When Linda meets Preeti: The Validation of Behavioral Biases in India

This paper presents results from an experiment testing 10 of the core biases from the behavioral economics literature amongst two distinct ‘non-WEIRD’ (Western Educated Industrialized Rich and Democratic) population groups: low-income Indians, and university students from an elite Indian university. The study tests for both the existence of the ‘behavioral bias’ for each measure with our ‘non-WEIRD’ sample and tests for heterogeneity across the socioeconomically distinct sub-samples. We find that both sub-samples display significant 'bias' in the majority of tests and across different categories of bias, suggesting that behavioral biases are not peculiar to Western samples. We further find that the patterns of bias are the same for each sub-sample for most measures, but that there are notable exceptions for a small subset of measures. In most of these cases, the student sample, closer to typical samples for this type of research, shows stronger bias than the low-income sample.