Inequality in Access to Higher Education in India between the Poor and the

Rich: Empirical Evidence from NSSO Data

Jandhyala BG Tilak

(Council for Social Development)

Pradeep Kumar Choudhury

(Jawaharlal Nehru University)

Paper prepared for the 36th IARIW Virtual General Conference

August 23-27, 2021
Session 17: Inequalities in Opportunity for Income, Education and Health II
Time: Thursday, August 26, 2021 [14:00-16:00 CEST]
 Inequality in Access to Higher Education in India between the Poor and
the Rich: Evidence from NSSO Data

Jandhyala BG Tilak
Council for Social Development, New Delhi

Pradeep Kumar Choudhury

Assistant Professor of Economics
Zakir Husain Centre for Educational Studies
School of Social Sciences
Jawaharlal Nehru University, New Delhi
Email: [email protected]

Abstract
Rising inequalities in the society are indeed becoming an important concern of all.
Among inequalities in different spheres, inequalities in education, and inequalities in
higher education in particular are seen as too serious to ignore any more. The
available studies on inequality to access higher education in India have largely
examined the issue from gender and social category of the students; too little is done
by examining income as a determining factor. Using NSSO surveys, conducted in
2007-08 and 2013-14, an attempt is made here to examine the income inequality and
access to higher education in India. The analysis shows that the inequality in access to
higher education has increased substantially by household‘s economic status in the
last seven years. Though the overall gender inequality has come down significantly,
this is very high between the rich and the poor. The inequality in access to HE also
varies considerably between rural and urban regions. The logit results lead us to
conclude that rich income groups have a higher probability of attending higher
education institutions than others. The difference in the probability of participation
between men and women narrows down as one move from poorest to richest
quintiles. Recent debates on higher education in India have raised a variety of
interesting policy related issues and through this empirical study the author has
highlighted a few of them, particularly the interaction between income inequality and
access to higher education, with the aim to facilitate a more informed policy discourse
on this.

Keywords: Income Inequality, Access, Higher Education, India

JEL Codes: I22, I23, I24, I25
The Problem

The role of higher education in national development is well recognised all over the
world. It is seen as a lever of social transformation as it is about enhancing knowledge
and skills of people. According to the human capital theory1 investment in higher
education makes a vital contribution to accelerate the process and the rate of
economic growth through enhancing human skills and productivity. Subsequent
research has shown that higher education critical for boosting economic growth,
improving income distribution, reducing poverty and reducing social and economic
inequalities, as it is regarded as the primary engine of upward mobility. It plays an
important role in promoting many dimensions of development of nations with respect
to social progress, human development, political stability and various other facets of
development (Tilak, 2003; 2007; 2018). Further, in the globalised knowledge
economy (a catalyst for the increased market demand for higher education) the types
of skills and knowledge required are increasingly acquired in higher education
institutions. Higher education has now become a necessary qualification to enter into
and compete for a decent job in the knowledge economies (Varghese and Malik,
2016). From human development perspective, investment in higher education is not
just a step towards improvement of productivity and better income distribution, but
quite importantly, an action towards fostering higher autonomous citizens who will be
able to decide more intelligently on the alternative lifestyle they could have (Comim,
2007: 96). In all, as the Task Force on Higher Education and Society (2002)
observed, ―Higher education is no longer a luxury; it is essential for survival. Higher
education is the modern world‘s ‗basic education.‖ In short, there are both economic
and non-economic incentives for expansion of higher education. Accordingly, many
developing countries have been experiencing fast expansion of their higher education
system and are fast entering a stage of massification. But much of the expansion in
higher education is taking place in the private sector in most developing countries,
while historically such an expansion took place in advanced countries in public sector
and mainly through public efforts; there has been a virtual halt in the growth of public
higher education, reducing the relative size of the public sector to a negligible level
(Tilak, 2013: 41).

1
See Theodore W. Schultz (1961) for an elaborate discussion on the fundamental aspects
human capital theory.
 Similar to the global trends, the higher education sector in India has seen a
massive expansion during the seven decades following independence and particularly
in the recent decades from early 1990s. There were only 0.26 million students in
higher education enrolled in 750 colleges and 30 universities in India in 1950-51. This
has increased to about 34.6 million students in 39,071 colleges and 11,923 ‗standalone
institutions‘ in 2015-16 (MHRD 2016). The gross enrolment ratio2 (GER) in higher
education, as estimated by the MHRD based on data collected from institutions of
higher education through the All-India Survey of Higher Education, has gone up
almost sixty times – 0.4 per cent in 1950-51 to 25 per cent in 2016-17 (UGC, 2015;
MHRD 2017). With this, India has grown into one of the largest systems of higher
education in the world; it is the second largest after China.

While the expansion of higher education sector has helped the country to reach
a stage of massification (which is to be celebrated), it is equally important to analyse
and identify the winners and losers in the process of expansion. Did the expansion of
the system lead to the widening of access to higher education among under-
represented groups and regions or has it widened inequalities? A major concern that is
highlighted often in the studies and policy debates include unequal access to and
participation in higher education among different socioeconomic groups of
population. There are visible disparities between regions, widening inequalities
between poor and non-poor and between social groups; and this is viewed as a
growing social concern. The groups that lag behind include women, scheduled castes,
scheduled tribes, ‗other‘ backward classes,‘ Muslim, and the poor from all groups,
particularly from rural areas (Thorat, 2016: 33). The enrolment rates of these groups
continue to be low, compared with their counter parts. For example, in 2016-17, as
against the overall gross enrolment ratio 25 per cent, it is 21 per cent for scheduled
castes and 15.4 for scheduled tribes. Similarly, the gross enrolment ratio is 26 per cent
among men and 24.5 among women, showing, of course, no big difference between
men and women. Between different states/union territories, the ratio ranges from 5.5
per cent in Daman & Diu and 56.1 per cent in Chandigarh; among the major states it

2
Gross enrolment ratio in higher education is the ratio of students enrolled in higher education
to total population in 18-23 age group. The enrolment ratio is called ‗gross‘ as it does not adjust for
students for age-group; it considers all students irrespective of age group in the numerator, while the
denominator includes only population of the age-group 18-23. This is considered the most standard
and widely used indicator of development of higher education.
varies between 14.4 per cent in Bihar and 46.9 per cent in Tamil Nadu (MHRD
2017).3

Like in many other developing countries of the world, higher education sector
in India was accompanied by fast growth of the private higher education institutions,
particularly during the last quarter century (Tilak 2009). Also, within the private
sector, it is the ―for-profit‖ higher education segment (which is largely market-
driven) is growing fast and the philanthropy and charity based private higher
education seems to be disappearing (Tilak, 2006; 2013; Varghese, 2015). The
contribution of private sector in higher education has raised equity, quality and
efficiency concerns, equity concerns being very serious, as students from lower
income families hardly access these institutions as these institutions charge exorbitant
levels tuition and other fees. Further, students from poor families face greater
difficulty in accessing limited seats available in elite public institutions, such as the
Indian Institutes of Technology, National Institutes of Technology, Indian Institutes
of Management, etc., due to tough entry level all-India competition4. The
representation of students in elite public higher education institutions is largely
confined to economically well-off families. The rising income inequality5 has
increased the challenges to access higher education (specifically quality higher
education) in India for the students from poor households and as a result, they are
persistently under-represented in institutions of higher learning.

In this context, this paper has been an attempt to unravel some specific inter-
related dimensions of inequality in participation in higher education by economic
status of the households. The importance of examining the linkages between
economic status and participation in higher education also lies with the fact that a
substantial proportion of the increase in economic inequality is linked with the
increase in the returns to education and low level of intergenerational mobility. More

3
The state-wise statistics on gross enrolment ratio in higher education in 2015-16 and 2016-17
are given in Tables A1 and A2 in the Appendix.
4
The public higher education institutions in India follow certain affirmative action policies to
admit students from some social groups such as scheduled castes, scheduled tribes, and other backward
class. However, there is hardly any such policy for admitting the students based on their economic
status.
5
The findings of the World Inequality Report 2018 (as reported in The Hindu, 14 December,
2017) reveals that the income share of India‘s top 1 per cent rose from approximately 6 per cent in
1982-1983 to around 23 per cent by 2014 and that of the top 10 per cent increased from 10 per cent to
56 per cent during this period.
clearly, a vicious circle is clear: the barriers to access to higher education among low-
income students widen the income inequality, which in turn widens the inequality in
access to higher education. Given these facts, it is important to examine how far
students from poor households are able to access higher education in India. We look
at the problem of unequal access to higher education by gender and region (rural-
urban) in the backdrop of economic inequalities. Inequality in higher education is
examined in terms of gross enrolment ratio6, gross attendance ratio7 (GAR) and higher
education attainment (HEA).8

Examining issues relating to unequal access to higher education in India, many

scholars (e.g., Chanana, 1993, 2016; Dhesi 2000; Sundaram, 2006; Hasan and Mehta,
2006; Raju, 2008; Salim 2008; Srivastava and Sinha, 2008; Sinha and Srivastava,
2008; Azam and Blom 2009; Ghuman, Singh and Brar 2009; Sundaram 2009;
Chakrabarti 2009; Basant and Sen, 2010, 2014; Srinivasan 2010; Khan and
Sabharwal, 2012; Tilak 2015; Thorat, 2016; Wankhede, 2016) have analysed the
variations in participation and attendance in higher education across different social
groups (caste and religion), gender, location of the households. These and several
other studies have found that enrolment ratio significantly varies between boys and
girls, and gender is an important factor in determining the access to higher education.
There has been a phenomenal growth in the number of female students enrolled in
higher education in India since independence. Currently women constitute 47 per
cent of total enrolments in higher education in 2016-17 (MHRD 2017). But gender
inequality persists in rural areas, among scheduled and non-scheduled population, and
even among the poor and even rich families. This received attention of some scholars
in recent years, who have studied gender inequality in higher education across social
groups, location of the household, discipline of study, type of institution etc. (e.g.,
Rao 2007; Raju 2008; Srivastava and Sinha 2008; Salim 2008; Ghuman, and Singh

6
Gross enrolment ratio in higher education is the ratio of students enrolled in higher education
to total population in 18-23 age groups.
7
Gross attendance ratio in higher education is the ratio of students attending higher education
to total population in 18-23 age group. Similar to gross enrolment ratio, the numerator (number of
students attending) does not make any adjustment for age group. As explained later, NSSO provides
data on attendance rate, not on enrolment ratios; but the scholars who used NSSO database, use these
two terms synonymously, of course, not very inappropriately.
8
HEA is defined as percentage of higher educated population in the total adult (normally 15+
age group) population.
and Brar 2009). The participation in higher education (measured in terms of gross
enrolment ratio) of women in urban areas is four times higher than those in rural
areas. Women in rural areas have remained doubly deprived; being women and living
in rural areas (Raju 2008). The status of women belonging to different disadvantaged
social groups such as scheduled castes and scheduled tribes in higher education
appeared to be worse than that of those belonging to forward castes. For example, the
gross enrolment ratio for scheduled tribe women is 12.9 per cent, as compared to the
overall gross enrolment ratio among women of 23.5 per cent (MHRD 2016).
Similarly, the participation of Muslim females in higher education was six per cent as
compared to nine per cent for Hindu females, 13 per cent for Sikh females and 16 per
cent for Christian females in 2005 (Thorat 2008; Srivastava and Sinha 2008).

As caste is a very important phenomenon in India, many studies have focused

their attention on inequalities in higher education between by social groups – caste
and religion (Chanana 1993; Kaul 1993; Hasan and Mehta 2006; Rao 2006; Dubey
2008; Srivastava and Sinha 2008; Thorat 2008; Sundaram 2009; Biswas et al 2010;
Basant and Sen 2010, 2014). As many of these studies found, the participation of two
disadvantaged caste groups namely scheduled castes and scheduled tribes in higher
education have improved over time, but in absolute terms, the rates were only about
half of the participation of non-scheduled population in 2004. The ‗other‘ backward
classes have higher participation rates than scheduled castes and scheduled tribes, but
lower than that of general category students (Azam and Blom 2009). The study by
Basant and Sen (2014) also using NSSO data concludes that Hindu upper castes have
higher probability of participation in higher education and Muslims and ‗other‘
backward classes have lowest chances. The study by Hasan and Mehta (2006) shows
that enrolment ratio in higher education among scheduled castes and scheduled tribes
in urban areas are slightly above their respective shares in total population, but it is
not so in case of rural areas. After controlling for completion rate in higher
secondary education, economic status is found to be a better predictor of college
attendance than social identity in urban India, while for rural areas the group identity
does matter. Wankhede (2016) has argued that the social backwardness of these
groups results into social sufferings and economic exploitation with high degree of
dependence on upper castes, which further leads toward educational backwardness.
The discipline-wise distribution of students from different social groups reveals a few
important aspects. Apart from overall rates of participation, we note significant
differences in the enrolment of students by discipline of study. Ghuman, Singh and
Brar (2009) found, based on a primary survey in rural Punjab found that as high as
three-fourth of total students from rural background studying in different professional
education programmes belonged to forward castes, leaving only one-fourth of total
space for the socially disadvantaged sections of the society. Differences exist in the
enrolment of students by different religious groups such as Hinduism, Islam,
Christianity, Sikhism, Jainism etc. The highest enrolment is among students
belonging to Hindu religion followed by Christian, Sikh and Jain. Students of Muslim
religion are least represented. Nivedita Sarkar (2016) also analysed, based on NSS
data, similar differences in women‘s participation in higher education across
disciplines.

