry faces an import tariff, the Y industry faces an export tax and vice versa. A representative consumer has a well-behaved utility function tJ(X, Y) The consumer demand for each good can be written, Xd(q,,qy,I) and yd(q,,qy,I), where I denotes income.
I turn to the foreign country. The transformation curve does
not change over periods,






8


Q*y= H (QX, K' Kff,T_ Tt) (8)



whereQ~y(Q*,) is the output of the agricultural (manufactured) good and K*

(T*) is the existing stock of capital (land) The foreign country exports (imports) the manufactured (agricultural) good. it faces the reciprocal

demand function (offer curve)



M Y (D(*, d+ K Td + Tf, L LnC), (9)

where

M*Y foreign import of good Y, E foreign export of good X,

L, the amount of the labor devoted to the lobbying,

t relative tariff rate, (1 + t)/(l + s)

The preferences of the consumers are summarized as the social utility function, W(C*,x,Cy) which depends on the consumption of both goods.

Since perfect competition prevails in every market, each individual

firm has no market power. The foreign firm cannot affect the tariff

level through the DFI. To examine the quid pro quo DFI, we assume that the foreign government controls the total amount of the DFI. The

government determines the level of DFI to maximize its national welfare. Thus, there are three players in the game we consider: the foreign government, the host-country representative manufacturing firm and the host-country representative agricultural firm.

There are two periods in the model. The timing of the game is as

follows:

Period 1. 1. The foreign government chooses DFI which remain fixed.

2. There exists no lobbying and free trade prevails.

Period 2. 3. Both domestic industries lobby and the endogenous

tariff rate is determined.






9


4. Given this tariff rate, trade occurs.

The foreign government moves as the Stackelberg leader, while the two host-country firms move as simultaneous Stackelberg followers. The structure of the model can generate quid pro quo DPI, and it is identical

to the one used by Bhagwati (1987) The foreign government does not necessarily choose DPI to maximize its national welfare under free trade

(the first period welfare) It could endure a first-period welfare loss for the purpose of obtaining the larger second period welfare.








2.3. The Equilibrium





This section analyzes the equilibrium of the model. I solve for the equilibrium in the usual backward fashion. In the second period, given a level of DPI, and the domestic prices of goods and labor, each domestic firm chooses its labor inputs for production and lobbying simultaneously in order to maximize its myopic (one-period) profit. The profit of the manufacturing firm is



maxH =[I +t(L,, L,)]P, F(K, L )-wL, wLm. (10)




where w is the wage rate. I impose the Cournot assumption with respect to

the lobbying behavior, that is, each firm decides its lobbying activity, taking the rival's lobbying activity as given. The first order conditions for the above problem are



(1 + t) PIFL W= 0,1 (11)






10


t'pxF w = 0. (12)



where FL = aF/aL,.

The second order condition is assumed to be satisfied.




M = (I+t)FLLFt" (t'F,)2 > 0, (13)



where FLL is the second order derivative. The economic problem of the agricultural sector is




maxl- = [+ S(L, Lm)] Py G(, Ly) wLy wL. (14)




I obtain the first-order conditions



(l+t)pyGL w = 0. (15)

s'pG w = 0. (16)



The second order condition is



A = (I+s)GLLs"G (s'GL)2 > 0. (17)



The equilibrium in the labor market is given by the full employment condition.



Lx + L + + La (18)



where L stands for the fixed endowment of labor. Equations (11), (12),

(15), (16), and (18) describe the equilibrium of the production sector. These five equations determine the five endogenous variables L, Ly Lm La,






11


and w. We assume that this system of equations has a unique solution for a given K, T, P,, Py, and L. The production level of each good, the endogenous tariff rate, and the domestic prices are determined based on these variables. Finally, by substituting the values of the endogenous variables into the demand equation, I can calculate the demands for both goods, imports and exports.

Given the equilibrium of the host country and the level of DFI, the foreign government chooses outputs of goods and the volume of trade in order to maximize its national welfare, W(C*x,C*y). It solves the

following problem:



MaX W2 (C2, *2
*2 -2 -2
S.,t. C. =Q,, E (19)

C;*2 *;2 TJ2,Kd+ Kf, Td T~-~
C =Q 2 + (D(E;2 K d+K/ + Tf L L, )




where the superscript denotes period 2. The maximum possible welfare level as a function of direct foreign investment in the first period is denoted by W2(Kf,Tf).

Consider now the first period. Since there is no lobbying, the equilibrium is described by the specific-factor model with direct foreign investment. Given the level of DFI, the equilibrium values of L,, Ly and w in the home country are defined by



p.Fj(K,L.) w = 0, (20)

PYG1(T,LY) w = 0, (21)

L, + Ly = L. (22)



The foreign country's reciprocal demand function in the case of free trade is simplified as






12


M*y= D(E,, Kd + Kf Td +T L 1)



= W ( Ex, Kd + Kf T, + T, ). (23)



The foreign government has a intertemporal utility function



V = W (CIx, C*'y) + pW2 (C*2x, C*2y) (24)



where p is the discount factor. The optimization problem is



Max V

s.t. C x x Ex,

C'ly = Q*y + ( E*x, Kd + Kf Td + T ). (25)



The foreign government chooses the level of DFI. Given this level, all remaining variables are determined.





2.4. The Basic Comparative Statics



I start with the comparative statics of the second period production equilibrium with respect to DFI and mobility of labor.






13


Proposition 1



(A) The effects of DFI on the manufacturing (importing) sector.



(1-1) The labor input of the manufacturing industry for production and the total labor input of the manufacturing industry increase, if the following condition is satisfied.

Z = t'FK (l+t) F,, 0.


( 1-2) The labor input of the agricultural industry for production decreases.

(1-3) The wage rate goes up.

(1-4) The output of the manufacture (agriculture) industry increases (decreases).

(1-5) The difference of labor input for lobbying activity between two industries expands.



(B) The effects of DFI on the agricultural (exporting) sector.



(1-6)The labor input of the agricultural sector for production and the total labor input of the agricultural industry increases, if the following condition is satisfied. N = sIGL (l+s)GLT 5 0. (1-7) The labor input of the manufacturing indusstry for production decreases.

(1-8) The output of the agricultural (manufacturing)industry increase

(decreases).

(1-9) The wage rate goes up. (1-10) The difference of labor input for lobbying activity between two industry shrinks.






14


(C) The effects of immigration of labor.



(1-11) The labor input for production of each industry, the rental rates for capital and land, the wage rate, the difference of labor input for lobbying activity between two industries does not change. (1-12) The labor input for lobbying of each industry increases equally.



The condition in (1-1) means that the effect of DPI in the manufacturing sector on the marginal value product of the productive labor ( (l + t) p, Flk ) exceeds that on the marginal value product of the lobbying activity ( t' F ) In other words, DFI favors the productive labor more than the lobbying labor in the manufacture industry. It is

natural that this condition is sufficient for the increase in L,. This argument applies to the condition in (1-6). The result (1-1) implies that the total labor input of agriculture decreases, since the total labor supply is fixed. The comparative statics' properties in the usual specific factor model does not change in this framework. See Jones(1971).

The result (1-10) implies that the tariff rate on the importing sector decreases as the result of DFI in the exporting sector. The reduction in the tariff rate in the second period encourages DIFI in the first period, that is, quid pro quo DPI. The result (1-5) means that the endogenous tariff rate on the importing industry and the tax rate on the exporting industry increases when DPI occurs in the importing sector increases. The DPI expands the distortion of the relative price (1+t)/(l+s)p,,/py. Since the importing industry is more protected as a result of the DPI and foreign export becomes difficult, I need another factor in order to raise

the quid pro quo DPI. Both (1-11) and (1-11) imply that the immigration of labor only changes the labor input for lobbying activity. Since both the tariff and tax rates do not change, every variable affecting national

welfare, such as production, export, import, and prices, remain constant






15


after the change in the immigration of labor. An increment of labor is absorbed in the rent-seeking lobbying sector. Although the lobbying sector (L, + La) expands, its impacts on t and s are identical.








2.5. Optimal Direct Foreign Investment





I characterize the optimal DFI for the foreign government. In the second period, the government solves the problem (19) The

necessary conditions for a maximum are



W2x/W2y = -H 2X, (26)

w2X/w2y = 2(27)



The optimal policy requires that the consumer price (W2x/W2v) the producer price (-H2X), and the trade price ( 2(D) be the same. The impacts of DFI on the second period's utility are




dW W2 -H2 + )2 2 LT 2f(2
dKf I- a Kf 8a




dW_- U12 2 a2 + D2 a(29
dTf a Tf jf



The optimal conditions for the whole problem are:



WIX/Wly = -Hlx, (30)






16


W /W1y= (31)



dV/dk, = dW'/dKf + pdW2/dKf



= dW1y(-Hk+Yk) + pW2y (-H2k k (D218Ln/aKf + 2&T/aKf)



= 0. (32)



dV/dT, = dW/dTf + pdW2/dT,



= dWy (-Ht++t ) + pW2(H2+ (2 21aLn/aT + 2/T)



= 0. (33)

Two additional terms emerge in the second period effect because of the

reallocation of labor ( -021aLn/aKf -0218Ln/aTf ) and the change in the tariff rate.

To examine conditions (32) and (33), we assume that the home

country has a utility function U(Cx, Cy) = U0(Cy) + Cx. This quasi-linear utility function implies that the host country's demand for good Y is independent of income. With c(D9) denoting the output of good Y the consumption of good Y in the second (first) period and the results of the comparative statics, I obtain



Tk < 0, T > 0, Ok < 0, OT > 0, DL > 0, OT > 0. (34)



Before examining a quid pro quo DFI, I have another route which induces DFI from the foreign government although it has no incentive in the myopic context. This is the effect through a change in the labor






17


reallocation between the production and the lobbying activity. If DFI increases the amount of labor devoted to the production aL/aKf < 0, aL,/Tf < 0 ) DFI increases the exports of the foreign country through this productive labor expansion (- cDL8Ln/8Kf > 0,

- q2aLf/OTf > 0 ) and increases the second period utility. This effect differs from the original quid pro quo effect the first period DFI reduces the second period protection. But the economic implication of this "productive labor expansion effect" is the same as the original one. We state:



Proposition 2

If the "productive labor expansion" effect (D- c21 Ln/aKE,

S16L /aTf ) is sufficiently large, the foreign government has a incentive for DFI.



2.6. Quid Pro Quo Direct Foreign Investment





This section examines the possibility of quid pro quo DFI in the importing and exporting sector. Consider DFI in the exporting sector. Since the DFI reduces the tariff rate in the importing sector, it promotes the export of the home country ('DtdT/dTf > 0). This makes a gain in the second period [See (32)]. Even though the DFI causes a loss in the first period, the home government makes DFI to obtain this gain.



( Proposition 3 )

The foreign government has an incentive to engage in quid pro quo DFI in the exporting sector.






18


The above proposition is novel. Bhagwati et al. (1987) analyze the HOS model. They assume a single capital and can not examine in which sector a quid pro quo DFI appears. Dinopoulos(1989), (1992), and Zhao (1990) examine quid pro quo DFI in the importing sector in the partial equilibrium framework.

The impact of DFI in the importing sector on the first period's utility dW1/dKf is negative. When the foreign government is myopic, it has no incentive to invest in the host country. Since more DFI causes more protection (dt/dKf>0) and higher tariff rates reduce the export of the foreign country (TD < 0), I conclude that the impact of DFI in the importing sector through the lobbying activity is negative. Like Dinopoulos (1992) and Zhao (1991), no quid pro quo DFI occurs.

I seek another political process which causes quid pro quo DFI in the original sense. Bhagwati (1985) states:

"Such "image building" can influence Congress to withstand the protectionist pressures from the import-competing industry." Following the above statement, I introduce this effect into the tariff function in the importing sector. The function is now reformulated as



t = t(Lm La, Kf ) (35)



The tariff/tax rate depends not only on the difference of the labor input for lobbying between industries but also on the level of DFI, itself. I

assume



tI > 0, t1l < 0, t2 < 0, t22 > 0, t12 < 0, (36)



where tl =o am'L), t2 =o/aKf and so on.






19


The higher the level of DFI, the lower the tariff rate, but its marginal effect decreases as DFI increases. The marginal effect of each factor decreases as its rival factor increases.

The effect of DFI on the tariff rate is



dt/dKf = tld(Lm-La)/dKf + t2. (37)



If DFI decreases Lm La, the tariff rate decreases as the result of DFI. Simple manipulation leads to



d 1
dK (Lm La) = {(1 +s)pGLL +s' pyGL} x[-(1 +t)pxFLL {tl2pxF+tIlPxFk} dKf detDY
+tIPXFL(t2PxFL +(I+t)pxFLK}J (38)



where D' is the coefficient matrix of the reformulated system and det D' < 0. If both t12pxF + t1pFK and t2pxFL+(l+t)pxFLK are negative, I get d(Lm La)/ dKf < 0. The former inequality means that the impact of DFI on the marginal value of Lm, tlpxF, through the political process (t12pxF) exceeds that through the direct productive process (tlpxFk), while the latter inequality implies that the same property holds on the impact with respect to the marginal value of Lx. With these inequalities, the

endogenous tariff rate decreases as DFI increases. DFI improves the

second period welfare by reducing the tariff rate (Gdt/dKf > 0). The

foreign government uses DFI the future benefit even though it generates a first period welfare loss.






20


I summarize the argument as follows:



(Proposition 4)

Suppose thnat the tariff function is t =t(Lm La, Kf) satisfying (39). The foreign government obtains an incentive to engage in quid pro quo DPI in the importing sector if the following inequalities hold



(a) t12p,(F + tPxFk < 0.

(b) t2PXj + (l+t) pXFlk < 0.








2.7. Concluding Remarks



The present chapter has formulated a specific-factor, two-period, two-country general equilibrium model with an endogenous tariff determined by lobbying activity of each industry. The lobbying activity

uses a productive resource, labor. The wage rate is determined by the

labor market equilibrium, and is not assumed to be fixed. I have examined

three problems: (1) the impacts of DFI in the exporting and importing sector on the second period equilibrium, (2) which DFI reduces the

protection encourages quid pro quo DPI, (3) the possibility of new type of quid pro quo DFI.

















CHAPTER 3
LEARNING-BY-DOING AND THE PATTERN OF INTRAINDUSTRY TRADE




3.1. Introduction


Traditional static models of intraindustry trade based on

monopolistic competition and static scale economies cannot determine the pattern of intraindustry trade (Krugman(1979), Lancaster(1979), Helpman and Krugman(1985). Bhagwati(1982) has proposed a novel explanation of intraindustry trade patterns under the label of "biological" model of trade in similar products. Countries with identical technological capabilities (i.e. Japan and the U.S.) could develop different products in autarky because of differential tastes. Learning-by-doing fixes the comparative advantage of each country within each industry, and when trade starts these countries exchange similar products developed in autarky. The term "biological" captures the feature that technology plays, which is analogous to "genotype" in biology, whereas each product develop in autarky is like the "phenotype" in biology, with differential national tastes corresponding to the environment which shapes the type of similar products a country cab produce.

Feenstra(1982) and Dinopoulos(1988) have developed models of

intraindustry trade based on differential national tastes. However, these models do not examine the role of learning-by-doing in determining the initial comparative advantage of each country. The present paper complements the above mentioned studies by developing a simple dynamic general equilibrium model of intraindustry trade based on learning-by21






22


doing. The model with building blocks for the context Krugman(l987) who examined the role of learning-by-doing in industrial targeting.

In the present model, two types of consumers in each of the two countries are considered. Each type has preferences over a particular set of similar products. The distribution of the two types of consumers differs across the two countries. Current productivity of labor, which is the only factor of production, increases over time because of learning-by-doing considerations.

The dynamic nature of the model allows us to determine the duration of autarky based on transport costs. In the absence of transport costs (or any barriers to trade), the location of production is indeterminate. Transportation costs provide a realistic and natural way to generate an initial period without trade, which is necessary to allow the interaction between national differential tastes and learningby-doing. This research investigates the effects of domestic income transfers across different types of consumers, tariffs and import quotas on the pattern and volume of trade.

The rest of the chapter is organized as follows: Section 3.2 develops the model. Section 3.3 analyzes the autarkic equilibrium. Section 3.4 examines how the pattern of intraindustry trade evolves over time. Section 3.5 investigates the impact of income transfers, tariffs and import quotas on the pattern and volume of trade. Section

3.6 concludes.






23


3.2. The Model






There are two types of consumers in the home country and the foreign country. Each type (type 0 and type 1) has the CES utility function:





type 0. XpX) Xi'0

i=M

type 1. UYY' E i)
j=1



where X,(y1) is consumption of the i-th product by type 0 (1) consumer and 0 < p < 1. Type 0 individual consumes only differentiated product x, while type 1 consumers only differentiated product y. The number of the goods each consumer purchases, n and m, respectively, are fixed.

There is one unit of labor (consumer) in each country. We

consider the distribution of consumers as mirror images of each other. That is, the home country has s(1-s) fraction of the type 0(1) consumer, while the foreign country has l-s(s) fraction of the type 0(l) consumer. We assume that s > 0.5 which means that the home country has a stronger national demand for the type 0 goods than type 1 goods and the reverse pattern of preference prevails in the foreign country except for s=0.5. Both countries have the same tastes when s=0.5.

The only factor of production is labor. At the initial stage,

each country has the same technology exhibiting constant returns to the production of each good.






24


xi = k(0)li, yj = k(0)lj, (1)



where li(l)is labor developed to production of the xi(yj)and k(0)> 0, is the production coefficient at the initial stage.

I assume that there is learning-by-doing in each industry. At time t, the production technology of each industry is given by



xi(t) = ki(t)li(t), yj(t) = kg(t)lj(t), (2)



where ki(t)[ kj (t)] is the production coefficient of goods, xi[yj]. The coefficient ki is embodied in the learning-by-doing effect. Each firm can learn from its experience and improve its technology. The accumulated production of each firm represent its experience.

There is no spillover regarding learning effects across firms or countries. I define the learning-by-doing effect as



ki(t)=[ oxj(s)ds+ko-,

1 (3)
k (t)[ = Yi(s)ds+ k o ],




where 0 < s < 1. This learning effect is assumed to occur at the industry level and to work as external economy for each firm. Perfect competition is compatible with this phenomenon.






25


3.3. Trade at The Initial Eauilibrium



First, I derive the initial equilibrium of each country. Since each country has identical production technology and each good appears symmetrically, the initial equilibrium is given by



The home country:



p= w/k (0), xi = sk (0)/n yj = (1-s) k(0) /m l= s/n, li = (l-s)/m. (4)



where pi is the price of the i-th good, w is the wage rate, x1 ( yj ) is the output of the i (j) -th good belonging to type 0 (1) and li ( lj ) is the labor employed in the i(j)-th sector of the type 0(l) goods.



The foreign country:



=~ w*/k (0) xj = (1-s) k(0) /n y~j = sk (0) /m 1i (1-s) /n, 1%j = (1-s) /m. (5)



Reflecting the pattern of taste of each country, the home country produces more quantity of type 0 goods than the type 1 goods, while the foreign country produces more quantity of type 1 goods than the type 0 goods.

I proceed to examine the pattern of trade by assuming "iceberg" type transportation costs: A fraction of g (<1) of any good shipped to the other country actually arrives. A fraction of 1-g of any goods evaporates during transportation. Suppose that the home (foreign) country exports the i(j)-th goods, the following inequalities must hold.






26




pi/g < P'i P*j/g < pj. (6)



Combing these two inequalities and using (5), we have the inequality g2 > 1. This contradicts the assumption g < 1. There occurs no trade at the initial stage. This result is natural, since this economy is Ricardian and both countries have the same technology at the initial stage. Note that no trade occurs even if there are no transportation costs.







3.4. The Dynamic Behavior of The Economy



In this section, I investigate the dynamic pattern of trade. Since no-trade occurs at the initial stage, a no trade period lasts. During this period, I can divide all goods into type 0 and type 1, since the initial equilibrium is symmetric. Within each group, prices, outputs, and labor inputs should be identical. The equilibrium at time t is

Home country:



p0(t)ko(t) = w(t), p1(t)k1(t) = w(t), x(t) = sk0(t)/n y(t) = (l-s)k1(t)/m (7)



where the subscript 0(l) denotes type 0(l). Foreign country:



p%0(t)k*o(t) = w*(t), p*1(t)k*1(t) = w*(t) x*(t) = (l-s)k*o(t)/n y*(t) = sk*1(t)/m. (8)






27




Since the home country produces more output of type 0 goods than the foreign country, we have







As time moves on, the learning-by-doing effect changes the production coefficients in both countries. The initial difference in outputs between countries causes the change in the production coefficients, which results in the difference in outputs in the following period. The home country develops a comparative advantage in type 0 production, while the foreign country develops a comparative advantage in type 1 goods through this process.

I expect to be a switching point T, the time when dynamic

comparative advantage overcomes the barrier of the transportation cost and trade starts. The home country exports type 0 goods and the foreign country exports type 1 goods. If trade starts at time T, both countries can export thus, the following two inequalities must hold:



po (T) /g < p*%(T) p~j (T) /g < p, (T). (10)



Using (9), we rewrite (10) as



k, (T) /k0 (T) < g2k*l (T) /ko (T) .(1



Inequality (11) means that net of transportation costs, the home country has comparative advantage in type 0 goods, while the foreign country has a comparative advantage of type 1 goods. I define the function H-(t) to analyze the pattern and the timing of trade.






28







if H(t) takes negative value at some t, trade occurs. At t=O, H(O)=l-g > 0. Thus, trade does not occur at the initial equilibrium. H(t) is continuous in t. This implies that a no trade period prevails in some range of time near time 0.

Since the shape of H(t) is ambiguous, two cases are possible:

(1) there is a time t with H(t) = 0. (2)H(t) is positive for all t. Case

(2) means that no trade occurs forever. Therefore, we focus on the case

(1). There may be multiple time intervals such that H(t)=0. Consider the first time t=T with HMT=0 and assume the regularity condition dH(t)/dt <0 at t=T. Take the time t+h, where h is a sufficiently small positive number. From the regulatory condition, H(t+h) is negative, which means the home (foreign) country has comparative advantage in the type 0(l) goods. The home country exports type 0 goods, while the foreign country exports the type 1 goods. We assume that complete specialization occurs. The type 0 goods are produced only in the home country, while the type 1 goods are produced only in the foreign country. Once specialization prevails, the learning -by-doing effects appears in neither type 1 goods of the home country nor the type 0 goods of the foreign country. So the coefficient of these goods do not change. On the other hand, the coefficient of exported goods increases as time passes. H(t) remains negative after the time t+h. This shows that the time T is the switching point from the no trade period to the trading periods. This switching takes place just one time and there is no reversal of the trade pattern.



Proposition 1

There is a unique switching point, T. Before T there is no trade. After T, there is trade based on taste difference across countries.






29


The difference in tastes in each country generates a difference in technologies. In our Ricardian economy, technological differences determine the trade pattern. Each country exports goods with larger domestic demand. Consider the case of s=0.5. In this case, both countries have the identical tastes. At the initial stage, both countries produce the same amount of all goods. Since both countries have the same production coefficients, trade does not occurs at any time. This is the subtle situation, because even if s deviates from 0.5 by a very small amount, trade can occur eventually.

To describe the trading equilibrium, we normalize the home wage to

1



Poko = 1, (12)

p*lk*l = 1, (13)

1 = 3 + (1-3)W*. (14)



Equation (12) is the zero profit condition of the home country's firm and (13) shows the zero profit condition of the foreign firm. Equation (14) is the market clearing condition for type 0 goods.

Equations (12), (13) and (14) determine three variables p, p* and w*. Equation (14) implies that w = 1, i.e. the same wage rate prevails in both countries. This property comes from the assumption that the distribution of preferences across the two countries are the mirror images of each other.

I examine the effects of the transportation cost and the learningby-doing parameter on the switching point. Differentiating the equilibrium condition H(T) = 0 with respect to the transportation costs

and assuming the regularity condition H'(T) < 0, we obtain



dT/dg = 2gk*l (T) /H1 (T) k*0 (T) < 0. (15)






30


The inequality (15) means that the length of the no trade period increase as the transportation cost increases.







3.5. The Effects of Government Policies





In this section, I examine two types of government policy, income tax (transfer) and trade policy such as a tariff and a quota. Without loss of generality, we focus on of home-country policies.

First, I analyze the impact of income tax (transfer).

We consider the income tax and transfer policy at the initial stage. The government imposes wz income tax on type 0 consumers and gives this tax revenue to type 1 consumers. Since we assume that the labor supply is fixed, income tax is equivalent with the pure transfer program. Once this policy is introduced, the outputs of the both goods are



type 0 goods: (s-z)/n, type 1 goods: (1-s+z)/m. (16)



If the output of the type 1 goods produced at the home country exceeds that of the foreign country, the inequality



1 s + z > s (17)



must hold. This inequality implies that z>2s-1. This also means that the home country produces less the type 0 goods than the foreign country does.

At the initial stage, trade does not occur. I assume that the

government implements this tax-transfer policy for some time interval.






31


During the no trade period, the home country is developing comparative advantage in type I goods over the foreign country. After some time passes, trade occurs but the trade pattern is completely reversed. After trade starts, the government does not need to adopt this policy because the pattern of production is fixed. I have



Proposition 2

The government of the home country can completely reverse the

existing pattern of trade by taxing the type 0 consumer by wz (z>2s-1) and transferring this to the type 1 consumer during the no trade period.



This income transfer program is useful to reverse the pattern of trade as long as it is implemented during the no trade period. However, once trade occurs and the home country imports type 1 goods, this policy becomes useless, since the transfer the type 1 consumer receives is spent on imported type I goods. This transfer only contributes to the improvement in the technology of the foreign country.

Next, I investigate the impact of trade policy. Consider a tariff imposed on type 1 goods by the home country. This tariff is applied at the same rate, v, to all the imports. For simplicity, the government introduces the tariff at the initial stage. Th 43 tariff is irreverent during the no trade period. We assume that the tariff revenue is distributed equally to all consumers. Since this tariff revenue transfer is neutral to the income distribution among consumers, the patter of production is not affected. The manner in which the tariff revenue is distributed is irrelevant during the no trade period.

If trade takes place at time T, the following inequalities must hold.



po (T) /g < p*0 (T) (1+v) p*j (T) /g < p, (T) (18)






.32




Following the Previous argument, I define the function N(t) as



N(t) =(1+v)k,1(t)/k0(t) < g2k*,(t) /k% Ct) (19)



The time N(t)=0 gives the switching point from autarky to the trade Period. At the initial stage, N(t) = 1 + V _g2 > 0. We also assume the corresponding regularity condition



N'(T) < 0 if N(T) =0.



