By Ankit Vyas
Low learning outcomes in Indian schools cause dropouts and prevent students from acquiring the skills needed to be a part of India’s economy. Teacher absenteeism and poor quality of in-service teacher training have a negative impact on learning outcomes. Another aspect that affects learning outcomes is the curriculum which burdens students with a large amount of information, most of which is not connected to their experiences and hence is not engaging for them. To improve outcomes, remediation programs by teachers as well as volunteers have been found to be effective. Training programs based on teachers’ needs are more effective and have a positive impact on learning outcomes. Better learning outcomes would increase retention rate and enable more people to contribute productively to the economy.
Introduction to the issue
The Indian education system is free and compulsory for grades 1-8. Education is also a fundamental right, through the Right to Education Act, 2009. Sarva Shiksha Abhiyan, the flagship program of the Government of India for Universalisation of Elementary Education has been successful in increasing enrolment, which now stands at 96% at the primary level. (Annual Status of Education Report, 2012) However, the enrolment rate has not translated into positive learning outcomes for the students. ASER (2012) found that 54% of grade 5 students surveyed were unable to read a grade 2 passage. In a separate study, the Planning Commission (2010) found that only 41.7% of students at a second grade level were able to read alphabets in their local language. Poor learning outcomes are not restricted to literacy. 29.1% students in grade 5 were unable to do subtraction, according to ASER’s 2010 study. Gupta, Williams and Leslie argue that the prevalence of rote learning methods across India, which focus on content memorization, is linked to these poor outcomes. Further, the National Curriculum Framework (2005) holds that content in the curriculum does not engage most students, especially those who come from socio-economically disadvantaged and from marginalized communities. This is because the curriculum is not based on and does not include the lived experiences of students. This has a negative impact on learning outcomes.
Wright, Horn and Sanders (1997) posit that teacher effectiveness is the most important determinant of learning outcomes in the classroom. In India, teacher effectiveness is impaired by two factors, absenteeism and lack of training. Kremer, Muralidharan, Chaudhury, Rogers (2005) found that 25.2 per cent or roughly one in four teachers were absent in rural areas. This absence rate reflects the absence rate from the school, as opposed to the classroom. Actual absence from teaching duties could be considerably higher. (Rogers & Vegas, 2009). Studies show that teacher absence does affect learning outcomes negatively. (Miller, Murnane and Willet, 2007)
A lack of in-service training exacerbates the issue. Uttar Pradesh, the largest state in terms of population and amongst the poorest, trains only 12.5% of its teachers. At an all India level, close to 60% of all Indian teachers receive little or no training at all, throughout their career. (DISE, 2011) In addition, the quality of in-service training leaves much to be desired and does not cater to the needs of the teacher. Instead, a universal approach to training is taken. (NCERT, 2012) Poor quality of teacher training has been found to affect learning outcomes negatively. (Aggarwal, 2000)
Evidence on importance of issue
Studies attribute much of the failure in higher grades and dropouts to poor performance in Math and English, two subjects whose outcomes at a primary level are below par and leave students incapable of picking up higher level concepts. (NCF, 2005). Retention at the elementary level is only 38.37% in states where elementary is I-VIII. (SSA 16th Joint Review Mission, 2012). Out of the 40% who make it to upper secondary, it narrows down to a meager 18% at tertiary level. (World Bank, 2010). The low quality of education ensures that even vocational education is beyond the reach of these students. Bhatt (2013) argues that the opportunity cost of two-thirds of India’s children not completing primary education comes to $100 billion per year.
In a study, Sikdar and Mukherjee (2007) found that 44% males and 36% females cited lack of quality in education as the reason for dropping out. At a secondary level, this figure was still significantly high, wherein 41% males and 30% females cited lack of quality as the reason for dropping out. McKay and Mills (2004) argue that the experience in developing and developed countries of East-Asia makes it clear that no macro-economic policy can succeed without investment in quality education and training. The educational achievement of a nation’s youth also has a direct impact on GDP. Hanushek (2008) found each additional year of schooling to increase the 40 year growth by 0.37 percentage points. This translates into a 10% growth over a 40 year period.
According to the MHRD report (2012), India needs to develop 500 million skilled workers by 2022. Low enrolment in vocational training, which stands at only 3% of the population, indicates that the current growth of skilled workers will be insufficient to meet the needs of the Indian economy in a decade. (National Skill Development Corporation, 2008) The evidence clearly indicates the need for definite measures to improve quality of learning outcomes.
