Engaging with homelessness in the City of Tshwane: Ethical and practical considerations

ABSTRACT

Policies and practices aimed at developing more engaged universities that are responsive to the needs of society have become key features of the higher education landscape of most countries. Visions of universities ‘engaged’ in matters of local importance increasingly require academics to reframe their scholarship as some form of ‘engagement’. This requirement has been addressed in many different disciplines and has been met with ambivalence. Academics who see engagement as a new form of ‘public good’ find it enhancing of their teaching and research activities, while others view engaged work as unnecessary and problematic ‘third mission’ activities that impede on ‘normal’ academic work. This article aims to contribute to these debates by interrogating the paradoxes of action and inaction. Drawing on recent experiences in reviewing a policy on homelessness for a municipality in South Africa, the article seeks to bring the ambiguities and challenges of engagement into greater visibility.

KEYWORDS:

• Community engagement
• homelessness
• categorisation
• policy
• power

1. Introduction

Policies and practices aimed at developing more engaged universities that are responsive to the needs of industry, government and ‘external social partners’ have in the past two decades become key features of the higher education landscape of most countries. Visions of universities ‘engaged’ in matters of local importance and acting as knowledge resource centres increasingly require academics to reframe their scholarship as some form of ‘engagement’. This requirement has been addressed in many different disciplines and has been met with ambivalence. Academics who see community engagement as a new form of ‘public good’ find it enhancing of their teaching and research activities as well as their disciplinary commitments and scholarly reputations. In contrast, others view engagement as unnecessary, and often highly problematic, ‘third mission’ activities that impede on ‘normal’ academic work (see Shore & McLauchlan, 2012). This article aims to contribute to these debates by interrogating the paradoxes of action and inaction. Drawing on recent experiences in reviewing a policy on homelessness for the City of Tshwane, a municipality in South Africa, the article seeks to bring the ambiguities and challenges of engagement into greater visibility.

Launched in December 2014, a partnership between University of Pretoria, University of South Africa, Tshwane Homelessness Forum (which represents several non-governmental organisations (NGOs) and homeless groups) and the City of Tshwane aims to develop ‘greater visibility, awareness, and understanding’ to the issue of homelessness in the City of Tshwane, to make ‘policy recommendations’ and to identify potential ‘strategies for preventing and/or reducing homelessness’ (University of Pretoria et al., 2015). As an academic anthropologist, I found this partnership an interesting development and decided to join a small group within the partnership tasked to review the Tshwane Homelessness Policy adopted in 2013. My position in the group may be described as one of a ‘participant observer’ who over a period of six months in 2015 regularly engaged with colleagues and a variety of other stakeholders on the issue of homelessness in the City of Tshwane. My interpretations in this article rest on what I have experienced in this process and should not be seen as a scientific judgment of some sort.

Identifying the public spheres within which homelessness resides was one of the first tasks in the policy review process. These spheres involve complex networks of actors, institutions and subjects in myriad connections between different sites and across different levels of society: local government officials, NGO practitioners, churches and other charitable institutions, academics and ‘the homeless’. Locating these different stakeholders in time and space is an important exercise. These are not contained domains but sites of interaction and contestation that present opportunities for unravelling ideologies, interests and knowledge which shape decision-making processes, whether in terms of policy or action. First steps in this direction were taken with a number of exploratory research projects conducted as part of the policy review process. The first part of this article describes some of the research findings by highlighting the conflicting understandings of homelessness as expressed in discussions about definitions and categorisations, causes and effects, and potential interventions. In the description of these findings, the term ‘informants’ is used to protect the anonymity of those who shared their ideas but also to avoid stereotypical notions that people in particular job categories necessarily think and act in certain ways. For example, during the policy review process I became aware of a general perception that law enforcement officers necessarily treat homeless individuals as disreputable vagrants while social workers see them as welfare ‘targets’. It is crucial not to view the different actors as bounded, autonomous and unified sources of intention and power. Guarding against reification equally applies in making reference to my academic colleagues and other participants in the Tshwane Homelessness Programme. They would describe the programme and the research findings differently than I do.

The second part of the article outlines potential avenues for future research in order to consider alternative ways of thinking and doing in responding to homelessness. It emphasises the merits of a research strategy that incorporates ‘studying down’, ‘studying up’ and ‘studying through’ (more on each of these research approaches later). The article concludes by referring to aspects of the debate on the desirability and feasibility of academics adopting a more engaged role in society.

