The Effects of Trade Liberalization on
Gender Wage Gap in China
Jingyi Han 11815116
Msc. in Economics - Development Economics
Business and Economics
University of Amsterdam
Master thesis
Supervised by Prof. Erik Plug, PhD
August 2018
Statement of Originality
This document is written by Student Jingyi Han who declares to take full
responsibility for the contents of this document.
I declare that the text and the work presented in this document are
original and that no sources other than those mentioned in the text and its
references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the
supervision of completion of the work, not for the contents.
Contents
Abstract:...4
1.Introduction:...5
2. Literature Reviews:... 7
3. Methodology... 10
4. Data...15
4.1 Data Resource... 15
4.2 Description of the model and relative variables...16
4.3 The tendency of Chinese gender wage gap...17
5. Results... 19
5.1 Results of regional trade liberalization degree...19
5.2 OLS regression of trade liberalization on gender wage gap...21
5.3 Robustness Check...26
6. Conclusion and Discussion...29
6.1 Conclusion... 29
6.2 Discussion...29
Abstract:
In the past 30 years, the world has been experiencing a significant change in the process of globalization and trade liberalization. At the same time, China was on the implement of the opening up policy and the social-economic reformation. Moreover, China was recognized to have achieved many improvements especially in the economic aspect. The changes which brought about by the economics transition have led to many massive structure adjustments in the Chinese labor market. One of the significant issues was the widening inequality income distribution between genders. The rising income inequality will not only decelerate the development of labor market but also react on the trade liberalization. To avoid the side effects result from the exaggerated gender wage gap, studying the relationship between trade liberalization and income distribution is critical. However, there are limited existing studies on the issue of how trade liberalization affects gender wage gap in China; and many of them have two shortages: 1) using the macro-level data; 2) unclear definition of the measurements of the trade liberalization index. In this article, I start with these two aspects to research how trade liberalization affects the gender wage distribution. Firstly, combining the Mincer wage equation and the multivariate regression analysis to do the OLS regression, I found that the wage gap became larger after 2001 when China went on the faster track of trade liberalizations as compared to before 2001. Secondly, introducing an index into OLS regression to measure the provincial trade liberalization degree through their industrial structure, I get the conclusion that the gender wage gap widens faster in the high trading states than those in low trading states. This result means that, in China, the higher level trade liberalization degree will widen the gender wage gap.
1. Introduction:
Under the big wave of globalization and the promotion of the reform and opening up policy, China has achieved rapid economic growth and a deepened trade liberalization progress since the 1980s. This economic achievement and trade openness have inevitably brought lots of shifts that changed the labor market structure. The most noticeable one is that the principal value of the limited economy which focuses on the efficiency has gradually replaced the value of the market economy which concentrates on the egalitarianism. This transformation altered the labor demand and supply in China because of various provincial governments focusing on multiple industries: some female-leading industries, while others require male labor most. What’s more, the salary determination system has turned from the single-set mechanism which determined by the central government in the past to the market-oriented system which government only have limited power to adjust employees’ salaries. The market-determinate-wage mechanism is helpful to accommodate the changes of the supply and demand in the labor market and the returns of the investment on human capital. The improvement of productivity caused by trade liberalization and economic development considerably raised the nominal and real wage of Chinese workers. On the other hand, the imbalance of the labor supply and demand and other institutional problems which appeared during the economic transition have led to the widening inequality on the income distribution among the different social groups, especially between different gender groups(Ding S, Dong X, and Li S, 2009). The widening gender wage gap has been observed by many scholars in China especially in the past 30 years.
However, most of the current studies, which on the issue of trade liberalization, focus on how trade liberalization affects poverty, economic development or growth. Parts of them concentrate on how trade liberalization impact on the wage distribution in industrial or export sectors and on the regional level. The relationship between trade liberalization and gender wage differences has not been considered as a serious.one. Gender wage gap as one of the most critical aspects can reflect the gender equality
(Kingdon, 2002). Significant wage inequality between genders infringes the workers’ legitimate rights and benefits and hinders the further improvement of labor productivity which will result in the social instability. To study the gender wage gap will be helpful to improve the production motivation and labor productivity of female workers and promote the efficiency of the labor market during the progress of economic transition.
Besides, most existing studies on the impact of trade liberalization on income distribution are usually using the macro-level data (Holz, 2004; Milannovic and Squire, 2007), which restricted by the reliability of data resource, the degree of freedom and the exogeneity between the outcome variables and dependent variables, for example, there is a causality between the increasing GDP and the market openness degree. And the theoretical supports of the defined measurements of trade liberalization degrees are obscure. With the enrichment of micro-level data, it becomes particularly important that how to construct the clear and valuable appropriate regional trade liberalization indicators.
Considering the two aspects, in this article, I use the data of China Nutrition and Health Survey(CHNS) data which is a micro-level the household survey and get the observations through screening and matching the database, to ensure that all samples are surveyed in the same continuous period. I first research the average gender wage difference in China from 1989 to 2011; and then measure the change of the gap under the influence of the roundly market liberalization in China, by combining the Mincer wage function with the multivariate regression analysis and controlling the power of education, age, and marriage on the wage. After that, referring to the model that Zhang Mingzhi, Liu Durou, Deng Ming proposed in 2014, I introduce an index to indicate provincial trade liberalization degree in China by associating the weighted labor employment indicators with the import tariff rate. The principle of this idea is to reflect trade liberalization degree among areas regarding their local industrial structures. With this index, running the OLS regression shows the effects of different levels of trade liberalization in various regions on gender wage gap. The results show
that the income inequality distribution between genders in China increasingly become wider after 2001, and the gender wage gap widens faster in the high trading states than those in low trading states.
