¹ A. K. Sen is the most influential advocate of this concept of development. For the most up–to–date account of Sens theoretical framework and its application to development issues, see Dreze, J. and Sen, A. K., Hunger and Public Action (Oxford: Oxford University Press, 1989).Google Scholar In a recent article. Sen has expressed concern over the impact of economic reforms on health in China. The article has provoked a debate: see Peter Nolan and John Sender, Death rate, life expectancy and Chinas economic reforms: a critique of Sen, A. K., World Development, Vol. 20 (1992), pp. 12797–1303 and A. K. Sen, Life and death in China: a reply, World Development, Vol. 20 (1992), pp. 1305–1Google Scholar

² In recent years there have been numerous articles in Chinese journals (e.g. Jiaoyuyanjiu) and newspapers (e.g. Jiaoyu bao) discussing such problems as the lack of funding, falling enrolment, increasing drop–out rates and child labour. In the English literature on the subject, see B0rge Bakken, Backwards reform in Chinese education, The Australian Journal of Chinese Affairs, Vol. 19/20 (1988), pp. 127–163; Suzanne Pepper, Chinese education after Mao: two steps forward, two steps backward and begin again, The China Quarterly, No. 81 (1980), pp. 1–65; Suzanne Pepper, Chinas Education Reform in the 1980s (China Research Monograph, Institute of East Asian Studies, University of California at Berkeley, Centre for Chinese Studies, 1990); Suzanne Pepper, Regaining the initiative for educational reform and development, in Lo chi–kin, Suzanne Pepper and Tsui Kai–yuen (eds.), China Review 1995 (Hong Kong: The Chinese University Press, 1995); D. Davis, Chinese social welfare: policies and outcomes, The China Quarterly, No. 119 (1989), pp. 577–597; World Bank, China: Lang–term Development Issues and Options (Baltimore: The John Hopkins University Press, 1985); World Bank, China: Strategies for Reducing Poverty in the 1990s (Washington D.C.: World Bank, 1992); Lo Leslie Nai–kwai, The changing educational system: dilemma of disparity, in Joseph Cheng and M. Brosseau (eds.), China Review 1993 (Hong Kong: Chinese University Press, 1994). There are studies questioning the seriousness of these problems: see Han Qinglin, Woguo zhongxiaoxue xuesheng liushi zhuangkuang de fenxi yu duice (Analysis of and response to students dropping out of primary and secondary schools in China), Jiaoyu yanjiu (Education Research), 1993. A not too negative assessment is that of Peter Nolan, The Political Economy of Collective Farms (Cambridge: Polity Press, 1988).Google Scholar

³ See e.g. Mankiw, N. G., Romer, D. and Weil, D. N.. A contribution to the empirics of economic growth, Quarterly Journal of Economics, May 1992, pp. 407–437; a recent Work Bank publication has shown that the contribution of education to the economic growth of the East Asian newly industrialized nations is significant, see, World Bank, The East Asian Miracle: Economic growth and public ploicy (New York: Oxford University press for the world bank, 1993).CrossRefGoogle Scholar

⁴ The proxies of educational attainment reviewed in this article are by no means perfect. I It is very difficult, if not impossible, to design a measure of educational attainment which can command universal approval. The indicators employed here are chosen on pragmatic grounds. Similar proxies are commonly used in many previous studies and can help to answer some of the questions raised. For example, the length of schooling proxies may to a certain extent help us find out whether parents shorten the length of schooling of their children as a result of economic reform.

⁵ Studies on the rates of return of schooling suggest that the marginal benefit of education k declines with the years of schooling. For a survey, see e.g. Paul Schultz, T., Education investments and returns, in H. Chenery and T. N. Srinivasan (eds.), Handbook of Development Economics (Amsterdam: Elsevier Science Publishers B. V., 1988), pp. 543–630.Google Scholar

⁶ Before 1992, the official net enrolment rate is defined as the share of children between and 11 enrolled in school. After 1992, no uniform definition for school–age children is adopted. The definition varies across provinces, see, e.g., State Education Commission (ed.), Zhongguo jiaoyu shiye tongji nianjian (Beijing: Renmin jiaoyu chubanshe, 1993), p. 17.

⁷ For an authoritative discussion of rural–urban classification, see Kam–wing, Chan, Urbanization and rural–urban migration in China since 1982, Modern China, Vol. 20 (1994), pp. 127–163.Google Scholar

⁸ See Kai–yuen, Tsui, Decomposition of Chinas regional inequalities, Journal of Comparative Economics, Vol. 15 (1993), pp. 600–627.Google Scholar

⁹ See Lo Nai–kwai, The changing educational system.