Access to higher education differs considerably between the students residing

in rural and urban areas. Regional – rural-urban disparities in higher education arise
due to natural clustering of institutions of higher education in and around metropolitan
and urban areas (Sinha, 2008; Agarwal, 2009). Students from rural areas do not have
many options to choose, which affects their participation in higher education. On the
other hand, people from urban areas are having a moderate access to a variety of
educational institutions and hence, they seem to be able to access education according
to their choice. Furthermore, it is not only the availability of opportunity that matters
to participate in higher education, a number of socio-economic factors of the
households are also important. The rate of participation of people in urban areas in
higher education is three times higher than that of the rural population in 2004-05
(Raju 2008). Though the enrolments in rural areas increased faster than enrolments in
urban areas during the last two decades, the students from rural areas still form only
30 per cent of the total enrolments in higher education in India (Azam and Blom
2009). Describing socio-economic profile of the students entering into higher
education, Hasan and Mehta (2006) based on 55th round of National Sample Survey,
reported that out of the total students enrolled in colleges, as high as 63 per cent were
from urban areas and the rest 37 per cent are from rural areas. Using 50th, 55th and
61st rounds of National Sample Survey data, Dubey (2008) has shown that the
probability enrolment in higher education was lower by three per cent for women in
rural areas and 0.3 per cent lower for women in urban compared to men.
 Besides examining the disparities in access to higher education by gender,
social groups (caste and religion), and location, a few studies (e.g., Salim 2008; Raju,
2008; Basant and Sen 2010; and Srinivasan 2010; Khan and Sabharwal, 2012;
Borooah, 2016) have examined the unequal access to higher education by some other
important socio-economic and institutional characteristics such as occupation of the
parents, economic status of the households, educational level of the parents,
household size, type of institutions etc. However, survey of literature points out that
although a few studies have mentioned that economic status of the household is a
major barrier to access higher education, academic interest to examine it in detail has
been relatively limited.

Among the recent studies, Basant and Sen (2014), Tilak (2015), Thorat
(2016), Wankhede (2016), Borooah (2018), Deshpande (2018), Kundu (2018), Sinha
(2018), Thorat and Khan (2018) have examined several dimensions of inequalities in
higher education (gender, caste, religion, region) and concluded that inequalities
between the rich and the poor are the highest and moreover they are increasing even
with the expansion of higher education sector in India.

Tilak (2015) has examined the growth and inequalities in higher education in
India using data from several NSS rounds between 1983 and 2009–10. The study was
primarily concerned with inequalities in higher education by gender, by social
groups—caste and religion, by region—rural and urban and by economic groups of
population classified by monthly per capita household consumption expenditure.
Considering important indicators on higher education, such as the gross enrolment
ratio, transition rate, and higher education attainment, Tilak examines whether
inequalities in higher education have increased or declined overtime. The study also
throws light on the groups that have improved most over the years in their higher
education status and on the decline or increase of inequalities between groups. Tilak
reports that gender inequalities in higher education have been reduced substantially;
there was good improvement in inequalities between scheduled and non-scheduled
population; but rural-urban inequalities are high and have not diminished much; and
inequalities between the rich and the poor are highly striking, and they have widened
over the years.

Hence it may be in order to focus on inequalities between the rich and the poor
in their access to higher education. Tilak (2015), like many others, has, however, not
examined inequalities between sub-groups of population like between women among
scheduled castes versus men among scheduled castes or between women among
scheduled castes versus women among other groups. By considering various groups
with reference to economic class, the present study attempts at deepening the
understanding of the inequalities in participation in higher education in India. It aims
to examine the heterogeneity in access to higher education by gender, social groups
(caste and religion) and location of the household (rural/urban), considering economic
class as the reference point. The relationship between economic status of the
household and their attendance in higher education is analyzed by gender, social
groups, and location of the household (rural/urban). In this study monthly per capita
consumption expenditure of the household is used as proxy for their economic status.
We note that in a few other studies (e.g., Dreze and Kingdon 1999; Duraisamy 2001;
Nagarajan and Madheswaran 2001; Tilak and Sudarshan 2001; Chakraborty 2006;
Srinivasan 2010), the economic condition of the households is measured not just in
terms of family income; rather they take into consideration a number of other factors
like the ownership of land, assets of the family, type of house the households live in
(pucca or kuchha) etc. Average monthly per capita consumption expenditure of the
households, data on which are regularly collected and provided by NSSO,9 are
extensively used by researchers as well as policy makers while measuring the
economic status of households.

Further, the study examines the variations in the household expenditure on

higher education by socio-economic groups. It is argued that the quality of higher
education accessed by the students of poor and non-poor households varies
substantially and this is largely due to the differences in their spending on higher
education. Even if some poor households send their wards to higher education, they
spend significantly less on it, as compared to the non-poor households, which might
affect quality, continuation, and performance of students in the studies. Similarly,
literature on household spending on higher education also reports existence of gender
bias in the household expenditure on higher education, more prominently among poor
families. The present study uses disaggregated individual-specific observations
available in the latest two education rounds database of the National Sample Survey

9
NSSO does not collect data on household or individual income.
Organization (NSSO) -- the 71st (January-June 201410) and 64th (July 2007 – June
2008) rounds. Inequality in access to higher education by social and religious groups
is equally important to examine but we do not look at those in detail though some
references are made in the discussion. After all, it may be safely assumed that the
lower quintiles include majority of the students belonging to scheduled caste and
scheduled tribe students. Thus inequality in attending higher education and family
expenditure on higher education by economic status of the household11 is analysed
here by gender, social groups, location of the household (rural/urban) and institution
type.

The following section briefly discusses the data set used for the analysis. It
also spells out the method used for the analysis. The inequality in access to higher
education --measured in terms of the gross attendance ratio, and higher education
attainment that is percentage of higher educated people in the total population -- by
economic status of the households in India is analysed in detail in the subsequent
sections. Taking economic status as cross-cutting reference for all dimensions, gender
and rural-urban differences are analysed. The last section provides a summary of the
major findings of the study along with some important policy implications.

Note on Data and Method

This paper uses the disaggregated individual specific unit level data available in the
latest two education rounds the National Sample Survey Organization (NSSO) --
the 71st round conducted in January-June 2014, and the 64th round conducted in
July 2007 – June 2008. The 64th round (Participation and Expenditure in
Education) covers a sample of 1, 00,581 households (63,318 rural households and
37,263 urban households). The 71st round (Education in India) includes a sample
of 65,926 households (36,479 rural households and 29,447 urban households) from
all over India. Unlike the more ‗general‘ or ‗normal‘ rounds, the focus of these two
rounds of data was to collect information on three important issues related to
10
We refer this to as 2013-14, as the survey conducted during January to June 2014 covers a
major part of the academic year 2013-14.
11
The economic status of the household is measured in terms of the quintiles based on the
average monthly per capita consumption expenditure (MPCE) of the households. The first quintile
includes bottom 20 per cent of the population, the second quintile includes 21-40 per cent of the
population and so on. Quintile one is the poorest group, while quintile 5 covering 81-100 per cent of
the population is the richest group.
education, in addition to many other household level characteristics in detail: (a)
participation years in education, (b) family expenditure, often referred to as private
expenditure, incurred by households on education, (c) incentives provided by the
government and (d) the extent of educational wastage in terms of dropout and
discontinuation along with causes of the same. The surveys also give information
on number of adults who have acquired higher education (or completed level of
higher education).

In this study, we have used the original unit level data, rather than
confining to the published tables brought out by the NSSO in its reports. The
availability of unit level data has allowed us to carry out the analysis in depth at a
disaggregated level. Further, the NSSO data used for the study helps us to analyse
by economic classes. Note that the data available from Ministry of Human
Resource Development, Government of India, the University Grants Commission,
and other government organisations, do not give us this information. Also, the
National Sample Survey (NSS) data are considered better not only because they are
highly reliable, but also in scope and detail than others, as they provide household
level information on several parameters that help us to examine in depth some of
the issues relating to inequality in participation in higher education. Inequalities in
participation in higher education are analysed here using gross attendance ratio.
While gross enrolment ratio is used more commonly to measure the participation,
the NSSO survey, because of its household approach, considers current attendance.
We believe that the gross attendance ratio is better than gross enrolment ratio, due
to likely differences between enrolment and attendance. As no data are available
on differences between enrolment and attendance, many scholars that we
mentioned above have used gross attendance ratio to be synonymous with gross
enrolment ratio. Higher education here includes graduation and higher levels of
education. Diploma courses after graduation are included in higher education, but
diploma courses after higher secondary level (but below degree level) are not
considered.

The analysis covers three major dimensions: First, inequalities in access to

higher education (measured in terms of the gross enrolment ratio and gross attendance
ratio) and higher education attainment are analysed. The trend and pattern of
attendance in higher education by different socio-economic and institutional factors
(gender, caste, location of the household, and type of institution) are discussed using
descriptive statistics. In all the cases economic status of the household is taken as
cross-cutting core category. Second, inequalities in household spending on higher
education are analysed. The variations in the household spending on higher education
are shown by gender, location (rural-urban), and type of institution for each
consumption expenditure quintile. Third, using the unit level data of 2013-14, the
predicted probabilities of attending higher education is analysed for persons aged 18-
23 years using logit model. The dependent variable for the logit estimation is a
dummy variable and takes value 1 for the persons who are in the age group of 18-23
and are currently attending higher education and, 0, if they (of the age group 18-23)
are currently not attending higher education. The analysis considers gender, location
of the household (rural/urban), social groups (caste and religion), income quintile and
household size as explanotory variables. To examine the hetrogeneity in the prdicted
probabilities of attending higher education , the statistical analysis is separtey made
by gender, location of the household and quintiles. However, as the NSS data are
based on sample surveys and observations become fewer as one moves to smaller and
smaller sub-groups, some of the results given here need to be interpreted with caution.

Trends and Pattern in Participation in Higher Education in India

Gross Enrolment Ratio

First, as per official statistics, the gross enrolment ratio in higher education in India is
24.5 per cent in 2015-16 (MHRD 2016). However it varies widely between states,
gender and social category. In some of the states/union territories such as Chandigarh,
Delhi, Kerala, Puducherry, Tamil Nadu, and Telangana, the gross enrolment ratio is
higher than the national average while the corresponding ratio is below the national
average in Bihar, Chhattisgarh, Jharkhand, and Odisha. Among the major states Bihar
figures at the bottom with 14.3 per cent gross enrolment ratio while Tamil Nadu
comes at the top with the gross enrolment ratio of 44.3 per cent (see Table A1, in
appendix12). Table 1 shows that economically better-off states (with Net State
Domestic Product per capita higher than national average) have achieved high gross
enrolment ratio (higher than national average), while poor states (with low NSDP per
12
Tables and figures are given in the end of the paper.
capita) are with the low gross enrolment ratio in higher education, with very few
exceptions. This clearly reveals the positive relationship between economic conditions
and participation in higher education in India at macro level.

The gross enrolment ratio for men in higher education is 25.4 per cent, while
it is 23.5 for women in India in 2015-16 (Table A1), showing no significant
difference. But in the states like Uttar Pradesh, Haryana, Himachal Pradesh, Jammu &
Kashmir, and Punjab the partcipation rate of women is higher than that of men.
Similarly, there are also variations in gross enrolment ratio in higher education by
social catogories. The representation of scheduled castes and scheduled tribes in
higher education is quite low, as compared to ratio for all. The gross enrolment ratio
among scheduled castes is 19.9 per cent and that among scheduled tribes is 14.2 per
cent, while the average of all is 24.5 per cent in 2015-16. Female students belonging
scheduled tribes are ssociated with the lowest gross enrolment ratio ratio, which is
12.9 per cent (Table A1).

Based on NSSO data, estimates on gross enrolment ratio in higher education

during the period 1983-84 to 2009-10 by gender, region, social groups (caste and
religion) and household expenditure quintiles are presented in Table 2. While only 7.6
per cent of the 18–23 age group population attended higher education in 1983–84, in
2009-10, 23.1 per cent attended, i.e., in about 26 years, the ratio increased by three
times. The gross enrolment ratio among men increased from 10.9 per cent in 1983–84
to 27 per cent in 2009–10: it increased by 2.5 times in about two decades and a half.
In contrast, only 19 per cent of the female in the relevant age group were enrolled in
higher education in 2009–10. But what is strikingly clear is: there has been a rapid
progress in the enrolment ratio among women, compared to men. The gross enrolment
ratio among women increased by more than four times. As a result, gender
inequalities in gross enrolment ratio have come down very significantly during this
period, indicating strong trends towards convergence. This may due to different
policies brought out by the Government of India to provide girls better access to
education both in school and higher education level.

The enrolment ratios of scheduled castes and tribes have consistently been
very much below those of non-scheduled population or the total population on
average. But both scheduled castes and scheduled tribes have made significant
advancement in participation in higher education, as the enrolment ratios of the
respective population groups increased by four to five times in about two decades and
a half between 1983–84 and 2009–10. The growth was relatively faster in case of
scheduled tribes, though in absolute terms their enrolment ratio is less than that of the
scheduled castes; and as a result, the differences between scheduled castes and
scheduled tribes have come down; and also the differences between the scheduled
population and non-scheduled population declined. However, it must be added that:
(a) the enrolment ratios among both the scheduled castes and scheduled tribes are low
and (b) still significant inequalities persist between scheduled and non-scheduled
population groups. The enrolment ratio in 2009–10 was nearly 12 per cent among the
scheduled tribes and 15 per cent among the scheduled castes, compared to 23 per cent
for all (Table 2).

Inequalities in gross enrolment ratio between various religious groups are

much higher. Estimates on gross enrolment ratio are available for Hindus, Muslims,
Christians and ‗others‘. The enrolment ratio is the highest among the Christians and
the least among the Muslims (Table 2). This is the same situation consistently
throughout the period between 1983–84 and 2009–10. Enrolment ratio among
Muslims was only 14 per cent in 2009–10, while it was 24.2 per cent among Hindus
and 37 per cent among Christians. The enrolment ratio among ‗Others‘ that includes
Jains, Sikhs, etc., is also high — 28 per cent in 2009–10. While there has been
improvement in case of all the four groups between 1983–84 and 2009–10, the inter-
group inequalities by religion did not decline much. In fact, the gap seemed to have
widened.

In contrast to inequalities by gender, caste and religion, rural–urban disparities

seem to be very high in the enrolment ratios. While 39 per cent of the relevant age
group population in urban areas attended colleges/universities in 2009–10, it is only
16.5 per cent population who attended in rural areas. The ratio in urban areas was
nearly 4.5 times higher than the ratio in rural areas in 1983. In 2009–10, this came
down to 2.3 times, suggesting narrowing down of rural–urban disparities.