This paper is structured as follows. The next section is the existing literature reviews on trade liberalization and the gender wage difference in both developed and developing economies. The third part outlines the methodology and the empirical models used in this article. The fourth section discusses the data and the measurement of local captures. The results and conclusion of this study are presented in the final two chapters.
2. Literature Reviews:
The theoretical foundation of the relationship between gender wage difference and trade liberalization is from two theories, one is the H-O-S theory and the other one is the Labor Market Discrimination theory. Both argue that with the expansion of trade, the total labor demands will increase and the relatively raise of workers’ wages will narrow the gender wage difference. On the other hand, Becker(1957) pointed out that the marginal productivity and working time do have essential effects on workers’ salaries. The factors affect the marginal productivity including the level of education, working skills and working experience and so on. Only excluding these indicators can we say that the gender discrimination influences labors’ income.
Currency research literates on trade liberalization and gender wage difference mainly concentrate on some certain countries or particular industry sectors, however, the conclusions of these papers present opposite directions from the two theories (Goldberg and Pavenik, 2007). To summarize, the gender wage gaps are still significant in most countries, and when doing transnational comparisons or some country-specific case studies, the trends are considerable different among countries (Xi Y, Zhang X and Cao L, 2013).
The literatures on the developed economies, Melly(2005) used household survey data of American and British to analyze the gender wage difference. They got the results
show that in these two countries’ labor markets, gender discrimination can explain parts of the wage gap between female and male; Also the discrimination effects on America's labour market are more detriment than that in Britain. Black and Brainerd(2004) analyzed the gender wage distribution of the export-oriented manufacturing industry in American since the 1970s. The research shows that international trade has enhanced the international competition and narrowed the existing gender wage gap in the United States. However, in 2009, Saure and Zoabi used the data of the same region from 1990 to 2007 to study the trade effects of NAFTA and found that the US gender wage gap was widening. Gibb, Fergusson and Horwood(2009) decomposed the gender wage gap by considering the human capital, jobs characteristics and family responsibilities based on the studies of domestic households in New Zealand; the results show that these factors can only explain 66.4 % of the wage undistributed between genders, discrimination is another reason which resulting in the gender wage difference of New Zealand. Besides, Cassels, Vidyattama and Miranti ‘s research in 2009 has indicated that if the gender wage gap lowers 1%, Australia's GDP will grow by 0.5%, it also illustrates that the widening gender wage gap plays a particular role on a country or region’s macroeconomic development.
For the developing countries, Menon and Rodgers(2009) studied India's household data from 1983 to 2004 and concluded that trade liberalization had widened the gender wage gap in India's manufacturing industry. Seguino(2000) analyzed the gender wage gaps in South Korea and Taiwan, China, and found that the wage inequality between genders in South Korea was decreasing, while Taiwan’s wage inequality was rather expanding. The following study by Berik(2004), after an empirical analysis of the manufacturing industry in Korea and Taiwan, indicated that with the deepening process of market openness, the gender wage gaps in both places were expanding, and more obviously in Taiwan.
In terms of the gender wage gap in the China, by using the Chinese household and provincial level data from 1995 to 2002, Braunstein and Brenner(2007) drew the
conclusion that foreign direct investment had a positive effect on the wage of both male and female, and women benefited more from FDI in 1995 than men, while the advantage disappeared in 2002. Ding(2009) separated the period of China's trade liberalization and economics transition into two parts from 1988 to 1995, and from 1995 to 2002. During both two periods, the trend of the gender wage difference kept widening. The most representative study about China is Hering and Poncet(2010), they thought the gender wage differences could explain by geographical differences related to market access by constructing a geological economics model. Liubin and Lilei(2012), based on the research ideas and methods of Hering and Poncet, adopted the CHIP1 data and concluded that trade liberalization had widened the gender wage gap.
The existing papers on China's trade liberalization and gender wage gap have offered various perspectives and measurement methods. However, comparing with that frontier literature abroad, there are still some shortcomings gathered on two aspects: first, the relevant foreign researches often use the time series data of the same individuals, so that they can do some dynamic analysis to study the relationship between trade liberalization and the change of wage distribution. The current influential studies on China mainly use the cross-sectional samples of one single year, so it is impossible to gain the dominant control of trade liberalization on the wage changes of the same individual. Secondly, regarding the selection of indicators to measure the regional trade liberalization degree, most studies in China used the signs such as local import penetration rate or some similar approximate ones. A prominent problem of such indicators is that it is hard to avoid their endogenous associations with GDP or other independent variables raised by Milanovic and Squire(2007). Unlike trade dependence degree or import and export penetration rates, tariffs rate combined with industrial structures I use in the following parts have little inherent problems which has been proved by Brandt et al. in 2012.
3. Methodology
To answer the research question how the trade liberalization affects the gender wage gap in a step by step procedure, this section introduces three phases and all of them are based on the micro-level data of certain individuals in a continuous period from 1989 to 2011 in 11 provinces in China. Firstly, I use the Mincer wage function combined with multivariate regression analysis to estimate the effects of gender difference on the income distribution in the Chinese labor market since 1989. Then I divide the survey years into two stages to compare the unequal wage distribution of gender at the different period. The selected threshold is the year 2001 when China formally entered the WTO and is considered as a milestone of China’s trade liberalization process; this step serves to estimate whether the improvement of the degree of trade liberalization leads to the changes of gender wage gap. The final section studies how different levels of regional market openness affect the gender wage gap, through running the regression with an index which can reflect the regional trade liberalization degree for the period after joining the WTO. The index links the tariff of different imported goods which used as a sign of trade openness with the weighted labor employment rate which can reflect the provincial industrial structures impacted by trade openness. Through this step, I can indicate and verify that the different levels of trade liberalization in different provinces will play various roles on gender wage gap.
