Survey at an IIT Campus Shows How Caste Affects Students' Perceptions

In order to further the discussion of appropriate policy interventions to reduce caste inequality, we need to understand better how caste affects individuals  in their economic and social lives, how caste values affect perceptions, and the social and individual behaviours based on such perceptions that perpetuate inequality and deprivation for certain caste groups.

It is widely accepted that caste disparities continue to exist in India. After Independence, the state moved towards creating a “casteless society.”  Caste was formally abolished and a system of quotas or reservations was created for the formerly “untouchable” caste groups (Scheduled castes or SCs and Scheduled tribes (STs)) in university admissions, government jobs, and political posts.  Despite affirmative action, the gap in average socio-economic and education status between high and low caste groups remains large as research documents.   

While previous research has examined inequality in income and education, it has not explored much about whether such inequality is solely due to denial of social and economic rights and opportunities in the past, or whether there are channels of social exclusion and psychological impact of discrimination that persist in the present.  To further the discussion of appropriate policy interventions to reduce caste in¬equality, we need to understand better how caste affects indi¬viduals in their economic and social lives, how caste values affect perceptions, and the social and individual behaviours based on such perceptions that perpetuate inequality and deprivation for certain caste groups.  

This article emphasises the education gap in the caste system.  Although education is supposed to be a leveler of inequality in opportunity, access to reasonable quality of education greatly differs by caste. The fraction of population belonging to SC or ST groups gets smaller the higher the level of educational attainment (Chakravarty and Somanathan 2008).  For example, in urban India in 1999–2000, individuals belonging to the SC/ST groups constituted 18.3% of the population in the 17–25 age group, but barely 11.3% of them had completed high school (Chakravarty and Somanathan 2008).  The proportion of individuals from SC and ST groups in college graduates was only 7.4% (Sundaram 2006).   These statistics show that the education system has a long way to go towards achieving caste equality. Educational attainment is one of the channels through which caste gaps continue to persist.  But it cannot be assumed that a difference in educational attainment is the only factor that keeps individuals from disadvantaged caste groups from moving forward socially and economically.  

Once students complete their studies and enter the labour market, they are likely to face discrimination.  Deshpande (2011) shows that discrimination is common in the workplace, in part because employers value “family background.”  Madheswaran and Attewell (2007) examined discrimination in the labour market using National Sample Survey (NSS) data.  They found that in urban salaried jobs, employees belonging to SC/STs received 30% lower wages on average compared to those from other caste groups. Fifteen percent of the wage differential was unexplained by education attainment and work experience. Thorat and Attewell (2007) designed a field experiment and found that companies discriminated by caste and religion in how often they contacted job applicants who had submitted identical resumes. Banerjee et al (2007) conducted similar experiments and found lower discrimination in the call-centre industry and none in the software industry.

Using placement data of MBA graduates from Indian Institute of Management-Ahmedabad, Chakravarty and Somanathan (2008) find that graduates from the SC or ST categories get significantly lower wages than those in the general one.  This difference disappears once their lower grade point average (GPA) scores are accounted for, suggesting that the large wage difference is due to the lower academic performance of SC/ST candidates.  The authors conclude that in the absence of any serious attempt to equalise school-level education opportunities, the current policy of reservations at elite educational institutions does not suffice to equalise labour market outcomes even for the minority of SC/ST candidates who benefit from them.

In this article, we examine the academic performance of students from SCs or STs compared to other caste groups at one of the top engineering institutes, the Indian Institute of Technology (IIT) located in Banaras Hindu University (BHU).  The three caste categories, General, OBC (Other Backward Classes), and SCs/STs are in hierarchical order according to the beliefs of the caste system, with the SCs/STs at the bottom and General Caste at the top.  Caste discrimination faced by SC/STs is undeniably the worst.  Therefore, we are primarily interested in looking at the gap in performance of SC/ST students compared to general caste students although we also report comparisons with the OBC category.  A significant gap exists in the academic performance of SC and ST students compared to those in the general caste group exists.  The gap remains even after controlling for different socio-economic backgrounds of students.  

In the survey, students from the SC/ST categories report facing negative attitudes from fellow students.  A majority of students from these categories believe that students from the general caste category have higher or same academic ability as others and that those from the reserved caste category have same or lower ability than others.  Most students from the general caste have similar beliefs about inferior abilities of students from lower caste groups and superior abilities of students from general caste groups.  We posit that such beliefs and perceptions create a psychological barrier to academic performance of students from lower caste groups. 

