Outsourcing the provision of traditionally publicly provided services has become commonplace in most industrialized nations. Despite its prevalence, there still is no consensus in the academic literature on the magnitude (and determinants) of expected cost savings to the government, nor the sources of those savings. This article considers the arguments for (and against) outsourcing and then examines the empirical evidence pertaining to whether any observed savings occur and whether they persist over time. In addition, we examine the existing evidence for the 'redistribution hypothesis' and the 'quality-shading hypothesis', which critics have used to argue that outsourcing lowers government expenditure by lowering wages and conditions and/or lower quality services. Finally, we consider the impact of contract design on outsourcing outcomes. While the power of incentives is a strong theme in economics, recent work has suggested that high-powered incentives may be suboptimal for many public sector services, because they may crowd out intrinsic motivation, particularly in instances where agents are highly motivated. We discuss the implications of this insight for the efficiency of public sector outsourcing.