Investable poverty: Social investment states and the geographies of poverty management

Abstract

The management of poverty is undergoing significant changes with the rise of social investment states. In this context, we examine how governmental concern about the long-term public cost of poverty is increasingly modulating the selection, sequencing and targeting of interventions that seek to manage poverty. Using examples drawn from the management of homelessness in Anglo-American cities, we outline a research agenda related to the objectification, economisation, and subjectification of ‘investable poverty’. These emergent developments at the intersection of social investment and poverty management invite geographers and others to rethink where, when and how poverty management occurs.

Keywords

governance inequality policy poverty social services welfare

I Introduction

Poverty management (sometimes referred to as poverty governance) is an enduring and paradoxical feature of capitalist society. On one hand, poverty management functions to alleviate the immediate hardships of poverty so as to preserve life and quell social disruption. On the other hand, poverty management simultaneously seeks to minimise dependency by coercing the poor onto the lower rungs of the labour market. While concerns about hardship, disorder and dependency still dominate welfare thinking, ‘investment’ is increasingly central to the way poverty management is understood and practised within liberal democracies of the Global North. Consider the following three examples. First, in 2012, the Greater London Authority launched a social service program to address street-based homelessness through the use of a ‘Social Impact Bond’ (Cooper et al., 2016). Social Impact Bonds are financial instruments involving private ‘impact investors’ financing time-limited social service interventions for populations associated with high life-time costs to public agencies, such as the homeless, unemployed youth, and prisoners (Rosenman, 2019; Berndt and Wirth, 2018). With the achievement of negotiated outcomes thought to result in reduced public costs into the future, government repays the private investors, plus profit. If the negotiated outcomes are not achieved, financial burden is said to be absorbed by the investors. Second, a 2012 article in the Stanford Social Innovation Review proposes ‘life bonds’ as a way to alleviate poverty (Attard, 2012). The proposal would involve a tradable bond being linked to 10 per cent of the national population living below the poverty line, its value increasing as bonded lives achieve health, education and other milestones. On maturity, ‘[p]roceeds from the realization of the value of a life bond would be electronically deposited into the registered bank account of the bond owner’ (Attard, 2012; see also Kish and Leroy, 2015). Third, in 2011, the New Zealand government committed itself to reducing the ‘forward fiscal liability’, or life-time public cost, of the welfare-recipient population through an actuarial ‘investment approach’ (Baker and Cooper, 2018). The approach involved focusing on initiatives – such as benefit sanctions, behavioural conditions and targeted services – that might reduce the life-time public cost of the welfare-recipient population, thereby offering fiscal returns-on-investment, absorbed into public coffers.

In recent years a seemingly inexhaustible field of experimentation in investment-focused poverty management has emerged, particularly within the Anglo-American world. Ranging from abstract thought experiments to the grounded work of social service innovation, poverty management has been, and continues to be, transformed by notions and instantiations of the ‘social investment state’ (Morel et al., 2012; Hemerijck, 2017a). Prospering in a post-2008 environment of prolonged fiscal restraint and contraction, the emergent Anglo-American variant of the social investment state melds ‘human capital’ development – the supply-side inculcation of marketable skills and attitude – with labour market attachment and fiscal conservatism (Chertkovskaya et al., 2013). As a rationality of government, social investment involves the state making and/or facilitating return-seeking ‘investments’ in people who are in poverty or at risk of being in poverty. Reimagining certain types of public spending as productive, forward-looking social investments, rather than reparative, present-bound social spending, the social investment state represents, according to Deeming and Smyth (2015: 298), ‘the very latest justification for social policy to guide the development of the economy and society in the twenty-first century’.

Political scientists and public policy researchers have produced a significant body of literature on the social investment state (for edited collections, see Liddle et al., 2012; Hemerijck, 2017b). However, most studies focus on high-level policy discourse at the expense of empirically grounded, theoretically informed analyses of how rationalities of social investment are reconfiguring spaces, practices and subjects of poverty management. Geographers, by contrast, have been central contributors to transdisciplinary literature on poverty management, attending to institutions, policies, programs and practices, both formal and informal, that endeavour to regulate the behaviour of poor people, minimise their most disruptive elements, and thereby ensure social order (Wolch and DeVerteuil, 2001; Greene, 2014; Evans and DeVerteuil, 2018). This is a field of research that is directly engaged with the messy actualities and translations of policy discourse into action, but has not explicitly examined the relationship between social investment rationality and contemporary poverty management.

This paper serves two main purposes. First, we seek to bring geographical research on poverty management into conversation with literature on the social investment state to better understand the emergence and implications of investment-inflected poverty management. Although our intention is, mainly, to bring an understanding of social investment to the geographical poverty management literature, there are reciprocal benefits associated with such a conversation. Second, we show that governmental awareness and concern about the long-term public cost of poverty and poor people increasingly modulates poverty management apparatuses in the Anglo-American world. In particular, considerations of long-term cost have come to modulate the targeting of poverty management interventions (i.e. the eligibility criteria of interventions), the selection of those interventions (i.e. the type and mixture of interventions), and, importantly, the sequencing of those interventions (i.e. the timing and duration of interventions). We argue that emergent social investment rationalities – buffeted by fortuitous alignment with projects of market-favouring governance and fiscal austerity – is encouraging practitioners to rethink poverty management: from present-focused ‘crisis management’ to future-focused ‘investment management’. Such changes make it essential for geographers and other critical social scientists to rethink where, when and how poverty management occurs.

The paper proceeds as follows. In the next section, we evaluate the ongoing ‘turn’ to social investment through a historical discussion of transdisciplinary literature on poverty management, highlighting the contributions of geographers and explaining why the literature’s widespread interpretation of poverty management as a present-focused crisis management apparatus is increasingly divorced from the emergent realities of investment-focused poverty management. Following that, we discuss the history and ideology of social investment states, and explain how poverty management is being rethought and reconfigured, unevenly, as a future-focused investment management apparatus in the Anglo-American context. The final substantive section proposes three avenues for future research – cohering around the objectification, economisation and subjectification of investable poverty – that would allow for a better understanding of contemporary poverty management. We draw from an emerging literature on chronic homelessness and ‘Housing First’ programs for illustrative examples.

II Geographies of poverty management

In situating the rise of social investment rationalities within established academic understandings of poverty management, a historical and geographical view is essential. The modern history of poverty management within the liberal democracies of the Global North – and Anglo-American nations in particular – can be understood in relation to three general epochs: (1) the pre-Keynesian welfare state, or ‘old’ liberalism of early industrial capitalism, (2) the inter- and post-war Keynesian welfare state, and (3) the post-Keynesian or neoliberal workfare state. This final era, emerging in the 1970s, we argue, is witnessing the emergence of a more future-focused, investment model of poverty management. Although the exact form poverty management takes varies considerably from place to place and over time, historical analyses of poverty management within liberal democracies find durable practices and imperatives, particularly in the way welfare states endeavour to relieve hardship and, simultaneously, discipline the poor through requiring them to work (Piven and Cloward, 1993; Hennigan, 2018). Scholars have likewise shown that poverty management has and continues to operate through both relatively ‘caring’ (i.e. charitable relief) and ‘controlling’ (i.e. punitive enforcement) policies.

Poverty management of early industrialising capitalism (the late 19th and early 20th centuries) generally relied on a patchwork of localised relief and punitive practices. Without a significant national policy or funding, such relief tended to manifest on an ad-hoc basis (Katz, 1996). Local organisations reacted to the increasing dislocations of the working class, most starkly evident in transnational migration and urbanisation. Such dislocations, paired with recurrent economic crises, produced unprecedented levels of urban pauperism. The relief arrangements were varied: food donations, shelter, incarceration, indentured servitude, or, most common, placement in poorhouses. For those deemed unable to work (the blind, the old, those maimed by industrial production), more permanent support emerged (e.g. pensions) (Piven and Cloward, 1993). In many cases, these relief arrangements maintained starkly paternalist tactics, endeavouring to ‘cure’ the poor of their mendicancy through inculcating a supposedly absent work ethic (Hopper, 2003). Such relief, while oftentimes instrumental to capital, nonetheless included caring impulses, either in terms of the distribution of food, clothing or shelter (Cloke et al., 2010; Katz, 1996). Other, more forthrightly controlling practices, though, accompanied such charitable efforts. Various laws meted out severe punishment, institutionalisation, and incarceration for the unemployed, beggars and the untethered working class seen to be shirking productive labour (see DiPastino, 2005; Schweik, 2010).

