GLOBALIZATION AND URBAN-RURAL INEQUALITY: EVIDENCE FROM CHINA

GLOBALIZATION AND URBAN-RURAL INEQUALITY: EVIDENCE FROM CHINA A Thesis submitted to the Faculty of the Graduate School of Arts and Sciences of Georgetown University in partial fulfillment of the requirements for the degree of Master of Public Policy in Public Policy By Yichi Zhang, B.A. Washington, DC April 12, 2012

Copyright 2012 by Yichi Zhang All Rights Reserved ii

GLOBALIZATION AND URBAN-RURAL INEQUALITY: EVIDENCE FROM CHINA Yichi Zhang, B.A. Thesis Advisor: Andreas Kern, Ph.D. ABSTRACT A long standing debate among academics and policy-makers concerns whether globalization leads to a higher degree of inequality. In this regard, China has attracted particular attention. In contrast to the existing literature that has concentrated on either overall inequality or coastal-inland regional inequality in China, this paper analyzes the impact of foreign trade and foreign direct investment on urban-rural income inequality. The key hypothesis is that if openness to foreign trade and foreign direct investment increases, then urban-rural income inequality will also increase. The paper examines the relationships empirically using panel data in both random-effects and fixed-effects models, for 31 provinces in China over the period 1978-2008. The study measures globalization as both the trade-to-grp ratio and the FDI-to-GRP ratio, and urban-rural inequality as the urban-rural income ratio. Findings indicate that the urban-rural income gap has been widening since China began to integrate into the world economy, but globalization is not the principal reason for the increase. The study finds no significant relationship between foreign trade and urban-rural income inequality in China. In fact, findings indicate that foreign direct investment has actually helped to reduce inequality. iii

Many thanks to my professors at the Georgetown Public Policy Institute, particularly my thesis advisor, Andreas Kern. I am also grateful to Mike Barker, Jeff Mayer, and many friends for their assistance. And a special thank you to my parents, for their unconditional love and support. Yichi Zhang iv

TABLE OF CONTENTS I. INTRODUCTION... 1 II. BACKGROUND... 3 III. LITERATURE REVIEW... 5 IV. CONCEPTUAL MODEL AND ANALYSIS PLAN... 8 V. DATA DESCRIPTION... 12 VI. ESTIMATION RESULTS... 14 VII. ROBUSTNESS CHECKS... 19 VIII. CONCLUSION... 21 IX. APPENDIX A DIAGRAM... 23 X. APPENDIX B SUPPLEMENTARY TABLES... 24 XI. REFERENCES... 32 v

INTRODUCTION Jimmy Carter has often said If you re totally illiterate and living on one dollar a day, the benefits of globalization never come to you. As globalization has been integrating the economies around the world through the channels of international trade and capital flows across borders, the question of whether globalization increases inequality and makes the poor worse off is under heated debate. The pro-globalization position argues that the wave of globalization since 1980 has been a positive force for poverty reduction and income equality, especially in developing economies (Dollar, 2005). On the other hand, opponents of this view assert that Globalization has dramatically increased inequality between and within nations (Mazur, 2000). Given the importance of this topic and such a divergence of viewpoints, academics have conducted a number of studies to analyze the impact of globalization on income inequality, including especially cross-country studies (e.g. Dollar, 2005; Alderson and Nielsen, 2002), as well as single-country studies (e.g. Hanson and Harrison, 1999). However, the number of studies on China is limited although China s economic development history offers a perfect case to study this subject. Since 1978 when China implemented its well-known Reform and Opening-up Policy, China has clearly been striving to integrate itself with the other economies throughout the world. In virtue of its continuing efforts to embrace globalization, China has become the nation with the largest trading volume and FDI inflows in the developing world (Zhang and Zhang, 2003). However, China s successful globalization has been accompanied by growing domestic inequality. According to the World Bank s World Development Indicators, estimated overall income inequality in China, as measured by the Gini index, increased from 29.1 in 1981 to 42.5 1

in 2005 (The World Bank, 2005). This raises the question whether globalization merely coincided with the widening income gap in China, or contributed to it? Existing literature on inequality in China has focused on costal-inland regional inequality (Zhang and Zhang, 2003). Very few studies have focused on urban-rural inequality and the effects of globalization on urban-rural inequality. The World Bank has estimated that 75 percent of the change in the overall income inequality in China can be attributed to the change in the urban-rural income inequality between 1984 and 1995 (The World Bank, 1997). Thus it is important and valuable to explore the relationship between globalization and urban-rural income inequality in China. This study examines the impact of foreign trade and foreign direct investment on urbanrural income inequality in China, using both random-effects and fixed-effects models with panel data for 31 provinces between 1978 and 2008. The study measures globalization as both the trade-to-grp ratio and the FDI-to-GRP ratio, and urban-rural inequality as the urban-rural income ratio. Based on the estimation results, the paper proposes policies to help the lagging rural areas catch up with more prosperous urban areas. Data used in the study come from Comprehensive Statistical Data and Materials on 60 Years of New China, published by National Bureau of Statistics of China. 2

