NITI Aayog

Choose Report Type

Publication Date

Thu, 03/19/2026 - 12:03

Report Upload

Download (5.21 MB)

PDF Text

Disclaimer
Views expressed in the paper are personal Education and Skilling for Employment:
From Credentials to Learning Outcomes

By

Arvind Virmani
1

Abstract
This paper examines India’s educa?on and skilling system within the context of the country’s
employment eco-system. This is an incredibly complex task. The ﬁrst challenge is in comparing
India with its 1.4 billion people and 36 States and Union Territories (UTs), with countries of
diﬀerent sizes and levels of per capita GDP (PCGDP). The second challenge is to ﬁnd consistent
and comparable data for this vast and diverse Na?on to evaluate the evolu?on of educa?on
and skills over ?me. This challenge is mul?plied manifold when comparing the 28 States and
8 UTs, with diverse educa?on, skilling and tes?ng systems, and diﬀerent levels of Per capita
State domes?c product (PCNSDP).
Though the focus of the paper is on Educa?on, Skilling and Employability, the analysis cannot
be divorced from some of the unique features of India’s labor market. Only 23% of workers
are regular wage or salaried employees, with a substan?al frac?on of these in the informal
sector. 58% of these workers are “self-employed” – be?er described as “micro-
entrepreneurs”, whose skilling needs and requirements, are seldom recognized in the
employment literature. The remaining 19% of “casual workers” have their own unique needs
for educa?on, skills and job placement. Much of literature focuses on the top end of regular,
formal jobs, cons?tu?ng roughly 10% of workers in India, but for most of India’s micro-
entrepreneurs, “Skills and Jobs are two sides of the same coin.”
This paper conceptually simpliﬁes the employment ecosystem into a set of three pyramids;
The educa?on pyramid, with three layers of primary, secondary and ter?ary educa?on; the
Skills pyramid with un/low skilled, semi (medium) skilled and (highly) skilled; and the Jobs
pyramid with low/minimum wage, average wage, and high wage employees.
The educa?on pyramid is a rela?vely recent phenomenon in India, with the educa?on system
be?er characterized as a cylinder in earlier years. The rela?ve neglect of schooling in the ﬁrst
50 years aLer independence, is reﬂected in low adult literacy, and low share of adult
popula?on with diﬀerent levels of schooling.

1
Dr Arvind Virmani is Member, NITI Aayog. Author acknowledges the dedicated work of Jasleen Sharma, Junior Research Assistant in the
oﬃce of Member(av), in collec?ng & processing the data, material & references. Any views expressed in the paper are those of the Author
and should not be a?ributed to NITI Aayog or GOI.
The change in a?en?on in past 15 years is reﬂected in the be?er youth literacy rates. This in
turn is due to the improvement in the primary enrolment and comple?on rates over the
decades. With access to primary educa?on having improved, the focus has shiLed to
improving quality of learning outcomes. Learning outcomes depend on pedagogy, teaching at
the right level, repeated tes?ng to sharpen focus on slow learners, and teacher re-training.
Interna?onal compara?ve analysis shows, that na?onal learning outcomes correspond to our
Per Capita GDP, but it will be a challenge to raise them to High Income PCGDP levels in 25
years.
One intriguing ﬁnding is that in a cross-country compara?ve context, almost all educa?onal
indicators are posi?vely corelated with per capita GDP. In contrast, in an inter-State
compara?ve context (within India), almost all educa?on indicators are un-corelated with Per
capita Net Domes?c product (PNSDP). Despite considerable variability across States & UTs
and across ?me, the paper ﬁnds li?le or no correla?on between minimum learning outcomes
and PCNSDP. There is no evidence that varia?on in school infrastructure, teacher pay, or
subsidies to child’s family, improve learning at primary & lower secondary levels.
Some States have made notable progress in improving learning outcomes in primary schools,
during the past decade (2014-24). Children (Std V) with Minimum reading proﬁciency (MRP)
has increased by 11.7% points in UP, 10.9% points in Jharkhand and 7.6% points in Odisha.
Children with Minimum Arithme?c proﬁciency (MAP) have increased by 13.7% points in UP
and 10.4% points in Odisha. This is far above the all-India increase of 0.6% points in MRP and
4.6% points in MAP during the same period. Share of children with MAP are however 63% of
children with MRP, showing that pedagogy is a greater challenge in arithme?c/math than in
reading.
The New Educa?on Policy (NEP, 2020) has recognized the links between educa?on and job
skilling. Besides exploring the links between educa?on, skills and jobs, this paper also explores
the missing links between stake holders, market par?cipants and State governments. There is
an urgent need to deepen coopera?on between Govts (C&S), Industry and NGOs/NPOs to
improve employability. Private markets are func?oning adequately to connect the top levels
of the three pyramids. They only need a posi?ve environment (EODB) from State & Central
Governments. Market connec?vity and eﬃciency, is poor at the middle levels of the three
pyramids, and virtually absent at the bo?om of the pyramids, despite eﬀorts by Govts and
NGOs. Much greater eﬀort is necessary, by State & Local Govts and NGOs, supported by
Central Govt, and Private Industry with its experience in job skills and CSR funds. Every skill
advisory commi?ee at the State and local level should have a majority of representa?on from
the most dynamic entrepreneurs and exporters of the State & local area respec?vely. By
strengthening the middle of the skilling pyramid, we can make India the human capital
supplier to the World: “Skill for India, Skill for the World.”

i

Table of Contents

1. Introduc?on ................................................................................................................... 1
2. Overview ........................................................................................................................ 4
3. Framework and Methodology ...................................................................................... 6
3.1 Na?onal Educa?on Policy (NEP 2020)...................................................................... 11
3.2 Methodology: Interna?onal Comparison ................................................................ 12
4. Legacy of State Eﬀorts ................................................................................................. 14
4.1 Adult Literacy ........................................................................................................... 14
4.2 Adult educa?on levels ............................................................................................. 15
4.3 Youth Literacy .......................................................................................................... 18
5. Primary Educa?on ....................................................................................................... 19
5.1 Enrollment & Comple?on ....................................................................................... 19
5.1.1 Pre-primary .................................................................................................. 21
5.2 Learning Outcome Surveys ..................................................................................... 22
5.2.1 Survey Design and Comparability ................................................................ 23
5.3 Interna?onal Comparison of Outcomes (MRP) ...................................................... 25
5.3.1 Methodology for calcula?ng MRP ................................................................ 25
5.3.2 Minimum Reading Proﬁciency (MRP) ........................................................... 26
5.4 Minimum Learning Outcomes (ASER) .................................................................... 27
5.4.1 Minimum Learning Proﬁciency: India 2024 .................................................. 28
5.4.2 Minimum Learning Proﬁciency: 2006 to 2024 ............................................. 29
5.5 Learning Outcomes: Oﬃcial Surveys ...................................................................... 37
5.5.1 Learning Outcomes: NAS and FLS ................................................................. 37
5.5.2 Change in Learning (NAS & FLS) ................................................................... 41
5.5.3 Learning Diﬀerence across loca?on & gender .............................................. 41
5.6 Learning Outcomes in States .................................................................................. 44
5.6.1 Learning Outcomes in States (ASER) ............................................................ 44
5.6.2 Minimum Reading Proﬁciency: NAS ............................................................ 48
5.6.3 Enrolment Rates across States (2024-25) .................................................... 51
5.7 Programs and Pla?orms .......................................................................................... 53
5.7.1 DIKSHA ......................................................................................................... 53
ii

5.7.2 NIPUN Bharat ............................................................................................... 54
5.7.3 PM eVidya .................................................................................................... 54
5.7.4 E-Pathshala ................................................................................................... 55
5.7.5 Samagra Shiksha Scheme ............................................................................. 55
5.7.6 PRATHAM: Teaching at the right level (TaRL) ............................................... 56
5.7.7 Educate Girls (NGO) ..................................................................................... 57
5.7.8 SwiLPAL ........................................................................................................ 58
5.7.9 Other AI tools ............................................................................................... 58
5.8 Summary and Conclusions ...................................................................................... 58
6. Secondary Educa?on ................................................................................................... 61
6.1 Transi?on from Primary to Secondary School ........................................................ 61
6.2 Enrollment & Comple?on Rates ............................................................................. 62
6.2.1 India’s schooling output rela?ve to China ................................................... 64
6.2.2 Secondary Enrolment Rates across States ................................................... 65
6.3 Learning Outcomes and Tradi?onal Results ........................................................... 69
6.3.1 Learning Outcomes: Lower secondary school ............................................. 69
6.3.2 Comparison of Grade 8 learning outcomes (NAS viz ASER) .......................... 74
6.3.3 Learning Outcomes & Scores: Grades 8 and 10 (NAS) ................................. 75
6.3.4 PRS (2024) results for Grades 6 and 9 .......................................................... 80
6.4 Programs & Pla?orms ............................................................................................. 80
6.4.1 SWAYAM ....................................................................................................... 81
6.4.2 NISHTHA ....................................................................................................... 82
6.4.3 Na?onal Digital Library of India ................................................................... 82
6.5 Summary & Conclusion ........................................................................................... 83
6.5.1 Summary ...................................................................................................... 83
6.5.2 Conclusion .................................................................................................... 87
6.5.3 Sugges?ons .................................................................................................. 89
7. Ins?tu?onal Issues ...................................................................................................... 89
7.1 Pupil Teacher Ra?o (PTR) ........................................................................................ 89
7.2 Gender parity........................................................................................................... 91
7.3 Public vs Private Schools ......................................................................................... 91
iii

7.3.1 Annual Status of Educa?on Report ............................................................... 92
7.3.2 Na?onal Achievement Surveys (NAS) 2021 .................................................. 94
7.3.3 PARAKH Rashtriya Sarvekshan (PRS) (2024) ................................................. 96
7.4 Digital Access in Schools ......................................................................................... 96
8. School Dropouts need Job Skills ............................................................................... 100
9. Employment Eco-system: Skills for Employability ................................................... 103
10. Skill for India, Skill for the World .............................................................................. 105
10.1 Voca?onal Educa?on: Interna?onal Comparison ................................................ 106
10.2 Schemes to improve Job skills .............................................................................. 108
10.3 Skilling Ins?tu?ons ................................................................................................ 110
10.3.1 Ministry of Skill Development & Entrepreneurship .................................... 110
10.3.2 Industrial Training Ins?tutes (ITIs) .............................................................. 111
10.3.3 Schemes for Upgrading Skilling Infrastructure ........................................... 115
10.3.4 Polytechnics ................................................................................................ 116
10.3.5 Employability of ITI and Polytechnic educated ........................................... 117
10.3.6 Non-Govt Organiza?ons (NGOs) in Skilling ................................................. 118
10.4 Training the Trainers: NSTI and CITS ..................................................................... 118
10.5 Training and Skilling in Firms ................................................................................ 120
10.6 Skilling & Employment Portals & Placement ﬁrms .............................................. 121
10.7 Skill Policy and Planning ........................................................................................ 124
10.8 Summary and Conclusion ..................................................................................... 128
10.8.1 Summary .................................................................................................... 128
10.8.2 Conclusion .................................................................................................. 131
11. Ter?ary Educa?on ..................................................................................................... 132
11.1 Introduc?on ........................................................................................................... 132
11.2 Ter?ary School Enrollment .................................................................................... 133
11.2.1 Courses and Programs ............................................................................... 134
11.2.2 Enrolment Rates across states ................................................................... 135
11.3 Ter?ary A?ainment: Adult (25+) ........................................................................... 137
11.4 Employability of the Creden?aled ........................................................................ 138
11.5 Ins?tu?onal Capacity ............................................................................................ 139
iv

11.6 Programs and Ini?a?ves ....................................................................................... 141
11.7 Summary & Conclusion ......................................................................................... 143
12. Compara?ve Scale of Human Capital ....................................................................... 145
12.1 Interna?onal comparison of Educated Adults ....................................................... 146
13. Research & Development ......................................................................................... 147
13.1 Doctoral Comple?on ............................................................................................. 147
13.2 R&D Expenditure ................................................................................................... 147
13.3 Researchers & Technician in R&D ......................................................................... 148
13.4 Firms that do R&D and Innovate .......................................................................... 149
13.5 Innova?on .............................................................................................................. 149
13.6 Patent applica?ons & Hi-Tech Exports.................................................................. 150
13.7 Programs and Ini?a?ves ....................................................................................... 151
13.8 Summary ................................................................................................................ 155
14. Summary and Conclusion ......................................................................................... 157
14.1 Summary ................................................................................................................ 157
14.2 Conclusions ............................................................................................................ 160
15. Appendix: Legacy Issues in Literature....................................................................... 162
16. References ................................................................................................................. 165

Figures
Figure 1: Educa?on structures: Pyramid, Trapezium and Cylinder ............................................. 7
Figure 2: Long term trend in Adult Educa?on (% of 25+ pop with each qualiﬁca?on) ............... 9
Figure 3: Educa?on, Skill, and Job Pyramids: Interlinking challenge......................................... 10
Figure 4: Adult Literacy Rate (Age 15 and above) ..................................................................... 15
Figure 5: Adults with Primary Educa?on (% of popula?on 25+) ............................................... 16
Figure 6: Adult with Secondary Educa?on (% of popula?on ages 25+) .................................... 17
Figure 7: Literacy rate, youth total (% of people ages 15-24) ................................................... 18
Figure 8: Primary Enrollment & Comple?on Rate .................................................................... 20
Figure 9: School enrollment, Pre-Primary (GER) ....................................................................... 21
Figure 10: Minimum reading proﬁciency at end of primary (% of students) ............................ 26
Figure 11: Minimum Learning Proﬁciency: Std III (% of children) ............................................ 34
Figure 12: Minimum Learning Proﬁciency: Std V (% of children) ............................................. 35
Figure 13: Minimum Reading Ability FLN (Std 3): 2024 & 2014 ............................................... 46
Figure 14: Minimum Subtrac?on Ability FLN (Std 3): 2024 & 2014.......................................... 46
Figure 15: Minimum Reading Ability Primary (Std 5): 2024 & 2014 ......................................... 47
v

Figure 16: Minimum Division Ability Primary (Std 5): 2024 & 2014 ......................................... 47
Figure 17: Inter State varia?on of Primary Enrolment Ra?o with Per Capita NSDP.................. 51
Figure 18: Progression to secondary school ............................................................................. 62
Figure 19: Secondary School enrollment: NER/GER ................................................................. 63
Figure 20: Secondary comple?on rate, total (% of relevant age group) ................................... 63
Figure 21: Inter State varia?on of Lower Secondary Enrolment Ra?o with PCNSDP ................ 66
Figure 22: Inter State varia?on of Lower Secondary Enrolment Ra?o with PCNSDP ................ 68
Figure 23: Minimum Learning Proﬁciency: Std 8 (% of children) ............................................. 71
Figure 24: Minimum Reading proﬁciency (% of std 8) in 2024 & 2021 .................................... 73
Figure 25: Minimum Arithme?c proﬁciency (% of std 8 students) in 2024 & 2014 ................. 73
Figure 26: Performance of Students (%) by School Type .......................................................... 92
Figure 27: Computer and Internet facility in Schools (% of Schools) ........................................ 97
Figure 28: Reten?on Rate Across Grades (% cohort) .............................................................. 102
Figure 29: Class-wise Dropout Rates: In-Class & Between-Class ............................................ 103
Figure 30: Employment Ecosystem: Educa?on, Skilling, Placement & Jobs ........................... 104
Figure 31: Students in Voca?onal & Technical Programs (%) ................................................. 106
Figure 32: Adults with Post-Secondary & Short Cycle (ter?ary) training (% of pop 25+)........ 108
Figure 33: Ins?tu?onal Framework of MSDE .......................................................................... 111
Figure 34: Students in Industrial Training Ins?tutes (ITI): Enrolled & Passed ......................... 112
Figure 35: Students enrolled in ITIs across States & UTs ........................................................ 114
Figure 36: Availability of Polytechnics Seats, across States and UTs ....................................... 117
Figure 37: Skill India Digital Hub (SIDH): One pla?orm for all skilling needs .......................... 122
Figure 38: Services through eShram ...................................................................................... 122
Figure 39: Role of Govt(C&S), Private sector and NGOs, in an Inclusive Skilling System ........ 125
Figure 40: Ter?ary Schooling: College enrollment, (% gross) ................................................. 133
Figure 41: Ter?ary State wise Enrolment Rates at diﬀerent level, 2021-22 ........................... 136
Figure 42: Adults with Bachelor & Master Degree (% of popula?on 25+) ............................. 138
Figure 43: University & College Density (per lakhs): 2021-22 ................................................ 140
Figure 44: University & College Density (per lakhs): NE states & UT’s ................................... 141
Figure 45: Research and development expenditure (% of GDP) ............................................ 147
Figure 46: Researchers & Technicians in R&D (per billion people) ......................................... 148
Figure 47: High-technology exports (% of manufactured exports) ........................................ 151

Tables
Table 1: Educa?on Indicators: 1971, 1981, 2000, 2011 and 2018 .............................................. 8
Table 2: Projec?on of Adults age 25+ with diﬀerent Qualiﬁca?ons based ................................. 9
Table 3: Improvement in Literacy Rates (%) ............................................................................. 14
Table 4: Improvement in Educa?onal A?ainment Rate (%) ...................................................... 16
Table 5: Benchmark adult educa?on rates by Per capita GDP ................................................. 17
Table 6: Improvement in Primary enrollment and comple?on rates (%) ................................. 20
vi

Table 7: Major learning assessment surveys in India ............................................................... 23
Table 8: Founda?onal Literacy & Numeracy skills (% of students) ........................................... 28
Table 9: Learning outcomes in Primary (% of students) ........................................................... 29
Table 10: Minimum Reading Proﬁciency, MRP (% of children): 2006 to 2024 .......................... 29
Table 11: Minimum Arithme?c proﬁciency, MAP (% of children): 2006 to 2024 ..................... 30
Table 12: Eﬀect of RTE on Minimum Reading Proﬁciency (% of children) ................................ 31
Table 13: Eﬀect of RTE on Minimum Arithme?c Proﬁciency (% of children) ............................ 31
Table 14: Eﬀect of Pandemic on Minimum Reading Proﬁciency (% of children) ...................... 32
Table 15: Eﬀect of Pandemic on Minimum Arithme?c Proﬁciency (% of children) .................. 33
Table 16: Minimum Learning Proﬁciency (% of children) ......................................................... 36
Table 17: Minimum Learning Proﬁciency (% of students): Grade 5/Standard 5 ....................... 39
Table 18: Minimum Learning Proﬁciency (% of students): Grade 3 / Standard 3 ..................... 40
Table 19: Minimum Learning Proﬁciency (% of students): Grade 3 & 5 ................................... 41
Table 20: Students (%) at diﬀerent levels of proﬁciency: NAS 2021 ......................................... 42
Table 21: NAS 2021: Average Scores & Correct Answers (%)- Primary ..................................... 42
Table 22: Students (%) at diﬀerent proﬁciency levels in grade 3 .............................................. 43
Table 23: Average scores in grade 3 (PRS 2024) ...................................................................... 43
Table 24: Min Learning outcomes in Std 5(%), across States-2024 (& change from 2014) ...... 45
Table 25: Min Reading Proﬁciency across States – % of students in Grade 5, NAS 2017 ......... 48
Table 26: Min Reading Proﬁciency across States (% of students in Grade 3, FLS 2022) .......... 49
Table 27: Cross-correla?on Matrix- Std/Grade 5 MRP across States ........................................ 50
Table 28: Correla?on Matrix - Grade 3 MRP across States: ....................................................... 51
Table 29: Compara?ve Performance of States in Primary Enrolment ...................................... 52
Table 30: Survey of percep?on of secondary teachers on Diksha content & pedagogy ........... 54
Table 31: Improvement in Secondary enrollment & Comple?on (%) ....................................... 61
Table 32: Numbers of Youth comple?ng diﬀerent educa?on levels ......................................... 64
Table 33: Youth Comple?ng diﬀerent levels - India, China, MIC4 ............................................. 65
Table 34: Compara?ve Performance of States in Lower secondary Enrolment (ANER’) .......... 66
Table 35: Compara?ve Performance of States in Lower secondary Enrolment (ASER’) ........... 67
Table 36: Compara?ve Performance of States in Upper Secondary Enrolment ....................... 69
Table 37: Learning outcomes in 2024 & change 2006 to 2024 (% of students) ....................... 70
Table 38: Eﬀect of RTE on Learning outcomes (% of students) ................................................ 70
Table 39: Eﬀect of Pandemic on learning outcomes (% of students) ....................................... 71
Table 40: Min learning outcomes across Stats-2024 and change 2014-24 in bracket) ............ 72
Table 41: Minimum Learning Proﬁciency (% of students): Grade/Standard 8 .......................... 74
Table 42: Learning Outcomes in Grade 8 [NAS (2021) & ASER (2022)] .................................... 75
Table 43: Grade 8 students (%) mee?ng Basic performance level: NAS ................................... 75
Table 44: Class 10 Basic Language ability in States/UTs in 2021 ............................................... 76
Table 45: Proﬁciency in Class 10 Math in 2021 ........................................................................ 77
Table 46: Distribu?on of Grade 8 & 10 students by performance level (%) ............................. 77
vii

Table 47: NAS (2021): Students (%) mee?ng basic performance level in Grade 10 ................. 78
Table 48: NAS 2021: Average Scores & Correct Answers (%) ................................................... 79
Table 49: PRS 2024: Average scores (2024)- Secondary ........................................................... 80
Table 50: Pupil-teacher ra?os & per cent of Trained Teachers ................................................. 89
Table 51: Interstate correla?on of MRP & MAP with PTR & TT ................................................ 90
Table 52: Improvement in Percent of Trained Teachers (%) ..................................................... 91
Table 53: Female-to-Male Ra?o in Educa?on of Indian children .............................................. 91
Table 54: Trend by School Type: 2010 - 2024 ............................................................................ 93
Table 55: Eﬀect of RTE on Minimum Learning Proﬁciency (% of students) .............................. 94
Table 56: Students (%) by performance levels, 2021 ............................................................... 95
Table 57: Overall Performance of Students: Average Scores & Correct Answers ..................... 96
Table 58: Overall performance of students: Average scores (2024) ......................................... 96
Table 59: Computer and Digital Ini?a?ves ................................................................................ 98
Table 60: Eﬀect of Computer and Digital Facili?es on Learning: Cross-State correla?on ......... 99
Table 61: Schools with Tinkering Labs (%) .............................................................................. 100
Table 62: Schools having ICT Labs (%) .................................................................................... 100
Table 63: Reten?on Rate by level of educa?on (%) ................................................................ 101
Table 64: Dropout Rate in Primary and Secondary school (%) ............................................... 102
Table 65: Students enrolled in ITI by Ins?tu?on type & course dura?on ............................... 112
Table 66: ITI enrollment in engineering & non-engineering streams (%) ............................... 113
Table 67: Employability (Below Graduate), 2019-2025 .......................................................... 117
Table 68: Skilling and cer?fying Trainers (2022) ..................................................................... 119
Table 69: CraL Instructor Training .......................................................................................... 120
Table 70: Firms oﬀering formal training (% of ﬁrms) .............................................................. 120
Table 71: Ter?ary enrolment Gross rates & numbers ............................................................. 134
Table 72: Comparison of Youth enrolled in Ter?ary educa?on ............................................... 134
Table 73: Enrolment, Share & Growth in Regular & Distance Programs: 2021-22 ................. 135
Table 74: Performance of States in Ter?ary Enrolment: All Types (2021-22) ......................... 137
Table 75: Employability of Graduates from University & Professional programs ................... 139
Table 76: Share of diﬀerent type Universi?es, & Change (2017-18 to 2021-22) .................... 140
Table 77: Number of Adults (25+) with diﬀerent levels of Educa?on ..................................... 145
Table 78: Interna?onal comparison of adults with diﬀerent levels of educa?on ................... 146
Table 79: Firms that spend on R&D (% of ﬁrms) ..................................................................... 149
Table 80: Global Innova?on Index Ranking, 2025 .................................................................. 150
Table 81: Patent applica?ons, Residents /Million popula?on................................................. 150

1

1. Introduc?on
In the context of India’s goal of sustained, fast, inclusive growth to achieve a Viksit Bharat,
educa?on and skilling are cri?cal for employment and wage growth. Educa?on builds human
capital which has been recognized as an important complement to Physical Capital. Educa?on
implies learning, and only learning outcomes ma?er for subsequent acquisi?on of job skills
and produc?ve employment. Educa?on cannot be measured by the cer?ﬁcates issued for
primary school, lower secondary school and upper secondary school, if they merely indicate
the number of years spent in school, if the majority of students don’t have minimum
proﬁciency in Reading and Arithme?c. Such cer?ﬁcates do not necessarily create human
capital.
Since 1960, economists have emphasized the importance of schooling in economic
development and growth. Human capital was commonly measured in terms of average years
of schooling of the labor force, in the empirical literature on produc?on and produc?vity. In
the early 2000s economists began to ques?on the use of simple measures like primary or
secondary school enrollment or years of schooling (formal comple?on rates), in determining
growth. Several researchers found that the link between average years of educa?on & similar
indicators, and actual learning, was tenuous. They hypothesized that there was a gap between
years of schooling and accumula?on of human capital. As Pritche? (2001) put it, “the true
problem is that years of schooling do not reﬂect learning.” “The educa?on system has failed,
so a year of schooling provides few (or no) skills.” Interna?onally comparable test results
showed large diﬀerences between average marks of children in many developing countries,
and those from East Asian & OECD countries. This suggested low quality of schooling in the
former, rela?ve to the la?er.
2

Quality of educa?on and cogni?ve ability of labor force was recognized as a key diﬀeren?ator
of the quality of human capital. In par?cular, the performance on interna?onally comparable
tests of math and science was shown to be highly signiﬁcant factor in economic growth. Since
then, the importance of interna?onal tests of language and math learning outcomes, such as
PISA, TIMSS and NAEP have received more a?en?on across the World. Over the last decade
OECD and others have tried to directly measure literacy, numeracy & problem-solving ability
(PIAAC) of adults across the OECD. Based on this data, researchers have constructed learning
adjusted years of schooling.
3
And other indices, which combine quan?ty & quality of
schooling. This data also allows researchers to directly link adult workplace skills to
produc?vity and employment outcomes.
4

2
“Low quality of schooling is consistent with the macroeconomic evidence and is obviously consistent with the household evidence of li?le
or no wage increment from addi?onal schooling.” Pritche? (2001)
3
World Bank/Angrist et al. (2021).
4
Survey of Adult Skills-PIAAC [OECD (2023)]
2

A recent review by Nancy Birdsall (2025) of the economic growth of the “miracle growth” East
Asian economies like South Korea, Taiwan, Malaysia and Thailand has re-emphasized this
point. To quote “the key to the unusually high total factor produc?vity growth of the East Asian
economies, which dis?nguished them from most other developing countries, was the
“technological learning” associated with their export push (as exporters learned from and
through interac?ons with impor?ng ﬁrms abroad). But behind technological learning at the
factory level, was the high level of good-quality schooling of managers and workers. In other
words, a key lesson from East Asia is the emphasis, going back decades, on high-quality
primary and secondary educa?on and on secondary and post- secondary technical educa?on.”
(Birdsall (2025))
The goal of Primary educa?on is to ensure that students master literacy and numeracy by
grade three (approximately age 10) so they can access the broader school curriculum. Children
who cannot read by grade three struggle to catch up, eventually falling so far behind that no
learning occurs at all. The core ingredients of this element include learning assessments of
each child, reading and math assistance for students in grades one to three who need
addi?onal support, and an assessment at the end of grade three to iden?fy children at risk.
5

India’s NEP has adopted some of the lessons learnt, but there are signiﬁcant gaps in our
understanding of the quality of educa?on and of learning outcomes across India. Public
knowledge about skilling and skill development systems is even more limited. This paper uses
available data to present a more detailed view of India’s Educa?on system, with par?cular
focus on learning outcomes. It also explores India’s skilling system and the links between
educa?on, skill development and employment
Educa?on and employment are typically dealt with in separate silos. Voca?onal and
Technical educa?on is oLen another separate silo. Job placement & matching skills to jobs is
seldom the primary responsibility of any public ins?tu?on. Lifelong learning is now a buzz
word, but few understand the importance of “learning by doing” and the links to the quality
and appropriateness of educa?on and skilling.
Skilling is both a necessary response to the limita?ons of the current educa?on system as well
as a job creator in itself. Poor founda?onal learning leads to dropouts at both primary and
later stages. Even those who formally complete school or higher educa?on oLen lack prac?cal,
job related, competencies. Therefore, improving the quality and relevance of educa?on is
essen?al not only to reduce dropout rates, but also to ensure that young people can acquire
relevant, job-ready skills. At the same ?me, students who do leave the system early, must be
provided with alternate skilling pathways, so that no young person remains idle or excluded
from the economy. The large share of self employed in employment (58%) is an addi?onal
important reason for emphasizing voca?onal educa?on & training.

5
World Development report (WDR) 2019.
3

This paper explores the system of schooling, skilling, placement and employment, a system
riddled with informa?on and co-ordina?on problems, incomplete and missing markets.
Historically, controls and restric?on on provision of educa?on by private commercial and non-
commercial organiza?ons, constraints on size of ﬁrms, labor regula?ons and even
cons?tu?onal requirements and judicial interpreta?ons, have contributed to slow
development of markets for human capital. This paper a?empts to ﬁll the informa?on gap by
examining general, voca?onal educa?on & training, technical & professional educa?on, job
skilling, job placement, learning by doing & ﬁrm provided training, and wage growth, as parts
of an integrated view of the employment eco-system. It also explores cri?cal parts of the
system for crea?ng employment, higher paying jobs and wage growth.
As India aims to become a developed country by 2047 under the PM’s vision of “Viksit
Bharat”, improving the produc?vity of its young popula?on is essen?al.
6
While access to
educa?on has expanded, learning outcomes remain weak. There is a gap between the
creden?al or degree and the learning expected of students with those qualiﬁca?ons. There is
also a gap between what is learned in school & college and the basic requirement of jobs;
many students earn degrees but lack the job skills needed to secure matching jobs. As a result,
formal educa?on alone does not guarantee be?er wages or economic mobility.
Sec?on 2 provides an overview of the paper, an execu?ve summary for policy makers. Sec?on
3 provides a framework within which the educa?on & skilling system is analyzed. The New
Educa?on Policy NEP (2020) is a key part of the framework, as it is designed to address these
issues. The NEP is summarized in this sec?on. Sec?on 4 takes stock of 75 years of educa?on
since independence; in terms of the educa?on levels of the adult popula?on compared
against global benchmarks. Sec?on 5 focuses on primary educa?on, presen?ng interna?onal
comparisons on primary school enrollments, comple?on rates and founda?onal learning. It
also explores rural learning outcome, dispari?es across states. Sec?on 6 moves to secondary
educa?on; Star?ng with the transi?on from primary to secondary educa?on it then analyzes
enrollment and comple?on rates using available data on learning outcomes. Sec?on 7
highlights school level issues including pupil teacher ra?o, gender parity, school types and the
state of digital infrastructure such as internet and computer access in schools. Sec?on 8
presents available data on student reten?on and school dropout rates, as a point of transi?on
for many youths in India, from school to jobs, and the corresponding unﬁlled need for training
for low skill jobs. Sec?on 9 provides an overview of the complex employment eco system.
Sec?on 10 addresses the skilling landscape by assessing the scale and structure of voca?onal
educa?on, formal training by ﬁrms, and the role of ins?tu?ons like ITIs and NCVT. Sec?on 11
covers ter?ary educa?on, with a focus on enrollment, comple?on, and the quality of skills
training in rela?on to employability. Sec?on 12 focuses on adult educa?on, par?cularly the
size and scalability of India’s educated labor force in compara?ve perspec?ve. Sec?on 13
examines research and innova?on through indicators such as doctoral comple?ons, R&D

6
Reference is to labor produc?vity and the role of educa?on & job skills in enhancing it.
4

expenditure, researcher and technician density, ﬁrm-level investments, innova?on outputs,
and patent ac?vity. Sec?on 14 concludes the paper.
2. Overview
The paper starts with the analysis of the post-independence structure of educa?on during
the ﬁrst 50 years or so and the legacy it leL in the form of adult literacy and educa?on. Partly
as the result of the cylindrical structure of educa?on with a narrow base and over emphasis
on higher educa?on, adult literacy rate, and the share of adults with primary and secondary
educa?on were s?ll below the level appropriate for our per capita GDP. Ter?ary educa?on
was, in contrast at or above the level expected at our per capita GDP.
Posi?ve adjustment in the educa?on system in the last 15 years or so are reﬂected in Youth
literacy rates, which were 14% higher than the minimum expected at our Per capita GDP level.
This in turn reﬂects the good performance in Pre-primary and primary educa?on enrolment.
Primary enrolment and comple?on rates were signiﬁcantly above the benchmark rates for our
per capita GDP. Somewhat surprisingly the average quality of teaching and learning as
measured by minimum reading proﬁciency was also above the interna?onal compara?ve
benchmarks. There is however a challenge before us, to raise the minimum learning outcomes
to the benchmarks appropriate for the per capita GDP levels which we are aiming to reach by
2047. This challenge is much higher for raising Minimum arithme?c proﬁciency than for raising
Minimum reading proﬁciency.
The next stage in the educa?on pyramid is the reten?on rate/drop-out rate of students
between primary and secondary educa?on. In interna?onal compara?ve perspec?ve India’s
reten?on rates are above the benchmark rates for its per capita GDP. The enrolment rate in
secondary school is just above the benchmark rate. State enrolment rates are found to be the
only educa?on indicator corelated with PCNSDP, in both lower secondary and upper
secondary school. The comple?on rates in both lower secondary school and upper secondary
school are signiﬁcantly above the benchmark for our per capita GDP. Only a modest eﬀort is
needed to achieve the lower secondary comple?on rates expected for a high-income country,
the achievement of benchmark levels for upper secondary comple?on rates will be more
challenging. All India rates would have to rise from the current 63% to 79% in 20 years.
The limited informa?on available on Minimum learning proﬁciency in Lower secondary school
(Std VI to VIII), however, introduces a note of cau?on. In rural areas, 42% of Std VI, 36% of Std
VII and 29% of Std VIII students cannot read a Std II level text. Though we don’t have direct
data for urban areas we es?mate that MRP may be 10%-12% points be?er than in rural areas.
The situa?on is even more challenging with respect to minimum arithme?c proﬁciency (MAP);
64% of rural students in Std VI, 59% in Std VII and 54% in Std VIII cannot do division.
Interna?onal comparisons of Indian skilling parameters were a lot less encouraging. The share
of secondary students enrolled in Voca?onal and Technical training was 13% points below the
5

interna?onal benchmark for a country at India’s per capita GDP. Interna?onal compara?ve
data on the share of adults who have taken post-secondary training and share of adults who
have taken ter?ary, short cycle courses show a somewhat be?er picture. Compared to the
benchmarks for its Per capita GDP, India’s adult popula?on is only (-)6% points below the
benchmark for post-secondary and (-)3% points below the bench mark for ter?ary courses. A
number of schemes and portals have been launched in the past ten years to remedy this
situa?on, but raising them to UMIC and HIC benchmarks in ﬁve and 25 years respec?vely, will
require a herculean eﬀort.
Firms providing formal training to their employees, is 7.7% in India, 8.7% in Vietnam and 8.4%
in Indonesia. This could be partly due to more compe??ve labor markets for corporate
employees. The incen?ves can be changed by allowing Companies and large ﬁrms to use a
much greater frac?on of their funds to train both employees and to upgrade ITIs and
Polytechnics. This will be cri?cal, if we want Indian private sector to compete with Thailand,
Malaysia and Mexico, with 18%, 24% and 38% of their ﬁrms have employee training programs.
India ‘s ter?ary educa?on is compara?vely well placed in interna?onal compara?ve frame.
India’s ter?ary enrolment rate is almost at the exact benchmark for its per capita income. Its
Adult popula?on with Bachelor’s degree is slightly above, and Master’s degree a li?le below
the benchmark for its per capita income. Given its enormous popula?on this makes its total,
ter?ary educated adult popula?on, the second highest in the World aLer China.
A cri?cal goal of the educa?on and skilling system is employment. The Global employability
Test (GET) is one way to measure its eﬀec?veness. Employability of those trained by this
system can be divided into three categories. Those gradua?ng from Voca?onal ins?tu?ons
(VET) like Industrial Training Ins?tutes (ITI) and Polytechnics have the lowest employability,
and those gradua?ng from professional ins?tu?ons such as Engineering (BE/BTech,
Computers (MCA) and Management (MBA), have the highest employability. Graduates of the
University-college system (BA, BCom, BSc, MA) fall in middle of the employability spectrum.
At the top of the system are the Professional Ins?tu?ons with employability in the 70-78%
range in 2025. Next come the Universi?es(-colleges), with employability in the 55-58% range.
Then come the ITIs at just over 40% and ﬁnally the Polytechnics at a li?le under 30%. The
good news is, that Employability has improved across the board during the past six years for
which data is available. The simple average increase of employability is 10% points for
Polytechnics, 20% points for universi?es, and 28% points for professional ins?tu?ons (2019-
2025).
The benchmark used in this paper are generally a func?on of per capita GDP. Absolute
numbers of educated and skilled popula?on, labor and work force is relevant when we
deﬁning compara?ve advantage rela?ve to both our comparator countries and vis-à-vis other
countries. Future compara?ve advantage depends on the rela?ve numbers of school &
university educated being graduated from its schools & colleges respec?vely. Even with India’s
school comple?on rates lower than higher income comparator countries, the compara?ve
6

numbers are higher than China’s and oLen exceed it. The number of children comple?ng
primary educa?on in India (137 mil) were 1.25 ?mes those in China and more than 2.5 ?mes
the combined total of Vietnam, Indonesia, Thailand and Mexico (MIC4). The number of
children comple?ng lower secondary school (65 mil) were 1.2 ?mes those of China. The
number of children comple?ng upper secondary school (98 mil) were also 1.2 ?mes those
China and 3.7 ?mes those of the MIC4 (Vietnam, Indonesia, Thailand & Mexico) combined.
The number of students enrolled in ter?ary educa?on (52 mil) was more than double those
of the four comparators combined.
The current stock of adults with diﬀerent levels of educa?on, also compares well with relevant
comparator countries. Broadly India has a larger number of adults with Ter?ary educa?on
(Bachelors & Masters) than China, while China has a higher number of adults with primary &
secondary educa?on; India’s adult popula?on with bachelor’s degree was 1.4 ?mes China’s,
while its adults with secondary educa?on were 0.85 China’s. India’s stock of adults with
primary, lower and upper secondary educa?on was 1.8 to 2 ?mes the corresponding numbers
for MIC4 (combined).
3. Framework and Methodology
The development of Indian educa?on was framed by the cons?tu?on. Educa?on including
Universi?es was entry No. 11, in List II (State list) of the cons?tu?on. The 42
nd
amendment of
the cons?tu?on in 1976, eﬀec?ve January 1977, deleted this entry and inserted Educa?on as
entry 25, in list II (concurrent list) of the cons?tu?on. Parliament obtained the right to set All
India norms, while all other rights remained with the States as long as they do not collide with
union law (Art 254). Entries 63-65, List I (Union list) in which Parliament had the right to
declare ins?tu?ons of Na?onal importance, establish agencies for specialist training &
research, and legislate academic standards in ins?tu?ons of higher educa?on and research,
con?nued unchanged.
This meant that Primary and Secondary educa?on (primary, middle and high school) was the
sole responsibility (and right) of the States, from independence in 1947 to 1976, when
educa?on was moved to the concurrent list of the cons?tu?on. Thus for 30 years, Indian States
had the complete and full responsibility to provide educa?on to its residents and the na?onal
result in primary & secondary educa?on, were a mere aggrega?on of diﬀerent results
obtained by States each with own goals and priori?es. The cons?tu?onal posi?on was
therefore one of decentraliza?on, similar to the USA in some ways, but very diﬀerent in others.
The USA had an even more decentralized system for educa?on than India, with local
authori?es playing a big role in primary and secondary educa?on, ﬁnanced by local taxes. USA
successfully built a pyramidal system of educa?on, with a wide base, a narrower middle and
a highly eﬀec?ve and eﬃcient high end (Figure 1, leL panel). At the other end of the
development policy spectrum, the USSR, with a very centralized na?onal educa?on system,
7

also created a pyramidal structure of educa?on, but with a much narrower middle than the
USA.
7

Figure 1: Educa?on structures: Pyramid, Trapezium and Cylinder

In contrast to both USA and USSR, India created a cylindrical structure of educa?on in the
early decades (right end of Figure 1). This was partly due to the cons?tu?onal division of
responsibili?es, in which Na?onal government had a substan?al role in higher educa?on, but
States had absolute authority and control over primary and secondary educa?on for four-ﬁve
decades. The adult literacy rate was 40.8%, and youth literacy rate was 53.8% in 1981, despite
the shiL of Educa?on from the State list to the Concurrent list of the cons?tu?on in 1977. The
adult popula?on with “lower secondary” educa?on, was 16.2% of all adults 25 years old &
over (1981). The low youth literacy rate reﬂected the primary school enrolment rate of 61%
in 1971 (Table 1).
The USA’s success in Primary and secondary educa?on was driven by local Govts and
communi?es. They created the Govt (public) educa?on system, using local taxes to create
universal access to Primary educa?on by 1910 and near universal secondary educa?on (~73%)
by 1940, from ~19% in 1910 and ~50% in 1930. At the State level India’s educa?on system,
with Primary & secondary educa?on centralized in the State capitals, with no decentraliza?on
to local rural and urban levels, was more akin to the USSR’s centralized na?onal system. USSR
raised the primary enrolment rate from 40% in 1914 (Tzarist Russia) to 60% in 1928 (post
revolu?on recovery) to 95% in 1932.
8
India in contrast had a primary enrolment rate of 79.5%,
and a primary comple?on rate of 75.3% in 2000, 50 years aLer independence. India did not

7
Like Eiﬀel tower
8
USSR was much less successful in raising the secondary enrolment rate which was 15-18% in 1930 and ~21% in 1940, with only10-15%
comple?ng secondary School.
Pyramid Trapezium Cylinder
Lower Secondary
Primary
Tertiary
Upper
Secondary
A
B
C
D
A
B
C
D
Note: A. Tertiary, B. Upper Secondary, C. Lower Secondary, D. Primary
8

have universal primary educa?on, and its secondary school gross enrolment rate of 47.2%
(2000), was half way between that of the USA and USSR in their educa?on system
development phase (1930-1940).
Table 1: Educa?on Indicators: 1971, 1981, 2000, 2011 and 2018

Data source: World Development Indicators, December 2025. Historical data is available for selected years for each variable.
USA’s private sector (including endowments and trusts) was welcomed and encouraged to
play a role in USA’s highly successful higher educa?on system, Indian Courts have historically
restricted private for-proﬁt educa?on.
9
The USSR succeeded in crea?ng a highly educated
(STEM) industrial workforce and in achieving speciﬁc technological milestones, especially in
defense and space.
10
India can learn much from the experience of USA’s highly successful
higher educa?on system, but the USSR and PRC experience is also relevant, because much of
India’s higher educa?on and R&D system is s?ll administered by Government.
Table 2 illustrates the evolu?on of India’s educa?on system from a cylindrical structure
through a Trapezoid to a Pyramid, based on regression equa?ons for adult educa?on, shown
in Figure 2. Figure 2 uses available data on adult qualiﬁca?ons (primary, lower secondary,
upper secondary and Bachelor) in diﬀerent years to run an intertemporal regression. The best
ﬁ?ng equa?ons are then used to interpolate & extrapolate the data (upper panel of Table 2).

9
In Unni Krishnan v. State of Andhra Pradesh (1993), the Supreme Court ruled that educa?on is a fundamental right under Ar?cle 21 but
cannot be commercialized or treated as a business for proﬁt, prohibi?ng "proﬁteering" while allowing only reasonable surplus for expansion.
In T.M.A. Pai Founda?on v. State of Karnataka (2002), it reinforced this by permi?ng private unaided ins?tu?ons autonomy but manda?ng
they avoid proﬁt mo?ves, aligning with Ar?cle 30 (minority rights) and public policy against educa?on as trade.
10
The state provided Universal Access at the Base, rigorous Selec?on for Advancement, Structured Academic and Career Ladder, and Elite
Ins?tu?ons at the Top. Compe??ve exams and specialized tracks (such as physics-maths schools and Olympiads) enabled academically giLed
students to rise through the system. Advancement through ins?tu?onal hierarchies provided structured mobility for those who met
performance standards.
Education Indicator (% of relevant group)19711981200020112018
Adult Literacy (% of pop 15+) 40.861.069.373.7
Youth literacy (% of pop 15-24) 53.876.486.194.6
Primary education, pop25+ (%) 39.051.459.6
Lower Secondary education, pop25+ (%) 5.016.2 37.646.1
Upper Secondary education, pop25+ (%) 9.826.932.3
Bachelor Education, pop25+ (%) 9.110.8
School enrollment, primary (% net)61.0 79.590.4
Primary completion rate, (% of relevant age) 75.393.3
School enrollment, secondary (% gross) 47.267.176.0
School enrollment, secondary (% net)
Lower Sec Completion, (% of relevant age) 77.4
School enrollment, tertiary (% gross) 9.923.328.4
9

Table 2: Projec?on of Adults age 25+ with diﬀerent Qualiﬁca?ons based

Source WDI, Dec 2025.Authors calcula?ons based on intertemporal regressions in ﬁg 2.
Figure 2: Long term trend in Adult Educa?on (% of 25+ pop with each qualiﬁca?on)

Source WDI, Dec 2025. Lines represent intertemporal regressions. Equa?on & R^2 are shown.
In 1971 the structure was a double cylinder (Figure 1, right panel), with the ra?o of Bachelors
to Primary and Upper secondary/primary an es?mated 0.31 and 0.29 respec?vely, with lower
secondary/primary at 0.98 (rows 7, 8 & 9, Table 2).
11
By 1981 these ra?os had become 0.15
for bachelors/primary and 0.28 for upper secondary/primary, transi?oning to a double
trapezoidal structure (Figure 1, middle panel). Pyramidal structure began to emerge during
2000-2011 with Bachelor/primary ra?on of 0.14, upper secondary/primary ra?o of 0.27 and

11
Literally a cylinder on top of a trapezium.
197119811991200020112019
Qualification
Bachelor Degree1.72.63.95.68.712.0
Upper Secondary1.84.95.810.822.434.7
Lower Secondary5.814.322.830.439.746.5
Primary 6.017.428.939.351.961.1
Bachelor Degree0.290.150.130.140.170.20
Upper Secondary0.310.280.200.270.430.57
Lower Secondary0.980.820.790.770.770.76
Primary 1.001.001.001.001.001.00
Ratio of estimated value to Primary
Estimated Value based on regression
y = 1.1485x + 4.8079
R² = 0.9876
y = 0.8467x + 5.0017
R² = 0.9889
y = 1.7133e
0.0611x
R² = 0.9506
y = 1.6536e
0.0404x
R² = 0.9226
0
10
20
30
40
50
60
70
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
Adult Education: % of 25+ pop with each qualification
Primary Lower Sec Upper Sec Bachelor
Trend:Primary Trend:LowSec Trend:UpSec Trend:Bac
10

lower secondary/primary ra?o of 0.77 in 2000 and ra?os of 0.17, 0.43 and 0.77 respec?vely
in 2011 (Table 2, lower panel).
As we survey the human capital eco-system in India today, educa?on, job skills and
employment opportuni?es can be viewed as three pyramids. The educa?on pyramid is built
on a base of primary educa?on, with secondary educa?on in the middle and ter?ary
educa?on at the top and researchers at its pinnacle or ?p. Similarly, the Skills pyramid is built
on basic/low skills, with “semi-skilled” in the middle, and skilled on top (Figure 3).
Figure 3: Educa?on, Skill, and Job Pyramids: Interlinking challenge

The educa?on and skill pyramids cannot grow if either the founda?ons are weak or the
connec?vity between educa?on and job skills at each level, is poor. Finally, the job pyramid
consists of minimum/low wage, Average/medium wage and high wage employment (ﬁgure
3). Each level of the employment pyramid must be linked to the appropriate levels of the skill
pyramid and the educa?on pyramids to generate the most produc?ve employment, on
average. Real wage growth within each layer, also depends on “on-the-job training” and
“learning by doing,” and movement between layers on upskilling and re-educa?on.
The connec?vity between the Ter?ary level of educa?on, skilled persons and High wage jobs
is far be?er than for semi-skilled worker and individuals with primary educa?on or less. The
demand from large ﬁrms creates proﬁtable opportuni?es for private placement agencies and
skill providers. The semi-skilled (medium skills) level is less proﬁtable but much more
important in India than in other middle-income countries (MICs). Self-employed workers
cons?tute 57.7% of all workers, of which about 11.8% are professionals. Thus 45.9% of them
need sound secondary educa?on and training in semi-skilled occupa?ons. As regular salary
and wage employment cons?tutes only 23.2 per cent of the labor force, while 19 % are casual
labor, the la?er need be?er primary educa?on and basic skills to operate eﬃciently in a
modern economy. Labor market linkages in the informal sector are either weak or non-
existent, and even within the formal sector, the eﬃciency of job matching mechanisms varies
across semi-skilled and skilled occupa?ons.
Secondary
Primary
Tertiary
Skills
Semi-skilled
Basic
Skilled
High
Average Wage
Minimum Wage
High Wage
V. High
JobsEducation
R & D
11

The above framework seems to underly the Na?onal Educa?on Policy, given that Voca?onal
educa?on and Training was covered by it, and the emphasis NEP placed on training & ac?vi?es
outside the boundaries of the conven?onal, formal academic curriculum. It is therefore
necessary to understand the goals and sugges?ons of NEP (2020), before placing Indian
educa?on and skilling system in an interna?onal compara?ve framework.
3.1 Na?onal Educa?on Policy (NEP 2020)
India’s Na?onal Educa?on Policy (NEP 2020) is a policy framework that aims to address many
of the persistent challenges in the Indian educa?on system. It structured Indian educa?on
system into four key stages: Founda?onal Stage: Pre-primary to Class 2, Preparatory Stage:
Class 3 to 5, Middle Stage: Class 6 to 8, Secondary Stage: Class 9 to 12.
Founda?onal Learning and numeracy (FLN) were iden?ﬁed as the highest priority, with the
goal of having every child able to do basic reading, wri?ng and arithme?c by Grade 3.
12
A
Na?onal Mission on FLN was launched, with States & UTs expected to prepare implementa?on
plans with targets, and track progress. Suppor?ng measures included ﬁlling local language
teacher vacancies, daily prac?ce of reading and arithme?c, extra “bridge” modules in early
grades, and intensive early-year play-based modules. NCERT/SCERT were to develop ac?vity-
based workbooks, and a na?onal repository of FLN materials to be made available on the
DIKSHA pla?orm. Peer-tutoring and community volunteer programs were to be mobilized to
reinforce FLN. Libraries (physical and digital) were to vastly expanded with high-quality, age-
appropriate books (including local-language transla?ons) to build a reading culture.
13

NEP 2020 emphasized, that “pedagogy must evolve to make educa?on more experien?al,
holis?c, integrated, inquiry-driven, discovery-oriented, learner-centered, discussion-based,
ﬂexible, and enjoyable. The curriculum must include basic arts, craLs, humani?es, games,
sports and ﬁtness, languages, literature, culture, and values, in addi?on to science and
mathema?cs.” “Experien?al learning was to be adopted, including hands-on learning, arts-
integrated and sports-integrated educa?on, story-telling-based pedagogy,” was to be adopted
at all stages. among others, as standard pedagogy within each subject, and with explora?ons
of rela?ons among diﬀerent subjects.
14
To close the gap in achievement of learning outcomes,
classroom transac?ons were to shiL, towards competency-based learning and educa?on. The
assessment tools were to be aligned with the learning outcomes, capabili?es, and disposi?ons
as speciﬁed for each subject of a given class.”
NEP (2020) recommended a number of ac?vi?es to make learning more enjoyable and create
job-oriented skills. (1) Grades 6-8 students to take a fun course, which gives a survey and
hands-on experience, of a sampling of important voca?onal craLs, such as carpentry, electric
work, metal work, gardening, po?ery making, etc., as decided by States and local communi?es

12
The original, pre-covid, target date was 2025.
13
Greater emphasis needed on library of teaching/training videos and videos of experiment to s?mulate interest in natural phenomena.
14
Nature study and explora?on of plants, insects, animals, and geology (rocks, need- weather to develop scien?ﬁc explora?on.
12

and as mapped by local skilling needs. 2) A prac?ce-based curriculum to be designed by NCERT
while framing the NCFSE 2020-21(grades 6-8). 3) All students were to par?cipate in a 10-day
bagless period during grades 6-8, where they intern with local voca?onal experts such as
carpenters, gardeners, po?ers, ar?sts, etc. 4) Similar internship opportuni?es to learn
voca?onal subjects were to be made available to students throughout Grades 6-12, including
holiday periods. 5) Voca?onal courses through online mode to be made available. 6) Bagless
days to be encouraged throughout the year for various types of enrichment ac?vi?es involving
arts, quizzes, sports, and voca?onal craLs. 7) Children to be given periodic exposure to
ac?vi?es outside school through visits to places/monuments of historical, cultural and tourist
importance, mee?ng local ar?sts and craLsmen and visits higher educa?onal ins?tu?ons in
their village/Tehsil/District/State
15
. 8) Develop a clear ac?on plan with targets and ?melines,
to have at least 50% of learners through the school and higher educa?on system, exposed to
voca?onal educa?on.
NEP (2020) recommended that teachers be given more autonomy in choosing aspects of
pedagogy, so that they may teach in the manner they ﬁnd most eﬀec?ve. Teachers should
also pay a?en?on to socio-emo?onal learning - a cri?cal aspect of any student’s holis?c
development. Teachers would be recognized for novel approaches to teaching that improve
learning outcomes. Each teacher was expected to par?cipate in at least 50 hours of CPD
opportuni?es every year for their own professional development, driven by their own
interests. CPD opportuni?es would cover the latest pedagogies regarding founda?onal
literacy and numeracy, forma?ve and adap?ve assessment of learning outcomes,
competency-based learning, and experien?al learning, arts-integrated, sports-integrated, and
storytelling-based approaches.
A Na?onal Assessment Centre, PARAKH (Performance Assessment, Review, and Analysis of
Knowledge for Holis?c Development), was set up as a standard-se?ng body under MHRD, to
set norms, standards, and guidelines for student assessment and evalua?on for all recognized
school boards of India, and help school boards to shiL their assessment pa?erns towards
mee?ng the skill requirements of the 21st century. It is also to guide the State Achievement
Survey (SAS) and undertake the Na?onal Achievement Survey (NAS), monitoring learning
outcomes in the country. PARAKH is also mandated to advise school boards regarding new
assessment pa?erns and latest researches, promote collabora?ons between school boards,
sharing of best prac?ces among school boards, and for ensuring equivalence of academic
standards among learners across all school boards.
3.2 Methodology: Interna?onal Comparison
The paper includes numerous interna?onal comparisons using data from World Bank, World
Development Indicators (WDI). A common methodology is followed in this comparison. For
each indicator (Y), the data is collected for all countries (i) and for the latest available year (t);

15
Op cit
13

e.g., enrolment rates, comple?on rates, literacy rate, propor?on of popula?on with primary
educa?on. For each country the Per capita GDP at PPP in constant 2021 interna?onal prices,
is collected for the same year (Xit), to create matched pairs of the indicator and PCGDP for
same year. Most social indicators (Y) are posi?vely corelated to Per Capita GDP (X), i.e., Y =
f(X).
16
Assuming a polynomial func?on, this can be wri?en as,
Y = A + B X + C X
2
, (1)
And the es?ma?on equa?on is,
Yit = A + B Xit + C Xit
2
+ Eit, where Eit is the error term, t is the year, i = 1 …………. N (2)
The cross-country regression for the latest available year t, yields es?mate of A, B and C. Using
the es?mates, we can calculate the “expected value” of the indicator (Yo) for any speciﬁc value
of PcGDP (Xo), as
Yo = A + B Xo + C Xo
2
(3)
The Gap between the actual value of the indicator and the expected value, can be calculated
using equa?on (2), as
GAPi = Ei = Yi – (A + B Xo + C Xo
2
) (4)
These can also be depicted on an XY graph, with a regression trend line through the data
points. This line represents the “expected value” of the indicator at each level of Per capita
GDP at PPP. The GAP is the ver?cal distance between the data point and the line, with points
above the line having a posi?ve gap and those below the line having a nega?ve gap.
In this paper we also use two key threshold levels for PcGDP PPP in 2021 interna?onal dollars:
One is $12,800 (X1), the PcGDP PPP of Indonesia aLer it became an Upper middle country
(UMIC) in 2019, and the second is $37,000 (X2), the PcGDP PPP (2021) of Romania when it
became a High-Income country (HIC) in 2019. By inser?ng these in the right side of equa?on
(3) in place of X, we can calculate the values of the indicators as Y1 and Y2, respec?vely. These
are used as minimum benchmarks for a lower income country like India, aspiring to become
a UMIC or HIC.
The latest year for which data is available for a country (j) is some?mes very old (e.g., 2013),
rela?ve to that of most countries (generally 2023). In this case we es?mate a value for the
indicator for the country j by assuming the GAPj (2013) which prevailed in 2013, remains
unchanged ?ll 2023, when the PCGDP is X3. The es?mated value of indictor is as follows,
Yj (2023) = GAPj (2013) +( A + B Xj3 +C Xj3
2
), where Xj3 = PCGDP-PPPj (2023) (5)
The paper also uses a baseline projec?on of India’s PCGDP PPP to determine when it will
cross the above men?oned UMIC & HIC thresholds of per capita GDP PPP of $12,800 and
$37,000 respec?vely. PCGDP projec?on used in this paper, is based on an average growth of

16
In this illustra?on we use a polynomial func?onal form. In the paper we have also used linear, log and exponen?al func?ons.
14

