Foreign Investment, Urban Unemployment, and Informal Sector

Transcription

1 Journal of Econoic Integration 20(1), March 2005; Foreign Investent, Urban Uneployent, and Inforal Sector Shigei Yabuuchi Nagoya City University Haid Beladi North Dakota State University Gu Wei Kansai Research Institute Abstract In this paper we extend the Harris-Todaro odel to include an urban inforal sector. The effects of exogenous inflow of foreign capital on outputs, uneployent and welfare are exained in the odel where the product of the inforal sector is used as an industrial input in the urban foral sector. We also analyze the iplications of the capital inflow for incoe distribution. JEL Classification: F21, R23, O17 Key words: Harris-Todaro odel; Foreign investent; Welfare I. Introduction The relationship between protection and capital accuulation has often been discussed, particular attention being paid to its welfare iplication. Johnson (1967) and Bhagwati (1958, 1968, 1973) deonstrated the possibility of iiserizing growth under soe distorted situations. The essence of their results is now well Corresponding address: Shigei Yabuuchi, Departent of Econoics Nagoya City University Yaanohata, Mizuho Nagoya Japan Tel/Fax: Eail: nagoya-cu.ac.jp 2005-Center for International Econoics, Sejong Institution, All Rights Reserved.

2 124 Shigei Yabuuchi, Haid Beladi, Gu Wei known as the following proposition: The capital inflow ust reduce the host-country welfare if the country iports the capital-intensive coodity under tariff protection and foreign capital receives the full value of its arginal product. [Uzawa (1969), Haada (1974), and Brecher and Alejandro (1977)]. Several authors have extensively studied the validity of the proposition. Yabuuchi (1982) confired their results and extended it by deriving a sufficient condition for tariff iiserization. Brecher and Findlay (1983) and Casas (1985) exained the issue in the context of a odel with sector-specific inputs. The analysis was extended also to the case of iport quotas by Dei (1985). In a related study, various aspects of duty free zones and export processing zones have been discussed by trade theorists such as Rodriguez (1976), Hailton and Svensson (1987), Miyagiwa (1986, 1993), Young (1987, 1992), Young and Miyagiwa (1987), Beladi and Marjit (1992) and Yabuuchi (1997). In addition, there are soe attepts to exaine the effects of foreign capital inflow under tariff protection in the presence of uneployent. Khan (1982) showed the validity of the proposition even in the obile capital Harris-Todaro odel (hereafter HT odel), while Grinols (1991) questioned the validity in the presence of an inforal sector. Chandra and Khan (1993), Chao and Yu (1991, 1994), Gupta (1994, 1995), and Beladi and Marjit (2000) reexained the arguent in various situations. On the other hand, with differing ephases, Yabuuchi (1993), Chao and Yu (1996), Din (1996), and Beladi and Yabuuchi (2001) investigated the effects of tariff and subsidy policies in the presence of international capital obility and uneployent. In the present paper, we eploy the sae odel as Beladi and Yabuuchi (2001) and exaine the different issues, that is, the effects of exogenous inflow of foreign capital on the anufacturing output, uneployent, welfare and incoe distribution. The odel extends the Harris-Todaro odel to include an urban inforal sector where the product of the inforal sector is used as an industrial input in the urban foral sector. Grinols (1991) exained the issue by using a siilar odel. The differences between our odel and Grinols are twofold. First, the inforal good is a pure interediate good in our odel, while it is a final consuption good in his odel. Second, there is no explicit uneployent in his odel. However, we introduce urban uneployent as well as an inforal sector. Our ain

