International migration of labour and skilled –unskilledwage inequality in a developing economyShigemi Yabuuchi a ,Sarbajit Chaudhuri b,⁎a Department of Economics,Nagoya City University,Yamanohata,Mizuho,Nagoya 467-8501,Japanb Department of Economics,University of Calcutta,56A,B.T.Road,Kolkata 700050,IndiaAccepted 7June 2006AbstractThe paper develops a three sector general equilibrium structure with diverse trade pattern and imperfection in the unskilled labour market to analyze the consequences of international mobility of skilled and unskilled labour on the skilled –unskilled wage inequality in the developing economies.It shows that the effects of international migration of labour on the wage inequity crucially depend on both the relative capital intensities between the low-skill and high-skill sectors and the institutional nature of the markets for unskilled labour.The analysis finds that an emigration (immigration)of either type of labour is likely to produce a favourable (an unfavourable)effect on the wage inequality.In particular,the result of emigration (immigration)of skilled labour on the relative wage inequality is counterintuitive.These results have important policy implications for an overpopulated developing country like India.©2006Elsevier B.V .All rights reserved.JEL classification:F13;J31Keywords:Skilled labour;Unskilled labour;Wage inequality;Emigration (immigration)of labour;Labour market imperfection;Diverse trade pattern1.IntroductionThe last two decades have witnessed a rapid growth of the global economy,reflected in reduced trade barriers,increased international trade,highly mobile capital and labour andthe Economic Modelling 24(2007)128–137⁎Corresponding author.23Dr.P.N.Guha Road,Belgharia,Kolkata 700083,India.Tel.:+91335410455,+91335575082;fax:+913328441490.E-mail addresses:yabuuchi@econ.nagoya-cu.ac.jp (S.Yabuuchi),sarbajitch@ ,sceco@caluniv.ac.in (S.Chaudhuri).0264-9993/$-see front matter ©2006Elsevier B.V .All rights reserved.doi:10.1016/j.econmod.2006.06.006rapid transmission of technology across national borders.Globalization perpetuates emigration from developing countries in the following way.It stimulates consumerism and consumption and raises expectations regarding the standard of living.The widening gap between consumption expectations and the available standard of living within structural constraints of the developing countries,combined with easy access to information and migration networks,in turn create tremendous pressure for emigration.Trade liberalization in the less developed countries,according to the conventional wisdom,was expected to lower the skilled –unskilled wage inequality following increases in the prices of the export commodities as these are generally exporters of commodities that are intensive in the use of unskilled labour.But empirical studies 1strongly suggest that the wage inequality has increased in many Latin American and South Asian countries including India.The scanty theoretical literature explaining the deteriorating wage inequality in the Southern countries includes works of Feenstra and Hanson (1996),Marjit,Broll and Sengupta (2000),Marjit,Beladi and Chakrabarti (2004)and Chaudhuri and Yabuuchi (in press).They have shown how trade liberalization and inflows of foreign capital might produce unfavourable effects on the wage inequality in the South given the specific structural characteristics of the less developed countries,such as features of labour markets,structures of production,nature of capital mobility etc.Unfortunately,economists have so far paid very little attention in analyzing the consequences of emigration of workers from developing economies on the skilled –unskilled wage rge-scale international migration of workers from a developing country,irrespective of whether skilled or unskilled,is expected to produce significant effects on the wage inequality.An exception 2in this regard is the paper of Marjit and Kar (2005)which has examined the consequence of emigration of