ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Gender Differentials in Education: Exploring the Capabilities Approach

Motivation and freedom of choice are mediated through the institutions of caste and religion, which restrict freedom, particularly of women. This paper discusses whether the capabilities approach provides any advantage in addressing questions of inequality that may be also mediated through such institutions. With the help of empirical data, the capabilities approach is used to identify "conversion factors" that are not typically addressed in the utility approach. A comparison is made between knowledge generated through the use of traditional data sources to measure access and returns to education with the knowledge about the dynamics of capabilities formation generated through the use of a mixture of data within the capabilities approach. This has the potential to help public policy decisions to improve education design and outcomes for girls from disadvantaged sections.


Gender Differentials in Education: Exploring the Capabilities Approach

Jeemol Unni

Motivation and freedom of choice are mediated through the institutions of caste and religion, which restrict freedom, particularly of women. This paper discusses whether the capabilities approach provides any advantage in addressing questions of inequality that may be also mediated through such institutions. With the help of empirical data, the capabilities approach is used to identify “conversion factors” that are not typically addressed in the utility approach. A comparison is made between knowledge generated through the use of traditional data sources to measure access and returns to education with the knowledge about the dynamics of capabilities formation generated through the use of a mixture of data within the capabilities approach. This has the potential to help public policy decisions to improve education design and outcomes for girls from disadvantaged sections.

This paper was presented at the annual conference of the Human Development and Capability Association on “Equality, Inclusion and Human Development”, New Delhi, 10-13 September 2008. The author acknowledges the comments of the participants at the conference, and G Omkarnath and Elizabeth Hill for reviewing an earlier draft of the paper.

Jeemol Unni ( is with the Gujarat Institute of Development Research, Ahmedabad.

he mainstream approach in economics to education is the human capital model. This is based on the concept of maximisation of the utility function. As opposed to this A martya Sen and Martha Nussbaum have proposed a capabilities approach to view various states of well-being, including access and use of education. Sen’s capabilities approach shifts the conceptual terrain from resources and opportunities as ends in themselves, to focus on the substantive freedom people have to convert resources into “valued doings and beings”.

In this paper, with the help of empirical data, we use the capabilities approach to identify “conversion factors” that are not typically addressed in the utility approach. We juxtapose the two app roaches to examine how institutions such as caste and religion mediate access and returns to education of men and women. The effort is to discuss whether, the capabilities approach provides any advantage in addressing questions of inequity that may be mediated through such institutions. The main innovation in this paper is a comparison between knowledge generated through the use of traditional data sources to measure access and returns to education with the knowledge about the dynamics of capability formation generated through a mixture of traditional, quantitative and some qualitative data within the capabilities approach.

1 Utility Approach versus the Capabilities Approach

The term utility is used to mean different things and loosely implies anything having value. Utility has been viewed as “happiness” as well as “desire-fulfilment” and both can be taken as guides of a person’s well-being. Sen (1987) took exception to the view of utility as a real-valued (i e, numerical) representation of choice. He questioned this approach of the binary relation of choice reflecting the person’s well-being: “an approach that goes back to the origin of the ‘revealed preference’ school”. This approach is deficient in analysing the motivation behind a person’s choice. It makes the heroic assumption that the binary relation underlying choice must be the ordering of the person’s own well-being.

Sen (1983) introduced the concept of a person’s capabilities to be and to do things of intrinsic worth, i e, resources adequate to achieve a specified set of functionings. Functionings or “states of being” could range from elementary states (being nourished, going to school) to complex personal states and activities (participation/appearing without shame). Sen’s capability approach emphasised that what mattered to people was “the actual living that people manage to achieve”. He argued that “when we make

Economic & Political Weekly

february 28, 2009 vol xliv no 9


interpersonal comparisons of well-being we should find a measure which incorporates references to functionings, that is what is achieved, but also reflects the intuition that what matters is not merely achieving the functioning but being free to achieve it” (Unterhalter 2003).

He favoured the capability to function as the criterion for assessing the standard of living, and by implication poverty, rather than the utility that might be derived from using that capability. Thus, Sen eschewed the welfarist approach to poverty with its underlying assumption that the evaluative criterion is the utility that people derive from goods and services. However, he neither offered a practical criterion for evaluating the various capabilities to function nor sought any aggregation of the social values of the separate capabilities (Kingdon and Knight 2003).