The first step base on the Mincer wage function and multivariate regression analysis function is to measure the existing gender wage difference since 1989 in the Chinese labor market. The function as follows:
Gender
Age
Age
Edu
Marr
P
o
w
^
2
r
ln
0 1 1 2 3 4 5The left side is the outcome variable which indicates the wage level and using logarithm as usual; The right side includes the variable of gender, the variables that reflect the labors’ productivity and personal characteristics and the error term.0is
coefficient of gender and
is error term.The independent variable, gender, is a dummy variable; the control variables are those can not only affect the wage level but also reflect the individual characteristics like workers’ age, educational level, and marital status. In addition, the province fixed effects were also introduced into this model to distinguish the difference between regions.
As shown in the formula, the wage level depends on gender, the observed individual characteristics and some unobserved factors. The estimated value of 1 is a critical parameter to show the difference between gender, because if 1 less than 0 means that the wage level of the female is lower than male when the other conditions are the same; and if 1 great than 0 means that men earn less than women in the labor market.
In the second step, I separate the observed period into two parts before and after 2001 which is the remarkable time of trade liberalization. Because in 2001, China formally became a member of WTO and since then China has held its promise to construct a rule-orientated market and trade-liberalized economy. Due to this transformation which brings a significant amount of foreign direct investment and creates a vast amount of job opportunities, entering the World Trade Organization is usually viewed as the turning point of Chinese trade liberalization. As the time passes, the market openness degree is deepening and the competition is becoming even fiercer. How does the process of trade openness affect the gender wage gap? Will the gender wage gap widen or shrink? Therefore, this function introduces the time dummy to compare the change of the gender wage difference in the different period.
The model of the second step is: Before 2001:
Gender Age Age Edu Marr pro
w 1 2 3 4 5 1 1 1 ^2 ln After 2001:
Gender
Age
Age
Edu
Marr
pro
w
1 1 1 2^
2
3 4 5The model aims to compare the value of 1and observes how the effect is of trade
liberalization on gender wage distribution. According to the H-O-S theory and labor market discrimination theory, the assumed null hypothesis is that the trade liberalization will narrow the gender wage gap due to the increasing job positions and the highly employment discrimination cost. From the first phase, the estimated value of gender wage gap 1 usually lower than zero according to some available papers which show that female earn less than male. And if indicates that female earn less more than the man after 2001, from that we can know that the gender wage gap has widened after entering the world trade organization, and vice verse.
After ensuring that it is the trade liberalization which affects gender wage gap in the second step, to further prove the relationship between trade liberalization and gender wage difference, I construct a regional trade liberalization indicator into the previous model which classifies the surveyed provinces according to their sensitivity to the process of trade liberalization. This idea is that if the trade openness affects the gender wage gap, the effects on high trading states should more profound.
In this part, how to construct a consistent regional trade liberalization measurement is critical, and it has been discussed for a long time in the research field of trade liberalization and wage distribution. The groundbreaking achievements are credited to Topalova(2007) who made a pioneering attempt and is viewed as a representative of the construction of regional trade liberalization index. However, due to a lack of academic supports and there is no theoretical discussion on the relationship between regional trade liberalization and wage change in this study, the empirical research achieved by Topalova on the measurement construction is not rigorous. Kovak(2013) built another regional trade liberalization index which depends on the weighted employment structure with using the specific elements of Topalova model and modeled the relationship between income distribution and regional trade development. Zhang Mingzhi, Liu Durou, Deng Ming(2014), base on the theory and method of the Kovak, combined the background of the greatest tariff concession in
2 1 1
1
)
1
ln(
)
1
ln(
)
1
ln(
2000 i t i t iTariff
Tariff
Tariff
d
China in the 21st century to gain the indicator which is targeting on analyzing the relationship of China regional trade liberalization and individual income distribution. The indicator used to measure the Chinese Regional Trade liberalization degree in this paper combine the ideas of Kovak(2013) and Zhang Mingzhi, Liu Durou and Deng Ming(2014).
The Regional Trade Liberalization degree is:
represents the trade liberalization degree of province r in the year t. means the import tariff of industry i in the year t.
is the difference of the tariff in the industry i between the year t and the primary year.
According to the research question, the primary year I chose is 2000 which a significant point in the progress of Chinese trade liberalization, and China has experienced a great tariff concession from that year on.
The change of import tariff which based on the year 2000 is:
The tariff data summarized from WTO annual report(country for China) and used the part of the Applied-Non MFN(duty for Most Favored Nations) which covers the most types of goods that China import and directly shows the industrial structure.
represents the effects on regional wage level brought by the inflation of the products’ prices. Jones(1975) thought that the changes in products’ prices could affect the regional labor employment structure. Feenstra(2004) proposed that considering the adverse effect of trade liberalization on the presence of tariff concession on product prices, when the tariff rate becomes lower, the degree of trade liberalization increases, the protection of the domestic market decreases, and the increasing competition from foreign products will inevitably cause the decreasing of the price of each product in this country. The existing research(Topalova, 2007; Hasan et al., 2007; Kovak, 2013) use the import tariff rate to replace the price change rate so that
ri t i Tariff t r
RTL
i t i ri t r d Tariff RTL
ln(1 ))
1
ln(
t iTariff
d
the tariff after weighing the labor employment structure can measure the trade liberalization degree. Although the trade openness makes all regions have the same concession of the tax, but the employment structure of labor varies from region to region, the trade liberalization degree will differ among areas. Besides, if the workers of one area gather densely in a specific industry, the small changes of the tariff in the products will have considerable influence on the income distribution.
