COSTS OF EDUCATION IN CHINA: ISSUES OF RESOURCE MOBILIZATION, EQUALITY, EQUITY AND EFFICIENCY By: Tsang, Mun C., Education Economics, 09645292, 1994, Vol. 2, Issue 3 Database: Business Source Premier ABSTRACT This paper is a survey of the costs of education in one Asian country, China. It focuses on three areas of education costs: (1) national expenditures on education; (2) unit costs of education; and (3) educational cost functions. It relates the analysis of education costs to four enduring policy issues in education in China: resource mobilization, inequality, inequity and inefficiency. The analysis is based on primary and secondary data sources and on existing studies in both the Chinese and English literatures. Whenever appropriate, comparative discussion is made by drawing on cost studies on other (mostly Asian) countries. The paper is aimed at providing a review of the current state of knowledge, and identifying knowledge gaps and areas for future research on education costs in China. Methodology and the Chinese Context Many decisions in education are concerned with the costs of education. Educational cost analysis can reveal the resource implications of an education policy, assess the financial feasibility of an education reform, provide diagnosis of past and current resource utilization in education, project future resource requirements for education, and evaluate the relative efficiency, inequity and inequality effects of alternative educational interventions. Studies of education costs can contribute to improved decision-making, planning, and monitoring in education (Tsang, 1988). The costs of education refer to the economic value of the inputs used in education. The cost or economic value of an education input is defined as its opportunity cost and is measured by the value of the input in its best alternative use. This definition of cost implies that the costs of education consist not only of public educational expenditures on

1

personnel, school facilities, supplies and equipment, but also parental and/or students' expenditures on education (direct private costs of education (DPCs) such as tuition and other education-related fees and expenditures on textbooks, uniform, school bag, writing supplies, transportation, boarding, etc.), students' foregone opportunities (indirect private costs, such as foregone earnings), as well as private contributions to education (contributions in cash or in kind, by individuals, parents or private organizations). In China, public expenditures on education are found in three accounts: operating expenditure on education (OEE) by the education bureaucracy (previously the Ministry of Education (MOE), and now the State Education Commission (SedC) and its provincial and local affiliates), capital expenditure on education (CEE) by the education bureaucracy, and education expenditure (mostly on vocational/technical education) by other central ministries (EEOCM). Data on OEE and CEE are available since 1950, but a continuous series for EEOCM is available only after 1980. Since the decentralization of education finance in the 1980s, public expenditures on education have been made by governments at the central, provincial and local levels. Thus, data on national expenditures on education since 1986 have to be compiled from an aggregation of data from various administrative levels. Tuition is free in basic education (5-6years of schooling before 1985, and 9 years of schooling after 1985); but parents still have to pay for other school fees (e.g. for sports and other school services) and other direct private costs of schooling. For a long period of time, cost recovery in post-basic levels was very limited. More attention to increased cost recovery in upper-secondary and higher-education levels has been made in recent years. Since 1987, the State Statistics Bureau has published periodic surveys of household expenditures in urban areas, but data on household education expenditure are incomplete. A comprehensive coverage of DPCs has yet to be implemented in such surveys. Parental and community contributions (known as `social contributions' in China) have long been a significant source of the financing of school construction, especially at the primary level. But no systematic effort was made to collect and report information on social contributions before 1986.

2

In addition to fiscal decentralization, the financing reform in education in 1980s also promoted the diversification of resource mobilization for education (Tsang, 1993), especially from non-government sources. Education institutions at various levels are encouraged to generate their own resources (e.g. revenue from production and sale of products by school), intensify their effort to gather social contributions and contributions from overseas Chinese, and collect fees from students. At the local level, education surcharges and levies have been imposed in the second half of the 1980s to support school expenditures. Currently, `self-generated' school revenue is used to purchase school equipment, provide welfare benefits for school staff and for repair or replacement of school buildings; social contributions are primarily used for school construction; school fees are used to support non-personnel school expenditures; and education surcharges and levies are used on non-personnel inputs, the repair or replacement of school building and on school equipment. Thus, in addition to government expenditures on education (known as budgeted education expenditures, including OEE, CEE and EEOCM) supported through the government's budget, there are extra-budgetary education expenditures (EBEE, sum of self-generated school revenue, social contributions, school fees, and education surcharges and levies) that are outside the government's budget. Since 1986, the government has reported both budgeted and extra-budgetary expenditures on education. In the Chinese context, one can obtain an estimate of total non-government expenditure on education by adding total non-fee DPC (i.e. total DPC minus tuition and other school fees) to EBEE. This paper attempts to relate the analysis of education costs to four prominent and enduring issues in education in China. The first issue is concerned with resource mobilization for education. For a long period of time, China has devoted a relatively low proportion of its fiscal and national resources to education. The limited education resources are reflected in low teacher salaries, poor physical conditions of schools, lack of equipment, etc. (Tsang, 1993). Since the promulgation of modernization policies in 1978, education has been regarded by the government as the strategic foundation for economic-oriented national development. Mobilizing additional resources for education is a key challenge for decisionmakers at all administrative levels. In the transition from a centralized planned economy to a `socialist-market economy', non-government

3

involvement in economic and educational production is encouraged. The government has tried to elevate the role of non-government costs in educational development in China. Thus resource mobilization for education is concerned not only with the fiscal effort of the government over time, but also with the mix of government and nongovernment resources for education. The other three issues concern inequalities, inequities and inefficiency in education. In the past four decades, national development in China had been subject to divergent ideologies and contrasting views within the leadership of the Chinese Communist Party (CCP); Chinese society was marked by large-scale tumultuous social changes. Education, being part of the state dominated by the CCP, was subject to the tensions within the CCP leadership. Thus educational developments have been highly politicized (Tsang, 1991). One view of educational development emphasizes ideological inculcation in communist ideal ('redness'), social equality and equity, and development through mobilization of the masses. Another view emphasizes a technical approach to development, efficiency and the importance of skills in economic production ('expertise', or social efficiency). Educational developments in China have oscillated between redness and expertise, and between the concerns for equality and efficiency. The expertise view has been the dominant one since 1978. Fluctuations in education costs have to be understood in the broader context of social change outside the education sector. Because of the limitation in length, the scope for this paper is restricted to the analysis of education costs in three areas: national expenditures on education, unit costs of education, and education cost functions. Some types of education costs and their analysis (such as capital versus recurrent costs, personnel costs versus non-personnel costs, education costs by administrative levels, allocation of education expenditures by level of schooling, etc.) have been omitted. Also, because of data limitations, most of the analyses concern the recent 5-year plan periods (1981-1991). The paper is partly based on previous cost studies on China conducted by the author and by other researchers, but it also contains new analyses using primary and secondary data sources. To understand the Chinese findings in international

4

perspectives, the paper provides a comparative analysis by drawing upon cost studies on other especially Asian, countries. The rest of the paper is divided into four sections. The second, third and fourth sections deal, respectively, with national expenditures on education, unit expenditures on education, and education cost functions. The last section provides a brief summary and discussion of areas for further research on costs of education in China. National Expenditures on Education National expenditures on education consist of expenditures from both government and non-government sources on the country as a whole. This section is focused on three analytical tasks: it (1) examines the trends in government and non-government expenditures over time; (2) estimates the total national expenditure on education in recent years; and (3) identifies the dilemma facing education decision-makers in the mobilization of non-government resources for education. Information for these analyses is based mostly on data from government sources. Because of the relatively long historical period covered (which increases the likelihood of inconsistency and other problems of data quality), the expenditure estimates given here are meant to indicate trends and order of magnitude. Resource Mobilization Table 1 gives public expenditures on education by the education bureaucracy for the period 1950-1991. Total public expenditure by the education bureaucracy (i.e. OEE plus CEE) increased from 0.41 billion yuan in 1950 to 41.83 billion in 1991 in current prices (see column (3) of Table 1), or an increase of 101 times! In 1950 prices, the 1991 figure was still 35 times that of 1950. The annual growth rate averaged 11.9% in nominal terms and 9.1% in real terms. As expected, OEE constituted the bulk of the total public expenditure by the education bureaucracy (averaging 89% during the whole period). Government allocation to CEE has historically, been low. Between 1950 and 1983, CEE averaged only 2.0% of total public capital expenditure (computed from State Education Commission 1984, p. 372).