PCGDP PPP (const 2021 Int$), of 6.4% in current decade, 5.4% in next decade and 4.4% in third
decade, with an overall three-decade average of ~5.5%. In this base case, India will cross the
UMIC PCGDP PPP threshold of $12,800 around 2030 and the HIC threshold of $37,000 around
2050 [Virmani (2024)]. If growth rates are higher than this base case, then every educa?on
and job skilling challenge outlined in this paper, will be even tougher than stated in the paper.
4. Legacy of State Eﬀorts
75 years have passed since independence the literacy and educa?on levels of the popula?on
today are the outcome of policies pursued by the State Govts and (to a lesser extent) by the
Central Government programs to nudge the States to improve the educa?on system. This
sec?on analyses the results of 75 years of adult educa?on in terms of literacy (sec 4.1) and
qualiﬁca?on of the popula?on (sec?on 4.2). The posi?on of India is also compared with other
countries with varying per capita GDP, to see where we stand and what improvements are
needed in our quest to become Viksit/Developed. The failure of most States to build the base
of the educa?on pyramid during the ﬁrst 30 years of independence, and the slow accelera?on
by many States during next 20 years, is reﬂected in the shor?alls we see in the educa?on
a?ainments of the adult popula?on today.
4.1 Adult Literacy
The adult literacy rate (ages 15 and above) increased from 69.8% in 2012 to 81.7% in 2023 an
improvement of 11.9 per cent points. The youth literacy rate (ages 15-24) also rose from
90.0% to 97.0% during the same period (+7.1), indica?ng near-universal literacy among
younger age groups (Table 3).
Table 3: Improvement in Literacy Rates (%)

Source: World Development Indicators, December 2025.
Figure 4 is the plot of literacy levels of every country in the World rela?ve to its Per Capita
GDP, for the year in which each country’s latest available literacy data. The do?ed regression
line can be seen as the expected level of literacy for each level of per capita GDP. When viewed
against this benchmark, India’s adult literacy rate is below the benchmark. The current
literacy rates are a legacy of 65 years of post-independence educa?on, inside and outside the
schooling system. As low literacy is oLen associated with low and stagnant wages, the
elimina?on of absolute poverty (1% in 2023-24), reduc?on in LMIC poverty rate (or vulnerable
popula?on) to 15% (2023-24), and reduc?on of mul?-dimensional poverty rate to 15% (2019-
21), is noteworthy.
17
It suggests that Government social welfare programs, such as subsidized

17
Bhalla and Bhasin (2025), EPW, July 13, 2024, Vol LIX, No 28; and 2025. Mul?-dimensional poverty from NITI.
Education IndicatorsChange
Literacy ratesRates Rates
Adult total (% of people ages 15+)202381.7201269.811.9
Youth total (% of people ages 15-24)202397.0201290.07.1
Latest available yearYear for comparison
15

food, cooking gas, toilets, housing, and electricity, water & telecom connec?ons, have played
an important role.
However, we need to move towards universal literacy to improve the produc?vity and wages
of the bo?om quarter of the adult popula?on. To achieve this, the literacy rate needs to be
raised by at least 10 per cent points to a?ain the 87% points level (UMIC, purple diamond) in
ﬁve years, and by 22% points to 99% by 2047 (pink diamond). This challenge can be met, given
that the youth literacy rate is already 99% (Figure 4).
Figure 4: Adult Literacy Rate (Age 15 and above)

Source: World Development Indicators, December 2024.
4.2 Adult educa?on levels
Adult educa?on levels reﬂect the enrolment and comple?on rates in primary, secondary and
ter?ary levels over 75 years, as well as short- and long-term courses oﬀered over the years.
The educa?onal a?ainment of the popula?on aged 25 years is available only for years in which
surveys are connected by diﬀerent countries. The per cent of popula?on (25+) with primary
educa?on increased by 11.4% points, since 2016 to 68.1% in 2023 (Table 4). Over the last 12
years from 2011 to 2023, the adult popula?on with lower secondary educa?on has increased
by 15.1% points to 52.7%, and upper secondary by 14.5% points to 41.3%. The propor?on of
popula?on (25+) with post-secondary educa?on, however increased by only 6.1% over the
same period, to 16.1% (2023).
In 2023, 3.1% of popula?on (25+) had Master’s degree or equivalent, 14.9% had Bachelor’s
degree or equivalent and 41.3% had passed secondary school (Table 4). The detailed
breakdown for ter?ary short cycle courses, Bachelor’s degrees and Masters degrees is not,
however, available for the same years, in the World Bank data set. However, the available data
suggests that increase in popula?on of those with Bachelor’s degrees or equivalent, is higher
India (77%)
Indonesia
Vietnam
China
Mexico
Thailand
UMIC (87%)
HIC (99%)
R² = 0.5221
50
60
70
80
90
100
0500010000150002000025000300003500040000
% of people ages =>15
PcGDP , PPP (const 2021 int $)
Adult Literacy Rate
16

than the average increase in ter?ary educa?on, while increase in per cent with short cycle
courses is about average and the acquisi?on of master’s degrees is much lower.
Table 4: Improvement in Educa?onal A?ainment Rate (%)

Source: World Development Indicators, July 2025.
We use the same interna?onal data to compare adult educa?on in India with other countries,
organized by per capita GDP of each country in the year for which latest survey data for adult
educa?on is available. This is plo?ed in Figure 5, with the cross-country regression (dashed)
line, deno?ng expected or average level of the index for each level of per capita GDP.
18
The
per cent of adults with primary educa?on in India is lower than the expected level for its per
capita GDP. This compares with Indonesia, Vietnam and China which are well above the
percentage of primary educated popula?on expected for their Per Capita GDP, and Mexico,
which is on benchmark line (Figure 5). The sharp increase in per cent of youth who have
completed primary educa?on in India, and the 99% comple?on rate for children, means that
the per cent of adult popula?on with primary educa?on will con?nue to rise.
Figure 5: Adults with Primary Educa?on (% of popula?on 25+)

Source: World Development Indicators, December 2024

18
Note: The interna?onal comparison is based on WDI December 2024, the data for India in Table 4, is from WDI July 2025, and is diﬀerent
for primary educa?on in 2023.
Education IndicatorsChange
Educational attainment, total (%) Population 25+Rates Rates
Primary 202368.1201656.711.4
Lower secondary 202352.7201137.615.1
Upper secondary 202341.3201126.914.5
Post-secondary 202316.1201110.06.1
Short-cycle tertiary20128.920056.42.5
Bachelor's or equivalent 202314.9201810.84.1
Master's or equivalent20233.120182.40.7
Latest available yearYear for comparison
India (66%)
VietNam
Indonesia
China
Thailand
Mexico
UMIC (77%)
HIC (93%)
R² = 0.4731
35
45
55
65
75
85
95
105
01000020000300004000050000600007000080000
% of 25 yrs+
PcGDP , PPP (const 2021 int $)
Adult Primary Education(≥25 yrs)
17

It will be a challenge to raise the share of the primary educated popula?on to 78.6% in next 5
years and to 93.0% in next 25 years, the minimum level expected for an Upper middle-income
country (UMIC) and a high-income country (HIC) respec?vely (Figure 5 &Table 5). This can
however play a key role in raising the produc?vity and wages of the vulnerable popula?on,
and reduce the need for Welfare payments.
Table 5: Benchmark adult educa?on rates by Per capita GDP

Data Source: World Development Indicators, December 2024
Note: Data includes projec?ons for 2030 (UMIC, $12,800) and 2050 (HIC, $37,000) based on target per capita GDP levels.
The gaps in secondary school achievement of the adult popula?on, are less than for primary
educa?on. For lower secondary, the gap between the actual ra?o in 2023 and the appropriate
level for India’s per capita GDP, was less than 5% in absolute value. Thailand (-)17.4%,
Indonesia (-)7.7% had even larger gaps than India, while Mexico (-)3.5% was less, and Vietnam
and China were close to the level expected at their PCGDP, while (Figure 6, leL panel).
Figure 6: Adult with Secondary Educa?on (% of popula?on ages 25+)

Source: World Development Indicators, December 2024
To raise the share of adults with lower secondary educa?on to Upper-Middle-Income Country
(UMIC) benchmark of 62.9% would require an increase of 11% points. To increase it to High-
Income (HIC) benchmark of 81.4% by 2050, it would require a more daun?ng increase of 29%
points.
The shor?all from benchmark is higher for upper than for lower secondary, but the GAP is
less nega?ve than all the peer compe?tors; India’s upper secondary a?ainment for adults
Year PCGDPPrimaryLower SecUpper SecBachelorMasters
2022$8,54562.849.831.612.13.2
2023$9,16065.652.233.713.23.5
Benchmarks/Targets (based on PcGdp regression)
2023$9,16074.057.141.113.04.2
2030$12,80078.662.946.815.65.4
2050$37,00093.081.465.023.99.2
India (52%)
VietNam
Indonesia
China
Thailand
Mexico
UMIC (63%)
HIC (81%)
R² = 0.5631
40
45
50
55
60
65
70
75
80
85
90
0500010000150002000025000300003500040000
% of 25 yrs+
PcGDP , PPP (const 2021 int $)
Lower Sec Education (≥25 yrs)
India (34%)
VietNam
Indonesia
China
Thailand
Mexico
UMIC (47%)
HIC (65%)
R² = 0.5505
20
25
30
35
40
45
50
55
60
65
70
0 10000 20000 30000 40000
% of 25 yrs+
PcGDP , PPP (const 2021 int $)
Upper Sec Education (≥25 yrs)
18

(33.7% in 2023) is below the level expected at its per capita GDP (Figure 6, right panel).
Vietnam (-)8.6%, Indonesia (-)9.1%, China (-)21.7%, Thailand (-)16.0% and Mexico (-)15.3%
are all below their expected levels. To reach the UMIC benchmark of 46.8% by 2030, India
needs to increase its rate by about 13% points, and to meet the HIC benchmark of 65% by
2050, the required rise is around 31% points. To raise the propor?on, of adults with secondary
educa?on to the minimum benchmarks for a High-income County is a very challenging, 25-
year, task.
As far as the ter?ary educa?onal a?ainment of the adult popula?on is concerned, India’s
performance is very good at the bachelor’s level and somewhat inferior at the Masters level
(Table 5), post-secondary and short-cycle educa?on are covered in Sec?on 10 (Skilling),
Bachelor's and Master's levels will be discussed in greater detail in in Sec?on 11 on Ter?ary
Educa?on, and Doctoral educa?on in Sec?on 13 on R&D
19
.
4.3 Youth Literacy
Youth literacy reﬂects the recent achievements in primary educa?on in the last decade.
These rates are compared with other countries in (Figure 7) Youth literacy of countries rela?ve
to the level expected at their level of Per capita GDP, based on cross-country regression. India’s
youth literacy rate in 2023 was 97.0%, which is 6.7% points above the level expected at its per
capita GDP. All comparator countries are also above their expected level, with Indonesia
+7.4% points above its benchmark, Vietnam +6.0% above its benchmark, China +4.1%,
Thailand +1.6% and Mexico +2.0%. India’s performance is therefore be?er than all its
comparators, rela?ve to the benchmark for each.
Figure 7: Literacy rate, youth total (% of people ages 15-24)

Source: World Development Indicators, December 2024

19
Doctoral comple?on rate for India has been revised down for 2023 from 3.5% to 0.1%.
India (97%)
Indonesia
VietNam
China
Thailand
Mexico
UMIC (93%)
HIC (100%)
R² = 0.4378
80
85
90
95
100
0500010000150002000025000300003500040000
(% of people ages 15-24)
PcGDP , PPP (const 2021 int $)
Youth Literacy Rate
19

Since the percentage for India exceeds the UMIC (2030) benchmark of 92.9%, sustaining
eﬀorts at their current level would be suﬃcient to achieve the 100% HIC benchmark, well
before the 2050 deadline. These results also provide a posi?ve signal for the future of adult
literacy, as the rise in youth literacy will result, ipso facto, in a gradual rise in the percent of
adults in popula?on.
As youth literacy largely reﬂects recent performance of the educa?on system, we can start by
looking Primary educa?on.
5. Primary Educa?on
Conven?onal iden?ﬁca?on of educa?on with general human development and human
rights, is encapsulated by the following statement: “Educa?on is a human right, a powerful
driver of development, and one of the strongest instruments for reducing poverty, improving
health, promo?ng gender equality, peace, resilience, and adapta?on to climate change. It
delivers large, consistent returns in terms of income, and is a cri?cal factor to ensure equity
and inclusion.”
20
Much less a?en?on has been paid in the post war era, to the role of
educa?on in crea?ng Human Capital as an important factor of produc?on, and driver of
economic produc?vity and real wage growth. Accoun?ng for and focusing on the direct
linkage between educa?on and jobs is a rela?vely recent phenomenon, and s?ll paid less
a?en?on in many educa?on departments around the World and by academics & educators
within India. NEP (2020) is a policy eﬀort the change this situa?on.
India, with its vast and diverse popula?on, operates one of the largest educa?on systems in
the world. Spanning from early childhood educa?on to higher educa?on, it caters to over 24.8
crore students across 14.7 lakh schools, supported by 98 lakh teachers
21
. While this scale
represents a signiﬁcant achievement in terms of access, less a?en?on has been paid to quality
of educa?on and learning outcomes. The diversity across States and UTs, therefore, also
means greater variance in learning outcomes This sec?on, focuses on learning outcomes in
terms of cross-country comparisons, and evolu?on of outcomes over ?me in India.
5.1 Enrollment & Comple?on
Primary enrolment and comple?on rates in India are now fairly sa?sfactory. As in other
countries, the GER is much larger than the NER. The setback during the pandemic is being
reversed. Data on net enrolment is not available for India, aLer 2013. So, we can only
determine the change in gross enrolment since then. The Primary enrolment rate (gross)
increased from 112.4% in 2014 to 120.5% in 2024 (+8.1), including over-age and under-age
students enrolled in primary grades (Table 6). The share of students enrolled in private primary

20
World Bank, Educa?on Factsheet, November 2024.
21
UDISE+ 2023-2024. UDISE data is only available from 2018-19.
20

schools also grew from 35.0% in 2014 to 43.8% in 2019 (+8.8), sugges?ng a gradual shiL
toward private schooling within the primary educa?on system (Table 6).
Table 6: Improvement in Primary enrollment and comple?on rates (%)

Source: World Development Indicators, 2025.

In contrast, the primary comple?on rate declined slightly from 99.6% in 2014 to 94.5% in
2024. There are however, many data gaps in this record, with available data showing high
vola?lity. Two waves of pandemic, in 2020 and 2021, and the lockdown for some months
during the ﬁrst wave, contributed to the data gaps and the decline in comple?on rates.
However, the average comple?on rate, increased marginally from an average of 97.4% during
2005-2014, to an average of 97.8% during 2015-2024.
The latest available year for Net Enrolment in primary schools for India is 2013. We can
compare its performance with other countries. India’s primary net enrolment rate was 92.3%
in 2013 (Figure 8), which is 6.8 points above the expected level at its per capita GDP. Vietnam
(+10.9), Indonesia (+2.6) and Mexico (+6.3) are also above their expected levels in 2013 (latest
available year). The base for India in Figure 8, is 2013; therefore, we es?mate the value for
NER in 2023 assuming that the gap remains 6.9% points. This gap is then added to the
expected benchmark for India at its 2023 per capita GDP. This es?mate (94.4%) is above the
HIC benchmark of 93.9%. We can therefore set a higher target of 99% to 100% to be at level
of the most advanced countries by 2047.
Figure 8: Primary Enrollment & Comple?on Rate

Source: World Development Indicators, April 2025; Note: 1. The year 2013 was chosen as the base year for primary
enrollment, as it is the most recent year for which data is available in India.
2. Data points with primary comple?on rates over 100% have been excluded for accuracy.
Education IndicatorsChange
Primary Enrolment rate (% gross) 2024120.52014112.48.1
Primary enrolment in private schl (%)201943.8201435.08.8
Primary completion rate(relevant age grp)202494.5201499.6-5.1
Latest available yearYear for comparison
India
(92%)
Vietnam
Indonesia
Mexico
UMIC (89%)
HIC (94%)
R² = 0.2331
80
82
84
86
88
90
92
94
96
98
100
020000400006000080000100000
Primary (%net)
PcGDP , PPP (const 2021 int $)
Primary School Enrollment
India (93.5%)
UMIC (84%)
HIC (92%)
R² = 0.4144
60
65
70
75
80
85
90
95
100
01000020000300004000050000600007000080000
(% of relevant age group)
PcGDP , PPP (const 2021 int $)
Primary Completion Rate
21

India’s primary comple?on rate was 93.5% in 2023 (Figure 8, 2
nd
panel), which is 11.9% points
above the level expected at its per capita GDP. China (+15.2), Vietnam (+13.5), Indonesia
(+13.0), Thailand (+9.9) and Mexico (+9.7) are also above their expected levels. India is already
above the UMIC (2030) benchmark of 84.1% and the HIC (2050) benchmark of 92.3%. The
greatest focus on primary schooling since 1999 has produced results in terms of improvement
of formal creden?als. Focus has now to shiL to learning outcomes in line with NEP (2020).
5.1.1 Pre-primary
NEP 2021 has laid emphasis on Pre-primary educa?on to create the founda?on for further
learning by children. This is par?cularly important for under-privileged infants and children
whose family and home environment may not be as s?mula?ng as that of a normal middle-
class family. Pre-primary educa?on enrollment in India has shown considerable varia?on over
recent years.
22

The Gross Enrollment Ra?o (GER) was 60.5% in 2017, peaked at 62.8% (2019), and declined
to 51.7% by 2021 likely due to disrup?ons caused by the COVID-19 pandemic. Given this
ﬂuctua?on, 2019 has been used as reference year for interna?onal comparison, represen?ng
India’s highest pre-pandemic enrollment level. In total, 147 countries reported pre-primary
enrollment data in 2019, while 8 addi?onal countries had submi?ed data in 2018. These were
included in the comparison using their corresponding per capita GDP for the reported year.
India’s pre-primary GER was 62.8% in 2019 (Figure 9), which is 8.3 points above the level
expected at its per capita GDP.
Figure 9: School enrollment, Pre-Primary (GER)

Source: World Development Indicators, April 2025. Note: The year 2019 was selected as the baseline since it represents the
period before the pandemic. For 8 countries without data for 2019, data from 2018 was used instead.

22
Gross enrollment ra?o is the ra?o of total enrollment, regardless of age, to the popula?on of the age group that oﬃcially corresponds to
the level of educa?on shown. Preprimary educa?on refers to programs at the ini?al stage of organized instruc?on, designed primarily to
introduce very young children to a school-type environment and to provide a bridge between home and school.
India (63%)
Vietnam
Indonesia
China
Thailand
Mexico
UMIC (63%)
HIC (83%)
R² = 0.4228
50
55
60
65
70
75
80
85
90
95
100
0 20000400006000080000100000
Pre-Prim (% gross)
PcGDP , PPP (const 2021 int $)
Pre-Primary School Enrollment
22

Vietnam (+34.3), China (+19.0) and Thailand (+5.8) are also at a pre-primary enrolment rate,
which is higher than the benchmark for their per capita GDP. Indonesia (-)1.0 and Mexico (-)
2.4 were below their benchmark or minimum expected levels for their PCGDP in the year for
which the data was available.
The data for India in Figure 9 is for 2019, therefore we es?mate the pre-primary enrolment
rate for 2023 assuming that the gap remains the same (8.3% points). This gap is then added
to the benchmark/min expected (57.2%) for India at its 2023 per capita GDP. Based on this
es?mate of 65.5%, we calculate that we are already above the UMIC benchmark of 63.4%,
and need an increase of 17.5% points to reach the benchmark of 83% for HIC.
WDR (2019) review of the literature highlighted the following: A study of preschools in a
Nairobi slum found that while par?cipa?on was high, the curriculum and pedagogy were not
age-appropriate, limi?ng learning, as children aged 3 to 6 were exposed to academic-oriented
instruc?on and even sat for exams. Similarly, in Peru, the Wawa Wasi program oﬀered safe,
community-based daycare and nutri?ous meals for children aged 4 to 6 in impoverished areas,
but it failed to improve children’s language skills. In Indonesia, playgroup programs showed
posi?ve eﬀects on children’s language and other development. Similarly, in Tonga, organizing
playgroups for children up to age 5 signiﬁcantly improved early-grade reading skills. Among
structured approaches, the Montessori model with its mul?-age classrooms, student-chosen
learning ac?vi?es, and minimal instruc?on has been more eﬀec?ve than tradi?onal models in
improving children’s execu?ve func?ons.
Good models for suppor?ng literacy and numeracy by Grade 3 are available, and they are both
cost-eﬀec?ve and scalable, even where resources are limited. In Liberia and Malawi, training
teachers to be?er evaluate their students, combined with addi?onal materials signiﬁcantly
improved learning in early grades. In Singapore, all students are screened at the onset of grade
1. Children who do not a?ain the appropriate early literacy skills are supported through the
Learning Support Program. These are straigh?orward approaches: they train teachers to
assess their students through ongoing, simple measurements of their abili?es to read, write,
comprehend, and do basic arithme?c. Children who need addi?onal support receive targeted
materials and undertake targeted ac?vi?es. These models have been tested with success in
Ghana, India, Jordan, and Kenya, and serve as a basis for precise design and budget es?mates.
5.2 Learning Outcome Surveys
Formal comple?on of primary educa?on does not guarantee that all children have a basic
knowledge of the curriculum prescribed for primary school. Worse, all those having a
cer?ﬁcate of primary school comple?on, may not meet the minimum proﬁciency levels (MPL)
including reading and learning, as per interna?onal or domes?c standards. These minimum
levels are cri?cal, both for low skill jobs in a modern economy, and for further study above the
primary level.
23

The World Development Indicators (WDI), report the share of pupils who are below the
Minimum Proﬁciency Level (MPL) in reading, at the end of primary school. MPL is a benchmark
developed by the Global Alliance to Monitor Learning (GAML). MPL is deﬁned as the ability
of a child to independently and ﬂuently read simple, age-appropriate texts, iden?fy explicitly
stated informa?on, interpret key ideas, and provide basic explana?ons or personal opinions
about the content. In other words, a pupil reaching MPL is able to read and understand a short
passage of narra?ve or factual text suitable for their grade level. This benchmark is used
interna?onally to track progress toward SDG indicator 4.1.1b, which monitors the propor?on
of children achieving minimum learning proﬁciency in reading and mathema?cs at the end of
primary educa?on.
In 2002, NCERT carried out a na?onal-level learning assessment of Class 5 students. The
results were published in 2006 and marked the ﬁrst oﬃcial a?empt to collect such large-scale
data on learning outcomes. Around 90,000 students were tested in language, mathema?cs,
and environmental science. According to Kingdon (2007), the na?onal average scores from
this NCERT test were low: students scored about 46.5% in math, 50.3% in science, and 58.6%
in language. These results broadly conﬁrmed the ASER (2006) survey’s ﬁnding that learning
levels in India were low, among children in Grade 5.
We use the latest available surveys (ASER, NAS, FLS, PRS) to create a picture of learning
outcomes in India. This requires an understanding of the design of these surveys, so that we
can compare results and decide on the u?lity of outcomes which appear to be inconsistent
between surveys.
5.2.1 Survey Design and Comparability
Since 2011, India has mul?ple large-scale assessments, each with diﬀerent objec?ves and
methods (Table 7). The Annual Status of Educa?on Report (ASER), led by ASER center,
Pratham, since 2005, is a household survey in rural India. It uses a two-stage sampling design
with Census 2011 as the frame; ﬁrst selec?ng 30 villages per district using Probability
Propor?onal to Size (PPS), then randomly sampling 20 households from each village.
Table 7: Major learning assessment surveys in India

Note: Colored survey has been used for comparison in this study.
The Na?onal Achievement Survey (NAS) was conducted periodically from 2001 to 2021. It
has now been updated and replaced by, what is referred to as the PARAKH Rashtriya
Sarvekshan (PRS), ﬁrst conducted in 2024 by NCERT. PRS also uses a two-stage PPS sampling
YearClass/Std 3Class/Std 5Class/Std 6Class/Std 8Class/Std 9Class/Std 10
2016ASERASERASERASER
2017NASNAS NAS NAS
2018ASERASERASERASER
2021NASNAS NAS NAS
2022FLS;AserASERASERASER
2024PRS;AserASERPRS;AserASERPRS
24

design based on the UDISE+ 2022-23 database, star?ng with schools, then sec?ons, and ﬁnally
students; up to 30 students per grade are surveyed. The Founda?onal Learning Study (FLS),
carried out in 2022, followed a mul?stage PPS sampling design, selec?ng schools by
management type and alloca?ng samples based on enrollment, with students drawn from
Grade 3 or newly admi?ed Grade 4.
The focus and mode of assessment also vary across these surveys. ASER is a one-on-one
household survey assessing rural children based on basic reading and arithme?c abili?es and
tes?ng ?me is about 7-8 minutes per child in one of the 19 Indian languages. PRS, building on
the earlier NAS, is a wri?en school-based assessment designed to evaluate competency-based
learning outcomes, covering founda?onal stage competencies in Grade 3, language,
mathema?cs, and environment in Grade 6, and language, mathema?cs, science, and social
science in Grade 9. It is conducted as a pen and paper test with OMR for data capture, las?ng
90 minutes for Grade 3 and 6 and 120 minutes for Grade 9, in 23 languages. The FLS (Grade
3) was a one-on-one oral and performance-based school test focusing on founda?onal literacy
(oral comprehension, decoding, ﬂuency, reading comprehension) and numeracy (number
opera?ons, frac?ons, pa?erns, data handling), with tes?ng ?me up to 35 minutes per student
per subject in one or more of 20 languages.
23

The surveys diﬀer in their comparability and accessibility of data. ASER has used standardized
and consistent tools since its incep?on, making its results comparable over ?me, but it is
limited to rural areas.
24
NAS 2021 is comparable to NAS 2017 in principle, but as the former
was taken in the middle of the Covid epidemic, its results are distorted by this once in a
century pandemic, and is therefore not useful for determining Medium-Long term trends. PRS
2024 was administered for the ﬁrst ?me using a new and more comprehensive strategy, and
its results are not comparable with earlier NAS rounds (Kar et al (2024). The FLS was
conducted once to establish benchmarks for founda?onal literacy and numeracy and does not
have cycles for comparisons over ?me; its data and reports are available, but the tools are not
in the public domain (Kar et al (2024)).
Two decades ago, Kingdon (2007) concluded that, “Inter State comparisons in India, are
hampered by the absence of na?onal standardized tests for secondary school. Each State
Board set its own syllabus and exam papers, so comparing pass rates between states was not
useful. For instance, in 2004, the high school pass rate was 37% in Manipur and 80% in Andhra
Pradesh, but because each state uses diﬀerent exams, these numbers can't be directly
compared”
More recently, Johnson and Parrado (2021) examined how learning outcomes are measured
in India and found that while numerous assessments and various state-level surveys have
emerged, there is limited coordina?on between them. The study highlights that most
assessments lack consistency in frameworks, subjects, and repor?ng formats, making

23
In 36 States and UTs, this meant one of 19 regional languages plus English.
24
Data from oﬃcial surveys (NAS & PRS) shows that there is li?le Urban-Rural diﬀerence in MRP and MAP.
25

comparisons over ?me or across systems diﬃcult. The authors emphasized the need for a
coherent na?onal learning assessment strategy that links classroom prac?ce with policy. They
also noted that while ASER has been regularly conducted since 2005, other assessments like
NAS have faced implementa?on challenges. Overall, the paper argues that improving learning
in India requires not just be?er teaching but also be?er measurement of learning outcomes
through coordinated and reliable assessment systems.
5.3 Interna?onal Comparison of Outcomes (MRP)
The World Development Indicators (WDI) deﬁne "Minimum Reading Proﬁciency (MRP)" as the
share of pupils at the end of primary schooling who cannot independently and ﬂuently read
simple, short narra?ves or expository texts, locate explicitly stated informa?on, or interpret
key ideas. This measure reﬂects learning depriva?on among children who complete primary
educa?on but lack fundamental reading skills.
25
In this sec?on we present the results of our
cross-country compara?ve analysis using this data. This allows us to judge the performance of
Indian Primary educa?on rela?ve to other countries at diﬀerent level of Per Capita GDP.
5.3.1 Methodology for calcula?ng MRP
WDI uses the 2017 Na?onal Achievement Survey (NAS) for Grade 3 and 5 in English language
and Mathema?cs for calcula?ng the Minimum Proﬁciency Level (MPL) in India. A policy
linking exercise carried out by UNESCO Ins?tute for Sta?s?cs and NCERT in 2019 mapped NAS
performance levels to the global MPL benchmark. The India Policy Linking Pilot Workshop
(Nov 2019) was part of a global eﬀort to establish a method for repor?ng on SDG Indicator
4.1.1. Led by UIS with support from MSI, DFID, and Gates, and hosted by NCERT and MHRD in
New Delhi, the pilot aimed to link India’s 2017 Na?onal Achievement Survey (NAS) for Grade
3 and 5 in English and mathema?cs to global benchmarks.
The policy linking method, developed through the Global Alliance to Monitor Learning
(GAML), is based on the Global Proﬁciency Framework (GPF). The GPF provides four Global
Proﬁciency Levels (does not meet, par?ally meets, meets, exceeds minimum proﬁciency), with
descriptors by grade and subject. Before benchmarking, a content alignment exercise was
carried out to check how well NAS items matched the GPF. Two four-day workshops were held
(Nov 12-15 for Grade 3, Nov 18-21 for Grade 5). The NAS assessments used had 25 mul?ple-
choice items, equated using Item response theory, and were administered in representa?ve
schools in 2017. Four groups of 18 panelists (72 in total mostly teachers and head teachers,
plus experts) par?cipated.
The process followed the Policy Linking Toolkit, involving three main tasks: Firstly, Checking
alignment of NAS assessments with the GPF. Each item was rated on a three-point scale:
Complete Fit (C) signiﬁes that all of the content required to answer the item correctly is

25
WDI reports the percentage of pupils below minimum reading proﬁciency (MRP). To present learning outcomes, we subtract this from
100% to get percentage of pupils above MRP.

26

contained in the subconstruct, i.e., if the learner answers the item correctly, it is because they
completely use knowledge of the subconstruct; Par?al Fit (P) signiﬁes that part of the content
required to answer the item correctly is contained in the subconstruct, i.e., if the learner
answers the item correctly, it is because they par?ally use knowledge of the subconstruct; No
Fit (N) signiﬁes that no amount of the content required to answer the item correctly is
contained in the subconstruct, i.e., if the learner answers the item correctly, it is because they
do not use knowledge of the subconstruct.
Secondly, matching NAS items with global proﬁciency levels and descriptors and ﬁnally
applying the Yes-No Angoﬀ method, where panelists judged how a minimally proﬁcient
learner would perform on each item. Two rounds of ra?ngs were conducted. In Round 1,
Class3 English showed the highest percentage at meets or exceeds (60.6%), while Class 3
mathema?cs was lowest (36.5%). Round 2 conﬁrmed consistent results, with Class 5 English
the lowest (35.6%). Reliability measures (Standard errors of measurement (SEM) < 1.0, inter-
and intra-rater consistency > 0.70) indicated strong consistency, par?cularly for Class 5
[UNESCO (2019)].
5.3.2 Minimum Reading Proﬁciency (MRP)
India’s share of pupils achieving minimum reading proﬁciency was 46.3% in 2017, which was
5.3% points above the benchmark/minimum expected level for its per capita GDP in 2017
(Figure 10, leL panel). Vietnam, China, Thailand, were well above their benchmarks, Indonesia
was marginally above its bench mark by 0.3% points, while Mexico was below its expected
level by (-)14.1% points (Figure 10, leL panel).
Figure 10: Minimum reading proﬁciency at end of primary (% of students)

Source: World Development Indicators, December 2024
As the data for India in Figure 10 is for 2017, we es?mate a MRP for 2023 (51.6%) assuming
that the gap between benchmark and actual remains the same as in 2017 (i.e., 5.3% points).
This gap is then added to the expected expected/benchmark for India at its 2023 per capita
GDP to obtain 46.3% in 2023. Based on this we es?mate the improvement needed to reach
India (46%)
Indonesia
VietNam
China
Thailand
Mexico
UMIC (54%)
HIC (79%)
R² = 0.7074
30
40
50
60
70
80
90
100
01000020000300004000050000600007000080000
End of Primary (%)
PcGDP , PPP (const 2021 int $)
Minimum Reading Proficiency
India (44%)
Indonesia
VietNam
China
Thailand
Mexico
UMIC (51%)
HIC (76%)
R² = 0.7214
0
10
20
30
40
50
60
70
80
90
100
0 20000 40000 60000 80000
Childrn out of school (%)
PcGDP , PPP (const 2021 int $)
Above minimum reading proficiency-Adj (%)
27

UMIC & HIC benchmarks. This shows that we need to improve Minimum reading proﬁciency
by ~ 2.6% points to reach the UMIC benchmark of 54.2% and by ~27.6% points to reach the
HIC benchmark of 79.2% for HIC. This is a challenge which can be met by each State focusing
on Founda?onal Literacy and Numeracy (FLN) and Minimum Learning Proﬁciency in Reading
(MRP) and Arithme?c (MAP).
The above data is for children who are in school. There is also a small share of children who
are not in school. India’s overall end-of-primary reading proﬁciency (adjusted for out-of-school
children) was 43.9% in 2017 (Figure 10, 2
nd
panel), which was 6.2% points above the
benchmark/min expected level at its 2017 per capita GDP. Vietnam (33%), China (+26% pts),
Thailand (+19% pts) were above their benchmark, Indonesia at its min expected level, while
Mexico was below (-)11.7% pts its benchmark.
The World Development Indicators (WDI), reported India’s minimum reading proﬁciency at
the end of primary using NAS 2017 (Sec?on 5.3.1). However, no such eﬀort has been made by
interna?onal agencies to use subsequent Indian surveys to es?mate changes in Minimum
Learning Proﬁciency as per interna?onal benchmarks. We examine large scale Indian surveys
to extract some results which are consistent with interna?onal benchmarks. These surveys
include the Annual Status of Educa?on Report (ASER), the Na?onal Achievement Survey
(NAS), Founda?onal Learning Study (FLS) and PARAKH Rashtriya Sarvekshan (PRS).
5.4 Minimum Learning Outcomes (ASER)
The Annual Status of Educa?on Report (ASER), is based on a survey of households in rural
India, carried out by ASER Center, an autonomous research unit of the PRATHAM Research
Founda?on (NGO/NPO). Its aim is to assess children aged 5-16 years, in basic reading and
arithme?c skills, rather than by school standard. This is the only survey which is available, from
the mid-2010s to the current ?me, using a broadly consistent methodology and sampling
procedure. Given the small diﬀerences in rural-urban performance, indicated by oﬃcial
surveys (sec?on below), it is a reasonable measure of the minimum learning proﬁciency of
primary school age children, at the na?onal level.
26
We use this to assess the posi?on of
learning outcome in language and arithme?c as of 2024, and the trends over past two
decades. As the data covers all Std from I to VIII, we focus on founda?onal learning (Std III)
and primary level (Std V).
ASER’s rural household-based survey is the most reliable domes?c benchmark because its
tests are clearly deﬁned and consistent over ?me. In language, ASER classiﬁes children by
whether they can read le?ers, Words, Std I level text, or Std II level text, while in mathema?cs
it assesses number recogni?on (1-9 and 10-99), subtrac?on, and division. The percentage of
students who are at Std II level text reading level are assumed to also meet the Std I level

26
NAS 2021: Diﬀerence between Total and Rural, is (1) For MRP: in class 3 is-2.5%, in Class 5 is +2.4%, and class 8 is +5.3%. (2) For MAP: in
class 3 is -4.5%, in class 5 is -3.8% and class 8 is +3.8%; of the Rural percentages from NAS 2021.
28

reading level. Similarly, the students who can do division are assumed to also be able to do
subtrac?on. In this paper we use Standard (Std), Class and Grade, interchangeably.
5.4.1 Minimum Learning Proﬁciency: India 2024
As of 2024, across founda?onal learning, 68.2% of Std I age children can read le?ers, 52.2%
of Std II age children can read/understand words and 47% of Std III age children can
read/understand Std I level text. The ﬂip side of this coin, is that, 30.8% of Std III age children
cannot read words, and 15.1% of Std II age children cannot read le?ers. The starkest indicator
of the creden?alism is that 8.2% of Std III age children, cannot even read le?ers (Table 8).
In arithme?c, 73.7% of Std I age children can recognize single digit numbers (1-9), 55.9% of
Std II age children can recognize two-digit number (10-99), while only 33.7% of Std III age
children can subtract. The ﬁrst two outcome are superior to that in reading but the last is
weaker by 10% points. This suggest that teachers are unable to adequately link the concept
of subtrac?on to real world/physical examples which children can relate to (Table 8).
Table 8: Founda?onal Literacy & Numeracy skills (% of students)

Source: ASER Report, 2024, Note; The ﬁgures represent cumula?ve learning levels.
In Std IV and V, it’s more appropriate to set the minimum language proﬁciency in terms of
reading standard II level text and the minimum learning in arithme?c, to ability to do division.
Reading levels are consistently higher than arithme?c skills across primary grades. By Std V,
nearly 48.7% Std II level text (Table 9). In arithme?c, while basic subtrac?on improves by 22%
points by Std V, more complex opera?ons like division remain weak rela?ve to reading ability
(Std 2 level text). Std V children who can who can read do division are 37% less than those
who can do division. This gap compares with the minimum learning gap of 28% in Std III. The
widening of the gap, suggests that schools need to pay more a?en?on to the pedagogy of
teaching founda?onal numeracy & arithme?c.
Standard => Std I Std IIStd IIIStd IStd IIStd III
Reading (% of children)
Letter68.2 31.815.18.2
Word29.452.2 47.830.8
Std 1 level text 12.229.047.0 53.0
Arithmetic (% of children)
Recognise no. 1-9 73.7 26.310.65.5
Recognise no. 10-9933.855.9 44.129.2
Subtraction 7.218.733.7 66.3
Can read Can't read
Can do Can't do
29

Table 9: Learning outcomes in Primary (% of students)

Source: ASER Report, 2024, Note; The ﬁgures represent cumula?ve learning levels.
Muralidharan and Singh (2021) review shows, that NEP reﬂects many insights from research,
such as focusing on early learning (FLN), improving teacher mo?va?on, and reorganizing small
schools. They suggested three principles for be?er implementa?on; focus use of independent
measurement, rigorous evalua?on, and cost-eﬀec?ve strategies.
5.4.2 Minimum Learning Proﬁciency: 2006 to 2024
ASER provides consistent data on children’s learning since 2006, making it possible to track
trends from 2006 to date. As Std I, III, V are landmark standard in the development of child,
this sector depicts inter-temporal changes in minimum learning proﬁciency. From 2006 to
2014, there was a decline in reading levels across primary; the share of Std I students who
could recognize le?ers, declined by (-)10% points, the share of Std III children who could
recognize words declined by (-)17% points, and the propor?on of Std V children who can read
Std I text declined by (-)13.9% points. The only excep?ons to this broad decline were small
gains in Std I level text reading among Std I children, and Std II level text reading among Std III
children (Table 10).
Table 10: Minimum Reading Proﬁciency, MRP (% of children): 2006 to 2024

Source: ASER reports 2024, 2014 and 2006
Between 2014 and 2024, recovery in Std III and Std V happened across the board, while in Std
I it is more nuanced. In Std I, le?er recogni?on improved strongly (+17%) aLer the earlier fall.
In Std III, there was an improvement in word recogni?on (+9%), reading of Std I level text
(6.8%) and reading of std II level text (3.4%). Similarly, there was an improvement in Std V, but
less than in Std III (last line, Table 10)
Standard => Std IV Std V Std IVStd V
Reading (% of children )
Word80.084.220.015.8
Std 1 level text 62.070.038.030.0
Std 2 level text 40.048.760.051.3
Arithmetic (% of children)
Recognise no. 11-99 81.485.918.614.1
Subtraction 47.355.852.744.2
Division21.530.778.569.3
Can readCan't read
Can do Can't do
MRP(%)
LetterWordStdItextStdIITextLetterWordStdItextStdIIText
Std I68.229.412.25.351.321.19.04.5
Std III91.869.247.027.085.260.240.223.6
Std V95.984.270.048.794.381.567.248.1
Change
Std I16.98.33.20.8 -10.4-2.3 2.41.9
Std III6.69.06.83.4-8.5-17.0-7.7 3.7
Std V1.62.72.80.6-3.6-11.5-13.9-4.9
Can readCan read
2014 to 2024 2006 to 2014
20242014
30

Minimum Arithme?c Proﬁciency (MAP) shows a similar trend of decline between 2006 and
2014, followed by recovery from 2014 to 2024. From 2006-2014, the propor?on of children in
all three standards with ability to subtract and divide fell across the board. The decline was
somewhat higher for Std III students (~50%) than for Std V (Table 11). The decline in Std V was
higher in terms of percent points than for Std III, but it was higher in percent terms for Std III
than for Std V, subtrac?on dropped by (-) 29% points (-36%) while division fell by (-)19% points
(-42%). Std III saw a decline of (-)23% points (-48%) and division a decline of (-)8% points (-
51%).
Table 11: Minimum Arithme?c proﬁciency, MAP (% of children): 2006 to 2024

Source: ASER reports 2024, 2014 and 2006. Note; Recogni?on of no. by single digit (1-9) and double digit (11-99). is NA for
year 2006, so we cannot determine change between 2006 and 2014.
From 2014-2024, there is clear improvement. In Std III, gains were visible across all
indicators, with children who can subtract increasing by 8.3% points (+33%) and those who
can divide improving by +4% points (54%). Improvements are smaller for Std V, with those
who can subtract improving by 5.2% points (10%) and those who can divide increasing by 4.6%
points (18%) of children.
5.4.2.1 Eﬀect of Right to Educa?on Act (RTE,2009)
The Right of Children to Free and Compulsory Educa?on Act or Right to Educa?on Act (RTE,
2009) was enacted by the Parliament of India in 2009 and came into eﬀect on April 1, 2010
establishing free and compulsory educa?on as a fundamental right for children aged 6 to 14
and se?ng minimum standards for elementary schools. As the cons?tu?on already enjoined
Government to provide educa?on to every child in India, the important change was in terms
of the wages and beneﬁts of teachers and increased administra?ve procedures to be followed
by Govt, private aided and private (NGO/NPO) schools. The goal of the Act was to improve the
quality of teaching and access to educa?on. The results belied these expecta?ons.
Table 12 shows reading outcomes before and aLer the Right to Educa?on Act (RTE, 2009), i.e.,
for 2006, 2010, and 2018. Before RTE (2006-2010), reading levels showed small
improvements in early grades as le?er recogni?on (+4.2% points, Word recogni?on (+1.4%
points) and reading ability of Std I level text (+1.2% points) increased. In Std III and V, there
were decline in per cent of children reading words (-)1.9% points and (-)1.8% points,
MAP(%)
1-911-99SubtractDivide1-911-99SubtractDivide
Std I73.733.87.22.057.723.84.51.1
Std III94.570.833.711.490.160.725.47.4
Std V97.785.955.830.796.180.750.626.1
Change
Std I16.010.02.70.9-3.3-1.0
Std III4.410.18.34.0-23.1-7.8
Std V1.65.25.24.6-28.7-19.2
2014 to 2024 2006 to 2014
Recognize nos Recognize nos
20242014
CanCan
31

respec?vely and Std I level text (-)2.2% points and (-)2.6% points, respec?vely. ALer RTE came
into eﬀect, the 2010-2018 period saw either a decline instead of improvement or faster
decline in reading ability across all standard, with two excep?ons. Minimum reading ability of
Std I children (Std I level text +3.2% points) improved and so did the minimum reading ability
of Std V children (Std II level text, +7.2% points).
27

Table 12: Eﬀect of RTE on Minimum Reading Proﬁciency (% of children)

Source: ASER reports 2018, 2010 & 2006
Similarly, arithme?c skills in Table 13 were already weakening as Std III dropped in subtrac?on
(-)12%, and Std V fell sharply in both subtrac?on and division (-)9%. ALer the RTE, the decline
con?nued. Std, I decline in number recogni?on while Std III saw declines in subtrac?on. Std V
shows the steepest fall, especially in subtrac?on (-)18% and division (-)8% except for Std 1
number recogni?on (+3.5%) and Subtrac?on (+0.4%).
Table 13: Eﬀect of RTE on Minimum Arithme?c Proﬁciency (% of children)

Source: ASER reports 2018, 2010 & 2006
The goal of the Right to educa?on Act was to improve the access to quality educa?on. The
learning outcomes were if anything contrary to these expecta?ons. RTE (2009) seems to have
had a broadly nega?ve impact on learning outcomes, aﬀec?ng both minimum reading
proﬁciency and minimum proﬁciency in arithme?c. While the pre-RTE period (2006-2010)
showed declines with stability in a few grades/standards, the post-RTE years (2010-2018)
showed greater declines in most grades.

27
This was partly due to a reduc?on of students, due to the closing of NGO/NPO schools.
MRP(%)
StandardLetterWordStdItextStdIITextLetterWordStdItextStdIIText
Std I 57.324.711.05.865.924.87.83.4
Std III87.965.344.527.294.175.345.720.0
Std V 94.182.469.450.397.991.278.553.4
Change
Std I -8.6-0.1 3.22.44.21.41.20.8
Std III-6.2-10.0-1.2 7.20.4 -1.9-2.2 0.1
Std V -3.8-8.8-9.1-3.10.0-1.8-2.6 0.4
Pre- RTE Act: 2006-2010
Per cent of cildren who can readPer cent of cildren who can read
Post-RTE Act: 2010-2018
In Year 2018In year 2010
MAP(%)
1-911-99SubtractDivide1-911-99SubtractDivide
Std I 64.327.25.92.065.823.75.52.1
Std III92.565.628.18.594.373.336.49.4
Std V 96.682.852.327.897.990.170.335.9
Change
Std I -1.53.50.4 -0.1 -2.30.0
Std III-1.8-7.7-8.3-0.9 -12.1-5.8
Std V -1.3-7.3-18.0-8.1 -9.0-9.4
Post-RTE Act: 2010-2018 Pre- RTE Act: 2006-2010
Recognize nos Recognize nos
In Year 2018In year 2010
Children who canChildren who can
32

5.4.2.2 Eﬀect of Covid 19, Pandemic
Another event which aﬀected school educa?on was the COVID 2019 pandemic. No ASER
survey was done during the pandemic. Data is therefore only available for 2018, 2022 and
2024. We therefore compare trends during 2022-2024 with the trends during 2018-2021.
During the pandemic years 2020-2021, reading outcomes declined across all standard and
all reading metrics. Among Std I children the share of children recognizing le?er and words
declined by (-)1.2% points and (-)4% points respec?vely. Share of Std III children who could
read Std I level text fell by (-)8.9% and those who could read Std II level text declined by
(-)6.7% points. The corresponding deteriora?on in Std V was (-)6.7% points for Std I level text
and (-)7.5% for Std II level text (Table 14).
Table 14: Eﬀect of Pandemic on Minimum Reading Proﬁciency (% of children)

Source: ASER reports 2018, 2010 & 2006
In the post-pandemic period, there was strong, across the board, recovery in minimum
reading proﬁciency. The recovery during 2022 to 2024 was greater than the decline during
2018-2022 in all but a few cases. The main excep?on was the share of Std V children who
could read Std II level text-these increased by only 5.9% points, aLer declining by (-)7.5%
points. The biggest per cent jump (40%) was in Std I, with children who could read words
increasing by 8.6% points and those who could read Std I level text improving by 3.4% points.
The second highest per cent jump (30%) was in Std III, with reading of Std I level text improving
by 11.4% points and of Std II level text improving by 6.5% points (Table 14).
Similarly, arithme?c skills declined across all primary standard and metrics from 2018-2022
In Std I recogni?on of both single- and two-digit numbers fell by 1-2%, Std III-V saw decreases
across subtrac?on, and division (Table 15). Post-pandemic (2022-2024), there was a clear
recovery. Students in Std I recognizing numbers (1-9 & 10-99) improved (by +11.3% & 8.2%
points). Children in Std III and Std V able to subtract increased by 7.8% &5.9% points
respec?vely, while those able to do division increased by +3.1% & 5.1% points respec?vely.
MRP(%)
LetterWordStdItextStdIITextLetterWordStdItextStdIIText
Std I68.229.412.25.356.120.88.84.5
Std III91.869.247.027.085.658.035.620.5
Std V95.984.270.048.794.079.162.742.8
Change
Std I12.18.63.40.8-1.2-3.9-2.2-1.3
Std III6.211.211.46.5 -2.3-7.3-8.9-6.7
Std V1.95.17.35.9-0.1-3.3-6.7-7.5
In year 2022
Post-Pandemic: 2022-2024During Pandemic: 2018 to 2022
In year 2024
Percent of children who can readPercent of children who can read
33

Table 15: Eﬀect of Pandemic on Minimum Arithme?c Proﬁciency (% of children)

Source: ASER reports 2018, 2010 & 2006
5.4.2.3 Long Term Trends in MLP
ASER is the only survey which can be used to compare na?onal trends over two decades.
Though ASER surveys started in 2005, there were major changes in 2006, which make the
2005 data non-comparable with later surveys. As 2006 survey was the ﬁrst survey with
improved methodology, it may have taken a year to stabilize, so we have to be somewhat
cau?ous in using the 2006 data when its widely diﬀerent from 2007. Figure 11 depict the
trends in minimum learning proﬁciency among Indian Primary school children from 2006 to
2024. The ﬁrst panel in Figure 11 & Figure 12 is the Minimum Learning Proﬁciency (MLP) in
speciﬁc Std (III or V) over the years. This shows how the outcomes have changed over the
years.
The second panel in Figure 11 & Figure 12 is based on the performance of iden?ﬁed cohorts.
A cohort refers to a group of students who entered Std I in the same year and are tracked as
they progress through grades. Thus the 2006 cohort is always show in cohort year 2006, which
ever grade we are comparing. Thus, this cohort was in Std III in 2008 and in Std V in 2010, In
the second panel of both Figure 11 & Figure 12 , its performance is the percentage shown for
2006. Similarly, the cohort which was in Std I in 2012 is always shown above 2012, though it
was in Std III in 2014 and in Std V in 2016.
Std III is an appropriate level to check how many students have achieved func?onal literacy
and numeracy (FLN) and iden?fy the students who have not achieved FLN, so teachers and
school administrators can pay special a?en?on to them. Minimum reading proﬁciency,
measured by ability to read and understand Std I level text. The share of children achieving
minimum reading proﬁciency (Std I level text), shows three phases. A declining trend from
2006 to 2012 with a sharper decline in later years, a rising trend from 2012 to 2018, and then
the sharp setback from the pandemic followed by recovery (blue diamonds & line, Figure 11 ,
top panel). Minimum arithme?c (subtrac?on) proﬁciency follows a similar pa?ern, but the
decline in the ﬁrst phase is more pronounced, and longer, with the trough reached in 2014.
the recovery in the second phase is minimal (green triangles and line, Figure 11, top panel).
The eﬀect of the pandemic is however less, but recovery from pandemic good.
MAP(%)
1-911-99SubtractDivide1-911-99SubtractDivide
Std I73.733.87.22.062.425.65.81.7
Std III94.570.833.711.490.362.725.98.3
Std V97.785.955.830.796.381.749.925.6
Change
Std I11.38.21.40.3-1.9-1.6-0.1-0.3
Std III4.28.17.83.1-2.2-2.9-2.2-0.2
Std V1.44.25.95.1-0.3-1.1-2.4-2.2
During Pandemic: 2018 to 2022Post-Pandemic: 2022-2024
In year 2022
Recognize nos Recognize nos
In year 2024
Children who can Children who can
34

Figure 11: Minimum Learning Proﬁciency: Std III (% of children)

Data source: Annual Status of Educa?on Reports 2006-2024
The results for cohorts, studying in Std III, are in lower panel in Figure 11. There was a decline
in minimum reading proﬁciency (Std I level txt) from cohort 2006 to cohort 2010, while
minimum arithme?c proﬁciency (subtrac?on) declined from cohort 2006 to cohort 2012. In
both cases there was an uptrend thereaLer, which peaked with cohort 2016. The Pandemic
resulted a sharp decline in minimum reading proﬁciency of the 2020 cohort and a smaller
decline in minimum arithme?c proﬁciency of this cohort. Both recovered sharply to put the
cohorts back on the phase 2 uptrend in share of students mee?ng minimum learning
standards. The gap between the minimum reading ability and minimum arithme?c ability is
starkly illustrated by the blue and green trend lines. This gap was fairly narrow in 2006,
expanded ?ll around 2014 and then narrowed (Figure 11, top panel). When we look at the
performance across cohorts (lower panel, Figure 11) the gap was fairly large even for the 2006
cohort, so its expansion for the middle cohort was less, before it narrowed again. It is therefore
clear, that the founda?on for numeracy is much weaker than for language, and bigger
improvements are needed in pedagogy of teaching arithme?c.
R² = 0.5125
R² = 0.911
22
27
32
37
42
47
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
% of children
Year
Minimum Learning Outcome: Std 3
Std 1 text Subtraction
R² = 0.4262
R² = 0.7995
20
25
30
35
40
45
50
55
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
% of children
Cohort year
Minimum Learning Outcome: Std 3 (Cohort)
Std 1 text Subtraction
35

Most countries deﬁne Std V as the comple?on of primary educa?on. We use the share of
students who can read Std level II text as a measure of minimum reading proﬁciency and the
share of children who can do division as the measure of minimum arithme?c proﬁciency.
Figure 12 shows the trends in minimum learning outcomes for Std V students during 2006-
2024. We again ﬁnd three phases in learning outcomes. A sharply declining trend from
2006/7-2012, a very slowly rising trend from 2012 to 2018, and then the pandemic disrup?on
of a sharper fall and quick recovery (Figure 12, 2
nd
panel). The trend for the cohort based
minimum learning outcomes is signiﬁcantly diﬀerent from the normal annual trends.
Figure 12: Minimum Learning Proﬁciency: Std V (% of children)

Data source: Annual Status of Educa?on Reports 2006-2024
The cohort learning outcomes for Std V shows a decline in minimum reading (Std II level text)
and arithme?c (division) performance from 2006 cohorts to 2007 cohort. Minimum learning
outcomes then slowly improved from cohort 2007 to 2014. Because of the pandemic the
performance was poor, while cohort 2020 reﬂects the post pandemic recovery. The gap
between minimum arithme?c proﬁciency and minimum reading proﬁciency follows the same
?me pa?ern as in Std III, but are much wider, reinforcing the need for changes in pedagogy
and related teacher training
Std I, founda?onal literacy and numeracy (LN) trends over ?me were similar to Std III
founda?onal trends were similar. Share of children recognizing/understanding words, decline
from 23.9% in 2007 to 19.1% in 2012. It then recovered to 24.7% by 2018. The pandemic shock
reduced this to 20.8% in 2022 with an equally sharp recovery to 29.4% in 2024 a share higher
than in 2007 (Table 16).
The same pa?ern was seen for minimum numeracy (10-99). Share of children
recognizing/understanding numbers (10-99), declined from 25.5% in 2007 to 21.0% in 2012.
It then recovered gradually to 27.2% by 2018. The pandemic shock reduced this to 25.6% in
2022 with an equally sharp recovery to 33.8% in 2024, higher than in 2007 (Table 16).
28

28
When compared with Std 3 student huge improvement can be seen in both reading and arithme?c. Improvement over grades could be
partly due to dropouts, as students who are unable to acquire basic learning skills are more likely to drop the school.
R² = 0.7388
R² = 0.8983
20
25
30
35
40
45
50
55
60
65
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
% of children
Year
Minimum Learning Outcome: Std 5
Std 2 Text Division Trend (Std 2 Text)Trend (Div)
R² = 0.2935
R² = 0.4735
20
25
30
35
40
45
50
55
60
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
% of children
Cohort year
Minimum Learning Outcome: Std 5 (Cohort)
Std 2 Text Division
36

Table 16: Minimum Learning Proﬁciency (% of children)