According to Kovak (2013) who reconstructs :ri
j rj j i ri ri 1 1is the cost share of the particular factors in the industry i which equals one minus the share of labor cost. Also, the share of labor cost equals to the ratio of the total wage of industry i to its added value. means the percentage of the labor employed in the sector i accounted to the total employment in the province r. The data for labor employment was collected from the National Bureau of Statistics of China.
In this paper, I chose eight kinds of industries which provide more than 80% of labor employment in China. There are 1) Agriculture, Forestry, Animal husbandry and Fishery, 2) Mining Industry, 3) Manufacturing industry, 4) Production of electricity, gas, and water, 5) Construction industry, 6) Transportation and Warehousing, 7) Telecommunications, Information transmission, Computer Services and Software and 8) Wholesale and Retail. The products’ tariff rates are collected from the WTO annual report from 2001 to 2011, and because of the different initials of HS Codes2 represent different types of industries, I use the HS Code of the version 2017 to classify the products into eight industries. From averaging the products’ tax rates, I got the tariff of each sector in the different years.
I gained the degree of trade liberalization for a specific area though summing the value of and then using this value to time the changes of duty
i ri ln(1 t) i Tariff d
in different industries. This measurement combines the tariff and the sign of trade liberalization with the labor employment structure which possess the regional industries characteristics.
By the results of the regional trade liberalization degree, the eleven provinces CHNS surveyed was distinguished into two parts: trade sensitivity areas and trade less-sensitivity areas in the period of entering WTO (after 2001).
The model with trade dummy variable became:
Trade
pro
Marr
Edu
Age
Age
Gender
w
6 5 4 3 2 1 1 1^
2
ln
3Trade = 1, if the province industries’ structure is sensitive to trade liberalization; Trade = 0, if the province industries’ structure is insensitive to trade liberalization. So if the gender wage gaps in the provinces which were significantly affected by trade liberalization would be wider than those less affected after 2001( 13 12 ),
this can verify that the trade openness is the prime culprit that expanding the wage difference among gender.
4. Data
4.1 Data Resource
The data used is the CHNS household survey which is a follow-up project studied between the China disease control and prevention center and the population center of the North Carolina University of the United States. CHNS aims to explore the influence of China social-economics transformation and the one-child policy implementation on the nationals’ health and nutrition. The survey started in 1989 and included ten years(1989, 1991, 1993, 1997, 1999, 2000, 2003, 2006, 2009, 2011 and 2013) and covered the urban and rural areas of 11 provinces in China. It collected the demographic characteristics, education condition, income level and health condition. Besides, there are some detailed community data which contain the food market, medical institutions, and other social services. Therefore, many empirical studies
nowadays use CHNS data to analyze some issues about Chinese labor market, and CHNS is applying widespread from the economics to sociology field.
4.2
Description of the model and relative variables
The model I used is to combine the Mincer wage function and multivariate regression analysis function:
Gender
Age
Age
Edu
Marr
P
o
w
^
2
r
ln
0 1 1 2 3 4 51) The outcome variable is the average monthly wage for the last survey year and then taking logarithm. CHNS collected the data on hourly pay, days wage, and the wage per piece completed work. Considering the time that female and male spend on the family is difficult to count, the monthly salary will summary the average level of efforts that people put on their work. Otherwise, the work of hourly wage, days wage, and the per piece-wage is apt to use the workers with low education level which is not representative for the whole labor market in China. Also, there are many missing values in the variable working hours and it is difficult to measure hourly earning, so directly using monthly wage can keep enough sample sizes and data credible.
2) Education level is an important indicator to reflect human capital resource. The higher educational level will bring more added value to labors’ occupation, and the people with high-level education usually have higher productivity and therefore get a high wage. CHNS have gathered the data of individuals’ education level as primary school, secondary school, high school and university. In this article, I obtain the years of schooling by adopting the data of the age that people completed education and then using it to minus six which is the legal age for children to go to school.
3) Marriage is another considerable factor which may affect the wage distribution between genders in the labor market. Individuals, especially women, need to spend more time and energies on the family after marriage. In China, many companies even will not employ single women because they think these women will cost more time during pregnancy as compared to those who already have children.
China, the eastern part is economically developed, while the central and western areas are in a little bit backward. Due to the imbalance in the economy, the demands of labor markets are various, same as the human capital and employment competition, all of these factors will affect income. Otherwise, the backward of the economic development caused by trade openness will also influence the trend of gender wage gap in one region, the gender discrimination usually more severe in economic developing areas than in developed areas. So I add province dummies to solve the differences among regions in China. There are elevens provinces CHNS surveyed which cover five regions in China and can reflect their characteristics. The CHNS institution started to do the surveys for Chongqing province in 2011 while the data collection for Shanghai province began in 2013.
4.3
The tendency of Chinese gender wage gap.
Graph 1.
The graph above shows the trend of the monthly wage between female and male from 1989 to 2011 separately. We can see that the total wage level in the Chinese labor market is growing since 1993, and from then on, an apparent wage difference occurred between female and male. However, the gap was negligible before 2001 as compared to the period after, and it is continuously expanding especially after the
year 2006. These changes in the wage gap coincide with the deepening process of the opening up policy and trade liberalization in China. This pattern indicates that the effects of the trade liberalization on income distribution between genders in China have become a controversial issue. Because of its paradoxical performance: increasing the total wage level on the one hand, but widening the gender wage gap on the other hand.
From the table 4.1, we can directly compare the gender differences among their wage, educational level, and marital status. The average salary of women workers($111.45) was lower than men($140.12) in the Chinese labor market, and the wage gap extended in the period after 2001 when men can earn $227 on average, while the wage of women were only $166.3 was about one third lower than men’s.
