Empirical analysis on 31 provinces' (municipalities') panel data from 2001 to 2008
Tongwei Xie, Research Institute of Finance and Economics, Shanghai University of Finance and Economics, Shanghai, China
Purpose – This article aims to analyze inter-provincial disparities of rural education and the convergence rate, and to discuss the effects of compulsory education reform after 2001.
Design/methodology/approach – The article estimates the rural average education years and education Gini coefficients of China's 31 provinces (municipalities) beside Hong Kong, Macao and Taiwan from 2001 to 2008, and applies the method of “economic convergence”.
Findings – The results show that after the reform of rural compulsory education, inter-provincial rural education disparities and educational equality have been significantly improved and trend to convergence; nevertheless the convergence rate on inter-provincial disparities of education equality declines. The defects of the education input system – “county as principle” – has been a factor that restricts the coordinated development of rural education.
Practical limitations – This paper suggests that it is necessary for China's provincial and central government to afford the expense of compulsory education. China's present investment system would also worsen inter-provincial inequities of education.
Social implications – Education equality is one of the basic social priorities. In China education equality has been improved; however it could be better if China's provincial and central government afforded the expenses of compulsory education.
Originality/value – This paper applies the method of “economic convergence” to analyze China's rural education disparities among its regions.
Average educational years; Educational Gini coefficients; Convergence; China; Education; Educational policy; Rural areas.
International Journal of Educational Management
Emerald Group Publishing Limited
Coordinated regional development of education is the reflection of educational justice and equality. Uncoordinated educational development among provinces in China has become one of the significant issues of regional development. In 2001, China's government carried out reform of the investment system for compulsory rural education. “County as principle” replaced “classification management”. This reform has promoted the development of compulsory rural education in China and has to some extent alleviated the problem of insufficient investment in education, and inter-provincial differences in rural education have also been significantly reduced. However, recent years' practice shows that the “county as principle” investment system still cannot solve the problem of rural education poverty. Under the conditions of farmers' low income and the “county as principle” investment system, a county government's finances become the most important determinant of education investment. In fact, most counties cannot ensure adequate investment in education. Counties' governments have assumed the burden of compulsory education inputs, while central government and the provincial government share little responsibility. This investment system discourages the further development of compulsory rural education. It is estimated that at present all of China's investment in compulsory education, the county governments' share is 78 percent, provincial governments' share no more than 1 percent, and central government's share much less. Therefore, inter-provincial disparities and inequality of education development are directly influenced by regional economic differences.
On the other hand, because of higher agricultural productivity and rapid economic development, East China is demanding a larger labor force, and the registered system of permanent residence of recent years is not so strict as previously. All of this results in a high level of rural labor mobility. The rural laborers who move to East China mainly come from Central and Western China, and they are better educated than those who stay in local regions. Additionally, in China, the investment system for compulsory rural education has obvious territorial characteristics. This means that the education investment for rural laborers who move to East China comes from the local governments of Central and Western China. As a result, investment and benefit in rural education in these regions are significantly asymmetrical. This phenomenon will eventually reduce local governments' incentives for investment in rural education, and could also enlarge inter-provincial education development disparities and regional inequities.
In order to analyze these changes in inter-provincial differences in rural education after reform of the education investment system and how the inter-provincial differences converge, after estimating the rural average education years and education Gini coefficients of 31 provinces (municipalities) beside Hong Kong, Macao and Taiwan from 2001 to 2008, this article in addition applies the method of “economic convergence” to analyze inter-provincial disparities of rural education and the convergence rate. The article also comments on the effects compulsory education reform – i.e. “county as principle” – after 2001.
2. Literature review
Since 1978 China's education has been significantly improved, but the issue of unbalanced education development also highlights that scholars are concerned about the issue. Wei and Yang (1997) found that regional differences in education development decreased year by year. Wang et al. (1998) and Wu (1999) pointed out that economic development disparities among regions result in an imbalance in education development. Zhu (2003) calculated the education Gini coefficient of inter-regional investment from 1998 to 2000, with the results showing that the inequality of education investment has increased in recent years. Tu's (2005) study showed that China's regional differences in education could be mainly seen from two aspects:
- the scale of regional of higher education development and educational investment; and
- China's region differences in education has been enlarged.
Du (2005) calculated the education Gini coefficient and found that inter-provincial education disparities had been reduced. Zhai (2006, 2007) also found that between 1995 and 2004 basic education disparities declined after his theoretical and empirical analysis. Yang and Li (2007) estimated the education Gini coefficient of China's 31 provinces from 1996 to 2004, and pointed out that China's education inequality had improved not only within regions, but also among regions. Sun (2008) also measured the education Gini coefficient in different periods, and found that China's educational development reduced education inequality within each region.