BACKGROUND Since 1978 when China adopted the Reform and Opening-up Policy, China has been making efforts to integrate into the globalized world. China s accession to the WTO in 2001 manifested its importance in the global economic arena as a major participant in the process of globalization. The World Development Indicators estimate that China s degree of dependence on foreign trade rose from 13.67% in 1978 to 70.57% in 2006. Meanwhile FDI in China has also increased from 386 million dollars in 1982 to 121,418 million dollars in 2007 (The World Bank, 2007). Over the past three decades, China s successful globalization has fueled unprecedentedly rapid economic growth. It is estimated that, between 1978 and 2002, the average growth rate of GDP per capita in China was 8.1% (Hu, Hu, and Chang, 2006). Such phenomenal growth has resulted in a dramatic rise in households income and a sharp decline in the poverty rate from 64% at the beginning of the reform to 10% in 2004 (Dollar, 2007). However, at the same time, various disparities have emerged within China. Urban-rural income inequality has risen, as the fruits of globalization have not been evenly shared between the urban and rural populations. The wage gap between skilled urban professionals and unskilled rural workers has been widening, as well as the gap in the provision of public goods (Dollar, 2007). According to the World Development Indicators, estimated overall income inequality in China, as measured by the Gini index, increased from 29.1 in 1981 to 42.5 in 2005 (The World Bank, 2005). Applying different datasets and approaches, Khan and Riskin estimated that the Gini coefficient in China rose from approximately 30.0 in 1988 to 45.2 in 1995 (Khan and Riskin, 2001). In regard to urban-rural inequality in particular, Ravallion and Chen argue that in 3

spite of sporadic declines, there is an increasing trend of absolute inequality measured by the difference between urban and rural mean income, as well as relative inequality measured by the ratio of urban to rural mean income (Ravallion and Chen, 2007). Moreover, the current set of public policies might have exacerbated rather than mitigated the problem. Li argues that massive international trade and FDI inflows have been mostly directed to urban areas in order to boost the manufacturing industries underpinning China s export-led growth strategy, while the rural population benefited very little from globalization (Li, 2010). Additionally, Dollar believes that the household registration (hukou) system has posed great challenges to rural-to-urban migration, restricting the opportunities of rural labors to increase their wages and income. The decentralized fiscal system in China might also have limited local governments abilities to redistribute, leading to under-provision of public goods such as education, health and social welfare in rural areas (Dollar, 2007). To help readers understand the political actors clearly, it may be helpful to provide an overview of the Chinese administrative divisions, as shown in Figure 1. The People's Republic of China administers 33 provincial level divisions, including 22 provinces, five autonomous regions, four municipalities, and two special administrative regions. In each provincial level division, the population is further divided into urban population and rural population. Addressing urban-rural inequality requires the involvement of multi-level political actors. It requires the central government to adopt sagacious policies and the local government to ensure that these policies are effectively implemented. 4

LITERATURE REVIEW In light of the importance of the subject and the divergent positions among academics, many studies are conducted to assess the role of globalization in income inequality. Some of these studies take a global comparative perspective. Dollar, for example, as a loyal defender of globalization, spares no effort in arguing for the profound influence globalization has on economic growth and poverty reduction. He argues that inequality within countries is not growing generally. Moreover, global inequality has actually witnessed a modest degree of decline in the age of globalization. However, inequality in very populous countries such as China, India and the United States has increased since 1980 (Dollar, 2005). On the other hand, Mazur argues that the benefits of globalization have been reaped disproportionally by a few industrial nations, resulting in concentrated wealth rather than uniform growth. In this sense, he says Globalization has dramatically increased inequality between and within nations (Mazur, 2000). Much of the literature on the relationship between globalization and inequality has focused on developed countries. For example, Alderson and Nielsen performed an empirical analysis of 16 OECD nations over the period 1967-1992, and found that North-South trade, direct investment, and immigration contributed substantially to income inequality. They found positive effects of all three globalization factors on inequality and concluded that globalization leads to rising income inequality in advanced industrial countries (Alderson and Nielsen, 2002). Due to the different positions of developed countries and developing countries in the global economy, how globalization affects the income inequality within a country might be different too. Existing literature on China has focused on mainly costal-inland regional inequality. 5

For example, Zhang and Zhang employed an approach to quantify the contribution of globalization, using a panel dataset consisting of 28 provinces over the period 1986-1998. They found that foreign trade and inward FDI played an important role in worsening China s regional inequality (Zhang and Zhang, 2003). Applying different data and methodologies, other studies observe similar relationships are also observed in other studies (e.g. Kanbur and Zhang, 2005; Wan, Lu and Chen, 2007). Yet, despite the change in urban-rural income inequality accounts for 75 percent of the change in overall income inequality in China between 1984 and 1995 (The World Bank, 1997), very few studies have investigated the effect of globalization on urban-rural inequality in China. Analyzing the data from the China Statistical Yearbook, Li documented a positive relationship between globalization and inequality. In particular, he found that the wage gap between urban and rural areas increased from 2.57 in 1978 to 3.11 in 2002, because FDI has been highly concentrated in urban areas. He noted that while the total amount of FDI in China has been growing, foreign investment directed to rural agricultural sectors has actually declined (Li, 2010). On the other hand, Wei and Wu, using data for roughly 100 Chinese cities over the period 1988-1993, produced different results. Applying both first-difference model and fixed-effects model, using the minimum distance to China s major seaports as the instrumental variable for openness to control for potential endogeneity, they documented a negative relationship between trade openness and urban-rural inequality, and predicted that cities with a higher degree of openness to trade would have a greater decline in urban-rural inequality (Wei and Wu, 2001). Smeeding argues that, in addition to the influence of globalization, domestic policies and institutions also play an important role in determining the level and trend of inequality within 6

high-income and middle-income countries (Smeeding, 2002). Hence, in following analysis, I also investigate the impact of various domestic factors on urban-rural inequality, including urbanization rate, investment differentials, and geographical features. In contrast to related prior work, this paper uses a new dataset from Comprehensive Statistical Data and Materials on 60 Years of New China, published by National Bureau of Statistics of China. The paper also constructs new measures of globalization and inequality, such as trade-to-grp ratio, FDI-to-GRP ratio, and urban-rural income ratio, for 31 provincial level divisions that China currently administers, over the period 1978-2008. It analyzes the mechanisms through which foreign trade, foreign direct investment and domestic factors impact urban-rural inequality in China. Employing a new dataset, and looking empirically at the urbanrural dimension of China s inequality, the study sheds new light upon the widely-held belief that globalization leads to higher inequality. 7