Note that during autarky, the outputs of all goods and all production coefficients are not affected by the tariff. The impact on the switching time is given by



dT=- 1 kiT>.(20)
dv N'(T) k,(T)



The introduction of a tariff delays the switching point.

Since autarky lasts longer in the tariff-ridden economy than in free trade, the volume of trade is larger in the tariff-ridden economy than in free trade economy at the switching point. After trade occurs, the pattern of trade becomes the same as that of the free trade. The tariff protects the type 1 industry in the sense that it delays the introduction of foreign competition.

I proceed to investigate a quota. The government sets the

uniform import quota of type 1 goods at the level of h. During the no trade period, the quota does not affect the pattern of production. The switching point does not depend on the level of the quota, since the pattern of trade is determined by the relative price of the both goods.






33


The level of quota does not affect this relative price, while the tariff does affect it. This means that the switching point does not change by the introduction of the quota. At the switching point T, the volume f

the imports of type 1 goods equals the min [yf,, h], where yf is the import under free trade at T. In each case, there is the time which the quota is binding. The foreign production of the type 1 goods increases over time due to the learning-by-doing effect, the imports will reach the restriction at some -ime even though h is set at any high level.

Assume that h
Let assume that h>yf at the switching point T. Since the quota is not binding, the home country imports type 1 goods and does not produce it. As time passes, the imports increase and reach the limit h at some

time T1. After this time the domestic industry can survive and continue to produce. In this case, the domestic production of the type 1 goods is discontinued from T to T1.

I summarize the arguments as follows;



Proposition 3

The tariff delays the time when the trade takes place, but

domestic production of the protected industry vanishes after this time. On the other hand, the quota does not affect the above time but protects domestic production after the introduction of trade.



Notice that the qualitative results do not depend on the assumption that the tariff revenue and the rent from the uuota auction are equally redistribution to the consumers. For example, if all the tariff revenue






34


is given to the type 1 consumers, the revenue goes to the foreign industry. Domestic production of type 1 goods can not be revived in the future. The manner of redistribution changes the level of the outputs but does not affect the pattern of trade.








3.6. Concluding Remarks



The chapter has developed a two-country dynamic Ricardian model of trade in similar products along the lines suggested by Bhagwati(1982). The model was based on differential national tastes, learning-by-doing and transportation costs. We analyzed the role of these three features in determining the pattern of intraindustry trade. A country exports those similar products which correspond to a larger domestic market. An income transfer policy can reverse the initial pattern of trade in similar products. A tariff or an import quota delays the time when trade starts and reduces the volume of trade. The simple structure of the model allows several possible extensions and generalizations.















CHAPTER 4


TRADE LIBERALIZATION IN AN OPEN ECONOMY WITH A UNIONIZED SECTOR




4.1. Introduction




During the last two decades, the world economy has experienced three distinct and parallel structural changes. First, zhe process of trade liberalization has intensified through multilateral and regional trade agreements ( i.e. NAFTA, Europe 1992,and the Uruguay round). Second, there has been an acceleration of technological progress

e.g. see Berman, Bound and Griliches (1994) ).And third, there has been a rise in wage income inequality due to a decline in the demand for less skilled workers ( i.e. see Richardson (1995) for a survey

Several economists have voiced their concerns -regarding the possible causal relationship between trade liberalization and the decline

in the demand for less skilled workers. ( e.g. Wood(1995) and Borjas and Ramey (1994) among others ) Other economists have focused on the effects of exogenous technological progress on the rise of wage income inequality. Ce.g. Richardson(1995), Berman et al. (1994), and

Bhagwati (1995) among others ) At the core of the debate on where technology or trade are responsible for the decline in the demand for less skilled workers is the Stolper and Samuelson (1941) mechanism that relates factor prices to commodity prices. According to this mechanism, a

reduction in the domestic price of the importable commodity reduces the wage of the factor is used intensively in the production of the importable. An implication of this mechanism is that trade liberalization




35







36


that causes a decline in the domestic price of 4iport competing

industries that use less skilled workers intensively will eventually reduce the relative demand and the wage of these workers.

The present chapter analyzes the validity of --he Stolper and Samuelson (1941) mechanism in economies that are characterized by imperfectly competitive product and factor markets. For this purpose, I construct a general equilibrium model of a small open economy with the

following features: The export producing sector is perfectly competitive and utilizes two factors of production, skilled and unskilled labor. The

import -competing sector consists of a unionized firm that is protected from foreign competition through an import quota. The domestic monopolist

utilizes both skilled and unskilled labor under a constant returns to scale technology. However, unlike the labor demand for skilled workers which is assumed to be competitive, we assume that the unskilled workers

in the import -competing sector have formed a labor union. The union bargains with the domestic monopolist over the negotiated wage and the employment of unskilled labor. The bargaining process is modeled as an efficient Nash bargaining game that results in simultaneous determination of employment and the wage.

The model is used to analyze the effects of trade liberalization

i.e. an increase in the quota ) and Immigration on income distribution and domestic prices. I identify conditions on the parameters of the model

that establish the robustness of the Stoiper- Samuelson mechanism in the presence of sector-specific labor union. I also characterize the sufficient conditions for the existence of a unique equilibrium.

The model combines two standards of literature that have dealt with similar issues. Batra(1972), Melvin and Warne(1973) and Ikeda(1990) have analyzed the general equilibrium behavior of a domestic monopolist assuming perfect competition in factor markets. Brander and Spencer(1988) and Mezzetti and Dinopoulos(1991) have developed partial-equilibrium






37


models to examine the effects of trade liberalization in unionized labor demand. In addition, Dinopoulos and Lane(1992), (1997) and Lane and Dinopoulos(l995) have constructed general equilibrium models with sectorspecific factors (instead of allowing factor mobility among all production factors )to analyze the effects of trade liberalization on state-owned firms that are modeled as labor unions.

The rest of this chapter is organized as follows: I formulate our model in section 4.2. In section 4.3, uniqueness of equilibrium is proven following Batra(1972) The impacts of bargaining parameters are examined in section 4.4. In section 4.5, I investigate the effects of trade liberalization on factor rewards and whether Stolper-Samuelson

mechanism works as a result. I also derive the impact of immigration on factor rewards. The final section is devoted to concluding remarks.










4.2. The Model



Consider a small open economy with two sectors. A unionized

sector (X) which the wage and employment are determined through Nash bargaining, and a competitive sector (Y) I designate the production

function by X = F(L,, H,, ) and Y = G(Lyr Hy ) where Li and Hi denote

unskilled labor and skilled labor employed in the i-th sector with i

=x or y. The production functions exhibit constant to scale and can

be written in the intensive form,



x= X/H,, =f (lx,), (1)

y = Y/Hy = g (1y), (2)






38


where i,(ly) shows the unskilled-skilled labor intensity in the unionized(competitive) sector. I assume that the representative consumer is characterized by a homothetic utility function which enables us to write the inverse demand function as:



p = u(z), u'(z)>0, (3)




where z=Yc/Xc, the relative demand, and subscript c denotes consumption. We normalize the price of good y to the unity and there p = Px is the price of good X. Moreover, for expositional simplicity, the price elasticity of demand for X denoted by E, and assumed to be constant and larger than one. If the utility function is C.E.S. type, the demand elasticity satisfies the above assumptions. [ Batra (1971),Melvin and Warne (1973), Konishi, Okuno- Fujiwara, and Suzumura (1990) adopt this assumption. I

I next describe the bargaining process and suppose that only the unskilled labor in sector x can establish a labor union and

negotiate with the management about level of their employment(Lx) and wage(r,). Following Mezzetti and Dinopoulos (1991), i assume that output x is produced by a domestic monopolist. Skilled labor and unskilled

labor working in the competitive sector can not form unions and are forced to accept the prevailing competitive wages, w and ry. The behavior of the labor union is modeled through a utility function that depends on employment and the excess wage(r. ry),



U = (r, r')8L .x (4)






39


y(8) represents the elasticity of U with respect to excess wage(employment). Following the literature, we assume that the union is employment-oriented, that is, y > 0.

The management of the firm is interested in maximizing its profit, l,




H" = pX r.Lx wHy. (5)



Both parties engage in efficient Nash bargaining. The threat point is a pair of -wLx and 0. This pair is obtained under the assumption that the union strikes ( that is, x = 0 ) if there is disagreement during the bargaining process. The generalized Nashbargaining product, G thus is constructed as



W = (pX-rL) I-a U, (6)



where parameter a is the bargaining power of the union and 0 a: 1.

I assume that under free trade, this economy imports the good of the unionized sector and exports the good of the competitive sector. The government imposes an import quota Q as trade restriction measure. Domestic consumption of the two goods is Xc = F(Lx,,HX) + Q and Y, = G(LyHy) Q. I assume that the consumption of the goods Y is positive under the quota, that is, G Q > 0.






40


4.3. Uniqueness of Equilibrium





In this section, I focus our attention to issues related to the uniqueness of momentary equilibrium relative to a set of bargaining

parameters and the level of quota. Uniqueness of equilibrium is a precondition for the following comparative statics analysis.

The competitive firm hires unskilled and skilled labor to maximize its profit, given the output price (normalized to one) and the factor prices. The optimization conditions for the firm are



=y g' (1y) (7)

w g g(1y) l'g' (l'). (8)




The above conditions give the relation between ly, and the relative

factor price in the competitive sector my~ = ry/w, as ly,= ly (coy) Differentiating ly. with respect to wo,, we have




if' = al / awy = (g lyg' ) /gg" <0 (9)





Unskilled labor employment and the wage in the unionized sector are determined through a efficient Nash bargaining process. The solution to this bargaining problem is obtained by maximizing the generalized Nash product with respect to the negotiated wage and employment. The two first order conditions of this problem are;






41


Contract curve.



9),= Y_0af f. (10)




Nash bargaining curve



0, ft +fI' )
(6_) +fxf-11)



where m is the relative factor price in the unionized sector,

r,/w and k = ya/(1-+(xy) is a parameter. Combing the Contract curve and Nash bargaining curve, I obtain the basic relation between coy and lx.




f' xB, (12)
Y- f -lxf'




where

B = [ ef/(g 1)ixf'] + 1 1 > 1, (13)



and



1 = a( y O)l(1 x + cxy ). (14)




Since s > 1 and f > 1,f', B is larger than one. Comparing the Nash bargaining curve (11) and (12) and k > P, I find that the unionized unskilled workers earn a higher wage, r,, than do the unskilled workers in the competitive sector. The skilled labor in the unionized sector is employed to maximize profits. The first order conditions for profit maximization is






42




w = p(l 1/s) ( f 1,xf ). (15)



I turn to the factor market equilibrium to derive a relation between kx and my. The supply of the both factors, L and H, are fixed. The full employment conditions are described as:



S+ hy =1, (16)

hlx +hyly = 1, (17)



where hs(hy) = Hx(Hy)/H and 1 = L/H is the unskilled:skilled labor ratio of the whole economy. Solving the full employment conditions (16), (17) with respect to hx ,hy and inserting these two variables into the definition of relative demand we get



g(lx 1) (lx ly)q
z= (18)
f(l-ly)+(lx -ly)q



where q = Q/H and ghy q > 0 must hold Equations (9), (12) and (18) determine z as a function of oy, given parameters H Q and 1. It is useful for intuitive purpose to investigate this functional relation. From (12), lx can be written as l = lx (msy). Differentiating (12), we nave



l'x = 81x / &my = M(f lxf')/ff" <0 (19)




where

Y = (20)
1-fl+pfE- fEx(f -lxf') /lXf






43


and




E=->1, (21)
6-1



Note that 0 < M < 1 since E > 1. The unskilled labor intensity of the unionized sector decreases as the relative wage in the competitive sector goes up. Note that (19) relates lx and the relative wage prevailing in the competitive sector, not in the unionized sector. In our model, the relative wages differ across the sectors because of the bargaining in the unionized sector. Taking into account of (19), differentiation of (18) with respect to ly leads to



dz/doy = N,(l-ly)dlx/daoy + N2(lx-l)dly/day, (22)

where




N g(f -lxf'+kf')+q(g-f+f'[lx -ly]) (23)
NI = (23)
[f ( ly) + (lx ly)qj2



and



f(g+g'[l-ly ])+q(g'[1x -ly ]+g-f) (24)
N2 = (24)
(Il- ly ] + (lx ly)q)2




After some algebraic manipulation, I find that N, and N2 take positive values.

( Lemma 1 )

If ly > (<) lx then, z'(coy) > (<) 0






44


Following Batra(1972), Melvin and Warne(1973), I define the consumption price pe through the utility maximizing behavior of the representative consumer.



P u = z(O). (25)



Differentiating totally the consumption price function, I get



P=C dp = u'(z)z'(Oy ). (26)
dOy



I can obtain



( Lemma 2 )

Suppose that ly > (<) lx. Then, p,' > (<) 0.



The consumption price function is upward(downward)sloping with respect to the relative wage in the competitive sector, if the competitive(unionized) sector is unskilled labor intensive.

I also define the production price pp by the optimization behavior of each firm, (8) and (13).



E(g- lyg')
p ----- (27)
(f -lxf,)



Noting that kx = kx(oy) and E takes a constant value, we have




dp = D[ B(Mx ly) + S( -P), (28)
P dp y =






45


where




D>E (g-lIgf 0, (29)
(B + lwy)(1 + lroy(f lfg)



(f lf')
S= MlxOyEflx lxf' >0. (30)




f'(f -l.f')
PX f lf f" > 0. (31)




px is the elasticity of substitution between skilled and unskilled labor in the unionized sector. If the production function is the Cobb-Douglas type, this elasticity takes the value one. I can prove



(Lemma 3)

(3-1)Suppose that ly > i, and px_>l. Then, dp,/doy is negative.

(3-2) Suppose that ly < 1. and p. <1. and if Ml,, > ly. Then, dp,/dCoy is positive.



The assumption that Ml. > ly implies that the ranking of value intensity is the same as the physical one despite the wage differential that is induced by unionization. The production price function is downward(upward) sloping with respect to the relative factor price in the competitive sector, if the competitive (unionized ) sector is unskilled labor intensive.

Finally, I will establish the relative factor prices of

unionized labor in the competitive sector is negatively related to the






46


factor intensity in the whole economy. Consider equation (18) which includes not only the factor intensity, 1, but also the total quantity of skilled labor, H. To find the relation between o y and 1, we have to investigate two cases: (1) L is fixed and H changes, (2) L changes and H is fixed. I examine the first case.

Differentiating (18) with respect to 1, given the fixed cy, we have



Il a [qgh + f (g-lq1I)](y -I)
=67 [f (Ily) +q(l _ly )]2 (32)




We can show that the sign of z' is the same as that of ly lx. The relative demand function and the consumption price function shift downward ( upward ) if i, >(<) ly. The production price function does not change since it is independent of the total factor supply. Considering these observations, we get



Lemma 4 )

dmo/dl < 0 and dp/dl <(>) 0 ,if 1. >(<) ly.



I prove the same result for the second case. This lemma means that the relative unskilled wage in the competitive sector declines if the relative supply of unskilled labor increases, which seems to be plausible. Batra(1972) proves the lemma in the closed economy with C.E.S. utility function. Ikeda(1990) extends the Batra's result to a tax distorted economy. Homma(1977) examines this relation in the economies with homothetic utility function .

Lemma 4 implies that there exists a monotonic functional relation between 1 and cy which enables us to assert that our model has a






47


unique equilibrium. Combing Lemmas 1,2,3, and 4, 1 realize that the intersection of the consumption price function and the production one gives the unique equilibrium relative factor price in the competitive sector. I am now in a position to state the uniqueness property.



( Proposition 1 )

There is a unique equilibrium under the following two cases.

(sl) The competitive sector is unskilled labor intensive.

(s2) The unionized sector is unskilled labor intensive and

mi. > ly where M is given by (20) .



Proposition 1 identifies two restrictions that are fairy general and establish the uniqueness of equilibrium despite of the presence of sector-specific labor union. All remaining variables are also uniquely determined based on the relative factor price in the competitive sector.





4.4. The Effects of The Bargaining Parameters



In this section, I investigate the effects of the bargaining parameters on the relative wages of the both sectors. The bargaining parameters only appears in (12) through parameter 0. 1 easily prove that P increases(decreases) if a, y ( 0 )increases The unskilled labor J'ntensity of the unionized sector is a function of the relative wage in

the competitive sector and the bargaining parameters: 1,, = !,.( wy : P ) Keeping the relative wage constant and differentiating (12) with respect

to 1,, and P, we obtain






48


dlx It
d = Ifi > 0. (33)




The lx function shifts upwards as 0 increases. The relative demand equation (18) is a function ) Differentiating this with respect to lx and 0, we get



z = dz/dp = N-(l-l1)dlx/dp. (34)



From (34) I have



( Lemma 5 )

Suppose that lx >(<) ly. Then, zP >(<) 0.



If the unionized (competitive) sector is unskilled labor intensive, the relative demand goes up (down). Next, I examine the change in the consumption price function. The change is given by




pf = --- = u'(z)z (35)
C(35



We have that pf >(<) 0, if lx >(<) ly. The consumption price function shifts upwards(downwards) if the unionized(competitive) sector is unskilled labor intensive.

Differentiating the production price function (27) I derive



p p 7 El (g- iyg')lf"
P P (<0. (36)
PP d6 (f-1~,f'






49


I show that the production price function moves downwards as3 increases. Then, I prove



Proposition 2)



Suppose that the unionized sector is unskilled labor intensive. Then, the relative wage in the competitive sector increases if the bargaining power of the labor union,a, the preference for the

employment,y, increase and the preference for the excess wage,e decreases.



If the competitive sector is unskilled labor intensive, the change in relative wage is ambiguous.

I turn to examine the effects on the relative wage of the unionized sector. The Nash bargaining curve (11) gives the relation between the relative wage of the unionized sector and the bargaining parameters. I get,




dwx dcoy(37) d,8 d,8



where








Since X. is larger than P3, F is greater than one. The direction of change in the relative wage of the unionized sector is identical with that in the competitive sector. Using (35), I have the impact on the relative

wage gap across sectors, ()) (0y







50



d (-1)dcoy 39
Tf8 d/8



Since F-i is positive ,the wage gap changes the same as the relative wage in the competitive sector. I summarize these as follows:



Proposition 3)



Suppose that the unionized sector is unskilled labor intensive. (3-1) The relative wage of the unionized sector increases if the bargaining power of labor union,a, the preference for employment, increase and the preference for wage,0, decreases. (3-2) The relative wage gap between two sectors expands if the bargaining power of labor union,a, the preference for employment,Y, increase and the preference for excess wage,@ decreases.








4.5. The Effects of Trade Liberalization and Immigration






In this section, I investigate impacts of measures of trade liberalization on income distribution among three factors. We also examine the effects when foreign unskilled or skilled labor immigrate to the economy.

In the present model, trade liberalization is captured by an expansion of the existing import quota. Since factor intensity of each sector does not directly depend on the quota, the relative demand z can







51


be as z = z (coy:Q) For given cw,, I find that relaxation of the quota decreases the relative demand,




ZQ =& (Y If I y )+ g~, -1)] 0.(40)




The consumption price function (25) is



P. P lu (Z (oy~:Q1 ) ). (41)



The move of this function resulting from quota relaxation is given by









The consumption price function shifts downward if the government relaxes the import quota. The production price is determined only by the optimization behavior of each firm the labor union, and is independent of the quota level. Therefore, the production price function does not change. We establish



Lemma 7)

Suppose that the unionized( competitive ) sector is unskilled labor intensive. The quota relaxation decreases( increases ) the relative wage of unskilled workers in the competitive sector. The price of good X declines as a result of quota expansion.



Considering (35) and (37), I have the impacts on the relative

wage of unionized sector and the wage gap moves the same as the relative






52


wage in the competitive sector. Differentiating (7) and (8) with respect to the quota, we get



dr do ,
d- = (4
d d(43)



- = 1 (44)
dQ Y dQ



Consider the case where the unionized sector is unskilled labor intensive and the government relaxes its quota. The price of the good X declines and the relative wage of the unskilled labor in the competitive sector increases. (43) and (44) state that the wage of the skilled labor which is not employed intensively in the unionized sector goes up, while the unskilled labor wage in the competitive sector which is used intensively in the sector declines. This relation between the output price change and the factor price change implies that the traditional Stolper-Samuelson theorem holds in our model despite of unionization. If the competitive sector is unskilled labor intensive, this theorem also holds as for these factor prices.

I turn to examine the change of the wage of unionized labor. The wage is determined through efficient Nash bargaining



r,=p[Af / 1, +(1- A)f'/ El. (45)



The impact of the quota is given by

dr_ = dp ___+(1)f'/E]+p (f -f_) (1-A)f"El dx (46)







53


The first term of RHS of (46) is negative. The relative wage of the competitive sector decreases if 1,> ly. The second term is negative

since the brace in the second term and 1',, are negative. The increase in the price of X raises the wage of unionized labor which is intensively hired in the unionized sector. The Stolper-Samuelson relation holds. We point out that an interesting case may occur. Suppose that lx < ly. In this case, the output price of X decreases while the relative wage in the unionized sector increases. The two terms of RHS of (46) take opposite signs. I can not derive determinate conclusion. This implies that the Stolper-Samuelson relation may not hold if the effect through relative wage change( the second term ) is dominated by the price change effect ( the first term) I summarize these observations



Proposition 4)



4-1) Stolper-Samuelson theorem holds with respect to the wages of the skilled labor and the unskilled labor in the competitive sector. (4-2) Stolper-Samuelson theorem holds with respect to the wage in the unionized labor if the unionized sector is unskilled labor intensive.



I turn to investigate the effects of immigration on the income

distribution among three factor prices. Consider that foreign unskilled labor immigrate to the economy. This immigration increases the total amount of domestic unskilled labor, L and the unskilled: skilled labor ratio of the whole economy, 1. From Lemma 4, 1 can show that this immigration decreases (increases) the price of good X if the unionized sector(competitive) sector is unskilled labor intensive and decreases the relative wage of unskilled labor in the competitive sector independent of the factor intensity among two sectors.







54


The impacts on the skilled labor's wage and the unskilled labor's wage employed in the competitive sector are given



dr,,=g I daco<0 (7





dw do
-g";- >0. (48)
dl ~ di



The unskilled labor's wage in the competitive sector decreases, while the skilled labor's wage increases. This results does not depend the

intensity ranking of the two sectors. The both wages are determined by marginal productivity of factors. Since the unskilled labor supply rises in the economy, the skilled labor becomes less abundant while the unskilled labor does abundant. These change in the two factor price seem to be plausible.

The change in the unskilled labor's wage working in the unionized sector is



dr_ dp [A I-f+PA(l f'1) +(___f""dc,(9
dldl 12 dl



If the unionized sector is unskilled labor intensive, both the output price and the relative wage in the competitive sector go down.

Then, the wage of unskilled labor in the unionized sector decreases. However, if the competitive sector is unskilled labor intensive, the output price increases and the relative wage decreases. The total change in the wage in the unionized sector depends on the magnitudes of two effects. If the effect of output price change dominates that of the relative wage change, the unskilled wage in the unionized sector

increases. The unionized unskilled labor with monopoly power can earn






55


higher wage than the competitive wage, reflecting its marginal

productivity. The counterintuitive result comes from this bargaining power. I s umm arize the observations



( Proposition 5 )



(5-1) Immigration of unskilled labor increases the wage of skilled labor and decreases the wage of unskilled labor in the competitive sector.

(5-2) Immigration of unskilled labor decreases the wage of unskilled labor in the unionized sector if the unionized sector is unskilled labor intensive.



If the skilled labor immigrates to the economy, the unskilled: skilled labor ratio in the whole economy declines. The effects of unskilled labor immigration are immediately derived from proposition 5.





4.6. Concluding Remarks



I have formulated a two-by-two open general equilibrium model with bargaining between unskilled labor and the management of firm. The two parties negotiate over the wage and employment. The unionized firm has monopoly power in the product market. The other factor prices, (the wage of unskilled labor working in the competitive sector and that of skilled labor,) are determined by perfect competition. The government has imposed an import quota on the unionized sector.

First, I have examined uniqueness of the equilibrium. In our bargaining model, the physical factor ranking between two sectors does not necessarily coincide with the nominal one, because the sectorspecific bargaining generates a discrepancy between unskilled labor's






56


wages across sectors. I have established uniqueness of equilibrium under the condition that the order of the above two rankings is same. Second,

I have investigated the impact of the bargaining parameters on the relative factor prices of unionized and competitive sector. If the unionized sector is unskilled labor intensive, the relative wages of both

sectors increase as the union becomes more employment-oriented and its bargaining power rises. Third, I have examined the effects of trade liberalization (quota expansion) I have shown that the Stolper-Samuelson type relation holds as a result of the trade liberalization. Finally, I have derived the effects of immigration. Immigration of unskilled labor

favors the skilled labor's wage, while it reduces the unskilled labor's wage in the competitive sector. If the unionized sector is unskilled labor intensive, this immigration decreases the wage of the unionized workers.
















CHAPTER 5
CONCLUSIONS




The model in the chapter 2 shows that the level of DFI is substitutive (complementary) to the lobbying activity and protection when DFI occurs in the exporting (importing) sector. The DFI in the exporting sector reduces the level of protection, which encourages the foreign government to engage quid pro quo DFI. On the other hand, the DFI in the importing sector leads to a higher tariff rate. However if DFI increases the host-country government resistance to lobbying and this effect

overcomes the effect of lobbying activity, the tariff rate is reduced by DFI and there is scope for quid pro quo DFI. I show another channel by which DFI influences protection: the amount of labor devoted to

productive activities may increase as the result of DFI. The foreign country exports more and obtains a higher level of welfare. The foreign

government chooses DFI taking this "productive labor expansion effect" into account. This effect emerges in the general equilibrium framework of

the model in which the tariff is determined through resource-using lobbying activities.

Chapter 3 establishes the existence of unique switching point, T. Before T there is no trade. After T, trade occurs based on taste difference across countries. The government of the home country can completely reverse the existing pattern of trade by taxing the type 0 consumer and transferring this to the type 1 consumer during the no trade period. The tariff delays the time when the trade takes place, but domestic production of the protected industry vanishes after this time.





57






58


Cn the other hand, the quota does not affect the above time but protects domestic production after the introduction of trade.

In chapter 4, 1 prove uniqueness of the equilibrium of the model

-,nder the conditions that the physical factor ranking between two sectors

cincides with the nominal one and the magnitude of the elasticity of substitution between skilled and unskilled labor in the unionized sector. Next, I investigate the effects of trade liberalization(quota expansion). The Stolper-Samuelson type relation holds as a result of the trade liberalization. Finally, I derive the effects of immigration. Immigration of unskilled labor favors the skilled labor's wage, while it reduces the unskilled labor's wage in the competitive sector. If the unionized sector

unskilled labor intensive, this immigration decreases the wage of the unionized workers.












