Graphics & Design


of 26
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Similar Documents
TUITION, PRIVATE DEMAND AND HIGHER EDUCATION IN CHINA WENLI LI, WEIFANG MIN (GRADUATE SCHOOL OF EDUCATION, PEKING UNIVERSITY, BEIJING, ) JUARARY 2001 Abstract All of higher education institutions in China have adopted cost-recovery policy since This study analyzes the impact of college costs, expected return to education and family education and financial background on the probability of individual enrollment in higher education in China, especially in urban areas. The data are from Urban Household Survey of the State Statistical Bureau of China, which was collected in August Using price-response measures, this study examines the cost sensitivity varying among different income groups. After that, this study analyzes the willingness to pay for higher education and financial resources for students' educational expenditure by using a college student survey data, which was collected in December This study finds that the main part of financial resources is coming from family while students are learning at college and the gap of willingness to pay among different income groups is becoming larger and larger with the increase in tuition. Finally, the conclusions are drawn and policy implications are discussed, aimed at informing enrollment projections and tuition policy choices in Chinese higher education system. Introduction Since the People s Republic of China was established in 1949, Chinese higher education has developed greatly. In 1949, there were 205 regular higher education institutions and about 116,500 full-time students. In 1998, the number of regular higher 1 education institutions reached 1,022 and full-time students number was about 3,408,800 head counts. According to the revised statistical indicator of Ministry of Education, the gross enrolment ratio of Chinese higher education was 10.5% in 1999 (Chinese Education Statistical Yearbook, 1999). Chinese regular higher education has expanded faster than before since In 1978 the admission number to regular higher education was about 400,000. In 1998, this number was about 1,080,000. The average annual increase rate was 6.11%. But in 1999, the admission number was about 1,560,000, which was 144% of that in In 2000, the admission number reached about 2,200,000, which was 141% of that in The expansion of higher education may continue in the following years (see table 1). In order to guarantee educational quality, higher education expansion needs adequate financial resources support. However, Chinese higher education has been facing serious financial tension since 1980s. In order to fill the gap between financial demand and supply, Chinese government and departments the concerned have been solving the problem through two paths: one is to enlarge financial resources and explore more channels to raise educational funds, the other is to improve the utilization efficiency of educational resources. Among many channels of fund-raising, cost-recovery policy implementation has been regarded as having theoretical support and practical value. Chinese higher education institutions were totally funded by the government before the implementation of opening-up and reform policy. Chinese central government announced a document of Decision to Reform Educational Structure in In this governmental document, the section related to tuition policy declared that higher education institutions could enroll a small number of student who pay tuition. In fact, a few of institutions have begun to put this soon-to-be policy into practice before From the mid 1980s to 1992, the two-track enrolment of paying tuition and no paying tuition existed at the same time in Chinese higher education system. The majority of students did not need to pay tuition and boarding fees. A small amount of students paid tuition and fees for higher education enrollment. In 1992, the State Education Commission of China made a proposal of one-track enrollment policy. Since 1993, more and more higher education institutions have changed the enrollment policy from 2 two-track to one-track. In 1997, each college student began to pay a certain percentage of recurrent expenditure per student for higher education. Cost recovery policy has been implemented in all regular higher education institutions in China. Tuition rose year by year. In 1997, average tuition in regular higher education institutions was 1,620 yuan, while average institutional recurrent cost per student was 8,350 yuan, and total enrollment number was about 3,170,000 in Chinese regular higher education institutions. In 1998, average tuition increased to 1,974 yuan, while average institutional recurrent cost increased per student to 11,020 yuan and total enrolment number reached 3,408,800. In 1999, average tuition reached 2,769 yuan, while average institutional recurrent cost per student increased to 14,400 yuan and total enrolment number was above 4,000,000. From Table 1, we can see that tuition increased 40.3% from 1998 to In 1999, the percentage of tuition to GDP per capita reached 42.4%, and the percentage of tuition to disposal income per urban resident was 47.3%, while the percentage to net income per rural resident exceeded 100% and reached 125.3%. Decision makers need to consider students and their families willingness and ability to pay for higher education when making decision to increase tuition and boarding charges. Lots of studies (Radner & Miller, 1975; St.John, 1990; Bershadker, 1998) found that private costs appeared to be a barrier to higher education. Apart from private educational costs, many other factors influence higher education enrollment decisions. The parental education and living environment may affect individual attitudes to higher education. The return to higher education may incite individual s expectation to higher education. The family s financial resources may affect one s ability to pay for higher education. This study conducts empirical analysis on the effect of financial incentives and family background on higher education enrollment choice, with particular focus on student response to private higher education cost. This study also examine the different private cost response among different income groups and compare the impact on individual enrollment choices of private cost, expected earnings and family background including parental education and financial resources. 