Social Dumping and International Trade

Naoto Jinji July 15, 2005

Abstract In this paper, I investigate the eects of social dumping in a North-South trade model when rms strategically interact in the output market. The South rm practices social dumping due to its monopsonistic power in the labour market. I show that, contrary to a common complaint by rms in developed countries, social dumping by the South rm is benecial to the North rm. The South rm, on the other hand, may be better o by not practicing social dumping. North consumers suer from social dumping. Imposing social clause taris or labour standards results in conferring a strategic advantage on the South rm, whereas it improves social welfare in the North. Keywords: social dumping; monopsony; oligopsony; labour standards; social clause taris; Cournot oligopoly. JEL classication: F12; F13; J42; J80; L13. Very preliminary and incomplete!! Please do not quote without authors permission. Comments and suggestions are welcome.

Financial

support from Japan Society for the Promotion of Science under the Grant-in-Aid for Young Scientists (B)

is gratefully acknowledged. Faculty of Economics, Okayama University, 3-1-1 Tsushima-Naka, Okayama 700-8530, Japan. Phone & Fax: +81-86251-7525. E-mail: jinji@e.okayama-u.ac.jp

Introduction

Social dumping refers to a situation in which rms that are located in countries where labour standards are weak produce and export goods by using unduly cheap labour under poor working conditions (Corden and Vousden, 2001). In the case of multinational enterprises (MNEs), social dumping also means the decision of a home rm to serve the domestic market through a plant located in a foreign country, where workers protection does not meet home standards and labor costs are thus signicantly lower (Cordella and Grilo, 2001: p. 645). Social dumping is not just the choice of rms. It is argued that the governments of developing countries often set lax labour standards to create a competitive cost advantage for their own industries (Sinn, 2001: p. 3) or to attract MNEs. Such behaviour of the governments may result in a race to the bottom. In order to prevent social dumping and protect rms located in developed countries from the threat of unfair competition arising from social dumping, labor unions in European countries and other developed countries and human rights activists argue that market access in the North should be conditioned on raising labor standards in the South (Golub, 1997: p. 20). The legal linkage between labour standards and trade restrictions is sometimes referred to as social clauses. Adoption of a particular type of social clauses in international trade agreements has been proposed and discussed.1 However, developing countries argue that such social clauses are disguised protectionism. It is important to know the sources of social dumping to examine the eects of social dumping. Maskus (1997) and Martin and Maskus (2001) point out that one potential source of social dumping is the monopsonistic labour markets in developing countries. Monopsony and oligopsony in labour markets have recently been a hot issue in the literature (Boal and Ransom, 1997; Bhaskar, Manning, and To, 2002). This is partly because empirical studies support such structures in labour markets. The main purpose of this paper is to examine the eects of social dumping in inter1 See

Leary (1996) for a survey of social clauses in international trade agreements. There is also a large literature on

labour standards and trade. See, for example, Abe and Zhao (2005), Bagwell and Staiger (1999, 2001), Bhagwati (1995), Brown (2001), Brown, Deardor, and Stern (1996, 1998, 2003), Dehejia and Samy (2004), Kok, Nahuis, and de Vaal (2004), Martin and Maskus (2001), Maskus (1997), Shelburne (2004), Srinivasan (1995), and Suranovic (2002).

national markets when social dumping is based on monopsonistic labour markets. The potential importance of the interaction between the strategic relationship of rms in the output market and the monopsonistic power in the labour market is emphasized. In order to examine this issue, I construct a simple duopoly model with asymmetric labour markets. Two rms located in dierent countries in which conditions of labour markets are dierent. In one country, the labour market is perfectly competitive, whereas in the other country the labour market is monopsonistic. This asymmetry in the labour markets aects the strategic relationship of rms in the output market. I also examine the eects of various policies, such as taris and labour standards, by the government of the country in which the output market is located. I only consider the case in which the output market is located in the country whose rm does not practice social dumping. The major results in this paper are as follows. First, I show that the monopsonistic power of one rm in the labour market causes the rm to conduct social dumping in the sense that the wage rate paid to workers is below the marginal value product of labour. When rms are price takers in the output market, the rm gains a competitive advantage from social dumping. However, it is shown that, when the output market is under Cournot competition, the rm that practices social dumping may ironically suer from its own social dumping. Under some plausible conditions, it could earn higher prots by not practicing social dumping. The rival rm that does not conduct social dumping, on the other hand, actually benets from the other rms social dumping. Second, since social welfare in the country in which the output market is located is reduced by the foreign rms social dumping, its government may have an incentive to implement some policies to improve domestic welfare. I show that any of the three policy instruments, namely, ad valorem taris, social clause taris, and labour standards, can improve domestic welfare. However, the eects of these policies on domestic and foreign rms prots are dierent. An ad valorem tari works as a tool of shifting rents from the foreign to the domestic rm. A social clause tari, on the other hand, works against the domestic rm. It may or may not increase prots of the rm that practices social dumping, depending on the conditions in the labour market. Labour standards, which

