Report No. 86 South Asia Region, Education Global Practice Value for Money from Public Education Expenditure on Elementary Education in India April 2016 Discussion Paper Series South Asia Region, Education Global Practice Value for Money from Public Education Expenditure on Elementary Education in India APRIL 2016 Discussion Paper Series Discussion Papers are published to communicate the results of the World Bank’s work to the development community with the least possible delay. The typescript manuscript of this paper therefore has not been prepared in accordance with the procedures appropriate to formally edited texts. Some sources cited in the paper may be informal documents that are not readily available. The findings, interpretations, and conclusions expressed herein do not necessarily reflect the views of the International Bank for Reconstruction and Development/The World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the Governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. TABLE OF CONTENTS ACKNOWLEDGEMENTS ............................................................................................................. i ABBREVIATIONS AND ACRONYMS ....................................................................................... ii EXECUTIVE SUMMARY ............................................................................................................ 1 Key determinant of VFM: Changes in learning levels.................................................................................. 9 Are we getting value for money? What is the expected change in productivity from a change in learning levels? ......................................................................................................................................................... 10 Calculation of VFM – How are private schools doing? .............................................................................. 11 Low VFM from public education spending: Source identification: ........................................................... 12 Low learning levels ..................................................................................................................................... 12 High public expenditure on education ........................................................................................................ 13 Non-productive expenditures in education ................................................................................................. 14 Inefficiency due to non-genuine enrolment ................................................................................................ 15 Low teacher attendance rates ...................................................................................................................... 16 VFM from changes in access to schooling ................................................................................................. 16 The Small Schools syndrome: Is maintaining small, inefficient schools leading to inequity? ................... 17 Convergence for Greater Efficiencies: Downward Integration with Pre-School Education and Upward Integration with Secondary Education: ....................................................................................................... 18 Towards Composite Schools: Convergence between SSA and Rashtriya Madhyamik Shiksha Abhiyan (RMSA) for increased efficiency: The Rajasthan experience..................................................................... 20 Conclusions and recommendations............................................................................................... 24 Benefits of Investing in Pre-Primary Education ........................................................................... 31 Financial and Economic Assessment of the Composite Schools Model in Rajasthan ................. 51 VFM Calculation .......................................................................................................................... 58 Appendix 1: Learning Outcomes: VFM Calculation .................................................................... 59 Appendix 2: Changes over time in Total Number of Schools, Total Enrolment and Average Enrolment per School ................................................................................................................... 63 Appendix 3: Some Statistics of Government Schools in India’s States as per 2014 – 15 DISE Data ............................................................................................................................................... 82 Appendix 4: Vimala Ramachandran’s evidence on teacher salary levels in India: Actual take home salaries of teachers# (in INR).............................................................................................. 89 TABLES Table 1: Models for Investing in Pre-Primary Education (Please see Appendix 5) ..................... 20 Table 2: Schools with electricity connection ................................................................................ 21 Table 3: Schools with library facility............................................................................................ 22 Table 4: Schools with computer availability for teaching learning purposes ............................... 22 Table 5: Teacher related parameters ............................................................................................. 22 Table 6: Average number of instructional days (upper primary) ................................................. 23 Table 7: Monitoring and Supervision ........................................................................................... 23 Table 8: Adjustment factors .......................................................................................................... 23 Table 9: Effective Efficiency of composite schools over elementary schools ............................. 24 Table 10: Models for Investing in Pre-Primary Education ........................................................... 34 Table 11: Schools with electricity connection .............................................................................. 52 Table 12: Schools with library facility.......................................................................................... 53 Table 13: Schools with playground .............................................................................................. 53 Table 14: Schools with computer availability for teaching learning purposes ............................. 53 Table 15: Teacher related parameters ........................................................................................... 54 Table 16: Average number of instructional days (upper primary) ............................................... 54 Table 17: Monitoring and Supervision ......................................................................................... 54 Table 18: Adjustment factors ........................................................................................................ 55 Table 19: Effective Efficiency of composite schools over elementary schools ........................... 55 Table 20: Total change in learning outcome factor ...................................................................... 56 Table 21: Change in actual income derived from school completion via composite school over elementary school in Rajasthan .................................................................................................... 57 Table 22: Performance in different content areas, NCERT’s National Achievement Survey (Class-5), 2011 and 2015 .............................................................................................................. 76 Table 23 :Annual Status of Education Report, 2010 to 2014 ....................................................... 77 Table 24:Estimates of primary-school teacher salaries as a ratio of per capita GDP ................... 77 Table 25: Temporal change in number of schools, total enrolment and average enrolment per school, in Govt and Private schools By state (2005-06 to 2014-15) ............................................ 79 Table 26:The number of ‘small’ government schools (with a total enrolment of 20 or fewer), their pupil teacher ratios and per pupil expenditure, 2014-15 ...................................................... 81 AUTHORS Geeta Kingdon Professor UCL Institute of Education, University College London Shabnam Sinha Senior Education and Institutional Development Specialist World Bank Venita Kaul Director School of Education Studies & Center for Early Childhood Education and Development, Ambedkar University, Delhi with Gaurav Bhargava and Kartik Pental Ernst & Young Management Consultants New Delhi ACKNOWLEDGEMENTS The Study was initiated as a part of the World Bank’s support to the project India: Elementary Education: SSA III to understand the value for money from public-funded elementary education by analysing public education expenditure. The Bank would like to thank Seema Bansal, Garima Batra and Shubhaluxmi Ganguly (Boston Consulting Group) for their help in finalizing the research on early childhood education (ECE). Mamata Baruah provided necessary support in formatting and publishing. i ABBREVIATIONS AND ACRONYMS ASER Annual Status of Education Report AWC Anganwadi Centres AWW Anganwadi Workers BRC Block Resource Coordinator CPC Child Parent Center CRC Cluster Resource Coordinator CAG Comptroller and Auditor General DISE District Information System on Education ECCE Early Childhood Care and Education GOR Government of Rajasthan GTR Grade to Teacher Ratio ICDS Integrated Child Development Services IECEI Indian Early Childhood Education Impact IRT Itemized Response Theory LEP Learning Enhancement Program MHRD Ministry of Human Resource Development MP Madhya Pradesh MWCD Ministry of Women and Child Development NAS National Achievement Survey NCTE National Council for Teacher Education NCERT National Council of Education, Research And Training NISA National Independent Schools Association OOSC Out of School Children PISA Programme for International Assessment PPE Per Pupil Expenditure PTR Pupil-Teacher Ratio RMSA Rashtriya Madhyamik Shiksha Abhiyan ROI Return on Investment RTE Right to Education SC Scheduled Caste SSA Sarva Shiksha Abhiyan ST Scheduled Tribe TLM Teaching Learning Materials UDISE Unified District Information System for Education VFM Value For Money ii FOREWORD Investing early in children is the key to enhancing the India’s competitiveness at the global level. India’s right to free and compulsory elementary education (RTE) to all children of the age group of 6-14 years of age rightly focuses on the early years and has demonstrated impressive gains. Almost 98% of habitations having access to an elementary school; pupil-teacher ratios have improved significantly and Sarva Shiksha Abhiyan ( SSA) as the vehicle for the RTE has taken huge strides in developing infrastructure for elementary schools across the country. While these are certainly reasons to celebrate, the concern at present is the persisting low levels of learning in the early grades. The recently concluded nationally driven National Achievement Test (NAS), 2015 for Class V that provided an opportunity to compare achievement levels over a period of time shows a decline in learning outcome levels. This holds true across States and across subjects. It is time to take stock of the return on the investments in school education to test its efficiency and effectiveness. The World Bank’s study Value for money from Public Education Expenditure on Elementary Education in India has been undertaken as part of its ongoing support to Sarva Shiksha Abhiyan. The study reflects findings that have profound implications for the school education sector; the decrease in Class V learning outcome levels is expected to reduce students’ expected wage earnings by up to 11.9 percent, significantly reducing India’s labour force productivity and economic growth. This is a phenomenon that links integrally to the key issue: have young children been provided the right start in life? Are they taught by teachers who are qualified, motivated and accountable? While India pays its teachers three times more than China, it needs to enhance the presence of these teachers in the classrooms and actually teaching. Recent studies show that greater accountability from teachers and improved presence in classrooms would reduce teacher costs by $1.5 billion per annum. The Report has raised some major emerging issues in the education sector while evaluating the value for money as a reflection of its being used efficiently to yield a high return in terms of children’s access to quality education and learning outcomes. Key funding of the study and major recommendations are included in the Report. An infographic based summary of the report is present below for easy readability. I hope this study will be useful for academics, policy makers, educational practitioners and people interested in the education sector in India. Keiko Miwa Practice Manager South Asia Region Education Global Practice The World Bank Group iii EXECUTIVE SUMMARY The World Bank has been supporting the Universalization of Elementary Education (UEE) program for India through its support to the flagship program of the Government of India, the Sarva Shiksha Abhiyan (SSA). The Bank’s ongoing support to SSA is a little over $ 1 billion. With reducing national level budgetary allocation and scarcity of resources for publicly-funded education, it is important to analyse the efficiency and effectiveness of the investments in the sector. It is critical that financial investments made have been used efficiently and yield a high return in terms of children’s access to education and learning outcomes. With this objective, the World Bank as a part of its ongoing support to SSA, undertook a Value for Money (VFM) analysis that could assess the returns from public education expenditure. This paper attempts to calculate and benchmark the economic value of any increases in children’s access to schooling and in students’ learning levels that may result from increases in public education spending over time. Some insights about key issues and recommendations are presented in Infographics for easy readability below: The reduction in total enrolment has rendered 1 Elementary Education in India: Some Facts Dwindling Enrolment: Increase in Small Schools  In the last 9 years, while number of Government schools increased by 1.4 lakh, their total Enrolment fell by 6.7 Million students. The number of private school rose by 1.7 lakh and their total enrolment rose by 35.5 million.  The reduction in total enrolment has rendered many Government schools economically unviable. There are 96,965 ‘small’ schools in India (those with 20 or fewer students as a whole, for classes 1 to 5 or 1 to 8). Learning Outcome Levels and Associated Costs  Half of students enrolled in Class V, cannot read a Class II level text; for children enrolled in government schools in Class V, reading levels between 2010 and 2012 are increasingly getting lower and stuck (ASER 2014)  During the period 2011-12 to 2014-15, the aggregate learning outcome levels for Class V students declined by 6 to 33 points (National Achievement Survey- NCERT). During the same time period the annual per pupil expenditure increased by up to 253 percent  Cost per unit of learning achievement is INR 338 in government schools and INR 63 in private schools. Labour Force Productivity  Every ‘one standard deviation’ positive increase in cognitive skills can increase a students’ expected wage earnings by 18 percent.  The decrease in Class V learning outcome levels between 2011-12 and 2014-15 is expected to reduce students’ expected wage earnings by up to 11.9 percent. Teacher Salaries and Absenteeism  Government school teachers in India are given a remuneration which is three times what China pays its public school teachers and 25 times what private schools in India remunerate their teachers.  The fiscal cost of teacher absence in India is around US $ 1.5 billion (or INR 9800 crore) per annum. Early Childhood Education  Adding 2 years of pre-school education to SSA (Right to Education) can generate a return as high as INR 25 for every INR 1 invested- with about 30 to 60 percent additional income ntary Education in India: Some Facts Elementary over one person's lifetime 2 The Small Schools Phenomenon Number of Small Schools (computed from UDISE raw data) Mean Annual per Pupil Teacher-Salary-Cost in Small Schools The reduction in total enrolment may make many Government schools economically unviable. There are 96,965 ‘small’ schools in India (those with 20 or fewer students as a whole, for classes 1 to 5 or 1 to 8). These nearly 1 lakh ‘small schools’ have a PTR of mere 6.7 students per teacher, a mean teacher salary cost per pupil of INR 85,872 p.a, and with the teacher salary bill going up to of INR 9,700 crore pa. While recognizing that some of these ‘small’ Government schools are in remote or hilly areas, the PTR of 6.7 students per teacher suggests (given the number of teachers allocated) that over time, student numbers have dwindled, leading to low cost-effectiveness. The RTE Act’s requirement for establishing more Government schools increases wastage, given the trend of the emptying of Government schools. To improve cost-effectiveness, in Rajasthan, Maharashtra and Chhattisgarh, about 23,700 Government schools have been merged with other Government schools or been closed down in 2014-15. 3 Emptying of Government Schools Average Enrolment per Government School Average Enrolment per Government School (UDISE 2005-06) (UDISE 2014-15; Except AP for which 2013-14) In the last 9 years, while number of Government schools increased by 1.4 lakh, their total enrolment fell by 6.7 million students. The number of private school rose by 1.7 lakh and their total enrolment rose by 35.5 million. Average enrolment per Government school fell by 24 students from a low base of 131 students per Government school in 2005, implying a 20% reduction in mean school size in 9 years. Between 2007 and 2014, Government schools with 50 or fewer students as a proportion of all Government schools increased from 24.3% to 32.3% (from 2.3 lakh to 3.5 lakh small schools). Since RTE obligated state Governments to establish neighborhood schools, 65,065 new Government schools have been established, reducing the mean size per Government school from an already low base. Currently, mean size per school (class 1 to 8) is 106 students per Government school. 4 Comparison of Change in Learning Outcome Levels vis-à-vis Change in Annual Per Pupil Expenditure at Government Schools Expenditure at Government Schools • There has been a sizeable decrease in the learning • The decrease in learning outcome scores is expected to outcome scores of Class V students. have an adverse impact on the income that students The decrease in the aggregate learning outcome scores can expect to derive upon joining the workforce. (for mathematics, comprehension and environmental sciences) varies between 6% to 33 percentage % • The decrease in expected income would be a result in a drop in productivity due to poor learning outcomes. • The average decrease in learning outcome scores for mathematics, comprehension and environmental • While the decrease in expected income varies from sciences is 18, 18 and 13 points respectively. State to State, it ranges from 2.0 percent to 11.9 percent. • As per the 2015 NAS, many of the States that were recorded to be doing better than the National average • In absolute terms, the expected income for students in 2012 are found to have reported some of the lowest who completed Class V in 2015 will be up to INR 1,827 learning outcome scores. per annum less than what students completing Class V in 2012 would expect to earn. 5 Causes for Low Learning Achievement Levels Large educational inputs related expenditure, i.e. on items that have little relationship with student learning levels; example: expenditure to reduce pupil teacher ratios by large scale hiring of teachers to meet 1:30 Pupil Teacher Ratio (PTR) to meet requirements of the Right to Education (RTE) Act. Stretched fiscal capacity of Government through increasing teacher salaries while these salaries are already high compared to other developing countries; up to 10 times the salaries of contract teachers and up to 25 times the salary in private schools. Inefficiency of maintaining small schools due to diminishing student population. Lack of access to quality pre-school education (as per RTE Act) leading to lack of preparedness of young children on cognitive, con-cognitive and socio-emotional skills constraining their ability for smooth transition to formal schooling in Class I. Private schools with pre-school have greater enrolment trends and retention levels Wastage may be due to considerable proportion of non-genuine enrolments and high levels of student absence – need to have third party assessment of enrolment data and attendance rates for Government schools. Low teacher attendance rates and need for enhanced accountability. Need for greater autonomy and efficiency in the social accountability mechanisms through the School Management Committees that have been put in place under the Sarva Shiksha Abhiyan (SSA). 6 Introduce Two Years of Pre School Education in Government Schools Adding 2 years of pre-primary education in the government schooling system in India can generate a return as high as INR 25 for every INR 1 invested. Students who undergo pre-primary schooling have improved educational outcomes with 30 percent more students completing high school, along with lower repetition and wastage in school. This leads to a skilled workforce and increased labour productivity with 30 percent to 60 percent additional income over an individual's lifetime. Additionally, other benefits to society in the form of reduced crime and improved health outcomes are seen. Finally, benefits of pre-primary education surpass costs required to set up the system within less than 10 years and lead to an additional GDP of approximately 4.3 percent by 2050 and 7 percent by 2074. Pre Primary Education Works! 