Navigation – Plan du site

Social Impact Bonds: A New Public Management Perspective

Benjamin Le Pendeven

Résumés

Les Social Impact Bonds (SIBs) sont un nouvel outil financier pour les projets socialement innovants. Ce concept apparut en 2010 en Grande-Bretagne avec un programme de réinsertion pour les détenus de la prison de Peterborough. Depuis, cette technique financière s’est développée dans de nombreux pays dans le monde. Dans cet article nous avons analysé les SIBs avec le cadre du Nouveau Management Public (NPM) proposé par Dunleavy et al. (2005). Au travers de ce cadre analytique nous avons essayé de comprendre dans quelle mesure ce nouvel outil financier vient ou nous s’inscrire dans le cadre du NPM au regard de ses trois principales catégories : le démantèlement, la compétition, et l’incitation. Pour cela, nous avons mené 62 entretiens avec des praticiens et experts des SIBs dans des organisations publiques, privées et sociales de plusieurs pays. Les entretiens furent menés los d’une période particulièrement riche, lors de laquelle les premiers SIBs anglo-saxons arrivèrent à terme et où le gouvernement français décida de massivement investir le sujet avec un appel à projets dédié.

Haut de page

Texte intégral

Introduction

1Social Impact Bonds (SIB) first appeared in 2010 in the United Kingdom in the wake of the “Big Society” initiative, a political platform launched by David Cameron when he was leader of the Conservative Party. The underlying idea was that information technology and innovations in public management can empower citizens and help them monitor the efficiency of public services. The government would thus be able to transfer the supervision of parts of current public services to civil society or the private sector.

2An SIB is a new financial tool designed to provide a solution for social issues: homelessness, rehabilitation of former prisoners, health, academic failure, etc. It involves cooperation between financial actors (banks, investment funds, or charities), the public sector (central government or local authorities) and social services providers (NGOs, social businesses or regular private firms). Its contract-based structure is intended to help individuals on the brink of social exclusion. In other words, parties involved in an SIB pull together their resources and expertise for one common purpose: tackling social issues in an innovative fashion. The Social Impact Bond is a financial innovation for solving social issues, but not in itself a social innovation or a social service. It is an innovation because it is the first mechanism combining a payment-by-result system based the achievement of social objectives while saving social costs for the public sector, private financing, and the outsourcing of social actions to non-public organisations. This kind of consortium had never before been used for public services.

3Unlike many financial tools employed in the social economy, SIBs can generate profits for the risk-taker. While not the only financial solution with this characteristic, it is unusual, and has prompted debate in the Third and the Public Sector.

4The SIB remains an emergent phenomenon, but is rapidly growing. In March 2017 Stievenart and L’Huillier (2017) estimated the number of SIBs to be about 74 in 18 countries. In an older study, Rizzello and Carè (2016) made a complete dynamic analysis of the introduction of SIBs from 2010 to 2016 (Appendix 1 and 2).

5This new tool for tackling social problems emerged in the context of the economic crisis that started in 2008. Social expenditures were rising rapidly but states and public administrations were unable to address all the social problems that they faced. The credit crunch (also affecting NGOs and social firms) and decreased public spending (Tchernonog, 2012) increased the need for new sources of finance for social issues. Political leaders had to streamline public spending and increase “individual and community self-reliance to ensure future economic and employment growth” (Joy and Shields, 2013). Public spending for social issues is under pressure, despite very real social needs. One of our interviewees, an economic counsellor to the French President, confirmed this: “we would like to solve new social problems. Unfortunately, the French State has no cash to invest in them. All financial innovations like SIBs are therefore welcome to fill this gap. They provide private money, and favour social innovations”.

6Historically social benefits have been the principal mechanism to alleviate poverty and deal with inequalities (Alesina et al., 2001). However, the anticipated social impact has often not resulted, leading to greater inequalities – either as expressed by the gap between the Gini coefficient before and after taxes, or assessed with more refined methods (Piketty et al., 2009). To better tackle these pressing issues, many observers assume that public policy involving social problems should be more innovative.

7The SIB therefore appeared to be an attractive option. Warner (2013) points out that “SIBs are the intellectual descendants of new public management’s emphasis on markets and performance management. One of the goals of SIBs is to introduce the rigors of private sector investors, the standards of positivistic scientific evaluation, and the discipline of market mechanisms to determine payment and the allocation of risk”. SIBs also “broaden the concept of the New Public Management language of alternative service delivery to include alternative service funding. The SIB model of lean and decentralized government is dichotomized against the welfare state form of top-down, universal, and institutionalized care in hospitals, jails, and shelters that deliver poor value for money” (Joy and Shields, 2013). For some authors SIBs epitomize the return or expansion of New Public Management (NPM) and the introduction of market-based mechanisms into the public sector (Warner, 2013; Joy and Shields, 2013), or they are seen as a new proof of neo-liberal practices in the welfare state (Cooper et al., 2016). “NPM” often has a negative connotation (Mathiasen, 1999; Dunleavy and Hood, 2009).

8The SIB concept is a source of theoretical debate but also an innovative solution for social issues (Schinckus, 2017). We approach the SIB phenomenon through an NPM perspective. We remain neutral in our usage and employ the term “NPM” descriptively. We seek to answer the following research question: to what extent is the Social Impact Bond concept an expression of New Public Management? This assumes the proximity of SIBs and NPM, a feature that is now under discussion (Warner, 2013; Joy and Schields, 2013). We decided to use the New Public Management framework published by Dunleavy et al. (2005) because it presents an updated version of the main characteristics identified by Hood (1991), with some additional characteristics constructed after decades of NPM deployment in several countries. This framework is widely used and validated by academics, and also by official agencies (the OECD, for example).

9We began by studying all the academic literature on Social Impact Bonds, together with official reports such as those of the OECD and national governments. We also examined existing SIBs and their impact, seeking to understand the concrete professional experiences of practitioners. Our interview data is mainly drawn from France. This choice was made for five principal reasons. First, the country has a strong welfare state and important social expenditures, but at the same time is encountering increasing social problems. Second, France launched a Call for Projects in March 2016 in which the first “Contrats à Impact Social” were signed at the end of 2016; and this approach was a rich source of understanding about NPM and SIBs Third, France has a strong tradition of Public-Private Partnerships. Fourth, the Social Economy (NGOs, social firms, impact investing funds, etc.) is well-developed in France. And finally, French citizens are in comparison with other countries very anti-market (IFOP, 2011; Spector, 2017).

10In the first part of this paper we present current thinking about SIBs, and introduce the NPM perspective as a conceptual framework for analysing the SIB. In the second part we explain the methodology we use to answer the research question; and in the third part we present our main results in terms of the three NPM characteristics (disaggregation, competition, and incentivisation), also enlarging the NPM concept with the characteristic of “innovation”. The fourth part presents our general conclusions.

1. Literature Review and Definition

11Social Impact Bonds are well documented in the mainstream media, but this coverage only describes the rationale of SIBs, details cases and covers the broadest issues (see Appendix 3 for details). Moreover, the coverage is primarily limited to English-speaking countries, where the vast majority of SIBs have been launched. Nonetheless, awareness of the very existence of this tool to the public at large has developed, and demonstrates the interest in SIBs their relevance to the current social context.

12We first present a literature review and definition of Social Impact Bonds, then a literature review related to New Public Management.

1.1. Social Impact Bonds

13The mechanism of SIBs is different from usual public procurement practice: it is not a public tendering process, but involves the co-construction of an innovative social policy by a public entity, investors and a social actor. There is no invitation to tender nor competition between potential providers; instead, it is created as a negotiated contract. An SIB is a set of contracts between different organisations, not a specific social enterprise with its own identity and existence.

14The stakeholders agree on a contract and a structure whose rationale is the following:

- SIB stakeholders agree upon programme outcomes programme that can be measured (for example, reducing unemployment in a particular population to 4%, or reducing to 10% homelessness in a specific area).

- The service provider implements a social programme to resolve an identified social issue. As a consequence of the preventive actions of this social provider public sector social expenditure could be rationalised. The involvement of social actors has several motivations: they can develop new expertise and competencies, they diversify their sources of funding, and they heighten their reputation (Brooking, 2015, Warner, 2013).

- A financial institution puts up capital to fund the social programme. It bears the risk of losing part or all of the investment if the outcomes are not achieved. Indeed it will lose money if the social programme developed by the social service provider does not achieve the objectives (example: SIB of Rikers Island, where Goldman Sachs, as financial sponsor, lost $ 1.2M and Bloomberg Philanthropies $ 6M). Financial institutions have many reasons for becoming involved: communication goals (for some commercial banks, for example); pursuit of statutory objectives (charities, for example); developing new ways of achieving social objectives (for impact investors, for example); or to experiment with a new financial tool with potential future diversification in view (Socially Responsible Investment departments of commercial banks, for example).

- An intermediary (such as a social consulting firm) can be involved in the elaboration of the SIB structure (see Figure 1). This would then be central to the network of contracts and sometimes has responsibility for raising capital.

- An evaluator is commissioned. This may be an accounting or consulting firm, or an academic economic body. This stakeholder assesses social impact during and/or at the end of the SIB.

