Jump to ContentJump to Main Navigation
Contested Politics of Educational Reform in IndiaAligning Opportunities with Interests$

Manisha Priyam

Print publication date: 2015

Print ISBN-13: 9780198098874

Published to Oxford Scholarship Online: March 2015

DOI: 10.1093/acprof:oso/9780198098874.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 24 February 2017

(p.262) Appendix

(p.262) Appendix

Research Methods

Source:
Contested Politics of Educational Reform in India
Publisher:
Oxford University Press

This research uses a variety of primary and secondary data sources in order to understand the complexities of implementation of educational reform in the Indian states of AP and Bihar. The primary data sources include interviews with education policymakers and policy implementers, teachers, and their unions; school and household surveys in selected sub-district units; and focus group discussions with SEC members on the working of educational decentralization. The secondary data sources include an analysis of national policies on education reform for nearly two decades (1990–2009); budgetary resources for education of the Centre and the states (from 1990 to 2007–8); and data on educational participation provided by the NSS (from 1993–4 to 2007–8).

The tools used are both qualitative and quantitative. Qualitative tools include interviews and focus group discussions. Quantitative tools include an analysis of budgetary data for school education, NSS data on educational participation, and school and household surveys in two selected districts in each state.

A: Key Informant Interviews—Policymakers, Reform Implementers, Teacher Union Leaders, Political Representatives

By far, the most important source of information was interviews with key informants, especially policymakers and stakeholders. Since the policy process in a federal context goes through various stages, from (p.263) setting out or ‘diagnosing’ the issues, debating ideas, formulating an agenda and authorizing policy to its implementation, and thereafter the consolidation of change, the number of actors are many and they work at various levels. Table A1 lists the actors interviewed at various levels and the themes on which information was obtained from each of these interviews.

Table A1 Interviews at Various Levels

Level

Actors/Stakeholders

Theme

National

Policymakers and experts.

Political economy context of reform policy design and implementation.

State

Policymakers and experts, including NGOs involved in educational reform.

  • Political economy context in the states.

  • Critical issues in school education policy in the states—policymaking and implementation challenges.

  • Policies for teacher management and school decentralization.

  • Attitude of political leaders to reforms.

  • Issues in implementation of reforms in the states.

Teacher union representatives.

  • History, internal structure of unions.

  • Teacher issues and grievances.

  • Attitude to education reforms.

  • Relationship with other unions.

  • Relationship with government.

  • Relationship with political parties and the political role of unions.

District

Reform implementation team.

  • Perception of key issues in the implementation of reforms in the district.

Teacher union representatives

Same as teacher union representatives at the state level.

Block/Mandal

Educational administrators—BEOs/MEOs.

  • Role perception in terms of the needs of school teaching.

  • Academic support provided by them to teachers and schools.

  • The frequency/adequacy of their presence in schools.

Office bearers of Panchayati Raj Institutions.

  • Their understanding of the key challenges before school education in their area.

  • Their role in the working of schools.

  • Their attitude towards school committees (interface of political decentralization with educational decentralization).

Mandal/block-level representatives of teacher unions if present.

A very thin presence of union representatives was found at this level.

Political representatives.

  • Association with schools—how often they visit schools?; and what they actually do for the schools?

  • Knowledge of school reform programmes.

  • Attitude towards school reform programmes, and working of school or village education committees.

  • Relationship with panchayats.

Village

Focus group discussions with community members (including parents, non-parents, socially disadvantaged castes, and women) and members and office bearers of school/village education committees.

  • Participatory nature of the committees, including process of elections and special representation for weaker social groups.

  • Functioning of committees—regularity of meetings, issues taken up, decision-making process.

  • Effectiveness of the committees in taking up school-related issues, including teacher supervision.

  • Relationship with the institutions of political decentralization.

  • Political interference in the working of committees.

Teachers in schools.

  • Challenges faced by them in teaching in poor rural areas.

  • Academic support given to them by the educational administration team.

  • Their expectations on the nature of academic support required by them to improve teaching.