Sen shifts the argument from resources and opportunities to achieve functionings, e g, going to school, to the need to look at the freedom people have to formulate capabilities “valued doings and beings” and thus convert resources into functionings they value. “Thus agency and freedom to make up one’s mind about schooling as a valued end and convert one’s aspirations regarding schooling into valued achievements lie at the heart of Sen’s capability approach and distinguish it from other positions” (Unterhalter 2003).

The capability approach to education requires us to think about gendered constraints on functionings and freedoms in educational institutions such as schools. Further, even when girls complete 12 years of schooling they encounter problems of discrimination in the food they eat, in access to employment and at the workplace, and their contribution to family decision-making or public debate. Measures such as “years of schooling” used extensively in human capital theory by no means capture these values. What if women do not claim education because they reflect what the society expects of them? The approach encourages “an examination of the ways in which gendered institutions prevent women converting capabilities into functionings”.

While it is easy to point out the difficulties with the traditional measures of education, it is difficult to visualise a measure using the capabilities approach. As we know in most multi-ethnic developing societies, certain groups and women are silenced. We attempt to identify “conversion factors”, e g, institutions of caste and religion that may prevent women from converting resources to “functioning”, in this case education. Further, the intrinsic value of education could be measured as access to media, newspapers, representation in local governments, participation in discussions, freedom from discrimination, etc.

2 Human Capital Approach

The human capital approach has at its core the maximisation of utility functions. This is a very popular approach to analysing access and returns to schooling. Traditionally some of the questions asked through this approach are who is schooled and what are the returns to education? Also through inclusion of independent variables in the participation functions it seeks to ask the question whether institutions such as caste and religion affect those who obtain education, and with separate earnings functions whether there are differences in the returns to education by gender and by these institutions.


The determinants of schooling analysed through this approach begin with a utility function where parents are assumed to maximise utility subject to a budget constraint. A demand for schooling function is derived from this maximisation of utility function, which depends on the price or cost of schooling, income and a set of control variables – mainly household, child and parent’s characteristics. A probability function of enrolment of a child in school and a grade attainment function help to understand the determinant of schooling. Using this method for Tamil Nadu, Doraisamy (2001) found that parent’s education and the household’s prosperity emerge as important determinants. The availability of schools and higher adult literacy in the village show better schooling outcomes. A girl’s participation in schooling is negatively affected by the presence of young siblings, and the number cattle in the household, while having an educated mother improves her chances. Teacher training had an impact on grade attainment of students.

Juxtaposing the choice among children of school enrolment and work participation Doraisamy (2007) engaged a similar utility maximisation approach and calculated the maximum likelihood probit estimates. While boy’s education was favoured, the parent’s education and household income had positive impact on enrolment over work participation. She concludes that the greater the resources of the household, the higher the demand for children’s schooling and less the need for child work to supplement incomes. The influence of institutions like caste and religion is investigated through dummy variables, with children from scheduled castes (SCs) and scheduled tribes (STs) less likely to attend school. The religion dummy indicates a uniform negative effect for child schooling compared to the reference category of other religions.

The major limitation of this method is that it assumes that children have choice. The motivation behind the obtained outcome is not investigated, except to say that certain characteristics of parent’s and the household can affect the outcome. The actual perception of the child or his/her value for education and the motivation for why SC/ST children are less likely to attend school are not studied in this method.

Another question that the economics of education literature using the utility approach asks is what are the returns to education? The traditional literature has observed a decreasing r eturns to education and higher returns to female education (Psacharapoulos G 1994). This has been used to justify policies for higher investments in elementary schooling and women’s e ducation. However, similar literature in India noted increasing returns to education or higher level of education had higher r eturns (Unni 1996; Kingdon and Unni 2001). The returns to f emale e ducation have been found to be higher than male e ducation in India as well.

