International Review of Economics and Finance 42 (2016) 103–115

Contents lists available at ScienceDirect

International Review of Economics and Finance

journal homepage: www.elsevier.com/locate/iref

The impacts of capital market distortion on wage inequality,

urban unemployment, and welfare in developing countries
Jiancai Pi ⁎, Xuyang Chen
Department of Economics, School of Business, Nanjing University, 22 Hankou Road, Nanjing 210093, PR China

a r t i c l e i n f o a b s t r a c t

Article history: This paper employs the general equilibrium approach to investigate how the change of rural–
Received 7 April 2015 urban capital market distortion influences skilled–unskilled wage inequality and urban unem-
Received in revised form 22 October 2015 ployment in developing countries. Our basic model shows that a decrease in capital market dis-
Accepted 22 October 2015
tortion will reduce the wage inequality and the urban unemployment rate. Our extended
Available online 28 October 2015
model confirms the main results of our basic model. Furthermore, this paper also analyzes
the effect of capital market distortion on social welfare, and finds that an alleviation of capital
JEL classification: market distortion will generate a welfare gain in both basic and extended models.
J31
© 2015 Elsevier Inc. All rights reserved.
J64
P23
Keywords:
Capital market distortion
Wage inequality
Urban unemployment
Social welfare
General equilibrium approach

1. Introduction

The rural–urban capital market distortion and population migration are typical features of developing countries. Rural–urban
migration, providing urban areas with abundant labor forces, stimulates economic growth of developing countries. However, com-
pared with urban manufacturing industries, agricultural sectors in rural areas have to pay a higher interest rate of capital, which
results in the inefficiency of resource allocation and thus makes the rural–urban capital market distortion become a stumbling
block to economic development. Many economists empirically explore the economic impacts of capital market distortion. For in-
stance, Zhuang (1996) shows that substantial distortions existed in the 1983 Chinese economy, where capital was underpaid in
most non-agricultural sectors. He finds that there would be an increase of almost 15% in real gross domestic product (GDP) if
China were switched to the free market system where all the distortions are eliminated. Bhattacharyya and Kumbhakar (1997)
investigate the agricultural sector in India's West Bengal by using a sample of 289 rice growers. They find that the farmers suf-
fered an output loss of 11.7% due to the capital market distortion. Lin and his coauthors (e.g., Chen & Lin, 2014, Lin, 2003,
2009, Lin, Cai, & Li, 1994) conduct their analyses in a more systematic and comprehensive way, and they empirically reveal

⁎ Corresponding author. Tel.: +86 25 83621121.

E-mail addresses: pi2008@nju.edu.cn (J. Pi), jscxyang@163.com (X. Chen).

http://dx.doi.org/10.1016/j.iref.2015.10.045
1059-0560/© 2015 Elsevier Inc. All rights reserved.
104 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115

that the heavy-industry-oriented development strategy in many developing countries (e.g., China and India) will lead to capital
market distortion and other factor-price distortions, and that such kinds of distortions will result in poorer economic performance,
higher unemployment level, and larger urban–rural inequality.
There are many factors that affect the capital market distortion. Chaudhuri and Dastidar (2011); Chaudhuri and Gupta
(1996), and Gupta and Chaudhuri (1997) find that the corruption of the bank official will lead to capital market distortion.
Bose (1998) points out that the small cultivators' heterogeneous likelihood of default and asymmetric information about
such kind of risk will worsen the availability of loans for the informal sector. Chaudhuri (2001) suggests that if the formal
interest rate is very low, not the small farmers but the bank official and the moneylender will catch the dissipated rent.
Chaudhuri and Gupta (2014) find that international factor mobility will raise the degree of capital market distortion. More
importantly, the issue about how the improvement of capital market distortion influences the economy has captured
much attention of development economists. For instance, Khan and Naqvi (1983) argue that in a two by two Harris–Todaro
economy, a reduction of capital market distortion will decrease urban unemployment and raise national welfare. Chao and
Yu (1992) claim that Khan and Naqvi's findings are not necessarily held if land is introduced to their model. However, if
labor is divided into skilled and unskilled types and an intermediate sector is introduced to consider the upstream and down-
stream relationship in the producing process, will the corresponding results be amended? This question still remains
unsolved in the existing literature.
In this paper, we use the term “capital market distortion” in the sense of Chao and Yu (1992), Gupta (1997b) and Khan and
Naqvi (1983) that is, there exists a capital rental differential. The distortion in our sense is referred to as the rural–urban capital
market distortion, which manifests as a far higher interest rate of capital in the rural area than in the urban area. Just as in Gupta
(1997a), the distortion derives from the fact that the rural sector usually uses capital from the unorganized informal credit market
while the urban sector can employ capital provided by formal lenders.1 This phenomenon emerges in many developing countries
(e.g., China and India). For example, Romatet (1983) provides the empirical evidence by investigating into the capital market
distortion of India's Calcutta.
On the one hand, along with the pace of globalization, many developing countries suffer from the growing skilled–unskilled
wage inequality (see e.g., Anwar & Sun, 2012; Feenstra & Hanson, 2003; Wood, 1997). This phenomenon has attracted more and
more attention. The existing literature tries to address such an issue from different perspectives. The first strand of literature
(see e.g., Anwar, 2006, 2010; Anwar & Rice, 2009; Beladi, Chakrabarti, & Marjit, 2010, 2011; Das, 2002, 2005; Marjit, Beladi, &
Chakarabarty, 2003; Moore & Ranjan, 2005; Pan & Zhou, 2013; Zhang, 2012, 2013) focuses on trade, factor mobility and techni-
cal progress, which shows that intra- and inter-industry trade, international factor flow, and technological advance can condi-
tionally increase the wage inequality. The second strand of literature (see e.g., Anwar, 2008; Anwar & Sun, 2015; Chaudhuri &
Yabuuchi, 2007; Pan, 2014; Pi & Zhou, 2012, 2013) concentrates on the role of public inputs, institutional arrangements, and
economic policies in changing the skilled–unskilled wage inequality, which highlights that the aforementioned channels will
generate their influences. For example, Chaudhuri and Yabuuchi (2007) find that although there are some factors that worsen
the wage inequality, labor market reforms can narrow down the wage gap through raising the wage rate of unskilled labor under
some condition. Nevertheless, to the best of our knowledge, the impact of rural–urban capital market distortion on wage in-
equality is largely ignored by the existing literature. Since the rural–urban capital market integration policy is always regarded
as one of the most important development policies in developing countries, it is worthy to analyze the impact of such a policy on
wage inequality.
On the other hand, urban unemployment in developing countries is explored by the existing studies from different angles,
which show that different development policies exert different impacts on unambiguously or conditionally reducing urban unem-
ployment. Beladi and Naqvi (1988) stress the role of the improvement in the agricultural production technique. Gupta (1997a)
highlights the effect of enhancing the agricultural output price. Chaudhuri (2000) focuses his attention on the wage and price
subsidies to the rural sector, and Yabuuchi and Beladi (2001) consider such an issue with an eye to the wage subsidy in a
model different from Chaudhuri (2000). Chaudhuri and Banerjee (2010) emphasize the impacts of FDI in agricultural land on
unemployment. In addition, Chaudhuri and other scholars (e.g., Chaudhuri, 2007, Chaudhuri, Yabuuchi, & Mukhopadhyay,
2006) also analyze the role that other development polices play (e.g., international factor mobility policies) in reducing urban
unemployment. However, the existing studies neglect to consider the influence of rural–urban capital market distortion on
urban unemployment in the presence of two types of labors and the intermediate product. This paper complements the existing
studies in this direction.
In order to fill the current research gap, we follow the work of Khan and Naqvi (1983), and make some modifications to get
closer to reality. We divide labor into skilled and unskilled types and introduce an intermediate sector in order to consider the
upstream and downstream relationship in the production process. We show that the mitigation of rural–urban capital market dis-
tortion will decrease the wage inequality and the urban unemployment rate, and that the total unemployed population can also
descend if the economy participates in the exportation of high-skill products. What is more, we also study the effect of rural–
urban capital market distortion on social welfare.
The rest parts of this paper are organized as follows. We construct the basic model in Section 2. Section 3 gives the extended
model. Concluding remarks are provided in Section 4.