increase the wage paid by the rm of engaging in social dumping, hurt the domestic rm and help the foreign social-dumping rm. The results in this paper brings out the striking contrast between social dumping and ecological dumping,2 which refers to a situation in which a government uses lax environmental standards to support domestic rms in international markets (Rauscher, 1994: p. 823). The existing studies have shown that ecological dumping is typically seen when the output market is imperfectly competitive and that ecological dumping actually confers a competitive advantage on domestic rms in international markets (Conrad, 1993; Barrett, 1994; Kennedy, 1994; Rauscher, 1994). By contrast, I show that social dumping based on the monopsonistic labour market may be harmful to the rm that practices social dumping and is benecial to the rival rms. This surprising result holds when the output market is imperfectly competitive but does not when it is perfectly competitive. A number of existing studies are related to this paper. First, Naghavi (2005) is most closely related to this paper. Like this paper, he investigates the consequences of asymmetric labour market in a North-South trade model. In the Southern country, the labour market is under oligopsony, while it is perfectly competitive in the Northern country. Unlike this paper, however, he assumes that the North rm can choose its plant location in either country and focuses on the eects of oligopsony in the Southern labour market on the North rms location choice. He shows that the North rm is not always attracted to the Southern country, in which wage is lower. However, he does not analyze how the market outcomes dier, depending on the conditions in the Southern labour market. He does not examine the eects of tari policies, either. Maskus (1997) and Martin and Maskus (2001) also analyze the eects of monopsony in the labour market. These two papers also examine the eects of taris on imports from the country in which the labour market is under monopsony. However, since they do not consider the strategic interaction between rms in the output market, the relationship between imperfect competition in the output market and social dumping in the input
2 Ecological

dumping is alternatively called eco-dumping or environmental dumping.

market is not claried by their analysis.3 Corden and Vousden (2001) examine the eects of improving labour standards in the export sector of developing countries. Although their main purpose is to analyze the effects of improving labour standards in the export sector on the wage dierential between the export and import sectors, they also explore the interaction between the monopsonistic labour market and terms-of-trade eects. However, the strategic interaction between rms in the output market is again not considered. Cordella and Grilo (2001) analyze the eects of imposing social clause taris on social dumping. However, their main concern is about rms location choices and how social clause taris work for preventing domestic rms from relocating to the country of low wages.4 They assume that the wages in the domestic and foreign countries are xed and do not consider the role of oligopsonistic structure in labour markets. The interaction of oligopoly in the output market and oligopsony in the factor markets is analyzed by Okuguchi (1998, 2000). He considers the model in which rms that play Cournot competition in the output market are oligopsonists in the factor market. His model is dierent from mine in that rms share the same factor market. In my case, since rms are located in dierent countries, they do not share the same factor market. The remainder of the paper is organized in the following way. Section 2 sets up the model. Section 3 examines the eects of social dumping. As a benchmark, the case in which rms are price takers in the output market is rst analyzed. After that, the case of Cournot competition in the output market is investigated. Section 4 analyzes the eects of policies by the government of the country in which the output market is located. Section 5 concludes.
3 In

the trade literature, the eects of the monopsony and oligopsony power in primary factor markets (Feenstra,

1980; Markusen and Robson, 1980; McCulloch and Yellen, 1980) and in intermediate good markets (Devadoss and Song, 2003a, b) have been examined in general equilibrium models with perfectly competitive nal good sectors. Kuroda (2004) examines the eects of local content protection in a small open economy with monopsonistic local intermediate good market. See also Bhagwati, Panagariya, and Srinivasan (1998, Chapter 24). 4 Leahy and Montagna (2000) investigate the issue of social dumping and the location choice by MNEs from the viewpoint of the governments in developing countries. They show that the governments of developing counties have an incentive to engage in social dumping in the sense of banning labour union in the short run to attract MNEs and extract higher rents in the long run.

The Model

There are two countries: Country N and Country S. In each country, one rm is located. Call these rms rm N and rm S. These rms produce a homogenous good. Let y N and y S be rms N and Ss output, respectively. Labour is the only production factor. I assume identical labour supply curves in the two countries. The inverse labour supply in each country is given by w(li) = + li , i = N, S, (1)

where li is employment by rm i, w(li ) is wage rate when the employment level is li , > 0, and > 0. For simplicity, I assume that one unit of labour is required to produce one unit of output. That is, y i = li , i = N, S. The labour market in Country N is competitive and hence rm N is a price taker in the labour market. The labour market in Country S is, on the other hand, monopsonistic.5 Firm S realizes that it faces an upward-sloping labour supply curve. The output market is duopolistic. I assume that rms compete in quantities in Cournot fashion. For simplicity, I assume that the output market exists only in Country N. The inverse demand in the market in Country N is given by p(y) = a y, where y y N + y S and a > 0. Firm is prot, i (y N , y S ), is given by i (y N , y S ) = (p(y) w(y i))y i, Consumers surplus in Country N, CSN , is given by CSN =
0 2 y

(2)

i = N, S.

