Bulletin of Monetary Economics and Banking, Vol. 22, No. 2 (2019), pp. 177 - 194

p-ISSN: 1410 8046, e-ISSN: 2460 9196

FINANCIAL STRUCTURE FOUNDATION OF THE URBAN–

RURAL INCOME GAP IN CHINA:

AN INVESTIGATION FROM THE PERSPECTIVE

OF THE DOUBLE DUAL STRUCTURE

Quan-Jing Wang1, Gen-Fu Feng1, Chun-Ping Chang2

1School of Economics and Finance, Xi’an Jiaotong University, Shaanxi, China. 2 Shih Chien University, Kaohsiung, Taiwan. Email: cpchang@g2.usc.edu.tw

ABSTRACT

This article is an empirical analysis of the relations between financial structure and the urban–rural income gap (URIG) in China’s economic transition, based on the country’s double dual structure. We employ data of 31 provinces in China from 2001 to 2016 to empirically study the influence of financial structure on the URIG. We find an inverted U-shaped relation between financial scale and the URIG, a positive impact of urban and rural financial structure on the URIG, and an inverted U-shaped relation between the mismatch of financial resources and the URIG. These findings show that selection of the optimal proportion of the state-owned economy through ownership reform and the promotion of financial development and optimization of the allocation of financial resources are two effective ways to reduce the URIG.

Keywords: Urban–rural income gap; Financial structure; Double dual structure; Financial resources mismatch.

JEL Classification: H73; O16; R58.

Article history:
Received	: January 19, 2019
Revised	: July 15, 2019
Accepted	: July 16,	2019
Available online : July 30,		2019

https://doi.org/ 10.21098/bemp.v22i2.1079

178	Bulletin of Monetary Economics and Banking, Volume 22, Number 2, 2019

I. INTRODUCTION

The increasing gap between rich and poor in China is a social problem of growing concern. Since 2000, China’s Gini coefficient has continued to exceed the threshold of 0.4, which is considered alarming. Income inequality has become the main source of the downward pressure on China’s economy (Chen and Fleisher, 1996; Lee et al., 2017). The ever-widening gap between the rich and the poor has reduced the total consumption of residents, restricted the improvement of the quality of labor, and triggered social instability, all of which have reduced the speed and quality of economic growth (Lu and Chen, 2004). The gap has become a major obstacle to the balanced development of China’s economy. The internal clarification of the gap between the rich and the poor in China has important theoretical and practical significance for improving China’s distribution status, increasing its economic growth rate, and achieving more balanced and persistent economic development.

In China, income inequality is mainly due to the urban–rural income gap (Molero-Simarro, 2016, hereafter denoted as URIG). For example, the per capita disposable income of urban residents in China was 33,616 yuan in 2016, whereas that of rural residents was 12,363 yuan, a difference of 2.72 times. From the trends in the per capita income gap between urban and rural residents and the Gini coefficient of China from 1990 to 2015, we see that these trends are similar, implying that the URIG and Gini coefficient change simultaneously. Further investigation of the URIG between different regions in China reveals differences. In 2015, the per capita income differences between urban and rural residents in eastern, central, and western China were 3.37, 3.44, and 4.24 times, respectively. It can be seen that provinces with a large URIG are mainly concentrated in the western region. This means that the URIG in underdeveloped areas is more serious when we consider regional differences.

Why, then, is China’s URIG still so high? Why is there such a big difference in the URIG between different regions? To answer these questions, we must analyze China’s basic economic system and economic structure and consider its economic characteristics. During the transition period from planned economy to market economy, the Chinese economy has maintained the characteristics of double dual structure. First, in China, the urban and rural sectors coexist, constituting the urban–rural dual structure of the country’s economy. Second, the urban sector comprises two parts, the state-owned economy and the non–state-owned economy, constituting another dual structure of the Chinese economy. This double dual structure thus provides an important perspective from which to understand the persistence of URIG and inconsistent performance in different regions.

There is a significant difference between the urban–rural dual sectors in the ability of residents to gain access to financial resources. In the early stage of financial development, urban residents have more convenient access to financial resources than rural residents. They also have rich financial knowledge and can better use financial resources to enhance their income, thus widening the income gap between urban and rural areas. However, eventually, rural residents’ access to financial resources improves, and the income gap between urban and rural areas narrows. We thus observe an inverted U-shaped relation between financial development and the URIG (Kim and Lin, 2011). In addition, in the context of the urban–rural dual structure, finance also presents urban–rural structure

Financial Structure Foundation of the Urban–Rural Income Gap in China:	179
An Investigation from the Perspective of the Double Dual Structure

characteristics, with financial resources concentrated in the urban sector (Zhang and Chen, 2015). The financial urban–rural structure lies at the heart of the ability of the real economy to obtain financial services, resulting in imbalance at the level of industrial structure and economic growth between urban and rural sectors, which can eventually lead to greater income inequality through the channel of labor mobility and the “slumping effect” (Cheong and Wu, 2014; Johansson and Wang, 2014). Finally, according Zhang and Chen (2015), financial development provides external financing support for the construction of local public facilities. The financial structure determines the local urbanization process and will indirectly affect the URIG through urbanization (Su et al., 2015).