Efforts to study the impact of institutions of caste and religion on returns to education using a similar methodology (Unni 2007) found that returns were higher for the SC and ST population c ompared to the Muslims. This was attributed to the fact that a ffirmative action in India has favoured jobs for the SC/ST p opulation, while the Muslim population did not benefit from any such backing.

february 28, 2009 vol xliv no 9


While these results appear intuitively correct, it fails to answer the question as to why such high returns do not encourage higher female participation in schooling and also greater participation by the SC and ST population, as we shall see later. Again the critique remains the same as for child schooling. That is, the utility approach concentrates on the outcome of education subject to some constraints, but does not investigate the motivation for or whether the person actually has the ability to choose. What are the institutional constraints placed on women and on particular caste and religious groups that prevent them from accessing education and jobs with better returns?

In a recent paper Kingdon and Theopald (2008) have used the human capital utility approach further to ask the question whether the demand for schooling actually depends on economic returns to education in the local labour market. It is argued that economic returns can have a positive substitution effect, that is higher return would mean greater participation in schooling. However, economic returns to education could have a negative income effect due to higher opportunity cost of labour in the current period. That is, household may not be willing to forego the extra incomes from labour of children for a longer term higher returns to education. In India, they noted that higher economic returns had a positive substitution effect for girls and, while the negative income effect was stronger for boys in poorer households. This they attributed to the fact that girl’s education was substituted by household work, while boys participated in economic activity and earned supplementary incomes for the household.

This result does not appear intuitively correct, nor is it empirically substantiated by the data on school attendance. The higher female returns to schooling should have encouraged greater school attendance among girls, but instead it is the boys who have higher attendance rates. In 2004-05 the attendance of boys in the age groups 5 to 14 years was 84.7% while for girls it was 79.2% (NSSO 2006). Similarly the net attendance ratios for 2005-06 in primary school (standards 1 to 5) was 73.2% for boys and 70.5% for girls, while it was 56.8% for middle, secondary and higher secondary school for boys and 45.6 for girls (NFHS-3 2007).

While the utility approach does provide some broad insights and direction of relationships of social and economic constraints on education, it is deficient in explanations for the motivation of the results obtained. It might also lead to wrong diagnosis for policy without an understanding of what drives the relationships observed.

3 Capabilities Approach to Education

What is it that the capabilities approach allows us to ask or analyse that gives us a better understanding of the economic and social dimension of education? In the utility approach it is assumed that people are free to choose, given the constraints of budget. The capabilities approach discusses the ability of the person to choose what he or she values, that is, what is the motivation behind this choice. The argument in the latter approach is shifted from resources and opportunities to acquire education to discussing the motivation for the choice or the freedom to formulate the capabilities. That is, the focus

Economic & Political Weekly

february 28, 2009 vol xliv no 9

shifts to how to convert resources into functionings (outcomes) that people value.

The starting point for the capabilities approach is a focus on how the social context sets the conditions for individual freedoms. In the case of education how free are men and women, boys and girls to access education of their choice given the institutional dominance of caste and religion in India? The advantage of the capabilities approach is that it invites a wider range of more searching questions with regard to gender equality than just a focus on outcomes.

One of the challenges to the capabilities approach is to devise and measure “conversion factors”. What variable can define the institutional dominance of caste/religion in the motivation of boys and girls to pursue education? A related issue noted by Sen is the type of data used for assessing well-being, in this case e ducation. He suggests three types of data: market purchase data, responses to questionnaire and non-market observations of personal states (Sen 1987). The utility approach uses market p urchase data for a bundle of goods and services that is bought and sold in the market, including education. Non-market items such as fertility, or the choice of having children, are modelled using the non-market observation of personal states, e g, number of children.

The third method of responses to questionnaires has been viewed with mistrust in economic literature on the ground that such information is subjective. However, this method has been increasingly used in recent years even in the utility approach (for instance, Kingdon and Knight (2003) use it to measure well-being). Non-market state and the questionnaire method are useful types of data to construct variables to explain the motivations behind value of education. In the rest of the paper we shall attempt to garner some evidence using mainly these two types of data to better understand whether gender differentials in c apabilities in education are mediated through institutions of caste and religion.

4 Indicators of Institutional Mediation Behind Low Value Placed on Education

While there could be many reasons for the low value placed on accessing education and the gender bias in it, in spite of high economic returns, in this paper we focus only on the institution of religion and caste that operate to restrict freedoms of individuals to choose. The society in India is stratified by social groups in terms of caste and religion. This system is hierarchical and certain historically vulnerable groups have remained at the bottom of the economic and social pyramid.