7 Recommendations Revisiting Norms • Close/Consolidate unviable small schools and encourage Public Private Partnerships. • Greater convergence between SSA and RMSA to reduce transation cost at national and state levels • Closing private schools in the name of non-compliance may be revisited and they may be assisted for reform Pre-School education in govt. schools • Two years of pre-school education may be brought under the ambit of SSA urgently as inherent inefficiencies in provision of pre-school education is leading to low learning levels in primary school children and migration of students to private schools that provide pre-school education along with primary schools. Learning Outcome Levels • Provide performance based support to states under SSA through ranking of states based on learning outcomes • India may participate in international assessments starting with PIRLS (Progress in International Reading Literacy Study) testing reading achievement of students in their fourth year of schooling and TIMMS (Trends in International Mathematics and Science Study) that tests 4th and 8th grade students to underatnd its position internationally Teacher Salaries and Absenteeism • Reduce teacher costs to the central government: with more efficient sharing of cost of teacher salaries between national and state governments • Increases in teacher salary may be linked to teacher performance • Link salary hikes to teachers accountability and performance with a mild form of performance related pay, as was done in Mexico as negotiated with teacher unions • Strengthen local leaders and School Management Committees for assessment of teacher attendance and performance 8 INTRODUCTION Since resources for publicly-funded education are scarce, it is important that every Rupee is used efficiently and yields a high return in terms of children’s access to education and learning outcomes. Value for Money (VFM) analysis shows what return the taxpayer gets from public education expenditure. This paper attempts to calculate and benchmark the economic value of any increases in children’s access to schooling and in students’ learning levels that may result from increases in public education spending over time. Influential research shows that it is not (average) years of education of the population but rather the (average) cognitive skill levels of the population that explains the economic growth rates of countries (Hanushek and Woessman, 2008). For this reason, the paper focuses first on VFM from government investment in terms of increasing children’s learning levels. Section 2 investigates how students’ literacy and numeracy skill levels change with changes in per pupil education expenditure by government, and then measures the extent to which those resulting changes in cognitive skill levels influence individuals’ productivity in the labour market. It also measures the cost per achievement unit in the government and private school sectors, to compare the VFM from education expenditure in these two schooling sectors. Section 3 examines the various underlying sources of VFM in school education and considers how VFM could be improved. Section 4 considers the VFM from government education expenditure in terms of increased access to schooling denoted by the drop in the number of out of school children (OOSC) between 2009 and 2014. Section 5 concludes. Key determinant of VFM: Changes in learning levels VFM shows the amount of output produced for each rupee spent. A state is said to be achieving high VFM for a given amount of educational expenditure, if it increases learning levels more than other states/countries, or, conversely, if it produces a given amount of learning at a lower cost than other states/countries. The paper follow two different approaches to measuring VFM. Firstly, it examines the change in productivity / earnings that a typical student would expect to get in the labour market, from the change in learning level that he/she experienced between 2012 – 2015, as measured in the National Council of Education, Research and Training’s (NCERT) National Achievement Survey (NAS) of class 5 at these two points of time. Then, this is put together with the increase in government education spending per pupil over the same period 2012-15, to yield VFM. The second approach is to compare the learning output per rupee in the government schooling sector with a suitable comparator. Evidence on two components is needed to calculate the VFM in this approach – on government school pupils’ learning levels (output) and on government schools’ per student cost (per pupil expenditure (PPE)). Putting these two together, i.e., dividing the annual per-pupil-expenditure by the number of learning units, will yield the rupee cost per learning unit in the government school sector. Ideally, one would like to compare VFM from government education spending in India with the VFM from government education spending in other countries; for that, on the output side, we would need data on students’ learning levels on the same test taken by students in India and in other countries. We do not have such data for any recent year since India has not taken part lately in any international tests and, even when it participated in the Programme for International Assessment (PISA) test, only two educationally advanced states took part – Himachal Pradesh and Tamil Nadu. 9 Thus, instead, to see whether VFM from government education spending is high or low, the study compares it with the VFM from private school spending. To do this, the learning output measure for government and private schools within India was required. Since the NCERT’s NAS does not test private school students, the paper uses the learning achievement levels measured in the Annual Status of Education Report (ASER) survey. This exercise was undertaken for 8 major states of India. PPE on government elementary schools is calculated from the state governments’ published budgets 1, and private schools’ total fee level in elementary (class 1 to 8) including the school’s tuition fee, examination fee, development fee and other compulsory payments, is taken as their PPE. The median of private school fee in each state is obtained from the National Sample Survey of 2014 (NSS, 2014). Are we getting value for money? What is the expected change in productivity from a change in learning levels? The recently concluded NAS 2015 for Class V students provides the first opportunity to compare achievement levels over a period of time, since the same class was also tested in 2011- 12. Class V is the first class for which learning outcome scores are available at two separate points in time and that maintains a consistent Itemized Response Theory (IRT) based assessment method. Table 1 contains the results. The mean achievement levels of class V children in eight sample states in Reading, Maths and EVS in 2012 and 2015 are set out in Appendix 1. The comparison of learning levels over the two years in Table 1A shows a decline in learning outcome levels and this downward trend is observed to hold true across States and across subjects. The literature suggests a positive link between cognitive skill levels and the labour market earnings a student can expect to derive after completing schooling. A drop in learning outcome levels means a drop in the level of cognitive skills and in turn a drop in the individual’s (future) productivity in the labour market. This drop in productivity would reduce the remuneration that profit-maximising employers are ready to pay. In order to facilitate comparison over time and across States, the NAS assumes a normal distribution for learning outcome scores and, by construction, keeps the expected mean score at 250 achievement points with a standard deviation (SD) of 50 achievement points around that mean. Aslam et al (2010) show that a 1 SD increase in cognitive skills increases wages by 18% in India; this is roughly in the range of size of effects from cognitive skills onto wages reported in Table 1 of Hanushek and Woessmann (2008) for Ghana, Kenya and Pakistan. Thus, it is assumed that for every one standard deviation reduction in the learning outcome score, a student’s expected earnings upon joining the workforce dip by a factor of 0.18. We multiplied the reduction in the mean test score in terms of standard deviations in the penultimate column of Table 1a by the factor of 0.18, and this gives the percentage change in earnings as a result of the reduction in cognitive skills that a student can expect to derive upon joining the workforce. The analysis uses change in the aggregate learning outcome achievement levels, i.e. average test scores for ‘Comprehension’, ‘Mathematics’ and ‘Environmental Sciences’ seen in column 1 Government expenditure on SSA and Mid Day Meals is removed, in order to exclude expenditure on free uniforms, books, scholarships for SC/ST children and on mid day meals. This is to keep Government PPE comparable to the private schools’ PPE, since private schools do not spend on these items. 10 5 of Table 1a. The state wise calculations given in last column clearly highlight how the drop in learning levels is expected to impact the future earnings of students. The drop is expected to be in the range of 2.04 percent in Gujarat to as high as 11.9 percent in Uttar Pradesh. Table 1b shows that in absolute terms a child completing Class V in 2014-15 should expect to draw earnings which are INR 312 (in Bihar) to INR 1,827 (in Tamil Nadu) less than what a child completing Class V in 2011-12 expected to earn. The estimates for average annual income of a Class V student have been determined using NSSO 67th round data 2. While the drop in learning levels and therefore the expected drop in economic productivity/ earnings is a worrying sign, the gravity of the VFM problem is further exacerbated by the fact that during the same time period (2011-12 to 2014-15), state governments have increased their expenditure on elementary education. Estimates of PPE of Government elementary schools are set out in Appendix Table 1k, whose note explains the method of estimating the PPE. The change in annual PPE between 2011-12 and 2014-15 are given in Table 1c. During this time period, the annual PPE has increased by upto 150% (in Madhya Pradesh), and by an average of about 81%, if we average across the 8 major states (last row of Table 1c). Only Gujarat has decreased annual PPE, by 15 percent. To summarise, there has unfortunately been negative VFM from increases in government expenditure on education over the period 2011 to 2014 since over this period, student learning levels fell instead of rising and, when this reduction in cognitive skills is monetised, it shows a negative impact on expected earnings/productivity. Calculation of VFM – How are private schools doing? Tables 2a and 2b show the VFM from education expenditure on government and private schools for 8 different states of India, for literacy and numeracy areas respectively. VFM is essentially a measure of the efficiency or cost-effectiveness of expenditure. Table 4a shows that there are large variations in the cost-effectiveness of government spending across the different states, with Uttar Pradesh being an outlier. Consider the case of Madhya Pradesh (MP) in Table 2a, for illustration. PPE in MP government school system is Rs 9384 per annum and in private schools Rs 3700 per annum. Thus, government schools’ PPE is 2.5 times private schools’ PPE. However, the learning units are higher in private schools: 58% of private school students and 28% of government students of class 5 could read a class 2 level text in 2014-15. Thus government schools’ learning output is just about half that in private schools. Putting the output and expenditure items together, we find that the cost per unit of achievement is Rs 338 in government schools and Rs 63 in private schools, implying that private schools are 5.3 times as cost effective as public schools, or that government schools are one fifth as efficient in producing output as private schools. However, studies show that a large part of the observed learning advantage of private school students is explained by their more educated and better-off home backgrounds. When home 2 A national estimate has been adjusted to a State level by using State wise per capita income estimates (2004-05 constant prices) as an adjustment factor. In order to do this, the per capita earnings estimates for the State have been divided by the per capita earnings estimates for India as a whole. The resulting factor has been used to adjust the National level estimate for expected income for a Class V pass out (INR 19,226). The figures for 2014-15 have not been adjusted for inflation so as to maintain comparison with the 2011-12 figures (as adjustment using Consumer Price Index may have nominally negated the negative impact of fall in learning outcome levels). 11 background is strictly controlled for 3, the raw public-private learning gap greatly falls but is usually not eliminated. In Desai et al (2008) and French and Kingdon (2010) both using national data on India, about 20 to 30% of the private school achievement advantage remained even after controlling strictly for home background. The last row of Table 1a shows that if only 25% of the raw public-private achievement gap of MP is attributed to superior private school quality (e.g., lower teacher absence rates), then private schools are 3.25 more efficient than government schools, rather than 5.3 times. If all the raw public-private learning difference is due to the superior home backgrounds of private school students, and private school quality is no better that government school quality – then the public-private gap in VFM collapses to the public-private gap in PPE only, i.e. private schools are only 2.5 times as cost-effective as government schools, or to put it another way, VFM from government education expenditure is only 40% as much as the VFM from private schools’ education expenditure, as seen in row G in Table 2a. In summary, there is very low VFM from government expenditure on education, in terms of producing the valued outcome of ‘learning’ among students. The private schooling sector gets significantly higher VFM in this respect. Together with the earlier finding in section 2.1 where VFM was in fact negative due to reductions in learning levels over time, this paints a fairly grim picture of overall VFM being achieved from public education spending, and it calls for a deep and honest probe into the factors behind these low returns. Section 3 below scrutinizes the most major ingredients behind the low-VFM crisis. Low VFM from public education spending: Source identification: There are several identifiable factors behind the observed low VFM that need attention in order to improve the efficiency of government spending on education. The first of these is low learning outcomes from education. Low learning levels With the appearance of the NCERT’s NAS 2015 of class V, there is clear evidence of poor and falling learning levels in government funded elementary education. The NAS, carried out only in government schools, shows low and falling cognitive skill (literacy and numeracy) levels. To take one example, in the 2010 NAS, only 45% of class 5 students could identify the correct answer to the Multiple Choice Question “How much greater is 555 than 198?” a subtraction word problem around the class 2 level of difficulty. By 2015, 40% of class 5 students who could answer this same question. Table 3 shows a consistent across-the-board decline in learning achievement level (of the order of 5 percentage points) in all subject areas. Secondly, the national ASER 2014 also indicates a similar story of low and falling learning levels. It shows that among class 5, only 42% of government and 62% of private school students could read a class 2 level text, and only 21% of government and 39% of private school students 3 Through a regression technique called ‘household fixed effects’ which compares only the achievement levels of children within the same household who go to private and Government schools. Since the home background (parental education; household wealth, etc.) is the same for different children of the same family, this is a more powerful way of eliminating the influence of family background, and yields a much purer private school effect than the simple cross-section Ordinary Least Squares regression analysis. Desai et al (2008) use India Human Development Survey of 2005-06; French and Kingdon (2010) use three years’ of ASER survey. The latter also do village level panel data analysis. 12 could do simple division, a class 2 or 3 level competency. Even among class 8 students in government schools, only 46% could do the division sum. Some ASER evidence is presented in Table 4. Thirdly, in the international PISA test of 15 year old children’s performance, India stood 73rd out of 74 participating countries, when it had put forward two of its educationally advanced states (Himachal Pradesh and Tamil Nadu). If VFM is to increase, the single most important reform is to increase children’s learning levels. The RTE Act’s best ideas for how to improve quality is to insist on lower pupil teacher ratios (30:1 in primary, and lower than that in upper primary), to ensure that teachers have certification of training qualifications, and to ensure the availability of basic infrastructure, but this inputs based approach to quality is not evidence-based (see section 3c below). The RTE Act does not prescribe any measures to strengthen teacher accountability and effort. On the contrary, the recognition requirement of the RTE Act is compelling the (5 times higher value yielding) private schools to close down due to non-compliance with the physical inputs requirements of the RTE Act or with other recognition conditions. . According to the National Independent Schools Association (NISA), 15,083 private unaided schools have closed down so far due to RTE non-compliance. Closure of private unaided schools often negatively impacts children’s right to education (RTE) (in a context where the child population is increasing and state governments are closing down unviably small government schools), while also reducing the overall efficiency of the system by reducing the number of high-value- yielding private schools and continuing to operate the lower-value-yielding government schools which, in some cases themselves do not fulfil the RTE infrastructure norms, since there is no provision in the RTE Act for the closure of government schools that do not comply with RTE norms. High public expenditure on education There are various ways of benchmarking the size of public expenditure on education in India. One way is to compare it with that in other countries, e.g. comparing India’s “PPE on education as a proportion of the country’s per capita GDP” with the same quantity in other countries. Another way is to compare government schools’ PPE with private schools’ within India. China and India comparison of public education expenditure Because of data availability, we compare not total education expenditure but rather its largest component, namely teacher salary expenditure, and it shows that education is produced three times as expensively in India as in China. Table 5.4 in Drėze and Sen (2013) is reproduced here as Table 5; it presents ‘teacher salary as a multiple of the country’s per capita income’ for India and several other countries. The authors find that the ratio of teacher salary to per capita GDP is 0.5 in Indonesia and 0.9 in China but 3.0 in India, i.e. average teacher salary expenditure is three times the average national per-capita-income in India, but that in China average teacher salary expenditure is less than the average national per-capita-income. In other words, China spends only one-third as much on teacher salary as India, when expressed as a multiple of national per capita income. This was before the wage inflation generated by the Sixth Pay Commission, whereby teacher salaries approximately doubled in one go (Kingdon, 2010). 13 Public and private school comparison of per-pupil-expenditure Another way of benchmarking the size of public expenditure on education is to compare PPE in government schools with that in private schools. Tables 2a and 2b already showed PPE in public and private schools for 8 major states of India. Table 2a showed that among these eight states, PPE in government schools is anything from 1 to 13 times as much as in private schools. Non-productive expenditures in education VFM can also be low due to educational expenditures on unproductive inputs, i.e. on items that have no relationship with student learning levels. One example is expenditure to reduce pupil teacher ratios. The RTE Act 2009 obliges schools to maintain a maximum pupil teacher ratio of 30 in primary classes, and lower than that in upper primary classes (due to specialist subject teachers). But evidence internationally shows that pupil teacher ratio is not consistently related with student learning (Hanushek, 2003; Altinok and Kingdon, 2014). Reducing pupil teacher ratio is a very expensive reform without demonstrated return in terms of increased student learning, especially in the context of high student absence rates and the alleged fake/over-stated enrolment figures in government schools 4. Another example of increasing spending on non-productive inputs is the increase in government teacher salaries via Pay Commission recommendations. Much of the fiscal capacity of government to increase education expenditure is tapped for increasing teacher salaries when these salaries are already high compared to other developing countries, and are also upto 10 times the salaries of contract teachers and upto 25 times the teacher salary in private schools, which is the market-clearing wage 5. While it is proper/just that equity concerns (equality of salary between contract and regular teachers within government schools) have led to the ‘regularisation’ of contract teachers in many Indian states, three separate studies with data from 5 states showed that learning levels among children taught by contract teachers were no less than among children taught by regular teachers 6 even though their salary was upto one- tenth of regular teachers’ salary, indicating that higher salaries are not a learning-enhancing input. The fact that learning levels of children attending private schools are not lower (and could be modestly higher) than of children attending government schools, despite teacher salaries being upto 1/25th of government teacher salaries again goes to show that increases in teacher salary are not a learning-related expenditure. Some expenditures that are arguably more quality-related have not been made mandatory 7. The inefficiency of using bureaucratically set high minimum wages rather than market clearing wages, can be addressed by following the suggestion of a professional development ladder for all teachers, suggested by Pritchett and 4 Though there are no hard figures, the SchoolTELLS survey of UP and Bihar found that 15% and 35% of enrolments were fake, i.e. included names of children who were never found in the school in 4 unannounced survey visits to the sample schools. The over-reporting of enrolment could be due to the economic incentives to obtain extra food-grains from the Mid-day meal scheme, or extra cloth from the free uniform scheme, or to siphon off funds intended for SC/ST children’s scholarship of Rs 350 per year. 5 That is, the wage at which the supply of educated persons equals the demand for educated persons in a local labour market. In rural Uttar Pradesh currently, mean teacher pay is around Rs 1500 per month in 2014-15 (inflating SchoolTELLS data using the inflation index), and the average salary of a primary (class 1 to 5) teacher in Government school is Rs 39,683 pm (Ramachandran, 2015). Thus, Government teacher salary in rural UP is more than 25 times private school teachers’ salary. Average teacher salary (of primary teachers) after 15 years’ experience is estimated at Rs 43,080 per month, in 2015-16 (based on Appendix Table 4). 6 Muralidharan and Sundararaman (2010); Goyal and Pandey (2010); Atherton and Kingdon (2010). 7 For example, investments in school leadership training; increasing teacher competence; monitoring and inspection expenditure; learning surveys; increased parental information about school quality; research and innovation; teaching-learning materials; computers; student exchanges; etc. 14 Murgai (2008). Another idea is to link salary hikes to increased accountability, or a mild form of performance related pay. Inefficiency due to non-genuine enrolment Another problem of wastage is that there is a considerable proportion of non-genuine enrolments and high levels of student absence. Some questions have been raised – from time to time – about the veracity and trustworthiness of enrolment data from the self-reported District Information System on Education (DISE). Newspapers regularly report scams related to fake enrolment numbers. To take two recent examples from Uttar Pradesh, in November 2015, the Comptroller and Auditor General (CAG) and the Mid-Day Meal Authority (MDMA) carried out a joint survey of enrolment figures in UP government schools and found that students’ names were entered in the enrolment registers of more than one government school, and that in almost every elementary school, total enrolment was inflated by at least 10% 8. Secondly, in September 2015, the DISE enrolment data for the Lucknow district were reviewed by the District Magistrate who ordered for a survey to be carried out by the district Basic Education Officer. The survey showed that 18% of students in Lucknow were “absent for long period” and the District Magistrate ordered the cancellation of the admission of many of the elementary school children whose names were in the enrolment registers 9. This is fairly consistent with the findings of the SchoolTELLS survey of 80 rural primary schools in 5 districts of Uttar Pradesh 10 where each school was visited 4 times in the year 2007-08, and it was found that 15% of students in the enrolment registers were never present in the school in any of the four survey visits, i.e. 15% of the total primary school enrolment was apparently fake. And this is disregarding the absenteeism among children who are not fake enrolments 11. It has been widely suggested that there are economic incentives for government schools to over-report enrolments since grains for mid-day meals, cloth for school uniforms, scholarship money for SC/ST students, and the number of teachers appointed, all these increase with the reported number of enrolled children in a school, and there are no penalties for over-reporting enrolments. A de facto pupil teacher ratio of 30:1, based on attendance rate, would be a desirable policy correction. 