The mechanism of how financial development influences income inequality can be described as follows: First, the financial sector accelerates the efficiency of capital, which promotes the flow of rural savings into urban investment. Due to the continuous improvement of the quality and quantity of human capital investment in urban areas, the proficiency of financial capital can rapidly grow there. The income of urban residents therefore exhibits a quickly grows. Second, due to the lack of capital in rural areas, industry there cannot be upgraded, such that more labor is concentrated on agriculture. Due to rural residents’ ability to work only in the rural sector, their income growth is also relatively slow, due to the better ability to work of urban residents, this eventually increases the level of URIG. In addition, given a lack of financial resources, the government’s unequal education policy between urban and rural areas increases the relative cost of human capital investment of rural residents above that of urban areas.

This paper makes the following contributions. First, different from the previous literature on the impact of financial development on the URIG, our work portrays the financial structure along the following three dimensions: financial development, the urban–rural financial structure, and the mismatch of financial resources. This approach allows us to comprehensively examine the relation between financial structure and the URIG. Second, unlike previous studies, this paper investigates the influence of financial structure on income inequality, as well as robustness tests, handling potential endogeneity using an instrumental variable (IV) two-stage least squares (2SLS) approach. Third, the full sample is split into three subsamples for the eastern, central, and western regions, respectively, allowing us to examine the differences in the relation between financial structure and the URIG between the different regions. This more accurate analysis can offer more targeted policy recommendations for the government.

The remainder of this paper is organized as follows. Section II reviews the literature on the relations between financial structure and income inequality. Section III describes the research design, providing detailed information on the data, variables, and measurement models. Section IV presents analyses of static effects, robustness tests, and the empirical results for the three subsamples. The last section draws our conclusions and provides policy recommendations.

II. LITERATURE REVIEW

The relations between financial development and the income gap between urban and rural residents has been given more importance by scholars and the

180	Bulletin of Monetary Economics and Banking, Volume 22, Number 2, 2019

governments (de Haan and Sturm, 2017; Ghossoub and Reed, 2017; Baiardi and Morana, 2019). Mookerjee and Kalipioni (2010) point out that the availability of financial resources lies at the heart of income inequality. Moreover, as Lee et al. (2017) proposed, financial development could affect income inequality due to unequal access to financial resources. Hsieh et al. (2019) find that income inequality increases with financial deepening but decreases with a more market-oriented financial system.

There are three viewpoints on the relations between financial development and income inequality. First, Greenwood and Jovanovic (1990) note an inverted-U relation between financial development and income inequality. Similarly, Kim and Lin (2011) find that the benefits of financial development on income distribution arise only when the country has reached a threshold level of financial development. By using a dynamic panel data model and data on 35 developing countries, Tan and Law (2012) provide strong evidence of a U-shaped relations between financial development and income inequality.

However, on the contrary, Galor and Zeira (1993) find that financial development exerts a negative effect on income inequality, since financial development provides the poor greater access to financial resources. In line with these authors’ viewpoint, Hamori and Hashiguchi (2012), Kunieda et al., (2014), and Naceur and Zhang (2016) also illustrate this idea. Similarly, other researchers claim that regional inequality can be alleviated by speeding up financial reforms to improve access to finance for the inland provinces (Zhang et al., 2007; Wang et al., 2015). By using data on 49 countries, Li et al. (1998) point out that financial development plays a major role in reducing income inequality.

Another strand of literature, however, claims that financial development promotes income inequality (e.g., Gregorio, 1996; Li and Yu, 2014; Dabla-Norris et al., 2015; Jauch and Watzka, 2016). Using a cross-country panel data of developing countries, Seven and Coskun (2016) find that, although bank development contributes to economic growth, it also exerts a positive impact on the growth of the Gini coefficient.

We now turn to the literature on the relations between financial development and income inequality in China. As Zhang et al. (2007) show, the rising regional disparity in post-reform China can be partly attributed to the unequal financial development between the coastal and inland provinces. Since the reform in the 1980s, the importance of financial structure in income distribution in China has grown and received wide attention from scholars in the context of financial marketization (Gimet and Segot, 2011; Johansson and Wang, 2014; Zhang and Chen, 2015; Lee et al., 2017; Wen et al., 2018). Employing data from China and structural vector autoregression, Zhang and Chen (2015) find that the inverted-U relations between financial development and income inequality in China. Furthermore, by using a regression-based inequality decomposition approach and data on China from 2006 to 2016, Lee et al. (2017) show that financial development is a crucial factor in regional inequality.

Scholars have investigated how urban–rural financial structure influences income inequality. For instance, from a macro perspective, Zhang and Chen (2015) point out that the URIG is gradually expanding due to the differential growth in income between urban and rural residents. Therefore, the urban–rural financial

Financial Structure Foundation of the Urban–Rural Income Gap in China:	181
An Investigation from the Perspective of the Double Dual Structure

structure mechanism acting on the URIG is in line with the mechanism of the impact of financial development on income growth (Li et al., 2018). Theory shows that urban–rural financial structure mainly affects the income of urban and rural residents through the following channels.