In an apparently liberal and democratic society the access to education and capacity to stay within the formal educational system to acquire higher levels of education are limited for the SC/ST and Muslim communities. There is systemic curtailment of freedom by agencies operating on these communities to choose a way out of poverty through education. Besides belonging to the lower castes, the choice of education is further restricted by gender and location. Being a girl/woman living in rural areas as well as belonging to the SC/ST community ensures that she is triply crippled. While only 15% of boys of the age group of 5 to 15 years

0.2 0.1 88.8 4.4 0.4 31.4 60.2 59.2 29.7 0.4 0.5


Figure 1: Share of Boys and Girls in Principal Work, Attending Education and Domestic Duties (Usual Principal Status) by Age Group, Location and Gender (2004-05)

Figure 1.1 (Rural Boys) Figure 1.3 (Rural Girls)
90.0 90.0
80.0 80.0
0.2 87.0 80.3 5.4 0.2 0.6 45.3 47.1 0.9 60.2 30.3 0.5
0.1 76.9 77.0 10.8 0.6 4.9 22.2 33.0 41.4 27.2 39.7 24.0

% to total population % to total population



% to total population

% to total population



40.0 40.0




5-9 10.-14 15-19 5 and above


Age Group Figure 1.2 (Urban Boys) Figure 1.4 (Urban Girls)






5-9 10.-14 15-19 5 and above

Age Group

70.0 70.0

60.0 60.0

50.0 50.0

40.0 40.0

30.0 30.0

20.0 20.0

10.0 10.0


5-9 10.-14 15-19 5 and above

87.6 88.3 57.7 49.8 27.4 14.7 29.7 9.2 5.5 2.4 0.4 0.1

5-9 10.-14 15-19 5 and above

Age Group

Computed using NSS, unit level data 61st Round on Employment and Unemployment, 2004-05

Age Group

and adjusted for census population .Source: NCEUS 2007, Figure 6.3.

to jobs with lower earnings and status. The number of years of schooling that the men and women workers in these groups were able to access reduced systematically from 5.4 years for men to

1.2 for women located in rural areas and belonging to SCs and STs. Compared to men the choice of having a good education gets progressively worse for women in the unorganised sector, as they carry the triple burden of gender, rural location and belonging to the SC/ST community.

Using the capabilities approach we look for “conversion factors” or variables and measures for the motivation behind two questions: (a) why the poor school attendance? (b) why the poor educational attainments?

4.1 Reasons for Poor School Attendance

School Infrastructure: While most large villages in India have a primary school, there are many biases in the development of

Principal worker.

Attending Educational Institution. Attending Domestic Duty, etc.

school located at a distance from the village and this is a gross curtailment of freedom of persons belonging to this community. The Sachar Committee Report further showed that the availability of roads leading to the village also declined as the village size declined and the share of Muslim population increased. Lack of infrastructure restricts the freedom of choice of parents to send their children to school and acts as a negative “conversion factor”.

Domestic Work and Child Labour: It is well known that young girls are engaged in helping their households with domestic duties and the care of siblings. In 2004-05, while nearly 40% (10.8%) and 50% (5.5%) of girls in rural and urban areas, respectively, in the age group of 5 to 19 years (10 to 14 years) were engaged in domestic duties as their major activity, hardly any boys were so engaged (Figure 1). This definitely curtailed the freedom of girls to attend school. The boys, more than 50%, if not attending school, were more likely engaged in economic activities

Table 1: Education and Work by Gender, Location and Social Group (2004-05)

Indicator Male Female Rural Female Rural Female SC/ST

Share of out of school children in 5 to 14 age group 15.5 20.7 23.5 28.5

Mean years of schooling (all workers) 5.4 2.5 1.9 1.2

Mean years of schooling (unorganised non-agr workers) 6.1 3.7 2.9 2.0

Share of all workers up to primary education (including illiterates) 55.6 80.7 85.0 90.7

Source: NCEUS 2007, Table 5.1.