The rise of Keynesian policy in Anglo-American nations drastically reshaped poverty management. This Keynesian shift was a reaction to the crisis of the Great Depression, an era of unforeseen and enduring levels of unemployment and poverty which often led to violent confrontations between unemployed workers (and their families) and the police/military. To quell disorder, the state quickly intervened by offering substantial funding for relief and job programs at the national scale. Further bolstered and enabled by wartime economies, Keynesian policy included a suite of state fiscal and monetary policies, including a ‘welfare state’ that largely embraced demand-side solutions. Such solutions most notably included the legal protection of organised labour, which, through the militant struggle of these unions, demanded even more from capital. In the United States, relief policies were significantly and eventually expanded to reduce racialised poverty (and concomitant urban riots) by buoying the consuming power of those left out of the burgeoning postwar economy, namely ethnic minorities (Katz, 1996: 260). This epoch also witnessed the expansion of state-subsidised housing and other social entitlements, often functioning to de-commodify the labour of workers, particularly the elderly and, to varying degrees, women with children. Paired with the liberalisation of relief was a general turn away from the most obviously punitive and controlling policies, particularly the criminalisation of homelessness and other sorts of visible unproductivity (Mitchell, 2011). Rather than resulting from long-term, technocratic statecraft, then, these Keynesian reforms were primarily responses to recurrent protests, the political fear of violent urban riots and widespread labour unrest (Piven and Cloward, 1979).

Whereas the expansion of relief through the interwar era resulted from the protest of the working class, the contraction of relief is attributable to the newly organised pressures of the capitalist class (Soss et al., 2011: 29). Well-organised and well-funded, pro-business groups reacted to the 1970s crisis of Keynesian political economy by refocusing upon supply-side state intervention while safeguarding capitalist profits (Harvey, 2005). However unevenly, this meant reshaping the poverty management landscapes in countries such as the United States, the United Kingdom, and Canada through two general processes. First, the ‘roll-back’ of the Keynesian welfare state through multiple rounds of deregulation, devolution, privatisation and retrenchment, and, second, the ‘roll-out’ of more paternalistic and supervisory workfare systems alongside revamped penal policies (Peck and Tickell, 2002).

The two-part neoliberal overhaul of poverty management is particularly clear when it comes to regulating the labour market. Means-tested programs formed the foundation of liberal welfare states by providing a relatively caring safety net for unemployed, underemployed, or unemployable individuals. Experiencing the profit squeeze of the 1970s, capitalists demanded a reduction in wages, something accomplished by cutting people from the relief rolls and forcing them into the market (Piven and Cloward, 1993: 348). To this end, the reforms ‘rolled-back’ means-tested benefits and ‘rolled-out’ time-limited ‘workfare’ programs (Peck, 2001). These workfare-style welfare practices have been further recalibrated through experimentation with paternalistic ‘conditional-cash transfers’ that make assistance contingent on behaviours such as educational participation (Peck and Theodore, 2015) as well as the vast spread of ‘job training’ programs (Lafer, 2002). These practices existed alongside others to rehabilitate the poor, including an array of drug treatment sites providing care and containment (DeVerteuil and Wilton, 2009).

Neoliberal reforms to poverty management coincided with deep political-economic shifts in national economies. Just as relief programs were slashed, well-paying jobs became more scarce and union power was severely reduced. Alternatively, capital relocated to less-unionised or otherwise cheaper locales, further undercutting working class wages and, in many cases, the tax base of deindustrialising cities. Unsurprisingly, people fell deeper into poverty within places that no longer had the capacity to offer adequate relief. Crime and other sorts of ‘urban disorder’ predictably emerged. Whereas the urban unrest of the 1960s and 70s was eventually met with expanded relief, this time around it primarily inspired the full force of punitive revanchism. Due to these processes, it is increasingly common to consider how the welfare state and the penal state operate in tandem to manage poverty through the ‘joint action of the punitive welfare-turned-workfare and a diligent and belligerent penal bureaucracy’ (Wacquant, 2009: 290–1), evidenced by the growth of mass incarceration (Gilmore, 2007). The inverse practice of banishing the unwanted is the practice of containing them through ghettoisation (Dear and Wolch, 1987; Wilson, 1987; Wacquant, 2009) amid increased policing and surveillance (Goffman, 2009; Stuart, 2011).

Much of the geographical scholarship on poverty management has focused on this recalibration of poverty management at national and state/provincial scales noting transitions in the welfare state and the convergence or divergence of welfare systems globally (Mohan, 2003; Hamnett, 2011). Developing alongside this scholarship has been an important focus on the local and, more particularly, urban geography of this recalibration in the Anglo-American context (Cope, 2001; DeVerteuil et al., 2002; Peck, 2001; Wolch and Dinh, 2001). Locally-oriented accounts resonate with narratives of urban neoliberalisation, particularly accounts that analyse relationships between urban entrepreneurial governance and attempts to erase visible traces of social disorder at the neighbourhood level (MacLeod, 2011), as well as accounts of how previous welfare settlements, such as public housing and general assistance, continue to serve as an integral foundation for poverty management practices (DeVerteuil, 2015; Fairbanks, 2009). In our view, attempts to understand the situatedness and complexity of poverty management landscapes in the context of urban neoliberalisation must grapple with the way this management manifests as an assemblage of seemingly contradictory practices, such as providing relief and disciplining the poor, at the municipal and neighbourhood scale.

These practices are woven together in historically contingent and spatially distinct ways to form poverty management apparatuses (see Evans and DeVerteuil, 2018). In using the term apparatus, we are referring to what Foucault (1980) called a dispositif: an ensemble of heterogeneous practices that operate in tandem, at a given moment in time, forming a strategic response to an urgent need or problem. ‘Apparatus’ usefully captures the coherence of poverty management landscapes as they adapt to various political pressures across political scales. This coherence is produced as the durable practices and imperatives noted above – particularly efforts to deter, discipline and provide relief to the poor – intersect with local conditions, interests and struggles to temper hardship and quell social disruption. Indeed, recent accounts have focused attention on the ways in which these variegated practices – controlling practices, such as banishment, ghettoisation and incarceration, and caring practices, such as charitable sustenance, welfare assistance, and rehabilitation – augment each other at the local level (Stuart 2014; Hennigan and Speer, 2018) and in the process engender poverty management apparatuses (Evans and DeVerteuil, 2018; DeVerteuil et al., 2009; Wolch and Deverteuil, 2001).

Working in this theoretical vein, several scholars have argued that neoliberal poverty management constitutes a kind of ‘crisis management’ apparatus that contains, controls, and minimises the ‘spillover’ of social disorder in the here-and-now (see Gowan, 2010; Fairbanks, 2011; Stuart, 2014). Gowan (2010: 187) traces the rise of the ‘homeless archipelago’, ‘the rough dormitory-style accommodations of the emergency shelters […] islands of deprivation, mundane and ubiquitous yet socially set apart’. Stuart (2014) examines the policing of these marginal urban spaces in cooperation with helping agencies providing shelter and counselling, chronicling a shifting logic towards ‘recovery management’. Fairbanks’ (2009, 2011) ethnography of the ‘unregulated’ recovery-house system in postwelfare Philadelphia provides insight into the powerful role of recovery practices (and discourses) and the continued significance of welfare (general) assistance which functions as the bedrock of an informal poverty management system that manages unruliness but pushes risk, vulnerability and suffering onto poor subjects. These in-depth, situated accounts, each documenting local manifestations of what we are calling poverty management apparatuses, are important touchstones for the theorisation of poverty management as a concept. These accounts also suggest that urban poverty management is largely a present-focused form of ‘crisis management’: efforts to contain or ameliorate the immediate social and economic spillovers of poverty. Fairbanks (2009: 268), for example, links this orientation to the ‘managed persistence’ of urban poverty, ‘a spatial governmentality that conceals and displaces, rather than corrects disorder and extralegality’.

