Dropout and Retention Among School Children in the Context of the Mid-Day Meal Programme in India

Abstract

This study examines the effect of the world’s largest school-feeding programme, the Mid-Day Meal (MDM) programme, on dropout and retention among schoolchildren in India. Data from the India Human Development Survey (IHDS) Rounds 1 (2004–05) and 2 (2011–12) were used. The sample included individual-level information on schoolchildren aged 5–10 years in IHDS-1 who turned 12–17 years in IHDS-2. Bivariate and logistic regression analysis was used to examine school dropout and retention rates by MDM consumption patterns and factors associated with them among schoolchildren. The findings suggest that schoolchildren having MDM support in their younger years had a lower dropout rate. Participants who received early and persistent MDM support were more likely to be retained in school. Participants residing in urban areas or living farther from school were more likely to drop out, while those receiving scholarships were associated with higher retention in school.

JEL Codes: H52, I21, I22, I28, Q18

Keywords

School Feeding Mid-Day Meal Programme Schoolchildren Dropout Retention India

1. Introduction

Education is a cornerstone of societal progress, and ensuring universal access to quality education is a fundamental goal for nations around the globe (United Nations, 2015). Without large investments in human capital, no nation can see sustainable economic growth (Sridevi & Nagpal, 2019). The Indian government has implemented major programmes and policies to improve access to education, such as the Sarva Shiksha Abhiyan (SSA) in 2001 and the Right of Children to Free and Compulsory Education (RTE) in 2009. As a result, enrolment for the age group 6–14 has been 96 per cent or above since 2009, and this proportion increased to 96.9 per cent in 2016 (Annual Status of Education Report [ASER], 2017). These interventions targeting school-aged children have integrated children into the formal education system. However, achieving universal education in India is severely hampered by rising dropout rates and low retention rates among school-aged children (Kumar et al., 2023; Mehta, 2023a, 2023b; Paul et al., 2021).

The Unified District Information System for Education (UDISE, 2022) defines the dropout rate as the percentage of students from a cohort enrolled at a given level in a given school year who are no longer enrolled at any grade in the following school year. Retention rate is defined as the percentage of a cohort of pupils enrolled in the first grade of a given level of education in a given school year who are expected to reach the last grade of the level. Both dropout and retention rates are critical to achieving universalisation of school education. In 2005, in lower-middle income countries, around 99 million children of primary-school age were not enrolled, and of those enrolled, only 78 per cent completed primary school. The majority of children who failed to complete primary school were in Sub-Saharan Africa and South Asia (Grantham-McGregor et al., 2007). In India, in 2021–22, the dropout rates were 1.4 per cent at the primary level (Classes 1–5), 3.0 per cent in upper primary (Classes 6–8), and 12.6 per cent at the secondary level (Classes 9–10) (UDISE, 2022: Table 6.13). The retention rate was 95.4 per cent for primary (Classes 1–5), 81.2 per cent for elementary (Classes 1–8), and 64.7 per cent for secondary (Classes 1–10) schoolchildren (UDISE, 2022: Table 6.15).

A variety of interrelated factors may influence learning, grade progression, and ultimate dropout rates. As per the fundamental theory of human capital, individuals arrive at decisions about whether or not to continue their education by comparing the marginal expected costs of doing so with the marginal expected benefits of completing their education (Adelman & Szekely, 2016). Studies show that children engaged in paid or domestic work were more likely to remain absent or drop out of school (Agrawal, 2014; Cardoso & Verner, 2006; Goulart & Bedi, 2008; Kumar et al., 2023). Families do not send their children to school because their children are required to work to support the family, either at home or outside the house. In an effort to achieve universal primary education, the Government of India introduced the National Programme for Nutrition Support for Primary Education (NP-NSPE) in 1995 as a centrally sponsored scheme. On 28 November 2001, the Supreme Court directed all the state governments/union territories to implement the Mid-Day Meal (MDM) scheme, in which every child in every government and government-aided school was to be served a cooked meal with a minimum content of 300 kilocalories and 8–12 grams protein per day for a minimum of 200 days per year (Ministry of Education [MoE], n.d.). The majority of the Indian states began providing cooked and warm meals by 2003, and eventually, around 120 million students were covered under the MDM programme by 2006, which is now regarded as the world’s largest school feeding programme (Kingdon, 2007).

The fundamental aim of this programme is to increase school enrolment, retention, and attendance of children in India by providing free cooked meals for lunch on working days to children in primary and upper primary classes in government, government-aided, local body, and Education Guarantee Scheme (EGS) schools. Alternate innovative education centres, madrassas, and maqtabs supported under the Sarva Shiksha Abhiyan and the National Child Labour Project Scheme schools run by the Ministry of Labour and Employment are also covered by the MDM programme (MoE, n.d.). Studies have reported that MDMs have substantially contributed to overcoming classroom hunger and have been a huge help to low-income families, relieving them of the responsibility of providing one-time meals to their children (Garg & Mandal, 2013; Sharma & Saini, 2015). The MDM programme has also led to an increase in school participation rates from marginalised households (Ramachandran, 2019).

While studies have examined the enrolment benefits of the MDM programme (Jayaraman & Simroth, 2011; Saxena & Mittal, 1985), MDM’s longitudinal effects on children’s dropout and retention rates as pupils transition from MDM beneficiaries to non-beneficiaries and vice versa have not been investigated, especially at the national level. This study seeks to add to the existing research and contribute to the larger literature on school feeding, with a special focus on the effect of school feeding on dropout and retention. Employing a longitudinal dataset from Rounds 1 and 2 of the India Human Development Survey (IHDS-1 and IHDS-2), the study aimed to examine the prevalence of dropout and retention in schoolchildren aged 5–10 years in IHDS-1 and 12–17 years in IHDS-2 as they transition from MDM beneficiaries to non-beneficiaries and vice versa. The secondary aim was to determine the factors associated with dropout and retention among schoolchildren in India.

2. Ethics Statement

This study employed a publicly available longitudinal secondary dataset that had no information that may lead to the respondents’ identification. The Inter-University Consortium for Political and Social Research (ICPSR) data repository retains all the IHDS datasets that were utilised in this study (Desai & Vanneman, 2008, 2015).

3. Materials and Methods

3.1 Data Source

We utilised data from the IHDS, which is a collaborative programme by researchers from the National Council of Applied Economic Research, New Delhi, and the University of Maryland. This is a nationally representative, multi-topic survey. IHDS collects extensive information on socio-demographic characteristics, education, fertility, health, agriculture, energy use, and the utilisation of large public programmes like the Integrated Child Development Services and Public Distribution System at the national level.

Data for Round 1 IHDS-1 was collected in 2004–05 and for Round 2 IHDS-2 it was collected in 2011–12. The survey encompassed 41,554 households from 1,504 villages and 970 urban neighbourhoods in Round 1 (Desai et al., 2010) and 42,152 households from 1,503 villages and 971 urban neighbourhoods in Round 2 (Desai et al., 2015) across India. Around 83 per cent, that is, 34,621 households, were re-interviewed in Round 2 along with some split households residing in the same community.