5

Total public expenditure by the education bureaucracy averaged only 7.8% of total government expenditure (a measure of `fiscal effort', see column (5)) and 2.4% of national income (a measure of `national effort', see column (6)) during 1950-1991. There were also significant variations in these two measures across the past seven plan periods (see Figure 1). These fluctuations are correlated with changes in national income in the four decades (not shown in Table 1). In particular, the obvious decline in both measures occurred during the Third and Fourth Plan Periods (1966-1970 and 1971-1975), the turbulent years of the Culture Revolution. But there was a marked increase in the two measures after the promulgation of modernization policies in 1978. For example, the fiscal-effort measure averaged 6.5% during 1950-1979 and 11.1% during 1980-1991. The national-effort measure averaged 2.2% during 1950-1979 and 2.9% during 19801991. While Table 1 gives total public expenditure by the education bureaucracy only, Table 2 gives total public expenditure on education (TPEE, column (l)), extra-budgetary expenditure on education (EBEE, column (5)) and the sum of TPEE and EBEE (see column (6)) in 1981-1991. TPEE equals the sum of OEE, CEE and EEOCM. During the 1981-1991 period, EEOCM averaged only 6.7% (see column (2)) of TPEE, indicating that the bulk of public education expenditure was spent by the education bureaucracy. Between 1981 and 1991, TPEE grew by an average annual rate of 13.8% in nominal terms and 6.5% in real terms. But TPEE grew at slower rates between 1986 and 1991 (11.6% in nominal terms and 2.0% in real terms). Column (5) in Table 2 shows that EBEE has grown rapidly during the 1986-1991 period. The implied average growth rate per annum was 27.3% in nominal terms and 16.3% in real terms. Thus EBEE grew much faster than TPEE during the same period. What is also striking is that EBEE's share of the sum of EBEE and TPEE grew from 23.5% in 1986 to 37.2% in 1991 (see column (7)). Non-government resources have thus become an increasingly important source of education expenditure in China since the financial reform in education. It is interesting to note that TPEE as a proportion of GNP for the Seventh Plan Period (1986-1990) averaged only 2.5%, lower than the 2.7% average for the Sixth Plan Period

6

(1981-1985). However, with the support of non-government sources, total budgeted and extra-budgetary education expenditure rose to an average of 3.4% of GNP for the 19861990 period (see column (8)).EBEE figures were not available for 1981-1985; they were expected to be much smaller than those for the 1986-1991 period. Assuming that EBEE averaged 15% (i.e. about half the rate for 1986) of TPEE during 1981-1985, total budgeted and extra-budgetary expenditure on education had an estimated average of 3.1% of GNP in the same period. In other words, the sharp rise in EBEE in the 19861990 period more than compensated for the decline in TPEE as a proportion of total government expenditure. Table 3 provides more details on the breakdown of EBEE by subsector of education. It shows that primary-secondary education claimed an average of 75% of EBEE during 1986-1991. And, in primary-secondary education, social contributions, and education surcharges and levies were the two most important sources of support for EBEE. EBEE contains most of the non-government expenditure on education in China, but it does not include non-fee DPCs (such as parental spending on textbooks, writing supplies, school bags, transportation costs, school uniform, boarding, etc.). Non-fee DPCs are not reported in government reports, but can be estimated from surveys of household education expenditure. Two recent studies provide a preliminary analysis of parental expenditure on schooling. In exploring the potential of household financing of education in China, Chen (1992) provided two analyses of household education expenditure. In the first analysis, Chen used data collected by the State Statistics Bureau on urban households. The data contain education expenditures on three items only: tuition and other school fees, stationery and writing supplies, and magazines and newspapers. They do not include expenditures on school bags, uniforms, transportation, boarding, etc. Also, part of the expenditure on magazines and newspapers may not be related to schooling. The results show that: (1) household education expenditure as a proportion of household income increased from 2.1% in 1987 to 2.4% in 1988 and 2.7% in 1989, and (2) higher-income households spent a lower proportion of their income on education. The second analysis was based on data that Chen and his colleagues collected from 320 households in the

7

Beijing metropolitan area in 1990. The data included household expenditure on schooling (DPCs) and household contributions to school. The analysis found that: (1) Beijing households spent between 2 and 4% of their income on education in 1990, (2) household expenditure on a secondary student was higher than that on a primary student (the ratio varied from 1.3 to 4.1 for different subsamples), and (3) there were large variations in education expenditures among `city/town' households, `near-rural' households and `remote-rural' households. The other study (Tsang, 1990) provides approximate estimates of DPCs at three schooling levels in two provinces (Guizhou and Shaanxi) in 1988. The data were collected by interviewing parents and school staff in selected urban and rural areas in the two provinces. The results indicate that: (1) for Guizhou province, DPCs at the primary level constituted 1.2% of household income in urban areas and 2.1% of household income in rural areas; the corresponding ratios were 1.9 and 4.2% for lowersecondary education, and 3.1 and 7.0% at the upper-secondary level; (2) for both provinces, DPCs were higher for higher levels of schooling; (3) for both provinces, DPCs for urban areas were higher than DPCs for rural areas at the same education level; and (4) non-fee-related DPCs as a proportion of total DPC averaged 84% at the primary level, 82% at the lower-secondary level, and 60% at the upper-secondary level. In addition, this study finds that the burden of private costs of schooling can adversely affect the demand for education, especially in poor areas. In Taijing County (per capita income of about US $50) in Guizhou, for example, 673 of the 1034 students who dropped out of primary school in 1988 reported economic difficulty to be the primary factor for their departure. A recent experiment that cut the costs of textbooks and school fees by half in two primary schools in this county resulted in marked improvement in attendance. In particular, female enrolment increased from 20 to 50% of the total enrolment in these schools. Total non-fee DPC can be estimated from these studies. At the primary-secondary level, total non-fee DPC equals about 80% of total DPC. In other words, total non-fee DPC equals four times of total fee-related DPC at this level. Since enrolment in higher education is about 1% of that of primary-secondary education and that non-fee DPC per higher-education student is roughly ten times of non-fee DPC per primary-secondary

8

student (an assumption), then non-fee DPC at the higher-education level can be estimated to be about 10% of non-fee DPC at the primary-secondary level. Total non-fee DPC equals the sum of non-fee DPCS at the primary-secondary and higher-education levels. By adding total non-fee DPC to EBEE, one obtains an estimate of the total nongovernment expenditure on education. And by adding total non-government expenditure on education to TPEE, ones obtains an estimate of the total national expenditure on education (TNEE). Table 4 provides estimates of TNEE for the 1986-1991 period. Once again, Table 4 documents the increasing importance of non-government financing of education in this period. Government expenditure on education as a proportion of TNEE decreased from 67.4% in 1986 to 52.6% in 1991. International Comparison, Inequality and Inequity Table 5 compares the TPEE of China with other countries in 1985. It shows that the Chinese government spent 13.1% of its total expenditure on education; this level is below the 16% average for developing countries but above the 12.5% average for Asian countries. In the same year, TPEE amounted to 2.8% of GNP for China; this level was less than the 4% average for developing countries and the 3.3% level for Asian countries. For both TPEE measures, the level of public investment in education for China in 1985 was at least comparable to those of other Asian countries of similar level of per-capita GNP; but China still lags behind Asian countries with a high level of per-capita GNP. From a historical perspective, the 1980s were the `golden years' in public mobilization of resources for education in China. Nevertheless, many educational institutions in many parts of China today remain in very poor condition because of the low level of public investment in education in the three decades before 1980. In order to support its transition to a socialist market economy and to raise its standard of living, China still needs to continue to raise its TPEE ratios in the 1990s to get close to the levels for higher-income countries. From a resource mobilization standpoint, one encouraging development in China in the 1980s was the increase in education expenditures financed by non-government sources. Preliminary evidence indicates Chinese households spend about 3% of their income on