We argue, though, that the steady rise of social investment rationality is seeing poverty management apparatuses reconceptualised and potentially reconfigured beyond (though still including) present-focused ‘crisis management’ toward future-focused ‘investment management’ within the current epoch of the post-Keynesian or neoliberal workfare state. On the one hand, investment management offers a starkly novel assemblage of practices as it deploys financial instruments (bonds, investments, etc.) paired with less-novel but still recent neoliberal discourses of ‘human capital’, ‘efficiency’ and cost-benefit analysis (Soss et al., 2011: 40). On the other hand, investment management harkens back to the well-established imperatives of pre-Keynesian, liberal poverty management. Specifically, this includes supply-side intervention to simultaneously reduce the number of people using welfare services while producing low-wage workers, either through moral correction (i.e. work ethic) or psychosocial rehabilitation and skills training (i.e. human capital) (Peck, 2001). Social investment strategies, like older forms of poverty management, involve contradictory impulses to both assist and, through the welfare axioms of less eligibility and work requirements, compel the able-bodied poor back into the labour market.

Yet unlike older forms of poverty management, which engaged those contradictory impulses for the purposes of short-term stabilisation of capitalist accumulation and social order, social investment rationalities have turned the gaze of, and impetus for, poverty management toward ways to invest public monies for longer-term public cost savings. What remains crucial is to understand how investment-oriented practices, such as the Social Impact Bond and considerations of ‘forward fiscal liability’ (described below), are rearranging existing poverty management apparatuses, and how such rearranging might alter socio-spatial relations both within and beyond immediate issues of poverty. In the following section, we begin to examine such issues by historicising different variants of the social investment state and clarifying the distinctive character of emerging social investment states in Anglo-American nations.

III Social investment states

Notions of social investment have become increasingly important to the way states conceive of, and justify, interventions that seek to address poverty. Cantillon and Van Lancker (2013: 554) note that ‘it is safe to say that the social investment way of framing problems and solutions concerning social protection is now widely shared across countries’, while Laruffa (2018: 689) states that ‘[s]ocial investment has arguably become the dominant approach through which to frame welfare reform among academics and international organisations’. While we believe such declarations of dominance are over-blown, it is certainly apparent that state agencies, often in partnership with non-profit and for-profit organisations, are engaging in a wide range of experiments with policies, programs and products that frame and deploy (types of) social expenditure as return-seeking investments, targeting citizens who are in poverty or at risk of poverty. In the more specific context of Anglo-America, these efforts are enabled by a range of accounting and financial technologies that are used and promoted by state agencies (see Table 1), including (but not limited to) analyses of fiscal liability and return on investment, programs of hypothecated reinvestment, and outcome-based contracting (some of which, like Social Impact Bonds, involve private finance). Whether referred to as the ‘developmental welfare state’ (a term more common in the Global South) or the ‘social investment state’ (more common in the Global North), Peck and Theodore (2015: 104) describe an emergent state form founded on ‘a new logic of social assistance […] different from the welfare-state rationalities of needs-based entitlement and universal coverage’. Rationalities of the social investment state, by contrast, foreground ‘socioeconomic promotion over social protection and long-term human-capital investment over temporary relief’ (2015: 104).

Table 1.

Selected technologies of the Anglo-American social investment state.

Technology	Possible actors involved	Description	Illustrative example
Fiscal liability calculation	– State only	Actuarial analysis of current and future fiscal impost (or ‘liability’) for the population and/or sub-populations, such as welfare-recipient populations	– Actuarial ‘investment approach’ to welfare in New Zealand (Baker and Cooper, 2018) – Generational accounting in Australia (Spies-Butcher and Stebbing, 2018)
Reinvestment strategies	– State only – State in partnership with private actors (non-profit and/or for-profit)	Redirecting funds from one area of state spending (e.g. prisons) to address the factors that lead to demand for that area of spending (e.g. concentrated disadvantage)	Justice reinvestment in the United States (Tonry, 2011)
Return on investment calculation	– State only – State in partnership with private actors (non-profit and/or for-profit)	Analysis of potential or actual impacts (individual, social, fiscal, economic, etc.) associated with a particular (set of) programs/interventions	Social return on investment (SROI) analysis in the United Kingdom (Arvidson et al., 2013)
Outcome-based contracting (i.e. Pay for Success / Pay for Performance)	– State only – State in partnership with private actors (non-profit and/or for-profit)	States’ payment to service providers contingent on evidence of outcomes (may or may not involve private investors providing up-front finance to service providers, as with Social Impact Bonds)	Social Impact Bonds in the United Kingdom (Edmiston and Nicholls, 2018)

Despite significant changes associated with the increasing prominence of social investment as a rationality of government – particularly in the Anglo-American context, where the ‘productive’ purpose of social spending has long been denigrated, and where much experimentation is occurring – geographers have been relatively slow to respond with research that charts and/or critiques those changes. There is, however, nascent interest in the embedding of private-financial instruments and subjectivities within and through poverty management practices (Rosenman, 2019; Loomis, 2018; Mitchell, 2017; Lake, 2016), and strong interest, more broadly, in the dialectical interaction of states and financial processes (Christophers, 2016; Peck and Whiteside, 2016; Bryant and Spies-Butcher, 2018). For their part, political scientists and policy scholars have compiled a sizeable literature focused specifically on social investment states, in Europe and North America primarily. In the following, we seek to bring these more specific accounts to the attention of a geography readership, outlining the character, history, and ideology of social investment state variants, before discussing the apparent implications of Anglo-American social investment rationalities for poverty management apparatuses.

Social investment discourse draws a distinction between different forms of social expenditure, with ‘productive’ social investments distinguished from ‘consumptive’ social spending (Baker and Cooper, 2018). Social investments are viewed as productive when the type and/or method of social expenditure develops human capital and, thus, generates returns-on-investment (e.g. social, fiscal, economic) over a person’s life. This focus on investment-returns changes the traditional temporal logic of the liberal, Anglo-American welfare state. Jenson (2012: 66) notes that, for the Keynesian welfare state, ‘the here-and-now was the most important timeframe and social citizenship focused on inequalities […] and challenges of the present that would be addressed in the present’. By contrast, for the social investment state, social expenditure ‘must not simply be consumed in the present to meet current needs […] it must be an investment that will pay off, generating returns in the future’ (2012: 66). Hemerijck (2017a: 19) similarly emphasises this future-orientation, noting that ‘[i]n essence, social investment is an encompassing strategy of developing, employing, and protecting human capital over the life course for the good of citizens, families, societies, and economies’. He goes on to identify three different ways that social investments in human capital are made: (1) by building the ‘stock’ of human capital over individual and population-level life courses; (2) by mitigating negative aspects associated with ‘flows’ of contemporary labour-market and life-course transitions; and (3) by maintaining ‘buffers’ of income protection and economic stabilisation.

In general terms, social investment states come in two main varieties (Morel et al., 2012; Mahon, 2013). ‘Heavy’ social investment has been pioneered by Nordic welfare states since the 1960s and involves strong systems of social protection alongside universal social programs (Deeming and Smyth, 2015). ‘Light’ social investment is characterised by weaker systems of social protection and more targeted social programming. The latter, more frugal, variety of social investment was born, intellectually, within liberal Anglo-American welfare states during the ‘Third Way’ ascendency of the 1990s and quickly established itself within international institutions, such as the European Union and the Organisation for Economic Cooperation and Development, even if actual implementation at national and sub-national levels proved, at first, elusive (see Diamond and Liddle, 2012; Mahon, 2013). ‘Light’ social investment emerged at a time when strategies of roll-back neoliberal reform – in particular, the retrenchment and marketisation of activities associated with the social state – began to confront practical and electoral limitations. ‘[S]traightforward neoliberalism had hit an ideational, political and economic wall’ (Jenson, 2010: 65), providing a context and pretext for the harsher edges of state restructuring to be hewn. For policy intellectuals and international institutions, social investment offered a way to shift the focus of reform from present-focused, wholesale withdrawal of social expenditure to future-focused, targeted social expenditure, provided it offered a ‘pay-off’ by preparing citizens for integration into a changing, flexible, knowledge-dominated labour market (Jenson, 2010). For a time, the post-1990s variety of social investment state was largely an abstract notion: part of elite policy consensus, but fitfully applied. Since the Global Financial Crisis of 2008, however, a number of conditions have crystallised – including programs of fiscal austerity and resulting crises of social reproduction, combined with enduring public expectations for social programs, and new-found data analytics capacities – moving social investment from the realm of abstract, elite consensus and increasingly toward actual policy experimentation and institutionalisation (Baker and Cooper, 2018).