3.2 Study Design

We utilised panel data from Round 1 (2004–05) and Round 2 (2011–12) of IHDS, and our final sample size was 9,818 children (5,068 boys and 4,750 girls). We restricted the data to individual-level information of children aged 5–10 years in IHDS-1, who then turned 12–17 years in IHDS-2 and were attending government, government-aided, EGS, or madrassa schools in IHDS-1. We included schoolchildren from these age groups because the MDM scheme covers primary schoolchildren (Classes 1–4), and since 2008, the programme covers all children studying in government, local body, and government-aided primary and upper primary schools and EGS schools or alternate innovative education centres including madrassas and maqtabs supported under SSA of all areas across the country (MoE, n.d.). Details of the final sample of re-interviewed respondents and the sample selection process are presented in Figure 1. See Desai and Vanneman (2008, 2015) for details on the sample of re-interviewed households.

Figure 1.

Source: Authors’ calculation based on data from IHDS-1 (2005) and IHDS-2 (2011).

3.3. Description of Variables

Data on whether any child or adolescent is enrolled in school and the respective year of schooling of the respondents was recorded in IHDS-1 and followed up in IHDS-2.

3.3.1 Outcome Variables

The first outcome variable was dropout, which is defined as a respondent aged between 5 and 10 years who was enrolled in primary school in IHDS-1 and dropped out of school in IHDS-2 when the respondent was aged between 12 and 17 years. The second outcome variable was retention, which was assessed using the year of schooling in IHDS-1 and IHDS-2. A child in a certain grade in IHDS-1 is expected to progress to a corresponding higher grade in IHDS-2. In the context of our data from IHDS-1 and IHDS-2, retention was defined as the percentage of children enrolled in a specific grade in primary school in IHDS-1 who continued their education and progressed to a higher grade in IHDS-2 (when the respondent was aged between 12 and 17 years).

Considering the 7-year gap between IHDS-1 and IHDS-2, the expected corresponding higher class for children in Grade 1 in IHDS-1 would be Grade 8 in IHDS-2; similarly, for children enrolled in Grades 2, 3, 4, and 5, it would be Grades 9, 10, 11, and 12, respectively. The participants were categorised into three groups according to their responses to the question on the year of schooling from IHDS-1 and IHDS-2; Group 1 (less-than-expected grade level in IHDS-2), Group 2 (promoted to expected or higher-than-expected grade in IHDS-2), and Group 3 (dropped out in IHDS-2).

3.3.2 Explanatory Variables

In IHDS-1, MDM consumption was defined as those who received grain, dalia, and/or a variety of meals. In IHDS-2, MDM consumption was defined as those who received school meals regularly or irregularly. Participants were categorised into four groups according to their response to the question on MDM consumption in IHDS-1 and IHDS-2: Group 1 (no MDM support in IHDS-1 and IHDS-2), Group 2 (MDM support in IHDS-1 only), Group 3 (MDM support in IHDS-2 only), and Group 4 (MDM support in both IHDS-1 and IHDS-2).

The socio-demographic characteristics of the sample are given in Table 1, including sex, household size (stratified as less than/equal to 4 members, 5–7 members, and greater than/equal to 8 members), education of parents/household adults (all are illiterate, at least one completed primary education, at least one completed secondary education, at least one completed higher education), place of residence (urban or rural), social group (high-caste Hindu, OBC, Dalit, Adivasi, Muslim, other religion), region (North, Central, East, Northeast, West, South), received scholarship, received free books, and distance to school (less than/equal to 1 km, 2–5 km, greater than/equal to 6 km).

Table 1.

Socio-demographic Characteristics of the Sample, IHDS-2 (2011–12)

Background Variables	Total (N = 9,818) % (N) IHDS-2 (2011–12)
Sex
Boys	51.3 (5,068)
Girls	48.7 (4,750)
Household size
Less than or equal to 4 members	27.5 (2,672)
5–7 members	53.0 (5,363)
Greater than or equal to 8 members	19.5 (1,783)
Asset group
Poorest	22.2 (1,741)
Second	23.5 (2,112)
Middle	28.7 (2,877)
Fourth	16.9 (1,945)
Richest	8.6 (1,140)
Education of adult members in the household
All are illiterate	20.8 (1,846)
At least one completed primary school	6.8 (691)
At least one completed secondary school	51.4 (5,089)
At least one completed higher education	21.0 (2,192)
Religion
High-caste Hindu	16.7 (1,778)
OBC	37.8 (3,384)
Dalit	25.1 (2,448)
Adivasi	7.0 (834)
Muslim	12.1 (1,217)
Other religion	1.3 (155)
Place of residence
Rural	83.9 (7,761)
Urban	16.1 (2,057)
Region
North	30.9 (3,368)
Central	8.2 (1,109)
East	26.1 (1,930)
Northeast	2.4 (315)
West	15.0 (1,341)
South	17.5 (1,755)
Received scholarship
No	67.2 (6,486)
Yes	32.8 (3,160)
Received free books
No	53.3 (5,018)
Yes	46.7 (4,744)
Distance to school
Less than or equal to 1 km	53.2 (5,133)
2–5 km	34.0 (3,327)
Greater than or equal to 6 km	12.8 (1,337)

Source: Authors’ computation based on data from IHDS-1 (2005) and IHDS-2 (2011).

Note: For some variables, the ‘N’ is not additive to the total ‘N’ because of missing cases.

Descriptive characteristics of the study sample are shown in Table 1. In this study sample of 9,818 respondents, 51 per cent (n = 5,068) of the study participants were boys and 48.7 per cent (n = 4,750) were girls. Around 53 per cent (n = 5,363) of the study participants had a household size of 5–7 members. Middle (28.7 per cent) and second (23.5 per cent) asset quintiles represented the largest proportions, while the richest quintile comprised only around 9 per cent of the sample. Around 51 per cent of the study participants had at least one parent or adult household member who had completed secondary-level education. The highest proportion of study participants were OBCs (37.8 per cent) and Dalits (25.1 per cent), followed by high-caste Hindus (16.7 per cent) and Muslims (12.1 per cent). A majority of the sample resided in rural areas (84 per cent; n = 7,761). The Northern region formed the bulk of the study sample, followed by the Eastern and Southern regions. Around 67 per cent of the participants did not receive a scholarship compared to 33 per cent who did. Approximately 53 per cent of the participants did not receive free books in school compared to 47 per cent who received free books. Over half the participants (53.2 per cent) lived within 1 km of their school, 34 per cent lived between 2 and 5 km away, and 13 per cent lived more than 6 km from school.

3.4 Statistical Analyses

Bivariate analyses were performed to estimate the prevalence of dropout and retention among children and adolescents aged 5–10 years in IHDS-1, who then turned 12–17 years in IHDS-2 by transition in MDM consumption as well as for the categories of independent variables. Two separate binary logistic regression models were then fitted to discern the extent to which MDM consumption pattern and socio-demographic factors explain the associations with dropout and retention. In the first regression model, the outcome variable was dropout and the model was controlled for MDM support in IHDS-1 and socio-demographic variables, whereas in the second regression model, the outcome variable was retention and the model was controlled for change in the MDM consumption pattern and socio-demographic variables. The equation for the logistic regression is as follows:

I n \frac{p_{i}}{(1 - p_{i})} = β_{0} + β_{1} x_{1} + β_{2} x_{2} + \dots β_{K} x_{k}

(1)

where P_i denotes the probability of the i-th observation belonging to the outcome category of interest, and β₀ … β_K are regression coefficients indicating the relative effect of a particular explanatory variable on the outcome variable. Further, individual weights were used to make the estimates nationally representative. The data analyses were performed using STATA version 15.0 (StataCorp, 2017).