9

education, a level comparable to the levels for other developing countries (Psacharopoulos & Woodhall, 1985). The findings on household education expenditure in China are consistent with those of recent studies of other developing countries by Tsang et al. (1990; Tsang & Kidchanapanish, 1992) and other analysts (Schiefelbein, 1986; Wolff, 1985; Bray & Lillis, 1987; Tilak, 1985). A review (see Tsang, 1994a) of these studies indicated that: (1) private costs constitute a substantial proportion of the total cost of education; (2) private education expenditures are an important source of funding for quality-related inputs (such as textbooks) in education; (3) families from higher socio-economic backgrounds (such as higher-income and urban families) spend more than families from lower socioeconomic backgrounds (i.e. lower-income and rural); (4) not only are private educational expenditures disequalizing, they also increase inequity as their share of household income is negatively correlated with the level of household income; and (5) private costs of education can have a negative impact on the demand for education (including female education), especially for families from lower socio-economic backgrounds. Thus, for education decision-makers, there is a dilemma between resource mobilization and quality-improvement objectives on the one hand, and equality and equity objectives on the other hand. Increased resource mobilization for education from private sources can increase inequality and inequity in education. One potential policy option for mitigating the dilemma is to target a larger proportion of additional public resources at disadvantaged populations and areas, while continuing mobilization of private resources for education in all areas (Verspoor & Tsang, 1993). The adoption of this policy option depends critically on the level of political commitment for education for disadvantaged populations in the context of given country. In summary, in terms of both fiscal effort and national effort, total government expenditure on education in the 1980s was substantially higher than in the previous three decades. Non-government expenditure on education increased rapidly in the 1980s, both in absolute amount and as a proportion of TNEE; they accounted for almost one-half of total national resources for education in 1991. However, increased resource

10

mobilization from non-government sources has been accompanied by educational inequality and inequity. Unit Expenditure on Education This section explores three issues related to unit expenditure on education in China: (1) whether unit education expenditures have increased over time in real terms; (2) the extent of disparities in unit expenditures on education; and (3) the determinants of regional disparities in unit expenditures on primary and secondary education. In most of the analyses and studies, `unit' expenditure refers to `per-student' expenditure. Resource Mobilization While Tables 1-4 document the increase in resources for education, they do not indicate whether per-student education expenditure has increased over time in real terms. Table 6 (columns (1)-(3)) gives per-student operating expenditure on education by the government (PSOEE) for three levels of education for the period 1981-1991 in 1981 prices. It shows that PSOEE did increase in real terms between 1981 and 1991 for all three levels of education. However, the growth pattern for the three educational levels were very different (columns (4)-(6)). For both primary and secondary education, PSOEE was on an upward trend. PSOEE increased by 130% for primary education and 90% for secondary education. The arithmetic average of the annual growth rates for the 10-year period was 8.9% for primary education and 7.4% for secondary education. But for higher education, PSOEE had large fluctuations between 1981 and 1991. It increased by 2.3% only in the 10-year period, and the arithmetic average of annual growth rates for the period was 0.4%. As a result of these growth patterns, the unit-cost ratios between the three levels of education have converged over time (see columns (7) and (8)). In other words, increased public expenditure on education in the 1980s benefited primary and secondary education, but not higher education. Using figures on EBEE in Table 3 for primary and secondary education, a rising trend is also found for per-student total budgeted and extra-budgetary expenditure for primary and secondary education. For the period 1986-1991, the value of this unit cost in yuan

11

(1981 prices) per student was respectively 106, 111, 119, 139, 155, and 169 (a similar unit cost cannot be computed for higher education because no separate EBEE is reported for higher education). This implies an average real rate of growth of 9.8% per year for primary-secondary education! In short, per-student expenditures on primary and secondary education had increased substantially in real terms in the 1980s. Disparities in Unit Expenditures on Education While resources for education have increased significantly during the 1980s, large disparities or inequalities also exist in education. Table 7 gives unit costs by level of education by administrative region (provinces, metropolitan areas and autonomous regions) in 1989. Consider first primary education. Column (1) shows that per-student budgeted expenditure on primary education ranged between 46.4 yuan/student for Hubei province and 297.9 yuan/student for Shanghai, and the coefficient of variation was 48.6%. The ratio of maximum per-student expenditure to minimum per-student expenditure was 6.4. This ratio was even larger (12.2) for per-student extra-budgetary expenditure on primary education (see column (2)). Per-student total (budgeted and extra-budgetary) expenditure on primary education ranged between 75.4 yuan/student Guizhou and 392.9 yuan/student for Shanghai, and the coefficient of variation was 42.0% (see column (3)). Large disparities also exist for secondary education (see the maximum-minimum ratios and coefficients of variation in columns (4)-(6)). The regional disparities in per-student budgeted expenditures were relatively the smallest at the higher-education level (see column (7)). But even at this level, the highest spending region (Shandong at 4136 yuan/student) outspent the lowest spending region Giangxi at 1906 yuan/student) by 117%. Column (8) gives the per-capita GNP for each region and again reflects the large disparities in per-capita GNP among regions in China. In fact, the relative dispersion of per-capita GNP is larger than that of any one of the seven perstudent expenditure measures in Table 7. The last row in Table 7 indicates that, for primary and secondary education, unit costs were generally highly correlated with percapita GNP such that regions with higher per-capita GNP had higher per-student expenditures. The correlation was much smaller for higher education.

12

Disparities in unit costs at the county level were reported in a recent study of education expenditures by Jiang (1992). The sample consists of 374 counties from nine regions (Hebei, Liaoning, Shanghai, Jiangsu, Jiangxi, Henan, Hubei, Guizhou and Gansu) such that counties of different levels of economic development were represented. Table 8 gives the average spending level by per-capita income for four measures of unit expenditures: per-capita budgeted operating expenditure on education, per-capita extrabudgetary expenditure on education, sum of per-capita budgeted and extra-budgetary expenditures on education, and per-student budgeted operating expenditure on primary education. It shows that for these unit-expenditure measures, the ratio of the highest spending level to the lowest spending level was about 2:1. Also, for counties with percapita income above 300 yuan per person, all four unit-expenditure measures rose with income. However, for both per-capita budgeted expenditure on education and perstudent budgeted expenditure on primary education, counties with the lowest income level (less than 300 yuan per person) did not have the lowest spending level. This might be due to a smaller student/teacher ratio for the lowest-income counties (because of remoteness and lower population density). Disparities in unit costs are also found at the school level in primary and secondary education. It is an official policy of the government to focus more of its resources on a small number of primary and secondary schools (this is also true for universities). Thus, in urban areas, `key' primary and secondary schools have more resources than their regular counterparts. In rural areas, `center' primary schools have more resources than regular primary schools. These key or center schools generally have higher-performing students. More generally, the government has norms (for per-student non-personnel expenditures and unit capital expenditures) such that city schools are favored over town schools which in turn are favored over rural schools (Tsang, 1990). These practices are aimed at assuring an appropriate level of educational quality for some portion of the student population so that an adequate supply of high-performing students is available for promotion to higher levels of education. These are controversial practices which reflect the tension between equality and social efficiency goals for education in China. Besides inequalities in norms for non-personnel and capital expenditures for primary and secondary schools in different areas, primary and secondary teachers also have very

13

different compensation depending on their employment status (Tsang, 1990). Gongban teachers are employees of the government; they receive a monthly salary, have urban residence (thus having government subsidies for food and their children can go to urban schools) and are eligible for a government pension when they retire. Minban teachers are not government employees; they are hired directly by the local community. Minban teachers receive living subsidies (about 70% of the monthly salary of gongban teachers), have rural residence (thus no food subsidies and their children cannot go to urban schools) and are not eligible for government pension. The compensation for minban teachers has improved over time in the 1980s, but it still lags significantly behind that for gongban teachers. Finally, inequalities in primary and secondary education are reflected in parental spending on schooling per child (per-student DPC). Chen (1992,p. 8) estimated perstudent DPC for households residing in different parts of the Beijing Metropolitan Area in 1990. At the primary level, per-student DPC was 203 yuan/student for `city/town' households, 122 yuan/student for `nearby rural' households, and 64 yuan/student for `remote-rural' households. At the lower-secondary level, per-student DPC was, respectively, 239, 121 and 147 yuan/student for the three categories of households, and at the upper-secondary level, per-student DPC was, respectively, 267, 304 and 265 yuan/student for the three categories of households. Tsang's study (1990) of per-student DPCs in Shaanxi and Guizhou documents the variation between rural and urban areas, and between levels of schooling (see Table 9). Unit costs of vocational-technical education (VTE) in China are related to the type of institution and school subject or curriculum. Currently, there are three types of VTE institutions at the upper-secondary level (Tsang, 1991): secondary vocational schools (SVSs, run by education bureaucracies), secondary technical schools (STSs, run by non-education ministries, departments and enterprises) and skilled workers' schools (SWSs, run by the Ministry of Labor). A study by the World Bank (1987,p. 49) of institutions in the provinces of Liaoning, Hubei and Shaanxi found that, in 1985, perstudent recurrent expenditure averaged 306 yuan per student for SVSs, 1217 yuan per student for STSs, and 1065 yuan per student for SWSs. In contrast, the per-student recurrent expenditure was 146 yuan per student for upper-secondary general schools