Source: Author

(p.264)

(p.265) Further, the names of national policymakers, state-level policymakers, and implementers were available to me from my experience of work with the GoI’s DPEP during 1993–5, and from having served on various review missions of the GoI to the states implementing the reform programme. Field officers of the UNICEF, the officers of the DoE in the GoAP and GoB, and the MV Foundation in AP also provided me with the names of relevant persons to be interviewed in the two states.

B: Note on Analysis of Data on Educational Finance Data (Chapter 2)

The expenditure of the Centre and the states on education has been taken from GoI’s Analysis of Budgeted Expenditure on Education, 1990–1 to 2006–7.

Budget Classification of Expenditure on Education

The budget classifies education expenditure under two heads: ‘General Education’ and ‘Technical Education’. I have used the expenditure classified under ‘General Education’, and in the text, use of the term ‘education expenditure’ of the centre and the states refers to expenditure classified under ‘General Education’ and incurred by the DoE alone. There are further sub-categories of this expenditure, namely, expenditure on higher education, elementary education, adult education, language development, and others.

(p.266) Revenue and Capital Expenditure

Budgetary expenditure for each sector is classified into two broad categories: expenditure incurred on the revenue account; and expenditure incurred on the capital account. ‘Revenue Account’ disbursements show expenditure on salaries and allowances, contingencies, grants-in-aid, and maintenance outlays. ‘Capital Account’ disbursements show incremental capital expenditure on construction and equipment. Bashir (2000) notes that almost 99 per cent of the expenditure of the education department is incurred on the ‘Revenue Account’. But this does not mean that no investment expenditure is made in education. The capital investment in education mainly comes from other departments of the government, such as the Department for Rural Development (undertakes school construction under poverty alleviation and employment generation programmes). Also, some of the ‘Revenue Account’ expenditure such as grants-in-aid to local bodies allows investment expenditure to be incurred (Bashir 2000: 3). I have used only the expenditure incurred by DoE and not by other departments on education. I have, therefore, mainly analysed the revenue expenditure of the education department of the states and the centre.

Plan and Non-plan Expenditure

The Indian central and state budgets make a distinction between Plan and non-Plan expenditure. Mooij and Dev (2004: 99) estimate that nearly one-quarter of the aggregate government expenditure is on Plan heads and is ‘officially intended for new programmes and initiatives’. Further, they estimate that about three-quarter of the aggregate expenditure is for non-Plan heads, and this includes the recurrent costs for all government departments (mainly salaries), interest repayments, and subsidies.

C: Analysis of Data on School Participation, NSS, 1993–4 to 2011–12 (Chapter 3 and Conclusion)

  1. 1. To explain the changes in school participation over time (1990s to present), by gender, social groups, and economic status, I have used NSSO data from five rounds, which include three (p.267) ‘quinquennial’ rounds on ‘Employment and Unemployment Situation of India’ (50th, 55th, 61st, and 68th rounds) and the 64th round on education.

  2. 2. The year, sample size in terms of number of households, and persons for each round of survey are reported in Table A2.

  3. 3. Weights: The in-built weighting system, provided by the NSSO for every round, has been used for generation of estimated numbers.

  4. 4. Data for ‘New Bihar’: The state of Bihar was bifurcated in 2000 into Bihar and a new state of Jharkhand. To ensure the comparability of NSS data for 50th and 55th rounds (when Bihar had not been bifurcated) with that of the 61st round, all the data for the state

    Table A2 Sample Size and Estimated Number of Persons for National Survey Rounds, 1993–4 to 2011–12

    NSSO Round

    Year*

    Sample Size in Households

    Sample Size by Number of Persons

    Estimated Number of Persons**

    All India

    50

    1993–4

    115,354

    564,740

    778,297,682

    55

    1999–2000

    120,514

    596,281

    916,607,400

    61

    2004–5

    124,680

    602,833

    971,897,367

    64

    2007–8

    100,581

    445,960

    1,011,228,200

    68

    2011–12

    101,724

    456,999

    1,088,255,655

    Andhra Pradesh

    50

    1993–4

    8,552

    37,332

    62,143,594

    55

    1999–2000

    53,380

    73,139,000

    61

    2004–5

    8,428

    34,310

    71,894,223

    64

    2007–8

    6,971

    26,313

    73,884,700

    68

    2011–12

    6,898

    25,658

    79,930,528

    Bihar

    50

    1993–4

    7,468

    34,577

    51,939,329

    55

    1999–2000

    50,471

    69,849,050

    61

    2004–5

    5,754

    30,844

    71,069,853

    64

    2007–8

    7,031

    34,147

    75,951,500

    68

    2011–12

    4,581

    23,508

    89,824,603

    Source: NSSO, various rounds.