were out of school, as for girls it was nearly 21%. As though the lack of freedom to choose education if one was a girl was not enough, girls living in rural areas (24%) and belonging to SC/ST households (28.5% were out of school) had even more restricted choices (Table 1). This discrimination at an early age then b ecomes a handicap when they enter the labour market, leading infrastructure mediated through social groups, which are first more difficult to identify and then to measure. The Sachar Committee set up by the Government of India to study the social, economic and educational status of Muslims had prepared certain special tabulations for the population Census of 2001 in order to study the availability of infrastructure to the Muslim population. They divi ded the villages into Muslim population of less than 10%, 10% to 39% and more than 40% and also by the size of villages. It was found that there was little bias in the availability of primary schools in large villages, but as the size of the village became smaller it was seen that systematically the availability of a school in the village with a share of more than 10% and 40% of Muslims went down drastically (Sachar Committee Report 2006: Ch 7). This would have a clear impact on the choice of children going to school, especially girl children. Parents would hesitate to send young girls to a

february 28, 2009 vol xliv no 9



c ompared to girls (27% and 15% in urban and rural areas) in agegroups 5 to 19 years (NCEUS 2007).

Girls and boys from tribal households and Muslim and SC households were most likely to be out of school, only partly due to being engaged in economic activity (Figure 2). Being engaged in economic activity and domestic duties are “conversion factors” that need to be factored into the choice of children’s education.

Further, the level of education of the mother also appears to influence this choice. According to the National Health and Fertility Survey in 2005-06, children in the age group of 5 to 14 years were most likely engaged in paid or unpaid work or household chores for more than 28 hours per week, in households where the mother had no education (NFHS-3 2007: Table 2.21). As the years of mother’s education rose the percentage of children so engaged

Figure 2: Incidence of Out of School Children and Child Labour (5 to 14 Years) across Socio-religious

Groups (2004-2005) 4.10 and 4.11). Here again we see that the institutions of caste and religion bear down on the women in her fertility decisions and curtail her freedom of choice in the sphere of education. Age at marriage and birth of first child are good measures of “conversion factors”.

Perceptions of Parents with Regard to Children’s Education: The questionnaire method can be used to obtain information on the perceptions of people with regard to education and also how and when

Hindu ST Muslims Hindu SC All Hindu OBC Other Religions Hindu Upper

education and the costs involved become a burden

Socio-religious group

on the households. In Gujarat in 2000 we noted

Computed using NSS, unit level data 61st Round on Employment and Unemployment, 2004-05 and adjusted for

28 26 24 22 20 18 16 14 12 10 8 6 4 2 0 % of Child Workers 27.6 6.7 4.4 24.2 22.6 3.4 17.9 3.4 16.6 3.2 10.9 2.5 7.9 1.7 Child Labour Out of School Children (Labour Pool)

freedom to study and acquire formal education, the relative differences across communities may be reflecting the fact that Muslims and OBC households are more likely to be self-employed rather than in salaried jobs. All the same, it helps to understand the motivation behind not sending children to school.

Age at Birth of First Child: The additional burden borne by girls is early marriage and early childbearing. This has a negative effect on the chances of the girls completing school and continuing into higher education. In 2005-06, the age at first birth for women in the age group of 25 to 29 was the lowest at 19.1 years for ST,

19.2 for SC, 19.5 for Muslim, 19.9 for all Hindus (including SC/ST),

21.3 for Christians and the highest was 23.7 for Jain women. Teenage pregnancy (women in ages 15 to 19 years) was 16% among tribal women, 15.5% among SCs, 12.4% among all Hindus

and 11.8% among Muslims (NFHS-3 2007: Tables

census population. Source: NCEUS 2007, Figure 6.2.

declined, with hardly any children working in either kinds of work if the mother had more than 12 years of education. While this does show a link between mother’s education and perhaps her agency in getting her children into school, it also partly reflects that children from affluent households do not need to work just as women in such households are most likely to have acquired an education. Further, women from Muslim and SC/ST households were least likely to have studied up to Standard 6.