(3)

p(t)dt (4)

= y /2.

Then, social welfare in Country N, W N , is measured by the sum of rm Ns prot and consumers surplus, i.e., W N = N + CSN . When the government of Country N imposes any taris on imports, tari revenue is added to the social welfare measure.
5 In

the subsequent analysis, I also consider the case of the competitive labour market in Country S as a benchmark.

The government of Country N implements various policies, which will be examined in section 4. I only consider the case in which the government commits to a certain policy before rms act. Throughout the paper, the government of Country S is assumed to be passive and allow its domestic rm to practice social dumping unless Country N requires labour standards on its imports.

Social Dumping

In this section, I analyze the eects of rm Ss practicing social dumping. As will be shown below, the strategic interaction between rms in the output market plays an important role to determine the eects of social dumping. In order to make this point clear, I rst consider a benchmark case in which the output market is competitive. 3.1 The competitive output market

Suppose that rms are price takers in the output market. In this case, rm Ns output is determined by yN = a yS . 1+ a yN . 1 + 2

(5)

Firm Ss optimal output is, on the other hand, given by yS = (6)

Assuming interior solutions, outputs and prots in Nash equilibrium (NE) are given by
N yp =

2(a ) , 3 + 2

S yp =

a , 3 + 2
2

(7) (8)

N p = 0,

S S p = yp

respectively, where the subscript p indicates equilibrium variables in the case of price takers in the output market. In order to have interior solutions, I assume that a > 0 holds. Since rm N is a price taker both in the factor market and in the output market, it earns zero prots in this case. Firm S, on the other hand, earns positive prots because

it has a market power in the factor market. Consumers surplus in this case is given by CSN = p 9(a )2 . 2(3 + 2)2 (9)

In order to understand the eect of social dumping by rm S, consider a case in which rm S is a price taker in its factor market. In this case, rm Ss output is determined by yS = a yN . 1+ a , 2+ p = 0, S (10)

Outputs and prots in NE are then given by yp = N a , 2+ p = 0, N yp = S (11) (12)

respectively, where a tilde () indicates variables in the case where the labour market in Country S is perfectly competitive. Consumers surplus in this case is given by
N CSp

2(a )2 = . (2 + )2

(13)

From (7) to (13) I obtain the following proposition: Proposition 1 When the output market is perfectly competitive, rm S earns higher prots by practicing social dumping. Firm Ss output and employment level are lower when it practices social dumping. Although Firm Ns output is higher when rm S engages in social dumping, its prot is unaected by rm Ss social dumping. Consumers in Country N suers from social dumping by rm S.
S Proof. Since p > 0 = p , rm N earns higher prots by social dumping. On the other S N S N S hand, p = c = 0. For output levels, it is easily shown that yp yp = (a )(1 + N )/(3 + 2)(2 + ) < 0 and yp yp = (a )/(3 + 2)(2 + ) > 0. Moreover, it follows N N

that CSN CSp = (a )2 (2 + 7)/2(3 + 2)2 (2 + )2 < 0. p This result is just as expected. The rm that practices social dumping benets from social dumping. This is because rm S has a monopsony power in the labour market in Country S, whereas rm N does not have any market power in the labour market. Firm N located in Country N, on the other hand, does not care about whether rm S
8

conducts social dumping, as long as the output market is competitive. However, since consumers surplus is lower in the presence of social dumping, Country N suers from social dumping by rm S. 3.2 Cournot competition in the output market

I now turn to the case of Cournot competition in the output market. In this case, from the rst-order condition (FOC) for prot maximization, rm Ns reaction function is given by y N (y S ) = a yS . 2+ a yN . 2(1 + )

(14)

Firm Ss reaction function is, on the other hand, given by y S (y N ) = (15)

Assuming interior solutions, outputs and prots in NE are respectively given by

N yc =

(a )(1 + 2) , 2 2 + 6 + 3
2

S yc =

(a )(1 + ) , 2 2 + 6 + 3
2

(16) (17)

N N c = yc

S S c = (1 + ) yc

where the subscript c indicates equilibrium variables in the case of Cournot competition in the output market. Consumers surplus and social welfare in Country N in this case are respectively given by CSN = c WcN (a )2 (2 + 3)2 , 2(2 2 + 6 + 3)2 (a )2 (17 2 + 20 + 6) = . 2(2 2 + 6 + 3)2 (18) (19)

As in the previous subsection, as a benchmark I consider a case in which rm S is a price taker in its factor market. In this case, rm Ns reaction function does not alter and hence is given by (14). Firm Ss reaction function in this case is, on the other hand, given by y S (y N ) = a yN . 2+