    Note: (*) NSSO years are July–June’; (**) Estimates generated by multiplying sample numbers with weights provided by NSSO.

    (p.268) of Bihar refer to ‘New Bihar’, and include NSS data for the region Jharkhand within the state of Bihar.

  5. 6. Quintiles of Monthly Per Capita Expenditure (MPCE): A household’s monthly expenditure per month is taken as a proxy of the living status of the household. Computed on a per person basis, the MPCE is the basis for official calculation of poverty in India. Quintiles based on MPCE for each round have been generated separately for each of the states; in this case, separately for the states of AP and Bihar. Within each state, rural, urban, and overall (rural + urban) quintiles have been generated separately. Quintiles are generated separately for rural and urban areas for the following reason: the monthly per capita expenditure levels in rural areas are generally much lower than in the urban areas. If quintiles were generated based on overall population, then lower quintiles would be highly skewed towards rural areas and higher quintiles would be biased towards urban areas. Moreover, living standards of rural and urban areas are also different.

  6. 7. The 55th round of NSS consists of two rounds—‘original’ and ‘revisit’. The revisit round consists of an additional sample size of 25 per cent of the original sample. Households covered in the original round of survey were not covered in the revisit; the additional 25 per cent were therefore new households, covered only in the revisit schedule. Basic household information such as expenditure, caste, religion, and type (of household) was not covered in the revisit round. For the household characteristics that are not available for the revisit round, columns for such information show up as ‘non-reported’.

  7. 8. For the 55th round, data analysed include original and revisit round, so we have a larger sample.

  8. 9. Lack of Sample and Missing Values: In case of lack of sample in the raw data for a certain variable, the corresponding space in the table (reporting for that variable) has been kept blank. In case of missing values for certain indictors, those records with missing values in the raw data have not been excluded both in the numerator as well as denominator while calculating ratios/percentages. This methodology basically assumes that the missing values are equally distributed across all (p.269) categories so as to avoid a bias in calculation. In most of the cases, missing values (if at all they are present) constitute less than 0.5 per cent of the total sample size for the particular state.

Definitions of Key Terms

  1. 1. Enrolment—Individuals reported by households as currently attending any educational institution and course of study are shown as ‘enrolled’ by the NSSO data. Official figures on enrolment, published by the government’s DoE, are collected on the basis of figures reported by the schools themselves. These figures are likely to have a possible bias in terms of reporting every name on the school register as ‘enrolled’. Very often therefore, the official gross enrolments exceed the child population for that age. Ghost reporting can be cross-checked by changing the source of information on enrolment from school to household reporting provided in the NSS. The exact question asked in the NSS survey is ‘current status in educational institution and course of study’. This question is asked for all persons below 30 years. Those who report ‘currently not attending’ have been further classified as: (i) ‘never attended’ and (ii) ‘ever attended but discontinued study’ (in the 55th round) and ‘ever attended but currently not attending’ (in the 61st round). These two classifications of not attending persons are not available for the 50th round.

  2. 2. Dropout—Individuals categorized as ‘ever attended but discontinued study’ (in the 55th round) and ‘ever attended but currently not attending’ (in the 61st round) have been considered as ‘dropout’ and the dropout ratio has been calculated as the ratio of the number of children who have dropped out to total number of children for the respective age groups being considered.

  3. 3. Never enrolled—Individuals categorized as ‘never attended’ have been considered as ‘never enrolled’ and the never enrolled ratio has been calculated as the ratio of number of never enrolled children to total number of children in the respective age groups. Accordingly, the total out-of-school children constitutes all those currently not attending, that is, (i) ‘never attended’ and (ii) ‘ever attended but discontinued’ taken together.