4. Results

Around 17 per cent of the participants did not receive MDM support in IHDS-1 and IHDS-2. About 46 per cent of the participants received MDM support at an early age in IHDS-1, and around 7 per cent of the participants received MDM support at a later age in IHDS-2. Approximately 30 per cent of the participants had consistent MDM support throughout IHDS-1 and IHDS-2 (Figure 1).

4.1 Dropout Rates Among School-Aged Children

Table 2 compares the prevalence of dropout among school-aged children in IHDS-2 by MDM consumption status in IHDS-1 and place of residence in IHDS-2. The results indicate that MDM support plays a crucial role in reducing school dropout rates. In both rural and urban areas, children who had access to MDM in IHDS-1 showed lower dropout rates than those who did not receive MDM.

Table 2.

Prevalence of Dropout Rates among School-Aged Children by Place of Residence and MDM Consumption, IHDS-1 (2004–05) and IHDS-2 (2011–12)

Place of Residence in IHDS-2	MDM Consumption Status in IHDS-1	%
Place of Residence in IHDS-2	MDM Consumption Status in IHDS-1	Attending School in IHDS-2	Dropped Out of School in IHDS-2
Rural	MDM beneficiary in IHDS-1	98.6	1.4
Rural	MDM non-beneficiary in IHDS-1	94.9	5.1
Urban	MDM beneficiary in IHDS-1	96.5	3.5
Urban	MDM non-beneficiary in IHDS-1	91.3	8.7

Source: Authors’ computation based on data from IHDS-1 (2005) and IHDS-2 (2011).

Note: MDM = Mid-Day Meal.

The dropout rate for rural children with MDM support was just around 1 per cent compared to about 5 per cent for those without MDM. In urban areas, the dropout rate was approximately 3 per cent for MDM beneficiaries compared to 9 per cent for non-beneficiaries. Despite the lower dropout rates among MDM beneficiaries compared to non-beneficiaries in both settings, the overall dropout rates are higher in urban areas. It is important to note that establishing causal directions is challenging; low income may act as both a cause and a consequence of dropping out. The higher dropout rates in urban areas could reflect a higher proportion of urban poor population and additional challenges faced by children in these settings, such as instability from migration, increased work opportunities, social challenges, and differing perceptions about the value of education.

Table 3 presents the results of logistic regression analyses of the determinants of dropout among government schoolchildren from IHDS-2 (2011–12) after controlling for MDM support, sex, household size, education of adult members in the household, social group, place of residence, region, scholarship, free books, and distance to school. Participants who received MDM support at an early age in IHDS-1 had lower odds of dropping out of school in IHDS-2 compared to those who did not receive any MDM support in IHDS-1 (OR: 0.29; CI: 0.21–0.39).

Table 3.

Determinants of Dropout Among Government Schoolchildren (IHDS-2, 2011–12)

Independent Variables	Dropout in IHDS-2
Independent Variables	Odds Ratio	95% Confidence Interval
MDM support in IHDS-1
No MDM support in IHDS-1^Ref.
Received MDM support in IHDS-1	0.29	(0.21–0.39)
Sex
Boys^Ref.
Girls	1.04	(0.77–1.41)
Household size
Less than or equal to 4 members^Ref.
5–7 members	0.83	(0.59–1.17)
Greater than or equal to 8 members	0.88	(0.54–1.44)
Education of adult members in the household
All are illiterate^Ref.
At least one completed primary	1.11	(0.39–3.13)
At least one completed secondary	1.64	(0.90–2.99)
At least one completed higher	2.28	(1.21–4.30)
Religion
High-caste Hindu^Ref.
OBC	1.06	(0.70–1.61)
Dalit	0.94	(0.58–1.52)
Adivasi	0.35	(0.14–0.92)
Muslim	0.69	(0.37–1.27)
Other religion	0.95	(0.42–2.15)
Place of residence
Rural^Ref.
Urban	2.35	(1.62–3.41)
Region
North Ref.
Central	1.40	(0.70–2.8)
East	0.35	(0.22–0.57)
Northeast	0.27	(0.08–0.87)
West	0.80	(0.50–1.26)
South	1.00	(0.65–1.54)
Received scholarship
No^Ref.
Yes	0.81	(0.55–1.19)
Received free books
No^Ref.
Yes	0.10	(0.06–0.17)
Distance to school
Less than/equal to 1 km^Ref.
2–5 km	2.72	(1.73–4.26)
Greater than/equal to 6 km	16.72	(10.84–25.76)

Source: Authors’ computation based on data from IHDS-1 (2005) and IHDS-2 (2011).

Notes:

(a) Respondent was considered a dropout if the child was attending school in IHDS-1 and not attending school in IHDS-2.

(b) Ref. denotes reference category.

(c) Statistical significance denoted by boldface fonts are for p < .05, p < .01, and p < .001, respectively. A significance level of 0.05 indicates a 5 per cent risk of concluding that an association exists between the dependent and independent variables.

(d) 95% confidence interval is given in parentheses.

(e) MDM = Mid-Day Meal.

In terms of demographics, participants from households where at least one adult member had completed higher education were more likely to drop out than children from households where all adult members were illiterate (OR: 2.28; CI: 1.21–4.30). The educational background of at least one adult member in a household serves as a household-level indicator rather than reflecting parental education levels. This adult member may or may not be a parent of the participant. Thus, the impact of having one educated adult may be minimal, since their education may not necessarily influence the educational outcomes of children in the household. Participants belonging to the Adivasi social group were the least likely to drop out (OR: 0.35; CI: 0.14–0.92). The data also showed that Adivasi children were far less likely to enrol in school. This suggests that while Adivasi children face barriers to initial enrolment, those who do attend school may exhibit greater persistence in their education than their high-caste counterparts. Participants living in urban areas had higher odds of dropping out of school than their rural counterparts (OR: 2.35; CI: 1.62–3.41). Adolescents residing in the Eastern (OR: 0.35; CI: 0.22–0.57) and Northeastern (OR: 0.27; CI: 0.08–0.87) regions of India were less likely to drop out of school than their counterparts in the Northern region. Students who lived 2–5 km and 6 km or more away from the school had higher odds of dropping out than those who lived 1 km or less from the school. Before the RTE Act was implemented in India, children living in small communities often faced difficulties accessing schools within walking distance (1–3 km). To address this issue, small schools known as EGS centres, or alternative schools, were established under the Education Guarantee Scheme (Govinda & Bandyopadhyay, 2008).

Participants receiving free books were less likely to drop out of school than those who did not (OR: 0.10; CI: 0.06–0.17). According to India’s 2009 RTE Act, all children between the ages of 6 and 14 (Classes 1 through 8, or primary school) must receive free education, and thus, all children enrolled in Grades 1–8 receive free textbooks from the central and state governments.