1
Gupta (1997a) considers three sectors in his model, i.e., the urban formal, urban informal and rural sectors. The urban informal sector and the rural sector share a
common interest rate of capital, while the urban formal sector employs capital at a lower interest rate.
 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115 105

2. The basic model

Consider a small open developing economy consisting of three sectors: two urban sectors and one rural sector.2 Specifically,
the urban low-skill sector X employs unskilled labor (U1), capital (K1), and an intermediate product (Y) to produce a tradable
commodity X. The urban high-skill sector Y, acting as an upstream industry for sector X, uses skilled labor (S) and capital (K2)
to manufacture a non-tradable intermediate product Y. The rural sector Z utilizes unskilled labor (U3) and capital (K3) to produce
a tradable agricultural good Z, which is treated as numeraire. All the commodity and factor markets are perfectly competitive.
The production functions are given as below:
8
> 1
< X ¼ F ðU 1 ; K 1 ; Y Þ
2
Y ¼ F ðS; K 2 Þ ; ð1Þ
>
: 3
Z ¼ F ðU 3 ; K 3 Þ

where F i(i = 1 , 2 , 3) is an increasing function with respect to each factor and features the property of strict quasi-concavity and
homogeneity of degree one.
In our model, skilled labor is only employed by the intermediate sector, which is similar to land as a specific factor for the
agricultural sector in Chao and Yu (1992).3
Since perfect competition generates the zero profit conditions, we can obtain:

pX ¼ aUX w þ aKX r U þ aYX pY ; ð2Þ

pY ¼ aSY wS þ aKY r U ; ð3Þ

1 ¼ aUZ wU þ aKZ rR ; ð4Þ

where pX and pY are the prices of X and Y, respectively. The price of Z is normalized to 1, and pX is given internationally due to
tradability. w, wS, and wU are the wage rates of urban unskilled labor, urban skilled labor, and rural unskilled labor, respectively.
w is downwards rigid and protected by the urban minimum wage acts. rU and rR are the interest rate of capital in urban and rural
areas, respectively. aij (i = U , S , K,j = X , Y , Z) is the amount of the ith factor used by the jth industry to produce one unit of the
corresponding product.
The rural–urban capital market distortion in developing countries manifests as a far higher interest rate of capital in the rural
area than in the urban area. That is, we usually have rR N rU. Following the setting in Khan and Naqvi (1983) and Chao and Yu
(1992), we formally characterize the rural–urban capital market distortion as:

r R ¼ αr U ; ð5Þ

where α N 1 is exogenously given and measures the extent of rural–urban capital market distortion.4 Concretely, the greater the
value of α is, the higher degree of distortion the capital market has.
The Harris–Todaro labor allocation mechanism (see Harris & Todaro, 1970) is given by:

w ¼ ð1 þ λÞwU ; ð6Þ

where λ is the Harris–Todaro urban unemployment rate, which is calculated by the ratio of urban unemployment to urban
employment in sector X. The wage of urban skilled labor, however, is supposed to adjust freely based on the supply–demand
relationship in sector Y, which leads to the full employment of skilled labor.
The market-clearing conditions yield:

aSY Y ¼ S; ð7Þ

aKX X þ aKY Y þ aKZ Z ¼ K; ð8Þ

aUX X ð1 þ λÞ þ aUZ Z ¼ U; ð9Þ

2
Different from Gupta (1997a), who considers three sectors consisting of an urban formal sector, an urban informal sector and a rural sector, this paper does not take
the urban informal sector into account. Gupta (1997a) focuses on the urban informal sector and the informal capital market, but this paper concentrates on the rural–
urban capital market distortion and skilled–unskilled wage inequality.
3
Here, it is necessary to mention Chao and Yu (1992) in methodological sense. By introducing land into the Harris–Todaro model, Chao and Yu (1992) aim at exam-
ining how welfare will change compared with the findings of Khan and Naqvi (1983), but they neglect to discuss the wage inequality. Since we concentrate on the issue
of wage inequality, our model focuses on labor types instead of introducing another specific factor (i.e., land).
4
Khan and Naqvi (1983) and Chao and Yu (1992) set rR =(1+t)rU, which is equivalent to Eq. (5).
106 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115

aYX X ¼ Y; ð10Þ

where S, K, and U are the skilled labor, capital, and unskilled labor endowments, respectively.
So far, the establishment of our basic model is completed. Eqs. (2) to (10) determine nine endogenous variables, wS, wU, rU, rR, pY, X,
Y, Z, and λ, the exogenous variable is α, and all the others are parameters.
Compared with the agricultural sector Z that massively employs unskilled labor for cultivation, sector X is an urban
manufacturing sector that is characterized by the large-scale utilization of capital. Therefore, it is convictive to assume that sector
Z is more unskilled-labor intensive than sector X, which leads to the following condition:5

λKX λUZ −ð1 þ λÞλUX λKZ N 0; ð11Þ

where λij is the allocative share of factor i in the jth sector (e.g., λKX ¼ XaKX =K).
Taking total differentiation to Eqs. (2)–(10) and solving the system (A-1) for endogenous variables, we can get:6