3 Foreign Investent, Urban Uneployent, and Inforal Sector 125 essages are as follows. Foreign investent into the urban foral sector increases the output of the sector under protection. The investent reduces uneployent and iproves welfare under certain conditions. The results show sharp contrast to the conventional wisdo. We derive the precise conditions for the results and identify the. In addition, the investent redistributes incoe fro capital to labor in the econoy as a whole. The paper is organized as follows. Section 2 presents the odel and basic assuptions. We exaine the effects of an exogenous inflow of foreign capital on output, uneployent and welfare in section 3. The iplication of the investent for incoe distribution is discussed in section 4. The final section includes concluding rearks on the findings. II. The Model and its Assuptions We consider a sall open econoy that consists of three sectors, the urban foral sector ( anufacturing sector ), urban inforal sector (interediate sector i) and rural sector (agricultural sector a). Production of agricultural and inforal goods requires labor and sector-specific capital. The anufacturing sector utilizes labor, sector-specific capital and the inforal good. Each sector utilizes strictly quasi-concave and constant returns to scale technology, and factors exhibit positive but diinishing arginal products. Thus, production functions in the urban anufacturing sector, X, urban inforal sector, X i, and rural agricultural sector, X a, are given by, X F ( K, L, R), X i F i ( K i, L i ), X a F a ( K a, L a ), (1) (2) (3) where K j and L j denote the eployent of capital and labor, respectively, in the jth sector (j, i, a). A distinguishing feature of our odel is the presence of an urban inforal sector that produces X i and is used as a pure interediate input (R) in the production of X. The agricultural good is exported at the noralized price of unity. The anufacturing good is an iportable good with world price p but whose doestic price p is increased by an ad valore tariff t, i.e., p(1+t)p.

4 126 Shigei Yabuuchi, Haid Beladi, Gu Wei Following Harris and Todaro (1970), it is assued that the real wage rate in anufacturing (w ) is rigid owing to soe political and/or institutional considerations while wage rate in the agricultural sector (w a ) is flexible. It is also assued that the inforal wage rate is constant and given w i αw, (4) where 0 < α < 1, represents the distortionary wage differential between the two sectors. As an exaple of this sort of wage differential, Beladi and Yabuuchi (2001) suggested the relation between the subcontract and parent firs as discussed in detail below. The subcontract firs are characterized by a low wage rate suppressed by the parent fir. The characteristic feature of the original HT odel is that in the labor arket equilibriu, the rural wage rate (w a ) equals the expected wage incoe in the anufacturing sector, which is by definition the anufacturing wage rate (w ) ties the probability of finding a job in the anufacturing sector. Let L u be uneployed in the urban area. Since the probability of finding a job in the anufacturing sector is L /(L +L u ) in the original HT odel without the inforal sector, then the expected wage in the anufacturing sector is w L /(L +L u ). Thus, the labor allocation echanis between rural and urban areas is expressed as w L /(L +L u ) w a. Introducing the inforal sector to the HT fraework would odify the labor allocation echanis such that in the labor arket equilibriu, the rural wage rate (w a ) equals the expected wage incoe in the urban area (that is, the anufacturing and inforal sectors). Since the probability of finding a job in the anufacturing sector is L /(L +L i +L u ) in the present case, then the expected wage in the anufacturing sector is w L /(L +L i +L u ). Siilarly, the expected wage in the inforal sector is w i L i /(L +L i +L u ). Thus, the expected wage in the urban area (the anufacturing and inforal sectors) is (w L +w i L i )/(L +L i +L u ). Therefore, the labor allocation echanis between rural and urban areas is expressed as (w L +w i L i )/(L +L i +L u )w a, or alternatively (w +µw i )w a (1+µ+λ), (5)

5 Foreign Investent, Urban Uneployent, and Inforal Sector 127 where and µl i /L and λl u /L. The labor arket equilibriu is expressed, in addition to (4) and (5) above, as w pf L ( K, L, R), (6) w i qf L i (, ), K i L i (7) w a F a L ( K a, L a ), (8) where F j L j, (j,i,a) and q is the price of the inforal good. Siilarly, in the capital arket equilibriu, we have F L j r pf K ( K, L, R), (9) r i qf K i (, ), K i L i (10) r a F a K ( K a, L a ), (11) where r j is the rental of capital in each sector (j,i,a). In the inforal good arket, the arginal value product of the interediate good equals its price. Thus, we have q pf R ( K, L, R). (12) The eployent condition in the labor arket is L a + L ( 1 + µ + λ) L. (13) The labor allocation in the econoy is shown in Figure 1, where, ii and aa are the arginal value products of labor in the doestic anufacturing, inforal and agricultural sectors, respectively. 00 is the aount of labor endowent. First, L is deterined at point M since w is fixed. Fro (4) and (5) we have (1+αµ)w w a (1+µ+λ) or ( 1 + αµ )w L w a ( L + L i + L u ) w a ( L L a ).