3 Data The data set used in this study is a representative sub-sample of Urban Household Survey database of China, collected by Urban Survey Organization of State Statistical Bureau of China in August Courtesy of USO of SSB, we obtained access to a random and equidistance subsample from a sample pool, the data of which were collected from seven representative provinces. The seven provinces are Beijing, Zhejiang, Guangdong, Hubei, Liaoning, Sichuan and Gansu, which respectively represent areas with a high, medium or low level of economic development. The selected data set we obtained includes 7,636 households and 24,219 individuals. The working sample of this study is the college-aged males and females who have already graduated from upper-secondary schools. The working sample is divided into two groups: college group and non-college group. Individuals of college group are those who have graduated from upper-secondary school and are learning at higher education institutions. Individuals of the non-college group are those who have graduated from upper-secondary school but do not go to college. All individuals of the working sample are at the age of years old. The working sample has 911 valid cases after data has been cleaned. Model To examine higher education enrollment decisions, this study use individual choice model described by Becker (1990, p ). Assume that each upper-secondary school graduate (indexed by i) has two status choices following upper-secondary education: enrolling in higher education institutions (Y=1) or not enrolling in higher education institutions (Y=0). Hypothesize that student utility depends on at least ability, family income, education quality and cost of student s college options and other chance factors. The utility of the work/unemployment option may depend on the student s potential wage and chance of unemployment as well as other chance factors. Although utility cannot be observed, we could observe the individual behavior: enrolling in higher education or not enrolling in higher education (work/unemployment). 4 Assume that the unobservable utilities are related linearly to their explanatory variables and chance factors. The individual utility function of enrolling in higher education is as follows: u ρ = β...(1) 5 + β ε 1, * i1 X i i The individual utility function of not enrolling higher education (work/unemployment) is as follows: u ρ = β...(2) 3 + β ε 0, * i0 X i i Where the β s are unknown parameters, the X s are explanatory variables which may influence individual utility, the ε s are random unobservable factors. The individual probability function of enrolling in higher education is: * * * Pi = Pr ob( ui 1 ui0 0) = Pr ob( u 0) = Pr ob( Yi = 1) ρ ρ = Pr ob( β 0 + β1x + ε 0) = Pr ob[ ε ( β 0 + β1x )] ρ ρ = 1 Pr ob[ ε ( β + β X )] = 1 F[ ( β + β X )]...(3) Where F is a cumulative distribution function. If the distribution of ε is symmetric, ρ then = F β + β )...(4) P i ( 0 1X Assume that the random component ε is distributed logistically. The probability of enrolling in higher education could be written as Zi e ρρ Pr ob( Yi = 1) =, Z X Zi i = β 0 + β, (5) 1+ e or 5 ln 1 P i P i ρρ = β + βx (6), 0 Where Prob ( Y i =1) and P i are the probability of enrollment in higher education, X ρ is the vector of independent factors. Logistic regression techniques can be used to estimate the parameters of formula (5) or (6). Variables The model of this study is a logistic distribution one. Dependent variable is individual status of enrolling in higher education or not. If one has been enrolled in higher education, the dependent variable equals to one. If one chooses to enter labor market, the dependent variable equals to zero. The following describes explanatory variables: (1) Expected costs: There are two types of college costs: direct costs and indirect costs. Direct costs include tuition, boarding fees, textbooks, additional transportation and living costs for schooling. Indirect costs are earning forgone, which is given up by individual for enrolling in higher education. Usually, some students can get financial aids during their learning periods. The financial aids should be subtracted from college costs of attendance. In China, there are six kinds of financial aids for students: scholarships, grants, loans, specially targeted grants for financial emergency, reduction or exemption of tuition and fees, and work-study. Because loans should be returned after students graduate from college, a particular amount of 40 percent of loans has been calculated in financial aids. It is unfortunate that there is no financial aids information in Urban Household Survey. In order to solve the problem, we did a College Student Survey in Beijing in Dec The College Student Survey data set has 3,721 individuals, random sampling from students of three four-year universities, three two-or-three-year 6 colleges, eight vocational colleges and three people-run postsecondary education institutions. We obtained an average financial aid of yuan per student each academic year from this data set, and subtracted it from the total college costs. (2) Expected lifetime earnings: According to human capital theory, more years of learning will lead to higher earnings. Assume that lifetime earning expectation is an important financial incentive to one s decision to enroll in higher education. In theory, the ideal method to obtain one s lifetime earnings is to make a longitudinal study to get the individual realized lifetime earnings following his graduation from college. Meanwhile, such information as individual realized costs, educational achievements and learning ability and so on can be obtained through longitudinal studies. In fact, because of limited time and funds, it is hard for researchers to make a large-scale longitudinal study in China (Chen & Min, 1998). Therefore, cross-sectional data have to be used to take the place of historical data. Although there may exist systematic errors, it is still a kind of calculation method. Supposing that an individual student and his family would like to make decision by referring to the current situation of labor market, we obtain individual expected lifetime earnings by combining the realized earnings of different ages with the same education background. (3) Parental education: In general, education background of one s parents may influence the individual attitudes towards higher education. Parents who obtained more years of learning may pay more attention to children s education, build and facilitate a better learning environment for children at home, and have higher willingness to pay for higher education. This study divides individual parental education into six brackets ranging from elementary school to four-year higher education. (4) Parental residence: Before the implementation of the reform and opening-up policy in 1978, labors had no opportunity to move their work sites from rural to urban areas. Basically, residents who had rural residences should only work in rural areas. In 1980s, the economic condition was becoming better and better. 