Charitable assets: social outcomes, financial values, and the new, nonprofit funding regime

ABSTRACT

Amidst growing signs of inequality, poverty, and marginalization, there have been calls for a new approach to funding the social sector organizations tasked with addressing these challenges. Rather than paying for services, governments and philanthropists are being encouraged to fund programs based on their ‘outcomes.’ This paper explores this growing movement around ‘outcomes-based funding’ (OBF) suggesting that outcomes in this context embody a distinctly financial logic and reflect an effort to turn the work of charities and nonprofits into a type of pseudo asset. The paper teases out these dynamics and their implications based on one particular form of OBF, the social impact bond, a financial instrument which uses private capital to fund social programs and calculates returns based on program outcomes. While SIBs have struggled as a market, the operations underlying these projects and informing the production of outcomes as investable assets have been carried forward into non-SIB work informing flows of public and philanthropic capital and embodying the practices of the larger OBF ecosystem. As a window into the new, outcomes-based nonprofit funding regime, the paper offers a unique lens and set of critical tools for exploring the relationship between capital, the social sector, and poverty governance.

KEYWORDS:

• Outcomes-based funding
• social sector
• financialization
• assetization
• poverty management

Introduction

Within the world of social programs, it is increasingly difficult to escape references to ‘outcomes.’ Talk of outcomes is ubiquitous, extending from public sector contracting, to philanthropy, to markets in ‘social’ and ‘impact’ investing, and ultimately informing what is presented as a new paradigm for funding social programs: ‘outcomes-based funding’ (OBF). The core premise animating this new world of outcomes is that funding for social sector agencies working in fields such as homelessness, early education, and employment training should be based not on the services these agencies provide, but rather the outcomes they produce – that is, their ability to produce quantifiable and monetizable change in social conditions. A growing ecosystem of advisory firms and support programs has emerged around this vision, dedicated to helping providers become more outcomes-focused.

This paper explores this growing outcomes mindset and its implications for the social sector. In one respect, the focus on outcomes is nothing new, first emerging in the 1960s and representing a signature of the neoliberal and New Public Management reforms of the 1980s and 1990s where outcomes were synonymous with an economic logic of efficiency and cost-effectiveness. And yet, outcomes in the contemporary context reflect a slightly different logic, one which is distinctly financial rather than economic in nature and which is informed by an effort to capture the value of social programs and to articulate that value proposition as a return on investment. The paper teases out this financial logic based on the results of an in-depth study of SIBs and OBF in Canada, the US, and the UK. Drawing from literature on the social studies of valuation and assetization, which has taken up the question of value and the distinct forms of valuation associated with the asset form (Muniesa et al. 2017; Doganova 2018; Birch and Muniesa 2020), it is argued that outcomes in the context of OBF are informed by an ‘asset rationale’ and reflect an effort to turn social programs into a type of charitable asset. The analysis draws from an exemplary case of OBF, the social impact bond, a financial model that uses private capital to fund social programs and deliver returns, paid by government, based on the value of the outcomes produced by providers. While SIBs have struggled as a market, and many SIB advisory firms have moved away from the model (Williams 2020), the techniques used to develop these projects and to turn outcomes into assets have been carried forward into non-SIB work and inform the efforts of key actors in the larger OBF ecosystem. The SIB experiment thus offers a fascinating window into the financialized form of outcomes and an opportunity to trace the financial logic animating the OBF regime. Teasing out this logic is essential to understanding the impact and limitations of this funding system ultimately laying the groundwork for a larger critical project on the social sector and its place in contemporary forms of ‘poverty governance’ (Soss et al. 2011).

Based on these objectives, the paper is organized as follows. The first section provides an overview of the new world of OBF juxtaposing the economic logics informing the outcomes of the past with the financial logics animating those in the present. These financial logics are framed first in terms of a problematic of capital allocation and second in terms of the treatment of outcomes as an indicator of the value of social programs and basis for transforming charitable work into a financial asset. The next section introduces SIBs as a test case for this vision of outcomes-based financing and details three key operations – framing/indexing, targeting/matching, and managing/governing – which have been central to SIB design and have been carried forward into non-SIB work. In view of these operations, the third section makes the case that outcomes are defined by both a distinct form, that of a socio-technical artifact, and a derivative logic, and that it is these features which inform the transformation of charitable work into a pseudo asset which increasingly informs capital flows within the social sector, part-and-parcel of a financialized nonprofit funding regime. The paper concludes by tracing the implications of the analysis suggesting that the view of outcomes as assets opens up a larger set of questions and critiques around the social sector and the nature of social problems work in the contemporary context.

The (new) world of outcomes

In 2017, the New York-based Nonprofit Finance Fund, a Community Development Financial Institution (CDFI), in partnership with the Federal Reserve Bank of San Francisco, published a book entitled, Investing in Results. Consisting of over eighty contributions from key actors in the US social and public sector, including government, philanthropy, advisory firms, and nonprofits, the book advocated for a fundamental shift in how social sector organizations are funded (Bartlett et al. 2017). Rather than funding on the basis of outputs as indicators of agency activity, the argument was that providers should be compensated according to the outcomes produced as a result of these efforts. For example, in the case of homelessness, funding would be based on the outcome of the number of individuals transitioned into permanent supportive housing instead of the output of the number of individuals temporarily housed in a shelter. Outcomes are thus seen to provide a better indication of not only the positive change, but the value produced by social programs. The intent behind Investing in Results was to rally key actors around this vision and to lay the foundation for the transition to the new world of outcomes.