8 http://epaperbeta.timesofindia.com//Article.aspx?eid=31813&articlexml=UP-schools-drawing-funds-for-non- existing-students-02112015004030 9 http://epaperbeta.timesofindia.com//Article.aspx?eid=31813&articlexml=BSA-SURVEY-18-primary-students- in-city-skip-29092015002036 10 Rural parts of districts Agra, Shrawasti, Mahoba, Bijnor and Lucknow. 11 Surveys by the MHRD and the ASER suggest that just over half the children who enrol have a tenuous connection with the school in UP. The ASER survey for 2015 shows student attendance rates in UP government schools as 55.1% in primary and 54.7% in upper-primary schools. Thus, when UP elementary schools show a pupil teacher ratio of 33 according to their enrolment data, this amounts to about 17 pupils per teacher actually present in school any day. 15 Low teacher attendance rates Low teacher attendance is another large source of wastage in the publicly funded part of the schooling system. Muralidharan et al. (2015) find that the fiscal cost of teacher absence in India is around US $ 1.5 billion (or Rs. 9800 crore) per annum. They advocate improving governance by hiring staff to increase the frequency of monitoring would be over ten times more cost effective in increasing teacher-student contact time than hiring more teachers. VFM from changes in access to schooling One important source of VFM comes from children’s increased access to schooling, since schooling has economic and non-economic benefits. A cost-benefit analysis of the return on the Sarva Shiksha Abhiyan (SSA) expenditure – over the period 2003 to 2009 – by Kingdon and Atherton (2010) showed a significant positive economic return from the public expenditure on the SSA program as this period saw an increase in school access. An important (inverse) measure of schooling participation is the number of OOSC. Estimates of OOSC in India are available from the IMRB surveys commissioned by the MHRD via EdCil in 2005, 2009 and 2014, and the annual ASER surveys. Their findings are summarised in Table 1. They show that between 2005 and 2009, the proportion of OOSC aged 6-14 fell from 6.9% to 4.3%, close to the corresponding numbers of 6.6% and 4.0% respectively in ASER. In absolute terms the number of OOSC fell from 13.5 million to 8.2 million over this period. However, the pace of enrolling OOSC slowed down after 2009. Between 2009 and 2014, the OOSC fell from 4.3% to 3% as per IMRB and from 4.0% to 3.3% as per ASER. In absolute terms, the number of OOSC fell during these 5 years from 8.2 million to 6.1 million, a decrease of about 20 lakh children who are out of school. While the number of OOSC fell by about 21 lakh, unfortunately 18.5% of the OOSC who enrolled never actually attended school, and another 37% dropped out of school, mostly by the end of grade 2, thus 55.5% of OOSC did not remain in school (IMRB, 2014). This reduces the number of OOSC who had meaningful schooling participation to 44.5% or 9.4 lakh children. To measure the economic return to staying on in school for these 9.4 lakh youngsters, one would consider the boost in earnings from extra years of schooling. If these students are retained in school until class 10 or beyond, they will enjoy higher earnings and higher productivity and will contribute to economic growth. While there appears to be little economic return to increments in education in the Indian labour market until lower secondary class 10 level of education (Colclough et. al., 2010), perhaps because education below class 10th does not lead to secure literacy and numeracy skills and employers are unwilling to pay higher wages for mere completion 5 or 8 years of schooling without learning cognitive skills 12, a major advantage of elementary education is that it permits access to high school and beyond, where there are significant economic returns. Then there are also the non-economic benefits of basic education. While the literature suggests that even the non-economic returns to education, such as higher civic sense, lower fertility rates, lower infant mortality, better child health and education, etc. are greater from secondary education than from elementary education, again elementary education permits access to the (high return) secondary and further levels of education. Moreover, government must be credited for establishing schools in remote rural 12 But this phenomenon is not confined to India alone. Colclough et. al. show that the economic return to primary schooling has collapsed over time in most developing countries. 16 areas where private schools are unlikely to operate, even though there is no established methodology for monetising the non-economic benefits of education or for monetising the benefits to remote communities of having school access. Changes in enrolment patterns across government and private schools can lead one to question the extent to which the OOSC have been absorbed by government schools, since as Table 2 shows government school enrolments have been falling sharply and private schools enrolments rising strongly 13. However, it is still likely that the OOSC who joined school mostly joined government school. Despite the difficulty of measuring the benefits from reduction in OOSC, and despite the size of benefits being compromised due to the low quality of schooling, there are nevertheless likely to be positive economic and non-economic returns to schooling participation that is facilitated by government education spending. The Small Schools syndrome: Is maintaining small, inefficient schools leading to inequity? Table 6 shows that government schools have been emptying over time: between 2005 and 2014, in the 20 major states of India, while the number of government schools increased by 1,38,422, their total enrolment fell by just under 6.7 million students, so that ‘average enrolment per government school’ fell by 24 students [from a base of 131 in 2005] 14, i.e. mean school size fell by nearly 20%. To understand Table 6, it is useful to see the underlying data in the Appendix Table 2. To take the example of MP, the number of government elementary schools in MP rose from 104,671 in 2005 to 114,360 in 2014-15, i.e. 9,689 new government schools were established in MP in this 9 year period, an increase of 9%. However, the number of students studying in government schools fell by 21,71,597, i.e. government school enrolment fell by 20%. Thus, average government school size fell from 104 students per school in 2005 to 76 students per school, a reduction of 28 students per school, or 27%. Table 6 shows just the changes over time for each of the 20 major Indian states. The emptying of government schools has reduced school size so much that many government schools are rendered economically unviable, with fewer than 20 students in the entire school as a whole. Table 7 shows, for each state, the number of ‘small’ government schools that have a total enrolment of 20 or fewer students. The last two rows of Table 7 show that in the listed 20 major states of India, there are 96,965 small schools and that these have 187,396 teachers and 1,254,608 students. Thus, the pupil teacher ratio in these schools is a mere 6.7 students per teacher, and the mean teacher-salary cost per pupil is Rs. 7156 per month. The teacher salary bill of these schools is nearly a staggering Rs. 9,700 crore per annum. The underlying data for each state, from which Table 7 is constructed, is given in Appendix Table 3. If these were all schools in remote areas, it would be equity-related expenditure that is facilitating access to schooling in remote areas. However, the reality of a pupil teacher ratio of 6.7 students per teacher suggests that at one time they had more students (hence the number of teachers allocated) but that, over time, student numbers have dwindled, thereby lowering the pupil teacher ratio. Thus, some part of this major phenomenon of small schools is due to 13 The underlying DISE data on total Government and private school numbers and their total enrolments, from which Table 2 has been made, are given in Appendix Table 2. 14 Over the same period, the number of private (recognised) schools increased by 1,70,064 and their total enrolment increased by nearly 35.5 million students. 17 reductions in student numbers and is indicative of a large problem of poor cost-effectiveness of education expenditure. That is presumably why some states have rationalised small schools to some extent during 2014-15, e.g. between Rajasthan, Maharashtra and Chhattisgarh, about 23,700 government schools have been merged with other schools or been closed down (newspaper reports). While the Government of India’s school location policies (reiterated in the RTE Act 2009) have long aimed to ensure a school either within a habitation, or within easy walking distance, little attention has been paid to the fact that school location policies affect not only schooling access but also many important aspects of school quality such as school size, the number of teachers, and the socio-economic composition of the student body. Anjini Kochar (2007) shows that habitation size determines the availability of schools in scheduled caste and tribe (SC/ST) habitations, and also determines the number of teachers, and that these in turn determine schooling attainment of children. Where neighbourhoods are segregated by caste, the availability of neighbourhood schools may lead to caste-based segregation in schools. Kochar shows that this increases the schooling of upper castes but reduces that of SCs. Thus, school location policies, through their effect on school quality, imply that the benefits of school access differ across castes within any given region. In this context of emptying of government schools, unviably small government schools, and the negative effects of caste-segregated schools, there is an inherent inefficiency built into the RTE Act’s prescription that governments must establish neighbourhood primary schools within a 1 km radius, and upper primary schools within a 3 km radius, of every population cluster. In a context where parents have been abandoning government schools, average government school enrolments have fallen secularly, and thousands of government schools are being closed down due to being economically unviable, creating yet more government schools would be very wasteful. This dissonance requires correction with an amendment of the RTE Act. Convergence for Greater Efficiencies: Downward Integration with Pre-School Education and Upward Integration with Secondary Education: Improving the Quality of Elementary Education through Introduction of Pre-School Education: SSA, a centrally sponsored scheme, is currently the main vehicle for RTE- for realizing the vision of providing quality education to all: which has demonstrated some success, as evident from the fact that the net enrolment rate at the primary level has risen to about 90 percent in 2011-12 and the annual dropout rate has declined from 9.4 percent in 2007-8 to 5.3 percent in 2012-13. (MHRD, 2013). While these are certainly reasons to celebrate, the concern is the persisting low levels of learning in the early grades, with a significant number of children in grade 5 reported to not being able to read even grade 2 text ( ASER 2013), resulting in children moving from one grade to another without learning even the basics of reading and writing. Responding to this concern, the MHRD has launched a new scheme titled “Padho Bharat Badho Bharat” with a focus on quality improvement in grades 1 and 2, for which specific funding is provided under SSA. While this focus on the early grades is well justified since a weak foundation can only lead to cumulative deficits in learning in higher grades, the diagnosis of the problem needs further examination. 18 Shifting Focus to the Child and School Readiness: Recent research in India in three states on about 12 thousand six year olds has provided convincing evidence that a major factor for low learning levels in early grades is inadequate school readiness with children coming into school in Grade 1 with inadequate cognitive and language competencies which are prerequisites for the primary school curriculum (Indian Early Childhood Education Impact (IECEI, 2014). With the legislative and policy mandate for all children to come into school at age 6, a significant number of children are now coming in from home environments which primarily have an oral culture with very little exposure to literacy environments or even to print material at home, with few role models which could inspire children in this context and with limited opportunities to develop their vocabulary in the school language. As a result they lack some basic cognitive and language competencies due to their limited experiences at home in the early years, which influence their learning levels in school. Preschool Education: A Profitable investment: Given this context, research across the globe, including in India, is now confirming that even one year of preschool education when children are 4 and 5 years old can lead to a significant increase in their school readiness levels at the time of entry to grade 1, which in turn contributes to enhanced levels of learning in primary grades. (IECEI, 2015) A study by NCERT (1994) on about 38,000 children in eight states provided evidence of impact of preschool participation on retention rates in primary schools which improved by about 15-20 percent as compared to students who had not participated. An analysis of short and long-term returns to investment in pre-primary education, if integrated into the government system in India, indicates significant potential benefits (The World Bank, 2015). The analysis examines Models 1, 2 and 3 for pre-primary education using a combination of revitalizing programs, mobilizing existing resources and integrating current working systems. Cost benefit analysis is undertaken in order to evaluate the cost to the system for pre- primary education inputs against potential short-term and long-term savings for government and society. The analysis of Model 1, 2 and 3 ( table presented below) represent a continuum from most comprehensive intervention to a convergence model to a focused but succinct package program to improve all early learners' school readiness. 19 Table 1: Models for Investing in Pre-Primary Education (Please see Appendix 5) Model Advantages Model 1: • Comprehensive and high-quality TWO YEARS OF PRE- • Newly developed infrastructure, teacher training PRIMARY EDUCATION and preschool curriculum in Government Schools • Situated in Government school system in order to ensure continuous transition from pre-primary to primary grades • Return on Investment (ROI) International evidence: ROI ratios from cost-benefit analysis range from $7 to $16 per dollar invested illustrating the high return from investing in pre- primary education • ROI: India specific: Adding these 2 years of pre- primary can generate a Cost-Benefit Ratio of 25, or Rs. 25 saved for every Rs. 1 invested. Model 2: • Optimizing / integrating existing resources CONVERGENCE between • Adding new resources, improving infrastructure existing government • Improved efficiency of existing systems departments to strengthen • ROI: Convergence model can generate ratio of preschool education at 10, or a savings of Rs. 10 for every Re. 1 invested Anganwadi Centers (AWCs) in pre-primary education. Model 3: • Highly focused and intense preschool pre-literacy SCHOOL READINESS and pre-math instruction PACKAGE for 30-60 days • Minimal costs with potential of ensuring school of preschool intervention at readiness for all learners the start of Class 1 • ROI: Model 3 yields a cost-benefit ratio of 75, or a savings of Rs. 75 for every Rs. 1 invested in pre-primary education. However, due to the minimal costs incurred in this specific model, benefits are not relevant in absolute terms. Towards Composite Schools: Convergence between SSA and Rashtriya Madhyamik Shiksha Abhiyan (RMSA) for increased efficiency: The Rajasthan experience: The Government of Rajasthan (GOR) has taken a strategic decision to consolidate some of its elementary schools and secondary/senior secondary schools into composite schools. The consolidation is being worked out within the guidelines set under the RTE Act. This case has been examined to study the merits of this move by comparing the effectiveness, efficiency and impact of Elementary schools vis-à-vis composite schools. Based on the realization that that a lot of elementary schools were actually created by separating primary and/or upper primary classes from existing composite schools, the GOR initiated its move on consolidation. While creating a public school network that was difficult to sustain and was inefficient in operations. A number of schools were operating with poor enrolment, poor monitoring and inadequate supervision mechanisms. 20 Rapid expansion of the public schools network has led to a sizeable financial outlay towards infrastructure expansion with major outlay towards development of school buildings, classrooms, toilets and drinking water facilities. Resources for outcomes related indicators, monitoring and quality improvement were getting constrained. The GOR decided to place elementary grades in the same ecosystem that houses secondary and senior secondary classes in order to – • Place them under the regular care and supervision of the secondary/senior secondary school Principal for concurrent monitoring and reduced unaccounted teacher absenteeism. • Provide elementary school teachers with access to subject teachers and focussed academic guidance (more qualified, trained and experienced) working with secondary and senior secondary teachers improve the Grade to Teacher Ratio (GTR) 15 thereby increasing the number of effective learning days. • Provide students from elementary schools with access to facilities such as better equipped libraries, playgrounds, functional computers etc. • Facilitate greater retention through improved quality of education and stronger parent- teacher association (a result of improved monitoring); and improved transition rates through unconstrained sharing of student data/information. Effectiveness: Composite schools benefit from having a larger institutional setup in terms of infrastructure, teacher availability and classroom availability. It is well documented that an enabling teaching learning environment leads to better learning outcomes for children, and in this sense certain parameters considered in this analysis have been charted out for both composite and elementary setups. Unified District Information System for Education (UDISE) data from 2013-14, shows that composite schools perform better than elementary school on various supporting infrastructure/facilities related parameters such as availability of electricity connections, libraries, playgrounds and computers. Data given in Table 1 shows that while 84.4 percent of composite schools have access to electricity, the corresponding figure for elementary schools stands at 52.4 percent. Table 2: Schools with electricity connection 2013-14 Variable/Parameter Composite Elementary With electricity connection 84.4 percent 52.4 percent Without electricity 15.6 percent 47.6 percent connection While 84.4% of composite schools have electricity connection, 52.4% schools have them at the elementary level. . 15 Grate to Teacher Ratio refers to average number of Classes/Grades a teacher has to manage at any given point of time. 21 Table 3: Schools with library facility 2013-14 Variable/Parameter Composite Elementary With Library 84.0 percent 80.4 percent Without Library 16.0 percent 19.6 percent Similarly, while 84 percent of composite schools have a library, the corresponding figure for elementary schools is 80.4 percent. Most composite schools have a larger library setup and many have a dedicated librarian. On the other hand, most elementary schools do not have a library room but rather a small collection of books that are stored in a cupboard. In order to improve the quality of education being imparted in public schools, the Government has been making efforts to leverage information and communication technology enabled teaching-learning transactions. However, implementing the same requires the schools to have access to computers. Table 4: Schools with computer availability for teaching learning purposes 2013-14 Variable/Parameter Average number of computers available for teaching and learning purposes Composite 3.0 Elementary 0.5 While composite schools house an average of three computers, the corresponding figure for elementary schools is 0.5 with many elementary schools having no computers. This could also be because a large percentage of elementary schools do not have electricity connections. More importantly, composite schools seem to have more effective classroom transaction processes. The percentage of teachers with a graduate degree is higher at composite schools. The GTR estimates show that at a given point in time a teacher in an elementary school teaches an average of almost two classes/grades. The corresponding figure for composite schools is closer to a teacher a class/grade. Table 5: Teacher related parameters 2013-14 Percentage of teachers Percentage of teachers Type of School with graduation and with Professional PTR GTR above qualification Composite 89.4 percent 96.6 percent 32 1.31 Elementary 81.4 percent 96.7 percent 20 1.83 The difference in GTR when factored in to adjust the average instructional days reveals that the ‘effective instructional days’ in composite school is 159 and the corresponding figure for elementary schools is 99. 22 Table 6: Average number of instructional days (upper primary) 2013-14 Type of School Average number of Average number of effective instructional days instructional days Composite 208 159 Elementary 180 99 The benefits of concurrent monitoring through the Secondary/Senior Secondary school Principals lead to composite schools reporting a lesser need for BRC (Block Resource Coordinators) and CRC (Cluster Resource Coordinators) visits leading to more efficient use of these resources. Table 7: Monitoring and Supervision 2013-14 (Rounded Off) Average no. Average no. Average Average of visits by of visits by Distance Distance Type of School Block Cluster from Cluster from Block Resource Resource Resource Headquarters Centre Centre Centre (in (in Km.) officer officer Km.) Composite 1 1 19 3 Elementary 2 3 24 4 Increased efficiency: Having larger institutional setup in terms of infrastructure availability and human resources, composite schools have the ability to cater to a larger number of students. Working closer towards full capacity enables these schools to put their resources to more efficient use. An analysis of SSA budget allocation on elementary education reveals that 50.1 percent of allocation is towards teacher salaries and trainings, 22.4 percent allocation is towards infrastructure creation and maintenance and 27.5 percent allocation is towards other elements. The Pupil-Teacher Ratio (PTR) in composite schools is higher than the PTR at elementary schools. Similarly, the number of students per school at composite schools is higher than the corresponding figure for elementary schools. Factors derived from these two data points when used to adjust the allocation figures for teacher related and infrastructure related components, provide for an estimate of the efficiency with which composite schools work. Table 8: Adjustment factors 2013-14 Type of School PTR Average strength per school Composite 32 301 Elementary 20 187 Adjustment factors (composite/elementary) 1.6 1.6 Using the above given factors to adjust the budget allocation for infrastructure and teacher related components reveals that composite schools provide for a 38 percent more efficient use of funds invested in infrastructure and teacher related components. In terms of percentage points, investing in composite schools leads to a 27.3 percentage point higher efficiency. 