4.2 Retention Rates

Figure 2 shows grade progression by MDM consumption status from IHDS-1 to IHDS-2. The five panels show grade progression by MDM consumption status, one for each grade cohort showing the MDM versus non-MDM distributions across grades. Considering the 7-year gap between IHDS-1 and IHDS-2, the expected corresponding higher grade for children in Grade 1 in IHDS-1 is Grade 8 in IHDS-2, and this pattern continues for subsequent grades. Of the 1,982 MDM beneficiaries studying in Grade 1 in IHDS-1, around 32 per cent progressed to Grade 8 in IHDS-2, 20 per cent to Grade 7, 17 per cent to Grade 6, and 7 per cent to Grade 5. In the case of MDM non-beneficiaries, of the 458 children studying in Grade 1, 22 per cent progressed to Grade 8, 21 per cent to Grade 7, 16 per cent to Grade 6, and 12 per cent to Grade 5. Among MDM beneficiaries studying in Grade 2 in IHDS-1, about 35 per cent progressed to Grade 9 in IHDS-2, whereas only around 27 per cent of MDM non-beneficiaries studying in Grade 2 in IHDS-1 progressed to Grade 9 in IHDS-2. Around 46 per cent MDM beneficiaries in Grade 5 reached Grade 12 compared to 42 per cent of MDM non-beneficiaries. This pattern persists for children studying in Grades 1, 2, and 5, where MDM beneficiaries consistently outpace non-beneficiaries in transitioning to the corresponding higher classes.

Figure 2.

Source: Authors’ calculation based on data from IHDS-1 (2005) and IHDS-2 (2011).

Table 4 shows the grade level achieved after 7 years by MDM beneficiaries and MDM non-beneficiaries. The percentages represent the distribution of students based on their grade in IHDS-1 and their subsequent grade in IHDS-2 categorised into three groups: students studying in a lower-than-expected grade in IHDS-2, those who were promoted to an expected or higher-than-expected grade in IHDS-2, and those who dropped out of school.

Table 4.

Grade-wise Progression and Schooling Status by MDM Consumption Status, IHDS-1 (2004-05) and IHDS-2 (2011-12)

Grade in IHDS-1 (2004–05)	MDM Consumption Status	Grade in IHDS-2 (2011–12) (%)
Grade in IHDS-1 (2004–05)	MDM Consumption Status	Lower-Than-Expected Grade Level	Promoted to Expected or Higher-Than-Expected Class in IHDS-2	Dropped Out of School in IHDS-2
Grade 1	MDM beneficiary in IHDS-1	55.4	44.4	0.2
Grade 1	MDM non-beneficiary in IHDS-1	59.6	39.6	0.7
Grade 2	MDM beneficiary in IHDS-1	47.1	52.9	0.0
Grade 2	MDM non-beneficiary in IHDS-1	48.1	50.2	1.8
Grade 3	MDM beneficiary in IHDS-1	42.0	57.6	0.4
Grade 3	MDM non-beneficiary in IHDS-1	36.5	61.1	2.4
Grade 4	MDM beneficiary in IHDS-1	57.2	41.1	1.7
Grade 4	MDM non-beneficiary in IHDS-1	52.2	44.1	3.8
Grade 5	MDM beneficiary in IHDS-1	48.8	42.7	8.5
Grade 5	MDM non-beneficiary in IHDS-1	48.3	40.2	11.4

Source: Authors’ computation based on data from IHDS-1 (2005) and IHDS-2 (2011).

Note: MDM = Mid-Day Meal.

Among MDM beneficiaries studying in Grade 1 in IHDS-1, about 44 per cent were promoted to an expected or higher grade in IHDS-2, while around 55 per cent were studying below the expected level. A small proportion (0.2 per cent) of MDM beneficiaries dropped out of school. In contrast, among MDM non-beneficiary students who were in Grade 1 in IHDS-1, about 40 per cent were promoted to an expected or higher grade in IHDS-2, and around 60 percent were studying below the expected level. Approximately 1 per cent of these students dropped out. Similarly, among MDM beneficiaries studying in Grade 2 in IHDS-1, about 53 per cent were promoted to an expected or higher grade in IHDS-2, whereas among MDM non-beneficiaries in Grade 2, about 50 per cent were promoted to an expected or higher grade in IHDS-2.

Overall, the table shows a consistent pattern for dropouts: MDM beneficiaries had a lower percentage of dropouts compared to non-beneficiaries. This trend remains consistent among children in Grades 1, 2, and 5, as MDM beneficiaries consistently outperform non-beneficiaries in progressing to the corresponding higher classes.

MDM non-beneficiaries in Grades 3 and 4 showed varied patterns. A higher percentage of MDM non-beneficiaries progressed to an expected or higher grade in IHDS-2 compared to MDM beneficiaries. About 58 per cent of MDM beneficiaries in Grade 3 progressed to an expected or higher grade in IHDS-2, compared to 61 per cent of non-beneficiaries. Similarly, around 41 per cent of MDM beneficiaries in Grade 4 progressed to an expected or higher grade, compared to 44 per cent of non-beneficiaries.

In Grade 5, approximately 43 per cent of MDM beneficiaries were promoted to an expected or higher grade in IHDS-2, while 40 per cent of non-beneficiaries were promoted to an expected or higher grade in IHDS-2. The dropout rate among MDM non-beneficiaries was also higher as compared to MDM beneficiaries.

However, the dropout rate notably increased to about 8 per cent among MDM beneficiaries and 11 per cent among non-beneficiaries. A considerable percentage of participants from this cohort dropped out of school. This increase in dropout highlights a critical educational challenge between Grades 5 and 10. According to IHDS-1 data, 50 per cent of males and 57 per cent of females who complete Grade 5 do not advance to Grade 10, with 34 out of 100 males and 29 out of 100 females discontinuing their education during this period.

The results reveal that MDM beneficiaries generally had better educational outcomes compared to non-beneficiaries, as evidenced by lower dropout rates and higher promotion percentages. The consistent pattern of lower dropout rates among MDM beneficiaries suggests that MDM support plays a crucial role in reducing school dropouts and improving academic progression. The stage between Grades 5 and 10 is critical, since it represents a transition point where a substantial percentage of students are at risk of discontinuing their education.

4.3 Effect of Changes in MDM Status

Children may have access to MDM support at different stages of their schooling. Table 5 presents the results of logistic regression analyses of the determinants of retention among government schoolchildren from IHDS-2 (2011–12) after controlling for change in the MDM consumption status and demographic variables.

Table 5.