The average age of people in the surveys was 38 and ranged from 36 to 42 which is thought of the golden age for work. Also, female workers were three years younger than their male counterparts. For the education level in China, people surveyed usually spent 13 years in school before they entered the labor market. After 2001, the years of schooling have averagely increased by four years for both female and male
Table 1.
The Characteristics of Observations
Average Female Male
Mean Before2001 After2001 Mean Before2001 After2001 Mean Before2001 After2001 Wage $102.94(12.45) $32(3.66) $285.47(54.51) $111.45(12.99) $25.46(4.96) $166.7(38.05)$140.12(15.28) $44.42(4.52) $227.87(39.78) Age 38.51(0.07) 36.32(0.09) 42.12(0.17) 36.35(0.11) 34.28(0.14) 39.29(0.15) 39.94(0.10) 37.64(0.13) 43.15(0.01) Edu 13.13(0.04) 11.72(0.05) 15.10(0.10) 13.09(0.07) 11.38(0.09) 15.22(0.10) 13.22(0.05) 11.95(0.07) 15.04(0.07) Marriage 1.89(0.00) 1.82(0.00) 2.00(0.01) 1.89(0.01) (0.01)1.80 2.01(0.01) 1.89(0.00) 1.83(0.01) 1.98(0.01) No. of obs 26,833 15,692 4,135 10,155 6,125 4,488 15,573 9,098 6,653
one of the crucial reasons why the wage level in China keeps raising in recent 20 years. However, differing from the wage level, the education degree of men and women was almost the same no matter the time. Moreover, in the period after 2001, women even accepted slightly more education than men, due to the one-child policy that may improve female’s social status but the wage gap was still significant. In the CHNS surveys estimating the marital status, ‘1’ means never married, ‘2’ means married and ‘3’ means divorced, the majority of surveyed people in China was in the marriage, and the situations were consistent between gender. Otherwise, the sample sizes of the male were more massive than female about five thousand in totals.
5. Results
5.1 Results of regional trade liberalization degree
The method to measure the degree of regional trade liberalization is based on the level of the year 2000 by weighting the labor employment factors and import tariffs which consider as to reflect their industrial structures’ sensitivities to the market openness.
Graph 3
From the above graphs, we can conclude that in the years before 2001, the average trade liberalization degree was up-and-below zero in China, and this situation started to change in 2004. As becoming the members of WTO, the trade liberalization degree had shown a significant improvement, and each province in China experienced an increase in the marketization and trade openness. However, the trade degree varies in the region level, the trading level in some areas grew faster while others moved gently. These changes tended to be stable after the year 2006 because we notice that the line of 2009 overlap with that of 2011. As Shengbin(2011) pointed out that during 2001 to 2006, the Chinese government cut down the tariff rates almost in all tax subjects, the policies implemented were no later than the December of 2006. Among them, the duties on manufactured goods decreased from 14.3% to 8.9 %; the rates on automobiles and their parts, textiles, and clothing decrease about a half. In comparison to the tax rates after 2006, there was no more great reduction, so the process of trade liberalization was not changed a lot after that.
Shanghai, Jiangsu and Shandong provinces which are located in the eastern coastal area of China had a higher level of trade liberalization degree up to 0.15 in 2011, indicating that they are sensitive to the market openness. In these areas,
which are significantly affected by the wave of the tariff concession and opening up policy. Although the degrees in other provinces also showed an ascending direction, like Chongqing, Henan, Heilongjiang, their degree responded to the trade openness was not distinct and less than 0.08 which was one half compare to the degrees of trade sensitive regions. The pillar industries in the areas which are in the lower trade liberalization level, for example, in Henan Province, they are center in the primary and secondary sector, the impact of trade liberalization on these industries has been modest.
According to the calculated trade liberalization index, I separate those provinces into trade sensitive and less sensitive types. Shanghai, Jiangsu, and Shandong where the trade liberalization degree is higher than others are the trade-sensitive provinces, and the others are less trade-sensitive provinces. Otherwise, the graph 3 demonstrates a significant difference in the extent of trade liberalization between the trade-sensitive regions and less trade-sensitive regions. Also, it shows that the trade-sensitive provinces tend to have a similar trend and the less trade-sensitive areas have another.
5.2 OLS regression of trade liberalization on gender wage gap
Table 2 indicates the OLS results of gender wage gap in China in the different period with the provinces fixed effects, years dummies and trade liberalization degrees separately. The first column in Table 2 estimates the whole level of gender wage gap in China and includes all provinces and surveys years from 1989 to 2011; we can conclude that female workers earned about 21% less than males on average in China when controlling the variable of age, education, and marriage. In the labor market, age and marital status can explain the inequality of income distribution slightly, but one additional year of schooling will exactly increase people’s wage level. The changes of the gender wage gap in the past 30 years became divergent after distinguishing the period with using the mark of China entering the World Trade Organization. Before 2001, Chinese women earned about 19% lower than man; and at the same time, the influence of age and education on wage was declining in contrast to those in column 1. During that period (before 2001), we know that the
education level and age for both men and women was almost at the same lower level3,
So this wage difference mainly came from the gender discrimination or the institutional problems in the Chinese labor market and society. Besides, the adverse effect of marriage on wage for gender has grown.
After 2001, when China was on the way to the rapid development of trade liberalization, the gender wage difference increased significantly to 25% on average,
Table 2.