(20)

Outputs and prots in NE are then given by yc = N a , 3+

2

yc = S

a , 3+
2

(21) , (22)

c = yc N N

c = yc S S

respectively, where a tilde ( ) indicates, as in the previous subsection, variables in the case where the labour market in Country S is perfectly competitive. Consumers surplus and social welfare in Country N in this case are respectively given by
N CSc

WcN

2(a )2 = , (3 + )2 3(a )2 = . (3 + )2

(23) (24)

From (16) to (24), the following proposition is obtained. Proposition 2 Under Cournot competition in the output market, (i) rm N benets from rm Ss social dumping; (ii) rm S earns higher prots by not practicing social dumping if is low; (iii) the employment by rm S is lower when rm S practices social dumping; and (iv) Country N suers from social dumping by rm S.
N Proof. (i) From (17) and (22), it yields that c c = (a )2 (4 2 + 13 + 6)/(2 2 + N S 6 + 3)2 (3 + )2 > 0. (ii) From (17) and (22), it follows that c c = (a )2 ( 4 + S S S 5 3 + 6 2 2 3)/(2 2 + 6 + 3)2 (3 + )2 . It is shown that c (resp. >) c if S (resp. >) , where 0.6658. (iii) Using (16) and (21), it yields that lc c = lS S yc yc = (a )(2 + )/(3 + )(2 2 + 6 + 3) < 0. (iv) Comparing (19) with (24) S yields W N W N = (a )2 2 (7 2 + 22 + 9)/2(2 2 + 6 + 3)2 (3 + )2 < 0. c c

A surprising result in this proposition is that rm N benets rather than suers from rm Ss social dumping. Firm S, on the other hand, is not always better o by practicing social dumping. In particular, if is lower than approximately 0.6658, rm S earns higher prots by not practicing social dumping. This makes a sharp contrast with the result in the case of competitive output market. The strategic interaction between two rms in the output market makes this dierence.
10

When rm S has a monopsony power in the labour market in Country S, it takes into account the fact that, given rm Ns output, it can increase its prots by reducing employment from the competitive level. However, since the output is proportional to the employment level, this choice decreases rm Ss output level for a given level of rm Ns output. This means that rm N is now facing a less-aggressive rival and hence that it can expand its output because outputs are strategic substitutes. Consequently, rm Ns output in NE is higher and rm Ss output in NE is lower when rm S has a monopsony power in the labour market. Figure 1 depicts reaction curves of the two rms. RN indicates rm Ns reaction curve. RS and RS indicate rm Ss reaction curves with and without social dumping, respectively. As drawn in the gure, by practicing social dumping, rm Ss reaction curve shifts inward, changing the Nash equilibrium point from Ec to Ec . As a result, y N in NE becomes higher and y S in NE becomes lower. Although rm Ss social dumping causes its output to contract, it increases the markS up. With social dumping, the mark-up is given by c pc wc = (a )(1 + )2 /(2 2 + 6 + 3). Without social dumping, on the other hand, the mark-up is given by c

S pc wc = (a )/(3 + ). Thus, it follows that (a )( 2 + 2 + 1) > 0. c c = (2 2 + 6 + 3)(3 + ) Firm Ss monopsony power in the labour market becomes higher as the labour supply becomes less elastic, i.e., as becomes higher. It holds that (c c )/ = (a)(6 4 + 38 3 + 78 2 + 54 + 9)/(2 2 + 6 + 3)2 (3 + )2 > 0. Thus, when is low, the eect of a reduction in the output dominates that of an increase in mark-up by practicing social dumping. Hence, rm Ss prot is lower under social dumping. As becomes higher, the eect of an increase in mark-up by social dumping tends to dominate that of a reduction on the output, yielding higher prots for rm S by social dumping. Although rm Ss social dumping is good for rm N, it is not good for Country N as a whole, because consumers in Country S suer from rm Ss social dumping, as in the case of competitive output market.

11

Taris and Labour Standards

In the previous section, I show that when the output market is imperfectly competitive, social dumping by rm S benets rm N. However, since social welfare in Country N is lower when rm S practices social dumping, the government of Country N may have an incentive to implement policy to improve social welfare in its own country. In this section, I consider three policy instruments. I rst consider the usual taris on imports. Although the case of ad valorem taris is examined, qualitative results do not alter when specic taris are considered. Second, I consider so-called social clause taris. These taris are imposed as a fraction of the dierence between the competitive and the actual level of wage in Country S.6 Finally, I consider labour standards. In this section, I only consider the case of Cournot competition in the output market with monopsonistic labour market in Country S. 4.1 Ad valorem taris

I rst consider ad valorem taris on imports. Let t be an ad valorem tari imposed by the government of Country N on imports from abroad. When a tari is imposed, rm Ss prot is given by S (y S , y N ; t) = (1 t)p(y) w S (y S ) y S . From the FOC, rm Ss reaction function in this case is given by y S (y N ) = (1 t)(a y N ) . 2(1 t + ) (26) (25)

Firm Ns reaction function remains the same, which is given by (14). Then, assuming interior solutions, outputs and prots in NE are respectively given by
N yt =

(a )(1 + 2) t(a 2c) , 2 2 + 2(3 t) + 3(1 t)

2

S yt =

(a )(1 + ) t(a + a + ) , (27) 2 2 + 2(3 t) + 3(1 t)

2

N N t = yt
6 Note

S S t = (1 t + ) yt

(28)

that the social clause tari considered in this paper is slightly dierent from that in Cordella and Grilo (2001).