LIST OF REFERENCES





Baldwin, R. and P.R. Krugman, 1983, Market access and international
competition: A simulation study of 16K random access memories, in
R. Feenstra, ed. Empirical Research in International Trade,
Cambridge, MIT Press, 308-321.

Batra, R.N, 1972, Monopoly theory in general equilibrium, Journal of
Economic Theory 4, 355-371.

Berman, E, Bounnd. J.and Griliches, Z, 1994, Changes in the demand for
skilled labor with U.S. manufacturing: Evidence from the annual
survey of manufacturers, Quarterly Journal of Economics, 109,367397.

Bhagwati, J.N, 1982, Shifting comparative advantage, protectionist
demands, and policy response, in J. Bhagwati, ed. Import
Competition and Response Chicago, University of Chicago
Press,48-65.
Bhagwati, J., 1985, Protectionism: Old wine in new bottles, Journal of
Policy Modelling, 7, 23-34.

Bhagwati, J., 1986, Investing abroad, Esmee Fairbairn Lecture, Mimeo.

Bhagwati, J., 1987, VER's quid pro quo DFI and VIEs: Political-economytheoretic analysis, International Economic Journal, 1, 1-14.

Bhagwati,J., 1995, Trade and wages: A malign relationship?, Columbia
University, mineo.

Bhagwati, J., R. Brecher, E. Dinopoulos and T. N. Srinivasan, 1987, Quid
pro quo foreign investment and welfare: A political-economytheoretic model, Journal of Development Economics, 14, 127-138.

Bhagwati, J., E. Dinopoulos and K. Wong, 1992, "Quid pro quo foreign
investment, American Economic Review, 82, 245-253.

Bhagwati, J.N. and T.N. Srinivasan, 1983, Lectures on Trade Theory

Bhagwati, J.N. and T.N. Srinivasan, 1971, The theory of wage
differentials: Production response and factor price
equalisation, Journal of International Economics 1, 19-35.

Borjas, G, and V, Ramey, 1994, Time-series evidence on sources of trends
in wage inequality, American Economic Review Paper And
Proceedings, 84, 10-16.

Brander, J.A, and B.J. Spencer, 1983, A reciprocal dumping model of
international trade, Journal of International Economics 15, 313321.






59






60


Brander, J.A. and B.J. Spencer, 1988, Unionized oligopoly and
international trade policy, Journal of International Economics
24, 217-234.

Brock, W. and S. Magee, 1978, The economics of special interest
politics: The case of the tariff, American Economic Review, 68,
246-250.

Dinopoulos, E., 1984, The optimal tariff with revenue-seeking: A
contribution to the theory of DUP activities, in D. Collander
(ed.), Neoclassical Political Economy, Cambridge, Mass., Ballinger
Publishing Company, 129-140.

Dinopoulos, E, 1988, A formalization of the biological model of trade
in similar products, Journal of International Economics 25, 95110.

Dinopoulos, E., 1989, Quid pro quo foreign investment, Economics and
Politics, 1, 145-160.

Dinopoulos, E., 1992, Quid pro quo foreign investment and VERs: A Nash
bargaining approach, Economics and Politics, 4, 43-60.

Dinopoulos, E. and T. Lane,1992, Market liberalization policies in
a reforming socialist economy, IMF Staff Papers 39, 465-494.

Dinopoulos, E. and T. Lane, 1997, Piecemeal reforms in a transitional
economy, mimeo.

Dinopoulos, E, T.R. Lewis, and D.E.M. Sappington, 1993, Optimal
industrial targeting with unknown learning-by-doing, Forthcoming
in Journal of International Economics ,

Dinopoulos, E., and Kar-yiu, Wong, 1991, Quid pro quo foreign investment
and policy intervention, in K. Koekkoek and C. Mennes (eds.),
International Trade and Global Development, New York,
Routledge, 162-190.

Dixit, A.K, 1988, Optimal trade and industrial policy for the U,S.
automobile industry, in R. Feenstra, ed. Empirical Methods for
International Trade, Cambridge, MIT Press, 137-158.

Feenstra, R.C, 1982, Product creation and trade pattern: A theoretical
note on the biological model of trade in similar product, in J.
Bhagwati, ed. Import Competition and Response, Chicago, University
of Chicago Press, 235-244.

Feenstra, R.C, and J. Bhagwati, 1982, Tariff seeking and the efficient
tariff, in J. Bhagwati (ed.), Import Competition and Response,
Chicago, University of Chicago Press, 245-258.

Findlay, R. and S. Wellisz, 1982, Endogenous tariff, the political
economy of trade restrictions and welfare, in J. Bhagwati (ed.), Import Competition and Response, Chicago, University of Chicago
Press, 223-234.

Freeman, R.B, 1995, Are your wages set in Beijing?, Journal of Economic
Perspectives 9, 15-32.






61


Helpman, E, 1981, International trade in the presence of product
differentiation, economies of scale, and monopolistic competition: A Chamberlinian-Hecksher-Olhin approach, Journal of International
Economics 11, 305-340.

Homma, M. 1977, A comparative static analysis of tax incidence, Journal
of Public Economics 8, 53-65.

Ikeda, H. 1990, Monopoly and tax incidence: A general equilibrium
analysis, Discussion Paper, Tezukayama University, Japan.

Jones, R. W., 1971, Distortions in factor markets and the general
equilibrium models, Journal of Political Economy 79, 557-572.

Konishi, H., Okuno-Fujiwara, M. and K. Suzumura, 1990, Oligopolistic
competition and economic welfare: A general equilibrium analysis of entry regulation and tax-subsidy, Journal of Public Economics
42, 67-88.

Krugman, P.R, 1979, Increasing returns, monopolistic competition, and
international trade, Journal of International Economics 9, 469479.

Krugman, P.R, 1981, Intra industry specialization and the gains from
trade, Journal of Political Economy 89, 959-974.

Krugman, P.R, 1987, A narrow moving band, the Dutch disease, and the
competitive consequences of Mrs. Thather: Notes on trade in the
presence of dynamic scale economies, Journal of International
Economics 27, 41-55.

Lancaster, K, 1980, Intra-industry trade under perfect monopolistic
competition, Journal of International Economics 10, 151-175.

Lane, T.D. and E. Dinopoulos, 1992, Fiscal constraints on marketoriented reform in a socialist economy, IMF Staff Papers, 39,184203.

Mayer, W., 1984, Endogenous tariff formation, American Economic Review,
74,970-985.

McDonald, I.M. and R. Solow, 1981, Wage bargaining and employment,
American Economic Review 71, 896-908.

Melvin, J.R. and R.D. Warne, 1973, Monopoly and the theory of
international trade, Journal of International Economics 3,
117-134.

Mezzetti, C. and E. Dinopoulos, 1991, Domestic unionization and
import competition, Journal of International Economics 31, 79100.

Nash, J., 1950, The bargaining problem, Econometrica 28, 155-162.

Richardson, J.D, 1995, Income inequality and trade: How to think, What
to conclude, Journal of Economic Perspectives 9, 33-55.

Santoni, M, 1996, Union-oligopoly sequential bargaining: Trade and
industrial policies, Oxford Economic Papers 48, 640-663.






62


Stolper, W. and P. Samuelson, 1941, Protection and real wages, Review of
Economic Studies 9, 58-73.

Wong, Kar-yiu., 1989, Optimal threat to trade restriction and quid pro
quo foreign investment, Economics and Politics, 1, 277-300.

Wood, A, 1995, How trade hurt unskilled workers, Journal of Economic
Perspective, 9, 57-80.

Zhao, L., 1991, Endogenous protection and direct foreign investment,
University of Florida, mimeo.















BIOGRAPHICAL SKETCH




Makoto Okamura was born in 1956 in Japan. After studying at Keio University, he obtained a Master of Arts in economics at Osaka University in 1982. He entered the PH.D program in economics at University of Florida in 1990.
















































63











I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Elias- :-I~~f!uios, Chair
Professor of Economics

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Richard Romano
Professor of Economics

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Bin Xu
Assistant Professor of Economics


I certify that I hane read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.




Jao~i s Seale
Associate Professor of Food and Resource Economics



This dissertation was submitted to the Graduate Faculty of the
Department of Economics in the College of Business Administration and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy.

May, 1997



Dean, Graduate School




Full Text
12
M*y = O (E*x, Kd + Kf Td + Tf L 1 )
= T( E*x, Kd + Kf Td + Tf ) (23)
The foreign government has a intertemporal utility function
V = W1 (C'1x, C*1y) + pW2 (C*2X, C*2y), (24)
where p is the discount factor. The optimization problem is
Max V
s.t. C'1* = Qnx E*x,
C*1y = Qny + ¥( E*xr Kd + Kf Td + Tf ) (25)
The foreign government chooses the level of DFI. Given this level, all
remaining variables are determined.
2.4. The Basic Comparative Statics
I start with the comparative statics of the second period
production equilibrium with respect to DFI and mobility of labor.


ACKNOWLEDGMENTS
I would like to express deep appreciation and gratitude to Dr.
Elias Dinopoulos, chairman of my supervisory committee. Without his
invaluable suggestions and encouragement, I could not have completed
this dissertation. I also gratefully acknowledge insightful comments by
Dr. Richard Romano, Dr. Bin Xu and Dr. James Seale. I really appreciate
valuable assistance from my friends, Mr. Shinji Yane and Mr. Sang H.
Lee.
IV


10
t'PxF w = 0. (12)
where FL = 5F/&LX.
The second order condition is assumed to be satisfied.
M = (1+t) FL1Ft" (t' FL)2 > 0, <13)
where FLL is the second order derivative. The economic problem of the
agricultural sector is
max EL = [l + s( La-Lj] PyG(T, Ly)-wLy-wLa- (14)
LL>
I obtain the first-order conditions
(1+t)PyGt, w = 0. (15)
s'pyG w = 0. (16)
The second order condition is
A = (l+s)GLLs"G (s'GJ2 > 0. (17)
The equilibrium in the labor market is given by the full employment
condition.
Lx + Lm + Lji + La = L, (18)
where L stands for the fixed endowment of labor. Equations (11), (12),
(15) (16) and (18) describe the equilibrium of the production sector.
These five equations determine the five endogenous variables Lx Ly Lm La,


THREE ESSAYS ON INTERNATIONAL TRADE AND WELFARE
By
MAKOTO OKAMURA
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1997

Copyright 1997
by
MAKOTO OKAMURA

I dedicate this dissertation to memory of the late Hiroshi
Okamura, my father and Misuzu Okamura, my mother.

ACKNOWLEDGMENTS
I would like to express deep appreciation and gratitude to Dr.
Elias Dinopoulos, chairman of my supervisory committee. Without his
invaluable suggestions and encouragement, I could not have completed
this dissertation. I also gratefully acknowledge insightful comments by
Dr. Richard Romano, Dr. Bin Xu and Dr. James Seale. I really appreciate
valuable assistance from my friends, Mr. Shinji Yane and Mr. Sang H.
Lee.
IV

TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iv
ABSTRACT vi
CHAPTERS
1 INTRODUCTION 1
2 ENDOGENOUS TARIFF FORMATION AND DIRECT FOREIGN INVESTMENT:
GENERALEQUILIBRIUM ANALYSIS 4
2.1 Introduction 4
2.2 The Model 6
2.3 The Equilibrium 9
2.4 The Basic Comparative Statics 12
2.5 Optimal Direct Foreign Investment 15
2.6 Quid Pro Quo Direct Foreign Investment 17
2.7 Concluding Remarks 20
3 LEARNING-BY-DOING AND PATTERN OF INTRAINDUSTRY TRADE 21
3.1 Introduction 21
3.2 The Model 23
3.3 Trade at the initial equilibrium 25
3.4 The Dynamic Behavior of the Economy 26
3.5 The Effects of Government Policies 30
3.6 Concluding Remarks 34
4 TRADE LIBERALIZATION IN AN OPEN ECONOMY WITH A UNIONIZED SECTOR 35
4.1 Introduction 35
4.2 The Model 37
4.3 Uniqueness of Equilibrium 40
4.4 The Effects of Bargaining Parameters 47
4.5 The Effects of Trade Liberalization and Immigration 50
4.6 Concluding Remarks 55
5 CONCLUSIONS 57
LIST OF REFERENCES 59
BIOGRAPHICAL SKETCH 63
v

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
THREEE ESSAYS ON INTERNATIONAL TRADE AND WELFARE
By
MAKOTO OKAMURA
May, 1997
Chairman: Elias Dinopoulos
Major Department: Economics
I formulate a two-period, two-sector, specific factor, general
equilibrium trade model with endogenous tariff formation. The tariff rate
depends on labor devoted to lobbying in each of the two industries. We
examine the possibility of quid pro quo direct foreign investment (DFI).
It is shown that more DFI in the exporting sector reduces the level of
protection and quid pro quo DFI occurs. Quid pro quo DFI takes place if
DFI in the importing sector increases directly the ability of the host-
country government to resist protection.
I construct a two-country Ricardian dynamic general
equilibrium model with learning-by-doing and transportation cost. The
two countries differ only in the distribution of the consumer's
preference. At the initial time, no trade occurs due to transportation
costs. There exists a unique time when trade begins. After that time,
each country exports the goods for which it has a stronger preference
vi

than does the other country. A government can completely reverse the
pattern of trade by a tax-transfer policy to change the distribution of
income. A tariff delays the time when trade begins, while a quota does
not affect the timing of trade.
Finally, I formulate a open economy model in a two-by-two
general equilibrium setting. Unskilled workers establish a labor union
and bargain with a firm about their employment level and wage, while
skilled workers are employed in a competitive market. Both parties are
engaged in Nash bargaining. The labor union is assumed to be employment
oriented. I establish that trade liberalization decreases the wage on
unskilled workers in the unionized sector if the unionized sector is
labor unskilled labor intensive. I also show that immigration of
unskilled labor decreases the wage of unskilled unionized labor if the
unionized sector is unskilled labor intensive.
Vll

CHAPTER 1
INTRODUCTION
According to traditional theories of international trade,
typically, in the original Hechsher-Ohlin-Samuelson general equilibrium
framework, three assumptions are imposed: First, capital movement from
one country to another country is seen essentially as s substitute to a
commodity trade. Direct Foreign Investment(DFI) is induced flow from the
capital abundant country with low reward to the capital scare country
with high reward to seek higher return. Second, technologies of trading
countries are assumed to be given. That is, no technical changes occur
at these countries. Third, in both product markets and factor markets,
perfect competition prevail. In this dissertation, I construct three
models to examine and extend these three assumptions.
Chapter 2 of the dissertation analyzes political-economic aspect
of DFI. There have been two distinct approaches, which attempts to
incorporate various political-economy considerations. The first strand of
literature has focused on product markets by modeling resource-using
lobbying or by modeling voting behavior. The level of protection is
determined through the self-interest maximizing behavior of economic
agents. The second strand of literature is associated with "Quid Pro Quo"
Direct Foreign Investment (DFI) This strand examines how DFI can be used
to defuse the threat of protection (or the actual level of protection)
based on political-economy considerations DFI aims at diffusing future
protection even if it is associated with current losses. I construct a
1

o
two-period, two-sector, specific factor model with DFI and endogenous
lobbying to integrate these two strands. In this model, I examine in
which sector, the exporting or importing sector is engaged in quid pro
quo DFI to defuse future protection of host country.
Chapter 3 builds a simple dynamic Ricardian general equilibrium
model with learning-by-doing effects of technologies. The learning-by-
doing effect means that each firm can learn from its experience and
improve its technology. Two types of consumers in each of the two
countries are considered. Each type has preferences over a particular
set of similar products. The distribution of the two types of consumers
differs across the two countries. Current productivity of labor, which
the only factor of production, increases over time because of learning-
by-doing considerations. The dynamic nature of the model allows us to
determine the duration of autarky based on transportation costs.
Transportation costs provide a realistic and natural way to generate an
initial period without trade, which is necessary to allow the
interaction between national differential tastes and learning-by-doing.
I investigates the effects of domestic income transfers across different
types of consumers, tariffs and import quotas on the pattern and volume
of trade.
Chapter 4 of the dissertation constructs a general equilibrium
model of an open economy with unionized sector. The export producing
sector is perfectly competitive and utilizes two factors of production,
skilled and unskilled labor. The import-competing sector consists of a
unionized firm that is protected from foreign competition through an
import quota. The domestic monopolist utilizes both skilled and unskilled
labor under a constant returns to scale technology. However, unlike the

3
labor demand for skilled workers which is assumed to be competitive, I
assume that the unskilled workers in the import-competing sector have
formed a labor union. The union bargains with the domestic monopolist
over the negotiated wage and the employment of unskilled labor. The
bargaining process is modeled as an efficient Nash bargaining game that
results in simultaneous determination of employment and the wage. I
analyze the effects of trade liberalization { i.e. an increase in the
quota ) and immigration on income distribution and domestic prices. I
identify conditions on the parameters of the model that establish the
robustness of the Stolper- Samuelson mechanism in the presence of sector-
specific labor union. I also characterize conditions for the existence of
a unique equilibrium.

CHAPTER 2
ENDOGENOUS TARIFF FORMATION AND DIRECT FOREIGN INVESTMENT: A GENERAL
EQUILIBRIUM ANALYSIS
2.1. Introduction
There have been two distinct approaches to the theory of
endogenous protection, which attempts to incorporate various political-
economy considerations in the formation of tariffs, quotas and other
instruments of protection. The first strand of literature has focused on
product markets by modeling resource-using lobbying (i. e., Findlay and
Wellisz (1982)) or by modeling voting behavior (i. e., Mayer (1984)). In
both cases, the level of protection is determined through the self-
interest maximizing behavior of economic agents. The second strand of
literature is associated with "Quid Pro Quo" Direct Foreign Investment
(DFI) This strand examines how DFI can be used to defuse the threat of
protection (or the actual level of protection) based on political-economy
considerations. Bhagwati (1985, 1986, 1987) has identified the new type
of DFI which aims at diffusing future protection even if it is associated
with current losses. Bhagwati et al. (1987), Dinopoulos (1989, 1992),
Wong (1989) Dinopoulos and Wong (1991), and Zhao (1991) have developed
models of Quid Pro Quo DFI.
These two distinct approaches to endogenous protection aim at
understanding the instruments and channels which determine the level of
protection, but have ignored each other's implications. For example,
although Findlay and Wellisz (1982) have emphasized the resources which
are utilized in the lobbying process within the context of general
equilibrium, Bhagwati et al. (1987) do not model the resource cost of
4

5
lobbying. In addition, by focusing on product markets, the first strand
the literature has abstracted from factor mobility influences on the
level of protection. Moreover, since the Quid Pro Quo DFI literature
examines the potential for defusing (or reducing ) existing protection,
it has opposite implications to the resource using lobbying literature
which studies the conditions under which protection emerges. Therefore,
there is some scope for integrating the two approaches in order to
analyze how endogenous protection based on resource using lobbying is
affected by DFI.
Quid Pro Quo DFI corresponds to a novel form of foreign
lobbying which utilizes resources in the present (by incurring a welfare
loss for the foreign country) in order to reduce protection in the host
country in the future (with the associated benefits) Therefore, it is
possible to build a model of endogenous protection with domestic lobbying
which uses resources and foreign lobbying which causes DFI in a quid pro
quo fashion. I construct a two-period, two-country, specific-factor
general equilibrium trade model with direct foreign investment.
Following Findlay and Wellisz (1982), a tariff formation function is
added in the second period. Free trade prevails in the first period.
The two-period, two-country framework is adopted from Bhagwati et al
(1987) .
By adopting the two-sector specific-factor model, I can examine
in which sector, the exporting or the importing, does quid pro quo DFI
take place. This paper shows that the level of DFI is substitutive
(complementary) to the lobbying activity and protection when DFI occurs
in the exporting (importing) sector. This reduces the level of
protection, which encourages the foreign government to engage quid pro
quo DFI. On the other hand, the difference between lobbying activities
expands as the result of DFI in the importing sector, which leads to a
higher tariff rate. However, if DFI increases the host-country government

6
resistance to lobbying and this effect overcomes the effect of lobbying
activity, the tariff rate is reduced by DFI and there is scope for quid
pro quo DFI. Another channel by which DFI influences protection is also
shown: the amount of labor devoted to productive activities may increase
as the result of DFI. The foreign country exports more and obtains a
higher level of welfare. The foreign government chooses DFI taking this
"productive labor expansion effect" into account. This effect emerges in
the general equilibrium framework of the model in which the tariff is
determined through resource-using lobbying activities.
Section 2.2 formulates the model. Section 2.3 characterizes the
equilibrium. Section 2.4 develops basic comparative statics. Section 2.5
calculates the optimal DFI levels. Section 2.6 analyzes the possibility
of quid pro quo DFI and section 2.7 offers the concluding remarks of the
chapter.
2.2. The Model
I formulate a two-country, specific-factor, general-equilibrium
model with endogenous tariff formation using building blocks from Findlay
and Wellisz (1982). There are two countries, foreign and domestic, in the
world. A manufactured good and an agricultural good are produced under
perfect competition and traded in both countries.
First I describe the host country. The manufactured (agricultural)
good is produced by a specific factor, capital (land), and a mobile
factor, labor. I denote the manufactured (agricultural) good by X (Y) .
The production functions are
X = F(K,LX), (1)
Y = G (T, Ly) ,
(2)

7
K = Kd + Kf, T = Td + Tf. (3)
The host country government imposes a tariff (subsidy) on each industry.
The domestic price of each good is
qx
= (1
+ t
)Px /
(4)
%
= (1
+ S
) Py
(5)
where
qx
(qy)
is
the domestic
price of
X (Y) and px (py) is the world
price
of
X (Y) .
The tariff
(subsidy)
rate is determined endogenously
through lobbying by each industry. Each industry hires lobbyists who put
political pressure to the government to obtain protection. The lobbying
activity utilizes only the following endogenous protection function which
determines the size of the tariff as a function of lobbying
t = t( Lm La
),
s = s (La -
71m ) t
(6)
where Lm (La)
is a
lobbying
input by manufactured (agricultural)
sector
We assume that
the
shape of
the tariff (subsidy) function, is as
follows:
t (o)
0,
t' > 0,
t" < 0,
S (o) =
o,
s' >0,
s" < 0.
(7)
This implies that for a given rival's lobbying behavior, the marginal
product of the lobbying by each industry is positive but decreasing.
These equations mean that when the X industry faces an import tariff, the
Y industry faces an export tax and vice versa. A representative consumer
has a well-behaved utility function U(X, Y). The consumer demand for each
good can be written, Xd(qx,qy,I) and Yd(qx,qy,I), where I denotes income.
I turn to the foreign country. The transformation curve does
not change over periods,

8
Q> H(Q*x,K* Kf,T" Tf), (8)
where Qy(Q*x) is the output of the agricultural (manufactured) good and K*
(T*) is the existing stock of capital (land). The foreign country exports
(imports) the manufactured (agricultural) good. It faces the reciprocal
demand function (offer curve)
M*y = where
M*y : foreign import of good Y,
E*x : foreign export of good X,
L : the amount of the labor devoted to the lobbying,
T : relative tariff rate, (1 + t)/(l + s)
The preferences of the consumers are summarized as the social utility
function, W(C*x,Cy) which depends on the consumption of both goods.
Since perfect competition prevails in every market, each individual
firm has no market power. The foreign firm cannot affect the tariff
level through the DFI. To examine the quid pro quo DFI, we assume that
the foreign government controls the total amount of the DFI. The
government determines the level of DFI to maximize its national welfare.
Thus, there are three players in the game we consider: the foreign
government, the host-country representative manufacturing firm and the
host-country representative agricultural firm.
There are two periods in the model. The timing of the game is as
follows:
Period 1. 1. The foreign government chooses DFI which remain fixed.
2. There exists no lobbying and free trade prevails.
Period 2. 3. Both domestic industries lobby and ~he endogenous
tariff rate is determined.

9
4. Given this tariff rate, trade occurs.
The foreign government moves as the Stackelberg leader, while the
two host-country firms move as simultaneous Stackelberg followers. The
structure of the model can generate quid pro quo DFI, and it is identical
to the one used by Bhagwati (1987) The foreign government does not
necessarily choose DFI to maximize its national welfare under free trade
(the first period welfare). It could endure a first-period welfare loss
for the purpose of obtaining the larger second period welfare.
2.3. The Equilibrium
This section analyzes the equilibrium of the model. I solve for
the equilibrium in the usual backward fashion. In the second period,
given a level of DFI, and the domestic prices of goods and labor, each
domestic firm chooses its labor inputs for production and lobbying
simultaneously in order to maximize its myopic (one-period) profit. The
profit of the manufacturing firm is
max n* =[l + t(Lm -La)]PxF(K,Lx)-wLx-wLm. (10)
Lx.Ly
where w is the wage rate. I impose the Cournot assumption with respect to
the lobbying behavior, that is, each firm decides its lobbying activity,
taking the rival's lobbying activity as given. The first order conditions
for the above problem are
(1 + t)pxFL w 0,
(11)

10
t'PxF w = 0. (12)
where FL = 5F/&LX.
The second order condition is assumed to be satisfied.
M = (1+t) FL1Ft" (t' FL)2 > 0, <13)
where FLL is the second order derivative. The economic problem of the
agricultural sector is
max EL = [l + s( La-Lj] PyG(T, Ly)-wLy-wLa- (14)
LL>
I obtain the first-order conditions
(1+t)PyGt, w = 0. (15)
s'pyG w = 0. (16)
The second order condition is
A = (l+s)GLLs"G (s'GJ2 > 0. (17)
The equilibrium in the labor market is given by the full employment
condition.
Lx + Lm + Lji + La = L, (18)
where L stands for the fixed endowment of labor. Equations (11), (12),
(15) (16) and (18) describe the equilibrium of the production sector.
These five equations determine the five endogenous variables Lx Ly Lm La,

11
and w. We assume that this system of equations has a unique solution for
a given K, T, Px, Py, and L. The production level of each good, the
endogenous tariff rate, and the domestic prices are determined based on
these variables. Finally, by substituting the values of the endogenous
variables into the demand equation, I can calculate the demands for both
goods, imports and exports.
Given the equilibrium of the host country and the level of DFI, the
foreign government chooses outputs of goods and the volume of trade in
order to maximize its national welfare, W(C*x,C*y). It solves the
following problem:
(19)
where the superscript denotes period 2. The maximum possible welfare
level as a function of direct foreign investment in the first period is
denoted by W2(K£,T£) .
Consider now the first period. Since there is no lobbying, the
equilibrium is described by the specific-factor model with direct foreign
investment. Given the level of DFI, the equilibrium values of Lx, Ly and
w in the home country are defined by
pxFi(K,Lx) w = 0
(20)
PyGi(T,Ly) w 0,
(21)
L* + Ly L.
(22)
The foreign country's reciprocal demand function in the case of free
trade is simplified as

12
M*y = O (E*x, Kd + Kf Td + Tf L 1 )
= T( E*x, Kd + Kf Td + Tf ) (23)
The foreign government has a intertemporal utility function
V = W1 (C'1x, C*1y) + pW2 (C*2X, C*2y), (24)
where p is the discount factor. The optimization problem is
Max V
s.t. C'1* = Qnx E*x,
C*1y = Qny + ¥( E*xr Kd + Kf Td + Tf ) (25)
The foreign government chooses the level of DFI. Given this level, all
remaining variables are determined.
2.4. The Basic Comparative Statics
I start with the comparative statics of the second period
production equilibrium with respect to DFI and mobility of labor.