Regression Results of Gender Wage Gap
Average Before2001 After2001 After 2001[1] [2] [3] Low TradingProvinces High TradingProvinces Gender -0.210***(0.008) -0.188***(0.010) -0.252***(0.012) -0.228***(0.014) -0.313***(0.021) Age 0.033***(0.002) 0.028***(0.002) 0.038***(0.003) 0.030***(0.04) 0.046***(0.006) Age^2 -0.0004***(0.000) 0.000***(0.000) 0.000***(0.000) 0.000***(0.00) 0.000***(0.000) Edu 0.014***(0.000) 0.004***(0.001) 0.032***(0.001) 0.033***(0.001) 0.027***(0.001) Marriage (0.007)0.002 (0.012)-0.005 (0.009)0.004 (0.010)-0.006 (0.017)0.032 Province yes yes yes yes yes
Year yes
Before 2001 yes
After 2001 yes yes yes
cons (0.048)4.286 4.091***(0.048) 5.923***(0.080) 5.964***(0.009) 6.182***(0.138) No.of obs. 25,728 14,890 10,838 7,447 3,391
R-squared 0.7927 0.6133 0.403 0.4019 0.4141
Adj
R-squared 0.7925 0.6128 0.402 0.4006 0.4124
Statistical significance: *p <.05; **p <.01; ***p <.001; Standard errors in the parentheses. Column 1,the average gender wage gap between 1989 and 2011;Column 2, the gender wage gap in the period before 2001; Column 3,the period after 2001;
which verify that trade liberalization negatively impacts the income distribution on genders. To estimate this relationship in a more profound way, I divide the provinces according to their trade sensitivity; for the area where insensitive to trade, the gap was 22.8% and closed to the whole level. However, in the sensitive areas which directly affected by the market openness, the uneven income distribution between genders has considerably risen to 31%; this means that when the females do the same job as male, the wage of the female was about one third less than the men’s salaries. According to the theories mentioned in section 3, the null hypothesis in this article is that as the deepening of the progress of the trade liberalization, the gender wage gap will narrow. This is because as the demands of workers increase mentioned in H-O-S theory, the cost of labor discrimination will increase when facing cut-throat competition, the employers will give up their gender preferences. From the results, we cannot accept the null hypothesis, and the p-values are statistically significant, so it is the trade liberalization that leads to the opposite effect on the gender wage gap. The sample size of observations in before 2001 is more lager than the period after, and for the trade less sensitive provinces the survey's data is more abundant, mainly due to the division, only three zones are defined for the trade-sensitive areas. The smaller sample sizes for high trading provinces lead to less precise estimate because the R-square change from 0.8 to 0.5.
Table 3 ranks the provincial gender wage gap after 2001 from low to high under the background of trade openness. The disparities among provinces approached 19%, the wage difference in Shanghai was 17% while in Jiangsu, it up to 36%. Shandong and Jiangsu, as trade sensitive provinces, the income gap between female and male workers arrived in 36% outdistanced the whole level. The income distribution in Shanghai, the remaining area defined sensitive to trade progress, was only 16.4%. This is due to the survey data that CHNS collected in Shanghai started in 2011 and as compared to the other two trade sensitive regions, the population, and areas covered by Shanghai were smaller. Another reason is after the year 2006. Shanghai has updated its industrial structures into service business which is not included in the
index when measuring the trade liberalization degree. Finally, for insensitive trade provinces, the uneven income distribution between genders concentrates on 20%~23% and 10% less than those in trade-sensitive areas.
Table 3.
Provincial Gender Wage Gap
After 2001 Shanghai -0.164**(0.048) Heilongjiang -0.189***(0.039) Guangxi -0.190***(0.033) Hunan -0.221***(0.043) Henan -0.225***(0.045) Guizhou -0.229***(0.047) Hubei -0.242***(0.044) Beijing -0.246***(0.050) Chongqing -0.254**(0.080) Liaoning -0.291***(0.040) Shandong -0.356***(0.041) Jiangsu -0.359***(0.031) Statistical significance: *p <.05; **p <.01; ***p <.001; Standard errors in the parentheses.
Graph 44.
The graph above displays the gender wage gap for various provinces transversely according to the survey years, all areas present a deepening gender income disparity with time and there is a more severe process in the trade-sensitive regions. The tendency of these two type provinces almost in a similar direction from 2000 to 2009, but there are many differences in the change of the gender wage gap between this two-type provinces. In 2000, female in Henan, when they did the same jobs and have had the same educational background as their male competitors, could earn almost the same salary as male, and it is about 10% lower for female’s salaries as compared to males’ in Heilongjiang and Liaoning provinces. However, in another area of China, Jiangsu, the income was 28% less in 2000 and extended to 46% in 2009, while at the same year (2009) in Henan, Liaoning, and Heilongjiang, the gap was 25%. Although in Shandong province the starting gap in 2001 was not as extensive as it in Jiangsu, the difference still gathered at 20% and moved to 43% in 2009.
From the result, we can conclude that the income distribution was more unequally between genders in the provinces with a high market openness level as compared to the areas not influenced severely by trade liberalization.
4 Trade sensitive provinces: Shandong and Jiangsu; Trade insensitive provinces: Liaoning, Heilongjiang and
5.3 Robustness Check
After proving that the improvement of market openness is disadvantage to the gender wage gap by comparing the difference around entering WTO, the results pointed above also showed that after 2001 the regions with high trading level usually have a more widening gender income inequality than those with low trading degree. During this part of the robustness check, I recalculated the regional trade liberalization index for the years before 2001 by removing the agricultural and construction from the original six industries5 to build a new industrial structure which gives priority to manufacturing. The idea comes from Topalova(2013) and Hasan et al. (2007) who argued that the process of trade liberalization would not have a direct impact on other productive sectors than manufacturing. Then concerning the index reacquired, Jiangsu, Liaoning, and Hubei become the trade-sensitive provinces during the period before 2001. The column 2 and 3 in Table 4 indicate the gender wage gap before 2001 in trade-sensitive provinces and less sensitive regions. The gaps in the trade-sensitive areas nearly equal to that in the trade closed areas at 18.5%. The results demonstrate that before 2001 the income inequalities between genders were in a stable position among China, compared to those after 2001, the gap widened a lot on the average provincial level and also enlarged the difference between regions. For trade less-sensitive areas, the gap changed from 18% to 22%; and in trade openness regions, the disparity up to 31%.
