Does affirmative action incentivise schooling?

Affirmative action policies in India have been shown to help bridge the gaps across caste groups. Yet, we know little about how such policies affect incentives for students that are still in school. I analyse the impact of reservations in government jobs and colleges for ‘Other Backward Classes’ (OBCs). The policies make it easier for OBCs to benefit from quotas if they meet certain educational requirements. 

India has a long history of affirmative action policies for Scheduled Castes (SCs) and Schedules Tribes (STs). Yet, in the 1990s, when the government widened the scope to include (OBCs), it was met with protests and riots and forever changed the socio-political landscape of the country. 

Affirmative action and school drop out 

Some excellent research on SC-STs already shows direct evidence of strong positive economic impacts on students who benefit from college reservations (Bertrand et al. 2016, Bagde et al. 2012). But absent from both the academic and political conversation is whether such quotas incentivise lower-caste students to stay in school for longer, given that they know they will be eligible to benefit from these polices in the future.

Different classes of government jobs require different levels of education. Simple clerical jobs may require individuals simply to finish secondary school, whereas higher-level posts may require a high school or college degree. Here, quotas make it easier for lower-caste students to make to college or into government jobs. 

The effects of such quotas on students in school are ambiguous, however. On the one hand, students may drop out early (say, at secondary school) knowing that thanks to reservations they can get a simple clerical job. On the other hand, they may stay in school for longer (and finish high school) with the hope of getting the higher-level position, which has also become easier to obtain. 

The findings

Federal job reservations and OBC educational attainment

Between 1991 and 1993, the federal government implemented reservations in government jobs. In my work (Khanna 2018), I use National Sample Survey (NSS) data to study the impact of OBC reservations on educational attainment, by leveraging variation across two dimensions: (a) I compare OBCs to non-OBCs, and (b) I compare those who were young enough in 1993 to change their schooling decisions to older cohorts who were too old to return to school.

Until about 1993, OBCs had, on average, more years of education than SC-STs but less than upper-caste students. But once job reservations for OBCs were implemented, the education gap between OBCs and upper-caste students began to shrink. 

Between 1993 and 2000, the average OBC student had attained 0.8 more years of education, over and above any gains made by the upper-caste. More students were not only getting more schooling but were also crossing important educational thresholds such as finishing middle or secondary school.

Reservations in jobs and colleges for OBCs at the state level

Even as national-level reservations in government jobs for OBCs started in the 1990s, many states had OBC quotas in jobs and colleges for quite a few years. Interestingly, the fraction of seats reserved and the size of the OBC population varies widely from state to state. Some southern states, for instance, not only had a longer history of reservation but also reserved a greater fraction of seats for OBCs (relative to the size of the OBC population). 

When I compare the gains in OBC reservations across regions, I find that the gains are more pronounced in states that reserved a greater fraction of seats. After a point, however, this relationship begins to taper off and flip signs, indicating that at very high levels of reservations, there may even be some detrimental effects on OBC education.

A case study: Haryana and the power of incentives 

In 1990, the state of Haryana established a commission of academics to identify which sub-castes were to benefit from reservations. The academics conducted a large state-wide socio-economic survey, created an ‘index of backwardness,’ and classified any sub-caste with a score above a certain cut-off to be eligible for reservations in state-run colleges and jobs. 

Before this policy was implemented, sub-castes who were just above the cut-off had similar levels of education compared to sub-castes just below the cut-off. In fact, when comparing sub-castes that just lost out from reservations to those that just made it into the benefits group, they were similar on all observable demographics. 

Using data from eight years after the policy was implemented, I find that the only thing that differed between these sub-castes was that younger members who were just above the cut-off (and hence eligible for reservation) had about 1.2 more years of education than their counterparts in sub-castes just below this threshold. It would seem that reservations had incentivised these students to stay in school for longer. 

Impacts on social groups with no affirmative action policies

Together, these results indicate that after reservations, OBCs that would expect to benefit from reservations after they finished school were more likely to get more education. Unfortunately, such gains were not seen for other social groups excluded from affirmative action policies, even though they started out at similar levels of education as the OBCs. Muslims did not catch up with the non-Muslims, and the poorer sections of upper-castes were still lagging behind the rich. 

This may support the notion that the improvements in educational attainment by the OBCs were unlikely to have been from other policies that targeted all social groups. For instance, there were no additional gains for OBCs in regions that spent more money on schooling infrastructure. Furthermore, poorer OBCs experienced more than double the gains of their richer counterparts. This result could be explained by the fact that the policy prohibits OBCs whose parents are in certain professions, and above a certain income threshold, from benefiting from reservation. 

Implications for government policy

Policymakers have a few tools in their affirmative action toolbox: 

• whether or not to implement reservations, 
• what fraction of seats to reserve, and 
• which sub-castes should be made eligible. 

While affirmative action policies may incentivise students to stick around in school longer, there are a number of important considerations before quantifying the costs and benefits. For example, larger class sizes and various impacts on the quality of education (due to peer effects and other factors) need to be kept in mind. Furthermore, if there are not enough government jobs, reservations may not be meaningful and have detrimental effects on excluded social groups (including the upper-castes) who now have access to fewer seats. 

For a while, these policies excluded certain other groups (like poorer sections of the upper-caste) who also fare poorly on socioeconomic indicators. This year, the Indian government has considered expanding reservations to low-income candidates from upper-caste groups, but a restricted number of government jobs led some commentators to suggest that these may simply be pre-election political campaign promises. As such, the underlying politicisation of these issues by various groups is unlikely to be in everyone’s best interest. 

As researchers, to understand what policies are best for bridging the opportunity gap, we need to compare many different types of social welfare policies. For now, however, it is clear that affirmative action in the 1990s helped OBC students attain more years of education and look forward to better job prospects in the future.

References

Bagde, S, Epple, D and Taylor, L (2016). “Does Affirmative Action Work? Caste, Gender, College Quality, and Academic Success in India”, American Economic Review 106(6).

Bertrand, M, Hanna, R and Mullainathan, S (2010). “Affirmative action in education: Evidence from engineering college admissions in India”, Journal of Public Economics 94(1-2): 16-29.

Khanna, G (2018), "Does Affirmative Action Incentivize Schooling? Evidence from India", Working Paper.