One of the key features of scholarly literature on ‘light’ social investment has been disagreement over its political-ideological status (White, 2012; Mahon, 2013). Peck (2011: 170) states that some analysts see the social investment state as ‘a turn toward social amelioration, or even a “softening” of the original Washington consensus line’, while others insist ‘it represents an attempt to secure legitimacy for further rounds of neoliberalisation, by way of devolved or outsourced governance and market-oriented governmentalities’. Similarly, Smyth and Deeming (2016: 674) claim that critics present social investment ‘as a subordination of social policy goals to the market economy; while its champions [position] it as an element of a “post neo-liberal consensus” aimed at reconnecting markets to social goals’. Evaluating these positions is an empirical and theoretical task. Empirically, there is a significant amount of research on policy intentions and strategies, but little on the implementation of social investment ideas and their effects on citizens, which makes assessments of ‘actually existing’ social investment difficult. Theoretically, taking a position on whether (and how) social investment ideas run with or against the neoliberal grain depends, to a large degree, on one’s conceptualisation of neoliberalism. Post-neoliberal diagnoses tend to employ a restrictive conceptualisation of neoliberalism, associated exclusively with ‘roll-back’ reforms that reduce state capacity, while those that see social investment states as deeply connected to neoliberalism see the latter as an adaptive project of market-favouring reform that involves shifting (and often enhanced) state capacities and the production of marketised citizen-subjectivities. Consistent with the second position, Laruffa (2018: 702) argues that social investment ideas ought to be seen as both a partial departure from the ‘less state’, retrenchment mode of neoliberal reform and an experimental step towards a more socially-aware mode of neoliberalism that ‘aims at investing in human capital so that individuals become attractive to employers’. This return of a productive and interventionist role for the state is different from the Keynesian welfare state and its focus on present-focused redistribution based on rights of social citizenship (Jenson and Saint-Martin, 2003). Within social investment discourse, as Jenson (2012: 66) points out, ‘it is acceptable for the state to have a significant role, but only when it is behaving like a good business would, seeking to increase the promise of future returns’. Rather than state-citizen relations being defined by rights and responsibilities, contemporary social investment states reconceptualise citizens as economic assets (or liabilities) in which to invest (or disinvest) with the expectation of increasing fiscal, social and/or economic returns (or, at least, mitigating economic loss).

Expectations of future returns licences states to distinguish between citizens and spatially-demarcated populations based on their investability. Contemporary social investment rationality sees the ‘social’ – and, in particular, the relations between state and citizen – reinterpreted as an economic object and, in so doing, transforms citizens from political actors (with political and social rights) into economic subjects (with different capacities for human capital development and, thus, different potential for investment returns). Within the citizen population, those who are in poverty or at risk of being in poverty are the focus of particularly intense attention because such people are understood to be costly to government (e.g. greater dependence on social assistance payments and publicly-funded services) and their potential for future human capital gains is undermined by conditions of poverty (e.g. poorer health, social and educational conditions than others; greater barriers to employment). Anglo-American social investment states are incentivised to understand and address poverty and its management differently: rather than seeing poor populations as ‘rabbles’ (people regarded as deviant and/or bothersome) to be managed efficiently and/or expediently in the present, they are reframed as a source of public and/or private economic value creation into the future. For those who can be recuperated into the labour market – either because they are young enough, healthy enough or otherwise tractable enough – social investment rationality suggests that states should invest in those people to the extent that they become able to secure their welfare beyond the state in the future. For those who cannot be recuperated into labour, social investment rationality dictates that the long-term costs of managing them be minimised (Willse, 2015; Cantillon and Van Lancker, 2013). With the rise of social investment rationalities, we argue that poverty management apparatuses are being reconceptualised and reconfigured from the present-focused ‘crisis management’ logic familiar in geographical scholarship to future-focused ‘investment management’ logic based around understanding, economising and addressing poverty in ways that minimise future costs and thereby enable returns on investment. Considerations regarding future cost have begun to modulate the selection and sequencing of poverty management practices, influencing decisions about which spatially demarcated poor populations are targeted, when those populations are targeted, and what mixture of practices are deployed to address different groups. Linked to the social investment state, investable poverty has become central to poverty management apparatuses but, at present, geographers and allied social scientists have not sufficiently grasped these changes.

IV Investable poverty

In this section, we narrow our focus to ‘light’ social investment, a variety born within liberal Anglo-American welfare states. While acknowledging that there can be a variety of ways that poverty management practices (e.g. banishment, containment, incarceration, charity, workfare, rehabilitation) coalesce in a given urban context, we focus on one apparatus in particular – the homeless apparatus – and explore the ways it is being made amenable to future-focused investment management. In this case, Dear and Wolch’s (1987) ‘service-dependent ghetto,’ what Gowan (2010: 187) later renamed the ‘homeless archipelago’ – enclaves of emergency shelters forming ‘islands of deprivation, mundane and ubiquitous yet socially apart’ – is being recalibrated through calculations of fiscal liability and return on investment as well as forms of outcome-based contracting. These engender a new problematisation, what we term ‘investable poverty’, that has emerged in conjunction with the targeting of specific groups with new interventions. For instance, in 2016, the city of Denver launched its ‘social impact bond initiative’, which used the Housing First model to reduce the costs of the chronically homeless population. Acquiring investments from banks, socially conscious investors, and non-profit foundations, the initiative expects a net return (profit) of $1 million, the city paying out ‘$15.12 for each day spent in housing minus the number of days that participant spends in jail’ over a five-year period (DSIBI, 2016). To understand the application and implications of social investment rationality for contemporary homeless management apparatuses, it is crucial to understand how emerging forms of poverty knowledge and practice render homelessness, as in Denver, investable. We identify three interrelated ‘domains’ that are involved in the realisation of investable poverty: objectification, economisation and subjectification.

1 Objectifying the homeless

Before homelessness can be translated into investable poverty, the homeless must first be rendered visible as objects of potential investment. Here, the term objectification has a double meaning. First, it refers to the formation of homelessness as an object of knowledge. Forming objects of knowledge requires definition, classification and enumeration: what is homelessness and, by extension, who and how many qualify as homeless? Such questions are answered using agreed-upon definitions and calculative practices (Baker and Evans, 2016). Hence objectification’s second meaning: it refers to the objective analyses of homelessness, analyses yielding comparable measurements, generalisable results, and predictive models. Importantly, this latter meaning of objectification is not independent of the former meaning. It is through the objective analysis of homelessness, and all of the practices involved, that the homeless are rendered visible as objects of what Alice O’Connor (2002) has called ‘poverty knowledge’ and, by extension, potential investment. We argue there is an important epistemic geography linked to poverty knowledge and related processes of objectification, one that invites geographers’ attention. Homelessness in particular has long been subject to ‘bureaucratic ignorance’ and ‘statistical neglect’, hence the ongoing struggle over homeless statistics (see Marquardt, 2016: 304). In this regard, we pose two guiding questions: how is homelessness made knowable and what kind of object are the homeless made into as a result?