skilled and unskilled labour on the wage inequality in an otherwise 2×3specific factor model of Jones (1971).They have shown that unskilled (skilled)emigration worsens (improves)the wage inequality under the necessary and sufficient condition that the distributive share of the intersectorally mobile factor (i.e.capital)of the skilled sector is greater (lower)than that of the unskilled sector.Their results point out an important implication between emigration and the wage inequality.However,these results are completely reversed if the relative distributive shares of capital are opposite.Besides,the assumption that both the sectors use the same type of capital may not be quite realistic in the context of a developing economy.3Moreover,labour market imperfection,especially that of unskilled labour and diverse trade pattern,which are the two salient features of the developing economies have not been taken care of in their model.One of the prominent features of the developing economies is the existence of imperfection in the unskilled labour market.Unskilled workers are employed in different sectors of a developing economy.Workers employed in the organized (formal)sectors receive relatively high contractual 1See Robbins (1995,1996a,b),Wood (1997),Khan (1998)and Tendulkar et al.(1996)in this context.2However,mention should be made of another paper by Kar and Beladi (2004)who in terms of a Heckscher –Ohlin –Samuelson-specific factor model have shown that both skilled and unskilled emigration from low-income countries,as an outcome of trade liberalization or stronger wage bargaining by ‘skilled ’labour unions,lower the wage inequality unambiguously and independently of factor intensity assumptions.However,the focus of the paper is not exclusively on the consequence of international migration of skilled/unskilled labour on wage inequality but to analyze the consequences of skill formation and migration of skilled labour on economic welfare.3In the literature on trade and development,a developing economy is typically depicted as an exporter of primary agricultural commodities and an importer of manufacturing goods.The production activities of these two types of commodities are entirely different and require two different types of capital.Hence,in a two-sector small open economy setting the assumption of homogeneous capital may be a limitation.However,in a higher dimensional setup with diverse trade pattern like the present one the use of homogeneous capital in the two non-agricultural sectors may be justified.129S.Yabuuchi,S.Chaudhuri /Economic Modelling 24(2007)128–137130S.Yabuuchi,S.Chaudhuri/Economic Modelling24(2007)128–137(unionized)wage while their counterparts engaged in the informal sector earn a lower competitive wage.The unionized wage is positively related to the competitive informal wage.As unskilled workers earn two different wages in the two different segments of the labour market,the average unskilled wage should be a weighted average of the two wages with weights being the proportions of unskilled labour employed in the two segments of the unskilled labour market.There are theoretical papers e.g.Carruth and Oswald(1981),Agenor and Montiel(1995),Marjit and Beladi (2002)and Marjit(2003)which have dealt with labour market distortion in the developing countries for different purposes.The purposes of the present paper are to construct a three sector general equilibrium structure which incorporates the diverse trade pattern and the imperfection that exists in the unskilled labour market of the developing countries and provide a sound theoretical foundation that can be used to analyze the consequences of rapidly increasing international mobility of different factors of production,trade related policies and also policies like labour market reform on the skilled–unskilled wage inequality in the liberalized regime.The analysis finds that an emigration (immigration)of either type of labour is likely to produce a favourable(an unfavourable)effect