6 128 Shigei Yabuuchi, Haid Beladi, Gu Wei Figure 1. Labor Allocation Mechanis. Thus, the labor allocation between urban and rural areas is given at the intersection of aa and the rectangular hyperbola QQ through point N. This is because the nature of the rectangular hyperbola ensures that the area 0(1+αµ)w NL is equal to the area 0BAL a. Therefore, it can be seen that 0 L a is eployed in the agricultural sector and 0L a stays in the urban area. Now let us focus our attention to the labor allocation within the urban area. Fro (4) and (7), the eployent in the inforal sector is deterined at point E, that is, the intersection of the arginal value product of the inforal sector ii and the given w i (αw ). Thus, the figure shows that L a L i is eployed in the inforal sector and L i L reains uneployed. It is assued that capital is sector specific in the rural and inforal sector so that K a K a and K i K i, whereas the total eployent of capital in the anufacturing sector (K ) is given by K K + K, (14) where K j is the fixed endowent of doestic capital in the jth sector (ja, i,) and K is the inflow of foreign capital. We assue that the inforal good is a pure interediate good and the output of the sector is used up in the urban anufacturing sector. Thus, we have

7 Foreign Investent, Urban Uneployent, and Inforal Sector 129 R X i. (15) It ay be ore realistic to consider consuption goods and/or services as inforal products in developing countries. Here, however, we focus our attention to another aspect of inforal sectors. We suppose relatively industrialized developing econoies that have still dual structure, such as Asian NIEs and India. In these econoies, the inforal sector consists of any subcontract firs. They produce various parts and sei-processed coponents for the parent firs. Such subcontract firs are observed around achine and autoobile akers in various countries. These activities also have typical characteristic features of inforal sectors, such as sall scale, easy entry and aong others a low-wage rate suppressed by the parent firs as pointed out by Beladi and Yabuuchi (2001). Since the inforal good is non-traded in the present setup, the price (q) is deterined through the equilibriu condition for the inforal good arket. This copletes the specification of our odel. We have fourteen unknown variables, w i, w a, r, r i, r a, L a, L, L i, µ, L u, λ, qr, and K, which are solved by twelve equations (4) - (15) and the definitions of µ and λ for given paraeters, w, p, K a, K i, L, α and K. F RL Finally, we suppose that F RK 0, that is, the interediate input is technically independent of either labor or capital in the anufacturing production. This is partly justified by the fact that four tires are used to produce a car and one Brown tube is used for a TV set. III. Foreign Investent and Uneployent Differentiating (4) - (15) and the definitions of µ and λ, and arranging ters with respect to the iportant variables, we obtain 0 pf LL 1 pf KL i i 0 µqf LL F L 0 i qf LL L i F L µ + λ 1 L L 0 0 ( 1 + µ + λ)f LL ( w i w a ) w a a dr dl dl a dµ dλ pf LK pf KK dk. (16)

8 130 Shigei Yabuuchi, Haid Beladi, Gu Wei Solving this for dλ with respect to dk, we have dλ dk M { µ ( w i Ω) ( 1 + µ + λ) 2 F a LL L }pf LK i qf LL F i L, (17) where is the value of the deterinant of the coefficient atrix (16), ΩpF LL i qf LL L F L i > 0, and Ω w a ( 1 + µ + λ)f a LL L. Siilarly, we have the effects of foreign investent on eployent as follows. dl dk F LK F LL > 0, (18) dl a dk pf i LK L w qf LL F i L < 0, (19) dl i dk L dµ dk + µdl dk 0. (20) Thus, the following proposition is iediate. Proposition 1. Foreign investent into the urban foral sector increases the eployent of the sector and decreases that of the rural agricultural sector, while it leaves the eployent of the inforal sector constant. Since the inforal wage rate is fixed owing to soe institutional reason stated above, the eployent in the sector is constant. This iplies that the output is also constant. The capital inflow raises the arginal value product of labor in the anufacturing sector. This is shown as the shift of the curve fro to in Figure 1. This increases the anufacturing eployent to L. On the other hand, the agricultural eployent decreases to L a according to the labor allocation echanis. Now we exaine the effect of the investent on uneployent. By definition, we have dl u dk λdl dk + L dλ dk. (21) Substituting (17) and (18) into (21), we obtain dl u dk ( Ω w )pf LK i qf LL L F i L (22) This leads to the following proposition.