CONCEPTUAL MODEL AND ANALYSIS PLAN Krugman (1991), in his New Economic Geography Theory, argued that a country would be divided into an industrialized core and an agricultural periphery, because manufacturing firms tend to locate in concentrated clusters in order to minimize transportation costs and realize economies of scale. The burgeoning manufacturing industry in the cities would promote economic activities and raise wages of urban labor, while disadvantaged rural labor would benefit little from international trade. Accordingly, we would expect a widening urban-rural income gap due to increasing foreign trade. However, the impact of foreign trade on exacerbating urban-rural income inequality could be mitigated through two channels. First, expanding manufacturing industry in urban areas could lead rural unskilled labor to migrate to the city so as to address the shortage of manufacturing labor, which would lead to the rise in wages of remaining rural agricultural labor because of excess demand for agricultural products. This dynamic would narrow the urban-rural income gap by driving up the wages of rural labor force. Second, migrant workers would possibly send remittances back to their households and relatives living in the rural areas, which would also reduce urban-rural income inequality. Similarly, the effect of FDI inflows on the urban-rural income gap could also be viewed from two directions. FDI is always directed to urban areas where companies are located and business is developed. Expansion of existing business in the host country due to FDI inflows could increase urban-rural inequality by raising the wages of urban skilled workers. However, the spillover effects of inward FDI between sectors (Schoors and van der Tol, 2002) could also 8

promote agricultural innovations and boost agricultural productivity, which in turn raises the income of rural labors and narrow the urban-rural income gap. The baseline hypothesis I test is that if openness to foreign trade and foreign direct investment increases, then urban-rural income inequality will increase too. In order to examine the relationship between globalization and urban-rural inequality in China, I construct three main variables of interest, traderatio, fdiratio, and incomeratio for the 31 provincial level divisions that China currently administers for which data are available, over the period 1978-2008. Specifically, traderatio and fdiratio denote the ratio of foreign trade and foreign direct investment to gross regional product, which measures the degree of openness. incomeratio measures the ratio of nominal per capita annual income in urban and rural households. In addition, incomeratioreal measures the ratio of real per capita annual incomes, which are the nominal incomes deflated by CPI indices respectively, in urban and rural areas. Through adjusting for price differentiations between urban and rural areas, the variable reflects the real living standards and well-being of the urban and rural populations. Additionally, I generate a series of indicators that control for other socio-economic characteristics, including the ratio of urban-to-rural population that measures the urbanization rate, the ratio of urban-to-rural investment in fixed assets that measures investment inequality, per capita GRP that measures the macroeconomic situation, initial inequality, and a coastal dummy that indicates whether a province is a coastal province. To decide between fixed-effects model and random-effects model, I perform a Hausman test rejecting the null hypothesis that the coefficients in these two specifications are the same. This suggests that the random effects are correlated with the other regressors, which would 9

produce biased estimates (Wooldridge, 2002). However, the advantage of a random-effects model is that it allows us to analyze the effects of province-specific fixed characteristics, such as initial inequality and geographical conditions. Hence, I decided to implement both randomeffects and fixed-effects panel regressions, with year fixed effects controlled. I let logincomeratio k,t denote the logarithm form of incomeratio for province k in year t. Then I hypothesize that inequality is determined by a number of factors, and the random-effects and fixed-effects models are illustrated as follows: logincomeratio k,t = β 0 + β 1 logtraderatio k,t + β 2 logfdiratio k,t + β 3 h k,t + β 4 z k + β 5 d1979 t ~ d2008 t + u k,t (1) This random-effects framework decomposes all the factors that contribute to urban-rural inequality into four categories: logtraderatio k,t and logfdiratio k,t are the degree of openness to foreign trade and foreign direct investment in logarithm form for province k in year t; h k,t denotes the set of socio-economic variables for a province that do change over time; z k is a summary of the fixed characteristics of a province that do not change over time; d1979 t ~ d2008 t control for year fixed effects such as macroeconomic fluctuations; and u k,t is a composite random variable, assumed to be independently and identically distributed. logincomeratio k,t = β 0 + β 1 logtraderatio k,t + β 2 logfdiratio k,t + β 3 H k,t +β 4 dprov1 k ~ dprov30 k + β 5 d1979 t ~ d2008 t + v k,t (2) The fixed-effects model allows us to control for z k in Equation(1), which denotes the fixed characteristics that are constant over time but correlated with independent variables, thus eliminating certain omitted variable bias. H k,t in Equation(2) is a set of variables other than foreign trade and FDI that affect urban-rural inequality. I include dprov1 k ~dprov30 k in the model 10

as 30 province-dummy variables that absorb the effects particular to each province, omitting one province as baseline category. v k,t is a composite random variable, assumed to be independently and identically distributed. Most importantly, the effects of globalization on urban-rural inequality are captured by the parameters β 1 and β 2. 11

DATA DESCRIPTION Data used in this study comes from Comprehensive Statistical Data and Materials on 60 Years of New China, published by National Bureau of Statistics of China. Comprehensive Statistical Data and Materials on 60 Years of New China is a comprehensive book containing extensive national-level and provincial-level data on social and economic development since the founding of People s Republic of China. The book has about 300 indicators, and data are presented in 16 sections: Population and Employment, National Account, Investment in Fixed Assets, Public Finance, Price Index, Agriculture, Industry, Construction, Transportation, Postal and Telecommunication, Domestic Market, Foreign Trade, Tourism, Banking, Education, Culture and People's Livelihood. My sample includes 31 provincial level divisions that China currently administers, for the period 1978-2008. The central variables under investigation are Total Imports and Exports, Foreign Direct Investment, Gross Regional Product, Per Capita Annual Disposable Income of Urban Households and Rural Households, for each province and each year respectively. To help readers interpret the variables correctly, I provide clear definitions in Table 1. The subsequent statistical analysis focuses on the trend of the urban-rural inequality measure from 1978 to 2008. The starting point is chosen because in 1978 China started to integrate into the globalized world economy by implementing the Reform and Opening-up Policy. The ending point is determined by data availability. Before conducting statistical analysis, I undertook the following procedures: eliminating provincial level divisions whose economies are not integrated in the country, and combining data of provinces whose jurisdiction are divided during the sample period. 12