13
( Proposition 1 )
(A) The effects of DFI on the manufacturing (importing) sector.
(1-1) The labor input of the manufacturing industry for production and
the total labor input of the manufacturing industry increase, if the
following condition is satisfied.
Z = tFK (1+t) Flk < 0.
(1-2) The labor input of the agricultural industry for production
decreases.
(1-3) The wage rate goes up.
(1-4) The output of the manufacture (agriculture) industry increases
(decreases).
(1-5) The difference of labor input for lobbying activity between two
industries expands.
(B) The effects of DFI on the agricultural (exporting) sector.
(1-6)The labor input of the agricultural sector for production and the
total labor input of the agricultural industry increases, if the
following condition is satisfied.
N = s'GL (1+s) Glt <0.
(1-7) The labor input of the manufacturing indusstry for production
decreases.
(1-8) The output of the agricultural (manufacturing)industry increase
(decreases).
(1-9) The wage rate goes up.
(1-10) The difference of labor input for lobbying activity between two
industry shrinks.

14
(C) The effects of immigration of labor.
(1-11) The labor input for production of each industry, the rental rates
for capital and land, the wage rate, the difference of labor input for
lobbying activity between two industries does not change.
(1-12) The labor input for lobbying of each industry increases equally.
The condition in (1-1) means that the effect of DFI in the
manufacturing sector on the marginal value product of the productive
labor ( (1 + t) px Fllc ) exceeds that on the marginal value product of the
lobbying activity ( t'pxFk ). In other words, DFI favors the productive
labor more than the lobbying labor in the manufacture industry. It is
natural that this condition is sufficient for the increase in Lx. This
argument applies to the condition in (1-6). The result (1-1) implies that
the total labor input of agriculture decreases, since the total labor
supply is fixed. The comparative statics' properties in the usual
specific factor model does not change in this framework. See Jones(1971).
The result (1-10) implies that the tariff rate on the importing sector
decreases as the result of DFI in the exporting sector. The reduction in
the tariff rate in the second period encourages DFI in the first period,
that is, quid pro quo DFI. The result (1-5) means that the endogenous
tariff rate on the importing industry and the tax rate on the exporting
industry increases when DFI occurs in the importing sector increases.
The DFI expands the distortion of the relative price (1+t) / (1+s)px/py.
Since the importing industry is more protected as a result of the DFI and
foreign export becomes difficult, I need another factor in order to raise
the quid pro quo DFI. Both (1-11) and (1-11) imply that the immigration
of labor only changes the labor input for lobbying activity. Since both
the tariff and tax rates do not change, every variable affecting national
welfare, such as production, export, import, and prices, remain constant

15
after the change in the immigration of labor. An increment of labor is
absorbed in the rent-seeking lobbying sector. Although the lobbying
sector (Lm + La) expands, its impacts on t and s are identical.
2.5. Optimal Direct Foreign Investment
I characterize
the optimal
DFI for
the
foreign
government.
In the second period, the
government
solves
the
problem
(19). The
necessary conditions for a maximum are
w2x/w2y = -h2x.
(26)
w2x/w2y = o2x.
(27)
The optimal policy requires that the consumer price (W2x/W2y) the producer
price (~H2X), and the trade price ( 02x) be the same. The impacts of DFI
on the second period's utility are
clW2
dKf
W\
-Hi
of O2
dL
' dKf
Or
dr
dKf
(28)
dW2
dTf
Wi
-Hi
O/
- of
8Ln
dTf
+ O i
dr
dKf
The optimal conditions for the whole problem are:
WVW1,, = -H1X,
(29)
(30)

16
WVW1, = -v|/x
(31)
dV/dkf = dWx/dK£ + pdW2/dKf
= dWxy (-Hk+\|/k) + pW2y (-H2k +02k 2x5Ln/3Kf + 02t5t/9Kf)
0.
(32)
dV/dTf = dW1/dTf + pdW2/dTf
= dWxy(-Ht+V|/t )+ pW2y(-H2t+ 0.
(33)
Two additional terms emerge in the second period effect because of the
reallocation of labor ( -2i<5Ln/<3Kf -02idLn/3Tf ) and the change in the
tariff rate.
To examine conditions (32) and (33), we assume that the home
country has a utility function U(Cx,Cy) = U0(Cy) + Cx. This quasi-linear
utility function implies that the host country's demand for good Y is
independent of income. With d>(v|/) denoting the output of good Y the
consumption of good Y in the second (first) period and the results of the
comparative statics, I obtain
VF)C < 0, ¥t>0, $> < 0,
0,
0, <1>t > 0.
(34)
Before examining a quid pro quo DFI, I have another route which induces
DFI from the foreign government although it has no incentive in the
myopic context. This is the effect through a change in the labor

17
reallocation between the production and the lobbying activity. If DFI
increases the amount of labor devoted to the production
( dJjn/8Kf < 0, oLn/0Tf < 0 ) DFI increases the exports of the foreign
country through this productive labor expansion (- <&2L5Ln/3Kf > 0,
- 2L3Ln/3Tf > 0 ) and increases the second period utility. This effect
differs from the original quid pro quo effect the first period DFI
reduces the second period protection. But the economic implication of
this "productive labor expansion effect" is the same as the original one.
We state:
( Proposition 2 )
If the "productive labor expansion" effect ( <3?2idLn/3Kf,
0213Ln/3Tf ) is sufficiently large, the foreign government has a incentive
for DFI.
2.6. Quid Pro Quo Direct Foreign Investment
This section examines the possibility of quid pro quo DFI in
the importing and exporting sector. Consider DFI in the exporting sector.
Since the DFI reduces the tariff rate in the importing sector, it
promotes the export of the home country (Otdt/dTf > 0) This makes a gain
in the second period [See (32)]. Even though the DFI causes a loss in the
first period, the home government makes DFI to obtain this gain.
( Proposition 3 )
The foreign government has an incentive to engage in quid pro quo
DFI in the exporting sector.

18
The above proposition is novel. Bhagwati et al. (1987) analyze
the HOS model. They assume a single capital and can not examine in which
sector a quid pro quo DFI appears. Dinopoulos (1989) (1992), and Zhao
(1990) examine quid pro quo DFI in the importing sector in the partial
equilibrium framework.
The impact of DFI in the importing sector on the first period's
utility dwVdKf is negative. When the foreign government is myopic, it
has no incentive to invest in the host country. Since more DFI causes
more protection (dt/dKf>0) and higher tariff rates reduce the export of
the foreign country (3>T < 0), I conclude that the impact of DFI in the
importing sector through the lobbying activity is negative. Like
Dinopoulos (1992) and Zhao (1991), no quid pro quo DFI occurs.
I seek another political process which causes quid pro quo DFI in
the original sense. Bhagwati (1985) states:
"Such "image building" can influence Congress to withstand the
protectionist pressures from the import-competing industry."
Following the above statement, I introduce this effect into the tariff
function in the importing sector. The function is now reformulated as
t = t(Lm L Kf ) (35)
The tariff/tax rate depends not only on the difference of the labor input
for lobbying between industries but also on the level of DFI, itself. I
assume
ti > 0, tU <0, t-2 < 0, t22 > Of ti2 < 0,
(36)
where tj = 3t/5( Ln, L*), t2 =5t/5Kf and so on.

19
The higher the level of DFI, the lower the tariff rate, but its marginal
effect decreases as DFI increases. The marginal effect of each factor
decreases as its rival factor increases.
The effect of DFI on the tariff rate is
dt/dKf = txd(Lm-La) /dKf + t2. (37)
If DFI decreases Lm La, the tariff rate decreases as the result of DFI.
Simple manipulation leads to
d 1
= + +S> PyGL}y-[-(1 + t)PxFLL{tl2PxF + tipxFk}
+hPxFJhPxFL +(i + t)PxFuc}J (38)
where D' is the coefficient matrix of the reformulated system and det D'
< 0. If both ti2pxF + tipFK and t2pxFL+(1+t) pxFLK are negative, I get d(Lm -
La) / dKf < 0. The former inequality means that the impact of DFI on the
marginal value of Lm, txPxF, through the political process (ti2pxF) exceeds
that through the direct productive process (tipxFk) while the latter
inequality implies that the same property holds on the impact with
respect to the marginal value of Lx. With these inequalities, the
endogenous tariff rate decreases as DFI increases. DFI improves the
second period welfare by reducing the tariff rate (0Tdt/dKf > 0) The
foreign government uses DFI the future benefit even though it generates a
first period welfare loss.

20
I summarize the argument as follows:
( Proposition 4 )
Suppose that the tariff function is t = t (Lm La, Kf) satisfying (39).
The foreign government obtains an incentive to engage in quid pro quo DFI
in the importing sector if the following inequalities hold
(a) t12pxF + tiPxFk < 0.
(b) t2PxFi + (l+t)pxFlk < 0.
2.7. Concluding Remarks
The present chapter has formulated a specific-factor, two-period,
two-country general equilibrium model with an endogenous tariff
determined by lobbying activity of each industry. The lobbying activity
uses a productive resource, labor. The wage rate is determined by the
labor market equilibrium, and is not assumed to be fixed. I have examined
three problems: (1) the impacts of DFI in the exporting and importing
sector on the second period equilibrium, (2) which DFI reduces the
protection encourages quid pro quo DFI, (3) the possibility of new type
of quid pro quo DFI.

CHAPTER 3
LEARNING-BY-DOING AND THE PATTERN OF INTRAINDUSTRY TRADE
3.1. Introduction
Traditional static models of intraindustry trade based on
monopolistic competition and static scale economies cannot determine the
pattern of intraindustry trade (Krugman(1979), Lancaster(1979), Helpman
and Krugman(1985). Bhagwati(1982) has proposed a novel explanation of
intraindustry trade patterns under the label of biological model of
trade in similar products. Countries with identical technological
capabilities (i.e. Japan and the U.S.) could develop different products
in autarky because of differential tastes. Learning-by-doing fixes the
comparative advantage of each country within each industry, and when
trade starts these countries exchange similar products developed in
autarky. The term biological captures the feature that technology
plays, which is analogous to genotype in biology, whereas each product
develop in autarky is like the phenotype in biology, with differential
national tastes corresponding to the environment which shapes the type
of similar products a country cab produce.
Feenstra(1982) and Dinopoulos(1988) have developed models of
intraindustry trade based on differential national tastes. However,
these models do not examine the role of learning-by-doing in determining
the initial comparative advantage of each country. The present paper
complements the above mentioned studies by developing a simple dynamic
general equilibrium model of intraindustry trade based on learning-by-
21

22
doing. The model with building blocks for the context Krugman(1987) who
examined the role of learning-by-doing in industrial targeting.
In the present model, two types of consumers in each of the two
countries are considered. Each type has preferences over a particular
set of similar products. The distribution of the two types of consumers
differs across the two countries. Current productivity of labor, which
is the only factor of production, increases over time because of
learning-by-doing considerations.
The dynamic nature of the model allows us to determine the
duration of autarky based on transport costs. In the absence of
transport costs (or any barriers to trade), the location of production
is indeterminate. Transportation costs provide a realistic and natural
way to generate an initial period without trade, which is necessary to
allow the interaction between national differential tastes and learning-
by-doing. This research investigates the effects of domestic income
transfers across different types of consumers, tariffs and import quotas
on the pattern and volume of trade.
The rest of the chapter is organized as follows: Section 3.2
develops the model. Section 3.3 analyzes the autarkic equilibrium.
Section 3.4 examines how the pattern of intraindustry trade evolves
over time. Section 3.5 investigates the impact of income transfers,
tariffs and import quotas on the pattern and volume of trade. Section
3.6 concludes.

23
3.2. The Model
There are two types of consumers in the home country and the
foreign country. Each type (type 0 and type 1) has the CES utility
function:
_l_
i=n p
type 0. (*,*) = (2>f)
M
=m 1
type i. u(yltym) = ,y,
i=i
where X^y,) is consumption of the i-th product by type 0(1) consumer
and 0 < p < 1. Type 0 individual consumes only differentiated product
x, while type 1 consumers only differentiated product y. The number of
the goods each consumer purchases, n and m, respectively, are fixed.
There is one unit of labor (consumer) in each country. We
consider the distribution of consumers as mirror images of each other.
That is, the home country has s(l-s) fraction of the type 0(1) consumer,
while the foreign country has l-s(s) fraction of the type 0(1) consumer.
We assume that s > 0.5 which means that the home country has a
stronger national demand for the type 0 goods than type 1 goods and the
reverse pattern of preference prevails in the foreign country except for
s=0.5. Both countries have the same tastes when s=0.5.
The only factor of production is labor. At the initial stage,
each country has the same technology exhibiting constant returns to the
production of each good.

24
Xi = k (0) It, y-j = k (0) lj,
(1)
where 1 (1^) is labor developed to production of the Xj.(yj)and k(0)> 0, is
the production coefficient at the initial stage.
I assume that there is learning-by-doing in each industry. At
time t, the production technology of each industry is given by
(2)
Xi(t) = ki(t)li(t), yj(t) = kj (t)lj(t),
where k (t) [ kj(t)] is the production coefficient of goods, Xi[yj].
The coefficient ki is embodied in the learning-by-doing effect. Each
firm can learn from its experience and improve its technology. The
accumulated production of each firm represent its experience.
There is no spillover regarding learning effects across firms or
countries. I define the learning-by-doing effect as
(3)
where 0 < e < 1. This learning effect is assumed to occur at the
industry level and to work as external economy for each firm. Perfect
competition is compatible with this phenomenon.

25
3.3. Trade at The Initial Eauilibrium
First, I derive the initial equilibrium of each country. Since
each country has identical production technology and each good appears
symmetrically, the initial equilibrium is given by
The home country:
Pi = w/k (0) Xi = sk(0)/n y-j = (l-s)k(0)/m ,
li = s/n, 1-j = (l-s)/m. (4)
where pA is the price of the i-th good, w is the wage rate, xi ( yj )is
the output of the i(j)-th good belonging to type 0(1), and li ( lj ) is
the labor employed in the i(j)-th sector of the type 0(1) goods.
The foreign country:
P*i = w*/k(0), x*i = (l-s)k(0)/n y*3 = sk(0)/m ,
l*i = (l-a)/n, l*j = (1-s) /m. (5)
Reflecting the pattern of taste of each country, the home country
produces more quantity of type 0 goods than the type 1 goods, while the
foreign country produces more quantity of type 1 goods than the type 0
goods.
I proceed to examine the pattern of trade by assuming iceberg
type transportation costs: A fraction of g (<1) of any good shipped to
the other country actually arrives. A fraction of 1-g of any goods
evaporates during transportation. Suppose that the home (foreign)
country exports the i(j)-th goods, the following inequalities must hold.

26
Pi/g < p* p*j/g < Pj. (6)
Combing these two inequalities and using (5), we have the inequality
g2 >1. This contradicts the assumption g < 1. There occurs no trade at
the initial stage. This result is natural, since this economy is
Ricardian and both countries have the same technology at the initial
stage. Note that no trade occurs even if there are no transportation
costs.
3.4. The Dynamic Behavior of The Economy
In this section, I investigate the dynamic pattern of trade.
Since no-trade occurs at the initial stage, a no trade period lasts.
During this period, I can divide all goods into type 0 and type 1, since
the initial equilibrium is symmetric. Within each group, prices,
outputs, and labor inputs should be identical.
The equilibrium at time t is
Home country:
p0(t)k0(t) = w(t), p! (t) ki (t) = w(t),
x(t) = sk0(t)/n y (t) = (l-s)ki(t)/m (7)
where the subscript 0(1) denotes type 0(1).
Foreign country:
p*o(t)k*0(t) = w*(t), p*x(t)k*i(t) = w*(t)
x*(t) = (1-s) k*0 (t)/n y*(t) = sk*i(t)/m.
(8)

27
Since the home country produces more output of type 0 goods than the
foreign country, we have
ko(t) > k*0(t), x(t) > x*(t), ki(t) < k\(t) y (t) < y*(t). (9)
As time moves on, the learning-by-doing effect changes the production
coefficients in both countries. The initial difference in outputs
between countries causes the change in the production coefficients,
which results in the difference in outputs in the following period. The
home country develops a comparative advantage in type 0 production,
while the foreign country develops a comparative advantage in type 1
goods through this process.
I expect to be a switching point T, the time when dynamic
comparative advantage overcomes the barrier of the transportation cost
and trade starts. The home country exports type 0 goods and the foreign
country exports type 1 goods. If trade starts at time T, both countries
can export thus, the following two inequalities must hold:
Po(T)/g < p*o (T) p*i(T)/g < Pi(T). (10)
Using (9), we rewrite (10) as
ki(T)/k0(T) < g2k\(T)/k*0(T) (11)
Inequality (11) means that net of transportation costs, the
home country has comparative advantage in type 0 goods, while the
foreign country has a comparative advantage of type 1 goods. I define
the function H(t) to analyze the pattern and the timing of trade.

28
H (t) = k1(t)/k0(t) < g2k\(t)/k*0(t) .
If H(t) takes negative value at some t, trade occurs. At t=0,
H(0)=l-g > 0. Thus, trade does not occur at the initial equilibrium.
H(t) is continuous in t. This implies that a no trade period prevails in
some range of time near time 0.
Since the shape of H(t) is ambiguous, two cases are possible:
(1) there is a time t with H(t) = 0. (2)H(t) is positive for all t. Case
(2) means that no trade occurs forever. Therefore, we focus on the case
(1). There may be multiple time intervals such that H(t)=0. Consider the
first time t=T with H(T)=0 and assume the regularity condition dH(t)/dt
< 0 at t=T. Take the time t+h, where h is a sufficiently small positive
number. From the regulatory condition, H(t+h) is negative, which means
the home (foreign) country has comparative advantage in the type 0(1)
goods. The home country exports type 0 goods, while the foreign country
exports the type 1 goods. We assume that complete specialization occurs.
The type 0 goods are produced only in the home country, while the type 1
goods are produced only in the foreign country. Once specialization
prevails, the learning -by-doing effects appears in neither type 1 goods
of the home country nor the type 0 goods of the foreign country. So the
coefficient of these goods do not change. On the other hand, the
coefficient of exported goods increases as time passes. H(t) remains
negative after the time t+h. This shows that the time T is the switching
point from the no trade period to the trading periods. This switching
takes place just one time and there is no reversal of the trade pattern.
Proposition 1
There is a unique switching point, T. Before T there is no trade.
After T, there is trade based on taste difference across countries.

29
The difference in tastes in each country generates a difference in
technologies. In our Ricardian economy, technological differences
determine the trade pattern. Each country exports goods with larger
domestic demand. Consider the case of s=0.5. In this case, both
countries have the identical tastes. At the initial stage, both
countries produce the same amount of all goods. Since both countries
have the same production coefficients, trade does not occurs at any
time. This is the subtle situation, because even if s deviates from 0.5
by a very small amount, trade can occur eventually.
To describe the trading equilibrium, we normalize the home wage to
1.
Pok0 1,
(12)
p\k*i = 1,
(13)
l=g+ (l-s)w*.
(14)
Equation (12) is the zero profit condition of the home countrys
firm and (13) shows the zero profit condition of the foreign firm.
Equation (14) is the market clearing condition for type 0 goods.
Equations (12), (13) and (14) determine three variables p, p* and w*.
Equation (14) implies that w = 1, i.e. the same wage rate prevails in
both countries. This property comes from the assumption that the
distribution of preferences across the two countries are the mirror
images of each other.
I examine the effects of the transportation cost and the learning-
by-doing parameter on the switching point. Differentiating the
equilibrium condition H(T) = 0 with respect to the transportation costs
and assuming the regularity condition H(T) < 0, we obtain
dT/dg = 2gk*1 (T)/H' (T) k*0 (T) < 0.
(15)

30
The inequality (15) means that the length of the no trade period
increase as the transportation cost increases.
3.5. The Effects of Government Policies
In this section, I examine two types of government policy,
income tax (transfer) and trade policy such as a tariff and a quota.
Without loss of generality, we focus on of home-country policies.
First, I analyze the impact of income tax (transfer).
We consider the income tax and transfer policy at the initial stage. The
government imposes wz income tax on type 0 consumers and gives this tax
revenue to type 1 consumers. Since we assume that the labor supply is
fixed, income tax is equivalent with the pure transfer program. Once
this policy is introduced, the outputs of the both goods are
type 0 goods: (s-z)/n, type 1 goods: (l-s+z)/m. (16)
If the output of the type 1 goods produced at the home country exceeds
that of the foreign country, the inequality
l-s + z>s (17)
must hold. This inequality implies that z>2s-l. This also means that the
home country produces less the type 0 goods than the foreign country
does.
At the initial stage, trade does not occur. I assume that the
government implements this tax-transfer policy for some time interval.

31
During the no trade period, the home country is developing comparative
advantage in type 1 goods over the foreign country. After some time
passes, trade occurs but the trade pattern is completely reversed. After
trade starts, the government does not need to adopt this policy because
the pattern of production is fixed. I have
Proposition 2
The government of the home country can completely reverse the
existing pattern of trade by taxing the type 0 consumer by wz (z>2s-l)
and transferring this to the type 1 consumer during the no trade period.
This income transfer program is useful to reverse the pattern of
trade as long as it is implemented during the no trade period. However,
once trade occurs and the home country imports type 1 goods, this
policy becomes useless, since the transfer the type 1 consumer receives
is spent on imported type 1 goods. This transfer only contributes to the
improvement in the technology of the foreign country.
Next, I investigate the impact of trade policy. Consider a
tariff imposed on type 1 goods by the home country. This tariff is
applied at the same rate, v, to all the imports. For simplicity, the
government introduces the tariff at the initial stage. This tariff is
irreverent during the no trade period. We assume that the tariff revenue
is distributed equally to all consumers. Since this tariff revenue
transfer is neutral to the income distribution among consumers, the
patter of production is not affected. The manner in which the tariff
revenue is distributed is irrelevant during the no trade period.
If trade takes place at time T, the following inequalities must
hold.
p0(T)/g < p*0(T), (1+v) p\ (T) /g < pi(T).
(18)

32
Following the previous argument, I define the function N(t) as
N(t) = (1+v) ki(t) /k0(t) < g2k*! (t)/k'0 (t) (19)
The time N(t)=0 gives the switching point from autarky to the trade
period. At the initial stage, N(t) = 1 + v -g2 >0. We also assume the
corresponding regularity condition
N'(T) < 0 if N(T) = 0.
Note that during autarky, the outputs of all goods and all production
coefficients are not affected by the tariff. The impact on the switching
time is given by
dT 1 kx(T)
= > 0. (20)
dv N'(T) k0(T)
The introduction of a tariff delays the switching point.
Since autarky lasts longer in the tariff-ridden economy than in
free trade, the volume of trade is larger in the tariff-ridden economy,
than in free trade economy at the switching point. After trade occurs,
the pattern of trade becomes the same as that of the free trade. The
tariff protects the type 1 industry in the sense that it delays the
introduction of foreign competition.
I proceed to investigate a quota. The government sets the
uniform import quota of type 1 goods at the level of h. During the no
trade period, the quota does not affect the pattern of production. The
switching point does not depend on the level of the quota, since the
pattern of trade is determined by the relative price of the both goods.

33
The level of quota does not affect this relative price, while the tariff
does affect it. This means that the switching point does not change by
the introduction of the quota. At the switching point T, the volume :f
the imports of type 1 goods equals the min[yf,, h] where yf is the
import under free trade at T. In each case, there is the time which the
quota is binding. The foreign production of the type 1 goods increases
over time due to the learning-by-doing effect, the imports will reach
the restriction at some _ime even though h is set at any high level.
Assume that h The home country produces the type 1 goods by yf-h. If the government
auctions the quota, it can earn the monopoly rent. We assume that this
rent is redistributed equally to the consumers. The home country can
produce and reduces the production by h.
Let assume that h>yf at the switching point T. Since the quota is
not binding, the home country imports type 1 goods and does not produce
it. As time passes, the imports increase and reach the limit h at some
time Tx. After this time the domestic industry can survive and continue
to produce. In this case, the domestic production of the type 1 goods is
discontinued from T to Tx.
I summarize the arguments as follows;
Proposition 3
The tariff delays the time when the trade takes place, but
domestic production of the protected industry vanishes after this time.
On the other hand, the quota does not affect the above time but protects
domestic production after the introduction of trade.
Notice that the qualitative results do nor depend on the assumption that
the tariff revenue and the rent from the quota auction are equally
redistribution to the consumers. For example, if all the tariff revenue

34
is given to the type 1 consumers, the revenue goes to the foreign
industry. Domestic production of type 1 goods can not be revived in the
future. The manner of redistribution changes the level of the outputs
but does not affect the pattern of trade.
3.6. Concluding Remarks
The chapter has developed a two-country dynamic Ricardian model of
trade in similar products along the lines suggested by Bhagwati(1982).
The model was based on differential national tastes, learning-by-doing
and transportation costs. We analyzed the role of these three features
in determining the pattern of intraindustry trade. A country exports
those similar products which correspond to a larger domestic market. An
income transfer policy can reverse the initial pattern of trade in
similar products. A tariff or an import quota delays the time when trade
starts and reduces the volume of trade. The simple structure of the
model allows several possible extensions and generalizations.