2. Conflicting understandings of homelessness

2.1. Defining homelessness and categorising homeless people

Throughout the policy review process, the need to define homelessness was a recurring theme in conversations with different informants but it was difficult to agree on what constitutes homelessness. The diversity, mobility and transience of the homeless population complicated attempts at defining or estimating the prevalence of homelessness. There was general agreement that definitions of homelessness are dictated by the scope and field procedures of data collection or intervention strategies. The most obvious definition, and one that dominated perceptions, was the notion that homelessness refers to a state of ‘being without shelter of any kind’. Informants often referred to the increasing numbers of people who are living on the streets, under bridges and highway overpasses, and alongside streams and other vacant public land.

Some informants viewed homelessness primarily in monetary terms as the absence of an income and a house. But many others recognised that homelessness meant more than ‘house-less-ness’ and included concepts such as marginalisation and social exclusion. They observed that there are many people who have a place to sleep but do not have a ‘home’ in the accepted sense of the word; that is, a sense of safety, privacy, stability and the ability to access and control a space for social relations and ‘decent living’. However, in the context of the enormous housing needs in South Africa, most informants suggested that people who are in ‘inadequate accommodation’ (i.e. overcrowded spaces or dwellings with improper structures and a lack of basic facilities which render them unfit for human habitation) and those who are ‘at risk’ of homelessness (e.g. those currently ‘squatting’ with relatives or friends) should not be considered ‘homeless’. People living in emergency and temporary shelters were regarded as in between the two extremities of highly visible and relatively concealed or ‘hidden’ homelessness, and were thus considered part of the city’s homeless population.

Definitions of homelessness were complemented by the ascription of attributes which are seen as characteristic of (and often explanatory of the behaviour of) homeless people. It became clear that the category of ‘the homeless’ has a range of associations including mental illness, substance abuse, deviance, social dysfunction, laziness and criminality. These associations often result in the problematisation of homeless people rather than homelessness itself. Some informants easily found willingness to accept relief in a shelter for a period longer than a year as evidence of laziness or a lack of moral fibre. A government official, for example, asked me: ‘what kind of a person wants to stay forever in Number 2 Struben Street?’. Her reference was to an infamous shelter run by the municipality on the outskirts of Pretoria’s city centre which most informants described as an overcrowded, dirty place characterised by violence, theft and drug abuse. Many informants lamented that the conduct of homeless persons is inconsistent with expectations of ‘responsible citizenship’. Homelessness is represented as inherently problematic for ‘the homeless’ but also for the broader society of citizens who deem themselves law abiding individuals. As a result, shielding the community against the crimes of ‘the homeless’ becomes a key priority in most action plans.

The negative images and associations of homeless people resulted in some informants being less concerned with the definition of homelessness and more focused on defining who is ‘legitimately’ homeless. They often argued that a person who has a home ‘somewhere’ cannot be considered homeless. They insisted that ‘we all come from a place that we call home’ and that ‘home is where a person wants to be buried someday’. Such perceptions feed the belief that homelessness is indeed a choice or a way of life. Unsurprisingly, notions of ‘home’ were connected to citizenship. Non-South Africans were portrayed as ‘outsiders’ and became scapegoats for the country’s high crime rates, especially those related to drug abuse and house robberies. For these informants, the primary purpose of the development of definitions was to differentiate between ‘the deserving’ and ‘the undeserving’. Such efforts at categorisation become fundamental to how the homeless and homelessness are perceived and they ultimately shape government and other responses to assist or govern this population. These notions are based on particular understandings of the causes of homelessness.

2.2. Understandings of the ‘pathways into homelessness’

During fieldwork, discussions of the reasons for people becoming, and remaining, homeless concurred with descriptions in the literature on homelessness. The most common explanations for homelessness fell into two broad categories, namely individualistic and structural explanations. Individualistic explanations focused on the personal characteristics and behaviours of homeless people and suggested that homelessness is a consequence of some form of individual fault or deviance. Typical problems associated with homeless persons were drug and alcohol abuse, problem gambling, mental illness, limited social skills, low levels of education, and a history in state care or correctional facilities. The question on the extent to which these same characteristics are found in the population who do have homes has not entered into the discourse on homelessness. Instead, people who are homeless are portrayed as ‘others’ who are fundamentally different from the rest of society because they fail to behave in accordance with expectations. These constructions of difference allowed a number of informants to observe that the phenomenon of homelessness was not part of ‘our culture’. Worth noting here is the explanatory power given to the concept of ‘culture’.