Hong Kong and Macau, which were former British and Portuguese territories respectively, are two special administrative regions of China. Since their reversions to China in 1997 and 1999, these two special administrative regions have been responsible for all issues except diplomatic relations and national defense. In addition, the Xinjiang Production and Construction Corps, XPCC for short, is a semi-military governmental organization in the Xinjiang Uyghur Autonomous Region of China. With strategic functions and unique political status, XPCC has been self-governing and exercising a high degree of autonomy (Vicziany and Zhang, 2004). In order to investigate urban-rural inequality in comparable provincial level divisions, I eliminate Hong Kong, Macau, and the XPCC from the data set, because comprehensive data is not available for these three divisions due to the uniqueness of their political governances and economic systems. In addition, Hainan province is included in Guangdong province because data for Hainan are not available before 1988 when it became a separate province. 13

ESTIMATION RESULTS Since 1978 when the Reform and Opening-up Policy was introduced, China has been a major participant in the process of globalization. At the same time, China has witnessed unprecedentedly rapid economic growth, accompanied by widening urban-rural inequality. Table 2 indicates the pattern of globalization, economic growth, and urban-rural inequality in China, taking the mean values of 31 provinces over the period 1978-2008. The degree of China s globalization is captured by the rapid increase in foreign trade and FDI flows. During the period of thirty years, the ratio of trade to GRP, which is usually defined as trade openness, has increased ninefold, from 3.62% to 33.63%. The importance of FDI for the Chinese economy, which is reflected in the share of FDI flows in GRP, increased from almost zero in 1978 to 6.01% in 1994, and then dropped to 2.84% in 2008. China s boom in trade and FDI has fueled the rapid economic growth, which is reflected in the huge increase in average GRP from 11.2 billion yuan to 1055.9 billion yuan. However, the economic gains of rapid growth have not been distributed evenly between urban and rural areas. As shown in Table 2, urban-rural inequality, which is measured by the ratio of per capita annual disposable income of urban households to rural households, has risen significantly from 1.64 in 1983 to 3.00 in 2008. Although various factors could have caused the increase in China s urban-rural inequality, the strong correlations between the indices in the table suggest that foreign trade and FDI played an important role. Nevertheless, it would be premature to conclude from these statistics that greater openness has contributed to greater inequality. To further investigate the link between globalization and urban-rural inequality by province, in Table 3, I present the pattern of 14

globalization, economic growth, and urban-rural inequality in China, taking the mean values of 31 years between 1978 and 2008 for each province in Table 3. First, for any given variable, there is a tremendous amount of heterogeneity across space. For example, trade openness ranges from 4.19% in Henan to 71.57% in Shanghai, while the FDIto-GRP ratio ranges from 0.25% in Xinjiang to 6.40% in Tianjin. Another striking feature of the table is that coastal provinces show a larger openness to foreign trade and FDI than inland provinces. Meanwhile, urban-rural inequality is actually smaller in more globalized coastal provinces than in inland provinces, which might be regarded as weak evidence that a higher degree of openness may have reduced, rather than increased, urban-rural inequality. Table 2 and Table 3 show different relationships between globalization and urban-rural inequality. Table 2 suggests that urban-rural inequality in China has increased substantially in the past three decades, accompanied by an increase in the degree of openness to the world economy. Therefore, it is reasonable to argue that increasing degree of openness leads to increasing inequality. Conversely, Table 3 tells a different story that urban-rural inequality is actually smaller in coastal provinces that are more open and globalized than in inland provinces, which implies that globalization leads to decreasing inequality. However, we cannot simply infer causation from these two tables. Thus results of more systematic empirical analysis are provided in Table 4. The regression results reported in Table 4 show that the coefficient on trade openness is negative but not statistically significant, and the coefficient on FDI openness is negative and statistically significant at the 1 percent level. As shown in Column 6 of the table, a 1 percent 15

increase in the degree of openness to FDI is associated with a 0.016 percent decrease in the urban-rural income ratio. The empirical results suggest that increasing openness to international trade does not have a significant impact on urban-rural inequality in China. This might be due to the fact that the two opposing forces that drive the inequality up and down have cancelled each other out. On the one hand, international trade usually takes place in urban areas because manufacturing industry aims to achieve economies of scale, thus excess demand for urban workers would drive up their wages and widen the urban-rural income gap (Krugman, 1991). On the other hand, expansion of manufacturing business in the city would lead to excess demand for agricultural products, which would raise wages of rural labor and narrow the income gap. Besides, the remittances that migrant workers send back to rural areas would also reduce the inequality. On the whole, the interaction of such mechanisms would deliver an ambiguous effect of international trade on urban-rural inequality in China. The empirical results imply that FDI inflow has significantly reduced China s urban-rural income inequality, which is in line with the theoretical arguments that the spillover effects of inward FDI between sectors (Schoors and van der Tol, 2002) promote agricultural innovations and boost agricultural productivity, which in turn raises the income of rural labors and narrows the urban-rural income gap, even if the volume of FDI directed to rural areas is comparatively small. The empirical results of my regression do not support the argument that the urban areas in China are economically connected to advanced economies, thus may enjoy better advantages from globalization than rural areas (Zhang and Zhang, 2003). Instead, provinces that experience 16

