'ml~~~~~l
H. I N 1Z ~~~~~~~~~~~~-                    ,          7,:41 :-
tZ -I'"rgr                                     I. ;.  �.i'-94"          - r, i 
W~~~ZA                                                                           *3v- V <a @
>~~r     -'.;-                                '             ,t      '-'           '    .
~44  i,  .r               IEM -- 
66~ ~ ~~~~-
=  i ; x s s   s*, L n L\t-     r jji              W;t .r. ^~~'~  L  r : - *"'sS
1 , s _ ~~~~.  '"*'4ttw.A% .#  |   .t. ''-,JU 
tI -'4~~~~~~~~~r)
zt,~~~~~~~~~~~~  ,~~~~~~~~'q~  C==
Ln                                                                             luul~~~~~~~~'4~YA  f- 
"4 -'44.,,  A~  -.i-.rPQw
LLLJ~~~~~~~~~~~~Ca                                          "F 4






THE WORLD BANK OPERATIONS EVALUATION DEPARTMENET ODE
24589
INDIA: Education Sector Development
in the 1990s
A Country Assistance Evaluation
Helen Abadzi
Director-General, Operations Evaluation: Robert Picciotto
Director: Gregory Ingram
Manager: Ruben Lamdany
Task Manager: Gianni Zanini
2002
The author is the Senior Evaluation Officer working in the area of education in  The World B   Ic  �� :31
the OED Sector and Thematic Evaluation Group (OEDST) of the World Bank.    e  o    ar         71
This paper is available upon request from OED.                         Washington, D. C.



OPERATIONS EVALUATION DEPARTMENT
ENHANCING DEVELOPMENT EFFECTIVENESS THROUGH EXCELENCE
AND INDEPENDENCE IN EVALUATION
The Operations Evaluation Department (OED) is an independent unit within the World Bank: it reports
directly to the Bank's Board of Executive Directors. OED assesses what works, and what does not; how
a borrower plans to run and maintain a project; and the lasting contribution of the Bank to a country's
overall development. The goals of evaluation are to leam from experience, to provide an objective basis
for assessing the results of the Bank's work, and to provide accountability in the achievement of its
objectives. It also improves Bank work by identifying and disseminating the lessons learned from
experience and by framing recommendations drawn from evaluation indings.
OED Working Papers are an informal series to disseminate the findings of work in progress to
encourage the exchange of ideas about development effectiveness through evaluation.
The findings, interpretations, and conclusions expressed here are those of the author(s) and do
not necessarily reflect the views of the Board of Executive Directors of the World Bank or the
governments they represent.
The World Bank cannot guarantee the accuracy of the data included in this work. The
boundaries, colors, denominations, and other information shown on any map in this work
do not imply on the part of the World Bank any judgment of the legal status of any territory
or the endorsement or acceptance of such boundaries.
Contact:
Operations Evaluation Department
Partnerships & Knowledge Programs (OEDPK)
email: ecampbellpage@worldbank.org
email: eline@worldbank:org
Telephone: 202-458-4497
Facsimile: 202-522-3125
http:/www.worldbank.org/oed
2



Acronyms
DEA            Department of Economic Affairs
DIET           District Institutes of Education
DOE            Department of Education
DPEP           District Primary Education Project
EFA            Education For All
GDP            Gross Domestic Product
GER            Gross enrollment ratio
IDA            International Development Association
NCERT          National Council of Educational Research and Training
NGO            Non-governmental organization
NIEPA          National Institute of Education Planning and Administration
NPE            National Policy on Education
OED            Operations Evaluation Department
PROBE          Public Report on Basic Education
SIEMAT         State Institutes of Educational Management and Training
TVET           Technical-vocational education and training
UPBEP          Uttar Pradesh Basic Education Project
UPE            Universal Primary Education






Contents
Preface                       .............................................................i
Executive Summary ............................................................ iii
1.  Self-Reliance In Project Implementation ............................................................. 1
2.  Completed Projects: Technical-Vocational Education And Training ....................3
3.   Ongoing Projects - District-Based Primary Education ...........................................4
Significant Achievements .................................................................5
Some Concerns ..................................................................7
A. Local-Level Social Processes and Obstacles ................................. ...................9
B. Low Quality of Education .................................. ............................... I I
C. Uneven Quality of In-service Training, Modest Institutional Development ..... 12
D. Issues of Data Reliability, Analysis, and Objectivity ..................................... 13
E. Supervision Concerns .................................................................  15
4.  Evaluation Of IDA        Assistance ............................................................    16
Outcomes .................................................................                     16
Institutional Development Impact ..........................................                    17
Sustainability .......................................... 17
5.  Sectoral Issues - Directions For The Future ......................................                  18
Open Discussion of DPEP Successes and Obstacles ..........................................     18
FinancialAccountability and Local-Level Concerns in DPEP ......................................... 19
Needfor Integrated Sectoral Dialogue ....................................................      20
Better Use of International Experiences and Technical Expertise ................................... 21
6. Recommendations ...............................................                                      21



Box 1: Field Observations Obstacles To Student Performance Despite Best Intentions ................8
Table 1: OED Ratings of Technical Vocational Education Projects ..............................................4
Figures:
1. Progress in Literacy Rates for Selected Countries (1980-1986) .................................................1
2. Progress in the Gross Domestics Product Share Devoted to Education (1980/1986) .................2
3. Enrollments have increased ...............................................................6
4. Dropout declined under Uttar Pradesh Basic Education .............................................................6
Annexes
A. Background, Policy Framnework, And Bank Education Lending History In India ..........      ........ 23
B. Summary of Discussion at CAE workshop on March 29, 2000 ............................................... 31
C. OED Responses to Selected Government Comments .............................................................. 33



Preface
This paper is one of the background papers prepared as an input to the India
Country Assistance Evaluation (Task Manager: Mr. Gianni Zanini) by the Operations
Evaluation Department (OED) of the World Bank. Findings are based on a review of
project appraisal and completion reports, sector reports, and a number of other documents
produced by the Borrower, the Bank, OED, and various educational researchers. Bank
staff, government officials, and local professionals were interviewed at both headquarters
and in the field office. Their valuable assistance is gratefully acknowledged.
Earlier versions of the paper (prepared by Ms. Sukhdeep Brar) were supervised by
Aklilu Habte (consultant) and were reviewed by the Bank's India education sector team.
A revised paper was discussed at a workshop in New Delhi on March 29, 2000, chairnd
by Professor Denzil Saldanha, and with the participation of central and state government
officials, academics and members of policy research institutes, and other representatives
of civil society. Workshop participant views, field visit observations, interviews with
officials and donor staff, and two sets of comments from the Ministry of Human
Resource Development were taken into account in preparing this final version of the
paper.
The author is grateful for all comments received, which have been taken into
account in this revised version. However the views expressed in this paper remain
entirely those of the author. They do not necessarily reflect the views of OED or the
World Bank.



ii



Executive Summary
I .   World Bank lending in education has taken place through a unique working relationship
in line with India's principle of self-sufficiency and domestic development. Until the late 1980s,
the govemment of India strongly resisted extemal funding for education programs. Subsequently,
the goal of universal elementary education resulted in demand for additional resources, leading
the Department of Education (DOE) to review its policy on external funding in education. The
Bank's continued efforts towards a dialogue with DOE aimed at confidence building also
contributed to this change in policy. Since 1980, the Bank's investments in education in India
have grown from an almost negligible amount to $2 billion. The Bank has approved four
vocational and technical education and training (TVET) projects and six basic education projects.
Completed Projects - Technical and Vocational Education and Training
2.    In addition to civil works and equipment, the TVET projects focused on improving the
quality of instruction and on empowering institutions to carry out the tasks needed to train
students who would be valuable to industry (curricular revisions, updating teacher skills, self-
maintenance of equipment, improved linkages with industry, income generation, financial, and
academic autonomy). They also focused on increasing female participation in TVET.
3.     All four TVET projects are completed and audited. Although the Vocational Education
Project (Cr. 2008-IN) fell short of some of its targets, the Technician I and II projects achieved or
surpassed theirs. Altogether, about 1,100 institutions were constructed or received additions,
equipment, and fumiture. Overall, their capacity increased more than 50 percent, by roughly
100,000 student places, and expansion often exceeded targets. The Operations Evaluation
Department (OED) has rated project performance as satisfactory or highly satisfactory, though
substantial improvements are still needed in industry linkages, quality of trainers and academic
flexibility.
Ongoing Projects - Primary Education
4.     The primary education projects, which focus on local-level delivery of inputs through
innovative means, are still under implementation. Therefore, evaluative statements about them
are tentative. There is much state variance in achievement of various project outcomes, but joint
review mission reports and other materials show achievements in the following areas:
*      A very high level of govemment commitment at the central and state level. Donor staff
have been very much impressed by the eagerness of various officials to implement the
very large and complex district-based projects. Overall, about 50 million children have
been involved.
*      Large enrollment increases in the most deprived areas. For example, in the Uttar Pradesh
Basic Education Project (UPBEP), enrollments in the 1991-2000 period reportedly
increased by 67 percent at the primary level and by 74 percent at the lower secondary
(also called upper primary). The gross enrollment ratio increased from a baseline level of
66 percent to 107 percent (including overage children), although the target level was only
78 percent.



iv
*      Large increases of female enrollments. Girls' enrollment reportedly increased by 97
percent in the UPBEP and by 39 percent in DPEP; girls' attendance was 77-87 percent in
project areas. Dropout decreased.
*      Textbooks that are actually used in classes were provided on a large scale.
*      In-service training of about 100,000 teachers through block and cluster resource centers.
Instructional methodology was emphasized and child-centered methodologies were
disseminated to teachers and parents.
*      Large numbers of teacher appointments, particularly locally residing para-teachers.
*      Formation of over 171,000 village education committees, some of which are very active.
*      Construction of thousands of schools, toilets, and drinking water fountains through local
village authorities and through suitable local designs.
*      Establishment of block and cluster resource centers as well as district institutes of teacher
education.
5.     Areas of concern for primary education include:
*      Less than impressive learning outcomes. Baseline to mid-term comparisons showed
strongest improvement for grade I and minimal improvement for the higher primary
grades. Overall, achievement in the poorer districts remained far below expected
minimum standards.
*      Fewer teachers in areas where need is greatest; some teachers may be reassigned near
their homes, leaving the poorest children under-served.
*    ' Uneven quality of in-service training and modest development of training institutions.
*      School construction problems in poorer areas and in the earlier project stages.
*      Limited function of village education committee activities in the poorer areas.
6.    In addition, it has been difficult to verify program effects due to:
*      Uncertainty regarding the provision of inputs, partly because donors are usually unable to
visit project sites unannounced and observe operations as they happen under everyday
conditions;
*      Uncertainty regarding outputs, because donors have limited access to data and studies;
studies are executed by a few chosen institutions;
*      Some questions regarding the reliability and methodology of data collection and
reporting;
*      Agreements between government and donors that remain unwritten, omission of some
controversial issues in aide-memoires; and
*      Inadequate procedures for ensuring financial accountability.
7.     Nevertheless, the government has tackled, with great energy, extremely difficult and
complex problems. It would be impossible to improve quickly on all fronts, and some
imperfections are inevitable. OED rates the outcome of the educational program as satisfactory,
with high relevance and substantial efficacy and efficiency. Institutional development is rated as
modest and sustainability as likely.



v
8.     Specifically for primary education, OED suggests:
*      Independent verification of enrollments, project benefits, andflow offunds at the local
level. NGOs or independent academics might undertake these activities and report to the
government and the donors, thus giving the government the opportunity to mitigate
problems early. Donors should visit many school sites unannounced. Data sets from
independent studies (e.g. health surveys and PROBE) should be analyzed to compare
sampled DPEP and non-DPEP districts.
*      Increased availability of the studies to donors and qualified academics, improved quality
of reports and statistical analyses. More extensive data analyses at the local, district, and
state levels, particularly by independent academics.