Literature Review Curriculum
The National Curriculum Framework (2005) reviewed the curriculum and made a few observations and recommendations.. One theme that resurfaces frequently in it is that of knowledge not being linked to real life.(NCF 2005) The experiences of the students, especially of those from socio-economically disadvantaged backgrounds are not reflected in the curriculum and hence not considered important enough to be discussed in the classroom by the teacher. Therefore, students are not engaged by the curriculum as it is disconnected from the experiences they have had and from the knowledge that they already possess. UNICEF (2000) also stresses on the need for the curriculum to take local context into consideration. A well, organized curriculum that allows teachers to adapt to students needs, increases learning outcomes. (Lockheed & Verspoor, 1991).
The curriculum not being linked to real life also affects the ability to read, negatively. Students are unable to engage with the text and extra reading is not encouraged. Children are encouraged to read only what is being tested. (NCF, 2005) The MHRD (1993) report on education “Learning without Burden” made some observations that are still relevant today. The crux of the problem, according to the report was the curriculum load, a point which was reiterated by the NCF in 2005 as well. This essentially meant that a lot was taught but little was understood. (Learning without Burden, 1993)
NCF(2005) links the school curriculum and examinations to a lack of quality. According to the report, the curriculum overburdens the students with information and the examinations further exacerbate this by forcing students to memorize the content, rather than engage with it. A rigorous academic regime is imposed on the students by the teachers who are constrained by the emphasis on completing the syllabus, which is considered an end in itself. (Learning without Burden, 1993). With exit examinations becoming even more important as determinants of economic success, the emphasis on completing the syllabus has increased even more. (NCF, 2005) For first-generation learners, the curriculum burden is even harder to bear, due to a lack of support and might be one of the reasons for poor outcomes and eventual dropout.
The classroom context, in which children learn in India, was examined by the NCF report. It shows that participation of children in class in terms of asking questions or discussing is discouraged. There are no incentives or opportunities to take initiative. Further, the NCF report illustrates that students are expected to reproduce textual knowledge, in a format required by the teacher. (NCF 2005) Predominantly, classrooms are teacher-led and no time is scheduled for independent activities by students.
Studies indicate that student-led classrooms lead to better learning outcomes. Jones looks at a student- led classroom as a place where the students are active participants in their learning and design their own learning goals and activities. (Jones, 2007) The simplest form of student-led learning is where short periods of student activity are scheduled into the lecture. Ruhl, Hughes and Schloss (1987) illustrated this with a simple study involving two classrooms at a college level, one which learnt through the traditional lecture method, without a break during a class of 45 minutes while the other was allowed three breaks of 2 minutes each in a 45 minute lecture. During this break, the students were allowed to discuss and compare notes with their partner. At the end of the class, students were asked to write all that they could remember from the class, in three minutes. The class which had breaks had an average of 108 correct facts as opposed to 80 correct facts in the classroom with the traditional method. A second evaluation was done through a multiple-choice test. The class which had breaks had an average score of 89 as compared to 80 in the other class. (Ruhl et al, 1987) However, this study involved college students, which affects its reliability in predicting impact of student-led learning at an elementary level.
In Indian classrooms, the knowledge source and generation is restricted to the classroom, teacher and textbook. The NCF (2005) encourages the application of the constructivist method wherein students are encouraged to construct knowledge for themselves by asking questions, discussing and linking knowledge back to their experiences. In this process, the teacher’s role is to facilitate the process of knowledge construction by encouraging discussion. Rajendra Kumar conducted a study on the effectiveness of constructivist methods in teaching math in grade 5. Three different schools were selected for the study. In each school, one section was chosen as the experimental group and taught through constructivist methods while the other section was taught through the traditional method. The constructivist methods allowed students to interact freely with each other to co-operate and to make sense of how the knowledge was linked to their real life experiences. The traditional method involved passive learning through the teacher’s lecture, the form that is most prevalent in India. The effectiveness of the constructivist method was measured through the use of pre-tests and post-tests. During pre-testing, no significant difference in achievement in mathematics was found in the two classes. The post-test revealed that the mean score of the experimental group (36.01) was significantly greater than the control group. (24.70). This supports the National Curriculum Framework’s assertion that constructivist teaching leads to better learning outcomes and is one aspect that should be integrated into the curriculum for better implementation. Training teachers on these methods is crucial to build their capacities as facilitators of this process.