Homelessness is objectified using a number of calculative practices – namely, counting, tracking, and modelling – that bring homelessness, in all its complexity, under the eye of positivist science (Stanhope and Dunn, 2011). The practice of counting the homeless, using point-in-time street counts, is perhaps the most established technique (Jocoy, 2013). Overlaid with these enumerative techniques is the practice of tracking interactions between homeless individuals and services over time using longitudinal surveys and administrative data. These tracking studies have proven to be a powerful ‘epistemological grid’ for quantifying a highly mobile population in lieu of the inadequacy of traditional census-taking (Marquardt, 2016). An important touchstone here is the work of Dennis Culhane and Randall Kuhn, two American psychologists who used administrative data to analyse patterns of homeless shelter use over time and showed how homeless people cluster into three groups: the transitionally homeless, the episodically homeless, and the chronically homeless (Culhane and Kuhn, 1998; Kuhn and Culhane, 1998). This step towards classifying the homeless has had significant repercussions. It diverted policy attention towards the chronically homeless who, while smaller in numbers, were shown to consume the most shelter resources (Del Casino and Jocoy, 2008). Importantly, it also energised efforts to count and track the homeless using technologies such as the Homeless Management Information System (HMIS), a database management system developed in the US that ‘provides fixed inquiry fields, including demographic details, categories of need, and kinds of service, in which standardized measures are input’ (Willse, 2015: 110). Equivalent systems bearing family resemblances to US database systems have been developed in Canada, Western Europe and Australia. These systems have become necessary amidst funding regimes that have problematised double-counting, prioritised evaluation, and thus require agencies to use universal data elements in their administrative processes (Baker and Evans, 2016). Finally, the objectification of homelessness has been bolstered by research seeking to model the aetiology of homelessness and, conversely, identify risk factors that can guide prevention efforts. This ‘epidemiological approach’ (Somerville, 2013: 389) reflects a perennial interest in causality. Homelessness is modelled by mapping correlations between causes, such as structural factors and individual vulnerabilities, and effects, such as the incidence of homelessness (Corinth and Rossi-de Vries, 2018; Hanratty, 2017).

Using objective techniques such as counting, tracking and modelling, this poverty knowledge objectifies homelessness in a very particular way: as a population (Evans, 2012; Willse, 2015). Seeing the homeless as a population is significant. Instead of persons to be addressed individually, the homeless – and the chronically homeless in particular – are visible en masse as a collective with its own definitive characteristics, patterns and dynamics. Harkening back to Foucault (1990) and his analysis of biopolitics, this form of embodiment permits investments in issues that surface at the level of population, such as rates of service use, housing retention, morbidity and mortality. Moreover, the constitution of the homeless as a population enables particular styles of governing the homeless, ones that target resources and services to particular sub-populations, such as the chronically homeless, while discounting other factors, such as racism and the commodification of housing (Sparks, 2012; Willse, 2015).

This ‘biopoliticization of homelessness’ (Willse, 2015: 111) has attracted some attention in geography (see Evans, 2011, 2012) and is certainly relevant to the turn towards social investment (see Cooper et al., 2016). However, if the reconfiguring of poverty management via social investment rationalities is to be fully understood, more research on the epistemic geographies linked to these processes of objectification would surely help. Attending to the objectification of homelessness reveals the practices that bring homelessness within new domains of intervention. This objectification process is itself ‘an eventful, material and place-based phenomenon’ (Evans et al., 2016: 250) raising additional lines of questioning: How are homeless populations, as objects of knowledge, socially and materially constituted? What kinds of sites, places, networks and spaces inform how we know what we know about homelessness? How does the epistemisation of homelessness dovetail with patterns of governance of various kinds and at various scales? What has been a ‘splintered geography of calculative access’ (Marquardt, 2016: 308) is now being knitted together through an array of database management technologies that allow counting across a range of service sites and scales (see Willse 2015). These objectification processes are undoubtedly ‘generative political events’ with geographies unto themselves (see Whatmore, 2009). Moreover, as Marquardt (2016: 307) has chronicled, evidence can exhibit ‘political performativity’ at certain scales of governance. The ways in which the objectification of homelessness make possible the biopolitisation of homelessness as a form of investment-focused poverty management and how this unfolds (rather unevenly) across space and place begs further attention. Inquiries along these lines take on particular relevance amidst wider calls, in population geography for example, to interrogate the political valences attached to surplus life and the bio-economy (see Tyner, 2013, 2015, 2016).

2 Economising the homeless

Once poor populations are rendered knowable as epistemic objects, investment-oriented poverty management involves reinterpreting those populations as an economic proposition. Calculations, representations and narratives of cost–benefit and risk–reward insert populations (such as the homeless) and segments of those populations (such as the chronically homeless) into economies of public value, wherein future actions, experiences and outcomes associated with population members become sites of public value creation and maintenance (see Table 1). Studies of homeless governance have begun to show how the management of homeless people – across government and voluntary organisations – has been reshaped by the demands of cost containment and reduction (Willse, 2015; Hennigan, 2017). With the aid of objectifying social science, homeless management efforts are increasingly tailored to sub-populations that are stratified and prioritised according to their cost to the public. The potential cost savings associated with reducing the number of chronically homeless persons has propelled the international expansion of select programs targeted specifically towards the chronically homeless sub-population (Baker and Evans, 2016).

Departing from the treatment-led orientation of traditional homeless service systems, housing-led or ‘Housing First’ programs have quickly become orthodoxy for jurisdictions wishing to address chronic homelessness. Much of this popularity derives from its particular design, wherein clients are immediately offered housing without first having to achieve employment, treatment, or sobriety. Such low barriers render it especially effective at enticing the long-term, hard-to-reach, and thus expensive homeless population into housing. Governments across the political spectrum see Housing First programs as a sound investment: spending public money now to save public money later. In many places, investment in programs focusing on this high-cost segment of the homeless population has gone hand-in-hand with efforts that warehouse or abandon less costly and, accordingly, less investable segments of the homeless population (Mitchell, 2011; Sparks, 2012). For example, under the mayoralty of Michael Bloomberg in New York City, Housing First programs were expanded markedly, helping to reduce the number of chronically homeless persons, yet over the same period, the non-chronic portion of the homeless population ballooned (Baker and Evans, 2016). Here, we see resources directed toward fiscally lucrative chronic homelessness and neglect of the remaining majority of the homeless population.

Such efforts reshape homelessness and its management in terms of public-economic value, paving the way for financial technologies that commodify the prospect of value realisation, within and beyond the public sector. This is apparent, for example, with the rise of ‘outcome-based’ or ‘performance-based’ contracting, whereby public sector agencies make payments to (typically) non-government service providers on the achievement of client-level outcomes (Rees, 2014). In place of traditional procurement, based on payment for the delivery of services, outcomes marketplaces are created, where non-government service providers compete to offer public agencies the most cost-efficient way to achieve outcomes that correlate with creation of public value. Since their inception in 2010, Social Impact Bonds have joined performance-based contracting with the world of ‘social finance’, a term that refers to investments and financial technologies that combine the receipt of economic returns with the achievement of positive social outcomes. Involving investors from philanthropic and corporate sectors, the ‘existence of social finance is predicated on the financial enclosure, or financialization, of previously-public or non-market welfare provision’ (Rosenman, 2019: 141–2). This is a task that, crucially, requires the active involvement of the state in making certain poor populations available for investment-focused experimentation and providing the assurance of its ‘full faith and credit’ that investors will be repaid. Alongside prisoners and unemployed youth, homeless populations have been among the most popular targets of Social Impact Bonds within the Global North. Analysing a Social Impact Bond-financed program for chronically homeless people in London, Cooper et al. (2016: 64) discuss how the high-cost segment of the homeless population are refashioned from rights-bearing citizens into securitised commodity-objects, monitored by abstract metrics and disciplines of investment management (see also Berndt and Wirth, 2018). Commenting on the British context more broadly, Edmiston and Nicholls (2018: 65) attest to the disciplining effects of investment logics and private-investment actors, noting that some service providers believed ‘the degree of micro-management built into the SIB was actually reducing their flexibility to autonomously pursue their social mission’.

The economisation of investable poverty is increasingly central to contemporary poverty management but has largely escaped the attention of poverty management scholars within geography and beyond. However, there are opportunities for productive, two-way engagements with scholars investigating the geography of markets, which have burgeoned in recent years (Berndt and Boeckler, 2011; Bryant and Spies-Butcher, 2018; Cohen, 2018; Gallagher, 2018). These otherwise diverse accounts foreground markets as ‘heterogeneous arrangements of people, things and sociotechnical devices’ that are embedded within – but also constitute – economic, political, social and environmental processes operating at multiple scales (Berndt and Boeckler, 2011: 559). Instead of assuming the forms, functions and effects of markets, geographical researchers are orienting themselves toward marketisation ‘as a process through which goods and services are assigned value, made ready for exchange, and circulated’ (Cohen, 2018: 901). Informed by studies of marketisation, key questions relate to the rationale and manner in which poverty management is economised: How and why do certain types of poverty and spatially demarcated poor people come to be understood as worthwhile investments? What processes, actors, relations and practices are required for investments in the poor to be made and matured? Focusing on these questions would uncover the ways in which markets or economies of investable poverty are ‘sites of struggle’ (following Cohen, 2018) that are reshaping the means and ends of contemporary poverty management.