on the wage inequality.In particular,the result of emigration(immigration)of skilled labour on the relative wage inequality is counterintuitive.It also shows that that the effects of international migration of labour on the wage inequity depend not only on the relative capital intensities between the low-skill and high-skill sectors but also on the institutional nature of the markets for unskilled labour.Therefore,unlike Marjit and Kar(2005),the results of the present paper may not change completely with the relative capital intensity conditions.The results have important policy implications for an overpopulated developing country like India.2.The model and resultsWe consider a small open developing economy with three sectors.Sector1produces a primary agricultural commodity,X1using unskilled labour,L,and land,N.Sector2produces a high-skill manufacturing commodity,X2,with the help of skilled labour,S,and capital,K.Sector3uses unskilled labour and capital to produce a low-skill manufacturing product,X3.So land and skilled labour are specific factors in sectors1and2,respectively.Capital is mobile between sectors2and 3.Unskilled workers employed in Sector3earn a unionized wage,W*,while their counterparts in the agricultural sector earn a competitive wage,W,with W*>W.All other markets are perfectly competitive.The diverse trade pattern of the economy is reflected in the fact that it exports the primary agricultural and the high-skill manufacturing commodities while it is a net importer of the low-skill manufacturing commodity.Product prices,P i;(i=2,3),are given internationally. Commodity1is chosen as the numeraire.A general equilibrium of the system is represented by the following set of equations:Wa L1þRa N1¼1ð1ÞW S a S2þra K2¼P2ð2ÞW*a L3þra K3¼P3ð3Þa N1X1¼Nð4Þa S 2X 2¼Sð5Þa L 1X 1þa L 3X 3¼Lð6Þa K 2X 2þa K 3X 3¼K ð7Þwhere a ji is the factor coefficient of factor j in sector i ,W S is the wage rate of skilled labour,R is rental to land,r is rental to capital.Eqs.(1),(2)and (3)are the three competitive industry equilibrium conditions in the three sectors.On the other hand,Eqs.(4)–(7)are the full-employment conditions of the four factors of production.The formal sector faces a unionized labour market.The relationship for the unionized wage rate is specified as 4:W *¼W *ðW ;U ÞðþÞðþÞð8Þwhere U is a parameter denoting the extent of bargaining power of the trade union.W *(.)satisfies the following properties:W *=W for U =0,W *>W for U >0;and,E W =((∂W */∂W )/(W /W *))≤1.Eq.(8)states that in the absence of any bargaining power of the trade unions 5,the unskilled wage rates are equal in sectors 1and 3.However,the unionized unskilled wage rate in sector 3,W *,exceeds the competitive unskilled wage rate,W ,when there is at least some power to the trade unions.The unionized wage is scaled upward as the competitive wage rate rises.Also with an increase in the bargaining power,the unions bargain for a higher wage.6There are eight endogenous variables in the system:W ,W *,W S ,R ,r ,X 1,X 2and X 3.The parameters of the system are:P 2,P 3,U ,N ,K ,L and S ,which are exogenously given.It is easy to note that this production structure does not possess the decomposition property.From Eqs.(4)–(7),it is easy to derive the following equation.½fða L 1=a N 1ÞN g þða L 3=a K 3Þf K −ða K 2=a S 2ÞS g ¼L :ð9ÞThe five input prices,W ,W S ,W *,R and r are determined by solving Eqs.(1–3)(8)and (9simultaneously.Once the factor prices are known the factor coefficients,a ji s ,are also known.X 1and X 2are obtained from Eqs.(4)and (5),respectively.Finally,X 3is found from either Eq.