¹⁰ For details of these surveys, see State Statistical Bureau, Department of Social Statistics, Zhongguo ertong tongji ziliao (Statistics of Children in China) (Beijing: Zhongguo tongji chubanshe, 1989), State Statistical Bureau, Office of Survey on the Situations of Children in China and Department of Social Statistics, Zhongguo 1987 nian ertong qingkuang diaocha (Statistics on the 1987 Survey of the Situations of Children in China) (Beijing: Zhongguo tongji chubanshe, 1990) and State Statistical Bureau, 1992 nian Zhongguo ertong qingkuang chouyang diaocha: Guojia zuizhong baogao (1992 National Sample Survey on the Situation of Children: National Final Report) (Beijing: State Statistical Bureau, 1993).Google Scholar

¹¹ The length of schooling is represented by two proxies: the share of literate population without completing primary education and the share of literate population going on to high school. For details, see below.

¹² For a detailed discussion of the restructuring in the early years of the reform era, see e.g. Pepper, Chinese education after Mao.

¹³ State Council, A resolution on problems pertaining to universal education.

¹⁴ For reasons behind the fiscal decline see e.g. Wong, Christine, Christopher Heady and Wing Woo, Zhongguojingji gaige yu caizheng guanli (Chinese Economic Reform and Fiscal Management) (Beijing: Zhongguo caizheng jingji chubanshe, 1993).Google Scholar

¹⁵ As stated in Guanyu jiaoyu tizhi gaige de jueding (Resolution on the reform of education system) issued by the Central Committee of the Communist Party in 1986, the responsibility of basic education is handed over to local governments. Another document issued by the Ministry of Finance in 1987 further clarified the role of local government in the provision of basic education in rural areas. See Guanyu nongcun jichu jiaoyu guanli tizhi gaige ruogan wenti de yijian (On the proposals with respect to a number of questions pertaining to the reform of rural education system), 15 June 1987. The principle ofdifangfuze fenji guanli is also enshrined in Zhonghua renmin gongheguo yiwu jiaoyufa (Law on Compulsory Education in the Peoples Republic of China).

¹⁶ For a detailed account of the restructuring of basic education, see Pepper, Chinese education after Mao, and Chinas Education Reform in the 1980s

¹⁷ I put quotation marks around the word negative because it is controversial whether such an effect reduces the welfare of rural households. From the perspective of some economists, economic reforms opens up more opportunities and the family as a unit can better allocate its resources to maximize the familys well–being. Of course, this is a rather naive view because the family is viewed as a homogenous whole. A more sophisticated view of the family may lead us to very different conclusions; see e.g. Sen, A. K., Economics and the family, Asian Development Review, No. 1 (1983), pp. 14–26 for an interesting discussion.Google Scholar

¹⁸ Davis, Chinese social welfare, p. 481Google Scholar

¹⁹ Ibid n. 9.

²⁰ Xizhi, Peng, Demographic Transition in China: Fertility Trends since the 1950s (Oxford: Clarendon Press, 1991), p. 112.Google Scholar

²¹ Using the same set of data, Pepper in Chinas Education Reform in the 1980s concludes that the decline was induced by the reversal of the educational strategy away from mass enrolment.

²² Yingjie, Liu (ed.), Zhongguo jiaoyu dashidian 1949–1990 (Chronology of Major Education Events in China) (Zhejiang: Zhejiang jiaoyu chubanshe, 1993).Google Scholar

²³ State Statistical Bureau, Department of Social Statistics, Zhongguo ertong tongji ziliao, and Office of Survey on the Situations of Children in China and Department of Social Statistics, Zhongguo 1987 nian ertong qingkuang diaocha.

²⁴ The surveys used different definitions of school–age children. In the 1983 survey it includes children between 7 and 14. The 1987 survey changed the definition to includechildren between 6 and 14. There are pros and cons with using these figures. They are better than the official net enrolment rate which records students enrolled at the beginning of the school year. However, it is not clear whether figures from sample surveys do reflect the national situation. In particular, the 1987 sample survey includes figures from only nine provinces: Neimenggu, Heilongjiang, Zhejiang, Shandong, Hubei, Guangdong, Sichuan, Yunnan and Ningxia.