This is far from an isolated effort. Throughout the social and public sector, in a variety of national contexts, there have been calls for the funding of social programs to be tied to outcomes, and the embrace of OBF as the key to addressing intractable social challenges. This is reflected in new forms of public sector contracting including: outcomes-based commissioning and Payment-by-Results (UK), Pay-for-Success and evidence-based policymaking (US), and social outcomes contracts (EU). Talk of outcomes also pervades philanthropic circles, informing efforts to develop and scale promising programs and providers, and animates the growing ranks of nonprofit advisory firms operating in the spaces between government, philanthropy, and the charitable sector and dedicated to transitioning nonprofits towards outcomes. In view of this expanding discourse, infrastructure, and practice of OBF, a key question is exactly what this represents and how it is impacting the social and public sector.

In one respect, the focus on outcomes is nothing new. The language of outcomes has long been used to inform, assess, and rationalize public sector spending. With roots in program budgeting and systems analysis dating back to the 1960s, talk of outcomes expanded dramatically in the 1980s (Bovaird 2014), a hallmark of neoliberal reforms and public sector outsourcing and marketization (Lowe and Wilson 2017). In this context, outcomes reflected a concern with accountability and transparency. They provided a way for agencies to ‘communicate their worth to external critics’ (Barman 2016, 43). Most notably, outcomes embodied an economic logic of efficiency and cost-effectiveness, an ‘economic style of reasoning’ (Berman 2022) geared towards the elimination of waste and the reduction of the costs associated with government programs. While these managerial and economic impulses remain, the way that outcomes are invoked by the likes of NFF and other advocates of OBF reflects a different logic, one that is distinctly financial in nature. To understand this new world of outcomes, we need to tease out this financial logic.

The problematic of capital allocation

One indication of this financial logic is the fact that outcomes are increasingly tied to a problematic not of economic inefficiency, but rather capital allocation. A frequent refrain in accounts of OBF is that the failure of governments and philanthropists to make significant headway in addressing social challenges such as poverty is not due to the complex nature of these problems, nor the shortcomings of past approaches, but rather the failure to properly allocate public and philanthropic capital to the programs that deliver the best results. By extension, the ‘market’ for nonprofit or social capital is deemed to be ‘irrational’ and inefficient, lacking a clear indicator of the value of social programs and thus a mechanism for directing capital to programs and providers. Consistent with this logic of capital allocation, rather than managers seeking out greater economic efficiencies, governments are encouraged to think like investors and to maximize the value and return on their investments (Mennicken and Muniesa 2017; Muniesa and Doganova 2020). The same is true of philanthropy. While implicit within philanthrocapitalism (McGoey 2015) and ‘millennial philanthropy’ (Mitchell and Sparke 2016), this investor mentality has become more pronounced. It is reflected in new approaches such as strategic capital pools,¹ in which multiple funders come together to invest in a group of promising providers, an embodiment of the ‘financial logic of leverage’ (Mitchell and Sparke 2016) and the principles of portfolio diversification and risk mitigation. More generally, the continued growth in the size and scale of philanthropy (McGoey 2015) has meant that there is more capital in need of allocation and thus in search of measures of both value and performance.

This problematic of capital allocation has become more pronounced within the post-financial crisis landscape. The austerity programs that followed the financial crisis of 2007–2009, combined with political climates hostile to public spending, have created additional pressures on governments to rationalize resources. At the same time, concerns around the sustainability of public spending have meant that service users in contexts such as homelessness (Willse 2015) and criminal justice (Aviram 2015) are increasingly viewed not only as cost burdens, but also as a source of prospective savings revealing a financial logic of investing in the present in order to save in the future. Another current running through the post-crisis landscape is social and impact investing which is explicitly rooted in a logic of capital allocation. The core premise of these markets is that financial capital can help to solve societal challenges, and thus produce both a financial and social return, if only it can be ‘unlocked’ and properly deployed. Informed by this logic, philanthropists are not only seeking to redeploy their core assets towards ‘impact’ investments, they are also treating grants as investments (Hebb and MacKinnon 2016, 454). This is further evidence of a financial logic that is distinct from the business logics typically highlighted in accounts of philanthrocapitalism (McGoey 2015; Mitchell 2017).² The mantra of social investment has also extended into the state contributing to the growth of the ‘social investment state’ and further solidifying the view of public spending as a ‘productive investment’ rather than a cost (Laruffa 2018, 689; Baker, Evans, and Hennigan 2020). Finally, the post-crisis landscape is defined by new confluences of public and philanthropic capital including the use of public funds to leverage private investments, and the efforts by philanthropy to engage state agencies as long-term funders allowing them to cash out of their investments and move on to other organizations (Jenkins 2011). Rather than separate funding streams, forms of public and philanthropic capital are thus increasingly intertwined and strategically enjoined.

Ultimately, a key consequence of this financial logic of capital allocation and investor mentality within philanthropy and government is the need for a clear indicator of which providers and programs are most deserving of funding as well as the return on these investments. In other words, there is a growing demand for a measure of the performance and value of social sector organizations. Outcomes have been repurposed to play this role informing investments and coordinating new circuits of public and philanthropic capital. Thus, it is within the post-crisis landscape that outcomes have taken on a different meaning morphing from an economic logic of efficiency and cost effectiveness to a distinctly financial logic of capital allocation. This informs the vision, championed in Investing in Results, of a social capital market in which forms of capital are seamlessly blended and flowing towards outcomes as indicators of social value, ‘Nonprofits shouldn’t have to convince funders that their programs are valuable in order to be funded. If the value is self-evident then their funding should be routine. A market for social outcomes … would be an adaptable, program-agnostic method of distributing resources based on value’ (Galloway 2017, 493).