14

(USGSs). Using data from 74 institutions in Shanghai in the 1984/198 year, Dougherty (1990,p. 390) estimated that per-student current expenditure was, respectively, 364, 562, 561 and 242 yuan per student for SVSs, STSs, SWSs and USGSs. Both studies found that SWSs were slightly less expensive than STSs; but both SWSs and STSs were much more expensive than SVSs, which in turn were more expensive than USGSs. A study of 54 STSs in Liaoning documents the variation in unit costs by school subject (Li et al., 1988, p. 286). Per-student total school cost (sum of per-student operating cost and per-student annualized capital cost), in yuan per student in 1984, was 1695 for engineering, 2130 for agriculture, 922 for health, 1615 for finance and economics, 1462 for politics and law, 2099 for sports and physical education and 7892 for arts. While Table 7 documents disparities in per-student operating expenditure among regions at the higher-education level, studies by Chinese researchers have shown that disparities in unit expenditures also exist for higher-education institutions within a region and for different types of higher-education institutions across the country. For example, perstudent total institutional cost (sum of per-student operating cost and per-student annualized capital cost) ranged between 1272 yuan per student and 3723 yuan per student among eight specialized higher-education institutions in Jiangsu province in 1983 (Li et al. 1988, p. 231). For higher-education institutions under the administration of central ministries, per-student total institutional cost was 2320 yuan/student for comprehensive universities, 2456 yuan/student for normal universities, 2700 yuan/student for agricultural universities, 2957 yuan/student for medical universities and 3213 yuan/student for engineering universities (Li et al. 1988, p. 212). Determinants of Unit Expenditure: Regional Analysis To further understand regional disparities in per-student budgeted expenditures (see Table 7), ordinary least-squares regression was used to estimate `determinants' equations for primary and secondary education. The dependent variables were average per-student budgeted expenditure on primary education of region, and average perstudent budgeted expenditure on secondary education of region. The explanatory variables were per-capita GNP of region, average student-staff ratio (at primary and secondary levels) of region, average annual pay (salary plus benefits and subsidies) of gongban teachers (at primary and secondary levels) in region, budgeted education

15

expenditure as a percentage of total government expenditure of region, and percentage of non-agricultural population of region. The results are shown in Table 10. For primary education, equation (2) shows that per-capita GNP, average pay, budgeted education expenditure as percentage of total government expenditure and percentage of non-agricultural population were all significant factors. The value of R2 is 0.90. For secondary education, equation (4) shows that only the estimated coefficients of percapita GNP and average pay were significant at the 5% level. Since per-capita GNP and percentage of non-agricultural population were highly correlated with each other (r = 0.88), equation (5) was estimated without percentage of non-agricultural population. The explanatory power of equation (5) was very close to that of equation (4). Additional analyses were conducted which indicated that multi-collinearity existed in equation (4). Equations (1) and (3) indicate that per-capital GNP of region is highly predictive of average per-student budgeted expenditure at both levels of education. For both levels of education, regions that have higher per-capita GNP and that have higher pay for gongban staff spent significantly more than regions with lower per-capita GNP and lower pay. Budgeted education expenditure was a significant factor for primary education, but not for secondary education. It should be pointed out that China is a very diverse country and large economic/cultural/geographical differences often exist within a region. Research at the sub-regional levels may yield more policy-relevant context-specific findings. Widening Disparities in Education Over Time? The above studies clearly document the significant inequalities or disparities in education in China. One common feature of these studies is that they show educational inequalities or disparities at one point in time only. Because of lack of comprehensive time-series data, it is difficult to provide direct quantitative evidence on how inequalities or disparities in education have changed over time in China. Most education observers and analysts (including this author) probably agree that, because of fiscal decentralization and diversification of financing sources, education disparities in the

16

1980s had widened relative to the pre-1980 period. Indirect evidence suggests that educational disparities had widened since the latter half of the 1980s. Table 4 indicates that non-government expenditures had increased in importance (in both amount and as a proportion of TNEE) over time during 1986-1991. Since such expenditures are disequalizing, it is highly probable that disparities in per-student non-government education expenditures had also increased during the same time period. Thus, regional disparities in per-student total education expenditure were very likely to have increased over time since the second half of the 1980s. With continuing fiscal decentralization and increasing reliance on non-government resources education disparities will likely continue to widen in the 1990s, unless there is some drastic intervention by the government. But further research on this topic is necessary to provide more direct and convincing evidence. International Comparison Table 6 shows that, using per-student operating expenditure by the government as a measure, higher education was 52.4 times as expensive as primary education and secondary education was 2.7 times as expensive as primary education in 1985 in China. In other Asian countries around the mid-1980s, higher education averaged 19.4 times as expensive as primary education and secondary education averaged 1.9 times as expensive as primary education (computed from Tan & Mingat, 1989, p. 56). Thus, the Chinese government spent relatively little on primary education and higher education was relatively very expensive. As explained in the next section, the high relative cost of higher education in China was due to small institutional size and very low student-toteacher ratios in higher education in In a review of five studies on the US and eight studies covering 17 developing countries, Tsang (forthcoming) found that: (1) vocational schools are more expensive than academic schools, comprehensive schools or diversified schools. Technical schools are the most expensive and academic schools are the least expensive; (2) the ratios of unit costs of VTE schools to those of academic schools vary widely; and (3) relative costs depend critically on the length of schooling, they vary by school subject and they can change significantly over time. These three findings for other countries also apply to

17

China. The review also found that vocational/technical schools averaged 1.6 times as expensive as academic schools in the US; and vocational/technical schools averaged 2.5 times as expensive as academic schools in countries in Africa, Asia and Latin America. These averaged ratios are comparable to those for Shanghai. But, in contrast, vocational/technical schools in Liaoning, Hubei and Shaanxi were relatively much more expensive than academic schools in the same regions. Educational disparities by region, (urban versus rural) areas, level of schooling, subject/curriculum and by type of institution are found for China as well as for other developing countries (see review in Tsang, 1988, pp. 201-204). The existence of such disparities in China is not surprising. What should be noted are the large disparities in education in China. For most of the time before 1978, the Chinese leadership had taken pride in pursuing socialism with a strong egalitarian orientation. But with the adoption of modernization policies and the transition to a socialist-market economy, educational (and economic) disparities have widened to substantial levels in the 1980s. Cost Functions and Economies of Scale This section reviews studies of educational cost functions to examine the issue of inefficiency in education in China. Inefficiency in Resource Utilization Educational cost functions relate the costs of education to the output of education, given prices and technology of education (Verry, 1987). An important application in the estimation of educational cost functions is the determination of economies (or diseconomies) of scale in educational production. If there are economies (or diseconomies) of scale, additional units of education output can be produced at relatively lower (or higher) additional costs and efficiency in the utilization of education resources will be raised (lowered). In practice, educational cost functions are estimated by relating some measures of education costs to some measure of scale (such as enrolment, or volume of training) and other control variables, including educational quality (for example,