They dene the social clause tari as the fraction of the dierence between the domestic and the foreign wages, assuming constant wages. In my case, since rms do not relocate, their denition of social clause tari is not appropriate.

12

where the subscript t indicates equilibrium variables in the case in which an ad varolem tari is imposed. Consumers surplus and tari revenue in Country N, TRN , in this case are respectively given by CSN t TRN t {(a )(2 + 3) t(2a + 2a )}2 = , 2(2 2 + 2(3 t) + 3(1 t))2 t{2a 2 + + (a + )(1 + 3) t(a + a + )}{(a )(1 + ) t(a + a + )} = . (2 2 + 2(3 t) + 3(1 t))2

N Social wefare in Country N in this case is given by WtN = t + CSN + TRN . t t

The eects of a small ad valorem tari are shown in the following proposition: Proposition 3 A small ad valorem tari on imports of goods by Country N has the usual rent-shifting eect. That is, it raises rm Ns prots and reduces rm Ss prots. It also improves Coutry Ns social welfare.
N Proof. Dierentiate t with respect to t and evaluate the derivative at t = 0 to yield N t t

=
t=0

2(a )(1 + 2)(2a 2 + 2(a + 2) + 3) > 0. (2 2 + 6 + 3)2

S Similarly, dierentiate t with respect to t and evaluate the derivative at t = 0 to obtain S t t

=
t=0

(a )(1 + )2 (4a 3 + 2(7a + 3) 2 + 2(7a + 8) + 3(a + 3)) < 0. (2 2 + 6 + 3)2

Moreover, dierentiate WtN with respect to t and evaluate the derivative at t = 0 to yield WtN t
5 4

=
t=0

(a ) > 0. (2 2 + 6 + 3)2

where 4a +2(8a+3) +4(9a+4) 3 +3(13a+10) 2 +6(3a+4)+3(a+2) > 0.

Proposition 3 shows that a small ad valorem tari in this context works in the same way as in the standard models of strategic trade policy.7 It has a strategic eect of rent-shifting and also can be used as a tool to improve domestic welfare.
7 With

respect to the eects of taris in the standard models of strategic trade policy, see, for example, Helpman and

Krugman (1989, Chapter 6).

13

4.2

Social clause taris

I now turn to the case in which the government of Country N imposes a social clause tari on imports from Country S. When a social clause tari is imposed, rm Ss eective wage is given by
S w = w S + (w S w S )

= (1 )w S + w S ,

(29)

where is a social clause tari rate and w S is the wage rate when the labour market in Country S is competitive, given the imperfectly competitive output market. w S is given by S w S = + yc a + 3 , = 3+ where yc is given by (21). S
S S When a social clause tari is imposed, rm Ss prot is given by = (p w )y S .

From the FOC, rm Ss reaction function in this case is given by y S (y N ) = (3 + )(a y N ) (a ) . 2(3 + )(1 + (1 )) (30)

Then, assuming interior solutions, outputs and prots in NE are respectively given by (a )(2(1 ) 2 + (7 5 ) + 3) , (3 + )(2(1 ) 2 + 2(3 2 ) + 3) (a )((1 ) 2 + 2(2 ) + 3) S y = , (3 + )(2(1 ) 2 + 2(3 2 ) + 3)
N y = N N = y 2

(31) (32) (33)

,
2

S S = (1 + (1 )) y

(34)

where the subscript indicates equilibrium variables in the case in which a social clause tari is imposed. Consumers surplus, tari revenue, and social welfare in Country N in this case are

14

respectively given by CSN = TRN WN (a )2 (3(1 ) 2 + (11 7 ) + 6)2 , 2(3 + )2 (2(1 ) 2 + 2(3 2 ) + 3)2 (a )(2 + ) 2 (1 ) = , (3 + )(2(1 ) 2 + 2(3 2 ) + 3) (a ) = , 2 (2(1 ) 2 + 2(3 2 ) + 3)2 2(3 + ) (35) (36) (37)

where (a )(3 + )2 (17 2 + 20 + 6) + 4 3(3 + )(2 + )2 3 2 (8 4 + 60 3 (17a 17 154) 2 (82a 82 150) 99a + 99 + 36) 2 + 2(3 + )(2 4 + 10 3 (17a 17 15) 2 (51a 51 6) 24(a )) . Using (31) to (37), I obtain the following proposition concerning the eects of a small social clause tari: Proposition 4 A small social clause tari by Country N reduces rm Ns prots, whereas it improves Country Ns social welfare. Moreover, it increases rm Ss profits if is low and reduces them if is high. It always raises employment by rm S.
N Proof. Dierentiate with respect to and evaluate the derivative at = 0 to yield N