The first channel is the resource allocation effect. The ability of financial development to integrate labor, capital, and other factors is an important part of this channel’s function. Therefore, with the expansion of urban–rural financial structure, more financial resources are allocated to the urban sector, which promotes the growth of urban residents’ income; the massive outflow of factors thus makes the income growth of rural residents more vulnerable. The second channel is the effect of technological innovation. Since technological innovation lies at the heart of economic development, the ability of financial development to promote technological innovation has become an essential consideration in the interpretation of financial functions. Technology innovation is a new source of income growth. The unbalanced development of urban–rural finance leads to innovative capital financing being mainly focused in the urban sector, with the opposite effect on the residents’ income. The third channel is the effect of financing. Given a deteriorating urban–rural financial structure, rural residents are subject to credit constraints far higher than in the urban sector, as well as constraints due to the financing of production activities and human capital, thus limiting their income growth. The fourth channel involves transaction costs. As urban financial institutions continue to expand, the level of urban financial development continues to be high, which helps reduce the institutions’ transaction costs, since innovative services launched by the financial sector mainly benefit urban residents. Therefore, overall, the imbalance between urban and rural financial development has completely different effects on the incomes of urban and rural residents, broadening the income gap between urban and rural areas (Wen et al., 2018).

III.DATA AND METHODOLOGY

A. Data Set

This paper focuses on the impact of financial structure on the URIG gap. Since the URIG is also affected by other factors, we control for these variables, presented below.

We use the urban–rural income gap (denoted by Urgap) as our dependent variable. Although there are many different inequality indicators, Cheong and Wu (2014) suggest that the most common is the Gini coefficient. Furthermore, previous studies attribute the increase in China’s Gini index to urban–rural inequality (Wu and Perloff, 2005; Benjamín et al., 2007; Li, 2010). Su et al. (2015) point out that the URIG is a general indicator for analyzing income inequality in China, due to the dual structure in the transition period from planned economy to market economy.

Hence, in line with Molero-Simarro (2017), we also use the URIG to capture income inequality in China, which is calculated by the proportion of the per capita income of urban residents to that of rural residents. The URIG is measured by the ratio of the per capita income of urban residents to that of rural residents. To eliminate the influence of inflation, urban and rural consumer price indexes are used to calculate

182	Bulletin of Monetary Economics and Banking, Volume 22, Number 2, 2019

the real per capita income of urban and rural regions, respectively. A higher value of this indicator implies a greater income gap between urban and rural areas.

The independent variables involve financial structure. We use the following three indicators to measure regional financial structure. 1) Our financial scale (Fir) is similar to the measure of Lee et al. (2017), who calculate the proportion of the deposits and loan balance of local financial institutions to the regional gross domestic product (GDP) to capture the regional financial scale. 2) We measure the urban–rural financial structure (Urf) according to the method of Lu and Chen (2004), that is, as the ratio of urban loans to total urban–rural loans. Urban loans are then measured by subtracting agricultural loans from total amount of loans. A higher Urf value indicates more financial resources allocated to the urban sector.

3)Our third measure is the mismatch of financial resources (Fm). Since data on loans for the state-owned sector are difficult to obtain, we calculate the ratio of the financing costs of regional private enterprises to that of state-owned enterprises. This indicator indirectly reflects the proportion of state-owned sector financing. A higher degree of financial resource mismatch indicates a larger proportion of financing in the state sector.

We use the following explanatory variables in our analysis.

1.Urbanization (Urban). By applying of bootstrap panel Granger causality estimations, Su et al. (2015) suggest that urbanization does Granger-cause an URIG. Similarly, Lee et al. (2017) and Wu and Rao (2017) find that the promotion of urbanization expands the URIG. Following Wu and Rao, we use the proportion of the urban sector population to the total population, Urban, to capture the influence of urbanization on the URIG.

2.Industrial structure (Ind). This index is generally used to denote industrial structure in the literature and includes the proportion of the added value of the tertiary industry to the GDP, the proportion of the added value of the secondary industry to the GDP, and the ratio of the added value of the tertiary industry to that of the secondary industry (Cheong and Wu, 2014). Cheong and Wu (2014) point out that the uneven distribution of industrialization can greatly exacerbate regional disparity. In line with other studies, we calculate the ratio of the added value of the secondary industry to the total GDP in the province to capture industry structure. A higher score for this indicator means that the industrial structure is progressing toward the manufacturing industries.

3.Foreign direct investment, or FDI (FDI). In addition to promoting the inflow of technology and capital, FDI changes the host country’s dependence on capital, labor, and other factors, which affects the proportion of each factor in the initial distribution, eventually exerting a significant influence on the URIG.

Former research also provides evidence supporting this viewpoint. Lessmann (2014) points out that, during the earlier period after the economic reform in the 1980s in China, the increase in FDI led to greater income inequality, but this influence vanished—and perhaps even reversed—after the 1990s.

Following Yu et al. (2011), we employ the proportion of regional FDI to the

GDP to control for the potential influence of FDI on the URIG.

4.Infrastructure level (Road). Infrastructure, especially traffic facilities such as railways and roads, is an important factor in improving the mobility of

Financial Structure Foundation of the Urban–Rural Income Gap in China:	183
An Investigation from the Perspective of the Double Dual Structure

materials between urban and rural sectors. As Calderón and Servén (2004) point out, better regional infrastructure facilitates the transfer of materials between urban and rural areas, which will narrow the URIG. Therefore, we calculate the regional infrastructure level by the regional road length per capita.

5.Economic development level (Pergdp). Kuznets (1955) claims that, as the level of economic development increases, the national income gap will first increase and then shrink. Furthermore, Benabou (1996) associates income inequality with growth and finds income inequality convergence in various countries.