Determinants of Retention among Government Schoolchildren from IHDS-2 (2011–12)

Independent Variables	Retention in IHDS-2
Independent Variables	Odds Ratio	95% Confidence Interval
Change in MDM consumption status
Group 1: No MDM support Ref.
Group 2: MDM support at IHDS-1 only	2.63	(1.34–5.19)
Group 3: MDM support at IHDS-2 only	0.42	(0.20–0.89)
Group 4: MDM support at both IHDS-1 and IHDS-2	1.86	(0.85–4.08)
Sex
Boys^Ref.
Girls	0.82	(0.49–1.39)
Household size
Less than or equal to 4 members^Ref.
5–7 members	1.19	(0.62–2.30)
Greater than or equal to 8 members	0.77	(0.36–1.61)
Education of adult members in the household
All are illiterate^Ref.
At least one completed primary	2.58	(0.71–9.34)
At least one completed secondary	2.31	(1.31–4.07)
At least one completed higher	2.37	(1.18–4.75)
Social group
High-caste Hindu^Ref.
OBC	0.31	(0.11–0.86)
Dalit	0.85	(0.26–2.80)
Adivasi	0.17	(0.06–0.53)
Muslim	0.25	(0.09–0.71)
Other religion	-	–
Place of residence
Rural^Ref.
Urban	0.69	(0.38–1.26)
Region
North Ref.
Central	6.12	(1.55–24.14)
East	6.94	(2.72–17.65)
Northeast	0.66	(0.30–1.42)
West	4.59	(1.41–14.94)
South	2.80	(1.27–6.19)
Received free books
NoRef.
Yes	1.03	(0.58–1.81)
Received scholarship
NoRef.
Yes	2.15	(1.09–4.26)
Distance to school
Less than or equal to 1 kmRef.
2–5 km	1.52	(0.85–2.70)
Greater than or equal to 6 km	1.15	(0.51–2.61)

Source: Authors’ computation based on data from IHDS-1 (2005) and IHDS-2 (2011).

Notes:

(a) Respondent was considered retained if the child was interviewed in IHDS-1 and IHDS-2 and was promoted to a higher class in IHDS-2 as compared to IHDS-1.

(b) Ref. denotes reference category.

(c) Statistical significance denoted by boldface fonts are for p value < .05, p value < .01, and p value < .001, respectively. A significance level of 0.05 indicates a 5 per cent risk of concluding that an association exists between the dependent and independent variables.

(d) 95% confidence interval is given in parentheses.

(e) MDM = Mid-Day Meal.

The results reveal that participants who received MDM support at an early age in IHDS-1 had higher odds of being retained in school compared with those who did not receive any MDM support (OR: 2.63; CI: 1.34–5.19). The positive influence of early MDM support on school retention indicates that providing nutritional assistance at the outset of a child’s education can establish a robust foundation for sustained academic engagement. In contrast, participants who received MDM support at a later age had lower odds of being retained in school than those who did not receive any MDM support (OR: 0.42; CI: 0.20–0.89). Participants aged 12–17 years are mainly enrolled in Grades 7 through 12, while MDM support is provided only up to the 8th grade. This discontinuation of MDM may account for the lower retention rates among those receiving MDM support at a later age in IHDS-2.

Adolescents from households where at least one adult member in had completed secondary education (OR: 2.31; CI: 1.31–4.07) or higher (OR: 2.37; CI: 1.18–4.75) were more likely to be retained than children from households where all adults were illiterate. The findings of this study present a contradiction: adolescents from households with at least one adult who has attained higher education are more likely to drop out of school (as observed in Table 3), yet they are also more likely to remain enrolled compared to children from households where all adults are illiterate. These findings highlight how the educational attainment of adult household members, along with their expectations and the socio-economic context, might collectively influence educational outcomes. The level of education among adults in the household can affect the pressures put on adolescents and the resources available to them, ultimately impacting whether those adolescents stay in school or drop out.

Adolescents belonging to OBC (OR: 0.31; CI: 0.11–0.86), Adivasi (OR: 0.17; CI: 0.06–0.53), or Muslim (OR: 0.25; CI: 0.09–0.71) social groups were less likely to be retained compared to high-caste Hindu children. Adolescents residing in the Central (OR: 6.12; CI: 1.55–24.14), Eastern (OR: 6.94; CI: 2.72–17.65), Western (OR: 4.59; CI: 1.41–14.94), and Southern (OR: 2.80; CI: 1.27–6.19) regions of India were more likely to be retained than their counterparts in the Northern region. Participants receiving a scholarship had higher odds of being retained than those who did not receive any scholarship (OR: 2.15; CI: 1.09–4.26). Scholarships provide financial support and incentives that can be crucial for continuing education, especially for children attending government schools.

5. Discussion

In this study, we assessed the impact of India’s school feeding programme on school dropout and retention. We found that students who received a MDM had lower odds of dropping out of school after 7 years than students who did not receive any MDM support, and MDM support is particularly effective when it begins in the early years of schooling.

This is in line with studies in India. The positive effect of early MDM support on school retention suggests that providing nutritional support early in a child’s education may create a strong foundation for ongoing academic engagement. Providing children with a hot cooked meal daily positively affects their retention rates, as demonstrated by various community-level studies in India that highlight the benefits of MDM consumption on retention (Agarwal et al., 2016; Laxmaiah et al., 1999; Nath & Nath, 2015; Saxena & Mittal, 1985). This finding is consistent with international studies that showed a reduced dropout rate among schoolchildren covered under SFPs (Ahmed, 2004; Tan et al., 1991; Wang et al., 2021; Zenebe et al., 2018). Studies have demonstrated that school meals improve enrolment and attendance, which in turn reduces dropout rates (Ahmed, 2004; Alderman et al., 2012; Buttenheim et al., 2011; Dreze & Goyal, 2003; Kazianga et al., 2012; Poppe et al., 2019; Tan et al., 1999; Vermeersch & Kremer, 2005).

However, our further analyses present a more nuanced picture of the effects of MDM support on dropout and retention. Participants who received MDM support at a later age had lower odds of being retained in school compared with those who did not receive any MDM support.

We extended our analysis to the effects of MDM on retention. Did MDM help students progress through school? Students in Grades 1, 2, and 5 who received MDM support were more likely to attain the expected grade level after 7 years; that is, they were in Grades 9, 10, and 11. However, this was less likely for students in Grades 3 and 4.

The results show that incentives are effective in retention. Adolescents receiving a scholarship had higher odds of being retained compared with those who did not receive any scholarship. Participants receiving free books were less likely to drop out of school compared to those who did not, which aligns with the existing literature that highlights the financial burden of textbooks as a barrier to education (Jhangiani & Jhangiani, 2017; Martin et al., 2017; Stein et al., 2017). Providing free textbooks in government schools is beneficial for the children because high textbook costs can discourage parents from sending their children to school, often pushing them to choose work over education. Adolescents receiving a scholarship were more likely to be retained in school compared with those who did not receive any scholarship. Scholarships provide financial support and incentives that can be crucial for continuing education, especially for families facing economic challenges. Student retention theories suggest that a student’s financial situation can impact their decision to stay in college (Tinto, 1988). Financial aid is often seen as a key factor in helping students stay in school and succeed.

The study found that demographics play a role in dropout and retention. Adolescents belonging to households where at least one adult member had completed higher education were more likely to drop out of school. Since these adult members may or may not be the respondents’ parents, the significance of this indicator in measuring educational persistence is limited. This counterintuitive result might also suggest that along with the educational attainment of the adult members in the household, parental participation at home and school is seen as a significant factor of education and development among children in both high-income and lower-middle income countries (Desforges & Abouchaar, 2003; Islam, 2017). The contribution of family members to the continuation and universalisation of primary education in India has received less attention from research (Paul et al., 2021).

Our finding that Adivasi students had a lower dropout rate than high-caste Hindu students contrasts with that of the existing literature, which often highlights higher dropout rates among marginalised communities. Data from IHDS-1 show that very few Adivasi children were enrolled in school. This suggests that, despite facing barriers to initial enrolment, Adivasi children who do attend school demonstrate greater persistence in their education compared to their high-caste peers. This result may be attributed to several factors, including community support for education and targeted interventions in certain regions.