(p.270) D: Note on Selection of Field Sites for School and Household Survey, Interviews with Teachers and their Unions, and Focus Group Discussions on Educational Decentralization

Selection of Districts

First, a two-stage selection process was used in each state to identify the districts and then, within them, the mandals or blocks, as the sub-district units in AP and Bihar respectively are known. The criterion for selecting the districts, followed by the sub-district units, was that of contrasting educational outcomes. In the first stage, all districts in each state were ranked on the basis of literacy rates of the 2001 census (a robust indicator of educational outcomes) and three clusters were made (clusters of districts with above, close to, and below state average literacy rates). Thereafter, one district was selected from the top cluster with higher than state average of literacy rate, and the second was selected from the bottom cluster with poor educational indicators.

Districts East Godavari in AP and Gaya in Bihar were selected from districts with literacy rates above the state average; and Mahabubnagar in AP and Purnia in Bihar were selected from the lowest cluster of districts with literacy rates below the state average.

Selection of Blocks or Mandals

Selection of sub-district units (two in each of the selected districts) was the second stage of the selection process. In each district, two sub-district units were selected on the same principle of better and poor educational indicators. A total of four districts and eight blocks were selected in this way. In AP, the selected mandals were Jadcherla and Dharur in Mahabubnagar and Prathipadu and Razule in East Godavari. In Bihar, the selected blocks were Belaganj and Fatehpur in Gaya and Baysi and Kasba in Purnia.

Within each block or mandal, the following were undertaken:

  1. 1. Interviews with teachers and their union representatives (this forms the data for Chapter 4).

  2. 2. A school and household survey to compare the condition and working of schools, nature of educational participation, and (p.271) household involvement in schooling of children (this forms the data for Chapter 6).

  3. 3. Community responses on the nature of the local state and working of school decentralization were elicited through focus group discussions (this forms the data for Chapter 5).

School and Household Survey

While data from the NSS provided the research with robust indicators of household participation in schools, and of changes in educational participation over time, they were unable to shed light on the manner in which schools function in local settings. The sample survey for schools and households was undertaken by me with the objective of filling this gap.

For the survey, I had the assistance of a research team of five field researchers provided by the NCIS, New Delhi and Hyderabad. Permission of the concerned officers in the GoAP and the GoB was obtained prior to visiting the field sites in September–October 2005 in Bihar and February–March 2006 in AP.

  1. 1. Selection of schools: The schools were randomly selected for each sub-district unit from school lists (for primary school or elementary schools with a primary section) provided by the DPEP office. The total number of schools on the list was divided by the ‘quota’, that is, 10 schools for every mandal/block. The quotient ‘n’ was used for counting on the school list and schools that fell on the ‘n’th number count were selected; 10 per cent oversampling was done in case schools were found to be closed at the time of visit. A survey questionnaire was administered for every school which was answered by the headmaster or, in his absence, any other teacher. For every school, teachers were also interviewed separately on the challenges they faced with respect to teaching. A total number of 39 schools were surveyed in AP and 40 schools were surveyed in Bihar.

  2. 2. Selection of households: The households were randomly selected from the school records for primary school Grades I–V, 10 households for every school (making it a total of 100 households in each block/mandal, 200 households in every district, and 400 in each state). The random sample of households refers to those selected (p.272) randomly from the school register. Since data on ‘out-of-school’ children (dropouts or never enrolled children) were not available in the school, such households were traced by asking at the village level. The purposively selected sample includes households additionally identified in this manner. The total sample refers to both randomly selected and purposively included households.