Using the questionnaire method we conducted surveys of 1,236 and 1,407 men and women in rural and urban areas in Gujarat in 2000 and again in 2008 in order to gauge the securities and insecurities they faced in the labour market (henceforth referred to as the GIDR Surveys, 2000 and 2008).1 This was modelled on the basis of the International Labour Organisation’s (ILO) Socioeconomic Security Programme (ILO 2004). In 2008, only about 8% of the households with children in the age group 5 to 14 years reported any child as working. The proportion was the highest among Muslim households (18%), followed by Other Backward Classes (OBCs) (nearly 11%), while it was the lowest, about 5%, among upper caste Hindu households. When the children were engaged in work nearly 75% of the respondents reported that having them to help was important for the business. Nearly half of the child workers did not attend school, and another 20% admitted that child work did interfere with the schooling of children. While this is a further measure of how child work can restrict that the majority of parents viewed the education

of children as a source of security for themselves in the future (GIDR Survey 2000). When we further queried about what level of education they aspire for their boys and girls if they were attending school, we found a gender difference in the responses. The mean level of education for the boys was graduation or above, while for girl children they only aspired to send them on average up to the higher secondary level. There could be a clear cultural reason for these responses, since most parents expected to live with their male children in their old age.

Poor households live from one crisis to another and have to garner resources to be able to stay alive. This has been clearly brought out in the ILO based socio-economic security surveys in many countries (ILO 2004: Table 4.5). Some households responded to a financial crisis by pulling their children out of school. In 2008 only about 3% of households with children in the age group 5 to 14 years reported that paying for the costs of children’s schooling constituted a financial crisis. However, 4.2% of these households said that they had to pull their children out of school in response to a crisis (GIDR Survey 2008). In most cases this was due to a serious illness in the household. While the proportion of parents having to resort to this was mercifully small, we noted a gender difference in that girl children were more likely to have to give up schooling. This points to one of the m otivations behind poor educational attainment among the poor and particularly among the girls.

Economic & Political Weekly

february 28, 2009 vol xliv no 9


On the positive side, however, the interest of parents to keep children in school was also brought out by another 3.3% of respondents with children of school going age who reported that they borrowed money or sold personal assets to keep children in school (GIDR Survey 2008).

4.2 Why Poor Educational Attainments

The quantity, level and quality of education in the country are in a deplorable situation even today. While communities that are lagging behind need to catch up with their education level, we find that they are falling behind even further. Besides, children are forced to drop out of school for various reasons and also end up having poor quality educational attainments.

Being Left Behind: In the era of globalisation education and skills are at a premium and this is crucial for communities if they wish to stay ahead. However, we found that in recent years, the Muslim community was probably getting left behind with the lowest reduction in illiteracy and growth of graduates compared to other communities. While the SC/ST reduced illiteracy by 6.9 percentage points and OBCs by 6.5, the Muslims were able to reduce illiteracy only by 4.8 percentage points from 1999-2000 to 2004-05. Further, the SC/ST and Muslims hardly increased above graduate education (0.3%), while OBCs and the forward communities did so by 1 and 2 percentage points, respectively, from 1999-2000 to 2004-05 (Sengupta et al 2008).

Dropout Rates: Another feature that indicates the lack of freedom of choice among children to attend school and the low educational attainments is the dropout rates. While there are a number of studies estimating the actual dropout rates, repetition rates, etc (Mehta 2007) using various secondary and primary sources of data, there are fewer efforts to understand the causes for such poor performance at the primary level.

It is well-known that girls tend to dropout of school after attaining puberty. However, the dropout rate is also mediated through the institutions of caste and religion. The status of children not attending school is of two kinds, those who never attended school and those who did and then dropped out. Nearly 25% of Muslim children in 2004-05 were either never enrolled in school or have dropped out, this was followed by nearly 20% of children from SC/ST households (Sachar Committee Report 2006: Figure 4.8). Less than 5% of children from upper caste Hindu household suffered this fate.

While quizzed about the main reasons for not attending school or dropping out among children in the age group 6 to 17 years, 36% of the boys and 21% of the girls said they were not interested in studies. It is difficult to interpret the reasons for the lack of interest, but it could reflect the low quality of education due to which children were unable to cope with the rigours of the educational system or that they could not see improved prospects of obtaining employment after acquiring education of such low quality. The next major reason given by 18% was that it cost too much. Further, while 15% of the girls said that they were required for household work, 7% of the boys said they were required for work on the family farm or enterprise (NFHS-3 2007:

116 Table 2.10). Thus, economic compulsion in terms of either lack of funds or the need to supplement incomes, or just the perception that it did not improve job prospects also act as barriers to acquiring education. Poor educational attainment due to early dropout from the educational system needs to be understood in terms of the reasons noted above.