Outcomes as assets

Informed by a problematic of capital allocation and the perspective of the investor, outcomes also assume a distinctly financial form, resembling a type of asset. Assets have been the focus of a group of scholars who, drawing from the social studies of economization and valuation, have sought to identify the distinguishing features of the asset form and explore what is involved in turning things into assets, what is described as a process of capitalization (Muniesa et al. 2017) or assetization (Birch and Muniesa 2020). In outlining a research agenda for the study of assetization, Birch and Muniesa (2020) argue that assets represent a defining feature of ‘technoscientific capitalism,’ a form of capitalism rooted in a shift from the commodity whose value lies in the moment of exchange between buyers and sellers to the asset whose value lies in future revenue streams. Assets are rooted in a distinct form of valuation. They are informed by the perspective of an investor weighing the risks and opportunity costs of different investment options, and they are tied to financial devices (e.g. business models) and conventions (e.g. discounting and net present value) which help to make the future imaginable and actionable in the present (Doganova and Muniesa 2015; Muniesa et al. 2017; Doganova 2018).

While they do not resemble assets in any strict sense (there are no obvious revenue flows to speak of), outcomes resemble the asset form and thus represent a type of financial asset. The very juxtaposition of outputs and outcomes, so central to accounts of OBF, mirrors the shift from commodities to assets described by Birch and Muniesa (2020). Like commodities, outputs are retrospective, transactional, and tied to services rendered. In contrast, outcomes are transformational and future-oriented, a reflection of the future value produced by providers. Outcomes thus share the temporal features and speculative logics of assets (Adkins, Cooper, and Konings 2020). They are tied to the perspective of an investor seeking to maximize returns and mitigate risks, and they are actualized through devices such as financial models, cost-benefit calculations, and data analytics. Outcomes may thus be viewed as a kind of pseudo asset, a reflection of the extension of the ‘asset rationale,’ if not the exact asset form, into the social and public sector (Chiapello 2015; Mennicken and Muniesa 2017; Muniesa and Doganova 2020), informing what Mennicken and Muniesa (2017) describe as a ‘shift in the vernaculars of “value creation” … The state does thus not “pay” or “fund” any longer. Instead, it “invests” in an accountable manner. It is the idea of a prospective benefit, whose “value” ought to be articulated in the terms of a return of investment, which is key (Muniesa et al. 2017)’ (10). Thus, in the context of the post-crisis drive towards OBF, outcomes embody a distinctly financial logic manifested not only in the focus on capital allocation and an investor mentality, but also a financialized form of valuation (Chiapello 2015) which allows outcomes to serve as indicators of social value and focal points for capital flows. The question is how these outcomes are produced and what it means to turn outcomes, and thus social programs, into assets. The balance of the paper takes up this question using a specific vantage point within the world of OBF: the social impact bond (SIB).

Turning outcomes into assets: the SIB experiment

In March 2010, a new model for funding social programs was introduced in the UK. Termed a ‘social impact bond,’ the idea was that investors would provide upfront capital for a social program and governments would deliver a return depending on the degree of success and based on the cost savings from reduced future demand on public services. While conceived prior to the financial crisis, the SIB model gained considerable traction in the post-crisis environment resonating with the currents of austerity and social investment – a way for governments to achieve fiscal savings, and for investors to realize both financial and social returns. With investor returns directly tied to the production of outcomes, SIBs are the perfect expression of OBF and the transformation of social programs into assets. Indeed, as noted by the CEO of Social Finance US, a key SIB advisory firm, the hope was that SIBs would ‘provide the right infrastructure to make social outcomes an investable asset’ (Palandjian 2012). As a financial instrument, SIBs are rooted in an investor perspective and are valued based on future revenue flows. They are also informed by financial tools and devices such as financial models, cost-benefit analyses, and forms of discounting and calculations of net present value. These financial features have been highlighted in the growing SIB literature (Chiapello 2015; Cooper, Graham, and Himick 2016; Neyland, Ehrenstein, and Milyaeva 2019), and they inform the view that SIBs are indicative of the ‘financialization’ of social and public services (Warner 2013; Sinclair, McHugh, and Roy 2021).

However, while SIBs clearly represent a financial instrument and a conduit for the flow of financial capital into the social sector, SIB practitioners have struggled to translate this vision into reality and to establish a viable market in these investments. The number of projects is impressive as is the global scope of the enterprise; and yet, the rate of growth (particularly in the UK and US) has been much slower than expected (Arena et al. 2016; Maier and Meyer 2017). Moreover, more return-motivated investors have largely steered clear of the market with SIBs supported by large infusions of public and philanthropic capital.

These struggles, and the limits of the SIB market, were documented as part of a larger study of SIBs and OBF in Canada, the US, and the UK. Funded by a three-year grant from the Social Sciences and Humanities Research Council of Canada (SSHRC), this project sought to chart the growth of OBF as exemplified by SIBs and to examine its implications for the social sector. The study was comparative, focused on the UK, US, and Canada as key contexts for SIB and OBF experimentation, and drew from documentary research, attendance at several SIB-themed conferences, and semi-structured interviews. Given the focus on the SIB market more broadly, rather than individual projects (or clusters of projects), the research started with the specialized advisory firms responsible for developing SIBs and building the market (e.g. Social Finance US and UK, and Third Sector Capital Partners) and then expanded outwards to include other key actors within the SIB ecosystem: government officials; social sector providers; philanthropists; and investors. The research also sought to engage with actors outside of the immediate SIB ecosystem, including nonprofit consultancies (e.g. Bridgespan) and philanthropic foundations committed to the larger philosophy of OBF but pursuing strategies other than SIBs including venture philanthropy and data-based approaches. Based on these target groups, an initial pool of respondents was developed using various sources (e.g. reports; press releases from SIB projects; rosters of conference attendees) with additional respondents secured through snowball sampling. In total, 196 interviews were conducted between May 2016 and September 2019, primarily in the cities of Toronto, Boston, and London, homes to the principal SIB advisory firms as well as many investors and providers and thus the de facto centers of the Canadian, US, and UK SIB markets. The breakdown of respondents is provided in Table 1. These interviews were digitally recorded, transcribed, and analyzed using a thematic coding system developed and refined during the course of the study.

