The Social Impact Bond as a Multi-Task Principal-Agent Problem: A Qualitative Illustration of the Parallel Between Theory and Practice

Abstract: The Social Impact Bond (SIB) has become an increasingly popular tool for governments all over the world to fund social programs with private investments. Due to the complex set of interactions between parties and a limited understanding of the drivers of efficiency in those contracts, the growth of the market has not taken off as expected. The purpose of this thesis is therefore to apply the multi-task principal-agent theory on SIBs to draw predictions on the optimal incentive structure for such contracts. In order to provide an illustration of the transferability of these predictions and given the current lack of empirical data on SIBs, this thesis will further qualitatively examine the parallel between theory and practice by reviewing all 138 globally issued SIB contracts by today. The results indicate a weak parallel between theory and practice which suggests that the contracts are either inefficiently designed, or the high complexity of incorporating the government as stakeholder and the challenges of dealing with social issues, has created too flexible and simplified contracts which makes the application of a general economic theory challenging. To the author's knowledge, this paper is the first to apply the multi-task principal-agent problem to the unique context of a SIB, and the contribution of this paper is therefore this informal theoretical development. Consequently, the results have important implications for the understanding of the SIB and the development of the market as well as prepares for empirical testing to take place once more data is available.