Data source: ASER 2006-2024; RBI: Macro-Economic Aggregates (2011-12 At Constant Prices) (Amount in ₹ Crore)
The row 15 of Table 16 also shows the correla?on between per capita GDP and learning
proﬁciency of children in Std I, III and V. We ﬁnd learning outcomes in Std I are highly corelated
with per capita GDP, for both reading proﬁciency of Std I level text (0.89) and arithme?c
proﬁciency in numbers 10-99 recogni?on (0.77). The correla?on with word recogni?on is
however moderate (0.34). Overall, however, the posi?ve correla?on suggests improvement
during 2006-2024, as PCGDP increased. In Std III, the correla?on between PCGDP and reading
proﬁciency of Std II level text is reasonably high (0.67), but is nega?ve for Std I level text,
subtrac?on and division is nega?ve. A similar nega?ve correla?on is found for Std V learning
outcomes (Table 16). These suggest that learning outcomes in primary Std I, III & V, have not
kept pace with the rising per capita GDP.
Row 16 shows the correla?on across nearby columns of reading & arithme?c. The correla?on
between MRP & MAP is fairly high in primary school, for the levels we have chosen to reﬂect
minimum learning proﬁciency for each std: 0.85 in Std I, moderately good 0.57 in Std III and
0.85 in Std V.
The diﬀerence in correla?on between Arithme?c levels and language levels along with the
fact that minimum arithme?c proﬁciency is generally lower than minimum language
proﬁciency, suggests the following: Higher level learning in arithme?c (math) is much more
dependent on learning in previous levels, than in reading (Language). Arithme?c involves
abstrac?ons and logical reasoning using these abstrac?ons. Every level is therefore built on
the founda?on of the previous lower level. Those unable to understand the lower level have
li?le chance of understanding the higher level. Language on the other hand is the instrument
of everyday communica?on. As much is, or can be, learned from listening to family & third-
party interac?ons as from formal schooling. So, language proﬁciency depends as much on the
StdPCGDP
YearWordNos 10-99Std 1 textSubtractStd 1 textSubtractStd 2 textDivideStd 2 textDivide(2011-12)
12345678910
1200623.4 6.67.847.918.519.915.253.045.356964
2200723.925.56.95.449.042.321.611.258.742.460466
3200823.222.06.84.450.338.722.210.856.237.061468
4200924.224.78.16.046.639.019.810.652.838.065394
5201024.823.77.85.545.736.420.09.453.435.969994
6201122.221.36.94.440.329.918.86.748.227.671609
7201219.121.07.14.238.626.321.46.746.824.874599
8201320.621.98.04.440.126.121.67.447.025.678348
9201421.123.89.04.540.225.423.67.448.126.183091
10201622.225.09.85.042.427.625.18.447.825.994751
11201824.727.211.05.944.528.127.28.550.327.8105448
12202220.825.68.85.835.625.920.58.342.825.6116892
13202429.433.812.27.247.033.727.011.448.730.7133501
14Avg23.024.68.45.443.730.622.29.450.331.7
15CorPcG0.340.770.890.26-0.36-0.19 0.67-0.22-0.62-0.57 1.00
16CorAR0.820.850.330.420.57-0.04-0.04 0.600.85
Std 3Std 5Std 1
37

environment in which the child lives at home as from school. Therefore, many children can
leapfrog reading levels, while some may even regress if their home environment is poor in
communica?on skills.
Research by Evans and Popova (2016) conﬁrm that changes in pedagogy, curriculum and
tes?ng are cri?cal to be?er learning outcomes. Eﬀec?ve teaching in classrooms is essen?al for
keeping students in school and helping them develop skills and learn eﬀec?vely. A large body
of evidence from school interven?ons in low- and middle-income countries clearly points to
more eﬀec?ve teaching, through improved pedagogy, as the most impac?ul way to improve
learning outcomes.
5.5 Learning Outcomes: Oﬃcial Surveys
In this sec?on we present some of the results of the Na?onal Achievement Survey (NAS 2017,
2021), The Founda?onal Learning Study (FLS 2022) and the Parakh Rashtriya Sarvekshan
survey (PRS 2024). As noted in Sec?on 5.2.1 NAS (2017) and NAS (2021) have the same design,
but as the la?er represents the outcome of a once in a century pandemic it is not possible to
discern trends by comparing the two. Similarly, though PRS (2024) design and implementa?on
is an update and improved version of the NAS, but comparable data on student performance
in Class/Grade 3 is not available in the format used to classify students in by performance level
in NAS 2021 & 2017. FLS (2022) study was conducted in 20 languages in 36 States and UTs. It
aimed to establish a baseline for achieving NIPUN Bharat goals, by se?ng benchmarks in
literacy and numeracy (Kar, Kamat & Guria (2024). It has been implemented only once.
NAS divides students into Below Basic, Basic, Proﬁcient, and Advanced, based on diﬀerent sets
of ques?on, percent of correct answers, average test scores, etc. The Founda?onal Learning
Study (FLS) groups students into Below Par?ally Meets, Par?ally Meets, Meets, and Exceeds
Minimum Proﬁciency based on global benchmarks. In these surveys, the distribu?on of
students across categories is mutually exclusive, adding up to 100% within each subject and
grade. It is understood that the students who are advanced in language (or arithme?c) are
also proﬁcient in language (or arithme?c, respec?vely) in this sec?on we aim to benchmark
these surveys against ASER, and to extract any other informa?on which is not available in ASER
surveys, such as performance diﬀerences across Urban-Rural and Gender.
5.5.1 Learning Outcomes: NAS and FLS
To make the surveys comparable, diﬀerent methods were used. First, simple and weighted
averages were created from state-level data available in each survey, and these were cross-
checked with the total averages reported by the survey itself. Then, interstate correla?ons
were run across categories, along with mul?ple regressions at the state level, to test the
consistency of results. Some of these are reported in the sec?on on Interstate comparisons.
Here we report the Na?onal level analysis to map NAS results on minimum learning outcomes
in Primary classes, with the results from ASER, recognizing that the la?er covers only rural
areas. Since these surveys are from diﬀerent years, it is diﬃcult to compare their results
38

directly. Nevertheless, we use surveys from nearby years, such as ASER 2022 with NAS 2021,
and NAS 2017 with ASER 2016 & ASER 2018. NAS results for percent of students reported as
proﬁcient & advanced in language or arithme?c are compared with ASER for per cent of
children who can read Std I level text & Std II level text, and those able to do subtrac?on and
division, respec?vely.
ASER reports, separately, the percentage of children able to read Words, Std I and Std II level
text in language, and those able to recognize numbers (10-99), subtract, and divide in
mathema?cs. We assume that a student who can read a Std 1 level text can also recognize
words, and a student who can do subtrac?on can also recognize numbers 10-99. NAS reports
separately the percentage of students at Proﬁcient and Advanced performance levels. We
assume that a student who is advanced is also proﬁcient. In our use of the term “Proﬁcient”,
therefore includes students who at a higher, “advanced” level.
Table 17 presents Grade 5 (end of primary) results. In Minimum Reading Proﬁciency (MRP)
reasonably close match between the “proﬁcient” level deﬁned in NAS and the ability to read
Std II level text, in ASER. Based on na?onal averages (i.e., all India), as per ASER 2022, 42.8%
of children can read standard II level text, and as per NAS 2021, 42.0% of student are proﬁcient
in language. Similarly based on na?onal averages, in minimum arithme?c proﬁciency, 25.6%
could do division as per ASER 2022, while 25.0% were Proﬁcient in mathema?cs as per NAS
2021. This implies that NAS benchmark of Class 5, proﬁciency in language, is equivalent to
ASER’s (Std I-VIII) benchmark of ability to read Std II level text, and the NAS benchmark of
proﬁciency in Class 5 mathema?cs is equivalent to ASER (Std I-VIII) benchmark of ability to
division. The small diﬀerences in these measures between ASER 2022 and NAS2021, may be
due to a combina?on rural-urban diﬀerences and diﬀerent survey year.
The second closest ASER bench mark for language (Std I level text) at 62.7% is oﬀ by almost
50% from the NAS “proﬁcient”, i.e., it cannot be explained by rural-urban diﬀerence or the
diﬀerence in year of survey. Similarly, the alterna?ve benchmark for arithme?c from ASER
(subtrac?on) at 49.9% is also oﬀ by almost 50%, from “proﬁcient” in math as per NAS 2021, a
diﬀerence too large to be due to rural-urban diﬀerence or diﬀerence in year of survey (Table
17)
Though NAS 2017 is an eight-year-old survey, it holds special signiﬁcance, because it was the
basis of the interna?onal comparison of Minimum Reading Proﬁciency presented in World
Bank, World Development indicators. As there is no ASER survey for 2017, ASER 2018 and
2016 are used to compare with NAS 2017. As per NAS 2017, 46.3% of students were
“Proﬁcient” (Table 17). This was (-)4% points less than the 50.3 per cent of children who could
read Std II level text (ASER 2018) and (-)1.5% point lower than the 47.8% of children who could
read Std II level text (ASER 2016). Though the gaps are higher than for NAS2021-ASER2022
comparison, they are small enough to be accounted for by rural-urban diﬀerences (Table 17).
39

Table 17: Minimum Learning Proﬁciency (% of students): Grade 5/Standard 5

Source: ASER Reports: 2022, 2018 and 2016; NAS survey 2021 and 2017
The picture is however very diﬀerent for Minimum Math/Arithme?c proﬁciency. In
mathema?cs, as per NAS 2017 43.7% of students are Proﬁcient is 16% points higher than the
27.8% who can do Division in 2018, and 18% points higher than the 25.9% who can do Division
in 2016 (ASER)). If we compare students proﬁcient in Mathema?cs (43.7% NAS 2017) with
children who can do Subtrac?on or higher (52.3% ASER, 2018, 52.3% ASER 2016), then the
NAS is (-) 8.6 % (-)6.8% lower than ASER 2018 & 2016 respec?vely. This shows that the NAS
2017 result for math proﬁciency, lies between ASER’s criteria Minimum Arithme?c Proﬁciency
of children who can do Subtrac?on or higher, and children can do Division (Table 17).
To conclude, for Class 5 language, NAS “Proﬁcient” level (including student at advance level)
for language is close to ASER benchmark of children who can read Std II level text. For
Mathema?cs, the comparable standard is expected to be Division, but the NAS results for
2021 is equivalent to this, but the NAS 2017 result is between MAP subtrac?on (including
division) and MAP division. This suggests that, the weakness observed in NAS 2017 math
assessment, were corrected in NAS 2021, and hopefully have been further improved in
Parakh.
Table 18 presents Grade 3 results on minimum learning proﬁciency across FLS, ASER, and
NAS. The categories diﬀer by survey. FLS reports the percentage of students who Meet Global
Minimum Proﬁciency (GMP), and those who Exceed GMP. We assume that those students
who exceed GMP have already met the GMP standard. For 2022, FLS and ASER can be directly
compared. As per FLS, 33.0% of students meet or exceed GMP in language, which is close to
35.6% of children who can read Std I level text or higher, as per ASER. There is a slightly higher
diﬀerence of 4% points between the 54.0% students who meet or exceed GMP for language,
as per FLS, and the 58.0% of children who can read words or higher text as per ASER. This
implies that FLS benchmark of “exceeds Global Minimum Proﬁciency” in language is
equivalent to the ASER benchmark of ability to read Std I (or higher-level) text (Table 18).
Grade 5Proficiency Levels Students %Level Students %
Can Read Std1 or higher text
62.7
Can do Substraction or higher
49.9
Can Read Std 2 level text42.8Can do Division25.6
NAS 2021Proficient (incl Advanced)42.0Proficient (incl Advanced) 25.0
Can Read Std1 or higher text
69.4
Can do Substraction or higher
52.3
Can Read Std 2 level text50.3Can do Division27.8
NAS 2017Proficient (incl Advanced)46.3Proficient (incl Advanced) 43.7
Can Read Std1 or higher text
66.4
Can do Substraction or higher
50.5
Can Read Std 2 level text47.8Can do Division25.9
ASER 2022
Language Mathematics
ASER 2018
ASER 2016
40

Table 18: Minimum Learning Proﬁciency (% of students): Grade 3 / Standard 3

Source: PARAKH 2024, FLS and ASER 2022
In mathema?cs, the standards used in FLS for Grade 3, does not match any speciﬁc ASER Std
(one to one). As per FLS, 52.0% students meet or exceed GMP, but as per ASER this number
lies between the 62.7% of children who can recognize numbers (10-99) or perform higher level
mathema?cs, and the 25.9% of children who can do subtrac?on or higher. This implies that
the FLS standard for mathema?cs lies in between the ASER standard of “recognize number
10-99 and can-do subtrac?on (right panel, Table 18).
Comparing NAS 2021 with ASER 2022, in language/reading proﬁciency, NAS es?mates 39.0%
of Class 3 students are Proﬁcient, 3.4% point more than the 35.6% of students with minimum
reading proﬁciency (Std I level text) as per ASER. This implies that NAS Class 3 benchmark of
Proﬁcient in language, is equivalent to the ASER benchmark of, “can read Std I level text.” In
mathema?cs, the standards are again very diﬀerent, as we found for FLS. As per NAS 2021,
42.0% of students are Proﬁcient, which is between the 62.7% of students who can recognize
numbers (10-99), and 25.9% of children who can do subtrac?on as per ASER.
NAS 2017 can be compared with ASER 2018 and ASER 2016. In language, 47.0% of Class 3
students are Proﬁcient, about +2.5% higher than the 44.5% who can read Std I level text as
per ASER 2018, and +4.6% higher than the 42.4% show can read Std I level text as per ASER
2016. A fairly good match of Minimum learning outcomes in language. In mathema?cs, NAS
53.0% of students at Proﬁcient, which lies between ASER Minimum Arithme?c proﬁciency
(MAP) among Std III students who are proﬁcient in two-digit numbers 10-99 (65.6% & 63.2%
in ASER 2018 & 2016), and per cent of those who can do subtrac?on (28.1% & 27.1% in 2018
& 2016). This conﬁrms that the NAS standards for minimum language proﬁciency is more or
less equivalent to the ASER standard for minimum learning proﬁciency in language., while the
NAS standards of proﬁciency in mathema?cs is weaker than ASERs.
To conclude, for Grade 3 language NAS “Proﬁcient” and FLS’s Minimum Global proﬁciency
(GMP) students broadly match ASER’s Std 1 level text reading standard, for mathema?cs, NAS
proﬁcient and FLS’s GMP standards, lie in between ASER’s arithme?c standard for recogni?on
Grade 3Proficiency Levels Students %Level Students %
Meets or Exceeds GMP (+)
54.0Meets or Exceeds GMP(+) 52.0
Exceeds Global Min Profeciency
33.0
Exceeds Global Min Profeciency
10.0
Can read Words or higher text
58.0
Can recognize Nos 10-99 or do highr math
62.7
Can Read Std1 or higher text
35.6
Can do Substraction or higher
25.9
NAS 2021Proficient (incl Advanced)39.0Proficient (incl Advanced) 42.0
Can read Words or higher text
65.3
Can recognize Nos 10-99 or do highr math
65.6
Can Read Std1 or higher text
44.5
Can do Substraction or higher
28.1
NAS 2017Proficient (incl Advanced)47.0Proficient (incl Advanced) 53.0
Can read Words or higher text
62.3
Can recognize Nos 10-99 or do highr math
63.2
Can Read Std1 or higher text
42.4
Can do Substraction or higher
27.6
ASER 2016
Mathematics
FLS 2022
ASER 2022
ASER 2018
Language
41

of two-digit numbers (10-99) and the standard for subtrac?on. This suggests that NAS & FLS
math standard is either weaker than ASER’s arithme?c standards, or is less stringently applied.
5.5.2 Change in Learning (NAS & FLS)
When comparing NAS 2021 with NAS 2017, the results show a clear decline in learning
outcomes, largely reﬂec?ng the impact of the COVID-19 pandemic. In 2017 the percentage of
students performing at the proﬁcient level in Classes 3 and 5 was higher in Language and
Mathema?cs. By 2021, these percentages had dropped sharply. The decline was more severe
in mathema?cs than in language. For example, in Class 3 mathema?cs, the share of proﬁcient
students fell by (-)11.6 percentage points (Table 19). In Class 5 mathema?cs, the drop was
(-)18.6 percentage points between 2017 and 2021. The Founda?onal Learning Study (FLS)
2022, though not directly comparable with NAS, suggests signs of recovery aLer the
pandemic. For instance, in Class/Grade 3 language, while 39% of students reached proﬁcient
level in NAS 2021, the FLS 2022 recorded 54% (Table 19).
Table 19: Minimum Learning Proﬁciency (% of students): Grade 3 & 5

Note* Proﬁcient includes students at advance level; Meets (including exceeds).
Source: Na?onal Achievement Survey 2021 & 2017 and FLS 2022
5.5.3 Learning Diﬀerence across loca?on & gender
This sub-sec?on analyses the results of NAS, FLS and Parakh surveys to determine Rural-
Urban and Male-Female diﬀerences in learning outcomes. This has a bearing on how
indica?ve are ASER’s rural surveys are for Urban students, and for students as a whole.
Table 20 shows that in Class 3, student learning proﬁciency in mathema?cs (42%) is far higher
than in any ASER survey. As analyzed in the earlier sub-sec?on, the standards for proﬁciency
in math seem to be weaker or less stringently applied than in ASER surveys. The percent of
students who are proﬁcient in language (30%) are approximately the same as in ASER surveys,
as benchmarks are equivalent.
This anomaly between NAS and other survey results with respect to Math proﬁciency, is not
seen in Class 5 results. This indicates a be?er applica?on of standards to tes?ng/assessment
of students. Geographical and gender diﬀerences in Class 3 student proﬁciency are generally
quite small. Learning outcomes for rural students are be?er than those of urban students, in
Class 5 Math (+13%) and Class 3 language (+5%), but worse in Class 5 language (-)5%. Language
learning outcomes for girls are clearly be?er than for boys in both Class 5 (+16%) and Class 3
(+8%), and the same in Class 5 Math. As indicated earlier, Class 3 Math results are unreliable.
Grade/Subject
Class III
Proficient*AdvancedProficient*Advanced
Meets*Exceeds
Proficient*Advanced
Language47.215.139.013.054.033.015.020.0
Mathematics52.916.242.012.052.010.010.0-2.0
Class V
Proficient*AdvancedProficient*Advanced
Language46.311.842.010.0
Mathematics43.612.725.06.0
NAS 2017 NAS 2021 FLS 2022Change:2022-2021
42

Table 20: Students (%) at diﬀerent levels of proﬁciency: NAS 2021

Source: Na?onal Achievement Survey (2021) Technical Report
Besides the distribu?on of students over learning levels, NAS 2021 also gives student
performance in terms of average scores and per cent of correct answers averaged over all
students in each grade. As already noted, Class 3 language scores are unreliable, so we will
ignore them.
In Class 5, both average scores and per cent of correct answers are higher for Language than
for Math (Table 21) for Class 3 and 5 from NAS survey. Across both Class, Language learning
outperforms Mathema?cs in average scores and correct answers. This is consistent with
earlier results on proﬁciency in language and Math (% of students who are proﬁcient).
Table 21: NAS 2021: Average Scores & Correct Answers (%)- Primary

Source: Na?onal Achievement Survey 2021; Note: Scores are out of 500, converted to percent.
In Class 5, urban students outperform rural students on average scores and correct answers
in Language, and average scores in Mathema?cs. Rural students are be?er only in correct
answers. In Class 3, rural students outperform urban students in correct answers in language,
Student(%) by performance levelsSubject/Category LanguageMathsLanguageMaths
Total
Proficient (including advance)42.025.039.042.0
Urban
Proficient (including advance)43.023.038.039.0
Rural
Proficient (including advance)41.026.040.044.0
Boys
Proficient (including advance)39.025.037.042.0
Girls
Proficient (including advance)44.025.040.042.0
Class 5 Class 3
Class 5
TotalUrbanRuralBoysGirls
Average score (%)
Language 61.862.660.860.661.0
Mathematics 56.856.256.056.056.2
Correct Answers (%)
Language 55.056.055.054.056.0
Mathematics 44.043.044.044.044.0
Class 3
TotalUrbanRuralBoysGirls
Average score (%)
Language 64.664.264.063.664.6
Mathematics 61.259.860.460.060.2
Correct Answers (%)
Language 62.061.062.061.063.0
Mathematics 57.056.058.057.057.0
43

while urban students have a small edge in average scores (0.2% points). In language, girls
outperform boys in both Class 5 and 3, across average scores and correct answers. In
mathema?cs, boys and girls achieved the same percentage of correct answers in Class 5.
Overall, girls tend to perform be?er in language, while boys and girls have comparable
performance in mathema?cs (Table 21).
The Founda?onal Learning Study (FLS) 2022, conducted under the NIPUN Bharat Mission, set
na?onal benchmarks for founda?onal skills at the end of Grade 3 to assess tes?ng English
reading with comprehension and Numeracy. Results are reported as the percentage of student
who Meets or Exceeds the Global Minimum Proﬁciency (GMP) level, with the “meets”
category including those who also exceed.
Table 22 shows that in English, results are uniform across total, boys, and girls, with 54% of
students mee?ng or exceeding GMP. In contrast, Numeracy scores are slightly varying, with
boys performing be?er (53%) than girls (51%). These FLS ﬁndings are not directly comparable
with NAS 2021, yet they highlight a diﬀerence: 54% of students in Grade 3 meet or exceed
GMP in English (FLS), against only 39% found proﬁcient in Language in NAS 2021 (Table 22)
Table 22: Students (%) at diﬀerent proﬁciency levels in grade 3

Source: Founda?onal Learning Study (FLS) 2022.
PARAKH Rashtriya Sarvekshan (PRS 2024), reports average scores for founda?onal stages as
Grade 3, it assesses Language and Mathema?cs, with results presented across gender and
loca?on. As PRS is based on a new, comprehensive design, its ﬁndings are not directly
comparable with earlier NAS rounds. It is unclear whether the deﬁciency in assessment of
Grade 3 math proﬁciency, no?ced in NAS survey for Class 3, has been rec?ﬁed, given that
average scores in language are only 4% higher than in mathema?cs (Table 23).
In Grade 3, Language scores are higher than Mathema?cs across all groups, with an overall
average of 64% in Language compared to 60% in Mathema?cs (Table 23). Rural students
perform marginally be?er than urban students in both subjects. By gender, girls outperform
boys, scoring 65% in Language compared to 63% for boys, while achieving the same score in
Mathema?cs (60%).
Table 23: Average scores in grade 3 (PRS 2024)

Source: PARAKH Rashtriya Sarvekshan, 2024
Grade 3
Global Proficiency LevelsTotalGirlsBoysTotalGirlsBoys
Meets or exceeds GMP54.054.054.052.051.053.0
Exceeds GMP 33.033.033.010.010.010.0
English Numeracy
Grade/Subjects
Grade 3TotalUrbanRuralMaleFemale
Language 6463646365
Mathematics 6059606060
Average Score(%)
44

In conclusion, the available data suggests that the geographical diﬀerences in learning
outcome between Rural and Urban areas, vary in both direc?ons, and are small enough to fall
within the range of sta?s?cal errors in the survey. We can use the ASER rural survey to broadly
represent na?onal learning outcomes.
5.6 Learning Outcomes in States
ALer looking at the overall na?onal picture, the next step is to see how learning outcomes
vary across states. For this, Std III results are used to reﬂect founda?onal learning, while Std V
results are taken to assess primary learning. The interstate comparison helps to understand
the diﬀerences and gaps in learning levels across states.
5.6.1 Learning Outcomes in States (ASER)
We start by examining learning outcomes across States for 2024 based on the ASER rural
survey. There is virtually no correla?on between per capita NSDP of a State and its primary
educa?on performance in terms of learning outcomes. For Std 5, the correla?on of State
PCNSDP with percent of children with Min reading proﬁciency (MRP) is only 0.16, and with
percent of children with Minimum Arithme?c Proﬁciency (MAP) in nega?ve (-)0.08. This
means that there is no dependence of learning outcomes on how well oﬀ/developed or
poor/underdeveloped the State or UT is. This is consistent with the results of Angrist et al.
(2023) who found no signiﬁcant improvement in learning outcomes resul?ng from increased
expenditures on educa?on. While enrolment expanded, being in school did not guarantee
learning.
Next, we divide all 27 States and UTs into three categories (A, B, C), using Minimum learning
levels among Std V students. We use the all India mean and the standard devia?on of
Minimum reading proﬁciency (Std II level text) to classify States & UTs by percent of children
with MRP (Table 24, leL panel), and we use Minimum Arithme?c proﬁciency (division) to
classify States & UTs by per cent of students with MAP (Table 24, right panel). The numbers in
brackets against each state show the change between 2014 and 2024, in the percent of
children in the STATE with MRP & MAP respec?vely. The B level is created by taking a range of
0.67 ?mes the standard division, on each side of the all-India average given by ASER for MRP
and MAP respec?vely (Table 24).
In the case of MRP in Std V, there are nine States & UTs in A (above average) category, ten in
B (average) and eight in C (below average). Surprisingly, UP and Odisha are in the A category
having seen among the highest improvement since 2014, of 11.7% and 7.6% respec?vely.
Three other States in A category improved learning outcomes between 2014 and 2024
(Mizoram, Maharashtra & U?arakhand), while four States/UTs MRP deteriorated (Kerela,
Haryana, Punjab, Himachal (column 1,Table 24).
45

Table 24: Min Learning outcomes in Std 5(%), across States-2024 (& change from 2014)

Source: Authors calc based on ASER report 2024 & 2014.Note: All India Std 5 MRP (std2 txt) is 48.7%, Standard devia?on is
11.1%, and MAP (division) is 30.7% with SD =10.1%; Note: Cut-ﬀs are based on All India +/- 0.67* SD
In the case of Minimum Arithme?c proﬁciency (MAP) in Std V, there are six States/UTs in A
(above average) category, nine States/UTs in B(average), and ten States in C (below average)
category. U?ar Pradesh (UP), U?arakhand and Mizoram, are in A category with an
improvement of 13.7%, 9.6% and 4.7% respec?vely, between 2014 & 2024 (column 4, Table
24). Punjab, Himachal and Haryana are other States which are in A category for both MAP and
MRP (columns 1 & 4). Eight States/UTs have below average outcomes in MRP and the 11 states
have below average outcomes in MAP, with four state common (Assam, Karnataka, Tamil Nadu
and Tripura (columns 3 & 6, Table 24). Lagging States need to improve their performance, with
the help of NGOs such as Pratham, as UP and other States have done.
The compara?ve results for Minimum learning outcomes across States and UTs, for Std III
are also depicted in Figure 13 & Figure 14 and those for Std V in Figure 15 & Figure 16.
Minimum reading proﬁciency (MRP) is measured by the percentage of Std III children who can
read a Std I level text (Figure 13), and the minimum arithme?c proﬁciency (MAP) is measured
by the percent of Std III children who can do subtrac?on (Figure 14). The States are sorted by
minimum learning outcomes in 2024 (purple dots), but the data for 2014 (pink dots) is also
given so that the change between 2014 and 2024 is visible (line).

A: Above averageB: AverageC: Below averageA: Above averageB: AverageC: Below average
More than 56.1%41.3% to 56.1%Less than 41.3%More than 56.1%41.3% to 56.1%Less than 41.3%
Mizoram (+15.6%)Jharkhand (+10.9%)Assam (+4.9%)UP (+13.7%)Odisha (+10.4%)MP (+7.8%)
UP (+11.7%) Sikkim (+10.2%) J & K (-1.0%) Uttarakhand (+9.6%)Jharkhand (+9.1%)Assam (+5.0%)
Odisha (+7.6%)MP (+9.6%) Arunachal (-3.4%) Mizoram (+4.7%)Maharashtra (+8.8%)Meghalaya (+4.7%)
Maharashtra (+6.1%)Nagaland (+7.6%)Tripura (-4.6%) Punjab (+4.4%)Chhattisgarh (+7.7%)Karnataka (+0.8%)
Uttarakhand (+3.3%)Chhattisgarh (+2.0%)Tamil Nadu (-11.3%)Himachal (+0.2%)West Bengal (+2.5%)Tripura (-0.4%)
Kerala (-0.8%) West Bengal (+1.4%)Karnataka (-13.2%) Haryana (-8.6%) Bihar (+1.2%) Rajasthan (-1.7%)
Haryana (-4.6%) Rajasthan (+0.9%)Andhra (-19.7%)J & K (+0.2%) Gujarat (-1.8%)
Punjab(-5.5%) Gujarat (-0.3%) Telangana (-22.9%)Andhra (-1.4%) Nagaland (-5.0%)
Himachal (-8.4%)Bihar (-4.5%)Arunachal (-5.0%)Tamil Nadu (-5.0%)
Meghalaya (-15.5%)Telangana (-8.5%)Sikkim (-14.4%)
Kerala (-17.8%)
Minimum Reading Proficiency (MRP) in Standard 5 Minimum Arithmetic Proficiency (MAP)in Standard 5
46

Figure 13: Minimum Reading Ability FLN (Std 3): 2024 & 2014

Data source: Annual Status of Educa?on Reports 2014 & 2024
Figure 14: Minimum Subtrac?on Ability FLN (Std 3): 2024 & 2014

Data source: Annual Status of Educa?on Reports 2014 & 2024
Minimum reading proﬁciency (MRP) is measured by the percentage of Std V students who
can read a Std II level text (Figure 15), and the minimum arithme?c proﬁciency (MAP) is
measured by the percent of Std V students who can do division (Figure 16). The States are
sorted by minimum learning outcomes in 2024 (purple dots), but the data for 2014 (pink dots)
is also given so that the change between 2014 and 2024 is visible (line).

47

Figure 15: Minimum Reading Ability Primary (Std 5): 2024 & 2014

Source: Author’s calcula?on based on ASER 2014 & 2024 reports
Figure 16: Minimum Division Ability Primary (Std 5): 2024 & 2014

Source: Author’s calcula?on based on ASER 2014 & 2024 reports.
Much greater a?en?on needs to be paid by almost all States & UTs on Minimum learning
outcomes, than has been the prac?ce in the past. Excessive focus on comple?ng set
curriculum and pressure to move all students to the next class has led to creden?alism, where
a cer?ﬁcate comple?on is more important than actual learning, this can and must be changed.
48

Angrist et al. (2023) evaluated over 150 programs and iden?ﬁed two cost-eﬀec?ve
interven?ons: Firstly, structured pedagogy, which includes textbooks, teacher guides, training,
and coaching; and secondly Teaching at the Right Level (TaRL), which groups students by their
actual learning level, not by age or grade, with or without a technology component. This
highlights that small, targeted investments can generate large gains in learning outcomes and
long-term economic beneﬁts.
5.6.2 Minimum Reading Proﬁciency: NAS
As interna?onal comparisons of Indian learning were based on the Na?onal Achievement
Survey of 2017, we start by looking at the Minimum Reading Proﬁciency across States as per
NAS 2017 data. There is a nega?ve correla?on between per capita NSDP and Minimum
Reading proﬁciency across States. In cross-country comparison, learning outcomes improve
with per capita GDP in at PPP in constant interna?onal prices. There is no natural improvement
in learning outcomes as their per capita NSDP at constant prices, increases.
States are therefore distributed into three categories, using the all-India results, along with
the standard devia?on of State results, following the methodology used for ASER results in
Table 24. The Average category is sub-divided into two sub-categories- above and below the
na?onal average. The numbers in bracket are the changes in MRP between NAS 2017 and NAS
2022. There are thirteen States in the B category, four of which four were above the Na?onal
average and nine below the na?onal average (columns 3 & 4, Table 25),
Table 25: Min Reading Proﬁciency across States – % of students in Grade 5, NAS 2017

Source: Authors calculation based on NAS 2017 & 2021. Note 1: Cut-ffs are based on All India +/- 0.5*SD. Note 2:
All India Std 5 MRP (Language) is 46.3% in 2017, Standard deviation (SD) is 12.7%.
Nine States were in category A and thirteen in category C. There was a posi?ve change in
learning outcomes between 2017 and 2022 in four States in category B and eleven of 13 States
in category C (Table 25, 3rd & 4th column). The largest improvements were seen in Punjab
(27%), Puducherry (23%), Sikkim (16%) Jammu & Kashmir (14%) and Arunachal Pradesh (13%).
A: Above average C: Below average
More than 52.6% 46.3% to 52.6% 40% to 46.3% Less than 40%
Punjab (+27%)
Chandigarh (-9%)Maharashtra (-1%)Manipur (+7%)Puducherry(+23%)
Rajasthan (-10%)Assam (-8%)West Bengal (+4%)Sikkim (+16%)
Jharkhand (-15%)Gujarat (-12%)Goa (+2%) Jammu&Kashmir (+14%)
Himachal Pradesh (-18%)Tamil Nadu (-14%)Haryana (+2%)Arunachal (+13%)
Kerala (-20%) Madhya Pradesh (0%)A&N Island(+6%)
Uttarakhand (-23%) Bihar (-4%) Mizoram (+5%)
Karnataka (-27%) Tripura (-8%) Odisha (+4%)
Andhra Pradesh (-31%) Chattisgarh (-12%)Delhi (+3%)
Dadra &Nagar (-37%) Telangana (-19%)Meghalaya (+3%)
UttarPradesh (+2%)
Nagaland (+1%)
Daman & Diu (-5%)
Lakshadweep (-6%)
B: Average
49

The founda?onal learning study was also signiﬁcant as to tried to bench park founda?onal
literacy across the diﬀerent language in which Founda?onal literacy and Numeracy is imparted
to Primary students. As the New Educa?on Policy (2020) stated: “The ability to read and write,
and perform basic opera?ons with numbers, is a necessary founda?on and an indispensable
prerequisite for all future schooling and lifelong learning”. However, various governmental, as
well as non-governmental surveys, indicate that we are currently in a learning crisis: a large
propor?on of students currently in elementary school - es?mated to be over 5 crores in
number - have not a?ained founda?onal literacy and numeracy, i.e., the ability to read and
comprehend basic text and the ability to carry out basic addi?on and subtrac?on with Indian
numerals. A?aining FLN for all children will thus become an urgent na?onal mission, with
immediate measures to be taken on many fronts and with clear goals that will be a?ained in
the short term (including 100% founda?onal literacy and numeracy by Grade 3). The highest
priority of the educa?on system will be to achieve universal founda?onal literacy and
numeracy in primary school by 2025.”
In this background, the cross-State results on, “Oral reading ﬂuency with reading
comprehension”, from the Founda?onal Learning study (2022) are important. Table 26 (like
Table 25) uses the all India mean results (%), and the standard devia?on of state results, to
assign States & UTs to three categories A, B & C and subdivide B into two subcategories (Table
26). There are ten States & UTs in category A (>40%), ﬁLeen in category B (26% to 40%) and
ten in category C (<26%). Of the States and UTs in category B, seven are above the all-India
average (33%) and eight below it. The best performing State was Odisha with 65% of Grade 3
students proﬁcient in reading. In the worst performing State only 7% of Grade 3 students met
minimum reading proﬁciency levels. Five states had only 13%-17% students mee?ng MRP
levels (Table 26). It will be a big challenge for States to meet the NEP (2020) goal of 100% FLN.
Table 26: Min Reading Proﬁciency across States (% of students in Grade 3, FLS 2022)

Source: Authors calculations based on FLS State reports (2022). Note: All India Std 3 MRP(English) is 33.0%, Standard deviation is 14.0%.
This sec?on also examines the degree to which ranking of States by share of students with
Minimum Reading proﬁciency, in various surveys are consistent with each other. Table 17
A: Above average C: Below average
More than 40%33% to 40%26% to 33% Less than 26%
Ladakh GoaRajasthanMeghalaya
Mizoram A&N IsIandHimachalPradeshDadra&Nagar, Daman&Diu
Punjab SikkimAssam Puducherry
Tripura MaharashtraChhattisgarhTamilnadu
West BengalGujaratJammu & KashmirAndhra Pradesh
Bihar KeralaUttarPradeshNagaland
UttarakhandJharkhandArunachal PradeshManipur
Haryana KarnatakaChandigarh
Lakshadweep Telangana
Odisha Madhya Pradesh
Oral Reading Fluency with Reading Comprehension
B: Average
50

iden?ﬁed the learning outcomes from ASER and NAS surveys, which appeared to be
comparable to each other. A correla?on matrix is constructed with these, Std/Grade 5,
minimum reading proﬁciency outcome indicators for all States (Table 27).
Minimum Reading Proﬁciency, as measured by ability to read/understand a Std II level text is
highly corelated across all ASER surveys, ranging from a high of 0.9 between ASER 2016 and
ASER 2018, to a low of 0.78 between ASER 2022 and 2016 surveys. This indicates a fairly high
degree of consistency, in which varia?ons can reasonable a?ributed to States and UT
improving their ranking or deteriora?ng. The correla?on between NAS 2021 and NAS 2017 is
only 0.18. The highest correla?on across surveys, is 0.32 between NAS 2021 and ASER 2022,
with the correla?on between NAS17 and ASER 2016 marginally lower at 0.29. The correla?on
between NAS 2021 and NAS 2017 is surprisingly low at 0.18, probably because of the
disrup?on caused by the pandemic in 2020 and 2021.
29
This suggests that even though na?onal
averages matched, state-level pa?erns of performance diﬀer across surveys, reﬂec?ng
varia?ons in test design, proﬁciency standards, and learning assessments.
Table 27: Cross-correla?on Matrix- Std/Grade 5 MRP across States

Source: ASER reports 2022, 2018 and 2016; NAS 2017 and 2021
Table 28 presents a similar correla?on matrix for Minimum reading proﬁciency among
Std/Grade 3 students, based on the indicators iden?ﬁed in Table 16. This includes inter-State
data from FLS, ASER and NAS. Minimum Reading Proﬁciency, as measured by ability to
read/understand a Std I level text is highly corelated across all ASER surveys, ranging from a
high of 0.84 an 0.86. This indicates a fairly high degree of consistency, in which varia?ons can
reasonable a?ributed to States and UT improving their ranking or deteriora?ng. The
correla?on between NAS 2021 and NAS 2017 is 0.27, a li?le higher than the correla?on for
Grade 5 (Table 28).
Cross-survey correla?ons are less than half those between diﬀerent ASER surveys higher than
between the two NAS surveys (2021 & 2017). NAS 2021 and ASER 2022 have a correla?on of
0.44 for Std/Grade 3 results, signiﬁcantly higher than the corresponding for Std/Grade 5. FLS
2022 and ASER 2022 have also had a correla?on of 0.39, but FLS correla?on with NAS 2021 is
half that at 0.18. This indicates signiﬁcant diﬀerences in ranking of States for the same year
(Table 28). Even though na?onal averages matched reasonably well for ASER 2022 and FLS

29
The relevant cross-correla?ons between NAS Maths proﬁciency and ASER MAP (division) and MAP (subtrac?on) are all nega?ve, ranging
from (-)0,13 to (-)0.40, sugges?ng conscious eﬀorts to upgrade those students who are weak in maths to help them pass the grade and move
on to higher grades.
ASER 2022 NAS 2021 ASER 2018 NAS 2017 ASER 2016
ASER 2022 (Std2 Text) 1.00 0.32 0.85-0.02 0.78
NAS 2021 (Proficient) 0.32 1.00 0.18 0.18 0.23
ASER 2018 (Std2 Text) 0.85 0.18 1.00 0.18 0.90
NAS 2017 (Proficient)-0.02 0.18 0.18 1.00 0.29
ASER 2016 (Std2 Text) 0.78 0.23 0.90 0.29 1.00
51

2022 and NAS 2021, state rankings s?ll diﬀer signiﬁcantly across surveys, reﬂec?ng varia?ons
in test design, proﬁciency standards, and learning assessments.
30

Table 28: Correla?on Matrix - Grade 3 MRP across States:

Source: FLS (2022), ASER reports 2022, 2018 and 2016; NAS 2017 and 2021
5.6.3 Enrolment Rates across States (2024-25)
As data on comple?on rates are not available for states, the paper uses enrollment ra?o as
a proxy for comple?on rates, to analyses inter-State performance. Two types of es?mates are
available for primary school enrolment (ages 6-10). Age-Speciﬁc Enrolment Ra?o (ASER) and
Adjusted Net Enrolment Ra?o (ANER). ASER is the total number of pupils enrolled in primary
school (Classes 1 to 5) who are of ages 6 to 10, expressed as a percentage of the popula?on
aged 6 to 10. ANER diﬀers from ASER in primary, with ra?o higher than 100%, rounded down
to 100%.
Figure 17 plots Age-Speciﬁc Enrolment Ra?o (ASER) for 2024-25 across all States and Union
Territories, arranged according to their per capita Net State Domes?c Product (PC NSDP) for
2023-24, since PCNSDP data for 2024-25 was not available for many States.
Figure 17: Inter State varia?on of Primary Enrolment Ra?o with Per Capita NSDP

Source: UDISE 2024-25 & RBI, Na?onal Sta?s?cs Oﬃce (NSO), Latest Updated on: Aug 29, 2025

30
The relevant cross-correla?ons between FLS numeracy and ASER are nega?ve Nos 11-99, but posi?ve for subtrac?on. The corresponding
correla?ons for NAS 2021 are posi?ve but low. sugges?ng less problems in tes?ng an evalua?on in grade 3 Math than in grade 5 Math results
from NAS 2021. The results from FLS are also quite variable, and therefore non-comparable.
Grade 3
FLS 2022ASER 2022NAS 2021ASER 2018NAS 2017ASER 2016
FLS 2022 (Exceeds)1.000.390.180.22-0.11 0.21
ASER 2022 (>Std1 Text)0.391.000.440.840.050.84
NAS 2021 (Proficient)0.180.441.000.380.270.32
ASER 2018 (>Std1 Text)0.220.840.381.00-0.01 0.86
NAS 2017 (Proficient)-0.11 0.050.27 -0.01 1.000.19
ASER 2016 (>Std1 Text)0.210.840.320.860.191.00
y = 9.1102ln(x) -15.356
R² = 0.2877
60
65
70
75
80
85
90
95
100
105
2500075000125000175000225000275000325000
Age Specific Enrolment Rate
PcNSDP (At Constant Prices) (₹) 2023-24
Primary ASER
52

The regression trendline between ASER and PC NSDP is shown in Figure 17, with an R² of 0.29.
Sixteen States & UTs are above the regression trendline equa?on, out of which twelve have
already reached 100% enrolment at the primary level (Figure 17).
31

Using the regression equa?on from Figure 17, an expected/benchmark primary ASER is
calculated for each State based on its PCNSDP, and the gap is deﬁned as the diﬀerence
between actual and expected/benchmark enrolment. The States are then grouped by over
performance (posi?ve gap) and underperformance (nega?ve gap), with States having a gap of
+/- 2.6% [= (std.dev) 7.8%/3) points deﬁned as having li?le or no gap (Table 29).
Table 29: Compara?ve Performance of States in Primary Enrolment

Source: Author Calculation based on UDISE+ 2024-25 & RBI, Na?onal Sta?s?cs Oﬃce (NSO), Latest Updated on: Aug 29, 2025.
Note: PC NSDP is for 2023-24. Gaps are in bracket.
Among States & UTs with Per capita NSDP less than Rs. 10,000/-, Assam (+7.9), Tripura
(+10.7), Jammu & Kashmir (+12.9), Meghalaya (+13.2), and Manipur (+14.3) had higher
enrolment than their expected or benchmark rates. (1
st
row, 3
rd
column of Table 29). Among
States with per capita NSDP of more than Rs. 10,000/ Punjab (+8.1), Himachal Pradesh (+6.5),
and U?arakhand (+6.7) performed be?er than expected of their PC NSDP. Though major
States above the na?onal average of 83.2% include, Tamil Nadu (88.9%), Karnataka (90.8%),
Gujarat (91.7%), Andhra Pradesh (86.5%) and Odisha (85.8%), their enrolment ra?o is below
the rate expected of States at their level of per capita NSDP.
The correla?on between ASER and PCNSDP is a moderate 0.43, sugges?ng that be?er-oﬀ
States have higher primary enrolment rates. This compares with a correla?on of 0.16 between
Minimum Reading Proﬁciency (MRP) and Per capita NSDP and of -0.08 between MAP and
PCNSDP (sec?on 5.6.1). This reinforces our earlier conclusion that minimum learning
outcomes, depend on pedagogy, teaching at the right level, and regular assessment, not on
income or expenditures.

31
The log regression line goes above 100% at PCNSDP of ~32,000, so we truncate it at 100% to calculate gaps.
Negative GapLittle or No gapPositive Gap
UP (-18.2%)WB (-1.5%)Assam (+7.9%)
MP (-11.4%)ChhattisG (-1.3%)Tripura (+10.7%)
Bihar (-8.6%)Rajasthan (+0.1%)J&K (+12.9%)
Odisha (-3.7%) Meghalaya (+13.2%)
Jharkhand (-3.4%) Manipur (+14.3%)
Nagaland (-3.4%)
Haryana (-14.7%)Delhi (-2.6%)Telangana (+5.3%)
Sikkim (-7.9%)Goa (0%) MH (+5.9%)
Tamil Nadu (-6.0%)Puducherry (-0.1%)HP (+6.5%)
AP (-5.5%)Kerala (+0.6%)Mizoram (+6.6%)
Karnataka (-4.7%)A&Nislands (+0.6%)UttaraK (+6.7%)
Gujarat (-3.9%)Chandigarh (+1.9%)Punjab (+8.1%)
Arunachal (-3.1%)
PcNsdp >
₹10,000
PcNsdp <
₹10,000
53

5.7 Programs and Pla?orms
Central Govt has made an eﬀort over the past decade or so, to nudge states to improve their
primary learning outcomes. These include Central Sector schemes, digital pla?orms and Public
private partnerships (PPP) with industry and Non-Proﬁt organiza?ons (NPO/NGOs). Some
States and Private non-proﬁts have also ini?ated schemes and eﬀorts on their own or in
coopera?on with States. This sec?on presents some of these programs and pla?orms relevant
to Primary educa?on, along with available academic reviews.
32

5.7.1 DIKSHA
Digital Infrastructure for Knowledge Sharing (DIKSHA, 2017) is a na?onal digital pla?orm
which provides Open Educa?onal Resources for teachers, teacher educators, and students
across all stages of school educa?on. It oﬀers 7452 QR-linked energized textbooks, curriculum-
aligned 3.78 lakh digital e-content, including interac?ve lessons, prac?ce exercises, teacher
training modules, and assessment tools, available in mul?ple Indian languages (36). To date it
features 19,673 courses, records 14.58 Cr comple?on, and has issued 12.5 Cr cer?ﬁcate.
33
It
reached 2.5 crore teachers, 25 crore students, and 50 lakh parents in all 36 states and union
territories (UTs) as of 2023, with 10 million app downloads and 3 crore daily hits at the usage
peak in 2020-21 (MoE, 2023). Oﬄine access to the pla?orm, in the form of QR-coded books
and downloadable content on SD cards, widens the reach in low-connec?vity loca?ons,
accoun?ng for 60% of rural users (UNESCO, 2023). DIKSHA’s implementa?on in state
educa?on systems (36), such as 90% primary and 78% secondary teacher adop?on in Tamil
Nadu, and 95 lakh learning sessions in U?ar Pradesh, indicates robust localized scale
(Ramanujam, 2019; Times of India, 2024).
Banerjee and Biswas, (2025)] used mixed method study es?mates, cost eﬀec?veness and
impact using quan?ta?ve evidence
34
. Their “ﬁndings oﬀer DIKSHA's reach of 27.5 crore users
and SWAYAM's 3 crore enrolments, at cost eﬀec?veness of ₹545 and ₹3,333 per user,
respec?vely. Both pla?orms enhance teacher competency (20-30% improvements) and
students' pass percentages (10-15%), enabling economic inclusion for marginalized and rural
teachers”. They also conﬁrmed that “SWAYAM’s higher competency gains (27.5% vs. 22.5%),
with a sta?s?cally signiﬁcant diﬀerence, and that SWAYAM’s mean gain exceeds DIKSHA’s by
4.12-5.88%. This aligns with SWAYAM’s quality focus and structured MOOCs”
Compared to global pla?orms like Coursera (₹5,000-10,000 per course) or Khan Academy (free
but less localized), DIKSHA and SWAYAM oﬀer cost-eﬀec?ve, context-speciﬁc solu?ons.

32
Other schemes like NISHTHA and Na?onal Digital library of India cover all levels of educa?on while Padna Likhna Abhiyan covers adult
educa?on and are detailed in sec?on 5.6
33
As of 7 Aug, 2025- h?ps://diksha.gov.in/data/
34
The authors use data such as Ministry of Educa?on sta?s?cs, DIKSHA dashboards, SWAYAM course reports, Economic Surveys, and NITI
Aayog documents, along with some inputs from CIET-NCERT.
54

However, reliance on government funding risks sustainability compared to private pla?orms’
revenue models (Mishra & Panda, 2022; World Bank, 2022)
35
.
Roul and Mohalik, (2025) surveyed 119 secondary school teachers to study the percep?ons of
quality of e-content on Diksha pla?orm. They found, “posi?ve percep?ons regarding content
alignment, engagement strategies, and accessibility, but highlighted challenges in inclusivity,
assessment integra?on, and mobile compa?bility. No signiﬁcant diﬀerences were observed in
teachers’ percep?ons based on gender or loca?on”. Some of their interes?ng results are
summarized in Table 30.
Table 30: Survey of percep?on of secondary teachers on Diksha content & pedagogy

Source: Source: Mr. Rajkishore Roul, Prof. Ramakanta Mohalik, 2025
5.7.2 NIPUN Bharat
Several schemes to help implement NEP 2020 have been ini?ated by the central government,
and some older schemes have been modiﬁed. Among these are,
Na?onal Ini?a?ve for Proﬁciency in Reading with Understanding and Numeracy (NIPUN Bharat
2021). The goal is to ensure every child a?ains founda?onal literacy and numeracy (FLN) skills
by the end of Grade 3, preferably by age 9, by 2026-27
36
. NIPUN focuses on teacher training,
development of quality learning materials (including digital content on DIKSHA), and regular
tracking of each child’s progress. NIPUN is implemented na?onwide through a ﬁve-?er
mechanism, with each state preparing its own ac?on plan and ac?vely monitoring progress.
5.7.3 PM eVidya
PM eVidya launched in 2020 integrates mul?ple components such as DIKSHA providing
curriculum-aligned e-content and QR-coded textbooks, 200+ dedicated DTH TV channels for
classes 1 to 12, SWAYAM MOOCs, and radio broadcasts to ensure educa?on reaches students
even without internet access. Special provisions are included for children with special needs,
including content for visually and hearing-impaired learners.

35
Banerjee, Anindita, and Rakheebrita Biswas. "Empowering Educators, Bridging Divides: Evalua?ng Diksha and Swayam for Digital Teacher
Training and Economic Inclusion in India." Interna?onal Journal for Crea?ve Research and Thoughts (IJCRT) 13.7 (2025): 17.
36
FLN means reading, wri?ng & basic math. It also provides addi?onal support to Classes 4 and 5 who need it to achieve founda?onal literacy.
Statements Yes (%)No(%)
Undecided(%)
Contents are aligned with learning outcomes94.1%3.4%2.5%
Contents depth are appropriate to learner needs87.4%8.4%4.2%
Contents are carried activities for further learning84.0%7.6%8.4%
Assessment is integrated with e-contents 67.2%16.8%16.0%
Promotes learner engagement 77.3%16.0%6.7%
Helps in retaining learner interest 79.8%11.8%8.6%
Contents encourages learners creativity 80.7%13.4%5.9%
The pedagogy and content are synchronised 77.3%11.8%10.9%
Content pedagogy appropriate to the student’s level85.7%6.7%8.6%
55

Kumar et al. (2024) surveyed 681 students from classes 1-12, as well as teachers and parents
across 13 states & 2 UT’s using an online ques?onnaire. The study found “higher educa?onal
TV viewership in rural areas (55.4%) than urban areas (36.9%), with higher viewership among
younger children (5-10 years), showing that visual media is more used by younger students”.
“Most teachers reported that the e-content was interes?ng (69.5%), provided comprehensive
knowledge (67.9%), and supported teaching -learning (65.0%) and professional development
(69.5%)”. A high share of parents (73.3%) reported that e-content supported teaching
learning. Students found the content interes?ng (67.7%) and useful for learning (77.7%).
61.7% of teachers and 37.9% of students reported that e-content reduced the need for tui?on.
The authors suggested improving reach through age-speciﬁc and pedagogically aligned,
engaging content.
37

5.7.4 E-Pathshala
E-Pathshala (2014) provides free digital access to NCERT textbooks, audio, video, e-books,
and other mul?media resources for classes 1 to 12 in English, Hindi, and Urdu. Bansal (2022)
ﬁndings suggested that the quality of content on ePathshala app is accurate and useful for the
target audience. The design of the app is user-friendly with eﬀec?ve naviga?on and
integra?on system. The search and ﬁlter op?ons make the app easily accessible. However,
author pointed out some weaknesses such as insuﬃcient technical support (only email, no
phone-in or real ?me chat), and no demonstra?ve tutorial.
5.7.5 Samagra Shiksha Scheme
Samagra Shiksha Scheme (2018) integrates Sarva Shiksha Abhiyan (SSA), Rashtriya Madhyamik
Shiksha Abhiyan (RMSA), and the Teacher Educa?on program, into a single scheme to improve
pre-primary to senior secondary educa?on. It proposes to improve learning outcomes
through be?er school infrastructure, teacher training, digital tools, voca?onal training and
sports. It will also strengthen digital ini?a?ves such as Shala Kosh, Shagun, Shaala Saarthi and
DTH channels.
38

Sharma et al. (2018) surveyed 50 teachers from government secondary schools across ﬁve
blocks of Pratapgarh district, Rajasthan
39
. It found, “90% of teachers agree that DTH channels
and portal content are eﬀec?ve, and 86% observed improvement in students’ performance,
skills and knowledge”. 56% of teachers reported that smart classrooms improve teaching
quality and save ?me, and parents reported posi?ve views based on students’ performance.
The authors suggested improvement in internet connec?vity, provision of more teaching
materials, and strengthening technical support for teachers.

37
Kumar et al. (2024) reported that PM e-Vidya was eﬀec?ve in suppor?ng students’ learning during COVID-19.
38
PIB report, 19 July 2018
39
Total 50 sample were taken involving teachers and head masters of secondary and senior secondary government schools who are teaching
under the scheme. The sample were taken across 5 blocks of Pratapgarh district - Chho?sadri, Arnod, Pipalkhunt, Dhariyawad and Pratapgarh
(10 schools from each block).
56

5.7.6 PRATHAM: Teaching at the right level (TaRL)
Pratham’s ASER was one of the ﬁrst to iden?fy the gap between the no?onal and actual
learning levels in elementary schools (classes 1-8). To address this gap, Pratham developed
the Teaching at the Right Level (TaRL) method in the early 2000s. TaRL begins instruc?on at
the child’s actual learning level, not grade. Children are grouped by ability, and teachers use
simple, engaging ac?vi?es to build founda?onal reading and arithme?c skills. Regular
assessments track progress, and children are regrouped as they advance. This method helps
children catch up quickly and lays the founda?on for future learning.
TaRL is implemented in two ways, the direct model and the partnership model. In the direct
model, Pratham instructors run “Learning Camps” las?ng 30-50 days, with 3 hours of daily
instruc?on. These camps focus on founda?onal skills and are conducted during school hours,
vaca?ons, or in community se?ngs. In the partnership model, government teachers use TaRL
approach over (4-6 months) with dedicated ?me for 2-2.5 hours a day. Con?nuous monitoring
and support ensure eﬀec?veness.
Pratham’s TaRL program has been implemented at scale in U?ar Pradesh, Bihar, Haryana,
Andhra Pradesh, Karnataka, Delhi, and a few districts of Maharashtra, Madhya Pradesh,
Arunachal Pradesh and Daman and Diu Union Territory. The ALJ Poverty Ac?on lab (J_PAL) at
MIT has partnered with Pratham since 2001 to evaluated the TaRL approach, from the
perspec?ve of scalability. Six randomized evalua?ons across seven states in India showed that
the TaRL is consistently eﬀec?ve when implemented systema?cally. It shows the largest eﬀect
sizes recorded in educa?on research.
In U?ar Pradesh, high intensity, short dura?on, learning Camps for 40 days, during school
hours, in 2013-14, with addi?onal 10-day summer camp. improved learning outcomes in
language and mathema?cs by 0.6 to 0.7 standard devia?on. Further, the share of children
unable to recognize le?ers declined from 34% to 8% in the interven?on group, and the share
of children who could read a paragraph or story increased from 15% to 49%, about double the
rate compared to control schools. UP State later scaled up TaRL through the Graded Learning
Program (GLP) across all primary schools in 2019, improving reading and numeracy skills in
Grades 1-5. Within 3 months of implementa?on, over 1.7 million children in grades 4 and 5
were able to read basic grade 1 level Hindi text. J-Pal research shows that TaRL can deliver
signiﬁcant gains at less than $10 per child per year, making it aﬀordable for large-scale
adop?on.
In Haryana, the Partnership model was applied in 2012-13. Government teachers
implemented Read India TaRL during a dedicated school hour for Grade 3-5 across 400
schools, supported by trained Associate Block Resource Coordinators (ABRCs). This resulted
in a 0.15 standard devia?on gain in language scores (Banerjee et al., 2016). TaRL is recognized
globally as a highly cost-eﬀec?ve interven?on. The Global Educa?on Evidence Advisory Panel
(2023) listed TaRL among the “Great Buys” for improving learning outcomes.
57

Duﬂo, Dupas, and Kremer (2011) found that eﬀec?ve teaching prac?ces such as grouping
students by ability and providing targeted support have shown strong results in countries like
Ghana and Kenya. In rural Kenya, separa?ng primary students into groups based on their ini?al
ability led to sizable gains in math and language for both high and low achievers, allowing
teachers to teach at a level more appropriate to children’s needs. Similarly, Duﬂo and Kiessel
(2012) found that, in Ghana, supplemen?ng teachers with community assistants to help the
weakest students led to sizable gains in literacy and numeracy, especially when done aLer
school. Further, training teachers to teach students in small groups, targeted to their learning
level, boosted their literacy skills.
Muammar et al. (2023) examined the eﬀect of the combina?on of Indonesia’s “Innovasi
learning materials with Teaching at the Right Level (TaRL) approach, on ini?al reading skills of
ﬁrst-grade students in a public Islamic school in Mataram City, Indonesia. Ini?ally, a large
number of students struggled with basic reading. The researchers implemented the TaRL
method in two phases (cycles), each including stages of planning, teaching, observa?on, and
reﬂec?on. By adjus?ng teaching to match the students' actual reading levels, the number of
students reaching the desired reading competency increased signiﬁcantly from around 52% in
the ﬁrst cycle to nearly 87% in the second phase. Teachers also became more eﬀec?ve, and
student engagement improved. The study concluded that TaRL, especially when supported
with addi?onal learning materials, is an eﬀec?ve way to improve founda?onal reading skills
in early grades.
5.7.7 Educate Girls (NGO)
Educate Girls (2007) is an NGO which works through a network of village-based volunteers
and gender champions called “Team Balika”, who ensure that girls are learning well, by
oﬀering supplemental classes. The program focuses on “child-centric learning and teaching
techniques” that are “ac?vity-based” and centered on “playful learning.” These Crea?ve
Learning and Teaching (CLT) techniques include two main methods: “Catch Up” methodology
that uses an ac?vity-based learning approach to help children who are lagging behind, and a
“Peer Group Learning” methodology that involves children working and learning together. It
ensures that children ﬁnd learning fun and thereby take interest and ini?a?ve in the process.
Educate Girls also trains grade 3-5 teachers in CLT. The program uses pre-test and post-test
assessments across ﬁve learning levels (A, B, C, D and E), in Hindi, English and Mathema?cs
This structure is similar to ASER’s reading and arithme?c levels.
40

40
(E) No response. (D) Alphabet reading in Hindi and English/ Single digit iden?ﬁca?on in Maths (C) Word reading in Hindi and English/ Two-
digit iden?ﬁca?on in Maths (B) Sentence reading in Hindi and English/ Two-digit addi?on and subtrac?on in Maths A) Story reading in Hindi
and English/ Two-digit division and mul?plica?on in Math.
In 2014-15, Educate Girls has implemented CLT in 6 districts of Rajasthan. For measurement of CLT impact, pre and post tests have been
conducted in 108 control schools. 1,47,281 students beneﬁ?ed from the CLT program, including 68,435 boys and 78,846 girls in class 3rd, 4
th

and 5th and in Hindi, English and Mathema?cs.