3 Subjectifying the homeless

With the chronically homeless population defined as an inordinate cost to both governmental and non-governmental/charitable institutions, Housing First – as a policy, program, and/or philosophy – has become a popular means to confront this expensive and intractable population (Del Casino and Jocoy, 2008: 196). Given the lengthened time horizons of investment technologies, such as forward liability calculations and Social Impact Bonds (in the case of Denver, determined by bond contracts), their value is contingent upon keeping Housing First clientele housed and off the streets. As mentioned above, clients are not initially required to demonstrate sobriety, gain employment, or achieve ‘mental health’ to receive housing (Tsemberis, 2010). Though this renders Housing First housing especially appealing to many, without first addressing such issues – issues that are assumed to have caused their homelessness in the first place – there is risk that this population will voluntarily or involuntarily return to the street.

Such a risk amid these broader program goals demands a particular sort of subjectification or, to put it differently, a model of case management that encourages and/or coerces clients to act and think in certain ways. Denver’s bond initiative, for example, calls for a suite of case management techniques. Together, they are to:

(i) address barriers to housing stability, (ii) manage mental illness, (iii) reduce interaction with the criminal justice system, and (iv) improve health outcomes. Services will include intensive case management, crisis intervention, substance use counseling, mental health treatment, peer support, skills building, connection to primary care, and various other services identified as appropriate to the client’s goals. (DSIBI, 2016)

Taken together, this comprehensive array of services functions to cut costs by keeping clients in (relatively cheap) housing and off the (relatively expensive) streets.

Critical appraisals of Housing First likewise highlight the lengths to which case management will go to ensure ‘housing stability’. The first issue is that of involuntary or coerced treatment for clientele, of getting clientele to abide by these ‘wrap-around’ case management services. In their analysis of the Housing First archetype, ‘Pathways to Housing’, Atherton and McNaughton Nicholls (2011: 773) found that case managers ‘retain a degree of paternalism in providing support to clients’ who ‘may have little choice but to comply with the support staff when they [staff] hold access to their income’, including their very housing. Evans (2012: 196), in his analysis of a program similar to Housing First, describes how the urging of now-housed clientele to reduce their alcohol use ‘gave rise to a new site of enclosure, one that disciplines “deviant” bodies and behaviours in a more efficient way’ and subjects ‘residents to a different type of professional surveillance and observation’, a situation foreshadowed by Klodawsky (2009) in an early evaluation of Housing First. Hennigan (2017) offers an in-depth account of Housing First case management techniques in two programs. As with Evans (2012) and Atherton and McNaughton Nicholls (2011), Hennigan describes how Housing First case management is, through its control of housing and threat of eviction, able to subtly coerce the ex-chronically homeless individual toward self-reformation. Rather than a therapeutically laissez-faire form of case management, Housing First clientele were repeatedly urged to be healthier (physically and mentally), wealthier (through employment and/or state benefits), and socially supported – as outlined by a series of diagnostic tools and assessments characterising a ‘self-sufficient’ adult – in order to achieve housing stability (Hennigan, 2017).

Urging voluntary treatment and self-reformation for clientele presents a second, rather paradoxical, aspect to the subjectification of Housing First’s clients: that of time-limited – rather than permanent – housing. As is the case with Denver’s bond initiative, Housing First programs often operate through private, subsidised housing systems rather than the more robust forms of public housing akin to Keynesian welfare state. Moreover, some programs (including Denver’s bond initiative) seek to ‘scatter’ Housing First clientele around the city within private-market rentals. Accordingly, recipients’ tenancy depends upon continued funding and willing landlords (Baker and Evans, 2016). Such a system can and does routinely lead to fixed term tenancies wherein ostensibly ‘permanent’ housing is only ever guaranteed months at a time (Hennigan, 2017). Anderson-Baron and Collins’ (2018) recent article specifically explores this issue. Analysing the practices of Housing First programs in Alberta, Canada, they found that service providers, working with limited resources, pressured by funders, and eager to get more of the chronically homeless population off the street, regularly pressured clientele to find housing elsewhere. Of the ten agencies studied, eight included requirements or expectations that clientele ‘graduate’, or leave the program, thereby freeing space and resources for those still expensively living on the street (Anderson-Baron and Collins, 2018: 600). Such practices leave ex-Housing First clients at risk of future homelessness, particularly in areas with escalating rents (a key driver of homelessness in the first place). Further, and as stressed by Hennigan (2017), these time limits compel service providers to subject clients to greater forms of paternalistic reformation, reflecting not so much a moralistic desire to normalise as a genuine effort to prepare clients for the less-than-caring spaces of unsubsidised private housing, austere welfare systems, and a precarious low-wage labour market. Investing in Housing First as a means to reduce the ‘consumption’ of governmental services by the chronically homeless, in other words, often does not stop once individuals are placed in Housing First programs; it continues as those same individuals are pushed into the private housing market completely, until – in terms of housing at least – they do not cost the state a thing.

This form of subjectification, though, risks undercutting what makes Housing First an attractive and, in many senses, effective form of poverty management: its capacity to reduce the chronically homeless population rather than, as is often the case with emergency shelters, continuously (and expensively) cycling the population into and out of institutionalised spaces. This limitation, though specific to Housing First, presents questions pertinent to other forms of investable poverty management: How, when, and at what point might forms of subjectification begin to undercut the very foundations/principles that rendered it possible and attractive in economic terms? How do social investment strategies operate within a broader landscape of welfare retrenchment? In what other ways might forms of objectifying, economising, and subjectifying various poor populations interact, contradict, and change over time and space?

V Conclusion

Framed by the increasing prominence of investment thinking and investment practices in the management of poverty, this paper has discussed how governmental awareness about the long-term public cost of poor populations is beginning to reshape apparatuses of poverty management in the Anglo-American world. Investable poverty – understood and realised through practices of objectification, economisation and subjectification – has become an emergent concern for policy-makers, service providers and financiers, offering the tantalising possibility of achieving positive social outcomes through fiscally-prudent ‘social investments’. By bringing geographical accounts of poverty management into dialogue with literature on the social investment state, the paper situated recent changes within the long historical arc of poverty management efforts in Anglo-American nations. Where older forms of poverty management were generally short-term in their outlook – offering temporary, putative fixes for crises of capitalist accumulation and social disorder – ‘light’ social investment rationalities have augmented this short-term outlook with a future-oriented, long-term agenda that is consistent with the economic and moral goals of neoliberal poverty management.

Geographers are ideally positioned to analyse these changes, given the sizeable body of literature on poverty management that has amassed in recent years and their attention to varying logics, goals and spatialities that are brought to bear on poor populations. However, there remains much for geographers and scholars of poverty management to understand about the infiltration of social investment rationality into poverty management apparatuses. We conclude by reiterating a research agenda for deepening geographical understandings of poverty management apparatuses. First, the concept of local poverty management apparatuses, with its emphasis on heterogeneity, synthetic processes and hybridisation, invites geographers to further examine the formation of actually-existing apparatuses, their conditions of emergence, their spatial and historical contingency, and their interrelationships (see Evans and DeVerteuil, 2018). Of specific importance is addressing the complexity of these apparatuses in relation to ongoing processes of neoliberalisation while not discounting their potential as spaces of resistance and potential social transformation and chronicling the differences within the liberal democracies of the Global North and beyond (see DeVerteuil, 2015). Second, as argued in this paper, taking stock of poverty management in the 21st century will have to address the increasing significance of investable poverty and the social investment technologies that render it visible. We point to three domains that might help orient geographers as they trace how social investment rationalities and technologies are reconfiguring poverty management apparatuses. This includes unpacking: (1) geographies of knowledge production that render certain target populations visible, (2) geographies of social investment that engender public value creation and maintenance, (3) geographies of subjectification that paternalistically aim to reform the poor into self-sufficient subjects of the market. In pointing to these domains we do not mean to suggest that these geographies are the same everywhere nor do they unfold effortlessly, absent of any friction or tension. Of central importance is the political nature of knowledge (and non-knowledge) and the uneven geographies of poverty management that may result (see Marquardt, 2016). But by drawing attention to the centrality of investable poverty and offering a suite of analytical/empirical entry points (objectification, economisation and subjectification), the paper hopes to spark further conversations and analyses attuned to the changing practices and spatialities of contemporary poverty management.