Table 4.
Regression Results of Gender Wage Gap
Before 2001 After 2001 Average TradingLow
Areas High Trading Areas Average Low Trading Areas High Trading Areas Gender -0.188***(0.010) -0.183***(0.019) -0.189***(0.011) -0.252***(0.012) -0.228***(0.014) -0.313***(0.021) Age 0.028***(0.002) 0.025***(0.005) 0.031***(0.003) 0.038***(0.003) 0.030***(0.04) 0.046***(0.006) Age^2 0.000***(0.000) 0.000***(0.000) 0.000***(0.000) 0.000***(0.000) 0.000***(0.00) 0.000***(0.000) Edu 0.004***(0.001) (0.001)0.004* 0.004***(0.001) 0.032***(0.001) 0.033***(0.001) 0.027***(0.001) Marriage (0.012)-0.005 (0.026)-0.006 (0.013)-0.006 (0.009)0.004 (0.010)-0.006 (0.017)0.032 Province yes yes
Year
Before
2001 yes yes yes
After
2001 yes yes yes
cons 4.091***(0.048) 4.123***(0.093) 4.050***(0.055) 5.923***(0.080) 5.964***(0.009) 6.182***(0.138) No.of obs. 14,890 4,063 10,827 10,838 7,447 3,391
R-squared 0.6133 0.5792 0.6254 0.403 0.4019 0.4141
Adj
R-squared 0.6128 0.5781 0.6249 0.402 0.4006 0.4124
Statistical significance: *p <.05; **p <.01; ***p <.001;Standard errors in the parentheses.
Column 1, the gender wage gap in the period before 2001; Column 2, period before 2001 with less trade liberalization sensitive countries; Column 3, period before 2001 with trade liberalization sensitive countries; Column 4,the period after 2001;Column 5, period after 2001 with less trade liberalization sensitive countries; Column 6, period after 2001 with trade liberalization sensitive countries
Table 5 .
Provincial Gender Wage Gap with
Trade Liberalization Degree
Trade Liberalization Degree(<2001) Before 2001 Trade Liberalization Degree(>2001) After 2001 Jiangsu -0.0016 -0.227***(-0.022) 0.1025 -0.359***(-0.031) Liaoning -0.0026 -0.195***(-0.024) 0.0348 -0.291***(-0.04) Beijing -0.0027 0.0445 -0.246***(-0.05) Guangxi -0.0027 -0.140***(-0.027) 0.0627 -0.190***(-0.033) Hubei -0.0029 -0.203***(-0.025) 0.0434 -0.242***(-0.044) Chongqing -0.0030 0.035 -0.254**(-0.08) Guizhou -0.0035 -0.167***(-0.036) 0.071 -0.229***(-0.047) Hunan -0.0036 -0.188***(-0.031) 0.0501 -0.221***(-0.043) Heilongjiang -0.0042 -0.152**(-0.047) 0.0369 -0.189***(-0.039) Henan -0.0042 -0.209***(-0.033) 0.0397 -0.225***(-0.045) Shandong -0.0078 -0.175***(-0.027) 0.0959 -0.356***(-0.041) Shanghai -0.0102 0.1389 -0.164**(-0.048) Statistical significance: *p <.05; **p <.01; ***p <.001; Standard errors in the parentheses;
Ranking order according to the trade liberalization degree from high to low.
There is little difference between trade liberalization degrees among provinces before 2001 in table 5, and the trade degrees for all the surveyed provinces were less then zero which means that at that time China was on the position of the locked market.
(-0.0016) at an insignificant level in contrast to that after 2001(0.1025). Although the opening degrees of trade in all provinces before 2001 were generally low, we can still sum up from table 5 that a high degree of trade liberalization corresponded to a high gender wage gap. This conclusion also accommodates for the period after 2001; the trade liberalization degree could usually lead to the changes in gender wage distribution.
6. Conclusion and Discussion
6.1
Conclusion
The gender wage gap is a standard issue around the world. During the economic transformation, China's labor market has become complicated and deeply affected by the market and trade opening up progress. Based on the related theories and the data of Chinese nutrition and health survey, this article mainly use the OLS regression, from the representative of theory and practice, to study and analyze the relationship of trade liberalization on income distribution between genders especially for the period that China has become the membership of WTO. The research results show that the gender wage gap in China existed before trade liberalization improvement and enlarged after that. Education and other personal characteristics can explain parts of the reason for the gender wage gap, however, when in the process of trade liberalization, although female and male are likely to have the same level of education, the gender wage gap expanded more in the trade-sensitive provinces. It is the process of trade liberalization that result in the widening gender wage gap in China. Moreover, the wage difference between regions concerns the regional trade liberalization degree concluded that female in the trade sensitivity provinces tended to earn less more than male as compared to those who work in less-sensitive areas.