58

5.7.8 SwiLPAL
Adap?ve learning systems such as SwiLPAL (Personalized Adap?ve Learning), developed by
ConveGenius, help teachers iden?fy learning gaps and plan diﬀeren?ated instruc?on.
SwiLPAL is rooted in the science of learning and evidence-based pedagogy and is built on
mastery learning. Its Pedagogy is based on learning one micro-skill at a ?me, moving from
Learn to Prac?ce to Master to Advance. The Delivery System (Adap?ve Tech) iden?ﬁes the
right learning gaps, guiding teachers with personalized Teaching Learning Materials (TLM) and
insights. The Experience (Joyful Learning) uses game-like missions, badges, and streaks that
feel like play but work like science. SwiLPAL has reached 22 states, 176 districts, 15,000+
schools, and over 2 million students, at an es?mated cost of INR 1,700-2,100 per child per
year.
Its impact has been evaluated through an RCT (by M Cramer) in government schools of Andhra
Pradesh. The study found that aLer 17 months, the gain in test scores in the [PAL] schools
corresponded to almost an addi?onal 2 years of extra learning beyond what would have been
expected over the period, with students learning 2.3x faster compared to peers Importantly,
SwiLPAL integrates within regular school ?metables, and even aLer replacing two weekly
periods per subject, students performed just as well in school exams as their peers.
In Rajasthan (Mission Buniyaad PAL Program), ICT labs-based PAL program resulted in 2x
higher learning gains; 6 months of extra schooling in 6 months. In U?ar Pradesh (Outcome-
Linked PAL Program) resultant into 3x-8x higher learning gains with approximately 2.1 years of
addi?onal schooling in 22 months. It has also been implemented in states such as Gujarat,
Maharashtra, and Telangana.
5.7.9 Other AI tools
Tools such as PD 360 provide expert-led professional development resources that can be
integrated into formal training programs and coaching cycles. Research-based systems like
FACET Mul?-Agent AI illustrate how AI can assist teachers in designing diﬀeren?ated
instruc?on while maintaining pedagogical control. General AI assistants such as ChatGPT for
Teachers and Google Gemini for Educa?on support lesson planning, assessment design,
feedback genera?on, and content adapta?on.
The Teacher App in India provides modular micro-learning courses on pedagogy, classroom
strategies, and prac?cal teaching techniques. Tools suppor?ng early math thinking, such as ST
Math and QANDA, provide step-by-step problem solving and adap?ve feedback, which
teachers can use to guide conceptual discussions in classrooms. Eduaide.AI help teachers to
integrate real-world examples and prompts that encourage reasoning.
5.8 Summary and Conclusions
Basic educa?on, especially literacy and numeracy, forms the founda?on of the educa?on and
skill development system. Further educa?on and skills are built on this founda?on.
Founda?onal literacy and numeracy (FLN) are essen?al for employment in today’s economy,
59

as stated in NEP 2020. They are par?cularly important for casual labor which cons?tutes 1/5
th

of workers.
Comparison of India with interna?onal benchmarks which the paper has used, show that
India’s performance is reasonably good. The primary comple?on rate was 93.5% in 2023,
11.9% points above the benchmark for its Per capita GDP in 2023 (81.6%). Other available
indicators are somewhat older but paint a similar picture. The Net Enrolment Rate was 92.3%
in 2013, about 6.8% points above the benchmark (85.4%) for that year’s per capita GDP. India’s
share of pupils achieving minimum reading proﬁciency was 46.3% in 2017, 5.3% points above
the benchmark (41.0%), and reading proﬁciency, adjusted for out-of-school children, was
43.9%, 6.3% points above the benchmark (37.7%). India’s pre-primary Gross Enrolment Rate
was 62.8% in 2019, 8.3% points above the benchmark (54.8%).
Domes?c surveys show small gaps between urban and rural outcomes. In 2021, rural students
performed be?er than urban students by 12-13% in primary school. Rural students were also
be?er in class 3 language (+5%), but worse in class 5 language (-5%) In 2024, rural students
performed marginally be?er than urban students in Language and iden?cal in Mathema?cs
in Grade 3. We have therefore assumed that ASER Rural learning outcome surveys are broadly
reﬂec?ve of overall learning outcomes.
Based on ASER (2024), all India, primary learning outcomes are as follows. The number of
children with Minimum Reading Proﬁciency, deﬁned as the propor?on of children who can
read a Std II level text in Primary school (Std V), are 48.7%. Those with Minimum Arithme?c
Proﬁciency (MAP), deﬁned in terms of the propor?on of children who can do division, are
30.7%. There is thus an 18 percent points diﬀerence between MRP and MAP (i.e., 36%). We
also es?mate that MRP in 2024 is s?ll 2% above the interna?onal benchmark of 47.9% for
India at the Per capita GDP of $9818 in 2024.
41

NEP 2020 recognized the role of mathema?cs and mathema?cal thinking for India’s future
workforce, par?cularly in emerging ﬁelds such as ar?ﬁcial intelligence, machine learning, and
data science. The NEP therefore says that mathema?cs and computa?onal thinking should be
given increased emphasis, with a variety of innova?ve methods, including use of puzzles and
games, which make mathema?cal thinking more enjoyable and engaging (at the founda?onal
stage). The broader the arithme?c proﬁciency at the primary level, the more likely is this to
be achieved.
State level data (ASER 2024) shows that the following States have the highest number of
children with Minimum Reading Proﬁciency in Primary school (Std V): Mizoram (67.5%),
Himachal Pradesh (66.8%), Kerela (66%), U?arakhand (63.9%), Haryana (63.5%), Punjab (61%)
and Maharashtra (59.6%). The top ranked States for Minimum Arithme?c Proﬁciency (MAP)
are, Punjab (48.8%), Himachal Pradesh (47%), Mizoram (44.7%), Haryana (43.2%),
U?arakhand (39.8%), U?ar Pradesh (39.4%) and Andhra Pradesh (36.2%).

41
As a result of the pandemic MRP may have declined by 0.35% between 2017, the year of interna?onal comparison and 2024. Meanwhile
he Benchmark has increased to 47.9% because of the increase in per capita GDP at PPP (2021 int dollars) to 9818
60

The poten?al challenge is to raise the MRP in primary school (Std V) by at least 2.6% points
in the next 5 years to reach the UMIC benchmark of 54.2%, and by 27.6% points to reach the
minimum HIC benchmark of 79.2%, in 25 years. This will require the eﬀorts of all States. Given
that about 14 (of 26) states and UTs were below the Na?onal es?mate in 2024, increasing
Minimum Reading proﬁciency remains a challenging task.
Experience from India and other countries shows that improving learning requires three
main changes: revising curriculum to focus on measurable learning outcomes, retraining
teachers in new approaches, and conduc?ng regular tests to monitor student progress and
give targeted support. To improve these outcomes, the quality of teaching and learning needs
to improve, with the full backing and support of the school administra?on. Government
programs like NEP 2020, NIPUN Bharat, and NISHTHA aim to address this by focusing on
educa?on quality. Digital pla?orms such as DIKSHA help support learning and teacher training.
Research and personal experience suggest, that the State primary educa?on systems in India
should pay more a?en?on to the following elements of primary and pre-primary teaching to
improve learning outcomes.
(1) Age-appropriate teaching material is cri?cal for learning. This includes paper-based
material like ﬂash cards used in TaRL, but cannot be limited to it. It can include singing,
music, radio and sound recordings, TV/DTH/DTS and video recording, educa?onal toys,
and even local materials (e.g., beads) for coun?ng.
(2) Teachers have to be familiarized with these materials and taught how to use them for
age-appropriate teaching. School administra?ons have to be sensi?zed to these non-
conven?onal methods, so they support teachers in the use of these materials.
Research shows that children who are subject to a lot of sound, music and
conversa?on become more proﬁcient in language.
(3) Tes?ng and re-tes?ng is very important for iden?fying lagging students and providing
speciﬁc directed assistance to them., and measuring their progress.
(4) Social, Civic Sense and self-discipline is cri?cal to learning and for future success. The
Japanese method for enhancing these quali?es should be adapted to Indian
condi?ons, and adopted in pre-primary and primary school.
(5) Rote learning to thinking: Awareness of the world and the people around us is the
founda?on of self-learning. The next level is curiosity about the environment around
us. Curiosity needs to be aroused by poin?ng to the puzzles in the world around her
(nature, animal economic ac?vi?es, people). The next stage is asking ques?ons and
ﬁnding answers. It’s a process, which can start by just looking closely at nature (trees,
leaves, ﬂowers, twigs).
(6) Physical sports, are key to developing hand eye coordina?on, mind body integra?on
and learning how to socialize. The simplest least costly games like racing, marbles,
“gulli-danda”, “pithu,” rope pulling, sack races and hop-skip & jump can be as eﬀec?ve
for younger children, as “kabaddi”, football, cricket, basketball for older ones.
61

6. Secondary Educa?on
A strong educa?onal pyramid can only be built, if the founda?on provided by Primary
educa?on is strong. Secondary educa?on plays a crucial role in building upon the founda?onal
skills acquired during primary schooling and acts as a gateway to higher educa?on and
employability. The latest posi?on with respect to interna?onally comparable data on
secondary enrolment and comple?on rates is summarized below.
The Gross enrolment ra?on in secondary school in 2024 was 78.1% of student in relevant age
group (Table 31). The lower secondary comple?on rate (Class/Std 8) was 87.3%. Between 2014
and 2024, the secondary enrolment rate increased by 3.2% points, while the lower secondary
comple?on rate increased by 1.4% points. While the improvement is modest, it reﬂects
con?nued eﬀorts by States.
Table 31: Improvement in Secondary enrollment & Comple?on (%)

Source: World Development Indicators, July 2025
The ﬁrst subsec?on examines the transi?on from primary to secondary school through the
lens of progression rates. The second sub-sec?on compare India’s performance in secondary
educa?on across countries, using available data on secondary school enrollment and
comple?on rates. The third sub-sec?on analyses changes in minimum learning outcomes
(MRP & MAP).
6.1 Transi?on from Primary to Secondary School
The “progression rates”, from primary to secondary school, for the latest available year, is
plo?ed for each country at its level of Per capita GDP at PPP (in constant 2021 interna?onal
dollars), in the same year. A regression equa?on is then ﬁ?ed to the interna?onal compara?ve
data to represent the benchmark or expected rates at each level of per capita GDP (Figure
18).
42
India’s progression rate from primary to secondary school was 91.3% in 2017, which is
2% points above the benchmark/expected level at the per capita GDP of India in 2017. China
(+9.3), Vietnam (+9.3), Thailand (+3.8), and Mexico (+1.5) were also above their benchmark
levels, while Indonesia was slightly below (-)0.1.

42
Progression to secondary school refers to the number of new entrants to the ﬁrst grade of secondary school in a given year as a percentage
of the number of students enrolled in the ﬁnal grade of primary school in the previous year (minus the number of repeaters from the last
grade of primary educa?on in the given year).
Education IndicatorsChange
Secondary enrolment rate (% gross)202478.1201474.93.2
Lower sec completion rate 202487.3201485.91.4
Latest available yearYear for comparison
62

Figure 18: Progression to secondary school

Source: World Development Indicators, December 2024
The base for India in Figure 18 is 2017, therefore we es?mate the value for 2023 assuming
that the gap remains the same (2.1% points). This gap is then added to the expected
expected/benchmark (90.4%) for India at its 2023 per capita GDP to get an es?mate for 2023
(92.5%). Based on this we es?mate the improvement needed to reach UMIC & HIC
benchmarks. As the es?mated 2023 number is above the UMIC benchmark (purple diamond),
we need an improvement of ~5.4% points to reach the HIC benchmark of 97.9% for HIC by
2050.
43

6.2 Enrollment & Comple?on Rates
India’s Net Enrolment rate in secondary school was 61.6% in 2013, which is 5.8% points above
the level expected at its per capita GDP. Indonesia (+10.1), Thailand (+1.5) and Mexico (+2.7)
are also above their expected levels (leL panel in Figure 19). As the base for comparison in
Figure 19 is 2013 for India, we es?mate the value for 2023 (69.5%) assuming that the gap
remains the same (5.8% points). This gap is then added to the expected/benchmark for India
at its 2023 per capita GDP (63.7%). Based on the es?mate for 2023 (69.5%), we are already
above the UMIC benchmark. A 17.6%-point improvement is needed over the next 25 years to
reach the HIC benchmark of 87.1% by 2050.

43
The UMIC and HIC benchmarks are based on the Per capita GDP PP of Indonesia which became a UMIC in 2019, and Romania which
became a HIC in 2019.
India (91%)
China
VietNam
Indonesia
Thailand
Mexico
UMIC (92%)
HIC (98%)
R² = 0.3147
85
87
89
91
93
95
97
99
101
0100002000030000400005000060000
% of students
PcGDP , PPP (const 2021 int $)
Progression to secondary school (%)
63

Figure 19: Secondary School enrollment: NER/GER

Source: World Development Indicators, December 2024
Since Gross secondary enrollment (GER) is available for the latest year (2023) for India, we
also do an interna?onal comparison using GER. India’s gross secondary school enrollment
rate in 2023 was 78.9% (Figure 19, 2
nd
panel), 4.6 percentage points above the level expected
at its per capita GDP. China (+15%), Vietnam (+18%), Indonesia (+17%), Thailand (+14%), and
Mexico (+11%). India is already near the UMIC (2030) benchmark of 79.2%. To reach HIC
(2050) benchmark of 94.8%, and will require only a 15.9% points improvement over 25 years.
India’s lower secondary comple?on rate was 85.5% in 2023 (Figure 20), which is 16.2
percentage points above the expected/benchmark level at its per capita GDP of 2023.
Indonesia (+23), Mexico (+11), China (+26), Vietnam (+24) and Thailand (+17) are also above
their expected/benchmark levels.
Figure 20: Secondary comple?on rate, total (% of relevant age group)

Source: World Development Indicators, December 2024; UNESCO Global Educa?on Monitoring Report (2023)
India (62%)
Indonesia
Thailand
Mexico
UMIC (69%)
HIC (87%)
R² = 0.6848
20
30
40
50
60
70
80
90
100
01000020000300004000050000600007000080000
Sec (% net)
PcGDP , PPP (const 2021 int $)
Secondary school enrollment (% net)
India (79%)
China
VietNam
Indonesia
Mexico
UMIC (79%)
HIC (95%)
R² = 0.5363
40
50
60
70
80
90
100
0100002000030000400005000060000
Sec (% gross)
PcGDP , PPP (const 2021 int $)
Secondary school enrollment (% gross)
India (85%)
Indonesia
Mexico
UMIC
(74.5%)
HIC (91%)
R² = 0.5292
30
40
50
60
70
80
90
100
01000020000300004000050000600007000080000
(% of relevant age group)
PcGDP , PPP (const 2021 int $)
Lower Secondary Completion Rate
India (63%)
Vietnam
Indonesia
Thailand
China
Mexico
UMIC (58%)
HIC (79%)
R² = 0.6821
20
40
60
80
100
0 20000 40000 60000 80000
% of cohort
PcGDP , PPP (const 2021 int $)
Upper Secondary Completion Rate
64

India already meets the UMIC (2030) benchmark of 74.5%, To reach the HIC (2050)
benchmark of 90.9% we need a 5.4 percentage points rise over 25 years. This should happen
in the normal course as quality of Primary educa?on improves and learning outcomes
improve.
India’s upper secondary comple?on rate was 63.0% in 2023 (Figure 20, 2
nd
panel), was 12.3
percentage points above the expected/benchmark level at its per capita GDP in 2023. Thailand
(+0.5) and China (+17.8) are also above their expected/benchmark levels, while Vietnam(-)5.9,
Indonesia (-)18.4 and Mexico (-)10.8 are below their expected/benchmark levels. India is
already above the UMIC (2030) benchmark of 57.6%. To reach the HIC benchmark of 79.4%
by 2050 requires a 16.4 percentage points rise over 25 years.
6.2.1 India’s schooling output rela?ve to China
Even with India’s school comple?on rates lower than higher income comparator countries,
Vietnam, Indonesia, Thailand, China and Mexico, the absolute numbers are comparable to
those of China and oLen exceed it. Table 31 presents the absolute numbers of children
comple?ng primary school, lower and upper secondary school, using popula?on
es?mates/projec?ons from the Technical Group on Popula?on Projec?ons (TGoPP) and the
UN Popula?on Division (UNpop). As UNpop popula?on es?mates are larger than TGoPP
es?mates, the es?mated school comple?on numbers are also higher for the former
Es?mated School Completers = Comple?on Rate (% of cohort) * Cohort Popula?on (million).
In 2023, India had an es?mated 131 million primary, 61 million lower secondary, and 94 million
upper secondary completers, using TGoPP (Table 32).
44
The number of children comple?ng
primary school were 137 million, and those comple?ng lower and upper secondary school
were 65 million, and 98 million, respec?vely.
Table 32: Numbers of Youth comple?ng diﬀerent educa?on levels

Sources: WDI/UNESCO (comple?on rates), Technical Group on Pop projec?ons (2020). na = not available.

44
Age groups (In years) and Corresponding level of educa?on- Age 6-10 as Primary (I-V), Age 11-13 as Upper Primary (VI-VIII), Age 14-15 as
Secondary (IX-X), Age 16-17 as Senior Secondary (XI-XII) and Age 18-23 as Higher Educa?on. Source: Educa?onal Sta?s?cs at a glance, Ministry
of Educa?on (2018)
CompletionPop agedCompleters Pop agedCompleters
Rates (%)(Millions)(million)(Millions)(million)
Primary Education (population aged 6-11)
2023 93.5140131146137
2030 (proj)84.1133111136115
2050 (proj)92.3 nana121112
Lower secondary education (popuation aged 12 to 14)
2023 85.5 71617665
2030 (proj)74.5 69517153
2050 (proj)90.9 nana6256
Upper secondary education (population aged 18to 23)
2023 63.01509415698
2030 (proj)57.61418115086
2050 (proj)79.4 nana131104
Technical group on Pop
UN Population
65

These numbers are compared with those of China and other comparator countries in Table
32. At the primary level, India had 1.25 ?mes completers (137 million) than China (109 million)
and more than 2.5 ?mes the combined total of Vietnam, Indonesia, Thailand, and Mexico
(MIC4). At the lower secondary level, India’s 65 million completers were 1.2 ?mes those of
China 53 mil) and 2.4 ?mes those of the MIC4.
These comparisons highlight that, despite lower comple?on percentages, India contributes a
large share of the global pool of school completers due to its popula?on size. At the upper
secondary level, India’s comple?on rate is higher than that of Vietnam, Indonesia, and Mexico.
Number of children comple?ng secondary school in India are 1.2 ?mes China and 3.7 ?mes
those of the MIC4 (Table 33).
Table 33: Youth Comple?ng diﬀerent levels - India, China, MIC4

Sources: WDI/UNESCO (comple?on rates), Technical Group on Pop projec?ons (2020).
6.2.2 Secondary Enrolment Rates across States
As data on comple?on rates for secondary schools are not available for states, the paper
uses enrolment ra?o as a proxy for comple?on rates, to analyses inter-State performance.
Secondary school enrolment is divided into two levels: Lower Secondary and Upper
Secondary. Two types of es?mates are available for lower secondary school enrolment:
Adjusted Net Enrolment Ra?o (ANER) for Classes 6-8 and Age-Speciﬁc Enrolment Ra?o (ASER)
for age 11-13. ASER is the total number of pupils enrolled in lower secondary (Classes 6 to 8)
of ages 11-13, expressed as a percentage of the popula?on aged 11 to 13. ANER is total
number of pupils enrolled in the Class 6-8 or a higher level of school educa?on corresponding
oﬃcial age group, expressed as a percentage of the popula?on of that age group in a given
school year. We truncate ANER ra?o above 100% to 100% to obtain an ANER’.
Figure 21, plots ANER’ and ASER for 2024-25 across all States and Union Territories, arranged
according to their per capita Net State Domes?c Product (PC NSDP) for 2023-24, since PCNSDP
data for 2024-25 was not available for many States. The regression trendline between ASER
and per capita Net State Domes?c Product (PC NSDP) is shown in Figure 21 (green), with an
IndiaVietnamIndonesiaThailandChinaMexico
Primary Education
Completers (million)1371024510913
Completion Rates(%)93.5100100100100100
Pop aged 6-11 (mil)1461024510913
Lower secondary
Completers (million)655142537
Completion Rates(%)85.510098.610010093.3
Pop aged 12-14 (mil)765142537
Upper secondary
Completers (million)984114838
Completion Rates(%)63.052.840.968.486.758.1
Pop aged 18-23 (mi)15682669613
66

R² of 0.25. The regression trendline between ANER’ and PCNSDP is also shown in Figure 21
(pink), with a stronger ﬁt and an R² of 0.51.
Using the regression equa?on from Figure 21, an expected or benchmark lower secondary
ANER’ and ASER is calculated for each State based on its Pc NSDP, and the gap is deﬁned as
the diﬀerence between actual and expected enrolment. States are then grouped by
over-performance (posi?ve gap) and under-performance (nega?ve gap), with States having a
gap of +/- 3.4% [= (std.dev) 10.1%/3) points deﬁned as having li?le or no gap (Table 34)
45
. The
interstate comparison of enrolment ra?o using ANER’ are diﬀerent from those using ASER.
Figure 21: Inter State varia?on of Lower Secondary Enrolment Ra?o with PCNSDP

Source: UDISE 2024-25 & RBI, Na?onal Sta?s?cs Oﬃce (NSO), Latest Updated on: Aug 29, 2025.
Table 34: Compara?ve Performance of States in Lower secondary Enrolment (ANER’)

Source: Author’s Calc based on UDISE 2024-25 & RBI, Na?onal Sta?s?cs Oﬃce (NSO), Latest Updated on: Aug 29, 2025 Note: PC NSDP is for
2023-24. Gaps are in bracket.

45
Both ANER′ and ASER are given, as the consistency of the data was not clear. Their correla?on with PCNSD is 0.55 and 0.31 respec?vely.
y = -5E-10x
2
+ 0.0002x + 75.882
R² = 0.252
y = -9E-10x
2
+ 0.0004x + 47.319
R² = 0.5078
50
55
60
65
70
75
80
85
90
95
100
5000 55000 105000 155000 205000
Enrolment Rate (%)
PcNSDP (At Constant Prices; ₹) 2023-24
Lower Secondary Enrolment Rate (ASER & ANER')
ASERANER'
Poly. (ASER) Poly. (ANER')
Negative GapLittle or No gapPositive Gap
Nagaland (-16.2%)UP (-2.4%) MP (+3.9%)
WB (-5.6%)J&K (-1.8%)Odisha (+5.5%)
Jharkhand (-1.2%)Meghalaya (+5.9%)
Bihar (-0.8%)Rajasthan (+8.1%)
Manipur (+1.1%)ChhattisG (+13.4%)
Assam (+1.7%)Tripura (+14.2%)
Arunachal (-24.9%)Gujarat (-3.1%)AP (+3.5%)
Sikkim (-18.5%)Mizoram (-1.2%)Tamil Nadu (+4.5%)
Haryana (-10.4%)A&Nislands (+2.0%)Chandigarh (+9.3%)
Karnataka (-6.9%) Kerala (+10%)
Telangana (10.1%)
Delhi (+10.4%)
UttaraK (+11.0%)
MH (+11.1%)
Puducherry (+11.6%)
HP (+12.4%)
Punjab (+13.2%)
PcNsdp >
₹10,000
PcNsdp <
₹10,000
67

Using ANER’, Odisha (+5.5%), Rajasthan (+8.1), Chha?sgarh (+13.4), and Tripura (+14.2) were
above the enrollment rate (ANER’) expected of their PCNSD (among States with per capita
NSDP less than ₹10,000). Among States with per capita NSDP of more than Rs. 10,000/ Kerala
(+10), U?arakhand (+11), Maharashtra (+11.1), Punjab (+13.2) and Himachal Pradesh (+12.4)
performed be?er than expected at their PCNSDP (Table 34, 1
st
3 columns). Though major
States above the na?onal average of 77.9% include, Karnataka (84%), Gujarat (88%) and
Haryana (80%) their enrolment rate is below the rate expected of States at their level of per
capita NSDP’.
Table 35 presents the inter-State comparison of lower secondary age-speciﬁc enrolment Rate,
ASER. States are grouped based on the gap between actual and expected enrolment levels.
States with a gap within +/-2.6% [= (std.dev) 7.8/3] are considered to have li?le or no gap.
Tripura (+3.5), West Bengal (+7.8), U?ar Pradesh (+8.0), Meghalaya (+9.2), and Manipur (+11)
had an enrollment rate (ASER) above that expected of their PCNSDP. Haryana, Karnataka,
Maharashtra, Punjab, Himachal Pradesh show li?le or no gap (0.0), enrolment levels in line
with expecta?ons. Only Tripura and Meghalaya have a posi?ve gap under both ANER’ and
ASER.
Table 35: Compara?ve Performance of States in Lower secondary Enrolment (ASER’)

Source: Author’s Calcula?on based on UDISE 2024-25, Note: Gaps are in bracket.
Upper Secondary
The Age-Speciﬁc Enrolment Rate (ASER) is the total number of pupils enrolled in Upper
secondary (Classes 9 to 12) of ages 14-17, expressed as a percentage of the popula?on aged
14 to 17. The Adjusted Net Enrolment Rate (ANER) is the total number of pupils enrolled in
Negative GapLittle or No gapPositive Gap
Sikkim (-25.4%)Assam (-1.5%)Tripura (+3.5%)
Nagaland (-22%)Tamil Nadu (-1.4%)WB (+7.8%)
J&K (-14%) Kerala (-0.3%)UP (8.0%)
Bihar (-11%)Puducherry (-0.2%)Meghalaya (+9.2%)
Arunachal (-10.7%)Haryana (0.0)Manipur (+11%)
Gujarat (-9.7%)Karnataka (0.0)
MP (-8%) Mizoram (0.0)
Jharkhand (-6.6%)Chandigarh (0.0)
A&Nislands (-6.3%)Telangana (0.0)
AP (-3%) Delhi (0.0)
Rajasthan (-3%)UttaraK (0.0)
MH (0.0)
HP (0.0)
Punjab (0.0)
Goa (0.0)
Odisha (0.5)
ChhattisG (0.7)
68

the Class 9 to 12 or a higher level of school educa?on, corresponding to ages 14 to 17,
expressed as a percentage of the popula?on of 14 to 17-year-olds in a given school year.
Figure 22, plots both ANER and the ASER for 2024-25 across all States and Union Territories,
arranged in order of per capita NSDP for 2023-24, since PCNSDP data for 2024-25 was not
available for many States. The regression trendline between ASER and PCNSDP is shown in
Figure 22 (green), with an R² of 0.30. The regression trendline between ANER and Pc NSDP is
also shown in Figure 22 (pink), with a stronger ﬁt and an R² of 0.53.
46
On this basis, ANER is
taken as the more appropriate measure. As per ANER, nineteen States and Union Territories
lie above the regression trendline equa?on, i.e., are performing be?er than expected at its
respec?ve level of PCNSDP.
Figure 22: Inter State varia?on of Lower Secondary Enrolment Ra?o with PCNSDP

Source: UDISE 2024-25 & RBI, Na?onal Sta?s?cs Oﬃce (NSO), Latest Updated on: Aug 29, 2025
Using regression equa?on from Figure 22, an expected or benchmark ANER is calculated for
each State based on its Pc NSDP, and the gap is deﬁned as the diﬀerence between actual and
expected upper secondary enrolment. States & UTs are then grouped by over-performance
(posi?ve gap) and under-performance (nega?ve gap), with a gap of +/- 3.1% [= (std.dev)
9.2%/3] deﬁned as having li?le or no gap.
Among States with PCNSDP less than Rs. 10,000, West Bengal (+9.9), Tripura (+6.6), Manipur
(5.6), Rajasthan (+3.4%) had lower secondary enrolment rates higher than
expected/benchmark at their PCNSDP. Among States with per capita NSDP of more than Rs.
10,000/ U?arakhand (+4), Punjab (+7.9), Kerala (+18.6%) performed be?er than expected of
their PCNSDP. Though Telangana (63.8%) and Mizoram (61.2%) had enrolment rates above the
na?onal average of 52.2%, it was less than expected at their level of PCNSDP (Table 36).

46
The interstate correla?on between PCNSDP and ANER’ is 0.69 and between PCNSDP and ASER is 0.52
y = 12.775ln(x) -85.506
R² = 0.3019
y = 17.963ln(x) -152.48
R² = 0.5305
30
40
50
60
70
80
90
100
0 50000100000150000200000250000300000350000400000
Enrolment Rate (%)
PcNSDP (at constant prices; ₹) 2023-24
Upper Secondary Enrolment Rate (ASER & ANER')
ASERANER
Log. (ASER) Log. (ANER)
69

Table 36: Compara?ve Performance of States in Upper Secondary Enrolment

Source: UDISE 2024-25 & RBI, Na?onal Sta?s?cs Oﬃce (NSO), Latest Updated on: Aug 29, 2025
6.3 Learning Outcomes and Tradi?onal Results
As shown in Table 7, Surveys which measure learning outcomes in secondary school cover
diﬀerent grades. ASER surveys (2006-2024) cover Standards 6, 7 and 8, while NAS (2021 &
2017) surveys cover Class 8 and 10 only. The PRS (2024) survey covers Grade 6 and 9, but does
not give us data on minimum learning outcomes. As indicated earlier, ASER, presents the
percentage of students who can read a Std II level text in language and those who can solve
division problems in mathema?cs. ASER uses the same test level (Std II) for all grades, so it is
the most useful for ge?ng the current picture as well as changes over ?me. The next sub-
sec?on therefore covers the learning outcome from ASER surveys. ThereaLer, learning
outcomes for grade 8 are compared for NAS and ASER surveys.
6.3.1 Learning Outcomes: Lower secondary school
Das and Zajonc (2010), who used data from a 2005 World Bank-led survey of students, in
secondary school, in two Indian states where students were tested using a selec?on of 36
ques?ons from the TIMSS 1999 math test. They found that a large number of students failed
to meet even basic interna?onal benchmarks.
Table 37 tracks changes in student ability to read a Std II text (MRP) and perform basic
division (MAP), across secondary grades, with the available data for 2006 to 2024. In 2024,
Minimum reading proﬁciency (MRP) as measured by the percent of children in who can read
Std II level text, is 57.7% in Std VI, 64.4% on Std VII and 71.1 in Std VIII. This means that in
rural areas, 42% of std V, 36% of Std VII and 39% of Std VIII students cannot read a Std 2 text.
Similarly Minimum Arithme?c proﬁciency (MAP) as measured by the percent of children who
can divide is 36% in Std VI, 41.5% in Std VII and 45.7% in Std VIII. MAP is therefore signiﬁcantly
Negative GapLittle or No gapPositive Gap
Nagaland (-10.1%)Meghalaya (-2.0%)Rajasthan (+3.4%)
J&K (-9.0%)ChhattisG (-1.5%)Manipur (+5.6%)
MP (-5.3%)Odisha (-1.1%)Tripura (+6.6%)
Bihar (+0.3%)WB (+9.9%)
Assam (+0.4%)
Jharkhand (+0.8%)
UP (+1.2%)
Sikkim (-27.0%)Haryana (-2.9%)A&Nislands (+3.9%)
Gujarat (-21.5%)AP (-1.6%) UttaraK (+4.0%)
Arunachal (-14.8%)Tamil Nadu (-1.2%)Puducherry (+4.1%)
Telangana (-0.8%)Punjab (+7.9%)
Mizoram (-0.7%)Goa (+8.5%)
Delhi (+0.2%)Kerala (+18.6%)
HP (+1.1%)Chandigarh (19.0%)
Karnataka (+1.3%)
MH (+2.8%)
PcNsdp >
₹10,000
PcNsdp <
₹10,000
70

lower than MRP across Std, with the MAP in Std VIII 25.7% points lower than MRP in the
same grade, a gap which needs to be reduced.
Between 2006 and 2014, there is a sharp decline in both MRP and MAP across all lower
secondary grades. For Std VIII, reading fell by 9.1% and division by 31.7%, while similar
declines are evident in Std VII and Std VI. Between 2014 and 2024 was a recovery in Arithme?c
(MAP) by 1.6% points to 3.4% points in stds6 to 8. In contrast, the declining trend in MRP, seen
from 2006 to 2014, was slowed, but not reversed. This likely due to reduced verbal
interac?ons and connectedness which is more important for language learning.
Table 37: Learning outcomes in 2024 & change 2006 to 2024 (% of students)

Source: ASER reports 2024, 2014 and 2006
The eﬀect of Right to Educa?on Act (2009) is analyzed by comparing changes in the pre-RTE
period (2006-2010) with the post-RTE period (2010-18). The immediate period aLer the
introduc?on of RTE (2009) shows declines in both reading and arithme?c competencies (Table
38). Between 2006 and 2010, MRP increased in Std VI and Std VII, but declined during 2010
to 2018. MAP declined in both pre-RTE and post-RTE period, but the decline was less in the
post-RTE period. Similar pa?ern was observed for MRP in Std VIII.
Table 38: Eﬀect of RTE on Learning outcomes (% of students)

Source: ASER reports 2018, 2010 & 2006
As in the case of primary schooling, Pandemic also disrupted secondary schooling. We
measure the eﬀect of this disrup?on by comparing the change in learning out comes in the
post-pandemic period 2022-2024 with the change during the pandemic period. As there was
no ASER survey during the Pandemic, we use the change from 2018 to 2022 for the la?er.
StdStdIITextDivideStdIITextDivide
VI 57.736.058.832.2
VII 64.441.567.737.8
VIII71.145.774.644.1
Change
VI -1.13.8-7.8-26.3
VII -3.33.7-8.4-29.6
VIII-3.51.6-9.1-31.7
2024 2014
2014 to 20242006 to 2014
StdStdIITextDivideStdIITextDivide
VI 59.834.767.549.3
VII 67.739.076.257.8
VIII72.843.982.967.4
Change
VI -7.7-14.6 0.9 -9.2
VII -8.5-18.8 0.1 -9.6
VIII-10.1-23.5-0.8-8.4
2010
Post RTEA(2010 to 2018)Pre- RTEA (2006 to 2010)
2018
71

From 2018 to 2022, MAP declined in all three standards, but increased in Std VIII, while
declining less in Std VI and Std VII (Table 39). There is a clear recovery aLer the pandemic, in
both MRP and MAP, across all lower secondary grades. The recovery is stronger in MAP with
the per cent of children who can divide, larger in 2024 than in 2018.
47

Table 39: Eﬀect of Pandemic on learning outcomes (% of students)

Source: ASER reports 2024, 2022 & 2018
Figure 23, shows trends in Minimum Learning Proﬁciency in Std VIII during 2006 to 2024. The
percentage of students with minimum reading proﬁciency (Std II level text) was high in 2006-
2007, but has been on a declining trend since then. The steepest decline was from 2006-2007
to 2014, with the rate of decline less during the past decade (green line). Part of the decline
in the last decades is due to weaker founda?onal learning outcome in primary school, in
previous decade. The declining trend in the percent of Std VIII students who can divide (MAP)
is even sharper from 2006 to 2014. There has been a gradual but modest improvement from
2014 to 2024. As founda?onal learning is even more important for Arithme?c than for reading,
the sharper fall in ability to divide, more likely reﬂects poor MAP in primary school with
respect to ability to subtract & two-digit number recogni?on.
Figure 23: Minimum Learning Proﬁciency: Std 8 (% of children)

Data source: Annual Status of Educa?on Reports 2006-2024
We use the learning outcome data from ASER, by grade to construct synthe?c cohorts. We
then follow each cohort through diﬀerent grades. Thus, the cohort marked 2006 (is the cohort
which entered Std I in 2006, while 2017 is the cohort which entered Std I in 2017 (X axis in

47
This can be seen by adding the columns headed divide (3 & 5) in table 31.
StdStdIITextDivideStdIITextDivide
VI4.94.3 -7.0-3.0
VII2.33.7 -5.6-1.2
VIII1.61.1 -3.30.7
Post PandemicPre-Pandcemic
2022-2024 2018-2022
R² = 0.9156
R² = 0.9167
30
40
50
60
70
80
90
% of children
Year
Minimum Learning Outcome: Std 8
Std 2 TextDivision Trend (Std 2 Text)Trend (Div)
R² = 0.8302
R² = 0.7727
40
45
50
55
60
65
70
75
80
% of children
Cohort year
Minimum Learning Outcome: Std 8 (Cohort)
Std 2 Text Division Trend (Std 2 Text)Trend (Div)
72

right panel, Figure 23). Comparing the right panel with the leL panel of Figure 23 , shows that
MAP outcomes in Std VIII do not change much across cohorts (pink lines) during 2006-2024.
The MAP results across cohorts also lower than the raw data across years, though the
percentage diﬀerence may not be as signiﬁcant, as the former are on average, lower than the
la?er. The trends in MRP across cohorts are also ﬂa?er than the trends over ?me diﬀerence
between the trends in MRP (red lines in right & leL panel), and are much less signiﬁcant in
percentage terms, as the average is lower for the cohorts.
Overall, the declining trend, suggests that as access improved for previously excluded
children, pedagogy and teaching has not changed suﬃciently to address the weaker self-
learning ability of new entrants. Move to universal school educa?on of such a large and
diverse popula?on, requires changes in the conven?onal curricula and teaching methods.
Inter-State Comparison
ASER data can also be used to compare Learning outcomes across States and UTs, with respect
to both Minimum Reading Proﬁciency (MRP) and Minimum Arithme?c Proﬁciency (MAP). The
27 States and UTs are put into three categories (A, B, C), using the all India mean outcomes in
Std VIII, and the standard devia?on of Minimum reading proﬁciency (Std II level text) to
classify States & UTs by per cent of students with MRP (Table 40, columns 1 to 3). Similarly
Minimum Arithme?c proﬁciency (Division) is used to classify States & UTs by per cent of
students with MAP (Table 40, columns 4 to 6). The numbers in brackets against each state
show the change between 2014 and 2024, in the percent of students in the State with MRP &
MAP respec?vely. The B level is created by taking a range of 0.5 ?mes the standard division,
on each side of the all-India average given by ASER for MRP and MAP respec?vely (Table 40).
Table 40: Min learning outcomes across Stats-2024 and change 2014-24 in bracket)

Source: Authors calculations based on ASER report 2024 & 2014. Note 1: Cut-offs are based on All India +/- 0.5*Stdev.
Note All India Std 8 MRP (std2 txt) is 71.1%, Stdev is 8.7%, and MAP (divide) is 45.7% with Stdev =11.2%. stdev=standard deviation.
In the case of MRP in Std VIII, there are 12 States & UTs in A (above average) category, 9 in B
(average) and 6 in C (below average) category, in 2024. Four States/UTs have improved their
MAP since 2014 in A category, two in B category and only one in C category. In the case of
A: Above averageB: AverageC: Below averageA: Above averageB: AverageC: Below average
More than 75.5% 66.7% to 75.5%Less than 66.7%More than 51.3%40.1% to 51.3%Less than 40.1%
Mizoram (+7.9%)Uttar Pradesh (+4.3%)Assam (+2.4%)Uttar Pradesh (+11.5%)Odisha (+10.7%)Madhya Pradesh (+7.7%)
Arunachal (+3.6%)Madhya Pradesh(+1.3%)Tamil Nadu (-5.1%)Uttarakhand (+4.8%)Jharkhand (-0.1%)Chhattisgarh (+7.0%)
Odisha (+1.5%)Jharkhand (-0.9%)J&K (-5.2%) Bihar (+2.5%) Telangana (-2.8%)Assam (+4.6%)
Uttarakhand (+1.0%)Maharashtra (-2.3%)Karnataka (-8.5%)Punjab (-1.9%)Andhra (-8.0%)Maharashtra (+3.4%)
Chhattisgarh (0.0%)Bihar (-4.4%) Telangana (-19.1%)Himachal (-10.0%)Arunachal (-12.3%)Karnataka (+0.9%)
Gujarat (-1.7%)Tripura (-5.0%) Andhra (-25.4%)Haryana (-10.2%)Nagaland (-30%)Tamil Nadu (-2.0%)
Haryana (-2.6%)Rajasthan (-11.4%) Mizoram (-24.6%) Gujarat (-2.1%)
Kerala (-4.1%)Meghalaya (-12.6%) J&K (-3.4%)
Himachal (-7.7%)West Bengal (-27.6%) Tripura (-6.3%)
Punjab (-10.1%) West Bengal (-6.6%)
Nagaland (-10.6%) Rajasthan (-15.0%)
Sikkim (-14.8%) Kerala (-21.3%)
Meghalaya (-29.4%)
Sikkim (-35.5%)
Minimum Arithmetic Proficiency (MAP) in Standard 8Minimum Reading Proficiency (MRP) in Standard 8
73

MAP, the skew is in the lower direc?on, with 7 States & UTs in A, 6 in B and 14 in C category,
in 2024. Four States have improved their MAP since 2014 in category A, one in category B and
ﬁve in category C. The largest improvement between 2014 and 2024 are seen in UP (+11.5%),
and Odisha (+10.7%).
Figure 24: Minimum Reading proﬁciency (% of std 8) in 2024 & 2021

Data source: Annual Status of Educa?on Reports 2014 & 2024
Figure 25: Minimum Arithme?c proﬁciency (% of std 8 students) in 2024 & 2014

Data source: Annual Status of Educa?on Reports 2014 & 2024
The ranking of states and UTs by Minimum Reading proﬁciency (MRP), measured as per cent
of children in Std VIII who can read Std II level text in 2024, is shown in Figure 24. This ﬁgure
also contains the 2014 MRP. The ranking of states and UTs by Minimum Arithme?c proﬁciency
(MAP), measured as per cent of children in Std VIII in 2024, who can divide, is shown in Figure
25.
74

6.3.2 Comparison of Grade 8 learning outcomes (NAS viz ASER)
This sec?on analyses, to what extent the results NAS surveys for lower secondary school,
match those from ASER, given the limita?ons arising from diﬀerent years of the surveys and
the diﬀerent criteria and tes?ng procedures used. Table 41 presents Std/Class 8 results on
minimum learning proﬁciency based on ASER and NAS data. Since these surveys were
conducted in diﬀerent years and follow diﬀerent assessment frameworks, direct comparison
is challenging. However, approximate comparisons can be drawn between nearby years, such
as ASER 2022 with NAS 2021, and ASER 2016 and 2018 with NAS 2017. NAS shows the share
of students achieving Basic and Proﬁcient levels. This paper uses Basic and Proﬁcient levels to
include all those above the speciﬁc level. Thus, Basic level includes Proﬁcient and Advanced.
As NAS ques?ons become more diﬃcult for higher grades, the percentage of students in each
grade is expected to be lower than those for preceding grade.
Table 41: Minimum Learning Proﬁciency (% of students): Grade/Standard 8

Source: ASER 2022, 2018, 2016 and NAS 2021 & 2017
ASER presents the percentage of children who can read a Std II level text in language and those
who can divide in arithme?c. ASER uses the same test ques?ons & marking system for Std II
text and division for all grades, so the percentage of students achieving this level is expected
to be higher in the upper grades.
For Std/Class, ASER 2022 and NAS 2021 comparability is lower than we found for Std/Class 8.
In language, 69.5% of children could read a Std II-level text as per ASER 2022, but this share
diﬀers by (-)9.5% points for Basic level and (-)35.5% points from proﬁcient level in NAS 2021.
Similarly in mathema?cs, 44.6% of children could do division (ASER 2022), but this share
diﬀers by points (-)17.6% from those who were proﬁcient and by (-)28.4% points who achieved
basic levels. Thus, in both language and Math, the match between NAS and ASER is inferior in
Std/Class 8 than we found in Std/Class 5. Similar mis-matches are found between NAS 2017
and ASER 2016 or ASER 2018 in language. with NAS 2017 percentage for Basic language skills
higher than ASER 2016 Std II level text by +12.2% points (17%) and Proﬁcient level lower by
(-)34.7% points (-48%). However, there is almost a perfect match between NAS 2017 student
who are proﬁcient in math (39.2%), with those who can do division as per ASER 2016 (43.2%)
[Table 41].
Grade 8Reading/Language Students %Arithmetic/Math Students %
ASER 2022Can Read Std 2 level text69.5Can do Division 44.6
NAS 2021Basic (incl prof + advance)79.0Basic (incl prof + advance)73.0
Proficient (incl Advanced)34.0Proficient (incl Advanced)27.0
ASER 2018Can Read Std 2 level text72.8Can do Division 43.9
NAS 2017Basic (incl prof + advance)85.2Basic (incl prof + advance)82.1
Proficient (incl Advanced)38.3Proficient (incl Advanced)39.5
ASER 2016Can Read Std 2 level text73.0Can do Division 43.2
75

The cross-State correla?on coeﬃcient for Std/Class 8 results in language and Math/Arithme?c
conﬁrms the high variance in grade results from ASER 2022 and NAS 2021. The cross-survey
correla?on between the per cent of Std VIII children able to read Std II level text as per ASER
and State results for per cent of Class 8 students who can meet the basics is 0.25 (Table 42).
The cross correla?on for Class 8 math is even lower (0.15), between the percent of students
who are proﬁcient in Math across States (NAS 2021) and those who can divide ASER 2022).
This is consistent with the all-India analysis in Table 27.
Table 42: Learning Outcomes in Grade 8 [NAS (2021) & ASER (2022)]

Source: Na?onal Achievement Survey (2021), ASER (2022)
6.3.3 Learning Outcomes & Scores: Grades 8 and 10 (NAS)
The Na?onal Achievement Survey (NAS) evaluates learning outcomes of students across
grades and subjects at the na?onal level. The deﬁni?on of “Basic” and “proﬁcient”
performance level for each subject changes with Class. The test of performance becomes
more complex as students move to a higher grade. In the previous sec?on, we have shown
that the basic performance is best indicator of minimum language proﬁciency and the
proﬁcient level is the best indicator of minimum math proﬁciency, in Class 8. This sec?on
begins with Class 8 to understand the pandemic’s impact on basic learning levels, followed by
Class 10 outcomes and the shiL between the two stages. Then looks at conven?onal results
of secondary school performance such as average scores, and the percentage of correct
answers across subjects and categories.
Table 43 shows the percentage of Class 8 students who achieved at least the basic level of
learning (including proﬁcient and advanced) in NAS 2017 and 2021. It also shows the proﬁcient
level for Math as this was found to be more comparable to other surveys.
Table 43: Grade 8 students (%) mee?ng Basic performance level: NAS

Source: Na?onal Achievement Survey, 2021 & 2017
Subject
ASER 2022NAS 2021 ASER 2022NAS 2021
ProfeciencyStd II TextBasic+DivisionProficient+
All India (%) 69.579.044.627.0
States Avg(%) 72.981.144.324.1
Cross survey correl 0.25 0.16
Reading/LanguageArithmetic/Math
Grade 8
Subject TotalUrbanRuralBoysGirlsTotalUrbanRuralBoysGirls
Language 79.085.075.076.081.085.288.084.085.086.0
Maths 73.075.070.073.072.082.180.083.082.082.0
Math: Proficient27.026.026.027.027.039.532.042.039.040.0
Science 62.069.060.064.062.081.480.083.082.080.0
Social Science61.065.059.059.062.080.280.080.080.080.0
20212017
76

In 2017, overall performance was high across subjects, with over 80% of students mee?ng the
basic level in all areas: 85% in Language, 82% in Math, 81% in science, and 80% in Social
Science. The high scores were also seen in both urban and rural areas and across gender. The
percent of students who were proﬁcient in math was 39.5%
In 2021, student performance dropped, because of the Pandemic. The largest declines were
seen in science (-)19.4 points and Social Science (-)19.2 points, showing a stronger pandemic
eﬀect on these subjects. The fall in performance was visible across all categories: urban and
rural, boys and girls, but the rural urban gap widened more than the gender gap in language,
The geographical and gender gaps in Math were less aﬀected by the pandemic.
As NAS 2021 is the latest available survey with data on Class 10, it gives a picture of the
learning outcomes in Upper secondary school. States are therefore distributed into three
categories (A, B, C), using the all-India results along with the standard devia?on of State
results, following the methodology used for NAS 2017 in Table 25. NAS 2021 gives separate
results for English and for Modern Indian languages (MIL). The States & UTs are therefore
distributed in three categories for each, and presented in the two panels in Table 44 with
English on the leL and MIL on the right. For English, there are 21 States/UTs in category A, 9
in B category, and 5 in category C. For MIL there are only 6 States & UTs in category A, 17 in
category B and 14 in category C. This is partly because the na?onal average for English is 76.0%
and for MIL is 47%, half of that for English (Table 44). All the States in A category for MIL are
also in the A category for English. In contrast only two of the 5 States & UTs which are in the
C category of English are in the C category of MIL.
Table 44: Class 10 Basic Language ability in States/UTs in 2021

Source: Authors calculation based on NAS 2021. Note 1: Cut-offs are based on All India +/- 0.5*Stdev
Note 2: All India Std 10 MRP(English) is 76.0%, Stdev is 8.6%, and MRP (Modern Indian Language) is 47%, Stdev is 18.2%.
B: AverageC: Below AvgA: Above avgB: AverageC: Below Avg
71.7% to 80.3%< 71.7% > 56.1%37.9% to 56.1%< 37.9%
J&K HaryanaAssam KarnatakaRajasthanBihar Mizoram
MaharashtraManipurUttarakhandUttar PradeshDelhi Assam Nagaland
AndhraSikkimJharkhandBihar KeralaTripura Meghalaya
Daman & DiuDelhiTripura ChattisgarhHaryanaA&N IslandsManipur
HimachalNagalandWest BengalDadra & NHChandigarhUttar PradeshPuducherry
LakshadweepChandigarhGujarat PunjabAndhra Goa
RajasthanGoaMP JharkhandSikkim
TelanganaPuducherryMeghalaya Odisha Lakshadweep
KeralaPunjabMizoram TamilnaduArunachal
Arunachal Odisha ChattisgarhLadakh
Ladakh Tamilnadu UttarakhandTelangana
A&N Islands Daman & DiuKarnataka
West BengalDadra &NH
Gujarat J & K
MP
Maharashtra
Himachal
More than 80.3%
Minimum Reading Proficiency (MRP) in grade 10
EnglishModern Indian Language (MIL)
A: Above avg
77

Table 45 shows the States similarly categorized into three groups, based on State Math
proﬁciency in Class 10, 2021. This is because the math “proﬁciency” level in NAS 2021 is more
comparable with minimum arithme?c proﬁciency (MAP) than the “basic” level. There are 7
States and UTs in the A category (above average), 12 in B (average) and 18 in C category, below
the na?onal average. This is two more than in the C category of MIL (Table 44). The All-India
average of 47% for basic ability in MIL is much higher than the 23% of students who are
“proﬁcient” in Math in class 10, but much lower than the 70% who had “basic” math ability.
Table 45: Proﬁciency in Class 10 Math in 2021

Notes: Authors calc based on NAS 2021. Cutoffs based on India avg +/- 0.5*Stdev India Avg MAP (proficient) is 23.0%, its Stdev is 11.3%
Table 46 presents the sub-categories within students mee?ng the Basic (+) performance
levels: classiﬁed as Basic, Proﬁcient, and Advanced. While earlier tables discussed the overall
Basic (+) category, this table breaks it down further to show varia?on across subjects in Class
8 and 10.
Table 46: Distribu?on of Grade 8 & 10 students by performance level (%)

Source: Na?onal Achievement Survey, 2021
As students move from Class 8 to Class 10, the share of those performing below basic rises
sharply in all subjects, showing a clear decline in overall learning levels. The fall is most visible
in science, where students at below basic level increased from 37% to 74%, and in Social
Science, from 39% to 62%. Students at the advanced and proﬁcient levels declined in all
subjects, reﬂec?ng weak conceptual learning. Only a slight increase is seen in Mathema?cs
(basic level +1%) and Social Science (proﬁcient +1%). Overall, the distribu?on highlights
A: Above average
More than 28.7%
A&N IslandsUttarPradeshPuducherrySikkimDaman & Diu
Bihar Manipur Madhya PradeshLakshadweepKerala
Delhi WestBengalOdisha MizoramGujarat
Haryana Andhra Pradesh NagalandHimachal Pradesh
Rajasthan Jammu & Kashmir MeghalayaLadakh
ChandigarhJharkhand TamilnaduTelangana
Punjab Karnataka ChattisgarhGoa
Uttarakhand Dadra &NagarMaharashtra
Assam ArunachalTripura
B: AverageC: Below average
Minimum Math Proficiency (MAP) in class 10
17.3% to 28.7% Less than 17.3%
Grade
Levels <BasicBasicProficientAdvanced<BasicBasicProficientAdvanced
Language --- -21.045.022.012.0
Language (MIL)53.037.010.00.0--- -
English 23.017.041.018.0--- -
Maths 30.047.017.06.027.046.019.08.0
Science 74.017.08.01.037.035.019.08.0
Social Science62.023.013.02.039.041.012.08.0
Grade 10Grade 8
78

widening learning gaps as students advance to higher grades, especially in Science and Social
Science.
Table 47 presents the percentage of students in Class 10 mee?ng the basic performance level
in NAS 2021, along with the change from Class 8 to Class 10. Overall, performance tends to
decline as student progress to higher grades, with the lowest achievement recorded in science
(26%), followed by Social Science (38%). Across all subjects, urban students performed be?er
than rural. By gender, girls showed stronger outcomes in Language (MIL) and English, while
boys performed slightly be?er in Mathema?cs, Science, and Social Science.
Table 47: NAS (2021): Students (%) mee?ng basic performance level in Grade 10

Source: Na?onal Achievement Survey, 2021
When comparing across grades, it is important to note that Class 8 assessed “Language”,
whereas Class 10 included “Language (MIL)” and “English”, making direct comparison diﬃcult.
A performance gap is visible, with an overall 38% decline in Language scores from Class 8 to
Class 10. This is likely the result of a shiL from the local language/mother tongue in Class 8 to
Modern Indian Language (MIL) in Class 10, making it necessary for students to relearn in MIL
the basics they had already learned in language. The decline seems more pronounced in rural
areas (-)41% vs (-)34% in urban, sugges?ng a possible diﬃculty in ge?ng teachers who can
teach MIL. This would be a ﬁt case for using digital & e-learning aids for language teachers, so
they can smooth the transi?on.
Further a clear decline in Basic performance, is visible, with performance is weaker in Class
10 Science (-)36% and Social Science (-)23% than in Class 8. The former deteriora?on is
signiﬁcantly larger than the la?er. This suggests that science teachers (in both urban & rural
areas) are unable to improve their pedagogy to match subject complexity. In contrast, basic
learning in social science, which is more connected to general knowledge seems to deteriorate
less in urban areas than in rural, less for boys than girls and less overall than Science.
This reinforces the case for using digital and e-learning aids to improve the quality of science
(& other subjects) teaching in rural schools.
Grade 10 TotalUrbanRuralBoysGirls
Language -----
Language (MIL)47.054.043.045.047.0
English 76.084.073.076.077.0
Maths 70.072.068.072.068.0
Math: Proficient23.024.021.024.021.0
Science 26.033.023.027.025.0
Social Science38.046.034.039.038.0
Change from Grade 8 to Grade 10
Maths -3.0-3.0-2.0-1.0-4.0
Science -36.0-36.0-37.0-37.0-37.0
Social Science-23.0-19.0-25.0-20.0-24.0
79

Table 48 presents percentage of correct answers and average scores for Class 8 and Class 10.
Between 2017 and 2021, the overall performance of students declined across both Class 8
and Class 10. The average scores fell more sharply in science (-)4.8 and Social Science (-)4.6
for Class 8, showing a no?ceable learning loss during the pandemic period. In Class 10, the
drop was even greater in Mathema?cs (-)6.8 and Science (-)9.4. While average scores in
English (+4.8) and Language (MIL) (+1.2) improved, most other subjects saw declines. The
pa?ern was consistent for correct answers, which also showed clear deteriora?on from 2017
to 2021.
Table 48: NAS 2021: Average Scores & Correct Answers (%)

Source: NAS (na?onal) 2021 (All India-Class 8 &10)
Note: Scores are out of 500, converted to percent; MIL: Modern Indian Language.
In 2021, the Class 8 average scores remained around 50-60%, correct answer at 35-55%. In
Class 10, the Language (MIL) at 52% and English at 55.4% recorded higher average scores
Mathema?cs and Science were the lowest.
Across both grades, urban students consistently scored higher than rural students in almost
all subjects, across both average scores and correct answers. Among gender, girls performed
slightly be?er in Language-related subjects, while boys scored marginally higher in
Mathema?cs and Science.
Grade 8
TotalUrbanRuralBoysGirls
Average score (%)
Language 60.464.059.860.462.4
Mathematics 51.051.250.250.650.6
Science 50.052.049.450.450.4
Social Science 51.051.850.450.651.2
Correct Answers (%)
Language 53.058.050.052.054.0
Mathematics 36.037.036.036.036.0
Science 39.042.038.040.039.0
Social Science 39.036.034.039.039.0
Grade 10
Average score (%)
Language (MIL) 52.051.449.649.451.0
English 55.461.056.057.658.8
Mathematics 44.044.042.843.843.2
Science 41.243.640.842.042.2
Social Science 46.249.059.847.047.4
Correct Answers (%)
Language (MIL) 41.044.040.041.042.0
English 43.050.039.043.043.0
Mathematics 32.033.032.033.032.0
Science 35.037.034.035.035.0
Social Science 37.040.036.038.037.0
Year 2021
80

To improve outcomes in rural areas, focused eﬀorts are needed to ensure the availability of
subject-specialized teachers and to expand e-learning access for both students and teachers.
These measures can help bridge the urban-rural performance gap and strengthen learning
con?nuity across grades.
6.3.4 PRS (2024) results for Grades 6 and 9
PRS 2024 reports average scores for Grades 6 and 9 across subjects, by gender and loca?on.
Table 49 shows a decline in learning outcomes as students move from Grade 6 to Grade 9.
In Grade 6, average scores were 57 in Language and 46 in Mathema?cs, while Grade 9
students scored 54 in Language and only 37 in Mathema?cs. Science and Social Science at the
secondary level recorded similar averages around 40.
Urban students consistently outperformed than rural across both grades and subjects. Girls
performed slightly be?er in Language, whereas boys showed a marginal advantage in
Mathema?cs.
When compared with NAS 2021, PRS 2024 shows rela?vely lower average scores, par?cularly
in Mathema?cs and Social Science, where the gap ranges between 4-6 points. This diﬀerence
possibly reﬂect varia?on in tes?ng design and grades, since PRS assesses Grades 6 and 9 while
NAS covers Class 8 and 10. Despite these structural diﬀerences, the overall trend remains
consistent: Language con?nues to record the highest performance, while Mathema?cs
remains the weakest area across all grades. Urban-rural diﬀerences remain modest. However,
girls performing be?er in Language and boys showing slight advantage in Mathema?cs, a
pa?ern visible in both assessments.
Table 49: PRS 2024: Average scores (2024)- Secondary

Source: PARAKH Rashtriya Sarvekshan (PRS), 2024
6.4 Programs & Pla?orms
A number of programs and pla?orms have been launched by the central government to
improve the teaching and learning at the secondary and higher level. Schemes like Samagra
Shiksha Scheme, E-Pathshala, PM eVidya, DIKSHA, also extend to secondary and senior
secondary educa?on; discussed earlier in sec?on 5.7. Padhna Likhna Abhiyan (PLA) was
launched in 2020 to eliminate adult illiteracy in India by providing basic func?onal literacy to
Grade/Subjects
Category TotalUrbanRuralBoysGirlsTotalUrbanRuralBoysGirls
Language 57595555596064606062
Mathematics 46474547465151505151
The World Around Us4951484950
Grade/Subjects
Language 5458515256
Mathematics 37383637364444434443
Science 40423941404144414242
Social Science 40413939414649604747
Grade 6 (PRS 2024)
Grade 9 (PRS 2024)
Grade 8 (NAS 2021)
Grade 10 (NAS 2021)
81

adults aged 15 years and above. The target was to make 57 lakhs adult illiterates’ literate
within a year.
48
A few schemes which have been reviewed by independent researchers are
highlighted in this sub-sec?on.
6.4.1 SWAYAM
Study Webs of Ac?ve-Learning for Young Aspiring Minds (SWAYAM, 2017) is a massive open
online course (MOOC), providing free educa?on to learners from Class 9 up to post-
gradua?on. SWAYAM oﬀers 16530 courses with video lectures, downloadable reading
material, quizzes, and discussion forums for clearing doubts. As of 2025, SWAYAM had more
than 5.4 crore enrolments and had issued over 45.2 lakh total cer?ﬁcates
49
.
Raj et al. (2025) examined SWAYAM pla?orm data across nine na?onal coordinators and
found that “student enrolment, exam registra?on, and successful course comple?on are
strongly and posi?vely correlated”.
50
Their analysis further shows that a substan?al share of
learners progressed from enrolment to cer?ﬁca?on, indica?ng “70.6% of learners have used
SWAYAM courses eﬀec?vely”. The study also notes that “the majority of students have
enrolled in technical courses rather than other,” sugges?ng that “the learning popula?on in
India is largely demanding technical-based educa?onal programs.”
Learner-level evidence from Singh and Bhandari (2025), based on a survey of 154 higher-
educa?on students using SWAYAM-NPTEL (Na?onal Program on Technology Enhanced
Learning) courses
51
. The authors report that “71% of enrolled and eligible learners successfully
completed courses and received cer?ﬁca?on,” and that “a majority of learners perceived
improved comprehension of subjects and usefulness of cer?ﬁca?on.” Students also rated
course relevance and content quality posi?vely, with “over 80% ﬁnding courses relevant and
77% repor?ng that learning outcomes were fully or mostly achieved.” The authors also note
that “Out of 93 respondents, 65.6% found the pla?orm user friendly” & opera?onal. Also, of
69 respondents “62.3% of respondents believe that companies value SWAYAM-NPTEL
cer?ﬁca?ons,” sugges?ng conﬁdence in the pla?orm’s credibility and applicability to
employability. The study emphasizes that “improved digital infrastructure, and curriculum
integra?on are necessary to maximize the beneﬁts of the pla?orm.”
Beyond par?cipa?on and comple?on, Mehra and Kant (2024), using survey data from over
500 students and faculty across engineering and science ins?tu?ons, found that MOOC under
the SWAYAM “fails to provide avenues for face-to-face discussion, hands-on skill development,
and real-life learning experiences.” However, students and faculty agree that “MOOCs provide
ﬂexibility, and provide knowledge beyond classroom curriculum”.