Among the expenditure groups, the gross enrolment ratios are the lowest
among the bottom (poorest) quintile and highest among the top (richest) quintile. One
finds a very systematic pattern of increasing enrolment ratios by every increase in the
expenditure level of the households, with no single exception. In other words, the
enrolment ratio among the second quintile (from bottom) has been higher than the
bottom quintile; the ratio among the third (middle) quintile is consistently higher than
the ratio among the second quintile; and so on. The population belonging to the top
income quintile has the highest ratio. This pattern did not change at any point of time
that we studied between 1993–94 and 2009–10. More importantly, inequalities in
enrolment ratios between the poorest and the richest quintiles have increased over the
years, as the enrolment ratio among the poorest quintile declined between 1993–94
and 2004–05, while the same has increased in case of all other quintiles, and at a
disproportionate rate in case of the richest quintile. The ratio in case of the richest
group increased from 26 per cent in 1993–94 to 37 per cent by 2004–05, while the
ratio for the poorest declined from a bare 2 per cent to 1.8 per cent during this period.

Gross Attendance Ratio: 2007-08 and 2013-14

Now, based on the 64th and 71st rounds of NSS, we examine gross attendance ratio
and inequality in the same between different groups. Table 3 presents the estimates of
eh same, namely, the gross attendance ratio (age group 18-23) in higher education by
gender, location and type of institutions and by expenditure quintiles in 2007-08 and
2013-14. In 2007-08, the gross attendance ratio in higher education in India was 12.5
per cent which has gone up to 24 per cent in 2013-14.

We note a very systematic pattern in the attendance ratios by expenditure

quintiles: the ratios increase systematically by increasing economic status of the
households, with no exception. Not only the richest quintile is at the top and the
poorest quintile at the bottom in attendance ratios, the ratio of any quintile is higher
than the ratio of the preceding (lower) quintile. This is true both in 2007-08 and
2009-10 (Figure 1). Gross attendance ratio in higher education by economic status
of the households shows wide and increasing inequality between 2007-08 and 2013-
14. In 2007-08, the difference in the gross attendance ratio between poorest and
richest families is 29.5 per cent points and this gap has gone up to 43.5 per cent points
in 2013-14 (Table 3). Between 2007-08 and 2013-14, the gross attendance ratio for
the poorest families has increased by 5.3 per cent (2.9 per cent to 8.2 per cent) while
for the richest households it has gone up by 19.3 per cent (32.3 per cent to 51.6 per
cent). This shows that the inequality in access to higher education has increased
substantially by household‘s economic status in the last seven years, corroborating the
findings of Tilak (2015) for earlier years.

Pattern of gender inequality in access to higher education by economic status

of the households provides some interesting aspects. In both the years, 2007-08 and
2013-14, the gross attendance ratio in higher education among men is higher than
among women. The difference in the ratio between men and women was 2.7 per cent
points in 2007-08 which has marginally increased to three per cent points in 2013-14.
The inequality between the poorest and the richest households has increased
alarmingly both among men and women: the difference in the ratio among men
increased from 27.8 per cent points in 2007-08 to 42.5 per cent points in 2013-14.
Such a difference among women has gone up from 31.5 per cent points to 44.5 per
cent points during the same period. It shows that the inequality in attending higher
education between poor and rich households is very high, and it has increased during
the last seven years in case of both men and women. The inequality and the increase
in inequality – both are higher among women than among men. Further, gender
inequality in the attendance ratio also varies by location of the households
(rural/urban) and it is more so when the household‘s economic status is taken into
consideration. For example, in both 2007-08 and 2013-14, the gross attendance ratio
among women belonging to urban areas is higher than that among men. However, the
attendance ratio is higher among men compared to women in rural areas though the
difference between them has come down from 5.3 per cent points in 2007-08 to 4.3
per cent points in 2013-14.

Rural-urban13 inequalities in higher education are generally found to be very

high. We note from Table 3 existence of significant levels of rural-urban disparity in
gross attendance ratio in higher education in 2007-08 and 2013-14 as well. In urban
areas 23 per cent of the relevant age group population attended higher education
institutions, compared to 8.2 per cent in rural areas in 2007-08; the respective ratios
increased to 35 percent and 19 per cent in 2013-14. The difference between rural and
urban areas which was 14.7 per cent points in 2007-08 has increased marginally to 16
per cent points by 2013-14. Comparisons in the rural-urban variations in the gross

13
NSSO uses the location of the household, not location of educational institution, as the base to
classify the sample into rural or urban. Therefore, in the entire analysis here the regional (rural-urban)
classification is done according to the location of the households and not on the basis of the location of
the higher education institutions attended by the students.
attendance ratio by economic status of the households highlight some more interesting
aspects worth-noting. As one can expect, the gross attendance ratio in urban areas is
higher than that in rural areas for all expenditure quintiles in both 2007-08 and 2013-
14. The only exception is the third quintile in 2013-14. Interestingly, the extent of
rural-urban disparity in access to higher education is found to be highest for the
richest households and it is true in 2007-08 and 2013-14 as well. The rural-urban
difference in the gross attendance ratio in the top quintile was 14.7 per cent points in
2007-8 and 11 per cent points in 2013-14. We do not find much disparity between
rural and urban among the poorest – the bottom quintile. The attendance ratio in the
bottom quintile in 2013-14 was 7.9 per cent in rural areas and 10.1 in urban areas.
The rural-urban difference was 1.1 per cent points in favour of the urban population in
2007-08, which increased to 2.2 per cent points (Table 3). In case of both quintiles,
the gross attendance ratio in urban areas is 25 to 29 per cent higher than that in rural
areas. This shows that rural-urban disparities in access to higher education have
widened between 2007-08 and 2013-14 and it is more so among rich households:
inequalities between the richest and the poorest increased less in rural areas, and we
note a high degree of increase in urban areas.

The attendance in higher education also varies by type of institution. The

higher education institutions are classified into three broad categories in the NSSO
data – government, private-aided, which we refer to as ‗government-aided private‘,
and private unaided, which can be referred to simply as ‗private‘. Since government
aided private institutions are generally found to be well funded by the government and
also they follow government rules and regulations, often these two categories are
clubbed in the literature into one category under the label of ‗government‘. Besides
analysing separately we also combine here these two and present attendance rates in
higher education for all categories in Table 3 and later in Table 9. There is a
significant difference in the gross attendance ratio in higher education between private
and government higher education institutions in India in 2013-14. The gross
attendance ratio in higher education for private institutions is 7.7 per cent while it is
16.2 per cent for government institutions. These figures are 2.6 per cent and 9.8 per
cent respectively in 2007-08. Interestingly, while a higher proportion of students in all
quintiles attends private institutions than government (including government-aided)
institutions, attendance rate in private higher education institutions for the richest
households is ten times higher than the poorest households; it is only five times higher
in government institutions in 2013-14. More or less similar trend is also visible in
2007-08. It is clear that private higher education is accessible more to rich households
than the poor, partly reflecting the differences in costs of education (particularly
tuition and other fees) between these two types of institutions. Private institutions not
only charge higher levels of fees and other charges than government institutions,
students in private institutions might incur higher levels of out of pocket expenses
than those in the government institutions, as we see later. Also, a larger proportion of
the poor attend government institutions, due to the reservation policies adopted in the
government, and not in private higher education institutions.

Comparing across groups, we find the following order in access to higher

education in the bottom and the top expenditure quintiles. The order given in Table 4
is based on ratios in 2013-14. The order and the figures in Table 4 highlight a few
important aspects. In the bottom quintile, rural women are at the bottom in their
participation in higher education; and in contrast, urban women fare much better,
better than even urban males. This holds true in 2013-14 as well as in 2007-08. But
the order with respect to other groups changes between 2007-08 and 2013-14. In
2007-08 urban males in the bottom expenditure quintile were at the top. In the richest
quintile, men fare better than women in all groups, marking a big change between the
two time periods in gender inequalities in the bottom quintile.

Interestingly, all the seven lowest (in order) estimates of gross attendance
ratio in higher education listed in the table belongs to the poorest households that
shows that economic status of the household is a major barrier in accessing higher
education in India for all – men or women, rural or urban. The attendance ratio is the
lowest in 2007-08 and 2013-14 in case of women who belong to the poorest
households living in rural areas; they have multiple disadvantages of being women,
poor and rural.

We sum up in Table 5, the extent of inequalities between different groups and

the improvement or deterioration between 2007-08 and 2013-14 that has taken place.
It can be easily noted that inequality between men and women in attendance in higher
education is very low, while rural urban inequalities are high. Inequalities in access to
government versus private schools are higher; but the highest degree of inequalities
exist between the richest and poorest sections of the population. Despite some
improvement, the attendance ratio among the richest expenditure quintile is still above
6 times higher than the ratio among the bottom quintile.

For many students from disadvantaged socioeconomic backgrounds, the

challenge is not getting into college, but getting out with a degree (Conlin et al, 2007).
There exists persistent gap between the college attendance and graduation rates or
rates of completion of higher education, and this gap is higher particularly for the
students of low income families in India. Graduation or completion is a more serious
issue for the students of the poor households attending higher education, than others,
as their opportunity cost of attending colleges is higher than that of the students
belonging to well-off families. Completion or graduation rates are normally
calculated as a proportion of students enrolled at the beginning of the given course
who successfully complete it within the stipulated/recommended years of the course,
for example, completing B.Tech. course within four years. But the available data do
not allow us to estimate completion or graduation rates. Instead we can look at higher
education attainment -- percentage of adult population with higher education in the
total population.

Higher Education Attainment

While attendance ratio is a flow variable, and since all those who attend higher
education do not necessarily complete higher education – some may dropout, some
may not succeed in the final examination, or there can be fallouts for other reasons,
including mortality, this is not considered a highly reliable variable on the level of
education development, though it is extensively used due to relatively easy
availability of data on this. A better variable is ‗higher education attainment‘, defined
as percentage of higher educated population to the total population‘. This is a stock
variable that reflects cumulative growth in human capital formation through higher
education that has taken place over a period; and it is considered as reflecting better
the level of educational development.

Inequality in access to higher education finally gets reflected in the

educational levels of population. Accordingly, we find again high degree of
inequality in higher education attainment across different groups. Table 6 shows the
percentage of adult population, who have acquired higher education, by gender,
region and consumption quintiles. In the country as a whole, around 9 per cent of the
total adult population have higher education in 2013-14, which marks a small increase
in absolute terms from 6.3 per cent in 2007-08, but 45 per cent increase in relative
terms. This ratio in both time periods varies widely with the economic status of the
households. This percentage ranges from 2 percent for bottom consumption quintile to
25 percent for the top consumption quintile in 2013-14. These corresponding figures
are 0.9 per cent and 20 per cent respectively in 2007-08. Among the poor the ratio
more than doubled, while the ratio increased by 25 per cent in the richest quintile. All
this marks somewhat impressive improvement in reducing the gap, though there is
still huge gap between the top and the bottom quintiles. The 25 per cent higher
education attainment among the richest quintile in India is comparable to average
rates in some of the advanced countries of the world.

Gender inequalities are also wide in the higher education attanment both in
2007-08 and 2013-14. In 2013-14 around 11 per cent of male adult population have
attined higher education, while only 7.2 per cent among women have the sam. These
figures are 8 and 5 per cent respecteviely in 2007-08, meaning significant impovment
in case of both men and women, the latter performing rleatively better. The gender
differences by consumption quintiles revelas some interesting picture. Although
improvements are seen for both the genders among the poorest households, the
improvement is higher among women compared to men between 2007-08 and 2013-
14. The percentage of women in the bottom quintile who have completed level of
higher education was 0.35 per cent in 2007-08 which increased to around 1.2 per
cent in 2013-14, registering an increase of 3.5 times. Of the total, around 16 and 21
per cent of women belonging to highest consumption quintile have attained higher
education in 2007-08 and 2013-14 respectively; these figures are lower than those
relating to men, which are 22 and 28 per cent respectively in 2007-08 and 2013-14.
All this shows that the gender inequality in terms of higher education attainment has
decreased among the poorest groups quintile but increased among the richest top
quintile).
As in case of enrolment or attendance ratios, rural-urban disparities are higher
than gender inequalities in higher educaiotn attainment. In 2013, the higher education
attainment among the urban population was 4.5 times higher than among the
population in rural areas; inequality by gender, as we have just noted, was only 1.5
times in favour of men. The improvements made by the rural population, and thereby
in improvement in inequality between rural and urban population between 2007-08
and 2013-14 are also very small, compared to the relative improvement achieved in
gender inequality during the same period. While 2.8 per cent of the rural population
had higher education in 2007-08, the rate increased to 4.6 per cent by 2013-14 and in
case of urban population it increased from 15.3 per cent to 19.2 per cent during the
same period.
The higher education attainment among the adult population of the lowest
consumption quintile is 1.8 per cent in rural areas and the percentage of students
completed higher education in rural areas and 3.2 per cent in urban areas. These
figures were 0.8 per cent and 2.3 per cent respectively in 2007-08. Thus, the higher
education attainment among the adult population belonging to high-income families
in urban region is higher than that in rural regions for all the consumption quintiles.
The higher education attainment also varies by gender in both rural and urban regions
in both 2007-08 and 2013-14.

To briefly note, between the three groups, we note that gender inequalities are
low, but they have marginally increased between 2007-08 and 2013-13; rural-urban
inequalities are very high, and they marginally declined; and inequalities between the
richest and the poorest strata declined; but they continue to be the highest among all
the three groups. The top quintile has 13 times higher education attainment than the
bottom quintile in 2013-14, while the corresponding ratios are 4.4 between urban and
rural population and 1.7 between men and women. (Table 6).

We also look at the unequal distribution of higher educated population across

different quintiles. As shown in Table 8, the higher educated population is very
unevenly distributed. Higher educated among the poorest households constitute just
about two per cent of the total educated in the country and the richest households have
74 per cent in 2007-08 and these figures are 3.7 per cent and 62 per cent respectively
in 2013-14. However, the gap in the same between the bottom quintile and the top
quintile has come down from 72 per cent to 58 per cent between 2007-08 and 2013-
14. The narrowing of the gap is a welcome feature; nevertheless, it should be noted
that among the poorest groups the educated are very few. Secondly, the decline in the
gap is not because of any big improvement among the poor, but because of decline in
the rate among the rich. For these two reasons, the situation should be regarded as
highly unsatisfactory, requiring attention of all concerned.