Structural explanations located the causes of homelessness in the political economy or the broader economic and social environment. Informants saw escalating housing costs and a lack of access to social grants as contributing factors to the prevalence of homelessness. A key concern was an increase in unemployment as a result of negative labour market forces and decreased opportunities for workforce participation which many informants related to de-industrialisation and the in-migration of ‘foreigners’. Although informants acknowledged that these conditions have exacerbated the vulnerability of the working class to economic marginalisation, poverty itself seldom featured in their discourses on homelessness.

The majority of informants oscillated between individualistic and structural causes of homelessness. Only a few understood homelessness as the outcome of a dynamic interaction between individual attributes and actions and structural factors. They argued that structural factors create the conditions within which homelessness will occur and individual deficits increase people’s vulnerability or exacerbate the impacts of homelessness.

Informants’ perceptions of the causes of homelessness naturally impacted on their ideas about the policy responses that are needed to address homelessness.

2.3. Perceptions of the ‘pathways out of homelessness’

Informants’ ideas concerning appropriate responses to homelessness were functions of how they understood the causes of homelessness. Those who explained homelessness in terms of individual deficit and deviance argued that policies have to be designed to regulate, monitor, rehabilitate, guide or ‘manage’ homeless people. For them the solutions obviously lie in interventions directed at normalising individuals. Family reunification, for example, was regularly mentioned as the primary strategy for solving youth homelessness. Informants who referred to this tactic argued that the first priority should be to reunite young homeless persons (often called ‘street children’) with their families and through family arbitration address the problems which have prompted the young persons to leave their homes. These informants did not consider economic conditions or family circumstances. Instead, homeless youth were seen as deviant in having left their homes and school while their families were often deemed complicit by not accepting responsibility for them. Values of family and community were also emphasised by officials who believed that the solution to adult homelessness (and other social problems such as xenophobia) was also to ‘help people to find the road back home’. In addition, some informants wanted to educate homeless people who were not self-reliant to the values of hard work and enterprise. These views, however, were supplemented with a conventional belief in the inevitability of inequality between the rich and the poor. Several informants suggested responses that were constructed in terms of charity. Framing homelessness as a welfare issue oriented solutions away from any structural considerations.

In contrast, responses aimed at addressing structural causes were employed by those who understood homelessness as a consequence of factors such as unemployment and lack of access to housing and education. These informants were critical of the lack of government commitment to the implementation of the rights to education, employment and adequate housing. They pointed to the absence of a national policy on homelessness as an indicator of the state’s lack of intent. Some informants acknowledged social grants as part of the government response to address inequalities, but noted that compliance with administrative requirements becomes a significant barrier for homeless persons hoping to access social grants. Interestingly, these informants were in the minority because many others questioned the merits of social grants to the poor by arguing that grants act as a disincentive for people to become self-reliant.

Informants also differed on the provision of shelter and low-cost or affordable housing as a viable response to homelessness. To begin with, the notion of ‘low-cost’ or ‘affordable’ housing was open to different interpretations. Informants were quick to point to the ambiguity of meaning of the concept ‘affordability’ but, the term was used whenever housing as a solution to homelessness was addressed. Those in favour of housing support pointed out that not all homeless persons were unemployed but that lowly paid employment compromised the ability of workers to commute between their homes and their places of work, resulting in many rough-sleepers. Some also referred to the many homeless men seen on Pretoria’s streets pushing trolleys filled with recyclable materials. Although they observed that motorists often regard these garbage collectors as a nuisance or a danger on the City’s roads, some informants expressed appreciation for the entrepreneurial spirit of these homeless individuals and were keen to see the municipality supporting them in one way or another.

Opponents of housing support as an intervention strategy argued instead that the demand for government or low-cost housing far outweighs supply and that housing provision often encourages illegal practices such as the selling or renting of government-subsidised houses. They argued that the government is under no obligation to ensure social security and should only be concerned with the well-being of the economy. If the economy has negative social impacts, individuals have to demonstrate that they are deserving of assistance.

To a large degree, these discourses on the characteristics of homeless people, the causes of homelessness and the technologies that are proposed as solutions to homelessness point to the problematisation of the homeless instead of homelessness. There is clearly a need for alternative perspectives and visions that are grounded in research. Importantly, such research should be reflexive activities that do not reproduce hegemonic social relations, but rather cast light on them.