=
=0

2(a )2 (1 + 2)(3 + 2) < 0. (3 + )(2 2 + 6 + 3)3 (a ) > 0, (3 + )(2 2 + 6 + 3)3

Dierentiate WtN with respect to and evaluate the derivative at = 0 to obtain WN =

=0

where 4 5 + 32 4 + 96 3 + 2(3a 3 + 61) 2 + (11a 11 + 81) + 3(a + 6) > 0.

S Moreover, dierentiate with respect to and evaluate the derivative at = 0 to

obtain

=
=0

(a )2 (1 + 2)2 (2 3 + 8 2 + 7 3) . (3 + )(2 2 + 6 + 3)3 (resp. <) 0 if (resp. >) , where 0.3101. Finally, (a )(3 + 2) > 0. (3 + )(2 2 + 6 + 3)2

It is shown that

S / =0

S dierentiate y with respect to and evaluate the derivative at = 0 to yield S y

=
=0

15

This proposition implies that, unlike the case of ad valorem taris, a small social clause tari does confer a strategic advantage on rm S by reducing its market power in the labour market. In fact, a small social clause tari allows rm S to commit to a higher employment level, which in turn means a higher output level, for a given level of rm Ns output. Because of this, the tari hurts rm N. Nevertheless, it improves social welfare in Country N because it increases consumers surplus and also generates tari revenue. The eect of a small social clause on consumers surplus can be seen by dierentiating (35) with respect to and evaluating the derivative at = 0: CSN =
=0

(a )2 (1 + )(2 + 3)(3 + 2) > 0. (3 + )(2 2 + 6 + 3)2

Despite the strategic advantage conferred by the tari, rm S does not always benet from a small social clause tari. This is because its mark-up is reduced by the tari,
S which can be conrmed by dierentiating p w with respect to and evaluating

the derivative at = 0: =
=0

(a )(1 + )(2 3 + 10 2 + 14 + 3) < 0. (3 + )(2 2 + 6 + 3)2

Proposition 4 also implies that a small social clause tari may yield a Pareto improvement in the sense that both countries are better o. This is possible because a small social clause tari improves the economic eciency of the global economy by mitigating the monopsony distortion in the labour market in Country S. Since social clause taris partly aim to correct the distortion in the labour market abroad, it may be argued that tari revenues should be rebated to the country of origin. In that case, tari revenue is not counted as part of Country Ns social welfare. However, a small social clause tari still improves Country Ns social welfare, which is conrmed
N by dierentiating + CSN with respect to and evaluating the derivative at = 0:

N + CSN 4.3 Labour standards

=0

(a )2 2 (1 + 3)(3 + 2) = > 0. (3 + )(2 2 + 6 + 3)3

In this subsection, I consider labour standards as a policy instrument. Suppose that certain labour standards are imposed in international trade agreements. Or, alterna16

tively, suppose that Country N unilaterally requires non-discriminatorily certain labour standards to imports of goods from abroad. In either story, the eect of imposing labour standards is to increase wage paid by rm S.8 Let w be the wage rate when rms comply with the labour standards. When the
S S S labour standards are binding for rm S, w wc , where wc + yc is the wage

rate paid by rm S when it practices social dumping under Cournot competition in the
S output market. Note that yc is given by (16).

When the labour standards are imposed, rm Ss prot maximization problem is expressed as max
yS

S = (p(y) w S (y S ))y S

s.t. w S (y S ) w.

From the FOC, rm Ss reaction function in this case is given by N a + (2 + ) 2(1 + )w ay , if y N , 2(1 + ) S N y (y ) = w a + (2 + ) 2(1 + )w , if y N > . the possibility of hidden violations of the standards.

(38)

I assume that rm S perfectly complies with the labour standards and does not consider

Assuming that the labour standards are not binding for rm N, the reaction function
S of rm N is not aected and hence is given by (14). When w wc holds, outputs and

prots in NE, where both rms stay in the market, are respectively given by a + (1 ) w w S , ys = , (39) (2 + ) (w ){(a w) 2 + (a + 2 3w) + w} N N 2 S s = ys , , (40) s = 2 (2 + )
N ys =

where the subscript s indicates equilibrium variables in the case under labour standards. Consumers surplus and social welfare in Country N in this case are given by CSN = s WsN
8 Corden

((a 2 + w) + w )2 , 2 2 (2 + )2 = , 2 (2 + )2 2

(41) (42)

and Vousden (2001) take a similar approach. They argue that the eects of raising minimum wages, improving

labour conditions, and allowing the operation of labour unions can be captured by an increase in the real wage.