*      Assessment of the local-level processes that greatly influence project success. Study of
the social or other types of obstacles which arise (e.g. difficulties of operationalizing
village education committees, keeping teachers in remote areas, local microcorruption
customs) in search of remedies.
*      Eliminating some fundamental limitations of lower-income classrooms (e.g. densely
packed large and noisy classes of children who receive limited teacher attention) through
after-hours remediation of the weakest students and specific training to help teachers
focus on low performers early on.
*      Restructuring the district and state institutes of education which thus far have not been
able to provide suitable service or locate suitable staff. Study world experience in
locating such capacity in various administrative entities, such as local colleges of
education.
9.     For the lending program in sectoral terms, OED recommends:
*      An integrated approach and study of the effects of one subsector on another,
*      Increased financing of higher education through user fees and alumni contributions;
*      Direct lending to states;
*      Better use of donors' technical expertise and world experience in various methodological
and administrative issues.






1.     Self-Reliance In Project Implementation
1.1    Atmnirbharta means self-reliance in Sanskrit. Ever since Mahatma Gandhi boycotted textile
imports from the UK and advocated the swadeshi (domestic) spinning and weaving of cloth from
raw materials, atmnirbharta has been a core value in India's financial and social policy. For many
years, imports were limited, foreign capital was discouraged, and local industry was subsidized to
build many of the products that other developing countries imported from the industrialized world.
The resulting achievements, such as automotive production and proficiency in software
development, have been an important source of national pride.
1.2    For many years, India tried to increase access to and quality of primary education in a
swadeshi manner, with its own funds and very little donor involvement. But poverty, population
size, and linguistic diversity made it difficult to make more serious inroads with just its own funding
and methodology. During the British Raj, schools had been managed locally, but the government
promoted centralization after independence. Compared to other countries in the region, which had
received consistent donor assistance, India performed about average in improving literacy and
primary education enrollment (Figure 1). Its financing of education also improved with the years, in
line with other countries (Figure 2). However, India consistently fell behind other Asian countries
(such as China) in education and health indicators, an issue pointed out during the 1990 Education
For All conference in Jom Tien. (Annex A details national trends in the education sector).
Figure 1. Progress in Literacy Rates for Selected Countries (1980 to 1986)
100
o560                        <
Thle left tlemXn is wnOIs htrcy,
D4                                            the 0fb  is X n0s.
B/ miw Fmc
20
IM Gxt9 es wee use Ior Sri La na. China, and ThaiwO
bol# IMNO fi b.ue were rie
0
0            20           40           60           10            100
I llev,.i 4B*1Q*JP



2
Figure 2. Progress in the Gross Domestic Product Share Devoted to Education (1980/1986)
6
* 
Thaaiand  aodd
9 4
Q
X                                 0 India
O                 Ba Pakian
O3                          Chn
*  ~~~~~~~OlSn L*nS
2
Burkin Fao
0
O          1         2          3         4          5         6
1980 percent GDP on education
1.3   From the 1 970s it was evident that to overcome India's serious social disparities, all
citizens would need access to a high-quality education that the central government could not
afford to finance or monitor on its own. Decentralization was needed to move authority, fiscal
responsibility, and implementation capacity to the local levels, an effort that India undertook
consistently since 1986 through the revised National Policy on Education (Annex A). Financing
education extemally was politically problematic. India had welcomed bilateral donor assistance in
the form of grants for education, but it wanted to borrow from the World Bank only for balance
of payments or hardware (such as the construction of agricultural universities in Bihar and Assam
in 1972). The government was concerned that the Bank might impose conditions that could be
seen as objectionable by some parties on such a sensitive national issue as primary education. On
the other hand, India could not on its own get the financial and implementation impetus needed to
educate its fast-increasing population. The Department of Economic Affairs agreed to borrow as
a way to increase absorption of IDA funds that were not being used elsewhere due to low
implementation capacity. The 1991 economic crisis subsequently created an urgent need for
increased resources. The story of IDA lending in education is a story of maintaining atmnirbharta
while building swadeshi implementation capacity.
1.4   After considerable thought and discussion, the governent asked the World Bank to
finance educational projects, under the condition that government staff take the lead in preparing
and implementing them (Annex A). From 1989 to 2000, the Bank financed four vocational-



3
technical education projects and seven primary education projects (Annex A Table 1). In both
subsectors, projects stressed decentralization and autonomy; states were asked to prepare plans
and submit them for apprbval. The states that were willing and able to prepare plans acceptable
to the government were included in the projects. OED Review Framework
1.5    The Operations Evaluation Department (OED) typically evaluates projects after their
completion. All five technical-vocational education projects under implementation have been
completed. OED audited them in October-November 1999, and has issued specific statements
about their outcomes.'
1.6   All primary education projects are still underway, and measures that will determine
outcomes are still being collected. Focusing on intermediate implementation variables, OED
reviewed files, studied supervision reports, and interviewed mid-level government staff about
views regarding relevance and implementation. Preliminary review findings were presented at a
workshop held in Delhi on March 29, 2000, and participants' views were elicited. Subsequently,
more interviews were held with officials and donor staff, and a field visit took place in the
Saharanpur district of Uttar Pradesh on April 3, 2000. This report presents a synthesis of data and
views received thus far. Since the primary-education projects are yet to close, OED's statements
about their outcomes and impact remain tentative.
2. Completed Projects: Technical-Vocational Education And
Training
2.1    The government's strategy was to bring about the much-desired industrialization of the
country, particularly in high-technology areas, in order to produce human and material resources
that would compete in the world market. The projects also focused on increasing female
participation in vocational-technical education. Four projects were implemented (Annex A) in
1989-99; they concentrated on the lower levels of the formnal technical-vocational education and
training (TVET) system, which train technicians and craftsmen to support the work of degree-
level engineers. They also focused on states and institutions that were most willing to make
much-needed changes. Over 10 years, the Bank committed $855 million to TVET. Though the
amount was minuscule compared to the needs of the country, the two parties developed a
productive working relationship, and the Bank established itself as a credible interlocutor in
education.
2.2    The projects were complex, with much construction of specialized buildings and large-
scale procurement of complex equipment. They tested the implementation capacity of the state
departments of technical education, most of which gradually rose to the challenge. Although the
Vocational Education Project (Cr. 2008-lN) fell short of some of its targets, the Technician I and
II projects achieved or surpassed theirs. Altogether, about 1,100 institutions were constructed or
received additions, equipment, and furniture. Overall, their capacity increased more than 50
1. Bank Report No. 20437, Performance Audit Report, India, Vocational Education Project, Technician Education I
Project, Technician Education II Project, Electronics Industry Development Project, April 27, 2000, Operations
Evaluation Department, World Bank



4
percent, by roughly 100,000 student places. Expansion often exceeded targets. For example, the
advanced vocational training scheme was to benefit 17,000 students but benefited 84,000.
Female participation exceeded targets and more than doubled, from about  1-15 percent to 30
percent. The one problem found consistently was that students often did not have specific
textbooks and studied from notes, while many of the books that had been developed were either
unsuitable or not available for students to take home and study.
2.3     OED rated the outcomes of the Technician Education I and II projects as highly
satisfactory when the projects were audited in October 1999 (Table 1). The Electronic Industry
Development Project (Ln. 3093-IN), which was appraised by the Bank's industry division,
created a best practice model of sustainability (the self-sustainability scheme) by gradually
withdrawing funds from participating polytechnics and engineering colleges while fostering
income generation and staff development. Overall, the experience in TVET underlined the
importance of communications among states to share experiences, raising expectations and
praising success, which must continue if the institutions are to become more closely linked with
industry, more autonomous, and more financially self-sufficient. The Bank's experience in India
shows that it is possible to implement successful TVET projects. Government commitment,
ownership, and attention to execution can make it possible to overcome the obstacles of these
very complex operations.
Table 1: OED Ratings of Technical-Vocational Education Projects
Cnteria                     Vocatfonal       Technician      Technician      Eectronics
Training        Education I      Education I1    Industry
Development
Outcome                     Satisfactory     Highly satisfactory  Highly satisfactory  Satisfactory
Sustainability              Likely           Likely          Likely           Likely
Institutional Development Impact  Modest     Substantial     Substantial      Substantial
Bank Performance            Satisfactory    Satisfactory     Satisfactory     Satisfactory
Borrower Performance        Satisfactory     Satisfactory    Satisfactory     Satisfactory
Note: The rating scale for institutional development as well as Bank and Borrower performance is "highly satisfactory",
"satisfactory", "unsatisfactory", "highly unsatisfactory". The outcome rating scale also includes "moderately satisfactory" and
wmoderately unsatisfactory" levels. Sustainability is rated as likely, uncertain, or unlikely.
3.      Ongoing Projects - District-Based Primary Education
3.1     These projects reflect a clear policy to move a significant share of educational
decisionmaking to the district and sub-district level and were expected to create the institutional
framework to carry out long-term, cost-effective human resource development. The general
objectives were to build institutional capacity; improve the quality and completion rates of the
primary education system; and expand access to primary education-especially for girls, scheduled
castes, and scheduled tribes. Starting in 1993 in Uttar Pradesh, the projects were implemented
simultaneously in several groups of states as well as in single states, such as Bihar. (Annex A has
detailed discussion of policies and project descriptions.). The lending program was preceded by
studies and baseline data collection. An elaborate monitoring and evaluation component is
producing many data. Donors have coordinated closely in financing and supervising the projects.
The govemment in collaboration with the donor community has carefully crafted a supervision



5
design that is elaborate, yet manageable, so much so that it has received a recognition award by
the Bank's Quality Assurance Group (Annex A). Bank and government collaboration have been
excellent, and relationships have been warm.
Significant Achievements
3.2    The government commitment and mobilization have been overwhelming, both at the state
and the central levels. Some very competent officials have been assigned to implement the
program, and there has been much attention to administrative issues at the higher levels, linkages
with states, and assured flow of donor and counterpart funds to the districts. Reports to the donors
are detailed and thoughtful and show a great deal of work. Government commitment and activity
far exceed the levels usually seen in other low-income countries and have very much impressed
Bank staff.2 As a result, the Uttar Pradesh Basic Education Project (UPBEP), the District Primary
Education Project (DPEP), and follow-on projects have received considerable worldwide
publicity. The Human Development Network has prepared a case study on this effort. 3
3.3    Social conditions vary considerably among states and districts therein, and there is much
performance variance among various states and projects. Joint review mission reports and other
materials for projects under implementation show progress in the following areas.
*   Reports on enrollment increases are impressive, particularly in the poorer areas. For example,
in the Uttar Pradesh Basic Education Project enrollments, over the period 1991-2000, show a
68 percent increase at the primary level (Figure 3) and 74 percent increase at the lower
secondary (also called upper primary). Contrary to the national trend (see Annex A, pam. 3) the
gross enrollment ratio (GER) in Uttar Pradesh project areas increased from a baseline level of
66 percent to 107 percent (including overage children), although the target level was only 78
percent. In DPEP, enrollment reportedly increased by 21 percent from 1996 to 2000. The GER
increased from 90 percent to 106 percent. By comparison, the GER in equivalent non-project
districts was 97 percent.4
2. An example is the bimonthly DPEP newsletter, which highlights champions, reports activities of the participating
states and the results of various studies.
3..Education Reform and Management Case Study, June 2000.
4. Making a difference: Focus on Girls Education in Uttar Pradesh, India. UP Education for All Project Board.
Presentation made at the World Bank Human Development Week, March 2000. It is unclear whether enrollment rates
are stable throughout the year or reflect only early counts.