3. Alternative perspectives and visions: potential avenues for research on homelessness

3.1. ‘Studying down’: homeless people as objects of study and intervention

With homelessness as a subject, ‘the homeless’ easily become objects of study. In fact, there is a tendency to only ‘study down’ – to be preoccupied with the poor, the marginalised and the homeless. This tendency has a long history. It is understandable that most governments have in the past attempted (and are continuing in the present) to establish the size of their country’s homeless population. Processes of enumeration resulted in considerable debate over the definition of homelessness but also generated statistics on homeless people. As we have learnt from scholars such as Foucault (1989) and Rose (1991), numbers may have the appearance of neutrality but they do more than simply describe a pre-existing reality – they also contribute to the creation of that reality. In several countries, data banks emerged that contained typologies or profiles of homeless individuals in terms of their demographic features, deficiencies and needs. These data banks served to describe and classify homeless people but, significantly, also to pathologise them. This in turn provided the basis for action to ‘rehabilitate’ homeless individuals and thus to construct them as objects of intervention.

Contemporary research may well contribute to the perpetuation of these social constructs by trying to establish for a causal link between ‘the homeless’ and a particular characteristic such as a mental disorder or substance abuse. Data are often complemented with statistics that capture the incidence of particular individual traits or group practices such as drug abuse, gambling and involvement in crime. Unfortunately, in most instances such quantification does not register the degrees of desperation which force people to view these practices as their only options. Instead, defining, classifying and counting become critical tools for letting categories such as ‘the homeless’ assume significance in order to legitimate spending and tactics of rule.

An additional danger of focusing on particular pathologies associated with homelessness is that this may serve to homogenise features of homelessness and the situation of those classified as homeless. As a result, the homeless may continue to be subject to research and intervention (including regulation) to a far greater extent than that which applies to other citizens. To exacerbate the problem, homeless people often lack the entitlements and resources to reframe the terms of such engagements. Homelessness implies deprivation not only of housing but also of power. Various institutions have power over ‘the homeless’ – not only governments and do-gooders who define them and determine their needs, but also researchers who study them. This does not imply that homeless persons are mere objects or passive victims. On the contrary, they react creatively to their circumstances and negotiate multiple fields of power within which others are positioned as experts and they themselves as targets of such expertise.

In view of this description of past and present responses to homelessness, it is imperative that future research demonstrates the diversity of the experience of homelessness and the agency of homeless people. It is equally important to recognise that the emphasis on ‘the homeless’ as the problem and as the objects of study and intervention diverts attention from the social relations which produce homelessness as an attribute of people. Homelessness needs to be interpreted as a consequence of social relations, and these relations require investigation and transformation. For this purpose, we need to extend our gaze by not only ‘studying down’ but also ‘studying up’ – by focusing on the powerful, the elite and the wealthy.

3.2. ‘Studying up’: powers of the state, the ‘do-gooders’ and the university

Exploring the processes through which ‘the homeless’ as a category comes to have salience is fundamental to an understanding of homelessness. In the City of Tshwane, numerous powerful institutions play instrumental roles in making the population of homeless people visible, imaged, (mis)understood and able to be governed. A ‘studying up’ approach should highlight the importance of the (in)actions and strategies of the powerful as well as the powerless in determining homelessness as an outcome and an embodied experience. However, limited space allows for consideration of only three categories of powerful actors: the state, the ‘do-gooders’ and the university. In the process of writing about these actors it becomes evident that often ‘development is not policy to be implemented, but domination to be resisted’ (Mosse, 2004:643).

3.2.1. The state

In the past two decades, ‘the state’ has been studied ethnographically in and through its margins across the globe, revealing the fragile self-representation of the state or bureaucracy as well as the sites and practices where the state’s presence is experienced as problematic (cf. Scott, 1998; Ferguson & Gupta, 2002; Das & Poole, 2004). Documenting homeless people’s experiences of the state will contribute to these perspectives. The state occupies a central role in the constitution of homelessness and impacts the lives of ‘the homeless’ in ways that go beyond the arenas of housing or social welfare. In this regard, the relationship between law and other technologies of government and the experiences of homeless people is an important field of study. People who are homeless are often denied legal entitlements associated with the status of citizenship and appear to be more vulnerable to law enforcement (cf. Lynch, 2002; Walsh, 2005). Research needs to illuminate the ways in which the legal apparatus of South Africa’s criminal justice system impacts on people who experience homelessness.