larger openness tend to witness less severe urban-rural inequality, which suggest that the economic disparity between urban areas and rural areas has not been worsened because of globalization. Some might be concerned that the urbanization rate has an impact on urban-rural inequality. In Column 2, I add the log ratio of urban-to-rural population to measure the urbanization rate. The coefficient on this variable is significantly negative, which implies that more urbanized provinces would have smaller urban-rural inequalities. To test the hypothesis that if the investment in urban areas exceeds that in rural areas, the income gap between the urban and rural households may widen too, I add the log ratio of urbanto-rural investment in fixed assets as a control variable. As shown in Column 3 of Table 4, the coefficient is statistically significant and positive, suggesting that the more investment there is in urban areas relative to rural areas, the larger the urban-rural inequality is. As we are concerned that economic development may help reduce inequality because richer provinces are more capable of redistributing rents from urban areas to rural areas, I add the logarithm form of per capita GRP as another regressor. The coefficient on this variable is statistically significant at the 5 percent level and negative, which is consistent with our hypothesis that provinces with higher per capita GRP tend to have lower inequality. In consideration of the persistence of urban-rural inequality, I include the logarithm form of initial urban-rural inequality in 1978 as another control variable, and use data in the following years if data are not available in year 1978. The coefficient is statistically significant at 1 percent level and positive, suggesting that inequality between urban and rural areas seems to be persistent over time. This means the government needs to make structural changes and spare 17

more efforts in redistributing, in order to rescue those provinces with large initial inequality from this inequality trap. As shown in Table 3, more globalized coastal provinces seem to have smaller urban-rural inequality than inland provinces. After including a dummy variable for 11 coastal provinciallevel divisions, we can observe in Column 6 of Table 4 that the effect of being a coastal province on inequality is negative and significant at the 10 percent level, which suggests a moderate effect, but also that being a coastal province reduces urban-rural inequality as expected. It is important to note that after including these control variables, the coefficient on trade openness remains statistically insignificant and the coefficient on FDI openness remains negative and statistically significant at the 1 percent level. The R-squared in Column 5 indicates that the variation in the degree of globalization and other controls across provinces over time, explains 81.0 percent of the variation in urban-rural income inequality, implying a good fit. 18

ROBUSTNESS CHECKS The estimation results are quite robust to different model specifications. Table 5 presents the regression results using the exports-to-grp ratio and the imports-to-grp ratio as alternative measures for trade openness. According to the results, both openness to exports and imports have no significant effect on urban-rural inequality, while the effect of FDI openness remains negative and highly significant at the 1 percent level. In order to control for price dynamics impacting purchasing power in urban and rural areas, I construct a series of CPI indices for urban and rural areas respectively to adjust the nominal incomes for price differentiations between urban and rural areas. The results are reported in Table 6. In this replication, the sign and statistical significance of all our two variables of main interest remain unchanged, and the effect of FDI openness in helping reduce inequality becomes larger. This suggests that after controlling for different price levels between urban and rural areas, the effect of FDI inflows is amplified. In addition to my random-effects models, I also implement a province fixed-effects regression with year fixed-effects. The estimation results presented in Table 7 still show an insignificant effect of trade openness and a negative and significant effect of FDI openness. Coefficients of other control variables do not change much, except the log form of per capita GRP, which becomes insignificant. Also, in order to eliminate potential cyclical components in the variables, I break the 31- year time span into several 5-year windows, taking the mean of the values across five years. The regression results are presented in Table 8. Although this model specification loses some statistical significance on urban-to-rural population ratio and coastal dummy due to a sharp 19

decrease in number of observations, the sign and statistical significance of trade openness and FDI openness remain unchanged, suggesting the robustness of my model. The central empirical findings remain rather robust even after controlling for different potentially intervening effects. From this I conclude that rising urban-rural inequality in China is mainly driven by forces other than global economic integration. It seems that domestic factors, such as the persistence of urban-rural inequality and unequal investment, are playing more important roles. 20

CONCLUSION This study examines the empirical pattern of the relationship between globalization and urban-rural income inequality in China, using panel data in both random-effects and fixed-effects models, for 31 Chinese provinces over the period 1978-2008. The study measures globalization as both the trade-to-grp ratio and the FDI-to-GRP ratio, and urban-rural inequality as the urbanrural income ratio. Findings indicate that urban-rural income inequality has been increasing since China began to integrate into the world economy, but globalization is not the principal reason for the increase. This study finds no significant relationship between foreign trade and urban-rural income inequality in China. In fact, findings indicate that foreign direct investment has actually helped to reduce the inequality. However, in order to derive more robust results, limitations in this paper might be addressed by further research. First, a substantial amount of data is missing, due to the complicated political and economic history of China. Imputation of these missing values by combining other sources of data might deliver more robust and unbiased results. Second, data availability does not allow us to explore the underlying channels through which globalization could affect urban-rural inequality. The comprehensive treatment of the linkages examined in this study requires data that are not yet available. However, studies like this one can still add value to the existing literature. Nevertheless, the results of this study could have fruitful public policy implications for academics and policy-makers in the field of globalization and urban-rural development in China and elsewhere. 21

In recent years, China has implemented preferential policies that favor rural areas, and these economically lagging areas have made remarkable progress in catching up with the urban areas. The government should take advantage of the positive distributional effect of globalization in reducing inequality through diversifying the export base of the country and further orienting domestic and foreign capital towards rural areas. Although urbanization also appears to have positive effects, the study does not suggest loosening the registration system to mobilize migration, which would add pressure to existing industrial cores. Instead, the government should promote rural investment in strengthening the infrastructure, funding education, and improving health in rural areas, in order to reduce urban-rural inequality and to enable also the rural population to reap the benefits of globalization. 22

APPENDIX A FIGURE 1 Administrative divisions of China Source: http://en.wikipedia.org/wiki/file:china_administrative.png 23