6
Figure 3. Enrollments have increased
Uttar Pradesh Basic Education                   District Primary Education
.� 4            1              _ Girls,97.3%             5Gis,38.6%
E 2             J             - increase     -                            increase
2EF                    7   Boys, 51.5%__E___2___Boys, 10.3%
increase S.                                    crease
0 
91192        99/O0  Total increase,          99997         9100  Total increase,
67.5%                                         21.2%
Reports on girls' enrollment show substantial increases, particularly areas of traditionally
lowest enrollment. For example, enrollments reportedly increased by 97 percent in the UPBEP
and 39 percent in DPEP; overall, DPEP districts
are enrolling more girls than similar non-DPEP       Figure 4. Dropouts declined under
districts; girls' attendance was 77-87 percent in     Uttar Pradesh Basic Education
project areas. Enrollment of scheduled-caste and
tribal girls rose by 44 percent and 40 percent,
respectively. Girls' dropout in UPBEP was          60 1
reduced by 42 percent, and the gender gap was    ,40 - lGiris
reportedly eliminated in many districts (Figure  L20 2
4). Representation of women teachers in UPBEP       o -
districts rose from 18 percent to 28 percent.             1991         1998
Nevertheless in some areas as in Orissa, female
participation is not satisfactory.
*   Though there is much variance among grades and states, learning outcomes are reported as
improving. A midterm assessment study was conducted in 42 districts of seven states covering
50,000 students in 1,800 schools in 1994 and 1997. Grade I scores in language and
mathematics showed considerable improvement in three years, whereas grades 3 / 4 scores
showed less.5 Half the districts in 1997 scored at or above the 60 percent mastery criterion in
grade 1, but almost no district reached this criterion in grades 3 and 4 (partly perhaps as a result
of curricular changes). The gender gap in achievement is reportedly closing.
*   Instructional methodology has been emphasized, and there has been much dissemination of
child-centered training to teachers and parents. A change in instructional methodology may lie
behind the improvements in grade 1.
5. The government indicates that there are improvements in most districts, but the magnitude of the effect of the DPEP
is hard to ascertain from the available analyses since there is significant variation in results between districts.



7
* Textbooks have been provided on a large scale, and supervision missions have found them used
in classes (though not available for some children). Some areas have printed their own books,
while others have used state models. Textbook revision in Kerala has become a model for other
states.
* Extensive teacher training has taken place through block and cluster resource centers, with
about 100,000 teachers receiving some training each year.6 In project areas, 92 percent of
teachers reportedly received training, compared to 37 percent in non-project areas. Through
interviews, it was ascertained that training had influenced learning achievement mostly in
Maharashtra, Haryana, and Tamil Nadu, partly in Kerala and Madhya Pradesh, and moderately
in Assam. Direct grants for teaching materials are available to teachers who want to innovate.
*   Many teachers have been appointed in all states, particularly locally residing para-teachers.
Karnataka, West Bengal, Andhra Pradesh and Uttar Pradesh filled vacancies that had continued
for years. Though there are exceptions, primary-level pupil-teacher ratios in many districts
have dropped from 82:1 to 40:1. Perhaps teacher attendance is improving; one study carried
out in the Saharanpur district showed teacher attendance to be a more encouraging 78 percent7.
*   The hoped-for vehicles for local-level change have been the village panchayats and the village
education committees, of which at least 171,000 have been constituted during program
implementation by local-level appointment. They have received school improvement grants as
well as systematic training to understand their role in the process of education and to play this
role effectively. Women and scheduled castes are represented in the committees.
*   Much school infrastructure has been built, contracted by panchayats, which traditionally have
had competence in this area. For example, UPBEP helped build 8,388 schools, 12,264
additional classrooms, 14,654 toilets, 61,669 drinking water fountains.
*   State and district project offices have been established, as well as Block and Cluster Resource
Centers and the District Institutes of Education of Teachers (DIET). Despite many difficulties
(see below) these institutions have provided large-scale in-service training.
Some Concerns
3.4     The 9th and I Oth joint review mission supervision reports for the projects indicate several
issues and risks. OED interviews and a field visit (see Text Box 1) also highlight areas in need of
further study by the donor community and government. The following problems stand out, which
may impact areas of least literacy and/or greatest poverty.
6. In-service training was found to correlate 0.48-0.51 with student achievement in grade I and 0.34-0.37 in grades 3-4
(Midterm Assessment Survey, p. 88). The stronger correlation for grade I is consistent with learning improvements in
this grade, but it is unknown whether this correlation implies causation. Perhaps achievements are higher in areas that
find it easier to organize training and get teachers to attend.
7. Study referred to in: Making a Difference: Focus on Girls Education in Uttar Pradesh, India. UP Education for All
Project, Board Presentation made at the World Bank Human Development Week, March 2000. By contrast, PROBE, a
study undertaken by independent academics, showed that only 53% of teachers sampled in impromptu visits of 122
village schools (unknown whether they were under a Bank project) were found to be teaching.



8
Box 1: Field Observations - Obstacles To Student Performance Despite Best Intentions
On April 3, 2000 the OED mission visited two primary school and the district institute of education in
Saharanpur, Uttar Pradesh. This is one of the districts where the program has performed well, and which
the World Bank president visited in 1997. The primary schools visited served low-income children. The
school teachers had been forewarned of the visit a day earlier.
In supply terms, the classes were adequately prepared. Teachers had been trained to use methods that were
meaningful to younger children, such as the use of manipulative and sketched objects for reading and
arithmetic. The mission saw several items which the teachers had developed with the Rs. 500 each had
received for this purpose. Science was taught through the use of local materials and simple experiments.
(These methods require some preparation, so the extent to which these methods were used on a day where
no visitor was expected is unknown.) In two classes peer-teaching was observed, as children broke into
groups with one child teaching others. Students had textbooks in their possession, though some were torn
and unusable. Children in the lower grades wrote on slates with small pieces of chalk.
The usual classroom setup of Indian schools created inherent limitations, which clearly affected student
performance. The early grades had about 45 students, and the higher grades about 25. Children sat on the
floor tightly, and it was hard for a teacher to walk among them. Therefore the teacher stood en at the front
of the class and demonstrated or asked questions, which the students answered in unison. The classes were
held in close proximity in the open air or in rooms with open windows, where all the noise of the school
traveled freely." The deafening noise of the other classes could only be countered by children shouting their
own responses.
In effect, children's participation in large classes is voluntary. Many children, particularly those sitting at
the back, were not giving meaningful responses. In higher grades, children raised their hands to respond,
but teachers picked only among those who did. Therefore; if some children fell behind, teachers might not
notice it. An informal query of some students showed that they could not perform to presumed class
standards. Two third-grade girls did not know how to read at all; two others could read a text on national
geography aloud but did not know its vocabulary and had no comprehension of the material. Some first-
grade children in the back of the class, using tiny pieces of chalk, were unable to trace the letters on the
blackboard. Student records showed periodic evaluations, but many grades were low, 5 or 6 on a scale of
10. One would expect that this system would most severely affect students with short attention spans (due
to malnutrition) and without literate parents to help them at home. Maybe the students who had survived to
reach the higher grades were the ones best able to pay attention in a noisy environment and leam on their
own or with home help. Cumulative deficiencies despite much teacher training may be one reason why
achievement test gains were highest in grade I and minimal in grade 3 or 4. It is hard to imagine student
performance and classroom management of 300 in those remote areas which only have only one teacher.
Classroom time-on-task seemed a significant issue. Two teachers had gone to assist local board
examinations, and their students were sitting without doing structured work. Other teachers looked in
occasionally, but that was done at the expense of their own class time. (There was no system of substitute
teachers.) It was unknown whether students had been taught the curricular skills prescribed up to that time
of the year, but first graders were still reading without matras in April. It is known that the attainment level
of children completing four or five years of education is very low compared to minimum criteria.9 Given
this picture of relatively well-to-do schools, one can only imagine student performance and classroom
management of 300 in remote areas where the schools have only one teacher.
8. Research carried out in the US shows that children in open classrooms learn slightly less than children in contained
classrooms (John Slavin, Johns Hopkins University).
9. NCERT, Classroom Processes: A National Picture, Mid-Term Report, March 2000.



9
A. Local-Level Social Processes and Obstacles
3.5     While a great deal of thought and planning takes place at the central and state level, some
social processes at the grass-root level may be unwittingly unraveling part of the benefits. Such
processes may include the following:
3.6     Inactive Village Education Committees. The Uttar Pradesh Basic Education project
(UPBEP) financed the development of village education committees. A gender-oriented
beneficiary assessment commissioned by OED in December 1999 uncovered in Uttar Pradesh the
surprising finding that village education committees were dormant in all the sites consulted,
project and non-project alike. The researchers had lists of committee members, but most of those
who were interviewed (particularly women) did not know that they were on the committees or did
not know what their roles were. Also records indicated that in Sitapur district 1448 village
education committee members were trained, but members were unsure of any training given.
Many of the uninformed women stated that husbands were involved, instead. '� Supervision
missions and related research (e.g. PROBE) have also raised questions about the effectiveness of
village education committees in low-literacy areas. Some of the poorer members are unable to
participate in school meetings due to agricultural work and may be unable to monitor financial
matters. The difficulty of involving illiterate populations in school affairs has been documented
in Bangladesh and elsewhere."
3.7     Aside from spotty school monitoring, the inactivity of these committees raises a
governance issue. Certainly there are clear government guidelines regarding flow of funds, but it is
unclear who is managing school funds when the committees are dormant. The village pradhan and
the headmaster, who must co-sign checks, may do so with no other oversight. Without oversight,
some schools and teachers, who are eligible to receive funds for innovation and improvements, may
use the money for purposes other than instruction. Complaints about shoddy construction in some
areas, echoed in newspaper articles about village-level corruption may be isolated incidents, but
give further rise to concerns.12 The main cause for concern is the appointed status of these
committees, because influential local people may appoint complacent female relatives. Besides the
fact that schools may not function effectively without local oversight, there is a concern that
microcorruption may be fueled. Transparency International has reported that some panchayat heads
10. Gender Impact Assessment Study (Basic Education Project - Uttar Pradesh), PRAXIS - Institute for Participatory
Practices, November-December 1999. Districts covered were Nainital, Sitapur under UPBEP and Almora,
Baageshwar, and Lucknow were not under UPBEP or DPEP and hence provide a counterfactual without the project.
11. The Government disagreed with the assessment of the VECs in this paper. According to the Department of
Elementary Education and Literacy in the Ministry of Human Resource Development, VECs were in general working
well, although with varying degrees of functionality across states. It was also pointed out by the Department that this
paper overlooks the fact that in many DPEP states (e.g., Maharashtra, Gujarat, Karnataka, Kerala, and also in some
other states (e.g., Madhya Pradesh and Rajashtan), the panchayats have effectively taken control over elementary
education since the 73rd and 74th amendments to the constitution ushered in considerable decentralization. On the
VECs the author obtained different information, but the increased role of the panchayats is acknowledged.
12. E.g. Bhatia, Bela & Jean Dreze, For Development and Democracy, Frontline, March 6, 1998.



10
have admitted to misappropriating money for small development works, including schools. Though
many village poor can organize and confront abusers, others cannot.'3
3.8     Fewer Teachers in Areas Where They Are Needed Most. Despite many teacher
appointments and consistent government efforts, large numbers of positions still go unfilled. One
problem is the strong political and social pressure to move teachers close to their homes. The
mission heard concerns that some panchayat heads may be bribed to acquiesce to transferring
teachers away from the areas of need or to absentee teachers who may put someone else in their
stead for a small remuneration. Irregular teacher transfers result in overcrowded classes in those
areas where the students are most vulnerable. State or district averages tend to obscure this problem.
For example, research in Assam showed that in remote areas student-teacher ratios sometimes rose
to 1:181 to 1:300.'4 Teacher absenteeism is of considerable concern, particularly in areas of low
literacy and limited activity of village education committees. Effective mechanisms to control
teacher transfers and absenteeism are not yet in place. Kerala and Karnataka have need-based
placements and systematic teacher assessment, but most states have not followed suit. The
appointment of para teachers may solve residence problems. Despite the lower education that many
of them have, it has worked well world-wide. However para teaching may be seen by some Indian
educators as a way to give the cheapest education to the students who need it the most.15
3.9     Clearly, low-literacy districts present special challenges. It may be more difficult to
mobilize the poor in areas where traditional patronage is given to local persons of importance and
the population has been accustomed to exploitation. In addition, there seem to be local customs for
expropriating government benefits, which are not well understood by policy-makers.'6 Sensitive
social research is needed to understand why policies to give control to the very poor have not
worked as well as theyi were expected. The most deprived areas need more resources and detailed
presence in the field.