Footnotes

Acknowledgements

We wish to thank the reviewers and Christian Berndt for helping us refine the paper’s contribution. The paper benefited from feedback at the New Orleans AAG and seminars within the geography departments of the University of Sydney and Victoria University Wellington.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Tom Baker's contribution was supported by a New Zealand government-funded Marsden grant, administered by the Royal Society of New Zealand.

ORCID iD

Tom Baker

References

Anderson-Baron

Collins

(2018) Not a ‘forever Model’: The curious case of graduation in Housing First. Urban Geography 39(4): 587–605.

Arvidson

Lyon

McKay

Moro

(2013). Valuing the social? The nature and controversies of measuring social return on investment (SROI). Journal of Public Finance and Public Choice 4(1): 3–18.

Atherton

McNaughton

Nicholls

(2011) Housing first: Considering components for successful resettlement of homeless people with multiple needs. Housing Studies 26(5): 767–777.

Attard

(2012) Life bonds: Linking returns to life outcomes. Stanford Social Innovation Review. Available at: https://ssir.org/articles/entry/life_bonds_linking_returns_to_life_outcomes (accessed 27 April 2019).

Baker

Cooper

(2018) New Zealand’s social investment experiment. Critical Social Policy 38(2): 428–438.

Baker

Evans

(2016) ‘Housing First’ and the changing terrains of homeless governance. Geography Compass 10(1): 25–41.

Berndt

Boeckler

(2011) Geographies of markets: Materials, morals and monsters in motion. Progress in Human Geography 35(4): 559–567.

Berndt

Wirth

(2018) Market, metrics, morals: The Social Impact Bond as an emerging social policy instrument. Geoforum 90(January): 27–35.

Bryant

Spies-Butcher

(2018). Bringing finance inside the state: How income-contingent loans blur the boundaries between debt and tax. Environment and Planning A: Economy and Space. DOI: 10.1177/0308518X18764119.

10.

Cantillon

Van Lancker

(2013) Three shortcomings of the social investment perspective. Social Policy and Society 12(4): 553–564.

11.

Chertkovskaya

Watt

Tramer

Spoelstra

(2013) Giving notice to employability. Ephemera: Theory & Politics in Organization 13(4): 701–716.

12.

Christophers

(2016) Geographies of finance III: Regulation and ‘after-crisis’ financial futures. Progress in Human Geography 40(1): 138–148.

13.

Cloke

May

Johnsen

(2010) Swept Up Lives: Re-Envisioning the Homeless City. London: Wiley.

14.

Cohen

(2018) Between perfection and damnation. Progress in Human Geography 42(6): 898–915.

15.

Cooper

Graham

Himick

(2016) Social Impact Bonds: The securitization of the homeless. Accounting, Organizations and Society 55: 63–82.

16.

Cope

(2001) Between welfare and work: The roles of social service organisations in the social regulation of labour markets and regulation of the poor. Urban Geography 22: 391–406.

17.

Corinth

Ross-de Vries

(2018) Social ties and the incidence of homelessness. Housing Policy Debate 28(4): 592–608.

18.

Culhane

Kuhn

(1998) Patterns and determinants of public shelter utilization among homeless adults in Philadelphia and New York City. Journal of Policy Analysis and Management 17: 23–43.

19.

Dear

Wolch

(1987) Landscapes of Despair. Princeton: Princeton University Press.

20.

Deeming

Smyth

(2015) Social investment after neoliberalism: Policy paradigms and political platforms. Journal of Social Policy 44(2): 297–318.

21.

Del Casino

Jocoy

(2008) Neoliberal subjectivities, the ‘new’ homelessness, and struggles over spaces of/in the city. Antipode 40(2): 192–199.

22.

DeVerteuil

(2015) Resilience in the Post-Welfare Inner City: Voluntary Sector Geographies in London, Los Angeles and Sydney. Bristol: Policy Press.

23.

DeVerteuil

Wilton

(2009) Spaces of abeyance, care and survival: The addiction treatment system as a site of ‘regulatory richness’. Political Geography 28(8): 463–472.

24.

DeVerteuil

May

Von Mahs

(2009) Complexity not collapse: Recasting the geographies of homelessness in a ‘punitive’ age. Progress in Human Geography 33(5): 646–666.

25.

DeVerteuil

Woobae

Wolch

(2002) New spaces for the local welfare state? The case of general relief in Los Angeles County. Social and Cultural Geography 3(3): 229–246.

26.

Diamond

Liddle

(2012) Aftershock: The post-crisis social investment welfare state in Europe. In: Liddle

Palier

Palme

(eds) Towards a Social Investment Welfare State? Ideas, Policies and Challenges. Bristol: Policy Press.

27.

DiPastino

(2005) Citizen Hobo: How a Century of Homelessness Shaped America. Chicago: University of Chicago Press.

28.

DSIBI (Denver Social Impact Bond Initiative) (2016) Denver Social Impact Bond Initiative: Permanent Supportive Housing (Summary). Available at: https://www.denvergov.org/content/dam/denvergov/Portals/344/documents/Denver_SIB_Summary.pdf (accessed 27 April 2019).

29.

Edmiston

Nicholls

(2018) Social Impact Bonds: The role of private capital in outcome-based commissioning. Journal of Social Policy 47(1): 57–76.

30.

Evans

(2011) Exploring the (bio)political dimensions of voluntarism and care in the city: The case of a ‘low barrier’ emergency shelter. Health & Place 17: 24–32.

31.

Evans

(2012) Supportive measures, enabling restraint: Governing homeless ‘street drinkers’ in Hamilton, Canada. Social & Cultural Geography 13(2): 185–200.

32.

Evans

Collins

Anderson

(2016) Homelessness, bedspace and the case for housing first in Canada. Social Science & Medicine 168: 249–256.

33.

Evans

DeVerteuil

(2018) The political spaces of urban poverty management. In: Jonas

Miller

Ward

Wilson

(eds) Handbook on Spaces of Urban Politics. New York: Routledge.

34.

Fairbanks

(2009) How it Works: Recovering Citizens in Post-Welfare Philadelphia. Chicago: University of Chicago Press.

35.

Fairbanks

(2011) The politics of urban informality in Philadelphia’s Recovery House Movement. Urban Studies 48(12): 2555–2570.

36.

Foucault

(1990) The History of Sexuality. Volume 1, An Introduction. New York: Vintage.

37.

Foucault

(1980) Power/Knowledge: Selected Interviews & Other Writings 1972–1977. New York: Pantheon Books.

38.

Gallagher

(2018) The business of care: Marketization and the new geographies of care. Progress in Human Geography 42(5): 706–722.

39.

Gilmore

(2007) Golden Gulag: Prisons, Surplus, Crisis, and Opposition in Globalizing California. Los Angeles: University of California Press.

40.

Goffman

(2009) On the run: Wanted men in a Philadelphia ghetto. American Sociological Review 74: 339–357.

41.

Gowan

(2010) Hobos, Hustlers and Back-Sliders: Homeless in San Francisco. Minneapolis: University of Minnesota Press.

42.

Greene

(2014) Managing poverty, managing dissent: Homeless politics and collective action in London. Policy & Politics 42(3): 315–331.

43.

Hamnett

(2011) The reshaping of the British welfare system and its implications for geography and geographers. Progress in Human Geography 35(2): 147–152.

44.

Hanratty

(2017) Do local economic conditions affect homelessness? Impact of area housing market factors, unemployment, and poverty on community housing rates. Housing Policy Debate 27(4): 640–655.

45.

Harvey

(2005) A Brief History of Neoliberalism. Oxford: Oxford University Press.

46.

Hemerijck

(2017a) Social investment and its critics. In: Hemerijck

(ed.) The Uses of Social Investment. Oxford: Oxford University Press.

47.

Hemerijck

(ed.) (2017b) The Uses of Social Investment. Oxford: Oxford University Press.

48.