The recent suicide by Dalit scholar, Rohith Vemula, who was a doctoral candidate at the University of Hyderabad is a poignant reminder of the humiliating impact of caste discrimination in university campuses.  According to one estimate, 18 SC students at institutes of higher education chose to end their lives in the last four years (Hindu 2016).   In an experiment involving students in sixth and seventh grades in rural Uttar Pradesh, Hoff and Pandey (2013) found no caste gap in performance of students when caste identity was anonymous.  When caste identity was made salient, there was a drop in performance of low caste students and a significant caste gap emerged favouring the general caste category students.

Our results reiterate that reservation policy in education is far from successful in levelling the playing field for students belonging to disadvantaged caste groups.  Policy intervention has to begin sooner, in the early school years, to attempt to equalise opportunities in education.  In addition, negative attitudes, perceptions and stereotypes about the ability of students belonging to the SC/ST groups are a major hurdle.  Policy should recognise how such perceptions hold back individuals and groups, and seriously attempt to think of ways to alter these.

Survey design

Students studying chemical engineering at IIT-BHU were contacted to participate in a survey online, where 121 students completed the survey.  In the survey, students were asked about their GPA for last six semesters, parental education, parental income bracket, parental occupation, type of school attended prior to coming to IIT, language of instruction in school prior to IIT (English or Hindi), and whether the student took the IIT entrance test in English or Hindi.  They were asked to indicate their caste category: general, OBC, or SC/ST.  

They were also asked about the attitudes of teachers and fellow students towards them and their opinion of the academic ability of other students from the general caste category and reserved caste category.

Table 1 lists the variables in the survey and the average characteristics of students in the sample. Table 2 describes sample characteristics by caste category.   



Students in the general caste category have higher average GPA compared to students in the OBC and SC/ST categories.  Those from SC/STs have the lowest GPA.    

While there is not a large difference in the percentage of students who attended private versus government schools between the three caste categories, those in the general and OBC category are more likely to have attended an English medium school prior to joining IIT, and more likely to have taken the entrance exam to IIT in English as compared to the SC/ST students.  

In terms of socio-economic status, students from the general caste tend to be at the highest, those from the OBC are in the middle and SC/ST students are at the bottom.  For students from the general caste, both parents are more likely to have at least college-level education, they are more likely to be in a higher family income bracket, and more likely to have a mother who works as a professional.  Students in the SC/ST category have the reverse of these outcomes.  

The difference in the GPA between students in the general caste and SC/ST caste is about 0.84 points and significant below one percent level.  The difference in the GPA between students in the OBC and SC/STs is about 0.56 points and also significantly below one percent level (Table 3, column 2).  

Some of the difference in the GPA across caste categories is likely to reflect differences in family and class background. We analysed the difference in GPA after controlling for family characteristics.  A linear regression of GPA on caste category also includes student’s observed socio-economic characteristics.  After including controls for student background variables, the difference in GPA between general caste students and SC/ST students is smaller in magnitude, but stays significant below 1% significance level (columns 3–4, Table 3).   Parents’ education and mothers’ occupations are not significantly correlated to the GPA.  Having taken the IIT entrance exam in English positively correlated with GPA (column 4) and so was having attended a government school (versus private) prior to IIT (columns 3–4).  

How do we explain the caste gap in GPA between those from general caste category and SC/ST students after controlling for family background variables? We hypothesise that caste continues to affect student’s academic performance in other ways: unobserved socio-economic and psychological factors.  

Students from lower caste categories are likely to face humiliation and harassing attitudes from others in their daily lives.  The survey tried to capture some likely channels of such psychological influences through additional questions.  They were asked how they feel about the attitudes of fellow students and teachers towards themselves.  They were asked about their perception of the academic ability of other students who belong to general caste and reserved caste categories.   

Figures 1a, 1b and 2a, 2b present the responses:

When asked about teachers’ attitudes, most students from each of the three caste categories find teachers helpful or neutral (Figure 1 and Table 4A).  Note that however 13% of students in the SC/ST caste category felt teacher attitudes towards them were hostile.   

When asked about the attitudes of fellow students, 72% of general caste students said helpful, and 28% said neutral.  In the SC/ST category, 46% said that attitudes of fellow students was helpful, 33% said that it was neutral and 21% said that it was hostile (Figure 1b).  Responses of OBC students fell in the middle of general and SC/ST categories.  Note that 21% students in the SC/ST category found the attitudes of fellow students hostile compared to none in the general category.