APPENDIX B TABLE 1 Definitions of Main Variables province Province year Year incomeurban Per Capita Annual Disposable Income of Urban Households-Value (yuan) incomerural Per Capita Annual Net Income of Rural Households-Value (yuan) incomerationominal incomeurban / incomerural cpiurban2008 Consumer Price Index-Urban (2008=100) cpirural2008 Consumer Price Index-Rural (2008=100) incomeratioreal incomeurban / cpiurban2008 / incomerural / cpirural2008 grp Gross Regional Product (100 million yuan) trade Total Imports and Exports (100 million yuan) exports Exports (100 million yuan) imports Imports (100 million yuan) fdi Foreign Direct Investment (100 million yuan) traderatio Total Imports and Exports / Gross Regional Product exportsratio Exports / Gross Regional Product importsratio Imports / Gross Regional Product fdiratio Foreign Direct Investment / Gross Regional Product popurban Population-Urban (year-end)(10000 persons) poprural Population-Rural (year-end)(10000 persons) popratio Population-Urban / Population-Rural investfixedassetsurban Investment in Fixed Assets-Urban (100 million yuan) investfixedassetsrural Investment in Fixed Assets-Rural (100 million yuan) Investratio Investment in Fixed Assets-Urban / Investment in Fixed Assets-Rural grppc Per capita Gross Regional Product (100 million yuan) incomeratioinitial Initial Inequality coastal =1 if Coastal Province Source: Author s illustration based on Comprehensive Statistical Data and Materials on 60 Years of New China. 24

TABLE 2 GLOBALIZATION, AND URBAN-RURAL INEQUALITY Year Gross Regional Product (100 million yuan) Trade/GRP (%) Exports/GRP (%) Imports/GRP (%) FDI/GRP (%) Urban-Rural Inequality 1978 111.9975 3.615060 3.001110 0.575940 0 2.528664 1979 126.9841 4.312120 3.686520 0.553050 0.03072 2.240736 1980 141.6835 4.786540 4.058160 0.656730 0.08489 2.069197 1981 154.8028 5.574600 4.731950 0.683570 0.10505 1.909897 1982 172.1517 5.580990 4.840610 0.592410 0.11687 1.774128 1983 193.3879 17.14069 9.960960 6.855010 0.15565 1.641368 1984 230.1747 19.90188 11.37483 8.174420 0.25461 1.659826 1985 277.8569 20.66971 7.445230 13.22448 0.31555 1.810166 1986 311.7648 21.60430 7.932230 13.67207 0.42710 2.016446 1987 370.1011 19.83834 8.497190 11.34115 0.49350 2.054766 1988 467.8680 19.00869 8.028660 10.98004 0.86228 2.032465 1989 531.9721 18.32434 7.607320 10.71703 0.82469 2.164323 1990 595.8183 19.55031 10.06019 9.49012 0.88823 2.122699 1991 684.9865 21.10829 11.05086 10.05743 1.23337 2.302405 1992 838.3589 21.63809 11.18753 10.45058 2.36417 2.485250 1993 1103.854 21.04046 9.688920 11.35154 4.03466 2.695464 1994 1462.748 28.46643 13.87391 14.59252 6.00811 2.781892 1995 1855.975 25.19955 12.68726 12.51229 5.41992 2.704408 1996 2185.940 21.65675 10.88278 10.77397 4.85992 2.538954 1997 2462.560 20.27314 10.96066 9.312480 4.56492 2.458176 1998 2663.174 18.74800 10.15009 8.597910 3.97672 2.436291 1999 2845.958 18.72603 9.679230 9.046790 3.33013 2.572236 2000 3177.557 22.08051 11.28258 10.79795 3.06566 2.688284 2001 3502.165 21.56336 10.73594 10.82742 3.06384 2.795124 2002 3889.965 22.57104 11.38290 11.18814 3.28672 2.895826 2003 4492.929 26.90521 13.22139 13.68382 3.24437 2.977704 2004 5406.706 31.27191 15.17538 16.09653 3.03152 2.940264 2005 6384.076 32.24985 16.45966 15.79018 2.93583 2.958434 2006 7439.065 33.91811 17.54525 16.37287 3.01177 3.014320 2007 8891.169 34.46961 18.04267 16.42718 3.06908 3.029970 2008 10559.26 33.63009 17.20203 16.42789 2.83973 2.999896 Source: Author s illustration based on Comprehensive Statistical Data and Materials on 60 Years of New China. 25

TABLE 3 GLOBALIZATION, AND URBAN-RURAL INEQUALITY Province Coastal Provinces Gross Regional Product (100 million yuan) Trade/GRP (%) Exports/GRP (%) Imports/GRP (%) FDI/GRP (%) Urban- Rural Inequality Beijing 2385.094 200.328 48.6879 151.640 4.77177 1.75330 Tianjin 1333.677 54.6530 32.3255 22.3276 6.40218 1.78680 Shanghai 3361.799 71.5697 39.7443 31.8253 5.11605 1.67297 Liaoning 3266.597 28.8733 21.2147 7.65864 2.83599 1.98361 Hebei 3684.323 9.49661 7.34563 2.15099 1.27445 2.25456 Shandong 6575.538 20.2307 12.8702 7.36041 2.18195 2.34044 Jiangsu 6594.498 33.6854 19.7326 13.9528 3.84595 1.88643 Zhejiang 4754.743 26.1673 18.7638 7.40353 2.02469 2.03199 Fujian 2524.733 38.8208 22.9536 15.8671 6.40203 2.34762 Guangdong 8284.409 1.50823 0.61356 0.89467 6.14734 2.41844 Guangxi 1608.349 9.93679 7.02072 2.91607. 0.36162 Inland Provinces Anhui 2116.913 7.62666 4.94552 2.96841 1.05888 2.61909 Chongqing 1181.499 9.84850 5.41392 4.43458 1.49713 3.07006 Gansu 759.5683 5.36957 3.24633 2.35072 0.39486 3.30721 Guizhou 774.7384 4.33399 2.82664 1.50735 0.76217 3.28606 Heilongjiang 2196.553 8.71913 5.40210 3.31703 1.80057 2.05423 Henan 3870.906 4.19163 3.00810 1.18353 0.75690 2.54035 Hubei 2601.432 7.73842 4.89027 2.84816 1.88064 2.40104 Hunan 2562.592 5.95827 4.25568 1.70259 1.40248 2.63182 Inner Mongolia 1385.149 7.72962 4.19269 3.53693 2.02749 2.27122 Jiangxi 1509.238 6.74249 4.68411 2.05838 1.96714 1.98564 Jilin 1433.249 11.6607 6.22154 5.43920 2.45775 1.96743 Ningxia 232.8495 8.90678 6.87499 2.03179 1.15222 2.73854 Qinghai 211.2708 4.91949 1.16897 3.75054 2.35298 3.13699 Shaanxi 1422.515 9.21087 4.96384 2.97397 1.32494 3.11242 Shanxi 1506.757 6.31810 4.51335 1.99811 0.65109 2.49173 Sichuan 2921.505 5.21706 3.26529 1.95177 0.79774 2.76838 Tibet 91.52258 8.83183 3.99073 4.84110 0.34203 3.53479 Xinjiang 979.1252 11.7162 7.55132 4.16488 0.24899 2.77166 Yunnan 1401.862 7.01756 4.45548 2.56209 0.51681 3.48538 Source: Author s illustration based on Comprehensive Statistical Data and Materials on 60 Years of New China. 26