The gender variations in these rates rates by expenditure quintiles reveal that
between 2007-08 and 2013-14, the difference between the highest quintile and
bottom expenditure quintiles has come down by 12.4 per cent (from 67.9 per cent in
2007-08 to 55.5 per cent in 2013-14) while it has come down by 15.9 per cent (79.1
per cent in 2007-08 to 63.2 per cent in 2013-14) among women. The gap between the
rich and the poor has reduced in case of women as compared to men in the last seven
years. There are more women who have higher education in the top quintile than their
male counterparts in 2007-08 and also in 2013-14. Highest expenditure quintile
among women accounted for 80 per cent of the higher educated in 2007-08, which
came down to 66 per cent in 2013-14, and in both years, these figures are higher
than the corresponding rates for men (71 per cent and 60 per cent respectivley in
2007-08 and 2009-10).

Again, the estimates in Table 8 reveal that there exists a significant rural-urban
disparity in the distribution of higher educated population by expenditure quintile.
The gap has narrowed down between the richest and the poorest households between
2007-08 and 2013-14 in both rural and urban regions. In 2007-08, the gap between
these two quintiles was 38.5 per cent and 85.4 per cent for rural and urban households
respectively, which declined to 22.4 per cent (rural) and 77.1 per cent (urban) in 2013-
14. Of the total number of people who completed higher education in urban areas,
merely one per cent belongs to the poorest households, while the corresponding
estimate is about nine per cent in rural areas in 2013-14.

Household Expenditure on Higher Education by Quintiles

The above discussion of inequality in access to education reveals that the rate of
participation heir education varies widely with the socioeconomic characteristics of
the households, particularly the economic status. This section examines inequality in
educational expenditure by households by economic status. It is also argued that
inequality in household expenditure can results to inequality in educational outcomes
since those who are able to pay more can access better quality higher education.
Therefore, it is quite important to look at the variations in the household expenditure
on higher education, in addition to examining the inequality in accessing it. In early
1960s, public funding and philanthropic contributions for higher education were the
major part of the resource to this sector in India and the contribution from private
sources in terms of tuition fee and other payments from students were negligible
(Tilak, 1983). With the introduction of new economic reform policies in the beginning
of the 1990s, the trend shifted towards household funding of higher education,
particularly households bearing a higher proportion of costs (Panchamukhi, 1990;
Varghese, 2013). It is being increasingly realised that ignoring the importance of
household expenditure on education proves costly for educational planning in the long
run (Tilak 2000, 2002). It may be more the case in higher education; but there are
very few studies on the subject and those few are in school education
(Panchamukhi1990; Tilak 2000, 2002). It is widely observed that the expenditure
on education is positively related to the level of household income

We examine here the variations in household expenditure on higher education.

Table 9 provides, in some detail, annual average expenditure on higher education by
economic status of the households across different expenditure quintiles by gender,
location and type of institution. At the very outset, we note that there is a significant
increase in the annual household expenditure per student on higher education; it more
than doubled from ₹14532 in 2007-08 to ₹30887 in 2013-14. High level of household
expenditure on education represents high level of inequality in education; increasingly
high levels of family expenditure suggest increasing trends in inequalities.

Evidently, the average expenditure is higher for each successive expenditure

quintile in both years across all respondents; average total expenditure on higher
education is the lowest for the poorest households and highest for the richest
households. The extent of increase in the household expenditure on higher education
between 2007-08 and 2013-14 is also highest among rich households (Rs. 27,376) and
lowest for the poorest households (Rs. 6,176). Also, in both 2007-08 and 2013-14 the
top quintile households (quintile 5) spends about 4.3 times higher on higher education
as compared to the bottom quintile (quintile 1).

Looking at gender variations in the household spending on higher education

by economic status of the households we note some interesting points. The
expenditure is higher in case of male students than female students both in 2007-08
and 2013-14 as well. Furthermore, between 2007-08 and 2013-14, the expenditure of
education on both men and women on higher education more than doubled between
2007-08 and 2013-14. (Table 9). Gender bias in favour of men in household spending
on education has been documented in many studies conducted in different regions of
India (Panchamukhi, 1990; Kingdon, 2005; Chaudhuri and Roy, 2006; Azam and
Kingdon, 2013; Saha, 2013). In a recent study, Duraisamy and Duraisamy (2016)
show that parents spend 11 percent more on the education of sons than daughters.
Here we note that households spent nearly 10 per cent higher on men in 2007-08,
which increased to 18 per cent in 2013-14. In other words the gap in the expenditure
on higher education between expenditure on men and women increased over the
years. More interestingly, it increases with the increase in the economic status of the
households. However, the gap in expenditure on men between the top and the bottom
quintile groups has remained more or less at 3 times in favour of the richest group; but
in case of women it came down from 3.4 times to 2.6 times.

The rural-urban differences in household expenditure on different levels of

education are highlighted by many scholars (e.g., Panchamukhi, 1990; Tilak, 2000).
Annual average household expenditure on higher education by location of the
households reveal that urban households spend more on higher education than their
rural counterparts and this holds true for both time periods under study. This is
understood. Similar findings were reported in a recent study by Duraisamy and
Duraisamy (2016). In 2013-14, urban households have spent 1.93 times higher on
higher education while this figure was 1.73 in 2007-08. This reveals that rural-urban
gap in the household expenditure on higher education has increased between 2007-08
and 2013-14. The annual average household expenditure on higher education in rural
areas has gone up from ₹10,420 to ₹21,728 (2.1 times) while in urban areas it
increased from ₹18,071 to ₹41,979 (2.3 times) in this period.

In 2007-08, the average expenditure varied widely between the lowest to the
highest quintile classes in both rural and urban areas. The differences in the average
expenditure on higher education between top and bottom quintiles were 4.2 times in
rural areas (₹4,343 to ₹18,488) and 2.5 times in urban areas (₹8,632 to ₹21,300). This
shows that the inequality in household expenditure on higher education by economic
status of the households is higher in rural areas than in urban areas in 2007-08. But in
2013-14, we note a change in the pattern: difference in the household expenditure on
higher education between richest and porest households is higher in urban areas than
in rural areas. The richest households in rural and urban areas have spent 3.8 times in
2007-08 and 4.1 times higher than the poorest households on higher education in
2013-14, showing that between 2007-08 and 2013-14, the variations in household
expenditure on higher education between rich and poor households have decreased in
the rural areas whereas it increased in the urban areas.

The estimates for both 2007-08 and 2013-14 show, in cofirmity with widely
known facts, that the average household expenditure on higher education is highest
for the students attending private-unaided institutions and lowest for the government
institutions. In 2007-08, the annual average household expenditure for the students
attending private-unaided higher education institutions is Rs. 27,971 while it is Rs.
8,552 for the students who are attending government institutions. These figures are
Rs. 52, 245 and Rs. 15,000 respectively in 2013-14, meaning a doubling of
expenditures in both types of institutions during this period. In 2007-08, students
attending private-unaided higher education institutions have spent 3.2 times higher as
compared to the students attending government institutions while this was 3.5 times
higher in 2013-14. This means that the difference in the household expenditure by
type of institution has marginally increased between 2007-08 and 2013-14 which is
largely due to the increasing costs of education in government aided private and
private-unaided higher education institutions. This is quite apparent because the
course fees charged in the private-unaided institutions is considerably higher than the
government institutions. However, in both government and private institutions, the
costs of higher education are increasing rapidly; and in the private institutions to
unaffordable levels for a vast majority of the poor.

We note that in 2007-08 and also in 2013-14, the average household

expenditure on higher education increases with the increase in the economic status of
the households (successively quintile 1 to 5 for students attending different type of
institutions), with the exception in private-aided institutions in 2007-08 where the
bottom quintile population spent little higher than the second quintile. This variation
was observed to be larger for those attending private higher education institutions
(both aided and unaided) as compared to government institutions. In 2007-08, the
difference in the expenditure between poorest and richest households was found to be
highest for private (unaided) institutions (4.3 times) followed by government-aided
private (3.9 times) and government institutions (2.9 times). In 2013-14, the
corresponding figures are higher: by 4.4 times, 3.4 times and 2.9 times respectively.
 In absolute terms, the household expenditure on higher education increases
with the increase in the income level, and this is true across all socio-economic
categories of students – men and women, rural and urban, attending government or
private institutions.

It is obvious that bottom quintiles spend high proportions of their incomes (or
total household expenditure on all items, as measure here) on higher education than
non-poor and rich groups. The proportion ranged from 30 per cent among the poorest
to 16 per cent among the richest in 2007-08. In 2013-14 the corresponding
proportions were 27 and 20 per cent. Table 10 gives further details by rural-urban
and male-female categories. As shown in Figure 6, there is a clear and consistent
pattern: the proportion steadily declining by increasing expenditure quintiles in every
group. The absence of intersection of lines in 2013-14 further highlight the
hierarchical pattern between different groups: in all expenditure quintiles uniformly a
high proportion of household expenditure is accounted for the education of men,
followed by urban households; then come rural households and finally education of
women. In 2007-08, the pattern was not so clear.

The information available from NSSO on household expenditure on higher

education includes the expenses under five separate heads: (i) course fees (including
tuition fee, examination fee, development fee and other compulsory payments), (ii)
books, stationary and uniform, (iii) transport, (iv) private coaching and (v) other
expenditure. We note that a major part – about 60 per cent of the expenditure of any
quintile is accounted by tuition and other fee paid to the institutions (Table 11). As
higher income groups tend to go to high fee charging private institutions, they also
spend a higher proportion of the total expenditure on fees, 63 per cent by the richest
quintile, compared to be 48 per cent by the bottom quintile. Quite interestingly, items
such as uniform, transport and private coaching account for small proportions of total
expenditure of the higher quintiles than that of low expenditure quintiles. Top
quintile spends a higher proportion on ‗other‘ expenditure, while the bottom quintiles
spend the least on ‗other‘ expenditure.

Barriers to Educational Attendance: Empirical Estimates

Students from all groups, particularly the weaker sections face several problems in
accessing higher education. The problems are more in rural areas, women face more
problems and the poor face different kinds of problems. We make an attempt here to
estimate the probability of people belonging to different social and economic groups
attending higher education.

The predicted probabilities of attending higher education is analysed for persons aged
18-23 years using logit model. The dependent variable for the logit estimation is:

HE_ATTENDANCE = 1, if the person in the age group of 18-23 is

currently attending higher education;
= 0, otherwise, i.e., if the person (of the age group 18-23) is
currently not attending higher education

The probability of attending higher education is estimated as follows:

P / 1-P = e (α + βiXi)

where,

P = probability of attending higher education

1 - P = probability of not attending higher education.

P / 1-P = odds ratio in favour of attending higher education versus not attending
higher education.

Xi = set of explanatory variables.

The analysis considers gender, region (rural/urban), social groups (caste and
religion), expenditure quintile and household size as explanotory variables. To
examine the hetrogeneity in the prdicted probabilities of attending higher education,
the regression estimates are made separately for each expendiutre quintiles. They are
made considering the characterstics of individuals such as gender, social group – caste
and religion, and location of the household. Houhsoeld size is also cosnderd as a
contorl variable. Such equations are also estimated separately by gender, and region
(rual and urban). The variables chosen for the logit mode, their notatiion and
defintions are given in Table A3 in Appendix.

The results in Tables 12 and 13 give the estimates for six major factors that
cause an effect the on probabilities of higher education of 18-23 year olds: sex,
regional (rural-urban), and religion, economic status of the household and household
size. The logit results for the entire sample (equation 1 of the Table 12) show that the
probability of an individual participating in higher education is statistically
significantly associated with majority of the predictors. Looking at the results of
equation 1, we find that the chances of attending higher education are significantly
higher for men as compared to women. This supports the findings of several other
studies conducted on Indian higher education (Dubey 2008; Raju 2008; Srivastava
and Sinha 2008; and Sundaram 2006, 2009). The location of the household
(rural/urban) matters significantly in attending higher education in India. The value of
the marginal effect associated with the variable region reveals that the individuals
residing in urban area have 4.2 per cent higher chances of attending higher education
as compared to those who belong to rural areas. The study by Raju (2008) shows
similar results as it finds that the rate of participation in higher education in urban
areas is three times higher than that of the rural areas in 2004-05. Using earlier NSSO
rounds data, Dubey (2008) has shown that the probability of female enrolment in
higher education was lower by three per cent in the rural region and 0.3 per cent in the
urban region compared to males.

The social group variable is categorized here into four different castes/classes
(scheduled caste, scheduled tribe, ‗other‘ backward classes and others), and in the
regression analysis scheduled tribe category is considered as the base (reference)
category. The results show that there is a clear hierarchy among the people, with the
predicted probability of attending higher education in terms of the social group. The
chances of attending higher education are 7.3 per cent and 11.1 per cent higher for
‗other‘ backward classes and general category respectively, as compared to scheduled
tribes. There is no statistically significant difference between the probability of
scheduled castes and scheduled tribes in attending higher education. In case of
religion, we considered only three variables, HINDU, MUSLIM and ‗Others‘. There
is statistically significant difference in predicted probabilities between Hindus,
Muslims and others, in the chances of participating in higher education. It is highest
for Hindus and lowest for Muslims. More clearly, Muslims have 10 per cent less
probability in attending higher education as compared to Hindus. There is a
significant difference in the probability of persons in different quintile groups in
attending higher education. The results show that the predicted probability of higher
education attainment increases with the increase in the economic status of the
household. For example, the probability of attending higher education (marginal
effect in Table 12) is 41.4 per cent higher for 5th quintile individuals as compared to
the poorest (first) quintile group individuals; it is 21.5 per cent for the fourth quintile,
12.3 per cent for the third quintal and 5.5 per cent higher for the second quintile. The
association between economic status of the household and participation in higher
education is positive and strong and corroborates with the findings of other students in
India (Chakrabarti, 2009; Azam and Blom 2009; Tilak 2015).

The results of the predicted probabilities of attending higher education by

gender are more or less consistent with the overall results with some differences. The
results show that urban males have 2.7 per cent higher chances to attend higher
education as compared to rural males, while urban females have 5.9 per cent higher
chances as compared to rural females. Interestingly, there is no statistically significant
difference between the scheduled caste and scheduled tribe categories for male
sample, but the scheduled caste females have significantly higher chances for
attending higher education as compared to scheduled tribe females. Also, there are
large differences between the predicted probabilities of participation in higher
education for women as compared to men in case of all religious groups. For instance,
Muslim women have 11 per cent less probability of participating in higher education
while it is 10 per cent for Muslim men. The predicted probabilities of attending
higher education for different quintile groups differ by gender. The estimates of
marginal effect show that the probability of attending a higher education institution is
higher for men as compared to women in each quintile.