18

using the student-teacher ratio as a proxy). Several recent studies have estimated cost functions for three different subsectors of education in China. Besides comparing the per-student current expenditure of different types of secondary vocational/technical schools in the Shanghai metropolitan area, Dougherty also used both log-linear and linear models to explore the existence of economies of scale for these schools. The explanatory variables include dummy variables of school types (TECH for STSs, WORK for SWSs, VOC for SVSs and GEN for USGSs), interaction terms involving size (N, enrolment) and school type, and two other control variables (BOARD, dummy variable on whether a school is residential and EDHEALItH, dummy variable on whether a school's programs are oriented towards the service sector). The dependent variables include total current expenditure (and its personnel and nonpersonnel components) and per-student current expenditure (and its personnel and nonpersonnel components). Estimated equations based on the two models indicate economies of scale for the different types of vocational/technical schools in Shanghai. Economies of scale were more pronounced among secondary technical schools and secondary vocational schools; but were less so for skilled workers' schools (and secondary general schools). Equation (a) in Table 11 gives one of the estimated equations based on a log-linear model. Tsang and Min (1992) studied the expansion of higher education in China in the 1980s. They found that higher education expanded primarily through institutional multiplication (i.e. increased enrolment by establishing additional institutions), instead of institutional enlargement (i.e. increased enrolment in existing institutions). The results of this type of expansion are that higher education is characterized by the existence of many small institutions (over half of the 1075 institutions have less than 1500 students), and a low student:teacher ratio (5.2 in 1985). Based on data from 156 institutions in five provinces in 1988-1989, they estimated an equation relating per-student recurrent expenditure to enrolment, student-to-teacher ratio and three sets of control variables (one set of dummy variables on the type (curriculum-based) of institutions, a second set of dummy variables on provinces and a third set on administrative levels). The equation (see equation (b) in Table 11) indicates the existence of economies of scale and that per-student recurrent expenditure decreased significantly with increasing enrolment. Simulations based on this

19

equations shows that if institutional size had increased from 1922 (average in 1988) to 3000 (a reasonable target for size), and student:teacher ratio had increased from 5.25 (average in 1988) to 8.0 (a reasonable target for the student:teacher ratio), per-student recurrent expenditure in 1988 would have decreased by 17%. In fact, the government could have saved about 7 billion yuan (or 2 billion US$ in 1988 prices) over the 19771989 period. The existence of economies of scale in higher education in China was also reported by an earlier study by the World Bank (1986). Taking advantage of economies of scale is an important policy option for improving internal efficiency and accommodating further enrolment expansion in higher education in China in the 1990s (Tsang & Min, 1992; Min, 1991; Wang, 1989). In addition, cost recovery in higher education is relatively low in China; shifting the costs of higher education towards the student and their families is a potential option for mobilizing resources for higher education (Min, 1991; Li et al., 1988). In a case study of adult education in Shenzhen, China, Xiao and Tsang (1994) examined the costs and financing of three types of adult education institutions: institutions whose clients come from the local community, institutions whose clients are employees from enterprises; and institutions with clients from both the local community and enterprises. They found that recurrent expenditure per training hour in the 1986-1990 period was significantly related to the volume of training, after controlling for the trainee:trainer ratio, and type of institutions. Again, economies of scale exist so that the recurrent expenditure per training hour decreases with the total training hours offered by an institution (see equation (c) in Table 11). In short, available studies of educational cost functions indicate that there is significant under-utilization of resources in different subsectors of education in China. International Comparison There are very few empirical estimations of cost functions for VTE. The best known, but dated, study was carried out by Maton and Van de Vijere (1970). Their sample consisted of 31 vocational training institutions in Europe, Asia and Latin America. Their analyses show that unit costs (personnel as well as institutional costs per student and per hour)

20

were significantly and negatively related to the size of an institution. The study by Dougherty (1990) is a recent addition to an otherwise sparse literature. Analyses of economies of scale have been part of a longer tradition for examining efficiency (or inefficiency) in higher education in both developing and developed countries. Earlier cross-national studies of both developed and developing countries have documented the existence of economies of scale in higher education (Psacharopoulos, 1982; Lee, 1984). In addition, a review of 60 years of studies in the US by Brinkman and Leslie (1986) as well as more recent studies by Cohn et al. (1989) and De Groot et al. (1991) also conclude the existence of economies of scale in higher education in the United States. The study by Tsang and Min (1992) documents the inefficiency and illustrates the magnitude of potential savings in higher education in China. As pointed out in the previous section, higher education in China is relatively much more expensive than the average of other Asian countries. China's student:teacher ratio is very low compared to the Asian average of 15.2 (in 1985) and to the typical levels (10-15) for developed countries. The average size of Chinese institutions is also relatively small compared to most of the developed countries with large higher-education systems (Tsang & Min, 1992, p. 62). Note that the findings on China do not imply that Chinese higher education is expensive in absolute terms; they simply indicate that the scarce resources for higher education have not been efficiently utilized. The problem of low internal efficiency has become more widely known in recent years in China. There are hardly any published empirical studies of cost functions of adult education and training. Part of the reason for this lack of studies is the difficulty in measuring the costs of adult education and training (Tsang, 1994b). Finally, there are also no published empirical studies of cost functions of primary and secondary education in China. Some evidence of economies of scale has been documented for primary and secondary schools in the US, but no conclusive finding has yet been drawn for primary and secondary schools in developing countries (see review in Tsang, 1988, pp. 208-209). Research on cost functions of primary and secondary education in China will address an obvious gap in the current literature.

21

Summary and Discussion For three decades after the founding of the People's Republic of China, public expenditures on education were consistently low in terms of both fiscal-effort and national-effort measures. The substantial increase in government expenditure on education in the 1980s represents a major turning point in public investment in education. But public expenditure on education in China today is still relatively low compare to higher-income countries. In order to develop human resources in support of further economic development, the Chinese government has to keep on increasing public expenditures in education in the 1990s and beyond. Another distinguishing feature of the costs of education in the 1980s is the increasing importance of non-government expenditures on education. By 1991, total nongovernment expenditure accounted for almost one-half of total national resources for education. The increases in both government and non-government resources for education have resulted in high rates of growth in per-student expenditures in real terms, especially for primary and secondary education. The development of non-government sources has been an effective strategy for raising national investment in education in China. However, the significant progress in resource mobilization for education in the 1980s was also accompanied by large inequalities and inequities in education. There is also indirect evidence of widening disparities in recent years in China. There is certainly a dilemma between resource-mobilization and equality/equity objectives for education. This dilemma is rooted in the long-standing tension in divergent development objectives and strategies among leaders of the CCP. To the extent that extreme inequalities can provoke social unrest, there is a real need to reassess government efforts for mitigating the extent of inequalities in education. Currently, both the central and provincial government do provide inter-governmental education grants to local governments for equalization purposes; but such grants are too small to significantly alter such inequalities. Part of the difficulty for the central government is that, through a publicfinance reform since 1982, it has devolved much of its revenue-raising and spending powers to lower levels of government. The national government is presently engaged in

22

a fundamental overhaul of its system of taxation. The challenge of educational equalization lies in securing the necessary political commitment so that central and provincial governments will increase their equalization grants substantially. The promotion of educational equity requires a careful examination of how to redistribute the costs and benefits of education so that there are net gains for the disadvantaged populations, especially the rural and poor population groups. There is also substantial under-utilization of current resources for education. Since inefficiency in resource utilization can undermine the gains in resource mobilization, improving efficiency in education should be a high priority area for the government. Studies of educational inefficiency and proposed solutions to this inefficiency should receive more attention (Li et al., 1988). Finally, this paper has identified a number of research areas for furthering the understanding of education costs in China. These areas include, for example, household expenditures (including expenditures on education) and household demand for education (including female education), over-time analysis of educational disparities, cost functions of primary and secondary education, determinants of unit expenditures at sub-regional levels (such as county and town/township levels) and cost recovery in higher education. Table 1. Public expenditure on education by education bureaucracy, 1950-1991 Legend for Table

[A] = OEE (billions of yuan) current prices (1) [B] = CEE (billions of yuan) current prices (2) [C] = OEE + CEE (billions of yuan) current prices (3) = (1) + (2) [D] = OEE + CEE (billions of yuan) 1950 prices (4) [E] = OEE + CEE as % of government expenditure (5) [F] = OEE + CEE as % national income (6)

Year

[A]

[B]

[C]

23

[D]

[E]

[F]