9 Foreign Investent, Urban Uneployent, and Inforal Sector 131 Proposition 2. Foreign investent into the urban foral sector reduces urban uneployent if and only if < Ω. w The capital inflow increases eployent of the urban foral sector owing to the increase in the arginal product of labor in the sector. This increase in urban eployent increases the expected urban wage rate. Thus, soe of the rural workers go to the city, and then this reduces rural eployent. We have shown that the eployent of the urban inforal sector is unchanged due to the capital inflow. If the increase in eployent in the foral sector is larger than the igrants fro the rural area, therefore, soe of the igrants will be absorbed in the uneployent pool. Proposition 3 provides the precise condition for the result. The condition can be written as w < w a ( 1 + µ + λ)f a LL L. Thus, it iplies that (1) if the rural wage rate is relatively large with respective to the anufacturing wage rate, and (2) if the rural eployent is sensitive to the a change in the rural wage rate (i.e., F LL is large), the decrease in the rural eployent becoes sall with respect to a given increase in the anufacturing eployent. Therefore, the necessary workers in agriculture ust be supplied fro the uneployent pool. This leads to a decrease in urban uneployent. IV. Foreign Investent and Welfare Now we investigate the iplications of capital inflow for output and welfare. The deand side of the odel is represented by a quasi-concave social utility function. Let U denote the social utility that depends on the consuption deand for agricultural and anufacturing goods denoted by, D a and D, respectively. Thus, U U( D a, D ). (23) The balance of trade equilibriu requires that D a + p D X a + p X r K. (24)

10 132 Shigei Yabuuchi, Haid Beladi, Gu Wei Differentiating (23) and (24) yields (see the Appendix) ( R U a )( du dk ) tp ( dx dk ) + [L {( w i w a )( dµ dk ) w a ( dλ dk )} K ( dr dk )], (25) where R { 1+ ( 1 )t} ( 1 + t), is the arginal propensity to consue the anufacturing good, and U a U D a.. Coparative statics shows that dµ dk µf LK L F LL < 0, (26) dr dk p F { LL F KK ( F LK ) 2 } F LL 0, (27) since F LL F KK ( F LK ) 2 { } 0 under the assuptions that F RL F RK 0 and the production function is hoogeneous to degree one. First, we exaine the effect of foreign investent on the output of the anufacturing sector. Differentiating the production function, we have p( dx dk ) ( w + µw i )( dl dk ) + w i L ( dµ dk ) + r. (28) Substituting (18) and (26) into (28) yields ( dx dk ) F LK ( w L + r K ) pf LL L > 0. (29) Thus, we have the following proposition. Proposition 3. Foreign investent into the urban foral sector increases the output of the sector. Now we are ready to see the iplication of foreign investent for welfare. Substituting (17), (26) and (27) into (25), we obtain ( R U a )( du dk ) tp F LK ( w L + r K ) pf LL +w ( L ) 2 ( 1 + µ + λ)f a LL pf i LK qf LL F i L. L (30)

11 Foreign Investent, Urban Uneployent, and Inforal Sector 133 This leads to the following proposition. Proposition 4. Foreign investent into the urban foral sector is iiserizing if and only if t > L ( 1 + µ + λ)f a LL θ L p Ω, where θ L w L ( w L + r K ). Foreign investent increases the output of the anufacturing sector that is distorted by the tariff. This effect is captured by the first ter of the RHS of (30), and it is harful to the welfare. On the other hand, the investent has positive incoe effect since it increases the eployent of the high wage-anufacturing sector and decreases that of the low wage agricultural sector. Proposition 4 provides the precise condition for the iiserization. The higher the tariff (i.e., the larger the distortion), and the ore sensitive the rural eployent with respect to the change a in the rural wage rate (i.e., the saller ( F LL )), the larger is the likelihood of iiserization. If the tariff rate is high, the negative distortion cost ay doinate the positive incoe effect. If the sensitivity of the rural eployent is high, the decrease in the rural eployent is large. Thus, the positive incoe effect becoes sall. Conversely, if the tariff rate is saller than the critical value, foreign investent iproves welfare even under the situation supposed by Uzawa (1969), Haada (1973), and Brecher and Alejandro (1977) Grinols (1991) showed the possibility of welfare iproveent due to the econoic contribution by workers eployed in the inforal sector whose opportunity costs are considerably low. We have shown that there is another possibility of welfare iproveent owing to labor allocation in the odel with both an inforal sector and uneployent. V. Incoe Distribution In this section, we investigate the iplication of foreign investent for incoe distribution. Wage incoe is expressed as w a L a + w L + w i L i w a L since (w L + w i L i ) w a (L +L i +L u ). Thus, the change in wage incoe depends solely on the change in the rural wage rate since the endowent of labor (L) is constant. Fro (17), we have