(6)or(7).Unskilled workers in this system earn two different wages —either the unionized wage,W *,in sector 3or a lower competitive wage,W ,in sector 1.The average wage for unskilled labour is given byW A u ðW k L 1þW *k L 3Þð10Þwhere λL 1and λL 3denote the proportion of unskilled labour employed in sectors 1and 3,respectively.4The unionized wage function may be derived as a solution to the Nash bargaining game between the representative firm and the representative union in the low-skill manufacturing sector.For detailed derivation see Chaudhuri (2003).5The union power,U ,is amenable to policy measures.If the government undertakes labour market reform measures e.g.ban on resorting to strikes by the trade unions,reformation of employment security laws to curb union power,U takes a lower value.6For a better modeling of union behaviour one may go through Carruth and Oswald (1981).Using a two sector general equilibrium model with wage differential they have shown that an increase in the union wage raises the non-union wage in a small open economy and that this result is independent of the ranking of the sectors in terms of factor intensities.131S.Yabuuchi,S.Chaudhuri /Economic Modelling 24(2007)128–137Now,let us investigate the effects of changes in the endowments of skilled and unskilled labour on the wage parative statics yields(see Appendix I):W¼ðh N1h S2h K3B=DÞðLþGSÞ;ð11ÞW S¼ðE W h N1h K2h L3=DÞðLþGSÞ;ð12ÞW*¼ðE W h N1h K3h S2=DÞðLþGSÞ;ð13ÞR¼Àðh L1h S2h K3=DÞðLþGSÞ;ð14Þwhere‘∧’shows proportionate change of the variables,θji andλji are distributive and allocative share of factor j in sector i,respectively,Δis the determinant of the coefficient matrix of the derived equational system,D¼h S2h K3ðh N1A−h L1EÞþh N1E Wðh K2h L3Bþh S2h K3C−h S2h L3DÞ<0;ð15Þand A=λL1(S LL1−S NL1)<0,B=−(λK2λL3/λK3)(S KS2−S SS2)<0,C=λL3(S LL3−S KL3)<0,D=λL3 {(S LK3−S KK3)−(λK2/λK3)(S KK2−S SK2)}>0,E=λL1(S LN1−S NN1)>0,G=(λL3λK2/λK3)>0.S ji k is the degree of substitution between factors j and i in the k th sector,j,i=L,N,S,K;and,k=1,2,3. For example,S LK1≡(R/a L1)(∂a L1/∂R),S LL1≡(W/a L1)(∂a L1/∂W)etc.S ji k>0for j≠i.On the other hand,totally differentiating Eq.(10),it is easy to find that(see Appendix II) W A¼f aþgðS1LLÀS1NLÞg Wþð1ÀaÞW*þgðS1LNÀS1NNÞRÀg L;ð16Þwhereα=(WλL1/W A)>0;and,γ=((W−W*)λL1/W A)<0(as W<W*).Using Eqs.(11–14),one can show that(see Appendix III)ðWS ÀWAÞ¼ðG=DðÀÞÞS½E W h N1ðh K2h L3Àh K3h S2ÞþaðþÞh N1h S2h K3ðE WÀ1ÞðV0ÞþgðÀÞðS1LNÀS1NNÞðþÞh S2h K3þðL=DðÀÞÞ½E W h N1ðh K2h L3Àh K3h S2ÞþaðþÞh N1h S2h K3ðE WÀ1ÞðV0ÞþgðÀÞfðS1LNÀS1NNÞðþÞh S2h K3þDðÀÞg ð17ÞIt can be shown that(θK3/θL3)≥(θK2/θS2)is equivalent toθK3≥θK2sinceθK3+θL3=1and θK2+θS2=1.From Eq.(17)we can now establish the following proposition.Proposition1.An emigration(immigration)of skilled labour improves(worsens)the skilled–unskilled wage inequality if the capital's share is greater in the low-skill sector3than that in the high-skill sector2i.e.θK3≥θK2while an emigration(immigration)of unskilled labour improves (worsens)the wage inequality if(i)θK3≥θK2;and if,(ii)(E W≤λL3θS2θK3A/λL1θN1(θK2θL3B+θS2θK3C−θS2θL3D).We note that the conditions,as stated in Proposition1,are only sufficient in nature.But,the role of the labour unions is not clearly reflected in the conditions.However,from Eq.(17)it is quite possible to find out other sufficient conditions for the same results to hold where the response of the union is much more pronounced.For example,the result on emigration of skilled labour holds under the sufficient condition,[E WθN1θK2θL3+γ(S LN1−S NN1)θS2θK3]≤0,while that of unskilled labour remains valid if[E WθN1θK2θL3+γ(S LN1−S NN1)θS2θK3+Δ]≤0.We note thatγ<0and its 132S.Yabuuchi,S.Chaudhuri/Economic Modelling24(2007)128–137absolute value increases with an increase in the unskilled wage differential between the two sectors implying a greater bargaining strength of the trade union in sector 3.Hence,the possibility of the results of the present analysis to hold increases with an increase in the strength of the labour union to mark up wage over the competitive unskilled wage.Another point to be noted in this context is that in the absence of any distortion in the market for unskilled labour i.e.W *=W ,sector 1does no longer have any operational significance and the model reduces to the Marjit and Kar (2005)model.In this situation,the results only depend on the relative factor intensities.Proposition 