Sources of Statistical Discrimination: Experimental Evidence from Georgia

While discrimination in the marketplace is well documented in empirical studies, field experiments have been less successful in identifying the sources of discrimination. This paper documents experimental evidence of the extent and nature of discrimination in the Georgian land market. The experiment is designed to uncover sources of statistical discrimination due to different beliefs about foreign investors. Discrimination is measured by the difference in price offers to foreign and Georgian investors. We find that the magnitude of discrimination shrinks significantly once foreign investors signal their willingness to search and pay the lease price in advance. This suggests that discrimination is largely driven by stereotypes about search costs and payment reliability of foreign investors - leaving no or very little preference-based discrimination. Knowing the source of discrimination can be helpful for policy makers to frame anti-discriminatory legislation.