Innovation in public service delivery: essays on the social impact bond

Following the requirements of an article-based dissertation, this thesis offers three separate articles that explore the effect of the social impact bond (SIB) model on processes of public service delivery. SIBs are an outcomes-based approach to funding social services, where typically a private investor gives upfront capital to a service delivery organization and is then paid back by the government based on the achievement of outcomes. Despite over ten years of experimentation, there is a critical gap in our understanding of how SIBs work in practice and the effects of the model on public service delivery. This thesis draws on theories of new public management, i.e. the incorporation of private sector mechanisms in public service delivery, to investigate various aspects of the ‘SIB effect’ related to financial incentives, performance targets, risk transfer, multi-sector collaboration, and provider flexibility. In the first paper, I offer a conceptual framework (based on a review of the SIB literature coupled with themes from new public management) for how SIBs might impact public service delivery. I then apply this framework in a case study of a SIB launched in the UK, using in-depth interviews, document review, and observation of front-line staff. I find that the SIB had mixed effects in this case; it enabled rigorous monitoring of performance, availability of additional resources for the service, and outcome goal alignment across stakeholders. However, the SIB model also generated stakeholder conflict, feelings of resource burden and disproportionality by the government partner, loss of democratic accountability, and the pressuring presence of an underlying financial driver. In the second paper, I investigate whether using a prescribed and replicable (i.e. high-fidelity) intervention insulates providers and front-line staff from potential negative effects of the SIB model (such as perverse incentives for outcome gaming). After laying out a framework for how the SIB model might affect providers and front-line staff, I draw on a case study (using in-depth interviews, participant observation, and document review) to analyze how these effects change under a high-fidelity intervention. I find that while using a high-fidelity intervention with the SIB model mitigates incentives to game outcome payments, it also restricts the ability of the provider to innovate and deliver flexible, context-driven services. Using a high-fidelity intervention does not protect providers from pressure to meet output targets, which could potentially create other disincentives in service delivery. In the third paper, co-authored alongside Drs. Eleanor Carter and Mara Airoldi, we investigate how SIB projects have been designed in relation to two of the model’s original promises: shifting the focus of public service delivery to achieving impact and transferring risk of poor performance to external investors. By conducting qualitative content analysis on the SIBs launched to date in the US and UK, we analyze how the design of SIB projects varies across two key dimensions: degree to which payment is linked to impact and nature of the working capital. We find that SIB design in practice has deviated from the model’s original intent in several important ways. In the UK, but not the US, outcomes that projects pay out for are generally not subject to rigorous validation methods, calling into question whether payments are meaningfully linked to impact. In both the US and the UK, risk mitigation strategies in the repayment structures may limit the intended transfer of risk to external investors. Through these three papers, I make original and substantive theoretical and empirical contributions to the debate on how social impact bonds affect processes of public service delivery. I do this by: 1) providing an original conceptual framework from which to analyze the effect of SIBs on systems of public service delivery, 2) laying out the conceptual arguments for how the SIB model could affect provider and front-line staff, and extending this conversation to include differences between types of interventions, 3) furthering the field’s empirical understanding of how the SIB model impacts public service delivery and front-line staff through an in-depth case study, surfacing a unique and original discussion of the risks of disproportionality, and 4) offering the first systematic investigation of SIB design over time, comparing the SIB model in practice to its original promises. These contributions have practical implications for policymakers deciding how best to deliver public services