48
The scheme especially focused on women, Scheduled Castes (SCs), Scheduled Tribes (STs), minori?es, and other disadvantaged groups.
49
h?ps://swayam.gov.in/
50
with coeﬃcient of determina?on of about 98% and p value is about 0.001.
51
Postgraduate and undergraduate students across all courses using Stra?ﬁed random sampling used to achieve representa?on by ins?tu?on
type, academic subject, and educa?on level of 154 students based on structured ques?onnaires.
82

Gunjan’s (2025) study of teacher educa?on students (B.Ed. and M.Ed.) found that “76.7% of
B.Ed. students and 66.7% of M.Ed. students” agreed that SWAYAM MOOCs are beneﬁcial, and
that “80% of B.Ed. students and 66.7% of M.Ed. students” were mo?vated to enroll in courses
to improve their knowledge and skills. In terms of quality, “73.3% of B.Ed. students and 80%
of M.Ed. students” SWAYAM MOOCs provide quality instruc?on and content. The study
suggests that improvements in internet connec?vity, increased awareness of the pla?orm,
and suﬃcient training to use SWAYAM pla?orm can strengthen the eﬀec?ve use of SWAYAM
MOOCs.
6.4.2 NISHTHA
Na?onal Ini?a?ve for School Heads' and Teachers’ Holis?c Advancement launched in
(NISHTHA, 2019) to train 42 lakh elementary teachers, school heads, SCERTs, DIETs faculty,
and resource coordinators. NISHTHA-Online was launched on the DIKSHA pla?orm in 2020.
The program was extended to secondary and senior secondary teachers in July 2021, and to
Pre-primary to Class V teachers and school heads, for FLN (Sep,2021).
52

Shrivastava (2024) did an opinion survey 400 teachers from Gujara?-medium schools in
Anand district about NISHTA. Author found that 43% of teachers felt the content's quality was
suitable for their professional and academic growth, 52.5% of teachers found assignments to
be need-based and 59.2% rated the evalua?on process for teacher cer?ﬁca?on as suitable.
The study further notes that 58.1% of teachers found the ICT pla?orm and tools a?rac?ve and
useful. Darji & Sondagar (2025) highlighted that 69.5%” highlighted ﬂexibility in pursuing
courses, indica?ng posi?ves of the program design and delivery
53
.
Using course-wise par?cipa?on data, Aayushi and Mi?al (2025) surveyed teachers from
Muzaﬀarpur, Bihar from 2021 to 2024 and found that teacher comple?on rates among the 13
courses in the NISHTHA 2.0 program “peaked in 2021 (45.7%)” but declined sharply to “22.0%
in 2022, 21.8% in 2023, and 4.1% in 2024.” The authors a?ribute this decline to factors such
as “prior comple?on of courses, reduced post-pandemic urgency and professional
responsibili?es. The study highlights courses such as “Toy-Based Pedagogy” were viewed as
more relevant for elementary or primary teachers than secondary school teachers. However,
for older students, "Ac?vity-Based Pedagogy for Engaging Adolescents" or "Interac?ve and
Experien?al Learning Approaches" would make the course relevant.
54

6.4.3 Na?onal Digital Library of India
Na?onal Digital Library of India (NDLI 2018) acts as a single-window pla?orm that aggregates
metadata and provides full-text indexing of resources collected from various na?onal and

52
Training programmes for teachers, PIB report 2022
53
Darji & Sondagar (2025) selected school teachers from various talukas of Anand district and prepared opinionnaire with 50 statements
having 10 components. Provides mostly similar results as Shrivastava (2024).
54
NISTHA 2.0: The study involved teachers from the Muzaﬀarpur district in Bihar, eight blocks were randomly selected from a total of sixteen
using the lo?ery method, and teachers from these blocks were chosen through cluster sampling. Percentage analysis and the t-test were
used as primary sta?s?cal tools for analyzing quan?ta?ve data, while qualita?ve data from interviews were analyzed using thema?c analysis.
83

interna?onal digital libraries, academic ins?tu?ons, and repositories. It oﬀers learning
materials across subjects and grade levels in mul?ple Indian languages, making quality
content accessible to students, teachers, researchers, and lifelong learners across the country.
Newar and Borah (2022) examined awareness and u?liza?on of NDLI through a survey of 384
respondents comprising students, teachers, and research scholars across India. The authors
found that “the overall level of awareness of NDLI was low (39.3%)”, with students have
compara?vely low level of awareness on NDLI, compared to research scholars”. In terms of
u?liza?on: “54% of respondents were able to access the required content using the app”, and
“71% rated the performance of the NDLI app as good.” The authors indicate that increasing
improved connec?vity and targeted outreach could strengthen u?liza?on of the pla?orm.
User sa?sfac?on and quality of engagement of NDLI are explored by Bhar? and Prabha (2024)
surveyed 150 users and report that “62% of users rated NDLI as sa?sfactory,” ci?ng
apprecia?on for its “vast content base and centraliza?on of resources.” The authors note that
strengthening naviga?on and content discoverability, expanding regional-language access,
and improving interface design, content organiza?on and tutorials would enhance
u?liza?on.
55

The digital library needs to add a completely new sec?on containing the best videos for skilling
of Master trainers, for semi-skilled workers in engineering skills for Industry, and repair &
maintenance, and self-employed workers in a range of old and poten?al new services.
6.5 Summary & Conclusion
6.5.1 Summary
Secondary educa?on in India does well on interna?onal comparisons, based on the per capita
GDP linked bench marks we have constructed; All available indicators are above the levels
expected at India’s per capita GDP. The progression rate from primary to secondary educa?on
at 91.3% is 2 per cent points above its benchmark (89.3%), Net Enrolment Rate of 61.6% is
5.8% points above the benchmark (55.8%), and a Gross Enrolment Rate of 78.9% is higher
than benchmark by 4.6% points. Both the Lower secondary comple?on rate at 85.5%, and the
upper secondary comple?on rate at 63.0%, are above benchmark levels by 16.2% and 12.3%
points respec?vely.
India’s progression rate from primary to secondary school is es?mated at 92.5% in 2023. This
is above the benchmark for the PCGDP we are projected to a?ain in ﬁve years. However, our
comparator countries (China, Vietnam, Thailand and Mexico) also overperformed on this
indicator, rela?ve to the benchmark for their PCGDP, while Indonesia was just on track. The
5.4% points increase needed in the progression rate for India, to meet the minimum HIC
benchmark of 97.9%, is well within reach.

55
Primary data includes interviews and open-ended surveys involving 150 users, primarily students, educators, and researchers from urban
and semi-urban academic ins?tu?ons asked to rate various features of NDLI on a sa?sfac?on scale and provide feedback on their overall user
experience.
84

India’s Net Enrolment Rate (NER) in secondary school was 61.6% in 2013, 5.8% points above
the level expected at its per capita GDP level in 2013. The NER is es?mated at 69.5% in 2023,
which is also above the UMIC benchmark, expected to be reached in ﬁve years. Indonesia,
Thailand and Mexico are also above the NER levels expected for their per capita GDP, while
China & Vietnam’s NER data was not available.
Within India, the Net secondary school enrolment rate (NER) is one of the few educa?on
indicators which is posi?vely corelated to pe capita NSDP. The good NER performance of India
is the result of some States and UTs doing be?er with respect to adjusted NER, than is
expected at their level of Per capita NSDP. These are, (a) At Lower Secondary: Odisha,
Rajasthan, Chha?sgarh, & Tripura among the lower income States & UTs, and Kerala,
U?arakhand, Maharashtra, Punjab, and Himachal Pradesh among the higher income States &
UTs. (b) At Upper Secondary: West Bengal, Tripura, Manipur, and Rajasthan among lower
income and U?arakhand, Punjab, and Kerala among higher-income states.
India’s lower Secondary comple?on rate was 85.5% in 2023, which is 16.2% above its
benchmark. Compe?tor countries such as China, Vietnam, and Thailand are at 100%. Upper
Secondary comple?on rate was 63.0% in 2023, 12.3% points above its benchmark. Thailand
(+0.5) and China (+17.8) are also above their expected levels, while Vietnam (-)5.9, Indonesia
(-)18.4 and Mexico (-)10.8 are below the level expected at their PCGDP.
The number of Indians comple?ng school educa?on are comparable to those of China, despite
lower comple?on rates (%) in 2023. In India 137 million students completed primary school,
65 million completed lower secondary school and 98 million students completed upper
secondary school. India therefore had 1.25 ?mes the students comple?ng primary school than
China (109 million) and more than 2.5 ?mes the combined total of Vietnam, Indonesia,
Thailand, and Mexico (MIC4). In India 65 million students completed lower secondary school,
1.2 ?mes those of China (53 mil) and 2.4 ?mes those of the MIC4. Similarly, the number of
Indian children comple?ng upper secondary school were 1.2 ?mes China and 3.7 ?mes those
in the MIC4.
The challenge is to raise secondary enrolment and comple?on rates to the interna?onal
benchmarks for a high-income country in 25 years or less. This requires an increase of 5.4 %
points in progression from primary to secondary, an increase of 17.6% points in Net
Enrolment Rate, 15.9% points in Gross Enrolment Rate, 5.4% points in lower secondary
comple?on, and 16.4% points in upper secondary comple?on rates.
As schooling comes within the purview of the State and local Governments, a disaggregated
view and ac?on plan will be cri?cal to success. The adjusted NER in lower secondary schools
of major states like Karnataka, Gujarat, and Haryana, are above the na?onal average of 77.9%,
but were below the level expected at their Per capita NSDP. Much stronger eﬀorts are also
required in Nagaland, West Bengal, Arunachal Pradesh, Sikkim, Haryana, and Karnataka.
Similarly, adjusted NER at the upper secondary level shows that, Telangana and Mizoram,
though above the na?onal average of 52.2%, were below the level expected at their Per capita
85

NSDP. Greater eﬀorts are also needed in Nagaland, Jammu & Kashmir, Madhya Pradesh, West
Bengal, Arunachal Pradesh, Sikkim, and Gujarat.
Minimum learning outcomes are cri?cal to the evalua?on of the quality of teaching and
learning. Minimum Reading Proﬁciency (MRP) is measured by the per cent of children who
can read a Std II level text. 57.7% of Std VI children and 71.1% in Std VIII are found to sa?sfy
this criterion in 2024. States such as Mizoram, Arunachal Pradesh, Odisha, and U?arakhand
performed above the average in 2024 and also showed improvements since 2014 (Std VIII).
Minimum Arithme?c Proﬁciency (MAP) is measured by the percent of children who can do
division. In 2024, MAP is found to be only 36.0% in Std VI and 45.7% in Std VIII. The MAP of
children in Std VIII, was above the all-India average in U?ar Pradesh, Himachal Pradesh,
U?arakhand, Bihar, and Punjab. As the all-India average MAP is 25.7 per cent points less than
MAP in Std VIII, urgent a?en?on needs to be given to close this a gap between MRP and MAP.
Learning outcomes improve from primary to lower secondary educa?on, though it is unclear
how much of this is due to dropout of students who have fallen behind, and how much to
actual learning by those who could not read or do division in earlier classes. The share of
children with Minimum Reading Proﬁciency increases from 48.7% at the end of primary (Std
V) to 71.1% by the end of lower secondary (Std VIII). Share of children with Minimum
Arithme?c Proﬁciency increases from 30.7% in Std V to 45.7% in Std VIII. Improvements in
MRP are observed in Mizoram (+7.9%), Arunachal (+3.6%), Odisha (+1.5%), U?arakhand
(+1.0%), U?ar Pradesh (+4.3%), Madhya Pradesh (+1.3%) and Assam (+2.4%) during the ten
years from 2014 to 2024. Improvements in MAP are observed in U?ar Pradesh (+11.5%),
Himachal Pradesh (+10.0%), U?arakhand (+4.8%), Bihar (+2.5%), Odisha (+10.7%), MP
(+7.7%), Chha?sgarh (+7%), Assam (+4.6%), Maharashtra (+3.4%) and Karnataka (+0.9%)
from 2014 to 2024. It is encouraging that States like UP and Odisha were able to improve
minimum learning outcomes so signiﬁcantly over a decade.
Data availability for minimum learning outcomes at upper secondary levels (Classes 9-12) is
lower. From the available data, paper selects the metrics which align best with interna?onal
metrics. These metrics are somewhat diﬀerent from those for Std I to Std VIII (used above).
According to NAS (2021), basic performance of Class 10 students was rela?vely high in English
and Mathema?cs, but lower in Modern Indian language, Science and Social Science. The per
cent of Class 10 students mee?ng basic performance level, was 76% in English, 47% in modern
Indian language (MIL), 70% in Mathema?cs, 26% in science, and 38% in Social Science. At the
upper secondary level (Class 10) top 20% states in learning outcomes were, (a) Jammu &
Kashmir, Maharashtra, Andhra Pradesh, Himachal Pradesh, Rajasthan, Telangana, and Kerala
in English, (b) Rajasthan, Delhi, Kerala, Haryana, Chandigarh, and Punjab (above all-India
average) and Bihar & U?ar Pradesh (average), in Modern Indian Languages. (c) Bihar, Delhi,
Haryana, Rajasthan, Chandigarh, Punjab, U?ar Pradesh, and West Bengal, in mathema?cs.
Surprisingly the performance in Modern Indian language was 47%, which was 29% points
lower than in English.
86

Urban students performed be?er than rural students across secondary educa?on. The share
of rural Class 8 students mee?ng the basic performance level was lower than urban students
by about 11.8% in Language and 6.7% in Mathema?cs. In Class 10, the gap increased to about
13.1% in English, and 5.6% in Mathema?cs (2021). The gap was 20.4% in modern Indian
language (only available in Class 10).
Average scores and percent of correct answers are other metrics available in India, which do
not have interna?onal counterparts. These show that, in Class 8, average rural scores were
6.6% lower in Language and 2.0% in Mathema?cs than urban scores. Correct answers in rural
areas were 13.8% lower in Language and 2.7% lower in Mathema?cs, than in urban areas. In
Class 10, rural students had 3.5% lower scores in Language, 8.2% in English and 2.7% in
Mathema?cs. On average correct answers in rural areas were 9.1% lower in Language, 22%
lower in English, and 3.0% lower in Mathema?cs (2021). Similar results were found in 2024 in
Grade 6 and 9. Rural students scored about 6.8% lower in Language and 4.3% lower in
Mathema?cs in Grade 6 than urban students. Rural Grade 6 results were about 12.1% lower
in Language and 5.3% lower in Mathema?cs than urban.
In primary school Girls perform be?er than boys in Language. In 2021, girls had higher
proﬁciency in Language in both Class 3 (+8%) and Class 5 (+16%), and also performed be?er
in average scores and correct answers. In Mathema?cs, performance of boys and girls is
largely similar in NAS (2021), though boys perform slightly be?er in Numeracy in 2022. In
2024, girls in class 3 scored higher in Language and iden?cal scores in Mathema?cs. In
secondary school girls perform be?er in Language and boys in Mathema?cs. In 2021, a greater
share of girls in Class 8, met basic performance levels in language than boys (+6.6%). However,
in Class 8 mathema?cs the share of girls mee?ng basic performance levels were (-)1.4% less
than boys. The posi?on in Class 10 was similar, with girls’ basic performance higher in
Language (4.4%) and English (1.3%), while it was lower by (-)5.6% in Mathema?cs.
In Class 8, average scores and correct answers in Language, for girls were 3.3% and 3.8% higher
(respec?vely) than for boys. In Mathema?cs both were iden?cal for boys and girls. In Class 10,
average scores and correct answers in Language for girls were 3.2% and 2.4% higher,
respec?vely, than for boys, whereas in Mathema?cs they were lower by (-)1.4% and (-)3.0%.
In English, average scores for girls were 2.1% higher than for boys, while correct answers were
iden?cal for both. By 2024, in Grade 6 girls scored higher in Language by 7.3% and lower in
Mathema?cs (-) 2.1%; in Grade 9 girls scored also higher in Language (7.7%) and lower in
Mathema?cs (-) 2.7%.
The biggest challenge will be to raise the minimum learning outcomes in reading (MRP) and
Arithme?c (MAP). As per the last available survey on learning proﬁciency, 28.9% of rural Std
VIII children do not have minimum reading proﬁciency, and 54.3% do not have Minimum Math
proﬁciency (MAP). For Urban students, the number may be ~10% less for MRP and ~5% less
for MAP. Raising MRP and MAP is important, both for obtaining higher comple?on rates at
lower secondary level, and for successful transi?on to upper secondary level. Assam, Tamil
87

Nadu, Karnataka, Telangana, Andhra Pradesh, West Bengal, Nagaland, and Sikkim are below
the all-India average in minimum reading proﬁciency. A majority of states, require stronger
eﬀorts to raise arithme?c proﬁciency including Madhya Pradesh, Chha?sgarh, Assam,
Maharashtra, Karnataka, Gujarat, Jammu & Kashmir, Tripura, West Bengal, Rajasthan, Kerala,
Meghalaya, Tamil Nadu, Sikkim, Nagaland, Mizoram, and Haryana. This can be achieved by
improving subject-speciﬁc pedagogy, especially in mathema?cs and science, ensuring
availability of teachers who are trained in teaching these subjects, and expanding digital and
e-learning resources to support both teachers and students.
To address these challenges, several programs and schemes have been launched: SWAYAM
provides large number of free, online courses for learners from Classes 9 to 12. Research
studies show that 70.6% of learners have used these courses, and 71% of those who signed
up completed their courses and received cer?ﬁca?on. The courses seem to ﬁll the gap in
technical educa?on. NISHTHA is a na?onal ini?a?ve to train elementary school teachers
(Classes 1-8), school heads SCERTs, DIETs and resource coordinators. a teacher training
ini?a?ve to strengthen pedagogy; Surveys by scholars show that 58% of teachers found the
ICT pla?orm and tools a?rac?ve and useful. The Na?onal Digital Library of India (NDLI), is a
single window providing full text indexing of resources, na?onal and interna?onal digital
libraries, academic ins?tu?ons and repositories. The NLDI oﬀers digital learning resources to
support students and teachers. A scholarly survey has found that 62% of users rated NLDI as
sa?sfactory. The study suggested that there is considerable room for improvement. These and
other ini?a?ves by the Central Government are aiming to improve raise comple?on rates, and
strengthen learning outcomes across the secondary educa?on system, but more needs to be
done at the State and UT level.
6.5.2 Conclusion
Knowledge is global, but applica?on of knowledge is local. NEP (2022) rightly emphasizes
holis?c learning, technical courses and extracurricular ac?vi?es. Much remains to be done to
provide standard opera?ng procedures for teachers and administrators, teaching and learning
material for teachers and students, and re-training the teachers to focus on learning
outcomes. The challenge for India at its level of per capita GDP, given its cons?tu?onal
structure, is how to eﬀec?vely apply Interna?onal, Na?onal and inter-State research to
educa?on & skilling at the local level.
Research shows the importance of four aspects of learning at the secondary level. Namely
Learning by playing, learning by socializing, learning by doing and learning to think. There are
many overlaps between them. These have to be adapted to reach the large, dispersed and
diverse, non-urban popula?on of India. In our view, these are best imparted by maximizing
the use of tradi?onal games, local materials, living things & animals, and the environment, in
pedagogy, curriculum and teaching.

88

(1) Learning by Playing (Sports, physical and board games)
Sports are the means for students to learn self-discipline (to follow rules of the game) and
self-mo?va?on to be?er their performance. Well known games like cricket and soccer are
undoubtedly important for older children, but simple, old games using local materials can be
equally eﬀect for lower secondary children. These include marbles, pi?hu, gulli-danda,
kabaddi, rope skipping and hop-skip and jump. These can even help bystanders learn coun?ng
and how to accept and interact with children who are be?er than oneself in games.
(2) Learning by Socializing (Mul?player games, play ac?ng, singing, discussion, debate)
Students can also organize periodic drives to collect dona?ons of old toys, games, sports
material, educa?onal music, videos etc., par?cularly in Urban areas for use in their own and
other schools. Schools can have dona?on boxes for these things along with books & comic
books and tools and material for arts & craLs, and making simple electric and mechanical
equipment
(3) Learning by Doing (Arts, CraLs, hobbies)
Students can make the simple items needed for tradi?onal games. This is an educa?onal
project in itself. Older students can also try and make simpler toys for younger students based
on known designs and local materials. They could also list local craLsman as temporary
teachers to help students how to make toys and games with wood and other materials.
(4) Learning by Thinking (Awareness, Curiosity, Ques?oning, study/ research for answers).
Ability to think, is preceded by awareness, curiosity and ques?oning. The conven?onal
method of doing this is through story books, discussion and debate on issues. The simplest
way, especially in rural India, is by inves?ga?ng the world around the school and home, and
becoming aware of its varied details. This includes, Nature (Leaves, trees, bushes,
seeds/beads, crops; worms, insects, birds’ animals) & Geology (Rocks, soils, clay, building
material, ﬁelds, hillocks).
Awareness can be followed by simple measurement, comparison and experimenta?on
(research) an invaluable aid to learning how to think. Curiosity about natural phenomenon
can also be aroused by experiments captured on audio visual material. Curiosity about things
and events which are foreign to local society can be aroused by comics and audio-visual
material.
The ques?on and answers that follow are the tool for teaching secondary school students
how to think and do simple research to ﬁnd the answers to their ques?ons. These things will
not happen spontaneously, Pedagogy and Standard opera?ng procedures (SOP) need to
developed for teachers and administrators, and discussed among them.
89

6.5.3 Sugges?ons
There must be weekly periods assigned in secondary school for sports and games, arts &
craLs and learning ou?ngs. Teachers must be trained how to do these things, and how to
teach their students. Besides wri?en materials like comic books, songs, recorded stories, short
illustra?ve videos are excellent for teaching both teachers.
Children drop out at every level of secondary school. They must be provided an outlet for
acquiring job skills. State Governments should iden?fy hub schools where job skills are
provided, these can include craLs and voca?ons with actual or poten?al demand in the
locality. Par?cular a?en?on to locali?es with poor a?endance and/or high drop-out rates.
Companies should be allowed and encouraged to adopt schools near their factories, to
provide voca?onal educa?on and training appropriate to secondary school.
7. Ins?tu?onal Issues
This sec?on examines Ins?tu?onal factors and educa?onal infrastructure.
7.1 Pupil Teacher Ra?o (PTR)
Teachers play a key role in improving the quality of educa?on. The number of students per
teacher and the share of trained teachers are conven?onal indicators of ins?tu?onal quality.
Table 50 shows the Pupil-Teacher Ra?o (PTR) and Trained Teachers. Between 2018 and 2024,
the PTR has declined across all levels of schooling, indica?ng smaller class sizes and be?er
teacher availability, and the poten?al for more focused a?en?on on students requiring greater
a?en?on. The sharpest improvement is seen at the primary level (27 to 20), followed by
secondary (19 to 15) and higher secondary (27 to 23). The improvement in PTR is reﬂected in
an increase in the percent of children mee?ng Minimum Arithme?c Proﬁciency (MAP), but
Minimum Reading Proﬁciency (MRP) deteriorated in Std IV to Std VIII (Table 51). We also
es?mate the cross correla?on between minimum learning ability (MRP & MAP) and Pupil
Teacher Ra?o (PTR) across States. There is zero correla?on between PTR and MAP in Primary
educa?on (Std V) and Upper Primary/Lower secondary (Std VIII). The correla?on between PTR
and MRP is very low in both Std V (-0.09) and Std VIII (-0.15) (Table 51).
Table 50: Pupil-teacher ra?os & per cent of Trained Teachers

Data Source: UDISE+ 2018,2021 and 2024 Reports.
20182024Change20212024Change
Preprimary 67.653.3-14.3
Primary (1-5) 27.020.0-7.088.791.42.7
Upper Primary (6-8)19.017.0-2.088.691.73.1
Secondary (9-10)19.015.0-4.092.292.70.5
Higher Secondary (11-12)27.023.0-4.091.892.60.8
Pupil-teacher ratioTrained teachers (%)
90

Table 51: Interstate correla?on of MRP & MAP with PTR & TT

Source: UDISE+ Reports, 2018-2025 and ASER 2024, 2018
Glewwe and Muralidharan (2016) found that improving teacher accountability through
pedagogy, monitoring a?endance, performance-linked pay, contract teachers, and be?er
school management is more eﬀec?ve than increasing inputs like salaries, based on global
evidence in low-income countries, De Ree et al. (2018) found no improvement in student
learning outcomes in mathema?cs, language, or science, three years aLer a large salary
increase for teachers, despite increased teacher sa?sfac?on, reduced ﬁnancial stress, and
fewer second jobs among teachers (in Indonesia). They argued that in public-sector systems
with high job security and weak accountability, uncondi?onal pay hikes are unlikely to raise
produc?vity. Mbi? et al. (2019) conducted a randomized experiment in public schools in
Tanzania, tes?ng teacher incen?ves, uncondi?onal grants, and showed that teacher incen?ves
had some posi?ve eﬀects, uncondi?onal grants had no eﬀect, and the combined interven?on
produced the most signiﬁcant gains in student learning.
According to UDISE+ (Table 50) the share of trained teachers, has improved at all levels,
except the preprimary level, which shows a major decline (from 67.6% to 53.3%). This could
possibly be due to the implementa?on phase of NEP 2020, which introduced new norms and
qualiﬁca?on requirements for early childhood educators. While other small improvements:
Primary (+2.7), Upper Primary (+3.1), Secondary (+0.5), and Higher Secondary (+0.8). In
Primary school (Std V), the interstate correla?on between percent of trained teachers and
learning outcomes is higher than the correla?on between PTR and learning outcomes but s?ll
low; 0.18 and 0.22 for MAP and MRP respec?vely but s?ll low (Table 51). In Lower
Secondary/Upper Primary the correla?on between trained teachers and MRP is close to zero
(0.05), but higher than with PTR for MAP (0.18). Simple interstate regressions, with either
MRP or MAP as dependent variable and Pupil Teacher ra?o (PTR) and Trained Teachers (TT) as
independent variables, shows that the coeﬃcients are not signiﬁcant. This conﬁrms that these
variables have li?le or no eﬀect on minimum learning outcomes.
The UDISE+ deﬁni?ons do not match the ones used in WDI, and the years for which data is
available are also diﬀerent. According to WDI, the share of trained teachers has shown a clear
improvement across all levels of educa?on. The largest increase is at the primary level, from
69.8% in 2017 to 91.7% in 2024 a gain of 21.9% points (Table 52). At the secondary and upper
secondary levels, the increase of trained teachers was by 16% points each, over seven years
to 2024. Even in preprimary educa?on, trained teachers have increased from 83.8% of total
in 2020 to 92.3% in 2024.
MRPMAPMRPMAP
PTR-MRPTT-MRPPTR-MAPTT-MAPPrimary (Std V) 48.730.7 -1.62.9-0.090.180.030.22
Lower Secondary (Std VIII)71.745.7-1.71.8-0.150.050.010.18
2024
Change (2024-18)
Cross State Correlations
91

Table 52: Improvement in Percent of Trained Teachers (%)

Data Source: World Development Indicators, December 2025.
7.2 Gender parity
India has largely achieved gender parity in school educa?on, with female-to-male ra?os at
or above 1.00 across most indicators, including Primary and Lower Secondary comple?on,
School Enrollment, and Progression to Secondary educa?on (Table 53). Despite this overall
parity in access and comple?on, a minor gender gap persists in youth literacy, with the rate
for young females (96%) slightly lower than that for young males (98%) in 2023. Net
enrollment rates at both primary and secondary levels favor females over males (ra?o of 1.02),
while a small disparity remains in gross enrollment at the primary (0.99) and ter?ary (0.98)
levels.
Table 53: Female-to-Male Ra?o in Educa?on of Indian children

Data Source: World Development Indicators, December 2024
7.3 Public vs Private Schools
Many studies of Indian educa?on (Sec?on Error! Reference source not found.) have pointed
to the diﬀerence in learning outcomes between Private and Public schools. This sec?on
examines the latest available data, to determine the current situa?on and changes if any. Due
to diﬀerences in grading levels, learning levels, and categoriza?on of school types, results
cannot be directly compared across surveys. Each survey is assessed separately to understand
the diﬀerences between public and private schooling.
Change
Trained teachers (% of total teachers) Rates Rates
In preprimary education 202492.3202083.88.5
In Primary education 202491.7201769.821.9
In secondary education 202492.3201776.515.8
In lower secondary education 202492.0201776.615.4
In upper secondary education 202492.7201776.416.3
Latest available yearYear for comparison
Female-to-Male Ratio in Education of Indian children
YearMalesFemalesRatio
Literacy rate, youth (% of ages 15-24)202398.096.00.98
Primary School enrollment rate, (%)- gross 2023112.6111.40.99
Primary School enrollment rate, (%)- net 201391.693.01.02
Primary completion rate,(% of relevant age group)202393.393.71.00
Progression to secondary school (%)201791.091.61.01
School enrollment, secondary, (% gross) 202379.078.81.00
School enrollment, secondary, (% net)201360.962.41.02
Lower secondary completion rate, (% of relevant age group)202384.586.51.02
School enrollment, tertiary, (% gross)202333.432.80.98
Rate for
92

ASER (Rural) reports results for alternate years from 2010 to 2024, for government and private
schools across Std III, V and VIII, assessing MRP using Std II level text and MAP using
Subtrac?on (Std III) and Division (Std V & VIII). NAS 2017, provides data for Government and
Government Aided schools for Class 3, 5, and 8. NAS 2021, contains data for State
Government, Government Aided, Private, and Central Government schools. Similarly, PRS
2024 covers the same four categories but for Grades 3, 6, and 9.
Our results are consistent with earlier research that highlighted long-term gaps in quality and
eﬃciency across school types. Kingdon (1996) found that private unaided schools deliver
be?er learning outcomes at lower cost compared to government or private aided schools, due
to stronger teacher accountability.
7.3.1 Annual Status of Educa?on Report
In 2024, ASER data show that private school students performed be?er than their
government school in both reading and arithme?c at all grade levels. The gap between the
two school types is 12-15% in reading and slightly higher in arithme?c, especially at the lower
grades (Figure 26). Std V is generally considered the end of primary school, while Std VIII is
classiﬁed in India as the end of either upper primary or lower secondary. Std III of interest
because it is the goal post for measuring founda?onal literacy and numeracy (FLN). While
Minimum Learning Proﬁciency (MLP) improves as children move to higher grades, the
diﬀerence between private and public schools con?nues, with private schools’ minimum
learning outcomes consistently be?er than those in public schools (Figure 26).
Figure 26: Performance of Students (%) by School Type

Source: Annual Status of Educa?on Report, 2024
Inter-temporal data on reading and arithme?c proﬁciency in Std III, V & VIII, for government
and private schools conﬁrms that the private public performance gap has persisted from 2010
to 2024 (Table 54). The per cent of children with Minimum Reading Proﬁciency (MRP) and
Minimum Arithme?c Proﬁciency (MAP) in private schools is consistently higher than that in
government schools. The performance gap is wider in arithme?c than in reading, sugges?ng
12.1
14.5
12.5
0
10
20
30
40
50
60
70
80
Std III Std V Std VIII
% of Children
Private vs. Gov Schools in Reading Std II text
Govt Pvt Pvt-Gov
19.9
15.3
13.9
0
10
20
30
40
50
Std III Std V Std VIII
% of Children
Private vs. Gov Schools in Arithmetic Skills
Govt Pvt Pvt-Gov
93

a rela?ve greater deﬁciency in pedagogy and teachers’ ability to teach arithme?c in
government schools.
Table 54: Trend by School Type: 2010 - 2024

Source: Annual Status of Educa?on Reports, 2010-2024
Kingdon (2007) observed that India’s educa?onal progress, though remarkable in enrolment,
s?ll suﬀered from weak a?endance and low learning levels. She warned that while private
schools deliver be?er outcomes, this led to inequality, as access to private schooling is
concentrated among be?er-oﬀ families. The learning gap in ASER 2024 conﬁrms this challenge
for Government schools. Singh (2014) ﬁnds that the public-private gap varies across loca?ons:
in rural areas, private schools have large posi?ve eﬀects on English and moderate eﬀects on
Math and Telugu for younger children, while in urban areas private schools do not outperform
government schools.
Kingdon (2017) reviews the growth and role of private schools in India using DISE, NSS and
ASER data, showing that more students are moving from government to private schools
because private schools are more aﬀordable and oﬀer be?er value; most charge low fees and
get slightly be?er learning results. They paid lower teacher salaries to sustain lower fee
structures. Many small schools were struggling with strict RTE Act rules. Data collated from
public domain by Na?onal Independent schools’ alliance (NISA) on school closures due to RTE
Act showed that as of October 2016, 9382 private schools had received closure threat.
Another 7898 Private school actual closure no?ces, and further, 3332 private schools had
actually closed down, based on data from only a few states.
56

It is useful to view the inter-temporal trends in learning outcomes in private schools, as a
benchmark of exogenous factors which may have aﬀected the quality of teaching and learning.
Consider ﬁrst the outcomes in Minimum Reading Proﬁciency (MRP) as measured by the
percent of children in private school who can read Std II text. The pa?en of change in both Std
V and Std VIII is nonlinear (Table 54, columns 3 & 5). In primary school Std V), the sharp drop
in performance ?ll 2012, was steadily reversed from 2014 to 2018. It suﬀered a temporary

56
Actual school closure from govt documents is available from only three States.
Sub
Std 5 Std 8 Std3Std5Std8
YearGovtPvtGovtPvt GovPvtGovtPvtGovtPvt
12345678910111213141516
1201050.764.282.087.5-13.5-5.533.247.833.944.267.072.0-14.6-10.3-5.0
2201241.761.273.484.2-19.5-10.819.843.420.337.844.557.1-23.6-17.5-12.6
3201442.262.671.582.4-20.4-10.917.243.420.739.340.054.2-26.2-18.6-14.2
4201641.763.070.081.0-21.3-11.020.344.121.138.040.251.2-23.8-16.9-11.0
5201844.265.169.082.9-20.9-13.920.943.522.739.840.054.2-22.6-17.1-14.2
6202238.556.866.280.0-18.3-13.820.243.121.638.741.853.8-22.9-17.1-12.0
7202444.859.367.580.0-14.5-12.527.647.526.541.841.955.8-19.9-15.3-13.9
Diff Gov-Pvt Diff Gov-Pvt
Division Division
Std 5Std 8 Std 3Std 5Std 8
Read Std2 text
Subtraction
94

setback during the pandemic, but recovery followed as soon as pandemic was over. Govt.
schools broadly follow this pa?ern with some hiccups. In Lower Secondary school (Std VIII),
there is a decline between 2010 and 2016, an upward bump in 2018, and further decline in
2022. It is clear that a poten?al improvement in outcomes star?ng in 2017 was disturbed by
the pandemic.
Consider next the outcomes in Minimum Arithme?c Ability (MAP) as measured by the percent
of children in private school who can do Division. In both Primary (Std V) and Lower
Secondary/Upper Primary, there is trend decline in MAP from 2010 to 2016 and then the same
cycle of recovery from 2016 to 2018, pandemic setback ?ll 2022 a then a recovery in 2024.
FLN in private schools, as measured by the MAP (Subtract) in Std III, has a completely diﬀerent
pa?ern, falling sharply from 47.8% in 2010 to 2012, remaining between 43.1% -44.1% level ?ll
2022, and then recovering sharply to 47.5%.
Some of these wobbles are likely the eﬀect of the right to educa?on Act (RTE) as learning
outcomes declined across the board in both private and govt schools for Std III, V and VIII in
both MRP and MAP (Table 55). ALer 2014 there was a recovery (except in Std VIII, MRP in
Govt schools), but there was another setback during the pandemic in 2020 and 2021.
Table 55: Eﬀect of RTE on Minimum Learning Proﬁciency (% of students)

Source: Annual Status of Educa?on Reports, 2016, 2014 & 2024; Note: Diﬀ = Public - Private
These ﬁndings align with the evidence presented by Kingdon and French (2010), who
analyzed ASER 2005-2007 data and found that private school children consistently performed
be?er than those in government schools, even aLer controlling for family background. They
noted that the diﬀerence was not just due to student characteris?cs but also due to school-
level factors, such as be?er teaching prac?ces, lower teacher absence, stronger management,
and more eﬀec?ve accountability systems.
State governments need to pay much greater a?en?on to improving the quality of educa?on
in Govt. schools by improving pedagogy, training teachers to teach at the right level and
administrators to iden?fy, student; earning levels and improvements through repeated
tes?ng, and back teachers to improve minimum learning outcomes.
7.3.2 Na?onal Achievement Surveys (NAS) 2021
Na?onal Achievement Survey (NAS) evaluates students learning outcomes across diﬀerent
classes, subjects, and school management types. Table 56, presents the percentage of
Minimum Learning
Standard
Year GovtPvtDiffGovtPvtDiffGovPvtDiffGovtPvtDiffGovtPvtDiff
Year 2010 50.764.2-13.582.087.5-5.533.247.8-14.633.944.2-10.367.072.0-5.0
Change 2014-2010-8.5-1.6-6.9-10.5-5.1-5.4-16.0-4.4-11.6-13.2-4.9-8.3-27.0-17.8-9.2
Year 2014 42.262.6-20.471.582.4-10.917.243.4-26.220.739.3-18.640.054.2-14.2
Change 2024-20142.6-3.35.9-4.0-2.4-1.610.44.16.35.82.53.31.91.60.3
Year 2024 44.859.3-14.567.580.0-12.527.647.5-19.926.541.8-15.341.955.8-13.9
Division
Std 5 Std 8Std 3 Std 5 Std 8
Std 2 text Subtraction
95

students performing at or above the Proﬁcient level for Classes 3 and 5, and at or above the
Basic level for Classes 8 and 10, by school type. This 2021 survey contradicts the generally
accepted results for the Class 3 level relevant for FLN, but is broadly in agreement with results
from other surveys for Primary and Lower Secondary.
The last column of Table 56 shows the diﬀerences in share of students mee?ng Minimum
Learning Proﬁciency in selected Classes. This shows that Public Schools marginally
outperformed private schools in Class 3 in both Reading proﬁciency and Math proﬁciency in
2021. Equally surprising, the State Govt. schools also outperform Central Govt. schools by
1/3
rd
. They con?nued to do so in Math proﬁciency even Class 5, with State govts schools again
bea?ng Central Govt. schools. The results for Class 8 (Lower Secondary) and Class 10 (Upper
Secondary) are similar to those of other surveys, with Private schools outperforming State
Govt. schools. This is accompanied by a reversal of the rela?ve performance of State Govt.
and Central Govt. schools, with the la?er now performing be?er. Because the 2021 results are
strongly aﬀected by the Pandemic, it is possible that Private Schools and Central Govt. schools
were more generous in allowing teachers and students of Primary and Lower Secondary
school, to absent themselves from school during the Pandemic, than State Govt schools.
Table 56: Students (%) by performance levels, 2021

Source: Na?onal Achievement Survey, 2021
The posi?on with respect to average number of correct answers is very similar to that about
minimum learning outcomes, with State govt. schools having higher levels of correct answers
on average than private schools (last column of Table 57). Average scores were consistent with
other surveys, with only one excep?on; Average scores in Math in Class 3 (Table 57) for State
Govt schools were higher than in Private schools, with Central Govt schools performing worse
than State Govt schools.
Performance Level
Class 3
TotalState Gov.Central GovtGov. aidedPrivateStateGovt-Pvt
Language 39403139391
Mathematics 42443143422
Class 5
Language 4241433944-3
Mathematics 25282422235
Class 8
Language 6972867387-15
Mathematics 7370806578-8
Class 10
English 7672877282-10
Language (MIL)4742554153-11
Mathematics 7068786473-5
Proficient (incl advanced)
Basic (inc Proficient & Advanced)
96

Table 57: Overall Performance of Students: Average Scores & Correct Answers

Source: Na?onal Achievement Survey, 2021; Note: Scores are out of 500, converted to percent.
7.3.3 PARAKH Rashtriya Sarvekshan (PRS) (2024)
PRS (2024) survey did not es?mate the students mee?ng Minimum Learning Levels. It
presented only the average scores in Grades 3, 6 and 9 by type of school. The diﬀerence in
State Govt school performance rela?ve to Private schools is very similar to that in the NAS
survey of 2021, in FLN (Grade 3), Lower Secondary (Grade 6) and Upper Secondary (Grade 9),
with private schools out-performing States in the la?er two (last column, Table 58). As before
Central schools perform worse than State Govt schools in primary, but be?er in secondary
school in both subjects.
Table 58: Overall performance of students: Average scores (2024)

Source: PARAKH Rashtriya Sarvekshan, 2024
7.4 Digital Access in Schools
Virmani (2006) reviewed the social services provided by the Indian Govt, including educa?on
and health service, in the erstwhile Planning Commission, and highlighted the poor quality of
basic educa?on (1-8) and public health services (preven?ve & cura?ve) in rural areas where
2/3
rd
of the popula?on lived. He concluded that raising the quality of teaching and learning
to US & EU levels would take decades using conven?onal methods of expanding the training
& recruitment of high-quality teachers (doctors) and raise the teacher-pupil (doctor-pa?ent)
Class/Subjects
State Gov.Central GovtGov. aidedPrivate
State-Pvt
State Gov.Central GovtGov. aidedPrivate
State-Pvt
Class 3
Language64.661.863.064.8-0.263.056.062.062.01.0
Mathematics 61.057.858.860.60.458.052.057.057.01.0
Class 5
Language 60.462.660.663.8-3.455.056.053.057.0-2.0
Mathematics 56.057.055.056.8-0.845.044.042.043.02.0
Class 8
Language 57.064.659.866.2-9.248.059.048.060.0-12.0
Mathematics 49.651.848.652.4-2.836.039.031.038.0-2.0
Class10
Modern Indian Language48.853.450.051.6-2.839.045.040.043.0-4.0
Mathematics 42.047.041.845.6-3.632.036.029.034.0-2.0
English53.464.456.264.2-10.839.054.037.050.0-11.0
Average Scores Correct Answers
Grade/Subjects
State GovCentral GovGov. aidedPrivateState-Pvt
Grade 3
Language 646063640
Mathematics 615758601
Grade 6
Language 52695260-8
Mathematics 43614049-6
Grade 9
Language 48694959-11
Mathematics 33483339-6
Average across Subjects
Average across Subjects
Average across Subjects
97

ra?o. This was due to the fact that the rural popula?on of India itself was larger than the
popula?on of the USA & the EU combined. He proposed the universal use of online (internet)
resources, online teaching (diagnosis & prescrip?on) and expert systems, to rapidly close the
gap in teaching (medical treatment) quality.
Banerjee et al. (2007) evaluated two interven?ons in urban India: remedial educa?on and
computer-assisted learning. The remedial program hired young women (Bal-sakhis) to work
separately with the low-performing students. This produced clear learning gains, showing that
grouping students by ability and targe?ng instruc?on was more eﬀec?ve than reducing class
size and was also cost-eﬀec?ve. The computer-based math program provided fourth graders
with two hours of shared weekly computer ?me using educa?onal games. This signiﬁcantly
improved scores, indica?ng that even limited, curriculum-linked technology can enhance
learning outcomes.
Digital access in schools is gradually becoming a part of how students learn and teachers
teach, though progress is patchy, because many teachers and administrators are new to the
eﬀec?ve & appropriate use of digital tools in teaching and learning. Internet connec?on,
computers, laptops, tablets, projectors, and smart classrooms indicate how much technology
is used for learning. The presence of these tools also diﬀers between government, aided, and
private schools. Looking at changes over the years helps to understand where schools have
improved and where digital gaps s?ll remain.
Figure 27, presents the availability of func?onal computer facili?es for pedagogical purpose
has improved overall between 2018-19 and 2024-25.
Figure 27: Computer and Internet facility in Schools (% of Schools)

Source: Uniﬁed District Informa?on System for Educa?on Plus Note: Values for Private schools have been derived using the
absolute scores of Government-aided and Private unaided schools.
Government schools nearly doubled their access from 27.2 % to 52.7%, showing steady
expansion in computer facility. Private schools con?nued to lead with higher levels, rising from
51.9% to 72.3%. A dip around 2022-23 in both categories suggests some temporary decline,
possibly due to pandemic, before recovery in later years.
25
30
35
40
45
50
55
60
65
70
75
2018-192019-202020-212021-222022-232023-242024-25
% of school with computer
Schools with Functional Computer Facility
Private All management Government
5
15
25
35
45
55
65
75
2018-192019-202020-212021-222022-232023-242024-25
% of school with Internet.
Schools with Internet Facility
Private
All management
Government
98

Across all management types, the share increased from 33.5% to 57.9%, reﬂec?ng integra?on
of computers in classroom learning.
From 2018-19 to 2024-25, the share of schools with non-func?onal computers increased from
1.0% to 5.6% overall. The share of schools with internet access (Figure 27, 2
nd
panel) has
improved across all types between 2018-19 and 2024-25. Government schools saw a sharp
rise from 9.8% to 58.6%, especially aLer 2020-21, indica?ng strong digital expansion in the
post-pandemic years. Private schools consistently performed be?er, increasing from 42.4% to
76.4 % during the same period and maintaining a clear lead across the board. Overall, schools
under all management types grows from 18.7% to 63.5%. The biggest hike occurred between
2020-21 and 2022-23, sugges?ng greater focus on digital infrastructure despite the pandemic.
Between 2021-22 and 2024-25 (Table 59), access to computer and digital ini?a?ves in
schools improved across all categories. The data show progress in both the number of
func?onal digital devices and the percentage of schools equipped with ICT facili?es.
Table 59: Computer and Digital Ini?a?ves

Source: Uniﬁed District Informa?on System for Educa?on Plus; Note: Values for Private schools have been derived using the
absolute scores of Government-aided and Private unaided schools.
Overall, the share of schools with desktops or PCs rose by 9%, reaching about 35% in 2024-25.
The increase was stronger in private schools (up by 12.2%) than in government schools (7.5%).
Laptop and notebook availability also improved, rising by 6.9% overall, with private schools
again showing larger growth. Tablet availability recorded one of the most signiﬁcant increases,
up 17%, driven mainly by government schools where access rose by nearly 23%. Smart
classrooms also expanded rapidly across all school types up 15.9% overall, with private schools
showing a large gain of more than 20%. This marks a strong shiL toward technology-based
teaching environments.
There are many ways to display audio-visual material, including TVs, computers and ﬁlms/
projectors. Projector availability increased modestly by around 5% across all schools, while
the propor?on of schools using PCs for teaching and learning grew slightly by 2-4%. Access to
mobile phones for teaching improved as well, rising by about 8%, showing wider use of simple
digital tools at the classroom level. Digital library access, though s?ll limited, from 2.2% to
6.9% overall with both government and private schools repor?ng steady growth.
School Type
Percentage of schools, 2024-252021-22Change2024-252021-22Change2024-252021-22Change
Desktops/ PC's 34.925.99.024.016.57.561.549.312.2
Laptop/Notebook 19.812.96.911.25.55.740.530.69.9
Tablets available 26.09.017.032.39.422.912.78.54.2
PC's: Teaching Learning9.46.72.75.73.62.118.714.44.3
Projector Available 21.516.74.816.111.84.335.629.75.9
Smart Classroom 30.614.715.928.614.414.237.316.920.4
Mobile Phones for teaching25.617.77.923.216.66.632.021.110.9
Digital Library 6.92.24.75.41.24.210.94.66.3
AllGovernment Private
99

To see how eﬀec?vely these digital facili?es have been used, we corelate learning outcome
across States, with the availability of the digital facili?es in States. There are three categories
of facili?es. (1) Facili?es with a correla?on of about 1/3
rd
, (2) Facili?es with a correla?on of
around 0.2-0.25, and (3) facili?es with low or nega?ve correla?on. The most promising seem
to be computers and desktops for reading proﬁciency & lesser extent in arithme?c, in primary
school, and Internet access for arithme?c proﬁciency in lower secondary school (Table 60).
Interstate regressions however show an insigniﬁcant eﬀect of these digital facili?es on
diﬀerences in learning outcomes (MRP & MAP) across States in std V.
57
One of the results
from other studies is that teachers who cannot use digital facili?es are not in a posi?on to use
them to teach their students. Teacher training must include training on how to use digital
facili?es to improve learning outcomes of their students.
Table 60: Eﬀect of Computer and Digital Facili?es on Learning: Cross-State correla?on

Source: Uniﬁed District Informa?on System for Educa?on Plus, 2024-25
Muralidharan, Singh, and Ganimian (2019) evaluated Mindspark, a technology-aided
instruc?on program delivered through aLer-school centers in Delhi for low-income public-
school students. These centers ran six days a week for 90 minutes, combining self-paced
soLware use with small-group support from teaching assistant. The study included 619
students from public middle schools, half of whom were randomly given vouchers for free
access. The program was evaluated over 4.5 months using standardized paper-based tests,
made comparable across grades using Item Response Theory (IRT). The students with access
to Mindspark scored 0.37 SD higher in math and 0.23 SD higher in Hindi. Longer-term
projec?ons showing gains of 0.6 SD in math and 0.39 SD in Hindi with 90 days of a?endance.
The program beneﬁted all students, with academically weaker students gaining the most, and
improved competencies across arithme?c, word problems, frac?ons, geometry, pa?ern
recogni?on, and Hindi comprehension skills.
Sailer et al. (2021) study of teachers in Bavaria, Germany, found that while a minimum level
of technology is needed, the key factors for eﬀec?ve digital learning are actually teachers’ own
digital skills and their ability to teach with technology. In fact, teachers’ skills were much more
important than the amount of equipment available. He concludes that to unlock the beneﬁts
of digital learning especially more construc?ve, interac?ve, and problem-solving ac?vi?es for

57
The excep?on is, MAP (std 5) = 20.54 + 0.8 Internet - 0.66 Desktop/PC, with both coeﬃcients signiﬁcant at the 5% level. The nega?ve
coeﬃcient on computers is hard to explain!
India/State/ UT ComputerInternetDesktop/PCsSmart ClassroomLaptopTabletMobile
India57.963.534.9 30.619.826.025.6
Minimum Reading Proficiency (MRP)
Correlation (ASER: Std V) 0.320.160.32 0.260.08-0.050.03
Correlation (ASER: Std VIII) 0.220.040.29 0.130.11-0.06-0.15
Minimum Arithmetic Proficiency (MAP)
Correlation (ASER: Std V) 0.080.120.07 0.09-0.36-0.06-0.04
Correlation (ASER: Std VIII) 0.140.300.08 0.09-0.240.030.01
100

students’ schools and educa?on systems should focus on improving teachers’ training and
ongoing educa?on, speciﬁcally on how to use digital tools in teaching.
58

ALer the Central Govt introduced Atal Tinkering labs on an experimental basis, such labs have
been set up in many high schools since 2021-22. The share of schools with ?nkering labs rose
sharply from 1.8% in 2021-22 to 11.1% in 2024-5. The increase was 10.2% in private schools,
compared with a 7.1% rise in government schools (Table 61).
Table 61: Schools with Tinkering Labs (%)

Source: Uniﬁed District Informa?on System for Educa?on Plus
ICT lab facili?es at the higher secondary level have also increased. The share of schools with
ICT labs increased from 17.5% to 25.6%, while func?onal ICT labs rose from 14.0% to 21.2%,
showing a posi?ve change of more than 7%. Number of government schools with Upper
Primary, Secondary, and Higher Secondary sec?ons now have working ICT labs, reﬂec?ng
ongoing eﬀorts to expand digital infrastructure and promote hands-on learning environments
through ini?a?ves like Atal Tinkering Labs and ICT in Schools (Table 62).
Table 62: Schools having ICT Labs (%)

Source: Uniﬁed District Informa?on System for Educa?on Plus
8. School Dropouts need Job Skills
Paradoxically the focus on secondary school access and reten?on has detracted a?en?on from
the need to provide basic job skills to our vast casual labor force. Primary and secondary
school reten?on and dropout rates therefore are important for es?ma?ng the skilling
requirement of children at diﬀerent levels from primary, through lower secondary to upper
secondary. Reten?on rates measure movement from a given level to the next higher level,

58
This applies to all new materials, media and methods (3M).
School Type2024-252021-22Change
All Management11.11.89.3
Government 7.80.77.1
Government Aided9.24.15.1
Private Unaided15.04.810.2
Others 5.60.45.2
School Type 2024-252021-22Change
ICT Labs
Government 25.617.58.1
Government Aided 26.027.8
-1.8
ICT Labs (Functional)
Government 21.214.07.2
Government Aided21.221.00.2
101

while dropout rates measure those who have dropped out before comple?ng a given level.
Our analysis of minimum learning ability suggests, that the returns to further schooling may
be zero or even nega?ve, for those who do not meet minimum reading proﬁciency and/or
minimum arithme?c proﬁciency in the class they are in. We must not only raise MRP and MAP,
but also provide basic skills to those who have not acquired such proﬁciency.
59

Table 63 shows reten?on rates from 2018-19 to 2024-25 across key school stages under both
the earlier structure and the new NEP 2020 structure. Overall, reten?on has improved over
?me, par?cularly primary grades. In 2024-25, reten?on rates in Primary classes average 98.9%
in Class 2 and 92.4 % in Class 5. In Secondary school the reten?on rate is 82.8% in Class 8
(Lower Secondary), 62.9% in Class 10 and 47.2% in Class 12 (Higher Secondary). These rates
are at their peak for the past seven years for classes 5, 8 and 12 (Table 63).
Table 63: Reten?on Rate by level of educa?on (%)

Source: Uniﬁed District Informa?on System for Educa?on (UDISE+ Reports 2018-2025)
The pa?ern of reten?on rates across classes for 2024-25 is depicted in Figure 28. The
reten?on rate decreases slowly in Primary school (Classes 1-5), accelerates in Lower
Secondary (Classes 6-8) and then declined sharply in Upper Secondary (Classes 9-12). The
decline increases from 2.2% per year to 3.2% per year to 10.0% & 7.9% per year, respec?vely.
These students must be given basic job skills in school itself, or in nearby skilling centers,
related to local agriculture, industry, and services. NGOs can play an important role in
suppor?ng these students and connec?ng them with skilling opportuni?es.

59
Primary-level reten?on rate increased from 53.4% in 2003-04 to 58.1% in 2004-05, and further to 71.0% in 2005-06. It dropped to 70.3%
in 2006-07, before rising to 73.7% in 2007-08(DISE analy?cal Report 2007-08). Data is not available ?ll 2017 aLer which UDISE+ started UDISE
2012-13 could not be found.
Class 2Class 5Class 8Class 10Class 12
2018-19 86.371.258.2
2019-20 87.074.659.540.2
2020-21 95.380.961.542.8
2021-22 95.481.264.743.6
2022-2392.190.975.865.544.1
2023-2498.085.478.063.845.6
2024-2598.992.482.862.947.2
Avg (3yr)
96.389.678.964.145.6
Retention rate relative to class 1 entrants
102

Figure 28: Reten?on Rate Across Grades (% cohort)

Source: UDISE+ 2024-25, based on NEP structure; Note* Class X – 62.9% es?mated based on diﬀerent data set.
Dropout rates have declined in primary and secondary school (Table 64)
60
. In Primary school
(Classes 1-5), dropout rates declined sharply from 4.5% in 2018-19 to 0.3% in 2024-25, despite
a temporary rise to 7.8% aLer two years of pandemic. The data for Classes 3-5 is only available
for 3 years but follows the same intertemporal pa?ern as for Classes 1-3, with a post pandemic
rise to 8.7% followed by decline to 2.3% in 2024-25.
In Lower Secondary/Upper Primary (Classes 6-8), dropout rates follow the same pa?ern as in
Primary, declining from 4.7% in 2018-19 to 3.5% in 2024-25. The la?er are however
signiﬁcantly more than the 0.3% in primary school. Dropout rates also declined in upper
secondary school (Classes 9-10), from 17.9% in 2018-19 to 11.5% in 2024-25. For secondary
school as a whole (Classes 9-12) the dropout rates are now 8.2% (Table 64). Dropout rates
have therefore dropped signiﬁcantly in Primary and Secondary school.
Table 64: Dropout Rate in Primary and Secondary school (%)

Source: UDISE+ Reports 2018-2025

60
Dropout rate is deﬁned as propor?on of pupil from a cohort enrolled in a given level at a given school year who are no longer enrolled at
any grade in the following school year. Promo?on Rate + Repe??on Rate + Dropout Rate = 100
98.9
92.4
82.8
62.9*
47.2
45
50
55
60
65
70
75
80
85
90
95
100
II V VIII X XII
Cohort left in school (%)
Cohort Retention rate (est %) 2024-25
Cohort retention (%)
Level
Class(1 to 5)( 3 to 5)(6 to 8)(9 to 10)(9 to 12)
2018-194.5 4.717.9
2019-201.5 2.616.1
2020-210.8 1.914.6
2021-221.5 3.012.6
2022-237.88.78.116.413.8
2023-241.93.75.214.110.9
2024-250.32.33.511.58.2
Primary Secondary
103

Figure 29 shows the drop rates between classes calculated from the reten?on rates and the
within school dropout rates. Broadly, the drop-out rates between Primary and Lower
Secondary school and between the la?er and Upper Secondary are higher than the drop-out
rates within the Lower and Upper secondary schools.
The school and skilling system must both take account to these dropout rates. Providing basic
job skills to the drop outs is as if not more important than eﬀorts to keep them in school.
Eﬀorts to improve the percent of children with minimum learning proﬁciency, must go hand
in hand with provision of job skill to children who drop out before comple?ng Class 10.
Figure 29: Class-wise Dropout Rates: In-Class & Between-Class

Source: UDISE+ 2024-25, based on NEP structure.
Dropout between class is calculated using Transi?on rate from UDISE+ (2024-25).
9. Employment Eco-system: Skills for Employability
Figure 3 depicted the educa?on, skilling and employment systems as three pyramids which
must be linked to each other at every level (bo?om, middle and top), to produce eﬃcient
outcomes. Compared to the Educa?on system, built and evolved over 75 years the skilling
system is very fragmented and highly variable in quality and eﬃciency. This is partly due to
the fact that the economy consists of a large informal sector, with 57.7% of the workers self-
employed and another 19% working as casual laborer. Even within the 23.2% deﬁned as
regular workers, many of the them are employed in households or single worker enterprises,
so that the top of the employment pyramid consists of about 10% of the labor force.
61
To some
extent private markets have developed to train and place poten?ally skilled workers to high
wage jobs in companies, through placement and other services.