Studies show that a significant correlation exists between teacher training, especially in-service training and student performance. (Kingdon, 2006)(NCERT, 2012). Under the National Policy of Education, 1986, District Institutes of Education and Training (DIET) were set up in all the districts of the country. The purpose of the DIET was to provide in-service and pre-service training programs to teachers. The National Teacher Education Curriculum Framework found that out of 571 districts that DIETs were set up in, 529 were functional. (NCTE, 2009.) However, only 40% of primary school teachers receive in- service training. (DISE, 2011-12). Looking at the number of DIETs set up, a lack of institutions does not seem to be the cause behind this. All the DIETs are supposed to be staffed by a team of lecturers for providing in-service training to teachers. However, NCTE (2009) data indicates that a lack of qualified faculty in the DIETs severely reduces their capacity to provide training, which could point towards the deficit of trained teachers in the country.
Existing teacher training programs in India have been found to focus on just transmitting information. Even sessions on innovative pedagogies such as child-centered learning are delivered in the traditional lecture method. The role of the teacher, conveyed through these trainings is that of a source of knowledge, rather than that of a facilitator.(NCF, 2005) The National Council of Teacher education developed a framework for in-service training for teachers in 2009. It called for more interaction and experience sharing, along with designing programs based on teacher needs. It also stressed that the content of training be linked to teacher experiences and classroom realities, as opposed to a universal approach. A crucial aspect that NCTE touched upon was the need to monitor the effectiveness of the training in terms of outcomes in the classroom. A common theme across the recommendations of NCTE is that all training content and approaches should be based on classroom needs of teachers and their experiences. (NCTE, 2009)
Ramachandran, Bhattacharjea and Sheshagiri (2008) assert that currently, teachers’ needs are not considered nor are they involved in the designing of training programs. Follow-up is not conducted to evaluate effectiveness. The key issue here is that implementation hurdles are not considered and this makes teachers lose interest in the trainings as they cannot link them back to the realities they face in their classroom everyday.
Evaluation of Interventions to improve learning outcomes
Pratham, an educational NGO began the Read India program on a national scale in the year 2007 to improve learning outcomes of rural and urban children in basic reading and math skills. Specific groups of students were targeted and for them, clear learning goals were set, appropriate teaching-learning activities and materials were used and they were organized on the basis of their ability level. (Banerji & Walton, 2011) J-PAL South Asia conducted an evaluation of the Read India program. It utilized a randomized control trial to evaluate the program, where it looked at three alternative evaluations. The first alternative included teacher training, monitoring and support, supplementary learning materials for children and village volunteers, who worked after school hours to provide extra attention and teaching to children who were lagging behind. The second alternative included teacher training, monitoring and support, supplementary learning materials for children but no volunteers. The third intervention included only distribution of supplementary learning materials to schools. Additionally, the control group did not receive any intervention. The evaluation assessed learning outcomes of students through the three different alternatives. (Banerji & Walton, 2011)
It was found that that intervention which had teacher training, monitoring, materials and volunteers showed significant improvement in learning outcomes. For the interventions dependent purely on teachers and materials, there was little or no improvement. Outcomes in the control group were the worst. 50% of the students in the control group had no change in learning levels during the entire duration of the study period. (Banerji & Walton, 2011) The effectiveness of volunteers offers the government a policy alternative to aid the teachers, especially in understaffed schools in the country. In this context, the Shiksha Karmi project represents a reference point. The Shiksha Karmi project was started in areas of Rajasthan that were educationally backward and where schools were non-functional because of teacher shortage and absenteeism. This project involved local communities in identifying and recruiting local teachers from the community to teach in the school. (Ramachandran & Sethi, 2000)
Similar to the Read India program teachers, Shiksha Karmis worked in the spirit of voluntarism and were provided training by an external agency. Srivastava and Jain found that the learning outcomes in SK schools compared favorably with formal schools. (Srivastava & Jain, 1995). Connecting this to J-PAL’s evaluation on the impact of volunteers, the idea of Shiksha Karmis could be transplanted even to schools which have regular teachers, in a bid to improve learning outcomes.