- If the agreed outcomes are reached the public sector will pay reimburse the investors plus interest. Failure to achieve agreed outcomes will result in a partial or total loss on the part of financial institutions or charities. If the programme is successful it will result in financial savings for the public sector. These savings are in part used – for example 50% of the total – to reimburse the investor.

Figure 1. Social Impact Bond standard structure and execution process

Image 10000000000002F400000191821B5AE1.jpg

15SIBs offer opportunities for governments seeking to modernise the deployment of social expenditures, which accounts for the wide interest in the instrument. Given the questions raised about its methods, however, (Schinckus, 2015b), a number of analyses were carried out. Cooper, Graham and Himick (2014) list several public studies (UK Cabinet Office, 2012; Government of Massachusetts, 2012; Government of Canada, 2013; Government of New-Zealand, 2012; New South Wales Government, 2012). There has been no official publication on SIBs in France apart from the documents attached to the Call for Projects in March 2016. The French government mandated a special group as part of the Social Impact Investment Taskforce for the UK’s presidency of the G8. This team was composed of practitioners, experts and few academics, and in 2014 it published a special report on impact investing, which included some discussion of SIBs (Comité Français sur l’investissement à impact social, 2014). This report was rather favourable to SIBs since the report failed to identify major problems arising. Two French think-tanks have also published reports on SIBs: the Fondation pour l’Innovation Politique (De Kerorguen, 2013), whose paper is a global introduction to the topic; and the Institut de l’Entreprise (Le Pendeven et al., 2015), which presented a complete review of the SIB concept and its implementation in a French context, learning from foreign examples. In the EU, the European Commission (2013) explained and promoted this innovative financial method mainly on the grounds that it is a way of renewing public sector social policies using private money. While these reports might mention New Public Management as a global trend for public policies, they do not demonstrate a logic and direct link with between SIBs and NPM. The Brookings Institution (2015) also published a detailed report on on-going international SIBs. These reports provide, first and foremost, an overview of this new financial instrument and avoid any attempt to place this tool within a theoretical framework.

16The concept of the Social Impact Bond is a recent phenomenon, but it generated several academic papers published in the period 2015-2017. Since its organisation and logic are both debatable and innovative scholars have described its mechanisms and studied its characteristics (Pettus, 2013). A review of this literature show that academic articles on Social Impact Bonds tend to envision a positive future for this new financial tool.

17Dowling and Harvie (2014) argue that SIBs import a capitalist mindset into the Third Sector. Sinclair et al. (2014) suggest that, as an economic mechanism that generates profit, SIBs de-moralize social actions.

18Norman (2010) on the other hand gives a very complete account of the concept of Social Impact Bonds. Schinckus (2015b) understands the debate as the confrontation of different ethical perspectives: the deontological perspective (the public implication for social problems) or that of consequentialist ethics (the relevance and impact on social problems, rather than intrinsic properties). Schinckus presents a global analysis of the SIB concept, assessing the debate on ethical issues and including a special focus on the financial dimension of social issues after the 2008 financial crisis. For Joy and Shields (2013), the ethical debate is also related to the obligation to measure and monetize a social impact. These obligations naturally orient welfare actions towards the resolution of easily measurable social problems. Such mechanical and localized perspectives limit adequate understanding of social issues. We can show in this paper how the technical specification of a new financing tool like an SIB can direct social impact itself; and also why SIBs cannot be applied to all social matters.

19The SIB is a major new government intervention that problematises the public realm and all its associated representations. For McHugh et al. (2013) and Sinclair et al. (2014), SIBs reduces government accountability because they outsource welfare services. McHugh et al. (2013) think SIBs are “the latest stage in an ideological shift which favours removing delivery of social and welfare services from conventional public or third sector providers, and they mark a significant challenge to the traditional ethos and operation of the voluntary and community sector. The relatively favourable reception accorded SIBs by sections of the third sector has led some commentators to claim that there are influential elements in the UK third sector which appear to endorse further privatisation measures in welfare reform”. This indicates how some authors have made a link between New Public Management and Social Impact Bonds.

20While the SIB is a financial innovation, few articles focus on this dimension. Bolton and Saville (2010) note that investors earn a financial return from investing in a social innovation. Schinckus (2015a) focuses on asset-pricing, applying classical bond pricing methods to SIBs. Pauly and Swanson (2014) also worked on the financial dimension of SIBs, studying them as a pure vehicle for project finance paid according to performance and constructing financial simulations so that they might understand the way in which they differed from classical financial vehicles. Cooper et al. (2014) study the implications of SIBs for social outcomes from an accounting perspectives, detailing the importance of this approach for contractual agreements and performance goals. None of these papers theorize the SIB concept, but they do contribute to a better understanding of the characteristics and implications of SIBs.

21The first SIBs, and especially the one in Peterborough, galvanised academic comment. In her article in the Stanford Social Innovation Review, Fiennes (2013) initiated a debate with Helbitz and Bolton (2013), who criticised the Social Finance methodology employed for the selection of the sample in the Peterborough SIB. The authors identify a selection bias arising from the fact that prisoners volunteered for the programme. The Peterborough SIB sometimes serves as a case study for articles that focus on a specific dimension: Schinckus (2015a) deals with SIB asset pricing as already mentioned, and Schinckus (2015b) discusses the valuation perspective. Together, these papers give a complete overview of this emblematic case and its consequences, providing a valuable resource for the analysis of the NPM framework.

22Our review of the literature revealed that, except for Warner (2013) and Cooper et al. (2016), no academic paper approached Social Impact Bonds theoretically, and in particular they neglect the NPM perspective.

1.2. New Public Management

23The apparent breakdown of the Keynesian paradigm of state and society in the face of stagflation in Western economies during the 1970s led to the development of new ideas, such as public choice theory (de Vries, 2010). This approach assumes that the state is not run by benevolent administrators but rather by rational agents seeking to protect their own interest (Buchanan and Tullock, 1962; Niskanen, 1971), contrasting with Musgrave’s view of government (Musgrave, 1989). Public choice helped the concept of New Public Management become a dominant theme of the 1990s.

24New Public Management proposes that the quality of public administration can be enhanced through productivity gains, like any other private organisation. Instead of allocating bureaucratic resources for services that can be outsourced, privatisation can contain government growth (Hood, 1991; Christensen and Laegreid, 2002; Kamarck, 2014; Moore, 1995; Politt and Bouckaert, 2004). Others define NPM in terms of the hierarchical structure within different public organisations, and the nature of control mechanisms (Aucoin, 1996).

25Despite NPM being loosely defined, Dunleavy et al. (2005) proposed in their modernized framework that NPM is based on three main principles. We decided to use this framework in our research because it is a well-established, recognized and updated version of the seminal framework of Hood (1991):

(i) - Disaggregation, by a logic of agencification, better measures for the public actions, competition by comparison (between public services in various areas, or with foreign countries) or flexibilization.

(ii) - Competition. By introducing competition among potential providers, outsourcing policies, and intragovernmental contracting, the public sector can better optimise and make savings.

(iii) - Incentivisation. There is a strong shift from a system of rewards based on the professional ethos of civil servants to a money-based system. Incentivisation is more broadly a strong concentration of practices like public-private partnerships, private finance initiatives, and anti-rent seeking policies.

26NPM primarily aims to provide better public services for the same cost. This approach has been widely criticized (Coats and Passmore, 2008; Dunn and Miller, 2007; Oehler-Sincai, 2008) and has not entirely worked as intended, in some cases leading to a larger staff and loose control mechanisms (OECD, 2010). New datasets and new approaches to social issues are now available, and highlight both concepts of performance and efficiency of public services. As a consequence, the public sector is better able to control its costs when rigorous methods of evaluation are applied to public services (Social Finance, 2011).

2. Methodology

27The Social Impact Bond is a recent object of research (Warner, 2013), a new way to finance social programmes. Only a few SIBs have been completed (about 20 according to Social Finance, 2016) and there is a lack of long-term data. For this reason none of the papers presented in the literature review build their research on data: many are only descriptive, some are based on one case-study as source of illustration, and to our knowledge none of them uses an empirical and quantitative approach.

28SIBs were created and developed in a context of actual social public policies before there was any attempt to conceptualize them. In our opinion it was important to base our methodology both on theoretical sources and on a strong empirical foundation if we were to be able to understand whether SIBs are related to New Public Management and have been experienced as much. We consequently developed a qualitative and exploratory approach. To identify the features of NPM we used the framework of Dunleavy et al. (2005), considered in the literature as relevant and also a modern version of the original NPM list of characteristics (Hood, 1991; OECD, 2010).

29We first did a complete literature review for Social Impact Bonds, although only a few publications deal with this emerging topic. We completed our review with public documents from governments (mainly the UK, but also Australia, Canada, and France), NGOs, intermediaries (like Social Finance), or SIB official reports (such as the Peterborough SIB Final Report on cohort analysis, Jolliffe and Hedderman, 2014) since they involve the potential expression of NPM principles for public policy expectations and/or findings. As for our research question, taking the intuition of Warner (2013) that SIBs are an NPM tool, we also enlarged the scope of our literature review to New Public Management: we examined the papers and reports for the leading NPM principles, and questioned them with regard to SIBs.

