This dissertation includes an analysis of sentencing reform in the United States federal system and England and Wales as well as an analysis of more general criminal law reform in Chile which converted its criminal process from one that was predominantly inquisitorial to one that is more adversarial in nature. By focusing on the effect of select instances of criminal law reform on lower court decision-making, the dissertation includes an analysis of whether lower courts are responsive to legislation and higher court mandates and whether the intent behind criminal law reforms was achieved or whether the reform resulted in unintended consequences. The dissertation includes empirical analyses of how lower court judges respond to limitations or expansion of their discretion imposed by the legislature and sometimes higher courts. In the U.S. federal example, case level data as well as opinions of judges from surveys conducted in 1991 and 2008 are analyzed. The analysis of sentencing reform in England and Wales relies on a content analysis of the reform, interviews with judges and sentencing officials, and statistical information on various prison populations from 1983 to 2007. In the case of Chile, the effect of reform on lower court judges is tested using data about rates of convictions, acquittals, and case processing times in lower courts. Although focused on criminal law reform, the dissertation describes the relationship between some higher law-making body (the principal) and lower courts (the agents). The principal-agent framework is used to explain why higher law-making bodies delegate to lower courts in the first place and to explain the risks involved in delegating discretion to lower court judges whose preferences may differ from the higher law-making bodies. Finally, the dissertation explores the political nature of criminal law reforms by comparing the political motivations behind reform with the actual consequences of such reform. Often political objectives, such as appearing tough on crime and reducing prison populations, are mutually exclusive. As a result, the political nature of criminal law reform prevents legislators from creating cogent or effective criminal law policy. Further, in reducing judicial discretion, a cornerstone of much of the reform analyzed here, legislators are discounting and weakening their most experienced and effective agents in the battle against crime