The likelihood of dropping out of school was higher for those who lived in urban areas. Jayachandran (2007) found that a higher percentage of children in urban areas drop out of school to engage in economic activities to support their families. The percentage of girls who drop out to look after household responsibilities was higher than that of boys.

Adolescents in the Northeastern and Eastern regions of India were less likely to drop out of school. Literacy data from the 2011 Census shows substantial variation in literacy rates across the Northeast, from 67 per cent in Arunachal Pradesh to 93 per cent in Mizoram (Dikshit & Dikshit, 2014). Additionally, as highlighted by Marchang (2017), approximately 21 per cent of individuals aged over 7 years in the Northeast had completed matriculation but not graduation, which is lower than the 25 per cent national average. These figures suggest that the Northeast, despite having lower dropout rates, also has a distinctive educational profile with high literacy rates but a relatively smaller proportion of individuals with only matriculation. Nayak (2009) states that despite significant improvements in literacy, the Northeastern region was characterised by high rates of school dropout, inter-district differences, and rural–urban disparities. Although schoolchildren were less likely to drop out in the Northeast, a significant portion of the population still did not complete graduation. For Eastern regions, data from UDISE 2021–22 reveals that states like Bihar, Jharkhand, and Odisha had dropout rates of 5 per cent, 4 per cent, and 7 per cent, respectively, at the upper primary level, which are higher than the national average of 3 per cent. However, West Bengal stands out with a relatively low dropout rate compared to other Eastern states. The current finding suggests that while the Eastern and Northeastern regions show lower dropout rates overall, there are still states with higher dropout rates and challenges in educational attainment. Adolescents in the Central, Eastern, Western, and Southern regions of India are more likely to be retained in school compared with their counterparts living in the Northern region. Our finding aligns with the data from the UDISE report of 2021–22, which reveals that the Northern states of Uttar Pradesh, Jammu and Kashmir, Rajasthan, and Uttarakhand have retention rates lower than the national average, of approximately 95 per cent primary level.

The current study shows that participants living 2–5 km or 6 km or more from school were more likely to drop out compared to those living within 1 km. Long-distance travel to school negatively impacts attendance and can lead to higher dropout rates (Chugh, 2011; Huisman & Smits, 2009; Hunt, 2008). According to Muralidharan and Prakash (2017), the likelihood of enrolling in or finishing Grade 9 decreases steadily with increasing distance to the nearest secondary school, highlighting distance as a significant barrier to secondary school enrolment. A study by Duraisamy (2004) revealed that for every additional kilometre from a primary school, the likelihood of a girl attending school decreased by 2 per cent, while for boys, it decreased by 1 per cent. In response to these challenges, the RTE Act mandated that ‘neighbourhood schools’ be located within 1 km of walking distance for children in Grades 1 through 5 and within 3 km for children in Grades 6 through 8. However, while these standards provide a general guideline, careful school mapping and planning are necessary to ensure effective implementation (Govinda & Bandyopadhyay, 2008).

School-age children living in households where at least one adult member had completed higher education were more likely to be retained in school as compared to children who belonged to households where all adult members were illiterate. The findings of this study show an intriguing paradox: school-age children from households with at least one adult who has completed higher education are more likely to drop out of school (as observed in Table 3), while they are also more likely to be retained in school (as observed in Table 5) compared to children from households where all adults are illiterate. This counterintuitive result necessitates a deeper examination of the social and psychological dynamics at play. The retention rates for adolescents from educated households can be attributed to the supportive environment typically fostered in such contexts. When children observe their parents or other adult household members and school environment to be encouraging, they are more likely to stay in school (Dey, 2016). It helps create a foundation where children feel valued and motivated to continue their schooling, which can lead to higher retention rates.

Children belonging to OBC, Adivasi, and Muslim social groups are less likely to be retained in school compared to high-caste Hindu children. This result is consistent with the literature which shows that compared to other social groups, tribal students have a significantly higher school dropout rate (Chugh, 2011; Prakash et al., 2017; UNICEF, 2014). The Tribal Development Report, 2022, published in two volumes by the Bharat Rural Livelihood Foundation (BRLF) found that nearly half of all children enrolled in schools and belonging to tribal communities drop out even before they complete Grade 8 (Shah & Vijayshankar, 2022). This can be attributed to a variety of factors, including low parental involvement and support, inability to comprehend the medium of instruction, shortage of schools nearby, teacher postings and absenteeism, policy towards girls, parental education, social background, dependency ratios, employment opportunities, village development, quality of the local infrastructure, and the provision of MDM among others (Drèze & Kingdon, 2001; Joy & Srihari, 2014). Muslim participants were less likely to be retained in school than high-caste Hindu children. Despite a rise in institutions, enrolment, teachers, and physical facilities over time, the educational progress of Muslim students remains unsatisfactory in terms of literacy rates, enrolment, retention, and grade completion (Narula, 2014). It will therefore be necessary to develop and implement an educational programme that is ecologically valid, culturally sensitive, and economically feasible to ensure that tribal children participate in schooling.

The MDM programme has been very successful in reducing the number of dropouts; according to Drèze and Kingdon (2001), the percentage of girls who were not enrolled in school almost halved when the local schools offered MDMs. The MDMs have substantially contributed to overcoming classroom hunger and have been a huge help to low-income families, relieving them of the responsibility of providing a one-time meal to their children (Garg & Mandal, 2013; Sharma & Saini, 2015). Thus, the school feeding programme gives parents a compelling reason for enrolling their children in school, which in turn lowers dropout rates. Dietary interventions are more effective in childhood (ages 5–10) than in adolescence (ages 12–17) because there is opportunity for catch-up if the circumstances are improved, especially through early nutrition interventions.

Studies have demonstrated the long-term benefits of investing in early childhood education and nutrition, including increased school achievement, enhanced potential for employment, and better health outcomes in later life (Lundborg et al., 2022; Ruel & Hoddinott, 2008). Schools providing cooked meals are mostly government or government-aided schools where the cost of schooling is generally lower, which attracts children from the lower economic strata. Therefore, children from households with low incomes may not always prioritise attending school, particularly if they are required to work or help with household chores at home in order to support their family. Thus, school feeding can encourage children to consistently attend school and provide them a concrete incentive that they might not otherwise get at home.

In sum, this study has shown that MDM support was effective in reducing dropout rate and improving retention of children in school. Children consuming MDM at younger ages (5–10 years) were less likely to drop out of school. Adolescents in the age group 12–17 years residing in urban areas were more likely to drop out of school compared to their rural counterparts. Adolescents receiving free books were less likely to drop out of school. Adolescents living farther from school—specifically 2–5 km or 6 km or more—were at a higher risk of dropping out of school. Children consuming MDM at younger ages (5–10 years) were more likely to be retained in school. Adolescents receiving a scholarship were more likely to be retained in school. This is one of the few attempts, given the Indian context, at a thorough evaluation of a programme using a nationally representative dataset. These original findings, along with other studies on the positive effects of school meals on student enrolment, attendance, physical growth, and daily nutrient intake, provide empirical support for the benefits of the programme in India.