Poor Reading Skills: There is a lot of literature available on the poor quality of education imparted in schools. This further dampens the enthusiasm among poor households to send their children to school, since such education would not provide the children with jobs. The Annual Status of Education Report (ASER 2007) based on surveys conducted by Pratham on an annual basis come out with startling results. The ASER survey found that 38.4% of children in Standard I could not recognise letters in 2006, but this decreased to 31.9% in 2007. Thus, while India’s enrolments rates in primary school are close to 90%, the learning achievements, although improving slowly, are a major cause for concern.

The NFHS-3 (2007) data shows that the ability to read varied according to the social groups and gender. In the age group 15 to 49 years, the STs fared the worst with 44% of men and 25% of women although having completed Standard 6, were not able to read. Of the rest, nearly 40% of men and 66% of women could not read at all. The SCs were slightly better off with 58% and 34% of men and women having completed Standard 6 and of the rest 27% men and 56% women not able to read at all. The next worst group was the Muslims, with just slightly lower percentage of those who could not read at all.

Poor Access to Media and Reading Newspapers: The capabilities approach also helps to understand the nature of the intrinsic good of education by distinguishing those aspects that are linked to achieved functionings, for example the value of considering a range of different viewpoints from newspapers, television and radio (Unterhalter 2003).

Given the poor reading skills noted above, we do not expect much exposure to newspapers among the adults in the age group 15 to 49 years. In 2005-06, about 29% of ST, 45% of SC and Muslim and 54% of all Hindu men reported reading the newspaper or magazines at least once a week, while only 10%, 15%, 18% and 23% of women, respectively, in these social groups read at least once a week (NFHS-3 2007). Among other media such as television, radio and cinema, it was exposure to television that was the highest in India. About 39% and 30% of ST men and women, 59% and 51% of SC, 57% and 45% of Muslim, and 54% and 56% of all

available at

Ganapathy Agencies

3/4, 2 Link Street Jaffarkhanpet, Ragavan Colony Chennai 600 083 Tamil Nadu Ph: 24747538

february 28, 2009 vol xliv no 9


Hindu men and women watched the television at least once a week. It is interesting to note that there is much less gender differential in viewing television than in reading of newspapers. This does not require any particular skills. However, even here the gender differentials exist and are particularly high among Muslims. Other religious groups, particularly the Jains, Sikhs and Christians were much better off in their exposure to the print and electronic media. Besides getting exposure, reading the newspapers and viewing television can be considered as empowerment of women and can be used as measures of the “conversion factor”, that might impact on the choice of education.

Role of Governments: As an issue to further social justice the GIDR Survey (2008) enquired what the government should do to help those who could not meet their basic needs. While the majority of the respondents felt that the government should provide jobs, nearly 20% of the men and 14% of the women respondents thought the government should do something to help meet the basic needs of education of children. The suggestions given included reducing fees, giving bicycles to school-going children, and provide high school and degree colleges in the village or in the urban block. The government already has a programme of giving bicycles to girl students in Gujarat and this was perhaps a demand for universalising the scheme. Thus, there was an awareness and demand for schooling and higher education in the community that was so far unfulfilled.


The capabilities approach is considered superior to the utility approach. While the utility approach assesses the resources and opportunities to achieve functionings, such as going to school,


The male and female respondents in these surveys were chosen based on stratified random sampling, by the activity status. The areas chosen for study were biased towards low income households since the main theme of the survey was socio-economic security. The urban areas surveyed in both years were Ahmedabad city. The rural areas surveyed in 2000 were in Ahmedabad district while in 2008 it was Surendranagar district. The sampling design for the survey of 2000 is presented in Unni and Rani (2002).


ASER (2007): Annual Status of Education Report, Pratham, New Delhi aser07/aser2007.php.

Doraisamy, Malathy (2001): “Demand for and Access to Schooling in Tamil Nadu” in A Vaidyanathan and P R G Nair (ed.), Elementary Schooling in Rural India (New Delhi: Sage Publications).

– (2007): “Child Schooling and Child Work” in Abusaleh Sheriff and Maithreyi Krishnaraj (ed.), State Markets and Inequalities: Human Deve lop ment in Rural India (New Delhi: Orient Longman).