23 Table 9: Effective Efficiency of composite schools over elementary schools Aggregate Teachers Type of Miscellaneo Infrastructure percentage Difference in compone School us and resources budget Efficiency nt allocation 50.1 100.0 Elementary 27.5 percent 22.4 percent percent percent 27.3 percent 31.3 Composite 27.5 percent 13.9 percent 72.7 percent percent Another interpretation of the aforementioned statistic can be that all other outcomes being the same, a composite school spends INR 0.73 per child where an elementary school spends INR 1.00. The only additional cost under the composite school model is the INR 20 per day the Government provides to students in the form of transport assistance. Conclusions and recommendations This paper has attempted to measure the VFM achieved from publicly funded education, by looking at both the school access and learning outcomes of children. While it was not possible to monetize the benefits of schooling access, it is clear there will be positive returns from the increase in schooling participation inherent in the observed modest reduction of OOSC. Sadly, the economic returns from public education expenditure are negative in terms of falling learning levels and their deleterious impact on labour market productivity. This problem is compounded by the ‘double-whammy’ that learning levels fell during a period when government expenditure on education nearly doubled, in per pupil terms. The increase in per pupil government education expenditure over the 10-year period 2006-2015 is due partly to a strong increase in total government education expenditure (1.4 lakh schools established) and partly to a 24% reduction in government school enrolment and a 20% reduction in the average enrolment per government school (Table 6). VFM from the government school system is also low in comparison with that from the private school system. The reasons are partly to do with lower learning levels in the government school system but more majorly due to the huge cost dis-advantage of government schools which pay bureaucratically set high minimum wages, compared to private schools which pay market- clearing wages based on the supply of unemployed graduates who are willing to work for low salaries. Private schools have flexibility in the mix of inputs they use and they also elicit greater teacher effort and demand greater teacher accountability, as seen from the lower teacher absence rates in private than in government schools (Muralidharan et. al., 2008). Section 3 of the paper discusses the factors behind the low VFM. These include low learning levels, high public expenditure on education, non-productive expenditures on education, the inefficiency of maintaining small schools, inefficiency due to non-genuine enrolment numbers, and the high teacher absence rates. The recommendations are based around these areas as well. 24 Recommendations: Moving from block grants to per-student grants: One of the best ways to increase school and teacher accountability and thus to raise VFM from government education expenditure would be to shift from a "block grant" to a per-student grant to government and aided schools. Under a per-student grant system, there will be strong monitoring of student enrolment numbers, and schools would lose government grant money (and thus be forced to lose teachers) if their student enrolment numbers fall. In most countries, government funding grant to schools is on a per-student basis. Moreover, one can tease even greater VFM through adopting the practice in many developed countries to have efficiency and equity incentives built into the grant formula for schools. For example, in the UK, the per student grant for 2015-16 is fixed at £ 2880 for primary, £3950 for class VIII and IX, and £4502 for class X students, but the school gets, for each child of a designated group, additional per pupil grant for different categories of disadvantage 16. Rethinking school location policies: While India’s long-standing school location policies (reiterated in the RTE Act 2009) have aimed to ensure a school either within a habitation, or within easy walking distance, research suggests that this policy ends up segregating children into different schools by caste, and raises the schooling attainment of the upper castes while reducing that for the lower castes. In other words, school location policies, through their effect on school size/quality, imply that the benefits of school access differ across castes within any given region. Thus school location policies need to be rethought. Amendment of the RTE Act: In the context of emptying of government schools, unviably small government schools, and the negative effects of caste-segregated schools, there is an inherent inefficiency as well as inequity built into the RTE Act’s prescription that governments must establish neighbourhood primary schools within a 1 km radius, and upper primary schools within a 3 km radius, of every population cluster. In a context where parents have been abandoning government schools, average government school enrolments have been falling, and thousands of government schools are being closed down due to being economically unviable, creating yet more neighbourhood government schools would be very wasteful. This dissonance requires thoughtful correction with an amendment of the RTE Act. Moderation of teacher salary levels in the forthcoming Seventh Pay Commission: The fact that learning levels of children attending private schools are not lower (and could be modestly higher) than of children attending government schools, despite teacher salaries being upto 1/25th (or as little as 4%) of government teacher salaries, and given that PISA test learning levels in China are significantly higher than in India even though teacher salary as a multiple of per capita national income is only one-third as much in China as in India, indicates that increases in teacher salary are not a learning-related expenditure. Thus, moderation of teacher salary increases – especially in the forthcoming Seventh Pay Commission – will help to raise VFM in the government schooling system. Teacher salary moderation can also help to strengthen the School Development and Management Committees (SDMCs) since it will reduce the current huge economic-distance between the well-paid teacher and the typically 16 Additional per pupil grant for deprivation, between £882 and £1,870 per annum (full breakdown is given); for looked-after children – £1,004; for low prior attainment – primary: £669; secondary: £940; • English as an additional language – primary: £466; secondary: £1,130; a lump sum for every school – primary: £115,797; secondary: £125,155; additional sparsity sum for small schools vital to serving rural communities – primary: up to £44,635; secondary: up to £66,656. 25 lowly paid parent, since this economic gap hinders parents’ ability to hold teachers accountable, as suggested in the literature (Kingdon and Rawal, 2010). Another idea is to link salary hikes to teachers accepting increased accountability, e.g. accepting a mild form of performance related pay, as was done in Mexico in return for a large increase in teacher pay, as negotiated with teacher unions. Protect the low fee private schools from closure: Since private schools are about 5 times as cost-effective as government schools, and the vast bulk of private schools are the low fee private schools that cannot afford the costs of compliance with all the RTE recognition norms, it is important that government takes a facilitative rather than a punitive approach, protects such schools from closure due to non-compliance with the infrastructure requirements of the RTE Act or with the many other recognition conditions imposed by over-zealous state governments. Governments can help private schools through subsidies to become RTE compliant. Protecting them from closure will also protect children’s RTE in a context where state governments are closing down small government schools, and it will maintain the overall efficiency of the system compared with the counterfactual that the high-value-yielding private schools are closed down, and lower-value-yielding government schools continue to operate which, ironically, themselves do not fulfil the RTE infrastructure norms, since there is no provision in the RTE Act for the closure of government schools that do not comply with RTE norms. Invest in quality-related expenditures: Government must invest more in quality related expenditures which have been largely neglected hitherto, such as investments in school leadership training; increasing the very low levels of teacher competence (as seen in the dismal performance on Teacher Eligibility Tests, TETs), through relevant training; monitoring and inspection expenditure, to reduce teacher absence rates and to ensure the accurate recording of enrolment numbers; annual (non-high-stakes) measurement of child learning to enhance teacher accountability; publication of transparent information for parents/the public about learning levels in all schools, since such visible evidence on relative school quality of all the schools within a district can set up healthy competition between schools); research and innovation; teaching-learning materials; computers; student exchanges; etc. Creating a professional development structure for teachers: The inefficiency inherent in bureaucratically setting high minimum wages rather than paying market clearing wages to teachers, can be addressed by establishing a professional development ladder for all teachers. Pritchett and Murgai (2008) suggest that all teachers can start life on annually renewable contracts and a modest salary for a specified period (3 or 5 years), after which their work is assessed and, if found satisfactory, they are promoted to the position of a regular teacher with a higher salary and a permanent post. They can be offered one or two more professional development opportunities during their career, relating promotion and pay-rise to their performance appraisal. Strengthening elementary education through introduction of pre-school education: In view of the significant and proven benefits of investment in preschool education especially for 4 to 5 year olds (since 5 to 6 year olds are already in school in at least 23 states), and the cumulative benefit of addressing the entire early learning continuum for children’s learning, it is recommended that a preschool class may be added to existing primary schools across the country and the curriculum for the pre-primary and first three years of primary be developed in a bottom up manner to ensure continuity and developmental appropriateness. A two year Diploma Course on Preschool Education has already been developed and notified by National 26 Council for Teacher Education (NCTE) (2015) which could support teacher education for this stage. This recommendation is in alignment with the RTE (2009) Section 12 which encourages all state governments to endeavour to establish preschool classes for children from 3 to 6 years so as to help them develop school readiness for primary grades. Possibilities may therefore be considered of multiple models including relocating AWs to primary schools and ensuring continuity of curriculum, to setting up a one year class in schools prior to grade 1. Reducing transactions costs via consolidating the SSA and RMSA schemes: At present SSA and RMSA operate as two separate programmes, with two different state implementation agencies, leading to duplication of staff, buildings, vehicles, and efforts. Since many schools are composite schools (containing primary, upper primary, and secondary sections), it makes sense to consolidate the SSA and RMSA into a single centrally sponsored scheme. Savings can be used more productively. An honest, calm and evidence-driven rethink is needed about education policies. Bold and imaginative reforms are required to deal with the parlous situation of very low VFM that has long being achieved from government’s educational funding. One thing is certain: the payoffs from wisely chosen and courageous reform that increases quality and learning levels in government schools will be extremely high, both in terms of government popularity and, more importantly, in terms of individual productivity and national economic growth. 27 References Altinok, N. and G. Kingdon. “New Evidence on Class Size Effects: A Pupil Fixed Effects Approach”, Oxford Bulletin of Economics and Statistics, 74, No. 2; p203-234.April 2012. Aslam, M., A. De, G. Kingdon and R. Kumar (2010) “Economic Returns to Schooling and Skills – An analysis of India and Pakistan, Mimeo, RECOUP Project, Faculty of Education, University of Cambridge. July. Atherton, P. and G. Kingdon (2010) “The relative effectiveness and costs of contract and regular teachers in India”, WPS/2010-15, CSAE, Department of Economics, University of Oxford. Colclough, C., G. Kingdon and H. Patrinos. “The Changing Pattern of Wage Returns to Education and its Implications”, Development Policy Review, 28, 6, 733-747. 2010. Desai, S. A. Dubey, R. Vanneman and R. Banerji (2008) “Private Schooling in India: A New Educational Landscape”, India Policy Forum, Brookings-NCAER, New Delhi. French, R. and G. Kingdon (2010) “The relative effectiveness of private and governmentschools in rural India: Evidence from ASER data”, DoQSS Working Paper 1003, Institute of Education. Goyal, S. and Pandey, P. (2011), 'Contract Teachers in India', Education Economics, 2011. Hanushek, E. A. and L. Wöessmann (2008) “The role of cognitive skills in economic development”, Journal of Economic Literature 2008, 46:3, 607–668. Hanushek, Eric, (2003) “The failure of input based policies”, Economic Journal. 2003. IMRB (2014) “National Sample Survey of Estimation of Out-of-School Children in the Age 6- 13 in India”, Social and Rural Research Institute, IMRB, India Market Research Bureau and EdCil, Educational Consultants India Limited, Delhi. September 2014. Atherton, Paul and G. Kingdon. “VFM Assessment of DFID Aid to Education in India”, A cost benefit analysis of UK aid to SSA, RMSA, and the Technical Capacity Fund. Report to DFID, October 2010. Kochar, Anjini (2007) “Can Schooling Policies Affect Schooling Inequality? An Empirical Evaluation of School Location Policies in India”, India Policy Forum, 2007/08, Volume 4. New Delhi. Muralidharan, K. and Sundararaman, V. (2010), 'Contract Teachers: Evidence from India', Department of Economics, University of California at San Diego. Muralidharan, K., J. Das, A. Holla, A. Mohpal (2014) “The Fiscal cost of weak governance: Evidence from teacher absence in India”, National Bureau of Economic Research, NBER Working Paper No. 20299, July 2014. 28 NUEPA (2014) “Education For All: Towards Quality with Equity”, National University of Educational Planning and Administration, New Delhi. Pritchett and Murgai (2008) “Teacher compensation: Can decentralisation to local bodies take India from perfect storm through troubled waters to clear sailing?” India Policy Forum, 2006/07, Delhi. Ramachandran, Vimala (2015) “Teachers in the Indian education system: Synthesis of a nine- state study”, National University of Educational Planning and Administration, NUEPA, March 2015. Wöessmann, L., and M. West (2006), "Class-size effects in school systems around the world: Evidence from between-grade variation in TIMSS", European Economic Review, 50, p.695- 736. 29 PAPER I BENEFITS OF INVESTING IN PRE-PRIMARY EDUCATION 30 Benefits of Investing in Pre-Primary Education Government Focus on Primary Education High-quality and well-timed preschool education has been shown, through sound scientific evidence, to provide foundational learning during the vital early years influencing cognitive functioning, socio-emotional development, self-regulation and overall health (Sylva et al., 2010; Yoshikawa, 2013). Furthermore, research on investments in preschool suggests that early childhood development directly influences economic, health and social outcomes for individuals, society and the government (Heckman, 2011). The rationale for preschool education currently is less an issue of "Why invest in early learning?" but rather a question of "How to structure and implement an effective program?" alongside functioning primary school systems. In India, the Government is committed under the RTE Act, 2009, to provide free and compulsory elementary education of satisfactory quality for all children between the ages of 6 to 14 years. SSA, a centrally sponsored scheme, is currently the main vehicle for realizing this vision which has demonstrated some success, as evident from the net enrolment rate at the primary level which has risen to 88% in 2013-14 and the annual drop-out rate which has declined from 9.4% in 2007-8 to 4.7% in 2013-14 (DISE, 2013-14). While this increased efficiency in the education system is certainly a reason to celebrate, the concern is the persisting low levels of learning in the early grades with a significant number of children in grade 5 unable to master the basics, leading to a cumulative learning deficit (NCERT, 2014; ASER 2013). In response to this concern, the Ministry of Human Resource Development (MHRD), the department tasked with school education system, has launched a new scheme titled “Padho Bharat Badho Bharat” with a focus on quality improvement in grades 1 and 2 for which specific funding is provided under SSA. This additional intervention in the early grades is well-justified as weak foundational skills have been proven to lead to chronically low learning levels in higher grades (Karoly, L.M., et al, 2005). While poor learning outcomes can be attributed to multiple variables in the school system, there has traditionally been a focus on transforming the “teacher factor" – either teacher absenteeism, poor quality of teaching or lack of skills among teachers which is understood to be correctable through further training and monitoring. This paper argues for a shift in perspective from the teacher to the student, the centerpiece of the learning process, advocating for early exposure to preschool curriculum and instruction. Shifting Focus to the Child and to School Readiness Recent longitudinal research from three states in India, on approximately 12,000 children, provides convincing evidence that a significant factor for low learning levels in early grades is inadequate school readiness (ASER, 2013). Children are entering Grade 1 with inadequate cognitive and language competencies, many of which are basic prerequisites for primary school curriculum (IECEI, 2015). This data is critical in light of international research which suggests that children beginning school with sufficient cognitive, behavioural and social-emotional skills required are more likely to benefit from the learning experiences provided as well as experience long-term success (Graue, 1992; Meisels, 1999). The legislative and policy mandate in India is for all children to start school at the age of 6. However, a significant number of students come from homes where oral culture dominates and 31 there is minimal exposure to literacy-rich environments. These children often lack access to print materials and parental role modelling of language which has the potential to build their vocabulary. As a result, they lack basic cognitive and language competencies which adversely influence their learning levels in school. Furthermore, the current provisions for pre-primary learning occur at the level of the AWCs which have, in practice, focused on delivering health and nutritional with minimal attention given to pre-primary education. Additional resources, such as structured educational curriculum and pedagogical training for Anganwadi workers (AWW), are required in order for AWCs to effectively facilitate early childhood education delivery. Therefore, a critical need exists currently for children's school readiness to be developed through proper preschool curriculum and instruction during the early years. The Importance of the Early Years It is imperative to understand why the early years are a critical time for investing in preschool education in order to maximize school level outcomes. Recent multidisciplinary research from the fields of neuroscience, behavioural sciences, economics, education and child development suggests that 90% of brain growth occurs before a child is 6 years old and the quality of a child's environment, during these early years, strongly influences this brain development (Reiss et al., 1996). Research also indicates that the critical periods for development of language, cognitive and socio-emotional competencies in the form of neural connections occur in the first seven years of life (Doharty, 1997). Therefore it becomes important, during this period, that a child is given exposure to a supportive environment and vibrant educational experiences for the development of competencies, which despite neuroplasticity, become difficult and resource intensive to address at the school level or in later years. Furthermore, children from disadvantaged backgrounds who experience poverty during their preschool years have lower rates of school completion than children and adolescents who experience poverty only in later years (Brooks-Gunn, J & Duncan, G.J., 1997). For children lacking a supportive home environment due to socio-economic disadvantages or lack of knowledge among caregivers, the effects are visible in school in the form of low learning levels, poor self-regulated behaviour, attention and working memory deficits as well as issues of social maladjustment. Pre-Primary Education and Learning Outcomes Global evidence as well as India-specific research confirms that even one year of high-quality pre-school education, when children are between ages 4 and 5, leads to a significant increase in school readiness levels at the time of entry to grade 1, which in turn contributes to enhanced levels of learning in primary grades (IECEI, 2015). A study by NCERT of approximately38,000 children in eight states has provided compelling evidence on the impact of preschool participation on retention rates in primary schools, which improved by 20.5% for preschool participants in comparison to children who had not received preschool education (Kaul et al, 1994). Evidence also indicates that while participation in preschool has many benefits, these are considerably enhanced if the program is high-quality and leads to sustained impact on learning levels. A recent meta-analysis including evaluations of 84 preschool programs concluded that on average, children gain about a third of a year of additional learning across language, reading and math skills from high-quality preschool education. At-scale preschools in Tulsa and Boston have produced even larger gains of between a half and a full year of additional learning in 32 reading and math. Furthermore, gains are significantly higher for children from disadvantaged backgrounds or with special needs (Yoshikawa et al, 2015). A review of specific evaluations of programs designed to enhance cognitive ability in mathematics between ages 4 to 5 years (Case, Griffin and Kelly, 1999; Kaul et al 1991) has concluded that strong preschool programs are required to provide a sound foundation in the critical early years necessary to improve the mathematics performance of students at the school level. (McCain, M.N. and J.F. Mustard, 2002). Research further demonstrates that the sustainability of the benefit in terms of improved learning levels is greater if there is continuity in curriculum and pedagogy from the first two years of pre-primary to the first three primary grades, thus making these early years the “foundation fives” (Crouch, in Press) which need to be given special attention and duly strengthened and promoted. These five years are also referred to as the ‘early learning unit’ in India’s XIIth FYP and emphasized as the early learning continuum which, if appropriately scaffolder, can ensure a sound foundation for children, leading to better learning levels in schools. The XIIth FYP recommended introducing a minimum of a year of preschool education for 4 to 5 year olds in primary schools. Additional Benefits of Pre-Primary Education Research indicates that the benefits of quality pre-primary education clearly outweigh the costs and include gains beyond improved learning outcomes and higher incomes. (Yoshikawa, 2015). These benefits may accrue from reduced spending on remedial education at school level, less drop-outs, better social and academic competence and, in the long-term, from lower incidence of deviant behaviour and juvenile delinquency and savings from counselling and the criminal justice systems. Additionally, while long-term benefits can be expected from a better skilled population, increased economic productivity, higher individual earnings; supplemental benefits include improved health outcomes and lower crime rate, which would overall benefit the society and the government socially and economically. Gains from Preschool Education Cost benefit analysis of investing in pre-primary education suggests significant benefits for government and society. -30% higher graduation rate Established longitudinal studies have analysed adult -40% lower repetition rate outcomes of participants from influential preschool programs such as the High/Scope Perry Preschool and the -14% higher income/ person: Chicago Child Parent Centers (CPC) and indicated $156,490 more over lifetime (High/Scope Perry Preschool Study) impressive benefit to cost ratios of $7 ( in savings) to $1 ( in cost) (Masse and Barnett, 2002). Estimates from the -7–12% increase in future income Abecedarian Project, another influential preschool for each year of preschool (Chicago CPC intervention, produced a ratio of 2.5:1. A similar study in Study) Brazil demonstrated the benefit to cost ratio for investing in preschool education to be 2:1. Furthermore, returns to investment analysis indicated that program participants could expect a 7 to 12% increase in future income (Young, 2002). According to Lancet (2011), a 50% increase in preschool enrolment in low and middle income countries would generate an estimated $34 billion at 6. 