^r R λwU θUZ
¼ fð1 þ λÞθSY λKZ λUX ½ðSUY −SYY ÞθKX þ ðSYK −SUK ÞθYX  þ θSY λKX λUZ ½ðSYY −SKY ÞθKX ð12Þ
α^ Δ
þðSKK −SYK ÞθYX  þ σ Y ðθKX þ θYX Þ½θKY ðð1 þ λÞλUX λKZ −λKX λUZ Þ−λKY λUZ g;

where “∧” denotes the percentage change (e.g., ^r R ¼ drrRR), Δ is the value of the determinant of the coefficient matrix of the system
(A-1). Δ b 0 ensures the local stability of the equilibrium solutions.7
^
Notice that given Inequality (11), the last two terms in the curly brace in Eq. (12) are strictly negative. Thus, rα^R N 0 if
(SUY − SYY)θKX + (SYK − SUK)θYX b 0, which is equivalent to:

SUK S S S
N UY þ YK − YY ; ð13Þ
θKX θYX θKX θYX

SUK SUY SYK

where θKX , θYX , and θKX represent the well-known Allen–Uzawa partial elasticities of substitution.
It is obvious that Inequality (13) implies SθUK
KX
N SθUY
YX
þ SθYK
KX
, which means that in sector X, unskilled labor is more easily substituted
by capital in the relative sense. Hence, Inequality (13) depicts the technological feature of production of X. We assume that
Inequality (13) holds for the reason that in the traditional manufacturing industry in developing countries, unskilled labor can
always be readily replaced by capital, i.e., mechanized production can be adopted.8 Therefore, Inequalities (11) and (13) are
assumed to hold throughout this section.
From the discussion above, we can obtain:

^r R
N 0; ð14Þ
α^

^U
w θ ^r
¼ − KZ R b 0; ð15Þ
α^ ^
θUZ α

^
λ ^U
1 þ λw
¼− N 0: ð16Þ
^
α λ α ^

^ yields:
^ S , and X
Solving Eq. (A-1) for ^r U , w

^r U wU θSY θYX λKZ ½λð1 þ λÞθKZ λUX þ λσ Z λUZ 

¼ b 0; ð17Þ
α^ Δ

^ S −λwU ðθKX þ θKY θYX ÞλKZ ½ð1 þ λÞθKZ λUX þ σ Z λUZ 

w
¼ N 0; ð18Þ
^
α Δ

^ λw λ ½ðS θ −S θ Þθ −θ ðθ þ θ Þσ ½ð1 þ λÞθ λ þ λ σ 

X U KZ YY KX YK YX SY KY KX YX Y KZ UX UZ Z
¼ N 0: ð19Þ
^
α Δ

5
Although there are popularization and promotion of modern agricultural machines in rural areas, intensive cultivation using large quantities of peasants is still the
main mode of production in many developing countries such as Southeast Asian economies.
6
See Appendix A.1.
7
See Appendix A.2.
8
Some traditional low value-added sectors (e.g., clothing industry and leather industry) in developing countries often have such a characteristic.
 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115 107

Combining Eq. (15) with Eq. (18) yields:

^ S −w
w ^U
N 0: ð20Þ
^
α

As is analyzed previously, unskilled labor is the only partly idle factor because of wage rigidity. Hence, we can calculate the
change of total unemployment denoted by LU (here, LU = λ XaUX) with regard to the change of capital market distortion:

^L ^ X
λ ^ ^r ^
p w
U
¼ þ þ SUK U þ SUY Y ¼ U ðΦ1 þ Φ2 þ Φ3 Þ: ð21Þ
α^ ^ ^
α α α^ ^
α Δ

where Φ1 =λKZ θSY [(SUY −SYY)θKX +(SYK −SUK)θYX][(1+ λ)θKZλUX − λλUZ σZ], Φ2 = σY(θKX + θYX)[θKZ (1+λ)(θKYλKZλUX − θKYλKXλUZ −
λKYλUZ)−λθKYλKZλUZ σZ] b 0, and Φ3 = −(1+λ)λKXλUZ θKZ θSY [(SKY −SYY)θKX +(SYK −SKK)θYX] b 0.
^
Because of the uncertainty of the sign of Φ1, the sign of Lα^U (i.e., the change of total unemployment) is ambiguous. Nevertheless,
^LU
if σZ (i.e., the elasticity of substitution in sector Z) is relatively low (i.e., smaller thanð1þλÞθ
λλUZ
KZ λUX
), then we have α^ N 0.
According to Eqs. (14) to (21), Proposition 1 can summarize the main results discussed above.

Proposition 1. (i) An alleviation of rural–urban capital market distortion will increase (resp. decrease) the wage rate of unskilled labor
(resp. skilled labor), and thus the wage inequality will be narrowed down. (ii) The urban unemployment rate decreases with the
improvement of capital market distortion, whereas the change of total unemployment in absolute amount is ambiguous. Furthermore,
if the agricultural sector's elasticity of substitution is sufficiently low, then the total unemployed population will also descend with the
reduction of capital market distortion.

Proposition 1 can be explained as follows. A decrease in α (i.e., a mitigation of capital market distortion) will result in the
reallocation of capital in different areas, making the interest rate of capital in the rural area reduce and that in the urban area
rise. Faced with this situation, firms in sector X will seek the substitution for capital in order to cut down the production cost.
According to Inequality (13), unskilled labor is more technologically feasible to be substituted than the intermediate product Y.
Therefore, large quantities of newly increased jobs will arise in the urban sector X due to the substantial rise of the demand
for unskilled labor.
In sector Y, on the one hand, for each unit of Y, more skilled workers are employed to save the cost as the urban interest rate
of capital increases, which generates a skilled-wage-enhancing effect. On the other hand, in the production process of X, there ex-
ists a great convenience of the substitution between unskilled labor and capital due to SθUK KX
N SθUY
YX
þ θSKX
YK
. Therefore, with the extensive
use of unskilled labor in sector X, the demand for the third factor (i.e., intermediate product Y) in producing X also descends
sharply,9 generating a skilled-wage-reducing effect that dominates the skilled-wage-enhancing effect. That is, the total demand
for skilled labor (i.e., YaSY) descends. Therefore, the wage rate of skilled labor will decrease.
Now we turn to the agricultural sector. On the one hand, although some unskilled workers can be replaced by modern agri-
cultural machines because capital is cheaper in the rural area, Z is still an unskilled-labor intensive product, which needs large
quantities of unskilled workers. Thus, we can clearly expect a slight drop in the demand side of unskilled labor. On the other
hand, there exists a large outflow of unskilled labor from the rural area to the urban area, which results from the attraction of
large quantities of newly increased jobs in the urban sector X and the labor's motivation of earning a higher wage rate. This causes
a substantial reduction in the supply of unskilled labor in the rural area. That is, as for rural unskilled labor, the drop in the supply
side ultimately overwhelms the decrease in the demand side. Therefore, we derive that the wage rate of unskilled labor in the
rural area will rise.
As for the urban unemployment rate, according to the Harris–Todaro migration mechanism (i.e., w ¼ ð1 þ λÞwU ), the urban
unemployment results from the downward rigidity of the wage rates of urban unskilled labor. Furthermore, on the basis of the
migration mechanism, the urban unemployment rate decreases with the rise of rural unskilled wage rate since w is exogenously
given in our model. Because the mitigation of capital market distortion raises the rural unskilled wage rate, which has been
explained previously, the urban unemployment rate will descend.
However, what we should point out is that the change of total unemployment in absolute amount remains indefinable, which
rests on the relative sizes of the newly migrant population from the rural area and the newly increased job positions in sector X.10
ð1þλÞθKZ λUX
Particularly, if the agricultural sector's elasticity of substitution σZ remains relatively small (i.e., σ Z b λλUZ ), which implies that
there will not be adequate unskilled labor replaced by cheaper capital in sector Z. Then the newly migrant population released
from the rural sector will be fully assimilated by the newly increased job positions in sector X, leading to a decrease in total
unemployment.
^
In a two by two Harris–Todaro model constructed by Khan and Naqvi (1983), αλ^ N0 holds. However, in Chao and Yu (1992), the
change of the urban unemployment ratio becomes ambiguous when a sector-specific factor (i.e., land) is introduced to the