Table 1. Respondents by country and sector.

	Canada	US	UK	Total
SIB practitioners	4	18	10	32
Government officials	8	3	7	18
Providers	23	12	17	52
Investors	4	2	7	13
Philanthropists	2	11	10	23
Other	13	22	23	58
Total	54	68	74	196

The study confirmed the more skeptical view of SIBs as a market that had fallen short of expectations. While the reasons for these struggles have been explored elsewhere (Williams 2020), another key finding from the study concerns the evolution of the SIB space. Recognizing the limits of the market, and the inability to survive on SIBs alone, many advisory firms have moved away from the SIB model and are working directly with governments and foundations in a consulting capacity, advising them on how to better invest in social programs.³ Instead of designing standalone projects for outside investors, the focus is on enhancing the impact of existing flows of public and philanthropic capital by making them more outcomes-focused. A senior SIB practitioner suggested that this was the goal all along, ‘[Our mandate] was all about promoting a reallocation of government dollars towards the most efficacious programs … It was never about tapping new sources of financing’ (US Respondent #51). A member of another SIB advisory firm expressed a similar sentiment, ‘We would love all government funding to go toward outcomes. It doesn’t have to involve privately financed [SIBs], but if they can really begin to make resource allocation decisions based on outcomes … that would be extraordinary’ (US Respondent #52).

As part of this advisory work, SIB practitioners have sought to carry forward key elements of the SIB model, what a senior UK practitioner described as ‘disaggregating what it takes to design a good SIB and making sure that we use those individual building blocks to do lots of interesting other work’ (UK Respondent #31). What is critical is that these building blocks, while divorced from financial capital, retain the financial logics and sensibilities of the SIB model including a preoccupation with capital allocation and an investor perspective and, most significantly, the focus on outcomes as reflections of the financial value of social programs. These forms of advisory work thus continue to embody an ‘asset rationale.’ Exploring these extensions provides valuable insights into not only how outcomes are turned into assets, but also how these forms of assetization inform circuits of public and philanthropic rather than financial capital. There are three elements of the SIB model which have been carried forward into non-SIB work and which reflect these forms of assetization: (1) framing and indexing; (2) targeting and matching; (3) managing and governing.

Framing and indexing

A first element involves the framing of social problems, and the adoption of a frame that is more conducive to the production of quantifiable and monetizable outcomes. A key challenge of social programs, and social problems work, is that they are seeking to address issues that are complex and resistant to change. As acknowledged by a respondent with a background in evaluation,

One of the dirty little secrets in social science research … is very little stuff actually works. If you look at the number of fields where there are consistently effect sizes large enough to think you are changing the trajectory of people’s lives, the number of programs like that is vanishingly small …  (US Respondent #56)

Given this challenge, the production of outcomes necessarily involves a process of translation and abstraction, similar to that underlying forms of economization and marketization (Caliskan and Callon 2009), through which complex social issues such as homelessness and unemployment are rendered into more manageable frames and recoded in terms of discrete indicators which may stand as signs of progress. The focus is thus on smaller problem fragments and more limited trajectories of change.

Central to this effort is the identification of proxies or surrogates, measures of social problems which are more amenable to intervention and may be realized over shorter timescales yet are connected (at least in theory) with longer-term outcomes. Academic research, particularly lifecourse analysis and longitudinal studies, has been especially valuable to these efforts, helping to map and model ‘pathways’ of change and revealing correlations between shorter and longer-term outcomes (Albertson et al. 2018, 80). As one respondent explained, what is key is the ability to establish a ‘rational link’ between proxies and longer-term variables of interest, ‘ … as long as you can build a rational link between what you are measuring and the positive benefits that describe success in a given project, then that outcome metric will make sense’ (CDN Respondent #14). They provided the example of reductions in high school dropout rates as a proxy for improved employment outcomes. Another example is ‘housing stability’ which has been used in homelessness projects as a proxy for reduced interactions with the criminal justice system and emergency healthcare. These correlations allow funders to infer that improvements in proxy variables will automatically translate into longer-term, more meaningful outcomes. At the same time, with selected proxies tied to administrative categories and cost data, these project imaginaries also allow for the monetization of outcomes as future savings (e.g. for an employment training program, reductions in social assistance and additional tax revenue).

Ultimately, this process of identifying proxies and correlations resembles what Martin (2015) describes as ‘indexing.’ Developed in reference to a proposal for a ‘stock-market based approach to support of the nonprofit sector’ (100), this involves the use of discrete indicators to represent larger forms of change. He illustrates this using the example of education and the use of graduation rates to ‘stand for the more uncertain question of what it means to be educated and what education delivers to society’ (Martin 2015, 101). The emphasis is thus on indicators of inequality which are viewed as amenable to change within the confines of philanthropic efforts with the focus easily shifting to ‘some other site or symptom of inequality if the initial expression of the underlying structural problem appears to be too intractable to resolve through the available means’ (101). Outcomes, as proxies or surrogates, are rooted in the practice of indexing. Both SIBs and OBF inevitably begin with the work of constructing these indicators and assembling prospective pathways of change.