1950

0.38

0.03

0.41

0.41

5.9

1.0

1951

0.74

0.09

0.84

0.74

6.8

1.7

1952

0.90

0.26

1.15

1.03

6.5

2.0

1953

1.28

0.34

1.62

1.40

7.4

2.3

1954

1.38

0.39

1.77

1.49

7.2

2.4

1955

1.41

0.25

1.66

1.39

6.2

2.1

1956

1.65

0.34

1.99

1.67

6.5

2.3

1957

1.95

0.30

2.25

1.86

7.4

2.5

1958

1.98

0.26

2.24

1.84

5.5

2.0

1959

2.41

0.48

2.89

2.35

5.2

2.4

1960

3.18

0.79

3.96

3.13

6.1

3.2

1961

2.68

0.24

2.92

1.98

7.9

2.9

1962

2.41

0.10

2.51

1.64

8.2

2.7

1963

2.49

0.21

2.70

1.88

7.9

2.7

1964

2.78

0.35

3.13

2.27

7.9

2.7

1965

2.91

0.33

3.24

2.41

7.0

2.3

1966

3.44

0.20

3.64

2.71

6.7

2.3

1967

3.27

0.10

3.37

2.53

7.6

2.3

1968

2.75

0.08

2.83

2.12

7.9

2.0

1969

2.70

0.06

2.76

2.09

5.2

1.7

1970

2.76

0.06

2.82

2.14

4.3

1.5

1971

3.30

0.11

3.41

2.61

4.7

1.6

1972

3.85

0.24

4.10

3.15

5.3

1.9

1973

4.21

0.34

4.55

3.47

5.6

2.0

1974

4.60

0.38

4.98

3.78

6.3

2.1

1975

4.83

0.37

5.19

3.94

6.3

2.1

1976

5.05

0.38

5.43

4.10

6.7

2.2

1977

5.30

0.38

5.69

4.21

6.7

2.2

1978

6.56

0.65

7.21

5.30

6.5

2.4

1979

7.70

1.11

8.81

6.35

6.9

2.6

1980

9.42

1.40

10.82

7.37

8.9

2.9

24

1981

10.25

1.51

11.76

7.82

10.5

3.0

1982

11.57

1.75

13.32

8.69

11.6

3.1

1983

12.79

2.41

15.19

9.76

11.8

3.2

1984

14.82

3.16

17.98

11.24

11.6

3.2

1985

18.42

4.38

22.79

13.09

12.4

3.2

1986

21.43

3.19

24.62

13.34

10.6

3.1

1987

22.87

2.83

25.70

12.98

10.5

2.8

1988

27.56

3.00

30.56

13.02

11.3

2.6

1989

31.79

3.35

35.13

12.71

11.7

2.7

1990

35.65

2.96

38.62

13.68

11.4

2.7

1991

38.87

2.96

41.83

14.40

11.0

2.6

6.5

2.2

11.1

2.9

7.8

2.4

Average 1950-1979 Average 1980-1991 Average 1950-1991

Source: computer from Ministry of Education (1984), State Education Commission (1989, 1991, 1992), and State Bureau of Statistics (1992). Table 2. Total expenditure of education, 1981-1991 Legend for Table

[A] = TPEE (billions of yuan) current price (1) [B] = EEOCM as % of TPEE (2) [C] = TPEE as % of government expenditure (3) [D] = TPEE as % of GNP (4) [E] = EBEE (billions of yuan) current price (5) [F] = TPEE + EBEE (billions of yuan) current price (6)

25

[G] = TPEE + EBEE as % of government expenditure (7) [H] = TPEE + EBEE as % GNP (8)

Year

[A]

[B]

[C]

[D]

1981

12.66

7.1

11.4

2.7

1982

14.18

6.1

12.3

2.7

1983

16.07

5.5

12.4

2.8

1984

19.11

5.9

12.4

2.7

1985

24.09

5.4

13.1

2.8

1986

26.50

7.1

11.4

2.7

1987

27.70

7.2

11.3

2.5

1988

32.36

5.6

12.0

2.3

1989

39.77

8.1

13.2

2.5

1990

43.38

7.9

12.8

2.5

1991

45.97

8.3

12.1

2.3

6.0

12.3

2.7

7.2

12.1

2.5

6.7

12.2

2.6

Average 1981-1985 Average 1986-1990 Average 1981-1991

Year

[E]

[F]

[G]

[H]

1981

NA

NA

NA

NA

1982

NA

NA

NA

NA

1983

NA

NA

NA

NA

1984

NA

NA

NA

NA

1985

NA

NA

NA

NA

1986

8.13

34.63

26

23.5

3.6

1987

9.54

37.24

25.6

3.3

1988

12.72

45.08

28.2

3.2

1989

16.45

56.22

29.3

3.5

1990

18.25

61.63

29.6

3.5

1991

27.18

73.15

37.2

3.7

Average 1981-1985

3.2[*]

Average 1986-1990

3.4

Average 1981-1991

3.4[*]

Source: compiled from State Education Commission (1989, 1991, 1992) and State Bureau of Statistics (1992).

[*] Estimated. Table 3. Extra-budgetary education expenditure, 1986-1991 (billion yuan, current prices) Legend for Table

[A] = Surcharges and levies (1) [B] = Social contributions (2) [C] = School generated (3) [D] = School fees (4) [E] = Total primary/secondary (5) = (1) + (2) + (3) + (4) [F] = Higher education and others (6) [G] = Total for education (7) = (5) + (6)

Primary and secondary education

27

Year

[A]

[B]

[C]

[D]

[E]

[F]

[G]

1986

1.72

1.60

0.70

1.06

5.08

3.07

8.15

1987

2.64

1.62

0.90

1.10

6.26

3.28

9.54

1988

3.45

2.46

1.31

1.63

8.85

3.87

12.72

1989

5.46

3.58

1.79

2.81

13.64

2.81

16.45

1990

5.60

5.45

2.45

3.06

16.56

1.69

18.25

1991

7.52

6.28

3.72

3.24

20.76

6.42

27.18

Source: computed from State Education Commission (1989, 1991, 1992). Table 4. Total national expenditure on education (TNEE), 1986-1991 (current billion yuan) Legend for Table

[A] = Total government (TPEE) (1) [B] = Total non-government (2) [C] = TNEE (3) = (1) + (2) [D] = TPEE as % of TNEE (4) [E] = TNEE as % of GNP (5)

Year

[A]

[B]

[C]

[D]

[E]

1986

26.50

12.79

39.29

67.4

4.1

1987

27.70

14.38

42.08

65.8

3.7

1988

32.36

19.89

52.25

61.9

3.7

1989

39.77

28.81

68.58

58.0

4.3

1990

43.38

31.71

75.09

57.8

4.2

1991

45.97

41.44

87.41

52.6

4.4

Table 5. Public expenditure on education, China and other Asian countries

28

As % of total government

Per-capita

expenditure Country

Bangladesh

(1985)

As % of GNP (1985)

GNP, US$ (1985)

10.3

1.5

159

7.3

3.8

151

Burma

10.9

1.8

184

China

13.1

2.8

273

India

13.7

3.0

259

Indonesia

15.0

3.7

470

S. Korea

16.6

3.4

2040

Malaysia

16.0

6.0

1860

9.6

1.8

142

Papua New Guinea

17.9

6.9

621

Philippines

11.5

1.8

581

8.1

2.8

374

12.5

3.6

712

12.5

3.3

16.0

4.0

Bhutan

Nepal

Sri Lanka Thailand

Average Asia Developing countries[*]

Source: Tan and Mingat (1989) and Table 2.