CHAPTER 4
TRADE LIBERALIZATION IN AN OPEN ECONOMY WITH A UNIONIZED SECTOR
4.1. Introduction
During the last two decades, the world economy has experienced
three distinct and parallel structural changes. First, the process of
trade liberalization has intensified through multilateral and regional
trade agreements ( i.e. NAFTA, Europe 1992,and the Uruguay round).
Second, there has been an acceleration of technological progress
( e.g. see Berman, Bound and Griliches (1994) ). And third, there has
been a rise in wage income inequality due to a decline in the demand for
less skilled workers ( i.e. see Richardson (1995) for a survey ).
Several economists have voiced their concerns regarding the
possible causal relationship between trade liberalization and the decline
in the demand for less skilled workers. ( e.g. Wood(1995) and Borjas and
Ramey (1994) among others ). Other economists have focused on the effects
of exogenous technological progress on the rise of wage income
inequality. ( e.g. Richardson(1995), Berman et al. (1994), and
Bhagwati (1995) among others ). At the core of the debate on where
technology or trade are responsible for the decline in the demand for
less skilled workers is the Stolper and Samuelson (1941) mechanism that
relates factor prices to commodity prices. According to this mechanism, a
reduction in the domestic price of the importable commodity reduces the
wage of the factor is used intensively in the production of the
importable. An implication of this mechanism is that trade liberalization
35

36
that causes a decline in the domestic price of import competing
industries that use less skilled workers intensively will eventually
reduce the relative demand and the wage of these workers.
The present chapter analyzes the validity of the Stolper and
Samuelson (1941) mechanism in economies that are characterized by
imperfectly competitive product and factor markets. For this purpose,
I construct a general equilibrium model of a small open economy with the
following features: The export producing sector is perfectly competitive
and utilizes two factors of production, skilled and unskilled labor. The
import-competing sector consists of a unionized firm that is protected
from foreign competition through an import quota. The domestic monopolist
utilizes both skilled and unskilled labor under a constant returns to
scale technology. However, unlike the labor demand for skilled workers
which is assumed to be competitive, we assume that the unskilled workers
in the import-competing sector have formed a labor union. The union
bargains with the domestic monopolist over the negotiated wage and the
employment of unskilled labor. The bargaining process is modeled as an
efficient Nash bargaining game that results in simultaneous determination
of employment and the wage.
The model is used to analyze the effects of trade liberalization (
i.e. an increase in the quota ) and immigration on income distribution
and domestic prices. I identify conditions on the parameters of the model
that establish the robustness of the Stolper-Samuelson mechanism in the
presence of sector-specific labor union. I also characterize the
sufficient conditions for the existence of a unique equilibrium.
The model combines two standards of literature that have dealt
with similar issues. Batra(1972), Melvin and Warne(1973) and Ikeda(1990)
have analyzed the general equilibrium behavior of a domestic monopolist
assuming perfect competition in factor markets. Brander and Spencer(1988)
and Mezzetti and Dinopoulos (1991) have developed partial-equilibrium

37
models to examine the effects of trade liberalization in unionized labor
demand. In addition, Dinopoulos and Lane(1992), (1997) and Lane and
Dinopoulos(1995) have constructed general equilibrium models with sector-
specific factors ( instead of allowing factor mobility among all
production factors ) to analyze the effects of trade liberalization on
state-owned firms that are modeled as labor unions.
The rest of this chapter is organized as follows: I formulate
our model in section 4.2. In section 4.3, uniqueness of equilibrium is
proven following Batra(1972). The impacts of bargaining parameters are
examined in section 4.4. In section 4.5, I investigate the effects of
trade liberalization on factor rewards and whether Stolper-Samuelson
mechanism works as a result. I also derive the impact of immigration on
factor rewards. The final section is devoted to concluding remarks.
4.2. The Model
Consider a small open economy with two sectors. A unionized
sector (X) which the wage and employment are determined through Nash
bargaining, and a competitive sector (Y). I designate the production
function by X = F(LX, Hx ) and Y = G(Ly, Hy ) where Li and Hi denote
unskilled labor and skilled labor employed in the i-th sector with i
= x or y. The production functions exhibit constant to scale and can
be written in the intensive form,
x = X/Hx =
f dx) ,
(1)
y = Y/Hy =
g (ly) ,
(2)

38
where lx(ly) shows the unskilled-skilled labor intensity in the
unionized(competitive) sector. I assume that the representative consumer
is characterized by a homothetic utility function which enables us to
write the inverse demand function as:
p = u(z) U'(z)>0, (3)
where z=Yc/Xc, the relative demand, and subscript c denotes consumption.
We normalize the price of good y to the unity and there p = px is the
price of good X. Moreover, for expositional simplicity, the price
elasticity of demand for X denoted by s and assumed to be constant
and larger than one. If the utility function is C.E.S. type, the demand
elasticity satisfies the above assumptions. [ Batra (1971),Melvin and
Warne (1973), Konishi, Okuno- Fujiwara, and Suzumura (1990) adopt this
assumption. ]
I next describe the bargaining process and suppose that only
the unskilled labor in sector x can establish a labor union and
negotiate with the management about level of their employment(Lx) and
wage(rx). Following Mezzetti and Dinopoulos (1991), I assume that output
x is produced by a domestic monopolist. Skilled labor and unskilled
labor working in the competitive sector can not form unions and are
forced to accept the prevailing competitive wages, w and ry. The
behavior of the labor union is modeled through a utility function that
depends on employment and the excess wage(rx ry) ,
x ry)0Lrx
U = (r:
(4)

39
Y (8) represents the elasticity of U with respect to excess
wage(employment). Following the literature, we assume that the union is
employment-oriented, that is, y > 0.
The management of the firm is interested in maximizing its
profit, nx
n* = pX rxLx wHx.
(5)
Both parties engage in efficient Nash bargaining. The threat point is a
pair of -wLx and 0. This pair is obtained under the assumption that the
union strikes ( that is, x = 0 ) if there is disagreement during the
bargaining process. The generalized Nash bargaining product, G thus is
constructed as
W = (pX-rxLx) -Ua (6)
where parameter a is the bargaining power of the union and 0 < a < 1.
I assume that under free trade, this economy imports the good
of the unionized sector and exports the good of the competitive sector.
The government imposes an import quota Q as trade restriction measure.
Domestic consumption of the two goods is Xc = F(LX,HX) + Q and Yc =
G(Ly,Hy) Q. I assume that the consumption of the goods Y is positive
under the quota, that is, G Q > 0.

40
4.3. Uniqueness of Equilibrium
In this section, I focus our attention to issues related to
the uniqueness of momentary equilibrium relative to a set of bargaining
parameters and the level of quota. Uniqueness of equilibrium is a
precondition for the following comparative statics analysis.
The competitive firm hires unskilled and skilled labor to
maximize its profit, given the output price(normalized to one) and the
factor prices. The optimization conditions for the firm are
ry = g' (ly) (7)
w = g(1 y) lyg' (ly) (8)
The above conditions give the relation between ly and the relative
factor price in the competitive sector coy = ry/w, as ly= ly (coy) .
Differentiating ly with respect to coy, we have
1' y = Sly / SCDy = (g ~ 1 yg ) l^" <0 .
(9)
Unskilled labor employment and the wage in the unionized sector
are determined through a efficient Nash bargaining process. The solution
to this bargaining problem is obtained by maximizing the generalized
Nash product with respect to the negotiated wage and employment. The two
first order conditions of this problem are;

41
Contract curve.
r of'
CO = ~co,, -X-
r-o -v r-e f-ij
Nash bargaining curve
fs f
lj'(e-l) f-lj
(10)
(11)
where cox is the relative factor price in the unionized sector,
rx/w and A. = ya/ (1-a+ay) is a parameter. Combing the Contract curve and
Nash bargaining curve, I obtain the basic relation between coy and lx.
COy =
r
f-hf
xB ,
(12)
where
B = [pef / (e l)lxf'] + 1 (3 >1, (13)
and
8 = a( y 0)/(1 a + ay ). (14)
Since e > 1 and f > lxf', B is larger than one. Comparing the Nash
bargaining curve (11) and (12) and X > p, I find that the unionized
unskilled workers earn a higher wage, rx than do the unskilled workers
in the competitive sector. The skilled labor in the unionized sector is
employed to maximize profits. The first order conditions for profit
maximization is

42
w = p (1 1/s) { f lf' ) (15)
I turn to the factor market equilibrium to derive a relation
between kx and coy. The supply of the both factors, L and H, are fixed.
The full employment conditions are described as:
hx + hy =1, (16)
hxlx +hyly = 1, (17)
where hx(hy) = Hx(Hy)/H and 1 = L/H is the unskilled:skilled labor ratio
of the whole economy. Solving the full employment conditions (16), (17)
with respect to hx ,hy and inserting these two variables into the
definition of relative demand we get
g(lx -l)-(lx -l )q
z- (18)
f(l-ly) + (lx-ly)q
where q = Q/H and ghy q > 0 must hold Equations (9), (12) and (18)
determine z as a function of a>y, given parameters H Q and 1. It is
useful for intuitive purpose to investigate this functional relation.
From (12), lx can be written as lx = lx (a>y) Differentiating (12), we
have
l'x = dlx / dcoy = M(f lxf' ) /ff" <0 (19)
wnere
M =
1
1 ~ P+PE ¡3E y. (f lxf') / ll ff "
(20)

43
and
(21)
Note that 0 < M < 1 since E > 1. The unskilled labor intensity of the
unionized sector decreases as the relative wage in the competitive
sector goes up. Note that (19) relates lx and the relative wage
prevailing in the competitive sector, not in the unionized sector. In
our model, the relative wages differ across the sectors because of the
bargaining in the unionized sector. Taking into account of (19),
differentiation of (18) with respect to ly leads to
dz/dcay = Ni (l-ly) dlx/da>y + N2 (lx-l) dly/dcoy, (22)
where
g(f-lJ' + ¥') + q(g-f + f'nx -ly])
[f(l-ly) + (lx-ly)q]2
and
f(g + g'n-ly]) + q(g[lX-ly]+g-f)
(f[l-ly] + (lX~ly)q)2
(24)
After some algebraic manipulation, I find that Ni and N2 take positive
values.
( Lemma 1 )
If ly > (<) lx then, z' (coy) > (<) 0 .

44
Following Batra(1972), Melvin and Warne(1973), I define the
consumption price pc through the utility maximizing behavior of the
representative consumer.
Pc = u[z (coy) ] .
(25)
Differentiating totally the consumption price function, I get
' = dPc
dcoy
u'(z)z'(COy).
(26)
I can obtain
( Lemma 2 )
Suppose that ly > (<) lx. Then, pc' > (<) 0.
The consumption price function is upward(downward)sloping with respect
to the relative wage in the competitive sector, if the
competitive(unionized) sector is unskilled labor intensive.
I also define the production price pp by the optimization
behavior of each firm, (8) and (13).
E(g-lyg')
'P~ (f-hf)
(27)
Noting that kx = kx(coy) and E takes a constant value, we have
Pp =~= D[B(Mlx ly) + S(1 pJJ,
dcov
(28)

45
where
D-Ex
(g-lygf)
->0.
(29)
(B + lXCOy)( 1 + lyCOyXf ~ Ijg)
(30)
(31)
px is the elasticity of substitution between skilled and unskilled labor
in the unionized sector. If the production function is the Cobb-Douglas
type, this elasticity takes the value one. I can prove
(Lemma 3)
(3-1) Suppose that ly > lx and px> 1.
Then, dpp/dcoy is negative.
(3-2) Suppose that ly < lx and px< 1.
and if Mlx > ly. Then, dpp/dmy is positive.
The assumption that Mlx > ly implies that the ranking of value
intensity is the same as the physical one despite the wage differential
that is induced by unionization. The production price function is
downward(upward) sloping with respect to the relative factor price in
the competitive sector, if the competitive (unionized ) sector is
unskilled labor intensive.
Finally, I will establish the relative factor prices of
unionized labor in the competitive sector is negatively related to the

46
factor intensity in the whole economy. Consider equation (18) which
includes not only the factor intensity, 1, but also the total quantity
of skilled labor, H. To find the relation between coy and 1, we have to
investigate two cases: (1) L is fixed and H changes, (2) L changes and H
is fixed. I examine the first case.
Differentiating (18) with respect to 1, given the fixed coy, we
have
, az [ z ~ a~ [f(i-iy)+q(ix-iy)]2 (32)
We can show that the sign of z1 is the same as that of ly lx.
The relative demand function and the consumption price function shift
downward ( upward ) if lx >(<) ly. The production price function does
not change since it is independent of the total factor supply.
Considering these observations, we get
( Lemma 4 )
dcOy/dl < 0 and dp/dl <(>) 0 if lx >(<) ly.
I prove the same result for the second case. This lemma means that the
relative unskilled wage in the competitive sector declines if the
relative supply of unskilled labor increases, which seems to be
plausible. Batra(1972) proves the lemma in the closed economy with
C.E.S. utility function. Ikeda(1990) extends the Batra's result to a tax
distorted economy. Homma(1977) examines this relation in the economies
with homothetic utility function .
Lemma 4 implies that there exists a monotonic functional relation
between 1 and coy which enables us to assert that our model has a

47
unique equilibrium. Combing Lemmas 1,2,3, and 4, I realize that the
intersection of the consumption price function and the production one
gives the unique equilibrium relative factor price in the competitive
sector. I am now in a position to state the uniqueness property.
( Proposition 1 )
There is a unique equilibrium under the following two cases.
(si) The competitive sector is unskilled labor intensive.
(s2) The unionized sector is unskilled labor intensive and
Mlx > ly where M is given by (20).
Proposition 1 identifies two restrictions that are fairy general and
establish the uniqueness of equilibrium despite of the presence of
sector-specific labor union. All remaining variables are also uniquely
determined based on the relative factor price in the competitive sector.
4.4. The Effects of The Bargaining Parameters
In this section, I investigate the effects of the bargaining
parameters on the relative wages of the both sectors. The bargaining
parameters only appears in (12) through parameter ¡3. I easily prove that
P increases(decreases) if a, y (0 )increases The unskilled labor
intensity of the unionized sector is a function of the relative wage in
the competitive sector and the bargaining parameters: lx lx ( Keeping the relative wage constant and differentiating (12) with respect
to lx and p, we obtain

48
(33)
The lx function shifts upwards as p increases. The relative demand
equation (18) is a function p Differentiating this with respect to lx
and P, we get
z0 = dz/dp = Ni (1-ly) dlx/dp.
(34)
From (34) I have
( Lemma 5 )
Suppose that lx >(<) ly. Then, z0 >(<) 0.
If the unionized (competitive) sector is unskilled labor intensive, the
relative demand goes up (down). Next, I examine the change in the
consumption price function. The change is given by
(35)
We have that pf >(<) 0, if lx >(<) ly. The consumption price function
shifts upwards(downwards) if the unionized(competitive) sector is
unskilled labor intensive.
Differentiating the production price function (27) I derive
dpp El?(g-lvg')lxf"
-m- = ; 5 <0.
(36)

49
i show that the production price function moves downwards as (3
increases. Then, I prove
( Proposition 2)
Suppose that the unionized sector is unskilled labor intensive.
Then, the relative wage in the competitive sector increases if the
bargaining power of the labor union,a, the preference for the
employment,y, increase and the preference for the excess wage, 9
decreases.
If the competitive sector is unskilled labor intensive, the change in
relative wage is ambiguous.
I turn to examine the effects on the relative wage of the
unionized sector. The Nash bargaining curve (11) gives the relation
between the relative wage of the unionized sector and the bargaining
parameters. I get.
d dp
(37)
where
r l- + E-M(f-lxf')2/l2xff"
1 fi+ftE- fiE(f-lxf )2 /12xff" '
Since X is larger than p, F is greater than one. The direction of change
in the relative wage of the unionized sector is identical with that in
the competitive sector. Using (35), I have the impact on the relative
wage gap across sectors, cox coy ,

50
dcoy
(39)
dp
Since F-l is positive the wage gap changes the same as the relative
wage in the competitive sector. I summarize these as follows:
( Proposition 3 )
Suppose that the unionized sector is unskilled labor intensive.
(3-1) The relative wage of the unionized sector increases if the
bargaining power of labor union,a, the preference for employment^,
increase and the preference for wage,0, decreases.
(3-2) The relative wage gap between two sectors expands if the
bargaining power of labor union,a, the preference for employment,y,
increase and the preference for excess wage,0 decreases.
4.5. The Effects of Trade Liberalization and Immigration
In this section, I investigate impacts of measures of trade
liberalization on income distribution among three factors. We also
examine the effects when foreign unskilled or skilled labor immigrate to
the economy.
In the present model, trade liberalization is captured by an
expansion of the existing import quota. Since factor intensity of each
sector does not directly depend on the quota, the relative demand z can

be as z = z(coy:Q). For given coy, I find that relaxation of the quota
decreases the relative demand,
_Q. & Oy-lJ[fO-ly) + gQ*-W
Q H[f(l-ly)-(lx-ly)H/QJ2
(40)
The consumption price function (25) is
Pc = pc[u(z (coy:Q] ) ) .
(41)
The move of this function resulting from quota relaxation is given by
(42)
The consumption price function shifts downward if the government relaxes
the import quota. The production price is determined only by the
optimization behavior of each firm the labor union, and is independent
of the quota level. Therefore, the production price function does not
change. We establish
( Lemma 7 )
Suppose that the unionized( competitive ) sector is unskilled labor
intensive. The quota relaxation decreases( increases ) the relative wage
of unskilled workers in the competitive sector. The price of good X
declines as a result of quota expansion.
Considering (35) and (37), I have the impacts on the relative
wage of unionized sector and the wage gap moves the same as the relative

52
wage in the competitive sector. Differentiating (7) and (8) with respect
to the quota, we get
dry r,t/, d0)y
dQ 8 v dQ '
(43)
dw
dQ
-Kg"K
dWy
~dQ
(44)
Consider the case where the unionized sector is unskilled labor
intensive and the government relaxes its quota. The price of the good X
declines and the relative wage of the unskilled labor in the competitive
sector increases. (43) and (44) state that the wage of the skilled labor
which is not employed intensively in the unionized sector goes up, while
the unskilled labor wage in the competitive sector which is used
intensively in the sector declines. This relation between the output
price change and the factor price change implies that the traditional
Stolper-Samuelson theorem holds in our model despite of unionization. If
the competitive sector is unskilled labor intensive, this theorem also
holds as for these factor prices.
I turn to examine the change of the wage of unionized labor.
The wage is determined through efficient Nash bargaining
rx p[ty / lx +(1-X)f / E],
The impact of the quota is given by
!t*- = L[)f + (i-i)p/E] + p[X (flx~f)+(i-X)f"/E]l'x^-
dQdQ F i2 J dO
(45)
(46)

53
The first term of RHS of (46) is negative. The relative wage of the
competitive sector decreases if lx > ly. The second term is negative
since the brace in the second term and l'x are negative. The increase in
the price of X raises the wage of unionized labor which is intensively
hired in the unionized sector. The Stolper-Samuelson relation holds. We
point out that an interesting case may occur. Suppose that lx < ly. In
this case, the output price of X decreases while the relative wage in
the unionized sector increases. The two terms of RHS of (46) take
opposite signs. I can not derive determinate conclusion. This implies
that the Stolper-Samuelson relation may not hold if the effect through
relative wage change( the second term ) is dominated by the price change
effeet( the first term). I summarize these observations
( Proposition 4 )
( 4-1) Stolper-Samuelson theorem holds with respect to the wages of the
skilled labor and the unskilled labor in the competitive sector.
(4-2) Stolper-Samuelson theorem holds with respect to the wage in the
unionized labor if the unionized sector is unskilled labor intensive.
I turn to investigate the effects of immigration on the income
distribution among three factor prices. Consider that foreign unskilled
labor immigrate to the economy. This immigration increases the total
amount of domestic unskilled labor, L and the unskilled:skilled labor
ratio of the whole economy, 1. From Lemma 4, I can show that this
immigration decreases(increases) the price of good X if the unionized
sector(competitive) sector is unskilled labor intensive and decreases the
relative wage of unskilled labor in the competitive sector independent of
the factor intensity among two sectors.

54
The impacts on the skilled labor's wage and the unskilled
labor's wage employed in the competitive sector are given
dr,, da.,
= g"l'v<0.
dl y dl
(47)
dl
(48)
The unskilled labor's wage in the competitive sector decreases, while the
skilled labor's wage increases. This results does not depend the
intensity ranking of the two sectors. The both wages are determined by
marginal productivity of factors. Since the unskilled labor supply rises
in the economy, the skilled labor becomes less abundant while the
unskilled labor does abundant. These change in the two factor price seem
to be plausible.
The change in the unskilled labor's wage working in the
unionized sector is
dp W + (l- k)f'} + p[A + (1- k)f "]l'x d0}
dl
K
dl
(49)
If the unionized sector is unskilled labor intensive, both the output
price and the relative wage in the competitive sector go down.
Then, the wage of unskilled labor in the unionized sector decreases.
However, if the competitive sector is unskilled labor intensive, the
output price increases and the relative wage decreases. The total change
in the wage in the unionized sector depends on the magnitudes of two
effects. If the effect of output price change dominates that of the
relative wage change, the unskilled wage in the unionized sector
increases. The unionized unskilled labor with monopoly power can earn

55
higher wage than the competitive wage, reflecting its marginal
productivity. The counterintuitive result comes from this bargaining
power. I summarize the observations
( Proposition 5 )
(5-1) Immigration of unskilled labor increases the wage of skilled labor
and decreases the wage of unskilled labor in the competitive sector.
(5-2) Immigration of unskilled labor decreases the wage of unskilled
labor in the unionized sector if the unionized sector is unskilled labor
intensive.
If the skilled labor immigrates to the economy, the unskilled:skilled
labor ratio in the whole economy declines. The effects of unskilled labor
immigration are immediately derived from proposition 5.
4.6. Concluding Remarks
I have formulated a two-by-two open general equilibrium model
with bargaining between unskilled labor and the management of firm. The
two parties negotiate over the wage and employment. The unionized firm
has monopoly power in the product market. The other factor prices,(the
wage of unskilled labor working in the competitive sector and that of
skilled labor,) are determined by perfect competition. The government has
imposed an import quota on the unionized sector.
First, I have examined uniqueness of the equilibrium. In our
bargaining model, the physical factor ranking between two sectors does
not necessarily coincide with the nominal one, because the sector-
specific bargaining generates a discrepancy between unskilled labor's

56
wages across sectors. I have established uniqueness of equilibrium under
the condition that the order of the above two rankings is same. Second,
I have investigated the impact of the bargaining parameters on the
relative factor prices of unionized and competitive sector. If the
unionized sector is unskilled labor intensive, the relative wages of both
sectors increase as the union becomes more employment-oriented and its
bargaining power rises. Third, I have examined the effects of trade
liberalization(quota expansion). I have shown that the Stolper-Samuelson
type relation holds as a result of the trade liberalization. Finally, I
have derived the effects of immigration. Immigration of unskilled labor
favors the skilled labor's wage, while it reduces the unskilled labor's
wage in the competitive sector. If the unionized sector is unskilled
labor intensive, this immigration decreases the wage of the unionized
workers.

CHAPTER 5
CONCLUSIONS
The model in the chapter 2 shows that the level of DFI is
substitutive(complementary) to the lobbying activity and protection when
DFI occurs in the exporting (importing) sector. The DFI in the exporting
sector reduces the level of protection, which encourages the foreign
government to engage quid pro quo DFI. On the other hand, the DFI in the
importing sector leads to a higher tariff rate. However, if DFI increases
the host-country government resistance to lobbying and this effect
overcomes the effect of lobbying activity, the tariff rate is reduced by
DFI and there is scope for quid pro quo DFI. I show another channel by
which DFI influences protection: the amount of labor devoted to
productive activities may increase as the result of DFI. The foreign
country exports more and obtains a higher level of welfare. The foreign
government chooses DFI taking this "productive labor expansion effect"
into account. This effect emerges in the general equilibrium framework of
the model in which the tariff is determined through resource-using
lobbying activities.
Chapter 3 establishes the existence of unique switching point,
T. Before T there is no trade. After T, trade occurs based on taste
difference across countries. The government of the home country can
completely reverse the existing pattern of trade by taxing the type 0
consumer and transferring this to the type 1 consumer during the no
trade period. The tariff delays the time when the trade takes place, but
domestic production of the protected industry vanishes after this time.
57

58
Or. the other hand, the quota does not affect the above time but protects
domestic production after the introduction of trade.
In chapter 4, I prove uniqueness of the equilibrium of the model
under the conditions that the physical factor ranking between two sectors
coincides with the nominal one and the magnitude of the elasticity of
substitution between skilled and unskilled labor in the unionized sector.
Next, I investigate the effects of trade liberalization(quota expansion).
The Stolper-Samuelson type relation holds as a result of the trade
liberalization. Finally, I derive the effects of immigration. Immigration
of unskilled labor favors the skilled labor's wage, while it reduces the
unskilled labor's wage in the competitive sector. If the unionized sector
is unskilled labor intensive, this immigration decreases the wage of the
unionized workers.

LIST OF REFERENCES
Baldwin, R. and P.R. Krugman, 1983, Market access and international
competition: A simulation study of 16K random access memories, in
R. Feenstra, ed. Empirical Research in International Trade,
Cambridge, MIT Press, 308-321.
Batra, R.N, 1972, Monopoly theory in general equilibrium, Journal of
Economic Theory 4, 355-371.
Berman, E, Bounnd. J.and Griliches, Z, 1994, Changes in the demand for
skilled labor with U.S. manufacturing: Evidence from the annual
survey of manufacturers, Quarterly Journal of Economics, 109,367-
397.
Bhagwati, J.N, 1982, Shifting comparative advantage, protectionist
demands, and policy response, in J. Bhagwati, ed. Import
Competition and Response Chicago, University of Chicago
Press,48-65.
Bhagwati, J., 1985, Protectionism: Old wine in new bottles, Journal of
Policy Modelling, 7, 23-34.
Bhagwati, J., 1986, Investing abroad, Esmee Fairbairn Lecture, Mimeo.
Bhagwati, J., 1987, VER's quid pro quo DFI and VIEs: Political-economy-
theoretic analysis, International Economic Journal, 1, 1-14.
Bhagwati, J., 1995, Trade and wages: A malign relationship?, Columbia
University, mineo.
Bhagwati, J., R. Brecher, E. Dinopoulos and T. N. Srinivasan, 1987, Quid
pro quo foreign investment and welfare: A political-economy-
theoretic model, Journal of Development Economics, 14, 127-138.
Bhagwati, J., E. Dinopoulos and K. Wong, 1992, "Quid pro quo foreign
investment, American Economic Review, 82, 245-253.
Bhagwati, J.N. and T.N. Srinivasan, 1983, Lectures on Trade Theory
Bhagwati, J.N. and T.N. Srinivasan, 1971, The theory of wage
differentials: Production response and factor price
equalisation, Journal of International Economics 1, 19-35.
Borjas, G, and V, Ramey, 1994, Time-series evidence on sources of trends
in wage inequality, American Economic Review Paper And
Proceedings, 84, 10-16.
Brander, J.A, and B.J. Spencer, 1983, A reciprocal dumping model of
international trade, Journal of International Economics 15, 313-
321.
59

60
Brander, J.A. and B.J. Spencer, 1988, Unionized oligopoly and
international trade policy, Journal of International Economics
24, 217-234.
Brock, W. and S. Magee, 1978, The economics of special interest
politics: The case of the tariff, American Economic Review, 68,
246-250.
Dinopoulos, E., 1984, The optimal tariff with revenue-seeking: A
contribution to the theory of DUP activities, in D. Collander
(ed.), Neoclassical Political Economy, Cambridge, Mass., Ballinger
Publishing Company, 129-140.
Dinopoulos, E, 1988, A formalization of the biological model of trade
in similar products, Journal of International Economics 25, 95-
110.
Dinopoulos, E., 1989, Quid pro quo foreign investment, Economics and
Politics, 1, 145-160.
Dinopoulos, E., 1992, Quid pro quo foreign investment and VERs: A Nash
bargaining approach, Economics and Politics, 4, 43-60.
Dinopoulos, E. and T. Lane,1992, Market liberalization policies in
a reforming socialist economy, IMF Staff Papers 39, 465-494.
Dinopoulos, E. and T. Lane, 1997, Piecemeal reforms in a transitional
economy, mimeo.
Dinopoulos, E, T.R. Lewis, and D.E.M. Sappington, 1993, Optimal
industrial targeting with unknown learning-by-doing, Forthcoming
in Journal of International Economics ,
Dinopoulos, E., and Kar-yiu, Wong, 1991, Quid pro quo foreign investment
and policy intervention, in K. Koekkoek and C. Mennes (eds.),
International Trade and Global Development, New York,
Routledge,162-190.
Dixit, A.K, 1988, Optimal trade and industrial policy for the U,S.
automobile industry, in R. Feenstra, ed. Empirical Methods for
International Trade, Cambridge, MIT Press, 137-158.
Feenstra, R.C, 1982, Product creation and trade pattern: A theoretical
note on the biological model of trade in similar product, in J.
Bhagwati, ed. Import Competition and Response, Chicago, University
of Chicago Press, 235-244.
Feenstra, R.C, and J. Bhagwati, 1982, Tariff seeking and the efficient
tariff, in J. Bhagwati (ed.), Import Competition and Response,
Chicago, University of Chicago Press, 245-258.
Findlay, R. and S. Wellisz, 1982, Endogenous tariff, the political
economy of trade restrictions and welfare, in J. Bhagwati (ed.),
Import Competition and Response, Chicago, University of Chicago
Press, 223-234.
Freeman, R.B, 1995, Are your wages set in Beijing?, Journal of Economic
Perspectives 9, 15-32.