17

respectively, where ( 2 + 2 + 3)w 2 + 2((a 2) 2 (a + ) 3)w + 2 (3a2 8a + 6c2 ) + 2a + 3 2 . Now, the following proposition shows the eects of the labour standards that
S marginally increases w S from wc .

Proposition 5 Requiring labour standards to rm S, which marginally increases w S

S from wc , is benecial to rm S and harmful to rm N. It, however, improves Coutry Ns

social welfare.
S S Proof. Dierentiate s with respect to w and evaluate the derivative at w = wc to

obtain

S s w

=
S w=wc

(a )(1 + ) > 0. (2 + )(2 2 + 6 + 3)

N S Similarly, dierentiate s with respect to w and evaluate the derivative at w = wc to

obtain

N s w

S w=wc

2(a )(1 + 2) < 0. (2 + )(2 2 + 6 + 3)

S Moreover, dierentiate WsN with respect to w and evaluate the derivative at w = wc to

obtain

WsN w

=
S w=wc

(a )(1 + 3) > 0. (2 + )(2 2 + 6 + 3)

Proposition 5 implies that, similarly to the case of social clause taris, the imposition of labour standards confers a strategic advantage on rm S. This is because it allows rm S to commit to a higher employment level than the level in the monopsony case. Firm N suers from labour standards, because it responds to an increase in y S by reducing y N . Figure 2 depicts the case of labour standards. The thin line of RS indicates rm Ss reaction curve without labour standards. The thick line that bends at Ec indicates
S S rm Ss reaction curve with labour standards of w = wc , which is denoted by Rs . The

Nash equilibrium point remains at Ec . As labour standards are raised from the level
S S corresponding to w = wc , the at part of Rs shifts up, as is drawn by the thick dotted

18

line in the gure. As a result, the Nash equilibrium point moves to Es , yielding a lower y N and a higher y S . Unlike the case of social clause taris, however, rm S benets from labour standards
S that marginally increases w S from wc , regardless of the value of . The reason is that in

the case of labour standards rm S can commit to a certain level of employment without worrying about the response in the wage rate. In the case of social clause taris, by contrast, the magnitude of the response in the wage rate for changing the employment level still matters. As a result, labour standards have a stronger commitment eect and hence are benecial to rm S, regardless of . Similarly to the case of social clause taris, Country N is better o by requiring labour standards, despite the loss in rm Ns prots. This is because the gain in consumers surplus dominates the loss in rm Ns prots.

Conclusions

To be concluded.

19

References
[1] Abe, Kenzo and Zhao, Laixun. 2005. International labor standards and their harmonization. Unpublished manuscript. [2] Bagwell, Kyle and Staiger, Robert W. 1999. The simple economics of labor standards and the GATT. In Deardor, A. V. and Stern, R. M., eds. Social Dimensions of the U.S. Trade Policies. Ann Arbor, MI: University of Michigan Press [3] Bagwell, Kyle and Staiger, Robert W. 2001. The WTO as a mechanism for securing market access property rights: implications for global labor and environmental issues. Journal of Economic Perspectives 15: 69-88. [4] Barrett, Scott. 1994. Strategic environmental policy and international trade. Journal of Public Economics 54: 325-338. [5] Bhagwati, Jagdish N. 1995. Trade liberalization and fair trade demands: addressing the environmental and labor standards issues. World Economy 18: 745-759. [6] Bhagwati, Jagdish N., Panagariya, Arvind, and Srinivasan, T. N. 1998. Lectures on International Trade, 2nd edition. Cambridge, MA: MIT Press. [7] Bhaskar, V., Manning, Alan, and To, Ted. 2002. Oligopsony and monopsonistic competition in labor markets. Journal of Economic Perspectives 16: 155-174. [8] Boal, William M. and Ransom, Michael R. Monopsony in the labor market. Journal of Economic Literature 35: 86-112. [9] Brown, Drusilla K. 2001. Labor standards: where do they belong on the international trade agenda? Journal of Economic Perspectives 15: 89-112. [10] Brown, Drusilla K., Deardor, Alan V., and Stern, Robert M. 1996. International labor standards and trade: a theoretical analysis. In Bhagwati, J. N. and Hudec, R. E., eds., Fair Trade and Harmonization: Prerequisites for Free Trade?, vol. 1: Economic Analysis. Cambridge, MA: MIT Press, pp. 227-280.

