Many developing countries are considering curtailing legal remedies available to investors in bilateral investment treaties (BITs). This change will likely benefit developing countries by restoring a portion of their sovereign autonomy, but perhaps at the cost of a decline in foreign investment. As with any trade or investment policy, there are two primary challenges to evaluating whether and how the strength of legal remedies in a BIT affect investment: (1) studying a change in policy at the provision level requires provision-level information on a large number of BITs, and (2) investment policy is likely correlated with unobserved drivers of investment. To address the first challenge I introduce a new comprehensive database, created by me in partnership with the United Nations Conference on Trade and Development (UNCTAD), that contains provision-level information for over 2,500 BITs. I also identify a natural experiment to address the second challenge: an arbitration decision that endowed some investors with new and stronger legal remedies through an unanticipated application of the "most favored nation" principal. Using this database and natural experiment, I present robust evidence that stronger legal remedies in a BIT do not lead to more investment. I also present suggestive evidence that stronger legal remedies imposed on a host economy by an arbitration tribunal may lead to a decline in investment as host economies react to their increased exposure to arbitration and tighter constraints on their regulation of foreign capital.