Backlash or Bonding?: Effect of Integration Policies in the Context of Affirmative Action

Last registered on August 29, 2024

Pre-Trial

Trial Information

General Information

Title
Backlash or Bonding?: Effect of Integration Policies in the Context of Affirmative Action
RCT ID
AEARCTR-0013944
Initial registration date
July 18, 2024

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
August 29, 2024, 11:33 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Primary Investigator

Affiliation
Paris School of Economics

Other Primary Investigator(s)

Additional Trial Information

Status
In development
Start date
2024-07-21
End date
2025-12-31
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
India has a long history of implementing Affirmative Action (AA) policies, commonly known as reservations, which explicitly favor historically disadvantaged groups. Evidence suggests that while AA policies benefit disadvantaged groups, they may come at a cost to advantaged groups (Bertrand et al., 2010; Gulzar et al., 2020). Explicit Affirmative Action policies may imply trade-offs not only due to efficiency concerns but also can potentially induce behavioral responses from both advantaged and disadvantaged groups, in highly competitive settings like admission in public universities. In such settings, the impact of integration policies aimed at reducing behavioral responses, such as mixing students in hostel rooms to increase exposure to different caste groups, remains unclear. On one hand, such policies can reduce misperceptions about the disadvantaged groups, thereby decreasing prejudice and stereotypes (Boisjoly et al. 2006; Carrell, Hoekstra, and West 2019; Rao 2019; Corno et al. 2022). On the other hand, mixing can make stereotypes salient and perpetuate existing tensions, particularly in competitive contact (Lowe 2021; Tabellini 2020).

For this project, I will examine the impact of increased exposure to students from different social groups in public engineering institutes in India, induced by an integration policy that allocates students to hostel rooms. I pose two sets of questions, i) What is the effect of roommates' identity on academic and job market outcomes? In high competitive settings, do interactions with the advantaged students induce negative outcomes due to stereotypes and reduced self confidence? ii) Do social interactions with students from different social groups change attitudes towards AA policies and stereotypes, particularly for the advantaged groups? and do these attitudes play a role in explaining the effects on mental health of the disadvantaged groups?
External Link(s)

Registration Citation

Citation
Vanukuri, Balasai. 2024. "Backlash or Bonding?: Effect of Integration Policies in the Context of Affirmative Action." AEA RCT Registry. August 29. https://doi.org/10.1257/rct.13944-1.0
Experimental Details

Interventions

Intervention(s)
This project utilizes a natural experiment in which students are quasi randomly assigned rooms in a public engineering college in India. These institutes are fully residential, and students once assigned live in the same hostel until they complete their undergraduate degree.

Room allocation in the university follows two steps, first hostel allocation and second room allocation within hostel. Hostel allocation for first-year students is determined by an algorithm that results in quasi-random allocation of roommates from different groups. There is no formal mechanism to change rooms or hostels.
Intervention Start Date
2024-07-22
Intervention End Date
2025-06-30

Primary Outcomes

Primary Outcomes (end points)
Academic Performance, Job market outcomes
Primary Outcomes (explanation)
Academic Performance - GPA, number of failed courses, Job market outcomes: No of internships applied, No of internships done, Internship Stipend, first placement after college; Objective and Subjective measures of skills: Programming, English, Math : college grades and also self reported skill levels.

Secondary Outcomes

Secondary Outcomes (end points)
Mental Health, Opinions on redistributive policies , Academic Effort, Aspirations, Interactions with roommates, Participation in college clubs,
Secondary Outcomes (explanation)
Mental Health: PHQ-4 Index and Relative happiness measures, Aspirations: Expected Salary at the end of graduation, higher education/foreign mobility, Opinions on redistributive policies: Support policies specific to India, overall attitudes related to inequality and poverty, Academic Effort: No of hours spent studying.

Experimental Design

Experimental Design
This project utilizes a natural experiment in which students are quasi randomly assigned rooms in a public engineering college in India. These institutes are fully residential, and students once assigned live in the same hostel until they complete their undergraduate degree. Treatment is exposure to students from other social groups (disadvantaged vs advantaged groups), more specifically living with someone from different social category than yours in during the first year of hostel/college life. Typically a hostel room contains 2 or 3 students.
Experimental Design Details
Not available
Randomization Method
Randomization done in office by a computer
Randomization Unit
Hostel rooms
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
1400 hostel rooms
Sample size: planned number of observations
3500 Students
Sample size (or number of clusters) by treatment arms
9 hostels, 3400 students, approximately 1300 rooms: 650 homogenous social background rooms (1700 students), 650 mixed background rooms (1700 students).
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
I can detect a 0.1 standard deviation, difference between students form most disadvantaged social group living with their own group compared students living with students in the middle social group. A 0.24 standard deviation difference can be detected when most disadvantaged students live in the most advantaged students and finally a 0.22 deviation difference when students form the most disadvantaged social group living with all the other students. These calculations are more with ANCOVA specification with 50% of the variation is explained by the baseline controls at 80% power.
IRB

Institutional Review Boards (IRBs)

IRB Name
Paris School of Economics
IRB Approval Date
2024-07-11
IRB Approval Number
2024-027