20

[11] Brown, Drusilla K., Deardor, Alan V., and Stern, Robert M. 1998. Trade and labor standards. Open Economies Review 9: 171-194. [12] Brown, Drusilla K., Deardor, Alan V., and Stern, Robert M. 2003. The eects of multinational production on wages and working conditions in developing countries. NBER Working Paper No. 9669. [13] Conrad, Klaus. 1993. Taxes and subsidies for pollution-intensive industries as trade policy. Journal of Environmental Economics and Management 25: 121-135. [14] Cordella, Tito and Grilo, Isabel. 2001. Social dumping and relocation: is there a case for imposing a social clause? Regional Science and Urban Economics 31: 643-668. [15] Corden, W. Max and Vousden, Neil. 2001. Paved with good intentions: social dumping and raising labour standards in developing countries. In Drabek, Z. ed. Globalisation Under Threat. Cheltenham: Edward Elgar. pp. 124-143. [16] Dehejia, Vivek H. and Samy, Yiagadeesen. 2004. Trade and labour standards: theory and new empirical evidence. Journal of International Trade and Economic Development 13: 179-198. [17] Devadoss, Stephen and Song, Wongun. 2003a. Oligopsonistic intermediate input and patterns of trade. International Economic Journal 17: 77-97. [18] Devadoss, Stephen and Song, Wongun. 2003b. Factor market oligopsony and the production possibility frontier. Review of International Economics 11: 729-744. [19] Feenstra, Robert C. 1980. Monopsony distortions in an open economy: a theoretical analysis. Journal of International Economics 10: 213-235. [20] Golub, Stephen S. 1997. Are international labor standards needed to prevent social dumping? Finance and Development 34: 20-23. [21] Helpman, Elhanan and Krugman, Paul R. 1989. Trade Policy and Market Structure. Cambridge, MA: MIT Press.
21

[22] Kennedy, Peter W. 1994. Equilibrium pollution taxes in open economies with imperfect competition. Journal of Environmental Economics and Management 27: 49-63. [23] Kok, Michiel, Nahuis, Richard, and de Vaal, Albert. 2004. On labour standards and free trade. Journal of International Trade and Economic Development 13: 137-158. [24] Kuroda, Tomohiro. 2004. Local content protection reconsidered: the case of domestic monopsonist. Economics Bulletin 6: 1-9. [25] Leahy, Dermot and Montagna, Catia. 2000. Temporary social dumping, union legalisation and FDI: a note on the strategic use of standards. Journal of International Trade and Economic Development 9: 243-259. [26] Leary, Virginia A. 1996. Workers rights and international trade: the social clause (GATT, ILO, NAFTA, U.S. laws). In Bhagwati, J. N. and Hudec, R. E., eds., Fair Trade and Harmonization: Prerequisites for Free Trade?, vol. 2: Legal Analysis. Cambridge, MA: MIT Press, pp. 177-230. [27] McCulloch, Rachel and Yellen, Janet L. 1980. Factor market monopsony and the allocation of resources. Journal of International Economics 10: 237-247. [28] Markusen, James R. and Robson, Arthur J. 1980. Simple general equilibrium and trade with a monopsonized sector. Canadian Journal of Economics 13: 668-682. [29] Martin, Will and Keith E. Maskus. 2001. Core labor standards and competitiveness: implications for global trade policy. Review of International Economics 9: 317-328. [30] Maskus, Keith E. 1997. Should core labor standards be imposed through international trade policy? World Bank Policy Research Working Paper No. 1817. [31] Naghavi, Alireza. 2005. Asymmetric labor markets, southern wages, and the location of rms. Fondazione Eni Enrico Mattei. Nota di Lavaro 17.2005.

22

[32] Okuguchi, Koji. 1998. Existence of equilibrium for Cournot oligopoly-oligopsony. Keio Economic Studies 35: 45-53. [33] Okuguchi, Koji. 2000. Comparative statics for oligopoly, oligopsony and oligopsonistic oligopoly. Journal of Economic Research 5: 173-187. [34] Rauscher, Michael. 1994. On ecological dumping. Oxford Economic Papers 46: 822-840. [35] Shelburne, Robert C. 2004. Wage dierentials, monopsony labor markets, and the trade-labor standards debate. Journal of Economic Integration 19: 131-161. [36] Sinn, Hans-Werner. 2001. Social dumping in the transformation process? NBER Working Paper Series No. 8364. [37] Srinivasan, T. N. 1995. International trade and labor standards. In van Dijck, P. and Faber, G. eds. Challenges to the New World Trade Organization. Amsterdam: Martius Nijho/Kluwer. [38] Suranovic, Steven. 2002. International labour and environmental standards agreements: is this fair trade? World Economy 25: 231-45.

23

yS
a
a 2+ a 2(1+)

RN

~S yc ycS

~ Ec Ec ~N y N yc c
a 2+

~ RS RS
a

yN

Figure 1: Reaction curves with and without rm Ss oligopsonistic power

24

yS
a
a 2(1+)
S ` (w)/ =ys ` S (w)/ =yc

RN

Es` Ec ysN `ycN

a 2+

S Rs ` S Rs

RS
a

yN

Figure 2: The eects of labour standards

25