61
Helpman, E, 1981, International trade in the presence of product
differentiation, economies of scale, and monopolistic competition
A Chamberlinian-Hecksher-Olhin approach, Journal of International
Economics 11, 305-340.
Homma, M. 1977, A comparative static analysis of tax incidence, Journal
of Public Economics 8, 53-65.
Ikeda, H. 1990, Monopoly and tax incidence: A general equilibrium
analysis, Discussion Paper, Tezukayama University, Japan.
Jones, R. W., 1971, Distortions in factor markets and the general
equilibrium models, Journal of Political Economy 79, 557-572.
Konishi, H., Okuno-Fujiwara, M. and K. Suzumura, 1990, Oligopolistic
competition and economic welfare: A general equilibrium analysis
of entry regulation and tax-subsidy, Journal of Public Economics
42, 67-88.
Krugman, P.R, 1979, Increasing returns, monopolistic competition, and
international trade, Journal of International Economics 9, 469-
479.
Krugman, P.R, 1981, Intra industry specialization and the gains from
trade, Journal of Political Economy 89, 959-974.
Krugman, P.R, 1987, A narrow moving band, the Dutch disease, and the
competitive consequences of Mrs. Thather: Notes on trade in the
presence of dynamic scale economies, Journal of International
Economics 27, 41-55.
Lancaster, K, 1980, Intra-industry trade under perfect monopolistic
competition, Journal of International Economics 10, 151-175.
Lane, T.D. and E. Dinopoulos, 1992, Fiscal constraints on market-
oriented reform in a socialist economy, IMF Staff Papers, 39,184-
203.
Mayer, W., 1984, Endogenous tariff formation, American Economic Review,
74,970-985.
McDonald, I.M. and R. Solow, 1981, Wage bargaining and employment,
American Economic Review 71, 896-908.
Melvin, J.R. and R.D. Warne, 1973, Monopoly and the theory of
international trade, Journal of International Economics 3,
117-134.
Mezzetti, C. and E. Dinopoulos, 1991, Domestic unionization and
import competition, Journal of International Economics 31, 79-
100.
Nash, J., 1950, The bargaining problem, Econometrica 28, 155-162.
Richardson, J.D, 1995, Income inequality and trade: How to think, What
to conclude, Journal of Economic Perspectives 9, 33-55.
Santoni, M, 1996, Union-oligopoly sequential bargaining: Trade and
industrial policies, Oxford Economic Papers 48, 640-663.

62
Stolper, W. and P. Samuelson, 1941, Protection and real wages, Review of
Economic Studies 9, 58-73.
Wong, Kar-yiu., 1989, Optimal threat to trade restriction and quid pro
quo foreign investment, Economics and Politics, 1, 277-300.
Wood, A, 1995, How trade hurt unskilled workers, Journal of Economic
Perspective, 9, 57-80.
Zhao, L., 1991, Endogenous protection and direct foreign investment.
University of Florida, mimeo.

BIOGRAPHICAL SKETCH
Makoto Okamura was born in 1956 in Japan. After studying at Keio
University, he obtained a Master of Arts in economics at Osaka
University in 1982. He entered the PH.D program in economics at
University of Florida in 1990.
63

I certify that I
conforms to acceptable
adequate, in scope and
Doctor of Philosophy.
have read this study and that in my opinion it
standards of scholarly presentation and is fully
quality, as a dissertation for the degree of
Elias'" pirgbpeulos, Chair
Professor of Economics
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy. (
,/Zvn^>v>
Richard Romano
Professor of Economics
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Bin Xu
Assistant Professor of Economics
I certify that I hane read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Ja$es Seale
Associate Professor
Resource Economics
of Food and
This dissertation was submitted to the Graduate Faculty of the
Department of Economics in the College of Business Administration and to
the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
May, 1997
Dean, Graduate School



33
The level of quota does not affect this relative price, while the tariff
does affect it. This means that the switching point does not change by
the introduction of the quota. At the switching point T, the volume :f
the imports of type 1 goods equals the min[yf,, h] where yf is the
import under free trade at T. In each case, there is the time which the
quota is binding. The foreign production of the type 1 goods increases
over time due to the learning-by-doing effect, the imports will reach
the restriction at some _ime even though h is set at any high level.
Assume that h The home country produces the type 1 goods by yf-h. If the government
auctions the quota, it can earn the monopoly rent. We assume that this
rent is redistributed equally to the consumers. The home country can
produce and reduces the production by h.
Let assume that h>yf at the switching point T. Since the quota is
not binding, the home country imports type 1 goods and does not produce
it. As time passes, the imports increase and reach the limit h at some
time Tx. After this time the domestic industry can survive and continue
to produce. In this case, the domestic production of the type 1 goods is
discontinued from T to Tx.
I summarize the arguments as follows;
Proposition 3
The tariff delays the time when the trade takes place, but
domestic production of the protected industry vanishes after this time.
On the other hand, the quota does not affect the above time but protects
domestic production after the introduction of trade.
Notice that the qualitative results do nor depend on the assumption that
the tariff revenue and the rent from the quota auction are equally
redistribution to the consumers. For example, if all the tariff revenue


60
Brander, J.A. and B.J. Spencer, 1988, Unionized oligopoly and
international trade policy, Journal of International Economics
24, 217-234.
Brock, W. and S. Magee, 1978, The economics of special interest
politics: The case of the tariff, American Economic Review, 68,
246-250.
Dinopoulos, E., 1984, The optimal tariff with revenue-seeking: A
contribution to the theory of DUP activities, in D. Collander
(ed.), Neoclassical Political Economy, Cambridge, Mass., Ballinger
Publishing Company, 129-140.
Dinopoulos, E, 1988, A formalization of the biological model of trade
in similar products, Journal of International Economics 25, 95-
110.
Dinopoulos, E., 1989, Quid pro quo foreign investment, Economics and
Politics, 1, 145-160.
Dinopoulos, E., 1992, Quid pro quo foreign investment and VERs: A Nash
bargaining approach, Economics and Politics, 4, 43-60.
Dinopoulos, E. and T. Lane,1992, Market liberalization policies in
a reforming socialist economy, IMF Staff Papers 39, 465-494.
Dinopoulos, E. and T. Lane, 1997, Piecemeal reforms in a transitional
economy, mimeo.
Dinopoulos, E, T.R. Lewis, and D.E.M. Sappington, 1993, Optimal
industrial targeting with unknown learning-by-doing, Forthcoming
in Journal of International Economics ,
Dinopoulos, E., and Kar-yiu, Wong, 1991, Quid pro quo foreign investment
and policy intervention, in K. Koekkoek and C. Mennes (eds.),
International Trade and Global Development, New York,
Routledge,162-190.
Dixit, A.K, 1988, Optimal trade and industrial policy for the U,S.
automobile industry, in R. Feenstra, ed. Empirical Methods for
International Trade, Cambridge, MIT Press, 137-158.
Feenstra, R.C, 1982, Product creation and trade pattern: A theoretical
note on the biological model of trade in similar product, in J.
Bhagwati, ed. Import Competition and Response, Chicago, University
of Chicago Press, 235-244.
Feenstra, R.C, and J. Bhagwati, 1982, Tariff seeking and the efficient
tariff, in J. Bhagwati (ed.), Import Competition and Response,
Chicago, University of Chicago Press, 245-258.
Findlay, R. and S. Wellisz, 1982, Endogenous tariff, the political
economy of trade restrictions and welfare, in J. Bhagwati (ed.),
Import Competition and Response, Chicago, University of Chicago
Press, 223-234.
Freeman, R.B, 1995, Are your wages set in Beijing?, Journal of Economic
Perspectives 9, 15-32.


13
( Proposition 1 )
(A) The effects of DFI on the manufacturing (importing) sector.
(1-1) The labor input of the manufacturing industry for production and
the total labor input of the manufacturing industry increase, if the
following condition is satisfied.
Z = tFK (1+t) Flk < 0.
(1-2) The labor input of the agricultural industry for production
decreases.
(1-3) The wage rate goes up.
(1-4) The output of the manufacture (agriculture) industry increases
(decreases).
(1-5) The difference of labor input for lobbying activity between two
industries expands.
(B) The effects of DFI on the agricultural (exporting) sector.
(1-6)The labor input of the agricultural sector for production and the
total labor input of the agricultural industry increases, if the
following condition is satisfied.
N = s'GL (1+s) Glt <0.
(1-7) The labor input of the manufacturing indusstry for production
decreases.
(1-8) The output of the agricultural (manufacturing)industry increase
(decreases).
(1-9) The wage rate goes up.
(1-10) The difference of labor input for lobbying activity between two
industry shrinks.


LIST OF REFERENCES
Baldwin, R. and P.R. Krugman, 1983, Market access and international
competition: A simulation study of 16K random access memories, in
R. Feenstra, ed. Empirical Research in International Trade,
Cambridge, MIT Press, 308-321.
Batra, R.N, 1972, Monopoly theory in general equilibrium, Journal of
Economic Theory 4, 355-371.
Berman, E, Bounnd. J.and Griliches, Z, 1994, Changes in the demand for
skilled labor with U.S. manufacturing: Evidence from the annual
survey of manufacturers, Quarterly Journal of Economics, 109,367-
397.
Bhagwati, J.N, 1982, Shifting comparative advantage, protectionist
demands, and policy response, in J. Bhagwati, ed. Import
Competition and Response Chicago, University of Chicago
Press,48-65.
Bhagwati, J., 1985, Protectionism: Old wine in new bottles, Journal of
Policy Modelling, 7, 23-34.
Bhagwati, J., 1986, Investing abroad, Esmee Fairbairn Lecture, Mimeo.
Bhagwati, J., 1987, VER's quid pro quo DFI and VIEs: Political-economy-
theoretic analysis, International Economic Journal, 1, 1-14.
Bhagwati, J., 1995, Trade and wages: A malign relationship?, Columbia
University, mineo.
Bhagwati, J., R. Brecher, E. Dinopoulos and T. N. Srinivasan, 1987, Quid
pro quo foreign investment and welfare: A political-economy-
theoretic model, Journal of Development Economics, 14, 127-138.
Bhagwati, J., E. Dinopoulos and K. Wong, 1992, "Quid pro quo foreign
investment, American Economic Review, 82, 245-253.
Bhagwati, J.N. and T.N. Srinivasan, 1983, Lectures on Trade Theory
Bhagwati, J.N. and T.N. Srinivasan, 1971, The theory of wage
differentials: Production response and factor price
equalisation, Journal of International Economics 1, 19-35.
Borjas, G, and V, Ramey, 1994, Time-series evidence on sources of trends
in wage inequality, American Economic Review Paper And
Proceedings, 84, 10-16.
Brander, J.A, and B.J. Spencer, 1983, A reciprocal dumping model of
international trade, Journal of International Economics 15, 313-
321.
59


19
The higher the level of DFI, the lower the tariff rate, but its marginal
effect decreases as DFI increases. The marginal effect of each factor
decreases as its rival factor increases.
The effect of DFI on the tariff rate is
dt/dKf = txd(Lm-La) /dKf + t2. (37)
If DFI decreases Lm La, the tariff rate decreases as the result of DFI.
Simple manipulation leads to
d 1
= + +S> PyGL}y-[-(1 + t)PxFLL{tl2PxF + tipxFk}
+hPxFJhPxFL +(i + t)PxFuc}J (38)
where D' is the coefficient matrix of the reformulated system and det D'
< 0. If both ti2pxF + tipFK and t2pxFL+(1+t) pxFLK are negative, I get d(Lm -
La) / dKf < 0. The former inequality means that the impact of DFI on the
marginal value of Lm, txPxF, through the political process (ti2pxF) exceeds
that through the direct productive process (tipxFk) while the latter
inequality implies that the same property holds on the impact with
respect to the marginal value of Lx. With these inequalities, the
endogenous tariff rate decreases as DFI increases. DFI improves the
second period welfare by reducing the tariff rate (0Tdt/dKf > 0) The
foreign government uses DFI the future benefit even though it generates a
first period welfare loss.


Copyright 1997
by
MAKOTO OKAMURA


CHAPTER 3
LEARNING-BY-DOING AND THE PATTERN OF INTRAINDUSTRY TRADE
3.1. Introduction
Traditional static models of intraindustry trade based on
monopolistic competition and static scale economies cannot determine the
pattern of intraindustry trade (Krugman(1979), Lancaster(1979), Helpman
and Krugman(1985). Bhagwati(1982) has proposed a novel explanation of
intraindustry trade patterns under the label of biological model of
trade in similar products. Countries with identical technological
capabilities (i.e. Japan and the U.S.) could develop different products
in autarky because of differential tastes. Learning-by-doing fixes the
comparative advantage of each country within each industry, and when
trade starts these countries exchange similar products developed in
autarky. The term biological captures the feature that technology
plays, which is analogous to genotype in biology, whereas each product
develop in autarky is like the phenotype in biology, with differential
national tastes corresponding to the environment which shapes the type
of similar products a country cab produce.
Feenstra(1982) and Dinopoulos(1988) have developed models of
intraindustry trade based on differential national tastes. However,
these models do not examine the role of learning-by-doing in determining
the initial comparative advantage of each country. The present paper
complements the above mentioned studies by developing a simple dynamic
general equilibrium model of intraindustry trade based on learning-by-
21


I dedicate this dissertation to memory of the late Hiroshi
Okamura, my father and Misuzu Okamura, my mother.


32
Following the previous argument, I define the function N(t) as
N(t) = (1+v) ki(t) /k0(t) < g2k*! (t)/k'0 (t) (19)
The time N(t)=0 gives the switching point from autarky to the trade
period. At the initial stage, N(t) = 1 + v -g2 >0. We also assume the
corresponding regularity condition
N'(T) < 0 if N(T) = 0.
Note that during autarky, the outputs of all goods and all production
coefficients are not affected by the tariff. The impact on the switching
time is given by
dT 1 kx(T)
= > 0. (20)
dv N'(T) k0(T)
The introduction of a tariff delays the switching point.
Since autarky lasts longer in the tariff-ridden economy than in
free trade, the volume of trade is larger in the tariff-ridden economy,
than in free trade economy at the switching point. After trade occurs,
the pattern of trade becomes the same as that of the free trade. The
tariff protects the type 1 industry in the sense that it delays the
introduction of foreign competition.
I proceed to investigate a quota. The government sets the
uniform import quota of type 1 goods at the level of h. During the no
trade period, the quota does not affect the pattern of production. The
switching point does not depend on the level of the quota, since the
pattern of trade is determined by the relative price of the both goods.


30
The inequality (15) means that the length of the no trade period
increase as the transportation cost increases.
3.5. The Effects of Government Policies
In this section, I examine two types of government policy,
income tax (transfer) and trade policy such as a tariff and a quota.
Without loss of generality, we focus on of home-country policies.
First, I analyze the impact of income tax (transfer).
We consider the income tax and transfer policy at the initial stage. The
government imposes wz income tax on type 0 consumers and gives this tax
revenue to type 1 consumers. Since we assume that the labor supply is
fixed, income tax is equivalent with the pure transfer program. Once
this policy is introduced, the outputs of the both goods are
type 0 goods: (s-z)/n, type 1 goods: (l-s+z)/m. (16)
If the output of the type 1 goods produced at the home country exceeds
that of the foreign country, the inequality
l-s + z>s (17)
must hold. This inequality implies that z>2s-l. This also means that the
home country produces less the type 0 goods than the foreign country
does.
At the initial stage, trade does not occur. I assume that the
government implements this tax-transfer policy for some time interval.


CHAPTER 4
TRADE LIBERALIZATION IN AN OPEN ECONOMY WITH A UNIONIZED SECTOR
4.1. Introduction
During the last two decades, the world economy has experienced
three distinct and parallel structural changes. First, the process of
trade liberalization has intensified through multilateral and regional
trade agreements ( i.e. NAFTA, Europe 1992,and the Uruguay round).
Second, there has been an acceleration of technological progress
( e.g. see Berman, Bound and Griliches (1994) ). And third, there has
been a rise in wage income inequality due to a decline in the demand for
less skilled workers ( i.e. see Richardson (1995) for a survey ).
Several economists have voiced their concerns regarding the
possible causal relationship between trade liberalization and the decline
in the demand for less skilled workers. ( e.g. Wood(1995) and Borjas and
Ramey (1994) among others ). Other economists have focused on the effects
of exogenous technological progress on the rise of wage income
inequality. ( e.g. Richardson(1995), Berman et al. (1994), and
Bhagwati (1995) among others ). At the core of the debate on where
technology or trade are responsible for the decline in the demand for
less skilled workers is the Stolper and Samuelson (1941) mechanism that
relates factor prices to commodity prices. According to this mechanism, a
reduction in the domestic price of the importable commodity reduces the
wage of the factor is used intensively in the production of the
importable. An implication of this mechanism is that trade liberalization
35


37
models to examine the effects of trade liberalization in unionized labor
demand. In addition, Dinopoulos and Lane(1992), (1997) and Lane and
Dinopoulos(1995) have constructed general equilibrium models with sector-
specific factors ( instead of allowing factor mobility among all
production factors ) to analyze the effects of trade liberalization on
state-owned firms that are modeled as labor unions.
The rest of this chapter is organized as follows: I formulate
our model in section 4.2. In section 4.3, uniqueness of equilibrium is
proven following Batra(1972). The impacts of bargaining parameters are
examined in section 4.4. In section 4.5, I investigate the effects of
trade liberalization on factor rewards and whether Stolper-Samuelson
mechanism works as a result. I also derive the impact of immigration on
factor rewards. The final section is devoted to concluding remarks.
4.2. The Model
Consider a small open economy with two sectors. A unionized
sector (X) which the wage and employment are determined through Nash
bargaining, and a competitive sector (Y). I designate the production
function by X = F(LX, Hx ) and Y = G(Ly, Hy ) where Li and Hi denote
unskilled labor and skilled labor employed in the i-th sector with i
= x or y. The production functions exhibit constant to scale and can
be written in the intensive form,
x = X/Hx =
f dx) ,
(1)
y = Y/Hy =
g (ly) ,
(2)


45
where
D-Ex
(g-lygf)
->0.
(29)
(B + lXCOy)( 1 + lyCOyXf ~ Ijg)
(30)
(31)
px is the elasticity of substitution between skilled and unskilled labor
in the unionized sector. If the production function is the Cobb-Douglas
type, this elasticity takes the value one. I can prove
(Lemma 3)
(3-1) Suppose that ly > lx and px> 1.
Then, dpp/dcoy is negative.
(3-2) Suppose that ly < lx and px< 1.
and if Mlx > ly. Then, dpp/dmy is positive.
The assumption that Mlx > ly implies that the ranking of value
intensity is the same as the physical one despite the wage differential
that is induced by unionization. The production price function is
downward(upward) sloping with respect to the relative factor price in
the competitive sector, if the competitive (unionized ) sector is
unskilled labor intensive.
Finally, I will establish the relative factor prices of
unionized labor in the competitive sector is negatively related to the


8
Q> H(Q*x,K* Kf,T" Tf), (8)
where Qy(Q*x) is the output of the agricultural (manufactured) good and K*
(T*) is the existing stock of capital (land). The foreign country exports
(imports) the manufactured (agricultural) good. It faces the reciprocal
demand function (offer curve)
M*y = where
M*y : foreign import of good Y,
E*x : foreign export of good X,
L : the amount of the labor devoted to the lobbying,
T : relative tariff rate, (1 + t)/(l + s)
The preferences of the consumers are summarized as the social utility
function, W(C*x,Cy) which depends on the consumption of both goods.
Since perfect competition prevails in every market, each individual
firm has no market power. The foreign firm cannot affect the tariff
level through the DFI. To examine the quid pro quo DFI, we assume that
the foreign government controls the total amount of the DFI. The
government determines the level of DFI to maximize its national welfare.
Thus, there are three players in the game we consider: the foreign
government, the host-country representative manufacturing firm and the
host-country representative agricultural firm.
There are two periods in the model. The timing of the game is as
follows:
Period 1. 1. The foreign government chooses DFI which remain fixed.
2. There exists no lobbying and free trade prevails.
Period 2. 3. Both domestic industries lobby and ~he endogenous
tariff rate is determined.


15
after the change in the immigration of labor. An increment of labor is
absorbed in the rent-seeking lobbying sector. Although the lobbying
sector (Lm + La) expands, its impacts on t and s are identical.
2.5. Optimal Direct Foreign Investment
I characterize
the optimal
DFI for
the
foreign
government.
In the second period, the
government
solves
the
problem
(19). The
necessary conditions for a maximum are
w2x/w2y = -h2x.
(26)
w2x/w2y = o2x.
(27)
The optimal policy requires that the consumer price (W2x/W2y) the producer
price (~H2X), and the trade price ( 02x) be the same. The impacts of DFI
on the second period's utility are
clW2
dKf
W\
-Hi
of O2
dL
' dKf
Or
dr
dKf
(28)
dW2
dTf
Wi
-Hi
O/
- of
8Ln
dTf
+ O i
dr
dKf
The optimal conditions for the whole problem are:
WVW1,, = -H1X,
(29)
(30)


49
i show that the production price function moves downwards as (3
increases. Then, I prove
( Proposition 2)
Suppose that the unionized sector is unskilled labor intensive.
Then, the relative wage in the competitive sector increases if the
bargaining power of the labor union,a, the preference for the
employment,y, increase and the preference for the excess wage, 9
decreases.
If the competitive sector is unskilled labor intensive, the change in
relative wage is ambiguous.
I turn to examine the effects on the relative wage of the
unionized sector. The Nash bargaining curve (11) gives the relation
between the relative wage of the unionized sector and the bargaining
parameters. I get.
d dp
(37)
where
r l- + E-M(f-lxf')2/l2xff"
1 fi+ftE- fiE(f-lxf )2 /12xff" '
Since X is larger than p, F is greater than one. The direction of change
in the relative wage of the unionized sector is identical with that in
the competitive sector. Using (35), I have the impact on the relative
wage gap across sectors, cox coy ,


TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iv
ABSTRACT vi
CHAPTERS
1 INTRODUCTION 1
2 ENDOGENOUS TARIFF FORMATION AND DIRECT FOREIGN INVESTMENT:
GENERALEQUILIBRIUM ANALYSIS 4
2.1 Introduction 4
2.2 The Model 6
2.3 The Equilibrium 9
2.4 The Basic Comparative Statics 12
2.5 Optimal Direct Foreign Investment 15
2.6 Quid Pro Quo Direct Foreign Investment 17
2.7 Concluding Remarks 20
3 LEARNING-BY-DOING AND PATTERN OF INTRAINDUSTRY TRADE 21
3.1 Introduction 21
3.2 The Model 23
3.3 Trade at the initial equilibrium 25
3.4 The Dynamic Behavior of the Economy 26
3.5 The Effects of Government Policies 30
3.6 Concluding Remarks 34
4 TRADE LIBERALIZATION IN AN OPEN ECONOMY WITH A UNIONIZED SECTOR 35
4.1 Introduction 35
4.2 The Model 37
4.3 Uniqueness of Equilibrium 40
4.4 The Effects of Bargaining Parameters 47
4.5 The Effects of Trade Liberalization and Immigration 50
4.6 Concluding Remarks 55
5 CONCLUSIONS 57
LIST OF REFERENCES 59
BIOGRAPHICAL SKETCH 63
v


26
Pi/g < p* p*j/g < Pj. (6)
Combing these two inequalities and using (5), we have the inequality
g2 >1. This contradicts the assumption g < 1. There occurs no trade at
the initial stage. This result is natural, since this economy is
Ricardian and both countries have the same technology at the initial
stage. Note that no trade occurs even if there are no transportation
costs.
3.4. The Dynamic Behavior of The Economy
In this section, I investigate the dynamic pattern of trade.
Since no-trade occurs at the initial stage, a no trade period lasts.
During this period, I can divide all goods into type 0 and type 1, since
the initial equilibrium is symmetric. Within each group, prices,
outputs, and labor inputs should be identical.
The equilibrium at time t is
Home country:
p0(t)k0(t) = w(t), p! (t) ki (t) = w(t),
x(t) = sk0(t)/n y (t) = (l-s)ki(t)/m (7)
where the subscript 0(1) denotes type 0(1).
Foreign country:
p*o(t)k*0(t) = w*(t), p*x(t)k*i(t) = w*(t)
x*(t) = (1-s) k*0 (t)/n y*(t) = sk*i(t)/m.
(8)


31
During the no trade period, the home country is developing comparative
advantage in type 1 goods over the foreign country. After some time
passes, trade occurs but the trade pattern is completely reversed. After
trade starts, the government does not need to adopt this policy because
the pattern of production is fixed. I have
Proposition 2
The government of the home country can completely reverse the
existing pattern of trade by taxing the type 0 consumer by wz (z>2s-l)
and transferring this to the type 1 consumer during the no trade period.
This income transfer program is useful to reverse the pattern of
trade as long as it is implemented during the no trade period. However,
once trade occurs and the home country imports type 1 goods, this
policy becomes useless, since the transfer the type 1 consumer receives
is spent on imported type 1 goods. This transfer only contributes to the
improvement in the technology of the foreign country.
Next, I investigate the impact of trade policy. Consider a
tariff imposed on type 1 goods by the home country. This tariff is
applied at the same rate, v, to all the imports. For simplicity, the
government introduces the tariff at the initial stage. This tariff is
irreverent during the no trade period. We assume that the tariff revenue
is distributed equally to all consumers. Since this tariff revenue
transfer is neutral to the income distribution among consumers, the
patter of production is not affected. The manner in which the tariff
revenue is distributed is irrelevant during the no trade period.
If trade takes place at time T, the following inequalities must
hold.
p0(T)/g < p*0(T), (1+v) p\ (T) /g < pi(T).
(18)


22
doing. The model with building blocks for the context Krugman(1987) who
examined the role of learning-by-doing in industrial targeting.
In the present model, two types of consumers in each of the two
countries are considered. Each type has preferences over a particular
set of similar products. The distribution of the two types of consumers
differs across the two countries. Current productivity of labor, which
is the only factor of production, increases over time because of
learning-by-doing considerations.
The dynamic nature of the model allows us to determine the
duration of autarky based on transport costs. In the absence of
transport costs (or any barriers to trade), the location of production
is indeterminate. Transportation costs provide a realistic and natural
way to generate an initial period without trade, which is necessary to
allow the interaction between national differential tastes and learning-
by-doing. This research investigates the effects of domestic income
transfers across different types of consumers, tariffs and import quotas
on the pattern and volume of trade.
The rest of the chapter is organized as follows: Section 3.2
develops the model. Section 3.3 analyzes the autarkic equilibrium.
Section 3.4 examines how the pattern of intraindustry trade evolves
over time. Section 3.5 investigates the impact of income transfers,
tariffs and import quotas on the pattern and volume of trade. Section
3.6 concludes.