Regression equation 4 and 5 (Table 13) provides the results for rural and urban
youth respectively. The results shows that there is no statistically significant
difference between the men and women in urban areas, but in rural areas women have
significantly lower chances of attending higher education than men. Similarly, there is
no significant difference in the probability of attending higher education between
scheduled castes and scheduled tribes in urban areas, while scheduled castes that
belong to rural areas have significantly higher chances of attending higher education
as compared to scheduled tribes in rural area. Although economic status of the
household matters in attending higher education for both rural and urban youth,
higher quintile groups have higher predicted probabilities of attending higher
education in urban areas than in rural areas.

Economic status of the households is generally found to have a significant

influence on the participation of students in higher education. Due to continuous
increase in the costs of education, poor students are facing difficulty to participate in
education and higher is the intensity in case of technical and professional education.
Further, the effect of the family income on the participation of education differs by
gender, social category, religion, location of the household (rural or urban) etc.
Therefore, an attempt is made here to examine the effect of individual and household
factors on the probability of participation in higher education separately for each
consumption expenditure quintile. The economic status of the household came out to
be statistically significant for all the logit results (equations 1 to 5 of Table 12) and
hence, it is important to analyse it in detail to get a better picture on the predicted
probabilities of attending higher education.

According to Raju (2008), the gap in gross enrolment ratio in higher education
between the ‗poorest of the poor‘ and the ‗richest‘ is 20 times and it is much higher in
case of women, (28 times) as compared to 16 times for men. Similarly, Tilak (2015)
found that the gross enrolment ratios are the lowest among the bottom (poorest)
quintile and highest among the top (richest) quintile; and inequalities in enrolment
ratios between the poorest and the richest quintiles have increased over the years.
Gender differences in the probability of attending higher education are found here to
be statistically significant only in case of first three expenditure quintiles which reveal
that poor households differentiate between male and female children in sending their
wards to higher education while gender does not seem to matter among rich
households. Further, the difference in the probability of attending higher education
between men and women narrows as we move from poorest to richest households.

The results show that the variable REGION (rural-urban) is statistically

significant for the top four quintiles and in all the cases the probability of attending
higher education is higher for urban households than rural households. The chances
of participation of individuals in urban areas in higher education increases as
compared to those in rural areas, when we move from 3rd to 5th quintiles. The results
reveal that for poor households, location hardly matters in sending their children to
access higher education. The study on participation of rural and urban youth in higher
education in India by Chakrabarti (2009) also arrives at a similar conclusion: children
belonging to higher income households in urban areas had 16 per cent higher chance
of attending higher education than those belonging to lower income households while
this difference is marginal for rural households.

Regression results across all expenditure quintiles show that probability of

attending higher education is significantly higher for scheduled castes, ‗other‘
backward classes and forward castes as compared to scheduled tribes (taken as the
reference category). Again, the effect of social category varies widely by quintiles.
For example, the scheduled caste population of the bottom expenditure quintile has
significantly higher chances of participation in higher education as compared to
scheduled tribes belonging to the same bottom quintile; the coefficients are
statistically insignificant for other quintiles. Non-poor or rich households (3rd, 4th, and
5th quintiles) belonging to others (Socialgrp_Other) have significantly higher chances
of participation in higher education as compared to other social groups. Similarly, the
predicted probabilities of participation in higher education vary by religion and
consumption expenditure quintiles. Muslim youth who belong to these non-poor
quintiles have significantly low probability in attending higher education as compared
to Hindus. Economic status, particularly belonging to 3rd to 5th quintiles does not
seem to matter for Muslims in declining to go for higher education or not. However,
among the bottom 1st and 2nd expenditure quintile groups, individuals belonging to
‗other‘ religions have higher chances of attending higher education than Hindus,
while it is opposite for rich households.

Conclusions

Rising inequalities in the society are indeed becoming an important concern of all.
Among inequalities in different spheres, inequalities in education, and inequalities in
higher education in particular are seen as too serious to ignore any more. Higher
education, which is an important instrument for reducing inequalities in the society,
the higher education system is characterised with increasing inequalities by gender,
social groups, regional (rural and urban) and by economic status. Using unit level
data available from the 68th and 71st rounds of NSSO surveys, conducted in 2007-08
an 2013-14, an attempt is made here to examine a few dimensions of inequality in
higher education between different social groups (scheduled castes, scheduled tribes,
other backward class, others), religions (Hindus, Muslims, Others), regional (rural and
urban), and by economic classes (expenditure quintiles, particular the poorest and the
richest). We have estimated gross attendance ratio in higher education (which is
generally considered as close to gross enrolment ratio), and higher education
attainment – percentage of adult population with higher education in the total
population. These two – the flow and stock indicators of development are considered
to be together capturing the status of higher education somewhat comprehensively. A
comparative picture on the inequalities in access to higher education is presented by
analysing the status in 2007-08 and 2013-14, the reference years of the NSSO
surveys.

In terms of both the indicators, we note that there has been significant
improvement in higher education in India. The gross attendance ratio increased
between 2007-08 and 2013-14 from 12.6 per cent to 24 per cent. While only 63 in
every 1000 adults had higher education in 2007-08, this figure has increased to 93 by
2013-14 – an increase by 48 per cent in 6-7 years. While there has been improvement
in the status in higher education of every group, the growth has not been even across
various social, regional and economic groups of population.

According to our analysis, gross attendance ratio in higher education by

economic status of the households shows wide variations. In 2007-08, the difference
in the gross attendance ratio between poorest and richest families is 29.5 per cent and
this gap has gone up to 43.5 per cent in 2013-14. This shows that the inequality in
access to higher education has increased substantially by household‘s economic status
in the last seven years.

There exist significant rural-urban disparities in gross attendance ratios in

2007-08 and 2013-14. Also, the extent of rural-urban disparity in access to higher
education is found to be highest among the richest households. Further, merely two
per cent of the higher educated belong to the poorest households and 74 per cent the
richest households in 2007-08 and these figures are 3.7 per cent and 62 per cent
respectively in 2013-14.

Inequalities between men and women have come down significantly, but at the
same time gap between men in the top expenditure quintile in urban areas and the
women belonging to the bottom quintile in in rural areas is very high – the enrolment
ratio being 56 and 7 per cent respectively in 2013-14. We have analysed gender
inequalities and rural and urban inequalities – both across different expenditure
quintiles. Both with respect to enrollment ratio and higher education attainment, the
gap between men and women is very small: the difference between two is to the
extent of 3-4 points. In contrast, the gap between rural and urban areas is quite high,
with a difference of 15-16 points. Of all, the gap between the richest quintile and the
bottom quintile is maximum: 44 per cent points in gross attendance ratio and 23 per
cent points in their education attainment (Table 14). It is also important to note that
the gap has widened particularly between the poorest and the richest sections of
population. Earlier data also showed similar trends, as Tilak (2015) in a recent study
found similar widening of inequality in accessing higher education by economic class
in India.

Participation in higher education is also related to the household expenditure

on higher education. There is a significant increase in the annual average household
expenditure on higher education (more than two times) between 2007-08 and 2013-
14. The difference between expenditure incurred on higher education by rural and
urban households is quite high: urban households spend almost doubled the
expenditure that the rural households spend. In case of education of women and men,
men spend 17 per cent than women on higher education (2013-14). Urban
households spend, on average Rs. 42 thousand per annum per student. Evidently, the
average expenditure is found to be increasing by each successive expenditure quintile
in both time periods. The bottom quintile spends Rs. 11000 per student, while the top
quintile spends nearly four times higher. Further, the estimates for both 2007-08 and
2013-14 show that the average household expenditure on higher education is highest
for the students enrolled in private-unaided institutions and lowest for the government
institutions. In 2013-14, students enrolled in private-unaided higher education have
spent 3.5 times higher as compared to those in government institutions. High level of
household expenditure on education reflects high level of inequality in higher
education.

Thirdly, the econometric analysis attempted here lead us to conclude that the
probability of an individual participating in higher education is statistically
significantly associated with majority of the predictors chosen. Men have a higher
probability of attending higher education compared to women; ‗others‘ (other than
scheduled population and backward classes), and Muslims have a lower probability,
compared to their respective counterparts. Similarly rich income groups have a
higher probability of attending higher education institutions than others. When we
estimated regression equations by each quintile, results are similar with some
important exceptions. The gender differences in the probability of attending higher
education are statistically significant only among the first three expenditure quintiles
which mean that poor households differentiate between male and female children in
access to higher education, while the rich do not. The difference in the probability of
participation between men and women narrows down as one move from poorest to
richest quintiles. Similarly, the effect of other individual and household factors (caste,
location of the household, religion) varies widely for different quintile classes. The
analysis on the barriers to access higher education in this study has largely considered
the demand side factors and does not include supply side variables due to the
limitations of the NSSO data used in this study. Therefore, an extended study, with
the inclusion of supply side determinants to access higher education, may reveal the
picture better. Recent debates on higher education in India have raised a variety of
interesting policy related issues and through this empirical study the authors have
highlighted a few of them, particularly the interaction between income inequality and
access to higher education, with the aim to facilitate a more informed policy
discourse on this.

To conclude, this study has analysed the trend and pattern of the inequality in
access to higher education among different economic classes in India and the barriers
they face in their participation in higher education. Some factors have been examined
here. Further research should unravel the factors in more detail. However, it may be
tentatively concluded that since it is not women in general, but women in the bottom
economic strata, it is not the people in rural areas, but people belong to the bottom
expenditure quintile in rural areas, who suffered most, it may be necessary to focus on
economic criteria, rather than gender, region (or even caste) in policy discourse that
aim at improvement of educational status of the population and reduction in
inequalities in higher education. Development programmes based on economic
criteria, may be difficult to implement to some extent due to relatively less reliable
data on economic/income levels of the households, but have an advantage of
committing less ‗errors of commissions and omissions.‘
References

Agarwal, P. (2009), Indian higher education: Envisioning the future. Delhi: Sage
Publications.

Azam, Mehtabul and A. Bolm (2009), ―Progress in Participation in Tertiary Education

in India from 1983 to 2004,‖ Journal of Educational Planning and
Administration 23 (2): 125-67.

Azam, M., Kingdon, G.G., (2013), ―Are Girls the Fairer Sex in India? Revisiting
Intra-Household Allocation of Education Expenditure,‖ World Development 42
143-164.

Basant, Rakesh and Sen, Gitanjali (2010), ―Who Participates in Higher Education in
India? Rethinking the Role of Affirmative Action‖ Economic & Political Weekly,
45 (39): 62-70.

Basant, Rakesh and Sen, Gitanjali (2014), ―Access to Higher Education in India: An
Exploration of its Antecedents‖ Economic & Political Weekly, 49(51): 38-45.

Biswas, Gautam, K.L. Chopra, C.S. Jha and D.V. Singh (2010), Profile of
Engineering Education in India: Status, Concerns and Recommendations. New
Delhi: Norosa Publishing House.

Borooah, V. K. (2016), The Progress of Education in India: A Quantitative Analysis

of Challenges and Opportunity, Gewerbestrasse: Palgrave Macmillan.

Borooah V. K. (2018), ―Higher Education in India: Access, Quality and Structure‖,

In: Higher Education Report 2016 on Equity in N. V. Varghese, N. S. Sabharwal
and C. M. Malish (ed.), New Delhi: Sage Publications, pp. 169-201.

Chakrabarti, Anindita (2009), ―Determinants of Participation in Higher Education and

Choice of Disciplines: Evidence from Urban and Rural Indian Youth,‖ South
Asia Economic Journal 10 (2). (July/December): 371-402.

Chakraborty, Sudip (2006) ―Demand Side Factors of Children‘s School Participation:

An Exploration with Household Survey Data for Rural North Bengal,‖ Journal
of Educational Planning and Administration 20 (2) (April): 189-204.

Chanana, Karuna (1993), ―Accessing Higher Education – The Dilemma of Schooling:

Women, Minorities, Scheduled Caste, and Scheduled Tribes in Contemporary
India,‖ in Higher Education Reform in India: Experience and Perspectives (eds.
Suma Chitnis and Philip G. Altbach). New Delhi: Sage Publications, pp. 115-54.

Chanana, Karuna (2016), ―Gender Representation in higher education: going beyond

numbers,‖ in India: Higher Education Report. (eds.: N.V. Varghese and G.
Malik) London: Routledge/National University of Educational Planning and
Administration, 2016, pp.113-134.
Chaudhuri, K., Roy, S. (2006), ―Do Parents Spread Educational Expenditure Evenly
across the Two Genders? Evidence from Two North Indian States‖, Economic
and Political Weekly, 41 (52): 5276-5282.

Conlin D. S., Ed.; Rubenstein, R. (2007), Economic Inequality and Higher Education:
Access, Persistence, and Success (Ed.), New York: Russell Sage Foundation.

Comim, F. (2007), ―Concept of Development: The role of Education,‖ in Higher

Education and National Development ed. by Bridges D., Juceviciene P.,
Juceviciene R., McLaughlin T., Stankeviciute J., London: Routledge.

Deshpande, S. (2018), ―Access and Exclusion: Mapping Social Diversity in Education

in India‖, In: Higher Education Report 2016 on Equity in N. V. Varghese, N. S.
Sabharwal and C. M. Malish (ed.), New Delhi: Sage Publications, pp. 359-379.

Dhesi, Autar S. (2000).A Quantitative Analysis of Demand for Higher Education.

New Delhi: Indian Council of Social Science Research (ICSSR).

Dreze, Jean and G. G. Kingdon (1999), School Participation in Rural India,

Development Economics Discussion Paper Series No. 18, London.

Dubey, Amresh (2008), ―Determinants of Post-Higher Secondary Enrolment in

India‖, In: Higher Education in India: Issues Related to Expansion,
Inclusiveness, Quality and Finance, University Grants Commission, New Delhi,
pp. 139-98

Duraisamy, P. (2001), ―Effectiveness of Incentives on School Enrolment and

Attainment,‖ Journal of Educational Planning and Administration 15 (2) (April):
155-77.

Duraisamy P. and Duraisamy M. (2016), ―Contemporary Issues in Indian Higher

Education: Privatization, Public and Household Expenditures and Student Loan,”
Higher Education for the Future, 3(2). 144-163.