Targeting and matching

A second operation carried forward from SIB engagements involves the targeting of programs and the effort to create the right match between participants and services. In the context of SIBs, identifying the right target population is a critical first step as this will impact projected cost savings as well as success rates and investor returns. There are several key considerations which inform the construction of target populations. One of these is cost with the emphasis placed on the sub-groups within a population of service users that are deemed to be most costly as consumers of public services and thus to represent the greatest prospective cost savings. A perfect example is the ‘chronically homeless,’ a category developed based on service utilization and cost data and explicitly informed by an economic rationale (Willse 2015, 163). The members of this group are ‘frequent’ or ‘high utilizers’ of public services, most notably criminal justice and emergency medical services. Considerations of cost also inform criminal justice populations (Aviram 2015), and children’s services with ‘children-at-risk’ representing another ‘category of concern envisaged through cost’ (Neyland, Ehrenstein, and Milyaeva 2019, 254). Ultimately, members of these categories are viewed not simply as economic burdens, but as sources of savings and thus value if only they can be transitioned into less costly patterns of service use or, ideally, turned into taxpaying citizens.

Categories such as chronic homelessness and children-at-risk reveal another key principle underlying the construction of target populations: the focus on individuals who are in contact with multiple service domains. The very notion of chronic homelessness emerged from research which tracked the movement of homeless individuals across housing, healthcare, and criminal justice systems. The ability to monitor populations in this way has been enabled by the greater availability of administrative data in digital form and the development of integrated administrative databases, data infrastructures which link different government systems and are capable of tracking the movements of individuals across administrative space (Culhane et al. 2018).⁴ It is these visibilities which have allowed for the construction of new target populations defined by service utilization, and informed efforts to reduce these contact points. Many SIBs have been built around these kinds of target groups including several US projects targeting ‘dually-involved’ populations.⁵ As they have drifted away from SIBs, advisory firms such as Third Sector Capital have invested heavily in these data infrastructures.

Target populations are also informed by notions of risk. Most SIBs are geared towards groups that are deemed to be both costly and ‘high risk.’ The chronically homeless is again a case in point, an ideal target group given not only the scale of potential savings, but also the high likelihood that these costs will be incurred absent any intervention. The same is true of populations of former offenders where being high risk for reoffending is viewed not as a threat to safety or security, but rather a source of value as this group is more likely to reoffend, to be convicted of a more serious offense, and to receive a longer sentence, and thus to yield the greatest savings if this fate can be avoided (Third Sector 2013, 15). Risk is herein re-coded as a form, and source, of value, a reflection of a probabilistic or anticipatory logic (Adams, Murphy, and Clarke 2009). Increasingly, practitioners are seeking to refine these articulations of risk and risk-based categories using tools such as predictive analytics and machine learning. For a member of a foundation leading the charge on developing these capabilities, the question is ‘how can we disaggregate … how do you break down one target population into risk categories and use predictive analytics to do that?’ (US Respondent #8). The focus is thus on creating more precise articulations of risk and thus calculations of the prospective benefits of interventions.

Finally, the construction of target populations is informed by the effort to create the right match between participants and programs, the idea being that some programs work better for some individuals than for others and that program success depends on finding the right match between the two. This strategy follows from the results of previous evaluations which reveal that programs which yield weak effects when averaged across an entire population produce positive results for specific sub-groups. This wisdom, and the logic of segmentation, is increasingly baked into evaluation methodologies such as ‘micro-targeting’⁶ which seeks to ‘more finely understand for whom the intervention worked and then figure out how to deliver it just to those people and then the people for whom it didn’t work find something else’ (US Respondent #61). The construction of target groups thus involves a form of reverse engineering with groups defined based on available programs and success rates across different categories of participants, akin to what Lakoff (2007) describes as finding the right patient for the drug. This degree of specialization is viewed as critical to producing desired outcomes,

the best way to know that you can produce that outcome is by getting so specialized. It’s like we work with this population with this history of issues, and when we have all those factors line up we know we can produce the outcome. (US Respondent #26)

Ultimately, these forms of targeting and matching speak to the critical role of precision in the production of outcomes, the use of new data systems and analytical capacities (machine learning; predictive analytics) to carve out increasingly specialized target groups whose futures can be accurately anticipated and countered with the right programs. The value of programs lies in these projected and altered fates.

Managing and governing

A final element that is critical to the production of outcomes involves the management of programs and providers. While providers have long been subject to oversight by public and philanthropic funders, another distinguishing feature of SIBs is the intensity and granularity of this management function. Most SIBs have relied on the development of sophisticated data systems, dashboards, and visualization tools yielding ‘high frequency data’ allowing project partners to monitor, in as close to real-time as possible, key indicators including inputs (referral and enrolment rates) and outputs (program completions) believed to correlate with, and to be predictive of, future outcomes. In cases where these indicators are not tracking as expected, providers are pushed to make the necessary adjustments. Enabled by devices such as tablets and apps, this management imperative extends through all aspects of program delivery allowing for the monitoring not only of program cohorts, but also individual clients as they move through projected pathways of change. Staff are also subject to these managerial regimes and are thus enjoined with clients in the production of successful outcomes. This more intense style of performance management was described by respondents as the ‘secret sauce’ (US Respondent #10) of SIBs, the place ‘where the magic happens’ (UK Respondent #55), a way of governing providers and participants that is critical to managing the value of these investments and ensuring that results align with expectations. These same techniques have been carried forward into non-SIB work including ‘active contract management’ in which government agencies manage groups of providers across entire contracting streams (Liebman 2018).

