Although entrepreneurship is generally considered positive, it is unclear if policy should dis­courage unrealistically overconfident people from becoming entrepreneurs (e.g., Parker, 2007). Despite decreasing actual profits, biased behavior can trigger information externalities relevant to entrepreneurship (Shane and Venkataraman, 2000) and ultimately create advantages for populations with biased rather than unbiased members (Bernardo and Welch, 2001). Existing studies on information externalities focus on overconfidence in privately acquired information, but ignore the effects of being overly optimistic about one’s competences and idiosyncratic risks, i.e. optimistic overconfidence. Despite the importance for entrepreneurship policies, analyses on the optimal magnitudes of such biases, their prevalence within populations, and their relations to market size, are largely absent in the literature.