²⁵ See Office of the Population Census of the State Council and Population Statistics Branch of the State Statistical Bureau, Zhongguo 1990 nian renkoupucha ziliao. The illiteracy and semi–illiteracy rates used in this paper are from the 1990 census; see State Statistical Bureau, Population Census Office, Zhongguo disici renkou pucha de zhuyao shuju (Major Figures of the Fourth National Population Census of China) (Beijing: Zhongguo tongji chubanshe, 1993). For those aged 15 and above, the official definition of illiterate and semi–illiterate population includes peasants recognizing fewer than 1,500 Han characters, and workers in enterprises and administrative units, and urban residents recognizing fewer than 2,000 Han characters. If there are doubts the interviewer was supposed to ask the interviewees to read passages from a booklet given to them beforehand. In practice, the interviewer determined whether interviewees were illiterate or semi–illiterate by their reported years of schooling. I thank Zuo Xujin and Jean Hung for providing me with this information. While there are complaints on such aspects of the 1990 census as infant mortality rates, to the best of my knowledge, no major problem of figures on education levels has been reported. In a recent paper entitled Zhongguo 1990 nian renkou pucha shuju zhiliang de pinggu in State Council Office of Population Census and State Statistical Bureau, Department of Population Statistics (eds., Zhongguo 1990 nian renkou pucha–guoji taolunhui lunwenji {Proceedings of the International Conference on the 1990 Chinese Population Census) (Beijing: Zhongguo tongji chubanshe, 1993), Zhang Weimin and Cui Hongyan compare the 1990 data on educational levels with those of the 1982 census. They discover that the two sets are broadly consistent, suggesting that the quality of the data on educational levels is acceptable. There does not seem to be any strong incentive for the interviewees to over or under–state their educational levels. In any case, if the bias of the data is consistent upward or downward for all age cohorts, the corresponding trends will probably not be affected.Google Scholar

²⁶ For example, if the age of the cohort is 15, then the year of birth is 1975.Google Scholar

²⁷ See also Lavely, William, Zhenyu, Xiao, Bohua, Li and Freedman, R., The rise in female education in China: national and regional patterns, The China Quarterly, Vol. 121 (1990), pp. 61–93. Using data from the 1982 fertility survey, they also discovered that the 1959–61 famine disrupted the declining trend of the female illiteracy rate.CrossRefGoogle Scholar

²⁸ Literate population is equal to the total population minus the illiterate and semiilliterate population as defined by the 1990 census.

²⁹ The definition of literate population may be found in Disici quanguo renkou pucha banfa issued by the state council, details of which may be found in Office of Population Census of the State Council and Population Statistics Branch of the State Statistical Bureau, Zhongguo 1990 nian renkou pucha ziliao, Vol. 4, pp. 514–15. The literate population with primary education as the maximum level of education is divided into the following categories: zaixiao: those receiving education in school and are registered; biye: those who have completed the last grade of primary school and have passed the examination; yiye: those who have not completed the last grade of primary education (for whatever reason) or completed the last grade but have failed; qita: those working and receiving primary education at the same time, including those attaining primary education through private tuition.

³⁰ For details, see Pepper, Chinese education after Mao, and Lavely et al., The rise in female education in China.

³¹ See e.g. State Statistical Bureau, Department of Social Statistics. Zhongguo shehui tongji ziliao (Chinese Social Statistics) Beijing: Zhongguo tongji chubanshe, 1993Google Scholar

³² See Chan Kam Wing, Urbanization and rural–urban migration in China.

³³ Lavely et al., The rise in female education in China.

³⁴ Care must be exercised in defining the rural as opposed to the urban designation. In the 1980s, many counties gained urban status so as to secure the many benefits associated with it. However, many of these counties are in fact largely rural. To rectify this problem, the 1990 census provides two definitions for the rural–urban demarcation: see State Statistical Bureau, Guanyu 1990 nian renkuo pucha zhuyu shu–ju di gongbao (Communique on major statistics of the 1990 population census), No. 5, 18 December 1990. Under the second definition, city (shi) population is composed of population in all the districts (qu) under cities with districts and population of streets (jiedao renkou) in cities without districts. Population in towns (Then) includes population of resident committees (jumin weiyuanhui) in those towns under cities without districts and population of resident committees of towns administered by counties. Chan Kam Wing in Urbanization and rural–urban migration in China since 1982 explains the rationale behind such a classification. Unless otherwise stated, the rural–urban classification used here is that of the second definition created by the 1990 census.

³⁵ In this section, the counties and cities are grouped by provinces. For a different regional classification see The rise in female education in China in which Lavely et al. borrowed the macroregional model of G. William Skinner and studied regional variation in female education.

³⁶ Kai–yuen, Tsui, Chinas regional inequality, 1952–1985, Journal of Comparative Economics, Vol. 15, pp. 12.Google Scholar

³⁷ See Tsui Kai–yuen, Economic reform and interprovincial inequalities in China, Journal of Development Economics, Vol. 50 (1996), pp. 353–368.