‘Studying up’ will also assist us in disaggregating the concept of the state in order to address false assertions about the state as an ‘up there’ unified source of power and intention (see Mitchell 1991:78). The difficulty of working with government partly resides in the fact that ‘the state’ comprises many different parts that do not act as a whole or in a coherent and coordinated manner. Because the state is influenced by and in turn influences several institutions and individuals, who hold power over the homeless also vary over time and in place. Studying powers beyond that of the state is essential because the state shares the function of improving the welfare of the homeless population with a number of other players who may be depicted as ‘do-gooders’.

3.2.2. The ‘do-gooders’

As indicated earlier in the article, state and non-state responses to homelessness normally revolve around a vision of society free from deviant or pathological ‘elements’. Such a vision not only provides welfare or development programmes with a laudable objective but also serves to justify intervention. Working with (or at times, against) the state, several ‘non-governmental’ organisations in the City of Tshwane focus on ‘the plight of the homeless’. Some have already become case studies in research projects (see Sanchez, 2010) but many more need to be studied as part of ‘the powerful’ that affects the lives of homeless people.

Ethnographies need to piece together the social relations which form the ‘organisation’ and explore how everyday activities are coordinated and connected to institutional texts such as policy documents. Given the diverse backgrounds, traits and needs of homeless people, the ways in which these organisations respond to homelessness may or may not be appropriate. A number of case studies demonstrate how shelter staff exercise power over residents by implementing ‘disciplinary technologies’ (Foucault, 1989) to monitor and judge residents against societal norms for sane, moral or productive behaviour (cf. Williams, 1996; Desjarlais, 2000). Wallerstein (2014:19) argues that such responses compel shelter residents ‘to feel at once independent and subject’. Equally problematic is the tendency to medicalise the personal characteristics of residents (see Löfstrand, 2012) in order to ‘justify decisions in discriminating between service-worthy and “treatment-resistant” clients’ (Marvasti, 2002:623). It is thus important to examine the degree of fit or mismatch between homeless individuals and the help or ‘treatment’ they receive.

Such a research focus does not imply that these organisations are not doing important or commendable work but that their work is influenced by power and capital. Examining the hopes, intentions and actions of individuals who work within these organisations is an important research objective. It would be naive to imagine that the ‘do-gooders’ necessarily share ‘the will to improve’, as Tania Murray Li (2007:5) labels the mentality of those whose desire it is ‘to make the world better than it is’. Within the industry of ‘development’, organisations and individuals may use homelessness, like poverty, for private or personal gain. Maia Green (2006:1121) observes that, in politicised contexts, ‘poverty can be claimed not so much as a problem for some social categories, but as a potential asset by others who stand to gain from the inputs associated with the development relationship’. She refers to employment, travel and capacity-building opportunities as examples of such inputs. In similar fashion, homelessness is a label applied to a particular social category for whom it is a problem while being an asset for those who benefit from homelessness as a category of development thinking. Homelessness thus has the potential to foster, to paraphrase Green (2006:1114) slightly, ‘a new sub profession of anti-homeless specialists’.

The implication of all of this is that, apart from the state, many other parties seeking to be seen as advocates for the homeless may be interested in particular constructions of ‘the homeless’ and of interventions to address homelessness. Li (2007:7) reminds us that a crucial skill in development work is the ability to identify ‘deficiencies that need to be rectified’ and that institutions’ or individuals’ claim to expertise ‘depends on their capacity to diagnose problems in ways that match the kinds of solutions that fall within their repertoire’. The possibility therefore exists that institutions and individuals who seek funding security and growth may be preoccupied with enumerating and classifying the City of Tshwane’s homeless population to demonstrate the extent and nature of demand for their services. For the same reason they may be complicit in perpetuating the discourse that homelessness is a result of individual deficit and that ‘rehabilitation’ and management of homeless individuals are its only remedies.

The ‘do-gooders’ in the ‘homelessness industry’ are not the only powerful actors who shape government thinking and actions. Academics occupy a particular position vis-à-vis the state and non-state actors.

3.2.3. The university

The state often relies on expert knowledge to justify its position, actions and inactions. Informed by the thinking of Michel Foucault, a number of scholars argue that ‘objective knowledge’ supplied by ‘research’ are used by governments to justify their policies (cf. Boden & Epstein, 2006; Miller & Rose, 2008). As we have seen earlier in the article, knowledge offers an indispensable means for legitimising the way a population or a problem is framed and upon which action is based. Miller & Rose (2008:43) also point out that claims to scientific research enable authorities to shape conduct ‘not through compulsion but through the power of truth, the potency of rationality and the alluring promises of effectivity’. This is especially the case when those that are ascribed the title of researchers or ‘experts’ are academics drawn from reputable universities. Academics are recognised as authorities who speak ‘the truth’ and are believed to possess specialised knowledge which renders such knowledge with a specific power (see Rose & Miller, 1992:177).