12 134 Shigei Yabuuchi, Haid Beladi, Gu Wei a dw a dk F LL ( dl a dk ) > 0. (31) On the other hand, the change in capital incoe ( r K + r i K i + r a K a ) depends solely on the change in the rental of rural capital since each sector-specific capital a is constant and dr dk F KL ( dl a dk ) < 0. Furtherore, fro (11) and (19), we obtain a dr a dk F KL ( dl a dk ) < 0. The following proposition suarizes the results. Proposition 5. Foreign investent into the urban foral sector redistributes incoe fro capital to labor. Wage incoe of urban foral sector increases owing to the increase in the eployent, while that of urban inforal sector is unchanged. The change in wage incoe of the rural sector depends on the elasticity of labor deand in the sector with respect to the rural wage rate. Wage incoe of the rural sector increases if and only if the elasticity of labor deand is less than one since a d( w a L a ) dk L a ( 1 ξ L )( dw a dk ), where w a L a F LL Anyway, the result shows that the total wage incoe increases definitely owing to the increase in the eployent in the high wageanufacturing sector. On the other hand, the total capital incoe decreases through the decrease in the rental of rural capital. ξ L a a. VI. Concluding Rearks We have reexained the effects of foreign investent in the odel with an inforal sector and uneployent. We have paid particular attention to the possibility of iiserizing growth under soe distorted condition. It has been shown that the investent increases the output of the invested sector, it ay decrease uneployent, and it redistributes national incoe fro capital to labor. In addition, we have shown that there is a possibility of welfare iproveent contrary to the conventional wisdo. Thus, our result is ore favorable than

13 Foreign Investent, Urban Uneployent, and Inforal Sector 135 the traditional one for developing countries. In this sense, our analysis is consistent with uch epirical evidence that eerging econoies intend to industrialize introducing foreign capital intensively. However, our results depend on several assuptions that capital is specific to the respective sector, the inforal wage rate is constant ( w i αw ;0 < α < 1 ), and the interediate input is technically independent of either labor or capital in the anufacturing production ( F RL F RK 0 ). Therefore, soe qualification ust be needed to evaluate and apply the results obtained in this paper. At the sae tie, it would constitute an iportant agenda for the further research in this area to relax these assuptions. Appendix Here we derive the equation (25) in the text. Differentiating (23) and (24), we have du U a dd a + pdd ( dx a + pdx r dk K dr ) + tp dm, (A1) where M( D X ) is the iport deand for the anufacturing good. Differentiating the iport deand function M, and considering the anufacturing price (p) is constant, we obtain dm {( p) ( dx a + pdx r dk K dr ) dx } R, (A2) where is the arginal propensity to consue the anufacturing good, and R{1+(1-)t}/(1+t). Substituting (A2) into (A1), we have RdU U a ( dx a + pdx r dk K dr ) tp dx. (A3) Differentiating (1) - (3), and considering (6) - (8), we have dx a + pdx r dk K dr w a dl a + w dl + w i dl i K dr. Considering (5), we can show (A4)