61
Our 34.4% of the regular workers have Bachelor of higher degree, yielding 8% of workers. We add another t 2% to account for ter?ary
educated and/or highly skilled self-employed workers, like consultants.
2.3
3.5
8.2
1.4
7.8
13.4
0
2
4
6
8
10
12
14
16
2 to 33-55 to 66-88 to 99-12
% of drop-out
Dropout rate (%) 2024-25
Dropout (in class)
Dropout(btwn class)
104

Figure 30: Employment Ecosystem: Educa?on, Skilling, Placement & Jobs

Note: Nos in (brackets) is change (2019 to 2025).
SIC = Skill India Center, CTS= CraLsmen Training Scheme, SIIC= Startup Incuba?on & Innova?on Center,
The rest of the system is a complex, fragmented, and nascent one in which the Central
Government, the State Govts, NGOs and MSME service companies, have a cri?cal role.
Figure 30 is an a?empt to show the complexity of the Indian employment ecosystem, linking
schooling, job skilling, placement and employment. This ﬁgure can be viewed as an inversion
of Figure 3, with the lowest levels on top (instead of at the bo?om), with pyramids converted
to cylinders for clarity of wri?ng, and the interlinkages between the pyramids/cylinders spelt
out in greater detail.
The ﬁrst column of Figure 30 represents the diﬀerent levels of the educa?on system from
Primary to Masters, which was depicted in Figure 3 as the educa?on pyramid. The third
column represents the diﬀerent levels of the skilling system, represented in Figure 3 as the
skilling pyramid, arranged from lowest to highest level of skilling. Some of these layers do not
exist or exist in very rudimentary form (e.g., NGOs at the lowest level), which should be viewed
as aspira?onal, rather than exis?ng today. The second column of Figure 30 links the two
pyramids through two parameters; School dropouts and job/skill counselling, some of which
is aspira?onal and may not presently exist. Linking the educa?on and skilling system at the
lower levels of educa?on is the dropout rate discussed in the previous sec?on. The
percentages shown at the primary middle and higher secondary level, represent rough
es?mates of the percentage of children who leave the educa?on system and enter the job
market. All these need to be channeled into the job skilling system, for provision diﬀerent
basic skills but this is not happening today.
1.
Education3. Skilling 4. Employability6. Jobs / Employment
5.
Job Placement
Primary
Middle
Higher Sec
Post Sec (NT)
Tertiary(Short)
B.Arts
B.Com
B.Pharma
B.Sc
BE/B.Tech
MA/MS
MCA
MBA
NGO
SIC, CTS
ITI, SIIC
Polytechnic
NSTI
85%
78%
18%*
15%
16%
28%
MRP 49%
MAP 31%
MRP 64%
MAP 41%
41%
29% (+11)
55
%+
25)
54 % (+25)
56%(+20)
58%(+11)
72%(+14)
Pvt. Skilling
Centers
78%(+42)
71%(+28)
E-Shram
Quicker Job
SIDH
Naukri.com
Foundit
WhatsApp
Team Lease
Linkedin
Helpers
(Domestic, Farm, construction
workers, Hotel & Office cleaner
Assistants
(Machine assemblers & operators
Craft & related trades workers)
Technicians
(Engineering, Medical & Pharma)
Associate Professionals
Managers
(Retell, whole sale trade, sales/
marketing, Human-resource
manager, Hotel manager)
Technical Manager
Professionals
(Health, Legal, ICT professionals,
Scientist, Legislative & Senior
Officials)
Higher Management
Pvt. Skilling
Centers
Schooling, Skilling, Employment System
Note: No. in (+) brackets is the change from 2019 to 2025, SIC = skill India center, CTS=craftsmen training scheme SIIC=Startup Incubation & Innovation Center
2. Dropout
105

The sixth column of Figure 30 consist of diﬀerent jobs from the minimum wage to the high
wage ones, represen?ng a more detailed version of the employment pyramid in Figure 3. The
link between the skilling pyramid/system and employment pyramid/system is divided into two
columns. Column four of Figure 30, headed employability, which contains es?mates of the
“employability” of the those with higher secondary and ter?ary educa?on and semi-skilled
(orange). At the primary and lower middle school educated, the es?mates are based on
Minimum Reading Proﬁciency and Minimum Arithme?c Proﬁciency. Column ﬁve of Figure 30,
headed job placement, is part of the job market, consis?ng of placement agencies and job
portals. While private pla?orms like Quikr Jobs, Naukri.com, Foundit, Team Lease, and
LinkedIn are reasonably eﬃcient in connec?ng skills to jobs at the high-end. EkStep
founda?ons Government portals such as NCS, SIDH, and e-Shram are trying to connect the
middle and low end of educated-skilled to jobs, they remain largely aspira?onal.
Empirical research in developed countries ﬁnds that Voca?onal and Technical educa?on
(VTE) graduates have faster transi?on to employment compared to general educa?on
graduates [Eichorst et al (2015), OECD (2021] and face lower unemployment rates in countries
with strong voca?onal systems [Deissenger et al (2015)]. They typically earn higher ini?al
wages than those with general educa?on, due to job-speciﬁc readiness, but the gap closes
over ?me. [Hanushek al (2017)]. This is not however true in developing countries, including in
India, because poorly linked training ins?tu?ons face skill mismatch and training content lags
behind evolving industrial needs [(King & Palmer (2010), ILO (2020)]. Job-skilling in India is
therefore an even greater challenge than school educa?on.
Virmani (2021, 2024) recommended greater use of expert systems (e.g. E-Guru for skilling)
and AI, to bridge informa?on asymmetries and integrate fragmented skill markets. Such an
expert system can advise and coordinate between skill seekers and skill providers, and
between employers and poten?al employees. EkStep founda?on has created (2025) an
“Open network for Employment and Skilling Transforma?on” (ONEST), on which it has built
the BlueDot AI ini?a?ve bridge the local informa?on gaps. Blue Dot accelerator and
Bharat10M accelerators are designed to promote and support startups, which can narrow the
knowledge gaps in the employment market (supply and demand for employment).
10. Skill for India, Skill for the World
In many developing countries, access to educa?on at diﬀerent levels was thought to be the
primary challenge. It was assumed that all kinds of good outcomes would follow automa?cally
from school enrolment. Later this automa?city was modiﬁed and a?ributed to school
comple?on. Quality of educa?on was almost an aLerthought, but “Job skilling” is perhaps an
addendum or add on. In developed countries, markets exist to meet demand for skills. In India
markets for skills exist at the upper end of the skilling pyramid (Figure 3), but are incomplete
at medium-high level and for the semi-skilled, and virtually non-existent for basic & low-level
skills. Thus, in a country in which almost 58% of workers are self-employed (“Nano-
106

enterprises”), a large frac?on of micro enterprises are one worker enterprises and almost 20%
of workers are casual workers, government (Central, State and Local) have to play a big role in
es?ma?ng current and future demand for skills, ensuring that ins?tu?ons for genera?ng and
supplying these skills exist, and helping match supply and demand.
The Skill India Mission was launched in 2015, aims to skill, re-skill, and up-skill individuals
through various training programs conducted in skill development centers and ins?tutes
under schemes such as PMKVY, JSS, NAPS, and CTS. Under PMKVY around 1.64 crore
individuals have been trained and 1.29 crore cer?ﬁed; Under JSS scheme 31.4 lakh trained
and 31.0 lakh cer?ﬁed; Under NAPS 40.8 lakh trained and 6.8 lakh cer?ﬁed; Under CTS (ITIs)
92.7 lakh enrolled with 55.9 lakh cer?ﬁed.
The next sub-sec?on 10.1 gives some facts about India’s Voca?onal Educa?on and Training
system, and compares India’s performance globally. Sub-sec?ons 10.2 and 10.3 survey the
skilling ins?tu?ons set up by the Central and State Govts. Sub-sec?on 10.4 compares the
skilling performance of Indian ﬁrms with other countries. Sub-sec?on gives an overview of the
skilling & employment portals, career conceding and placement agencies
10.1 Voca?onal Educa?on: Interna?onal Comparison
Interna?onal data on Voca?onal Educa?on and Training (VET) allows us to compare the
performance of India, with other countries, at the school level. The share of secondary
students enrolled in voca?onal and technical (VT) programs was 1.8% in 2017, well below the
expected/benchmark level at its per capita GDP of $7327 in 2017 (Figure 31).
Figure 31: Students in Voca?onal & Technical Programs (%)

Source: World Development Indicators, December 2024
This was a bigger gap than seen in any educa?onal a?ainment indicator in India. And larger
gap than for Thailand, which was also well below its benchmark level. Indonesia, China and
Mexico were above their expected/benchmark levels. We need to note however that the R-
India (2%)
Indonesia
China
Thailand
Mexico
UMIC (18%)
HIC (26%)
R² = 0.1277
0
5
10
15
20
25
30
35
40
0 20000 40000 60000 80000
Voc (% of Sec Students)
PcGDP , PPP (const 2021 int $)
Vocational & Technical Education
107

square of the benchmarking equa?on is quite low.
62

According to NEP 2020, less than 5% of the 19-24 age group had received formal voca?onal
educa?on (TwelLh Five-Year Plan, 2012-2017), primarily because voca?onal educa?on in the
past focused largely on Grades 11-12 and on dropouts in Grade 8 and upwards. In response
to the limited par?cipa?on in voca?onal educa?on, it proposed early exposure to voca?onal
learning. NEP recommends that all students should undertake a hands-on experience during
Grades 6-8, providing prac?cal experience in voca?onal craLs aligned with local skill needs. It
suggests that similar internship opportuni?es be made available to students throughout
Grades 6-12, to learn voca?onal subjects. Online op?ons should also be used.
The base for India in Figure 31 is 2017, therefore we es?mate the percent of students in VET
programs for 2023 (3.3%) assuming that the gap remains the same (-)12.8% points. This gap
is then added to the expected/benchmark (16.1%) for India at its 2023 per capita GDP. Based
on this we es?mate the improvement needed to reach UMIC & HIC benchmarks. This shows
that we need to improve percentage of students enrolled in voca?onal-technical programs by
over 15% points to reach the UMIC benchmark of 18.4% and by ~22.4% points to reach the
HIC benchmark of 25.7% for HIC. This is one of the biggest challenges facing job skilling in
India today. Voca?onal and technical educa?on must be expanded at all levels to improve
workforce skills and employability.
NEP 2020 aims to have 50% of learners through the school and higher educa?on system
exposed to voca?onal educa?on. A clear ac?on plan with targets and ?melines was to be
developed. The policy further states that focus areas for voca?onal educa?on will be chosen
based on skills gap analysis and mapping of local opportuni?es. To oversee and coordinate
this eﬀort, the Ministry of Human Resource Development (MHRD) would cons?tute a Na?onal
Commi?ee for the Integra?on of Voca?onal Educa?on (NCIVE), consis?ng of experts in
voca?onal educa?on and representa?ves from across Ministries, in collabora?on with
industry.
Interna?onal data is also available on VET courses at the post-secondary (non-ter?ary) level.
India’s post-secondary training at 14.4%, 5.8 percent points below it’s expected/benchmark
of 20.3% for its per capita GDP in 2023 (Figure 32). This is be?er than its secondary enrolment
in Voca?onal and technical training programs.
63
It is also comparable to the performance of
Indonesia (-12.1%), Vietnam (-5.1%), China (-11.8%), Mexico (-9.8%) and Thailand (-6.8%),
whose performance gap was worse than India’s. India however needs to increase the per cent
of students in post-secondar training programs by 9.4% points in 5 years to reach the UMIC
(2030) benchmark and by 20.6% points in 25 years to reach the HIC (2050) benchmark.

62
Secondary educa?on, general pupils- Secondary general pupils are the number of secondary students enrolled in general educa?on
programs, including teacher training. Secondary voca?onal pupils: Number of secondary students enrolled in technical and voca?onal
educa?on programs, including teacher training. (Secondary school students enrolled in voca?onal -technical program =100* Secondary
voca?onal pupils/ Secondary educa?on, general pupils).
63
The % of popula?on ages 25 and over that a?ained or completed post-secondary non-ter?ary educa?on.
108

Figure 32: Adults with Post-Secondary & Short Cycle (ter?ary) training (% of pop 25+)

Source: World Development Indicators, December 2024
India’s short-cycle ter?ary share was 3.2% points below the expected/benchmark level of
15.6% in 2021 (Figure 32, 2
nd
panel), based on its per capita GDP of 2021.
64
India’s poten?al
compe?tors had a larger performance gap on this indicator: Vietnam (-5.5%),
Indonesia (-8.8%), China (-7.1%), Thailand (-7.2%) and Mexico (-5.9%) were also below their
expected/benchmark levels.
The number for India in Figure 32 (2
nd
panel) is 2021. Therefore, we es?mate the value for
2023 (13.5%) assuming that the gap remains the same (-3.2% points). This gap is then added
to the expected/benchmark (16.7%) for India at its 2023 per capita GDP. Based on this we
es?mate the improvement needed to reach UMIC & HIC benchmarks. This shows that we need
to improve short-cycle ter?ary share by about 6.2% points to reach the UMIC benchmark of
19.7% and by ~15.7% points to reach the HIC benchmark of 29.2% for HIC.
NEP 2020 proposed to integrate voca?onal educa?on into the educa?onal oﬀerings of all
secondary schools, in a phased manner over the next decade. To support this integra?on, the
policy proposed collabora?on between secondary schools and Industrial Training Ins?tutes
(ITIs), polytechnics, and local industry, along with the development of shared skill laboratories
to strengthen prac?cal training infrastructure.
10.2 Schemes to improve Job skills
The Central Govt has introduced many schemes to improve the Indian skilling system. Jan
Shikshan Sansthan (JSS) provides technical and employability skills (NSQF level 2 & 3) to non-
literates, neo-literates, and school dropouts up to 12
th
standard
65
. Pradhan Mantri Kaushal
Vikas Yojana (PMKVY) (2015) oﬀers Short-Term Training (STT) for school or college dropouts

64
The % of popula?on ages 25 and over that a?ained or completed short-cycle ter?ary educa?on.
65
Jan Shikshan Sansthan (JSS), the Shramik Vidyapeeth (SVP), was launched in 1967. The scheme was oﬃcially renamed Jan Shikshan
Sansthan in the year 2000.
India (14%)
Indonesia
VietNam
China
Mexico
Thailand
UMIC (24%)
HIC (35%)
R² = 0.4855
0
5
10
15
20
25
30
35
40
01000020000300004000050000600007000080000
% of Pop. 25+
PcGDP , PPP (const 2021 int $)
Post Sec Training
India (12%)
VietNam
Indonesia
China
Thailand
Mexico
UMIC (20%)
HIC (29%)
R² = 0.4933
0
5
10
15
20
25
30
35
40
01000020000300004000050000600007000080000
% of Pop. 25+
PcGDP , PPP (const 2021 int $)
Short Tertiary Courses
109

and unemployed youth, focusing on soL skills, entrepreneurship, and ﬁnancial & digital
literacy. It also includes Recogni?on of Prior Learning (RPL), where individuals are assessed
and cer?ﬁed based on their prior learning experience. Under Special Projects, training is
provided in speciﬁc areas and within government or corporate premises.
Under PMKVY around 1.3Cr trained, 1.1Cr cer?ﬁed, and 21 lakh placed.
66
3,300 schools,
colleges, and higher educa?on ins?tu?ons covered under PMKVY, beneﬁ?ng 4.26 lakh
candidates
67
.PMKVY 4.0 (2022-26) related to Industry 4.0, AI, robo?cs, mechatronics, IoT, and
drones. Deen Dayal Upadhyaya Grameen Kaushalya Yojana (DDU-GKY) launched in 2014,
provides placement-linked skill training (3-12 months) to rural youth, covering over 250 trades
across sectors such as retail, hospitality, health, construc?on, automo?ve, leather, electrical,
plumbing, and gems & jewelry.
Under the vision of NEP (2020), 1,200 voca?onal skill labs were set up across 400 Navodaya
Vidyalayas and 200 EMRS schools. In Sikkim, 246 Livelihood Schools oﬀer three-month
training programs for educated unemployed youth in organic farming. In secondary and senior
secondary schools there is hands-on, life-skill-oriented training. These are supported by
Human Resource Development Department (HRDD) grants, tools, seeds, compos?ng
materials, and technical support from line departments.
The CraLsmen Training Scheme (CTS) (1950), is a voca?onal training program aimed at
ensuring a supply of skilled workers for industry by providing employable skills to youth
through systema?c training. CTS courses oﬀered through Government and Private it is, NSQF-
compliant courses of 6 months to 2 years dura?on across 169 trades, classiﬁed into 86
engineering trades (Technical, mechanical, electrical, electronics, manufacturing,
construc?on, and industrial), 78 non-engineering trades (IT-enabled services, design, oﬃce
administra?on, healthcare, hospitality, tourism, agriculture, retail, and social services), and 5
trades for Persons with Disabili?es Currently, CTS operates through 14,643 ITIs (3,331
government and 11,312 private), enrolling about 26.58 lakh trainees in 1-year and 2-year
courses across 169 NSQF-compliant trades, with the objec?ve of mee?ng current and future
skilled manpower requirements of the economy.
PM Vishwakarma launched in 2023, supports ar?sans and craLspeople through cer?ﬁcate
recogni?on, skill upgrada?on (basic and advanced training), toolkit incen?ves, credit and
marke?ng support, and incen?ves for digital transac?ons. About 12.3 lakh candidates have
enrolled for basic training, and 12 lakhs have completed it.
SAMARTH launched in 2017, Scheme for Capacity Building in the Tex?les Sector provides entry
level training, upskilling, and reskilling, for jobs in tex?les and related sectors.
Several schemes have been introduced by the Center and states to improve the quality of
post- secondary skilling. One of these is Future Skills Prime (2018) oﬀers industry-backed

66
Source: Legacy PMKVY Dashboard: All
67
Human Capital for Viskit Bharat, “Skilling: Future Ready Workforce”, FiLh Na?onal CS Conference.
110

courses and cer?ﬁca?ons to build in-demand digital skills, aligned with Na?onal Occupa?onal
Standards (NOS) and Na?onal Skills Qualiﬁca?ons Framework (NSQF). States have also created
new ins?tu?ons to provide post-secondary skills. These are given in Sec?on 10.3.
10.3 Skilling Ins?tu?ons
India requires a strong network of skilling ins?tu?ons to address the growing challenge of
unemployability among youth, especially those who drop out of school or lack basic
educa?on. A large sec?on of young people enters the labor market without adequate skills,
making them unprepared for available jobs. Skilling ins?tu?ons play a crucial role in ﬁlling this
gap by providing structured, job-oriented training. A strong ins?tu?onal framework is
therefore essen?al to improve employability, create a steady pipeline of skilled workers, and
link learning with livelihood.
10.3.1 Ministry of Skill Development & Entrepreneurship
The Ministry of Skill Development and Entrepreneurship (MSDE) (2014) coordinates all skill
development eﬀorts in India. It works to reduce the gap between the demand and supply of
skilled workers and build a strong system for voca?onal and technical training. The Ministry
focuses on upgrading exis?ng skills and developing new ones to prepare youth for current and
future jobs.
MSDE oversees India’s na?onal skilling system through a set of specialized ins?tu?ons (Figure
33). The Na?onal Council for Voca?onal Educa?on and Training (NCVET) func?ons as the
regulator for uniﬁed regulator for skills training approving qualiﬁca?ons, recognizing awarding
and assessment bodies, monitoring implementa?on, and ensuring grievance redressal. The
Directorate General of Training (DGT) oversees Industrial Training Ins?tutes (ITIs), Na?onal
Skill Training Ins?tutes (NSTIs) and is the apex organiza?on for development and coordina?on
of programs for voca?onal training ins?tutes.
The Na?onal Skill Development Corpora?on (NSDC) acts as an implementer for various skilling
schemes of MSDE and other Ministries/Departments, such as, PMKVY, NAPS, PM
Vishwakarma, etc. and the Na?onal Skill Development Fund (NSDF) for raising funds from
Government and Non-Government sectors for skill development in the country. The Indian
Ins?tute of Entrepreneurship (IIE) provide training, research, and consultancy services in the
ﬁeld of SMEs, with a special focus on women entrepreneurship development. The Na?onal
Instruc?onal Media Ins?tute (NIMI) develops instruc?onal material for voca?onal training and
digital content; ongoing reforms within the skill ecosystem highlight the need to expand such
material in video and visual formats to support pedagogy prac?ces.
Together, these ins?tu?ons create the administra?ve, regulatory and industry-linked
framework that anchors MSDE’s skilling ini?a?ves.
111

Figure 33: Ins?tu?onal Framework of MSDE

Source: MSDE, Review of Output Outcome Monitoring Framework (OOMF), 2024-25
Skill India Centre, are located in every State and UT. They oﬀer short-term, NSQF-aligned
training across mul?ple interest areas such as IT-ITES, electronics, apparel and tex?les, beauty
and wellness, healthcare, construc?on, automo?ve, telecom, logis?cs, tourism and
hospitality, agriculture, retail, green jobs, BFSI, media and entertainment, power, capital
goods, and aerospace and avia?on. Courses are job-oriented and include roles like data entry
operator, electrician, plumber, healthcare aide, beau?cian, sewing machine operator, junior
soLware developer, solar PV installer, automo?ve service technician, retail sales associate,
oﬃce assistant, and related entry- to mid-level occupa?ons aligned with local labor demand.
Industrial Training Ins?tutes (ITIs), and Polytechnics are under the administra?ve and ﬁnancial
control of State Governments or Union Territory Administra?ons.
10.3.2 Industrial Training Ins?tutes (ITIs)
The capacity of Industrial Training Ins?tutes (ITIs) in India has expanded over ?me, but
outcomes in terms of student comple?on have varied across years. The number of students
enrolled increased from 1.3 lakh in 2019 to 14.5 lakh in 2023 and slightly to 13.3 lakh in 2024
(Figure 34). The number of students passing also rose from 0.1 lakh in 2019 to 6.5 lakh in 2022,
reaching 11.0 lakh in 2023 and 10.8 lakh in 2024. As a result, the pass rate improved from
7.1% in 2019 to 81.6% in 2024 (Figure 34). These outcomes must be interpreted cau?ously,
as student enrolment and comple?on were aﬀected by the pandemic during 2020 and 2021.
The sharp increase in enrolment and pass rates between 2019 and 2022 may be due to data
errors. These ﬁgures are currently under veriﬁca?on and may be subject to revision.
112

Figure 34: Students in Industrial Training Ins?tutes (ITI): Enrolled & Passed

Source: ITI Public Dashboard, Last Updated on 02/03/2026
Table 65 provides a breakdown of students by type of ins?tu?on, course dura?on, and loca?on
from 2020 to 2024. A notable pa?ern is the con?nued dominance of private ITIs in enrolment,
though government ITIs has also shown moderate growth (Table 65). The number of students
enrolled in Government ITIs increased from 1.2 lakh in 2020 to 5.8 lakh in 2024, with an annual
growth rate of 48% likely due to lower fees and be?er recogni?on of cer?ﬁcates.
Table 65: Students enrolled in ITI by Ins?tu?on type & course dura?on

Source: ITI Public Dashboard, Last Updated on 02/03/2026, Note: Data for 2019 has not been included, as the ﬁgures are not
comparable with subsequent years and show inconsistencies.
There was also a signiﬁcant shiL from long dura?on to short dura?on courses. The two-year
courses con?nue to account for the largest share of students, with enrolment increasing from
2.1 lakh in 2020 to 9.2 lakh in 2024.
7.1%
3.0% 8.8%
68.8%
76.0%
81.6%
0
10
20
30
40
50
60
70
80
90
0
2
4
6
8
10
12
14
201920202021202220232024
Student Passed (%)
(Stduents in Lakhs)
ITI Enrolled & Passed Students
Enrolled
Passed
Pass/Enroll
All India ShareCAGR(%)
202020212022202320242024
2024/2020
Institute Type
Government1.21.13.56.65.80.448
Private 2.02.05.97.97.50.638
Course Duration
6 months 0.00.00.00.00.00.045
1 year 1.10.90.94.54.10.338
2 year 2.12.28.59.99.20.744
Location
Urban 1.00.92.84.43.90.341
Rural 2.32.26.610.19.30.742
Total 3.23.19.414.513.3
Number of Students (in lakhs)
113

Enrolment in one-year courses also increased from 1.1 lakh to 4.1 lakh over the same period.
Short dura?on six-month courses con?nue to have small share of total enrolments, though
they have shown rapid growth over ?me, with an annual growth rate of about 45%. In terms
of loca?on, ITIs con?nue to serve a predominantly rural popula?on. Enrolment in rural areas
increased from 2.3 lakh in 2020 to 9.3 lakh in 2024, 70% of total enrolments, while urban areas
accounted for the remaining 30% (Table 65).
Table 66 shows a breakdown of enrolments by stream (engineering and non-engineering) and
type of occupa?on (industrial/service and grey/white-collar). Students enrolled in Engineering
trades con?nue to dominate ITI enrolment, 85% of total students in 2024, while non-
engineering trades account for the remaining 15%.
Enrolment in engineering trades increased from 2.6 lakh in 2020 to 11.2 lakh in 2024,
reﬂec?ng strong growth over ?me. Most students within the engineering stream are enrolled
in industrial work trades. Enrolment in Industrial work trades increased from 2.5 lakh in 2020
to 10.8 lakh in 2024 (Table 66). These trades are related to mining, manufacturing, and
construc?on, as well as repair and maintenance ac?vi?es in both domes?c and commercial
spaces. They include electrician, liL mechanic, painter, and refrigera?on and air condi?oning
technician. Grey/white-collar engineering trades, which typically include oﬃce-based
technical and support roles such as laboratory assistant, draughtsman, or surveyor, con?nue
to account for a small share of enrolments, though they have also grown over ?me (47%).
Table 66: ITI enrollment in engineering & non-engineering streams (%)

Source: ITI Public Dashboard, Last Updated on 02/03/2026, Note: Data for 2019 has not been included, as the ﬁgures are not
comparable with subsequent years and show inconsistencies.
Non-engineering trades account for a smaller share of total enrolments, increasing from 0.6
lakh in 2020 to 2.3 lakh in 2023 to 2.0 lakh in 2024 (Table 66). Within this stream, enrolment
in service work trades increased from 0.4 lakh in 2020 to 1.1 lakh in 2023, slightly to 0.9 lakh
in 2024. These include trades such as ar?sans using advanced tools, bamboo and wood
workers, sewing technology, and weaving technicians. Grey/white-collar non-engineering
trades include personal and service-sector occupa?ons such as spa therapy, tourist guide,
radiology technician, photographer, human resource assistant, interior designer, and data
entry operator, and have shown strong growth over ?me with an annual growth rate of about
43% as compare to Service work (27%).
All India ShareCAGR(%)
Stream 202020212022202320242024
2024/2020
Engineering 2.62.68.912.111.20.8544
Industrial work2.52.58.511.710.80.9744
Grey/White collar0.10.10.30.40.40.0347
Non-Engineering0.60.50.52.32.00.1535
Service work0.40.30.31.10.90.4627
Grey/White collar0.30.20.21.21.10.5443
Total 3.23.19.414.513.2
Number of Students (in lakhs)
114

Figure 35 gives a compara?ve analysis of the States in terms of the availability of ITIs and
enrolment in ITIs for the latest available year (2024). For interstate comparison, we divide the
number of ITIs by the popula?on of the State in crore, and the number of students enrolled
by the State’s popula?on in lakhs, and rank them by the la?er.
Enrolment is highest in Lakshadweep (494 students per lakh popula?on), followed by
Himachal Pradesh (255), Andaman & Nicobar Islands (207), Haryana (172), and Goa (138).
Among the larger states, U?ar Pradesh (134), Punjab (136), and Odisha (118) also record
rela?vely high enrolment levels per capita. In contrast, Assam (12), Nagaland (17), Mizoram
(22), and Manipur (22) record low enrolment per capita, indica?ng rela?vely limited
par?cipa?on in ITI training.
Figure 35: Students enrolled in ITIs across States & UTs

Source: ITI Public Dashboard, Last Updated on 03/03/2026 and PIB, “Opera?onal of Industrial Training Ins?tute”, 5/08/2024
A comparison of enrolment levels with the number of ITIs also reveals signiﬁcant varia?on
across states. Some states such as Himachal Pradesh, Rajasthan, Karnataka and U?arakhand
have a rela?vely large number of ITIs in rela?on to their enrolment levels. This suggests that
these states may need to consider consolida?on and be?er u?liza?on of exis?ng ins?tu?ons
rather than further expansion.
On the other hand, states such as Puducherry, Kerala, Telangana, and Madhya Pradesh also
show a dispropor?onate gap between enrolment levels and the number of it is (Figure 35).
Overall, the na?onal average stands at about 95 students enrolled per lakh popula?on and
107 ITIs per crore popula?on. The wide varia?on between states indicates uneven access to
ITI training opportuni?es across the country.

0
50
100
150
200
250
300
350
400
450
500
Assam
Nagaland
Mizoram
Manipur
Meghalaya
WB
Puducherry
Arunachal
Delhi
Tamil Nadu
Sikkim
Tripura
DNHDD
J&K
Chandigarh
CGH
UK
MP
Kerala
Telangana
Bihar
MH
Karnataka
Andhra
Jharkhand
Gujarat
Rajasthan
Odisha
Ladakh
UP
Punjab
Goa
Haryana
A&N Islands
HP
LakshaDw
Availability of Skilling: ITI per Population (2024)
Enrolled/Lakh popITI/Crore pop
115

At present, there are 14,615 ITI func?oning across India, comprising 3,316 Government and
11,299 Private ITIs.
68
During the last ﬁve years (2020-2024), a total of 433 new ITIs was
established: 146 Government and 287 Private. Among states, West Bengal (38 Government,
11 Private) and Odisha (24 Government) recorded the highest number of new government
ITIs. In the private side, expansion was led by U?ar Pradesh (101 Private, 35% share) and
Maharashtra (54 Private, 19% share), followed by Bihar (46 Private, 16% share).
Within the skilling ecosystem, Industrial Training Ins?tutes (ITIs) play a central role in providing
voca?onal and technical training. However, despite their expansion, there remains a need to
strengthen the quality, relevance, and infrastructure of ITIs to meet industry requirements.
Several ini?a?ves have been introduced to improve ins?tu?onal capacity and align training
with labor market demand.
10.3.3 Schemes for Upgrading Skilling Infrastructure
A number of measures have been taken to improve the ins?tu?onal infrastructure and
pedagogy of medium level skills and skills for semi-skilled jobs:
SANKALP (2018): Skill Acquisi?on and Knowledge Awareness for Livelihood Promo?on, is a
World Bank supported ini?a?ve aimed at improving short-term skill training by strengthening
ins?tu?onal frameworks and delivery systems.
PMSETU: Pradhan Mantri Skilling and Employability Transforma?on through Upgraded ITIs, A
CSS with investment 60Cr to upgrade 1,000 Government ITIs in a hub-and-spoke model (200
hubs and 800 spokes). The scheme includes advanced infrastructure, modern trades, digital
learning systems, innova?on centers, training-of-trainers facili?es, produc?on units, and
placement services, support from the World Bank and Asian Development Bank.
STRIVE (Skills Strengthening for Industrial Value Enhancement), World Bank assisted project
aimed at improving the relevance and eﬃciency of skills training through ITIs and
appren?ceships. Under the project, 500 ITIs (467 Government and 33 Private) were upgraded
with improved infrastructure, labs, equipment, and tools.
State govts have also made eﬀorts to ﬁll gaps in there skilling system by introducing new
ins?tu?ons. Examples of these are given below:
Honda Voca?onal Training Ins?tute (HVTI) in collabora?on with the “Automo?ve Skill
Development Council (ASDC)” at Tapukara, Rajasthan, is a state-of-the-art training center for
automo?ve skilling. It oﬀers specialized courses for 12th-grade pass-outs and IT graduates,
training over 3,000 youth annually with a 75% placement rate.
The “Kalike Jothege Kaushalya” ini?a?ve, implemented by the Karnataka Skill Development,
integrates skill-based learning within the formal educa?on system through a “50-hour, credit-
linked training per semester in partnership with leading industries, to 168 colleges across
seven universi?es under KJK 2.0, covering high-demand domains and beneﬁ?ng over 66,000

68
Industrial Training Ins?tutes in the country, posted On: 30
th
July 2025 by PIB Delhi
116

students through strengthened industry academia linkages, internships, appren?ceships, and
placements.
10.3.4 Polytechnics
In India, there are “Standalone Ins?tu?ons”, which are outside the purview of both
Universi?es/colleges and the school system, but require recogni?on from a statutory body.
These include polytechnics and ins?tutes for Nursing, Teacher training, Paramedical and Post-
graduate diploma in management (PGDM). Polytechnics cons?tute 32% of standalone
technical ins?tu?ons (12,000), with nursing colleges (30%) and teacher training ins?tu?ons
(28%). There are 3550 to 3,850 listed ins?tutes across 35 States and Union Territories, oﬀering
around 586 diploma courses with an approved intake of more than 11 lakh students. About
63.5% of Polytechnics are privately managed and 36.5% are government managed, mostly by
central govt. Most privately managed polytechnics are self-ﬁnancing. Privately managed
ins?tu?ons include Private State universi?es. Government Polytechnics ensure baseline
competence of faculty through higher pay and compe??ve exams. Their infrastructure though
func?onal is dated, resul?ng wide gap between industry requirement and learning of
students. Private polytechnics invest in modern laboratories and smart classrooms aligned
with contemporary standards. There is however a wider range in the quality of faculty, from
excellent researchers to poor teachers, so learning outcomes are correspondingly diverse.
Diploma programs are concentrated in core engineering disciplines such as Civil, Mechanical,
Electrical, Computer, Electronics and Communica?on, and Automobile Engineering. The
distribu?on of polytechnic colleges is uneven across states, with higher concentra?on in larger
states and limited presence in smaller and northeastern states. There are many gaps in
approval status and data reliability, so the data used is not as good as for ITIs. A?empts are
being made by AICTE to move from rigid regula?ons to transparent disclosure of data relevant
to student-parent choices. The Approval process Handbook (2024-27) mandates disclosure
department-wise disclosure of infrastructure, faculty creden?als and student faculty rates.
This will facilitate autonomy among be?er performing ins?tu?ons.
Figure 36 shows the availability of polytechnics and the availability of seats in polytechnics,
across Stats and UTs. It plots the number of Polytechnics per crore and the number of seats
per lakh of popula?on in State, and ranks them by the la?er. Tamil Nadu, Andhra Pradesh,
Goa, Maharashtra and Haryana are ranked on top. It is noteworthy that Maharashtra, Haryana
and Tamil Nadu are among the largest producers of manufactured goods. Other States
wan?ng to a?ract manufacturing need to improve the availability and quality of polytechnics
in their State.
117

Figure 36: Availability of Polytechnics Seats, across States and UTs

Sources: h?ps://list.polytechniccolleges.in/
10.3.5 Employability of ITI and Polytechnic educated
The eﬀorts of the Central and State Govts have improved the employability of the semi-skilled
emerging from job skilling ins?tu?ons, but much remains to be done. Table 67, shows
employable talent percentages across diﬀerent educa?onal domains from 2019 to 2026. The
share of graduates from Industrial Training Ins?tute (ITI) programs who are employable has
increased from 31.3% in 2022 to 46.0% in 2026. The share of graduates of Polytechnics who
are employable has increased from 18.1% in 2019 to 32.9% in 2026. Between 2022 and 2026,
the improvement in employability of ITI graduates 14.7% points) is greater than for
Polytechnics (11.5% points), increasing the performance gap since 2022. A greater eﬀort is
needed to improve the job skill related learning outcomes of Polytechnics.
Table 67: Employability (Below Graduate), 2019-2025

Source: Global Employability Report 2025 & 2026, Global Employability Test (GET)
There is a wide employability gap between States’ skilling systems, with Andhra Pradesh
graduates twice as employable as UP graduates of it is and polytechnics. Andhra Pradesh's
implementa?on of the NAIPUNYAM skill census portal (assess-learn-apply framework) and
cascading skill hub system tries to align polytechnic output with state labor market demand.
0
20
40
60
80
100
120
140
160
180
200
Meghalaya
Manipur
Assam
Mizoram
Nagaland
DNHDD
Bihar
Rajasthan
Chhattisgarh
Tripura
J&K
Arunachal
West Bengal
Madhya
Jharkhand
Uttar Pradesh
Himachal
Gujarat
Delhi
Chandigarh
Sikkim
Punjab
Kerala
Uttarakhand
Puducherry
Telangana
A&N
Odisha
Haryana
Karnataka
Maharashtra
Goa
Andhra
Tamil Nadu
Ratio
Availability of Skilling: PolyTechnic per Population
Seats/LakhPolyTechnics/Cr
Year20192020202120222023202420252026CAGR
Polytechnic18.132.025.021.427.622.429.032.99.0
Industrial Training Institutes (ITI)31.334.240.041.046.010.1
Aggregate47.346.245.946.250.351.354.856.42.5
Domains Wise Employable Talent 2019-2026
118

U?ar Pradesh needs to improve coordina?on between state educa?on directorates, industry
chambers, and training ins?tu?ons. The JEECUP (Joint Entrance Examina?on Polytechnic)
admissions system focuses on group-based tracking, which limits ﬂexibility. Data on
employment-outcome is cri?cal for upda?ng curriculum and supply-side planning to match
demand.
10.3.6 Non-Govt Organiza?ons (NGOs) in Skilling
NGO/NPO-led Skill Development Ini?a?ves, complement government eﬀorts by providing
short-term, employment-oriented training through community-based and industry-linked
programs. These are par?cularly important for less educated youth who need to be trained in
basic job skills for poten?al jobs at the local level and for semi-skilled jobs at the local &
community level. But private for-proﬁt skilling companies also play a cri?cal role in developing
future ready skills.
Nitya Founda?on Skill Training Program provides voca?onal training in areas such as
embroidery, art and craL, beau?cian, technician work, computer training, educa?on, and
tourism through experienced trainers. It has also trained youth under government schemes
like DDU-GKY, Seekho aur Kamao, and Gharib Nawaz Skill Development.
Jindal South-West (JSW) Skills School aims to improve the employability of youth and women
by oﬀering industry-focused voca?onal skill courses.
Both Satya Shak? Founda?on (SSF) and Amuja Founda?on (SEDI) provide a combina?on of
trade skills and soL skills training, along with on-the-job training (OJT), career counseling, and
placement support to enhance employability.
The “Telangana Academy for Skill and Knowledge (TASK)”, a NPO by the Government of
Telangana, strengthens industry-academia-government linkages by partnering with global
technology leaders, se?ng up “112 Advanced Technology Centers”, suppor?ng ini?a?ves like
Genome Valley and Pharma City appren?ceships, and enabling real-?me job matching
through the “T-GATE pla?orm”, thereby providing aﬀordable, high-quality training, faculty
development, and a pool of job-ready talent.
Na?onal Ins?tute of Informa?on Technology (NIIT Founda?on) started in 2004, runs Skill
Development Centers oﬀering short-term courses such as Professional Edge, Advanced Word-
Excel, Oﬃce Automa?on, CRM BPO, Retail & Inventory Management, BFSI, Digital Marke?ng,
Spoken English, Financial & Digital Literacy, and Career EDGE IT.
10.4 Training the Trainers: NSTI and CITS
The eﬀec?veness of Industrial Training Ins?tutes (ITIs) depends on the availability and
qualiﬁca?on of trainers. Only 46.9% (Table 68) of the 2.17 lakh sanc?oned trainer posi?ons
are ﬁlled across all ITIs. The shor?all is more acute in government ITIs (42.2%) compared to
private ITIs (49.7%). Among the ﬁlled posi?ons, only 18.4% of trainers (around 18,779
individuals) possess the CraL Instructor Training Scheme (CITS) cer?ﬁca?on while a large
119

majority (over 83,000 trainers) are yet to be formally cer?ﬁed. In terms of employment type,
regular posi?ons dominate (78.2%), but a notable share of trainers are on contract (16.1%) or
in other (5.8%) on temporary arrangements.
Table 68: Skilling and cer?fying Trainers (2022)

Source: NCVT MIS Instructor Dashboard (as of July 28,2025)
Eﬀorts are being made to increase the supply of trained faculty and to upgrade their
capabili?es. The Na?onal Skill Training Ins?tute (NSTI), provides short-dura?on, modular and
customized training programs for engineers, supervisors, technicians, industrial personnel,
and faculty of technical ins?tu?ons, like ITIs. The ins?tute focuses on higher-level skill
upgrada?on using industry-relevant equipment and training aligned with the requirements of
industries, government establishments, PSUs, and technical ins?tu?ons.
The CraL Instructor Training Scheme (CITS) for training instructors engaged in ITIs under the
CraLsmen Training Scheme (CTS). Opera?onal since the incep?on of CTS, CITS provides
training techniques of transferring hands-on skills to trainees. During 2024-25, CITS recorded
10,732 admissions against a sea?ng capacity of 17,475. The program oﬀers one-year, NSQF-
aligned courses across 55 trades, covering both engineering and non-engineering disciplines.
The CraL Instructor Training Scheme (CITS) plays a central role in preparing qualiﬁed
instructors for ITIs. In 2021-22, a total of 8,133 trainees were registered for CITS, out of which
6,916 appeared and 5,949 passed the examina?on (Table 69). As per the ins?tu?onal divide:
government ins?tutes account for nearly 89% of successful trainees (5,297), while private
ins?tutes contribute only 11% (652). By trade category, most candidates are from industrial
and service-oriented trades (76.5%), which include technical and opera?onal roles such as
ﬁ?er, electrician, mechanic, or related. The remaining 23.5% represent grey/white-collar
trades, covering roles such as laboratory assistant, draughtsman, or oﬃce-related instructor
training.
Position Sanctioned Filled %
Total 21772410220746.9
Gov 861543638142.2
Pvt 1315706538249.7
CITS certified
Yes1877918.4
No83493
Employment Type
Regular7990578.2
Contract1643716.1
Others59305.8
120

Table 69: CraL Instructor Training

Source: Directorate General of Training, CITS Dashboard (as of July 28,2025)
10.5 Training and Skilling in Firms
“Learning by doing” in industry is one of the most celebrated skilling avenues in the growth
of human capital, produc?vity and real wage growth. But formal training courses in ﬁrms, are
also important in crea?ng a culture of quality and learning in ﬁrms. According to Table 70, the
percentage of ﬁrms providing formal training to their permanent, full-?me employees in India
stood at 7.7% in 2022, down from 35.9% in 2014. This compares with 8.7% of ﬁrms in Vietnam
and 8.4% for Indonesia in 2023. Thailand recorded 18% in 2016. In contrast, Mexico reached
37.8% in 2023, and Malaysia reported 24% in 2019.
Table 70: Firms oﬀering formal training (% of ﬁrms)

Source: World Development Indicators, July 2025
Given the large informal sector organized ﬁrms are an important part of the skilling eco-
system. Industry par?cipa?on in training helps bridge the gap between classroom instruc?on
and workplace requirements. Several government ini?a?ves have been designed to promote
appren?ceships, internships, and digital skill development through ﬁrm-led models.
The Na?onal Appren?ceship Promo?on Scheme (NAPS), launched in 2016, builds on the
framework of the Appren?ces Act, 1961, which governs appren?ceship training in India.
Earlier, appren?ceship programs were largely conducted under the Na?onal Appren?ceship
Training Scheme (NATS for graduate and diploma students in technical ﬁelds), managed by the
Ministry of Educa?on. NAPS, introduced by the MSDE, expanded this system to include ITI
graduates, school dropouts, and non-technical trades. It provides basic and on-the-job
training along with a s?pend. Under the scheme, 45 lakh appren?ces have been engaged, 23
CITS TrainingRegistered Appeared Passed%
2021-22 Traineein Exam
Institute Type813369165949
Gov 59825826529789.0
Pvt 2151109065211.0
Trade
Blue collar 64835410455376.5
Grey/White collar16501506139623.5
Country YearPcGDP
Firms (%)
Lower Middle Income Countries
India 2022$8,5457.7
Viet Nam2023$13,4928.7
Upper Middle Income Countries
Indonesia2023$13,8908.4
Thailand2016$19,37018.0
Mexico2023$22,14337.8
Malaysia20193130124.0
121

lakhs have completed training, and 6 lakhs have been cer?ﬁed
69
. NAPS 2.0, launched in 2022,
con?nues to promote appren?ceship by transferring s?pend support directly to appren?ces
through Direct Beneﬁt Transfer.
PM’s Internship Scheme (2024) to provide internship opportuni?es to one crore youth in the
top 500 companies over ﬁve years. The 12-month internship includes at least six months of
on-the-job experience, with monthly assistance of ₹4,500 from the Government and ₹500
from industry.
10.6 Skilling & Employment Portals & Placement ﬁrms
A strong employment ecosystem requires eﬀec?ve linkages between training and actual job
opportuni?es (Figure 30, column 4). The Government has made a big eﬀort to ﬁll the gap in
job matching and labor market services, through crea?on skilling and employment portals
that facilitate the link diﬀerent types of workers to available jobs.
To address these coordina?on gaps and create a seamless transi?on from skilling to
employment, the Government of India launched the Skill India Digital Hub (SIDH) in 2023.
This ﬂagship ini?a?ve integrates Skill India and Digital India under a single digital public
infrastructure for educa?on, skilling, employment, and entrepreneurship. SIDH brings
together various ini?a?ves such as PMKVY, Jan Shikshan Sansthan (JSS), Appren?ceship, PM
Vishwakarma and ITI training processes onto one digital pla?orm. It enables the en?re
lifecycle of candidates from enrolment, training, assessment, and cer?ﬁca?on to placement
to be managed digitally and transparently.
By linking with Udyam, e-Shram, NCS, and ASEEM portals, SIDH ensures interoperability
between training, job matching, and social welfare databases. With over 5,000 courses, AI-
driven career recommenda?ons, and QR-enabled digital cer?ﬁcates, the pla?orm supports
evidence-based monitoring and personalized employability pathways. This convergence
marks a major step toward transforming India’s fragmented skilling landscape into a
connected and data-driven employment ecosystem.
The ci?zen-centric design (Figure 37) of the Skill India Digital Hub (SIDH), which connects key
components of the skill ecosystem: educa?on, skilling, employment, entrepreneurship, and
industry linkages. Each node represents a service layer integrated into SIDH, ensuring that
individuals can access courses, cer?ﬁca?ons, job opportuni?es, and entrepreneurial support
through a single uniﬁed digital pla?orm.
Employment services in India are supported by both government and private digital
pla?orms. The Ministry of Labor & Employment plays a central role through the Na?onal
Career Service (NCS), eShram which connects job seekers, employers, and training providers.
Alongside this, several private portals have emerged to match workers with opportuni?es
across sectors. Together, these pla?orms form an important bridge between skilling and
employment, helping youth access jobs, appren?ceships, and career informa?on more easily.

69
Data sourced from Appren?ceship Performance Dashboard (as of 10/Nov/2025)
122

Figure 37: Skill India Digital Hub (SIDH): One pla?orm for all skilling needs

Source: MSDE, Review of Output Outcome Monitoring Framework (OOMF), 2024-25
The e-Shram Portal (launched in 2021) is India’s ﬁrst na?onal database of unorganized workers
including migrant workers, construc?on workers, gig and pla?orm workers, street vendors,
domes?c workers, and others. Each registered worker is issued a Universal Account Number
(UAN) linked to Aadhaar, which enables portability of iden?ty and access across schemes.
Through this single UAN, workers can be connected to mul?ple services shown in the (Figure
38) such as skill development (Skill India), appren?ceship opportuni?es (NAPS), scheme
discovery (myScheme), pension beneﬁts (PM-SYM), and employment services (Na?onal
Career Service). The portal helps the Government deliver social security and welfare beneﬁts
more eﬃciently to this large and diverse workforce.
Figure 38: Services through eShram

Government ini?a?ves are designed to ﬁll gaps in the private eco system and to promote
inclusive growth. Private sector has to eventually play the primary role in genera?ng medium
and high-end skills needed by industry, and in connec?ng demand and supply of skills. Several
private employment portals such as QuikrJobs, Naukri.com, Foundit (formerly Monster India),
TeamLease, and LinkedIn provide large-scale digital marketplaces for job search and
123

recruitment. These pla?orms use technology for real-?me job matching, resume building, and
candidate tracking.
QuikrJobs (2008) is an online pla?orm for blue-collar, grey-collar, and entry-level jobs. It
connects employers and job seekers through lis?ngs for roles such as delivery staﬀ, data entry
operators, BPO/tele-callers, and technicians.
Naukri.com (1997) is online recruitment pla?orm operated by Info Edge that provides hiring
related services to corporates/recruiters, placement agencies and to job seekers in India and
overseas. The pla?orm oﬀers services such as resume database access, job lis?ngs, and
response management tools.
Foundit (formerly Monster) (1999) is an online job pla?orm that connects job seekers with
openings across roles and sectors. It oﬀers personalized job recommenda?ons using user
proﬁles, skills, and preferences. The pla?orm also supports career needs such as skill
development, mentorship, and ﬂexible work op?ons, helping users ﬁnd relevant opportuni?es
more eﬃciently.
Addi?onally, pla?orms like WhatsApp Business are increasingly being used by recruiters for
outreach, interviews, and informa?on sharing.
TeamLease Services (est. 2002) act as a supply chain company oﬀering solu?ons to employers
for their hiring, produc?vity and scale challenges. It provides hiring, staﬃng, appren?ceships,
and skill-development services to employers across sectors. Supports the full employment
cycle through three areas (3Es): Employment (large-scale staﬃng and trainees), Employability
(skill and training programs), and Ease-of-Doing-Business (HR services).
The company also runs TeamLease Skills University (TLSU) in Gujarat, India’s ﬁrst voca?onal
university, and operates major appren?ceship programs such as NETAP (Na?onal
Employability through Appren?ceship Program) oﬀering on-the-job training to appren?ces.
LinkedIn (launched in 2003) is the world’s largest professional networking pla?orm, with more
than 1 billion members across 200+ countries. It helps users build professional proﬁles,
connect with employers, search for jobs, and access learning and career resources. LinkedIn
operates through recruitment solu?ons, adver?sing, and premium subscrip?ons. Since 2016,
it has been part of MicrosoL. The pla?orm is widely used for hiring, networking, and
showcasing skills and work experience.
Indeed (2004) global pla?orm for job search and hiring. It operates in more than 60 countries
and allows users to search jobs, post resumes, and research companies. Powered by AI and
its large hiring.
EkStep's BlueDots AI ini?a?ve (2025), creates open digital rails- simple, AI-powered signals
that light up "Blue Dots" on maps to make hyperlocal livelihoods, jobs, skilling opportuni?es,
services, and welfare digitally discoverable at district and village levels. It tackles the
124

"discovery crisis" where abundant local supply and demand remain "digitally dark," using
vernacular voice AI to connect youth with nearby opportuni?es within two bus stops.
Together, these public and private systems form a digital bridge between skilling and
employment, improving visibility, access, and eﬃciency in India’s labor market.
10.7 Skill Policy and Planning
The skilling system poses a much greater challenge than the educa?on system, because we
are not even at the stage of universal access to job skilling, whereas the educa?on system has
already achieved it or is or close to it. We need not however follow the historical educa?on
strategy, of ﬁrst providing access and then focusing on pedagogy and learning outcomes
(sec?on 5.4 & 6.3). Nor should we build a cylindrical skilling system and then try to convert it
into a pyramid, as happened with educa?on (Figure 1). We must aim for a good quality
pyramid, which links to the educa?on and jobs pyramid at every level (Figure 3). Earlier sub-
sec?ons have tried to provide a data-based understanding of the Skilling eco-system, which
iden?ﬁes the gaps in the system. This allowed us to sketch a picture of the enormous challenge
in many dimensions of the skilling system. In this sub-sec?on we turn to an understanding of
the loca?onal and public policy aspects of the challenge.
The conﬂuence of Govt, Industry, NGO and skilling ins?tu?ons which operate at the Na?onal
level, are diﬀerent from those which operate at the State level, and those which operate at
the local level (District, City, Industrial area). These diﬀerences mirror the diﬀerences in
informa?on and co-ordina?on problems which plague the system at each level, with respect
to the goal of op?mal skilling outcomes and employability. Broadly speaking, the diﬀerent
levels of skills -- basic, medium (semi-skilled) and High (skilled) -- have ul?mately to be
addressed at the Local, State and Na?onal level respec?vely.
Figure 39 a?empts to capture and encapsulate the complexity of the skilling system from the
perspec?ve of public policy and planning. The author’s tour of skilling centers of diﬀerent
levels and across many States, has shown how weak the base of the skilling pyramid is, despite
a decade or more of Central programs to strengthen it! This is partly due to the fact that State
level oﬃcial, in middle of pyramid (Figure 39), s?ll give more weight to tradi?onal social
welfare issues, than to job skilling for employment genera?on and real wage growth.
To some extent this is the historical legacy of a system which ?ll recently, did not appreciate
the importance of learning outcomes at the bo?om of the pyramid (primary & lower
secondary school) for inclusion.
70
The fact that the governance system s?ll pays more
a?en?on to higher educa?on and futuris?c skills than to the semi-skilled/medium skills, adds
to the concern. Another indicator of the governing elite’s bias is the fact that CSR funds going
to job skilling are a ?ny frac?on of the CST funds going into health.