The emphasis on performance management, and managing towards specific outcomes, is also impacting how programs are evaluated. Formal evaluations based on randomized controlled trials (RCTs) have long been viewed as the gold standard in the social sector, valued for their epistemic rigor and perceived ability to offer the most rigorous measure of causality and thus true program impact. It was due to these capabilities that practitioners, particularly in the US, advocated for the use of RCTs in SIB projects with outcomes and thus investor payments dependent on the results of these evaluations (Williams 2021). However, SIB practitioners quickly discovered that RCTs were not only expensive and logistically challenging, but also made it difficult to manage towards specific outcomes given the lack of knowledge regarding the comparison group. A US SIB practitioner bemoaned the lack of feedback from RCTs and the management challenge this created, ‘[Provider x] needs better real-time data of proximate indicators so they can change their behaviors and manage. How do you manage otherwise?’ (US Respondent #30). Recognizing these limits, SIB advisory firms, along with other proponents of outcomes-based financing, have pushed for more data-informed methods such as A/B testing and rapid cycle evaluations imported from the private sector (Manzi 2012).⁷ Rather than one-off, episodic, retrospective assessments, the focus is on conducting a large number of smaller experiments yielding a continuous stream of data which can then inform ‘real-time’ performance management. Several respondents likened this move from RCTs to data-informed methods to a shift from ‘snapshots’ to ‘movies’ (US Respondent #51). Time and temporality are critical here with outcomes rooted in access to continuous data flows and the ability to shorten the distance between intervention and assessment thus blurring measurement and management and capitalizing on the reactivity of measures (Espeland and Sauder 2007).

Ultimately then, another key lesson of the SIB experiment is that the production of outcomes requires not only a more intense form of performance management, but also a shift in the methods and modes of assessment which are increasingly viewed through a managerial lens and driven by a distinct conception of value. In exploring an analogous transition in the context of pharmaceutical trials from RCTs to more flexible alternatives, Helgesson and Lee (2017) note how methods are informed by, and are constitutive of, particular forms of value with alternatives to RCTs privileging flexibility and adaptability over epistemic rigor. In the case of the social sector, it is a financialized form of value which is key and which helps to account for the embrace of methods which use the continuous present to govern towards the future. As noted by Birch and Muniesa (2020), a similar form of governance underlies assets with investors constantly seeking to manage and protect the value of their investments. In each of these respects, governance once again emerges as an essential ingredient, the ‘secret sauce,’ not only of SIBs but OBF more generally,

we look at it as the secret sauce, the thing that is going to change the world, is creating a constant feedback loop to the government about which programs are working … trying to understand if you do continuous monitoring can you improve overall performance of the community. (US Respondent #5)

Socio-technical artifacts, derivative logics, and the new, nonprofit funding regime

Taken together, the operations underlying the SIB model, and extending into the larger world of OBF, reveal two key features of outcomes. First, outcomes possess a distinct form, resembling what may be described as socio-technical artifacts. They are products of, and are derived from, socio-technical networks, capacities, and devices. This view of outcomes follows the central insight from the social studies of economization and valuation that economies, markets, and values are not discovered, but rather are actively performed through the enactment of socio-technical assemblages and the operationalization of calculative devices, what Chiapello (2020), borrowing from Thévenot (1984), describes as ‘investments in form.’ These investments involve the creation of a capacity to produce outcomes, what Power (1997) describes in the context of accounting as ‘auditability.’ This refers to the fact that organizations undergoing financial audits must first be translated into a form that is auditable – that is, they must be represented and rendered visible in the terms and categories of the audit. Similarly, the production of outcomes depends on the translation of social problems (framing/indexing), participants (targeting/matching), and providers (managing/governing) into a form that is conductive to the production of change and the ability to manage towards this end.

These translations have been enabled by new technical capabilities: forms of connectivity provided by integrated administrative databases; higher resolution and more granular views of individuals and their anticipated futures afforded by predictive analytics; and greater visibility of providers yielded by more sophisticated data management systems and forms of continuous assessment. Shifts in scale and across time are also key. The greater precision underlying indexing and targeting shifts the scales according to which problems and progress are defined, a form of magnification which allows smaller changes to appear more significant. Time is likewise critical both in terms of movements across time, projecting the effects of programs into the future while bringing the future back into the present through near-term proxies, and the immediacy of ‘real-time’ assessments which turn evaluation into a management device.

The practice of producing outcomes as socio-technical artifacts is nicely illustrated by a comment from a homelessness provider during a 2017 panel on OBF. Responding to a graph presented by a fellow panelist that framed progress on outcome measures in terms of changes in color, they commented, ‘Coming from the nonprofit world … to think about outcomes and graphics and charts and things in different colors is really exciting to me because when things change colors you know you are moving the needle in this outcomes-based world’ (Bartlett et al. 2017). In the new world of outcomes, progress is thus defined, not in relation to the problem itself, but the problem as rendered in the indicators, scales, and data points of the representational system. All of this helps to account for the constant drumbeat around ‘capacity’ and ‘capacity-building’ within the OBF ecosystem which, in essence, refers to the capacity to produce outcomes as socio-technical artifacts.

In addition to assuming a particular (socio-technical) form, outcomes are also animated by a distinctly financial logic. This includes the features of the asset as a way of valuing the future from the perspective of the present; however, it also encompasses what may be described as a ‘derivative logic.’ Within the SIB literature, authors have likened SIBs to a derivative with the value of these investments derived from the performance of the provider (Sinclair, McHugh, and Roy 2021). However, as Bryan and Rafferty (2014) argue, it is not necessarily the form of the derivative that is critical, but rather a derivative logic defined by a form of disaggregation and aggregation – the decomposition of things (in this case people) into a ‘a set of constituent elements or attributes’ (892) and the re-bundling of these attributes (e.g. risks) to create novel categories. The result is new opportunities for trading based on movements in these aggregated indicators. And yet, while Bryan and Rafferty (2014) equate this derivative logic with flows of financial capital, this very logic also defines the production of outcomes more generally and thus extends to circulations of public and philanthropic capital. The work of Martin (2015) is helpful in tracing the larger echoes of these derivative logics. He argues that the ‘social logic of the derivative’ is evident in non-financial domains such as philanthropy and public polling. In exploring this derivative logic, he points to a similar dynamic of disaggregation and aggregation, while also highlighting the role of risk as well as forms of leverage and arbitrage, ‘small interventions that make significant difference, of a generative risk in the face of generalized failure but on behalf of desired ends’ (52).