14 136 Shigei Yabuuchi, Haid Beladi, Gu Wei w dl + w i dl i L ( w i w a )dµ L w a dλ w a dl a. (A5) Thus, fro (A3), (A4) and (A5), we have RdU U a tp dx + [ L {( w i w a )dµ w a dλ} K dr ]. (A6) This leads to the equation (25) in the ain text. Acknowledgeent The authors would like to express their appreciation to an anonyous referee for any helpful coents and suggestions. Received 16 Deceber 2002, Accepted 25 Septeber 2003 References Beladi, H. and S. Marjit (1992), Foreign capital and protectionis, Canadian Journal of Econoics 25, Beladi, H. and S. Marjit (2000), A general equilibriu analysis of foreign investent and intersectoral linkage, Journal of International Trade and Econoic Developent 9, Beladi, H. and S. Yabuuchi (2001), Tariff-induced capital inflow and welfare in the presence of uneployent and inforal sector, Japan and the World Econoy 13, Bhagwati, J. N. (1958), Iiserizing growth: A geoetric note, Review of Econoic Studies 25, Bhagwati, J. N. (1968), Distortions and iiserizing growth: A generalization, Review of Econoic Studies 35, Bhagwati, J. N. (1973), The theory of iiserizing growth: Further application, in M. B.Connolly and A. K. Swoboda eds., International Trade and Money (University of Toronto Press) Brecher, R.A. and C.F. Diaz Alejandro (1977), Tariffs, foreign capital and iiserizing growth, Journal of International Econoics 7, Brecher, R.A. and R. Findlay (1983), Tariffs, foreign capital and iiserizing growth, Journal of International Econoics 7, Casas, F. R. (1985), Tariff protection and taxation of foreign capital: the welfare iplications for a sall country, Journal of International Econoics 19, Chandra, V. and M. A. Khan (1993), Foreign investent in the presence of an inforal sector, Econoica 60,

15 Foreign Investent, Urban Uneployent, and Inforal Sector 137 Chao, C. C. and E. S. H. Yu (1991), Iiserizing growth for a quota-distorted sall econoy under variable returns to scale, Canadian Journal of Econoics 24, Chao, C. C. and E. S. H. Yu (1994), Foreign capital inflows and welfare in an econoy with iperfect copetition, Journal of Developent Econoics 45, Chao, C. C. and E. S. H. Yu (1996), International capital obility, urban uneployent and welfare, Southern Econoic Journal 62, Dei, F. (1985), Welfare gains fro capital inflows under iport quotas, Econoics Letters 18, Din, Musleh (1996), International capital obility and developent policy in a dual econoy, Review of International Econoics 4, Grinols E. L. (1991), Uneployent and foreign capital: the relative opportunity costs of doestic labor and welfare, Econoica 58, Gupta, M. R. (1994), Foreign capital, incoe inequality and welfare in a Harris-Todaro odel, Journal of Developent Econoics 45, Gupta, M. R. (1995), Tax on foreign capital incoe and wage subsidy to the urban sector in the Harris-Todaro odel, Journal of Developent Econoics 47, Haada, K. (1973), An econoic analysis of the duty free zone, Journal of International Econoics 4, Hailton, C. and L. O. Svensson (1982), On the welfare effects of a duty free zone, Journal of International Econoics 13, Harris, J. R. and M. Todaro (1970), Migration, uneployent and developent: A twosector analysis, Aerican Econoic Review 60, Johnson, H. G. (1967), The possibility of incoe losses fro increased efficiency or factor accuulation in the presence of tariffs, Econoic Journal 77, Khan, M. A. (1982), Tariffs, foreign capital and iiserizing growth with urban uneployent and specific factors of production, Journal of Developent Econoics 10, Miyagiwa, K. (1986), A reconsideration of the welfare econoics of a free-trade zone, Journal of International Econoics 21, Miyagiwa, K. (1993), The locational choice for free-trade zones, Journal of Developent Econoics 40, Rodriguez, C. A. (1976), A note on the econoics of the duty free zones, Journal of International Econoics 6, Uzawa, H. (1969), Liberalization of foreign investents and the national econoy, Ekonoisuto 23, (in Japanese). Yabuuchi, S. (1982), A note on tariff-induced capital inflow and iiserization in the presence of foreign profits, Journal of International Econoics 12, Yabuuchi, S. (1993), Urban uneployent, international capital obility and developent policy, Journal of Developent Econoics 41, Yabuuchi, S. (1997), Direct investent, onopoly and welfare, Pacific Econoic Review 2,

16 138 Shigei Yabuuchi, Haid Beladi, Gu Wei Young, L. (1987), Interediate goods and the foration of duty-free zones, Journal of Developent Econoics 25, Young, L. (1992), Uneployent and the optial export-processing zone, Journal of Developent Econoics 37, Young, L. and K. Miyagiwa (1987), Uneployent and the foration of the duty-free zones, Journal of Developent Econoics 26,