Therefore, our paper uses the real GDP per capita to measure the level of regional economic development.

6.Trade dependence (Trade). Previous research shows a U-shaped relation between trade dependence and the URIG (Wei and Zhao, 2012). However, on the contrary, Jalil (2012) states that the Kuznets curve fits the relations between openness and income equality in the case of China, implying an inverted U-shaped relation between trade and income inequality. Lin and Fu

(2015) also point out that trade openness exerts a negative effect on income inequality in autocracies, while trade expansion has a positive effect on the

Gini index in democratic countries. Unlike previous studies with nonlinear findings, however, Mah (2013) declares that trade openness positively affects income inequality. In line with the author, we therefore use the proportion of the regional trade balance to the GDP to proxy for regional trade dependence.

7.Financial expenditure intensity (Govex). According to Zhang and Chen (2015), the increase in fiscal expenditures and financial development widens the

URIG in the short term, but eventually narrows it. Accordingly, we use the proportion of fiscal expenditures to regional fiscal revenue to measure the level of regional public expenditure.

In addition, it is worth noting that our paper uses the corresponding price

index to calculate the actual values for all the nominal variables. The empirical testing uses the logarithms of all the variables. At the same time, to eliminate the influence of extreme values, the dependent variable is winsorized at the top and bottom percentiles.

The data are from the China Statistical Yearbook, the China Finance Yearbook, and the China City Statistical Yearbook and the China’s economic and social development statistical database. After filtering and deleting observations missing values, we obtain a final balanced panel of data covering 31 provinces, autonomous regions, and municipalities in China from 2001 to 2016.

B. Methodology

We examine how financial structure affects the URIG by using the following measurement model:

(1)

186	Bulletin of Monetary Economics and Banking, Volume 22, Number 2, 2019

Table 2.

Panel OLS Estimator for Full Sample

This table reports the results on the effect of financial structure on urban- rural income gap. The t-statistics are in parenthesis; ‘Year’ and ‘Province’ represents the individual fixed effect and time fixed effect, respectively; Finally, ***,

**, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

	Independent Variable		Dependent Variable: Urgap
		(1)	(2)	(3)
	Fir	0.408***
		(4.38)
	2	-0.364**
	Fir	(-2.31)
			0.156***
	URF
			(3.90)
	URF2		-0.099
			(-1.56)	0.057***
	Fm
				(3.74)
	Fm2			-0.054*
		-0.164***	-0.061***	(-1.75)
	Urban			-0.022**
		(-2.73)	(-3.01)	(-2.29)
	IND	-0.245***	-0.009***	-0.027**
		(-2.80)	(-3.23)	(-2.02)
	FDI	-0.056	0.001	-0.007
		(-1.09)	(0.07)	(-0.91)
	Road	0.039	0.022	0.007
		(0.69)	(1.20)	(0.79)
	Pergdp	0.371***	0.141***	0.055***
		(4.89)	(5.44)	(4.36)
	Trade	0.360***	0.137***	0.052***
		(5.20)	(5.71)	(4.49)
	Govex	0.402***	0.149***	0.600***
		(5.83)	(6.20)	(5.03)
	Year
		yes	yes	yes
	Province	yes	yes	yes
	Cons	2.134***	0.903***	0.365***
		(4.12)	(3.81)	(4.44)
	N	496	217	496
	R2	0.238	0.540	0.737
	F	9.472	11.957	7.110

C. Robustness Test

C1. New Index of the URIG

To verify the reliability of the benchmark results, this section changes the measure of the dependent variable and conducts a robustness test. Some use the Taylor index to calculate the URIG (Braithwaite and Braithwaite, 1980). To avoid the regression bias caused by the index, we calculate the Taylor index (TL) of the URIG and carry out a panel ordinary least squares estimation with fixed effects for robustness. The larger the value of TL, the larger the URIG. The regression results are shown in Table 3.

Table 3 displays the results of the relations between financial structure and the URIG, based on the two-way fixed effects model employing a new dependent variable, the URIG (TL). The explanatory variable in columns (1) to (3) is similar to that in Table 2. After controlling for other factors and the effects of individual

Financial Structure Foundation of the Urban–Rural Income Gap in China:	187
An Investigation from the Perspective of the Double Dual Structure

provinces and time, we find the coefficient of Fir in column (1) is 0.055 and significantly positive at the 1% level; the regression coefficient of Fir2 is found to be -0.069 and significantly negative at the 5% level. These findings provide evidence of an inverted U-shaped relations between financial development and the URIG. The coefficient of Urf in column (2) is 0.090 and significantly positive at the 1% level, while the regression coefficient of Urf2 is -0.070 and nonsignificant at the 10% level. This result indicates that the urban–rural financial structure exerts a positive effect on the URIG. Finally, the coefficient of Fm in column (3) is 0.083 and significantly positive at the 5% level, while the coefficient of Fm2 is -0.050 and significantly negative at the 5% level. This result supports an inverted U-shaped relations between financial resource mismatch and URIG. The results in Tables 2 and 3 are similar, which provides strong evidence of the reliability of the benchmark findings.

Table 3.