TABLE 4 GLOBALIZATION, AND URBAN-RURAL INEQUALITY (Random Effects Model with Year Fixed Effects) Dependent variable: log ratio of urban/rural households disposable nominal income from 1978 to 2008 (1) (2) (3) (4) (5) (6) Trade/GRP in log -0.001 (0.006) -0.002 (0.006) -0.002 (0.007) 0.002 (0.007) -0.007 (0.007) -0.008 (0.007) FDI/GRP in log -0.019*** (0.004) -0.018*** (0.005) -0.025*** (0.005) -0.028*** (0.005) -0.018*** (0.006) -0.016*** (0.006) Log ratio of urban/rural population -0.031*** (0.010) -0.031*** (0.011) -0.025** (0.012) -0.022* (0.012) -0.026** (0.012) Log ratio of urban/rural Investment in Fixed Assets 0.049*** (0.013) 0.058*** (0.013) 0.039*** (0.013) 0.032** (0.014) Per capita GRP in log -0.137*** (0.037) -0.164*** (0.038) -0.118** (0.047) Initial Inequality in log 0.276*** (0.064) 0.281*** (0.060) Dummy for Coastal Province -0.096* (0.052) 0.470*** 0.450*** 0.301** -1.421*** -1.976*** -1.344** Constant (0.115) (0.118) (0.130) (0.465) (0.472) (0.607) No. of Obs. 706 601 521 521 429 429 No. of Province 29 29 28 28 24 24 R-squared 0.793 0.788 0.781 0.774 0.807 0.810 Note: Robust standard errors are in parentheses. * p<0.1, ** p<0.05, *** p<0.01 GRP is defined as gross regional product for each province. Trade ratio is defined as the ratio of total imports and exports to GRP. FDI ratio is defined as the ratio of foreign direct investment to GRP. Because trade, exports, imports and FDI are all reported in US dollars, we use official exchange rate to US dollars from Penn World Table to derive trade ratio, exports ratio, imports ratio, and FDI ratio. All of these ratios are in percentages. The urban-rural inequality is measured by the ratio of per capita annual disposable income of urban households to rural households. Source: Author s illustration based on Comprehensive Statistical Data and Materials on 60 Years of New China. 27

TABLE 5 GLOBALIZATION, AND URBAN-RURAL INEQUALITY (Random Effects Model with Year Fixed Effects) (Alternative Measure of Openness) Dependent variable: log ratio of urban/rural households disposable nominal income from 1978 to 2008 (1) (2) (3) (4) (5) (6) Exports/GRP in log -0.012 (0.009) 0.001 (0.009) 0.003 (0.011) 0.003 (0.011) -0.014 (0.010) -0.014 (0.010) Imports/GRP in log 0.012 (0.008) 0.001 (0.009) -0.003 (0.010) -0.001 (0.010) 0.007 (0.011) 0.006 (0.011) FDI/GRP in log -0.020*** (0.004) -0.018*** (0.005) -0.024*** (0.006) -0.027*** (0.006) -0.018*** (0.006) -0.016*** (0.006) Log ratio of urban/rural population -0.031*** (0.010) -0.030*** (0.011) -0.025** (0.012) -0.023* (0.012) -0.026** (0.012) Log ratio of urban/rural Investment in Fixed Assets 0.049*** (0.013) 0.057*** (0.013) 0.034** (0.013) 0.026* (0.014) Per capita GRP in log -0.129*** (0.038) -0.147*** (0.040) -0.092* (0.048) Initial Inequality in log 0.278*** (0.071) 0.285*** (0.069) Dummy for Coastal Province -0.119** (0.056) 0.507*** 0.469*** 0.318** -1.315*** -1.725*** -0.962 Constant (0.119) (0.123) (0.136) (0.473) (0.495) (0.621) No. of Obs. 706 601 521 521 429 429 No. of Province 29 29 28 28 24 24 R-squared 0.794 0.788 0.781 0.775 0.809 0.812 Note: Robust standard errors are in parentheses. * p<0.1, ** p<0.05, *** p<0.01 GRP is defined as gross regional product for each province. Trade ratio is defined as the ratio of total imports and exports to GRP. FDI ratio is defined as the ratio of foreign direct investment to GRP. Because trade, exports, imports and FDI are all reported in US dollars, we use official exchange rate to US dollars from Penn World Table to derive trade ratio, exports ratio, imports ratio, and FDI ratio. All of these ratios are in percentages. The urban-rural inequality is measured by the ratio of per capita annual disposable income of urban households to rural households. Source: Author s illustration based on Comprehensive Statistical Data and Materials on 60 Years of New China. 28

