Experimental Design
We have recruited 412 boys (270 Hindus and 142 Muslims) aged 13 to 18 in West Bengal, and randomly assigned them to:
1. Regular Camp (N = 120) – free 12-day camp. The camp will include talks and games designed to impart democratic and universalist values.
2. Ritual Camp (N = 120) – free 12-day camp held concurrently with the Regular Camp. This camp will include the same activities, but will additionally contain “Durkheimian” ritualized features (flag-raising, synchronized singing and dancing, common uniforms, etc.)
3. Control Group (N = 172) – these boys were not invited to attend either camp.
The two camps will take place from December 19th to 31st, with a day off on December 25th. They will take place in the same location (a stadium), with one camp taking the morning session each day, and the other camp taking the afternoon session – alternating each day.
Our primary interest is in comparing endline outcomes between these groups, with (1) vs. (3) capturing the effect of the camp, and (1) vs. (2) capturing the effect of collective rituals. We will also pool (1) and (2), comparing with (3), for a higher-powered test of the effects of the two types of camps.
Those assigned to the camps have been randomly assigned to groups of 10 (24 groups of 10 in total, across the two camps). Most of the camp activities will be done in these groups. For each camp, groups have been randomly assigned such that:
- 6 groups have 5 Muslims and 5 Hindus each
- 6 groups have 2 Muslims and 8 Hindus each
This cross-cutting variation allows us to test for the effects of intergroup contact, and explore whether these effects differ by whether the camp has rituals or not.
For our analysis, we will run OLS regressions with randomization strata fixed effects, and a baseline-measured dependent variable when available. We will use robust standard errors when the unit of observation is the individual, and we will cluster standard errors at the individual-level when there are multiple observations per individual.