Research Abstract A diverse workforce has long been associated with multiple firm benefits, but this is sometimes difficult to achieve due to employer discrimination. Although multiple institutional arrangements have been put in place to ban discriminatory behavior, the effects of such regulations remain relatively unexplored, often neglecting start-ups. We propose that institutional changes aiming to outlaw employment discrimination will trigger two main effects: they will (a) depress start-up founding rates through enhancement of wage-work appeal, and (b) increase the average start-up quality due to a higher threshold for leaving wage-work. We test our predictions by exploiting the staggered enactment of Employment Non-Discrimination Acts in the U.S. Consistent with our theory, we find that this institutional protection reduced the quantity of entrepreneurship but increased its quality. Managerial Summary Do laws that protect traditionally oppressed groups from discrimination in the workplace affect the propensity of workers to leave their company and start an entrepreneurial venture? Our study shows that legislative changes that ban the discrimination against members of the LGBT community in the labor market have a two-fold effect on entrepreneurship. First, they reduce the workers' propensity to start a new venture (thus reducing the loss of human capital) because of an increased relative appeal of the workplace. Second, they increase the overall quality of the startups founded in the region, increasing the overall welfare. Additionally, we find that these effects are more likely to be present when litigation against employment discrimination is more frequent and when the protected minorities are more prevalent.

Incumbents can enjoy a cost advantage vis-a-vis new entrants and so deter new firm entry by establishing and leveraging connections with the judicial system. Connected firms may in fact avoid fully complying with the costly norms implied by a law, a regulation, or a contract. At the same time, they can also credibly threaten to sue new entrants. Therefore, a change in the institutional environment that diminishes the ability of incumbent firms to establish judicial connections-i.e., an increase in judicial independence-can promote entrepreneurship. Exploiting reforms that change the way in which U.S. state judges are selected, we confirm that this is the case, and we show that this effect is more salient in states and industries where the likelihood of entering into a business dispute is higher. The paper also sheds some light on the mechanisms behind this effect.

Links with science have been argued to improve the (innovation) performance of firms. Yet we still know comparatively little about the project-level characteristics affecting the organization of such links. Our study, based on a sample of 52 projects carried out by a multinational company in the semiconductor industry, reveals that the knowledge attributes of a project help in predicting how the R&D activities will be organized. In particular, basic projects are likely to be developed through formal cooperative agreements with universities. Such projects also tend to be strategically less important. For strategically more important projects, in contrast, and for those where the knowledge to be developed is particularly novel to the firm, the firm is more likely to resort to formal contracting with a university for a specific component of the R&D project, usually early on in the project