Another study to analyse learning outcomes, was conducted on evaluation of teacher-led summer camps where the purpose was to remediate and teach students according to their learning levels. The students showed significant gains during these summer camps. (Banerji & Walton, 2011) What is implicit in this study is that teacher effectiveness increased substantially in an environment outside school that was free from constraints such as completing the syllabus, which has been cited as being one of the parameters that teachers are judged on. Banerji & Walton (2011) analyse that the teacher-led summer camps were effective also because of the grouping of children based on levels, with well- defined goals for each. Teachers need to be sensitized to de-linking the completion of syllabus to achievement of learning outcomes. This view is also expressed in the NCTE framework which talks about the need to move away from syllabus-based achievement testing. (NCTE, 2009)Integrating the process of teacher desensitization to syllabus completion within in-service training programs will make implementation more effective.
Addressing Teacher Absenteeism
In a study in India, Kremer et al. (2005) found that 25% of teachers were absent and only half were found teaching. Miller, Murnane and Willet (2007) have shown that teacher absences are directly linked to poor learning outcomes. Teacher absence was found to be more related to daily incentives to attend work. These included recency of inspections, infrastructure in school as well as closeness to a paved road. There was little evidence to suggest that attempting to strengthen school community ties will reduce absence. (Kremer et al., 2005) School Management Committees were formed, under the RTE, for the very purpose of strengthening school-community ties. (American India Foundation, 2009). However, the mere existence of school management committees is not sufficient for controlling teacher absence. Evidence shows that a lack of awareness about teacher absenteeism impairs SMCs from taking action to resolve the issue. A study on the functioning of SMCs, conducted in 3 clusters of Madhya Pradesh and Chattisgarh, found that 73% of the parents were unaware of teacher absenteeism in school. (Bandyopadhyay & Dey, 2011) Banerjee and Duflo conducted a randomized experiment using impersonal monitoring wherein a teacher was given instructions to take a picture of himself and the students every day at opening and closing time. The camera had a tamper-proof date and time function. Further, teachers were offered a bonus based on number of days actually attendend. In the comparison schools, teachers were just told that they could be dismissed for poor performance. There was no other monitoring mechanism. In the treatment schools, absence rate was cut to half, from 36% in the comparison schools to 18% in the treatment schools. Because of fewer absences, treatment schools were found to have taught for one more month than a comparison school. The study also conducted a cost-benefit analysis that found that the program was not too expensive, in principle, to be scaled up to the entire school system. The program cost only $6 child for an increase in over 30% of the number of days per child. More importantly, the study found that test scores in the treatment schools were 0.17 standard deviations higher than in the comparison schools. (Banerjee & Duflo, 2006). However, this conclusion should be treated with caution as studies show that teaching time by itself is a poor predictor of student achievement. Rather, it is effective use of time which is a more accurate predictor. (Reimers, 1993)
Evidence shows that remedial programs led by volunteers have had a significant impact on improving learning outcomes. A study by Banerjee examined the impact of an after-school program run by youth volunteers. These youth volunteers were provided a week’s training and then asked to conduct after- school reading camps. The study found that a child in the intervention village was 60% more likely to read than a child in a control village. (Banerjee, Banerji, Duflo, Glennerster, and Khemani, 2010)While remedial programs are successful in increasing student learning outcomes, remediation at the classroom level itself will maximize the effectiveness of classroom time. To achieve this objective, the Shiksha Karmi project should be extended to even areas where there are functional schools and teachers who come regularly. Here, the Shiksha Karmis should volunteer during school time and help the teacher with remediation.
A curriculum that does not engage students and increases their academic workload is an impediment to positive learning outcomes. There is a need to adapt the curriculum to meet the children’s needs and to teach them at their level. Following the National Curriculum’s framework’s recommendations, the curriculum development should be decentralized to contextualize the curriculum to meet the children’s needs. The academic workload needs to be reduced and more focus needs to be put on inclusion of constructivist activities in the curriculum. For teaching in a constructivist manner, the capacities of teachers will have to be built. As per the NCTE and NCF, regular in-service trainings, which are directed at the teacher’s classroom needs will be effective in increasing learning outcomes. For regularity to be increased, the institutional capacity of the District Institutes of Educational and Training needs to be augmented by hiring more qualified faculty to design and deliver the trainings to the teachers. Involving the teachers in the design of the training programs will increase the relevance and effectiveness of these trainings in developing competencies of the teachers to teach in the classroom, thereby improving learning outcomes. Teacher absenteeism has been a chronic problem in the Indian educational system that has for long affected the learning outcomes of children negatively. Increasing awareness amongst the School Management Committees about this issue could prove to be an effective method of increasing accountability and improving teacher attendance, in order to improve learning outcomes.