Within the context of OBF, indexing is a key expression of this derivative logic, this given its roots in a form of aggregation and the use of discrete indicators to stand in for larger social challenges. Indeed, Martin (2015) refers to indexing as an ‘explicit application of the derivative logic’ (101). The aforementioned forms of indexing also embody a form of leverage, another aspect of the derivative logic noted by Martin (2015), as small changes in proxy measures are imagined (by virtue of correlations projected over time and across space) to have much larger effects. Notions of risk are likewise salient here, central to the imagined trajectories of participants and the value of interventions.

Derivative logics are especially evident in the construction of target groups and the effort to carve up populations into precise tranches. Here individuals are reduced to indicators such as ‘risk,’ ‘cost,’ and ‘need,’ with these designations used to construct new categories that then inform decisions around services and programs. This follows Martin’s (2015) observation that, ‘derivatives treat people not as whole but in parts, less as subjects who must meet a threshold for participation than as attributes of risk that can be profiled, collected, and ranked’ (67). Integrated administrative databases are the perfect expression of these dynamics as determinations of being ‘in need’ or ‘at risk’ are based on the bundling of contacts with multiple service systems, a stitching together of identity fragments to form a composite character which may then be valued and matched to services based on anticipated pathways of change. As explained by a proponent of these data systems, the focus is on ‘linking data from different siloed agencies and entities so that they have a comprehensive picture of people within these target populations’ (US Respondent #8). Another respondent echoed this sentiment referring to how this approach was inspired by the private sector and the kinds of algorithms and customer profiles used by companies such as Amazon to make product recommendations,

If you go to Amazon, they don’t look at your data on what you are browsing in electronics as a separate instantiation of you as opposed to what you look at when you are looking at your movie choices … They say here is somebody and, from what they are interested in, they fit into these larger personality profiles. So here’s how we should interact with them and how we should recommend additional services. The government doesn’t do that. We create profiles within individual agencies basically ignoring what’s going on in other agency data so we only see one facet of any individual’s life … we’re wildly sub-optimizing by trying to optimize on one agency’s viewpoint of a human as opposed to seeing that human as a complete individual with multi-faceted needs (US Respondent #5).

While these strategies purport to offer a comprehensive view of individuals, the resulting figure resembles what Appadurai (2016), borrowing from Deleuze, refers to as the ‘dividual,’ another manifestation of the derivative logic and a reflection of forms of data analysis which ‘atomize, partition, qualify, and quantify the individual so as to make highly particular features of the individual or actor more important than the person as a whole’ (Appadurai 2016, 109). The figure and form of the dividual is reflected in the very nature of outcomes, as revealed by an exchange with a US SIB practitioner who, when asked who the client is in a SIB project, replied, ‘I think we really view the client in many cases as the outcomes for participants which I know is kind of a weird answer’ (US Respondent #45). A strange answer indeed, but one which is telling in terms of where the focus of these efforts ultimately lies. Instead of participants, the emphasis is on changes manifested in the form of the dividual, in reduced contacts with service systems or shifts from one administrative cost category (or tranche) to another. Similarly, program success is defined by the aggregate number of outcomes realized across a population independently of changes in the lives and life chances of program participants (see also Rosamond 2021). The form of the dividual, as an embodiment of the derivative logic, is thus critical to the production of outcomes as charitable assets.

Ultimately, it is through the form of the socio-technical artifact and the logic of the derivative that outcomes have assumed a distinctly financial form, a way to articulate the future value of social programs and to turn the work of social sector providers into a type of investable asset defined by notions of risk, probability, and links between the future and the present. This mode of valuation, or assetization, builds on ‘regimes of anticipation,’ transforming the anticipatory into the speculative and capitalizing on the futures imagined in theories of human capital (Adams, Murphy, and Clarke 2009). It is informed by the capacities of contemporary data analytics not only to project the future, but also to ‘model and financialise the propensities and tendencies of life’ (Amoore and Piotukh 2016, 9). These various elements are reflected in the operations – framing/indexing, targeting/matching, and managing/governing – which are central to the SIB model but which also resonate beyond the ‘SIB space’ informing the larger world of OBF. Coupled with the problematic of capital allocation, and the fact that both governments and philanthropists are increasingly thinking like investors, the view of outcomes as a type of asset points to a significant shift in the nonprofit funding regime,⁸ revealing a system of funding that is increasingly informed by a financial logic and defined by new circuits and flows of public and philanthropic capital constituted in terms of outcomes-based value chains. This is a system of funding which bears the imprint of ‘financialized’ (Langley 2021) or ‘technoscientific’ (Birch and Muniesa 2020) capitalism and reflects the ripple effects of the financial crisis, a sign of how these larger contexts have shaped the social sector not through new channels of financial capital (as imagined in accounts of SIBs), but rather the reconfiguration of existing capital flows. It is as the focal points of this new funding regime that outcomes have diverged from those of the past taking on a new financial life and allowing for the reimagining of the social sector through a distinctly financial lens.

Conclusion

During a conversation with a London-based charity consultant about innovation in the charitable sector, they referenced a nonprofit, Charity Water, which builds water projects in the developing South. This program is unique in its use of technology (gauges and cameras on water pumps) to track water flow allowing donors to monitor the impact of their donations in real-time,

They’ve partnered with Google … to install smart trackers on every water source they build so that donors … can go onto Google Maps, search this code, and it will zoom into street view of this well they built. And it will show the people using it. And then you can click on the link and it shows the water flow every day. So you can see at 10:00am this much water came out (UK Respondent #62).