[*] Average not based on 1985 only. Table 6. Per-student operating expenditure on education by government (PSOEE), 1981-1991 (1981 prices) Annual (yuan) per student

29

Primary

Secondary

Higher

education

education

education (3)

Year

(1)

(2)

1981

25.9

69.6

1752.8

1982

30.3

84.3

1886.1

1983

33.2

93.7

2112.3

1984

36.3

101.2

2094.9

1985

40.9

111.0

2140.1

1986

39.4

110.0

2090.2

1987

45.5

107.2

1758.2

1988

49.4

114.8

1675.8

1989

49.5

115.4

1464.3

1990

56.0

127.4

1610.1

1991

59.6

132.1

1793.5

Average 1981-1991

Annual growth ratee (%)

Year

Primary

Secondary

education

education

education

(5)

(6)

(4)

Higher

1981

9.3

13.1

-2.3

1982

17.0

21.2

7.6

1983

9.4

11.1

12.0

1984

9.2

8.0

-0.8

1985

12.7

9.8

2.2

1986

-3.6

-1.0

-2.3

1987

15.7

-2.5

-15.9

30

1988

8.4

7.0

-4.7

1989

0.3

0.5

-12.6

1990

13.0

10.4

10.0

1991

6.4

3.7

11.4

7.4

0.4

Average 1981-1991

Unit cost ratios

Secondary

Year

Higher

education

education

(7) = (2)/(1)

(8) = (3)/(1)

1981

2.7

67.6

1982

2.8

62.2

1983

2.8

63.6

1984

2.8

57.8

1985

2.7

52.4

1986

2.8

53.1

1987

2.4

38.6

1988

2.3

33.9

1989

2.3

29.6

1990

2.3

28.8

1991

2.2

30.1

2.6

47.1

Average 1981-1991

Source: computer from State Education Commission (1989, 1991, 1992), Table 3 and State Statistics Bureau (1992, p. 236).

31

Table 7. Per-student education expenditures by education level by region, 1989 Primary education

Budgeted Region

Extra-budget

(1)

(2)

Sum (3) = (1) + (2)

Beijing

246.0

61.2

307.2

Tianjin

203.0

95.8

298.9

Shanghai

297.9

95.0

392.9

Liaoning

127.5

74.1

201.5

99.8

83.8

183.5

Zhejiang

119.9

70.4

190.2

Guangdong

120.3

90.3

210.6

Jilin

117.3

74.4

191.7

Heilongjiang

111.1

62.2

173.4

Fujian

129.3

76.2

205.5

Shandong

68.3

62.8

131.0

Hebei

66.4

67.7

134.0

Shanxi

94.9

37.2

132.1

141.7

26.8

168.4

Hubei

46.4

56.6

103.0

Hunan

77.5

50.1

127.7

Hainan

144.8

36.3

181.1

Qinghai

145.7

35.9

181.6

Ningxia

110.6

7.8

118.5

Anhui

59.5

34.5

94.0

Jiangxi

68.2

24.5

92.7

Henan

48.6

53.2

101.8

Guangxi

98.2

50.5

148.7

Sichuan

78.8

48.4

127.2

Guizhou

64.5

10.9

75.4

Jiangsu

In. Mongolia

32

Yunnan

109.2

11.6

120.9

86.6

38.5

125.2

Gansu

102.7

33.3

136.0

Xinjiang

188.2

NA

166.3

52.5

166.2

SD

56.6

24.6

69.8

CV[*](%)

48.6

46.9

42.0

Maximum

297.9

95.8

392.9

Minimum

46.4

7.8

75.4

Max/min

6.4

12.2

5.2

Shaanxi

NA

Statistics

Mean

Correlation with PC-GNP

0.86

0.63

0.93

Secondary education

PerSum

Budgeted Region

Higher

capital GNP

Extra-

(6) =

education

budget

(4) +

budgeted

(5)

(4)

(5)

Beijing

582.1

170.1

752.2

3256.2

4304

Tianjin

426.1

182.4

608.5

3066.0

3335

Shanghai

586.2

202.1

788.3

3282.0

5489

Liaoning

268.3

151.9

420.2

2621.1

2396

Jiangsu

197.4

151.1

348.5

2974.4

1894

Zhejiang

222.5

130.9

353.4

2667.5

1885

Guangdong

251.0

180.6

431.6

3557.0

2195

Jilin

226.9

111.3

338.3

3302.1

1513

33

(7)

(yuan) (8)

Heilongjiang

219.9

72.2

292.1

2970.7

1670

Fujian

272.3

127.0

399.3

2575.2

1451

Shandong

176.8

166.7

343.5

4136.4

1480

Hebei

223.7

102.9

326.7

2645.5

1283

Shanxi

210.0

71.0

281.0

2832.8

1262

In. Mongolia

260.5

56.7

317.2

3280.0

1220

Hubei

186.5

127.3

314.2

2832.7

1342

Hunan

191.2

128.5

319.7

2300.7

1077

Hainan

283.8

93.4

377.2

3885.0

1371

Qinghai

278.1

22.7

300.9

2853.4

1381

Ningxia

240.2

37.0

277.2

2861.8

1239

Anhui

147.8

69.4

217.2

2122.7

1055

Jiangxi

149.1

70.5

219.6

1905.5

1003

Henan

158.5

122.4

280.9

2237.4

1012

Guangxi

204.6

98.7

303.3

2778.2

848

Sichuan

165.5

84.3

249.8

2496.8

938

Guizhou

144.5

29.2

173.7

2800.8

748

Yunnan

248.2

50.6

298.9

3064.6

871

Shaanxi

224.4

86.7

311.1

2611.7

1074

Gansu

186.8

43.3

230.1

2650.8

1007

Xinjiang

324.0

NA

3056.2

1510

NA

Statistics

Mean

250.2

105.1

352.7

2883.6

1650.1

SD

107.8

49.3

141.0

478.1

1040.6

43.1

46.9

40.0

16.6

63.1

Maximum

586.2

202.1

788.3

4136.4

5489.0

Minimum

144.5

22.7

173.7

1905.5

748.0

Max/min

4.1

8.9

4.5

2.2

7.3

0.91

0.71

0.96

0.34

1.00

CV[*](%)

Correlation with PC-GNP

34

Source: computed from data reported in State Education Commission (1990, pp. 40, 81-84).

[*] Coefficient of variation. Table 8. Disparities in unit education expenditures (yuan/unit) at country level, 1990 Legend for Table

[A] = Per-capita budget expenditure [B] = Per-capita extra budget expenditure [C] = Sum of per-capita budget and extra budget expenditure [D] = Per-student budgeted expenditure on primary education

Per-capita

[A]

[B]

[C]

[D]

income level

(1)

(2)

(3) = (1) + (2)

(4)

< 300 yuan

18.78

10.25

29.03

85.15

300-400 yuan

16.91

13.30

30.21

65.46

400-600 yuan

18.28

12.52

30.80

72.06

600-800 yuan

19.70

19.23

38.93

84.07

> 800 yuan

33.66

22.01

55.67

154.19

Mean of sample

19.78

14.60

34.48

82.84

Source: Jiang (1992, pp. 13-14). Table 9. Direct private costs of education in Shannxi and Guizhou, 1988 (yuan per student) Primary

Lower

35

Upper

education

Urban

secondary

Rural

Urban

Rural

secondary

Urban

Rural

Shaanxi province Tuition

0

0

0

0

12

10

Other school fees

7

5

18

14

50

40

21

16

43

30

30

30

8

5

15

10

20

15

Miscellaneous

15

15

20

20

25

25

Total[*]

51

41

96

74

137

120

Textbooks Writing supplies

Guizhou province Tuition

2

1.5

Other school fees

7

7

13

12

52

44

24

10

40

35

40

40

8

5

15

10

20

15

Miscellaneous

15

15

20

20

25

25

Total[*]

56

39

91

79

146

132

Textbooks Writing supplies

2.5

2

3.5

2.5

Source: Tsang (1990).

[*] Does not include boarding costs in some secondary schools which total about 150 yuan per year. Table 10. Determinants of per-student budgeted education expenditures, 1989 Primary education

Explanatory variables

(1)

Per-capita GNP of region

0.047[*]

Student: staff ratio

(2)

0.018[*] 0.65

36

Average annual pay of gongban staff

0.084[*]

Budgeted educational expenditure as % of government expenditure

4.65[*]

% Non-agricultural population

222.23[*]

Constant

39.41[*]

R2

0.74

No. observations

-217.35[*] 0.90

29

29

Secondary education

Explanatory variables

(3)

Per-capita GNP of region

(4)

0.094[*]

(5)

0.053[*]

Student: staff ratio

-4.45

0.070[*] -7.65

Average annual pay of gongban staff

0.12[*]

0.13[*]

4.66

1.06

Budgeted educational expenditure as % of government expenditure % Non-agricultural population

183.00

Constant

94.83[*]

R2

0.83

No. observations

-129.42[*] 0.91

29

-21.00[*]

0.90 29

29

[*] Significant at 5% level. Table 11. Cost functions of education, China Equation (a): estimated cost function of upper-secondary schools, log-linear model, 1984-1985

37

Dependent variable: total current expenditure (yuan)

Explanatory

Estimated

variables

coefficient

Constant

6.51[**]

TECH

1.27

WORK

0.50

VOC

1.86

TECH.logN

0.75[**]

WORK.logN

0.87[**]

VOC.logN

0.56[*]

GEN.logN

0.83[**]

BOARD

0.028[*]

EDHEALTH R2

-0.46[**] 0.74

No. of observations

74

Source: Dougherty (1990, p. 392).