1. ASER.( 2012) Annual Status of Education Report (Rural). Retrieved from http://img.asercentre.org/docs/Publications/ASER%20Reports/ASER_2012/fullaser20 12report.pdf
2. Planning Commission. (2010) Evaluation Report on Sarva Shiksha Abhiyan. Retrieved from http://planningcommission.nic.in/reports/peoreport/peoevalu/peo_ssa2106.pdf
3. National Curriculum Framework (2005)NCERT. Retrieved from http://www.ncert.nic.in/rightside/links/pdf/framework/english/nf2005.pdf
4. Wright, S.P. , Horn , S.P. and Sanders, W.L.(1997) Teacher and Classroom context effects on student achievement: Implications for Teacher Evaluation, Journal of Personnel Evaluation in Education 11: 57-67. Retrieved from http://www.sas.com/govedu/edu/teacher_eval.pdf
5. Kremer, M., Muralidharan, K., Chaudhury, N., Hammer, J., Rogers, F.H. (2005). Teacher Absence in India: A snapshot. Journal of the European Economic Association April–May 2005 3(2– 3):658–667. Retrievedfrom http://www.teindia.nic.in/Files/Articles/Articles_23feb12/jeea_teacher_absence_in_ind ia.pdf
6. Rogers, F, Vegas, E. (2009). No More Cutting Class? Reducing Teacher Absenteeism and Providing Incentives for Performance. World Bank Policy Research Working Paper. Retrieved from http://www- wds.worldbank.org/servlet/WDSContentServer/WDSP/IB/2009/02/26/000158349_2 0090226142341/Rendered/PDF/WPS4847.pdf
7. Miller, R., Murnane, R., Willet, J. (2007)Do Teachers Absences impact student achievement? Longitudinal evidence from one urban school district. National Bureau of Economic Research, 1 Retrieved from: http://www.nctq.org/nctq/research/1190901594955.pdf
8. DISE. (2011). Elementary Education in India: Progress towards UEE Analytical Tables 2011 2012. Retrieved from http://www.dise.in/Downloads/Publications/Publications%202011- 12/Analytical%20Tables%202011-12.pdf
9. NCERT. (2012). Impact of In-service Teacher Training on Classroom Transaction. Retrieved Fromhttp://www.ncert.nic.in/departments/nie/dtee/publication/print_material/SSA%20INSET.pdf
10. Yash Aggarwal. (2000). Quality Concerns in Primary Education in India. Where is the problem?DISE, 9. Retrieved from http://www.dise.in/Downloads/Reports&Studies/Quality%20Concerns%20in%20Pri mary%20Education.pdf
11. Sarva Shiksha Abhiyan (2012). 16th Joint Review Mission Results Monitoring Framework. Annex-3. Retrieved fromhttp://www.rteforumindia.org/sites/default/files/Results%20Framework_16th%20JRM_AM.pdf
12. World Bank. (2010). Retrieved from http://data.worldbank.org/indicator/SE.TER.ENRR
13. Winthrop, R, Bulloch, G, Bhatt, P, Wood, A. (2013). Investment in Global Education: A strategic imperative for business. Retrieved fromhttp://www.brookings.edu/~/media/research/files/reports/2013/09/investment%20 in%20global%20education/investment%20in%20global%20education%20final%20% 20web
14. Sikdar, S., Mukherjee, A. (2012). Enrolment and Dropout Rate in School Education. Economic & Political Weekly. Retrieved from: http://www.epw.in/commentary/enrolment-and- dropout-rate-school-education.html
15. McKay, M. and Mills, G (2004) ‘Education Lessons: East Asia’s Economic Transformation’ in P. Govender and S. Gruzd (eds) Back to the Blackboard: Looking Beyond Primary Education in Africa, NEPAD Policy Focus Report 1, Braamfontein, South Africa: The South African Institute of International Affairs.