Robustness Test: New Measurement of Urban- Rural Income Gap

This table reports the results on the effect of financial structure on urban- rural income gap which is measured by the TL index. The t-statistics are in parenthesis; ‘Year’ and ‘Province’ represent the individual fixed effect and time fixed effect, respectively; Finally, ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

	Independent Variable		Dependent Variable: TL
		(1)	(2)	(3)
	Fir	0.055***
		(3.10)
	2	-0.069**
	Fir	(-2.16)
			0.090***
	URF
			(3.97)
	URF2		-0.070
			(-1.53)	0.083**
	Fm
				(2.21)
	Fm2			-0.050**
		-2.682*	-2.297	(-2.15)
	Urban			-4.091***
		(-1.93)	(-1.53)	(-2.96)
	IND	-4.155**	-2.569	-6.477***
		(-2.03)	(-1.29)	(-2.83)
	FDI	-2.473**	-1.085	-2.346*
		(-2.30)	(-0.97)	(-1.67)
	Road	-0.204	1.244	-0.835
		(-0.16)	(0.95)	(-0.60)
	Pergdp	5.110***	4.751**	7.576***
		(2.90)	(2.47)	(3.82)
	Trade	5.193***	5.173***	5.164*
		(3.26)	(2.91)	(1.84)
	Govex	6.294***	5.995***	7.163***
		(4.03)	(3.35)	(4.70)
	Year
		yes	yes	yes
	Province	yes	yes	yes
	Cons	0.314***	0.305**	0.031
		(3.04)	(2.45)	(0.09)
	N	496	217	496
	R2	0.631	0.735	0.313
	F	7.563	6.703	7.048

188	Bulletin of Monetary Economics and Banking, Volume 22, Number 2, 2019

C2. Endogeneity

Because the dependent and independent variables involve regional economic performance, endogeneity could exist between the variables. We employ an IV method and a 2SLS estimation method to address the influence of endogeneity on our empirical results. We select the first-order lag and the growth rate of the financial structure as the IVs of financial structure. Specifically, the first-order lag of Fir (denoted L.Fir) and the growth rate of Fir (denoted Firr) are employed as IVs of financial scale, while the first-order lag of Urf (L. Urf) and the growth rate of Urf (Urfr) are used as instrumental ariables for the urban–rural financial structure; the first-order lag of Fm (L.Fm) and the growth of Fm (Fmr) are included as IVs of financial scale.

The selection of IVs requires a validity check. In the IV–2SLS regression, we report the test of the IVs in the last rows of Table 4. First, the estimated values and t-values of the IVs in the first-stage regression are reported. The regression coefficients of the IVs are significantly positive, which indicates that the instrumental variables employed are significantly positively correlated with financial structure. Furthermore, the Cragg–Donald Wald statistic obtained from the weak instrumental variable test is very high, indicating no problem with weak instrumental variables. The p-value for over-recognizing the Hansen test results is very large, suggesting that we cannot reject the null hypothesis that the IVs are exogenous. These indicators show that the IVs we select are appropriate. Turning to the regression results in Table 4, we can see that the results in the various columns for the 2SLS instrumental variable regression are similar to the basic regression results in Table 2.

Table 4 reports the empirical results for the IV–2SLS estimation. The results for the dependent and independent variables for each column in Table 4 are consistent with those in Table 2. After controlling for the other factors and the fixed effects of provinces and time, we find the coefficient of Fir (in column (1)) to be 0.233 and significantly positive at the 1% level, while the coefficient of Fir2 is -1.469 and significantly negative at the 1% level, implying an inverted U-shaped relations between financial development and the URIG. The coefficient of Urf (in column (2)) is 0.072 and significantly positive at the 1% level, while the coefficient of Urf2 is -0.119 and nonsignificant at the 10% level. This result suggests that the urban– rural financial structure has a positive effect on the URIG. The coefficient of Fm (in column (3)) is 0.039 and significantly positive at the 1% level, while the coefficient of Fm2 is -0.475 and negative at the 1% level. This result shows an inverted U-shaped relations between financial resource mismatch and the URIG. The regression results in Table 4 are similar to those in Table 2, verifying the reliability of the benchmark results.

Financial Structure Foundation of the Urban–Rural Income Gap in China:	189
An Investigation from the Perspective of the Double Dual Structure

Table 4.

Robustness Test- IV- 2SLS Estimations

This table reports the results on the effect of financial structure on urban- rural income gap using the IV-2SLS. The t-statistics are in parenthesis; ‘Year’ and ‘Province’ represent the individual fixed effect and time fixed effect, respectively; The ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Moreover, to save space, we do not report the coefficient of control variable, which is available upon request.

	Independent Variable		Dependent Variable: Urgap
		(1)	(2)	(3)
	Fir	0.233***
		(12.56)
	2	-1.469***
	Fir	(-3.53)
			0.072***
	URF
			(15.36)
	URF2		-0.119
			(-1.04)	0.039***
	Fm
				(9.97)
	Fm2			-0.475***
				(-5.21)
	N	465	186	465
	R2	0.166	0.225	0.154
	F	73.896	64.759	174.132
	Firr	0.921*
		(1.80)
	L.Fir	0.708**
		(2.17)
			0.623**
	URFr
			(2.04)
	L.URF		0.715**
			(2.21)
				0.938*
	Fmr
				(1.74)
	L.Fm			0.160***
				(9.09)
	Cragg-Donald Wald F	127.66	137.31	114.65
	Hansen-value	1.079	0.650	0.082
	p-value	(0.299)	(0.420)	(0.774)