[*] Significant at 5% level.

[**] Significant at 1% level. Equation (b): estimated cost function of higher-education institutions, 1988 Dependent variable: per-student recurrent expenditure (yuan/student)

Estimated Explanatory variables

coefficient[*]

1/(enrolment)

379648

38

Student:teacher ratio

-105

Teacher-training institutions

-468

Art institutions

1011

Other institutions

-523

Locally run institutions

-1492

Institutions in Yunnan

999

Institutions in Shanxi

684

Constant

2406

R2

0.55

No. of observations

156

Source: Tsang and Min (1992, p. 64).

[*] Coefficient shown here are all significant at the 5% level. Explanatory variables with estimated coefficients not significant at the 5% level are not shown in the equation. Equation (c): estimated cost function of adult education institutions, 1990 prices Dependent variable: recurrent expenditure per training hour (yuan/hour)

Estimated Explanatory variables

coefficient[*]

1/(no. training hours)

51468[*]

Trainee:trainer ratio

44.37

Community-oriented institution

-0.36

Mixed institution

2.02[*]

Constant

0.4

R2

0.70

No. of observations

22

39

Source: Xiao and Tsang (1994, available from authors).

[*] Significant at 5% level.

Figure 1. Public education expenditure ratios (Y), by plan periods (X), 1953-1990. References Bray, M. & Lillis, K. (1987) Community Financing of Education in Developing Countries: Issues and Policy Options (Oxford, Pergamon Press). Brinkman, P. & Leslie, L. (1986) Economies of scale in higher education: sixty years of research, The Review of Higher Education, 10, pp. 1-28. Chen, L. (1992) The impact of population growth on education and the potentiality of family education expenditure in China, paper presented at the Policy Seminar on Financing of Education in China, Dalian, China, 17-22 August 1992 (in Chinese). Cohn, E., Rhine, E. & Santos, M. (1989) Institutions of higher education as multi-product firms: economies of scale and scope, The Review of Economics and Statistics, 71, pp. 284-290. De Groot, H., McMahon, W. & Volkwein, J. (1991) The cost structures of American research universities, The Review of Economics and Statistics, 73, pp. 424-431. Dougherty, C.R.S. (1990) Unit costs and economies of scale in vocational and technical education: evidence from the People's Republic of China, Economics of Education Review, 9, pp. 389-394. Jiang, M. (1992) The development pattern of Chinese education finance at county level: observations on 374 counties, paper presented at the Policy Seminar on Financing of Education in China, Dalian, China, 17-22 August 1992 (in Chinese). Lee, K. (1984) Further Evidence on Economies of Scale (Washington, DC, World Bank).

40

Li, Y., Chen, L., Meng, M. & Wang, S. (Eds) (1988) Studies in Economies of Education (Shanghai, China, Shanghai People's Press) (in Chinese). Maton, J. & Van de Vijvere (1970) The comparative study of training cost: a possible approach, International Labour Review, 102, pp. 577-590. Min, W. (1991) A Comparative Study of Higher Education Development in Selected Asian Countries 1960-1990: Country Case Analysis of China (Beijing, China: Institute of Higher Education, Beijing University) Ministry of Education, China (1984) Achievement of Education in China (Beijing, China, People's Education Press). Psacharopoulos, G. (1982) The economics of higher education in developing countries, Comparative Education Review, 26, pp. 139-159. Psacharopoulos, G. & Woodhall, M. (1985) Education and Development: An Analysis of Investment Choices (Oxford, Oxford University Press). Schiefelbein, E. (1986) Education Costs and Financing Policies in Latin America: A Review of Available Research (Washington, DC, World Bank). State Education Commission, China (1989) Education Statistical Yearbook 1988 (Beijing, China, Beijing Industrial University Press) (in Chinese). State Education Commission (1990) Report on Regional Educational Expenditures in China, 1989 (Shanghai, China, Tongji University Press) (in Chinese). State Education Commission, China (1991) Education Statistical Yearbook 1990 (Beijing, China, Beijing Industrial University Press) (in Chinese). State Education Commission, China (1992) Education Statistical Yearbook 1991/1992 (Beijing, China, Beijing Industrial University Press) (in Chinese).

41

State Bureau of Statistics, China (1992) China Statistical Yearbook 1992 (Beijing, China, China Statistics Press) (in Chinese). Tan, J. & Mingat, A. (1989) Educational Development in Asia: A Comparative Study focussing on Cost and Financing Issues (Washington, DC, World Bank). Tilak, J. (1985) Analysis of Costs of Education in India, Occasional Paper No. 10 (New Delhi, India, National Institute of Educational Planning and Administration). Tsang, M. (1988) Cost analysis for educational policymaking: a review of cost studies in education in developing countries, Review of Educational Research, 58, pp. 181-230. Tsang, M. (1990) Financing of primary and secondary education in Shaanxi and Guizhou, Report prepared for the China Department, World Bank, Washington, DC. Tsang, M. (1991) The structural reform of secondary education in China, Journal of Educational Administration, 29, pp. 65-83. Tsang, M. (1993) Financial reform of basic education: the Chinese experience, paper presented at the Economics of Education Symposium, Manchester, England, 19-21 May 1993. Tsang, M. (1994a) Private and public costs of schooling in developing countries, in: Husen, T. & Posthlewaite, N. (Eds) International Encyclopedia of Education, Second Edition (Oxford, Pergamon Press), pp. 4702-4708. Tsang, M. (1994b) Costs of continuing education: industrialized countries, in: Husen, T. & Posthlewaite, N. (Eds) International Encyclopedia of Education, Second Edition (Oxford, Pergamon Press), pp. 1146-1152. Tsang, M. (forthcoming). The costs of vocational training, in Middleton, J. & Ziderman, A. (Eds) Vocational Education and Training in Developing Countries (Washington, DC, World Bank).

42

Tsang, M., Zaki, M. & Ghafoor, A. (1990) Household Educational Expenditures in Pakistan (East Lansing, MI: College of Education, Michigan State University). Tsang, M. & Min, W. (1992) Expansion, efficiency, and economies of scale of higher education in China, Higher Education Policy, 5, pp. 61-66. Tsang, M. & Kidchanapanish, S. (1992) Private resources and the quality of primary education in Thailand, International Journal of Educational Research, 17, pp. 179-198. Verry, D. (1987) Education cost functions, in: Psacharopoulos, G. (Ed.) Economics of Education: Research and Studies (Oxford, Pergamon Press), pp. 402-409. Verspoor, A. & Tsang, M. (Eds) (1993) Case Studies in Financing Quality Basic Education (Washington, DC, Education and Social Policy Department, World Bank). Wang, S. (Ed.) (1989) Introduction to Economics of Education (Beijing, China, Beijing Normal University Press) (in Chinese). Wolff, L. (1985) Controlling the Costs of Education in Eastern Africa: A Review of Data Issues and Policies, Staff Working Paper No. 702 (Washington, DC, World Bank). World Bank (1986) China: Management and Finance of Higher Education (Washington, DC, World Bank). World Bank (1987) Technical/Vocational Education for China's Development (Washington, DC, World Bank). Xiao, J. & Tsang, M. (1994) The costs and financing of adult education in Shenzhen, China, International Journal of Educational Development, 14, pp. 51-64. ~~~~~~~~ By MUN C. TSANG Mun C. Tsang, Department of Education, Michigan State University, Erikson Hall, East Lansing, MI 48824-1034, USA

43