Although the results are mainly in line with those of other studies, the study is limited because we did not consider school-related factors, such as school infrastructure, availability and quality of teaching, adequate water and sanitation facilities in schools, maternal characteristics, and school environment. Since the roll-out of MDM is state dependent, focusing on state-level analysis rather than broad regions may provide more useful insights.

6. Conclusion

Dropout and retention in schools are multifaceted issues influenced by various socio-economic, cultural, and educational factors (Gouda & Sekher, 2014; Jayachandran, 2007; Kumar et al., 2017, 2023). By ensuring that students have access to healthy meals, school feeding programmes contribute to better physical and cognitive development, which can positively influence retention. To improve retention and reduce dropouts among school-aged children, policies should also focus on children living in urban areas, ensuring equitable access to schools and addressing transportation barriers, and continuing and expanding the provision of free textbooks and scholarships to alleviate financial constraints. By addressing these factors, policies can create a more supportive educational environment and improve overall student retention. Understanding the dynamics of these factors, particularly in the context of the MDM programme, becomes crucial for crafting effective policies and interventions. Most government-run or government-aided schools that serve cooked meals draw children from lower socio-economic classes because their schooling costs are often lower. There is extensive literature that children attending government schools and belonging to the lower socio-economic strata are more likely to drop out, and thus, for vulnerable sections of the country, a scheme like this can serve its purpose (Kumar et al., 2023). The estimates from these large datasets help policymakers in establishing priorities and assessing the degree to which operational targets are achieved, thereby fostering target-based decision making. The MDM programme can make extensive use of the data available to monitor each stage of the programme through a longitudinal comparison of the indicators. Only then will the effectiveness of the intervention be suitably demonstrated for beneficiaries of the programme.

These findings could be taken to support a broad focus by the government, thus providing a foundation for potential new policy recommendations, tackling the high prevalence of dropout, and implementing programmes in the early years to improve retention among schoolchildren. The MDM programme should be expanded to include schoolchildren beyond the 8th grade, thereby providing crucial nutritional support and incentivising continued education for adolescents. Additionally, the results of this study are likely to encourage additional research on urban children, school mapping, and provision of free textbooks and scholarships, thus giving way to sub-programmes focusing on the factors emerging from the current study. The broader relevance of these findings could help the central and state governments identify regional- and state-specific measures, thereby benefiting educational outcomes and future employment among Indian children and adolescents.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Adelman

M. A.

, & Szekely

(2016). School dropout in Central America: An overview of trends, causes, consequences, and promising, interventions (World Bank Policy Research Working Paper No. 7561).

Agarwal

, Bano

, Chopra

, & Jain

(2016). Status of implementation of mid-day meal program in Meerut: A cross sectional study. IOSR Journal of Dental and Medical Sciences, 15(12), 39–42.

Agrawal

(2014). Educational inequality in rural and urban India. International Journal of Educational Development, 34, 11–19.

Ahmed

A. U.

(2004). Impact of feeding children in school: Evidence from Bangladesh. International Food Policy Research Institute.

Alderman

, Gilligan

D. O.

, & Lehrer

(2012). The impact of food for education programs on school participation in Northern Uganda. Economic Development and Cultural Change, 61(1), 187–218.

ASER (2017). Annual status of education report (rural) 2016. ASER Centre. https://asercentre.org/aser-2016/

Buttenheim

, Alderman

, & Friedman

(2011). Impact evaluation of school feeding programs in Lao PDR (World Bank Policy Research Working Paper No. 5518).

Cardoso

A. R.

, & Verner

(2006). School drop-out and push-out factors in Brazil: The role of early parenthood, child labor, and poverty (World Bank Policy Research Working Paper No. 4178).

Chugh

(2011). Dropout in secondary education: A study of children living in slums of Delhi. National University of Educational Planning and Administration, Occasional Paper, No. 37.

10.

Desai

, Dubey

, Joshi

B. L.

, Sen

, Sharif

, & Vanneman

(2010). India Human Development Survey Users’ Guide Release 03. University of Maryland and National Council of Applied Economic Research. https://www.icpsr.umich.edu/icpsrweb/content/DSDR/idhs-data-guide.html

11.

Desai

, Dubey

, & Vanneman

(2015). India Human Development Survey-II Users’ Guide Release 01. University of Maryland and National Council of Applied Economic Research. https://www.icpsr.umich.edu/icpsrweb/content/DSDR/idhs-II-data-guide.html

12.

Desai

, & Vanneman

(2008). India Human Development Survey (IHDS), 2005: Version 12 [Data set]. Inter-University Consortium for Political and Social Research. https://www.icpsr.umich.edu/web/DSDR/studies/22626/versions/V12

13.

Desai

, & Vanneman

(2015). India Human Development Survey-II (IHDS-II), 2011–12: Version 6 [Data set]. Inter-University Consortium for Political and Social Research. https://www.icpsr.umich.edu/web/DSDR/studies/36151/versions/V6

14.

Desforges

, & Abouchaar

(2003). The impact of parental involvement, parental support and family education on pupil achievement and adjustment: A literature review (Vol. 433). DfES.

15.

Dey

D. K.

(2016). The nature and extent of parent’s attitude towards school dropouts. Imperial Journal of Interdisciplinary Research, 2(7), 1690–1695.

16.

Dikshit

K. R.

, & Dikshit

J. K.

(2014). Population of the north-eastern states of India. In North-East India: Land, people and economy (pp. 421–456). Springer.

17.

Dreze

, & Goyal

(2003). Future of mid-day meals. Economic and Political Weekly, 38(44), 4673–4683.

18.

Drèze

, & Kingdon

G. G.

(2001). School participation in rural India. Review of Development Economics, 5(1), 1–24.

19.

Duraisamy

(2004). Gender, intra-family allocation of resources and child schooling in Tamil Nadu, India. Journal of Educational Planning & Administration, 18(4), 489–506.

20.

Garg

, & Mandal

K. S.

(2013). Mid-day meal for the poor, privatised education for the non-poor. Economic and Political Weekly, 48(30), 155–163.

21.

Gouda

, & Sekher

T. V.

(2014). Factors leading to school dropouts in India: An analysis of National Family Health Survey-3 data. IOSR Journal of Research & Method in Education, 4(6), 75–83.

22.

Goulart

, & Bedi

A. S.

(2008). Child labour and educational success in Portugal. Economics of Education Review, 27(5), 575–587.

23.

Govinda

, & Bandyopadhyay

(2008). Access to elementary education in India (Country Analytical Review). National University of Educational Planning and Administration (NUEPA), New Delhi.

24.

Grantham-McGregor

, Cheung

Y. B.

, Cueto

, Glewwe

, Richter

, & Strupp

(2007). Developmental potential in the first 5 years for children in developing countries. The Lancet, 369(9555), 60–70.

25.

Huisman

A. H. M.

, & Smits

J. P. J. M.

(2009). Keeping children in school: Household and district-level determinants of school dropout in 363 districts of 30 developing countries. SAGE Open. https://doi.org/10.1177/2158244015609666

26.

Hunt

(2008). Dropping out from school: A cross country review of the literature. Create Pathways to Access Series, Research Monograph, No; 16.

27.