4 - 17.6:1 rate of return. A review of research indicates that while there may be variation in exact ratios of benefits to cost, the “best current evidence suggests that the impact of quality preschool per dollar spent on cognitive and achievement outcomes is larger than the average impact of well-known educational interventions per dollar spent, such as class-size reductions in elementary schools” (Yoshikawa et al, 2015). 33 Cost-Benefit Analysis of Investing in Pre-Primary Education An analysis of short and long-term returns to investment in pre-primary education, if integrated into the government system in India, indicates significant potential benefits (The World Bank, 2015). The analysis examines Models 1, 2 and 3 for pre-primary education using a combination of revitalizing programs, mobilizing existing resources and integrating current working systems. Cost benefit analysis is undertaken in order to evaluate the cost to the system for pre- primary education inputs against potential short-term and long-term savings for government and society. International and India-specific evidence is synthesised in order to gather evidence for how economic gains accrue from investing in pre-primary education during the early childhood years. As data repeatedly advocates that preschool programs must focus on quality and duration, Model 1, 2 and 3 represent a continuum from most comprehensive intervention to a convergence model to a focused but succinct package program to improve all early learners' school readiness. Table 10: Models for Investing in Pre-Primary Education Model Advantages Model 1: TWO YEARS OF • Comprehensive and high-quality PRE-PRIMARY • Newly developed infrastructure, teacher training EDUCATION in and preschool curriculum Government Schools • Situated in Government school system in order to ensure continuous transition from pre-primary to primary grades Model 2: CONVERGENCE • Optimizing / integrating existing resources between existing government • Adding new resources, improving infrastructure departments to strengthen • Improved efficiency of existing systems preschool education at AWCs Model 3: SCHOOL • Highly focused and intense preschool pre-literacy READINESS PACKAGE for and pre-math instruction 30-60 days of preschool • Minimal costs with potential of ensuring school intervention at the start of readiness for all learners Class 1 Model 1: Two Years Pre-Primary Education Description, Costs and Advantages Model 1 for pre-primary education entails adding 2 years of pre-primary education in the current government school system in India, including newly trained educators, pre-primary curriculum and teaching methodology, to ensure high-quality, pre-primary learning for the early childhood years. Students entering government schools are introduced to pre-literacy and pre-math learning intended to develop cognitive, linguistic and socio-emotional school readiness skills prior to entering primary instruction. Advantages for Model 1 include a comprehensive and high-quality program; newly developed intervention built from the ground- up; a program situated in the Government school system in order to ensure continuous preschool to primary transition. Global Evidence Seminal studies such as High/Scope Perry Preschool, the Abecedarian and Chicago CPC Studies illustrate that preschool interventions yield the greatest gains when there is a focus on 34 high-quality, sustained curriculum and instruction that is delivered by trained educators in a well-resourced and structured environment (Schweinhart, L.J. et al, 2005; Masse, L.N., Barnett, W.S., 2002; Reynolds, A.J., 2001). Evidence from these longitudinal studies indicates significant benefits for those who attended preschool in comparison to those who did not. Program participants had better education attainment (30% higher graduation rate; 40% lower repetition rate), and higher income over a lifetime (14% higher income per person or an average of $156, 490; 7 to 12% increase in future income for each year of preschool). Furthermore, individual benefits such as higher income contributed to government savings from increased tax revenues over a lifetime (28% increased tax revenues associated with higher expected earnings), reduced government subsidies and spending on remedial or corrective education. ROI ratios from cost-benefit analysis range from $7 to $16 per dollar invested illustrating the high return from investing in pre-primary education. Finally, preschool education can be expected to contribute to GDP increase in 2080 of 3.5% to 4% from a preschool program started in 2005 (Dickens et al, 2006). India-specific Data Research in India also points towards higher returns from pre-primary education for disadvantaged populations. Preschool exposure improved learning outcomes with students' average test scores improving positively and significantly as pre-primary education quality improved (Kaul et al, 2014). Students were also likely to have better educational attainment with 20% higher retention for preschool students and 17% less likely to drop out by Class IV (Kaul et al, 1993). Finally, systematic challenges like child labour, social exclusion and economic backwardness can be overcome through Government sponsored preschool programs for children from disadvantaged backgrounds. Cost-benefit analysis for Model 1 suggests that adding these 2 years of pre-primary can generate a Cost-Benefit Ratio of 25, or Rs.25 saved for every Rs. 1 invested. Benefits from Cost-Benefit Analysis Adding 2 years of pre-primary stage to government elementary schools would lead to benefits for the individual, society as well as the Government and these benefits may potentially be magnified as they are sure to include disadvantaged student populations. Model 1 shows a requirement of an initial investment of Rs. 13,600 Crores and an annual spend of Rs. 12,650 Crores. This estimate is based on the understanding, as per DISE data, that there is a significant number of small schools across states with declining enrolments which could accommodate the additional pre-primary grades with minimal expense. The corresponding cost-benefit analysis for Model 1 suggests that adding these 2 years of pre-primary can generate a Cost- Benefit Ratio of 25, or Rs. 25 saved for every Rs. 1 invested. 35 Benefits of Pre-Primary Education Further benefits include a skilled workforce and increased labour productivity with 30% to 60% additional income over an individual's lifetime. Additionally, spill over benefits to society can be seen in the form of 22 percent reduction in crime (basis international research) and improved health outcomes finally, benefits of pre-primary education are expected to surpass costs required to set up the system, with savings of Rs 171,000 Crores expected within the first 10 years. Pre-primary investments also contribute to a significant increase in GDP. As more pre-primary educated cohorts enter the workforce with higher incomes, 4.3% additional GDP can be expected by 2050 and 7% additional GDP can be expected by 2074. Figure 1 provides a summary of the analysis for reference. Ease of Implementation Benefits for Model 1 highlight the great potential value of a comprehensive 2 year pre-primary education model. Accordingly, Model 1 entails significant requirement of resources, development and integration into the current Government school system in order for the program to function effectively and successfully. Challenges entail the entire movement of preschool delivery from one government department to another and the corresponding logistical shift in responsibilities and obligations that would require. Furthermore, considerable investment is required in terms of infrastructure, advisory, materials and training. However, transferring preschool education to the school system ensures an optimal leveraging of school related resources and transferrable expertise in curriculum and pedagogy as well as allows for the most efficient and logical transition of the preschool learner into the primary grades. 36 Model 2: Convergence Model Description, Costs and Advantages Model 2 focuses on strengthening pre-primary education within existing government AWCs with convergence and coordination between the Ministry of Women and Child Development (MWCD), under the Integrated Child Development Services (ICDS) program, and the MHRD. This model requires construction of new AWCs so that they may function as classrooms, adding a dedicated and trained pre-primary teacher in order to provide instruction at each AWC beyond the AWWs. Costs also include teacher training for pre-primary teachers and supervisors, teaching learning materials (TLM) and additional central costs. Advantages for Model 2 include integrating resources, improving infrastructure, bringing in new resources and improving efficiency of existing government programs. Evidence Early Childhood Care and Education (ECCE) for children below 6 years is currently under the charter of MWCD as the ICDS program, which is mandated to offer preschool education for 3 to 6 year-olds as one of its six services (supplementary nutrition; provision of non-formal pre- school education; nutrition and health education; immunizations; health check-ups; referral services). ICDS has a reach of 13.4 AWCs with an enrolment of 3.7 Crore children and is the sole provider of Government ECCE. However, research shows that less than half of AWCs have teaching learning materials and several states report nil yearly expenditures of pre-school kits allocated under the ICDS budget suggesting a lack of delivery of pre-school education. A recent study which explored the pathways followed by children from 3 to 6 years revealed that with the rising aspirations of parents for enrolling children early in organized education, and rapid expansion of private provisions, the overall trend in parental choice is of 4 year-olds being moved out of AWCs into either private preschools, if the parents can afford it, or to government primary schools as underage children (IECEI, 2015). About 20% children in government schools in Class 1 were found underage and were attending school without formal enrolment. Despite 1.3 million AWCs in place currently, the IECEI study found low school readiness levels among 5 year-olds at the time of entry to Grade 1. The AWCs, due to multiple responsibilities as well as infrastructure and human resource constraints, are not able to provide a systematically organized preschool education program which is required for children over 4 years. In addition, in at least 23 states of India the age for entry to Grade 1 is 5+ years, so that 5 to 6 year-olds are already not in AWCs but formally enrolled in the schools. Benefits from Cost-Benefit Analysis Model 2 calls for a revitalization of the current ICDS program utilizing resources available at both MWCD as well as MHRD with shared cooperation and costs between both government departments. The Convergence Model also utilizes continuing assumptions from Model 1 from high-quality pre-primary programs, such as High/Scope Perry Preschool and Chicago CPC, but it also integrates evidence from the U.S. Headstart program which, similar to ICDS, is a large- scale long-running government program offering a combination of pre-primary education, nutrition and health interventions. Data from the Headstart Impact Study (2010) indicates that Headstart provides 80% of the effects of well-known early childhood programs such as Perry Preschool and Abecedarian and Model 2 accordingly assumes 80% of benefits indicated in Model 1. Additionally, students who had attended Headstart were 17% more likely to complete high school than non-participants as opposed to 31% increase in high school completion in Perry Pre-school program and Model 2 adjusts for these differences as well. Additional spill over benefits also indicate that students who had attended Headstart were less likely to be arrested at age 22 (5% vs. 15%) suggesting savings in crime related spending. Overall the 37 Headstart program has significant benefits albeit lesser than standalone high-quality pre-school programs. The analysis for Model 2 indicates short-term savings from reduced government spending on remedial education (Rs. 12,333 Crores) as well as household tuition spending (Rs. 5,800 Crores). Furthermore, it predicts long-term savings from reduction in government subsidies (Rs. 1719 Crores in 2032; Rs. 23, 355 Crores in 2074); household savings from higher income (Rs. 12, 300 Crores in 2032, Rs. 37.2 Lakh Crores in 2074); and subsequently government savings from increased tax revenues (Rs. 145 Crores in 2032, Rs. 3.5 Lakh Crores in 2074). Overall, Model 2 yields a cost-benefit ratio of 10, or a savings of Rs. 10 for every Rs. 1 invested in pre-primary education. Ease of Implementation Model 2 or the Convergence Model leverages the capacity of multiple governmental departments thus improving delivery of existing programs which in return saves on costs as well as improves efficiency. Effective inter-sectoral coordination requires regulatory, operational and financial convergence between multiple ministries and can be challenging. However, despite the obstacles, the Convergence Model may have potential as MWCD released a curriculum framework in 2012 for ECE outlining pedagogical approaches and curriculum content along with the role of teachers and parents in early learning. In 2013, the newly adopted National ECCE policy set a vision for ECCE and stipulated an institutional arrangement focusing on convergence with the MHRD. Model 3: School Readiness Package Description, Costs and Advantages School readiness research suggests that children lacking in basic cognitive and language competencies perform adversely in early grades. Model 3, or the "School Readiness Package", attempts to mitigate these gaps by introducing pre-primary education, consisting of pre-literacy and pre-math learning, in first 30 to 90 days of Class 1. The package comprises an intense program of instructing and reinforcing basic skills in order to gain a threshold level of competency for young learners over a short, focused period of time. Model 3 requires minimal one-time costs of content development, printing of activity books, TLM, teacher training and additional central costs. Advantages for Model 3 include a focused, short-duration program that seeks to deliver an intervention of basic early skills to prepare all students to learn. Evidence Research suggests low learning levels in early grades may be due to a lack of school readiness. School readiness is a measure of how prepared a child is to do well in school from a cognitive, social and emotional perspective. A recent study conducted across three states in India – Assam, Telangana and Rajasthan – with 2500 six-year-olds suggests cognitive and language competencies are lacking for in-coming students resulting in adverse conditions for learning in primary grades (IECEI, 2015). Some of this may be attributed to first generation learners entering school as a result of RTE 2009. Model 3 offers young learners the opportunity to bridge the growing gap between pre-literacy and pre-math skills and primary learning. The model is an equitable intervention as it is offered to all students and reinforces skills, thus building on success, as well as teaches new basic skills. The School Readiness Package also utilizes underlying assumptions from Model 1 and particularly examines global and India-specific evidence in order to synthesize understandings 38 of how short-term education interventions may lead to significant but small education gains for students. Model 3 examines the Pratham Balsakhi Intervention, a Learning Enhancement Program (LEP) or remedial intervention, which resulted in an increase of 0.4 points out of 12.5 total points or 3.2% increase in reading and an increase of 0.3 points out of 6 total points or 5% increase in writing1 (approximately 4% average academic gain). In comparison to Model 1 where high- quality pre-school programs, such as the Perry Pre-school program, showed academic gains of 34% for participants, this indicates a small but significant gain. Hence, estimation of benefits in Model 3 is expected to be scaled down by a ratio of (4% by 34%) or 11.7% of Model 1 benefits. Benefits from Cost-Benefit Analysis Analysis of benefits in Model 3 reveal short-term savings from reduced household spending on tuition spending (Rs. 920 Crores). More significantly, it predicts long-term savings from reduction in government subsidies (Rs. 594 Crores in 2032; Rs. 2,816 Crores in 2074); household savings from higher income (Rs. 5,176 Crores in 2032, Rs. 8.3 Lakh Crores in 2074); and subsequently government savings from increased tax revenues (Rs. 101 Crores in 2032, Rs. 77,451 Crores in 2074). Overall, Model 3 yields a cost-benefit ratio of 75, or a savings of Rs. 75 for every Rs. 1 invested in pre-primary education. However, due to the minimal costs incurred in this specific model, benefits are not relevant in absolute terms. Finally, a comparison between models is presented below which synthesizes cost-benefit analysis across all 3 models in order to offer a comparison of savings for investments in pre- primary education. Models 1, 2, and 3 can be evaluated on the basis of three key metrics: the amount of investment required, years needed to recover costs and the cost-benefit ratio. Model 1 requires an investment of Rs. 14,021 Crores but recovers costs the quickest while Model 2 requires the steepest investment of Rs. 64, 102 Crores as well as the greatest duration of time to recover costs with a reasonable cost-benefit ratio of 10. Finally, Model 3 requires minimal costs of Rs. 1259 Crores and recovers cost by Year 12. However, cost-benefit ratio for Model 3 or the School Readiness Package is 75 and not relevant in absolute terms due to the minimal costs required in the model. The comparison of all 3 models suggests that 2 years of government pre-primary education would recover costs the most efficiently (by year 7) as well as present a meaningful cost-benefit ratio of 25. Annexure 1 provides further details and assumptions for cost-benefit analysis of investing in pre-primary education as per Model 1, 2 and 3. 39 Comparison between models Model 1: Model 2: Model 3: Key metrics Pre-primary under Convergence School Readiness MHRD Package Investment required (Cr)* 14,021 64,102 1,259 Recovered by Year 7 Year 19 Year 12 PV of costs (Cr) 134,429 198,459 5,640 PV of benefits (Cr) 3,357,973 1,885,658 427,436 Benefit-cost ratio 25 10 75* Benefit-cost ratio not relevant in absolute terms due to minor costs in Model 3 10 Recommendation Based on the review and the analysis of Model 1, 2 and 3, presented above, this paper strongly recommends that the MHRD initiate the process to include pre-primary education as part of the RTE (2009) and support states in introducing a pre-primary section in all their government primary schools for 4 to 6 year olds. It is also recommended that the curriculum for the pre- primary and first three years of primary be conceptualized and developed in a ‘bottom up’ manner ( and not top down as is often the case) to ensure continuity in learning as well as a sound early learning foundation for children. Again, this recommendation is in alignment with the RTE (2009) Section 11 which encourages all state governments to endeavour to establish preschool classes for children from 3 to 6 years in order to help them develop school readiness for primary grades. Possibilities may be considered of setting up multiple models as per feasibility, including relocating AWCs to primary schools and ensuring continuity of curriculum, or setting up a one to two years pre-primary section in schools prior to grade 1. In 2015, the NCTE announced a revised teacher education curriculum which supported pre- primary education and encouraged states to set up a cadre of teachers especially trained for pre- primary and early grades. This investment in pre-primary education which ensures significant returns, and which is specially intended to benefit children from marginalized communities, will contribute in a significant way towards narrowing the social equity gap and ensuring learning for all, one of the Sustainable Development Goals (2015) which India and the international community has made a strong commitment towards. 40 Details of Cost Benefit Analysis for India Cost-Benefit Analysis MODEL 1: Assumptions for Cost & Benefit Key assumptions for costs calculations Costs Key assumptions Construction of classrooms Additional classrooms required in only those primary schools with >30 PCR i.e. ~30%; 1 classroom per school; Rs. 5 lakh / classroom; Phased construction and intake over 5 years Cost of toilets Additional toilets not required Mid day meals MDM: Rs. 1,250/student; Other: Rs. 100/student Other: stationery, etc. (assuming that no uniforms / books etc. provided to students) Teaching learning materials Rs. 1,000 per classroom Teacher salaries and training Schools with >20 PTR will require additional teacher for pre-primary i.e. costs ~70% of schools 1 teacher per school; Rs. 10,000/month salary 30 days induction training @ Rs. 200 per day; 20 days in-service training @ Rs. 200 per day 11 Key assumptions for benefits calculations Benefits Key assumptions Immediate Savings: • ICDS Savings 1/6th of ICDS budget assumed to be spent towards ECE for 3-6 year olds (given 6 key tasks under ICDS charter) => cost savings of Rs 831 / student; Nutrition expense only towards 4, 5 year olds assumed saved = Rs 297/ student • Remedial Savings Savings in LEP line item under SSA budget - starting at 10% in 2017 to 70% savings by 2021 (once ECE kids are in class 1-5); 70% benchmark from Chicago CPC Study); excludes Padhe Bharat Badhe Bharat allocation which will be required for ensuring high learning levels in foundation years • Tuition Savings 45% primary school children take afterschool tuition (ASER); average tuition spend – Rs. 1,200 p.a.; savings of 70% for ECE participants (Chicago CPC Study) Long term Savings: 1. Class 12 completion rates increase over time, maxing at 60% (Brazil is • Increased Income around there. Australia is ~80%) 2. % of school / college grads entering workforce assumed at 60% (20% of women and 100% of men) 3. Starting salaries computed basis trades associated with the 3 segments – class 10 pass outs, class 12 pass outs, and college graduates 4. These segments would have otherwise earned as much as class 8 pass outs, class 10 pass outs and class 12 pass outs respectively. 