Hennigan

(2017) House broken: Homelessness, housing first, and neoliberal poverty governance. Urban Geography 38(9): 1418–1440.

49.

Hennigan

(2018) From Madonna to Marx: Towards a re-theorisation of homelessness. Antipode 51(1): 148–168.

50.

Hennigan

Speer

(2018) Compassionate revanchism: The blurry geography of homelessness in the USA. Urban Studies. DOI 10.1177/0042098018762012.

51.

Hopper

(2003) Reckoning with Homelessness. Ithaca: Cornell University Press.

52.

Jenson

(2010) Diffusing ideas for after neoliberalism: The social investment perspective in Europe and Latin America. Global Social Policy 10(1): 59–84.

53.

Jenson

(2012) Redesigning citizenship regimes after neoliberalism: Moving towards social investment. In: Morel

Palier

Palme

(eds) Towards a Social Investment Welfare State? Ideas, Policies and Challenges. Bristol: Policy Press, 61–87.

54.

Jenson

Saint-Martin

(2003) New routes to social cohesion? Citizenship and the social investment state. Canadian Journal of Sociology 28(1): 77–99.

55.

Jocoy

(2013) Counting the homeless: The culture of quantification in American social policy. Cultural Geographies 20(3): 397–403.

56.

Katz

(1996) In the Shadow of the Poorhouse: A Social History of Welfare in America (10th anniversary edn). New York: Basic Books.

57.

Kish

Leroy

(2015) Bonded life: Technologies of racial finance from slave insurance to philanthrocapital. Cultural Studies 29(5–6): 630–651.

58.

Klodawsky

(2009) Home spaces and rights to the city: Thinking social justice for chronically homeless women. Urban Geography 30(6): 591–610.

59.

Kuhn

Culhane

(1998) Applying cluster analysis to test a typology of homelessness by pattern of shelterutilization: Results from the analysis of administrative data. American Journal of Community Psychology 26: 207–232.

60.

Lafer

(2002) The Job Training Charade. Ithaca: Cornell University Press.

61.

Laruffa

(2018) Social investment: Diffusing ideas for redesigning citizenship after neo-liberalism? Critical Social Policy 38(4): 688–706.

62.

Lake

(2016) The subordination of urban policy in the time of financialization. In: DeFilippis

(ed.) Urban Policy in the Time of Obama. Minneapolis: University of Minnesota Press.

63.

Liddle

Palier

Palme

(eds) (2012) Towards a Social Investment Welfare State? Ideas, Policies and Challenges. Bristol: Policy Press.

64.

Loomis

(2018) Rescaling and reframing poverty: Financial coaching and the pedagogical spaces of financial inclusion in Boston, Massachusetts. Geoforum 95: 143–152.

65.

MacLeod

(2011) Urban politics reconsidered: Growth machine to post-democratic city? Urban Studies 48(12): 2629–2660.

66.

Mahon

(2013) Social investment according to the OECD/DELSA: A discourse in the making. Global Policy 4(2): 150–159.

67.

Marquardt

(2016) Counting the countless: Statistics on homelessness and the spatial ontology of political numbers. Environment and Planning D: Society and Space 34(2): 301–318.

68.

Mitchell

(2011) Homelessness, American style. Urban Geography 32(7): 933–956.

69.

Mitchell

(2017) Metrics millennium: Social impact investment and the measurement of value. Comparative European Politics 15(5): 751–770.

70.

Mohan

(2003) Geography and social policy: Spatial divisions of welfare. Progress in Human Geography 27(3): 363–374.

71.

Morel

Palier

Palme

(2012) Beyond the welfare state as we knew it? In: Morel

Palier

Palme

(eds) Towards a Social Investment Welfare State? Ideas, Policies and Challenges. Bristol: Policy Press.

72.

O’Connor

(2001) Poverty Knowledge: Social Science, Social Policy and the Poor in Twentieth-Century U.S. History. Princeton: Princeton University Press

73.

Peck

(2001) Workfare States. New York: The Guilford Press.

74.

Peck

(2011) Global policy models, globalizing poverty management: International convergence or fast-policy integration? Geography Compass 5(4): 165–181.

75.

Peck

Theodore

(2015) Fast Policy: Experimental Statecraft at the Thresholds of Neoliberalism. Minneapolis: University of Minnesota Press.

76.

Peck

Tickell

(2002) Neoliberalizing space. Antipode 34(3): 380–404.

77.

Peck

Whiteside

(2016) Financializing Detroit. Economic Geography 92(3): 235–268.

78.

Piven

Cloward

(1979) Poor People’s Movements: Why They Fail, How They Succeed. New York: Vintage Books.

79.

Piven

Cloward

(1993) Regulating the Poor: The Functions of Public Welfare. New York: Vintage Books.

80.

Rees

(2014) Public sector commissioning and the third sector: Old wine in new bottles? Public Policy and Administration 29(1): 45–63.

81.

Rosenman

(2019) The geographies of social finance: Poverty regulation through the ‘invisible heart’ of markets. Progress in Human Geography 43(1): 141–162.

82.

Schweik

(2010) The Ugly Laws: Disability in Public. New York: NYU Press.

83.

Smyth

Deeming

(2016). The ‘social investment perspective’ in social policy: A longue durée perspective. Social Policy & Administration 50(6): 673–690.

84.

Somerville

(2013) Understanding Homelessness. Housing, Theory and Society 30(4): 384–415.

85.

Soss

Fording

Schram

(2011) Disciplining the Poor: Neoliberal Paternalism and the Persistent Power of Race. Chicago: University of Chicago Press.

86.

Sparks

(2012) Governing the homeless in an age of compassion: Homelessness, citizenship, and the 10-Year Plan to end homelessness in King County, Washington. Antipode 44(4): 1510–1531.

87.

Spies-Butcher

Stebbing

(2018) Mobilising alternative futures: Generational accounting and the fiscal politics of ageing in Australia. Ageing & Society. DOI: 10.1017/S0144686X18000028.

88.

Stanhope

Dunn

(2011) The curious case of housingfirst: The limits of evidence based policy. International Journal of Law and Psychiatry 34(4): 275–282.

89.

Stuart

(2011) Race, space, and the regulation of surplus labor: Policing African Americans in Los Angeles’s Skid Row. Black Critiques of Capital: Marxism, Racism, and Exclusion 13(2): 197–212.

90.

Stuart

(2014) From ‘rabble management’ to ‘recovery management’: Policing homelessness in marginal urban space. Urban Studies 51(9): 1909–1925.

91.

Tonry

(2011) Making peace, not a desert: Penal reform should be about values, not reinvestment. Criminology & Public Policy 10(3): 637–649.

92.

Tsemberis

(2010) Housing First: Ending homelessness, promoting recovery, and reducing costs. In: Ellen

O’Flaherty

(eds) How to House the Homeless. New York: Russell Sage Foundation.

93.

Tyner

(2013) Population geography I: Surplus populations. Progress in Human Geography 37(5): 701–711.

94.

Tyner

(2015) Population geography II: Mortality, premature death, and the ordering of life. Progress in Human Geography 39(3): 360–373.

95.

Tyner

(2016) Population geography III: Precarity, dead peasants, and truncated life. Progress in Human Geography 40(2): 275–289.

96.

Wacquant

(2009) Punishing the Poor: The Neoliberal Government of Social Insecurity. Durham: Duke University Press.

97.

Whatmore

(2009) Mapping knowledge controversies: Science, democracy and the redistribution of expertise. Progress in Human Geography 33(5): 587–598.

98.

White

(2012) Must we all be paradigmatic? Social investment policies and liberal welfare states. Canadian Journal of Political Science / Revue Canadienne de Science Politique 45(3): 657–683.

99.

Willse

(2015) The Value of Homelessness: Managing Surplus Life in the United States. Minneapolis: University of Minnesota Press.

100.

Wilson

(1987) The Truly Disadvantaged: The Inner City, the Underclass, and Public Policy. Chicago: University of Chicago Press.

101.

Wolch

DeVerteuil

(2001) Landscapes of the new poverty management. In: May

Thrift

(eds) TimeSpace. London: Routledge.

102.

Wolch

Dinh

(2001) The new Poor Laws: Welfare reform and the localisation of help. Urban Geography 22: 482–489.