GoI (2006): “Social, Economic and Educational S tatus of Muslim Community in India”, Sachar Committee Report, Prime Minister’s High Level Committee, Cabinet Secretariat.

the capabilities approach assesses the motivations and freedom people have to formulate capabilities. It, therefore, encourages us to ask a wider range and more probing questions than the utility approach.

The capabilities approach, however, is still in its infancy when it comes to formal measurement and criteria for evaluation of various capabilities. In the case of education, do people have the freedom to choose and what motivates it? The starting point for this paper is that the motivations and freedoms of choice are mediated through institutions of caste and religion. It restricts the freedom, more so of women, and makes them place a low value on education. We explore indicators for the motivation behind poor school attendance and poor educational attainments based on the institutional context. We note that poor school infrastructure, domestic work, child labour, early marriage and childbirth and perceptions of parents motivated through institutions of caste and religion influence poor attendance of girl children in school. Similarly, caste and religion also motivate poor educational attainments through higher dropout rates, poor reading skills and poor access to media among girls. These explanations for the achieved functionings of schooling by gender help to identify “conversion factors” not typically addressed in the utility approach. The capabilities approach may have an important public policy role to play as research on the dynamics of capabilities failure could be used to improve education policy design and, in turn, educational outcomes for girls from disadvantaged caste and religious communities. However, the remedy for poor enrolment and educational attainment of children may not lie in education policy alone, but in a wider set of socio-economic polices.

ILO (2004): Economic Security for a Better World, ILO Socio-Economic Security Programme (Geneva: International Labour Organisation).

Kingdon, Geeta G and Jeemol Unni (2001): “Education and Women’s Labour Market Outcomes in I ndia”, Education Economics, 9(2), pp 173-95.

Kingdon, Geeta G and John Knight (2003): “Wellbeing Poverty Versus Income Poverty and Capabilities Poverty”?, CSAE WPS/2003-16, Global Poverty Reduction Group, Centre for the Study of African Economies (UK: University of Oxford).

Kingdon, Geeta G and Nicolas Theopold (2008): “Do Returns to Education Matter to Schooling Participation? Evidence from India”, Education Economics, forthcoming.

Mehta (2007): Student Flow at Primary Level – An Analysis Based on DISE Data (New Delhi: National University of Educational Planning and Administration).

NCEUS (2007): Report on Conditions of Work and Promotion of Livelihoods in the Unorganised Sector, National Commission for Enterprises in the Unorganised Sector, Government of India.

NFHS-3 (2007): National Family Health Survey, India, Vol I (Mumbai: International Institute of Population Studies).

NSSO (2006): “Status of Education and Vocational Training in India, 2004-05”, 61st Round, Ministry of Statistics and Programme Implementation (New Delhi: Government of India).

Psacharapoulos, G (1994): “Returns to Investment in Education: A Global Update”, World Development, 22(9): pp 1325-43.

Sen, Amartya (1983): “Poor, Relatively Speaking”, O xford Economic Papers, 35, pp 153-69.

– (1987): Commodities and Capabilities (Delhi: O xford University Press).

Sengupta, Arjun, K P Kannan and G Raveendran (2008): “India’s Common People: Who Are They, How Many Are They and How Do They Live?”, Economic & Political Weekly, 43(11), pp 49-63.

Unni, Jeemol (1996): “Returns to Education by Gender among Wage Employees in Urban India”, Journal of Educational Planning and Administration, April.

– (2007): “Earnings and Education among Social Groups” in A Sheriff and M Krishnaraj,

State, Markets and Inequalities: Human Development in Rural India (New Delhi: Orient Longman).

Unni, Jeemol and Uma Rani (2002): “Insecurities of Informal Workers in Gujarat, India”, SES Papers 30 (Geneva: International Labour Office).

Unterhalter, Elaine (2003): “Education, Capabilities and Social Justice”, background paper prepared for the Education for all Global Meeting Report, 2003-04, Gender and Education for All: The Leap to Equality, 2004/ED/EFA/MRT/PI/76, United Nations Educational, Scientific and Cultural Organisation.

Economic & Political Weekly

february 28, 2009 vol xliv no 9

Dear Reader,

To continue reading, become a subscriber.

Explore our attractive subscription offers.

Click here

Back to Top