30We then decided to interview several practitioners involved in the SIB movement in France and in three others countries (Finland, UK and Belgium) so that we might understand the concrete capacity for SIBs to be perform as part of a New Public Management framework: examining the dynamics of SIBs in public sector reforms and their working sequences and mechanisms, so that we might appreciate agents’ behaviour and practice (Hedström and Bearman, 2009; Depeyre and Dumez, 2007; Yin, 2012). Since NPM covers a wide range of practices and principles resulting from concrete operational changes (Hood, 1991, OECD, 2015), it was essential to obtain some qualitative and empirical feedback. We addressed four different populations:

(i) - government counsellors, senior officials and politicians. This population was interviewed because of their centrality to the SIB process in France and Europe. Their position in central administration and in regional and local authorities means they can observe the changes resulting from SIBs and/or NPM policies.

(ii) - NGOs, social firms and various social innovators. This population expected either to launch SIBs in France, and/or launched the first ones, and/or decided not to work on Social Impact Bonds, but nonetheless seriously considered this topic and opportunity. We interviewed them because of their involvement in the SIB community, in the ideological debates around SIBs (opinion pieces in the national press), and the feedback they gathered from other organisations in foreign countries.

(iii) - Financial investors. This category of interviewees was selected because they are the main risk-takers in the SIB mechanism. Unlike other kind of outsourcing of public services to private sector and civil society, these are a new kind of stakeholder. It was essential to collect their own perspective and testimony.

(iv) - Intermediaries and consultants in social finance and innovation. This final category is composed of experts on SIBs (in France and abroad) and of consultants involved in the structuration, evaluation and promotion of SIBs.

31We identified these practitioners from the media, from groups of SIB experts and/or specialists on social finance in France (official circles and professional associations), and by recommendations from the first persons we interviewed. Eventually some professionals contacted us independently because we were identified as being neutral specialists with regard to SIBs.

32In total we conducted 62 interviews, including 14 exploratory interviews (to calibrate the interview format and identify the most relevant topics for discussion) and 48 semi-structured interviews, with 31 different people. The interviews lasted from 22 to 107 minutes, conducted face to face, by phone, or by Skype. Details of the interviewees are in Appendix 4. Social Impact Bonds were a controversial topic and source of debate with practitioners in the social economy;1 even if those in favour of SIBs people are overrepresented (the majority were involved in the creation of the first SIBs in France), the sample also include many who were neutral or critical regarding Social Impact Bonds.

33The first interview was made in March 2015, and the last was conducted in June 2017. During this period the context for Social Impact Bonds changed quickly and the French government launched a Call for Projects generating SIB opportunities.2 The first French SIBs appeared in 2016 (Husson, 2016). This was a fruitful period to complete the study, also taking account of initial results from UK, USA, Australia, Canada and other countries with older SIB approaches. Given the lengthy period in which we studied the impact of SIBs on public policies and their potential relationship to NPM, we took the opportunity to meet 18 people several times (2 to 5 times each).

34For reasons of confidentiality and/or at the interviewees’ request (given the political sensitivity in the early phases of the introduction of SIBs in France), most interviews were not recorded. Instead we wrote up full research notes a few minutes after each interview, basing our analysis on these notes. Since we had no transcriptions it was not possible to code the material with NVivo software. We analysed the content of our research notes by aggregating content and quotations from each interview by topic, taking into account the NPM framework of Dunleavy et al. (2005). To complete these qualitative interviews, and to strengthen the validity of our approach, we followed the recommendation of Yin (2012) and looked for multiples sources of evidence after every interview: archival records, websites, press articles and other secondary documents. We thus consolidated our knowledge of the consequences of SIBs on public policy according to the principles of NPM.

3. Results

35As we explained in the section about the definition of NPM, we consider the definition and framework of Dunleavy et al. (2005) as as our point of reference. They modernized the accepted definition by Hood (1991) that formalises three chief integrating topics of New Public Management: Disaggregation, Competition, and Incentivization. In the following we analyse Social Impact Bonds through their spectrum.

3.1. Disaggregation

36The logic of “disaggregation” in NPM is construed in the sense of Dunleavy et al. (2005). Translated into public policy in general, and social public policy in particular, this amounts to either removing existing social mechanisms; dismantling (which may entail removal followed by reconstruction) them; or replacing traditional providers of the social service concerned with other organizations that decouple policy systems and generate new management procedures and practices. These last create a concern for “intermediate organizational objectives rather than service delivery or effectiveness”. Lastly, disaggregation furthers the devolution of public responsibility to agencies, fragmenting existing hierarchical administration into quasi-governmental bodies.

37All the same, in theory an SIB may contribute to the disaggregation of public policy by bringing about a real reduction in the need for public services; this can happen if preventive experimental devices are very effective in reducing social problems. However, from preliminary empirical results, Joy and Shields (2013) argue that “there is a consensus in the literature that the shift to SIBs will not reduce bureaucracy and cut public sector costs”. Further, by isolating the social problems addressed rather than considering them as part of a whole SIBs can also contribute to the disaggregation of administration to agencies that are dedicated to precisely identified problems. This potential disaggregation may also arise from SIBs if the objective of the experiments conducted was to create competition between the existing public policy and new systems run by the third sector or the private sector. The purpose here would be to lower the cost of social services provided and financed by public funds.

38However, beyond these theoretical possibilities the organizational principle of SIBs, and the empirical analysis of existing SIBs, demonstrate that SIBs do not increase disaggregation.

39The principal logic of the SIB does not seem consistent with this definition of disaggregation: in principle it finances exploratory projects and experiments, rather than replacing existing programmes. By implementing SIBs the public sector may, in addition to reimbursing the funds invested, have to pay interest. Therefore, because the logic is one of disaggregation of public powers via pure externalization, the operating mode would be different and would not cost as much. Here, the interest paid is intended to finance the risk taken, and not to be an advance on the cash flow or an externalization.

40Further, to our knowledge, and according to interviews conducted and analyzed as part of our empirical research, the logic of disaggregation does not seem to prevail today because SIBs that operate on different national markets are oriented toward social problems that are addressed very poorly, if at all, by the public powers. Therefore, to date, of the 74 SIBs identified in the world (Stievenart and L’Huillier, 2017), including the six French SIBs already signed and 8 others that have received ministerial approval, SIBs complement social policies rather than replace them. This limits the risk of privatization of existing public services and the demise of the welfare state, as some authors have reported (McHugh et al., 2013; Warner, 2012).

41In addition, the analysis of profoundly detailed case studies described in the literature (Cooper et al., 2016; Van Es et el., 2016; Schinckus, 2015a; Schinckus, 2017) including the literature most critical of SIBs (Marty, 2016; Dowling, 2017; Maier and Meyer, 2017) does not detect any such logic of disaggregation.

42The “agencification” phenomenon seen in “Leading-Edge Countries“ (Dunleavy et al., 2005) has massively occurred in recent decades (up to 40 agencies in New Zealand, for example), and has been recognized by the OECD as a moderately effective policy (OECD, 2004).

43The interviews we have conducted with the French government officials in charge of the social and solidarity economy sector show that SIBs are not designed to supersede former social programmes but rather to tackle issues that the French state has neglected. As additional empirical evidence, owing to the lack of competition and overlap between experiments (exploratory devices) and social services of the Welfare State, the services concerned often receive these innovations and test devices with great interest (UK Cabinet Office, 2012; Government of Massachusetts, 2012; Government of Canada, 2013; Government of New Zealand, 2012; New South Wales Government, 2012) especially when they see that their missions would expand should the SIB in question succeed.

44This has long been observed concerning the first SIB in history, in Peterborough. This case clearly illustrates that SIBs do not favour disaggregation or public disengagement: following the experiment in the reintegration of offenders serving short sentences, the Ministry Of Justice (MOJ) extended the prevention device to much of Great Britain, affecting 45,000 offenders in prison. This did not occur in the context of the creation of an agency, but rather in the MOJ setting. Members of the former SIB Lab at Harvard Kennedy School, Azemati et al. (2013) also posited that the most important criterion for rolling out an SIB is its potential for a large impact, in direct relation with the need for the social policies to be massively deployed. They argue that SIBs must, by definition, expand public policy to new objectives.

45Further, even if the public policy in question aims to lower the social costs assumed by the citizens (recidivism is costly) and improve social welfare by reducing criminality, the device extended the sway of the welfare state by putting it in charge of a new component: prevention of recidivism among offenders serving short-term sentences. The social dimension of the welfare state can then be reinforced.

  • 3 « The project was the first example of an SIB, a type of payment-by-results contract where investor (...)

46In addition, whereas social operators involved in SIBs were associations,3 actors involved in extending this device to public policy work in the public service. There is therefore no de-professionalization of the public sector nor an evolution toward managerialization in the internal practices of public services.

47Therefore, we cannot really consider the SIB to be a tool corresponding to the logics of NPM regarding disaggregation. A senior-official in the French economic administration told us of the potential for SIBs against indebtedness that “If the SIB is a success, if the NGO is efficient, we will develop an additional service in the Banque de France for structuring and developing our actions against this social problem”.

3.2. Competition

48In the context of NPM “Competition” is defined as follows: “Introducing purchaser/provider separation into public structures so as to allow multiple different forms of provision to be developed and to create (more) competition among potential providers. Increasing internal use was made of competition processes to allocate resources (in place of hierarchical decision making). The ‘‘core’’ areas of state administration and public provision were shrunk, and suppliers were diversified.” (Dunleavy et al 2005, p. 470).

