Abstract: For decades, the bundling of research services into commissions paid for the execution of securities trades has been the focus of policy discussion and academic debate. The practice whereby asset management firms use investor funds to cover research costs, known as “soft dollar” payments in the United States, resembles a form of kickback or self-dealing. The payments allow asset managers to use investor funds to subsidize the cost of their own research efforts even though those managers charge investors a separate and explicit management fee for advisory services. Why do soft dollars exist? Over the years, defenders of the practice have argued that soft dollars mitigate principal-agent problems between the investment manager and the broker, improve fund performance, and provide a public good in terms of the increased production of research on public companies. This Article evaluates these theoretical law-and-economics arguments through the lens of empirical academic research done in the past as well as an emerging new body of empirical studies exploring the impact of MiFID II, an E.U. Directive that severely restricted the use of soft dollar payments in European capital markets as of January 2018. The weight of empirical evidence suggests that the arguments in favor of soft dollars are not robust. MiFID II’s unbundling of commissions appears to have, on balance, improved European market efficiency by eliminating redundancy and producing information that is of greater value to investors.