12 41 Key assumptions for benefits calculations Benefits Key assumptions Long term Savings: • Increased Tax Revenue 1. Class 12 completion rates increase over time, maxing at 60% (Brazil is around there. Australia is ~80%) 2. % of school / college grads entering workforce assumed at 60% (20% of women and 100% of men) 3. Starting salaries computed basis trades associated with the 3 segments – class 10 pass outs, class 12 pass outs, and college graduates 4. Income tax slabs changed every 3 years assuming 3% p.a. real growth in tax slabs • Subsidy Savings 1. Subsidies linked to BPL – food, kerosene, and MNREGA – assumed to remain the same in absolute terms going forward (historically as well, real growth has been ~1.5%) 2. % BPL population reduced basis 30-year historical CAGR of ~2.2%; have taken a floor of 10% BPL families 3. BPL families will come out BPL status as ECE kids join the workforce (assuming 2 kids per family) – this will result in subsidy reduction 4. Subsidy savings capped at 25% 13 This investment can be recovered quickly in the short term itself, and will also generate high returns in the long term Benefits quantified as part of this study Short term benefits Long term benefits ICDS cost reduction as ECE for 4 Subsidy reduction from reduction in Reduced and 5 year olds moves out number of BPL families costs Reduced spending on remedial ed. • Better learning outcomes Government • Lower repetition, less drop-outs Increased tax revenues Increased • Growing skilled workforce Income • Higher employability • Higher income Reduced spending on tuitions with Reduced better learning outcomes costs Society Higher Income given higher labour Increased productivity and skilled workforce income • Higher school completion/ graduation rates • Higher starting salaries 6 42 Investment recovery expected within 7-8 years; cost savings of Rs. 26,600 Cr. annually in steady state ICDS Lower spending on Lower spending on Total Source of cost reduction remedial education tuitions (govt. + society) savings Offsets total (as much as 70% (~50% children of low investment in savings per global income families take first 10 years of studies1 ) tuitions) ~ 102,400 Cr. Total savings in 1st 10 years ~ 30,400 Cr. ~ 103,400 Cr. ~ 37,150 Cr. ~170,950 Cr. (2016-2025) Excludes "Padhe Bharat Badhe Bharat" budget, which will be required to Annual savings ensure high learning levels in steady state (once all children in primary classes 1-5 ~ 3,800 Cr. ~ 15,400 Cr. ~ 7,400 Cr. ~ 26,600 Cr. have 2 years of pre- school education) xx Benefit to Govt. xx Benefit to Society 7 Long term benefits begin with preschool students entering the workforce: ~25,800 cr. in 2032 and growing at ~14% annually Incremental Increased Tax Subsidy reduction Total benefit Income – skilled Revenue Source of workforce; higher benefit productivity (with higher incomes) (as families move out of (to Government (30-60% increase over BPL faster) and society) a person's lifetime) Benefits in 2032 – when 1st ~ 22,400 Cr. ~ 260 Cr. ~ 3,100 Cr. ~25,800 Cr. ECE cohort fully enters workforce Benefits in steady state ~2074 – when entire workforce up ~ 67.8 Lakh Cr. ~ 6.3 Lakh Cr. ~ 23,350 Cr. ~ 74.3 Lakh Cr. to 60 yrs of age has undergone ECE xx Benefit to Govt. xx Benefit to Society 8 43 Overall, in 2015 Rs. terms, for every Re. 1 invested, returns up to Rs. 25 will be generated! 56,446 36,531 . 201,223 137,256 431,456 ) y 2,869,785 2,926,518 ) 56,733 s 3,357,973 ) t OpEx= Benefit-cost 134,430 ) 123,588 ratio of 25 CapEx= 0 10,842200,000 400,000 600,000 3,400,000 In Crore Rupees ncome Gains Taxes on Earnings Subsidy Savings Remedial ICDS 10 MODEL 2: Assumptions for Cost & Benefit Model 2: Assumptions for key costs and benefits Costs Benefits 1. Construction of Anganwadis such that they • Estimation of benefits is based on a parallel can function as proper classrooms program, Head Start, a federal program in the – 54% 'kachha' Anganwadis will need this US which combines pre-primary education with health and nutrition, similar to ICDS. 2. Dedicated and trained pre-primary teacher in 1. Research indicates that HS provides 80% each AWC (over and above AWWs) of the effects of well-known early childhood programs such as Perry 3. Teacher training Preschool and Abecedarian1 a. Costs for training pre-primary teachers 2. Students who had attended HS were 17% more likely to complete high school than b. Per day payment model to MHRD non-participants2 assumed for additional training infra – [as against 31% increase in high required school completion in Perry Pre-school 4. Central costs program3] a. 1 pre-primary expert per state (within 3. Students who had attended HS were less ICDS org) likely to be arrested at age 22 (5% vs 15%)2  Overall program has significant benefits albeit 5. Pre-primary in-service training for all lesser than stand alone high-quality pre-school 50,000 supervisors in ICDS System programs 1. Deming, D. (2009) 2. Barnett, W.W. and Hustedt, J.T., (2005) 3. Schweinhart, L.J., et al. (2005) 2 44 Model 2 – Benefits for government and society Benefits quantified for Convergence Model (in steady state value) Short term benefits Long term benefits Long term benefits (in steady state) (in 2032) (in 2074) Reduced spending on Subsidy reduction = Subsidy reduction = remedial education = Rs. Rs. 1,719 Cr. Rs. 23,355 Cr. Reduced costs 12,333 Cr. (80% of model 1) Government Increased tax Increased tax Increased revenues = revenues = Income Rs. 145 Cr. Rs. 3.5 Lakh Cr. Reduced spending on Reduced tuitions = Rs. 5,800 Cr. costs (80% of model 1) Society Increased Higher Income = Rs. Higher Income = income 12,300 Cr. Rs. 37.2 Lakh Cr. 4 Model 2: Overall, in 2015 Rs. terms, for every Rs. 1 invested, returns up to Rs. 10 44,424 govt. 109,805 264,704 erms) 110,475 ciety 1,575,568 1,620,954 erms) 45,386 efits 1,885,658 erms) OpEx= 169,567 Cost Benefit-cost 198,458 erms) ratio of 10 CapEx= 0 28,891 500,000 1,000,000 1,500,000 2,000,000 In Crore Rupees on Income Gains Taxes on Earnings Subsidy Savings Remedial 5 45 MODEL 3: Assumptions for Cost & Benefit Model 3: Assumptions for key costs and benefits Costs Benefits • School Readiness Package Model will require • Research on short-term programs, similar to the minimal costs school readiness package, show significant but small gains in student achievement  One time Content development costs  Pratham Balsakhi Intervention, a LEP  Printing and distribution of materials intervention, resulted in an increase of 0.4 including activity books for English, Math points out of 12.5 total points or 3.2% increase and Regional Language content in reading and an increase of 0.3 points out of 6 total points or 5% increase in writing1 (~ 4%  Teaching Learning Materials per average academic gain) classroom  Teacher training on how to transact the • On the other hand, full fledged pre-school package (in-service) programs (e.g. Perry Pre-school program) shows  Central Costs academic gains of 34% for participants2  Hence, estimation of benefits expected to scale down by a ratio of (4% by 34%) = 11.7% of Model 1 1. Banerjee, A., et al., (2005) Deming 2. Schweinhart, L.J., et al. (2005) 6 Model 3 – Benefits for government and society (all benefits scaled down to 11.7% of model 1) Benefits quantified for School Readiness Package Short term benefits Long term benefits Long term benefits (in steady state) (in 2032) (in 2074) Subsidy reduction = Subsidy reduction = Rs. 594 Cr. Rs. 2,816 Cr. Reduced costs Government Increased tax Increased tax Increased revenues = revenues = Income Rs. 101 Cr. Rs. 77,451 Cr. Reduced spending on Reduced tuitions = Rs. 920 Cr. costs Society Increased Higher Income = Rs. Higher Income = income 5,176 Cr. Rs. 8.3 Lakh Cr. 8 46 Model 3: SRP Cost Benefit comparison *Benefits not relevant in . absolute 26,757 7,466 34,223 terms s) y 7,184 386,030 393,214 s) s 427,436 s) OpEx= 5,640 t Benefit-cost 5,640 s) ratio of 75* CapEx= 0 0 20,000 40,000 420,000 440,000 In Crore Rupees Income Gains Taxes on Earnings Subsidy Savings 9 47 References ASER (2013), Annual Status of Education Report (Rural) 2013. Available online: http://img.asercentre.org/docs/Publications/ASER%20Reports/ASER_2013/ASER20 13_report%20sections/aser2013fullreportenglish.pdf Brooks-Gunn, J & Duncan, G.J. (1997). The Future of Children CHILDREN AND POVERTY Vol. 7 • No. 2 – Summer/Fall 1997 Case, R., Griffin, S., & Kelly, W. (1999). Socioeconomic gradients in mathematical ability and their responsiveness to intervention during early childhood. In D. Keating & C. Hertzman (Eds.), Developmental health and the wealth of nations: Social, biological, and educational dynamics (pp. 125-149). New York: Guilford Press. Doherty, R.W. (1997), “The Emotional Contagion Scale: A Measure of Individual Differences”, Journal of Nonverbal Behaviour, 21, pp. 131-154. Eisenberg, E., Eggum, N.D., Spinrad, T.L., (2010). Emotion-Related Self-Regulation and Its Relation to Children’s Maladjustment. Annu Rev Clin Psychol. 2010 Apr 27; 6: 495–525. Graue, M. E., Social Interpretations of Readiness for Kindergarten, Early Childhood Quarterly, vol. 7, no. 2, June 1992, pp. 225–243. Headstart Impact Study: Final Report (January 2010) IECEI (2015) Quality and Diversity in Early Childhood Education. A View from Andhra Pradesh, Assam and Rajasthan. Available online: http://ceced.net/wp- content/uploads/2015/04/IECEI-Executive-Summary-Report.pdf Karoly, L.M., Kilburn, R. & Cannon. J.S., 2005. Early Childhood Interventions: Proven Results, Future Promise. Santa Monica: RAND. Kaul, V. (1991). Early childhood education programme. New Delhi: NCERT. Kaul, V; Ramachanran C. & Upadhyay, G.C. (1994). Impact of ECE on Retention in Primary Grades: A longitudinal study. NCERT, New Delhi. Kaul, V, AB Chaudhary, and S Sharma (2014), ‘Quality and Diversity in Early Childhood Education: A view from Andhra Pradesh, Assam and Rajasthan’, Centre for Early Childhood Education and Development, Ambedkar University, Delhi. The Lancet. Child Development in Developing Countries. (2011). Available online: http://www.thelancet.com/series/child-development-in-developing-countries-2 Masse, L. N., & Barnett, W. S. (2002). A benefit-cost analysis of the Abecedarian early childhood intervention. In H. Lewin & P. McEwan (Eds.), Cost effectiveness and educational policy: 2002 Year Book of the American Educational Finance Association (pp. 157–176). Larchmont, NY: Eye on Education 48 McCain, M.N., and Mustard, J.F. (2002). Early Years Study 3: Making Decisions, Taking Action. Toronto: Margaret & Wallace McCain Family Foundation. Meisels, S. J. (1999) Assessing readiness. In R. C. Pianta & M. Cox (Eds.), The transition to kindergarten (pp. 39 – 66). Baltimore: Paul Brookes. Morrison, F.J. (2010) Self-Regulation and Academic Achievement in the Transition to School. Child development at the intersection of emotion and cognition. Human brain development., (pp. 203-224). Washington, DC, US: American Psychological Association, x, 261 pp. http://dx.doi.org/10.1037/12059-011 NCERT (2014), NAS Class VIII Summary Report. Available online: http://www.ncert.nic.in/departments/nie/esd/pdf/NAS_8_cycle3.pdf Reiss A.L., Abrams M.T., et al. Brain development, gender and IQ in children: a volumetric imaging study. Brain. 1996 119(5):1763–1774. Reynolds, A. J. (2000). Success in early intervention: The Chicago Child-Parent Centers. Lincoln, NB: University of Nebraska Press. Schweinhart, L. J., Montie, J., Xiang, Z., Barnett, W. S., Belfield, C. R., & Nores, M. (2005). Lifetime effects: The High/Scope Perry Preschool Study through age 40. Ypsilanti, MI: High/Scope Press. Yoshikawa, H., Weiland, H., Brooks-Gunn, J., Burchinal, M.R., Espinosa, L.M., Gormley, W.T., Ludwig, J., Magnuson, K.A., Phillips, D., Zaslow, M.J. (2013) The Evidence Base for Early Childhood Education. Society for Research on Child Development. Young, Mary Eming, 2002. In M. Young (Ed.), From early child development to human development: Investing in our children's future (pp. 123-142). 49 PAPER-II FINANCIAL AND ECONOMIC ASSESSMENT OF THE COMPOSITE SCHOOL MODEL (CLASSES 1-10) IN RAJASTHAN 50 Financial and Economic Assessment of the Composite Schools Model in Rajasthan The GOR has taken a strategic decision to consolidate some of its elementary schools and secondary/senior secondary schools into composite schools. The consolidation is being worked out within the guidelines set under the RTE Act. This paper explores the merits of this move by comparing the effectiveness, efficiency and impact of Elementary schools vis-à-vis composite schools. The Background The Government’s decision to work towards developing more composite schools was based on the realization that a lot of elementary schools were actually created by separating primary and/or upper primary classes from existing composite schools. This decision was the Government’s initial response to the RTE. Subsequently, the Government also opened up a number of schools to cater to students who did not come in the prescribed radius of an existing facility. While tracking the achievements of this move, the Government realized that the State had ended up creating a public school network that was difficult to sustain and was inefficient in its operations. A number of schools were operating with dipping enrolments. Further, a number of schools were not being monitored and supervised at the desired level. As a result there were schools where teachers were found to be absent or taking unaccounted affecting the quality of classroom instruction. Further, leakages were discovered in the mid-day meal scheme. The responsibility for monitoring the operations of the elementary schools has been placed on the BRCs and CRCs. The number of schools and the geographic spread to be covered only allow for bi-annual monitoring visits. Therefore monitoring and supervision is not concurrent. The rapid expansion of the public schools network has led to a scenario where the Government has had to cope with a sizeable financial outlay towards infrastructure expansion. Most of the existing outlay has been towards the development of school buildings, classrooms, toilets and drinking water facilities. However, a large number of schools are yet to gain access to inputs that are required to support quality education. This includes but is not limited to inputs such as libraries, play grounds, electricity connections and computers. The Government’s decision to move towards consolidation via composite schools was founded on the realization that this model could address a lot of the aforementioned challenges; in turn leading to the provision of better quality education. Placing the elementary grades in the same ecosystem that houses secondary and senior secondary classes would: ● Place them under the regular care and supervision of the secondary/senior secondary school Principal. This would ensure concurrent monitoring and reduce unaccounted teacher absenteeism and leakages from the system. ● Provide elementary school teachers with access to subject teachers (more qualified, trained and experienced) who are working with secondary and senior secondary teachers; thereby providing them with access to ongoing academic support. 17 ● Improve the GTR thereby increasing the number of effective learning days. 17 Grate to Teacher Ratio refers to average number of Classes/Grades a teacher has to manage at any given point of time. 51 ● Provide students from elementary schools with access to facilities such as better equipped libraries, playgrounds, functional computers etc. ● Facilitate greater retention through improved quality of education & stronger parent-teacher association (a result of improved monitoring); and improved transition rates through unconstrained sharing of student data/information. Having identified the aforementioned benefits, the Government has moved towards ensuring that every Gram Panchayat has at least one composite school. Placing this composite school near the seat of the Panchayat also ensures better monitoring and supervision from the community’s end. In order to ensure that students’ whose schools have been consolidated do not suffer due to the distance between their place of residence and the composite school, the Government provided them with a daily travel allowance (linked with their attendance). Finally, the Government acknowledged that a few of the elementary school catering to remote villages and those with high student strength (multiple classes for each grade) should be left untouched under the consolidation plan. However, in order to further improve upon the effectiveness, efficiency and impact of these schools the Government has linked them to the composite school under a hub and spoke model. Under this setup, the Principal and subject teachers of each composite school are expected to help with the monitoring and academic supervision/training of teachers at these elementary schools. This model is yet to be fully operationalized and the Government is in the process of finalizing the modalities for the same. Effectiveness Composite schools benefit from having a larger institutional setup in terms of infrastructure, teacher availability and classroom availability. It is well documented that an enabling teaching learning environment leads to better learning outcomes for children, and in this sense certain parameters considered in this analysis have been charted out for both composite and elementary setups. Taking into consideration UDISE data from 2013-14, it has been observed that composite schools perform better than elementary school on infrastructure/facilities related parameters such as availability of electricity connections, libraries, playgrounds and computers. Data given in Table 1 shows that while 84.4 percent of composite schools have access to electricity, the corresponding figure for elementary schools stands at 52.4 percent. Table 11: Schools with electricity connection 2013-14 Variable/Parameter Composite Elementary With electricity connection 84.4 percent 52.4 percent Without electricity 15.6 percent 47.6 percent connection Similarly, while 84 percent of composite schools have a library, the corresponding figure for elementary schools is 80.4 percent. Herein, it is also important to understand that most composite schools have a larger library setup and many have a dedicated librarian. On the other hand, most elementary schools do not have a library room but rather a small collection of books that are stored in a cupboard. 52 Table 12: Schools with library facility 2013-14 Variable/Parameter Composite Elementary With Library 84.0 percent 80.4 percent Without Library 16.0 percent 19.6 percent Similarly, while 84 percent of composite schools have a library, the corresponding figure for elementary schools is 80.4 percent. Herein, it is also important to understand that most composite schools have a larger library setup and many have a dedicated librarian. On the other hand, most elementary schools do not have a library room but rather a small collection of books that are stored in a cupboard. Table 13: Schools with playground 2013-14 Variable/Parameter Composite Elementary With Playground 54.0 percent 42.9 percent Without Playground 46.0 percent 57.1 percent In order to improve the quality of education being imparted in public schools, the Government has been making efforts to leverage information and communication technology enabled teaching-learning transactions. However, implementing the same requires the schools to have access to computers. It is observed that while composite schools house an average of three computers, the corresponding figure for elementary schools is only 0.5. As a result it can be concluded that many of the elementary schools do not have computers. This can also be because a large percentage of elementary schools do not have electricity connections. Table 14: Schools with computer availability for teaching learning purposes 2013-14 Variable/Parameter Average number of computers available for teaching and learning purposes Composite 3.0 Elementary 0.5 More importantly, composite are observed to be more effective in terms of the quality of classroom transactions. The percentage of teachers with a graduate degree is higher at composite schools. Further, while the schools support a higher PTR, they still work within the limits prescribed under the RTE. The PTR at composite schools is 32 students to a teacher and the corresponding figure for elementary schools is 20. The GTR at composite schools is much lower than the figure observed for elementary schools. The GTR figure has been derived by multiplying the number of schools under ‘primary’, ‘upper primary’ and ‘primary + upper primary’, ‘primary + upper primary + secondary’, ‘primary’ + upper primary + secondary + senior secondary’ + ‘upper primary + secondary’ and ‘upper primary + secondary + upper secondary’ categories with ‘5’, ‘3’, ‘8’, ‘10’, ‘12’, ‘5’ and ‘7’ respectively. The latter set of figures is simply the number of grades in each type of school. 53 The resultant figures have been divided by the number of teachers working at each type of school. The GTR estimates show that at a given point in time a teacher in an elementary school is teaching an average of almost two classes/grades. The corresponding figure for composite schools is closer to a teacher a class/grade. As a result, it can be assumed that almost half the children in an elementary school are only passive ‘listeners’ and not ‘learners’. Table 15: Teacher related parameters 2013-14 Percentage of teachers Percentage of teachers Type of School with graduation and with Professional PTR GTR above qualification Composite 89.4 percent 96.6 percent 32 1.31 Elementary 81.4 percent 96.7 percent 20 1.83 The difference in GTR when factored in to adjust the average instructional days reveals that the ‘effective instructional days’ in composite school is 159 and the corresponding figure for elementary schools is 99. Table 16: Average number of instructional days (upper primary) 2013-14 Type of School Average number of Average number of effective instructional days instructional days Composite 208 159 Elementary 180 99 The benefits of concurrent monitoring through the Secondary/Senior Secondary school Principal are clarified by the fact that composite schools report a lesser number of BRC and CRC visits; and this is despite these schools being relatively closer to the BRC and CRC offices. Table 17: Monitoring and Supervision 2013-14 (Rounded Off) Average no. Average no. Average Average of visits by of visits by Distance Distance Type of School Block Cluster from Cluster from Block Resource Resource Resource Headquarters Centre Centre Centre (in (in Km.) officer officer Km.) Composite 1 1 19 3 Elementary 2 3 24 4 Efficiency Having larger institutional setup in terms of infrastructure availability and human resources, composite schools have the ability to cater to a larger number of students. Working closer towards full capacity enables these schools to put their resources to more efficient use. An analysis of SSA budget allocation on elementary education reveals that 50.1 percent of 54 allocation is towards teacher salaries and trainings, 22.4 percent allocation is towards infrastructure creation & maintenance and 27.5 percent allocation is towards other elements. As mentioned earlier, the PTR at composite schools is higher than the PTR at elementary schools. Similarly, the number of students per school at composite schools is higher than the corresponding figure for elementary schools. Factors derived from these two data points when used to adjust the allocation figures for teacher related and infrastructure related components, provide for an estimate of the efficiency with which composite schools work. Table 18: Adjustment factors 2013-14 Type of School PTR Average strength per school Composite 32 301 Elementary 20 187 Adjustment factors (composite/elementary) 1.6 1.6 Using the above given factors to adjust the budget allocation for infrastructure and teacher related components reveals that composite schools provide for a 38 percent more efficient use of funds invested in infrastructure and teacher related components. In terms of percentage points, investing in composite schools leads to a 27.3 percentage point higher efficiency. Table 19: Effective Efficiency of composite schools over elementary schools Aggregate Infrastructu Difference Type of Teachers percentage Miscellaneous re and in School component budget resources Efficiency allocation Elementary 27.5 percent 22.4 percent 50.1 percent 100.0 percent 27.3 Composite 27.5 percent 13.9 percent 31.3 percent 72.7 percent percent Another interpretation of the aforementioned statistic can be that all other outcomes being the same, a composite school spends INR 0.73 per child where an elementary school spends INR 1.00. The only additional cost under the composite school model is the INR 20 per day the Government provides to students in the form of transport assistance. Impact At this point it is also important to appreciate that composite schools are imparting better quality education and in this sense are delivering a higher impact at a relatively lower cost per student. This is primarily because of their higher effectiveness in terms of higher number of effective learning days, better supporting infrastructure and stronger monitoring arrangements. At the very outset, it is observed that over the past five years the overall enrolment in Government elementary schools in Rajasthan has decreased by 9.3 percent. However, in the last year itself the composite schools in Rajasthan have reported a 17.6 percent increase in enrolment (where the number of composite schools has largely remained unchanged). The Government also believes that the composite schools have ensured an improvement in retention and transition rates. Secondary school teachers have access to the elementary school records. They are able to track and follow up with parents of students who do not show up for enrolment in class 9. However, the Government MIS does not provide dropout figures and transition rates disaggregated by elementary and composite schools. 