Ghuman, Ranjit S., S. Singh and J.S. Brar (2009) Professional Education in Punjab:
Exclusion of Rural Students. Patiala: Publication Bureau, Punjabi University.

Hasan, Rana and A. Mehta (2006) ―Under-representation of Disadvantaged Classes in

Colleges: What do the Data Tell Us?‖ Economic and Political Weekly 41 (35):
3791-96.

Kaul, Rekha (1993). Caste, Class and Education: Politics of the Capitation Fee
Phenomenon in Karnataka. New Delhi: Sage Publications India Pvt Ltd.

Khan, Khalid and Sabharwal, N. S. (2012), Access and Equity in Higher Education:
Aspects of Gender, Caste, Ethnicity, Religion, Occupation and Economic Groups
in Rural and Urban Areas During Pre and Post Reform Periods, Working Paper
Series, Vol. 6, No. 4, Indian Institute of Dalit Studies, New Delhi.
Kingdon, G.G., 2005. ‗Where has all the Bias Gone? Detecting Gender Bias in the
Intra-Household Allocation of Educational Expenditure. Economic Development
and Cultural Change, 53 (2): 409-451.

Kundu, Amitabh (2018), ―Economic and Educational Inequalities in India: Access

and Outcome Indicators for Different Socio-religious Groups‖, In: Education
Report 2016 on Equity in N. V. Varghese, N. S. Sabharwal and C. M. Malish
(ed.), New Delhi: Sage Publications, pp. 63-91.

MHRD (2016) All India Survey on Higher Education (2016), All India Survey on
Higher Education report 2015-16, Ministry of Human Resource Development,
Government of India New Delhi.
(http://aishe.nic.in/aishe/viewDocument.action?documentId=227, last accessed
on 26 December, 2017)

MHRD (2017) All India Survey on Higher Education (2016), All India Survey on
Higher Education report 2016-17, Ministry of Human Resource Development,
Government of India New Delhi.
http://mhrd.gov.in/sites/upload_files/mhrd/files/statistics/AISHE2016-17.pdf

Nagarajan, R. and S. Madheswaran (2001) ―Determinants of Child Schooling in Rural

Maharashtra: A District Level Analysis,‖ Journal of Educational Planning and
Administration (15)(4): 463-85.

NSSO: National Sample Survey Organization, Report No. 532: Education in India,
2007-08: Participation and Expenditure, New Delhi.

NSSO: National Sample Survey Organization, Report No. 575: Education in India,
2014: New Delhi.

Panchamukhi, P.R. (1990) Private Expenditure on Education in India: An Empirical

Study, Unpublished report, Indian Institute of Education, Pune.

Raju, Saraswati (2008), ―Gender Differentials in Access to Higher Education‖, In:

Higher Education in India: Issues Related to Expansion, Inclusiveness, Quality
and Finance, University Grants Commission, New Delhi, pp. 79-102.

Rao, Srinivasa S. (2006), ―Engineering and Technology Education in India: Uneven

Spread, Quality and Social Coverage,‖ Journal of Educational Planning and
Administration 20(2): 205-25.

Rao, Srinivasa S. (2007). ―Neglected Terrain in the Quest for Equality: Women in
Elite Engineering and Technology Education,‖ in Women’s Education and
Development (ed. Jandhyala B G Tilak). New Delhi: Gyan Publishing House,
pp. 187-212.

Saha, A. (2013). ―An Assessment of Gender Discrimination in Household

Expenditure on Education in India.‖ Oxford Development Studies 41(2): 220-238.
Sarkar, Nivedita. 2016. The Economics of Participation in Higher Education. Ph.D.
Thesis. National University of Educational Planning and Administration, New
Delhi. (Unpublished)

Sinha, Sachidanand (2008), ―Identification of Educationally Backward Districts

Higher Education,‖ In: Higher Education in India: Issues Related to Expansion,
Inclusiveness, Quality and Finance, University Grants Commission, New Delhi,
pp. 56-78.

Sinha (2018), ―Regional Disparities in Availability, Access and Equity in Higher

Education in India‖, In: Education Report 2016 on Equity in N. V. Varghese, N.
S. Sabharwal and C. M. Malish (ed.), New Delhi: Sage Publications, pp. 131-
168.

Sinha, Sachidanand and Srivastava, Ravi (2008), ―Inclusiveness and Access of Social
Groups to Higher Education‖, In: Higher Education in India: Issues Related to
Expansion, Inclusiveness, Quality and Finance, University Grants Commission,
New Delhi, pp. 111-138.

Srivastava, Ravi and Sinha, Sachidanand (2008), ―Inter-Social Groups Disparities in

Access to Higher Education‖, In: Higher Education in India: Issues Related to
Expansion, Inclusiveness, Quality and Finance, University Grants Commission,
New Delhi, pp. 103-110.

Srinivasan, J. (2010). Demand for Education in India: Sub Regional Study. New
Delhi: Well Publication.

Sundaram, K. (2006): ―On Backwardness and Fair Access to Higher Education:

Results from NSS 55th Round Surveys 1999-2000‖, Economic & Political
Weekly, 41(50): 5173-82.

Sundaram, K. (2009) ―Fair access to Higher Education Re-Visited, Some Results for
Social and Religious Groups from NSS 61st Round Employment-Unemployment
Survey, 2004-05,‖ Journal of Educational Planning and Administration 23(2):
111-23.

Salim, A. Abdul (2008), ―Opportunities for Higher Education: An Enquiry into Entry
Barriers‖, in Higher Education in Kerala, eds.: by Nair, K.N and Nair, P.R.
Gopinath, Centre for the Development Studies.

Schultz, Theodore W. (1961) ―Investment in Human Capital,‖ The American

Economic Review 51(1): 1-17.

The Hindu, (2017), ―Income shares of top 1% surged: report‖, 14 December,

http://www.thehindu.com/business/income-share-of-top-1-surged-
report/article21665296.ece (last retrieved on 15th December, 2017).

Thorat, Sukhadeo (2008), ―Emerging Issues in higher Education: Approach and

Strategy of 11th Plan,‖ in Higher Education in India: Issues Related to
 Expansion, Inclusiveness, Quality and Finance, University Grants Commission,
New Delhi, pp. 1-26.

Thorat, S. (2016), ―Higher Education Policy in India: Emerging Issues and

Approaches,‖ India: Higher Education Report. (eds.: N.V. Varghese and G.
Malik) London: Routledge/National University of Educational Planning and
Administration, pp.15-40.

Thorat and Khan (2018), ―Private sector and Equity in Higher Education: Challenges
of Growing Unequal Access‖, In: Education Report 2016 on Equity in N. V.
Varghese, N. S. Sabharwal and C. M. Malish (ed.), New Delhi: Sage
Publications, pp. 92-130.

Tilak, Jandhyala B G (1983), ―Voluntary Contributions to Education in India,‖

Economic Analyst: Journal of the Punjab School of Economics 4 (2) (December):
61-78

Tilak, Jandhyala B G (2000), Household Expenditure on Education in India: A

Preliminary Explanation of the 52nd Round of the National Sample Survey.
National Institute of Educational Planning and Administration, New Delhi, July
(Mimeo).

Tilak, Jandhyala B G (2002), Determinants of Household Expenditure on Education

in Rural India. Working Paper No. 88, New Delhi: National Council of Applied
Economic Research

Tilak, Jandhyala B G (2002), ―Education Poverty in India,‖ Review of Development

and Change 7 (1) (January-June): 1-44

Tilak, Jandhyala B G (2003), ―Higher Education and Development,‖ in the Handbook

on Educational Research in the Asia Pacific Region (eds. J.P. Kleeves & Ryo
Watanabe) Dordrecht: Kluwer Academic Publishers, pp. 809-26

Tilak, Jandhyala B G (2006) ―Private Higher Education: Philanthropy to Profits,‖ In

Higher Education in the World: The Financing of Universities, Global
University Network for Innovation, Barcelona and Palgrave Macmillan, pp. 113-
25

Tilak, Jandhyala B G (2007), ―Post-Elementary Education, Poverty and Development

in India,‖ International Journal of Educational Development 27(4) (July): 435-
45

Tilak, Jandhyala B G (2009), ―Private Sector in Higher Education: A Few Stylized

Facts,‖ Social Change 39 (1) (March): 1-28

Tilak, Jandhyala B G (2013), ―Higher Education in the BRIC Member-Countries:

Comparative Patterns and Policies,‖ Economic & Political Weekly, 48(14): 41 -
47.
Tilak, Jandhyala B G (2015), ―How Inclusive is Higher Education in India?‖ Social
Change 45(2) (June 2015): 185-223.

Tilak, Jandhyala B G (2018), ―Preface‖ to the India Higher Education Report 2016 on
Equity in N. V. Varghese, N. S. Sabharwal and C. M. Malish (ed.), New Delhi:
Sage Publications.

Tilak, Jandhyala B G and R M Sudarshan (2001) Private Schooling in Rural India,

Working Paper Series No. 76, National Council of Applied Economic
Research (NCAER), New Delhi.

UGC (2015), Annual Report 2014-15, University Grants Commission, New Delhi.

Varghese, N.V., (2013), ―Private Higher Education – The Global Surge and India
Concerns,‖ in India Infrastructure Report 2012. Routledge, New Delhi, pp. 145-
156.

Varghese, N.V (2015), ―Challenges of Massification of Higher Education in India‖,

CPHRE Research Paper, National University of Educational Planning and
Administration, New Delhi.

Wankhede, Govardhan (2016), ―Perspective on Social group disparities in higher

education‖ in India: Higher Education Report. (eds.: N.V. Varghese and G.
Malik) London: Routledge/National University of Educational Planning and
Administration, pp. 97-112.
 Main Tables

Table 1: States/Union Territories grouped by Gross Enrolment Ratio and per

capita Net State Domestic Product, 2015-16

Gross Enrolment Ratio (GER)

High Low
Goa, Delhi, Sikkim, Chandigarh, Gujarat, Andaman &
Haryana, Puducherry, Nicobar Isles, Mizoram,
Net State Domestic Product

Maharashtra, Kerala, West Bengal

High Uttarakhand, Karnataka,
(NSDP) per capita

Telangana, Tamil Nadu, Andhra

Pradesh, Arunachal Pradesh,
Punjab, Himachal Pradesh
Jammu & Kashmir Chhattisgarh, Nagaland,
Rajasthan, Meghalaya,
Odisha, Madhya Pradesh,
Low Assam, Jharkhand,
Manipur, Uttarakhand,
Bihar, Tripura

Source: constructed by the authors, based on

Per Capita NSDP at current prices: Handbook of Statistics on Indian Economy 2016-17,
Reserve Bank of India; Gross Enrolment Ratio: MHRD (2016)
Notes: (a) NSDP per capita data was not available for the states such as West Bengal and
Tripura in 2015-16 in the Handbook of Statistics on Indian Economy 2016-17, and the NSDP
per capita for preceding year is considered.
 Table 2: Gross Enrolment Ratio in Higher Education, by Gender, Region, Social Groups (caste and religion) and Household
Consumption Expenditure Quintiles: 1983-84 to 2009-10

Gender Region Caste Religion Household Expenditure Quintiles All

----------------- ---------------- ------------------------------- ------------------------------------------------- ------------------------------------------- Population
Year Male Female Rural Urban SC ST OBC Non- Muslims Hindus Christians Others Q1 Q2 Q3 Q4 Q5
SC/T
1983-84 10.87 14.49 3.95 17.68 3.7 2.4 - 9.0 4.1 7.5 20.0 10.6 - - - - - 7.67
1987-88 11.82 5.37 4.77 19.56 4.0 3.0 - 10.2 4.4 8.8 17.0 11.4 - - - - - 18.57
1993-94 11.7 5.9 4.66 20.17 3.8 3.4 - 10.6 4.6 9.1 16.2 10.5 2.03 3.03 4.83 9.21 25.93 18.85
1999-2000 12.1 8.0 5.38 20.44 5.1 6.4 7.0 11.9 5.2 10.4 18.6 14.0 1.25 3.31 4.73 10.06 30.96 10.05
2004-05 14.8 10.4 7.51 23.79 7.9 7.3 10.1 14.6 7.6 13.2 20.8 14.7 1.80 4.10 6.11 11.87 36.75 12.59
2009-10 27.0 18.7 16.52 38.48 14.8 11.8 22.1 - 13.8 24.2 36.9 28.0 5.22 8.05 15.64 24.92 61.71 23.05

Source: Tilak (2015).

Table 3: Gross attendance ratio in Higher Education by Income Quintile, in Rural and Urban Areas and by Type of Education Institution, 2007-08
and 2013-14

Rural Urban Total Institution Type

Quintile ------------------------------ ----------------------------- -------------------------------- ------------------------------------------------------------------------
Private- Private- Government &
Male Female Total Male Female Total Male Female Person Government
Aided Unaided Private-Aided
2013-14
1 8.95 6.79 7.87 9.71 10.55 10.09 9.06 7.23 8.15 4.24 1.78 2.06 6.02
2 12.72 10.42 11.57 11.06 13.20 12.08 12.42 10.88 11.66 5.96 3.04 2.51 8.99
3 22.44 15.91 19.39 17.29 15.24 16.30 21.24 15.74 18.65 9.69 4.08 4.81 13.77
4 29.27 24.32 27.04 26.79 30.90 28.77 28.42 26.76 27.66 12.58 7.35 7.58 19.93
5 43.90 45.67 44.67 56.25 55.17 55.74 51.52 51.81 51.65 17.49 13.56 20.38 31.05
All 21.14 16.81 19.08 35.01 35.09 35.04 25.45 22.45 24.03 10.15 6.06 7.68 16.22

(2007-08)
1 3.86 1.85 2.81 4.56 3.24 3.91 3.91 1.94 2.89 1.65 0.95 0.20 2.60
2 5.64 2.99 4.34 4.83 5.85 5.33 5.55 3.31 4.45 2.28 1.33 0.73 3.61
3 7.48 4.63 6.08 9.89 8.84 9.41 7.97 5.42 6.74 3.67 1.95 1.03 5.62
4 12.99 8.93 11.05 15.12 13.36 14.27 13.72 10.47 12.16 6.92 3.36 1.64 10.28
5 29.45 22.15 26.22 32.98 39.06 35.66 31.75 33.11 32.35 13.72 9.94 8.31 23.66
All 10.14 6.24 8.23 22.31 23.73 22.96 13.87 11.15 12.56 6.04 3.75 2.58 9.79

Source: Estimated by the authors based on unit level of data available from NSSO.
Table 4: Order of Groups by Gross Enrolment Ratio among the Bottom the Top
Expenditure Quintiles

Category Gross Attendance

Ratio (%)
--------------------------
2007-08 2013-14

Bottom Quintile
1 Rural Female 1.85 6.79
2 All Female 1.94 7.23
3 Rural All 2.81 7.87
4 All persons 2.89 8.15
5 Rural Male 3.86 8.95
6 All Male 3.91 9.06
7 Urban male 4.56 9.71
8 Urban all 3.91 10.09
9 Urban female 3.24 10.55

Top Quintile
1 Rural Male 29.45 43.90
2 Rural All 26.22 44.67
3 Rural Female 22.15 45.67
4 All Male 31.75 51.52
5 All persons 32.35 51.65
6 All Female 33.11 51.81
7 Urban Female 39.06 55.17
8 Urban All 35.66 55.74
9 Urban Male 32.98 56.25

Source: Table 3.