6.2 Discussion
Regarding the results, the gender wage gap in China enlarge increasingly when the country in a high trading liberalization level and the inequalities among regions are
also significant. The reasons cause this situation is from three aspects; firstly, the gender discrimination based on culture have rooted in China for thousands of years, the female is supposed to return home regardless of their education and qualification. Although the campaigns for the equality between gender prevailed since the 1950s the early days of China, it’s generally viewed as going backward in the 21stcentury said by Julie Broussard, the director of UN Women. And, in some high trading areas like Zhejiang and Guangdong province, people has a serious preferences for sons than daughters, because they want boys to inherit their properties especially when they are in the good economic condition. As a result, women are in the inferior social status and those who have the same educational level in the workplaces earn lower salaries than their male counterparts. Secondly, on the rapid process of economic transformation, various areas have experienced vast adjustments of their industrial structures, they get rid of some industries like textile which center on female workers or replace them with machines; this results in the imbalance of supply and demand in the labor market and deepen the discrimination of the female. Finally, due to the inequality development among regions, the concentration of the high-level talented people in the economically developed areas lead to the high discrimination among genders, in the developing zones, people crowd into labor-intensive industries which preferred male workers.
The Chinese government should accelerate the pace of industrial restructuring and bring them into a steady rhythm to solve the existing problems. At the same time, the administration should strengthen the supervision of gender discrimination in the labor market to create a fair employment environment. Putting more effort into eliminating the economic inequality between areas and promoting employment for low-skilled female workers in the commercial developing region. Finally, from a cultural perspective, it is essential to change the traditional social division of labor and form a broad understanding of the contribution of women workers during the process of social and economic development in the whole society.
Reference:
1．Airola J., & C.John, 2008, Wage inequality in Post-reform Mexico, Journal of Income Distribution, 17(1):110-135.
2．Becker GS., 1957, The Economics of Discrimination, Chicago: University of Chicago Press.
3．Branko M., & Lyn S., 2007, Does Tariff Liberalization Increase Wage Inequality? Some Empirical Evidence, NBER Chapters, in Globalization and Poverty, National Bureau of Economic Research, Inc., pages 143-182.
4．Chen Y., Deamurger S, & Fournier M., 2004, Earning differentials and Ownership Structure in Chinese Enterprises, Economics Development and Cultural Change, 53(4), pages 933-985.
5．Cassels R., Vidyattama Y., & Miranti R., et al., 2009, The Impact of A Sustained Gender Wage Gap on the Australian Economics, Australian Journal of Labor Economics.
6．Cockburn J., 2008, Trade Growth and Gender in Developing Countries: A Comparison of Ghana, Honduras, Senegal and Uganda, Annual Conference on Global Economic Analysis.
7．Ding S, Dong X, & Li S, 2009, Women Employment and Family Income Inequality During China’s Economics Transition, Feminist Economics, 15(3), pages 163-190.
8．Elissa B., & Mark B.,2007, Foreign direct investment and gender wages in urban China, Feminist Economics, vol. 13, issue 3-4, pages 213-237.
9．Gunseli B., Yana R..& Joseph E Z., 2004, International Trade and Gender Wage Discrimination: Evidence from East Asia, Review of Development Economics, vol. 8, issue 2, pages 237-254.
10．Gibb S.J., Fergusson D.M., & Horwood L.J., 2009, Sources of the Gender Wage Gap in a New Zealand Birth Cohort, Australian Journal of Labor Economics.
11．Goldberg P., & N. Pavcnik, 2007, Distributional Effects of Globalization in Developing Countries, 2007, 45 (1) :39-82.
12．Hanson G., & A.Harriaon, 1999, Trade Liberalization and Wage Inequality in Mexico, Industrial Labor Relations Review, 52(2):217-288.
of Liberalization? American Economic Review, 103 (5): 1960-76.
14．Laura H., & Sandra P., 2010, Market Access and Individual Wages: Evidence from China The Review of Economics and Statistics, vol. 92, issue 1, pages 145-159
15．Lei L., Bin L., Bo H., & Lu X., 2012, The Impacts of Trade Openness on Urban Residence Income Distribution and the Gender Wage Gap, China Economic Quarterly.
16．Melly B., 2005, Decomposition of differences in distribution using quantile regression, Labour Economics, vol. 12, issue 4, pages 577-590.
17．Menon, N., & RodgersY., 2009, International Trade and the Gender Wage Gap: New Evidence from India's Manufacturing Sector, World Development, Elsevier, vol. 37(5), pages 965-981.
18．Oaxaca R., 1973, Male-Female Wage Differentials in Urban Labor Markets, International Economics Reviews, vol 14, issue 3.
19．Oostendorp R H., 2009, Globalization and the Gender Wage Gap, World Bank Economics Review, vol. 23, issue 1, pages 141-161.
20．Robert C F., Alan H., Marcel P T., & Haiyan D., 2009, Estimating Real Production and Expenditures across Nations: A Proposal for Improving the Penn World Tables, The Review of Economics and Statistics, MIT Press, vol. 91(1), pages 201-212.
21．Seguin & Stephanie, 2000, Gender Inequality and Economic Growth: A Cross-Country Analysis, World Development, Elsevier, vol. 28(7), pages 1211-1230.
22．S E. Black & E Brainerd., 2004, Importing equality? The impact of globalization on gender discrimination, Industrial and Labor Relations Review, ILR Review, Cornell University, ILR School, vol. 57(4), pages 540-559.
23．Saure P, & Zoabi H,2009, Effects of Trade on Female Labor Force Participation, Swiss National Bank Working Paper, Page 1-58.
24．Saure P., & Zoabi H., 2014, International trade, the gender wage gap and female labor force participation, Journal of Development Economics, vol. 111(C), pages 17-33.
25．Topalova P., 2007, Trade Liberalization, Poverty and Inequality: Evidence from Indian Districts, NBER Chapters, in Globalization and Poverty, National Bureau of Economic Research, Inc., pages 291-336.
26．Yanle X., Xiangwen Z., & Liang C., 2013, Trade liberalization and China's gender wage gap -- an empirical study based on instrumental variables and quantile regression methods, Journal of Shanxi University, 2013(7).
