Is cricket any match for India’s caste bias?

Caste-based discrimination continues to be widespread in India even though it was outlawed almost 70 years ago.

Members of the Dalit community, the 300 million people at the bottom of India’s caste ladder, have to put up with the worst of the consequences. They often face public humiliation and violence, lack access to good education, and are made to do the lowest jobs available.

India’s affirmative action policies have succeeded to some extent in empowering them, but it hasn’t weakened the pull of caste. Consider this: A Pew Research survey in 2021 found that Indians, on average, prefer to be friends with people from their own caste. “About one-quarter (24%) of Indians say all their close friends belong to their caste, and 46% say most of their friends are from their caste,” the survey found.

One way to reduce prejudice is by promoting interaction among people from different caste backgrounds. But many social psychologists have argued that effects of such measures depend on whether people are collaborating to achieve a common goal or are competing against each other.

A study tested this theory by organising cricket tournaments for boys and young men from different castes in Uttar Pradesh, a state where 39% of the upper and backward caste households practice ‘untouchability’ as per the India Human Development Survey-II. The popularity of cricket in India and among the study participants made it a common and natural source of social contact.

A total of eight cricket tournaments were organised between January and July 2017, one in each of the eight gram panchayats selected for the study.

The study was done in partnership with J-PAL South Asia and the Sarathi Development Foundation.

Matt Lowe and his team recruited 1,261 boys and young men aged between 14 and 30 years, of which 800 were randomly assigned to play in the tournaments. The remaining 461 served as back-up players. A subset of players of these back-up players were “low priority” back-ups - they served as a comparison group as they had the least interaction with participants from other castes.

Almost a third of the participants (35%) in every league were randomly assigned to teams that had players just from their caste. The remaining players were put in mixed-caste teams with varying caste composition. Each team played eight matches against randomly selected opponents.

This design helped the research team test three things. First, whether simply participating in the league and having contact with other caste groups reduced discrimination. Second, whether playing on the same team with members from other groups increased inter-caste bonhomie. Third, if playing against members of a different caste increased feelings of prejudice.

The interactions between the participants were recorded for each match. Within a month after each league ended, the researchers followed up with the players with additional surveys and exercises to examine if the tournament had any effect on their willingness to interact and make friends with players of different castes.

They also measured participant’s favouritism for their own caste by asking them to vote anonymously for one member of each team to receive professional cricket coaching. They were told to vote for the player they most preferred, which allowed Lowe and his team to see whether the voting was based on caste or sporting abilities.

Lastly, the researchers checked for the willingness of the participants to trade with people from other castes in the absence of strong financial incentives. They did this by giving each player a pair of mismatched gloves and flip-flops after the league ended. Each pair came with a coloured sticker that was a strong marker of the recipient’s caste. The players were offered money for making a successful trade with other participants to get a usable pair. Half of the players were randomly selected to receive a bonus if they traded for an object having a different coloured sticker.

A key finding of the study is that simply participating in the league succeeded in reducing caste bias, but it also confirmed what social psychologists have long theorised: The nature of interaction matters.

The effects of the cricket tournaments were stronger among participants who experienced collaborative contact. In other words, those who played in mixed-caste teams.

Collaborative contact increased cross-caste friendships and reduced favouritism towards members of their own caste during the voting exercise.

The degree of caste diversity in the teams was an important factor as well. Participants playing for teams with greater caste mix had more friends and were more likely to play in future with people from different castes.

Furthermore, cross-caste friendships persisted even when teams were not winning matches together, suggesting that these friendships were formed because players were working together.

Compared with the participants who experienced collaborative contact, playing against members of other castes - or adversarial contact - led to fewer cross-caste friendships.

Cross-caste trade was observed to be similar for both collaborative and adversarial contact, but it was higher for players from mixed-caste teams who received a bonus. This suggests that collaborative contact may not work in the absence of financial incentives for trading.

Participants who experienced adversarial contact also chose more players from other castes as teammates to play with in future. This indicates that they recognised the sporting abilities of other-caste members and were willing to play on the same team. However, this is only true when players were asked to choose teammates for a match with a reward for winning. If there were no stakes to a match, players chose fewer other-caste players.

This study provides evidence to suggest that contact with other social groups can reduce bias. However, as the study shows, just bringing groups members in contact with each other may not be sufficient and in some cases may even worsen relations.

It is collaborative contact, where people work together towards a common goal, that can reduce caste divisions. Local, short-term sports programmes can be one effective way to achieve that in India.

More generally, affirmative action policies that lead to more interaction between disadvantaged groups and dominant groups might improve relations if such interactions are collaborative in nature.

The paper can be accessed by clicking here.

(The study has been conducted by Matt Lowe, assistant professor of Economics, University of British Columbia)