Table 5 Inequalities in Gross Attendance Ratio

2007-08 2013-14 Change
Urban/Rural 2.79 1.84 0.95
Male/Female 1.24 1.13 0.11
Govt/Private 3.79 2.11 1.68
Q5/Q1 11.21 6.34 4.87
Note: Inequalities are measured as a simple
ratio
Govt includes Govt and Govt aided private
Source: Based on Table 3
Table 6 Higher Education Attainment (Percentage of adult population (above 15 years of
age) who acquired higher education, by Consumption Quintile, Region and Gender,
2007-08 and 2013-14

Rural Urban Total

Quintile ----------------------------- ----------------------------- --------------------------------
Male Female Total Male Female Total Male Female Person
2013-14
1 2.53 1.05 1.79 3.76 2.56 3.17 2.67 1.22 1.95
2 3.75 1.86 2.81 5.19 3.63 4.43 3.98 2.13 3.06
3 5.40 2.45 3.94 8.07 5.67 6.89 6.03 3.21 4.64
4 8.90 4.28 6.60 14.27 9.53 11.93 10.83 6.16 8.51
5 15.86 9.11 12.54 35.20 27.84 31.64 28.64 21.33 25.09
All 6.19 3.07 4.64 21.80 16.50 19.21 11.21 7.28 9.27

2007-08
1 1.29 0.26 0.77 3.02 1.64 2.33 1.40 0.35 0.87
2 1.92 0.54 1.23 2.45 1.47 1.96 1.97 0.64 1.31
3 3.15 1.15 2.15 5.06 2.99 4.05 3.49 1.46 2.48
4 5.57 2.04 3.82 9.15 5.35 7.31 6.77 3.12 4.97
5 11.84 7.26 9.61 28.07 21.46 24.96 22.39 16.28 19.48
All 3.95 1.65 2.80 17.75 12.68 15.32 7.95 4.67 6.33
Source: Estimated by the authors based on unit level of data available from NSSO.

Table 7: Inequalities in Higher Education

Attainment
2007-08 2013-14 Change
Urban/Rural 5.47 4.14 1.33
Male/Female 1.54 1.70 -0.16
Q5/Q1 22.39 12.87 9.52

Note: Inequalities are measured as a simple ratio.

Source: Based on Table 5.

45
Table 8: Distribution of Population (5+) having who Acquired Completed level of Higher
Education, by Monthly Per Capita Consumption Expenditure Quintile

Quintile Male Female Rural Urban Total

2013-14
1 4.16 2.96 8.88 1.02 3.70
2 6.47 5.44 13.80 2.07 6.07
3 12.35 10.15 21.89 6.12 11.50
4 17.35 15.34 24.16 12.64 16.57
5 59.66 66.11 31.26 78.15 62.16
Total 100.00 100.00 100.00 100.00 100.00

2007-08
1 2.72 1.20 6.16 0.57 2.15
2 3.17 1.82 7.57 0.72 2.66
3 6.63 4.17 13.21 2.74 5.70
4 16.82 12.51 28.38 9.99 15.2
5 70.66 80.3 44.68 86.00 74.28
Total 100.00 100.00 100.00 100.00 100.00

Source: Estimated by the authors based on unit level of data available from NSSO.

46
Table 9: Annual Average Household Expenditure on Higher Education by Monthly Per Capita Consumption Expenditure Quintiles (Rs.)

Government- Government &

Quintile Male Female Rural Urban Government Aided Private Pre-Unaided -Aided Private Total
2013-14
1 11147 10634 10629 12501 7715 12637 16241 9171 10922
2 13532 10085 11535 13833 8469 12958 18871 9984 11944
3 16993 12831 15068 16374 10231 14966 25914 11635 15341
4 23399 18795 18571 26065 12382 21563 34465 15766 21345
5 51680 43379 40335 51417 22928 43526 71460 31925 47876
All 33116 28094 21728 41979 15000 29677 52245 20486 30887

2007-08
1 5096 4096 4343 8632 4007 5343 7853 4495 4746
2 5198 4903 5129 4848 4219 5061 7970 4530 5091
3 6121 4691 5564 5562 4223 6079 8984 4867 5564
4 8643 7345 7308 9289 6163 7590 17590 6629 8109
5 21797 18926 18488 21300 11884 21125 34072 15767 20500
All 15080 13795 10420 18071 8552 15061 27971 11048 14532
Source: Author‘s calculations based on NSS unit level data 2007-08 and 2013-14, applying sample weights.

47
Table 10 Household Expenditure on Higher Education as % of Total Household Expenditure
Quintiles Male Female Rural Urban Total
2013-14
Bottom Quintile 29.72 23.02 24.10 28.79 26.86
2 25.58 19.54 21.23 24.67 22.97
3 23.61 17.58 18.16 22.62 20.89
4 22.31 17.31 18.04 21.62 19.98
Top Quintile 21.83 18.49 17.46 22.35 20.21
All Quintiles 23.54 18.56 18.80 23.13 21.24

2007-08
Bottom Quintile 32.65 27.40 29.56 31.16 30.62
2 30.06 20.28 28.99 23.39 25.74
3 23.03 18.91 15.29 24.09 21.33
4 17.22 15.66 12.89 18.15 16.48
Top Quintile 17.92 14.59 15.89 16.42 16.21
All Quintiles 21.90 17.25 18.18 20.63 19.74

Source: Author‘s calculations based on NSS unit level data 2007-08 and 2013-14.

48
Table 11. Household Expenditure on Higher Education, by items (%), 2013-14

Consumption Quintiles
--------------------------------------------------------------------
Items Q1 Q2 Q3 Q4 Q5 All
Course Fee 48.17 49.43 51.87 56.52 63.25 60.41
Books, Stationery &
Uniform 15.19 15.18 14.87 13.13 9.82 11.08
Transport 14.39 13.76 12.40 11.48 8.01 9.23
Private Coaching 14.71 12.15 11.65 11.05 8.89 9.64
Other Expenditure 7.54 9.48 9.21 7.81 10.04 9.64
Total 100.00 100.00 100.00 100.00 100.00 100.00
Source: Author‘s calculations based on NSS unit level data 2007-08 and 2013-14.

49
 Table 12: Predicted probabilities of attending higher education by persons aged between 18-23 years by gender and location

Variable All Male Female Rural Urban

------------------------- -------------------------- ------------------------ ------------------------ -------------------------
Coefficient Marginal Coefficient. Marginal Coefficient. Marginal Coefficient. Marginal Coefficient Marginal
Effect Effect Effect Effect Effect
Gender -0.086*** -0.016*** … … … … -0.198*** -0.035*** 0.028 0.006
(0.023) (0.004) … … … … (0.032) (0.006) (0.033) (0.007)
REGION 0.216*** 0.042*** 0.135*** 0.027*** 0.319*** 0.059*** … … … …
(0.024) (0.005) (0.031) (0.006) (0.036) (0.007) … … … …
Socialgrp_SC 0.072 0.013 0.009 0.002 0.155** 0.026** 0.168*** 0.027*** -0.058 -0.012
(0.048) (0.009) (0.063) (0.012) (0.073) (0.012) (0.061) (0.009) (0.078) (0.016)
Socialgrp_OBC 0.392*** 0.073*** 0.338*** 0.066*** 0.463*** 0.081*** 0.443*** 0.075*** 0.327*** 0.068***
(0.043) (0.008) (0.056) (0.010) (0.065) (0.010) (0.055) (0.009) (0.069) (0.014)
Socialgrp_OTHER 0.583*** 0.111*** 0.481*** 0.095*** 0.719*** 0.131*** 0.599*** 0.104*** 0.537*** 0.113***
(0.043) (0.008) (0.057) (0.011) (0.065) (0.011) (0.056) (0.009) (0.068) (0.014)
Religion_MUSLIM -0.568*** -0.104*** -0.516*** -0.098*** -0.638*** -0.110*** -0.578*** -0.094*** -0.548*** -0.114***
(0.035) (0.006) (0.046) (0.008) (0.054) (0.008) (0.053) (0.008) (0.047) (0.009)
Religion_OTHER -0.083** -0.016** -0.162*** -0.032*** 0.006 0.001 -0.049 -0.009 -0.119* -0.025*
(0.042) (0.008) (0.057) (0.011) (0.063) (0.011) (0.057) (0.010) (0.063) (0.013)
2nd_Quintile 0.436*** 0.055*** 0.448*** 0.058*** 0.419*** 0.050*** 0.426*** 0.049*** 0.477*** 0.064***
(0.057) (0.007) (0.077) (0.009) (0.087) (0.010) (0.067) (0.008) (0.113) (0.015)
3rd_Quintile 0.852*** 0.123*** 0.891*** 0.134*** 0.795*** 0.109*** 0.885*** 0.121*** 0.819*** 0.123***
(0.053) (0.007) (0.070) (0.009) (0.080) (0.009) (0.062) (0.008) (0.103) (0.013)
4th_Quintile 1.307*** 0.215*** 1.298*** 0.218*** 1.317*** 0.211*** 1.309*** 0.205*** 1.343*** 0.233***
(0.052) (0.007) (0.070) (0.009) (0.078) (0.010) (0.062) (0.009) (0.099) (0.013)
5th_Quintile 2.143*** 0.414*** 2.147*** 0.422*** 2.138*** 0.403*** 2.062*** 0.381*** 2.234*** 0.448***
(0.051) (0.007) (0.068) (0.009) (0.077) (0.010) (0.062) (0.009) (0.097) (0.012)
HH_SIZE -0.001 -0.001 0.013** 0.002** -0.019** -0.003** 0.003 0.001 -0.004 -0.001
(0.005) (0.001) (0.006) (0.001) (0.0075) (0.001) (0.006) (0.001) (0.007) (0.002)
Constant -2.298*** -2.275*** -2.408*** -2.301*** -2.113***
(0.067) (0.089) (0.102) (0.083) (0.124)
Log-Likelihood -23258.89 -13203.19 -10027.83 -12221.39 -11013.18
Pseudo-R2 0.122 0.111 0.138 0.097 0.114
Observations 41,240 41,240 22,794 22,794 18,446 18,446 23,035 23,035 18,205 18,205

Note: Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1

50
 Table 13: Predicted probabilities of attending higher education by persons aged between 18-23 years, by Consumption
Expenditure Quintiles

Variable Poorest (1st) Quintile 2nd Quintile 3rd Quintile 4th Quintile Top (5th) Quintile
------------------------ ------------------------- -------------------------- ------------------------- -------------------------
Coefficient Marginal Coefficient Marginal Coefficient Marginal Coefficient Marginal Coefficient Marginal
Effect Effect Effect Effect Effect
GENDER -0.212** -0.020** -0.187*** -0.025*** -0.211*** -0.037*** -0.064 -0.014 0.003 0.001
(0.092) (0.009) (0.071) (0.009) (0.053) (0.009) (0.048) (0.010) (0.035) (0.009)
REGION 0.148 0.014 0.193** 0.026** 0.093* 0.017* 0.173*** 0.038*** 0.304*** 0.074***
(0.109) (0.011) (0.079) (0.011) (0.056) (0.009) (0.048) (0.010) (0.037) (0.008)
Socialgrp_SC 0.351** 0.025** 0.192 0.020 0.123 0.018 0.057 0.011 -0.021 -0.005
(0.162) (0.011) (0.136) (0.014) (0.107) (0.015) (0.105) (0.021) (0.081) (0.020)
Socialgrp_OBC 0.614*** 0.050*** 0.544*** 0.066*** 0.489*** 0.078*** 0.403*** 0.085*** 0.262*** 0.064***
(0.150) (0.011) (0.125) (0.014) (0.097) (0.015) (0.096) (0.019) (0.069) (0.017)
Socialgrp_OTHER 1.039*** 0.101*** 0.768*** 0.100*** 0.725*** 0.125*** 0.579*** 0.126*** 0.432*** 0.105***
(0.166) (0.016) (0.135) (0.017) (0.101) (0.016) (0.097) (0.019) (0.067) (0.016)
Religion_MUSLIM -0.775*** -0.060*** -0.742*** -0.086*** -0.786*** -0.120*** -0.525*** -0.109*** -0.379*** -0.092***
(0.140) (0.009) (0.107) (0.010) (0.079) (0.010) (0.073) (0.014) (0.058) (0.014)
Religion_OTHER 0.847*** 0.118*** -0.069 -0.009 0.090 0.0174 -0.126 -0.028 -0.168*** -0.041***
(0.212) (0.037) (0.158) (0.022) (0.107) (0.021) (0.095) (0.021) (0.058) (0.014)
HH_SIZE 0.010 0.001 0.006 0.001 0.015 0.003 0.016 0.004* -0.032*** -0.008***
(0.015) (0.001) (0.013) (0.002) (0.010) (0.002) (0.010) (0.002) (0.009) (0.002)
Constant -2.561*** -1.965*** -1.511*** -1.080*** 0.0129
(0.168) (0.143) (0.111) (0.107) (0.083)
Log-Likelihood -1703.35 -2583.61 -4463.46 -5112.47 -9345.20
Pseudo-R2 0.0214 0.0178 0.0189 0.0127 0.0135
Observations 5,069 5,069 5,910 5,910 8,335 8,335 8,068 8,068 13,858 13,858
Note: Standard errors in parentheses; *** p<0.01, ** p<0.05, * p<0.1

51
 Table 14: Gap in Participation in Higher Education
Gross Higher
Attendance Education
Ratio Attainment
Gender (Male-Female)
2007-08 2.72 3.28
2013-14 3.00 3.93
Regional (Urban-rural)
2007-08 14.73 12.52
2013-14 15.96 14.57
Economic (Q5-Q1)
2007-08 29.46 18.61
2013-14 43.50 23.14

Source: Based on Table 3 and 6.

