Contracting-out, a way to reform governmental activities, is fundamental to contemporary governance. Work previously done directly by government employees and organizations is shifted to contractors, usually firms or non-profit organizations. This means a substantial translocation of resources and jobs required for the activities in question. Such significant redistribution often leads to political struggle rather than technical optimization. To uncover these politics, the dissertation analyzes eleven cases, from the UK, Sweden and Denmark in the 1980s and 1990s, in which contracting-out happened to greater or lesser extents. The cases come from three sectors of government activity: janitorial service, space provision and data management. The case studies are based on a large number of original interviews with key decision makers and participants from the three countries and sectors, findings from newly uncovered quantitative data, and archival records. The case studies show that well-regarded theories which predict the extent of contracting-out in fact fall short. Neither transaction cost economics, theories of partisan politics, nor models of the political economy of nations predicts the extent of contracting-out. How, then, can we explain when governments can contract-out, and when they cannot? Framing -- how governments justify contracting-out -- significantly affects the reform's success. Contracting-out is more likely when it is hitched to broad social or national goals. But when the reform is framed narrowly, as the goal in itself, it is less likely to occur. To contract-out, governments must also navigate challenges that vary primarily by sector. Contracting-out is difficult in sectors where interest groups can exert great influence and will not be happy with the reform in any shape. But contracting-out is likely in sectors without significant interest group stakes, if interest groups are excluded by the structure of decision-making, and if the shape of the reform can be molded to avoid harming the interests of mobilized groups.