TABLE 6 GLOBALIZATION, AND URBAN-RURAL INEQUALITY (Random Effects Model with Year Fixed Effects) (CPI Adjusted Measure of Income) Dependent variable: log ratio of urban/rural households disposable nominal income from 1978 to 2008 (1) (2) (3) (4) (5) (6) Trade/GRP in log 0.004 (0.008) -0.002 (0.008) -0.005 (0.009) -0.003 (0.007) -0.003 (0.009) -0.003 (0.009) FDI/GRP in log -0.026*** (0.005) -0.021*** (0.005) -0.021*** (0.006) -0.024*** (0.006) -0.022*** (0.006) -0.019*** (0.007) Log ratio of urban/rural population -0.044*** (0.012) -0.047*** (0.014) -0.043*** (0.014) -0.039*** (0.014) -0.042*** (0.014) Log ratio of urban/rural Investment in Fixed Assets 0.032** (0.016) 0.043*** (0.015) 0.042*** (0.015) 0.032** (0.016) Per capita GRP in log -0.149*** (0.044) -0.132*** (0.044) -0.068 (0.055) Initial Inequality in log 0.256*** (0.076) 0.265*** (0.077) Dummy for Coastal Province -0.122* (0.063) 0.571*** 0.472*** 0.406*** -1.521*** -1.491*** -0.602 Constant (0.133) (0.132) (0.147) (0.555) (0.549) (0.707) No. of Obs. 593 506 426 426 426 426 No. of Province 25 25 24 24 24 24 R-squared 0.676 0.715 0.700 0.693 0.695 0.699 Note: Robust standard errors are in parentheses. * p<0.1, ** p<0.05, *** p<0.01 GRP is defined as gross regional product for each province. Trade ratio is defined as the ratio of total imports and exports to GRP. FDI ratio is defined as the ratio of foreign direct investment to GRP. Because trade, exports, imports and FDI are all reported in US dollars, we use official exchange rate to US dollars from Penn World Table to derive trade ratio, exports ratio, imports ratio, and FDI ratio. All of these ratios are in percentages. The urban-rural inequality is measured by the ratio of per capita annual disposable income of urban households to rural households. Source: Author s illustration based on Comprehensive Statistical Data and Materials on 60 Years of New China. 29

TABLE 7 GLOBALIZATION, AND URBAN-RURAL INEQUALITY (Fixed Effects Model with Year Fixed Effects) Dependent variable: log ratio of urban/rural households disposable nominal income from 1978 to 2008 (1) (2) (3) (4) Trade/GRP in log 0.010 (0.006) 0.011* (0.007) 0.007 (0.007) 0.004 (0.007) FDI/GRP in log -0.013*** (0.004) -0.012*** (0.005) -0.019*** (0.006) -0.018*** (0.006) Log ratio of urban/rural population -0.024** (0.010) -0.024** (0.011) -0.028** (0.012) Log ratio of urban/rural Investment in Fixed Assets 0.045*** (0.013) 0.041*** (0.013) Per capita GRP in log 0.062 (0.051) 0.596*** 0.572*** 0.414*** 1.190* Constant (0.109) (0.114) (0.127) (0.646) No. of Obs. 706 601 521 521 No. of Province 29 29 28 28 R-squared 0.794 0.789 0.782 0.783 Note: Robust standard errors are in parentheses. * p<0.1, ** p<0.05, *** p<0.01 GRP is defined as gross regional product for each province. Trade ratio is defined as the ratio of total imports and exports to GRP. FDI ratio is defined as the ratio of foreign direct investment to GRP. Because trade, exports, imports and FDI are all reported in US dollars, we use official exchange rate to US dollars from Penn World Table to derive trade ratio, exports ratio, imports ratio, and FDI ratio. All of these ratios are in percentages. The urban-rural inequality is measured by the ratio of per capita annual disposable income of urban households to rural households. Source: Author s illustration based on Comprehensive Statistical Data and Materials on 60 Years of New China. 30

TABLE 8 GLOBALIZATION, AND URBAN-RURAL INEQUALITY (Random Effects Model with Year Fixed Effects 5-year Windows) Dependent variable: log ratio of urban/rural households disposable nominal income from 1978 to 2008 (1) (2) (3) (4) (5) (6) Trade/GRP in log -0.021* (0.011) -0.025** (0.012) -0.030*** (0.012) -0.006 (0.013) -0.011 (0.013) -0.011 (0.013) FDI/GRP in log -0.041*** (0.009) -0.044*** (0.009) -0.061*** (0.010) -0.054*** (0.010) -0.039*** (0.010) -0.039*** (0.011) Log ratio of urban/rural population -0.043** (0.020) -0.054*** (0.020) -0.032 (0.020) -0.026 (0.022) -0.027 (0.022) Log ratio of urban/rural Investment in Fixed Assets 0.089*** (0.023) 0.094*** (0.022) 0.083*** (0.022) 0.082*** (0.025) Per capita GRP in log -0.199*** (0.047) -0.146*** (0.056) -0.147** (0.075) Initial Inequality in log 0.246*** (0.085) 0.244*** (0.082) Dummy for Coastal Province -0.007 (0.078) 0.332*** 0.232** -0.130-2.376*** -1.875*** -1.889** Constant (0.107) (0.111) (0.137) (0.531) (0.669) (0.956) No. of Obs. 177 157 139 139 115 115 No. of Province 29 29 28 28 24 24 R-squared 0.853 0.840 0.846 0.845 0.876 0.876 Note: Robust standard errors are in parentheses. * p<0.1, ** p<0.05, *** p<0.01 GRP is defined as gross regional product for each province. Trade ratio is defined as the ratio of total imports and exports to GRP. FDI ratio is defined as the ratio of foreign direct investment to GRP. Because trade, exports, imports and FDI are all reported in US dollars, we use official exchange rate to US dollars from Penn World Table to derive trade ratio, exports ratio, imports ratio, and FDI ratio. All of these ratios are in percentages. The urban-rural inequality is measured by the ratio of per capita annual disposable income of urban households to rural households. Source: Author s illustration based on Comprehensive Statistical Data and Materials on 60 Years of New China. 31