^
9
In fact, by simple calculation, we have αY^ ¼ − λwUΔθKY ðθKX þ θYX ÞλKZ σ Y ½ð1 þ λÞθKZ λUX þ λUZ σ Z  N0, which confirms our inference.
10
Notice that the following identity holds: λ@ X @ a@ # # # ≡ ð Z # a# −Z @ a@ Þ−ðX @ a@ −X # a# Þ, where the superscripts ‘#’ and ‘@’ denote the equilib-
UX −λ X aUX
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} |fflfflfflfflfflfflfflfflfflfflffl
UZ
ffl{zfflfflfflfflfflfflfflfflfflfflfflUZ
ffl} |fflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflffl}
UX UX
the change of total unemployment newly‐migrant population newly‐increased jobs
rium solutions before and after the improvement of capital market distortion, respectively.
108 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115

agricultural sector in Khan and Naqvi's (1983) model. This paper argues that if labor is divided into two types and skilled labor is
specific to the urban intermediate-input sector, Khan and Naqvi's (1983) finding can be regained.
In the last part of this section, we investigate the effect of capital market distortion on social welfare. We use V(DX,DZ) to mea-
sure social welfare, where DX and DZ are the total demand for final products X and Z, respectively. Furthermore, function V is
strictly quasi-concave, and strictly increasing in DX and DZ. Notice that our assumed economy is faced with a budget constraint
as below:

pX DX þ DZ ¼ pX X þ Z: ð22Þ

The right side of Eq. (22) represents the national income when the economy reaches general equilibrium, thus this equation
characterizes the balance of trade.
By some calculations, we can obtain11:
   
dV  dX dZ dλ ðα−1ÞλK 3 rU wU
¼ V DZ pX þ ¼ V DZ −U 1 wU þ ðΩ1 þ Ω2 þ Ω3 Þ b 0; ð23Þ
dα dα dα dα Δ

where V ⁎ denotes the maximum social welfare under the balance of trade, V DZ ¼ ∂D
∂V
N 0,
Z

Ω1 = (1+ λ)θKZ θSY θYXλKXλUX(SYK − SKK) N 0,

Ω2 = (1+ λ)θKZλUX(θKX + θYX)(λKY + θKYλKX)σY N 0,
Ω3 = σZ {[(SYK −SKK)θYX + (SKY − SYY)θKX]θSYλKXλUZ + λUZ σY(θKX + θYX)(θKYλKX + λKY)}+(1+ λ)(SKY − SYY)θKX θSYλKXθKZλUX N 0.

Based on Eq. (23), we can obtain the following proposition.

Proposition 2. Social welfare will increase when the capital market distortion improves.

As has been analyzed previously, the alleviation of capital market distortion will narrow down the wage gap and reduce the
unemployment rate, which leads to the optimization of resource allocation. This will finally increase total output and national
income, and raise social welfare.

3. The extended model

In some developing countries, just as is pointed out by Beladi et al. (2010), some skilled-labor intensive industries also engage
in exporting.12 In accordance with this phenomenon and in order to test the robustness of our findings in Section 2, we introduce
a fourth sector M to build our new model, which is another high-skill sector producing a tradable final product M at an interna-
tionally determined price pM. Our notations in this section are identical with those in Section 2. As is known, developing countries
always export labor intensive products to developed countries for a better use of their cheaper labor force, thus it is natural to
assume that sector M is more labor intensive than sector Y. Hence, besides Inequality (11), we should add the following condi-
tions in this section:

θSM θKY −θKM θSY N 0; ð24Þ

λSM λKY −λKM λSY N 0: ð25Þ

In the following part of this section, we assume that Inequalities (11), (24) and (25) hold. Moreover, in Appendix B.2, from the
perspective of consistency with the local stability of the new equilibrium solutions, we further demonstrate the reasonableness of
these three conditions in our extended model.
The following equation should be added to the established model:

pM ¼ aSM wS þ aKM r U : ð26Þ

Furthermore, Eqs. (7) and (8) should be rewritten as:

aSY Y þ aSM M ¼ S; ð27Þ

aKX X þ aKY Y þ aKM M þ aKZ Z ¼ K: ð28Þ

Combining Eqs. (26) to (28) with Eqs. (2) to (6), (9), and (10), we can obtain the new system.