3.10    Possible School Construction Problems Formal reports show a satisfactory condition of
schools. '8, but informal reports are somewhat conflicting. There have been some complaints about
13. Bhatia, Bela & Jean Dreze, Campaign in Rural India. Transparency International Working Paper, Berlin, 19
September, 1998 (www.transparency.de). Possibly panchayat-level corruption misuses 30"/o of funds compared to the
usual 220/o, as reported in the article.
14. Primary Education Scenario in Darrang District of Assam: A Status Review. District Primary Education Program,
Assam and National Informatics Center, 1996. See also Annex B.
15. In the workshop held on March 29, 2000 participants emphasized this issue. At any rate, the Elementary Education
Bureau maintains that the para teachers have the same qualifications as other teachers and that there are training
programs in place for them.
16. Dreze, Jean. Comments on Mahal, Shrivastava, and Sanan's Paper. Governance, Decentralization, and Reform in
China, India, and Russia. Jean-Jacques Dethier (Ed.) Kluwer Academic Publishers, in press.
17. In its comments on an earlier draft of this paper, the Government stressed that "most states have already taken
measures: either through teacher rationalization (MP) or appointment of local persons as para-teachers (AP, MP, UP,
Rajasthan, etc.). The past two years have seen a significant improvement in teacher availability, not only through para-
teachers but also appointment of regular teachers against vacancies-Karnataka, West Bengal, AP and Uttar Pradesh
have filled vacancies that had continued for many years. A study on teacher absenteeism in Uttar Pradesh 1998 showed
high teacher attendance. This study was conducted through spot/surprise inspection in two districts of Uttar Pradesh.
18. In the case of DPEP schools, a government-commissioned third-party evaluation found that the quality varied from
average to very good. This result compares very favorably with PROBE study findings in sampled districts, which



school quality, e.g., that plaster peels off, doors fall down, brick walls collapse. Sometimes school
sanitation had not been adequately addressed, particularly in the earlier schools. Sometimes toilets
are constructed where there is no water, and they may be used as fodder storehouses or be limited to
teachers; students may use the fields, rendering them unsuitable for athletics. The consultants of the
OED gender impact assessment study found almost no functioning toilets in the Uttar Pradesh
districts they visited. Lack of facilities deemed necessary to enable girls to go to school may
contribute to dropout in the most vulnerable areas.
3.11   Push-out of Low-Caste Children. Government research shows that enrollments of low-
caste children are often commensurate with the population and that social equity has increased in
DPEP districts.'9 However, there are concerns that low-caste children may still be pushed out of
classes by teachers who are indifferent or who assign them tasks appropriate to their caste.20 For
example, the percentage of dalit dropouts is considerably higher in Kerala schools, despite the
attention and budgetary resources.2' Some teachers may be doing this without realizing it, and the
government has conducted sensitization training. Some officials have stated to the OED mission
their concerns that this is still happening in some schools financed by Bank projects, but no
assessment has yet been carried out about the incidence of this issue.
B. Low Quaity of Education
3.12    Perceived Low Quality of Some Government Schools The physical and staffing
limitations of schools, particularly those in the less advantaged areas, affect student information
processing and subsequent performance. Criterion-based achievement test results show limited
performance improvements in language and particularly in mathematics in higher primary grades.
For example, UPBEP end-of-project achievement tests showed language performance at grade 5 at
41.3 percent (compared to baseline performance of 41.5 percent) and math performance at 34.8
percent (compared to baseline performance of 33.3 percent; National Council of Educational
Research and Training - NCERT considers the ideal mastery criterion as 80 percent). Teacher
behaviors are being extensively studied in hopes of improving their impact on students.22
3.13    In light of teacher absenteeism, some parents who can barely afford to, pay Rs. 30-50 per
month to send their children to small private schools. 23 Overall, 4.8 percent of the students go to
schools that are not recognized but are perceived to be results-oriented, because teachers and
management are under private control (also because they teach English). In Haryana, 18.7 percent
included two DPEP districts (Public Report on Basic Education in India, 1999. New Delhi: Oxford University Press);
the PROBE study found that in 70% of the 122 government schools sampled, the roof was leaking.
19. The index of social equity for phase one districts shows that districts with an index greater than 95% have increased
from six in 1995-96 to 14 in 1998-99.
20. PROBE report, p. 49. Also Jean Dreze and Geeta Gandhi Kingdon, School Participation in Rural India, 2000 in
print
21. Dalit Dropouts more in Kerala Schools. Statesman, January 26, 2000.
22. Classroom Processes: A National Picture. Mid-Term Report, National Institute of Educational Research and
Training, March 2000.
23. Aggarwal, Yash. Primary Education in Unrecognized Schools in Haryana: A Study of DPEP Districts, National
Institute of Educational Planning and Administration, 2000.



12
of the students and in Punjab 15.5 percent attend such schools. Withdrawal of students from the
public system accounts to an extent for the recorded dropout rate in the higher grades. The costs
and benefits of students attending these largely unlicensed private schools deserve attention.24
3.14   Classes in government schools tend to be large and noisy and do not facilitate information
processing (see Text Box I on page 8). The current school organization and classroom
management may cause all the government planning, financing, and initiative to hit a brick wall.
To improve student performance, fundamental changes are needed, which the poorer Indian states
currently cannot afford: smaller and less dense classes, schools built to minimize noise, substitute
teachers and/or teacher assistants, provision of writing implements to younger children so that they
can use their time effectively. These changes are probably not possible in the next decade, and the
present system has been used traditionally, therefore it may seem normal to Indian educators. Yet,
there is an urgent need to study how to improve learning within the current physical limitations of
the schools and class hours. An NCERT classroom processes study underway studies teacher
behaviors but not what students do in response and how their performance is affected).25
3.15   One feasible improvement might be remediation. Students with particularly low
performance should be caught early and receive after-hours or summer classes by teachers who
would be paid through the village education committees to teach extra hours. Remedial measures
may lower dropout and repetition and increase learning achievement. Also, targeted educational
research is needed, perhaps in collaboration with a foreign institution that has expertise in
information-processing problems, to assess the relative importance of the various distractions
present in low-income classrooms in determining children's' performance and finding low-cost
remedies. 26
C Uneven Quality of In-service Training, Modest Institutional Development
3.16   The I 0th joint supervision report indicates that institutions which must impart guidance and
training need capacity building, and therefore insufficiently involved in the DPEP's pedagogical
renewal process. State Institutes of Educational Management and Training (SIEMATs), expected
to impart suitable teacher and administrator training, have hardly been established. District
Institutes of Education and Training (DIETs) also face problems. Block and cluster resource centers
lack the guidance that these institutions can offer.
3.17   Many SIEMATs and DIETs find it difficult to appoint or to attract suitable staff. The
difficulties of these institutions indicate that they may not have the optimal administrative
placement. 27 In many of the same towns, there are colleges of education that have qualified staff
24. Kingdon, Geeta. Private Schooling in India: Size, Nature, and Equity-effects,. Economic and Political Weekly, vol.
31, December, 1996.
25. NCERT, Classroom Processes: A National Picture.. Mid-Term ReporL March 2000.
26. The Department of Elementary Education and Literacy has drawn attention to improvements being made to the
teacher/pupil ratio, the competence of teachers, incentives for students in SC/ST communities, and the provision of
learning materials
27. Evaluation ofManagerial Structures and Processes under DPEP: A Synthesis Study,. Research, Evaluation, and
Studies Unit - DPEP. 1996.



13
and could assume some training responsibilities, but they have very limited contact with schools
and different budgetary sources.
3.18    Much teacher training takes place, and its effects may be seen in classrooms. Nevertheless,
supervision mission reports indicate limited insight among trainers regarding teaching methodology
and the rationale for training of trainers. Sufficiently qualified instructors typically come from
secondary teaching and may not be knowledgeable on effective instruction for small children or
how to bring it about through training in classes. Advanced training among trainers is rare. For
example, there is much discussion on child-centered learning, but there is a wide opinion difference
on what the term means.
3.19    The instructional problems of rural schools demand new solutions, and India could have
honed its instructional methodology by seeking inspiration from international best practices on
how to optimize learning for poor children. However, the government has not welcomed foreign
technical assistance in instructional methodology and in-service training issues and has relied on
local consultants. The Bank has concluded that they are well-trained professionals who may need
little further inputs. According to other sources, some may have outdated knowledge and no
experience with rural schools. Because the government signaled that they did not desire help in
this area, donors, including the Bank, have made limited inputs. (One notable exception has been
the joyful learning methodology.) This has led some observers in India to believe that the Bank
had little to offer in the area of instructional methodology.
3.20    Though in-service training is essential, there is a concern that teachers are taken too often
out of the classroom for it; in areas of particularly limited teacher availability, in-service training
may be one more factor that reduces the number of contact hours with students.
D. Issues of Data Reliability, Analysis, and Objectivity
3.21    The data which help gauge the impact of the program may have varying degrees of error.
Examples are:
3.22    Questionable enrollnent increases. There is some concern that the impressive
enrollment increases of the DPEP program may be to some extent inflated.29 Overall,
government primary-school enrollments apparently are inflated by 20 percent.30 Likely reasons
28. Raina, Vinod, External Funds,Jnternal Conflicts, Right to Education. Seminar 464., April 1998. However,
according to government comments, "India believes that contextual knowledge is far more important than international
expertise and has used government experience and NGOs like Bodh, Eklavya, Digantar, MV foundation, which have
considerable experience in the field... it needs to be recognized that there are enough best practices available in the
country."
29. Kingdon, Geeta G. and Jean Dreze, Education Statistics: Myths and Reality, The Hindu, March 6, 1998.
30. India: Policies to Reduce Poverty and Accelerate Sustainable Development. World Bank: Report No. 19471,
January 31, 2000. Poverty Reduction and Economic Management Unit, p. 104. Cited therein are: Government of India
(1998) Attending an Educational Institution in India: Its Level, Nature, and Cost, Fifty Second Round (July 1995-
June 1996). National Sample Survey Organization, Department of Statistics, Ministry of Planning, Statistics, and
Program Implementation, New Delhi. Also Public Enterprises Survey 1997-98, Department of Public Enterprises,
Ministry of Industry, New Delhi. Nevertheless, the Elementary Education Bureau defends the accuracy of the data and
states that the data reporting systems have been strengthened.



14
are fear of teacher transfers from favorable schools, pressure to show progress, food incentives
for teachers and students.3' Ironically, inflated enrollments may also inflate dropout rates, thoufh
attendance registers may also be inflated. The government has studied enrollment progressions 2
and verified enrollments based on a 5% sample check of 1998 data, which found that data on
enrollment were fairly accurate ranging from zero to five percent despite a lag of one year in
Assam, Karnataka, Kerala, Madhya Pradesh, Tamil Nadu and Himachal Pradesh. Overall the
quality of data was found to be dependable.  That is comforting, but since there is much focus
on enrollment increases, particularly for girls, in areas where there has been cultural resistance, it
may be useful to study different data sets for such evidence. For example, independent
verification may come by analyzing attendance data from the second round of the National
Family Health Survey (1997-1998), in comparison with the first round (1992-1993). Data from
the Public Report on Basic Education (PROBE) may also be analyzed to show progress in
education achievements and school conditions in the two or more districts where there is overlap
in the coverage of DPEP and PROBE and where data are comparable. The surveys for the
PROBE report might also be repeated in all districts initially sampled so as to evaluate the net
contribution of DPEP even more accurately. The 2001 census will show the extent to which
children of various ages and socioeconomic conditions attend school.