52
 Appendix Tables

Table A1: Gross Enrolment Ration in Higher Education (18-23 years), by States, 2015-16
All Scheduled Castes Scheduled Tribes
---------------------------------- ---------------------------- ---------------------------
State and UTs Male Female Total Male Female Total Male Female Total
A & N Islands 22.3 2.7 23.5 _ _ _ 11 13.6 12.3
Andhra Pradesh 34.7 26.9 30.8 28.6 22.4 25.5 27.4 19.8 23.4
Arunachal Pradesh 28.8 26.9 28.7 _ _ _ 34.4 33.2 33.8
Assam 16.2 14.7 15.4 17.5 16 16.8 20.8 18 19.3
Bihar 15.8 12.6 14.3 11.4 7.1 9.3 13.4 11.2 12.3
Chandigarh 48.4 70.4 57.6 28.6 37.8 32.7 _ _ _
Chhattisgarh 15.7 14.6 15.1 15.6 13.8 14.7 14.7 9.1 9.3
Dadra & N. Haveli 7.8 11.3 9.1 17.3 30.4 22.9 7.6 5.7 6.6
Daman & Diu 4.6 9.2 5.7 23.1 27.7 25.1 15.2 12.6 14
Delhi 43 48.2 45.4 30.2 28.6 29.5 _ _ _
Goa 25 30.9 27.6 27.7 26.7 27.2 17.3 24.1 20.6
Gujarat 22.9 18.3 20.7 27.7 23.1 25.5 13.4 13 13.2
Haryana 25.9 26.4 26.1 17.3 16.7 17 _ _ _
Himachal Pradesh 29.6 35.5 32.5 20 22.3 30.8 30.8 32.7 31.8
Jammu & Kashmir 23.5 26.2 24.8 13.6 17.9 15.7 10.2 8.8 9.5
Jharkhand 16.2 14.8 15.5 13.1 10.6 11.9 10.2 10.8 10.5
Karnataka 26.3 25.9 26.1 19.3 18.0 18.7 16.9 15.1 16.1
Kerala 26.6 35.0 30.8 16.4 28.5 22.4 13.6 19.2 16.5
Lakshadweep 4.1 10.2 7.1 _ _ _ 2.2 4.7 3.4
Madhya Pradesh 21.1 17.9 19.6 17 13.8 15.5 9.8 7.4 8.6
Maharashtra 31.9 27.6 29.9 31.9 27 29.6 18.1 11.4 14.7
Manipur 35.3 33.1 34.2 57.8 47.8 52.8 20.9 18.5 19.7
Meghalaya 20.4 21.1 20.8 55.3 44.3 50.1 15.7 18.4 17.1
Mizoram 25.2 23.0 24.1 192.6 96.7 158 25.6 23.5 24.5
Nagaland 14.2 15.6 14.9 _ _ _ 13.5 14.8 14.1
Odisha 21.5 17.8 19.6 16.5 12.9 14.7 10.7 8.2 9.4
Puducherry 44.2 42.1 43.2 33.2 31.7 32.5 _ _ _
Punjab 25.8 28.5 27.0 17.7 18.4 18.0 _ _ _
Rajasthan 21.8 18.5 20.2 16.7 13.4 15.2 16.9 13.5 15.2
Sikkim 36.7 38.5 37.6 36.2 22.5 29.1 20 28.8 24.5
Tamil Nadu 46.3 42.4 44.3 34.6 34.2 34.4 36.4 27.3 31.8
Telangana 39.3 33.4 36.3 38.1 34.2 36.1 39.2 28.7 33.9
Tripura 19.9 14.0 16.9 18 11.3 14.6 12.9 9.1 10.9
Uttar Pradesh 24.2 24.9 24.5 20.3 20.7 20.5 33.5 27.7 30.6
Uttarakhand 33.6 32.9 33.3 23.8 23.2 23.5 40.3 36.8 38.6
West Bengal 19.1 16.2 17.7 14.2 11.5 12.8 10.6 8.4 9.5
All India 25.4 23.5 24.5 20.8 19 19.9 15.6 12.9 14.2
Source: MHRD (2016)

53
Table A2. Gross Enrolment Ratio in Higher Education (18-23 Years) 2016-17
All Scheduled Castes Scheduled Tribes
Sl.
State/UTs ---------------------------- ---------------------------- --------------------------
No.
Male Female Total Male Female Total Male Female Total
1 Andaman & Nicobar
21.5 24.2 22.8 - - - 11.5 15.7 13.6
Islands
2 Andhra Pradesh 36.5 28.4 32.4 32.8 25.9 29.3 29.0 21.3 24.9
3 Arunachal Pradesh 29.3 28.5 28.9 - - - 31.8 30.6 31.2
4 Assam 17.9 16.6 17.2 19.5 18.5 19.0 23.9 21.2 22.5
5 Bihar 16.0 12.8 14.4 11.9 7.4 9.6 16.2 11.2 13.7
6 Chandigarh 47.3 68.8 56.1 29.7 38.4 33.5 - - -
7 Chhattisgarh 16.4 15.8 16.1 16.1 14.6 15.3 9.9 9.9 9.9
8 Dadra & Nagar Haveli 7.6 11.9 9.2 14.9 30.4 21.5 6.5 5.3 5.9
9 Daman & Diu 4.5 8.5 5.5 19.5 29.8 24.1 12.2 11.9 12.0
10 Delhi 42.8 48.4 45.3 28.9 30.7 29.7 - - -
11 Goa 25.0 31.9 28.1 23.6 26.0 24.7 19.5 25.7 22.5
12 Gujarat 22.9 17.3 20.2 31.6 21.8 26.9 14.9 12.6 13.8
13 Haryana 28.5 29.7 29.0 18.3 18.2 18.3 - - -
14 Himachal Pradesh 33.0 40.7 36.7 22.7 26.7 24.7 33.7 38.3 36.0
15 Jammu and Kashmir 23.6 27.7 25.6 13.7 18.8 16.1 11.0 10.0 10.5
16 Jharkhand 18.4 17.0 17.7 14.6 12.1 13.4 11.7 13.3 12.6
17 Karnataka 26.4 26.6 26.5 19.1 18.4 18.8 17.0 16.1 16.5
18 Kerala 28.3 40.1 34.2 17.0 30.2 23.6 15.4 21.0 18.3
19 Lakshadweep 4.1 10.6 7.3 - - - 2.1 5.3 3.7
20 Madhya Pradesh 20.9 19.0 20.0 18.3 16.1 17.3 10.4 8.9 9.7
21 Maharashtra 32.0 28.2 30.2 31.9 28.1 30.1 17.9 11.7 14.8
22 Manipur 35.3 34.7 35.0 60.9 54.1 57.5 21.0 19.4 20.2
23 Meghalaya 23.1 23.8 23.5 51.4 44.5 48.1 17.2 21.2 19.3
24 Mizoram 25.3 23.7 24.5 116.8 95.1 108.9 25.1 23.5 24.3
25 Nagaland 16.1 17.0 16.6 - - - 15.1 16.8 16.0
26 Odisha 23.0 18.9 21.0 20.1 14.7 17.4 13.1 9.7 11.3
27 Puducherry 41.8 44.5 43.1 30.9 33.0 31.9 - - -
28 Punjab 27.0 30.6 28.6 19.4 21.7 20.4 - - -
29 Rajasthan 21.6 19.3 20.5 17.4 14.5 16.1 19.5 16.2 17.9
30 Sikkim 33.9 40.8 37.3 27.7 24.9 26.3 21.4 32.7 27.1
31 Tamil Nadu 48.2 45.6 46.9 38.6 38.0 38.3 44.7 27.6 36.0
32 Telangana 38.0 33.6 35.8 34.9 33.3 34.1 37.2 28.3 32.7
33 Tripura 21.5 16.8 19.1 20.5 15.3 17.9 15.2 12.3 13.7
34 Uttar Pradesh 24.6 25.3 24.9 20.9 21.3 21.1 37.9 28.7 33.3
35 Uttrakhand 33.8 33.0 33.4 24.1 23.7 23.9 39.9 40.4 40.2
36 West Bengal 19.8 17.2 18.5 14.8 12.2 13.5 11.5 8.9 10.1
All India 26.0 24.5 25.2 21.8 20.2 21.1 16.7 14.2 15.4
Source: MHRD (2017).

54
Table A3: Notation and Definition of Variables

Notation of the Name of the variable Definition of the variable

variable

HE_ATTENDANCE Attendance in Higher Education 1, if the person in the age group

of 18-23 is currently attending
higher education
0, otherwise

GENDER Sex of the students (dummy 1, if the individual is Female

variable) 0, if the individual is Male

REGION Region 1, if the Individual‘s Residence is Urban

0, if the Individual‘s Residence is Rural

CASTE Caste of the students (dummy variables)

Socialgrp_ST Scheduled Tribe (Reference) = 1, if the student belongs to Scheduled Tribes

= 0, otherwise
Socialgrp_SC Scheduled Caste = 1, if the student belongs to Scheduled Castes
= 0, otherwise
Socialgrp_OBC Other Backward Class = 1, if the student belongs to Other Backward
Classes
= 0, otherwise
Socialgrp_OTHER Unreserved category = 1, if the student belongs to non-Scheduled
Castes, non-Scheduled Tribes and
non-Other Backward Classes
= 0, otherwise

RELIGION Religion of the students (dummy variables)

Religion_HINDU Hindu (Reference) = 1, if the student is Hindu

= 0, otherwise
Religion_MUSLIM Muslim = 1, if the student is Muslim
= 0, otherwise
Religion_OTHER Jain, Buddhist, Christian = 1, if the student is from other religion
= 0, otherwise

Expenditure
QUINTILES Economic status of the household (dummy variables)

Poorest (1st) Quintile 1st Quintile 1, if the Individual belongs to 1st Quintile
=0, otherwise
2nd_Quintile 2nd Quintile 1, if the Individual belongs to 2nd Quintile
=0, otherwise

55
3rd_Quintile 3rd Quintile 1, if the Individual belongs to 3rd Quintile
=0, otherwise
4th Quintile 4th Quintile 1, if the Individual belongs to 4th Quintile
=0, otherwise
Richest (5th) Quintile 5th Quintile 1, if the Individual belongs to 5th Quintile
=0, otherwise

HH_SIZE Household size Total number family members of the household

56
 Gross Enrolment Ratio

0
10
30
40
50
60

20
Daman & Diu
Lakshadweep
Dadra & N. Haveli
Bihar
Chhattisgarh
Nagaland
Assam
Jharkhand
West Bengal
Tripura
Madhya Pradesh
Gujarat
Rajasthan
Odisha
A & N Islands
Meghalaya
Mizoram
Uttar Pradesh
Figure 1: Gross Enrolment Ratio in Higher Education, 2016-17

ALL-INDIA

57
Jammu & Kashmir
Figures

Karnataka
Goa
Punjab
Arunachal Pradesh
Haryana
Maharashtra
Andhra Pradesh
Uttarakhand
Kerala
Manipur
Telangana
Himachal Pradesh
Sikkim
Puducherry
Delhi
Tamil Nadu
Chandigarh
Figure 2: Gross Attendance Ratio in Higher Education by Monthly Per Capita Consumption Expenditure (MPCE) Quintiles
(2007-08 and 2013-14)

60.0
Gross Attendace Ratio

50.0
40.0
30.0
20.0
10.0
0.0
1 2 3 4 5 All
Household Expenditure Quintiles

All Population (2013-14) All Population (2007-08)

58
Figure 3: Gross attendance ratio in Higher Education by Monthly Per Capita Consumption Expenditure (MPCE) Quintiles
and Gender (2007-08 and 2013-14)

60.00
Male(2013-14) Female(2013-14)
Gross Attendance Ratio

50.00
Male(2007-08) Female(2007-08)
40.00

30.00

20.00

10.00

0.00
1 2 3 4 5
Expenditure Quintiles
Source: Estimated by the authors based on unit level of data available from NSSO.

59
Figure 4: Gross attendance ratio in Higher Education by Monthly Per Capita Consumption Expenditure Quintiles and by
Region (2007-08 and 2013-14)
60.00

Rural(2013-14) Urban(2013-14)
Gross Attendance Ratio

50.00
Rural(2007-08) Urban(2007-08)
40.00

30.00

20.00

10.00

0.00
1 2 3 4 5
Expenditure Quintiles

Source: Estimated by the authors based on unit level of data available from NSSO.

60
Figure 5: Annual Average Household Expenditure on Higher Education by Monthly Per Capita Consumption Expenditure
Quintiles
By Gender All
60000 60000
Male (2013-14)
50000
Household Expenditure

Female (2013-14) 50000

Total (2013-14)

HH Expenditure
Male (2007-08)
40000 40000 Total (2007-08)
Female (2007-08)
30000 30000

20000 20000

10000 10000

0 0
1 2 Consumption
3 Quintiles4 5 1 2 3 4 5
Consumption Quintiles

60 80000
Rural (2013-14) Private-Unaided (2013-14)
50 Urban (2013-14)
Rural (2007-08) 60000 Government & Private Un-aided
Household Expenditure

HH Expenditure
Urban (2007-08) (2013-14)
40
Private-Unaided (2007-08)
30 40000
Government & Private Un-
20 aided(2007-08)
20000
10

0 0
1 2 3 4 5 1 2 3 4 5
Consumption Quintiles Consumption Quintiles

By Region by Type of Institution

61
Figure 6. Household Expenditure on Higher Education as % of Total Household Expenditure

2013-14 2007-08
31.0 35.0

29.0
30.0
27.0

25.0 25.0

23.0

20.0
21.0

19.0
15.0
17.0

15.0 10.0
Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

Male Female Rural Urban Total Male Female Rural Urban Total

62