C3. Results for Sub- Samples

According to the previous empirical analysis, there is an inverted U-shaped relation between both financial scale and financial resource mismatch and the URIG, whereas the urban–rural financial structure exerts a positive effect on the URIG. However, the degree of market-oriented reform and financial development are not the same across different regions of China, so there are great differences in financial structure. Most previous studies split the full sample of Chinese provinces into three subsamples (i.e., eastern, central, and western regions) to determine the influence of a particular factor on income inequality. For instance, Yu et al., (2011) and Lessmann (2013) investigate the influence of FDI on income inequality for eastern, central, and western zones; Su et al. (2015) and Wu and Rao (2017) also examine the relations between urbanization and income inequality in these three regions. Is the effect of financial structure on the URIG then, consistent across different regions? By clarifying these issues, we can more specifically formulate policies to narrow the income gap between urban and rural areas.

190	Bulletin of Monetary Economics and Banking, Volume 22, Number 2, 2019

For this purpose, in line with Cheong and Wu (2014), we construct three subsamples—the eastern, central, and western regions—to explore the impact of financial structure on the URIG among the different regions of the sample provinces. The three subsamples include 12, eight, and 11 individual provinces, respectively, for a period of 16 years. The three subsamples constitute long panel data. In the long panel data regression, the assumption of an independent and identical distribution of the perturbation terms can be relaxed. The comprehensive Feasible Generalized Least Squares (FGLS) method considers intra-group autocorrelation, inter-group heteroscedasticity, and contemporaneous correlation in long panel data regressions. We therefore regress the three subsamples, using comprehensive FGLS. The regression results are shown in Table 5.

First, we examine whether financial scale exerts different impacts on the URIG between the three subsamples. The coefficient of Fir in column (1) of Table 5 is 0.214 and significantly positive at the 1% level, while the coefficient of Fir2 is -0.344 and significantly negative at the 1% level, indicating an inverted U-shaped relations between financial scale and the URIG in the eastern region. The results of columns

(4)and (7) suggest that this relations also holds in the central and western regions. However, if we pay more attention to the absolute values of the coefficients, we find that the influence of financial scale on the URIG is greater in the eastern region than in the other two, and the positive effect of financial scale on the URIG also decreases less in the eastern region.

We further analyze the influence of urban–rural financial structure on the URIG. The coefficient of Urf in column (2) of Table 5 is 0.038 and significantly positive at the 1% level, while the coefficient of Urf2 is -0.206 and nonsignificant at the 10% level, which means that the urban–rural financial structure exerts a positive impact on the URIG in the eastern region. Similar findings are obtained for the central and western regions, as shown in columns (5) and (8), respectively. Furthermore, when we compare the coefficient of Urf in columns (2), (5), and (8), we find that the positive effect of the urban–rural financial structure on the URIG is largest in the central region, while that in the western region is the smallest.

Finally, we turn to the effect of financial resource mismatch on the URIG. We find that the coefficient of Fm, in column (3) in Table 5, is 0.116 and significantly positive at the 1% level, while the coefficient of Fm2 is -0.188 and significantly negative at the 1% level, indicating an inverted U-shaped relations between financial resource mismatch and the URIG in the eastern region. The results of columns (6) and (9) suggest that this relations between financial resource mismatch and the

URIG also holds in the central and western regions. However, if we pay more attention to the absolute values of these coefficients, we find that the influence of financial resource mismatch on the URIG is greatest in the western region, and the positive effect of financial resource mismatch on the URIG also decreases more in the western region than in the other two regions.

Financial Structure Foundation of the Urban–Rural Income Gap in China:	191
An Investigation from the Perspective of the Double Dual Structure

Table 5.

Further Research for Three Sub- Samples

This table reports the results on the effect of financial structure on urban- rural income gap for three sub-samples. The Z-statistics is reported in parenthesis; ‘Year’ and ‘Province’ represent the individual fixed effect and time fixed effect, respectively; The ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Wald value provides the fitting effect of this model. Moreover, to save space, we do not report the coefficient of control variable, which is available upon request.

	Independent			Dependent Variable: Urgap
			(I) East		(II) Center				(III) West
	variable
		(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	Fir	0.214***			0.073***			0.035***
		(12.39)			(16.76)			(9.53)
	Fir2	-0.344***			-0.418***			-0.715***
		(-3.91)			(-5.17)			(-3.24)
	URF		0.038***			0.041***			0.027***
			(6.97)			(6.56)			(5.97)
	URF2		-0.206			-0.616			0.696
			(-0.88)			(-1.43)			(1.48)
	Fm			0.116***			0.166**			0.468***
				(3.44)			(2.53)			(3.18)
	Fm2			-0.188***			-0.374**			-0.549***
				(-4.76)			(-2.10)			(-8.64)
	N	176	77	176	128	56	128	192	84	192
	Wald	147.339	45.838	134.854	85.272	75.350	272.861	783.467	294.631	318.267