3.23   Some incorrect data on school construction. Data on the condition of schools sometimes
do not coincide with the condition found in the field, and in some areas, construction may not
have taken place as indicated. For example, academics recall visiting schools in Nainital (Uttar
Pradesh) and in Chamba (Himachal Pradesh), which on paper were complete, but very few
students were in attendance and there was no boundary wall34 or washrooms. These may be
individual incidents, but systematic verification of the quality and reliability of monitoring data
may therefore be required in some areas.
3.24   Achievement data reliability. Student achievement has been studied extensively and data
are available for various levels and areas of the country.35 DPEP used sophisticated sampling
techniques to obtain for its districts a learning baseline in language and mathematics in 1994 and
to monitor progress in 1997. Numerous field assistants were used, but lack of money and time
limited the supervision of their work. Some reportedly did not visit the schools whose data they
recorded. The extent of the problem is unknown, and the Elementary Education Bureau (I)
denies that it happens. Applying sophisticated statistical procedures on unverified data may lead
to misleading results.
31. Public Report on Basic Education in India (PROBE), 1999, p. 91. According to some sources the data include
some underage, nominal, double, and fake enrollments.
32. Aggarwal, Yash, How Many Pupils Complete Primary Education in Five Years? National Institute of Educational
Planning and Administration, 2000.
33. Communication from the Department of Elementary Education and Literacy, Ministry of Human Resource
Development, October 20, 2000
34. Govemment comments indicate that boundary walls were sometimes not built due to a lack of funds.
35. See, for example, School Effectiveness and Learning Achievement at Primary Stage, NCERT, in A.K. Sharma (Ed.)
New Delhi, 1995.



15
3.25   National-level data are extensively analyzed by the Elementary Education Bureau, but
access to them has been mainly restricted to the two institutions, NCERT and NIEPA (National
Institute of Educational Planning and Administration). Few other consultants and academics
have been pennitted to analyze the data. These restrictions raise issues of openness and
objectivity. At the same time, the levels of analysis carried out have been inappropriate. Some
reports (e.g., the widely disseminated Mid-Term Assessment Survey) contain data by district only
and provide no national averages to compare against. Without means of central tendency and
appropriate statistical tests, it is very hard to make sense of figures pertinent to project
achievement. 36 The program has a large research component and spends significant amounts for
management information system, data analysis, and studies, but recent supervision mission
reports say little about the work of this centrally administered DPEP research program. It is
unclear how funds for it are spent.
3.26   Donor staff report that they have limited access to studies carried out on DPEP data.
Requests for lists of studies often go unanswered. The OED mission was also unable to get a list
of studies, despite repeated requests and assurances that they would be sent.
E. Supervision Concerns
3.27   Aside from data reliability issues, program effects are rather hard to ascertain. To foster
the high degree of client ownership that the government desired, the donors agreed to visit
educational institutions in various states only after obtaining permission from the Elementary
Education Bureau, which is sometimes withheld. Generally, they do not visit sites without
considerable advance notice; this makes sense logistically, but creates problems of objectivity.
Generally, districts determine which schools will be visited, and these (according to some
workshop participants) may be the better schools. Staff know several days in advance that the
donors and government will visit and therefore prepare well. With advance warning, long-
postponed repairs may be carried out, giving the impression that schools are better maintained
than they are, students and teachers may come who do not usually do so, and instructional
methods may be used that are deemed too-time consuming for everyday use. Thus, visitors may
overestimate project achievements in certain areas. Very often staff do stop by schools on the
road that are not on the official visit list, but it is unclear how representative these schools are of
district schools.
3.28   Supervision should be exchange of knowledge and not policing. If the same persons
visited the same areas repeatedly, progress (or lack thereof) might be easier to spot. But until
recently, there was a lack of continuity in the supervision staff, as various donors send different
staff and consultants each time, and even the same staff visit different districts. This supervision
mode may be a good learning experience for donor staff, most of whom do not know Indian
languages and cannot follow classroom instruction, but it may not be the best way to find out how
project schools work on ordinary days.
36. NCERT staff received training in the sophisticated technique of hierarchical linear modeling through DPEP, but not
in more common and useful techniques.



16
3.29    Upon return from supervision trips, government staff scrutinize and challenge the aide-
memoires of donor staff in lengthy pre-wrap-up sessions. Discussions often result in deleting
issues that might be embarrassing to the government. As a result, supervision reports do not
present all the important problems encountered in missions. Some donor staff defend this
practice as long as the government makes a commitment to improve the problems at hand.
However, it is hard for subsequent missions and for evaluators to follow up on unwritten
promises, particularly since there is alwavs some donor staff turnover. Also, if supervision
mission reports accentuate the positive aspects of the projects and project data are not widely
available for analysis, achievements may be exaggerated. Thus, donors may be criticized for
presenting an overly positive or even naive project view. 3
3.30    Supervision missions visit many sites throughout the country, ask a common set of
questions, and they could be collecting school data and establishing a data set for future analyses.
Thus far they have decided not to do so. However, such a data set might provide valuable
information on changes taking place as inputs are provided. Supervision logistics may be
rearranged so that many school sites are visited unannounced. Decisions on schools to visit in a
certain district could be taken on the day of the visits through a random selection process.
4. Evaluation Of IDA Assistance
4.1     Based on the evaluation of the closed projects and incomplete assessment of the ongoing
projects, this sectoral review has reached the conclusions that follow.
Outcomes
4.2     The relevance of the Bank's lending and sector work has been high. All projects have
been relevant to the strategy of developing human resources in India and to the country's
economic development needs. Though formal sector work was carried not out early, the
government's initiative contributed to relevant work that dealt directly with sectoral issues.38 The
TVET projects focused on the first areas open to IDA financing and the objectives were relevant
to industrial development. Project designs were appropriate to the objectives. All projects
focused on increased access, quality, female participation, and improved responsiveness at the
level of local institutions, and the primary education projects correctly targeted areas of low
female enrollments and financial inputs in the poorest districts of the country.
37. The Department of Elementary Education and Literacy has objected to this characterization of the supervision
process. It has stated, inter alia, that "DPEP has devised an elaborate Joint Supervision Mechanism involving
professionals and experts from various external funding agencies, national level resource institutions and local experts.
The supervision mechanism is very collaborative, the leadership rotates amongst all agencies and GOI, and States'
viewpoints are given due consideration." The Department also emphasized that supervision missions do "have wrap
ups with the States and with the Government of India, so that information gaps can be filled up and any misconceptions
cleared. But to suggest that such an elaborate system, with so many players from different countries and institutions
can be doctored is highly unfair and unwarranted".
38. Bank Report No. 15756-IN, India: Primary Education Achievement and Challenges, September 1, 1996



17
4.3    Efficacy has been substantial, as all projects have proceeded satisfactorily. The TVET
projects focused on states that had the willingness and the potential for industrial development,
rather than the less industrialized states. On the other hand, the primary education projects
focused on districts where the education of the poor might have the highest impact on improved
health, lower rate of population increase, and income potential. The numerous achievements of
the projects are catalogued in the audits of the TVET projects and in para. 3.2 of this report.
4.4    The use of IDA resources was efficient. The TVET projects were very complex
multistate operations that demanded a high level of government expertise in procurement, civil
works, and equipment acquisition, and complex planning of coordinated inputs. Despite the vern
unimpressive performance of the state departments of public works, there seemed to be little
wastage of resources or defective civil works. From the Bank's perspective, the primary
education projects cost little because they were appraised locally. Given the innovative, district-
based design, estimating the efficiency of funds use is hard. Perhaps greater savings might have
been achieved with large-scale procurement, but the benefits of empowerment and institutional
development likely to come about in the long run would likely have been lost. The impact of
these projects on the supply of trained human resources in the country is likely to be substantial i
the long run.
Institutional Development Impact
4.5   In the late 1980s few economists were working on education, there was no central or stat
capacity to develop and manage a program such as DPEP, and the state and central governments
did not have the framework for managing a major investment thrust in primary education. As
discussed in Annex A, many state and local institutions were asked to do a great deal of planning
and implementation work, and many rose to the challenge; district education offices are now
much more involved in local planning and management of education. It appears that dialogue
with donors, sector work, lending, and supervision greatly contributed to the development of this
capacity in India. Compared to the relatively small amount of money that the World Bank
invested in education of such a large country, the efforts are impressive.
4.6    Though institutional responses at the top were very satisfactory, changes at the base,
which determine how well schools are managed locally, are less certain. District and state
institutes of education remain unable to provide training leadership, and some districts with low
literacy seem unable to monitor implementation effectively. Perhaps the incentives that will
sustain project impact have not yet been changed. For this reason, institutional development is
rated as modest.
Sustainability
4.7    Sustainability of the technical-vocational as well as the primary education projects is
rated as likely. Although it is hard to sustain better-than-average programs when the areas aroun(
them continue to be deprived, the government is making many efforts to ensure sustainability.
Uttar Pradesh has approved taking over the entire burden of continuing good practices of UPBEP
after the project closes this September. An amount of over 125 crore has been provided in the
state education budget. Tamil Nadu has extended successful DPEP interventions in non-DPEP



18
districts with its own resources. Kerala has provided financial resources for training of all
teachers in the state since they will use DPEP textbooks. Rajasthan has committed financial
resources to carry on project intervention after project completion. So, sustainability is more
certain in some states than in others.
4.8     Since none of the basic education projects have closed and formnal evaluations have not
been undertaken, rating Bank assistance is difficult. Tentative ratings of the country sectoral
assistance as a whole during the 1990s are as follows:
AGGREGATE ASSISTANCE RATING                     Satisfactory
Outcome                                        Satisfactory
Relevance                                      High
Efficacy                                       Substantial
Efficiency                                      Substantial
Institutional development impact               Modest
Sustainability                                  Likely
Note: The scale for institutional development as well as Bank and Borrower performance is "highly satisfactory", "satisfactory",
'"unsatisfactory", "highly unsatisfactory". The outcome rating also includes "moderately satisfactory" and "moderately unsatisfactory"
levels. Sustainability is rated as likely, uncertain, and unlikely.
5. Sectoral Issues - Directions For The Future
5.1     A country with the size, income, and variability of India requires different strategies at
different stages. As implementation experience increases, the issues discussed in this section are
becoming increasingly important.
Open Discussion of DPEP Successes and Obstacles
5.2     The government's commitment, willingness to implement, and detailed preparation of
supervision missions, are most impressive and a cause for admiration. However, the evaluator is
presented with a dilemma, because the government controls inputs, processes, and outputs:
*       It is difficult to evaluate the provision of inputs, partly because donors are usually unable
to visit project sites unannounced and observe operation as it happens under everyday
conditions;
*       It is difficult to evaluate outputs and impacts, because donors have limited access to data
and studies; studies are executed by a few chosen institutions;
*       The reliability of the data, particularly of enrollments, has been questioned to some
extent;
*       The record of achievements is incomplete because some agreements between government
and donors remain unwritten, while some controversial issues are omitted from aide-
memoires.



19
5.3    The government's very impressive achievements will be much more credible if the
obstacles and the difficulties encountered are openly discussed. The difficulty and magnitude of
the task are well understood, as is the fact that the easier targets (e.g. enrollments in more
accessible areas) often are achieved first. The more difficult issues can be tackled more
effectively if they are presented and studied. For example, the inactivity of village education
committees in low-literacy areas and the social factors that contribute to it need to become a focus
for study and innovative thinking.
5.4    The Elementary Education Bureau has relatively few staff members, and they are
extremely busy with information they get from state project offices. Opportunities to visit the
field are relatively limited. The government may obtain more valuable information from donor
staff if supervisions are more impromptu and if donor interaction with states is direct. The
government may also benefit from data verification and analysis by independent groups and
researchers.
Financial Accountability and Local-Level Concerns in DPEP
5.5    The donor community has been careful to allow ihe government to implement the
projects with minimal intervention. However, concern for government sensitivity and client
ownership must be balanced with accountability and verification that the Bank funds are used for
the designated purposes. Unquestionably, the intentions of the central and state governments'are
good, but local circumstances may present unforeseen difficulties.