The above research by Chinese scholars show two opposite conclusions, namely that regional differences in education development have been either reduced or enlarged. According to the data used in this literature, research using data from after 2001 tends to see a decrease, while research using data from before 2001 tends to see an increase, because of the implementation of reform of the compulsory education system. However, there are few articles to apply the approach of “economic convergence” to analyze inter-provincial education disparities. Until now, only Sun (2008) has taken advantage of this method, and his study pointed out that China's inter-provincial education disparities had gradually been reduced. Therefore, this paper also utilizes the approach of “economic convergence” to analyze rural education disparities among China's provinces and the convergence speed after investment system reform of rural compulsory education in 2001.
3. Calculations on average education years and education Gini coefficient of China's rural areas
3.1 Methods of measurement
Education level is divided into five different qualifications (i.e. illiterate or semi-illiterate, primary, junior high school, senior middle school, college or above) and the education years are as follows:
- illiterate or semi-illiterate for 0 years;
- primary school for six years;
- junior high school for nine years;
- senior middle school for 12 years; and
- college or above for 16 years.
The mathematical formula that estimates the average years of education is defined as follows: Equation 1 In this formula, AEY represents average education years, i represents the number of different educational qualifications, EY i represents the education years of each education level, and P i represents the population percentage of each education level in total population.
The mathematical formula that estimates the education Gini coefficient is as follows: Equation 2 where EA i represents the educational achievement percentage of each group. TEA i represents the percentage of total educational achievements, and its formula is: Equation 3
3.2 Data sources
The data were taken from the China Statistical Yearbook and China Population Statistics Yearbook from 2002 to 2009. The statistic criteria was the population aged six years old and above, and according to the calculation formula, we obtain panel data of the average education years and education Gini coefficient from 2001 to 2008.
3.3 Estimation results
3.3.1 Average education years and education Gini coefficient of China's rural areas
As shown in Table I, China's rural average years of education increased from 6.752 in 2001 to 7.285 in 2008. With the development of China's rural education, China's rural education inequality has been greatly improved; the education Gini coefficient fell from 0.237 to 0.217. It is notable that in 2005 the level of rural education development declined, and education inequality increased.
3.3.2 Average education years and education Gini coefficient of provinces' (municipalities') rural areas
The rural average education years of every province (municipality) increased at different degrees, and rural education inequality was also obviously improved. However, plenty of provinces' (municipalities') education development fell back in 2002 and 2005. In 2005, 24 provinces' (municipalities') rural average education years declined: they were Tianjin, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Shanghai, Zhejiang, Anhui, Fujian, Jiangxi, Shandong, Henan, Hubei, Hunan, Guangxi, Hainan, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, and Ningxia. Provinces' (municipalities') education inequality also increased in both 2002 and 2005. In 2002, these provinces were: Beijing, Tianjin, Hebei, Inner Mongolia, Jilin, Heilongjiang, Jiangsu, Zhejiang, Anhui, Jiangxi, Hunan and Guangdong, Guizhou, Tibet, Shaanxi, Gansu, and in 2005 they were Beijing, Shanxi, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Shanghai, Fujian, Shandong and Henan, Hubei, Hunan and Guangdong, Guangxi, Hainan, Chongqing, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Qinghai, Ningxia, and Xinjiang.
4. Comparison of rural education development
In order to reflect the changes from 2001 to 2008 in educational disparities among provinces (municipalities), this paper shows four GIS charts of average education years and education Gini coefficient of provinces' (municipalities') rural areas.
4.1 Comparison of average education years
From 2001 to 2008, 31 provinces’ (municipalities’) rural education was significantly improved, and inter-provincial disparities in rural education were also reduced. In Figure 1, in 2001, Jiangxi is the only province whose average education years of education in rural areas is above eight years (blue), while Beijing, Tianjin, Shanxi, Jiangxi also achieved this in 2008. The rural average education years of Beijing, Tianjin, Hebei, Shanxi, Liaoning, Jilin, Heilongjiang, Henan, Hunan, Guangdong were between seven years and eight years (green) in 2001; this increased to 17 provinces in 2008 (Hebei, Inner Mongolia, Liaoning, Jilin, Heilongjiang, Shanghai, Jiangsu, Zhejiang, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, Hainan, Shaanxi, and Xinjiang). The average years of education in rural Inner Mongolia, Shanghai, Jiangsu, Zhejiang, Anhui, Fujian, Shandong, Hubei, Guangxi, Hainan, Chongqing, Sichuan, Shaanxi, Ningxia and Xinjiang was between six and seven years (yellow) in 2001. In 2008 this number dropped from 15 to eight provinces (Anhui, Fujian, Chongqing, Sichuan, Yunnan, Guizhou, Gansu, and Ningxia). The average years of education in rural Guizhou, Yunnan and Gansu was between five and six years (brown) in 2001; however in 2008 only the province of Qinghai saw this level. The rural average education years in Tibet and Qinghai was below five years (red) in 2001, while only Tibet showed this level in 2008.