V. CONCLUDING REMARKS

In this paper, we construct an economic model based on China’s economic double dual structure and theoretically analyze the relations between financial structure and the urban–rural income gap (URIG). After the theoretical analysis, we conduct empirical tests using the data of 31 provinces in China from 2001 to 2016 and employ such estimations as panel ordinary least squares, IV–2SLS, and FGLS. Specifically, we depict the financial structure along three dimensions: financial scale, the urban–rural financial structure, and the degree of financial mismatch. The results show an inverted U-shaped relations between financial scale and the URIG, as well as between financial resource mismatch and the URIG, and the urban– rural financial structure exerts a positive effect on the URIG. Finally, we split the total sample, by region, into three subsamples to explore the effect of financial structure on the URIG between different regions, employing comprehensive FGLS. The results show that, in the eastern, central, and western regions, the relations between financial structure and the URIG is consistent with that for the full sample. However, the influence of financial scale on the URIG is greatest in the eastern region, and the positive effect of financial scale on the URIG decreases less in the eastern region than in the other two regions. We find the urban–rural financial structure to have a positive effect on the URIG, and this effect is strongest in the central region, while that in the western region is the weakest. Finally, the influence of financial resource mismatch on the URIG is greatest in the western

192	Bulletin of Monetary Economics and Banking, Volume 22, Number 2, 2019

region, and this positive effect decreases more in the western region than in the other two regions due to its lower financial development.

The findings of this paper offer several policy implications. First, expansion of the financial scale eventually narrows the income gap between urban and rural areas. Local governments should promote the improvement of the financial system and the development of the financial market, which can narrow the income gap between urban and rural areas. Second, easing the barriers of access to financial resources between urban and rural sectors will narrow the income gap between urban and rural areas. The government should improve the allocation of financial resources between urban and rural areas. To develop rural finance, barriers to accessing financial resources in the rural sector must be reduced, and financial institutions encouraged to provide more services to rural regions. By formulating preferential measures such as interest rates and tax rates, the flow of financial resources can be guided from the urban sector to the rural sector, increasing the proportion of rural financial resources and narrowing the income gap between urban and rural residents. Third, the government must promote the process of urbanization and break down its barriers. The urban–rural structure is the main cause of the inequality of resources, knowledge, learning ability, and opportunities between residents. Breaking the barriers to urbanization is conducive to the optimization of the allocation of financial resources among all residents, narrowing the urban–rural financial structure and thus reducing urban– rural income inequality.

REFERENCE

Baiardi, D., and Morana, C. (2018). Financial Development and Income Distribution Inequality in the Euro Area. Economic Modelling, 70, 40-55.

Benabou, R. (1996). Inequality and Growth. NBER Macroeconomics Annual, 11, 11- 74.

Benjamin, D., Brandt, L., Giles, J., & Wang, S. (2008). Income Inequality During China’s Economic Transition. China’s Great Economic Transformation, 729-75.

Braithwaite, J., and Braithwaite, V. (1980). The Effect of Income Inequality and Social Democracy on Homicide: A Cross-National Comparison. The British Journal of Criminology, 20, 45-53.

Calderón, C., and Servén, L. (2004). The Effects of Infrastructure Development on Growth and Income Distribution. The World Bank.

Chen, J., and Fleisher, B. M. (1996). Regional Income Inequality and Economic Growth in China. Journal of comparative economics, 22, 141-164.

Cheong, T. S., and Wu, Y. (2014). The Impacts of Structural Transformation and Industrial Upgrading on Regional Inequality in China. China Economic Review, 31, 339-350.

Dabla-Norris, M. E., Kochhar, M. K., Suphaphiphat, M. N., Ricka, M. F., and Tsounta, E. (2015). Causes and Consequences of Income Inequality: A Global Perspective. International Monetary Fund.

De Gregorio, J. (1996). Borrowing Constraints, Human Capital Accumulation, and Growth. Journal of Monetary Economics, 37, 49-71.

Financial Structure Foundation of the Urban–Rural Income Gap in China:	193
An Investigation from the Perspective of the Double Dual Structure

De Haan, J., and Sturm, J. E. (2017). Finance and Income Inequality: A Review and New Evidence. European Journal of Political Economy, 50, 171-195.

Galor, O., and Zeira, J. (1993). Income Distribution and Macroeconomics. The Review of Economic Studies, 60, 35-52.

Ghossoub, E. A., and Reed, R. R. (2017). Financial Development, Income Inequality, and the Redistributive Effects of Monetary Policy. Journal of Development Economics, 126, 167-189.

Gimet, C., and Lagoarde-Segot, T. (2011). A Closer Look at Financial Development and Income Distribution. Journal of Banking & Finance, 35, 1698-1713.

Greenwood, J., and Jovanovic, B. (1990). Financial Development, Growth, and the Distribution of Income. Journal of political Economy, 98, 1076-1107.

Hamori, S., and Hashiguchi, Y. (2012). The Effect of Financial Deepening on Inequality: Some International Evidence. Journal of Asian Economics, 23, 353- 359.

Harris, R. D., and Tzavalis, E. (1999). Inference for Unit Roots in Dynamic Panels Where the Time Dimension is Fixed. Journal of Econometrics, 91, 201-226.

Hsieh, J., Chen, T. C., and Lin, S. C. (2019). Financial Structure, Bank Competition and Income Inequality. The North American Journal of Economics and Finance.

Im, K. S., Pesaran, M. H., and Shin, Y. (2003). Testing for Unit Roots in Heterogeneous

Panels. Journal of Econometrics, 115, 53-74.

Jalil, A. (2012). Modeling Income Inequality and Openness in the Framework of Kuznets Curve: New Evidence From China. Economic Modelling, 29, 309-315.