Islam

(2017). Parental involvement in education: Evidence from field experiments in developing countries. Discussion paper 02/17, Monash University, Department of Economics.

28.

Jayachandran

(2007). How high are dropout rates in India? Economic and Political Weekly, 42(11), 982.

29.

Jayaraman

, & Simroth

(2011). The impact of school lunches on primary school enrollment: Evidence from India’s midday meal scheme. The Scandinavian Journal of Economics, 117(4), 1176–1203.

30.

Jhangiani

R. S.

, & Jhangiani

(2017). Investigating the perceptions, use, and impact of open textbooks: A survey of post-secondary students in British Columbia. The International Review of Research in Open and Distributed Learning, 18(4), 172–192.

31.

Joy

, & Srihari

(2014). A case study on the school dropout scheduled tribal students of Wayanad District, Kerala. Research Journal of Educational Sciences, 2(3), 1–6.

32.

Kazianga

, De Walque

, & Alderman

(2012). Educational and child labour impacts of two food-for-education schemes: Evidence from a randomised trial in rural Burkina Faso. Journal of African Economies, 21(5), 723–760.

33.

Kingdon

G. G.

(2007). The progress of school education in India. Oxford Review of Economic Policy, 23(2), 168–195.

34.

Kumar

, Singh

A. J.

, & Handa

(2017). Literature survey on educational dropout prediction. International Journal of Education and Management Engineering, 7(2), 8.

35.

Kumar

, Patel

S. K.

, Debbarma

, & Saggurti

(2023). Determinants of school dropouts among adolescents: Evidence from a longitudinal study in India. PLoS One, 18(3), e0282468.

36.

Laxmaiah

, Rameshwar Sarma

K. V.

, Rao

D. H.

, Reddy

, Ravindranath

, Rao

M. V.

, & Vijayaraghavan

(1999). Impact of mid day meal program on educational and nutritional status of school children in Karnataka. Indian Pediatrics, 36(12), 1221–1228.

37.

Lundborg

, Rooth

D.-O.

, & Alex-Petersen

(2022). (2018). Long-term effects of childhood nutrition: Evidence from a school lunch reform. The Review of Economic Studies, 89(2), 876–908.

38.

Marchang

(2017). Out-migration from north eastern region to cities: Unemployment, employability and job aspiration. Journal of Economic & Social Development, 13(2), 43–53.

39.

Martin

M. T.

, Belikov

O. M.

, Hilton

, Wiley

, & Fischer

(2017). Analysis of student and faculty perceptions of textbook costs in higher education. Open Praxis, 9(1), 79–91.

40.

Mehta

C. A.

(2023a, August 4). Out-of-school children in the context of school education in India. Education for All in India. https://educationforallinindia.com/addressing-out-of-school-children-india/

41.

Mehta

C. A.

(2023b, September 10). Dropout rates in schools in India. Education for All in India. https://educationforallinindia.com/dropout-rates-in-schools-in-india/

42.

Ministry of Education. (n.d.). Pradhan Mantri Poshan Shakti Nirman—PM POSHAN. http://pmposhan-ams.education.gov.in/About_mdm.aspx

43.

Muralidharan

, & Prakash

(2017). Cycling to school: Increasing secondary school enrollment for girls in India. American Economic Journal: Applied Economics, 9(3), 321–350.

44.

Narula

(2014). Educational development of Muslim minority: With special reference to Muslim concentrated states of India. Journal of Education and Research, 4(1), 93–108.

45.

Nath

, & Nath

(2015). A study of the impact of Mid-Day-Meals programme on enrolment and retention of primary school children. International Journal of Applied Research, 1(10), 407–413.

46.

Nayak

(2009). Human development reports on north-east India: A bird’s eye view (MPRA Paper No. 17015). https://mpra.ub.uni-muenchen.de/17015/

47.

Paul

, Rashmi

, & Srivastava

(2021). Does lack of parental involvement affect school dropout among Indian adolescents? Evidence from a panel study. Plos One, 16(5), e0251520.

48.

Poppe

, Frölich

, & Haile

(2019). School meals and educational outcomes in rural Ethiopia. The Journal of Development Studies, 55(8), 1741–1756.

49.

Prakash

, Beattie

, Javalkar

, Bhattacharjee

, Ramanaik

, Thalinja

, & Isac

(2017). Correlates of school dropout and absenteeism among adolescent girls from marginalized community in north Karnataka, south India. Journal of Adolescence, 61, 64–76.

50.

Ramachandran

(2019). School mid-day meal programme in India: Past, present, and future. The Indian Journal of Pediatrics, 86(6), 542–547.

51.

Ruel

M. T.

, & Hoddinott

(2008). Investing in early childhood nutrition (IFPRI Policy Brief No. 8). International Food Policy Research Institute.

52.

Saxena

R. R.

, & Mittal

S. C.

(1985). Impact of mid-day meal programme on enrolment and retention at primary stage. Fourth Survey of Research in Education (1983–88), NCERT, 2, 1278.

53.

Shah

, & Vijayshankar

P. S.

(Eds.). (2022). Tribal development report: Livelihoods. Taylor & Francis.

54.

Sharma

M. L.

, & Saini

(2015). An evaluation study of mid-day meal programme in Jaipur. An International Journal of Applied Management & Technology, 2015, 133–144.

55.

Sridevi

K. V.

, & Nagpal

(2019). Trends in school dropout rate in India. Researchers’ Guild, 2(1), 2–24.

56.

StataCorp

L. P.

(2017). Stata statistical software: release 15.

57.

Stein

, Hart

, Keaney

, & White

(2017). Student views on the cost of and access to textbooks: An investigation at University of Otago (New Zealand). Open Praxis, 9(4), 403–419.

58.

Tan

J. P.

, Lane

, & Lassibille

(1999). Student outcomes in Philippine elementary schools: An evaluation of four experiments. The World Bank Economic Review, 13(3), 493–508.

59.

Tinto

(1988). Stages of student departure: Reflections on the longitudinal character of student leaving. The Journal of Higher Education, 59(4), 438–455.

60.

UNICEF. (2014). Global initiative on out-of-school children. UNICEF and UNESCO.

61.

Unified District Information System for Education (UDISE). (2022). Report on Unified District Information System for Education Plus (UDISE+) 2021–22 flash statistics. Department of School Education and Literacy, Ministry of Education, Government of India. https://dashboard.udiseplus.gov.in/#/reportDashboard/sReport

62.

United Nations. (2015). Sustainable Development Goal 4, Department of Economic and Social Affairs. https://sdgs.un.org/goals/goal4

63.

Vermeersch

, & Kremer

(2005). School meals, educational achievement, and school competition: Evidence from a randomized evaluation (Policy Research Working Paper No. 3523). Washington, DC: World Bank.

64.

Wang

, Shinde

, Young

, & Fawzi

W. W.

(2021). Impacts of school feeding on educational and health outcomes of school-age children and adolescents in low-and middle-income countries: A systematic review and meta-analysis. Journal of Global Health, 11, 04051.

65.

Zenebe

, Gebremedhin

, Henry

C. J.

, & Regassa

(2018). School feeding program has resulted in improved dietary diversity, nutritional status and class attendance of school children. Italian Journal of Pediatrics, 44, 1–7.