4.2 Comparison of education Gini coefficient
With the development of rural education, from 2001 to 2008, inequality in rural education in 31 provinces (municipalities) was also greatly improved. In Figure 2, in 2001 the provinces (municipalities) with a rural education Gini coefficient below 0.20 (blue) were Hebei, Shanxi, Inner Mongolia, Jilin, Heilongjiang, Shanghai; in 2008 they were Hebei, Shanxi, Inner Mongolia, Jilin, Heilongjiang, Shanghai, Hubei, Guangdong, Guangxi, and Hainan. Provinces (municipalities) with a rural education Gini coefficient between 0.20 and 0.30 (green) in 2001 were Beijing, Tianjin, Liaoning, Zhejiang, Anhui, Fujian, Jiangxi, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, Hainan, Chongqing, Sichuan, Guizhou, Gansu, and Tibet; in 2008 they were Beijing, Tianjin, Liaoning, Jiangsu, Zhejiang, Anhui, Fujian, Jiangxi, Shandong, Henan, Hunan, Chongqing, Sichuan, Guizhou, Yunnan, Gansu, Tibet, Ningxia and Xinjiang. Provinces (municipalities) with a rural education Gini coefficient between 0.30 and 0.40 (yellow) in 2001 were Jiangsu, Yunnan, Qinghai and Xinjiang, while only Qinghai had this level in 2008. Provinces (municipalities) with a rural education Gini coefficient above 0.40 (red) in 2001 were Shaanxi and Ningxia, while only Shaanxi showed this level in 2008. According to Figure 2, rural education equity in Northeast China and North China is better than in other regions, and in recent years Ningxia is the province that has seen the greatest improvement in education fairness.
5. Convergence of rural education disparity among provinces
5.1 σ convergence of rural education disparities
According to the definition of σ convergence, the changes of dispersion are the foundation to judge whether rural education disparity among provinces is converging. Here we use the variation coefficients of the average education years and the education Gini coefficient. The means and variations of the average education years and education Gini coefficient (Figures 3 and 4) show that after the reform of the investment system for rural compulsory education in 2001, China's inter-provincial education development disparities and inequalities shrank year by year (except for 2005), which shows significant σ convergence.
5.2 β convergence of inter-provincial rural education disparities
For the β convergence analysis, this paper uses Barro and Sala-i-Martin's (2005) model. The β convergence model is as follows: Equation 4 Using the panel data of the average education years and education Gini coefficient from 2001 to 2008, divided into three periods (2001-2004, 2005-2008 and 2001-2008), the convergence speeds derived are shown in Tables II and III.
In all three periods – i.e. 2001-2004, 2005-2008 and 2001-2008 – the inter-provincial disparities of education development showed convergence. The convergence rates were 0.071 percent, 0.099 percent, and 0.042 percent, respectively. The inter-provincial disparities of rural education inequality also show convergence, with convergence rates of 0.152 percent, 0.061 percent, and 0.107 percent, respectively. This is mainly because of the reform of the investment system for compulsory education after 2001, which increased investment in rural education, and rural education has been improved significantly. In 1985, China began to implement the reform of education finance and the hierarchical management system. In this system, the burden of compulsory education structure is irrational. Central and provincial government share little of the burden, but township government and farmers have to shoulder the burden of most of the investment in compulsory education. As the township government and farmers do not have sufficient finance, rural education has faced a serious shortage of funds. Since 2001, in order to address the input problem of rural compulsory education, the county government became the main investor in compulsory rural education. The convergence rates show that the reform effect of the financial system for rural compulsory education is obvious.
Further, we contrast the period 2001-2004 with 2005-2008: it is clear that the convergence rate of education development disparities in 2005-2008 is 30.77 percent faster than in the period 2001-2004, but the convergence rate of educational inequality disparities in the period 2001-2004 is 1.49 times faster than in the period 2005-2008. This difference indicates that the improvement effect of inter-provincial inequality disparities in education declined. To some extent, the reform of the rural compulsory education system after 2001 accelerated the development of education and improvements in education equality. However, this is still not enough. Provincial and central government invest relatively little in education. But county governments can afford adequate funds for the development of rural education. Therefore, rural education investment is decided by local government revenue, which is directly subject to local economic development. Also, economic development disparities among regions significantly affected inter-provincial differences in rural educational inequality. In Central and Western China, local governments could not provide adequate investment in rural education because the economy there is less developed than in Eastern China. Once the economic development disparities are decreased, the convergence rate of inter-provincial differences in education equality will also decrease.