49The study of the principles of an SIB arrangement demonstrates that:

- SIBs that propose that the public authority join the consortium are not substitutable for one another, contrary to the bids typically received in response to a call for tenders. Notably, they deal with very varied social topics, often with different approaches.

- Theoretically, substitutable SIB projects are possible (for example, two projects intended to reduce malnutrition by 20% in a given department), but in practice, to our knowledge no such projects exist.

The analysis of interviews confirms this premise. Substitutability is non-existent despite the fact, for example, that in calls for SIB proposals by the French state, many projects were submitted to combat one of the main social ills in France: unemployment. Therefore, of the six French SIBs signed, all concerned employment, yet each adopted a different approach: focus on training, EAV, entrepreneurship, fashion makeovers, job seeker mobility, etc.

- Competition therefore does not exist concerning the quality/price ratio for a need identified by the public power, and no calls for tenders have been arranged.

- Beyond the lack of substitutability of projects, the selection criteria are not transparent, which makes the usual concept of competition (in the broad sense) difficult to objectify for project leaders. This top-down selection bias is quite different from the transparency processes that can typically be observed in public procurement.

- Lastly, SIB processes include a dimension of co-construction between the public party and the private and association partners: joint work on indicators, priorities to set for the programme, the potential rate of return in case of success (and more broadly, construction of the incentive system of the device). This dimension seems especially remote from the precepts of pure and perfect competition: here, the public power co-constructs the proposal with the “suppliers,” without reference to classical ideas of competition.

50Therefore, given these characteristics and the classic definition of competition within key components of NPM advanced by Dunleavy et al. (2005), the SIB clearly does not seem to have developed from NPM.

51Further, in an SIB we can observe an unusual form of competition. In fact, it is an extension of the concept of competition cited above. There is no competition, strictly speaking, to gain access to the client’s money (state or communities) because in this case the client does not pay but only potentially shares the gains resulting from a reduction in its social spending. However, there is competition over access to other resources held by these public parties, and these are quite vital to SIB arrangements.

52This competitive tension is particularly consequential in cases where public decisions are symbolically or formally very centralized (Call for Projects in France, for example). The interviews clearly show that in France the 80 projects submitted were initially sorted based on the quality of the arrangement, the rigour of the process and pertinence as judged by the stakeholders.

53However, once this filter is applied, real multi-criterion selection was made to gain access to public parties’ resources. There is thus a form of resources-based competition that largely exceeds the financial dimension. Because the monetary, cognitive and organizational resources of the public institutions cited above are limited, the public parties must choose both (i) from among the social policies raised and (ii) the SIBs signed. It may seem surprising to imagine public institutions creating competition between subjects as diverse and incomparable as the integration of homeless people, the struggle against dependence, over-indebtedness or efforts to prevent long-term unemployment, yet this is an economic reality, at least in France: the Director in charge of social and solidarity economy in the Treasury General Direction of the French government explained to us that “At the Treasury General Direction, we cannot assist and be involved in all the relevant SIBs projects. Sometimes it’s sad because there are quite interesting initiatives, but if these SIB-based social innovations are to succeed we need to concentrate on only a few experiments. The experiences of our colleagues in the UK are similar, at the local or at the national level: they have to select SIB projects, because there were too many of them, and they do not have enough staff to follow all topics and consortiums”. Moreover, the co-founder of an accountability and consulting firm, specialist on SIBs of which the main job during the Call for Projects was to work with social innovators to prepare the files, said during the interview that “When I selected a consortium to help, I had to think about all the dimensions: will the administration and politicians be motivated by this social issue? Is the potential to scale interesting? Is it a politically sensitive topic? Even if the potential of cost-savings is attractive, does this service develops a clear social impact? Etc.”.

54First, there is competition between the social subjects addressed. Accordingly, factors that dictate the decisions made by public authorities may vary greatly, and are always very highly contingent on context: (a) economic-budgetary, (b) social, (c) political, (d) cultural, and (e) local institutional, coupled with (f) the more or less recent history of related social initiatives:

55(a) The level of public spending and especially public expenditures allotted to the various social subjects. Choice is often oriented towards the areas of greatest expenditure, where the potential gains in avoided expenditures may be the largest. Thus the President of a French Department (local government) testified that “RSA [e.n. Revenu de Solidarité Active, a subsidy for unemployed people without unemployment benefits] is a key topic for us. Apart from the SIBs, we try anything we can to fight unemployment. So we are very interested by Social Impact Bonds focused on employment, and we clearly prioritize this social goal”.

56(b) The most pressing social needs, or those considered priorities. Even when countries are a priori at the same level of development and fairly similar (for example countries in Western Europe), they all have distinct social problems, although social problems exist in many countries. For example, France has a long-standing unemployment problem (the unemployment rate has not declined below 9% of the workforce since 2009, whereas the rate in Germany has constantly been below 8% since 2009, and the rate decreased almost continually to 5.5% in 2017). Political and administrative decision-makers generally favour what they view as the most pressing social needs, or those that have recently appeared (for example, the digital divide that plagues the elderly), along with subjects that have received heavy media coverage. In Canada for example the topic of family breakdown is well-analyzed by the media, contributing to the decision to launch the SIB “Sweet Dreams”). It is also the case where civil society and citizens have exerted particular pressure. In Europe for example economic integration of migrants and refugees is pressing, which has led to the creation of a Belgian SIB, Duo for a Job, and a Finnish SIB led by the organization Sitra. The respondents interviewed expressed clear views on this subject: the most prominent social subjects were unemployment, considered the most serious ill in France, over-indebtedness as a social problem that is both important and overlooked by public powers. Several interviewees agreed on this, and all the 6 first French SIBs concern the fight against unemployment.

57(c) The political positions of elected officials and the contexts in which they operate, which naturally influence the themes chosen. Therefore, political contingencies may expedite or favour one theme over another, with the expectation that the results of the SIB will allow deployment of an effective and innovative new social policy with timing that benefits the authorities, for instance in the political (elections) or institutional (government planning of reforms) spheres. For instance, the first SIB in history, that of the Peterborough prison, allowed the British government to create a service to re-integrate offenders serving short sentences at a time of massive reform of the judicial and prison system.

58(d) The choice of social subjects to address via SIBs necessarily involves cultural aspects of the country. Depending on the habits of the population, the degree of acceptance of innovation and of intervention from associations or private businesses in solving social problems, public powers may favour certain SIB arrangements. For example, the leader of an SIB in Finland, explained to us that “In Finland, usually the public services are completely in charge of unemployment issues. So when we worked on our SIB with civil servants and politicians, we sought to explain that innovative services were not destined to be scaled to the Finnish population, and that it is not a first step to privatization. In my opinion, if we succeed in launching this SIB it will be because it is intended for a special community, that of recent immigrants”.

59(e) The institutional organization of countries that have used or may use SIB to solve some problems varies considerably. Notably, the level of commitment of countries in resolving social causes via the welfare state differs (in 2016: 31,5% of GDP in France, 19,3% in the USA, 21,5% in GB - OECD data). The institutional level at which decisions are made is critical. In France, for example, employment policy is managed dually, at both the national level (Pôle Emploi, or employment centre, training plans for the unemployed, etc.) and at the departmental level (benefit payments). Therefore, both types of actors are probably mobilized as final payers of an SIB oriented toward employment, whereas in other countries such as Australia the system of end payers is completely decentralized.

60(f) Lastly, public powers proposing a broad choice of SIB arrangements cannot ignore previous related experiments. Therefore, if recent initiatives have been made, but using subsidies and not an SIB framework, and especially if these initiatives have failed, local actors may have no interest in tests or innovations concerning a particular social initiative. For example, an Executive civil servant in a local authority in the south of France, in charge of promoting SIBs within the European programme “Alp’SIB”, explained that “Some Central-Europe countries tested various social services to work on school failure. They massively failed. Nowadays, they absolutely don’t want to work on this topic again. They are “allergic”; meanwhile in France, Italy or Germany for example, it is clearly a hot topic because we have never engaged deeply with it”.

61Consequently, despite the great heterogeneity of projects and social subjects addressed, a form of competition arises between organizations that propose SIB arrangements to the administration and to public and political powers.

62This competition could be observed regarding the emergence of French SIBs: 80 projects were received and examined, and the State was in charge of selecting (when it was the final payer) or giving its approval (when the payer final was a community) on the projects. Therefore, given the competition generated by the Call for Projects, proposals aim to be well-formulated (to pass the first selection filter regarding economic quality) and attractive (see the multi-criterion list mentioned above) to the public authorities, in order to improve their chances of being approved or assigned ministerial approval. Although this was not officially a subject of discussion, there was thus real competition between the projects. Such competition was particularly marked in projects that aimed to have the state as final payer: ministries examined each of the arrangements, actors, social innovations proposed, etc., passing them through the various filters mentioned above.

63As stated above, the “competition” dimension of NPM encompasses more than just the “economic” or “price” dimension. Political and administrative decision-makers should be find the SIB useful, given that they have their own personal and individual motivations.