5.6    The difficulty of obtaining verifiable assessments of inputs and outputs is coupled with
limited financial accountability. Bank supervisions typically do not have the closeness of contact
needed to detect mismanagement, an issue which should certainly interest the government. This
is a problem not only in education but in other sectors. OEDCM (an operations evaluation group
dealing with evaluation methods) studied the financial accountability of Bank projects in India
and concluded that the rules carefully laid out by the government may not always be followed;
therefore the projects are not automatically insulated against fraud, waste, and abuse of public
funds since these problems is growing.39
5.7    During the appraisal of the DPEP, an ingenious system was devised to funnel funds to
village education committees through nonprofit societies (controlled by the DPEP state project
offices) in order to prevent delays that might be caused by government mechanisms. As
Transparency International reports,40 it is possible that there is significant local-level
mismanagement in areas of low literacy and limited community involvement. There are clear
government guidelines regarding flow of funds, but the extent to which they are followed at the
lowest levels is unclear. No systematic research has been undertaken of community funds usage,
and supervision reports do not mention it. However, one Bank financial analyst who visited some
DPEP districts observed that accounting knowledge was very limited at the local level, and that
39. See background paper for this CAE on public financial accountability in India titled Evaluating Bank Assistancefor
public Financial Accountability in India in the 1990s: A Desk Review, by Vinod Saghal and Deepa Chakrapani, July
10, 2000. The Auditor General of India plans to carry out a performance audit of DPEP.
40. Transparency International, op. cit.



20
funds might not properly be accounted. Though amounts may be small, their misuse may be
visible to local residents. Thus critics may raise concerns that the World Bank is not interested in
mitigating corruption.
5.8    Beneficiary assessments need to be made to verify the extent to which the panchayats,
village education committees, teachers, and headmasters use the small amounts of money they
receive for school improvements. In other countries, such as Bangladesh, Transparency
International has conducted surveys that document the extent to which free textbooks and tuition
are in fact free and identify pockets or patterns of mismanagement. This or other organizations
could conduct such research in DPEP areas in hopes of identifying the social forces that must be.
strengthened in order to reduce the incidence of the problem wherever it happens. Understanding
how to strengthen village committees against abuse may help other projects that depend on
grassroots mobilization in very poor areas, such as the Andhra Pradesh District Poverty Initiatives
Project.
Need for Integrated Sectoral Dialogue
5.9    Decentralization ofproject leadership to the state level. The earlier district-based
projects are multi-state projects in which the central government leads the implementation. In
effect, the government finances the expansion of the system. In turn, the World Bank mainly
interacts with the central government and very little with state governments. Given the level of
decentralization that is already in place, it appears that the project leadership must move from the
center to the states, but little thought has been given to this issue. The nature of the Bank's
relationship and contact currently with the central government might evolve towards state
governments.
5.10   Taking a holistic sectoral view. Policy and investment programs should be tailored to the
needs of the whole sector. Millions of primary school students are graduating without the
availability of lower secondary schools to attend. (The donor community and the government are
studying this issue, but implementation is still to come.) Also, the relatively wide availability of
higher education at low student fees contrasts with the limited financing available for primary
education. For this reason, some Indian academics believe that the country does not need World
Bank money to implement primary education and that rationalization of higher education fees
might create the savings necessary to, substitute for donor funds. Higher education in particular
must be examined in light of the finances and access of other subsectors.
5.11   Higher education, particularly science-oriented (e.g., engineering and medicine) is very
socially desirable; engineering education functions as general university education for the well-
to-do and the better students (traditionally boys). As a result, parents put a lot of pressure on the
government to operate and finance about 400 public engineering colleges. Though most students
can pay and have high earnings upon graduation, these institutions still charge low fees. The issue
is politically sensitive worldwide, but some have found ways to meet the challenge. East Asian
countries, such as Malaysia and Singapore, provide many student loans and inform their students
about the true cost of education. Other countries with large numbers of universities, such as the
United States and Canada, limit public expenditures to these schools and encourage them to raise
funds from alumni. The extensive provision of higher education in India is widely recognized as a



21
cornerstone of its impressive economic development in the past decade. Nevertheless, spending
large amounts on those who can pay limits the amounts available to the multitudes who will never
access higher education.
5.12   The Bank has not entered into an extensive dialogue with the government on higher
education, but is preparing a project to support regional engineering colleges. A cornerstone of
the dialogue and lending might be the decrease of government financing for these institutions.
Like the self-sustainability scheme of the Department of Electronics, the government might
gradually decrease outlays to engineering colleges while enabling and stimulating them to find
funds elsewhere.
5.13   Focus on planning capacity in the government and in associated institutions. Aside from
the impressive project-related planning efforts of the central government, planning capacity at
lower levels is limited. District-level planning focuses on student and financial projections but is
too little involved in long-range planning or in the various education subsectors. Finally, there
seems to be little global strategic thinking and policy analysis at the highest level. For example,
the Bank has found it difficult to engage secretaries of secondary and higher education in a joint
dialogue, given the commonality of issues. The Bank and government need to work toward this
goal as an upcoming stage in their developing relationship.
Better Use of International Experiences and Technical Expertise
5.14   India has shown some reluctance to learn from the educational lessons of other
countries.4' The technical education projects made limited use of international technical
assistance, and the DPEP program has none. Foreign technical assistance should not necessarily
be used, but international expertise may be useful. For example, India could obtain institutional
twinning with universities that have developed expertise in researching information-processing
processing problems of poorer countries. Also, solutions to the administrative difficulties of the
state and district institutes of education may be sought in East Asia or Latin American models.
5.15 The donor community can serve as mediators to put government staff in touch with
organizations and governments in other countries that have dealt with the same issues and have
developed various solutions.
6.     Recommendations
6.1    For primary education, OED suggests:
*      Verification of enrollments, project benefits, andflow offunds at the local leveL NGOs
might undertake these activities and report to the government and the donors, thus giving
the government the opportunity to mitigate problems early. Donors should visit many
school sites unannounced. DPEP data could be compared to those of other studies. For
41. The Department of Elementary Education and Literacy disagrees with this statement. It believes that India has all
the knowledge needed for its particular circumstances.



22
example it may be possible to do a more credible analysis using school attendance data
from the second round of the National Family Health Survey (1997-8), esp. in
comparison with the first round (1992-3). The PROBE data might also be analyzed to
determine differences between DPEP and non-DPEP districts.
*      Increased availability of the studies to donors and qualified academics, improved quality
of reports and statistical analyses. More extensive data analyses at the local, district, and
state levels, particularly by independent academics.
*      Assessment of the micro- and local-level processes that greatly influence project success.
Study of the social or other types of obstacles which arise (e.g. difficulties of energizing
village education committees, keeping teachers in remote areas, local microcorruption
customs) in search of remedies.
*     Remediation of some fundamental limitations of lower-income classrooms (e.g. densely
packed large and noisy classes of children who receive limited teacher attention) through
after-hours remediation of the weakest students and specific training to help teachers focus
on low performers early on.
*     Restructuring the district and state institutes of education which thus far have not been
able to provide suitable service or locate suitable staff. Study world experience in locating
such capacity in various administrative entities, such as local colleges of education.
6.2    For the lending program in sectoral terms, OED recommends:
* An integrated approach and study of the effects of one subsector on another,
* Increased financing of higher education through user fees and alumni contributions;
* Direct lending to states;
* More extensive use of donors' technical expertise and world experience in various
methodological and administrative issues.



23                                   ANNEX A
Background, Policy Framework, And Bank Education
Lending History In India
I.    India's goal of attaining free and compulsory education for all children up to the age of
14 years has been an elusive one. The Indian constitution envisaged achieving it by 1960. The
goal was reaffirmed at the Karachi Conference in 1960 when the target date for achieving
Universal Primary Education (UPE) was reset at 1980. Subsequently, the target date has been
shifted repeatedly. '
National Trends in Education
2.     India has made some significant progress in increasing enrollment, increasing retention,
and reducing the gender gap through policies aimed at improving basic education quality and
access. Still, its progress in education has been moderate in comparison to some other countries,
particularly China. In 1997, for example, adult literacy rates in India were still a very low 39
percent for women and 67 percent for men, compared to 68 percent for women and 87 percent for
men in China (Annex A Table 2). While India already trailed China in literacy levels in 1970, its
subsequent gains also were smaller: India's overall literacy improved by 20 percentage points (to
53 percent), while China improved by 31 percentage points (to 83 percent).2 India's weak
education base contributes to is lag in other indicators of human development, which are now
recognized to have a strong correlation to education. It also continues to be among the countries
with the largest persisting gender gaps in enrollment at all levels.3
3.     Despite an increase in national enrollments in primary schools from 85 million in 1987 to
100 million in 1993, 32 million children ages 6 to 10 were still out of school. All-India
Educational Surveys conducted by the National Council of Educational Research and Training
(NCERT) actually indicate declining overall gross enrollment ratios (GERs) for the primary
sector. In 1986/87, the total GER for grades I to 5 was 95.8 (110 for boys, 79.8 for girls). This
rose to 104.5 in 1993/94 (115.3 for boys and 92.9 for girls). In 1997/98 the total GER declined to
89.7 (97.7 for boys and 81.2 girls).4 Official enrollment statistics, however, reflect enrollment at
the start of the academic year and do not take into account the children that enroll but never really
attend school or attend only intermittently. Thus, statistics often conceal rather than reveal the
real state of affairs. Besides, aggregate statistics in India hide major inter- and intra-state
differentials.
1. This annex is abstracted from a background paper prepared by Ms. Sukhdeep Brar, consultant (currently educator at
the Asian Development Bank) in December 1999.
2. World Development Indicators, World Bank.
3. Population Action Intemational, 1998, Educating Girls: Gender Gaps and Gains, Washington DC.
4. Selected Educational Statistics, 1993-94 and 1997-98, and Education In India, 1986-87, Department of Education,
Government of India, New Delhi.



24                                  ANNEX A
4.     Though enrollments rose significantly before 1993/94, dropout rates remained high at 35
percent. Expansion of access led to an increase in pupil-teacher ratios, which increased from 55:1
in 1980 to 64:1 in 1993 - among the highest in the world. During this period, expenditures on
education by the center and states have fluctuated. They rose sharply in the late 1980s, peaked in
1992/93, and have fluctuated since then. Although the share of total recurrent budgetary resources
devoted to education has been increasing in recent years, from 11.8 percent in 1986/87 to 13.4
percent in 1994/95, it is still below the average of 17.5 percent for all low-income countries
excluding China and India.5 In 1997/98, India's budgetary estimates of public expenditure on
education were 3.62 percent of gross domestic product (GDP), compared to an internationally
recommended norm of 6 percent (nonetheless, India's expenditures are comparable to those of
similar countries). With a relatively low per capita GDP, this translates into lower per capita
expenditures on education (Annex A Tables 3 and 4).
Policy Framework and Dialogue
5.    In 1986, the Government of India (GOI) adopted a revised National Policy on Education
(NPE) to meet the changing demands of India's economic, political, and social goals. The new
policy acknowledged that the general formulations in the 1968 policy did not get translated into a
detailed implementation plan accompanied by assignment of specific responsibilities and
financial and organizational support. Therefore, even though the goals of NPE were also defined
in broad, abstract terms, the Department of Education (DOE) followed this up with a
comprehensive program of action detailing implementation strategies. The NPE particularly
emphasized equalizing educational opportunity up to age 14 for those who had been denied
equality - women, scheduled castes and scheduled tribes, the handicapped, and some minority
groups - improvement of educational quality, and adequate institutional infrastructure to the
educationally backward areas. To ensure that the state governments accepted the policy
directives, DOE assumed much of the fiscal responsibility for achieving these objectives.6 This
action effectively opened the sector to external funding for basic education.