    Table A3 Characteristics of Sample (Schools and Households), Field Survey, AP and Bihar

    Literacy Rate

    No. of Schools Surveyed

    Total No. of Teachers in Schools Surveyed

    Total Enrolment in Schools Surveyed

    No. of Households Surveyed in Mandal/Block

    Andhra Pradesh

    District 1: Mandals selected in East Godavari district

    Pratipadu

    49.02

    10

    29

    1,429

    100

    Razul

    80.10

    10

    32

    779

    99

    District Total

    65.48

    20

    61

    2,208

    199

    District 2: Mandals selected in Mahabubnagar district

    Dharur

    25.86

    9

    18

    503

    100

    Jadcherla

    58.29

    10

    34

    570

    98

    District Total

    44.41

    19

    52

    1,073

    198

    Andhra Pradesh Total

    60.49

    39

    113

    3,281

    397

    Bihar

    District 1: Blocks selected in Gaya district

    Belaganj

    52.94

    10

    36

    892

    100

    Fatehpur

    40.13

    10

    25

    1,336

    100

    District Total

    50.45

    20

    61

    2,228

    200

    District 2: Blocks selected in Purnia district

    Kasba

    35.13

    9

    30

    2,646

    100

    Baysi

    28.03

    11

    48

    1,858

    110

    District Total

    35.10

    20

    78

    4,504

    210

    Bihar Total

    47.00

    40

    139

    6,732

    410

    Sources: GoI (2001) for literacy rates (column1); School and household survey in selected field sites, AP and Bihar, 2005–6.

    (p.273) A total number of 397 households were surveyed in AP and 410 households were surveyed in Bihar.

  3. 3. While randomness was compromised to the small extent of purposive inclusion of households with children out of school, this measure enhanced the stated objectives of the survey of inclusion of the community’s voice on a range of issues of school participation and institutional functioning.

E: Note on Focus Group Discussions with Community Members and Members of SEC in AP and VSS in Bihar (Chapter 5)

Method: The villages, where focus group discussions (FGDs) were held, were selected randomly from amongst the villages where schools had been selected for the survey. The villages selected were ‘feeder’ villages to the schools selected (that is, each selected school was earmarked for serving a particular village and children from the particular village were expected to attend the earmarked school). I have detailed the procedure for selection of schools for the survey in section ‘D’.

The names of SEC/VSS chairperson/adhyaksha and other committee members were gathered from the school. Time and place of the FGD was publicized in advance. Women, members from marginalized castes and minority communities, and parents of non-school-going children were specially mobilized to attend. The FGDs were held in both states after school hours, but before it was dark in Bihar (late afternoon) as there was no electricity. In AP, the SEC members would often converge near the school even during school hours, while in Bihar, we had to persuade the members to come together and speak about issues related to their schools and children.

In AP, resource persons from the NCIS, Hyderabad, led by Sudha, facilitated the mobilization of villagers to attend the meeting and helped with the local Telugu language. I had learnt some key Telugu words and was able to understand most of the conversation. The discussions were recorded by me in English, and one team member in Telugu. In Bihar, the NCIS, New Delhi team, led by Shafique Ahmed and Pravin Ramteke, facilitated mobilization of VSS members, especially encouraging the purposive inclusion of members from the weaker castes and (p.274) communities and women. I was conversant with the local dialects in both Gaya and Purnia, therefore no language assistance was required.

The discussions were structured around the following issues:

  1. 1. The process of elections and nature of membership—whether the elections were free, fair, and regular, and whether women and weaker sections were adequately represented.

  2. 2. Regularity of meetings and the participatory process.

  3. 3. Issues or grievances raised in the meetings.

  4. 4. Whether the schools function regularly and have adequate resources.

  5. 5. Teacher punctuality and regularity, teacher’s interaction with community members.

  6. 6. Working of incentive scheme MDM—quality, regularity.

  7. 7. Schools committees and politics, especially the role of panchayats and teacher unions.

  8. 8. Other issues and voices of the community.

The discussions were recorded in writing and later cross-tabulated around these themes for analytical purposes. On this basis, a final report was generated for discussions with each community.

Table A4 Total Number of FGDs Conducted and Number of Persons Who Attended the FGDs

District/Block/Mandal

No. of FGDs

No. of Persons

District/Block/Mandal

No. of FGDs

No. of Persons

East Godavari

7

74

Gaya

7

141

Razule

4

37

Belaganj

4

70

Prathipadu

3

37

Fatehpur

3

71

Mahabubnagar

6

36

Purnia

6

216

Jadcherla

3

20

Baysi

3

142

Dharur

3

16

Kasba

3

74

Andhra Pradesh

13

110

Bihar

13

357