11
See Appendix A.3.
12
A typical example is the software industry in India, which is of great competitiveness worldwide and creates large amounts of export value.
 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115 109

Solving the subsystem (B-1) for endogenous variables, we can get:13

^r U
¼ 0; ð29Þ
α^

^r R
¼ 1: ð30Þ
α^

Eqs. (29) and (30) show that the rural interest rate of capital will decrease with the alleviation of capital market distortion and
that the interest rate of capital in the urban area will remain unchanged.
Continuing to solve the subsystem (B-1) for other variables, we obtain:

^S
w θ ^r w^ θ ^r ^
λ ^U
1 þ λw w^ −w^U
¼ − KM U ¼ 0; U ¼ − KZ R b 0; ¼− N 0; and S N 0:
^
α ^
θSM α α^ ^
θUZ α ^
α λ α ^ ^
α

The above results confirm the validity of our previous conclusions in Section 2 concerning the wage inequality and the urban
unemployment rate. We should also point out that all the above findings do not depend on Inequality (13) in our basic model.
When focusing on the urban unemployed population (i.e., LU), we have:

^L ð1 þ λÞθKZ ½ðλKZ λUX −λKX λUZ ÞλSM −ðλKY λSM −λKM λSY ÞλUZ −λλKZ λSM λUZ σ Z
U
¼ :
α^ λθUZ ½ðð1 þ λÞλKZ λUX −λKX λUZ ÞλSM −ðλKY λSM −λKM λSY ÞλUZ 

Given conditions (11) and (25), it is clear that both the numerator and the denominator of the right term in the above equa-
tion are strictly negative, which implies that:

^L
U
N 0: ð31Þ
α^

Thus, dissimilar to our basic model, with the improvement of capital market distortion, total unemployment will definitely de-
crease if a skilled-labor intensive exportable sector is introduced.
All the results discussed above can be summed up by Proposition 3.

Proposition 3. When a skilled-labor intensive exportable sector is introduced, an alleviation of capital market distortion will decrease
the wage inequality, the urban unemployment rate, and the total unemployed population.

Proposition 3 can be interpreted as follows. When the capital market distortion mitigates, the reallocation of capital in different
areas will decrease the excessively high interest rate of capital in the rural area, which leads to the substitution of unskilled labor
by capital for the purpose of cost minimization in the production process of Z and at the same time causes capital outflow in the
urban area. Thus, the supply for capital in the urban area will decline, which implies the same reduction in the demand side of
urban capital as the interest rate of capital in the urban area remains unchanged. That is, capital will be partly replaced by un-
skilled labor in sector X, which leads to more job positions created for unskilled workers. Furthermore, dissimilar to our basic
model, the newly increased jobs in sector X will definitely fully assimilate the newly migrant population released from the
rural sector. This finally makes both the urban unemployment rate and the total unemployed population descend.
Lastly, because the prices of high-skill goods (i.e., Y and M) and corresponding factors (i.e., skilled labor and urban capital) are
all independent of the degree of capital market distortion,14 there exists the Rybczynski effect in these two sectors. Specifically,
capital outflow causes a decrease in the capital endowments in the urban area when the capital market distortion improves.
Thus, the skilled-labor intensive sector M will experience expansion and prosperity, while the capital intensive sector Y will cer-
tainly shrink.15
At the end of this section, we examine the change of social welfare with regard to capital market distortion. Function V(DX, DM, DZ)
is applied to characterize social welfare, while now the national budget constraint is given as follows:

pX DX þ pM DM þ DZ ¼ pX X þ pM M þ Z; ð32Þ

where all the notations and settings are identical with those in Section 2.

13
See Appendix B.1.
^ ^ ^
14
As for the price of Y, we have pα^Y ¼ θSY wα^S þ θKY rα^U ¼ 0.
Y^ ^ ^
15
In fact, by simple calculation, we have α^ ¼ − θUZ ½ðð1þλÞλKZλλKZUXλ−λ
SM ðð1þλÞλUX θKZ þλUZ σ Z Þ
KX λUZ ÞλSM −ðλKY λSM −λKM λSY ÞλUZ 
N0 and M
α
λSY Y
^ ¼ − λSM α
^ b0, which just verify the Rybczynski theorem in our
model. In accordance with our analysis, developing countries' share of world exports in many high-skill sectors such as communication industry has experienced a
sustained growth in recent years.
110 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115

Then the change of the optimal social welfare V⁎ is given by:16

" !#
dV  dλ Z^
¼ V DZ −U 1 wU þ ðα−1ÞK 3 r U −σ Z : ð33Þ
dα dα α^

Z^
Solving the new equilibrium system for α^ and using Inequalities (11) and (25), we have:

Z^ θKZ ½ð1 þ λÞλUX þ λUZ σ Z ðλKY λSM −λKM λSY þ λKX λSM Þ þ ð1 þ λÞθUZ λKZ λSM λUX σ Z
¼ b 0: ð34Þ
α^ θUZ f½ð1 þ λÞλKZ λUX −λKX λUZ λSM −ðλKY λSM −λKM λSY ÞλUZ g

^
λ wU ^
Given Inequality (34) and ^
α
¼ − 1þλ
λ α ^ N 0, we have:

dV 
b 0: ð35Þ
dα

Similar to our basic model, we can immediately get Proposition 4 based on Inequality (35).

Proposition 4. With the improvement of capital market distortion, there is a welfare gain when a high-skill tradable sector is intro-
duced into the economy.

Like our basic model, the economic intuition behind Proposition 4 is obvious. The improvement of capital market distortion, as
has been analyzed in Proposition 3, will optimize allocation of resources since the wage inequality, the urban unemployment rate,
and the total unemployed population all descend. This gives rise to a welfare-enhancing effect.
Besides the interpretation above, in the following part, we will illustrate the impact of capital market distortion on social
welfare from another perspective. To do this, we firstly rewrite the change of V ⁎ in another way:

!
dV

ðpX X þ pM M þ Z ÞV DZ ^
X ^
M Z^
¼ ΠX þ ΠM þ ΠZ ; ð36Þ
dα α ^
α ^
α α^

pX X ^ ^ ^
where Πi(i = X , M , Z) measures sector i's share in the economy (e.g., Π X ¼ p Xþp ), αX^ N 0, M
^ b 0, and α
α ^ b 0.
Z
X M MþZ
^
Therefore, given αb0, the change of social welfare is determined by three effects according to Eq. (36): one contraction effect of
sector X, and two expansion effects of sectors M and Z. Moreover, each sector's share in the economy measures the weight put on
its corresponding effect. Together with Inequality (35), it is clear that when the capital market distortion improves, the two ex-
pansion effects of the high-skill tradable sector and the agricultural sector will definitely dominate the contraction effect of the
urban low-skill sector, which leads to an increase in national income and social welfare.