64During our interviews regarding the French case many interviewees confirmed that state authorities play down the economic-financial dimension in the choice of winning SIBs because they do not want to be too open to criticism on the cost-cutting dimension of SIBs. Instead they focus on the “innovative” dimension, the scope of social issues addressed, and the quality of the arrangements (including the quality and reputation of the actors engaged).

65To summarize, there is no “Competition by selection” as in the classic framework of NPM and public procurement, which requires potential suppliers to compete relative to an identified and expressed need. Rather, there is “Competition by choice”, a form of extension of the notion of competition classically evoked in NPM, where competition is expressed in the capacity to obtain validation by the state or communities for a particular SIB arrangement. Evidently, public authorities cannot invest in dozens of SIBs simultaneously. Social Impact Bonds are consequently linked to the NPM’s characteristics of competition, but more broadly.

3.3. Incentivization

66The definition and components of Dunleavy et al. (2005) with respect to incentivization within NPM are clear. With its Payment by Results (PbR) system, the Social Impact Bond appears to correspond to the definition of NPM. The SIB is inherently a device based on PbR: the actors in the consortium are paid in part, or not at all in case of nonattainment, according to objectives previously set. This incentive mechanism is very powerful because it might involve a total loss for the financers of the arrangement (logic of high-risk/high-reward). Nonetheless, the originality of the SIB is the diversity of actors involved in the arrangement, all of whom have slightly different incentives.

67The principal motivated and incentivized parties are the private financiers of the arrangement: banks, foundations, impact investing funds, charities, etc. They assume most of the risk-benefit incentive: should the minimum objectives of the contract not be attained, they would lose all or part of the investment. If the debt objectives are reached or exceeded, they would receive premiums (interest) that may create real incentives, and offer very high IRR. The example of the Australian SIB Benevolent Society SBB is enlightening. Two tranches of SIBs were proposed: a portion of 7.5M AU$ with 100% guaranteed capital protection (by the New South Wales government), but with an annual interest rate of 0 to 10% in case of performance improvement of over 40%; and a portion of 2.5M AU$ with no guaranteed capital protection, but with a potential annual return on investment of 30% in the case of outperformance by at least 40%.

68Private actors or associations. Pressure and the incentive system are transferred: all SIB social actors are paid during the years that the service is delivered. Their risk in case of failure therefore remains limited. Nonetheless, the consortium of private financers often pays a premium. This creates a financial incentive - which may be particularly strong - for the SIB to function well. Further, these actors have a much stronger incentive than the gain of some additional euros or dollars: they have a reputation issue. Should the SIB succeed, or even in case of failure but in a well-executed mission, the service providers have much to gain. This arrangement may be particularly attractive if the device is deployed and scaled over many public services; the associations or private structures previously engaged in the SIB would probably be well positioned to delegate the public service proposed.

69Two interviewees mentioned this, both of them being managers of French associations. They indicated the strong incentive for these managers to demonstrate the pertinence of their proposal via the SIB, and thus position themselves well for its subsequent deployment elsewhere in France

70A priori, to our knowledge possible intermediaries and evaluators have no particular incentives to ensure that the SIB performs well after it is launched. Instead, they all have an incentive to reduce the cost of their actions and so make the SIB more attractive to the public authorities (better potential ROI in innovation).

71The Payment-by-Results (PbR) method relies on a rigorous evaluation of the performance of public employees based on pre-established outcomes (O’Connor and Neumann, 2006; NHS, 2012). According to Warner (2012), the rationale of the SIB is similar to that of PbR, since it provides an incentive for parties to align their interests in order to improve public service efficiency. So far PbR has mostly been applied to healthcare, albeit with mixed results (O’Connor and Neumann, 2006).

72This application raised criticism both on empirical grounds (Mullen et al., 2009) and with respect to public value creation (Warner, 2012). Despite these criticisms, our interviews confirm that PbR has been rather well received, also by the majority of social providers. The risk involved relates to the legal framework around PbR, which still remains unclear in France.

73Measuring performance through the PbR method has some merit. Public services are no longer a simple means to an end – eg. helping fragile populations with public money. PbR shifts the focus to real outcomes and the intrinsic quality of specific public policies (NCVO, 2011). To evaluate impact of social programmes – which differs slightly from the outcome – remains one of the biggest challenge faced by economists. For instance, how can we measure the benefit for society when a long-term unemployed person returns to the job market? The assumption is that we can measure the efficiency of a social programme according to agreed criteria. In their seminal 1978 paper Charnes, Cooper, and Rhodes presented the first model to measure the efficiency of public programmes. Their definition of efficiency comes from engineering, especially combustion engineering, where “efficiency is the ratio of the actual amount of heat liberated in a given device to the maximum amount which could be liberated by the fuel [being used].” In economic terms, efficiency “has to do with the comparison between input, and output or between costs and benefits. At a given input, the greater the output, the more efficient an activity is” (Afonso et al., 2006).

74Given the consensus in the academic literature on efficiency, it is possible to construct new indicators (Afonso et al., 2006). Applied to social policies and their impacts, refined techniques developed by Duflo, Kremer and Glennerster (2006) are now available. Stemming from medicine, techniques such as Random Control Trial (RCT) control the impact of a social policy on a given population, while comparing the results with another population that does not benefit from the social programme. Both cohorts should be otherwise similar on every count (age, gender, socio-economic background, etc.). Other techniques rely on historical data that are compared with the results of the SIB (Dermine, 2014). Most SIB performance measurement is based on historical data. They can be easily related to the impact of the SIB, especially when long time periods are involved. This is for instance the case of the Belgian SIB, Duo for a Job, which relies on public unemployment data related to immigration and the labour market to increase the population’s ability to find a job. According to one expert in the sector we interviewed, this is due to the cost of RCT. Besides, RCT is not always easy to implement.

75The Rikers Island SIB failure is a concrete illustration of this difficulty. The SIB consisted of a rehabilitation programme for inmates. However, the service provider was unable to distinguish the cohorts that were needed for the evaluation. Indeed, prisoners who were supposed to follow the programme did not regularly attend it, while others sometimes participated. This brought about a statistical bias that could not be overcome by the service provider. The Rikers Island programme was therefore halted after a one-year trial period (Brookings, 2015). This explains why measuring social impact implies a very high degree of transparency among different actors, and so it is important to avoid informational asymmetries. Otherwise, the SIB is likely to fail.

76The above shows how the SIB is completely aligned with the concept of NPM and its definition on the incentivisation dimension.

3.4. An extension of the New Public Management concept: Innovation

77Exploring Social Impact Bonds through the prism of NPM creates the capacity to question the principles and details of the NPM. Our research proposes one additional, but supplementary, element: innovation. This can enrich the concept of NPM, alongside the existing criteria of incentivization, competition and disaggregation (the framework of key components identified by Dunleavy et al. 2005) and the original definition of New Public Management by Hood (1991). However, this is supplementary since it is not applicable to all aspects of NPM in all situations.

78Social Impact Bonds are not intended to optimize state and/or local authority public policy, but are aimed at encouraging emerging innovations to social policy by the private and charitable sectors. SIBs are a means of generating social innovation: they enable the introduction of experimental approaches that would not otherwise have been tried or financed. Using the example of a Dutch SIB directed at youth unemployment, Van Es et al. (2016) demonstrate the innovative nature of the project, and how it was complementary existing social policy, and not a substitute for it.

79SIBs are instruments of innovation for the following nine reasons.

801. Decentralization. In many Western countries social policy is not the responsibility of central government, but rather of local or municipal authorities (for example in France Départements have responsibility for social policy, in Australia state governments like New South Wales). There are potentially hundreds of organisations that could in each country contribute to experimental projects. This multiplies the number of people involved, the social issues to resolve, local dynamics and creative ideas. This logic of dispersion is very often an incubator for innovation and experiment. One consulting firm specializing on the creation of SIBs expressed its surprise to us that many different organisations from several regions contacted them with a view to developing new projects.

81Moreover, SIBs are by definition involved in a logic of decentralisation since initiatives originate with practitioners, NGOs, dedicated businesses, cooperatives and so on. The French case, where there was a centralized Call for Projects by the Ministry of Economic Affairs, is an exception.

822. The diversity of stakeholders. This point is often raised in the literature on SIBs: SIBs several kinds of organisations collaborate on SIBs, and sometimes they are not used to working with other agencies, for example a private-equity fund (as private funding provider) and an NGO specializing in the care of the elderly. Academic studies have demonstrated that diversity of organisations and people increases innovativeness (Gassmann, 2001). We interviewed the co-founder of a consulting company specialising in impact investing whose job was to assist a group of local actors in south-west of France in designing an SIB. She said that “These organisations decided to work together, without any consideration for their different status, only because they were motivated to help fragile populations in their city, and because they can do it without the state. This diversity allowed them to propose really attractive and innovative approaches”. Cross-sector collaboration is not new and peculiar to SIBs, it existed long before the first SIB was developed; cross-sector collaboration was already seen as a positive approach to social problems (Bryson et al., 2006). Public and private organisations (i.e. government, businesses, non-profit, charitable and community organisations) would pool their resources to create complementarities and eliminate duplication with regard to significant social issues that would not be able to address on their own (Mahoney et al., 2009). SIBs are the typical tool for the efficient achievement of these three points.