70
Na?onal Educa?on Policy, NEP (2022) which does recognize is less than ﬁve years old.
125

Figure 39: Role of Govt(C&S), Private sector and NGOs, in an Inclusive Skilling System

At the local level, successful exporters and growing/expanding MSME ﬁrms, were found by
the author to be the most enthusias?c proponents of skilling. Quality output is the driver of
their success and their growth, as they need well trained semi-skilled workers (including
supervisors) to sustain the quality, as they expand. They are willing to help develop curricula
relevant to needs of their industry, donate old equipment & machinery and even depute
skilled workers to train the trainers, of the skilling ins?tu?on. The MSME producers and
exporters do not feel they have the power to inﬂuence the courses & pedagogy in the schools,
ITIs and Polytechnics around them. Local industry associa?ons have li?le interest in doing so,
if they are run by owners of large stagnant industries.
At one of the skilling centers visited by the author in a small town, 90% of the trainees were
women and half of them were planning to set up their own beauty parlors, mostly in
neighboring villages where they lived. The authors sugges?on of adding a small training
module on entrepreneurship at the end of the beau?cian training, which would familiarize
District Government officials
Coordination, Facilitation
Industrialists,Exporters
Demand:Skills
Supply:Trainers,M&E
CSRfunds
Schools
Collages
(+Retired)
Skilling Centers
ITIs
Polytechnics
(+Retired)
NGO
Self –
Employed
Placement
(Eg: SIDH, eSharm,
Naukri.com
TeamLease
Skilling
Collages,
Universities
State Governments
Demand & supply est. & Plan
Export, Industry & Services
Association and Companies
Polytechnics
Others*
Policy frame work &
incentives for skilling
National demand for skilled
& semiskilled workers
Skilling industry association/councils
Locational Pyramid of Skilling
Central Government
Skill Gaps:
International
Demand-Supply
Note: * Others = Paramedical, Teacher training, Nursing, PGDM
126

them with basic ideas of accoun?ng, ﬁnancial returns, raw material procurement and
marke?ng. Given that almost 58% of workers are self-employed, this module should be added
to skilling programs in personal care, Repair & maintenance of household durables and semi-
durables and repair & maintenance of housing/commercial real estate (electricians and
plumbers).
71

There are also NGOs opera?ng at the local level which understand the needs of the society
and work to close the gaps in learning and provide job skills. One NGO, which the author heard
oﬀ, is par?cularly striking and perhaps excep?onal. This NGO iden?ﬁed a need in the
community for semi-skilled, educa?on para-professionals, below the level of primary teacher,
and outside the formal system. It deﬁned 11 diﬀerent levels, for which it created courses
(pedagogy), standards and cer?ﬁca?on and trained & cer?ﬁed thousands of aspirants and
placed all of them in employment. This leads us to the aphorism: “Jobs and skills are two
sides of the same coin” Governments need to help and support NGOs which are adding job
skilling and job placement to their tradi?onal focus on basic educa?on for deprived groups.
District and local oﬃcials may have li?le knowledge/apprecia?on about the importance of
skilling for employment, wage growth and fast inclusive growth, nor are they trained on their
role in promo?ng it. There are no standard opera?ng procedures with respect to job skilling.
These oﬃcers however did try to implement Central Govt skilling programs as per the
instruc?ons coming down to them. Contact with the local ins?tu?ons and companies relevant
to job skilling and employment appeared to be minimal. The most basic job of any skill
development oﬃcer should be to start by carrying out an informal survey of the demand for
semi-skilled workers in the district/block and the supply of these skills by the local ITIs. The
second cri?cal role would be to facilitate informa?on ﬂows between local schools, colleges,
skilling centers, ITIs, Polytechnics, and industrial ﬁrms and educa?on-skilling-job oriented
NGOs, in the district. Only on the basis of these can they do the real job of coordina?ng the
supply and demand of semi-skilled manpower required in their area (pyramid base in ﬁg 35).
Any expenditures required for this purpose or for basic job skills, has to come from the State
budget. Another important role can be played at the local level is to convert Krishi Vikas
Kendra (KVKs) into skilling Centers for Viksit agriculture and Agro-processing. This has been
successfully done in the Satara district of Maharashtra.
The States therefore have a key role in State’s skilling policy, manpower planning and skill
budge?ng. Besides informa?on from the districts, they need to interact with State level
Universi?es, skilling colleges, Polytechnics, Industry associa?ons, placement companies and
large NGOs opera?ng in the State (middle panel, Figure 39). A State Manpower ins?tute could
also do regular survey of demand for skills from Agriculture, Industry and Services, and the
supply of skills by educa?on and skilling ins?tu?ons in the State. The manpower skill
requirements arising from capital expenditures and repair and maintenance of old equipment,

71
Chioda (2021) showed that a 3-week MBA program with a soL skill component of 75% and 25% hard skills component (in Uganda),
increased earnings by 30% over 3-5 years.
127

machinery and structures should also be input on the demand side. Tripar?te commi?ees of
senior State govt oﬃcers with industry and educa?on & Skilling ins?tu?ons need to be set up
by all States following the pa?ern of the most progressive States, and must meet every year
to review the previous year and plan for the next. A?empt must be made to include
representa?ves of expor?ng industries and those which are growing, and thus have the
greatest need for semi-skilled and skilled manpower. Private colleges and skilling ins?tutes
must be fully represented in the commi?ee.
The State govts need to focus on the gap in the availability of good quality job skilling trainers
in their State, by incen?vizing semi-skilled workers to become trainers and raising their status
through recogni?on and rewards. In my visits I discovered an expert trainer, close to
re?rement, who had made a?rac?ve, working models of complex machines and processes,
which are invaluable aid to teaching. A modern manufacturing company, had put up visual
representa?on of the process, at the worksta?on used to train the workers, assigned to that
job.
72
In some of the most advanced Govt Skill development ins?tutes visited, however, the
availability of training videos was s?ll very patchy. States need to set up a digital library of
videos which fascinate and inspire both trainers and the trained, and which can be used to
make learning more relatable and less abstract for the average youth. These must be
conveniently accessible from the districts. A fund could also be set up for purchasing tools,
equipment and machinery to replace old/outdated ones in the its ITIs and Polytechnics. A
forum should be set up where teachers from Govt and Private, ITIs and Polytechnics, can
interact with, and learn, from each other. Ini?ally an appropriate IAS oﬃcer could be seconded
to organize the forum.
The Central govt has ini?ated many programs for skilling, re-skilling and upskilling, set up key
pla?orms to link youth skilling and employment, and ini?a?ves for upgrading the quality of
infrastructure and training. Policy liberaliza?on related to internship and appren?ceship has
also helped in promo?ng “learning by doing,” and “on the job training.” Extension of
appren?ceship schemes to export oriented and dynamic MSMEs would also help. Large and
Medium companies are the best placed to help develop semi-skilled workers. They should be
allowed and encouraged, perhaps even mandated, to spend a larger propor?on of their CSR
funds on job skilling. The rules should be ﬂexible enough so that they can use these funds to
subsidize the training of poorer children in their own training ins?tutes. Companies should be
encouraged to use CSR funds for skilling of trainers from Industrial Training Ins?tutes (ITIs) and
Polytechnics, and for providing sabba?cal to Engineering Professors from local Colleges and
Universi?es.
Central government could set up a Na?onal Skill development Council, and encourage the
private ITIs, Polytechnics, skilling colleges and placement industry to form a Skill development
Associa?on (Figure 39). Either the Na?onal Skill development corpora?on or a manpower
ins?tute should be mandated to make periodic es?mates of the supply and demand for skills

72
This was the ﬁrst ?me they had done this to be able to train more workers, more quickly.
128

at every level (basic, semi-skilled and highly skilled). The results of the surveys/es?mates must
be available to the public to facilitate their educa?on, training and job planning. The same or
diﬀerent ins?tu?on could be charged to study global demographic trends and es?mate the
demand - supply gaps in semi-skilled labor, which can poten?ally be ﬁlled by Indian labor. gaps
that global demand. The govt should also consider switching from creden?al-based hiring to
skill-based hiring. Before this is done however, the set of skills required for each type of job,
would have to be deﬁned.
10.8 Summary and Conclusion
This sec?on summarizes the analyses of job skilling in India, based on the limited data
available at both interna?onal and na?onal levels. It covers interna?onal comparisons of
voca?onal and technical enrolment, post-secondary training, and short-cycle ter?ary
educa?on and the schemes related to them. The quality of skilling, measured by employability
is also shown. Skilling requires qualiﬁed trainers and quality infrastructure and some of the
gaps in these, including in States, are analyzed along with the schemes to improve these two
cri?cal aspects. It delves into the kind of trades for which training is being given. The adequacy
of training provided by ﬁrms to regular employees is also compared with other countries.
10.8.1 Summary
The share of secondary students enrolled in Voca?onal & Technical program was a low 1.8%
in 2017. This was 12.8% below the benchmark for India’s per capita GDP in 2017. Comparator
countries like Vietnam, Thailand, China and Mexico, had enrolment rates which were higher
than their benchmark for their level of Per capita GDP. We have to raise this share to 18.4% in
ﬁve years and to 25.7% in two decades, in line with the projected growth of PCGDP. Though
ITI seat availability averages 103 seats per lakh popula?on for the country as a whole, it varies
greatly across States, with only Himachal Pradesh, Karnataka, and Rajasthan having rela?vely
higher availability.
Implicitly recognizing this structural gap, the Na?onal Educa?on Policy (NEP 2020)
recommended integra?on of voca?onal educa?on into mainstream educa?on, in a phased
manner and the elimina?on of the “hard separa?on” between academic and voca?onal
streams. It recommends early and con?nuous voca?onal exposure from Grade 6 onwards,
including hands-on sampling of craLs, 10-day “bagless” internships with local ar?sans, and
industry exposure through Grades 6-12. NEP speciﬁed that focus areas for voca?onal
educa?on, be iden?ﬁed based on skills gap analysis and mapping of local economic
opportuni?es. It suggested, that “By 2025, at least 50% of learners through the school and
higher educa?on system shall have exposure to voca?onal educa?on, for which a clear ac?on
plan with targets and ?melines will be developed. Towards this end, secondary schools will
also collaborate with ITIs, polytechnics, local industry with and hub-and-spoke skill
laboratories”. To oversee this eﬀort, the NEP suggested the cons?tu?on of a Na?onal
Commi?ee for the Integra?on of Voca?onal Educa?on (NCIVE), consis?ng of experts in
129

voca?onal educa?on and representa?ves from across Ministries, in collabora?on with
industry.
Many schemes have been launched to address the daun?ng challenge of skilling the labor
force. These include the Skill India Mission and associated programs such as Pradhan Mantri
Kaushal Vikas Yojana, Jan Shikshan Sansthan, CraLsmen Training Scheme, PM Vishwakarma,
and SAMARTH. These aim to expand voca?onal par?cipa?on, strengthen training quality, and
improve the transi?on from skills to employment. NEP 2020 has recognized the importance
of VET, non-academic skills and holis?c educa?on. Raising the enrolment of secondary school
students in VET programs, however remains a massive challenge, requiring concerted eﬀort
by all States, in coopera?on with private industry, non-proﬁt organiza?ons and the Central
Government.
NEP 2020 promotes mul?disciplinary higher educa?on ins?tu?ons that oﬀer voca?onal
programs alongside academic degrees, including within four-year undergraduate programs. It
encourages short-term cer?ﬁcate courses, credit accumula?on and transfer, and alignment of
qualiﬁca?ons with the Na?onal Skills Qualiﬁca?ons Framework (NSQF), thereby addressing
the historical lack of ver?cal mobility for voca?onal learners.
Of India’s adults aged 25 & over, 14.4% had post-secondary training, which was 5.8% points
below benchmark for its per capita GDP in 2023. This gap was smaller than for secondary
school enrolment in VET programs. The post-secondary training gap was also smaller than for
comparator countries such as Indonesia, China, Mexico and Thailand. In India we need to
increase post-secondary training to 24% of adults, by the ?me we become an Upper-Middle
Income Country around 2030 and further to 35% by mid-century.
There was an increasing trend (2019-22) in shorter-dura?on courses, both 6 month (122%)
and one year (7.1) in Industrial Training Ins?tutes (ITIs). At the same ?me, enrolment in 2-year
courses declined. The la?er needs to be reviewed, and course content and pedagogy modiﬁed
accordingly.
ITI enrolment is classiﬁed into engineering and non-engineering trades. Engineering trades
are sub-categorized into industrial and grey/white-collar occupa?ons, while non-engineering
trades include service and grey/white-collar occupa?ons. Enrolment in engineering trades has
unfortunately declined by about 9.8%, largely in industrial work, though fortunately they s?ll
account for the majority of enrolments. Within engineering trade, enrolment in grey/white-
collar engineering trades expanded by 3.6% from 2019 to 2022. In contrast, enrollment in non-
engineering trades grew by 2.4%, led by a growth of 19.2% in enrollment in service work from
2019 to 2022. With the increased focus on manufacturing there is a need to strengthen
industrial-work technical/engineering skills, to facilitate employment in quality manufacturing
and manufactured exports.
To strengthen skilling infrastructure and training quality, several ini?a?ves such as SANKALP,
PM SETU and STRIVE, have been implemented by the Central Government, Some States have
also collaborated with private, industry related, NGOs such as the Honda India Founda?on.
130

They have introduced new, more relevant, skilling courses, upgrading infrastructure, and
training the trainers in modern, industrially demanded, skills. These eﬀorts are bearing fruit.
ITI capacity has increased from 18.2 lakh in 2014 to 23.2 lakh in 2022. There was a 21%
increase in student enrolment in Government ITIs, likely due to lower fees and improved
recogni?on of cer?ﬁcates. The employability of ITI graduates has improved from 31.3% in
2022 to 46% in 2026.
The eﬀec?veness of Industrial Training Ins?tutes depends on the availability and quality of
trainers, yet only 46.9% of sanc?oned trainer posi?ons are currently ﬁlled, with shor?alls
more acute in government ins?tu?ons. To strengthen trainer capacity, the Na?onal Skill
Training Ins?tute provides modular, industry-aligned programs aimed at upgrading
instruc?onal skills. However, only 18.4% of trainers hold CITS cer?ﬁca?on, and staﬃng
remains mixed, with 78.2% regular, 16.1% contractual, and 5.8% temporary trainers. To
address this, the CraL Instructor Training Scheme operates under the CraLsmen Training
Scheme to strengthen instructor capacity by training ITI trainers in techniques for eﬀec?vely
transferring hands-on voca?onal skills to trainees. NEP pointed to the need to train teachers
for voca?onal subjects, create standards, and set up a Na?onal Commi?ee for Integra?on of
Voca?onal Educa?on (NCIVE) to oversee implementa?on. Much more needs to be done to
increase the availability of teaching material and to improve the pedagogy for training the
trainers.
India’s short-cycle ter?ary gap was 3.2% points below the benchmark for its Per capita GDP,
The performance gap was, however, smaller than the corresponding gap of comparator
countries such as Vietnam, China, Indonesia and Thailand rela?ve to their benchmarks. The
enrolment in such courses will have to be raised by 7.4% points by 2030 and by 16.8% points
by 2050 to meet the benchmark for High income country. Polytechnics, have a key role in
improving technical educa?on at the ter?ary level. There are 3,850 Polytechnics ins?tutes
across 35 States and Union Territories, oﬀering around 586 diploma courses. Polytechnics are
concentrated in larger states such as Tamil Nadu, Andhra Pradesh, and Maharashtra. 63.5% of
polytechnics are privately managed as they are self-ﬁnancing, invest in modern laboratories
and smart classrooms aligned with contemporary standards. Several non-government
ini?a?ves including Nitya Founda?on, Jindal South-West, Satya Shak? Founda?on, Ambuja
Founda?on, Telangana Academy for Skill and Knowledge, and NIIT Founda?on, support
government eﬀorts through short-term, employment-oriented training and industry-linked
community programs. As result of all these public and private eﬀorts, employability of
Polytechnic graduates has improved from 18.1% in 2019 to 32.9% in 2026.
“Learning by doing” in industry remains an important skilling avenue, yet formal ﬁrm-level
training in India has declined, with only 7.7% of ﬁrms providing training in 2022. This is less
than in comparator countries like Vietnam, Indonesia, and Malaysia. Strengthening industry
par?cipa?on is therefore cri?cal to link training with jobs. Ini?a?ves such as the Na?onal
Appren?ceship Promo?on Scheme and the PM Internship Scheme aim to expand
131

appren?ceship and internship opportuni?es through ﬁrm-led training. Rules for treatment of
training expenses in income tax law, may need to be changed to facilitate training & re-training
od employees.
10.8.2 Conclusion
To strengthen India’s skilling ecosystem, the paper recommends a na?onal Job Skilling Policy.
This includes integra?ng employment and skill development ministries, launching an Annual
Skills and Employment Survey with a data bank and portal, projec?ng skill demand, and
crea?ng course material and teaching aids. Investments are required for training the trainers,
especially for low- and medium-skill occupa?ons and blue-collar work such as repair,
maintenance, and construc?on. Raising the social status of blue-collar skills through awards,
contests, role models, and community outreach is important. Training quality also depends on
the quality of tools and equipment; capital subsidies can help. Partnerships with Australia
(TAFE), Canada, and Germany can support master-trainer development. Recognizing skill
development as a service industry can expand access to bank credit and visas. A Skill
Development Industry Council and an AI-based expert system (E-Guru) can help connect skill
seekers, providers, and employers. The Government’s training portal, iGOT, can be used for
strengthening job skills of district skill development oﬃcers.
States have a key role in State’s skilling policy, manpower planning and skill budge?ng. Besides
collec?ng informa?on from the districts, they need to interact with State level Universi?es,
skilling colleges, Polytechnics, Industry associa?ons, placement companies and large NGOs
opera?ng in the State. A State Manpower ins?tute could also do regular survey of demand for
skills from Agriculture, Industry and Services, and the supply of skills by educa?on and skilling
ins?tu?ons in the State. The manpower skill requirements arising from capital expenditures
and repair and maintenance of old equipment, machinery and structures should also be input
on the demand side. Tripar?te commi?ees of senior State govt oﬃcers with industry and
educa?on & Skilling ins?tu?ons need to be set up by all States following the pa?ern of the
most progressive States, and must meet every year to review the previous year and plan for
the next. A?empt must be made to include representa?ves of expor?ng industries and those
which are growing, and thus have the greatest need for semi-skilled and skilled manpower.
There is a need to create libraries of audio-visual teaching materials, establishing tripar?te
commi?ees (Skill Department, Industry Body, Skill Development University), and se?ng up
ins?tutes for trainer training, curriculum development, and job pla?orms. States need
dedicated funds for tools and equipment and should establish skill development Centers in
hub schools. NITI Aayog may design a benchmark skilling system and pilot it in selected states.
Local commi?ees (SDO, ITI, industry) can iden?fy skills in demand and introduce new ITI
courses, trainers, and equipment with support from local industry and the state.
Training in skilling centers providing skilling’ in personal care, repair and maintenance of
household durables and semi durables, and repair & maintenance of housing/commercial real
estate (electricians & plumbers), should add a small training module on entrepreneurship as
132

58% of India’s workers are self-employed. This module would familiarize them with basic ideas
of accoun?ng, ﬁnancial returns, raw material procurement and marke?ng.
Government needs to help and support NGOs which are adding job skilling and placement to
their tradi?onal focus on basic educa?on. Govt should consider conver?ng Krishi Vikas Kendra
(KVKs) into skilling centers for Viksit agriculture and agro-processing.
Skill development oﬃcer should be trained in carrying out survey of demand for semi-skilled
workers in the district and block. They should also inves?gate the supply of these skills by it is
in the district/block. They should be armed with standard opera?ng procedures (SOPs), to
facilitate ﬂow of informa?on between local schools, colleges, skilling centers, ITIs,
Polytechnics, on one hand and industrial ﬁrms and educa?on-skilling-job oriented NGOs.
Global evidence from voca?onal educa?on and training (VET) shows its advantages. Only 3-
4% of youth received formal VET in 2004-05, yet VET graduates had lower unemployment
(11%) than general secondary graduates (14%). Workers with formal VET earned 13-16% more
than those with no or informal VET in 2018-19, and this wage advantage increases with age.
11. Ter?ary Educa?on
11.1 Introduc?on
According to AISHE (2021-22), India had 1,168 universi?es, of which 423 were State Public
Universi?es (37%), 53 were Central Universi?es and 153 were Ins?tutes of Na?onal
Importance. Private universi?es account for 34% of total, with 391 State Private Universi?es
and 81 Private Deemed Universi?es. 56% of universi?es specialized in general programs and
16.5% in technical programs. The remaining universi?es oﬀer medical (6.8%), agriculture
(4.6%), science (2.8%), law (2.3%), and management (2.0%) programs.
There are 47,845 colleges in India of which 21.5% are Government colleges, with 34.8% of
total enrolment. Private aided colleges account for 13.3% of colleges and 20.6% of enrolment,
while private unaided colleges are 65.3% of colleges and 44.6% of enrolment. Colleges are
also concentrated on general educa?on (59.5%), followed by teacher educa?on (8.7%),
engineering & technology (7.0%), nursing (4.3%), arts (2.7%).
Standalone ins?tu?ons are evenly distributed across polytechnics (32%), nursing ins?tu?ons
(30%), and teacher training ins?tutes (28%). Other ins?tutes include Paramedical (6%), PGDM
(3%), and ins?tutes under ministries (1%).
In this sec?on, we compare India’s ter?ary educa?on, such as enrollment, and share of adult
popula?on with bachelor’s and master’s degrees. To add to Human capital and factor
produc?vity, this educa?on must generate skills which are needed by industry and services.
So, we also look at the employability of the creden?aled. The sec?on concludes with a review
133

of Programs and Ini?a?ves aimed at improving access, quality, and employability outcomes in
higher educa?on.
11.2 Ter?ary School Enrollment
This sub-sec?on presents India’s ter?ary (college, university) gross enrolment ra?o (GER) in
an interna?onal compara?ve framework. India’s GER for ter?ary educa?on was 33.1% in 2023
(Figure 40) marginally below the expected/benchmark level at its per capita GDP of 2023. This
is a commendable achievement. Thailand (-4.4%) and Mexico (-4.7%) are below the expected
level for their respec?ve Per capita GDP levels. However, Vietnam (+1.4), Indonesia (+2.9)
performed somewhat be?er than their benchmark while China exceeded the level expected
at its per capita GDP by +23.3 per cent points.
Figure 40: Ter?ary Schooling: College enrollment, (% gross)

Source: World Development Indicators, December 2024
Though India is close to its current benchmark, ter?ary enrolment will have to be increased
by 7.5% points in next ﬁve years to meet the minimum UMIC benchmark of 40.7% (Figure 40).
To reach the HIC benchmark of 61.7% it will require an improvement of 28.6 percentage points
over twenty-ﬁve years. We cannot rest on our laurels, in our quest to make Bharat Viksit by
2047.
Table 71 presents India’s es?mated number of enrolled students at the ter?ary levels for 2023,
using popula?on projec?ons from both the Technical Group on Popula?on Projec?ons and
the UN Popula?on Division. Es?mated Student enrolled = Gross enrolment rate of cohort (%)
* Cohort (18-223) Popula?on (million). this yields an es?mate 50 million ter?ary students
using TGoPP, and 52 million, using UNpop.
India (33%)
VietNam
Indonesia
Thailand
Mexico
China
UMIC (41%)
HIC (62%)
R² = 0.5673
10
20
30
40
50
60
70
80
0100002000030000400005000060000
TER (% gross)
PcGDP , PPP (const 2021 int $)
Tertiary School Enrollment
134

Table 71: Ter?ary enrolment Gross rates & numbers

Sources: WDI/UNESCO (completion rates), UN population (2024 rev).
Technical Group on Pop projections (2020); na = not available
Although India’s ter?ary enrolment rate (33.1%) is below that of China and the MIC4
(Vietnam, Indonesia, Thailand, and Mexico), India’s absolute enrolment is s?ll 2.2 ?mes that
of the MIC4 combined (Table 72).
Table 72: Comparison of Youth enrolled in Ter?ary educa?on

Sources: WDI/UNESCO (completion rates), Population Division (2026), Vietnam & Mexico are 2022
11.2.1 Courses and Programs
The division of ter?ary educa?on between STEM courses and Humani?es and Arts is
emerging as a factor in determining produc?vity and growth. India has also introduced
distance learning, in addi?on to the conven?onal Physical facility-based learning in ter?ary
educa?on; these are classiﬁed as “distance mode” and “regular mode” respec?vely. The
distribu?on of ter?ary student across programs and mode, for the latest available year (2020-
21) is shown in Table 73. In regular mode, 24.1% of students were enrolled in B.A. programs,
12.6% in B.Sc. and 10.3% in B. Com. In 2020-21 (rows 1 to 3, Table 73). Total enrollment in
graduate courses con?nued to grow during the en?re period 2017-18 to 2021-22. During this
period, the fastest growing undergraduate programs were B. Pharma, B. Sc Nursing, BBA LLB,
and B.Ed. B.Sc. was the only program in which enrolment declined (-0.6).
EnrolmentPop agedEnrolledPop agedEnrolled
Gross (%)(Millions)(million)(Millions)(million)
Tertiary Enrolment Gross(%)
2023 33.1 1505015652
2030 (proj)40.7 1415815061
2050 (proj)61.7 nana13181
Technical group on Pop
UN Population
Pop aged 18-23yr
IndiaVietnamIndonesiaThailandChinaMexico
Tertiary Enrolment Gross(%)
Enrolled (million)523123726
Enrolment Rate(%)33.142.245.146.274.846.4
Pop aged 18-23(mil)15682669613
135

Table 73: Enrolment, Share & Growth in Regular & Distance Programs: 2021-22

Source: All India Survey on Higher Educa?on (AISHE), 2021-22
At the postgraduate level (regular), M.Sc. had a share of (2.2%) in 2021-22, but at 8.3% was
the fastest growing program (line 18, Table 73). The share of Masters of arts (MA) program in
post-graduate enrolment was however higher than the share of Bachelor of Arts (BA) in under-
graduate enrolment.
In the distance mode, enrolment was concentrated more heavily in B.A (42.6%) and M.A (18%)
and these were also the fastest growing programs between 2017-18 and 2021-22 (Table 73).
Enrolment in distance MBA and M. Com courses declined during this period. The demand for
distance educa?on seems to be related to the desire to get a (under-) graduate or post-
graduate degree to enhance creden?als, rather than to learn and train for speciﬁc jobs. This
in turn related to the prac?ce of creden?al-based hiring rather than job skill-based hiring for
basic and medium skill jobs.
11.2.2 Enrolment Rates across states
As State-wise comple?on data are not available for ter?ary educa?on, enrolment rates are
used to analyses inter-State performance. AISHE 2021-22 provides ter?ary enrolment data in
absolute numbers, which are converted into enrolment rates using the popula?on aged 18-
23 years for each State (Enrolment/Popula?on aged 18-23) *100. State-wise es?mates are
available by mode- Al, Regular, and Distance.
Figure 41 shows enrolment rates in All Type & Regular modes for 2021-22 across States and
Union Territories, arranged according to their per capita Net State Domes?c Product (PC
NSDP) for 2021-22. The regression trendline between All enrolment and per capita Net State
Raio:
Graduate programmesAbsShareCAGRAbsShareCAGRAbsShare CAGRDistance/
(Lakhs) (%)Fy22/Fy18(Lakhs) (%)Fy22/Fy18(Lakhs) (%)Fy22/Fy18Regular
12345678910
1Bachelor of Arts (BA)113.326.94.590.424.13.619.542.64.20.22
2B.Com43.410.32.038.710.32.24.39.4-1.9 0.11
3B.Pharma4.51.119.04.51.219.0 0.00
4Bachelor of Science (B.Sc) 49.711.80.847.212.60.72.24.8-0.2 0.05
5B.E/B.Tech 38.59.1 -0.638.510.2 -0.60.00
6BCA6.21.55.55.71.57.20.51.1-7.1 0.09
7B Ed15.23.69.914.94.010.00.30.75.60.02
8BBA7.21.710.96.61.811.80.51.10.70.07
9LLB4.61.17.64.61.27.70.00.0-4.2 0.00
10BSc Nursing 4.01.013.64.01.113.70.00.0-2.2 0.00
11M.B.B.S.3.20.87.33.20.97.3 0.00
12Post-Graduate programmes
13Master of Arts (MA) 20.94.97.211.43.06.18.218.05.00.72
14MBA6.51.52.85.31.45.71.22.6-6.4 0.23
15M.Com5.21.23.13.40.94.21.53.3-2.9 0.45
16M.Sc9.52.37.68.32.28.31.12.51.80.14
17Sub-total (Tertiary)331.778.73.7286.876.33.439.586.32.30.14
18Other Tertiary 89.821.35.289.123.76.66.313.79.60.07
19Total Tertiary 421.6100.04.0375.8100.04.145.7100.03.20.12
All programmes Regular Mode Distance mode
136

Domes?c Product (Pc NSDP) is shown in red in Figure 41, with an R² of 0.44. The R² for regular
mode is much less (0.29) sowing a weaker dependence on Per capita NSDP.
73

Figure 41: Ter?ary State wise Enrolment Rates at diﬀerent level, 2021-22

Source: Author's Calc based on 1. AISHE 2021-22, (Table- 6 (a) & (b), 2. Popula?on projec?on for India & States 2011-36
(Table 20) and for NE & UT- pop-proj. (Educa?on.Gov.in) 3. PCNSP: Na?onal Sta?s?cs Oﬃce (NSO), RBI (on Aug 29, 2025).
Using the regression equa?on from Figure 41, an expected or benchmark enrolment rate (all)
is calculated for each State based on its Pc NSDP, and the gap is deﬁned as the diﬀerence
between actual and expected enrolment. States are then grouped by over-performance
(posi?ve gap) and under-performance (nega?ve gap), with States having a gap of +/- 3.0%
[= (std.dev) 8.9%/3] points deﬁned as having li?le or no gap (Table 74). U?ar Pradesh (+4.5%),
Madhya Pradesh (+5.3%) and Manipur (+13.3%) were above the enrollment rate expected of
their PCNSD (among States with per capita NSDP less than ₹10,000). Among States with per
capita NSDP of more than Rs. 10,000/ Arunachal Pradesh (+1.9%), Delhi (+4.0), U?arakhand
(+5.2%), Himachal Pradesh (+5.7%), Tamil Nadu (+7.5%), Chandigarh (+20.4%) and Puducherry
(+24.3%) performed be?er than expected at their PCNSDP (Table 74, 1
st
3 columns).

73
The regression trendline between Distance Mode and PCNSDP has an R² of 0.31.
y = -6E-10x
2
+ 0.0003x + 7.513
R² = 0.4432
R² = 0.292
15
20
25
30
35
40
45
25000450006500085000105000125000145000165000185000205000
Enrolment Rates (%)
PcNSDP (at constant prices; ₹) 2021-22
Tertiary Enrolment Rates (%) 2021-22
All Type Regular
Trend: All Type Trend: Regular
137

Table 74: Performance of States in Ter?ary Enrolment: All Types (2021-22)

Source: Authors calc based on AISHE 2021-22.; Note: PC NSDP is for 2021-22. Gaps are in brackets.
Though major States above the na?onal average of 28.4% include Sikkim (38.8%), Goa
(35.7%), Karnataka (36.2%) and Mizoram (32.3%) their enrolment rate is below the rate
expected of States at their level of per capita NSDP’.
11.3 Ter?ary A?ainment: Adult (25+)
This sec?on compares the ter?ary educa?on levels of India’s adult popula?on, aged 25 and
over, with that of other countries. India’s share of adults (25+) with a Bachelor’s or equivalent
was 13.2% in 2023 (Figure 42), 0.2% points above the expected/benchmark level at its per
capita GDP. In contrast, Vietnam (-5%), Indonesia (-7%), Mexico (-2%) and Thailand (-2%) were
all below their expected/benchmark levels, with China showing the largest nega?ve gap
(-11%). India however, needs to increase the percent of students with bachelor’s or equivalent
by 2.4% points in 5 years to reach min UMIC benchmark, and by 10.7% points to reach min
HIC benchmark in 25 years.
India’s share of adults (25+) with a Master’s or equivalent was 3.5% in 2023 (Figure 42, 2
nd

panel) which is (-) 0.7% point below the expected/benchmark level at its per capita GDP.
India’s poten?al compe?tors had a larger performance gap on this indicator: Vietnam (-4.9%),
Indonesia (-5.0%), China (-5.9%), Mexico (-4.6%) and Thailand (-5.2%) are all below their
expected/benchmark levels. The share of adults with Masters or equivalent needs to be
increased by 1.9% points by 2030 (UMIC), and by 5.7% points by 2050 (HIC).
This comparison of higher educa?on suggests that historically, ter?ary educa?on was more
important for India’s service-oriented economy, than it was for the manufacturing focused
economies of East and South East Asia.
Negative GapLittle or No gapPositive Gap
Tripura (-8.0)J&K (+0.1)UP (+4.5)
ChhattisG (-7.8)WB (+0.7)MP (+5.3)
Assam (-7.3)Meghalaya (+0.8)Manipur (+13.3)
Odisha (-6.7)Rajasthan (+1.0)
Nagaland (-6.7)Bihar (+1.7)
Jharkhand (-4.1)
A&NIslands(-17.8)MH (-2.7)Delhi (+4.0)
Gujarat (-17.2)TelanG (+0.2)Arunachal (+1.9)
Goa (-8.1)Andhra (+1.9)UttaraK (+5.2)
Haryana (-7.5)Kerala (+2.7)HP (+5.7)
Punjab (-7.2) Tamil Nadu (+7.5)
Sikkim (-6.2) Chandigarh (+20.4)
Karnataka (-4.5) Puducherry (+24.3)
Mizoram (-3.2)
Performance of States in Tertiary Enrolment : All (2021-22)
PcNsdp <
₹10,000
PcNsdp >
₹10,000
138

Figure 42: Adults with Bachelor & Master Degree (% of popula?on 25+)

Source: World Development Indicators, December 2024
Na?onal Educa?on Policy 2020 (NEP 2020) noted that students comple?ng Grades 11-12 with
voca?onal subjects oLen lacked clear pathways to con?nue their chosen voca?ons in higher
educa?on. Admission criteria in general higher educa?on ins?tu?ons were not aligned to
accommodate voca?onal qualiﬁca?ons, resul?ng in limited ver?cal mobility for students from
the voca?onal stream. NEP 2020 proposed the integra?on of voca?onal educa?on within
Higher Educa?on Ins?tu?ons (HEIs) will oﬀer voca?onal educa?on either independently or in
partnership with industry and NGOs. The B.Voc. programs, voca?onal courses will be made
available within regular Bachelor’s degree programs, including four-year mul?disciplinary
degrees. HEIs may also oﬀer short-term cer?ﬁcate courses, integrate ‘Lok Vidya’ into
voca?onal curricula, and explore delivery through Open and Distance Learning (ODL) modes.
11.4 Employability of the Creden?aled
The issue of the employability of those who have obtained creden?als has come to the fore
in the past decade. Either the degree does not measure what they are supposed to have
learnt, or the course curriculum is so divorced from the skills required in jobs, that they are
not employable. Those who have graduated from professional courses (MCA, BE/BTech) are
more employable than those who have general educa?on degrees (BA, BCom, B Pharma, BSc).
MBAs have the highest employability of 72.3% in 2026, followed by BE/BTech at 70.2% and
MCA at 68.2%. Those with general educa?on have an employability of between 55.6% for BA
degree holders and 62.8% for BCom degree graduates, in 2026 (Table 75).
India (13%)
Vietnam
Indonesia
China
Mexico
Thailand
UMIC (16%)
HIC (24%)
R² = 0.5825
5
10
15
20
25
30
01000020000300004000050000600007000080000
% pop of 25 yrs+
PcGDP , PPP (const 2021 int $)
Adults with Bachelors
India (3.5%)
Vietnam
Indonesia
China
Mexico
Thailand
UMIC (5%)
HIC (9%)
R² = 0.4065
0
2
4
6
8
10
12
0100002000030000400005000060000
% pop of 25+ yrs
PcGDP , PPP (const 2021 int $)
Adult with Masters
139

Table 75: Employability of Graduates from University & Professional programs

Source: Global Employability Report 2025 & 2026; Global Employability Test (GET).
Employability among all post school degree holders (including from Polytechnics) has shown
gradual improvement between 2019 and 2026, rising from 47.3% to 56.4% in aggregate (Table
75). BCOMs have shown the fastest increase of 11.1% per annum (compound growth),
followed by MBAs with a compound annual increase of 10.4% over seven years (column CAGR,
Table 75). Average employability for general educa?on has increased at a faster rate (7.5% per
annum), than for professional degrees (6.4% per annum) suggests a steady recovery and
be?er alignment of higher educa?on with market needs.
The slow improvement in employability of BSc (3.7% per annum) and BE/BTech (3.0%) degree
holders, and STEM courses in general (4.9% per annum) suggest that there is great scope of
improving the quality of pedagogy, by replacing outdated instruments & lab equipment, and
introducing an appren?ceship/ internship program for Professors in industry. This issue may
also be connected to the diﬀerence in minimum learning outcomes in Arithme?c vs Reading
in school (sec?on 6) with the much lower MAP rela?ve to MRP.
Top performing States by Youth Employability, were U?er Pradesh (78.6%), Maharashtra
(75.4%), Karnataka (73.9%), Kerela (72.2%) and Delhi (71.2%) in 2026 (India Skills Report,
2026).
11.5 Ins?tu?onal Capacity
Table 76 presents the distribu?on of universi?es in 2021-22 and the change since 2017. State
Public Universi?es form the largest share (37%), with 423 ins?tu?ons, followed closely by
State Private Universi?es, which account for 35%. State Private Universi?es also show the
highest growth, adding 129 ins?tu?ons since 2017. Ins?tu?ons of Na?onal Importance make
up around 14% of the system and have increased by 52 during this period. In contrast, Deemed
Universi?es both private and government form a small share of the total and show marginal
change. Central Universi?es account for about 5% of all universi?es and have expanded
modestly. Overall, the growth in the university system is driven mainly by State Private
Universi?es and Ins?tu?ons of Na?onal Importance, while other categories have remained
Year
20192020202120222023202420252026CAGR
Bachelor of Arts (B.A.)
29.348.042.744.249.247.154.055.69.6
Bachelor of Commerce (B.Com)
30.147.040.342.660.648.155.062.811.1
Bachelor of Pharmacy (B.Pharma)
36.345.037.244.657.554.056.058.06.9
Bachelor of Science (B.Sc)
47.434.030.338.137.751.358.061.03.7
Master of Computer Applications (MCA)
43.125.022.429.330.664.671.068.36.8
Bachelor of Engineering/ Technology (B.E/B.Tech)
57.149.046.855.257.464.771.570.23.0
Master of Business Administration (MBA)
36.454.046.655.160.171.278.072.810.4
Average of
University Courses
35.843.537.742.451.350.155.859.37.5
Professional Programs
45.542.738.646.549.466.873.570.46.4
STEM
46.038.334.241.845.858.664.164.44.9
140

stable.
Table 76: Share of diﬀerent type Universi?es, & Change (2017-18 to 2021-22)

Source: All India Survey on Higher Educa?on (AISHE), 2021-22
Figure 43 shows the availability of universi?es and colleges per lakh popula?on (18-23 years)
across major states. The pa?ern is uneven, and in most states, college presence far exceeds
university presence. Large states such as Bihar, West Bengal, Jharkhand and Assam have very
low university density (0.2-0.7 per lakh youth), yet their college numbers are much higher
ranging from 7 to 15 per lakh youth. Even in states with rela?vely be?er access Delhi (1.3
universi?es; 8 colleges) and U?ar Pradesh (2.6 universi?es; 29 colleges) the gap remains.
Southern and western states such as Maharashtra, Karnataka, Andhra Pradesh, Tamil Nadu
and Gujarat show colleges outnumber universi?es by a large margin.
Figure 43: University & College Density (per lakhs): 2021-22

Source: All India Survey on Higher Educa?on (AISHE), 2021-22.
Overall, the na?onal picture suggests that India’s higher educa?on system has grown through
colleges far more than universi?es, leading to limited university-level capacity rela?ve to the
size of the youth popula?on.

No. of UnivShare(%)
2021-222021-22
NumbersShares (%)
State Public University 4233772-3.0
State Private University 391341294.4
Deemed University Private 817 1-2.0
Institue of National Importance15313521.9
Central University 535 8-0.5
Deemed University Government 333 0-0.9
Total11341002620.0
Number of Universities
Change: 2017-18 to 2021-22
0
5
10
15
20
25
30
35
40
0
10
20
30
40
50
60
70
80
College/Univ
University-College/LakhPop
Number of Institutions/Lakh (18-23 yrs) (2021-22)
Universities/lakhs pop Colleges/lakhs pop College/Univ
141

In contrast to major states, the North-Eastern states and some Union Territories (Figure 44)
show a completely diﬀerent pa?ern. The number of universi?es per lakh youth is extremely
high, possibly because the popula?on base is small, while the number of colleges remains
limited. As a result, these regions have more universi?es than colleges in per lakh terms.
States such as Arunachal Pradesh, Mizoram, Sikkim and Manipur report very high university
density but only a modest number of colleges. This creates very low college-to-university
ra?os (oLen below 1), the pa?ern is similar in small UTs like Puducherry and Goa. This
imbalance indicates that NE states and UTs require more colleges to create a balanced higher
educa?on structure. Unlike large states where colleges are far ahead of universi?es the
challenge here is to expand college capacity.
Figure 44: University & College Density (per lakhs): NE states & UT’s

Source: All India Survey on Higher Educa?on (AISHE), 2021-22.
11.6 Programs and Ini?a?ves
India’s ter?ary educa?on system is undergoing major reforms to expand access, improve
quality, and make learning more ﬂexible and employment-oriented. Recent policies and
ini?a?ves focus on increasing enrolment, strengthening universi?es and colleges, promo?ng
digital and mul?disciplinary learning, and improving governance and regula?on. Together,
these eﬀorts aim to build a higher educa?on ecosystem that supports both academic
inclusivity and the emerging needs of the labor market.
NEP 2020 proposed major reforms to expand and improve higher educa?on in India. The
policy aims to increase the Gross Enrolment Ra?o (GER) in higher educa?on from 26.3% in
2018 to 50% by 2035, including voca?onal and online learning. To give students more
ﬂexibility, NEP introduces Mul?ple Entry and Exit Op?ons, where an undergraduate program
allows a cer?ﬁcate aLer one year, a diploma aLer two years, and a three- or four-year
bachelor’s degree, with a research op?on in the fourth year. The policy also plans a new Higher
Educa?on Commission of India (HECI) as a single regulatory structure replacing bodies like the
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0
100
200
300
400
500
600
700
800
NagalandManipurGoaArunPMizoramPuducherrySikkim
College/Univ
University-College/LakhPop
Number of Institutions/Lakh (18-23 yrs) (2021-22)
Universities/lakhs pop
Colleges/lakhs pop
College/Univ
142

UGC and AICTE, with separate ver?cals for regula?on, accredita?on, funding, and academic
standards. NEP further promotes digital transforma?on through the Na?onal Educa?onal
Technology Forum (NETF) to support e-learning, AI-enabled tools, and blended teaching-
learning models. Overall, these reforms seek to improve access, ﬂexibility, quality, and
governance within the higher educa?on system.
As part of NEP 2020’s push for digital transforma?on in higher educa?on, SWAYAM serves as
na?onal Massive Open Online Course (MOOC) pla?orm, 5.15 crore enrolments, 16,530+
courses, and 388 universi?es permi?ng up to 40% credit transfer. In February 2024, the
SWAYAM Plus ini?a?ve was launched; it delivers industry-integrated courses in high-demand
areas such as AI, Data Analy?cs, and Healthcare, forming the base of the Digital University.
74

Academic Bank of Credits (ABC), pla?orm securely stores, accumulates, and transfers
academic credits earned by students across recognized educa?onal ins?tu?ons. It acts as a
digital “credit bank,” where each learner’s credits are recorded through their unique APAAR
ID, enabling ﬂexible learning and smooth credit mobility under the Mul?ple Entry-Exit system.
Hoque (2025), examined the integra?on of Academic Bank of Credits (ABC) with MOOCs
among Social Sciences 186 students of Bihar central university.
75
The study reports
improvement in research & cri?cal thinking, communica?on, and interdisciplinary
understanding and applica?on of theore?cal knowledge to real-world situa?ons. Students
also reported autonomy in designing learning paths and deciding what, when, and how to
learn. Students reported posi?ve outcomes in skill development, par?cularly data analysis and
prac?cal applica?ons. Author found that students “appreciate the ﬂexibility and personalized
learning opportuni?es, challenges such as managing workloads, perceived lack of depth in
some MOOCs and technical issues hinder the ABC-MOOC model’s eﬀec?veness. Concern
about limited interac?on, aﬀec?ng collabora?ve learning, also emerged.” Authors suggested
that improvements are needed in course quality, workload support, and infrastructure.
Rashtriya Uchchatar Shiksha Abhiyan (RUSA) 2013, designed to support State and State-aided
higher educa?on ins?tu?ons in improving access, equity, quality, and employability. The ﬁrst
phase of the scheme was launched in 2013 and the second phase was launched in 2018. Now,
in the light of the Na?onal Educa?on Policy, RUSA scheme has been launched as Pradhan
Mantri Uchchatar Shiksha Abhiyan (PM-USHA). Dobi and Tripathi (2020), found an
improvement in male & female GER in Govt Degree Colleges in Jammu & Kashmir
76
.
PM-USHA strengthens innova?ve pedagogy, learning, digital infrastructure, accredita?on, and
governance across colleges and universi?es. It promotes ﬂexible learning through Mul?ple
Entry-Exit, the Academic Bank of Credits, Choice Based Credit System (CBCS), and digital

74
Details discussed earlier in Sec?on 6.4.1.
75
By combining quan?ta?ve and qualita?ve data; sample comprised 186 students for a ques?onnaire survey and 40 students for semi-
structured interviews, speciﬁcally from Social Sciences programs such as Economics, Psychology, Sociology, Poli?cal Science, Hindi, History,
and English.
76
Based on 15 Government Degree Colleges from four districts of Jammu & Kashmir (Anantnag, Pulwama, Kulgam, Shopian), selected
through stra?ﬁed sampling by literacy levels; ins?tu?ons established before 2010-11 and oﬀering Arts, Science, and Commerce
undergraduate programs.
143

educa?on pla?orms such as SWAYAM. It also aims to enhance research, skilling, and industry
linkages so that higher educa?on becomes more mul?disciplinary, employment-oriented, and
aligned with NEP 2020 targets for quality and par?cipa?on. Shan? Priya & Vijayalakshmi
(2025), that ins?tu?ons par?cipa?ng in PM-USHA increased infrastructure investment by 30%
increase over ﬁve years. Funds have also been used for expansion of digital infrastructure such
as smart classrooms and e-libraries. 70% par?cipants reported improved pedagogical skills
aLer a?ending PM-USHA-sponsored training sessions
Vidya Lakshmi Portal (August 2015), is a single-window digital pla?orm that allows students
to apply for educa?on loans from mul?ple banks in one place. It connects students with banks
following Indian Banks’ Associa?on (IBA) guidelines, which require loan applica?ons to be
processed within 15-30 days. The portal simpliﬁes the loan process by providing a common
applica?on form, tracking facility, and access to government scholarship links, helping
students secure ?mely ﬁnancial support for higher educa?on. Gagan and Malagi (2025)
Survey data from 286 student respondents indicate, that (1) “Digital loan services are more
eﬀec?ve than tradi?onal methods”, (2) digital channels improve ?mely approval.
Na?onal Scholarship Portal (NSP) is a single-window digital pla?orm of the Government of
India that brings together scholarship schemes from the Centre, States, UGC, AICTE, and other
agencies. It allows students to register, apply, and track applica?ons online, while providing
scheme-wise eligibility and documenta?on details. By centralizing the en?re process, NSP
improves transparency, reduces duplicate entries, and ensures ?mely and eﬃcient
disbursement of scholarships to eligible students across the country. Mohapatra, impact
evalua?on report on Na?onal Scholarship Portal have contributed to improving access to
higher educa?on among disadvantaged groups.
Unnat Bharat Abhiyan (UBA) 2.0 is a ﬂagship rural development ini?a?ve of the Ministry of
educa?on, which connects higher educa?on ins?tu?ons (HEIs) with local communi?es. Under
UBA 2.0, 688 selected HEIs both technical and non-technical have adopted around 3,555
villages to support their development. Ins?tu?ons conduct surveys, iden?fy local challenges,
and work with Panchayats to prepare village development plans, which are ﬁnalized through
Gram Sabhas. IIT Delhi serves as the Na?onal Coordina?ng Ins?tute. The program focuses on
key areas such as sanita?on, water management, rural infrastructure, energy access,
educa?on, health, agriculture, skill development, and awareness about government schemes,
aiming to link academic exper?se with grassroots needs for sustainable rural development.
11.7 Summary & Conclusion
This sec?on summarizes the analyses of Ter?ary educa?on in India. It covers enrolment rate
at Interna?onal, na?onal and state-level, employability outcomes, ter?ary a?ainment, along
with the programs and schemes aimed at improving quality, pedagogy and alignment with
labor-market needs.
144

At the Inter-na?onal level, the Gross Enrolment Ra?o (GER) for ter?ary educa?on was 33.1%
in 2023, at around the benchmark expected at India’s per capita GDP. States which have
performed above the level expected of their Per capita NSDP (PCNSDP) include U?ar Pradesh,
Madhya Pradesh, Manipur, U?arakhand, Himachal Pradesh, and Tamil Nadu. The number of
students enrolled in India’s ter?ary educa?on is es?mated to be 50-52 million. This is less than
Chinas, but 2.2 ?mes that of Vietnam, Indonesia, Thailand and Mexico taken together, despite
their performance being be?er than ours for their level of PCGDP. The GER for ter?ary
educa?on must keep pace with the projected pace of per capita GDP growth. This requires an
increase in Ter?ary GER of 7.5% points, to 40.7 % by 2030, and an increase of 28.6% points to
61.7% by 2050.
Several programs and ini?a?ves aim to improve access, quality, and employability outcomes.
SWAYAM provides a MOOC (Massive Open Online Course) pla?orm for online learning. ABC
(Academic Bank of Credits) enables digital storage and transfer of academic credits. PM-USHA
supports infrastructure, governance, and quality improvements in higher educa?on.
Employability of graduates has improved from 47.3% to 56.4% on average, between 2019 to
2026. Professional degrees such as MBA (72.3%), BE/BTech (70.2%), and MCA (68.2%)
con?nue to show higher employability, while general degrees like BA (55.6%) and BCom
(62.8%) lag but are improving faster. The slow improvement in employability of BSc (3.7% per
annum) and BE/BTech (3.0%) degree holders, and STEM courses in general (4.9% per annum)
reﬂec?ng the need for be?er pedagogy, updated laboratories, and stronger industry linkages.
In terms of a?ainment, India’s share of adults (25+) with a Bachelor’s degree is 13.2%, slightly
above the level expected of its per capita GDP. This is be?er than the performance of Vietnam,
Indonesia, Mexico, and Thailand for their per capita GDP. The challenge will be to raise this to
15.6% (2030) and 23.9% by 2050, in line with the projected increase in PCGDP. The rela?ve
performance in Master’s degrees is less than for Bachelors, 3.5% fewer adults having an MA
than the PCGDP related benchmark. However, Vietnam, Indonesia, China, Thailand and
Mexico showing larger nega?ve gaps.
The Na?onal Educa?on Policy (NEP) 2020 outlines major reforms for ter?ary educa?on
expanding the Gross Enrollment Ra?o to 50% by 2035, integra?ng voca?onal educa?on within
ter?ary educa?on, and accelera?ng the adop?on of online and digital learning through
pla?orms such as SWAYAM. Together, these shiLs aim to improve access, quality, and
employment relevance across India’s higher educa?on system.
One of the recurring problems of Higher educa?on in India is its weak connect with the skills
required to become employable. The following steps can be considered to improve the link
between Higher educa?on, Job skills and formal employment.
1. Map every University to local industry and signiﬁcant regional economic actors.
2. Introduce Faculty Appren?ceships with industry, changes rules to make these eligible
for faculty Sabba?cal and approved refresher course work and/or retraining.
145

3. Make provision in all engineering departments for appointment of “Professor of
Prac?ce” from the set of re?red or serving engineers in Industry.
4. Liberalize the rules for se?ng up and running specialized colleges, such as for risk
analysis and risk management. Regulate Private colleges on the basis of learning
outcomes and employability (GET test), then resident the number of resident
professors.
5. SoL skill: Ability to discuss and debate, search for ques?ons and answers, is best learnt
by doing. Universi?es should promote forums to facilitate such ac?vi?es.
6. Na?onal and State universi?es should have as one of their primary goals, searching for
cost-eﬀec?ve, sustainable solu?ons to Na?onal and State problems, respec?vely. The
learning occurs by the a?empt to solve problems, even if few solu?ons are found
7. Complete liberaliza?on of Govt rules for STEM professors, conferences foreign visits
and foreign visitors.
12. Compara?ve Scale of Human Capital
In an interna?onal context, the scale of India’s Human capital is as, if not more, important
than a?ainment rates, as this determines interna?onal compara?ve advantage, and evolving
(dynamic) compara?ve advantage. Table 77 presents two sets of es?mates of educa?onal
a?ainment of India’s adult popula?on aged 25 years & over. Number of adults with given level
of educa?on are based on two sets of popula?on es?mates; One, the Technical Group on
Popula?on Projec?ons, and two the UN Popula?on Division. Adults at speciﬁc educa?on (mi)
= Popula?on aged 25 & over (mi) * Per cent of total adults with the speciﬁc a?ainment (%).
Table 77: Number of Adults (25+) with diﬀerent levels of Educa?on

Source: Author calcula?on based on World Development Indicators, December 2024 & Census of 2011, Popula?on projec?on for India &
States 2011-36, Report of the Technical group on popula?on projec?on, July 2020 (MoHFW).
CompletionPop agedCompleters Pop agedCompleters
Rates (%)(Millions)(million)(Millions)(million)
Primary Completed
2023 65.6795522820538
2030 (proj)78.6906712936735
2050 (proj)93.0nana11641083
Lower secondary completed
2023 52.2795415820428
2030 (proj)62.9906570936589
2050 (proj)81.4nana1164947
Upper secondary completed
2023 33.7795268820276
2030 (proj)46.8906424936438
2050 (proj)65.0nana1164757
Bachelors completed
2023 13.2795105820108
2030 (proj)15.6906142936146
2050 (proj)23.9nana1164278
Masters completed
2023 3.57952882029
2030 (proj)5.49064993651
2050 (proj)9.2nana1164108
Technical group on Pop
UN Population
146

In 2023, based on Technical Group es?mates, around 522 million adults had completed
primary educa?on, 415 million had completed lower secondary educa?on, and 268 million
had completed upper secondary educa?on (Table 77). Corresponding es?mates using UN
popula?on projec?ons are slightly higher, at about 538 million, 428 million, and 276 million,
respec?vely. Even at the ter?ary level, India had approximately 105-108 million adults with a
bachelor’s degree and nearly 28-29 million with a master’s degree (Table 77).
12.1 Interna?onal comparison of Educated Adults
Table 78 shows, that the number of adults with primary and lower secondary educa?on in
India is signiﬁcantly lower than in China (982 mil & 742 mil, respec?vely), but the gap is only
15% at the upper (overall) secondary school level. Number of Indian adults with primary,
lower secondary and upper secondary educa?on is 1.8 to 2 ?mes that of Vietnam, Indonesia,
Thailand, and Mexico combined (MIC4).
Table 78: Interna?onal comparison of adults with diﬀerent levels of educa?on

Sources: WDI/UNESCO (completion rates), Population Division (2026)
Note: China rates are 2020, Maters rates for Thailand 2022 & Mexico 2020
For bachelor’s-level educa?on, India’s adult a?ainment rate was higher than that of Vietnam,
Indonesia, and China, but lower than Thailand and Mexico. India has around 108 million adults
with a bachelor’s degree, which is 1.4 ?mes that of China (79 mil) and 2.4 ?mes that of the
MIC4. At the master’s level, India’s a?ainment rate (3.5%) is higher than that of Vietnam,
Indonesia, China, and Thailand, but lower than Mexico. In absolute numbers, India has
approximately 29 million adults which is 3.1 ?mes China’s and 6.3 ?mes the MIC4, which
together (China + MIC4) have only 13 million Master degree holders (Table 78).
IndiaVietnamIndonesiaThailandChinaMexico
Primary Education
Completers (million)538551393998264
Completion Rates(%)65.687.383.874.895.885.7
Pop aged (millions)8206316652102575
Lower secondary
Completers (million)42841942874252
Completion Rates(%)52.265.656.754.272.468.9
Pop aged (millions)8206316652102575
Upper secondary
Completers (million)27625652032631
Completion Rates(%)33.739.139.139.431.840.9
Pop aged (millions)8206316652102575
Bachelors education
Completers (million)10871597913
Completion Rates(%)13.211.29.317.47.717.1
Pop aged (millions)8206316652102575
Masters education
Completers (million)2901192
Completion Rates(%)3.50.70.72.00.92.5
Pop aged (millions)8206316652102575
147

13. Research & Development
Research and Development (R&D) forms the core of a knowledge-based economy and
depends on both the strength of the educa?on system and the availability of skilled human
capital. India’s R&D performance is shaped by the educa?on base, the rate at which students
complete doctoral programs, overall R&D expenditure, and the size and quality of its research
workforce, including researchers and technicians. Firm-level investment in R&D and the ability
to translate research into innova?on outcomes also play a crucial role. In recent years, the
Government has introduced ini?a?ves such as the Na?onal Research Founda?on (NRF) to
promote high-quality research and Startup India to encourage innova?on and
entrepreneurship. In this sec?on we compare India with some of its poten?al compe?tors,
using WDI and other interna?onal data.
13.1 Doctoral Comple?on
The paper found no rela?onship between the percent of adults in a country who had acquired
Doctorates, and the per capita GDP of the country. This sec?on, therefore compares India’s
performance with a selec?on of lower middle income and upper middle-income countries.
India’s doctoral comple?on rate was 0.08% in 2023, which is higher than Vietnam (0.07%) and
Indonesia (0.05%), but lower than that of Malaysia (0.10%), China (0.12%), Thailand (0.12%),
and Mexico (0.31%).
13.2 R&D Expenditure
India’s R&D expenditure was 0.6% of GDP in 2020 (Figure 45), 0.2 percentage points above
the expected level. Thailand (+0.2%) and China (+1.5%) were also above their benchmark
levels, while Vietnam (-0.3%), Indonesia (-0.4%), and Mexico (-0.7%) fall below their
benchmark levels (Figure 45).
Figure 45: Research and development expenditure (% of GDP)

Source: World Development Indicators, December 2024
India
(0.6%)
VietNam
Indonesia
Thailand
China
Mexico
UMIC (0.7%)
HIC (1.3%)
R² = 0.3378
0.0
0.5
1.0
1.5
2.0
2.5
01000020000300004000050000600007000080000
Expenditure (% of GDP)
PcGDP , PPP (const 2021 int $)
Expenditure on R&D
148

The base for India in Figure 45 is 2020, therefore we es?mate the value for 2023 (0.7%)
assuming that the gap remained the same (0.2% points). This gap is then added to the
expected expected/benchmark (0.5%) for India at its 2023 per capita GDP. Based on this we
es?mate the improvement needed to reach UMIC & HIC benchmarks. This shows that we are
likely already above the UMIC benchmark, and need an increase of 0.6% points to reach the
benchmark of 1.3% for HIC (2050).
13.3 Researchers & Technician in R&D
India had 0.3 researchers per billion people in 2020 (Figure 46), 0.4 percent point below the
expected level. Vietnam (-0.6/billion), Indonesia (-1.0/billion), and Mexico (-1.8/billion) fall
below their benchmarks, while Thailand and China had a smaller gap of -0.4/billion.
Figure 46: Researchers & Technicians in R&D (per billion people)

Source: World Development Indicators, December 2024
The base for India in Figure 46 is 2020, therefore we es?mate the value for 2023 (0.6%)
assuming that the gap remains the same (-0.4% points). This gap is then added to the expected
expected/benchmark (1.0%) for India at its 2023 per capita GDP. Based on this we es?mate
the improvement needed to reach UMIC & HIC benchmarks. This shows that we need to
improve researchers per billion people by about 0.9% points to reach the UMIC benchmark of
1.5% and by ~2.3% points to reach the HIC benchmark of 2.9% for HIC India has a much smaller
research workforce and must expand it signiﬁcantly to support future R&D and Innova?on.
India had 0.07 technicians in R&D per billion people in 2018 (Figure 46, 2nd panel), 0.13 below
the expected level. Vietnam (-0.24/billion), Indonesia (-0.33/billion), Thailand (-0.24/billion),
and Mexico (-0.43/billion) were also below their benchmark levels, with Indonesia and Mexico
showing the largest gap.
The base for India in Figure 46 (2
nd
panel) is 2018, therefore we es?mate the value for 2023
(0.2%) assuming that the gap remains the same (-0.1% points). This gap is then added to the
expected expected/benchmark (0.3%) for India at its 2023 per capita GDP. Based on this we
India (0.3%)
VietNam
Indonesia
Thailand
China
Mexico
UMIC (1.5%)
HIC (3%)
R² = 0.4975
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
01000020000300004000050000600007000080000
Researcher / billion people
PcGDP , PPP (const 2021 int $)
Researchers in R&D
India
(0.07%)
VietNam
Indonesia
Thailand
Mexico
UMIC (0.4%)
HIC (0.8%)
R² = 0.3684
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0 100002000030000400005000060000
Techinicians / Billion people
PcGDP , PPP (const 2021 int $)
Technicians in R&D
149

es?mate the improvement needed to reach UMIC & HIC benchmarks. This shows that we need
to improve technicians per billion people by about 0.2% points to reach the UMIC benchmark
of 0.4% and by ~0.6% points to reach the HIC benchmark of 0.8% for HIC.
This again illustrates the big challenge that India faces in training semi-skilled workers for all
sectors of the economy, Agriculture, Industry and Services, modern and tradi?onal. This can
also be seen as an opportunity to raise the produc?vity and wages in every sector of the
economy, formal and informal, employed or self-employed.
13.4 Firms that do R&D and Innovate
We found no rela?onship between the percent of ﬁrms that spend on R&D, and the per capita
GDP of the country.
77
The share of ﬁrms which have developed products, improved processes
and do R&D in India is 1.4% (Table 79). This is higher than Vietnam (0.9%), Indonesia (0.9%),
Thailand (0.8%), but lower than Mexico (2.4%) and Malaysia (8.1%). China at 18.2% was way
ahead of the pack (Table 79). This data suggests that Mexico and Malaysia are likely to be our
peer compe?tors in high tech manufacturing.
78