4. Concluding remarks

Based on Khan and Naqvi (1983), we investigate how the change of rural–urban capital market distortion influences skilled–
unskilled wage disparity and urban unemployment in developing countries. Our perspective is quite different from the existing
literature. We find that the improvement of capital market distortion will decrease the wage inequality and the urban unemploy-
ment rate. The latter shows the reestablishment of Khan and Naqvi's (1983) finding. However, the existence of a skilled-labor in-
tensive exportable sector does affect the final variation of the total unemployed population.
This paper suggests that governments in developing countries should actively promote the rural–urban capital market integra-
tion. In the era of globalization, a growing number of developing countries are engaged in global manufacturing and exporting
manufactured goods to developed countries. In this context, according to the results obtained in this paper, apart from the con-
traction in the skilled–unskilled wage gap, the alleviation of capital market distortion will also create more employment opportu-
nities to assimilate migrant workers, reducing the urban unemployment rate as well as the total unemployment in absolute
amount. More importantly, just as is analyzed by this paper, national income and social welfare will also increase with the
improvement of capital market distortion.
Here, we point out several further extensions for future studies. Firstly, the rural sector in our paper may be replaced by two
sectors. One rural sector uses capital with a higher interest rate, while the other rural sector employs capital with a lower interest
rate equal to that in the urban sector. Secondly, on the basis of the above extension, there may simultaneously exist an informal
sector in the urban area, which uses capital with a higher interest rate in contrast with the formal urban sector. Thirdly, the role of
foreign capital can be considered when there exists the capital market distortion. In this situation, foreign capital may act as a
distortion adjustor in reducing the wage inequality and the urban unemployment rate in developing countries.

16
See Appendix B.3.
 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115 111

Acknowledgments

We would like to express our sincere thanks to two anonymous reviewers and the editor Hamid Beladi for their useful and
detailed comments and suggestions on improving this paper. We also thank Yu Zhou and Jun Yin for their helpful discussions.
Pi acknowledges the financial support provided by the Program for New Century Excellent Talents in University, the Fundamental
Research Funds for the Central Universities, and the Collaborative Innovation Center for China Economy. Any remaining errors are
ours.

Appendix A. The basic model

A.1. Comparative static analysis

Differentiating Eqs. (2)–(10) and expressing the outcomes in the matrix form, we have:
2 32 3
θSY 0 θKY 0 −1 0 0 0 0 w^S
6
6 0 0 θKX 0 θYX 0 0 0 0 7 6^ 7
76 wU 7
6
6 0 θUZ 0 θKZ 0 0 0 0 0 7 6^ 7
76 r U 7
6
6 0 0 1 −1 0 0 0 0 0 7 6^ 7
76 r R 7
6
6 0 ð1 þ λÞwU 0 0 0 0 0 0 λwU 7 6 ^Y 7
76 p 7
6 −σ θ
6 Y KY 0 σ Y θKY 0 0 0 1 0 0 7 6^ 7
76 X 7
6λ σ θ
6 KY Y SY λKZ θUZ σ Z A −λKZ θUZ σ Z λKX SKY λKX λKY λKZ 0 7 6^ 7
76 Y 7
4 0 −λUZ θKZ σ Z ð1 þ λÞλUX sUK λUZ θKZ σ Z ð1 þ λÞλUX SUY ð1 þ λÞλUX 0 λUZ λλUX 54 Z^ 5
−1 λ^
20 3 0 SYK 0 SYY 1 0 0
ðA  1Þ
0
60 7
6 7
60 7
6 7
6 −α ^ 7
6 7
¼ 60 7
6
7;
60 7
6 7
60 7
6 7
40 5
0

where θij is the distributive share of the ith factor in the jth sector (e.g, θSY =wSaSY/pY), λij is the allocative share of factor i in thejth sector
^ −a ^KY
(e.g.,λUX ¼ XaUX =U), σi (i=Y,Z) is the elasticity of substitution between the corresponding factors in sector i (e.g., σ Y ¼ a^rSYU − w^ S ), Sij is the
partial elasticity of substitution between factors i and j in sector X (e.g., SUK ¼ ∂a UX r U
∂r aUX
), and A=λ KXSKK −σ YθSYλKY b 0.
U

A.2. Dynamic adjustment process and stability

The dynamic adjustment process is based on the principle of the excess demand function, which includes both the Marshallian
quantity adjustment process in product markets and the Walrasian price adjustment process in factor markets. The differential
equations of the dynamic adjustment process for the basic model can be specified as below:

X_ ¼ d1 ðpX −aUX w−aKX rU −aYX pY Þ; ðA  2Þ

Y_ ¼ d2 ðpY −aSY wS −aKY r U Þ; ðA  3Þ

Z_ ¼ d3 ð1−aUZ wU −aKZ r R Þ; ðA  4Þ

 
_ S ¼ d4 aSY Y−S ;
w ðA  5Þ

r_ U ¼ d5 aKX X þ aKY Y þ aKZ Z−K ; ðA  6Þ

r_ R ¼ d6 ðαr U −rR Þ; ðA  7Þ

 
_ U ¼ d7 aUX X ð1 þ λÞ þ aUZ Z−U ;
w ðA  8Þ
112 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115

p_ Y ¼ d8 ðaYX X−Y Þ; ðA  9Þ

λ_ ¼ d9 ½w−ð1 þ λÞwU ; ðA  10Þ

_ S ¼ dw
where “⋅” denotes differentiation with regard to time (e.g., w dt
S
), and parameter di is strictly positive measuring the speed rate
of adjustment.
The determinant of the Jacobian matrix of Eqs. (A-2)–(A-10) is given by:


0 0 0 0 −θKX 0 0 −θYX 0

0 0 0 −θSY −θKY 0 0 1 0

0 0 0 0 0 −θKZ −θUZ 0 0
0 1 0 −σ Y θKY σ Y θKY 0 0 0 0
d1 :::d9 S K UpX
j J1 j ¼ λKX λKY λKZ σ Y θSY λKY A −σ Z λKZ θUZ σ Z λKZ θUZ SKY λKX 0
wS r U XZ
0 0 0 0 1 −1 0 0 0
ð1 þ λÞλ
UX 0 λUZ 0 ð1 þ λÞλUX SUK θKZ λUZ σ Z −θKZ σ Z λUZ ð1 þ λÞλUX SUY λUX

1 −1 0 0 SYK 0 0 SYY 0
0 0 0 0 0 0 −ð1 þ λÞ 0 −1

Thus, given Eq. (A-1), it is obvious that:

d1 :::d9 S K UpX Δ
j J1 j ¼ ; ðA  11Þ
λwU wS r U XZ

where Δ is the value of the determinant of the coefficient matrix of Eq. (A-1).
According to the relevant theory of differential equation, a sufficient condition for the local stability of the equilibrium solution
of the system is that the real parts of all the characteristic roots of the Jacobian matrix are strictly negative, which implies that
sign|J| = (−1)N, where N is the number of rows or columns of the Jacobian matrix. Therefore, we have | J1 | b 0, which is equiv-
alent to the expression Δ b 0.