833. The presence of private funders. Social innovations developed in SIBs are financed by non-public institutions. We saw in the introduction that various sponsors can be involved: banks, charities, impact investing funds, etc. Such a diversity of potential funders also increases the capacity to generate innovation: each entity can have its own strategy and expectations, ensuring an effective impact upon social programmes. For example, some financial agencies will be oriented to high-risk investment, so they would finance the most innovative and risky SIBs, while a private bank would be more conservative, even if potential profitability is lower.

844. The logic of a legal structure supportive of innovation and risk-taking. Even if the literature neglects legal analysis (beside (Social Finance, 2011; Humphries, 2013, Dermine, 2014) and a few legal analyses in the specific French ecosystem - Le Pendeven, Nico et Gachet, 2015), the relevant legal structure is private and not public law. SIBs depend on contract law, and only partially on administrative law; this provides greater flexibility since public procurement rules and procedures inhibit innovation.

855. SIBs are often dedicated to social topics beyond the range of public sector activity. Social Impact Bonds can be applied to very different social issues (homelessness, over-indebtedness, unemployment, old-age dependency, etc.) and with a special emphasis on prevention. Private firms or NGOs can be in a position to propose innovative solutions.

866. For a few SIBs (a tiny minority), no public entity is involved. The great majority of international SIBs have a public entity involved in the contract (central government or a local authority, Brookings, 2015). Currently in France all the SIBs are publicly supported, but an endowment fund is being created. This fund aims to be complementary to the public sector as an end-payer. This lack of a need for public sector involvement is positive in stimulating the creativity of projects. Social innovators can then create programmes that are not a priority for politicians and civil servant (regarding the selection criteria presented above).

877. Communication about SIBs is very focused on innovation. The rhetoric of progress and innovation can itself be a tool used by project leaders to propose innovative social solutions. Our analysis of communications in six countries shows that private, associative and public promoters of SIBs emphasise their innovativeness and capacity for experimentation. The French Call for Projects did not primarily centre on potential cost savings and disengagement by the state, but focused mainly on the capacity to generate innovative social solutions and unfamiliar experiments for disadvantaged populations. This strategy of emphasising innovation is also identified by opponents of SIBs.4

88Beyond the purely public promotion of innovation, the innovation dimension often plays a decisive part. As we argued in our section on “Competition”, public institutions are subject to contradictory constraints, selecting the Social Impact Bonds in which they are involved according to many criteria. Othere things being equal, at a time when the entrepreneurial and innovative spirit has a high profile policy-makers tend to select the most innovative projects because of their potential for political communication.

898. New sources of finance. Because Social Impact Bonds do not substitute for other programmes or tools for the financing of social innovation they enable organisations with limited resources to submit projects that they would otherwise not be able to pursue. One of our interviewees in charge of social investments in a large French public financial organisation cited the case of where the existence of the SIB mechanism enabled her to arrnage the finance of a project that she had previously not been able to fund. For this project subsidies were inadequate (the project required more than € 1m.) and conventional private sector funding was not available. The existence of an SIB framework meant that disadvantaged people received care and a preventive social policy.

909. Prevention. In western countries social policy mainly relies on ex-post subsidy. Developing preventive programmes and actions is therefore a form of innovation. Many interviewees were really clear on this point: most socially preventive action would involve new initiatives. An interviewee from a French NGO that sought to combat the dependence of elderly persons pre-emptively by introducing dedicated sports programmes told us that “Our strategy is to prevent dependency. There is no public action in France on that social issue, nor anything developed by NGOs or social firms. With our programme we will be nearly the first to help seniors to keep fit, and clearly the first to do it with this kind of course”.

91For all these reasons we can say that SIBs generate innovations. While the recent literature on the finance of innovation notes the emergence of new financing tools (Block et al., 2017., Drover et al., 2017), to our knowledge just a very small part of these financial innovations enlarge the market of innovation financing with new projects. This is what SIBs do.

4. Conclusion

4.1. Findings and Discussion

92We have analysed Social Impact Bonds using the New Public Management framework of Dunleavy et al. (2005), evaluating the practices and consequences of this innovative tool. Our major conclusion is a response to our initial question: to what extent is the Social Impact Bond an outcome and a new element of NPM policy?

93Our results demonstrate that SIBs can in part be considered a tool for New Public Management. On the one hand SIBs clearly use incentive mechanisms (the system of targets, pay-back and interest payments). But on the other hand, SIBs do not contribute to the disaggregation of the public sector, and in some ways can be the basis for new preventive public social actions (for example the policy against recidivism developed by the Ministry of Justice in UK). For Dunleavy and his colleagues, the third dimension of their typology is competition. Here the relation of the SIB and NPM is balanced. Indeed, as usually conceived, SIBs do not fit the NPM’s definition: there is no competition over costs and prices between potential providers. Nevertheless, these mechanisms do generate competition. Indeed, as outlined in the results, we demonstrate that SIBs also generate competition between socially innovative projects for access to public sector resources. These resources are mainly not financial, but still generate a choice between projects.

94One contribution of our findings is an enlargement of the usual definition of NPM: we think that the SIBs can add to the typology of Dunleavy et al. (2005) the complementary dimension of innovation. SIBs can be a very useful tool to spur social innovation. Examples in western countries demonstrate the capacity for this new financial tool to support creativity and to enable projects lacking financial solutions. At a time when the public sector has to be innovative and contribute to a nation’s overall economic performance, the “innovation” dimension that we identified opens a new avenue in NPM practice.

95Our conclusion opens up a discussion about the organizational and managerial practices of the public sector that also involves the private sector. Indeed, after years of cost-cutting, optimization, the sale of non-strategic assets, and a renewed concentration on core activities, large companies have turned to the renewal of their practices, products and services by generating innovation. One of their goals is not to be disrupted by innovative organisations such as start-ups. The identification of an “innovation dimension” in the public sector could be a symbol of the same strategy developed in public organizations: to innovate so as not to be downgraded. Social Impact Bonds are thus a symbol of the innovativeness of public organisations, and of a proximity to the private organisations which seek to innovate.

4.2. Limits

96Naturally our paper has several limitations that can be overcome with further theoretical and empirical research.

97The first limitation is related to the novelty of the instrument we study: most of the SIBs introduced around the world are still running. Of course, as the cases of Peterborough and New-York show, significant problems can be discovered only at the end of a programme. And so as programmes terminate in the coming years more will be learned about the implementation of SIBs that would be important for NPM. Brookings (2015) lists the main existing challenges for Social Impact Bonds, including the public policy perspective.

98The second limitation is that our paper focuses on the material aspects of SIB implementation and outcomes, but without systematically analysing all existing SIBs. As shown in the literature review (Cooper et al., 2014; Humphries, 2013), specific aspects and local considerations can create real differences in the deployment of SIBs and any potential changes in public and social policy following their outcomes. Our conclusions about SIBs and NPM could well change, depending on local policy and behaviour. Our conclusions are systematic and based on the available data, but would be open to modification once more SIB programmes have terminated successfully or unsuccessfully.

99Third, Social Impact Bonds involve a new form of contracting that create new forms of network. These features are still evolving, with new forms (Development Impact Bonds linking developed and emerging economies) or new logics (the avoidance of reliance on public sector finance and administration by some SIBs through the engagement of insurance companies). Our conclusions are therefore valid only for the conventional form of SIBs, but newly-emeerging characteristics will in all likelihood call into question these initial theoretical analyses, and their placement in an NPM framework.

4.3. Future Research

100Social Impact Bonds are a recent phenomenon, and so academic papers dealing with the issue are still limited in number. This is a rich area for research. We studied existing SIBs, but over the next two years many SIBs om many countries will come to an end. We will take advantage of new data to improve our analysis, developing a systematic analysis of all the completed SIBs at world level, and the consequences they have had for social and public policy. This will improve our empirical findings about NPM and SIBs.

101In the next stage of our research we can examine new forms of SIBs: those funded by search funds or private-equity (through a bond issue, or private-equity fundraising); SIBs where some actors combine different SIB roles (investors also becoming intermediaries, for example); SIBs with an impact on non-social issues, but taking advantage of their risk averse mind-set to focus on public savings; SIBs without public sector among their stakeholders (like an Australian SIB whose model replaces the public sector by private insurance firms). All these new forms of SIBs, or changes to the SIB market, are likely to alter our understanding of the impacts of SIBs on social and public policy.

102We also know that Social Impact Bonds emerge in many countries to deal with diverse social issues (Brookings, 2015). Culture and values play an important role in the way that institutional systems are organised (Alesina, 2015). It should be kept in mind that cultural differences influence the way organisations are shaped (Hofstede, 1981) and how public policy is conducted, especially that regarding redistribution (Alesina, 2006). There are different expectations of institutions across countries, and some societies are more prone to base their social policies on efficiency principles. Future research could also compare social policies and concrete actions initiated following SIB projects in different countries and cultures.

Haut de page

Bibliographie

Afonso A., Schuknecht L., Tanzi V. (2006), « Public Sector Efficiency: Evidence for New EU Member States and Emerging Markets », European Central Bank, Working Paper Series, n° 581, January.

Alesina R. (2006), Combattre les inégalités et la pauvreté Les Etats-Unis face à l'Europe, Flammarion.