6.    In 1988, the Bank and the government agreed to make human resources a critical focus of
their dialogue and overall collaboration. The need to open the social sectors to IDA funding was
first recognized by the Department of Economic Affairs (DEA), Government of India, as a way to
increase the absorptive capacity for IDA funds. Traditionally, DEA viewed IDA funds as balance
of payments support for the economy, and as resource support for the plan expenditures. The
"undisbursed" commitments - IDA funds committed by the Bank but not disbursed or tied up in
slow disbursing projects - were creating the impression among donors that India did not need the
money. Given India's poverty, this was a contradiction. In reality, large undisbursed balances
were an indication of weak institutional capacity to use these resources to promote growth. The
macroeconomic crisis of 1991 created an urgent need for increased resources wherever possible.
Education was considered a sector that could absorb large amounts of IDA funds. DOE, however,
continued to resist the overtures for IDA funding.
5. UNDP, 1997.
6. Education, under the constitution, is on the concurrent list. The major responsibility lies with the states.



25                                  ANNEX A
7.     Pressure from DEA notwithstanding, three factors helped change DOE's perspective.
First, the declaration of 1990 as International Literacy Year led to the establishment of the
National Literacy Mission by DOE and the launch of the Total Literacy Campaigns. These helped
focus attention not only on eradication of adult illiteracy but also on corrective measures such as
increasing enrollment and retention at the primary school level. Second, the Jom Tien conference
of 1990, held in Bangkok, formulated the international goals of Education For All (EFA), with an
emphasis on universal elementary education by the year 2000, and brought problems of basic
education into sharper focus. Since the Bank was a cosponsor of the conference, it also provided
an opportunity for furthering dialogue and understanding between the Bank and DOE. Finally, as
a consequence of the NPE, several centrally sponsored interventions were implemented aimed
specifically at improving the quality of teacher training, upgrading the facilities in primary
schools, improving curriculum content, innovations in teaching, and learning materials. The NPE
also made sweeping statements for across-the-board improvements in all sectors of education.
Implementation of these recommendations required, among other things, a substantial increase in
resources. In recognition that the educationally backward states needed broad-based interventions
if India was to achieve its development goals, EFA projects were envisaged for Bihar, Rajasthan,
Uttar Pradesh, Andhra Pradesh, and Madhya Pradesh. At this juncture, DOE began exploring
avenues for external funding for these projects. In a significant shift in fiscal allocation policy and
in an effort to make IDA funds attractive to the ministries, DEA allowed external funding as an
additionality to the departmental/ministerial budget.7
8.    India's efforts to implement large-scale primary education to that point had been slow.
After the 1990 Education for All Conference, where the low comparative indicators of India
became quite apparent, India's Secretary of Education agreed to borrow from the Bank for
primary education only after his staff had prepared the best project possible. For a year, the best
intellects of NCERT (National Council for Educational Research and Testing) and NIEPA
(National Institute of Educational Planning and Administration) worked on the proposal that was
to become the Uttar Pradesh District Primary Education Project. The proposal, prepared by
people inexperienced with Bank designs, needed improvement but was very substantial. High-
level officials stated that the government's goal was to learn as much as possible from the Bank
in order to design projects well and become independent of it. Some high-level officials read
material about the Bank's work in education, studied Bank documents extensively, and grilled
staff on details. They also demanded that the various studies needed for the sector be carried out
by Indians rather than by international consultants. Bank staff worked with their Indian
counterparts to rationalize the number of components of the proposed projects and carry out
beneficiary assessments, while supervising consultants inexperienced in Bank sector work, who
carried out various studies. To learn more from Bank staff, government counterpart staff
"shadowed" them. This sector work was eventually published as "India: Primary Education
Achievement and Challenges."'' The material has guided policy and implementation plans since
then.
7. This did not, however, translate into increased expenditures on education.
8. Bank Report no 15756-lN, September 1, 1996.



26                                   ANNEX A
Lending Activities (1972-2000)
9.     India's strong resistance to IDA funds in the social sectors before the 1990s restricted
IDA flows into education to a 1972 education project in the agricultural universities of Bihar and
Assam. The objective of the project was to integrate education, research, and extension through
improved facilities and faculty training in the two universities. The high quality of the physical
facilities provided by the project was considered a major achievement upon project completion
and something that would not have been possible without IDA assistance.
Vocational and Technical Education
10.    From 1962 to 1987, lending to population, health, and nutrition (PHN), and education
constituted less than I percent of the total Bank group lending to India. There was neither
dialogue nor lending in these areas other than for population. In the 1980s, the first indirect
opportunity for the Bank's entry into the education sector came with the Vocational Training
project ($137.5 million), which was negotiated with the Ministry of Labor. This paved the way
for Technician Education Project I ($223 million), which was approved in 1990 and covered eight
states. That project was followed in March 1991 by Technician Education Project II9 ($255.7
million), which was identical to the first project but extended coverage to an additional eight
states and the Union Territory of Delhi. The objective of these projects was to support the ten-
year Technician Education Investment Program envisaged in the NPE. The projects were
designed in accordance with the prevailing paradigm with a heavy focus on civil works, furniture
and equipment, which accounted for 66 percent of the total assistance.
11.    The Technician Education I project closed on September 30t, 1998. The project
Implementation Completion Report (ICR) and a subsequent audit by OED indicate that the
project objectives were relevant, clearly defined, and their direct linkage to national policies was
very strong. The achievement of most project objectives was found to be highly satisfactory with
some targets exceeded and others showing some shortcomings. The sustainability of the project
has been assessed as highly likely. The main gains from the project include modernization of the
laboratories and workshops, systematic faculty development, and interaction between industry
and the polytechnics. There are considerable achievements in increasing access for women and
the handicapped. However, not all states have performed satisfactorily, and the flexibility that
was to be introduced in the curriculum with the introduction of the multi-point credit system
happened only to a very limited extent. Discussions with state officials indicated that interaction
with industry has been achieved, to some extent, in urban areas that have a strong industrial
presence. It remains negligible in the institutes that are situated in smaller towns. The quality of
instructors still needs substantial improvement. A lack of clear-cut policy support within the
system is considered a major reason for the non-achievement in some areas in some states.'0
9. To deal with the magnitude of India, projects overlap; a group of states is included in one project, while another
group is included in a very similar project that starts one or two years later. This is different from the serial projects
usually carried out, where one finishes and another starts.
10. Report of the Expert Group on Policy Initiatives for Technician Education for Sustaining Technician Education
Project.



27                                ANNEX A
Uttar Pradesh Basic Education
12.   Continued dialogue with DOE led to the Bank's entry into basic education in 1993 with
the approval of the Uttar Pradesh Basic Education Project (UPBEP I, $165 million). During the
evolution of the Bank's thinking on IDA funds for basic education, the inputs were conceived to
be somewhat broader and more policy-based than those incorporated in UPBEP 1. The UPBEP
was modeled on the Bihar Education Project and Lok Jumbish and was in turn replicated by the
District Primary Education Project (discussed below). The Bank did push for, and succeeded in,
incorporating some measures, such as those for teacher empowerment through provision of
discretionary funds for teaching-learning materials, which have proven successful. The Bank also
required that baseline and follow-up surveys of learning achievement be conducted, an area
hitherto neglected. On the whole, the Bank's assistance strategy at the time of entry was relevant,
clearly directed, and well timed. The Bank's major thrust in basic education came at a time when
its own experiences and DOE policy had enabled both to develop a coherent sectoral policy with
a commonality of objectives.
13.   UPBEP I, designed as an EFA project, was launched in 10 of the 63 districts of Uttar
Pradesh, India's largest state with a population of 180 million. In addition to direct inputs, the
project also envisage'd capacity building and improvement in the quality of education. The
districts chosen for initial implementation had female literacy rates below the national average of
39 percent and 8 of the 10 districts selected had female literacy below the state average of 26
percent. The project has run parallel with the decentralization efforts that have gained momentum
nationally. Extensive learning assessment was established in the first projects. Village education
committees (reporting to democratically elected panchayats) were asked to carry out
microplanning, survey the village children, and map attendance. Most of the funding was to be
spent at the block and cluster levels; village committees contracted for school construction;
teachers could spend money on materials they needed; and in-service training was carried out at
the district level as well. Recurrent expenditures were provided for through state budgets, but
project funds were distributed to the districts through registered nongovernmental organizations.
Because the execution was spread so widely, there was almost no large-scale procurement. While
this may have restricted economies of scale, it may have also minimized possible corruption,
since large sums of money were not accessible to anyone.
District Primary Education Program
14.   Funding for the District Education Programs at the end of 1999 amounted to about $2
billion in commitments and expenditures and included five donors. To facilitate the complex
implementation and supervision, the government asked all donors to put money in the same
program conduit, but permitted various donors to finance activities in specific states. Donors are
asked to supervise the program twice a year in a joint mission. They form teams, and rotate
supervision responsibilities, supervising each other's financed areas. They try to visit all districts
every two years. The supervision needs and deadlines have been an important contribution to
planning and implementation. Government staff prepare for these for weeks in advance and drive
their subaltems and agencies working for them to carry out the work so that they can present
achievements against targets. The system has developed clear, target-oriented forms that
succinctly track progress towards objectives and are a best practice for other Bank operations. As



28                                  ANNEX A
a result of carrying out this complex supervision work, the task team received a recognition by
the Quality Assurance Group of the Bank during FY00.
Box 1. DPEP Supervision
Supervision of the District Primary Education Program in India of six operations (including DPEP I and
Uttar Pradesh BEP 11 reviewed by the RSA 3 panel) is quite astonishing. It shouldn't be possible to
supervise effectively a program in 150 districts involving a half-dozen donors and a $1 billion of Bank
credits, but it is being done remarkably well under difficult and varied circumstances. The approach centers
on joint supervision by donors and central government, a regular six-month schedule of missions, very
clear objectives for each mission, site visits to each participating state, explicit monitoring of progress
impacts and outcomes, regular oversight by the Bank's country office between formal missions. The M&E
system is excellent. Concerns for the development objectives and long-term sustainability are foremost in
the minds of the recent missions.
Source: Rapid supervision appraisal form. Ouality Assurance Group of the World Bank, January 2000.
15.    Due to the extremely high pent-up demand for schooling in Uttar Pradesh, UPBEP I was
quickly followed by UPBEP 11 (US$ 59.4m), whose objective was to consolidate and strengthen
the inputs of UPBEP I. Experience in UPBEP I led to the implementation of the District Primary
Education Project (DPEP) - an ambitious project taking to scale the content and objectives of the
UPBEP. The DPEP I ($260.3 million) was launched in 1994 to cover 42 districts in seven states.
Although initial funds for DPEP were committed through the Social Safety Net (SSN)
Adjustment Program, which was approved in 1993 to mitigate the risk to the social safety net
from the budgetary cuts imposed by fiscal adjustment, the DPEPs are multidonor projects with
IDA being the largest donor but involving the European Community, DFID, the Dutch
government and United Nations Children's Fund (UNICEF). The DPEP bureau in DOE
coordinates the aid inflows. All donors are represented on the supervision missions for the
project. DPEP is the most intensive effort to operationalize the strategy for decentralized planning
for Universal Elementary Education (UEE). The districts selected under DPEP had women's
literacy rates below national average. An exception was made for Kerala where a district was
selected for DPEP coverage with women's literacy rates well above the national average but
below the state average.
16.    DPEP has been expanded through two additional phases in 1996 (DPEP II, $425.2
million) and 1997 (DPEP III, $152 million). Two other expansions are under preparation. One
will extend coverage to the 48 additional districts in Uttar Pradesh, leaving only seven districts
that will be funded exclusively by the state government. The other, to be implemented in
Rajasthan, became effective in October 1999. Thus, from a negligible amount of lending, the
share of the social sectors rose to 20 percent by 1991. As of June 1998, social sectors accounted
for 29 percent of World Bank lending in India with education accounting for 11 percent of the
total lending.
17.   It is easy to both overestimate and underestimate the Bank's input into basic education in
India. Even though the Bank is the largest foreign donor, foreign funding to basic education in
India constitutes only about 4 percent of the total public expenditure on basic education.