16. E. Hanushek et al. (2008). Education and Economic Growth,” Education Next 8, no. 2, 66 Retrieved from http://hanushek.stanford.edu/sites/default/files/publications/Hanushek%2BJamison%2BJamison%2BWoessmann%202008%20EdNext%208(2).pdf
17. Ministry of Human Development, Government of India (2012-13) Annual Report. Retrieved from http://mhrd.gov.in/sites/upload_files/mhrd/files/AR_2012-13.pdf
18. National Skill Development Corporation. (2008). Human Resource and Skill Requirements in the Education & Skill Development Services Sector (2022)- A Report, 6,Retrieved from http://www.nsdcindia.org/pdf/education-skill development.pdf
19. UNICEF. (2000). Defining quality in education. Working Paper Series,11. Retrieved from http://www.unicef.org/education/files/QualityEducation.PDF
20. Lockheed, Verspoor. (1991). Retrieved from http://courses.umass.edu/educ870/teacher_education/Documents/EQ1%20Motiv%20 Tchr%20Quality%20-%20Leu.pdf
21. Learning without burden. (1993) Report of the National Advisory Committee. Appointed by the Ministry of Human Resource Development. Retrieved from Http://www.teindia.nic.in/Files/Reports/CCR/Yash%20Pal_committe_report_lwb.pdf
22. Jones, L. (2007). The Student-Centred Classroom, 2. Retrieved from http://www.cambridge.org/other_files/downloads/esl/booklets/Jones-Student- Centered.pdf
23. Ruhl, K. L., Hughes, C. A., & Schloss, P. J. (1987). Using the pause procedure to enhance lecture recall. Teacher Education and Special Education, 10, 14-18, 3. Retrieved from http://www4.ncsu.edu/unity/lockers/users/f/felder/public/Papers/Prince_AL.pdf
24. Nayak, R. A Study on Effect of Constructivist Pedagogy on Students’ Achievement in Mathematics at Elementary Level. Retrieved from http://www.ncert.nic.in/pdf_files/Rajendra%20Kumar%20Nayak.pdf
25. Kingdon, G. (2006). Teacher Characteristics and student performance in India: A pupil fixed effects approach. Retrieved from http://economics.ouls.ox.ac.uk/14026/1/gprg-wps- 059.pdf
26. National Curriculum Framework for Teacher Education. (2009). Towards Preparing Professional and Humane Teacher, 81. Retrieved fromhttp://www.ibe.unesco.org/curricula/india/ii_te_fw_2009_eng.pdf
27. Ramachandran, V. , Bhattacharjea, S., Sheshagiri, K. (2008) Primary School Teachers. The Twistsand Turns of Everyday Practice. Educational Resource Unit. 57-62. Retrieved from http://azimpremjifoundation.org/sites/default/files/userfiles/files/Teacher%20bookle t.pdf
28. Banerji, R., Walton, M. (2011). What helps children learn? Evaluation of Pratham’s Read India program in Bihar and Uttarakhand. Retrieved fromhttp://www.povertyactionlab.org/publication/what-helps-children-learn-evaluation-prathams-read-india-program
29. Ramachandran, V. , Sethi, H. (2000). Rajasthan Shiksha Karmi Project An Overall Appraisal. NewEducation Division Documents no 7.5, 8, 24. Retrieved fromhttp://eruindia.org/files/Shiksha%20Karmi%202000.pdf
30. merican India Foundation. (2009). School Management Committee and The Right to EducationAct 2009. Retrieved from http://aif.org/wp-content/uploads/2012/08/SMCBook1.pdf
31. Duflo, E, Hanna, R, Ryan, S. (2010). Incentives Work: Getting teachers to come to school. PovertyAction Lab. Retrieved from http://economics.mit.edu/files/5995
The author has 4.5 years of experience in education, which involves a combination of teaching as well as research and communication. For three years, he was engaged with Teach For India where he started off on the Communications team and then transitioned to the Fellowship program, where he taught full-time in a low-income community in Mumbai. Currently, he is engaged in conducting research on ‘Innovative teaching practices in Municipal schools in Gujarat’ at the Indian Institute of Management, Ahmedabad. Ankit is passionate about using technology in the classroom and has co-founded VideosForKnowledge, a venture to create basic general knowledge videos to reduce the knowledge gap between high-income and low-income schools.