55 As per a paper by Dr Geeta Gandhi Kingdon, every additional instruction day leads to a 0.01 time standard deviation change in students learning outcome scores. The corresponding estimate for having a teacher who holds a graduate degree is 0.09. Assuming a 0.05 standard deviation change for facilitating provisions/infrastructure and adjusting by differences between elementary schools indicators and composite school indicators reveals that the average learning outcome score for students from composite schools can be up to 34 points more than the average learning outcome score for their peers from elementary schools. The learning outcome test used as a basis of the calculation is the NAS which assumed that scores are normally distributed with a standard deviation of 50 points. Therefore the impact of better access to electricity in composite schools translates into a 0.81 point improvement in learning outcome scores. The estimate has been derived by multiplying the difference in composite school and elementary school performance (32.2 Percent) with the learning outcome change factor (0.05) and the standard deviation for NAS scores (50). Table 20: Total change in learning outcome factor Learning Difference Change outcome between in change Parameters of effectiveness composite learning factor and outcome (Std. elementary factor Deviations) Electricity availability 32.2 percent 0.05 0.81 Library facility 3.6 percent 0.05 0.09 Playground availability 11.0 percent 0.05 0.28 Computer availability 2.5 0.05 2.80 Teachers with graduation and above 8.0 percent 0.09 0.36 Average number of effective instructional days 60 0.01 30.00 Total change in learning outcome factor 34.34 *Learning outcome impact factors and standard deviation have been taken from the paper “Teacher characteristics and student performance in India: A pupil fixed effects approach” Further, as per another paper by Dr Kingdon, a standard deviation change in learning outcome scores leads to 18 percent increase in the income that a student can expect to earn after completing upper primary schooling. Given that composite school students are expected to score 34.34 points better than peers studying in elementary schools and that NAS scores work with a standard deviation of 50 points, the expected improvement in income for a student completing upper primary education from a composite school is 12.4 percent. In an absolute sense this work out to be INR 4,397 per year. 56 Table 21: Change in actual income derived from school completion via composite school over elementary school in Rajasthan Income parameters Income (in INR) Expected income for Upper primary complete (2016) 18 37,112 Per capita National Income India 61,855 Per capita income Rajasthan 59,097 Per capita income adjustment factor 0.955 Expected income for Upper primary complete Rajasthan (Elementary) 35,457 Expected income for Upper primary complete Rajasthan (Composite) 39,854 Difference in income (Composite vis-à-vis Elementary) 4,397 Sustainability Basis the analysis presented through this paper it can be concluded that composite schools can better sustain the directives of the RTE, SSA and RMSA. They make for more efficient use of funds, are more effective n instruction and thus produce better academic results. Given the increase in enrolment, transitions rates and learning outcomes; composite schools are facilitating movement of students from elementary to secondary education. They are doing this while imparting better quality education and this makes for a more sustainable case in terms of benefits to students and society at large. 18 Calculated using a Mincer Regression based NSSO data 57 APPENDICES VFM Calculation 58 Appendix 1: Learning Outcomes: VFM Calculation Appendix Table 1A: Learning Outcome Scores for Class V Class Class Class V Class V Class 5 V Class V Class 5 Class 5 5 All Mathe Mathe All State Readi Readin EVS EVS Subjec matics matics Subjects ng g 2015 2012 2015 ts 2012 2015 2015 2012 2012 Tamil Nadu 278 259 279 264 288 267 282 263 Uttar 282 248 298 257 284 260 288 255 Pradesh Punjab 252 249 252 238 245 236 250 241 Odisha 253 232 257 237 253 249 254 239 Gujarat 251 243 256 250 250 247 252 247 Kerala 277 259 244 230 252 240 258 243 Madhya 250 229 265 236 264 238 260 234 Pradesh Bihar 228 208 242 235 236 226 235 223 Appendix Table 1B: Average Annual Earnings for Students Completing Class 5 State Per Capita Per Capita Annual Earnings for Income for State Income Students Completing Class 5 (2011-12; 2004-05 Adjustment (2011-12; 2004-05 constant constant prices) Factor prices) Tamil Nadu 89,050 1.44 27,679 Uttar 30,071 0.49 9,347 Pradesh Punjab 76,895 1.24 23,901 Odisha 41,876 0.68 13,016 Gujarat 87,175 1.41 27,096 Kerala 78,387 1.27 24,365 Madhya 37,979 0.61 11,805 Pradesh Bihar 22,582 0.37 7,019 Average Annual earnings for Students Completing 19,226 Class 5 (2011-12; 2004-05 constant prices) National Per Capita Income Estimate for India 61,855 (2011-12; 2004-05 constant prices) Appendix Table 1C: Annual PPE Change Calculation for Odisha, with SSA expenditure Year Annual PPE Expenditure Annual Aggregate Change 2011 - 12 6,971.4 2012 - 13 7,941.8 2013 - 14 9,047.2 2014 - 15 10,317.1 13.9% 2015 - 16 12,825.2 59 Cumulative Change Between 2011 - 12 and 2014 - 15 48% Appendix Table 1D: Annual PPE Change Calculation for MP, with SSA expenditure Annual Aggregate Year Annual PPE Expenditure Change 2011 - 12 5,064.5 2012 - 13 6,356.2 2013 - 14 7,977.5 2014 - 15 12,662.8 25.5% 2015 - 16 14,082.0 Cumulative Change Between 2011 - 12 and 2014 - 15 150% Appendix Table 1E: Annual PPE Change Calculation for Kerala, with SSA expenditure Annual Aggregate Year Annual PPE Expenditure Change 2011 - 12 25,730.0 2012 - 13 28,829.6 2013 - 14 32,302.6 2014 - 15 39,678.8 12.0% 2015 - 16 43,976.8 Cumulative Change Between 2011 - 12 and 2014 - 15 54.2% Appendix Table 1F: Annual PPE Change Calculation for Gujarat, with SSA expenditure Annual Year Annual PPE Expenditure Aggregate Change 2011 - 12 49,187.4 2012 - 13 49,215.7 2013 - 14 49,244.1 2014 - 15 41,805.7 0.1% 2015 - 16 49,329.1 Cumulative Change Between 2011 - 12 and 2014 - 15 -15.0% Appendix Table 1G: Annual PPE Change Calculation for Tamil Nadu, with SSA expenditure Annual Year Annual PPE Expenditure Aggregate Change 2011 - 12 17,731.3 2012 - 13 21,036.1 2013 - 14 24,957.0 2014 - 15 38,252.0 18.6% 2015 - 16 38,912.0 Cumulative Change Between 2011 - 12 and 2014 - 15 115.7% 60 Appendix Table 1H: Annual PPE Change Calculation for Punjab, with SSA expenditure Annual Year Annual PPE Expenditure Aggregate Change 2011 - 12 9,214.3 2012 - 13 10,971.2 2013 - 14 13,063.0 2014 - 15 17,158.0 19.1% 2015 - 16 20,535.0 Cumulative Change Between 2011 - 12 and 2014 - 15 86.2% Appendix Table 1I: Annual PPE Change Calculation for Bihar, with SSA expenditure Annual Aggregate Year Annual PPE Expenditure Change 2011 - 12 3,283.9 2012 - 13 3,989.0 2013 - 14 4,845.5 2014 - 15 6,249.2 21.5% 2015 - 16 7,966.7 Cumulative Change Between 2011 - 12 and 2014 - 15 90.3% Appendix Table 1J: Annual PPE Change Calculation for Uttar Pradesh, with SSA expenditure Annual Aggregate Year Annual PPE Expenditure Change 2011 - 12 9,787.2 2012 - 13 12,519.2 2013 - 14 16,013.95 2014 - 15 21,814.65 27.9% 2015 - 16 29,424.74 Cumulative Change Between 2011 - 12 and 2014 - 15 122.9% Appendix Table 1K: Estimated annual recurrent PPE in Government schools, by State, 2015-16 Annual PPE, Annual PPE, including SSA State excluding SSA expenditures expenditures Odisha 8,897 12,825 Madhya Pradesh 9,285 14,082 Kerala 39,268 43,977 Gujarat 47,045 49,329 Tamil Nadu 33,127 38,912 Punjab 16,166 20,535 Bihar 3,105 7,967 Uttar Pradesh 23,012 29,425 61 Note: The annual Government schools’ PPE has been calculated using State Government revenue expenditure on education figures and UDISE Government school enrolment data. To estimate the PPE on government elementary schools, we have taken the State government’s revenue expenditure on elementary education and subtracted from it line items related to government support ‘aided’ schools, and also subtracted Sarva Shiksha Abhiyan (SSA) and Mid-Day Meal (MDM) expenditure. Since most of the secondary schools also provide upper primary education, and thus have 7 classes – all the way from grade 6 to grade 12 – therefore we attribute (3/7)th of the Government expenditure on government secondary schools (i.e. excluding support to aided schools) to being expenditure on government-run elementary education. The resulting estimate for State Government’s expenditure has been divided by official enrolment figures for elementary education at Government schools in each state. Given that enrolment figures are only available up till 2013-14, the figures for 2014-15 and 2015-16 have been estimated on the basis of the aggregate growth rate for the preceding years, as set out in the Appendix Tables 1D to 1J. The last column in the table adds in per pupil SSA expenditure (excluding SSA expenditure on civil works). Thus MDM expenditure is not included either in the first or the last column. The PPE estimates also do not include government expenditure on teachers’ pensions. The PPE in elementary education in Government schools in Bihar is very low in comparison with that in other states. This is explained by two things. Firstly, the fact that average pupil teacher ratio was 54 in Bihar, 26 in India (DISE Flash Statistics, 2013-14, latest available in Nov. 2015). Secondly, in Bihar a very high proportion (64%) of teachers were fixed-pay (contract) teachers getting a salary of Rs. 8000 pm, so its salary expenditure is a fraction that of other states 19, for example, in 2014-15, average salary of govt. primary school teachers in 9 states was Rs.39,683 per month (Ramachandran, 2015). 19 As seen in Table 6, in 2014-15, average school size is 285 students per Government elementary school in Bihar, and only 108 in India. Bihar had an average of 249 students per Government elementary school in 2005-06, and this increased by 36 by 2014-15; The India average in the bottom row, is 131.7 students per Government elementary school in 2005-06, and this fell by 23.7 by 2014-15; thus there are 108 students per Government school in India by 2014-15. Moreover, Table 7 shows that there are far fewer ‘small’ schools in Bihar (169) than in other states (average 4848 schools), i.e. with fewer than 20 enrolled students as a whole. While 2014-15 raw DISE data is downloadable, the 2014-15 Flash Statistics are not yet available on the U-DISE website as in Nov 2015. Page 13 of DISE 2013-14 Flash Statistics shows 364,715 teachers in Government. schools. Out of these around 2,32,000 are ‘fixed-pay’ teachers (since their jobs are not annually renewable in Bihar, para teachers there are not called ‘contractual’ teachers) whose pay was Rs. 8000 per month in Bihar in 2012-13 (from which the later figures are extrapolated). 62 Appendix 2: Changes over time in Total Number of Schools, Total Enrolment and Average Enrolment per School Andhra Pradesh Avg. Enrolment Number of schools Enrolment per school Academic Year Govt. Private Govt. Private Govt. Private 2005-06 76,361 18,623 7,397,630 3,725,310 97 200 2006-07 80,836 20,096 7,332,625 3,974,194 91 198 2007-08 79,324 21,125 6,821,442 4,216,944 86 200 2008-09 79,550 21,753 6,520,838 4,389,525 82 202 2009-10 79,813 22,985 6,310,989 4,540,259 79 198 2010-11 79,358 24,472 6,191,110 4,640,434 78 190 2011-12 78,673 26,098 6,175,060 4,692,880 78 180 2012-13 77,046 27,052 5,994,514 4,717,074 78 174 2013-14 75,089 28,404 5,967,621 4,934,846 79 174 2014-15 -- -- -- -- -- -- Absolute -1,272 9,781 -1,430,009 1,209,536 -18 -26 change % change -1.7 52.5 -19.3 32.5 -18.6 -13.0 Source- DISE data, www.dise.in The 2014-15 figures for Andhra Pradesh appear not to be correct, so they have been omitted. Assam Avg. Enrolment Number of schools Enrolment per school Academic Year Govt. Private Govt. Private Govt. Private 2005-06 38,053 2,162 3,739,904 212,358 98 98 2006-07 53,261 10,735 4,557,504 865,207 86 81 2007-08 53,950 12,777 4,616,563 1,085,872 86 85 2008-09 60,147 8,395 4,943,370 922,648 82 110 2009-10 44,518 8,820 4,204,893 957,207 94 109 2010-11 44,371 9,488 4,097,714 1,045,304 92 110 2011-12 42,917 7,930 4,174,185 887,186 97 112 2012-13 42,993 8,399 4,045,328 942,701 94 112 2013-14 50,186 6,753 4,563,766 814,862 91 121 2014-15 50,063 15,078 4,522,912 1,330,366 90 88 Absolute change 12,010 12,916 783,008 1,118,008 -8 -10 % change 32 597 21 526 -8 -10 Source- DISE data, www.dise.in 63 Bihar Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 53,271 507 13,286,024 111,017 249 219 2006-07 54,034 850 14,872,679 247,868 275 292 2007-08 66,636 1,238 17,349,687 313,118 260 253 2008-09 67,656 93 18,675,114 34,175 276 367 2009-10 67,642 14 19,000,385 7,108 281 508 2010-11 67,934 397 19,564,714 97,726 288 246 2011-12 69,366 86 20,519,815 28,982 296 337 2012-13 69,911 429 18,828,627 161,321 269 376 2013-14 70,673 1,698 19,853,552 499,365 281 294 2014-15 71,140 8,056 20,266,079 1,867,028 285 232 Absolute change 17,869 7,549 6,980,055 1,756,011 36 13 % change 34 1489 53 1582 14 6 Source- DISE data, www.dise.in Chhattisgarh Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 46,944 4,403 4,197,692 609,818 89 139 2006-07 44,889 4,079 3,683,954 511,268 82 125 2007-08 45,754 3,954 3,840,468 567,554 84 144 2008-09 45,847 4,060 3,825,454 667,721 83 164 2009-10 46,266 4,642 3,766,823 748,912 81 161 2010-11 46,394 4,945 3,807,603 824,695 82 167 2011-12 47,210 5,504 3,789,376 946,583 80 172 2012-13 47,822 5,788 3,754,252 984,370 79 170 2013-14 47,468 5,650 3,564,881 1,009,144 75 179 2014-15 47,253 6,046 3,429,623 1,118,510 73 185 Absolute change 309 1,643 -768,069 508,692 -16 46 % change 1 37 -18 83 -18 33 Source- DISE data, www.dise.in 64 Gujarat Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 32,634 4,622 6,065,471 1,089,529 186 236 2006-07 32,935 5,537 6,083,940 1,456,921 185 263 2007-08 33,114 5,925 6,031,806 1,630,687 182 275 2008-09 33,182 5,924 6,006,917 1,705,360 181 288 2009-10 33,426 6,513 5,881,273 1,933,118 176 297 2010-11 33,552 7,191 5,916,978 2,228,365 176 310 2011-12 33,496 7,444 5,982,181 2,393,253 179 322 2012-13 33,767 8,972 6,215,390 3,003,059 184 335 2013-14 33,713 9,462 6,105,605 3,122,142 181 330 2014-15 33,673 9,965 5,935,018 3,206,626 176 322 Absolute change 1,039 5,343 -130,453 2,117,097 -10 86 % change 3 116 -2 194 -5 36 Source- DISE data, www.dise.in Haryana Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 12,771 788 1,918,883 165,494 150 210 2006-07 14,478 1,702 2,107,524 384,485 146 226 2007-08 14,735 3,008 2,172,495 774,792 147 258 2008-09 15,467 3,480 2,327,437 914,992 150 263 2009-10 15,155 3,424 2,303,923 1,032,830 152 302 2010-11 14,956 5,235 2,087,364 1,306,276 140 250 2011-12 15,021 5,675 2,135,714 1,511,674 142 266 2012-13 14,988 6,461 2,098,675 1,758,293 140 272 2013-14 14,974 6,450 2,067,684 1,821,312 138 282 2014-15 14,579 7,212 1,983,948 1,970,018 136 273 Absolute change 1,808 6,424 65,065 1,804,524 -14 63 % change 14 815 3 1090 -9 30 Source- DISE data, www.dise.in 65 Himachal Pradesh Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 14,464 1,549 908,359 163,461 63 106 2006-07 14,521 2,093 870,755 210,844 60 101 2007-08 14,973 2,224 852,496 231,544 57 104 2008-09 15,071 2,289 816,450 248,828 54 109 2009-10 15,091 2,317 777,455 258,662 52 112 2010-11 15,126 2,313 746,331 289,296 49 125 2011-12 15,001 2,384 695,417 310,444 46 130 2012-13 15,111 2,434 657,700 327,136 44 134 2013-14 15,219 2,497 628,831 340,113 41 136 2014-15 15,355 2,601 600,381 358,527 39 138 Absolute change 891 1,052 -307,978 195,066 -24 32 % change 6 68 -34 119 -38 30 Source- DISE data, www.dise.in Jammu & Kashmir Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 16,105 3,346 1,030,425 536,712 64 160 2006-07 16,471 4,240 1,025,436 603,494 62 142 2007-08 16,502 4,287 1,043,785 617,090 63 144 2008-09 20,866 4,549 1,240,578 667,246 59 147 2009-10 21,311 4,786 1,253,651 719,643 59 150 2010-11 22,180 4,914 1,213,365 784,681 55 160 2011-12 22,538 4,955 1,152,609 755,621 51 152 2012-13 23,103 5,028 1,113,305 745,796 48 148 2013-14 23,234 5,073 1,076,708 764,470 46 151 2014-15 23,378 5,165 1,025,747 827,299 44 160 Absolute change 7,273 1,819 -4,678 290,587 -20 0 % change 45 54 0 54 -31 0 Source- DISE data, www.dise.in 66 Jharkhand Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 35,584 627 5,167,981 139,153 145 222 2006-07 39,269 1,349 5,997,707 357,309 153 265 2007-08 39,820 2,124 6,183,622 536,050 155 252 2008-09 39,768 2,082 6,002,522 537,611 151 258 2009-10 39,625 2,250 5,757,524 766,409 145 341 2010-11 40,529 2,703 5,598,510 886,951 138 328 2011-12 40,343 2,475 5,390,338 848,648 134 343 2012-13 40,674 2,583 5,144,565 959,946 126 372 2013-14 40,666 2,335 5,021,552 900,852 123 386 2014-15 40,405 6,368 4,819,302 1,718,486 119 270 Absolute change 4,821 5,741 -348,679 1,579,333 -26 48 % change 14 916 -7 1135 -18 22 Source- DISE data, www.dise.in Karnataka Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 44,373 9,712 5,123,711 1,790,745 115 184 2006-07 44,842 10,522 5,523,379 2,365,873 123 225 2007-08 45,622 10,819 5,445,484 2,476,784 119 229 2008-09 46,199 11,318 5,148,278 2,660,022 111 235 2009-10 46,325 11,834 4,788,516 2,848,229 103 241 2010-11 46,553 12,903 4,625,327 3,043,197 99 236 2011-12 50,885 19,966 4,783,689 3,637,528 94 182 2012-13 46,218 14,742 4,621,231 3,774,358 100 256 2013-14 46,030 15,310 4,277,320 3,546,104 93 232 2014-15 45,639 15,989 4,135,898 3,626,129 91 227 Absolute change 1,266 6,277 -987,813 1,835,384 -24 43 % change 3 65 -19 102 -21 23 Source: www.dise.in 67 Kerala Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 4,737 6,644 1,170,023 2,130,523 247 321 2006-07 5,084 7,099 1,228,664 2,173,323 242 306 2007-08 5,115 7,311 1,242,177 2,283,533 243 312 2008-09 5,050 7,302 1,180,499 2,204,098 234 302 2009-10 5,098 7,327 1,160,923 2,195,075 228 300 2010-11 NA NA NA NA NA NA 2011-12 5,333 9,230 1,007,249 2,662,352 189 288 2012-13 4,946 10,081 948,567 2,930,334 192 291 2013-14 5,111 10,151 919,566 2,919,190 180 288 2014-15 4,571 11,847 869,939 3,067,622 190 259 Absolute change -166 5,203 -300,084 937,099 -57 -62 % change -4 78 -26 44 -23 -19 Source- DISE data, www.dise.in Madhya Pradesh Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 104,671 16,664 10,883,542 2,651,887 104 159 2006-07 105,620 20,238 11,065,004 4,117,305 105 203 2007-08 106,408 22,592 10,748,076 4,662,624 101 206 2008-09 109,757 22,989 10,662,881 4,907,812 97 213 2009-10 111,510 23,455 10,466,162 5,018,827 94 214 2010-11 112,014 23,801 10,653,880 4,702,519 95 198 2011-12 112,079 27,148 10,221,216 4,920,512 91 181 2012-13 112,895 27,227 9,913,184 4,971,038 88 183 2013-14 114,444 26,668 9,511,486 4,901,200 83 184 2014-15 114,360 28,152 8,711,945 4,789,781 76 170 Absolute change 9,689 11,488 -2,171,597 2,137,894 -28 11 % change 9 69 -20 81 -27 7 Source- DISE data, www.dise.in 68 Maharashtra Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 60,824 23,462 7,809,393 7,149,294 128 305 2006-07 61,893 24,537 7,998,048 7,358,485 129 300 2007-08 61,732 25,548 7,746,195 7,911,902 125 310 2008-09 65,984 26,069 7,696,935 8,221,269 117 315 2009-10 67,573 26,551 7,583,759 8,270,299 112 311 2010-11 68,972 28,253 7,421,942 8,656,256 108 306 2011-12 69,782 29,935 7,231,470 8,930,490 104 298 2012-13 69,541 25,002 6,985,891 9,178,586 100 367 2013-14 67,307 28,130 6,312,059 9,382,952 94 334 2014-15 67,382 29,702 6,185,668 9,611,864 92 324 Absolute change 6,558 6,240 -1,623,725 2,462,570 -36 19 % change 11 27 -21 34 -28 6 Source- DISE data, www.dise.in Odisha Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 48,387 3,494 5,493,342 348,851 114 100 2006-07 47,967 3,231 4,620,233 307,594 96 95 2007-08 53,667 5,768 5,799,475 541,387 108 94 2008-09 55,715 6,447 5,922,892 625,593 106 97 2009-10 53,041 3,732 5,496,308 493,204 104 132 2010-11 57,179 7,060 5,653,997 717,530 99 102 2011-12 58,023 7,202 5,565,229 739,071 96 103 2012-13 58,355 7,418 5,458,962 761,458 94 103 2013-14 58,412 7,611 5,357,699 798,715 92 105 2014-15 58,508 9,797 5,237,812 1,148,698 90 117 Absolute change 10,121 6,303 -255,530 799,847 -24 17 % change 21 180 -5 229 -21 17 Source: www.dise.in 69 Punjab Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 18,198 2,100 2,021,749 378,316 111 180 2006-07 18,828 2,122 2,213,348 488,924 118 230 2007-08 18,508 1,518 2,119,908 439,050 115 289 2008-09 19,326 2,549 2,141,047 686,031 111 269 2009-10 19,969 3,303 2,046,938 861,386 103 261 2010-11 20,238 3,204 2,168,656 918,187 107 287 2011-12 20,370 3,594 2,193,899 1,026,200 108 286 2012-13 20,214 4,370 2,155,102 1,094,456 107 250 2013-14 21,343 7,603 2,293,421 1,701,493 107 224 2014-15 19,607 9,416 1,865,431 2,156,170 95 229 Absolute change 1,409 7,316 -156,318 1,777,854 -16 49 % change 8 348 -8 470 -14 27 Source: www.dise.in Rajasthan Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 75,221 19,098 8,595,362 3,082,114 114 161 2006-07 79,500 21,465 8,661,863 3,800,368 109 177 2007-08 80,576 22,727 8,179,500 4,017,135 102 177 2008-09 81,058 24,027 7,855,141 4,407,590 97 183 2009-10 81,006 24,767 7,476,412 4,698,717 92 190 2010-11 77,532 26,216 7,104,179 4,778,560 92 182 2011-12 77,833 29,766 7,155,509 5,112,169 92 172 2012-13 78,870 31,948 6,818,584 5,541,084 86 173 2013-14 83,564 33,658 6,410,664 5,691,938 77 169 2014-15 69,943 36,311 5,940,328 6,085,868 85 168 Absolute change -5,278 17,213 -2,655,034 3,003,754 -29 7 % change -7 90 -31 97 -25 4 Source: www.dise.in 70 Tamil Nadu Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 34,539 17,035 5,197,684 4,557,013 150 268 2006-07 35,129 17,294 5,084,746 4,691,843 145 271 2007-08 35,336 17,971 4,899,895 4,942,858 139 275 2008-09 35,436 18,454 4,610,905 5,267,716 130 285 2009-10 35,616 18,812 4,460,474 5,464,087 125 290 2010-11 36,122 18,907 4,273,526 5,512,190 118 292 2011-12 36,575 18,966 4,226,225 5,529,293 116 292 2012-13 36,940 19,402 3,913,563 5,747,698 106 296 2013-14 37,000 19,735 3,858,172 5,530,163 104 280 2014-15 37,760 19,393 4,119,616 5,133,327 109 265 Absolute change 3,221 2,358 -1,078,068 576,314 -41 -3 % change 9 14 -21 13 -27 -1 Source : www.dise.in Uttarakhand Avg. Enrolment Number of schools Enrolment per school Academic Year Govt Private Govt Private Govt Private 2005-06 16,222 2,685 1,038,613 309,083 64 115 2006-07 16,155 3,006 1,079,319 333,749 67 111 2007-08 16,971 3,639 1,083,967 449,045 64 123 2008-09 17,288 4,295 1,064,243 534,229 62 124 2009-10 17,327 4,800 991,687 588,042 57 123 2010-11 17,345 5,024 941,232 661,995 54 132 2011-12 17,500 5,326 907,931 712,331 52 134 2012-13 17,460 5,600 864,675 773,816 50 138 2013-14 17,426 5,716 832,340 842,024 48 147 2014-15 17,478 6,187 788,258 919,388 45 149 Absolute change 1,256 3,502 -250,355 610,305 -19 34 % change 8 130 -24 197 -30 29 Source : www.dise.in 71 Uttar Pradesh Avg. Enrolment Academic Number of schools Enrolment per school Year Govt Private Govt Private Govt Private 2005-06 124,998 36,871 22,055,410 8,119,442 176 220 2006-07 128,962 40,007 23,091,898 9,070,616 179 227 2007-08 135,573 44,485 22,508,818 9,567,565 166 215 2008-09 140,267 46,474 21,487,653 10,870,648 153 234 2009-10 147,070 48,019 19,892,972 11,644,675 135 243 2010-11 151,494 49,548 19,689,829 12,237,360 130 247 2011-12 154,757 65,713 19,585,396 15,540,557 127 236 2012-13 163,078 73,656 18,619,853 17,802,029 114 242 2013-14 160,752 74,897 17,712,153 18,060,720 110 241 2014-15 160,942 82,072 17,096,321 19,711,749 106 240 Absolute change 35,944 45,201 -4,959,089 11,592,307 -70 20 % change 29 123 -22 143 -40 9 Source- DISE data, www.dise.in West Bengal Avg. Enrolment Academic Number of schools Enrolment per school Year Govt Private Govt Private Govt Private 2005-06 52,790 6,433 8,950,732 … 170 2006-07 52,831 14,434 8,636,483 … 163 2007-08 57,487 12,523 11,542,965 1,728,026 201 138 2008-09 57,344 13,427 11,324,973 1,632,555 197 122 2009-10 77,194 11,362 13,378,168 1,662,626 173 146 2010-11 79,131 8,708 13,438,105 1,024,570 170 118 2011-12 81,623 8,273 13,256,933 1,065,469 162 129 2012-13 82,259 9,445 12,972,344 996,723 158 106 2013-14 81,915 9,657 11,810,855 963,059 144 100 2014-15 82,444 13,128 11,525,341 1,489,769 140 113 Absolute change 29,654 6,695 2,574,609 -238,257 -30 -25 % change 56 104 29 -14 -18 -18 Source- DISE data, www.dise.in Note: For West Bengal, there is a problem with the private school enrolment data, which appears to be unduly high for the two years 2005-06 and 2006-07. 