38
where lx(ly) shows the unskilled-skilled labor intensity in the
unionized(competitive) sector. I assume that the representative consumer
is characterized by a homothetic utility function which enables us to
write the inverse demand function as:
p = u(z) U'(z)>0, (3)
where z=Yc/Xc, the relative demand, and subscript c denotes consumption.
We normalize the price of good y to the unity and there p = px is the
price of good X. Moreover, for expositional simplicity, the price
elasticity of demand for X denoted by s and assumed to be constant
and larger than one. If the utility function is C.E.S. type, the demand
elasticity satisfies the above assumptions. [ Batra (1971),Melvin and
Warne (1973), Konishi, Okuno- Fujiwara, and Suzumura (1990) adopt this
assumption. ]
I next describe the bargaining process and suppose that only
the unskilled labor in sector x can establish a labor union and
negotiate with the management about level of their employment(Lx) and
wage(rx). Following Mezzetti and Dinopoulos (1991), I assume that output
x is produced by a domestic monopolist. Skilled labor and unskilled
labor working in the competitive sector can not form unions and are
forced to accept the prevailing competitive wages, w and ry. The
behavior of the labor union is modeled through a utility function that
depends on employment and the excess wage(rx ry) ,
x ry)0Lrx
U = (r:
(4)


5
lobbying. In addition, by focusing on product markets, the first strand
the literature has abstracted from factor mobility influences on the
level of protection. Moreover, since the Quid Pro Quo DFI literature
examines the potential for defusing (or reducing ) existing protection,
it has opposite implications to the resource using lobbying literature
which studies the conditions under which protection emerges. Therefore,
there is some scope for integrating the two approaches in order to
analyze how endogenous protection based on resource using lobbying is
affected by DFI.
Quid Pro Quo DFI corresponds to a novel form of foreign
lobbying which utilizes resources in the present (by incurring a welfare
loss for the foreign country) in order to reduce protection in the host
country in the future (with the associated benefits) Therefore, it is
possible to build a model of endogenous protection with domestic lobbying
which uses resources and foreign lobbying which causes DFI in a quid pro
quo fashion. I construct a two-period, two-country, specific-factor
general equilibrium trade model with direct foreign investment.
Following Findlay and Wellisz (1982), a tariff formation function is
added in the second period. Free trade prevails in the first period.
The two-period, two-country framework is adopted from Bhagwati et al
(1987) .
By adopting the two-sector specific-factor model, I can examine
in which sector, the exporting or the importing, does quid pro quo DFI
take place. This paper shows that the level of DFI is substitutive
(complementary) to the lobbying activity and protection when DFI occurs
in the exporting (importing) sector. This reduces the level of
protection, which encourages the foreign government to engage quid pro
quo DFI. On the other hand, the difference between lobbying activities
expands as the result of DFI in the importing sector, which leads to a
higher tariff rate. However, if DFI increases the host-country government


be as z = z(coy:Q). For given coy, I find that relaxation of the quota
decreases the relative demand,
_Q. & Oy-lJ[fO-ly) + gQ*-W
Q H[f(l-ly)-(lx-ly)H/QJ2
(40)
The consumption price function (25) is
Pc = pc[u(z (coy:Q] ) ) .
(41)
The move of this function resulting from quota relaxation is given by
(42)
The consumption price function shifts downward if the government relaxes
the import quota. The production price is determined only by the
optimization behavior of each firm the labor union, and is independent
of the quota level. Therefore, the production price function does not
change. We establish
( Lemma 7 )
Suppose that the unionized( competitive ) sector is unskilled labor
intensive. The quota relaxation decreases( increases ) the relative wage
of unskilled workers in the competitive sector. The price of good X
declines as a result of quota expansion.
Considering (35) and (37), I have the impacts on the relative
wage of unionized sector and the wage gap moves the same as the relative


3
labor demand for skilled workers which is assumed to be competitive, I
assume that the unskilled workers in the import-competing sector have
formed a labor union. The union bargains with the domestic monopolist
over the negotiated wage and the employment of unskilled labor. The
bargaining process is modeled as an efficient Nash bargaining game that
results in simultaneous determination of employment and the wage. I
analyze the effects of trade liberalization { i.e. an increase in the
quota ) and immigration on income distribution and domestic prices. I
identify conditions on the parameters of the model that establish the
robustness of the Stolper- Samuelson mechanism in the presence of sector-
specific labor union. I also characterize conditions for the existence of
a unique equilibrium.


THREE ESSAYS ON INTERNATIONAL TRADE AND WELFARE
By
MAKOTO OKAMURA
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1997


36
that causes a decline in the domestic price of import competing
industries that use less skilled workers intensively will eventually
reduce the relative demand and the wage of these workers.
The present chapter analyzes the validity of the Stolper and
Samuelson (1941) mechanism in economies that are characterized by
imperfectly competitive product and factor markets. For this purpose,
I construct a general equilibrium model of a small open economy with the
following features: The export producing sector is perfectly competitive
and utilizes two factors of production, skilled and unskilled labor. The
import-competing sector consists of a unionized firm that is protected
from foreign competition through an import quota. The domestic monopolist
utilizes both skilled and unskilled labor under a constant returns to
scale technology. However, unlike the labor demand for skilled workers
which is assumed to be competitive, we assume that the unskilled workers
in the import-competing sector have formed a labor union. The union
bargains with the domestic monopolist over the negotiated wage and the
employment of unskilled labor. The bargaining process is modeled as an
efficient Nash bargaining game that results in simultaneous determination
of employment and the wage.
The model is used to analyze the effects of trade liberalization (
i.e. an increase in the quota ) and immigration on income distribution
and domestic prices. I identify conditions on the parameters of the model
that establish the robustness of the Stolper-Samuelson mechanism in the
presence of sector-specific labor union. I also characterize the
sufficient conditions for the existence of a unique equilibrium.
The model combines two standards of literature that have dealt
with similar issues. Batra(1972), Melvin and Warne(1973) and Ikeda(1990)
have analyzed the general equilibrium behavior of a domestic monopolist
assuming perfect competition in factor markets. Brander and Spencer(1988)
and Mezzetti and Dinopoulos (1991) have developed partial-equilibrium


41
Contract curve.
r of'
CO = ~co,, -X-
r-o -v r-e f-ij
Nash bargaining curve
fs f
lj'(e-l) f-lj
(10)
(11)
where cox is the relative factor price in the unionized sector,
rx/w and A. = ya/ (1-a+ay) is a parameter. Combing the Contract curve and
Nash bargaining curve, I obtain the basic relation between coy and lx.
COy =
r
f-hf
xB ,
(12)
where
B = [pef / (e l)lxf'] + 1 (3 >1, (13)
and
8 = a( y 0)/(1 a + ay ). (14)
Since e > 1 and f > lxf', B is larger than one. Comparing the Nash
bargaining curve (11) and (12) and X > p, I find that the unionized
unskilled workers earn a higher wage, rx than do the unskilled workers
in the competitive sector. The skilled labor in the unionized sector is
employed to maximize profits. The first order conditions for profit
maximization is


9
4. Given this tariff rate, trade occurs.
The foreign government moves as the Stackelberg leader, while the
two host-country firms move as simultaneous Stackelberg followers. The
structure of the model can generate quid pro quo DFI, and it is identical
to the one used by Bhagwati (1987) The foreign government does not
necessarily choose DFI to maximize its national welfare under free trade
(the first period welfare). It could endure a first-period welfare loss
for the purpose of obtaining the larger second period welfare.
2.3. The Equilibrium
This section analyzes the equilibrium of the model. I solve for
the equilibrium in the usual backward fashion. In the second period,
given a level of DFI, and the domestic prices of goods and labor, each
domestic firm chooses its labor inputs for production and lobbying
simultaneously in order to maximize its myopic (one-period) profit. The
profit of the manufacturing firm is
max n* =[l + t(Lm -La)]PxF(K,Lx)-wLx-wLm. (10)
Lx.Ly
where w is the wage rate. I impose the Cournot assumption with respect to
the lobbying behavior, that is, each firm decides its lobbying activity,
taking the rival's lobbying activity as given. The first order conditions
for the above problem are
(1 + t)pxFL w 0,
(11)


52
wage in the competitive sector. Differentiating (7) and (8) with respect
to the quota, we get
dry r,t/, d0)y
dQ 8 v dQ '
(43)
dw
dQ
-Kg"K
dWy
~dQ
(44)
Consider the case where the unionized sector is unskilled labor
intensive and the government relaxes its quota. The price of the good X
declines and the relative wage of the unskilled labor in the competitive
sector increases. (43) and (44) state that the wage of the skilled labor
which is not employed intensively in the unionized sector goes up, while
the unskilled labor wage in the competitive sector which is used
intensively in the sector declines. This relation between the output
price change and the factor price change implies that the traditional
Stolper-Samuelson theorem holds in our model despite of unionization. If
the competitive sector is unskilled labor intensive, this theorem also
holds as for these factor prices.
I turn to examine the change of the wage of unionized labor.
The wage is determined through efficient Nash bargaining
rx p[ty / lx +(1-X)f / E],
The impact of the quota is given by
!t*- = L[)f + (i-i)p/E] + p[X (flx~f)+(i-X)f"/E]l'x^-
dQdQ F i2 J dO
(45)
(46)


50
dcoy
(39)
dp
Since F-l is positive the wage gap changes the same as the relative
wage in the competitive sector. I summarize these as follows:
( Proposition 3 )
Suppose that the unionized sector is unskilled labor intensive.
(3-1) The relative wage of the unionized sector increases if the
bargaining power of labor union,a, the preference for employment^,
increase and the preference for wage,0, decreases.
(3-2) The relative wage gap between two sectors expands if the
bargaining power of labor union,a, the preference for employment,y,
increase and the preference for excess wage,0 decreases.
4.5. The Effects of Trade Liberalization and Immigration
In this section, I investigate impacts of measures of trade
liberalization on income distribution among three factors. We also
examine the effects when foreign unskilled or skilled labor immigrate to
the economy.
In the present model, trade liberalization is captured by an
expansion of the existing import quota. Since factor intensity of each
sector does not directly depend on the quota, the relative demand z can


27
Since the home country produces more output of type 0 goods than the
foreign country, we have
ko(t) > k*0(t), x(t) > x*(t), ki(t) < k\(t) y (t) < y*(t). (9)
As time moves on, the learning-by-doing effect changes the production
coefficients in both countries. The initial difference in outputs
between countries causes the change in the production coefficients,
which results in the difference in outputs in the following period. The
home country develops a comparative advantage in type 0 production,
while the foreign country develops a comparative advantage in type 1
goods through this process.
I expect to be a switching point T, the time when dynamic
comparative advantage overcomes the barrier of the transportation cost
and trade starts. The home country exports type 0 goods and the foreign
country exports type 1 goods. If trade starts at time T, both countries
can export thus, the following two inequalities must hold:
Po(T)/g < p*o (T) p*i(T)/g < Pi(T). (10)
Using (9), we rewrite (10) as
ki(T)/k0(T) < g2k\(T)/k*0(T) (11)
Inequality (11) means that net of transportation costs, the
home country has comparative advantage in type 0 goods, while the
foreign country has a comparative advantage of type 1 goods. I define
the function H(t) to analyze the pattern and the timing of trade.


CHAPTER 1
INTRODUCTION
According to traditional theories of international trade,
typically, in the original Hechsher-Ohlin-Samuelson general equilibrium
framework, three assumptions are imposed: First, capital movement from
one country to another country is seen essentially as s substitute to a
commodity trade. Direct Foreign Investment(DFI) is induced flow from the
capital abundant country with low reward to the capital scare country
with high reward to seek higher return. Second, technologies of trading
countries are assumed to be given. That is, no technical changes occur
at these countries. Third, in both product markets and factor markets,
perfect competition prevail. In this dissertation, I construct three
models to examine and extend these three assumptions.
Chapter 2 of the dissertation analyzes political-economic aspect
of DFI. There have been two distinct approaches, which attempts to
incorporate various political-economy considerations. The first strand of
literature has focused on product markets by modeling resource-using
lobbying or by modeling voting behavior. The level of protection is
determined through the self-interest maximizing behavior of economic
agents. The second strand of literature is associated with "Quid Pro Quo"
Direct Foreign Investment (DFI) This strand examines how DFI can be used
to defuse the threat of protection (or the actual level of protection)
based on political-economy considerations DFI aims at diffusing future
protection even if it is associated with current losses. I construct a
1


17
reallocation between the production and the lobbying activity. If DFI
increases the amount of labor devoted to the production
( dJjn/8Kf < 0, oLn/0Tf < 0 ) DFI increases the exports of the foreign
country through this productive labor expansion (- <&2L5Ln/3Kf > 0,
- 2L3Ln/3Tf > 0 ) and increases the second period utility. This effect
differs from the original quid pro quo effect the first period DFI
reduces the second period protection. But the economic implication of
this "productive labor expansion effect" is the same as the original one.
We state:
( Proposition 2 )
If the "productive labor expansion" effect ( <3?2idLn/3Kf,
0213Ln/3Tf ) is sufficiently large, the foreign government has a incentive
for DFI.
2.6. Quid Pro Quo Direct Foreign Investment
This section examines the possibility of quid pro quo DFI in
the importing and exporting sector. Consider DFI in the exporting sector.
Since the DFI reduces the tariff rate in the importing sector, it
promotes the export of the home country (Otdt/dTf > 0) This makes a gain
in the second period [See (32)]. Even though the DFI causes a loss in the
first period, the home government makes DFI to obtain this gain.
( Proposition 3 )
The foreign government has an incentive to engage in quid pro quo
DFI in the exporting sector.


55
higher wage than the competitive wage, reflecting its marginal
productivity. The counterintuitive result comes from this bargaining
power. I summarize the observations
( Proposition 5 )
(5-1) Immigration of unskilled labor increases the wage of skilled labor
and decreases the wage of unskilled labor in the competitive sector.
(5-2) Immigration of unskilled labor decreases the wage of unskilled
labor in the unionized sector if the unionized sector is unskilled labor
intensive.
If the skilled labor immigrates to the economy, the unskilled:skilled
labor ratio in the whole economy declines. The effects of unskilled labor
immigration are immediately derived from proposition 5.
4.6. Concluding Remarks
I have formulated a two-by-two open general equilibrium model
with bargaining between unskilled labor and the management of firm. The
two parties negotiate over the wage and employment. The unionized firm
has monopoly power in the product market. The other factor prices,(the
wage of unskilled labor working in the competitive sector and that of
skilled labor,) are determined by perfect competition. The government has
imposed an import quota on the unionized sector.
First, I have examined uniqueness of the equilibrium. In our
bargaining model, the physical factor ranking between two sectors does
not necessarily coincide with the nominal one, because the sector-
specific bargaining generates a discrepancy between unskilled labor's


61
Helpman, E, 1981, International trade in the presence of product
differentiation, economies of scale, and monopolistic competition
A Chamberlinian-Hecksher-Olhin approach, Journal of International
Economics 11, 305-340.
Homma, M. 1977, A comparative static analysis of tax incidence, Journal
of Public Economics 8, 53-65.
Ikeda, H. 1990, Monopoly and tax incidence: A general equilibrium
analysis, Discussion Paper, Tezukayama University, Japan.
Jones, R. W., 1971, Distortions in factor markets and the general
equilibrium models, Journal of Political Economy 79, 557-572.
Konishi, H., Okuno-Fujiwara, M. and K. Suzumura, 1990, Oligopolistic
competition and economic welfare: A general equilibrium analysis
of entry regulation and tax-subsidy, Journal of Public Economics
42, 67-88.
Krugman, P.R, 1979, Increasing returns, monopolistic competition, and
international trade, Journal of International Economics 9, 469-
479.
Krugman, P.R, 1981, Intra industry specialization and the gains from
trade, Journal of Political Economy 89, 959-974.
Krugman, P.R, 1987, A narrow moving band, the Dutch disease, and the
competitive consequences of Mrs. Thather: Notes on trade in the
presence of dynamic scale economies, Journal of International
Economics 27, 41-55.
Lancaster, K, 1980, Intra-industry trade under perfect monopolistic
competition, Journal of International Economics 10, 151-175.
Lane, T.D. and E. Dinopoulos, 1992, Fiscal constraints on market-
oriented reform in a socialist economy, IMF Staff Papers, 39,184-
203.
Mayer, W., 1984, Endogenous tariff formation, American Economic Review,
74,970-985.
McDonald, I.M. and R. Solow, 1981, Wage bargaining and employment,
American Economic Review 71, 896-908.
Melvin, J.R. and R.D. Warne, 1973, Monopoly and the theory of
international trade, Journal of International Economics 3,
117-134.
Mezzetti, C. and E. Dinopoulos, 1991, Domestic unionization and
import competition, Journal of International Economics 31, 79-
100.
Nash, J., 1950, The bargaining problem, Econometrica 28, 155-162.
Richardson, J.D, 1995, Income inequality and trade: How to think, What
to conclude, Journal of Economic Perspectives 9, 33-55.
Santoni, M, 1996, Union-oligopoly sequential bargaining: Trade and
industrial policies, Oxford Economic Papers 48, 640-663.


24
Xi = k (0) It, y-j = k (0) lj,
(1)
where 1 (1^) is labor developed to production of the Xj.(yj)and k(0)> 0, is
the production coefficient at the initial stage.
I assume that there is learning-by-doing in each industry. At
time t, the production technology of each industry is given by
(2)
Xi(t) = ki(t)li(t), yj(t) = kj (t)lj(t),
where k (t) [ kj(t)] is the production coefficient of goods, Xi[yj].
The coefficient ki is embodied in the learning-by-doing effect. Each
firm can learn from its experience and improve its technology. The
accumulated production of each firm represent its experience.
There is no spillover regarding learning effects across firms or
countries. I define the learning-by-doing effect as
(3)
where 0 < e < 1. This learning effect is assumed to occur at the
industry level and to work as external economy for each firm. Perfect
competition is compatible with this phenomenon.


11
and w. We assume that this system of equations has a unique solution for
a given K, T, Px, Py, and L. The production level of each good, the
endogenous tariff rate, and the domestic prices are determined based on
these variables. Finally, by substituting the values of the endogenous
variables into the demand equation, I can calculate the demands for both
goods, imports and exports.
Given the equilibrium of the host country and the level of DFI, the
foreign government chooses outputs of goods and the volume of trade in
order to maximize its national welfare, W(C*x,C*y). It solves the
following problem:
(19)
where the superscript denotes period 2. The maximum possible welfare
level as a function of direct foreign investment in the first period is
denoted by W2(K£,T£) .
Consider now the first period. Since there is no lobbying, the
equilibrium is described by the specific-factor model with direct foreign
investment. Given the level of DFI, the equilibrium values of Lx, Ly and
w in the home country are defined by
pxFi(K,Lx) w = 0
(20)
PyGi(T,Ly) w 0,
(21)
L* + Ly L.
(22)
The foreign country's reciprocal demand function in the case of free
trade is simplified as


29
The difference in tastes in each country generates a difference in
technologies. In our Ricardian economy, technological differences
determine the trade pattern. Each country exports goods with larger
domestic demand. Consider the case of s=0.5. In this case, both
countries have the identical tastes. At the initial stage, both
countries produce the same amount of all goods. Since both countries
have the same production coefficients, trade does not occurs at any
time. This is the subtle situation, because even if s deviates from 0.5
by a very small amount, trade can occur eventually.
To describe the trading equilibrium, we normalize the home wage to
1.
Pok0 1,
(12)
p\k*i = 1,
(13)
l=g+ (l-s)w*.
(14)
Equation (12) is the zero profit condition of the home countrys
firm and (13) shows the zero profit condition of the foreign firm.
Equation (14) is the market clearing condition for type 0 goods.
Equations (12), (13) and (14) determine three variables p, p* and w*.
Equation (14) implies that w = 1, i.e. the same wage rate prevails in
both countries. This property comes from the assumption that the
distribution of preferences across the two countries are the mirror
images of each other.
I examine the effects of the transportation cost and the learning-
by-doing parameter on the switching point. Differentiating the
equilibrium condition H(T) = 0 with respect to the transportation costs
and assuming the regularity condition H(T) < 0, we obtain
dT/dg = 2gk*1 (T)/H' (T) k*0 (T) < 0.
(15)


58
Or. the other hand, the quota does not affect the above time but protects
domestic production after the introduction of trade.
In chapter 4, I prove uniqueness of the equilibrium of the model
under the conditions that the physical factor ranking between two sectors
coincides with the nominal one and the magnitude of the elasticity of
substitution between skilled and unskilled labor in the unionized sector.
Next, I investigate the effects of trade liberalization(quota expansion).
The Stolper-Samuelson type relation holds as a result of the trade
liberalization. Finally, I derive the effects of immigration. Immigration
of unskilled labor favors the skilled labor's wage, while it reduces the
unskilled labor's wage in the competitive sector. If the unionized sector
is unskilled labor intensive, this immigration decreases the wage of the
unionized workers.


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
THREEE ESSAYS ON INTERNATIONAL TRADE AND WELFARE
By
MAKOTO OKAMURA
May, 1997
Chairman: Elias Dinopoulos
Major Department: Economics
I formulate a two-period, two-sector, specific factor, general
equilibrium trade model with endogenous tariff formation. The tariff rate
depends on labor devoted to lobbying in each of the two industries. We
examine the possibility of quid pro quo direct foreign investment (DFI).
It is shown that more DFI in the exporting sector reduces the level of
protection and quid pro quo DFI occurs. Quid pro quo DFI takes place if
DFI in the importing sector increases directly the ability of the host-
country government to resist protection.
I construct a two-country Ricardian dynamic general
equilibrium model with learning-by-doing and transportation cost. The
two countries differ only in the distribution of the consumer's
preference. At the initial time, no trade occurs due to transportation
costs. There exists a unique time when trade begins. After that time,
each country exports the goods for which it has a stronger preference
vi


54
The impacts on the skilled labor's wage and the unskilled
labor's wage employed in the competitive sector are given
dr,, da.,
= g"l'v<0.
dl y dl
(47)
dl
(48)
The unskilled labor's wage in the competitive sector decreases, while the
skilled labor's wage increases. This results does not depend the
intensity ranking of the two sectors. The both wages are determined by
marginal productivity of factors. Since the unskilled labor supply rises
in the economy, the skilled labor becomes less abundant while the
unskilled labor does abundant. These change in the two factor price seem
to be plausible.
The change in the unskilled labor's wage working in the
unionized sector is
dp W + (l- k)f'} + p[A + (1- k)f "]l'x d0}
dl
K
dl
(49)
If the unionized sector is unskilled labor intensive, both the output
price and the relative wage in the competitive sector go down.
Then, the wage of unskilled labor in the unionized sector decreases.
However, if the competitive sector is unskilled labor intensive, the
output price increases and the relative wage decreases. The total change
in the wage in the unionized sector depends on the magnitudes of two
effects. If the effect of output price change dominates that of the
relative wage change, the unskilled wage in the unionized sector
increases. The unionized unskilled labor with monopoly power can earn


53
The first term of RHS of (46) is negative. The relative wage of the
competitive sector decreases if lx > ly. The second term is negative
since the brace in the second term and l'x are negative. The increase in
the price of X raises the wage of unionized labor which is intensively
hired in the unionized sector. The Stolper-Samuelson relation holds. We
point out that an interesting case may occur. Suppose that lx < ly. In
this case, the output price of X decreases while the relative wage in
the unionized sector increases. The two terms of RHS of (46) take
opposite signs. I can not derive determinate conclusion. This implies
that the Stolper-Samuelson relation may not hold if the effect through
relative wage change( the second term ) is dominated by the price change
effeet( the first term). I summarize these observations
( Proposition 4 )
( 4-1) Stolper-Samuelson theorem holds with respect to the wages of the
skilled labor and the unskilled labor in the competitive sector.
(4-2) Stolper-Samuelson theorem holds with respect to the wage in the
unionized labor if the unionized sector is unskilled labor intensive.
I turn to investigate the effects of immigration on the income
distribution among three factor prices. Consider that foreign unskilled
labor immigrate to the economy. This immigration increases the total
amount of domestic unskilled labor, L and the unskilled:skilled labor
ratio of the whole economy, 1. From Lemma 4, I can show that this
immigration decreases(increases) the price of good X if the unionized
sector(competitive) sector is unskilled labor intensive and decreases the
relative wage of unskilled labor in the competitive sector independent of
the factor intensity among two sectors.