Moreover, currently, the number of China's mobile rural laborers is great. According to Chinese government statistics, the total has reached 230 million, of which over 150 million have been out of work more than six months. Those rural laborers are better educated than others who stay in their local areas. They are mainly from Central and Western China and move to Eastern China. Additionally, China's education investment system has obvious territorial characteristics, which means that the education input of those people comes from local governments in Central and Western China. However, Eastern China could take advantage of them without investment in their education. Thus, this will discourage local government investment in compulsory education; furthermore, it will worsen education inequality among provinces.
After the investment system for rural compulsory education was reformed in 2001, rural education developed significantly, and inequalities in rural education have been greatly reduced. Inter-provincial disparities of the education development and educational inequality show significant convergence. However, the convergence rates of inter-provincial disparities on education inequality have decreased. This is because the new investment system for rural compulsory education still cannot provide adequate funds, especially in Central and Western China. In addition, this investment system has obvious territorial characteristics; local government in Central and Western China cannot receive symmetrical returns because of large-scale rural labor mobility. This phenomenon not only reduces local governments’ incentives to invest in rural education, it also worsens inter-provincial inequalities in education. Therefore, China's education input system should be regulated again. In fact, in developed countries, most of the investment in primary education is provided by central government. It is necessary for China's provincial and central government to meet the expenses of compulsory education.
Figure 1Thirty-one provinces' (municipalities') rural average education years in 2001 (left) and 2008 (right)
Figure 2Thirty-one provinces' (municipalities') rural education Gini coefficients in 2001 (left) and 2008 (right)
Figure 3Means and variations of rural average education years among provinces from 2001-2008
Figure 4Means and variations of rural education Gini coefficient among provinces from 2001-2008
Table IAverage education years and education Gini coefficient of China's rural areas, 2001-2008
Table IIConvergence speeds of education development disparities
Table IIIConvergence speeds of education inequality disparities
Barro, R.J., Sala-i-Martin, X. (1995), Economic Growth, McGraw-Hill, New York, NY, .
Du, P. (2005), "Study on education differences among China's schools based on Gini coefficient", Education and the Economy, Vol. 3 pp.30-4.
Sun, B. (2008), "Convergence test on China's education development in the past thirty years after reform and opening policy", Journal of Education, Vol. 12 pp.14-18.
Tu, D. (2005), "Educational disparities research", Education Science, Vol. 2 pp.8-11.
Wang, S., Du, Y., Liu, Y. (1998), "Analysis on imbalance in educational development in China", Education Research, Vol. 6 pp.19-23.
Wei, H., Yang, D. (1997), "Decentralization and regional educational differences", China Social Sciences, Vol. 1 pp.98-112.
Wu, D. (1999), "Development of regional gaps in China's education – imbalance problems of education development", Education Research, Vol. 7 pp.22-6.
Yang, J., Li, X. (2007), "Education inequality, human capital accumulation and economic growth: an empirical study on China", Quantitative Economics & Technical Economics Research, Vol. 2 pp.37-45.
Zhai, B. (2006), "Balanced development of education: theory, indicators and calculation method", Education Research, Vol. 3 pp.16-28.
Zhai, B. (2007), "Empirical analysis on balanced development of basic education in China", Education Research, Vol. 7 pp.22-30.
Zhu, M. (2003), "An empirical study on education investment fair among China's provinces", Economic Issues, Vol. 2 pp.121-4.
Barro, R.J., Lee, J.W. (1993), "International comparisons of educational attainment", Journal of Monetary Economics, Vol. 32 No.3, pp.364-94.
Barro, R., Lee, J. (1993), "International measures of schooling years and schooling quality", American Economic Review, Vol. 86 No.2, pp.218-23.
Bernard, A.B., Durlauf, S.N. (1996), "Interpreting tests of the convergence hypothesis", Journal of Econometrics, Vol. 71 pp.161-73.
Kyriacou, G. (1991), "Level and growth effects of human capital: a cross-country study of the convergence hypothesis", Starr Center for Applied Economics, New York University, New York, NY, Research Report 91-26, .
Lucas, R. Jr (1988), "On the mechanics of economic development", Journal of Development Economics, Vol. 22 No.1, pp.3-42.
Mankiw, N., Romer, D., Weil, D. (1992), "A contribution to the empirics of economic growth in the neoclassical model", Quarterly Journal of Economics, Vol. 107 pp.407-37.
Thomas, V., Wang, Y., Fan, X. (2000), Measuring Education Inequality: Gini Coefficients of Education, The World Bank Institute, Washington, DC, .
Thomas, V., Wang, Y., Fan, X. (2003), "Measuring education inequality: Gini coefficients of education for 140 countries, 1960-2000", Journal of Educational Planning and Administration, Vol. XVII No.1, .
Zhai, B. (2008), Educational Equilibrium Theory – Analysis on Balanced Development of Basic Education in China, People's Education Press, Beijing, pp.46-50.
Tongwei Xie can be contacted at: email@example.com