Table 79: Firms that spend on R&D (% of ﬁrms)

Source: World Development Indicators, December 2025.
13.5 Innova?on
India ranks 38th out of 133 countries (Table 80) in the Global Innova?on Index (GII). Among
major compe?tors, India ranks 6 posi?ons above Viet Nam (44), 3 posi?ons above Thailand
(41), 17 ranks above Indonesia (54), and 20 ranks above Mexico (56). Malaysia is however
ranked 4 ranks above India. China is an outlies at rank 10, with 19
th
rank on inputs, but 5
th
oh
output. This may be connected to its highly successful policy of Reverse Engineering and
Developing (READ) technologies of foreign companies located in China, and companies,
startups and research ins?tu?ons possessing advanced technology, innova?ve solu?ons, in
High Income, technologically developed, countries (HITDCs)

77
Based on WDI, Dec 2024 data. The discussion is based on updated data from WDI, Dec 2025, which ?ghtens the deﬁni?on to ﬁrms which
develop products, process improvement and do R&D.
78
And also of UK, Japan and USA.
Country YearPcGDP%firms
India 2022$8,5941.4
Viet Nam2023$13,5460.9
Indonesia2023$13,8900.9
Thailand2016$19,3710.8
Mexico2023$21,9172.4
China 2024$23,84618.2
Malaysia2024$34,1168.1
150

Table 80: Global Innova?on Index Ranking, 2025

Source: GII Database, WIPO, 2025
13.6 Patent applica?ons & Hi-Tech Exports
We found no rela?onship between the Patent applica?ons, residents per million, and the
per capita GDP of the country. India reports 19 applica?ons per million (Table 81), higher than
Vietnam (11), Indonesia (5), Thailand (12), and Mexico (9), but lower than Malaysia (26). China
with 1,010 applica?ons per million, is far higher than all other selected countries, except Japan
(1,770). The United States (790), Germany (479), France (197), and the UK (173) all report
lower paten?ng levels than China.
Table 81: Patent applica?ons, Residents /Million popula?on

Source: World Development Indicators, December 2024.
India’s high-technology exports are 15% (Figure 47) of manufactured exports. The UMIC
benchmark for 2030 is 10%, and India is already above this level. The HIC benchmark for 2050
is 16%, which India is close to achieving. India’s High technology exports were higher than
Indonesia (9%) but lower than Mexico (18%), Thailand (26%) and China (27%). China and
Thailand are much higher, with a gap of about (+13%) compared to India.
Country PcGDP
Patent/Mi
India $8,05019
Viet Nam $12,04911
Indonesia$12,7575
Thailand $20,24512
China $20,4071010
Mexico $21,0329
Malaysia $29,82326
Japan $44,5491770
UK $52,872173
France $53,901197
Germany $62,950479
United States$71,318790
151

Figure 47: High-technology exports (% of manufactured exports)

Source: World Development Indicators, December 2024
13.7 Programs and Ini?a?ves
India’s research and innova?on ecosystem is supported by several na?onal programs
designed to strengthen scien?ﬁc talent, expand research capacity, and promote early-stage
innova?on. These ini?a?ves focus on nurturing young researchers, suppor?ng ins?tu?ons,
and encouraging industry academia collabora?on. Together, they aim to build a pipeline of
scien?ﬁc talent and enhance India’s overall R&D performance.
The Ins?tutes of Eminence (IoE) ini?a?ve is a Government of India program launched in 2017
to empower select universi?es and higher educa?on ins?tu?ons to become world-class
teaching and research hubs. It grants them greater autonomy, funding (for public ins?tu?ons),
and the ability to collaborate globally, with the goal of placing Indian universi?es among the
top-ranked ins?tu?ons worldwide. 10 Public and 10 private universi?es/ins?tutes have been
selected, with the public universi?es receiving a corpus of 1000 crore for the purpose.
As reported on the University of Hyderabad web site under the ?tle, “IoE – Ini?a?ves &
Impact, UoH Herald (2018)”, the IoE ini?a?ve has helped promote an “ecosystem of
innova?on, entrepreneurship and transla?onal research” through Start-ups and public-private
partnerships. Improvements are reﬂected in research output, with the average number of
publica?ons per Univ of Hyderabad faculty increasing from about 1-1.5 to nearly 3 annually,
including around 30% of publica?ons in Q1 journals.
Government of India (GOI) has also designated several universi?es and research ins?tu?ons
as Centers of Excellence (CoEs) in fron?er areas of science, technology, and innova?on. These
centers focus on specialized topics such as nanotechnology, renewable energy, advanced
manufacturing, AI, biotechnology, and energy storage. They are funded under schemes like
India (15%)
Indonesia
Thailand
China
Mexico
UMIC (10%)
HIC (16%)
R² = 0.1999
0
5
10
15
20
25
30
01000020000300004000050000600007000080000
% of manf. exports
PcGDP , PPP (const 2021 int $)
High-technology exports
Vietnam (42.7%)
152

the Training and Research in Fron?er Areas of Science and Technology (FAST) and the
Ins?tutes of Eminence (IoE) ini?a?ve.
GOI has also ini?ated Missions in fron?er areas like Quantum and AI. The Na?onal Quantum
Mission (NQM) was approved by the Union Cabinet on 19 April 2023, aims to build na?onal
capacity in quantum technologies. The Mission supports research and development in
quantum compu?ng, communica?on, sensing and metrology, and quantum materials and
devices. Implementa?on is through four thema?c hubs with support for startups and capacity
building. The Na?onal Mission for Ar?ﬁcial Intelligence (2024) to support the development
and deployment of AI applica?ons in priority sectors. The Mission focuses on building research
capacity, strengthening academia-industry collabora?on, and applying AI in areas such as
healthcare, educa?on, agriculture, and urban infrastructure.
INSPIRE-MANAK: Innova?on in Science Pursuit for Inspired Research-Million Minds
Augmen?ng Na?onal Aspira?on and Knowledge (2008), aims to iden?fy and encourage
scien?ﬁc talent among students aged 10-15 (Classes 6-10). The scheme invites around 10 lakh
idea submissions each year from over 5 lakh schools, from which 1 lakh ideas receive an
INSPIRE Award of ₹10,000. Schools can nominate up to ﬁve students annually through an
online system, and selec?ons are made based on the originality and quality of ideas. Selected
students par?cipate in district, state, and na?onal-level exhibi?ons, and par?cipants reaching
the Na?onal Level Exhibi?on & Compe??on receive mentorship at leading technical
ins?tu?ons. The top 60 projects are recognized as na?onal winners. The Na?onal Innova?on
Founda?on serves as the scheme’s resource partner.
Mir (2024) assessment of the INSPIRE-MANAK Scheme in Jammu and Kashmir, uses primary
data from district, state, and na?onal level compe??ons along with secondary sources
79
.
Findings show that about 1,000 student entries were submi?ed at the District Level Exhibi?on
and Project Compe??on (2021-2022), with 95 advancing to the State Level and 17 reaching
the Na?onal Level, indica?ng growing interest and awareness among students and teachers.
This aligns with the region’s broader performance over the past 14 years, where around
24,906 schools have par?cipated; despite having fewer schools than larger states
80
. Jammu
and Kashmir has secured a notable share of Na?onal Level Exhibi?on and Project Compe??on
awards, sugges?ng eﬀec?ve promo?on of scien?ﬁc temperament and crea?vity at the school
level.
NIDHI’s PRAYAS (2016) program supports young innovators in conver?ng their ideas into
proof-of-concept prototypes. It provides early-stage assistance that allows innovators to
experiment & create a workable product ready for incuba?on and commercializa?on. As a
pre-incuba?on ini?a?ve, PRAYAS is implemented through a Program Management Unit
(PMU), with selected incubators (31) designated as PRAYAS Centers (PCs). 401 innovators have

79
Oﬃcial documents, project documenta?on, and compe??on reports from SCERT and related nodal oﬃcers in Jammu and Kashmir are
used in the study. To obtain qualita?ve informa?on, a few interviews with par?cipants, instructors, and nodal oﬃcials were carried out.
80
Maharashtra (141,763) & Karnataka (185,211)
153

been supported, of which 306 companies were started in PCs, 214 prototypes were
developed, and 173 IPs were ﬁled. 135 innovators raised Rs 78 crore, and 202 innovators
generate 911 jobs.
Prime Minister’s Research Fellowship (PMRF, 2018) is a ﬂagship ini?a?ve designed to a?ract
India’s best young talent into doctoral program at IITs, IISc, IISERs, and top Central Universi?es
and NITs. It supports students pursuing cu?ng-edge research in priority areas of science and
technology. Fellows receive one of the highest doctoral s?pends in the country ₹70,000 to
₹80,000 per month along with a research grant of ₹2 lakh per year. Candidates are selected
through a rigorous process under discipline-wise nodal ins?tutes, with entry allowed through
direct admission or lateral entry from ongoing PhD program. PMRF aims to strengthen India’s
research ecosystem by enabling outstanding students to undertake high-quality, na?onally
relevant research within leading ins?tu?ons. Between 2018 and 2024, about 3,688 scholars
were admi?ed state-wide.
81

Bha?a (2025) analysis shows that PMRF Fellows have published 300+ papers in peer-reviewed
interna?onal journals and conferences (2022-2024). There were more than 45 patent
applica?ons, and several technology-oriented innova?ons, including three start-ups incubated
at leading ins?tu?ons (IIT Bombay, IIT Delhi and IISc). Fellows have collaborated with ISRO,
DRDO, and CSIR labs, to integrate academic with applied research. Remaining challenges
include, ins?tu?onal concentra?on (80% fellows 8 ins?tu?ons), Disciplinary gaps(public
health educa?on-job skilling, governance) and Post fellowship pathways. Awareness
campaigns, mentoring ini?a?ves in ?er 2, 3 ci?es and crea?on of a PMRF Alumni Council are
recommended by the Author.
PRISM, Promo?ng Innova?ons in Individuals, Start-ups and MSMEs (2014) supports
individuals and enterprises in transla?ng ideas into workable prototypes, processes or
technology solu?ons. It also supports to public-funded autonomous ins?tu?ons and
registered socie?es engaged in fostering innova?on, par?cularly for developing technologies
that beneﬁt MSME clusters. This is implemented through 12 TePP Outreach cum Cluster
Innova?on Centers (TOCICs)
82
. Priority areas include smart materials, waste-to-wealth
solu?ons, and water and sewage management. DSIR assessed the PRISM scheme through
survey & interviews of coordinators and innovators. The centers conducted 571
workshops/outreach ac?vi?es reaching 17,017 individuals, while 1,082 proposals were
received during 2015-2020. Most innovators developed prototypes in clean energy,
industrially u?lizable materials (40) and aﬀordable healthcare (37). With 83% of innovators
had completed the project, 42 patents generated and 38 startups noted. Majority of
innovators experienced technical challenges transforming prototypes into (fabricated)
products. Marke?ng challenges are even more daun?ng.

81
MoE, 2024-25
82
Technopreneur Promo?on Programme (TePP).
154

SPARC, Scheme for Promo?on of Academic and Research Collabora?on (2018) strengthen
India’s research ecosystem by enabling structured academic and research partnerships
between top Indian ins?tu?ons and globally ranked foreign universi?es. The scheme supports
joint research projects that include mobility of faculty and students, longer-dura?on visits by
interna?onal experts, and opportuni?es for Indian researchers to work in advanced
laboratories abroad. SPARC is designed to bring global exper?se to address na?onal
challenges, deepen bilateral research ?es, and enhance the interna?onal visibility and ranking
of Indian higher educa?on ins?tu?ons. SPARC approved 284 proposals involving about 100
top interna?onal universi?es, in phase I (2018-20).
According to Thapar (2025), 650 collabora?ve projects have been supported by SPARC,
resul?ng in almost 1700 research papers, around 80 patents and technology prototypes have
been developed and 900 workshops na?onwide. Some of ar?cles from SPARC were published
in high impact journals ACS Nano, Chemical Engineering journal and Environment
interna?onal. It has familiarized faculty and scholars in global research culture &
methodology. It has also led to an improvement of curriculum and crea?on of Indo-German
and Indo-Australian collabora?ve networks.
Anusandhan Na?onal Research Founda?on (ANRF), established under the ANRF Act 2023,
serves as the apex body for providing strategic direc?on to scien?ﬁc research in line with the
Na?onal Educa?on Policy. Its mandate is to seed, grow, and strengthen research and
innova?on across universi?es, colleges, research ins?tu?ons, and R&D laboratories in India.
With its crea?on, the Science and Engineering Research Board (SERB) has been subsumed into
ANRF. The Founda?on is designed to build strong linkages among industry, academia,
government departments, and research ins?tu?ons, and to create mechanisms for
par?cipa?on and support from industry and State governments to advance India’s research
and development ecosystem.
Atal Innova?on Mission (AIM), established by NITI Aayog in 2016, is the Government of India’s
ﬂagship ini?a?ve to promote a culture of innova?on and entrepreneurship through a holis?c
approach spanning schools, universi?es, research ins?tu?ons, industry, and the MSME sector.
AIM operates its program through real-?me MIS systems and third-party reviews to ensure
con?nuous improvement. At the school level, the Atal Tinkering Lab (ATL) program. focused
on STEM (Science, Technology, Engineering, and Mathema?cs), sets up dedicated innova?on
spaces for students from Grades 6 to 12, enabling hands-on learning with technologies such
as IoT, 3D prin?ng, robo?cs, electronics, and rapid prototyping; so far, 10,000 ATLs have been
established across 722 districts in India. At the higher-educa?on and enterprise level, AIM
supports Atal Incuba?on Centers (AICs) hosted in universi?es, ins?tu?ons, and corporates to
nurture early-stage startups by providing technical facili?es, mentorship, funding access,
partnerships, and co-working and lab spaces. AIM has opera?onalized 72 AICs, which
collec?vely support more than 3,500 startups, have generated over 32,000 jobs, and include
155

around 1,000 women-led enterprises across sectors such as health technology, ﬁntech, ed-
tech, space and drone technology, AR/VR, food processing, and tourism.
The Athena Infonomics (2023) assessment of Atal Tinkering Labs adopted a mixed-method
evalua?on, combining secondary analysis of 1,000 ATLs with primary stakeholder data to
review program implementa?on and outcomes. The ﬁndings (N=493), indicate posi?ve
learning impacts, 68.8% of schools reported improved academic performance, while 60%
observed enhancement in scien?ﬁc temper and 45.4% noted higher achievement in STEM
areas. Teaching prac?ces also shiLed, with about 71.8% schools repor?ng be?er student-
teacher engagement and 71.0% integra?ng technical basics into learning ac?vi?es. The impact
on student mindset was posi?ve towards science and technology in 74.4% schools, with more
students pursuing science for higher studies in 69.4%, be?er learnings among students in
59.6%, enhancement of 21st century skills in 58.2%, be?er ability to relate to school
curriculum in 42.2%, enhanced innova?on skills of students in 40.8%, and spirit of
entrepreneurship in students witnessed in 23.1%.”
Under Digital India Innova?on, MeitY key Digital India innova?on ini?a?ves include TIDE 2.0
Scheme, launched in 2019 to provide ﬁnancial and technical support (IoT, AI, Robo?cs, etc.)
through 51 incubators using emerging technologies and aimed at suppor?ng about 2,000 tech
startups; Domain-speciﬁc Centers of Excellence, with 26 CoEs established to drive capabili?es
in emerging technology areas; the SAMRIDH Scheme, launched in 2021 to strengthen
accelerators and scale soLware product startups, targe?ng 300 startups over three years; and
the Next Genera?on Incuba?on Scheme (NGIS) to handhold 300 soLware product startups
in Tier-2/3 ci?es through 12 loca?ons under NPSP 2019. Addi?onal ini?a?ves include the SIP-
EIT Scheme to reimburse interna?onal patent-ﬁling costs for startups and MSMEs, and Digital
India-GENESIS, a na?onal umbrella program to discover, support, and grow startups in smaller
ci?es through collabora?ve engagement with government and industry.
Startup India, launched in January 2016, aims to build a strong innova?on-driven ecosystem
by suppor?ng new enterprises through simpliﬁed compliance, a single-window portal, fast-
track patent processes, relaxed procurement norms, and easier exit.
13.8 Summary
This sec?on focused on the interna?onal compara?ve analyses of Indian R&D. It covered R&D
inputs, such as R&D expenditure, researchers & technicians in R&D, adult popula?on with
Doctoral degrees, and ﬁrms engaged in R&D. It also considered available measures of output
such as innova?on index, patents, high-technology exports. Programs and ini?a?ves aimed at
strengthening the research ecosystem are discussed.
India’s R&D expenditure was 0.6% of GDP in 2020, above the benchmark for its level of PCGDP.
It is es?mated at 0.7% in 2023. India’s R&D/expenditure ra?o is larger than those of Vietnam,
Indonesia, and Mexico. India’s R&D expenditure ra?o of 0.2% is above its benchmark, as is
Thailand’s (+0.2%, but China’s is above its PCGDP benchmark by 1.5%. India had 0.3
156

researchers per billion people in 2020 and is es?mated at 1.0 per billion in 2023. In terms of
the gap from its Per capita GDP benchmark(-0.4/bi) I was be?er than Vietnam (-0.6/bi),
Indonesia (-1.0/bi), and Mexico (-1.8bi), below that of Thailand (-0.4/bi and China(-0.4/bi) had
a smaller gap of -0.4/billion but remain above India. India had 0.07 per billion technicians in
R&D in 2018, (es?mated at 0.3 per billion in 2023), which was below its PCGDP benchmark by
(-)0.13/bi. All comparators were even more below the level expected for their PCGDP;
Vietnam (-0.24), Indonesia (-0.33), Thailand (-0.24), and Mexico -0.43), than India.
The propor?on of Indian adults (25+) who had completed a Doctorate was 0.08% (2023), a
rate higher than Vietnam (0.07%) and Indonesia (0.05), but lower than Malaysia (0.10%),
China (0.12%), Thailand (0.13%), and Mexico (0.31%). The share of ﬁrms engaged in R&D for
innova?on was 1.4% (2022), was more than Vietnam (0.9%), Indonesia (0.9%), Thailand (0.8%)
but lower than Mexico (2.4%) and Malaysia (8.1%). China, at 18.2%, was far ahead of all,
indica?ng a much stronger base of ﬁrms engaged in R&D. India ranked 38th in the Global
Innova?on Index for 2025. India ranked above Vietnam (44), Thailand (41), Indonesia (54), and
Mexico (56), but below Malaysia (34) and far behind China (5). India reports 19 Patent
applica?ons per million popula?on, higher than Vietnam, Indonesia, Thailand & Mexico, but
far below China (1,010). “High-technology” exports were 15% of manufactured exports,
higher than Indonesia (11%) but lower than Mexico (18%), Thailand (26%) and China (27%).
India needs to expand its research workforce to keep pace with the projected growth of per
capita GDP. This requires an increase of 0.9 researchers per billion to reach 1.5 per billion
researchers by 2030, and an increase of 2.3 researchers per billion to reach 2.9 per billion by
2050. The number of technicians needs to be doubled is 0.2 to 0.4 per billion popula?on in
ﬁve years and quadrupled to 0.8 per billion by 2050. High-technology exports are an area of
strength, already above UMIC benchmark and close to HIC benchmark of 16% for 2050.
The central Government has introduced a number of programs to a?ract scholars into
Doctoral programs, increase paten?ng ac?vity in Central Govt universi?es, and improve ﬁrm
par?cipa?on in university research. The PMRF (Prime Minister’s Research Fellowship) is
designed to support research scholars who join doctoral. Ini?a?ves like PRISM, NIDHI-PRAYAS,
INSPIRE-MANAK and Atal Innova?on Mission (AIM) help innovators and startups develop
prototypes and scale ideas. At the ins?tu?onal level, Ins?tutes of Eminence (IoE) empower
select universi?es to become world-class research hubs, ANRF (Anusandhan Na?onal
Research Founda?on) provides strategic direc?on to research, and SPARC promotes global
academic collabora?on. while fron?er programs such as the Na?onal Quantum Mission
(NQM) and Na?onal Mission on AI build advanced capaci?es. Digital India Innova?on
programs such as TIDE 2.0, SAMRIDH, and NGIS, along with Startup India, build the wider
startup ecosystem Together, these ini?a?ves aim to expand India’s research base, strengthen
innova?on, and align R&D with na?onal development goals. The tax subsystem needs to be
modiﬁed to increase incen?ves for deep tech R&D by corpora?ons.
157

14. Summary and Conclusion
14.1 Summary
The paper started with the analysis of the post-independence structure of educa?on during
the ﬁrst 50 years or so and the legacy it leL in the form of adult literacy and educa?on. Partly
as the result of the cylindrical structure of educa?on with a narrow base and over emphasis
on higher educa?on, adult literacy rate, and the share of adults with primary and secondary
educa?on were s?ll below the level appropriate for our per capita GDP. Ter?ary educa?on
was, in contrast at or above the level expected at our per capita GDP.
Posi?ve adjustment in the educa?on system in the last 15 years or so are reﬂected in Youth
literacy rates, which were 14% higher than the minimum expected at our Per capita GDP level.
This in turn reﬂects the good performance in Pre-primary and primary educa?on enrolment.
Primary enrolment and comple?on rates were signiﬁcantly above the benchmark rates for our
per capita GDP. Somewhat surprisingly the average quality of teaching and learning as
measured by minimum reading proﬁciency was also above the interna?onal compara?ve
benchmarks. This has been supported by ini?a?ves such as NIPUN Bharat, to ensure
founda?onal literacy and numeracy by Grade 3, pedagogical pla?orms like DIKSHA and PM
eVidya, which provide curriculum-aligned e-content, QR-coded textbooks, and mul?-channel
delivery through apps, TV, and radio.
There is however a challenge before us, to raise the minimum learning outcomes to the
benchmarks appropriate for the per capita GDP levels which we are aiming to reach by 2047.
States like Telangana, Karnataka, Tamil Nadu, Andhra Pradesh, Assam, Tripura and Arunachal,
where MRP is below the all-India average, need to pay much more a?en?on to this issue. This
challenge is much higher for raising Minimum arithme?c proﬁciency than for raising Minimum
reading proﬁciency.
The next stage in the educa?on pyramid is the reten?on rate/drop-out rate of students
between primary and secondary educa?on. In interna?onal compara?ve perspec?ve India’s
reten?on rates are above the benchmark rates for its per capita GDP. The enrolment rate in
secondary school is just above the benchmark rate. State enrolment rates are found to be the
only educa?on indicator corelated with PCNSDP, in both lower secondary and upper
secondary school. The States and UTs show a nega?ve gap between the secondary enrolment
expected at their Per capita NSDP, are Haryana, Karnataka, Sikkim, Arunachal Pradesh, West
Bengal and Nagaland. These States will have to step up their enrolment to help meet the
Na?onal enrolment rates required for achieving the Viksit Bharat. The comple?on rates in
both lower secondary school and upper secondary school are signiﬁcantly above the
benchmark for our per capita GDP. Pla?orms such as SWAYAM, oﬀering MOOCs across
disciplines, and teacher training programs like NISHTHA, have expanded opportuni?es for
both students and educators.
158

Only a modest eﬀort is needed to achieve the lower secondary comple?on rates expected for
a high-income country, the achievement of benchmark levels for upper secondary comple?on
rates will be more challenging. All India rates would have to rise from the current 63% to 79%
in 20 years.
The limited informa?on available on Minimum learning proﬁciency in Lower secondary school
(Std VI to VIII), however, introduces a note of cau?on. In rural areas, 42% of Std VI, 36% of Std
VII and 29% of Std VIII students cannot read a Std II level text. Though we don’t have direct
data for urban areas we es?mate that MRP may be 10%-12%. The States and with MRP below
the na?onal average (Andhra Pradesh, Telangana, Karnataka, Tamil Nadu, Assam and J&K) can
make a contribu?on by focusing on improving MRP. The situa?on is even more challenging
with respect to minimum arithme?c proﬁciency (MAP); 64% of rural students in Std VI, 59%
in Std VII and 54% in Std VIII cannot do division. Fourteen States have MAPs below the Na?onal
average (e.g. Tamil Nadu, Sikkim, Meghalaya, Kerala, Rajasthan, West Bengal) and are best
posi?oned to help raise MAP.
Paper also analyzed some ins?tu?onal issues like Parent teacher ra?os (PTR), Trained
Teachers (TT), Gender parity and Digital infrastructure. Available data on enrolment and
comple?on conﬁrms that gender parity has largely been a?ained in the Indian school system
There are s?ll some gaps in minimum learning outcomes between male and female students
at certain levels, par?cularly in minimum Arithme?c proﬁciency (MAP). The eﬀect of PTR and
trained teachers on Minimum learning outcomes (MRP and MAP) was not signiﬁcant. Digital
infrastructure is expanding in Indian schools, but the eﬀect of digital facili?es on MRP and
Map was insigniﬁcant. It is unclear whether, a) The appropriate audio-visual material has been
provided for eﬀec?ve teaching and learning, and (b) The informa?on/knowledge of teachers
in how to use the digital infrastructure eﬀec?vely has been suitable enhanced. The Teacher
App and AI-based tools such as Eduaide.AI, can support diﬀeren?ated instruc?on and
pedagogy. Tools such as ST Math and QANDA, can support early mathema?cs thinking, by
providing step-by-step problem solving and adap?ve feedback, which teachers can use to
guide conceptual discussions in classrooms.
Interna?onal comparisons of Indian skilling parameters were a lot less encouraging. Th share
of secondary students enrolled in Voca?onal and Technical training was 13% points below the
interna?onal benchmark for a country at India’s per capita GDP. Interna?onal compara?ve
data on the share of adults who have taken post-secondary training and share of adults who
have taken ter?ary, short cycle courses show a somewhat be?er picture. Compared to the
benchmarks for its Per capita GDP, India’s adult popula?on is only (-) 6% points below the
benchmark for post-secondary and (-)3% points below the bench mark for ter?ary courses. A
number of schemes and portals have been launched in the past ten years to remedy this
situa?on, but raising them to UMIC and HIC benchmarks in ﬁve and 25 years respec?vely, will
require a herculean eﬀort.
159

Firms providing formal training to their employees, is 7.7% in India, 8.7% in Vietnam and 8.4%
in Indonesia. This could be partly due to more compe??ve labor markets for corporate
employees. The incen?ves can be changed by allowing Companies and large ﬁrms to use a
much greater frac?on of their funds to train both employees and to upgrade ITIs and
Polytechnics. This will be cri?cal, if we want Indian private sector to compete with Thailand,
Malaysia and Mexico, with 18%, 24% and 38% of their ﬁrms have employee training programs.
India is compara?vely well placed interna?onal compara?ve frame, with respect to ter?ary
educa?on, but not so well in R&D. India’s ter?ary enrolment rate is almost at the exact
benchmark for its per capita income. Its Adult popula?on with Bachelor’s degree is slightly
above, and with Master’s degree a li?le below the benchmark for its per capita income. Given
its enormous popula?on this makes its total, ter?ary educated adult popula?on, the second
highest in the World aLer China.
The R&D picture is however, mixed. India’s total expenditure to GDP is above the benchmark
for its per capita GDP, the researchers per popula?on and technicians per popula?on are
signiﬁcantly below its interna?onal benchmark. Despite these handicaps, its High- tech
exports/total exports are close to the benchmark for a threshold high income country.
The share of Indian ﬁrms doing R&D for innova?on is 1.4%, compared to 0.8-0.9% for Vietnam,
Indonesia and Thailand, 2.4% for Mexico and 8.1% for Malaysia and 18.2% for China. Indian
patents per popula?on are higher than those of Thailand, Vietnam, Mexico and Indonesia, but
lower than those of Malaysia and far below those of China, which is just second to Japan. India
also ranks higher than most of these countries in the Global innova?on Index; India (38),
compared to Vietnam (44), Thailand (45), Indonesia (55) and Mexico (58). Malaysia (34) and
China (10) are ranked higher. A number policies and programs have been introduced to reduce
the big gap with China, but the tax incen?ves for private R&D could be improved further.
A cri?cal goal of the educa?on and skilling system is employment. The Global employability
Test (GET) is one way to measure its eﬀec?veness. Employability of those trained by this
system can be divided into three categories. Those gradua?ng from Voca?onal ins?tu?ons
(VET) like Industrial Training Ins?tutes (ITI) and Polytechnics have the lowest employability,
and those gradua?ng from professional ins?tu?ons such as Engineering (BE/BTech,
Computers (MCA) and Management (MBA), have the highest employability. Graduates of the
University-college system (BA, BCom, BSc, MA) fall in middle of the employability spectrum.
At the top of the system are the Professional Ins?tu?ons with employability in the 70-78%
range in 2025. Next come the Universi?es(-colleges), with employability in the 55-58% range.
Then come the ITIs at just over 40% and ﬁnally the Polytechnics at a li?le under 30%. The
good news is, that Employability has improved across the board during the past six years for
which data is available. The simple average increase of employability is 10% points for
Polytechnics, 20% points for universi?es, and 28% points for professional ins?tu?ons (2019-
2025).
160

The benchmark used in this paper are generally a func?on of per capita GDP. Absolute
numbers of educated and skilled popula?on, labor and work force is relevant when we
deﬁning compara?ve advantage rela?ve to both our comparator countries and vis-à-vis other
countries. Even with India’s school comple?on rates lower than higher income comparator
countries, the compara?ve numbers are higher than China’s and oLen exceed it. The number
of children comple?ng primary educa?on in India (137 mil) were 1.25 ?mes those in China
and more than 2.5 ?mes the combined total of Vietnam, Indonesia, Thailand and Mexico
(MIC4). The number of children comple?ng lower secondary school (65 mil) were 1.2 ?mes
those of China. The number of children comple?ng upper secondary school (98 mil) were also
1.2 ?mes those China and 3.7 ?mes those of the MIC4 (Vietnam, Indonesia, Thailand &
Mexico) combined. The number of students enrolled in ter?ary educa?on (52 mil) was more
than double those of the four comparators combined.
The above measures are more important for future compara?ve advantage, than the current
stock of adults with diﬀerent levels of educa?on. Broadly India has a larger number of adults
with Ter?ary educa?on (Bachelors & Masters) than China, while China has a higher number
of adults with primary & secondary educa?on; India’s adult popula?on with bachelor’s degree
was 1.4 ?mes China’s, while its adults with secondary educa?on were 0.85 China’s. India’s
stock of adults with primary, lower and upper secondary educa?on was 1.8 to 2 ?mes the
corresponding numbers for MIC4 (combined).
14.2 Conclusions
The paper’s analysis focused on factors that inﬂuence learning outcomes and the ins?tu?onal
condi?ons needed to translate educa?on into skills and employment. Primary educa?on
shows that while enrolment and comple?on are largely universal, a substan?al share of
children do not achieve minimum reading and arithme?c proﬁciency by the end of primary
school. Only about 46.3% of Grade 5 students meet minimum reading proﬁciency. Eﬀorts to
improve learning outcomes should focus on teaching prac?ces rather than further expansion
of inputs. ASER (2006-2024) shows that learning outcomes stagnated or declined despite
rising enrolment, teacher pay, and per capita GDP, indica?ng that input expansion alone has
not improved minimum learning outcomes.
Research evidence from many countries shows that structured pedagogy and Teaching at the
Right Level (TaRL) lead to gains in reading and mathema?cs by grouping children according to
their learning level and providing targeted instruc?on. Studies from India, Kenya, Ghana, and
Indonesia show that these approaches work when schools allocate speciﬁc ?me for
instruc?on, provide basic teacher support, and monitor implementa?on. The evidence
suggests that learning outcomes can improve through small, targeted changes in classroom
organiza?on and teaching methods, even without large increases in spending.
Learning outcomes also improve when government systems introduce performance-linked
incen?ves, regular monitoring, and contract-based arrangements for teachers. Randomized
161

studies show that teacher incen?ves raise student achievement more eﬀec?vely than
uncondi?onal spending on school inputs. Technology-based interven?ons also improve
learning only when they are curriculum-linked and supported by teachers or teaching
assistants; access to equipment alone does not lead to gains.
The Na?onal Educa?on Policy (NEP 2020), which places FLN, competency-based learning,
experien?al pedagogy, and voca?onal exposure within schooling at the core of educa?on
reform. Recent ini?a?ves such as NIPUN Bharat, DIKSHA, and PM eVidya reﬂect this shiL
toward structured pedagogy, teacher support, and system-level monitoring.
At the State level, there is no meaningful correla?on between per capita NSDP and learning
outcomes, with the correla?on between State PCNSDP and Grade 5 minimum reading
proﬁciency at 0.16, and with minimum arithme?c proﬁciency at -0.08, showing that higher
State income does not predict be?er learning performance.
States such as U?ar Pradesh (+11.7% points) and Odisha (+7.6% points) improvements in
reading outcomes from 2014 to 2024 despite lower income levels, while some higher-income
states show decline such as (Kerala (-0.8%), Haryana (-4.6%), Punjab (-5.5%), and Himachal
Pradesh (-8.4%)). Pratham developed, “Teaching at the Right Level (TaRL)” method, has been
used successfully in UP and Haryana. It begins instruc?on at the child’s actual learning level,
not grade. Children are grouped by ability, and teachers use simple, engaging ac?vi?es to build
founda?onal reading and arithme?c skills. Regular assessments track progress, and children
are regrouped as they advance. This method helps children catch up quickly and lays the
founda?on for future learning. Con?nuous monitoring and support ensure eﬀec?veness.
Survey comparisons across ASER, NAS, and FLS show that benchmarks for minimum reading
proﬁciency are broadly comparable, while benchmarks for minimum arithme?c proﬁciency
vary, especially at std III. As MAP is structurally weaker than MRP, Cohort analysis shows that
gaps between arithme?c and reading widen with grade progression, indica?ng weak
numeracy founda?ons. Mathema?cs learning therefore requires targeted and structured
support, par?cularly in the early grades. Evidence from the Mindspark program in Delhi shows
that technology-aided instruc?on, par?cularly in mathema?cs, improves student
performance across mul?ple math competencies and leads to larger learning gains in math
than in language. The gains are highest among weaker students when technology is combined
with small-group support from teaching assistants.
To reduce the gap between creden?als and learning, educa?on systems should strengthen
assessment integrity and reduce excessive reliance on exam-based creden?als. Evidence
shows that learning outcomes with limited comparability across surveys and over ?me,
weakens accountability and inter-state comparison. State-level rankings diﬀer substan?ally,
with cross-survey correla?ons for minimum reading proﬁciency ranging from 0.18 to 0.44,
reﬂec?ng diﬀerences in test design and proﬁciency thresholds. Decades old evidence from
Indian states showed that test scores may be inﬂated due to chea?ng and weak monitoring,
162

while reforms such as external grading, mul?ple test booklets, and technology-based
assessments reduce manipula?on and be?er reﬂect actual learning. At the same ?me,
governments can reduce incen?ves for creden?al inﬂa?on by placing less weight on formal
qualiﬁca?ons in public-sector recruitment and instead deﬁning job-speciﬁc skill requirements
and assessing candidates directly on these skills.
Weak founda?onal learning also aﬀects skill development. While voca?onal educa?on
improves employment outcomes in advanced economies, similar beneﬁts are limited in
developing countries due to weak linked training ins?tu?ons, skill mismatch and poor links
with industry. The framework in Sec?on 3 views educa?on, skills, and employment as linked
systems. Produc?ve employment requires alignment between educa?on levels, type of skills
acquired, and the structure of jobs available. In India, weak coordina?on across these systems
has created a disconnect between schooling, skills, and employability. Real wage growth
within each layer depends on on-the-job training and learning by doing, while movement
across layers requires upskilling and re-educa?on. The importance of skill development has
s?ll not been fully appreciated by most stakeholders in India. Formal recogni?on of skilling
and job placement as a cri?cal service, the development of skill development councils at the
na?onal and State level, and representa?on of exporters and rela?vely faster growing MSME
on local educa?on and skilling intui?ons and Government tripar?te bodies, can help change
percep?ons.
There is a need for greater use of audio-visual material from musical recording, through how
to videos (of instruments, equipment & machinery), to scien?ﬁc experiments for s?mula?ng
interest, thinking and innova?on. These are not just for students, but also teacher and
trainers, across the en?re spectrum from pre-primary, through primary, secondary school and
skilling from upper secondary to ter?ary levels. We also need more eﬀec?ve use of Expert
systems, the precursors to modern AI, to improve the quality of teaching, training, learning,
and to match the demand for skills by employers with the supply of skills acquired by
individuals. Expert systems and AI can also transform the output quality and incomes of the
self-employed, who cons?tute more than half the total Indian workforce (in agriculture,
industry and services).
15. Appendix: Legacy Issues in Literature
In the early 1950s India and China started with similar goals for educa?on, but China moved
faster by focusing on improving the quality of schooling early on. India focused more on
increasing access, which led to poor learning levels in schools. China started improving school
quality 20 years before India. In 2001, China changed its curriculum to reduce rote learning
and improve teaching. Teachers were trained with new methods and given incen?ves to work
in rural schools. Digital tools were used to improve access in remote areas. A clear na?onal
plan and regular inspec?ons helped keep track of progress. India can learn from the
experience of both market and non-market economies [Kumar and Varghese (2022)].
163

Kingdon and Muzammil (2003), found that the eﬀect of an?-chea?ng measures taken in UP
in 1992, was to reduce the pass rate in the high school exam from 57% in 1991 to 14.7% in
1992. It declined to 17% among regular candidates and 9% among Private candidates.
The PROBE report (1999), based on a 1996 survey. found widespread inac?vity or negligence.
In half the schools surveyed, no teaching was taking place. Teachers were oLen seen drinking
tea, reading comics, or simply doing nothing. This was not an excep?on but had become the
norm.
Kremer et al (2005) used na?onwide data collected in 2003, through three unannounced visits
to over 3,700 government primary schools across 20 Indian states, covering 98% of India’s
popula?on. They found that 25% of teachers were absent on any given day, and only about
half were teaching. There was wide state-level varia?on from 15% absence in Maharashtra to
42% in Jharkhand. The study found no link between higher salaries and lower absence; in fact,
older, be?er-paid, and more educated teachers were oLen more absent. In contrast, contract
teachers, despite earning less, had similar or slightly be?er a?endance. Daily incen?ves, such
as school inspec?ons, be?er infrastructure, and loca?on near roads, were more eﬀec?ve in
lowering absence. The existence of a Parent-Teacher Associa?on (PTA) or hiring local teachers
did not reduce absence. The authors highlight the need for reforms focused on monitoring,
accountability, and contract-based models, rather than simply increasing inputs like salary.
Kingdon (2007) found that, “the bar for passing is set very low, i.e., a student only needs on
average 33% marks in their various subjects in order to pass high school.” “Moreover, students
rely on ‘guess papers’ which are sold a few weeks before the exams, and that try to predict
exam ques?ons which are oLen remarkably close to the actual paper. There is frequent
leaking of papers in advance of examina?ons”.
Muralidharan and Sundararaman (2011) conducted a randomized controlled trial in rural
government schools in Andhra Pradesh from 2005 to 2007 to test whether teacher incen?ves
could improve learning outcomes. In incen?ve schools, teachers received performance pay in
the form of annual bonuses based on the average improvement in their students’ test scores
in math and language. In contrast, control schools received addi?onal spending on school
inputs like teaching materials, but no performance pay. The study found that students in
incen?ve schools performed signiﬁcantly be?er than those in control schools, and gains were
sustained in both incen?vized and non-incen?vized subjects. The interven?on was also shown
to be more cost-eﬀec?ve than simply providing more inputs. The ﬁndings suggest that well-
designed teacher incen?ves through performance pay can be a scalable way to raise test
scores and learning outcomes in the public educa?on system.
Singh (2015) compared learning outcomes in Ethiopia, India (Andhra Pradesh), Peru, and
Vietnam. He found that cross-country gaps in test scores appear by age 5 and widen
signiﬁcantly by age 8. The main reason is school-year produc?vity (deﬁned as the learning
gains per grade completed) is much higher in Vietnam compared to countries like Peru and
India.
164

Muralidharan et al. (2017) examined the issue of teacher absence in India using a na?onally
representa?ve panel dataset from 2003 and 2010, based on unannounced visits to schools.
They found teacher absence was 26.2% (2003) and 23.6% (2010). Thus, the problem remained
widespread despite improvements in input-based interven?ons like school infrastructure,
teacher qualiﬁca?ons, and student-teacher ra?os. Surprisingly, schools with lower student-
teacher ra?os had higher teacher absence, sugges?ng ineﬃciency in simply adding more
teachers. In contrast, schools that were visited more frequently by oﬃcials had signiﬁcantly
lower absence, highligh?ng the importance of school monitoring. The authors es?mated that
this weak governance leads to a ﬁscal loss of over $1.5 billion annually, and argued that
increasing administra?ve monitoring would be much more cost-eﬀec?ve than con?nuing to
expand inputs alone.
Banerjee et al. (2017) also compared Pratham’s experience in Bihar & U?arakhand (2008-10)
Haryana(2012-13), and U?ar Pradesh (2013-14). the team developed two versions that
worked even in regular government se?ngs. Their paper highlighted six main challenges when
scaling up: market eﬀects, spillovers, poli?cs, context diﬀerences, selec?on bias, and diﬃculty
in implementa?on. In TARL the biggest barrier was making the program ﬁt within the
government system. Success came only when schools changed how they grouped and taught
students and ensured ?me and staﬀ for the program. The authors suggest that scaling up
needs more than just good results in small pilots. It requires repeated tes?ng, strong local
partnerships, and understanding how the implemen?ng organiza?on works. In short, scaling
up a policy is not just about proof of scalability, but about making the system ready for change.
Muralidharan and Singh (2020) evaluated the large-scale MP Shaala Gunva?a program,
implemented across 1,774 schools in Madhya Pradesh (2014-16), to improve management
quality using global best prac?ces such as assessments, ra?ngs, and planning. Their RCT study
found that the “interven?on had no impact on either school func?oning or student
outcomes”. Authors concluded that results illustrate that “well-designed programs, that
appear eﬀec?ve based on administra?ve measures of compliance, may be ineﬀec?ve in
prac?ce.”
Singh (2020) inves?gated the reliability of administra?ve data in India's educa?on sector,
using data available from Madhya Pradesh and Andhra Pradesh. In 2016/17 MP govt had
introduced ‘mul?ple test’ booklets in grades 3,5 & 8 and asked for grade 8 answer scripts to
be sent to a diﬀerent school for grading, to counter test score manipula?on. Using direct audit
of this experiment by the MP govt in 2017, he concluded that results were “severely inﬂated
due to chea?ng, par?cularly severe for low-performing students”. He found that the
propor?on of correct responses to the same mul?ple-choice ques?ons was, on average, 38.9
% point (pp) higher in Math and 33.8 pp higher in Hindi in the oﬃcial test, from a base of
25.1% and 37.9% correct responses in the retest in the two subjects respec?vely. He also
found that “the diﬀerence between the propor?on of correct responses as reported in the
oﬃcial data and the retest, was present across the full distribu?on. The regression equa?on,
165

showed that, student-level average test scores, on the same mul?ple-choice ques?ons, are
higher by ~54% points in Math in Grade 7 on the oﬃcial assessment than the retest, but this
discrepancy is reduced by 50% in Grade 8. In Hindi, the mismatch is ~ 36% points in Grades 6
and 7, but is reduced by 27 percentage points in Grade 8. He concluded that “Mandated
mul?ple booklets and external grading sharply cut the discrepancy, indica?ng that chea?ng
(copying) by students and teacher assisted manipula?on of scores accounts for much of the
gap”.
In Andhra Pradesh, Singh (2020) “did not ﬁnd any such distor?ons in tablet-based tes?ng,
with only 2-5% students ﬂagged as chea?ng.” This study was an RCT of grade 4 students in
2400 AP schools, with 768 schools assigned to a benchmark oﬃcial grading, and 1694 schools
in which students took the same tests administered on tablets. Using the Angrist, Ba?s?n and
Vuri (2017) procedure, 38-43% of classrooms in the paper-based tes?ng arm are ﬂagged for
chea?ng. Students tested on paper, score higher, by ~28% points in mathema?cs, ~26% points
in English and 21% points in Telugu, than students who tested on tablets. In the retest audit,
students score 16-20 percentage points higher (on average) in the teacher-administered tests
than in the independently-proctored retest. The magnitude of this distor?on is comparable to
the analogous sample from M.P. in Grade 8.
Further, in government schools, students perform “18-24 percentage points worse on
tablets than on paper”, compared to 3-8 percentage points in private schools.” The diﬀerence
between tablet and paper assessments is lower with higher propor?on of girls in the
classroom: a classroom with only boys is predicted to have 32 percentage points lower scores
on tablets, but by 11-12 percentage points in a girls-only classroom in math and Telugu. The
coeﬃcient is small and sta?s?cally insigniﬁcant in English. “In 2016-17 and 2017-18, Singh
(2019) also conducted observa?ons of the test administra?on in 52 classrooms across 17
schools in four districts.” In these visits, “in 40% of classrooms, the teacher leL the classroom
for at least part of the test leaving students unsupervised, teachers commonly providing the
correct answers, & helped erase & correct them.” External monitors were rarely present
during assessment.”
16. References
Aayushi, and K. Mi?al. “Year-wise Analysis of School Responsiveness of School Teachers towards
NISHTHA 2.0 Online In-service Teacher Educa?on Program.” Asian Journal of Educa?on and Training,
vol. 11, no. 2, 2025, pp. 70–87. h?ps://doi.org/10.20448/edu.v11i2.6937.
Angrist, Noam, et al. “Improving Learning in Low- and Lower-Middle-Income Countries.” Journal of
Beneﬁt-Cost Analysis, 2023, h?ps://doi.org/10.1017/bca.2023.26.
Angrist, Noam, et al. Measuring Human Capital Using Global Learning Data. World Bank, 2021.
Arias, Omar, et al. The Skills Balancing Act in Sub-Saharan Africa: Inves?ng in Skills for Produc?vity,
Inclusivity, and Adaptability. 2019.
ASER Centre. Annual Status of Educa?on Report (ASER), h?ps://asercentre.org/.
Athena Infonomics. “Assessment of Atal Tinkering Labs (ATLs).”
h?ps://www.athenainfonomics.com/projects/assessment-of-atal-?nkering-labs-atls.
166

Banerjee, Abhijit V., et al. “Remedying Educa?on: Evidence from Two Randomized Experiments in
India.” The Quarterly Journal of Economics, vol. 122, no. 3, 2007, pp. 1235–1264.
Banerjee, Abhijit, et al. “Mainstreaming an Eﬀec?ve Interven?on: Evidence from Randomized
Evalua?ons of ‘Teaching at the Right Level’ in India.” 2016.
Banerjee, Abhijit, et al. “From Proof of Concept to Scalable Policies: Challenges and Solu?ons, with
an applica?on.” 2017.
Banerjee, Anindita, and Rakheebrita Biswas. “Empowering Educators, Bridging Divides: Evalua?ng
Diksha and Swayam for Digital Teacher Training and Economic Inclusion in India.” Interna?onal
Journal for Crea?ve Research and Thoughts, vol. 13, no. 7, 2025, p. 17.
Bansal, Tulika. “Analysing and Evalua?ng ePathshala Mobile Applica?on.” Online Interna?onal
Interdisciplinary Research Journal, vol. 12, no. 6, Nov.–Dec. 2022.
Bhar?, Kunwar Sanjay, and Shashi Prabha. “Assessing User Sa?sfac?on with Indian Digital Library
Pla?orms: A Case Study of NDLI.” IOSR Journal of Humani?es and Social Science, vol. 30, no. 5, 2025,
pp. 81–85. h?ps://doi.org/10.9790/0837-3005078185.
Bha?a, Diva. “Prime Minister’s Research Fellowship (PMRF).” IMPRI Impact and Policy Research
Ins?tute, 2025.
Birdsall, Nancy. “The World Bank’s East Asian Miracle: Too Much a Product of Its Time?” Journal of
Economic Perspec?ves, vol. 39, no. 4, 2025, pp. 127–148.
Chioda, Laura, et al. Making Entrepreneurs: Returns to Training Youth in Hard versus SoL Business
Skills. Working Paper no. 28845, Na?onal Bureau of Economic Research, 2021.
“Cost-eﬀec?ve Approaches to Improve Global Learning: What Does Recent Evidence Tell Us Are
‘Smart Buys’ for Improving Learning in Low- and Middle-Income Countries?” FCDO, World Bank,
UNICEF, and USAID.
Darji, Chirag, and Devangi Sondagar. “Eﬃcacy of Nishtha Programme on School Teachers: A Survey.”
AIJFR-Advanced Interna?onal Journal for Research, vol. 6, no. 5, 2025.
Das, Jishnu, and Tristan Zajonc. “India Shining and Bharat Drowning: Comparing Two Indian States to
the Worldwide Distribu?on in Mathema?cs Achievement.” Journal of Development Economics, vol.
92, no. 2, 2010, pp. 175–187.
De Ree, Joppe, et al. “Double for Nothing? Experimental Evidence on an Uncondi?onal Teacher
Salary Increase in Indonesia.” The Quarterly Journal of Economics, 2018.
Deissinger, Thomas. “The German Dual System as ‘Good Prac?ce’?” Journal of Voca?onal Educa?on
& Training, vol. 67, no. 4, 2015.
Duﬂo, Annie, and Jessica Kiessel. “Teacher Community Assistant Ini?a?ve (TCAI).” Policy Brief 4004,
Interna?onal Growth Centre, 2012.
Duﬂo, Esther, et al. “Peer Eﬀects, Teacher Incen?ves, and the Impact of Tracking: Evidence from a
Randomized Evalua?on in Kenya.” American Economic Review, 2011.
Eichhorst, Werner, et al. “A Road Map to Voca?onal Educa?on and Training Systems Around the
World.” ILR Review, vol. 68, no. 2, 2015.
Evans, David K., and Anna Popova. “What Really Works to Improve Learning in Developing
Countries?” 2016.
Glewwe, Paul, and Karthik Muralidharan. “Improving School Educa?on Outcomes in Developing
Countries: Evidence, Knowledge Gaps, and Policy Implica?ons.” 2016.
Gunjan. “U?liza?on of SWAYAM MOOCs by Teacher Educa?on Students: An Analysis of Awareness,
A?tudes, and Barriers.” Interna?onal Journal of Novel Research and Development, vol. 10, no. 5,
2025.
167

Hanushek, Eric A., et al. “General Educa?on, Voca?onal Educa?on, and Labor-Market Outcomes over
the Life-Cycle.” Journal of Human Resources, vol. 52, no. 1, 2017.
Interna?onal Labour Organiza?on. “What Is Skills Mismatch and Why Should We Care?” 2020,
h?ps://www.ilo.org/resource/ar?cle/what-skills-mismatch-and-why-should-we-care.
Johnson, Doug, and Emilio Parrado. “Assessing the Assessments: Taking Stock of Learning Outcomes
Data in India.” 2021.
Kar, Sudipto, et al. “The Landscape of Large-Scale Assessments in India.” ASER 2024, Annexure 9.
King, Kenneth, and Robert Palmer. Planning for Technical and Voca?onal Skills Development.
UNESCO-UNEVOC, 2010.
Kingdon, Geeta Gandhi. “The Quality and Eﬃciency of Private and Public Educa?on: A Case Study of
Urban India.” 1996.
Kingdon, Geeta Gandhi. “The Progress of School Educa?on in India.” 2007.
Kingdon, Geeta Gandhi. The Private Schooling Phenomenon in India: A Review. IZA, 2017.
Kingdon, Geeta Gandhi, and Rebecca French. “The Rela?ve Eﬀec?veness of Private and Government
Schools in Rural India.” 2010.
Kingdon, Geeta Gandhi, and Mohammed Muzammil. The Poli?cal Economy of Educa?on in India.
Oxford UP, 2003.
Kremer, Michael, et al. “Teacher Absence in India: A Snapshot.” Journal of the European Economic
Associa?on, vol. 3, 2005, pp. 658–667.
Kumar, Abhay, et al. “The Indian Educa?onal TV: Outreach, Percep?ons and Preferences.” Voices of
Teachers and Teacher Educators, vol. 13, no. 2, 2024, pp. 176–188.
Mbi?, Isaac, et al. “Inputs, Incen?ves, and Complementari?es in Educa?on: Experimental Evidence
from Tanzania.” The Quarterly Journal of Economics, 2019.
Mehra, Anurag, and Pramaths Kant. “A Compara?ve Evalua?on of MOOCs and Classroom Learning in
Engineering and Science in India.” 2024.
Ministry of Educa?on, Government of India. UDISE+. h?ps://udiseplus.gov.in.
Ministry of Educa?on. Na?onal Achievement Survey (NAS) 2017: Na?onal Report. 2018.
Ministry of Educa?on. Na?onal Educa?on Policy 2020. Government of India, 2020.
Ministry of Educa?on. Founda?onal Literacy and Numeracy Assessment: Na?onal Report. 2022.
Ministry of Educa?on. Na?onal Achievement Survey (NAS) 2021: Na?onal Report. 2022.
Mir, S. A. “Impact Assessment of the INSPIRE MANAK Scheme in Jammu and Kashmir.” Kashmir
Journal of Social Sciences, vol. 12, no. 1, 2024, pp. 135–141.
Muammar, S. R., and N. S. Ningsih. “Implemen?ng the Teaching at the Right Level (TaRL) Approach.”
2023.
Muralidharan, Karthik, and Abhijeet Singh. “Improving Public Sector Management at Scale?” 2020.
Muralidharan, Karthik, and Abhijeet Singh. “India’s New Na?onal Educa?on Policy.” 2021.
Muralidharan, Karthik, and Venkatesh Sundararaman. “Teacher Performance Pay.” Journal of Poli?cal
Economy, vol. 119, no. 1, 2011, pp. 39–77.
Muralidharan, Karthik, et al. “Disrup?ng Educa?on?” NBER, 2019.
Muralidharan, Karthik, et al. “The Fiscal Cost of Weak Governance.” Journal of Public Economics,
2017.
NCERT. PARAKH Rashtriya Sarvekshan (PRS) 2024: Na?onal Report. 2024.
Naveen Kumar, and Vinitha Varghese. “Elementary Educa?on in India versus China.” UNU-WIDER,
2022.
168

Newar, Purnima, and Rimakhi Borah. “Awareness on Na?onal Digital Library of India.” IJISM, vol. 20,
no. 4, 2022.
OECD. Learning for Jobs. 2021.
OECD. Survey of Adult Skills (PIAAC). 2023.
Pritche?, Lant. “Where Has All the Educa?on Gone?” World Bank Economic Review, vol. 15, no. 3,
2001, pp. 367–391.
PROBE Team. Public Report on Basic Educa?on in India. Oxford UP, 1999.
Raj, Vijaya, et al. “An Analysis of the Eﬀects of SWAYAM Learning in India.” Journal of Social Sciences
and Business, vol. 4, no. 1, 2025.
Roul, Rajkishore, and Ramakanta Mohalik. “Quality of E-Contents on DIKSHA Pla?orm.” 2025.
Sailer, Michael, et al. “Digital Learning in Schools.” 2021.
Sharma, Karishma, et al. “A Study of Samagra Shiksha Abhiyan.” IJRESS, vol. 8, no. 5, 2018, pp. 45–52.
Shrivastava, Pree?. “Study the Eﬃcacy of Nishtha Programme.” Interna?onal Journal of Novel
Research and Development, vol. 9, no. 6, 2024.
Singh, Abhijeet. “Private School Eﬀects in Urban and Rural India.” 2014.
Singh, Abhijeet. “Learning More with Every Year.” Journal of Development Economics, 2015.
Singh, Abhijeet. “Myths of Oﬃcial Measurement.” RISE Working Paper, 2020.
Singh, R., and R. S. Bhandari. “Students’ Awareness and Usage of SWAYAM–NPTEL.” Revista Review
Index Journal, vol. 5, no. 1, 2025.
Thapar, Madhur. “SPARC (2018): India’s Global Research Collabora?on.” IMPRI, 2025.
UNESCO Ins?tute for Sta?s?cs, et al. Se?ng Global Benchmarks for Grades 3 and 5. 2022.
University of Hyderabad. “Ins?tu?on of Eminence: Ini?a?ves and Impact.”
h?ps://herald.uohyd.ac.in.
Virmani, Arvind. “Lessons of Government Failure: Public Goods Provision and Quality of Public
Investment.” March 2006, Planning Commission ( h?ps://egrowfounda?on.org/research/lessons-of-
goverment-faliure-public-goods-provisions-and-quality-of-public-investment/ ).
Virmani, Arvind. India Vision, 2050. Policy Paper No 1/2021, EGROW Founda?on, May 2021 (
h?ps://egrowfounda?on.org/research/india-vision-2050/ ).
Virmani, Arvind. Viksit Bharat.: Unshackling Job Creators, empowering Growth Drivers, Working
Paper, NITI Aayog, July 2024 ( h?ps://www.ni?.gov.in/sites/default/ﬁles/2024-
07/WP_Viksit_Bharat_2024-July-19.pdf )
World Bank. World Development Report 2019. 2019.
World Bank. “Educa?on Factsheet.” 2024.
World Bank. World Development Indicators. 2025.

Other Publications by Dr. Arvind Virmani, Member, NITI Aayog.
Sl.
No.
Subject Month of
Publication
Link
1 Gondwana: Economic Integration of Indian
Ocean Region
Dec, 2025 https://niti.gov.in/sites/default/files/2026-
02/Gondwana-Economic-Integration-of-Indian-
Ocean-Region.pdf

2 Child Malnutrition & Mortality: Role of
Sanitation & Sewage Systems prepared
Nov, 2024 https://niti.gov.in/sites/default/files/2024-
12/Revised%20NITI%20Working%20Paper_1.pdf

3 Viksit Bharat: Unshackling job creators,
empowering growth drivers
July, 2024 https://niti.gov.in/sites/default/files/2024-
08/WP_Viksit_Bharat_2024-July-
31_chk%20with%20cover%20V2.pdf

4 Bharatiya Model of Inclusive Development May, 2023 https://niti.gov.in/sites/default/files/2023-
06/NITI_policy-paper_BMID_2023-May.pdf

5 NITI Working Paper on 'Bharatiya Model of
Inclusive Growth: Sabka Sath Sabka Vikas'
Dec, 2023 https://niti.gov.in/sites/default/files/2024-
02/NITI%20WORKING%20PAPER_Report_0.pdf

Education and Skilling for Employment From Credentials to Learning Outcomes