This initiative is the ideal embodiment of the new form and logic of OBF described in this paper. Rather than the well, or even the water itself, the value of the project is tied to the outcome of future water flow and the ability, thanks to the visibilities afforded by new technologies, to monitor this flow in real-time. A similar logic underlies the funding of social programs as embodied not only in the SIB experiment, but also the larger world of OBF. Within this system, governments and philanthropists are increasingly thinking like investors, seeking to allocate their capital to providers not on the basis of services rendered, but the future value produced by these services. Reflecting this sensibility, outcomes have been repurposed as indications of financial rather than economic value, the product of a distinct form and practice of financial valuation rooted in the intersections between the technical and the financial, between feats of socio-technical engineering and a derivative logic, essential ingredients in the transformation of social programs into investable assets and defining features of the post-crisis nonprofit funding regime.

This account of outcomes and the OBF regime offers a unique and distinctly critical lens for exploring the social sector with the emphasis placed not only on how social sector organizations are funded, and the impact of funding sources and logics on programming and service delivery, but also on how these funding practices are shaping the sector’s role in broader forms, relations, and practices of governance. An extensive literature has examined how charitable and voluntary organizations have been implicated in contemporary forms of ‘poverty governance’ (Soss et al. 2011). These accounts are largely informed by a neoliberal lens and focus on forms and logics of quantification, economization, and marketization (Willse 2015; Kurunmaki, Mennicken, and Miller 2016). More recently, with the advent of SIBs and the growth of social investment, emphasis has been placed on financial capital and the financialization of poverty management (Rosenman 2019; Baker, Evans, and Hennigan 2020). However, while acknowledging these currents, the present analysis instead points to the impact of a distinct form, logic, and practice of valuation, one which is financialized (or assetized) but is not tied to financial capital, flowing instead through circuits of public and philanthropic funding. This suggests that forms of governance within the post-crisis landscape are increasingly shaped and mediated by these articulations of value, by the challenge of capital allocation and the need to demonstrate the value of, and returns on, investments of public and philanthropic capital. A series of critical questions emerge from this perspective. What are the implications of public and philanthropic capital flowing along these outcomes-based value chains? To what extent is this contributing to the intensification of governance and surveillance, or alternatively, more progressive forms of programming with providers freed from the strictures of government contracts? What are the politics underlying these funding practices both in terms of valuation as a ‘political technology’ (Doganova 2018; Muniesa and Doganova 2020) and the internal politics of valuation (Williams 2021) within this space including forms of resistance and tensions between different groups of actors around the vision of OBF?.⁹ In view of other accounts of ‘investable poverty’ (Baker, Evans, and Hennigan 2020), the ultimate question is how these logics and funding modalities are providing for more limited articulations of value, confining the scope of social program work to discernible returns rendered as artifacts of data, management, and evaluation systems.

To answer these questions, and tease out these critical dynamics, further research is needed based on different vantage points within the OBF regime including forms of OBF that extend beyond SIBs. This includes case studies of individual providers and how they have been impacted by the new funding environment. Ideal candidates include Mayday Trust in the UK and Community Solutions in the US, both of which exemplify the logics of OBF. Research is also needed on new techniques and applications including: the impact of micro-targeting and greater specialization in service delivery; the use of machine learning and precision analytics to parse participant data and develop more refined and customized interventions and program pathways (e.g. First Place for Youth); and the development of sophisticated forecasting tools used to conduct ‘virtual policy experiments’ and predict where investments are likely to have the greatest impact (e.g. Social Genome Project). There are also new funding platforms such as Beam which allows donors to fund training programs for specific, named homeless individuals and to track their progress towards employment (Fearn 2019), a further reflection of the form of the dividual and the social logic of the derivative. Finally, there is the question of how these dynamics are playing out across different national contexts. The hope is that these kinds of inquiries, guided by the theoretical tools provided in this paper, will offer further, critical insights into the shifting relationships between capital flows, charitable work, and poverty governance in the post-crisis landscape.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Social Sciences and Humanities Research Council of Canada: [Grant Number 435-2016-1039].

Notes on contributors

James W. Williams

James W. Williams is an Associate Professor in the Department of Social Science at York University in Toronto, Canada. His research focuses on financial regulation as well as the financialization of social and public policy, including the growing influence of finance on the social and charitable sector.

Notes

1 Examples include: Blue Meridian Partners; Acumen; and Living Cities.

2 In their account of ‘millennial philanthropy,’ Mitchell and Sparke (2016) refer to the growing ‘reliance on financialized investment rationalities’ (732). However, this remains underdeveloped with the emphasis placed on market and neoliberal logics.

3 An example is the group at the Harvard Kennedy School which began its life focused exclusively on SIBs and aptly named the SIB Technical Assistance Lab, but was rebranded as the Government Performance Lab as it shifted towards government advisory work.

4 A key proponent of these forms of linked data is Actionable Intelligence for Social Policy, a data lab at the University of Pennsylvania (aisp.upenn.edu).

5 One project targeted individuals at the intersection of homelessness and child welfare systems (Cuyahoga), while another focused on the criminal justice and child welfare systems (Illinois).

6 ‘Micro-targeting’ has been used in contexts such as pharmaceutical trials, political polling, and marketing to precisely target programs or interventions.

7 These techniques are similar to the methods used by retailers such as The Gap to determine product placement.

8 The notion of a ‘funding regime’ is used by Richmond and Shields (2004) to describe the funding environment within the charitable sector. As it is used here, a funding regime is understood more in terms of the system of funding – i.e. the sources and circuitry of capital flows and the logics informing those flows.

9 A perfect example are ongoing battles around the value and merits of RCTs. While some funders have advocated for newer, data-based methods, others continue to push for RCTs.