Alesina R., Giuliano P. (2015), « Culture and Institutions », IZA Discussion Paper, n° 9246, August.

Alesina R., Glaeser E., Sacerdote B. (2001), « Why doesn't the US have a European-style welfare system? », Working paper n° 8524, Cambridge, MA.

Aucoin P. (1996), « The New Public Management: Canada in Comparative Perspective», Ottawa: Institute for Research on Public Policy.

Azemati H., Belinskyy M., Gillette R., Liebman, J., Sellman, and Wyse, A. (2013), « Social Impact Bonds: Lessons Learned So Far », Community Development Investment Review, Federal Reserve Bank of San Francisco: p. 23-33.

Barker M. (2014), « A Prisoner Rehabilitation Programme in Peterborough Seemed to be Working. Now it's been Cut », Independent, April 28.

Birkwood S. (2014), « Peterborough Prison Social Impact Bond Pilot Fails to Hit Target to Trigger Repayments », Third Sector, August 7.

Block, J.H., Colombo, M.G., Cumming, D.J., and Vismara, S. (2017), “New players in entrepreneurial finance and why they are there”, Small Business Economics, pp.1-12

Bolton E., Saville L. (2010), « Towards a new social economy: blended value creation through social impact bonds », London: Social Finance.

Brookings (2015), « The Potential and Limitations of Impact Bonds: Lessons From the First Five Years of Experience Worldwide », Global Economy and Development programme, Brookings Institution, July.

Bryson J. M., Crosby B.C. and Stone M.M. (2006), “The Design and Implementation of Cross-Sector Collaborations: Propositions from the Literature”. Public Administration Review, 66: 44–55

Buchanan J., Tullock G. (1962), The calculus of consent: Logical foundation of constitutional democracy, Indianapolis, IN: Liberty Fund.

Buysschaert A. (2015), « Social Impact Bond: état des lieux », Le Vif, July 28.

Cabinet Office (2013), « Social Investment: Brief Background Note on the UK Government's Approach », https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/266124/CO_Social_investment_background_one-pager.pdf.

Cahalane C. (2014), « Social impact bonds: is the dream over? », The Guardian, May 1.

Christensen T., Laegreid P., (2002), « New Public Management: Puzzles of Democracy and the Influence of Citizens », The Journal of Political Philosophy, 10(3), p. 267-295.

Coats D., Passmore E. (2008), « Public Value: The Next Steps in Public Service Reform »The Work Foundation, September.

Cohen D., Zelnick J. (2015), « What We Learned from the Failure of the Rikers Island Social Impact Bond », Non Profit Quarterly, August 7.

Cohen R. (2015), « A flagship SIB has Flopped, but the Fans Are Undeterred », Third Sector. September 17.

Comité français sur l’investissement à impact social, (2014), « Comment et pourquoi favoriser des investissements à impact social ? », September, Paris.

Charmes A, Cooper W.W., Rhodes E. (1978), “Measuring the efficiency of decision making units”,

European Journal of Operational Research, 2, pp.429-444

Cooper C., Graham C. and Himick D., (2014). « Social Impact Bonds: Can Private Finance Rescue Public Programmes? », Critical Perspectives On Accounting, July 7-9, Toronto, Canada.

David A. (2008), « La recherche-intervention, cadre général pour la recherche en management », in Albert David, Armand Hatchuel and Romain Laufer, Les nouvelles fondations des sciences de gestion, Paris : Vuibert.

De Kerorguen Y. (2013), « Un outil de finance sociale: les social impact bonds », Etudes Fondapol, Décembre.

Depeyre C., Dumez H. (2007), « La théorie en sciences sociales et la notion de mécanisme : à propos de Social Mechanisms ». AEGIS Le Libellio, 3 (2), pp.21-24

De Vries J. (2010), « Is New Public Management Really Dead? », OECD Journal on Budgeting. 1: p. 1-5.

Dermine T. (2014), « Establishing Social Impact Bonds in Continental Europe » John F. Kennedy School of Government, Harvard University, May.

Dowling E., Harvie D. (2014), « Harnessing the Social: State, Society and (Big) Society » Sociology, 48(5): p. 869–886.

Drover W., Busenitz L., Matusik S., Townsend D., Anglin C., and Dushnitsky G., (2017), “A Review

and Road Map of Entrepreneurial Equity Financing Research: Venture Capital, Corporate Venture Capital,

Angel Investment, Crowdfunding, and Accelerators”, Journal of Management, Vol.43, Issue 6, pp.1820-1853

Duflo E., Kremer M. and Glennerster M. (2006), « Using Randomization in Development Economics Research: A Toolkit », National Bureau of Economic Research, Technical Working Paper, 333, December.

Dunleavy P., Hood C. (2009), From old public administration to new public management,

Public Money & Management, 14(3)

Dunleavy P., Margetts H., Bastow S. and Tinkler J. (2005), « New Public Management is Dead: Long Live Digital Era Governance », Public Policy & Services, LSE, September.

Dunn W., Miller D., (2007), « Critique of the New Public Management and the Neo-Weberian State: Advancing a Critical Theory of Administrative Reform », Public Organization Review, 7(4), December.

European Commission. (2013), « Towards Social Investment for Growth and Cohesion – Including Implementing the European Social Fund 2014–2020 », Staff Working Document.

Fiennes C. (2013), « What the First Social Impact Bond Won’t Tell Us”. Stanford Social Innovation Review, April 3.

Foley S. (2015), « Social Impact Bonds Need Help to Fulfil Their Potential », Financial Times. July 13.

Forbes (2014), « Social Impact Bonds Are Going Mainstream », September 15.

Government of Canada (2013), « Social Finance », Retrieved 13 Feb 2013, from http://www.hrsdc.gc.ca/eng/consultations/socialfinance/index.shtml.

Government of Massachusetts (2012), « Massachusetts First State in the Nation to Pursue 'Pay For Success' Social Innovation Contracts », from http://www.mass.gov/anf/press-releases/2012/mafirst-to-pursue-pay-for-success-contracts.html

Government of New Zealand (2012), « Social finance and social enterprise », Retrieved 14 February 2013, from http://www.ocvs.govt.nz/work-programme/building-capacity/social-enterprise.html.

Hedström P., Bearman P. (2009), The Oxford handbook of analytical sociology. Oxford University Press.

Helbitz A., Bolton E. (2013), « Making the Case for Intervention », Stanford Social Innovation Review, April 25.

Hofstede G. (1981), « Culture and Organizations », International Studies of Management & Organization, Vol. 10, n° 4, Organizations and Societies (Winter, 1980/1981), p. 15-41.

Hood C. (1991) « A Public Management for All Seasons? », Public Administration, 69(1): p. 3-19.

Humphries K.W. (2013), « Not Your Older Brother's Bonds the Use and Regulation of Social Impact Bonds in the United States », Law & Comtempories Problems, 76: p. 433-452.

Husson S. (2016), “L’ADIE expérimente le premier contrat à impact social”, La Croix, November 24

Jolliffe D., Hedderman C. (2014), « Peterborough Social Impact Bond: Final Report on Cohort 1 Analysis », QinetiQ/University of Leicester, Prepared for Ministry of Justice.

Joy M., Shields J. (2013), « Social Impact Bonds: The Next Phase of Third Sector Marketization? », Canadian Journal of Nonprofit and Social Economy Research, 4(2): p. 39-55.

Kamarck E. (2014), The End of Government (As We Know It): Making Public Policy Work, John F. Kennedy School of Government, Harvard University.

Le Pendeven B., Nico Y., Gachet B. (2015), « Social Impact Bonds: un nouvel outil pour le financement de l’innovation sociale », Report Institut de l’Entreprise, novembre.

Mahoney J.T., McGahan A.M. and Pitelis C.N. (2009), « The Interdependence of Private and Public Interests », Organization Science, 20(6): p. 1034-1052.

Maier F., Meyer M. (2017), “Social Impact Bonds and the Perils of Aligned Interests”, Administrative

Sciences, 7(3), 24

Mathiasen D.G. (1999), “The new public management and its critics”, International Public Management Journal, 2(1), p.90-111

Marty F. (2016), « Les obligations à impact social : une nouvelle génération de PPP pour les politiques

sociales? », Discussion paper Series, EPPP DP, Paris

McHugh N., Sinclair S., Roy M., Huckfield L. and Donaldson C. (2013), « Social Impact Bonds: A Wolf in Sheep’s Clothing? », Journal of Poverty and Social Justice, 21(3): p. 247-57.

Ministry of Justice (2015), « 2010 to 2015 Government Policy: Reoffending and Rehabilitation », MoJ Policy paper, Updated 8 May.

Moore M. (1995), Creating Public Value: Strategic Management in Government, Harvard University Press.

Mullen K., Frank R. and Rosenthal M. (2009), « Can You Get What You Pay For? Pay-for-Performance and the Quality of Healthcare Providers », Working Paper n° 14886, NBER, Cambridge, MA.

Musgrave R., Musgrave P. (1989), Public Finance in Theory and Practice. 5th edition, McGraw-Hill Book Co, New York

National Council for Voluntary Organizations (NCVO) (2011), « Payment by Results », (Discussion Paper), London, U.K.: National Council for Voluntary Organizations, November.

NHS (2012), « A Simple Guide to Payment by Results », Department of Health.

