Measuring education inequality in primary schools of India: A district level analysis Chandrashekhar 1 & Bedanga Talukdar 2 The Government of India is keeping no stone unturned to provide access to elementary education to each child till fourteen years of age. Considering the nature of diversity and vast size of the country, noteworthy achievements have been arrived at. Nevertheless, the access to primary education for all has been a distant dream. Particularly, the education gap in various administrative regions has been staggering and needs immediate attention. Distribution of education services by means of upgrading school infrastructure to all districts of India is a robust approach to ensure access to elementary education to all. The present study aspires to develop a measure for education inequality at the district level to quantify the gap in education attainment. At the same time, the study endeviours to look into quality of education with respect to education inequality. The quality of education can be measured with quality of students, quality of teachers and quality of school infrastructure. We wish to see education inequality in the primary education in each district of India with respect to quality of student, teacher and school infrastructure. This is also one of the approaches to test the robustness of education inequality. Data and methods: The National University of Education Planning and Administration (NUEPA) provides data on various aspects of elementary education which includes enrolment-based indicators, school-based indicators, teachers-related indicators and facility based indicators, for all the 644 countries across the country. The NUEPA data is popularly known as District Information System for Education (DISE). The present study uses this data for the analysis of education inequality and other school related indictors. In addition, the Census of India, 2011 has been used to calculate the total population in each age group. Methodologies: To find out the education inequality, education Gini has been calculated with following technique. The formula for the calculation of Gini has been adopted, with little modification, from a study by Thomas, Wang and Fan in 2000. Methods for calculating education Gini: For the calculation of the education Gini, proportion of education in different grades have been taken from the DISC data and proportion of population in the respective age group have been taken from the Census of India, 2011. Following formula is used to calculate education Gini, E = ( 1 μ ) ( N N 1 ) ( n i=2 i=1 p i y i y j p j ) j=1 After expanding the above equation, we have E = (1/μ) {p2(y2-y1)p1 + p3(y3-y1)p1 + p3(y3-y2)p2 + p4(y4-y1)p1 + p4(y4-y2)p2 + p4(y4-y3)p4 + p5(y5-y1)p1 + p5(y5-y2)p2 + p5(y5-y3)p3 + p5(y5-y4)p4} where, 1 PhD Scholar, International Institute for Population Sciences, Mumbai, India. Email: shekhar_tanman@yahoo.com 2 PhD Scholar, International Institute for Population Sciences, Mumbai, India. Email: bedanga.talukdar@yahoo.com
E = Education Gini based on education attainment distribution, small population μ = average years of schooling pi and pj = proportion of population with certain level of schooling, here 1 st to 5 th standard of schooling yi and yj = years of schooling at different educational attainment levels, here it is 1 to 5 years of schooling n = number of levels in school attainment data; and, n=5 for this study N = Population size Gini is sensitive to population size (N) if the population size is too small; in this case the districts of India. The sensitivity is reflected by a factor of (N/N-1). In addition to calculate education inequality through education Gini, composite index for quality of student, quality of teacher and quality of school infrastructure have also been calculated. Choice of indicator and unit of analysis: The choice of relevant functioning and capabilities for any measure is a value judgment rather than a technical exercise (Sen, 2008). It needs mention here that selection of indicators has been carried out after going through extensive literature, considering the robustness of the indicator and value it adds to the development of the composite index. Student: Four indicators have been use to calculate the composite index of quality of student namely, net enrolment rate (NER), transition rate (TR), gender parity index (GPI) and survival rate (SR). Composite index for quality Teacher: To assess the quality of teachers, again four indicators have been selected; these are pupil teacher ratio (PTR), percentage of female teacher in the school, percentage of teachers received in-service training and percentage of graduate teachers. school infrastructure: To find out the quality of school infrastructure, we have used six indicators namely, percentage of puckka (bricked-wall) school, percentage of school with girl toilet, percentage of school with drinking eater facility, percentage of school with mid-day meal facility, percentage of school with electricity connection and percentage of school connected to all weather roads. Since the ranges of the values of indicators are large therefore, we have we have used the following formula to transform a variable into unit free index between 0 and 1. The formula used is as follows Di = (x min)/(max min) Here Di is the dimension index and x is value of the specific dimension. After the calculation of Di, the arithmetic mean has been calculated to form the composite index for each dimension. Results and findings: The results demonstrate that there is large scale variation both at the inter-state and intra-state in terms of education inequality at the primary level. It is found that the coastal districts of Kerala and Tamil Nadu have lowest education inequality at the primary level (Figure 2). On the contrary, the highest education inequality has been found in districts of Bihar, Uttar Pradesh and some districts of north eastern states. The southcentral districts of India have modest education inequality; whereas, the north Indian districts, coastal districts of Odisha and Telangana and districts of northeast India have relatively high education inequality. 2
This variation is by and large similar with quality of student, quality of teacher and quality of school infrastructure. We have found noteworthy similarity between education inequality at the district level and quality of student, quality of school infrastructure and quality of teacher. Thereby, school quality indicators are found to be high in districts where there is high education inequality at the primary level. To find out the association between quality indicators and education Gini, we carried out Pearson s correlation test of association. The result of the correlation illustrates that there is high correlation between education Gini and quality of student. For the other indicators like quality of teacher and quality of school infrastructure the association is positive however, the magnitude of the association is little less than that of quality of students (Table 1). The improvement in the education equality can be shown by shifting the education Lorenz curve for two districts (Figure 1). In the case of Paschim Champaran district, Bihar, the education Lorentz curve is steep and located away from the egalitarian line, resulting into a large education Gini. On the other hand, the Lorentz curve for Kasaragod district, Kerala, the education Lorentz curve shifted much closer towards the egalitarian line. This indicated that despite progress in expanding the primary education in districts of Bihar, more than half of the population in the early age did not receive primary education. However, in Kasaragod, Kerala, the Government of Kerala expanded the primary education much early which resulted into more enrolment of children in the primary schools. Table 1. Pearsons correlation coefficient for education Gini with composite index of quality of student, teacher and school infrastructure. Education Gini Composite index for quality of students Composite index for quality of school infrastructure Composite index for quality of teacher Education Gini 1 students 0.7006 1 school infrastructure 0.4703 0.397 1 teacher 0.3653 0.3544 0.6258 1 Figure 1. Lorentz curve for Paschim Champaran, Bihar and Kasaragod, Kerala, 2010-11 Conclusions: 3
There is wide variation in education inequality at primary education level in both inter-state as well as intrastate. The intra-state variation is more prominent in north India than in south India whereas inter-state variation is more visible in north Indian state. The education inequality is high in the south Indian states and relatively moderate in central part of India. The Lorentz curve clearly demonstrates the performance of districts with respect to education inequality; simultaneously, it reflects the gap in primary education at the district level. The findings of the study can be used to identify poor performing districts in terms of primary education. Figure 2. Map of India showing education Gini, student related indicator, teacher related indicator and school infrastructure related indicator at the district level, 2010-11. 4
Note: References have not been included because of page limitation. 5