A.3. Social welfare analysis

The optimization problem for the economy is given as follows:

Max V ðDX ; DZ Þ
DX ;DZ

s:t: pX DX þ DZ ¼ pX X þ Z

Therefore, we have:
 
dV  dX dZ
¼ V DZ pX þ : ðA  12Þ
dα dα dα

In the following part, we will analyze the problem from the perspective of profit maximization, whereas in Section 2, the
general equilibrium system is built by using cost minimization method.
Given Eq. (1), we have:
 
1 1 1
pX dX ¼ pX F U dU 1 þ F K dK 1 þ F Y dY ¼ wdU 1 þ r U dK 1 þ pY dY; ðA  13Þ

1
where F 1i ¼ ∂∂iF (i = U , K , Y), and the second equality holds because each factor is paid according to the value of its marginal
product under perfect competition.17
Similarly, we have:

pY dY ¼ wS dS þ r U dK 2 ; ðA  14Þ

17
The optimization problem of the firm in sector X is given by: Max pX F 1 −wU 1 −r U K 1 −pY Y. Solving the first-order conditions (FOCs) yields: pX F 1U ¼ w, pXF1K =rU,
U 1 ;K 1 ;Y 1
and pXF1Y =pY.
 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115 113

dZ ¼ wU dU 3 þ r R dK 3 : ðA  15Þ

Combining Eqs. (A-13), (A-14), (A-15), and (5) with Eq. (6), and noticing that S ¼ S, ð1 þ λÞU 1 þ U 3 ¼ U, and K 1 þ K 2 þ K 3 ¼ K,
we can obtain:

pX dX þ dZ ¼ −U 1 wU dλ þ ðα−1ÞrU dK 3 : ðA  16Þ

On the other hand, we have:

  h i  
dK 3 ¼ dðZaKZ Þ ¼ K 3 a^KZ þ Z^ ¼ K 3 ðw
^ U −^rR Þσ Z θUZ þ Z^ ¼ K 3 Z−σ
^ ^
Z rR ; ðA  17Þ

where the last equality is obtained according to Eq. (15).

^ and combining Eqs. (12), (A-12), (A-16), and (A-17) with Z,
Solving Eq. (A-1) for Z, ^ by rearrangement and simplification, we
can get Eq. (23) in Section 2.

Appendix B. The extended model

B.1. Comparative static analysis

Notice that the newly constructed model has the decomposable feature. Taking total differentiation of Eqs. (2) to (5), and (26),
we can get the following subsystem:
2 32 3 2 3
θKX 0 0 0 θYX ^r U 0
6 θKY −1 7 6 7 6 7
6 0 0 θSY 76 ^r R 7 6 0 7
6 0 θKZ θUZ 0 0 7 6 ^ 7 ¼ 6 0 7: ðB  1Þ
6 76 w U7 6 7
4θ
KM 0 0 θSM 0 54 w ^S 5 40 5
1 −1 0 0 0 ^Y
p −α^

Taking total differentiation of Eqs. (6), (9), (10), (27), and (28) to form the equations of the remaining system and substituting
the solutions of Eq. (B-1) (i.e., ^r U , ^r R , w
^ U, w
^ S , and p ^ Y,
^Y ) into them, we can obtain the remaining variables X, ^ Z,
^ M,
^ and λ.
^

B.2. Dynamic adjustment process and stability

In line with the extended model, the following differential equation is added to the adjustment process:

_ ¼ d ðp −a w −a r Þ:
M ðB  2Þ
4 M SM S KM U

In addition, Eqs. (A-5) and (A-6) should be replaced by:

 
_ S ¼ d5 aSY Y þ aSM M−S ;
w ðB  3Þ

r_ U ¼ d6 aKX X þ aKY Y þ aKM M þ aKZ Z−K : ðB  4Þ

Then, we combine Eqs. (A-2) to (A-4) with Eqs. (B-2) to (B-4), (A-7) to (A-10) (notice that d6, d7, d8, and d9 should be
substituted by d7, d8, d9, and d10, respectively) to form the new equations of adjustment process.
As in Appendix A.2, we can calculate the determinant of the corresponding Jacobian matrix:

d1 :::d10 pX pM SKU
j J2 j ¼ f−θUZ ½θKX θSM þ ðθKY θSM −θKM θSY ÞθYX ½ðð1 þ λÞλKZ λUX −λKX λUZ ÞλSM
r U wS XZM
−ðλKY λSM −λKM λSY ÞλUZ g:

Dissimilar to Appendix A.2, here we have |J2 | N 0 since the number of rows (or equivalently, columns) of the new Jacobian
matrix is 10.
Thus, we have:
[θKXθSM + (θKYθSM − θKMθSY)θYX][((1 + λ)λKZλUX − λKXλUZ)λSM − (λKYλSM − λKMλSY)λUZ] b 0, which is compatible with conditions
(11), (24), and (25).
114 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115

B.3. Social welfare analysis

The optimization problem for the economy is given by:

Max V ðDX ; DM ; DZ Þ
DX ;DM ;DZ

s:t: pX DX þ pM DM þ DZ ¼ pX X þ pM M þ Z

Thus, we can obtain:

 
dV  dX dM dZ
¼ V DZ pX þ pM þ : ðB  5Þ
dα dα dα dα

Here, Eq. (1) should be substituted by:

8 1
>
> X ¼ F ðU 1 ; K 1 ; Y Þ
>
< 2
Y ¼ F ðS2 ; K 2 Þ
; ðB  6Þ
>
> Z
3
¼ F ðU 3 ; K 3 Þ
>
: 4
M ¼ F ðS4 ; K 4 Þ

where S2 þ S4 ¼ S, ð1 þ λÞU 1 þ U 3 ¼ U, K 1 þ K 2 þ K 3 þ K 4 ¼ K.
By applying the same approach as in Appendix A.3, we have:

dX dM dZ dλ ^
K
pX þ pM þ ¼ −U 1 wU þ ðα−1ÞK 3 r U 3 ; ðB  7Þ
dα dα dα dα ^
α

^  
K 3
^ þ Z^
a ^ U ^r R
w Z^ ^r Z^ Z^
¼ KZ ¼ − σ Z θUZ þ ¼ −σ Z R þ ¼ −σ Z þ : ðB  8Þ
α^ α^ α^ α^ α^ ^
α α ^ α^

Combining Eqs. (B-5) and (B-7) with Eq. (B-8) yields Eq. (33) in Section 3.

