This thesis explores the effects of consumer observational learning on the darknet markets. To accomlish this a unique micro dataset is created from the Agora market. By exploiting variations in how sales data are presented a novel IV strategy is developed to estimate the causal impact of vendor’s previous sales on their future rate of sales. The attained estimates are zero after sensitivity checks and I can not produce robust estimates of a causal effect. A difference in differences estimation is explored with similar results.
In addition to the estimation of obervational learning effects this thesis also presents an introduction to the darknet market ecosystem and suggests methods for parsing darknet market scrapes into usable datasets.