7 Since the end of 1980s, a small number of rural labors have flown into urban areas to find employment opportunities. Although more and more rural residents are flowing into urban areas, there still exists some degree of inequality of educational opportunities. Table 2 reports the difference in enrollment opportunity between rural and urban students, coming from the data set of College Student Survey. (5) Socioeconomic status of one s family in terms of income backgrounds: The individual family income backgrounds may affect the ability to pay for higher education. This study adopts seven basic brackets divided along monthly expenditure per capita (in yuan). Table 3 is the consumption groups, which may substitute income groups and table 4 is the descriptions of selected variables in this study. Calculation of Cost and Earning Expectations Cost Expectations Divide the working sample into two groups: college and non-college sub-samples. For college student group, we cannot get their realized indirect costs from the survey data directly. For non-college subject group, we cannot get their realized direct costs from the survey data directly. In order to solve this problem, this study estimates expected direct costs for non-college subject group and expected earning forgone for college student group. Total private costs are equal to direct costs plus earning forgone. Assuming that indirect (direct) cost expectations of college (non-college) subsample are a linear function of individual and family characteristics and a random unobservable component, the cost equations could be described as the follows: Opp cost = α + α M (7) 0 1 γ DirectCost = µ N µ 1 γ 2...(8) 8 Where α 0 and µ 0 are constant components, γ 1 and γ 2 are random unobservable components, M and N are independent variables. Through regressing realized earnings of non-college sub-group on the characteristics in M, the estimates of α 0 and α 1 are achieved and used to compute the opportunity costs of college student sub-group. Through regressing realized direct costs of college student sub-group on the characteristics in N, the estimates of µ 0 and µ 1 are attained to compute the direct cost expectations of non-college sub-group. Finally, the total college costs of individuals are obtained by direct plus opportunity costs. Earning Expectations Considering the lifetime earning expectations, assume that individual lifetime earnings expectations are a linear function of exogenous characteristics and a random component. Earning = θ L θ1 γ 3...(9) Where the θ s are unknown parameters, γ 3 is random unobservable components, L is independent variables including regional factor and individual background. Use the selected data set of Urban Household Survey, which includes 7,636 households and 24,219 individuals. In order to see more clearly the economic value of education in recent Chinese labor markets, this study selects a typical sub-sample of youth labors with upper-secondary school education and above and at the age of forty years old and below. Excluding missing value of education information and abnormal cases, the selected sample has 4,498 valid cases. The selected sample is divided into four groups: four-year higher education graduates, two or three-year college graduates, secondary professional education and upper-secondary school graduates. Regressing earnings on the independent variables for each group by equation (9), the authors obtain θ 0 and θ 1 to compute expected earnings for each individual of each group. 9 Self-selection Bias Here, we need to discuss self-selection bias that might be raised by OLS techniques. If the reason for college students to enroll in higher education institutions is just because their salaries will be lower than non-college sub-group, and the reason of non-college students not to enter higher learning institutions is just because their direct costs will be higher than college student sub-group, the results obtained by OLS regression will overestimate cost expectations of college sub-group and underestimate cost expectations of non-college sub-group. Actually, the population of applicants for enrolling higher education is much larger than that of being admitted to colleges and university. Averagely, only one student can be admitted to higher education institutions from every two and a half applicants (Li and Min, 2000a). In order to correct self-selection bias, many scholars and researchers adopt switching models to replace OLS models. However, there is still doubt switching models. Lewis (1986) compared lots of research results by using OLS models and switching models, and found that switching models of correcting self-selection bias raised the variation of estimates. He thought that the correction methods made calculation more complicated and calculation procedure made results unstable. Therefore, our study does not use correction methods but OLS techniques to obtain cost and earning expectations. Table 5,6,7 and 8 are cost and earning expectations functions and regression results. Findings This study obtains private demand function for higher education by logistic regression techniques. The results are in table 9. Data in table 9 indicates that private costs, expected earnings, individual preferences and tastes in terms of family background are important determinants to higher education enrollment status. Among the listed independent variables, private costs, expected earnings, parental education and family income background significantly influence private demand for higher education. The coefficients of the four independent variables are statistically significant and have the expected signs. From table 9, it appears that private costs are negatively correlated with individual enrollment status in higher education institutions. However, the expected 10 earnings have statistically significant positive coefficients on individual enrolling in higher education. It means that the probability of individual choice to enroll in higher education is negatively correlated with the expected costs, but positively correlated with the expected earnings. Apart from this, parental education and family income background also have positive effects on individual choice to enroll in higher education. It implies that parents with higher education or higher earnings intend to build better learning environments for children. They have stronger willingness t
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!