72 Table 1a: Change in Earnings due to change in Learning levels for Class 5 Govt.- school Students Change in Change in Change in Change in Change in mean score Change in earnings Reading Mathe- mean score in terms of State EVS due to Comprehen matics all 3 across no. of Scores drop in sion Scores Scores subjects standard mean score deviations Tamil Nadu -19 -15 -21 -18 -0.36 -6.6% Uttar Pradesh -34 -41 -24 -33 -0.66 -11.9% Punjab -3 -14 -9 -9 -0.18 -3.1% Odisha -21 -20 -4 -15 -0.30 -5.4% Gujarat -8 -6 -3 -6 -0.12 -2.0% Kerala -18 -14 -12 -15 -0.30 -5.3% Madhya -21 -29 -26 -25 -0.50 -9.1% Pradesh Bihar -20 -7 -10 -12 -0.24 -4.4% Appendix 1 gives the list of learning outcome assessment scores for 2011 and 2015. NAS sets the standard deviation around the mean of achievement test score at 50. Table 1b: Absolute Change in Annual Earnings of Class 5 govt school Students, 2011-12 to 2014-15 Average Annual Average Annual Earnings Absolute Change in State Earnings in 2011-12 in 2014-15 Annual Earnings Tamil Nadu 27,679 25,852 -1,827 Uttar Pradesh 9,347 8,236 -1,110 Punjab 23,901 23,155 -746 Odisha 13,016 12,313 -703 Gujarat 27,096 26,543 -553 Kerala 24,365 23,078 -1,286 Madhya 11,805 10,728 -1,077 Pradesh Bihar 7,019 6,707 -312 Appendix 1 shows the per capita earnings adjustment figures. Table 1c: Change in Govt schools’ Per Pupil Expenditure, 2011-12 to 2014-15 State Change in Annual Per Pupil Expenditure Tamil Nadu 115.7% Uttar Pradesh 122.9% Punjab 86.2% Odisha 48.0% Gujarat -15.0% Kerala 54.2% Madhya Pradesh 150.0% Bihar 90.3% AVERAGE FOR THESE 8 STATES 81.5% Appendix 1 shows annual per pupil expenditure by state, and how the change in PPE is calculated. 73 Table 2a: Government and Private schools’ Value for Money Comparison (using data on children’s Literacy Outcomes) Uttar Tamil Madhya S. No. Variables Bihar Gujarat Kerala Punjab Odisha Pradesh Nadu Pradesh A Govt Per Pupil Expenditure (Rs.) 23012 3105 47044 33126 9384 39267 16166 8897 B Govt Achievement (Reading) 27 45 45 50 28 61 61 50 Govt Expenditure per C 859 70 1055 664 338 641 265 178 Achievement Units (Rs.) (c=a/b) D Private Per Pupil Expenditure (Rs.) 1800 4200 5400 10800 3700 8400 7900 7150 E Private Achievement (Reading) 61 88 64 40 58 71 74 77 Private Expenditure per F 29 48 84 269 63 119 107 93 Achievement Units (Rs.) (f=d/e) Govt./ Private Per Pupil G 12.8 0.7 8.7 3.1 2.5 4.7 2.0 1.2 Expenditure Ratio (g=a/d) Govt./ Private Numeracy Ratio H 0.44 0.51 0.70 1.24 0.48 0.87 0.83 0.65 (g=b/e) Private/Government Efficiency I 29.3 1.5 12.5 2.5 5.3 5.4 2.5 1.9 Ratio ( g = c/f ) Note: Rows B and E show the percentage of students of class 5 who can read a class 2 level text, as per the ASER data (see source below). Here per pupil expenditure in the govt. school system is calculated after removing the Sarva Shiksha Abhiyan expenditures (on free uniform, books, cash scholarships) and Mid Day Meal expenditure. This is to make it comparable to per pupil expenditure (fee levels) in private schools, which do not provide meals, uniforms, books or scholarships. Source: Row A: Authors’ calculations based on state government budgets, as summarised in Appendix Table 1K; Rows B & E: Annual Status of Education Report 2014; Row D: calculations based on raw 71st round National Sample Survey collected in summer/autumn 2014. 74 Table 2b: Government and Private schools’ Value for Money Comparison (using data on children’s Numeracy Outcomes) Uttar Tamil Madhya S. No. Variables Bihar Gujarat Kerala Punjab Odisha Pradesh Nadu Pradesh Govt Per Pupil A Expenditure (Rs) 23012 3105 47044 33126 9384 39267 16166 8897 Govt Achievement B 12 31 14 26 10 26 37 21 (Division) Govt Expenditure per C Achievement Units (Rs.) 1902 99 3384 1294 938 1534 436 434 (c=a/b) Private Per Pupil D 1800 4200 5400 10800 3700 8400 7900 7150 Expenditure (Rs.) Private Achievement E 39 72 35 26 29 50 54 45.40 (Division) Private Expenditure per F Achievement Units (Rs.) 47 58 155 414 130 169 147 157 (f= d/e) Govt./ Private Per Pupil G 12.8 0.7 8.7 3.1 2.5 4.7 2.0 1.2 Expenditure Ratio (g=a/d) Govt./ Private Numeracy H 0.31 0.43 0.40 0.98 0.35 0.52 0.69 0.45 Ratio (g=b/e) Govt./ Private Efficiency I 40.9 1.7 21.8 3.1 7.2 9.1 3.0 2.8 Ratio ( g = c/f ) Note: Rows B and E show the percentage of students of class 5 who can do simple (3 digit by 1 digit) Division, as per the ASER data (see source below). Here per pupil expenditure in the govt. school system is calculated after removing the Sarva Shiksha Abhiyan expenditures (on free uniform, books, cash scholarships) and Mid Day Meal expenditure. This is to make it comparable to per pupil expenditure (fee levels) in private schools, which do not provide meals, uniforms, books or scholarships. Source: Row A: Authors’ calculations based on state government budgets, as summarised in Appendix Table 1K; Rows B & E: Annual Status of Education Report 2014; Row D: calculations based on raw 71st round National Sample Survey collected in summer/autumn 2014. 75 Table 22: Performance in different content areas, NCERT’s National Achievement Survey (Class-5), 2011 and 2015 Content Area % correct % correct Change in answers in answers in % correct NAS Cycle 3 NAS Cycle 4 answers (2011) (2015) (2011 to 2015) Reading comprehension Locating information 54 49 -5 Grasp of Ideas /Interpretation 47 42 -5 Inference/evaluation 55 49 -6 Mathematics Operations 54 49 -5 Geometry 52 48 -4 Measurement 47 43 -4 Number system 51 45 -6 Environmental Science Family & environment 58 54 -4 Food 49 45 -4 Shelter 58 52 -6 Water 64 59 -5 Travel 49 46 -3 Real Life 44 40 -4 Source: NCERT (2015) “What students know and can do: A summary report of India’s National Achievement Survey: Class 5 (Cycle 4), 2015. 76 Table 23 :Annual Status of Education Report, 2010 to 2014 Table 24:Estimates of primary-school teacher salaries as a ratio of per capita GDP Country/state Reference year Estimated ratio of teacher salary to: Per capita GDP Per capita SDP OECD average 2009 1.2 -- Asian countries China 2000 0.9 -- Indonesia 2009 0.5 -- Japan 2009 1.5 -- Bangladesh 2012 ~1.0 -- Pakistan 2012 ~1.9 -- India 77 Nine Indian 2004-5 3.0 4.9 statesa Uttar Pradeshb 2006 6.4 15.4 Bihar 2012 5.9 17.5 Chhattisgarh 2012 4.6 7.2 Source: Table 5.4 in Chapter 5 of Dreze, Jean and Amartya Sen (2013) “An Uncertain Glory: India and its Contradictions”. Allen Lane, London. The authors cite the OECD (2011) for OECD countries average figure; Ciniscalco (2004) for China; estimates based on BRAC provided figures for Bangladesh and the Collective for Social Science Research (Karachi) for Pakistan. For India Nine states’ estimate, from Kingdon (2010), and for Uttar Pradesh, authors recalculated from Kingdon (2010) using Economic Survey data. For Chhattisgarh, authors’ own estimates based on enquiries from Education Departments and Planning Commission data on per capita SDP. Note: GDP = Gross Domestic Product SDP = State Domestic Product a : Andhra Pradesh, Bihar, Gujarat, Jammu & Kashmir, Madhya Pradesh, Maharashtra, Rajasthan, Uttar Pradesh, West Bengal. Figures in this row refer to all primary-school teachers (including contract teachers, who earn much lower salaries than regular teachers, and before the Sixth Pay Commission. b : Based on Sixth Pay Commission scales (fixed in 2009 with retrospective effort from 2006) The international figures apply to ‘statutory salaries of teachers’ after 15 years of service, at the primary level. Unless stated otherwise, Indian figures refer to regular teachers (as opposed to contract teachers). 78 Table 25: Temporal change in number of schools, total enrolment and average enrolment per school, in Govt and Private schools By state (2005-06 to 2014-15) Number of schools Total Enrolment Average Enrolment per school in govt. Absolute Change Absolute Change schools Absolute Change % Change % Change % Change STATE (2005 - 2014) (2005 - 2014) In baseline (2005 - 2014) year Gov Privat Priva Privat Govt Private Govt Private Govt 2005-06 Govt Private Govt t e te e Andhra -1,272 9,781 -2 53 -1,430,009 1,209,536 -19 32 97 -17 -26 -18 -13 Pradesh Assam 12,010 12,916 32 597 783,008 1,118,008 21 526 98 -8 -10 -8 -10 Bihar 17,869 7,549 34 1489 6,980,055 1,756,011 53 1582 249 36 13 14 6 Chhattisgarh 309 1,643 1 37 -768,069 508,692 -18 83 89 -16 46 -18 33 Gujarat 1,039 5,343 3 116 -130,453 2,117,097 -2 194 186 -10 86 -5 36 Haryana 1,808 6,424 14 815 65,065 1,804,524 3 1090 150 -14 63 -9 30 Himachal 891 1,052 6 68 -307,978 195,066 -34 119 63 -24 32 -38 30 Pradesh Jammu & 7,273 1,819 45 54 -4,678 290,587 0 54 64 -20 0 -31 0 Kashmir Jharkhand 4,821 5,741 14 916 -348,679 1,579,333 -7 1135 145 -26 48 -18 22 Karnataka 1,266 6,277 3 65 -987,813 1,835,384 -19 102 115 -24 43 -21 23 Kerala -166 5,203 -4 78 -300,084 937,099 -26 44 247 -57 -62 -23 -19 Madhya 9,689 11,488 9 69 -2,171,597 2,137,894 -20 81 104 -28 11 -27 7 Pradesh Maharashtra 6,558 6,240 11 27 -1,623,725 2,462,570 -21 34 128 -36 19 -28 6 Odisha 10,121 6,303 21 180 -255,530 799,847 -5 229 114 -24 17 -21 17 Punjab 1,409 7,316 8 348 -156,318 1,777,854 -8 470 111 -16 49 -14 27 Rajasthan -5,278 17,213 -7 90 -2,655,034 3,003,754 -31 97 114 -29 7 -25 4 79 Tamil Nadu 3,221 2,358 9 14 -1,078,068 576,314 -21 13 150 -41 -3 -27 -1 11,592,30 Uttar Pradesh 35,944 45,201 29 123 -4,959,089 -22 143 176 -70 20 -40 9 7 Uttarakhand 1,256 3,502 8 130 -250,355 610,305 -24 197 64 -19 34 -30 29 West Bengal 29,654 6,695 56 104 2,860,123 -764,967 32 -21 170 -30 NA -18 NA India simple 6921 8503 15 269 -336,961 1,777,361 -8 310 131.7 -23.7 20.4 -20.3 12.4 average 35,547,21 India TOTAL 138,422 170,064 -- -- -6,739,228 -- -- -- -- -- -- 5 Source: DISE state report card for each state for each year 2005-06 to 2014-15, downloaded from www.dise.in Note: The 2014-15 data on Andhra Pradesh’s number of schools seems problematic as the number of schools is dramatically smaller in 2014-15 compared to 2013-14. Thus, we have used only the figures upto 2013-14 for Andhra. The ‘India’ totals and averages in the last two rows refer to the 20 major states listed in the table. 80 Table 26:The number of ‘small’ government schools (with a total enrolment of 20 or fewer), their pupil teacher ratios and per pupil expenditure, 2014-15 Total Number Per Pupil salary Number Number of of Per pupil pupil Teacher expense of small Pupils in Teachers salary salary States Ratio in per govt these govt in these expense expense these annum Schools schools govt (Rs. pa) (Rs. per schools (Rs. schools month) crore) (a) (b) (c) ( d=b/c ) ( e )* ( f=e/b ) (g =f/12) Andhra Pradesh 6,508 84,292 7,380 11.4 382 45,263 3,772 Assam 3,893 52,166 9,499 5.5 491 94,134 7,845 Bihar 169 1,660 615 2.7 32 1,91,533 15,961 Chhattisgarh 5,502 71,416 10,729 6.7 555 77,668 6,472 Gujarat 1,374 20,445 2,876 7.1 149 72,724 6,060 Haryana 391 4,833 586 8.2 30 62,684 5,224 Himachal Pradesh 5,084 64,719 11,251 5.8 582 89,874 7,490 Jammu & Kashmir 7,058 90,413 17,793 5.1 920 1,01,741 8,478 Jharkhand 1,472 21,889 2,605 8.4 135 61,526 5,127 Karnataka 10,052 126,153 17,088 7.4 883 70,028 5,836 Kerala 1,713 5,460 1,713 3.2 89 1,62,197 13,516 Madhya Pradesh 8,872 121,546 14,958 8.1 773 63,622 5,302 Maharashtra 11,709 146,831 22,247 6.6 1,150 78,327 6,527 Odisha 1,251 17852 2,164 8.2 112 62,665 5,222 Punjab 4,406 61,948 8988 6.9 465 75,005 6,250 Rajasthan 7,996 105,307 14,921 7.1 771 73,252 6,104 Tamil Nadu 3,879 56,246 7,952 7.1 411 73,091 6,091 Uttarakhand 6,522 77,409 13,409 5.8 693 89,553 7,463 Uttar Pradesh 5,135 69,220 11,204 6.2 579 83,679 6,973 West Bengal 3,979 54,803 9,418 5.8 487 88,845 7,404 Total-20 major states 96,965 1,254,608 187,396 9689 Weighted average 4,848 62,730 9,370 6.7 484 74,232 7,156 Note: * (e = c x Rs. 43,082 x 12); that is, total teacher salary expense in these small schools in 2014-15 is equal to the Number of teachers in these schools multiplied by the Average teacher salary per month in elementary education in India, and this monthly figure is then multiplied by 12 to obtain the annual total salary expense. Average salary of primary and secondary government school teachers in 9 Indian states in July 2014, is presented in Table 6.3 in a NUEPA study by Professor Vimala Ramachandran (2015). This is reproduced here as Appendix 2. Ramachandran reports salaries of new appointees (0 years’ experience) and salary after 15 and after 25 years’ experience. We have taken the salary rate after 15 years’ experience as the average salary of teachers. While teachers of classes 1 to 5 get primary teachers’ salary rate, teachers of classes 6-8 get paid the same salary rate as secondary school teachers. Thus we have taken the simple average of primary and secondary teachers’ salaries after 15 years’ experience across the 9 states, as the measure of mean salary of teachers of elementary classes (1 to 8). This was Rs. 43,082 per month in the school session 2014-15. 81 Appendix 3: Some Statistics of Government Schools in India’s States as per 2014 – 15 DISE Data Andhra Pradesh Number Number Pupil Percentage Number of S. No. Enrolment of of Teacher of Schools Teachers Schools Students Ratio 1 0 (Zero) 225 0.49 177 - - 2 5 or less 513 1.11 464 1,162 3 3 10 or less 1,645 3.57 1,571 10,663 7 4 20 or less 6,508 14.11 7,380 84,292 11 5 50 or less 25,821 55.97 44,644 698,587 16 6 100 or less 35,796 77.59 84,550 1,423,106 17 7 200 or less 42,943 93.08 136,859 2,404,808 18 8 500 or less 45,967 99.64 180,420 3,256,256 18 Source: raw data of DISE downloaded from www.dise.in Assam Number Pupil Number of Percentage Number of S. No. Enrolment of Teacher Schools of Schools Teachers Students Ratio 1 0 (Zero) 157 0.31 154 - 0.0 2 5 or less 299 0.60 457 565 1.2 3 10 or less 922 1.84 1,945 5,799 3.0 4 20 or less 3,893 7.78 9,499 52,166 5.5 5 50 or less 18,831 37.61 51,785 586,209 11.3 6 100 or less 34,579 69.07 108,578 1,719,959 15.8 7 200 or less 45,705 91.29 165,909 3,258,766 19.6 8 500 or less 49,883 100 838,533 4,414,147 5.3 Source: raw data of DISE downloaded from www.dise.in Bihar Number Pupil S. Number of Percentage Number of Enrolment of Teacher No. Schools of Schools Teachers Students Ratio 1 0 (Zero) 53 0.07 305 - 0 2 5 or less 59 0.08 331 22 0.1 3 10 or less 74 0.10 363 153 0.4 4 20 or less 169 0.24 615 1,660 3 5 50 or less 1,556 2.19 3,797 57,136 15 6 100 or less 10,481 14.73 26,566 762,987 29 7 200 or less 33,918 47.68 106,621 4,196,426 39 8 500 or less 60,388 84.89 271,793 12,610,983 46 Source: raw data of DISE downloaded from www.dise.in 82 Chhattisgarh Number Pupil Number Percentage Number of S. No. Enrolment of Teacher of Schools of Schools Students Teachers Ratio 1 0 (Zero) 379 0.80 443 - 0.0 2 5 or less 681 1.44 1,028 1,137 1.1 3 10 or less 1,582 3.35 2,761 8,701 3.2 4 20 or less 5,502 11.64 10,729 71,416 6.7 5 50 or less 21,403 45.29 48,221 636,953 13.2 6 100 or less 36,648 77.56 98,510 1,736,016 17.6 7 200 or less 45,436 96.15 141,760 2,927,441 20.7 8 500 or less 47,180 99.84 154,750 3,382,334 21.9 Source: raw data of DISE downloaded from www.dise.in Gujarat Number Pupil Number Percentage Number of S. No. Enrolment of Teacher of Schools of Schools Students Teachers Ratio 1 0 (Zero) 8 0.02 60 - 0.0 2 5 or less 51 0.15 139 165 1 3 10 or less 223 0.66 495 1,590 3 4 20 or less 1,374 4.08 2,876 20,445 7 5 50 or less 7,411 22.01 16,748 231,964 14 6 100 or less 13,944 41.41 38,959 706,085 18 7 200 or less 22,518 66.87 88,496 1,962,218 22 8 500 or less 32,036 95.14 177,479 4,844,804 27 Source: raw data of DISE downloaded from www.dise.in Haryana Number Pupil Number Percentage Number of S. No. Enrolment of Teacher of Schools of Schools Students Teachers Ratio 1 0 (Zero) 21 0.14 8 - 0.0 2 5 or less 47 0.32 37 95 3 3 10 or less 133 0.91 145 806 6 4 20 or less 391 2.68 586 4,833 8 5 50 or less 2,942 20.18 7,708 98,086 13 6 100 or less 7,161 49.12 28,966 413,082 14 7 200 or less 11,852 81.29 65,660 1,077,347 16 8 500 or less 14,266 97.85 92,872 1,764,847 19 Source: raw data of DISE downloaded from www.dise.in 83 Himachal Pradesh Number Pupil Number of Percentage Number of S. No. Enrolment of Teacher Schools of Schools Teachers Students Ratio 1 0 (Zero) 2 0.01 2 - 0.0 2 5 or less 482 3.14 906 1,687 2 3 10 or less 1,633 10.63 3,364 11,251 3 4 20 or less 5,084 33.11 11,251 64,719 6 5 50 or less 11,709 76.26 33,956 281,016 8 6 100 or less 14,498 94.42 54,222 473,592 9 7 200 or less 15,254 99.34 63,762 569,737 9 8 500 or less 15,348 99.95 65,536 595,502 9 Source: raw data of DISE downloaded from www.dise.in Jammu & Kashmir Number Pupil Number of Percentage Number of S. No. Enrolment of Teacher Schools of Schools Teachers Students Ratio 1 0 (Zero) 48 0.21 283 - 0.0 2 5 or less 747 3.20 1,877 2,507 1 3 10 or less 2,299 9.83 5,725 15,373 3 4 20 or less 7,058 30.19 17,793 90,413 5 5 50 or less 16,409 70.19 51,852 399,559 8 6 100 or less 21,490 91.92 82,704 755,223 9 7 200 or less 23,240 99.41 95,981 980,728 10 8 500 or less 23,363 99.93 97,346 1,011,150 10 Source: raw data of DISE downloaded from www.dise.in Jharkhand Number Percentag Pupil Number of Number of S. No. Enrolment e of Teacher of Schools Teacher Students Schools Ratio s 1 0 (Zero) 91 0.23 397 - 0.0 2 5 or less 121 0.30 459 117 0.3 3 10 or less 245 0.61 654 1,205 2 4 20 or less 1,472 3.64 2,605 21,889 8 5 50 or less 12,706 31.45 22,236 433,910 20 6 100 or less 25,596 63.35 49,733 1,354,796 27 7 200 or less 33,631 83.23 76,208 2,488,021 33 8 500 or less 39,561 97.91 108,635 4,267,841 39 Source: raw data of DISE downloaded from www.dise.in 84 Karnataka Number Pupil Number of Percentage Number of S. No. Enrolment of Teacher Schools of Schools Teachers Students Ratio 1 0 (Zero) 544 1.20 500 - 0.0 2 5 or less 1,086 2.37 1,260 2,172 1.7 3 10 or less 3,200 7.01 4,480 19,648 4.4 4 20 or less 10,052 22.02 17,088 126,153 7.4 5 50 or less 22,520 49.34 46,999 541,381 11.5 1,201,57 6 100 or less 31,662 69.37 81,055 3 14.8 2,398,88 7 200 or less 40,075 87.81 130,244 9 18.4 3,896,84 8 500 or less 45,270 99.19 176,326 2 22.1 Source: raw data of DISE downloaded from www.dise.in Kerala Pupil Number of Percentage of Number of Number of S. No. Enrolment Teacher Schools Schools Teachers Students Ratio 1 0 (Zero) 3 0.06 33 - 0.0 2 5 or less 22 0.48 90 73 0.8 3 10 or less 79 1.73 312 561 1.8 4 20 or less 1,713 8.42 1,713 5,460 3.2 5 50 or less 1,448 31.68 7,399 42,687 5.8 6 100 or less 2,482 54.30 14,867 116,555 7.8 7 200 or less 3,331 72.87 24,583 237,500 9.7 8 500 or less 4,109 89.89 41,788 486,670 11.6 Source: raw data of DISE downloaded from www.dise.in Madhya Pradesh Pupil Number Percentage of Number of Number of S. No. Enrolment Teache of Schools Schools Teachers Students r Ratio 1 0 (Zero) 398 0.35 471 - 0.0 2 5 or less 858 0.75 1,201 1,546 1.3 3 10 or less 2,170 1.90 3,265 12,499 3.8 4 20 or less 8,872 7.76 14,958 121,546 8.1 5 50 or less 47,782 41.78 89,065 1,517,279 17.0 6 100 or less 87,161 76.22 185,653 4,333,540 23.3 7 200 or less 109,438 95.70 265,653 7,351,649 27.7 8 500 or less 114,204 99.86 292,218 8,603,286 29.4 Source: raw data of DISE downloaded from www.dise.in 85 Maharashtra Percentag Number Number Pupil Number of S. No. Enrolment e of of of Teacher Schools Schools Teachers Students Ratio 1 0 (Zero) 5 0.00 11 - 0.0 2 5 or less 1,154 1.70 2,020 4,131 2.0 3 10 or less 3,879 5.76 7,060 26,455 3.7 4 20 or less 11,709 17.37 22,247 146,831 6.6 5 50 or less 35,164 52.19 70,680 889,649 12.6 6 100 or less 46,508 69.00 108,829 1,723,121 15.8 7 200 or less 59,042 87.62 182,440 3,532,482 19.4 8 500 or less 66,686 99.00 254,074 5,663,642 22.3 Source: raw data of DISE downloaded from www.dise.in Odisha Pupil Number of Percentage Number of Number of S. No. Enrolment Teacher Schools of Schools Teachers Students Ratio 1 0 (Zero) 4 0.02 26 - 0.0 2 5 or less 65 0.33 102 233 2.3 3 10 or less 282 1.44 415 2,056 5.0 4 20 or less 1,251 6.38 2,164 17,852 8.2 5 50 or less 6,712 34.23 15,706 211,696 13.5 6 100 or less 12,868 65.62 43,623 661,415 15.2 7 200 or less 17,929 91.44 85,880 1,368,700 15.9 8 500 or less 19,512 100 108,487 1,794,324 16.5 Source: raw data of DISE downloaded from www.dise.in Punjab Pupil Number Percentage Number of Number of S. No. Enrolment Teacher of Schools of Schools Teachers Students Ratio 1 0 (Zero) 101 0.17 490 - 0.0 2 5 or less 265 0.45 806 615 0.8 3 10 or less 952 1.63 2113 6,445 3.0 4 20 or less 4,406 7.53 8988 61,948 6.9 5 50 or less 24,142 41.26 54078 754,196 13.9 6 100 or less 40,880 69.87 109150 1,956,517 17.9 7 200 or less 53,021 90.62 172848 2,427,832 14.0 8 500 or less 58,307 100 216319 5,111,775 23.6 Source: raw data of DISE downloaded from www.dise.in 86 Rajasthan Pupil Number of Percentage Number of Number of S. No. Enrolment Teacher Schools of Schools Teachers Students Ratio 1 0 (Zero) 273 0.37 1010 - 0.0 2 5 or less 842 1.20 1,972 2,085 1.1 3 10 or less 2,340 3.34 4,666 14,522 3.1 4 20 or less 7,996 11.43 14,921 105,307 7.1 5 50 or less 29,174 41.71 61,645 832,743 13.5 6 100 or less 47,920 68.51 143,760 215,871 1.5 7 200 or less 64,724 92.54 265,757 4,551,909 17.1 8 500 or less 69,846 99.86 315,634 5,872,861 18.6 Source: raw data of DISE downloaded from www.dise.in Tamil Nadu Number Pupil Number of Percentage Number of S. No. Enrolment of Teacher Schools of Schools Teachers Students Ratio 1 0 (Zero) 1 0.00 10 - 0.0 2 5 or less 157 0.42 327 597 1.8 3 10 or less 713 1.89 1,440 5,319 3.7 4 20 or less 3,879 10.27 7,952 56,246 7.1 5 50 or less 15,521 41.10 34,611 448,557 13.0 6 100 or less 23,735 62.86 73,579 1,048,850 14.3 7 200 or less 32,601 86.34 153,550 2,317,931 15.1 8 500 or less 37,215 98.56 232,594 3,625,485 15.6 Source: raw data of DISE downloaded from www.dise.in Uttarakhand Percentag Number Pupil Number of Number of S. No. Enrolment e of of Teache Schools Teachers Schools Students r Ratio 1 0 (Zero) 243 1.39 164 - 0.0 2 5 or less 948 5.42 1,405 2,569 1.8 3 10 or less 2,455 14.04 4,443 15,017 3.4 4 20 or less 6,522 37.31 13,409 77,409 5.8 5 50 or less 13,075 74.81 33,211 290,788 8.8 6 100 or less 15,804 90.42 48,518 479,809 9.9 7 200 or less 17,045 97.52 57,442 647,880 11.3 8 500 or less 17,433 99.74 61,016 753,280 12.3 Source: raw data of DISE downloaded from www.dise.in 87 Uttar Pradesh Number Number Pupil Percentage Number of S. No. Enrolment of of Teacher of Schools Teachers Schools Students Ratio 1 0 (Zero) 237 0.15 396 - 0.0 2 5 or less 514 0.32 941 1,043 1.1 3 10 or less 1,316 0.82 2,616 7,743 3.0 4 20 or less 5,135 3.19 11,204 69,220 6.2 5 50 or less 32,317 20.08 79,273 1,085,205 13.7 6 100 or less 87,105 54.12 248,249 5,136,582 20.7 7 200 or less 146,502 91.02 467,341 13,166,135 28.2 8 500 or less 160,500 99.72 531,737 16,788,300 31.6 Source: raw data of DISE downloaded from www.dise.in West Bengal Number Number Pupil Percentage Number of S. No. Enrolment of of Teacher of Schools Teachers Schools Students Ratio 1 0 (Zero) 216 0.26 1112 - 0.0 2 5 or less 422 0.51 1,510 705 0.5 3 10 or less 976 1.18 2,681 5,347 2.0 4 20 or less 3,979 1.82 9,418 54,803 5.8 5 50 or less 25,520 30.95 65,535 834,198 12.7 6 100 or less 52,226 63.34 160,334 2,766,202 17.3 7 200 or less 69,768 84.62 246,839 5,170,506 20.9 8 500 or less 77,727 94.28 329,329 7,599,835 23.1 Source: raw data of DISE downloaded from www.dise.in 88 Appendix 4: Vimala Ramachandran’s evidence on teacher salary levels in India: Actual take home salaries of teachers# (in INR) Primary Secondary Salary of Salary Salary Salary of Salary Salary State new after 15 after 25 new after 15 after 25 appointee years years appointee years years Tamil 15,345 28,660 50,140 26,370 48,750 84,410 Nadu 18,794 (R) 26,098 (R) 33,672 (R) 24,272 (R) 34,618 (R) 44,762 (R) Karnataka 21,814 (U) 30,198 (U) 38,892 (U) 28,102 (U) 39,978 (U) 51,622 (U) 28,650 (R) 39,780 (R) 44,400 (R) 37,494 (R) 57,523 (R) 78,637 (R) Jharkhand 31,600 (U) 43,260 (U) 48,100 (U) 39,208 (U) 60,160 (U) 82,247 (U) Odisha 14,031 26,659 27,347 25,625 37,806 43,034 Rajasthan 26,013 NA NA 28,331 NA NA Mizoram 16,504 NA NA NA NA NA Uttar 29,293 39,683 44,783 37,226 47,716 52,996 Pradesh 35,936 (R) 59,113 (R) 79,288 (R) 40,602 (R) 66,868 (R) 89,699 (R) Punjab^ 36,588 (U) 60,194 (U) 80,742 (U) 41,340 (U) 68,092 (U) 91,346 (U) Source: State reports ; R – Rural; U - Urban # Actual take home salary includes basic pay, grade pay, dearness allowances, HRA, city compensatory allowances, any other benefits and deductions (if any). Also, actual take home salaries for teachers might differ from district to district. The above is only a generalized indicator for each state. ^ Salaries are given for Mohali district because the salaries of teachers vary across districts. Source: Table 6.3 in Ramachandran, V. (2015) “Synthesis study of teachers in Nine states”, NUEPA, New Delhi. 89