1can now be intuitively explained as follows.An emigration of skilled labour leads to an increase in the skilled wage rate,W S .To satisfy the zero profit condition in sector 2,the return to capital,r ,falls.Producers in sector 2substitute capital for skilled labour.So,a S 2falls and a K 2rises.As r falls given the relative price of commodity 3,the unionized unskilled wage,W *,rises to satisfy the zero profit condition in sector 3.But,W *can increase only if the competitive unskilled wage,W ,rises.The return to land,R ,falls as a consequence.Producers in sector 1adopt more (less)land (unskilled labour)intensive techniques than before which in turn implies a contraction of sector 1both in terms of output and employment of unskilled labour,as land is specific to this sector.The released workers from sector 1are now absorbed in sector 3.Sector 3expands both in terms of output and employment.Thus,we find that the average unskilled wage increases due to (i)an increase in W ;(ii)an increase in W *;and,due to (iii)an increase (a decrease)in the proportion of unskilled labour employed in the higher (lower)wage-paying sector i.e.λL 3(λL 1).Therefore,the average unskilled wage,W A ,rises unambiguously.What happens to the skilled –unskilled wage inequality depends on the rates of increase in W S and W A .If θK 3>(=)θK 2the savings on capital cost in the low-skill manufacturing sector is more than (equal to)that in the high-skill sector,which in turn,implies that the rate of increase of the unionized unskilled wage,W *,is greater than (equal to)that of the skilled wage,W S .But,as we have mentioned above,that there are other two factors working positively on the average unskilled wage.Thus,the wage inequality improves under the sufficient condition as mentioned in Proposition 1.The presence of the other two factors implies that proposition 1may hold even under situations where the above sufficient condition does not hold.The effect of an immigration of skilled labour can be explained exactly in the opposite way.On the other hand,an emigration of unskilled labour raises the competitive unskilled wage,W ,in sector 1.The return to land decreases.Consequently,producers in sector 1increase (lower)the per unit use of land (unskilled labour).Sector 1contracts both in terms of output and employment.As W increases the unionized unskilled wage,W *,in sector 3also increases.So,the return to capital,r ,falls to satisfy the zero profit condition.This in turn raises the skilled wage,W S ,in sector 2.In both sectors 2and 3the capital intensity of production increases.Sector 2expands as it employs a specific factor,skilled labour.Sector 3releases capital to the expanding sector 2.Sector 3contracts both in terms of output and employment.What happens to the proportions of unskilled labour in sector 1and 3(λL 1and λL 3)is somewhat uncertain as the endowment of unskilled labour has fallen.At least,it can be shown that λL 1falls (λL 3rises)if and only if E W ≤λL 3θS 2θK 3A /λL 1θN 1(θK 2θL 3B +θS 2θK 3C −θS 2θL 3D ).However,this is only a sufficient condition for W A to increase as both W and W *have increased.Now,the rate of increase in W A is greater than that in W S under another sufficient condition that the low-skill manufacturing sector is not less capital intensive (in a special sense)7relative to the high-skill sector.Consequently,the wage inequality improves under the above two sufficient conditions.The consequence of an immigration of unskilled labour can be explained in the opposite manner.7See Jones and Neary (1984)in this context.133S.Yabuuchi,S.Chaudhuri /Economic Modelling 24(2007)128–137134S.Yabuuchi,S.Chaudhuri/Economic Modelling24(2007)128–137Thus,we have found that the outcomes of international migration of labour on the skilled–unskilled wage inequality depend not only on the relative capital intensities of the low-skill and high-skill sectors but also on the institutional nature of the markets for unskilled labour.Our results shows that in the present