Students were asked their opinion of the academic ability of other students in the general caste category.  43% of students in the general category respond that it is the same as others, 55% say more than others, and 2% say less than others.   Students in the SC/ST category have a similar response:  54% say same as others, 46% say more than others, and 0% say less than others (Figure 2a).  In other words, at least 98% students from general caste category as well as in SC/ST caste category feel that the former have higher or same ability as others.  Responses of students from the OBC category are similar.

Students were similarly asked about the ability of other students in the reserved category.  In the general caste category,  61% of students said it is less than others,  and 39% said same as others.  Students in SC/ST category responded as follows:  46% said less than others, and 54% said same as others (Figure 2b).   Students in the OBC category have similar responses.  More than half the students in general caste and about half the students in OBC and SC/ST category believe that students from the reserved caste category have lower academic ability than others.  

In other words almost everyone in both the general and reserved caste categories believes that general caste students have higher or same ability as others.   And a majority of students in both caste categories believe that reserved category students are less able.  Almost no one in any of the caste categories believes that the reserved category students have more ability than others.  

It is also possible that using the phrase “reserved category” in the survey question leads the respondent to associate students in this category with lower ability.  In this case, the answer to the question about the ability of reserved category students can be attributed to both mindsets as well as the effect of using the term “reserved category” in the question and we are unable to distinguish between the two channels of effect.    


We discuss results from a survey of students belonging to general caste, OBC and SC/ST at one of the elite engineering institutes in India, IIT-BHU.  We find that GPA is lower for students from SC/ST caste category.  The difference in GPA continues after controlling for socio-economic characteristics and family background which is not surprising.  

Students belonging to SC/ST group report facing hostile attitudes from teachers and fellow students.  A majority of general caste, OBC and SC/ST caste students believe that general caste students have higher academic ability and reserved caste students have lower ability.

Despite constitutional remedies like caste-based reservation in higher education and jobs, the effects of caste continue to persist.  The reservation policy has not succeeded in levelling the playing field in higher education.   Further, income gaps exist after students’ from lower caste categories graduate and find a job (Chakravarty and Somanathan 2008).  Policy intervention in education has to begin sooner, in pre and early-school years, to attempt to level the playing field.  

Negative attitudes, perceptions and stereotypes about the ability of SC/ST caste category are a major hurdle too.  Our survey indicates those from SC/ST caste category face several reminders of their caste identity in day to day life on a campus of higher learning.  In an experiment involving students in sixth and seventh grades in rural Uttar Pradesh, Hoff and Pandey (2012) find no caste gap in performance of students when caste identity is anonymous.  But when students are reminded of their caste identity, a significant caste gap in performance emerges favouring general caste students.  Hoff and Pandey (2008) find that low caste students had internalised the values of their discriminatory system, and heeded its “narrative” which kept them from achieving outcomes comparable to those from higher caste categories.  This validates discriminatory ideology and reproduces the effects of discrimination over time.  

The good news is research in the behavioural science reveals that perceptions and attitudes are malleable.  A study in social psychology found that when African-American students were encouraged to see intelligence as something fluid and changeable as opposed to fixed at birth, they obtained higher grades and enjoyed academics more (Aronson et al 2002). In another study, students from a minority background were asked to write about a value that was important to them while those in the control group were told to write about something least important to them (Cohen et al 2006).  Those in the first group increased their academic performance relative to the control group.  The authors attribute the result to the fact that students in the first group reaffirmed their self-worth and so were able to mitigate the anxiety or stress that minority students have to deal with.   Policymaking can attempt to recognise how negative perceptions hold back individuals and groups, and find ways to change them.   Interventions can target shifting perceptions and mindsets about castes and this needs to be explored. 

Acknowledgement: We wish to thank the students who participated in the survey. We give special thanks to Surabhi Agarwal, MPhil student at University of Hyderabad, for help in survey design, implementation, and useful comments. We also thank Karla Hoff for useful comments.

Must Read

Several unconstitutional laws in India repress its citizens. A reading list from the EPW Archives.
The prevention of torture has been one of the key human rights developments in the last decade. With India’s strong stake for a seat at the security council, the issue has assumed importance. However...
Reform is long overdue in the Indian Administrative Services, but how should the government go about it? 
This analysis of electoral data from the Karnataka Chief Electoral Officer’s website and the single-person household estimates from the Centre for Research and Debates in Development Policy, New...
Whatever be the policy instrument through which we adopt the principle of net neutrality, India’s de-facto position on this will emerge only as we begin to see new services and business models tested...
Back to Top