³⁸ In an important notice Guowuyuan guanyu choucuo nongcun xuexiao banxue jingfei de tongzhi (Notice of the State Council on the funding of rural education), Guofa 1984, No. 174, the State Council laid down the guidelines for the village (xiang) governments to levy education surcharge. A subsequent document of the State Council entitled Guowuyuan zhengshou jiaoyufei fujia de zanxing guiding (Temporary regulations of the State Council on the levy of education surcharge), Guofa 1986, No. 50, explained in concrete terms how the education surcharge should be levied and managed. For various reasons, many localities have not started levying the surcharge. Mismanagement is also rampant.

³⁹ According to the first State Council notice in n. 38, the rate of the surcharge is 1% of the product tax, value added tax and business tax. The rate was increased to 2% in 1990 see State Council, Guowuyuan guanyu xiugai zhengshou jiaoyufei fujia de zanxing guiding (On the State Councils amendment to the temporary regulations on the levy of education surcharge ) Guofa, 7 June 1990, and 3% from 1 January 1994 onwards.

⁴⁰ See Editorial Committee on the Chinese Education Yearbook, nianjian, Zhongguo jiaoyu 1992 (Chinese Education Yearbook 1992) (Beijing: Renmin jiaoyu chubanshe, 1993), p. 79.Google Scholar

⁴¹ For example, waves of peasant uprisings in the first half of 1993 prompted the central government to issue directives ordering local authorities to reduce tax burdens on farmers. The immediate response of some local governments was to cut the education surcharge, thus seriously jeopardizing educational funding; see Tong Hua, Nongcun jiaoyu buying chuxian de jishache (Rural education: it is not right to brake), Zhongguo jiaoyu bao (Chinese Education Daily), 19 March 1994

⁴² There are numerous reports of funds earmarked for education being diverted to other uses. According to a recent article in the Zhongguo jiaoyu bao (Wu Fusheng, Qianghua yiwujiaoyu ying xian qianghua fazhi guannian (Strengthening the rule of law is the first step to the strengthening of compulsory education), 17 April 1994), a large share of the revenue from education surcharge is not actually used in education.

⁴³ In Dui dangqian woguo zhongxiao xuesheng liushi wenti de tantao (An exploration of the reasons behind primary and secondary school students dropping out of schools), Jiaoyu yanjiu. No. 3 (1989), pp. 49–53, Xie Huanli points out that drop–out rates of some counties in such relatively rich provinces as Zhejiang and Jiangsu were higher than the national average.Google Scholar

⁴⁴ The county and city illiteracy and semi–illiteracy rates are from State Statistical Bureau, Zhongguo renkou tongji nianjian 1992 (Chinese Population Statistics Yearbook) (Beijing: Zhongguo tongji chubanshe, 1993), pp. 296–427. The adult literacy rate is denned as 100 minus the illiteracy and semi–illiteracy rate.Google Scholar

⁴⁵ The two censuses have different definitions of adult population. In this paper, the adult population refers to those aged 15 or above. All the data have been adjusted to ensure comparability.

⁴⁶ For other desirable properties of Theils entropy measure,see e.g. Sen, A. K.., On Economic Inequality (Oxford: Clarendon Press, 1973).CrossRefGoogle Scholar

⁴⁷ The mean values presented in Table 6 are not the same as the official adult literacy rates because our rate for each unit is equal to the literate population of IS or above divided by the total population and not by the population of 15 or above. Figures for all the counties and cities using the official definition have not been readily available.

⁴⁸ One referee raises the question of whether interprovincial migration may explain the different trends in eastern and interior provinces. While restrictions on labour mobility have relaxed considerably in the 1980s, the scale of long–distance migration was probably still not significant enough to explain the divergent trends in the 1980s. To understand the relationship between interprovincial migration and intraprovincial inequalities in educational attainment, it is necessary to know who the migrants were in the 1980s and their education levels. However, such knowledge is very limited.

⁴⁹ Let us suppose that the naive view of the family is adopted. If a family reduces the schooling of its children because there are more opportunities to the family elsewhere, then the naive view implies that family welfare has increased. On the other hand, a reduction of schooling due to supply–side factors (e.g. a fall in the public funds used in education) may reduce a familys welfare.

⁵⁰ One interesting research direction is to find out whether educational attainment is constrained by supply–side or demand–side factors. One can specify a demand function and a supply function of education. The interesting question is to determine whether the observed educational attainment is on the demand function or the supply function. The market disequilibrium framework may be adapted to this context. The model may then be applied to local (e.g. county) data. For a survey of the econometrics of disequilibrium model see e.g. Maddala, G. S., Limited Dependent and Qualitative Variables in Econometrics (Cambridge: Cambridge University Press, 1983), ch. 10.CrossRefGoogle Scholar