setup(i)the condition by Marjit and Kar is only a sufficient condition in the case of skilled emigration,and(ii)their condition is neither sufficient nor necessary in the case of unskilled emigration.This is why unlike the Marjit and Kar(2005)paper the effects on wage inequality here may not necessarily be the opposite under the two alternative capital intensity conditions.3.Concluding remarksIn this paper we have developed a three sector specific factor general equilibrium framework incorporating some of the essential characteristics of the developing economies e.g.market distortions of unskilled labour,diverse trade pattern etc.to analyze the consequences of emigration(immigration)of both skilled and unskilled labour on the skilled–unskilled wage inequality in the developing countries.The issue is important as the international mobility of labour has significantly increased in the liberalized regime and it has not so far drawn sufficient attention of the trade and development economists.8We have shown that the effects of international migration of labour on the wage inequality depend not only on the relative capital intensities of the low-skill and high-skill sectors but also on the institutional nature of the markets for unskilled labour and that an emigration(immigration)of either type of labour is likely to produce a favourable(an unfavourable)effect on the wage inequality.In particular,the result of emigration(immigration)of skilled labour on wage inequality is interesting and counterintuitive.9 Also,the results have important policy implications for all migrant receiving developing countries including India where illegal immigration of people(unskilled labour)from neighboring poor countries through border areas is a serious and mounting problem.10The governments of such countries have so far remained indifferent to such an influx of people for petty political interests. But,the present analysis finds that the governments should take appropriate measures to prevent the immigration of unskilled labour11in order to improve the skilled–unskilled wage inequality. AcknowledgementThe authors are indebted to two anonymous referees of this journal for their interesting and constructive comments on an earlier version of the paper.The usual disclaimer,however,applies.8It should,however,be mentioned that there exists a voluminous literature((e.g.Bhagwati and Rodriguez,1975, Rivera-Batiz(1982,1984),Thompson(1984),Djajic(1986),Quibria(1988))that examines the consequence of international migration of labour on welfare of the non-migrants.However,this literature does not make any distinction between migrant and non-migrant labour force from the viewpoint of skill and hence has not analyzed the outcome of emigration of labour on the skilled–unskilled wage inequality.9An interesting theoretical exercise might be to bring in the active role of labor unions in wage setting in the presence of skilled–unskilled emigration and analyze the consequence of labour market reform on wage inequality.We are thankful to one of the two anonymous referees for pointing this out.This has been done in Chaudhuri and Yabuuchi(in press).The paper finds that contrary to the common wisdom,a policy of labour market reform may raise the competitive unskilled wage and improve wage inequality under reasonable condition.10It is a well-known argument that unskilled(and illegal)immigration increases wage inequality by depressing local unskilled wage.This has been a serious and mounting problem in all migrant receiving countries.11Measures to encourage emigration of skilled labour also produce the same result.But such a policy may not be desirable from the view point of national welfare.Appendix A.Derivation of Eqs.(11)–(14)Totally differentiating Eqs.