However, the Bank has been a catalyst in making primary education a greater regional and local
priority. It is also significant as an enabling force for DOE to instill into the basic education



29                                                       ANNEX A
system, in a holistic manner, changes and innovations that it considered necessary but for which it
would have had difficulty allocating funds. In the case of UPBEP, it also allowed a region-
specific flow of funds, which would have been hard to justify politically for an internal allocation
of resources. Most important, it has shown the low-performing states that, given commitment, the
right inputs, and improved management, it is possible to significantly raise enrollment and
retention in schools. The strengthening and expansion of the Bank's country office in New Delhi
and the creation of the DPEP bureau in DOE has made closer supervision of projects possible,
which in turn has resulted in significant project achievements.
Annex A Table 1. The Lending Program in the Education Sector
Final         Credit         Disbursed            Cancelled or
Project name                FY    Creditl    Approval Closing          amount         ( million August undisbursed
Loan no. Year          Date          (S million)    1999)                (S million)
Completed Projects
Agricuftural Universities   1973  342         1973       12/31/82      12              12                  0
Technical Education l       1990 2130         1990       09/30/98       235            231.15              18.6
Technical Education II      1991  2223        1991       10/31/99       307.1          174.64              51.36
Vocational Training         1989  3045        1989       12/31/96       30             0                   30
Vocational Training         1989  2008        1989       12/31t98       250            121.42              116.14
Eiectronics Industry        1989 3093         1989       03/31/97       8              7.47                0.527
Development
Total                                                                   855                                534.67
Projects Under Implementatlon
Uttar Pradesh Basic         1993 25090        1993       09130/00       165            143.46
District Primary Ed         1995 26610        1995       03i31/02       260            122.31
District Prim Education II  1996  28760       1996       06/30/03       425.2          96.2
District Primary Ed. III (Bihar)  1998  30120  1998      09130t03       152            13.73
Rajasthan District Primary  1999  N0440       1999       12/31/04       85.7           0
Education
Uttar Pradesh Basic Ed II   1998  30130       1998       09/30/00       59.4           29.23
Andhra Pradesh Economic     2000              1999
Restruduring Program
Total                                                                   1.141.1        951.61               776.3
Total for all Lending



30                                          ANNEX A
Annex A Table 2. Comparative Literacy Rates
Country       Adult Literacy    Female Literacy   Male literacy
India         51.2              36.1              64.5
China         81                70.9              89.6
Pakistan      37                23.3              49
Bangladesh    38                24.3              48.4
Nepal         27                12.8              39.7
Indonesia     83                59                65
Malaysia      83                77.5              88.2
Philippines   94                80                75
Thailand      94                90.7              95.6
Source: UNDP. HDR 1997.
Annex A Table 3. Comparative Expenditures (1990)
Country          Real Per Capita GDP    Per Capita     Education
(PPP$)                GNP             (percent of GNP)
1994                  1994            1990
India            1348                   320            3.8
Indonesia        3740                   880            1.3
Sri Lanka        3277                   640            3.2
Philippines      2681                   950            2.4
China            2604                   530            2.6
Pakistan         2154                   430            2.7
Thailand         7104                   2410           3.8
Source: Human Devolopment ReporL 1997. UNDP.
Annex A Table 4. Comparative Expenditures (1995)
Country          Real Per Capita GDP    Per Capita      Education
(PPPS) -               GNP            (percent of GNP)
1994                  1994            1995
India            1348                   320            3.5
Indonesia        3740                   880            1.3-
Sri Lanka        3277                   640            3.1
Philippines      2681                   950            2.2
China            2604                   530            2.3
Pakistan         2154                   430            2.7-
Thailand         7104                   2410           4.2
SoLc: Hunan Deveopment Report 1997. UNDP.
The figures for expenditures on education as percentage of GNP are taken from the UNESCO World
Education Report., 1998 except those marked by an asterisk which are taken from the UNDP, HDR
and pertain to the year 1990.



31                                 ANNEX B
Summary of Discussion at CAE Workshop on Education
March 29, 2000
Technical and vocational education projects
*  Participants confirmed the substantial achievements in the four completed technical-
vocational education projects, which OED had rated satisfactory or highly satisfactory, but
considered that these institutions should focus more on industry linkages.
District Primary Education Projects
The discussions were focused primarily on the large, ongoing district primary education
projects (DPEPs).
The main observations and issues were:
* A very high level of government commitment at the central and state level. Donor staff whi
participated also expressed much praise for the eagerness of various officials to implement
the very large and complex district-based projects.
* Large enrollment increases in the most deprived areas, particularly of female enrollments.
Decreased dropout. Textbooks that are actually used in classes were provided on a large
scale. Participants agreed that there had been gains and improvements, but mentioned some
reservations on gender parity.
*  In-service training of about 100,000 teachers through block and cluster resource centers.
Some participants objected to taking teachers out of the class for this purpose and this
reducing the already small number of school days.
* Establishment of block and cluster resource centers as well as district institutes-of teacher
education. Instructional methodology was emphasized and child-centered methodologies
were disseminated to teachers and parents. Some participants stated that the trainers had vel
limited understanding of primary-level instruction, teachers' behaviors did not improve, and
the meaning of child-centered methodology was unclear and that everyone had his own
definition.
*  Large numbers of teacher appointments, particularly locally residing para-teachers. Some
participants were quite derogatory about these teachers, stating that the country gives the
cheapest education to its poorest children.
*  Formation of over 171,000 village education committees, some of which are very active.
Construction of thousands of schools, toilets, and drinking water fountains through local
panchayats.
Areas of concern for primary education included:
* Less impressive learning outcomes. Baseline to mid-term comparisons showed strongest
improvement for grade I and minimal improvement for the higher primary grades. Overall,
achievement in the poorer districts remained far below expected minimum standards.
Participants agreed.



32                                 ANNEX B
* Fewer teachers in areas where need is greatest; teachers may be reassigned near their homes,
leaving the areas of greatest need under served. Participants agreed.
* Uneven quality of in-service training and modest development of training institutions.
Uneven school construction quality, particularly in poorer areas and in the earlier project
stages. Some participants mentioned examples of inadequate schools.
* Limited function of village committee activities in the poorer areas. Some participants stated
that these committees depended on village relations, which might support special interests
and corruption.
* Questionable reliability of data collection and reporting. Some participants mentioned
examples.
* Government sensitivity regarding the unavoidable difficulties of this very impressive
program; limited access to donors of data and studies. The government spokesperson
disagreed with my objections and strongly disagreed that any data or studies were kept from
the donors. (She told me that she would give me everything available, but despite repeated
requests, I got nothing.)
* Inability of donors to visit project sites unannounced and observe operation as it happens
under everyday conditions. Participants agreed that it was important for the donors to see
sites unannounced; one participant said that Bank staff only saw "three to five star schools"
and did not have a good sense of reality. Government staff disputed this assertion.
* Rejection of donors' or other international technical expertise. Participants agreed that India
was ignoring important international lessons and processes in its projects. Government staff
stated that indeed they had the purpose of developing uniquely Indian methods with limited
outside influence.



33                                 ANNEX C
OED Responses to Selected Government Comments
Comments on the background paper from the Department of Elementary Education and
Literacy in the Ministry of Human Resource Development in the Government of India were taken
into account in this revision of the background paper. The purpose of this annex is to explain the
author's position on comments for which there was less than a full response in the main body of
the paper. The text follows the same order as the comments from the Ministry. It starts,
however, with a note on the issue of generalizations which was raised a number of times.
Generalizeability of various statements. Implementation results in India are not uniform; the areas
where problems have arisen and are discussed are often the poorest, and this is indicated in the
paper wherever needed. Sometimes government comments incorrectly attributed comments in
the paper as applicable to all of India.
Functioning of Village Education Committees (VECs) in poor areas (see paragraph 3.6). The
Government disagreed with the evaluation of VECs in this paper. The facts presented in main
body of the paper on VECs were limited to information from Uttar Pradesh. They were obtained
from a study of gender issues in Uttar Pradesh financed by the World Bank. In the view of the
author of this background paper the decentralization mandated by the 73rd and 74th amendments
to the Constitution is an important reason why extra attention and resources should be allocated to
the areas of particular poverty and illiteracy to ensure that VECs function as expected.
The Government cites the second in-depth review of DPEP, conducted by a third party, which
shows that increased participation in school management and school affairs is evident in almost
all the states. OED has had no access to this report and no means to ascertain the magnitude of
effect. The government also commented that DPEP has mobilized other grassroot-level bodies,
such as parents, teachers, mahila samakhya groups, that also play an active role in the proper
functioning of schools. The extent to which participation is consistent in the long term needs to
be better understood.
School construction quality (see paragraphs 3.10 and 3.23). The Government considered that
there was no basis for the criticism of school construction quality. The background paper
clarifies the prevalence of problems of low-literacy areas and adds a reference to an uncited
government-financed third-party study showing average to good construction quality. Again,
poor construction seems more likely in the areas where village education committees may also be
dormant. For example, the consultants who conducted the OED gender impact assessment study
found almost no functioning toilets in primary schools in the Uttar Pradesh districts they visited.
Teacher attendance and student-teacher ratios (see paragraph 3.8). The Govemment objected to
statements in the draft paper that there is a shortage of teachers and that teacher appointments are
motivated by factors other than educational improvement. Improvements are mentioned in the
main text. However the reported tendency to move teachers close to home through political
means may undermine the effort of the central government to improve teacher attendance .
Enrollment data reliability (see paragraphs 3.3, 3.12, 3.22, 3.24 and 3.25). The Government
objected to the criticism of enrollment data and education performance. The main body of the



34                                 ANNEX C
report did mention the sample taken by the government and more detail was added on learning
outcomes as reported in mid-term assessment studies. The paper also mentions the sources of all
documents indicating that enrollment data may be overestimated. The report echoes concerns
mentioned in other government documents, which are referred to in footnotes.
Availability of studies and research papers to donors and academics (see paragraph 3.26). The
main body of this paper echoes the frustration of many donor staff who try to get copies of studies
paid for by project funds and somehow cannot. OED was not able to get a list of studies, despite
promises that they would be supplied. Procedures regarding sharing this information certainly
exist, but their implementation is unclear. The Government maintained that studies and papers
were available.
Extent of Financial accountability (see paragraph 5.6). The Government disagreed with the
statement in the paper that rules to ensure financial accountability may not always be followed.
Clear government guidelines certainly exist and the Government has emphasized that State
societies have well-defined rules and regulations to ensure financial accountability. However, the
extent to which these rules are implemented, particularly at the lower levels, is unclear. Yearly
audits often find problems that are corrected, but at other times audits may merely check figures.
For this reason, OED undertook a study of the system (see separate background paper mentioned
in footnote 39) and found that financial accountability is weak. OED has been collecting several
such audit reports for a more detailed study.
Random choice of schools for supervision, unannounced visits (paragraph 3.30). The
Government has objected to unannounced random school visits by supervision missions.
Objection to this principle is highly unusual, and OED has not found another country that objects.
Findings from currently limited and pre-arranged supervision field visits cannot be generalized to
all project schools or conditions of daily operation.
Use ofpara teachers. (paragraph 3.8) The Government questioned OED's weak support for para
teachers. OED agrees that para teachers may be very effective. However, it also had to report
concerns of workshop participants about their quality (see Annex B).
Enrollment improvements and closing of gender and caste gaps (paragraph 3.11). The
Government objected to suggestion the SC and ST children do not receive quality education,
showed lower than average enrollments, and higher percentage dropouts. With the large
population and multiple studies, conclusions are difficult but evidence of problems is referred to
in the paper. Hopefully the year 2001 census data will show the extent to which education has
increased among the poorest, and different social groups.
Use of PROBE report (paragraph 3.22). The Government stressed that there was overlap between
coverage of the PROBE report and coverage of DPEP. Overlap between PROBE and DPEP
districts does exist, but is limited to only two districts and a slightly different time frame.
Therefore, the PROBE outcomes are of limited generalizeability to DPEP districts. Nevertheless,
the report has been used in the paper as a means of discussing the general state of government-
financed education.