Niskanen W. (1971), Bureaucracy and representative government, Transaction Publishers.

Norman J. (2010), The Big Society – The Anatomy of the New Politics. The University of Buckingham Press, Buckingham.

O’Connor R., Neumann V. (2006), « Payment by Results or Payment by Outcome? The History of Measuring Medicine », Journal of the Royal Society of Medicine. 99(5): p. 226–231.

OECD (2004), Modernizing government: The synthesis. Paris: OECD. Paper presented at the thirtieth meeting of the Governance Committee. Written by Alan Matheson, OECD publications. Paris

OECD (2010), « Value for Money in Government Public Administration after New Public Management », OECD publications. Paris

OECD (2015), « Social Impact Investment building the evidence base », OECD publications. Paris

Oehler-Sincai I. (2008), « Strengths And Weaknesses Of The New Public Management (NPM) – Cross-Sectional and Longitudinal Analysis », Paper presented in the Conference organized by the SOG and the QoG Institute, November 13-15, Romania.

Pauly M., Swanson A. (2014), « Social Impact Bonds in Nonprofit Health Care: New Product or New Package? », Working Paper n° 18991, NBER, Cambridge, MA.

Pettus A. (2013), « Social Impact Bonds », Harvard Magazine, July-august.

Philipps A., Anderson J. (2015), « What We Learned From the Nation's First Social Impact Bond », Huffington Post, February 7.

Piketty T., Saez E., Atkinson A. (2009), « Top Incomes in the Long Run of History », Working paper no. 15408, NBER, Cambridge, MA.

Politt C., Bouckaert G. (2004), Public Management Reform. Chapter XI. Oxford University Press.

Porter E. (2015), « Wall St. Money Meets Social Policy at Rikers Island », New York Times, July 28.

Rizzello, A., Carè, R., (2016), Insight into the Social Impact Bond Market: An Analysis of Investors”,

ACRN Oxford Journal of Finance and Risk Perspectives, 5.3: 145-171

Roman J., Milner J., Poethig E., Walsh K. (2015), « Putting Evidence First: Learning From the Rikers Island Social Impact Bond », The Huffington Post, Posted 6 July, updated August 13.

Schinckus C. (2015a), « The Valuation of Social Impact Bonds: An Introductory Perspective With the Peterborough SIB », Research in International Business and Finance, 35: p. 104-110.

Schinckus C. (2015b), « Financial Innovation as a Potential Force for a Positive Social Change: The Challenging Future of Social Impact Bonds », Research in International Business and Finance, Available online 26 November.

Schinckus, C., (2017), “Financial innovation as a potential force for a positive social change: The challenging

future of social impact bonds”, Research in International Business and Finance 39, pp.727–736

Sinclair S., McHugh N., Huckfield L., Roy M. and Donaldson C. (2014), « Social Impact Bonds: Shifting the Boundaries of Citizenship », Social Policy Review: Analysis and Debate in Social Policy, 26: 119.

Social Finance (2011), « A Technical Guide to Commissioning Social Impact Bonds », Social Finance report, November.

Stievenart E., L’Huillier, H. (2017), “Quel avenir pour les contrats à impact social?”, KIMSO Reports, Paris

UK Cabinet Office (2012), « Social impact bonds », Retrieved 12 February 2013, from https://www.gov.uk/social-impact-bonds

Warner M.E. (2012), « Profiting from Public Value? The Case of Social Impact Bonds », Creating Public Value Conference, September 20-22.

Warner M.E. (2013), « Private Finance for Public Goods: Social Impact Bonds”. Journal of Economic Policy Reform, 16(4): p. 303-319.

Yin Robert K. (2012), Applications of Case Study Research, 3rd ed., Thousand Oaks (CA), Sage Publications

Zelnick J.R. (2015), « Social Impact Bonds in Criminal Justice: A Deal We Can’t Refuse? », Prison Legal News, September, August 25.

Haut de page

Annexe

Appendix 1: Social Impact Bonds by geographical area and year of activation

2010

2012

2013

2014

2015

2016

TOTAL

UK

1

13

1

9

8

1

33

US

0

1

3

4

1

2

11

Australia

0

0

2

0

0

0

2

Canada

0

0

0

1

0

0

1

Continental EU

0

0

1

0

10

2

13

India

0

0

0

0

1

0

1

Israel

0

0

0

0

1

1

2

Korea

0

0

0

1

0

0

1

Peru

0

0

0

0

0

1

1

TOTAL

1

14

7

15

20

6

65

Source: Table from Rizzello and Carè, 2016

Appendix 2: Total capital raised by geographical area and year of activation (in USD)

2010

2012

2013

2014

2015

2016

Total per country

North America

-

9,600,000

21,160,000

35,274,210

5,900,000

26,200,000

30,471,669

UK

3,788,166

12,708,539

1,151,266

4,559,436

7,961,209

303,053

22,106,922

Continental Europe

-

-

866,950

207,006

18,675,950

2,255,159

22,005,065

Asia + Israel

-

-

-

8,923

18,785,950

5,150,000

98,134,210

Australia

-

-

22,106,922

-

-

-

110,000

Others

-

-

-

-

110,000

-

23,944,873

Total per year

3,788,166

22,308,539

45,285,138

40,049,575

51,433,109

33,908,212

196,772,739

Source: Table from Rizzello and Carè, 2016

Appendix 3: Examples of press articles about SIBs

Topic

Press articles

Observations a few years after the emergence of first SIBs

Forbes (2014)

Foley (2015) for the Financial Times

Buysschaert (2015) for the Belgian Journal Le Vif

Philipps and Anderson (2015) for The Huffington Post

Peterborough SIB presentations and explanations

Cahalane, 2014 for The Guardian

The Economist, 2014

Barker, 2014 for Independent

Birkwood, 2014 for Third Sector

Rikers Island SIB descriptions after failure

Roman et al. (2015) for the Huffington Post (Roman et al., 2015

Porter (2015) for the NY Times

Cohen (2015) for Third Sector

Cohen and Zelnick (2015) for Nonprofit Quarterly

Appendix 4: List of interviewed people

Category

Profile

Explorative interview

Semi-structured and deepening interview

TOTAL

Finance

CEO of an impact investing fund

1

2

3

Director and person responsible for SIB initiatives in a big co-operative bank

1

0

1

Director and person responsible for SIB initiatives in a big co-operative bank

1

1

2

CEO of a big co-operative bank

1

1

President of a private-equity fund, specialist on impact investing

2

2

Managing director of a private-equity fund, specialist on impact investing

2

2

Managing Director of a new fund for SIBs intended to develop businesses in difficult neighbourhoods

1

1

Persons in charge of social investments in a large French public financial organisation

2

2

Director of an innovation division in a large French public financial organisation

2

2

Investment director of a private-equity fund, specialist on impact investing

2

2

Public Sector

Economic counsellor of the French President of the Republic

1

5

3

Counsellor for social economy, Ministry of Economic Affairs and Finance

1

4

5

Person responsible for social and solidarity in the Treasury General Direction of the French government

1

2

3

Executive in a local authority in south of France, responsible for promotion of SIBs with the European programme “Alp’SIB”

1

1

Official representative for innovation, entrepreneurship and international economics for the French President of the Republic

1

1

2

Counsellor of the Ministry of Economic Affairs and Finance, responsible for economy and real-estate financing

1

0

1

President of a French Department (local government)

1

1

2

Services and experts involved with SIBs: consultants, evaluators, and lawyers

Associate in an international strategy consulting firm, expert on SIB who launched one of the first SIB in continental Europe

1

1

2

Founder of a consulting firm about social impact

1

3

4

Founder of a consulting firm about social impact

1

3

4

Public relations director in a global private company, author of a think-tank report on the "Big Society" movement in the UK

1

0

1

Co-founder of an accountability and consulting firm, specialist on SIBs

2

2

Master's degree student, author of a research report on SIB in Europe

1

1

Co-founder of a consulting firm specialising in social impact evaluation

1

1

Co-founder of a consulting firm specialising in social impact evaluation

2

2

Professor of Law in Assas University

1

1

2

Partner of a large consulting firm – lawyer associate in charge of the Social Economy business-unit

1

1

Partner in a law firm, specialist on public contracts

1

1

NGOs, social firms and social innovators

CEO of a big NGO, specialising in physical activities for seniors

1

1

In charge of partnerships in a big NGO, specialising in physical activities for seniors

1

1

NGO person involved in services for unemployed people

1

1

Person responsible for a Finnish SIB directed at refugee unemployment

1

1

Person responsible for innovation in a big NGO, specialising in physical activities for seniors

2

2

TOTAL: 33

14

48

62

Haut de page

Pour citer cet article

Référence électronique

Benjamin Le Pendeven, « Social Impact Bonds: A New Public Management Perspective », Finance Contrôle Stratégie [En ligne], NS-5 | 2019, mis en ligne le 27 février 2019, consulté le 22 août 2019. URL : http://journals.openedition.org/fcs/3119 ; DOI : 10.4000/fcs.3119

Haut de page

Auteur

Benjamin Le Pendeven

Audencia Business School
blependeven@audencia.com

Articles du même auteur

Haut de page

Droits d’auteur

Tous droits réservés

Haut de page