(1–3),(8)and (9)and using envelope conditions we get the following expressions in the matrix form 12h L 1000h N 10h S 20h K 2000h L 3h K 30−E W0100A B C D E 266664377775W W S W *r R 266664377775¼0000ðL þG S Þ266664377775;ðA1Þwhere ‘∧’shows proportionate change of the variables,θji and λji are distributive and allocative shares of factor j in sector i ,respectively,A =λL 1(S LL 1−S NL 1)<0,B =−(λK 2λL 3/λK 3)(S KS 2−S SS 2)<0,C =λL 3(S LL 3−S KL 3)<0,D =λL 3{(S LK 3−S KK 3)−(λK 2/λK 3)(S KK 2−S SK 2)}>0,E =λL 1(S LN 1−S NN 1)>0,G =(λL 3λK 2/λK 3)>0.S ji k is the degree of substitution between factors j and i in the k th sector,j ,i =L ,N ,S ,K ;and,k =1,2,3.For example,S LK 1≡(R /a L 1)(∂a L 1/∂R ),S LL 1≡(W /a L 1)(∂a L 1/∂W )etc.l S ji k >0for j ≠i ;and,S jj k <0.Solving the derived system (A1)by using the Cramer's rule one getsW ¼ðh N 1h S 2h K 3B =D ÞðL þGS Þ:ð11ÞOther results can be obtained in the similar manner.Appendix B.Derivation of Eq.(16)Eq.(10)can be rewritten asW A ¼k L 1W þk L 3W *¼k L 1W þð1−k L 1ÞW *;ðA2Þsince λL 1+λL 3=1by definition.Totally differentiating (A2),we havedW A ¼k L 1dW þk L 3dW *þWd k L 1−W *d k L 1:ðA3ÞSince λL 1=a L 1X 1/L ,we obtain using (A3)k L 1¼S 1LL W þS 1LN R þX 1ÀL ;ðA4ÞOn the other hand,from Eq.(A4),we haveX 1¼Àa N 1¼ÀðS 1NL W ÀS 1NN R Þ:ðA5ÞThus,substituting (A4)and (A5)into (A3),we obtainW A ¼a þg ðS 1LL ÀS 1NL ÞÈÉW þð1Àa ÞW *þg ðS 1LN ÀS 1NN ÞR Àg Lð16Þwhere α=W λL 1/W A and γ=(W −W *)/W A <0as W <W *.12Detailed derivations can be obtained from the authors on request.135S.Yabuuchi,S.Chaudhuri /Economic Modelling 24(2007)128–137Appendix C.Derivation of Eq.(17) Using(11)–(14)and(16),one can write.ðWS ÀWAÞ¼ðE W h N1h K2h L3=DÞðLþGSÞÀð1ÀaÞðE W h N1h K3h S2=DÞðLþGSÞÀaþgðS1LLÀS1NLÞÈÉðh N1h S2h K3B=DÞðLþGSÞþgðS1LNÀS1NNÞðh L1h S2h K3=DÞðLþGSÞþg LðA6ÞUsing the relations that(S LL1+S LN1=0=S NN1+S NL1)we can rewrite(A6)as follows.ðWS ÀWAÞ¼ðG=DÞS½E W h N1ðh K2h L3Àh K3h S2Þþah N1h S2h K3ðE WÀ1ÞþgðS1LNÀS1NNÞh S2h K3ðh N1þh L1Þ þðL=DÞ½E W h N1ðh K2h L3Àh K3h S2Þþah N1h S2h K3ðE WÀ1ÞþgðS1LNÀS1NNÞh S2h K3ðh N1þh L1Þþg D ðA7ÞA little manipulation yields the following expression.ðWS ÀWAÞ¼ðG=DðÀÞÞS½E W h N1ðh K2h L3Àh K3h S2ÞþaðþÞh N1h S2h K3ðE WÀ1ÞðV0ÞþgðÀÞðS1LNÀS1NNÞðþÞh S2h K3þðL=DðÀÞÞ½E W h N1ðh K2h L3Àh K3h S2ÞþaðþÞh N1h S2h K3ðE WÀ1ÞðV0ÞþgðÀÞfðS1LNÀS1NNÞðþÞh S2h K3þDðÀÞg ð18ÞReferencesAgenor,P.R.,Montiel,P.,1995.Development Macroeconomics.Princeton University Press,New Jersey. Bhagwati,J.N.,Rodriguez,C.,1975.Welfare-theoretical analyses of the brain drain.Journal of Development Economics 2,195–221.Carruth,A.A.,Oswald,A.J.,1981.The determination of union and non-union wage rates.European Economic Review16, 285–302.Chaudhuri,S.,2003.How and how far to liberalize a developing economy with informal sector and factor market distortions.Journal of International Trade and Economic Development12(4).Chaudhuri,S.,Yabuuchi,S.,in press.Economic liberalization and wage inequality in the presence of labour market imperfection.International Review of Economics and Finance.Djajic,S.,1986.International migration,remittances and welfare in a dependent economy.Journal of Development Economics21,229–234.Feenstra,R.C.,Hanson,G.H.,1996.Foreign investment,outsourcing and relative wages.In:Feenstra,R.,Grossman,G., Irwin,D.(Eds.),Political Economy of Trade Policies:Essays in Honor of J.N.Bhagwati.MIT Press,Cambridge,MA. Jones,Ronald,1971.A three-factor model in theory,trade and history.In:Bhagwati,J.,et al.(Eds.),Trade,Balance of Payments and Growth.North-Holland,Amsterdam,pp.3–21.Jones,R.W.,Neary,P.,1984.Positive theory of international trade.In:Jones,R.W.,Kenen,P.B.(Eds.),Handbook of Development Economics,vol.1.North-Holland,pp.1–62.Kar,S.,Beladi,H.,2004.Skill formation and international migration:welfare perspective of developing countries.Japan and the World Economy16(1),35–54.Khan,A.R.,1998.The impact of globalization in South Asia.In:Bhalla,A.S.(Ed.),Globalization,Growth and Marginalization.Macmillan.Marjit,S.,2003.Economic reform and informal wage—a general equilibrium analysis.Journal of Development Economics72,371–378.Marjit,S.,Kar,S.,2005.Emigration and wage inequality.Economics Letters88,141–145.136S.Yabuuchi,S.Chaudhuri/Economic Modelling24(2007)128–137。