Jauch, S., and Watzka, S. (2016). Financial Development and Income Inequality: A Panel Data Approach. Empirical Economics, 51, 291-314.

Johansson, A. C., and Wang, X. (2014). Financial Sector Policies and Income Inequality. China Economic Review, 31, 367-378.

Kim, D. H., and Lin, S. C. (2011). Nonlinearity in the Financial Development– Income Inequality Nexus. Journal of Comparative Economics, 39, 310-325.

Kunieda, T., Okada, K., and Shibata, A. (2014). Finance and Inequality: How Does Globalization Change Their Relationship? Macroeconomic Dynamics, 18, 1091- 1128.

Kuznets, S. (1955). Economic Growth and Income Inequality. The American Economic Review, 1-28.

Lee, W. C., Cheong, T. S., and Wu, Y. (2017). The Impacts of Financial Development, Urbanization, and Globalization on Income Inequality: A Regression-based Decomposition Approach (No. 651). Asian Development Bank Institute.

Levin, A., Lin, C. F., and Chu, C. S. J. (2002). Unit Root Tests in Panel Data: Asymptotic and Finite-Sample Properties. Journal of Econometrics, 108, 1-24.

Lessmann, C. (2013). Foreign Direct Investment and Regional Inequality: A Panel Data Analysis. China Economic Review, 24, 129-149.

Li, B., Hao, Y., and Chang, C. P. (2018). Does an Anticorruption Campaign Deteriorate Environmental Quality? Evidence from China. Energy & Environment, 29, 67-94.

Li, H., Squire, L., and Zou, H. F. (1998). Explaining International and Intertemporal Variations in Income Inequality. The Economic Journal, 108, 26-43.

Li, J., and Yu, H. (2014). Income Inequality and Financial Reform in Asia: The Role of Human Capital. Applied Economics, 46, 2920-2935.

194	Bulletin of Monetary Economics and Banking, Volume 22, Number 2, 2019

Li, Y. (2010). Analysis on the Disparity in Economic Growth and Consumption Between Urban Sector and Rural Sector of China: 1978–2008. Frontiers of Economics in China, 5, 559-581.

Lin, F., and Fu, D. (2016). Trade, Institution Quality and Income Inequality. World Development, 77, 129-142.

Lu, M., and Chen, Z. (2004). Urbanization, Urban-Biased Economic Policies and Urban-Rural Inequality. Economic Research Journal, 6, 50-58.

Mah, J. S. (2013). Globalization, Decentralization and Income Inequality: The Case of China. Economic Modelling, 31, 653-658.

Molero-Simarro, R. (2017). Inequality in China Revisited. The Effect of Functional Distribution of Income on Urban Top Incomes, the Urban-Rural Gap and the Gini Index, 1978–2015. China Economic Review, 42, 101-117.

Mookerjee, R., and Kalipioni, P. (2010). Availability of Financial Services and Income Inequality: The Evidence from Many Countries. Emerging Markets Review, 11, 404-408.

Naceur, S. B., and Zhang, R. (2016). Financial Development, Inequality and Poverty; Some International Evidence (No. 16/32). International Monetary Fund.

Seven, U., and Coskun, Y. (2016). Does Financial Development Reduce Income Inequality and Poverty? Evidence from Emerging Countries. Emerging Markets Review, 26, 34-63.

Su, C. W., Liu, T. Y., Chang, H. L., and Jiang, X. Z. (2015). Is Urbanization Narrowing the Urban-Rural Income Gap? A Cross-Regional Study of China. Habitat International, 48, 79-86.

Tan, H. B., and Law, S. H. (2012). Nonlinear Dynamics of the Finance-Inequality Nexus in Developing Countries. The Journal of Economic Inequality, 10, 551-563.

Wang, C., Wan, G., and Yang, D. (2015). Income Inequality in the PRC: Trends, Determinants, and Proposed Remedies. China’s Economy: A Collection of Surveys.

Wei, H., and Zhao, C. (2012). Effects of International Trade on Urban-Rural Income Gap in China. Finance & Trade Economics, 1.

Wen, J., Feng, G. F., Chang, C. P., and Feng, Z. Z. (2018). Stock Liquidity and Enterprise Innovation: New Evidence from China. The European Journal of Finance, 24, 683-713.

Wen, J., Yang, D., Feng, G. F., Dong, M., and Chang, C. P. (2018). Venture Capital and Innovation in China: The Non-Linear Evidence. Structural Change and Economic Dynamics, 46, 148-162.

Wu, D., and Rao, P. (2017). Urbanization and Income Inequality in China: An Empirical Investigation at Provincial Level. Social Indicators Research, 131, 189- 214.

Wu, X., and Perloff, J. M. (2005). China’s Income Distribution, 1985–2001. Review of Economics and Statistics, 87, 763-775.

Yu, K., Xin, X., Guo, P., and Liu, X. (2011). Foreign Direct Investment and China’s Regional Income Inequality. Economic Modelling, 28, 1348-1353.

Zhang, Q., and Chen, R. (2015). Financial Development and Income Inequality in China: An Application of SVAR Approach. Procedia Computer Science, 55, 774- 781.

Zhang, J., Wan, G., and Jin, Y. (2007). The Financial Deepening-Productivity Nexus in China: 1987–2001. Journal of Chinese Economic and Business Studies, 5, 37-49.