References

Anwar, S. (2006). Factor mobility and wage inequality in the presence of specialization-based external economies. Economics Letters, 93(1), 88–93.
Anwar, S. (2008). Labour inflow induced wage inequality and public infrastructure. Review of Development Economics, 12(4), 792–802.
Anwar, S. (2010). Wage inequality, increased competition, and trade liberalization: Short run vs long run. Review of International Economics, 18(3), 574–581.
Anwar, S., & Rice, J. (2009). Labour mobility and wage inequality in the presence of endogenous foreign investment. Economic Modelling, 26(6), 1135–1139.
Anwar, S., & Sun, S. (2012). Trade liberalisation, market competition and wage inequality in China's manufacturing sector. Economic Modelling, 29(4), 1268–1277.
Anwar, S., & Sun, S. (2015). Taxation of labour income and the skilled–unskilled wage inequality. Economic Modelling, 47, 18–22.
Beladi, H., & Naqvi, N. (1988). Urban unemployment and non-immiserizing growth. Journal of Development Economics, 28(3), 365–376.
Beladi, H., Chakrabarti, A., & Marjit, S. (2010). Skilled–unskilled wage inequality and urban unemployment. Economic Inquiry, 48(4), 997–1007.
Beladi, H., Chakrabarti, A., & Marjit, S. (2011). North-South outsourcing, immigration, and skilled wages: Through the lens of incomplete contracts. Review of
Development Economics, 15(3), 417–428.
Bhattacharyya, A., & Kumbhakar, S. C. (1997). Market imperfections and output loss in the presence of expenditure constraint: A generalized shadow price approach.
American Journal of Agricultural Economics, 79(3), 860–871.
Bose, P. (1998). Formal-informal sector interaction in rural credit markets. Journal of Development Economics, 56(2), 265–280.
Chao, C. -C., & Yu, E. S. H. (1992). Capital markets, urban unemployment and land. Journal of Development Economics, 38(2), 407–413.
Chaudhuri, S. (2000). Rural–urban migration, informal sector, urban unemployment, and development policies: A theoretical analysis. Review of Development
Economics, 4(3), 353–364.
Chaudhuri, S. (2001). Interaction of formal and informal credit markets in backward agriculture: A theoretical analysis. Indian Economic Review, 36(2), 411–428.
Chaudhuri, S. (2007). Foreign capital, welfare and urban unemployment in the presence of agricultural dualism. Japan and the World Economy, 19(2), 149–165.
Chaudhuri, S., & Banerjee, D. (2010). FDI in agricultural land, welfare and unemployment in a developing economy. Research in Economics, 64(4), 229–239.
Chaudhuri, S., & Dastidar, K. G. (2011). Corruption in a model of vertical linkage between formal and informal credit sources and credit subsidy policy. Economic
Modelling, 28(6), 2596–2599.
Chaudhuri, S., & Gupta, M. R. (1996). Delayed formal credit, bribing and the informal credit market in agriculture: A theoretical analysis. Journal of Development
Economics, 51(2), 433–449.
Chaudhuri, S., & Gupta, M. R. (2014). International factor mobility, informal interest rate and capital market imperfection: A general equilibrium analysis. Economic
Modelling, 37, 184–192.
Chaudhuri, S., & Yabuuchi, S. (2007). Economic liberalization and wage inequality in the presence of labour market imperfection. International Review of Economics and
Finance, 16(4), 592–603.
Chaudhuri, S., Yabuuchi, S., & Mukhopadhyay, U. (2006). Inflow of foreign capital and trade liberalization in a model with an informal sector and urban unemployment.
Pacific Economic Review, 11(1), 87–103.
Chen, B., & Lin, J. Y. (2014). Development strategy, urbanization and the urban–rural income gap in China. Social Sciences in China, 35(1), 5–20.
Das, S. P. (2002). Foreign direct investment and the relative wage in a developing country. Journal of Development Economics, 67(1), 55–77.
Das, S. P. (2005). Gradual globalization and inequality between and within countries. Canadian Journal of Economics, 38(3), 852–869.
Feenstra, R. C., & Hanson, G. H. (2003). Global production sharing and rising wage inequality: A survey of trade and wages. In K. Choi, & J. Harrigan (Eds.), Handbook of
international trade. London: Basil Blackwell.
 J. Pi, X. Chen / International Review of Economics and Finance 42 (2016) 103–115 115

Gupta, M. R. (1997a). Informal sector and informal capital market in a small open less-developed economy. Journal of Development Economics, 52(2), 409–428.
Gupta, M. R. (1997b). Foreign capital and the informal sector: Comments on Chandra and Khan. Economica, 64(254), 353–363.
Gupta, M. R., & Chaudhuri, S. (1997). Formal credit, corruption and the informal credit market in agriculture: A theoretical analysis. Economica, 64(254), 331–343.
Harris, R., & Todaro, M. (1970). Migration, unemployment, and development: A two sector analysis. American Economic Review, 60(1), 126–142.
Khan, M. A., & Naqvi, S. N. H. (1983). Capital markets and urban unemployment. Journal of International Economics, 15(3–4), 367–385.
Lin, J. Y. (2003). Development strategy, viability and economic convergence. Economic Development and Cultural Change, 51(2), 277–308.
Lin, J. Y. (2009). Economic development and transition: Thought, strategy, and viability. Cambridge: Cambridge University Press.
Lin, J. Y., Cai, F., & Li, Z. (1994). China's miracle: Development strategy and economic reform. Shanghai: Shanghai Joint Publishing Company and Shanghai People's
Publishing House.
Marjit, S., Beladi, H., & Chakarabarty, A. (2003). Trade and wage inequality in developing countries. Economic Inquiry, 42(2), 295–303.
Moore, M., & Ranjan, P. (2005). Globalization vs. skill biased technical change: Implications for unemployment and wage inequality. Economic Journal, 115(503),
391–422.
Pan, L. (2014). The impacts of education investment on skilled–unskilled wage inequality and economic development in developing countries. Economic Modelling, 39,
174–181.
Pan, L., & Zhou, Y. (2013). International factor mobility, environmental pollution and skilled–unskilled wage inequality in developing countries. Economic Modelling, 33,
826–831.
Pi, J., & Zhou, Y. (2012). Public infrastructure provision and skilled–unskilled wage inequality in developing countries. Labour Economics, 19(6), 881–887.
Pi, J., & Zhou, Y. (2013). Institutional quality and skilled–unskilled wage inequality. Economic Modelling, 35, 356–363.
Romatet, E. (1983). Calcutta's informal sector: Theory and reality. Economic and Political Weekly, 18(50), 2115–2128.
Wood, A. (1997). Openness and wage inequality in developing countries: The Latin American challenge to East Asian conventional wisdom. World Bank Economic
Review, 11(1), 33–57.
Yabuuchi, S., & Beladi, H. (2001). Urban unemployment, informal sector and development policies. Journal of Economics, 74(3), 301–314.
Zhang, J. (2012). Inflow of labour, producer services and wage inequality. Economics Letters, 117(3), 600–603.
Zhang, J. (2013). Factor mobility and skilled–unskilled wage inequality in the presence of internationally traded product varieties. Economic Modelling, 30(1), 579–585.
Zhuang, J. (1996). Estimating distortions in the Chinese economy: A general equilibrium approach. Economica, 63(252), 543–568.