The aim of this thesis is to examine the interaction between policy and implementation choices, by tracing the adoption of outcome-based payments in UK welfare-to-work policy. New Labour’s flagship welfare-to-work programme, the New Deal for Young People, was delivered by civil servants and private and voluntary sector providers working under highly-specified, fee-for-service contracts. The programme was then reformed to introduce elements of outcome-based payment, rewarding job placements. The Coalition government’s Work Programme extended this approach further. The use of financial incentives was presented by government, and largely accepted by the policy community, as a measure to improve efficiency and effectiveness in achieving policy outcomes. However, the literature based on transaction cost analysis and principal-agent theory has shown that the use of high-powered incentives in contracts to provide complex public services involves sacrificing some policy goals, even while it rewards others. The thesis uses process-tracing methodology to examine the factors which led successive governments to switch towards outcome-based payments. It finds evidence that the shift was only possible because of significant changes in policy, and that the outcome-based approach addressed political concerns beyond the aims of achieving efficiency and effectiveness in welfare-to-work policy.