42
w = p (1 1/s) { f lf' ) (15)
I turn to the factor market equilibrium to derive a relation
between kx and coy. The supply of the both factors, L and H, are fixed.
The full employment conditions are described as:
hx + hy =1, (16)
hxlx +hyly = 1, (17)
where hx(hy) = Hx(Hy)/H and 1 = L/H is the unskilled:skilled labor ratio
of the whole economy. Solving the full employment conditions (16), (17)
with respect to hx ,hy and inserting these two variables into the
definition of relative demand we get
g(lx -l)-(lx -l )q
z- (18)
f(l-ly) + (lx-ly)q
where q = Q/H and ghy q > 0 must hold Equations (9), (12) and (18)
determine z as a function of a>y, given parameters H Q and 1. It is
useful for intuitive purpose to investigate this functional relation.
From (12), lx can be written as lx = lx (a>y) Differentiating (12), we
have
l'x = dlx / dcoy = M(f lxf' ) /ff" <0 (19)
wnere
M =
1
1 ~ P+PE ¡3E y. (f lxf') / ll ff "
(20)


14
(C) The effects of immigration of labor.
(1-11) The labor input for production of each industry, the rental rates
for capital and land, the wage rate, the difference of labor input for
lobbying activity between two industries does not change.
(1-12) The labor input for lobbying of each industry increases equally.
The condition in (1-1) means that the effect of DFI in the
manufacturing sector on the marginal value product of the productive
labor ( (1 + t) px Fllc ) exceeds that on the marginal value product of the
lobbying activity ( t'pxFk ). In other words, DFI favors the productive
labor more than the lobbying labor in the manufacture industry. It is
natural that this condition is sufficient for the increase in Lx. This
argument applies to the condition in (1-6). The result (1-1) implies that
the total labor input of agriculture decreases, since the total labor
supply is fixed. The comparative statics' properties in the usual
specific factor model does not change in this framework. See Jones(1971).
The result (1-10) implies that the tariff rate on the importing sector
decreases as the result of DFI in the exporting sector. The reduction in
the tariff rate in the second period encourages DFI in the first period,
that is, quid pro quo DFI. The result (1-5) means that the endogenous
tariff rate on the importing industry and the tax rate on the exporting
industry increases when DFI occurs in the importing sector increases.
The DFI expands the distortion of the relative price (1+t) / (1+s)px/py.
Since the importing industry is more protected as a result of the DFI and
foreign export becomes difficult, I need another factor in order to raise
the quid pro quo DFI. Both (1-11) and (1-11) imply that the immigration
of labor only changes the labor input for lobbying activity. Since both
the tariff and tax rates do not change, every variable affecting national
welfare, such as production, export, import, and prices, remain constant


I certify that I
conforms to acceptable
adequate, in scope and
Doctor of Philosophy.
have read this study and that in my opinion it
standards of scholarly presentation and is fully
quality, as a dissertation for the degree of
Elias'" pirgbpeulos, Chair
Professor of Economics
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy. (
,/Zvn^>v>
Richard Romano
Professor of Economics
I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Bin Xu
Assistant Professor of Economics
I certify that I hane read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.
Ja$es Seale
Associate Professor
Resource Economics
of Food and
This dissertation was submitted to the Graduate Faculty of the
Department of Economics in the College of Business Administration and to
the Graduate School and was accepted as partial fulfillment of the
requirements for the degree of Doctor of Philosophy.
May, 1997
Dean, Graduate School


18
The above proposition is novel. Bhagwati et al. (1987) analyze
the HOS model. They assume a single capital and can not examine in which
sector a quid pro quo DFI appears. Dinopoulos (1989) (1992), and Zhao
(1990) examine quid pro quo DFI in the importing sector in the partial
equilibrium framework.
The impact of DFI in the importing sector on the first period's
utility dwVdKf is negative. When the foreign government is myopic, it
has no incentive to invest in the host country. Since more DFI causes
more protection (dt/dKf>0) and higher tariff rates reduce the export of
the foreign country (3>T < 0), I conclude that the impact of DFI in the
importing sector through the lobbying activity is negative. Like
Dinopoulos (1992) and Zhao (1991), no quid pro quo DFI occurs.
I seek another political process which causes quid pro quo DFI in
the original sense. Bhagwati (1985) states:
"Such "image building" can influence Congress to withstand the
protectionist pressures from the import-competing industry."
Following the above statement, I introduce this effect into the tariff
function in the importing sector. The function is now reformulated as
t = t(Lm L Kf ) (35)
The tariff/tax rate depends not only on the difference of the labor input
for lobbying between industries but also on the level of DFI, itself. I
assume
ti > 0, tU <0, t-2 < 0, t22 > Of ti2 < 0,
(36)
where tj = 3t/5( Ln, L*), t2 =5t/5Kf and so on.


47
unique equilibrium. Combing Lemmas 1,2,3, and 4, I realize that the
intersection of the consumption price function and the production one
gives the unique equilibrium relative factor price in the competitive
sector. I am now in a position to state the uniqueness property.
( Proposition 1 )
There is a unique equilibrium under the following two cases.
(si) The competitive sector is unskilled labor intensive.
(s2) The unionized sector is unskilled labor intensive and
Mlx > ly where M is given by (20).
Proposition 1 identifies two restrictions that are fairy general and
establish the uniqueness of equilibrium despite of the presence of
sector-specific labor union. All remaining variables are also uniquely
determined based on the relative factor price in the competitive sector.
4.4. The Effects of The Bargaining Parameters
In this section, I investigate the effects of the bargaining
parameters on the relative wages of the both sectors. The bargaining
parameters only appears in (12) through parameter ¡3. I easily prove that
P increases(decreases) if a, y (0 )increases The unskilled labor
intensity of the unionized sector is a function of the relative wage in
the competitive sector and the bargaining parameters: lx lx ( Keeping the relative wage constant and differentiating (12) with respect
to lx and p, we obtain


43
and
(21)
Note that 0 < M < 1 since E > 1. The unskilled labor intensity of the
unionized sector decreases as the relative wage in the competitive
sector goes up. Note that (19) relates lx and the relative wage
prevailing in the competitive sector, not in the unionized sector. In
our model, the relative wages differ across the sectors because of the
bargaining in the unionized sector. Taking into account of (19),
differentiation of (18) with respect to ly leads to
dz/dcay = Ni (l-ly) dlx/da>y + N2 (lx-l) dly/dcoy, (22)
where
g(f-lJ' + ¥') + q(g-f + f'nx -ly])
[f(l-ly) + (lx-ly)q]2
and
f(g + g'n-ly]) + q(g[lX-ly]+g-f)
(f[l-ly] + (lX~ly)q)2
(24)
After some algebraic manipulation, I find that Ni and N2 take positive
values.
( Lemma 1 )
If ly > (<) lx then, z' (coy) > (<) 0 .


40
4.3. Uniqueness of Equilibrium
In this section, I focus our attention to issues related to
the uniqueness of momentary equilibrium relative to a set of bargaining
parameters and the level of quota. Uniqueness of equilibrium is a
precondition for the following comparative statics analysis.
The competitive firm hires unskilled and skilled labor to
maximize its profit, given the output price(normalized to one) and the
factor prices. The optimization conditions for the firm are
ry = g' (ly) (7)
w = g(1 y) lyg' (ly) (8)
The above conditions give the relation between ly and the relative
factor price in the competitive sector coy = ry/w, as ly= ly (coy) .
Differentiating ly with respect to coy, we have
1' y = Sly / SCDy = (g ~ 1 yg ) l^" <0 .
(9)
Unskilled labor employment and the wage in the unionized sector
are determined through a efficient Nash bargaining process. The solution
to this bargaining problem is obtained by maximizing the generalized
Nash product with respect to the negotiated wage and employment. The two
first order conditions of this problem are;


46
factor intensity in the whole economy. Consider equation (18) which
includes not only the factor intensity, 1, but also the total quantity
of skilled labor, H. To find the relation between coy and 1, we have to
investigate two cases: (1) L is fixed and H changes, (2) L changes and H
is fixed. I examine the first case.
Differentiating (18) with respect to 1, given the fixed coy, we
have
, az [ z ~ a~ [f(i-iy)+q(ix-iy)]2 (32)
We can show that the sign of z1 is the same as that of ly lx.
The relative demand function and the consumption price function shift
downward ( upward ) if lx >(<) ly. The production price function does
not change since it is independent of the total factor supply.
Considering these observations, we get
( Lemma 4 )
dcOy/dl < 0 and dp/dl <(>) 0 if lx >(<) ly.
I prove the same result for the second case. This lemma means that the
relative unskilled wage in the competitive sector declines if the
relative supply of unskilled labor increases, which seems to be
plausible. Batra(1972) proves the lemma in the closed economy with
C.E.S. utility function. Ikeda(1990) extends the Batra's result to a tax
distorted economy. Homma(1977) examines this relation in the economies
with homothetic utility function .
Lemma 4 implies that there exists a monotonic functional relation
between 1 and coy which enables us to assert that our model has a


25
3.3. Trade at The Initial Eauilibrium
First, I derive the initial equilibrium of each country. Since
each country has identical production technology and each good appears
symmetrically, the initial equilibrium is given by
The home country:
Pi = w/k (0) Xi = sk(0)/n y-j = (l-s)k(0)/m ,
li = s/n, 1-j = (l-s)/m. (4)
where pA is the price of the i-th good, w is the wage rate, xi ( yj )is
the output of the i(j)-th good belonging to type 0(1), and li ( lj ) is
the labor employed in the i(j)-th sector of the type 0(1) goods.
The foreign country:
P*i = w*/k(0), x*i = (l-s)k(0)/n y*3 = sk(0)/m ,
l*i = (l-a)/n, l*j = (1-s) /m. (5)
Reflecting the pattern of taste of each country, the home country
produces more quantity of type 0 goods than the type 1 goods, while the
foreign country produces more quantity of type 1 goods than the type 0
goods.
I proceed to examine the pattern of trade by assuming iceberg
type transportation costs: A fraction of g (<1) of any good shipped to
the other country actually arrives. A fraction of 1-g of any goods
evaporates during transportation. Suppose that the home (foreign)
country exports the i(j)-th goods, the following inequalities must hold.


o
two-period, two-sector, specific factor model with DFI and endogenous
lobbying to integrate these two strands. In this model, I examine in
which sector, the exporting or importing sector is engaged in quid pro
quo DFI to defuse future protection of host country.
Chapter 3 builds a simple dynamic Ricardian general equilibrium
model with learning-by-doing effects of technologies. The learning-by-
doing effect means that each firm can learn from its experience and
improve its technology. Two types of consumers in each of the two
countries are considered. Each type has preferences over a particular
set of similar products. The distribution of the two types of consumers
differs across the two countries. Current productivity of labor, which
the only factor of production, increases over time because of learning-
by-doing considerations. The dynamic nature of the model allows us to
determine the duration of autarky based on transportation costs.
Transportation costs provide a realistic and natural way to generate an
initial period without trade, which is necessary to allow the
interaction between national differential tastes and learning-by-doing.
I investigates the effects of domestic income transfers across different
types of consumers, tariffs and import quotas on the pattern and volume
of trade.
Chapter 4 of the dissertation constructs a general equilibrium
model of an open economy with unionized sector. The export producing
sector is perfectly competitive and utilizes two factors of production,
skilled and unskilled labor. The import-competing sector consists of a
unionized firm that is protected from foreign competition through an
import quota. The domestic monopolist utilizes both skilled and unskilled
labor under a constant returns to scale technology. However, unlike the


20
I summarize the argument as follows:
( Proposition 4 )
Suppose that the tariff function is t = t (Lm La, Kf) satisfying (39).
The foreign government obtains an incentive to engage in quid pro quo DFI
in the importing sector if the following inequalities hold
(a) t12pxF + tiPxFk < 0.
(b) t2PxFi + (l+t)pxFlk < 0.
2.7. Concluding Remarks
The present chapter has formulated a specific-factor, two-period,
two-country general equilibrium model with an endogenous tariff
determined by lobbying activity of each industry. The lobbying activity
uses a productive resource, labor. The wage rate is determined by the
labor market equilibrium, and is not assumed to be fixed. I have examined
three problems: (1) the impacts of DFI in the exporting and importing
sector on the second period equilibrium, (2) which DFI reduces the
protection encourages quid pro quo DFI, (3) the possibility of new type
of quid pro quo DFI.


39
Y (8) represents the elasticity of U with respect to excess
wage(employment). Following the literature, we assume that the union is
employment-oriented, that is, y > 0.
The management of the firm is interested in maximizing its
profit, nx
n* = pX rxLx wHx.
(5)
Both parties engage in efficient Nash bargaining. The threat point is a
pair of -wLx and 0. This pair is obtained under the assumption that the
union strikes ( that is, x = 0 ) if there is disagreement during the
bargaining process. The generalized Nash bargaining product, G thus is
constructed as
W = (pX-rxLx) -Ua (6)
where parameter a is the bargaining power of the union and 0 < a < 1.
I assume that under free trade, this economy imports the good
of the unionized sector and exports the good of the competitive sector.
The government imposes an import quota Q as trade restriction measure.
Domestic consumption of the two goods is Xc = F(LX,HX) + Q and Yc =
G(Ly,Hy) Q. I assume that the consumption of the goods Y is positive
under the quota, that is, G Q > 0.


23
3.2. The Model
There are two types of consumers in the home country and the
foreign country. Each type (type 0 and type 1) has the CES utility
function:
_l_
i=n p
type 0. (*,*) = (2>f)
M
=m 1
type i. u(yltym) = ,y,
i=i
where X^y,) is consumption of the i-th product by type 0(1) consumer
and 0 < p < 1. Type 0 individual consumes only differentiated product
x, while type 1 consumers only differentiated product y. The number of
the goods each consumer purchases, n and m, respectively, are fixed.
There is one unit of labor (consumer) in each country. We
consider the distribution of consumers as mirror images of each other.
That is, the home country has s(l-s) fraction of the type 0(1) consumer,
while the foreign country has l-s(s) fraction of the type 0(1) consumer.
We assume that s > 0.5 which means that the home country has a
stronger national demand for the type 0 goods than type 1 goods and the
reverse pattern of preference prevails in the foreign country except for
s=0.5. Both countries have the same tastes when s=0.5.
The only factor of production is labor. At the initial stage,
each country has the same technology exhibiting constant returns to the
production of each good.


CHAPTER 2
ENDOGENOUS TARIFF FORMATION AND DIRECT FOREIGN INVESTMENT: A GENERAL
EQUILIBRIUM ANALYSIS
2.1. Introduction
There have been two distinct approaches to the theory of
endogenous protection, which attempts to incorporate various political-
economy considerations in the formation of tariffs, quotas and other
instruments of protection. The first strand of literature has focused on
product markets by modeling resource-using lobbying (i. e., Findlay and
Wellisz (1982)) or by modeling voting behavior (i. e., Mayer (1984)). In
both cases, the level of protection is determined through the self-
interest maximizing behavior of economic agents. The second strand of
literature is associated with "Quid Pro Quo" Direct Foreign Investment
(DFI) This strand examines how DFI can be used to defuse the threat of
protection (or the actual level of protection) based on political-economy
considerations. Bhagwati (1985, 1986, 1987) has identified the new type
of DFI which aims at diffusing future protection even if it is associated
with current losses. Bhagwati et al. (1987), Dinopoulos (1989, 1992),
Wong (1989) Dinopoulos and Wong (1991), and Zhao (1991) have developed
models of Quid Pro Quo DFI.
These two distinct approaches to endogenous protection aim at
understanding the instruments and channels which determine the level of
protection, but have ignored each other's implications. For example,
although Findlay and Wellisz (1982) have emphasized the resources which
are utilized in the lobbying process within the context of general
equilibrium, Bhagwati et al. (1987) do not model the resource cost of
4


CHAPTER 5
CONCLUSIONS
The model in the chapter 2 shows that the level of DFI is
substitutive(complementary) to the lobbying activity and protection when
DFI occurs in the exporting (importing) sector. The DFI in the exporting
sector reduces the level of protection, which encourages the foreign
government to engage quid pro quo DFI. On the other hand, the DFI in the
importing sector leads to a higher tariff rate. However, if DFI increases
the host-country government resistance to lobbying and this effect
overcomes the effect of lobbying activity, the tariff rate is reduced by
DFI and there is scope for quid pro quo DFI. I show another channel by
which DFI influences protection: the amount of labor devoted to
productive activities may increase as the result of DFI. The foreign
country exports more and obtains a higher level of welfare. The foreign
government chooses DFI taking this "productive labor expansion effect"
into account. This effect emerges in the general equilibrium framework of
the model in which the tariff is determined through resource-using
lobbying activities.
Chapter 3 establishes the existence of unique switching point,
T. Before T there is no trade. After T, trade occurs based on taste
difference across countries. The government of the home country can
completely reverse the existing pattern of trade by taxing the type 0
consumer and transferring this to the type 1 consumer during the no
trade period. The tariff delays the time when the trade takes place, but
domestic production of the protected industry vanishes after this time.
57


48
(33)
The lx function shifts upwards as p increases. The relative demand
equation (18) is a function p Differentiating this with respect to lx
and P, we get
z0 = dz/dp = Ni (1-ly) dlx/dp.
(34)
From (34) I have
( Lemma 5 )
Suppose that lx >(<) ly. Then, z0 >(<) 0.
If the unionized (competitive) sector is unskilled labor intensive, the
relative demand goes up (down). Next, I examine the change in the
consumption price function. The change is given by
(35)
We have that pf >(<) 0, if lx >(<) ly. The consumption price function
shifts upwards(downwards) if the unionized(competitive) sector is
unskilled labor intensive.
Differentiating the production price function (27) I derive
dpp El?(g-lvg')lxf"
-m- = ; 5 <0.
(36)


56
wages across sectors. I have established uniqueness of equilibrium under
the condition that the order of the above two rankings is same. Second,
I have investigated the impact of the bargaining parameters on the
relative factor prices of unionized and competitive sector. If the
unionized sector is unskilled labor intensive, the relative wages of both
sectors increase as the union becomes more employment-oriented and its
bargaining power rises. Third, I have examined the effects of trade
liberalization(quota expansion). I have shown that the Stolper-Samuelson
type relation holds as a result of the trade liberalization. Finally, I
have derived the effects of immigration. Immigration of unskilled labor
favors the skilled labor's wage, while it reduces the unskilled labor's
wage in the competitive sector. If the unionized sector is unskilled
labor intensive, this immigration decreases the wage of the unionized
workers.


34
is given to the type 1 consumers, the revenue goes to the foreign
industry. Domestic production of type 1 goods can not be revived in the
future. The manner of redistribution changes the level of the outputs
but does not affect the pattern of trade.
3.6. Concluding Remarks
The chapter has developed a two-country dynamic Ricardian model of
trade in similar products along the lines suggested by Bhagwati(1982).
The model was based on differential national tastes, learning-by-doing
and transportation costs. We analyzed the role of these three features
in determining the pattern of intraindustry trade. A country exports
those similar products which correspond to a larger domestic market. An
income transfer policy can reverse the initial pattern of trade in
similar products. A tariff or an import quota delays the time when trade
starts and reduces the volume of trade. The simple structure of the
model allows several possible extensions and generalizations.


62
Stolper, W. and P. Samuelson, 1941, Protection and real wages, Review of
Economic Studies 9, 58-73.
Wong, Kar-yiu., 1989, Optimal threat to trade restriction and quid pro
quo foreign investment, Economics and Politics, 1, 277-300.
Wood, A, 1995, How trade hurt unskilled workers, Journal of Economic
Perspective, 9, 57-80.
Zhao, L., 1991, Endogenous protection and direct foreign investment.
University of Florida, mimeo.


BIOGRAPHICAL SKETCH
Makoto Okamura was born in 1956 in Japan. After studying at Keio
University, he obtained a Master of Arts in economics at Osaka
University in 1982. He entered the PH.D program in economics at
University of Florida in 1990.
63


28
H (t) = k1(t)/k0(t) < g2k\(t)/k*0(t) .
If H(t) takes negative value at some t, trade occurs. At t=0,
H(0)=l-g > 0. Thus, trade does not occur at the initial equilibrium.
H(t) is continuous in t. This implies that a no trade period prevails in
some range of time near time 0.
Since the shape of H(t) is ambiguous, two cases are possible:
(1) there is a time t with H(t) = 0. (2)H(t) is positive for all t. Case
(2) means that no trade occurs forever. Therefore, we focus on the case
(1). There may be multiple time intervals such that H(t)=0. Consider the
first time t=T with H(T)=0 and assume the regularity condition dH(t)/dt
< 0 at t=T. Take the time t+h, where h is a sufficiently small positive
number. From the regulatory condition, H(t+h) is negative, which means
the home (foreign) country has comparative advantage in the type 0(1)
goods. The home country exports type 0 goods, while the foreign country
exports the type 1 goods. We assume that complete specialization occurs.
The type 0 goods are produced only in the home country, while the type 1
goods are produced only in the foreign country. Once specialization
prevails, the learning -by-doing effects appears in neither type 1 goods
of the home country nor the type 0 goods of the foreign country. So the
coefficient of these goods do not change. On the other hand, the
coefficient of exported goods increases as time passes. H(t) remains
negative after the time t+h. This shows that the time T is the switching
point from the no trade period to the trading periods. This switching
takes place just one time and there is no reversal of the trade pattern.
Proposition 1
There is a unique switching point, T. Before T there is no trade.
After T, there is trade based on taste difference across countries.


16
WVW1, = -v|/x
(31)
dV/dkf = dWx/dK£ + pdW2/dKf
= dWxy (-Hk+\|/k) + pW2y (-H2k +02k 2x5Ln/3Kf + 02t5t/9Kf)
0.
(32)
dV/dTf = dW1/dTf + pdW2/dTf
= dWxy(-Ht+V|/t )+ pW2y(-H2t+ 0.
(33)
Two additional terms emerge in the second period effect because of the
reallocation of labor ( -2i<5Ln/<3Kf -02idLn/3Tf ) and the change in the
tariff rate.
To examine conditions (32) and (33), we assume that the home
country has a utility function U(Cx,Cy) = U0(Cy) + Cx. This quasi-linear
utility function implies that the host country's demand for good Y is
independent of income. With d>(v|/) denoting the output of good Y the
consumption of good Y in the second (first) period and the results of the
comparative statics, I obtain
VF)C < 0, ¥t>0, $> < 0,
0,
0, <1>t > 0.
(34)
Before examining a quid pro quo DFI, I have another route which induces
DFI from the foreign government although it has no incentive in the
myopic context. This is the effect through a change in the labor


44
Following Batra(1972), Melvin and Warne(1973), I define the
consumption price pc through the utility maximizing behavior of the
representative consumer.
Pc = u[z (coy) ] .
(25)
Differentiating totally the consumption price function, I get
' = dPc
dcoy
u'(z)z'(COy).
(26)
I can obtain
( Lemma 2 )
Suppose that ly > (<) lx. Then, pc' > (<) 0.
The consumption price function is upward(downward)sloping with respect
to the relative wage in the competitive sector, if the
competitive(unionized) sector is unskilled labor intensive.
I also define the production price pp by the optimization
behavior of each firm, (8) and (13).
E(g-lyg')
'P~ (f-hf)
(27)
Noting that kx = kx(coy) and E takes a constant value, we have
Pp =~= D[B(Mlx ly) + S(1 pJJ,
dcov
(28)


than does the other country. A government can completely reverse the
pattern of trade by a tax-transfer policy to change the distribution of
income. A tariff delays the time when trade begins, while a quota does
not affect the timing of trade.
Finally, I formulate a open economy model in a two-by-two
general equilibrium setting. Unskilled workers establish a labor union
and bargain with a firm about their employment level and wage, while
skilled workers are employed in a competitive market. Both parties are
engaged in Nash bargaining. The labor union is assumed to be employment
oriented. I establish that trade liberalization decreases the wage on
unskilled workers in the unionized sector if the unionized sector is
labor unskilled labor intensive. I also show that immigration of
unskilled labor decreases the wage of unskilled unionized labor if the
unionized sector is unskilled labor intensive.
Vll


7
K = Kd + Kf, T = Td + Tf. (3)
The host country government imposes a tariff (subsidy) on each industry.
The domestic price of each good is
qx
= (1
+ t
)Px /
(4)
%
= (1
+ S
) Py
(5)
where
qx
(qy)
is
the domestic
price of
X (Y) and px (py) is the world
price
of
X (Y) .
The tariff
(subsidy)
rate is determined endogenously
through lobbying by each industry. Each industry hires lobbyists who put
political pressure to the government to obtain protection. The lobbying
activity utilizes only the following endogenous protection function which
determines the size of the tariff as a function of lobbying
t = t( Lm La
),
s = s (La -
71m ) t
(6)
where Lm (La)
is a
lobbying
input by manufactured (agricultural)
sector
We assume that
the
shape of
the tariff (subsidy) function, is as
follows:
t (o)
0,
t' > 0,
t" < 0,
S (o) =
o,
s' >0,
s" < 0.
(7)
This implies that for a given rival's lobbying behavior, the marginal
product of the lobbying by each industry is positive but decreasing.
These equations mean that when the X industry faces an import tariff, the
Y industry faces an export tax and vice versa. A representative consumer
has a well-behaved utility function U(X, Y). The consumer demand for each
good can be written, Xd(qx,qy,I) and Yd(qx,qy,I), where I denotes income.
I turn to the foreign country. The transformation curve does
not change over periods,


6
resistance to lobbying and this effect overcomes the effect of lobbying
activity, the tariff rate is reduced by DFI and there is scope for quid
pro quo DFI. Another channel by which DFI influences protection is also
shown: the amount of labor devoted to productive activities may increase
as the result of DFI. The foreign country exports more and obtains a
higher level of welfare. The foreign government chooses DFI taking this
"productive labor expansion effect" into account. This effect emerges in
the general equilibrium framework of the model in which the tariff is
determined through resource-using lobbying activities.
Section 2.2 formulates the model. Section 2.3 characterizes the
equilibrium. Section 2.4 develops basic comparative statics. Section 2.5
calculates the optimal DFI levels. Section 2.6 analyzes the possibility
of quid pro quo DFI and section 2.7 offers the concluding remarks of the
chapter.
2.2. The Model
I formulate a two-country, specific-factor, general-equilibrium
model with endogenous tariff formation using building blocks from Findlay
and Wellisz (1982). There are two countries, foreign and domestic, in the
world. A manufactured good and an agricultural good are produced under
perfect competition and traded in both countries.
First I describe the host country. The manufactured (agricultural)
good is produced by a specific factor, capital (land), and a mobile
factor, labor. I denote the manufactured (agricultural) good by X (Y) .
The production functions are
X = F(K,LX), (1)
Y = G (T, Ly) ,
(2)