Are private institutions, such as reputation, and public institutions, such as courts, complements or substitutes? Some argue that trust-based social networks that rely on reputation sap the ability of state agents to provide public goods, while others argue that strong reputation-based social networks are the key to improving state governance. Survey-based experiments conducted in Russia in 2005 and 2008 that manipulate the reputation of business partners and the ability of firms to use courts indicate that reputation has a powerful impact on trade. Courts also influence the decision to trade, but their impact is weaker. In addition, evidence from a survey experiment and a multivariate analysis finds that good courts and a good reputation are complements rather than substitutes. By using a survey-based experiment to manipulate reputation and the quality of courts, this essays mitigates endogeneity and unobservable selection problems that plague many studies of institutions.