However, the relationship between government and experts is not neutral and without bias. The production of knowledge may be consistent with dominant rationalities and visions of problems and solutions which attain legitimacy at particular times. This may frame what is researched and how it is researched, and may impact the use of ‘evidence’ in policy-making. The scope for equal partnerships between the state and its ‘experts’ is further limited by considerations of development aid and donor funding (cf. Crewe & Harrison, 1998; Lewis & Mosse, 2006). The rise of what some have termed ‘audit culture’ (Shore & Wright, 1999; Strathern, 2000) also means that increasing demands of ‘accountability’ and measurement of individual performance impact academics’ decisions regarding their involvement in research and other forms of engaged work.

Research must give insight into the varied ways in which ‘expert knowledge’ impacts the generation and use of ‘evidence’ on homelessness. This is especially relevant in cases where the boundaries between the research community and the policy-makers are blurred (as with the Tshwane Homelessness Programme). Questions of ‘ownership’ then become relevant: for example, who can lay claim to, speak on behalf of or represent themselves as ‘the university’ (Shore & Taitz, 2012)? It is crucial that we recognise that the alliance between the state and universities may give government officials the power to cloak their discourses on homeless people in academic clothing.

Apart from ‘studying down’ and ‘studying up’, we are also encouraged to engage in ‘studying through’.

3.3. ‘Studying through’: policies as ‘actants’

The literature on policy-making demonstrates that policies are not unproblematic and politically neutral technologies for problem-solving (see Wedel et al., 2005). Instead, they are ‘actants’ or objects in a network which perform tasks or shift action (Akrich & Latour, 1992:259; also see Shore et al., 2011).

A useful concept for studying policy and processes of policy-making is what Wright & Reinhold (2011) call ‘studying through’. This research approach entails that we ‘study through’ by constructing our analysis through a policy (or a conflict or a debate) that moves through different sites. It proposes that we ‘follow the policy’ through its development, implementation, contestation and movement across ‘policy worlds’. A ‘studying through’ approach recognises the actors and agency at work in different settings. This is a valuable aspect if we accept David Mosse’s (2004) argument that policy documents work as sites for coalition-building. They facilitate mobilisation and maintenance of political and organisational support across public and private sectors despite different agendas and points of view. Thus ‘development policy ideas are important less for what they say than for who they bring together’ (Mosse, 2004:649; original emphasis).

While policies are inherently forward-looking, ‘studying through’ also provides a methodology for constructing a ‘history of the present’ by tracing the events that resulted in the current state of affairs. Given the ahistoricism of explanations for homelessness, reflection on the continuity between current discourses on homelessness and the historical assumptions that inform them is a crucial contribution. We need to understand how the fact of being ‘homeless’ emerged, and changed, and how ‘the homeless’ became a particular subgroup of the South African population. Existing perspectives on recent national and local government responses to homelessness in South Africa (cf. Du Toit, 2010; Naidoo, 2010) constitute a good starting point for unravelling the relationship between policies and their effects.

Policies have effects that are often difficult to assess and which take time to manifest. The literature tells us that policies have institutional effects such as enhancing state capacity, expanding bureaucratic control and maintaining relations of power. In addition, policies have ideological effects because their workings involve ‘taking what is essentially a political problem, removing it from the realm of political discourse and recasting it in the neutral language of science’ (Shore & Wright, 1997:8). This appears to be a similar process to Tania Murray Li’s (2007) ‘rendering technical’ or James Ferguson’s (1994) ‘anti-politics machine’. Thus there is a potential danger that policies on homelessness may naturalise homelessness, objectify homeless persons and depoliticise interventions.

A further caveat in navigating the field of homelessness which a ‘studying through’ research approach may illuminate is the limited operational control which the state, ‘do-gooders’ or academics have over interventions and their impacts (see Quarles van Ufford, 1988). The literature abounds with arguments about the disjuncture between policy designs and their implementation or outcomes. James Scott argues that the discrepancy between policy and practice is partly due to planners’ preference for ‘administrative ease’ (Scott, 1998:242) or their ‘hubris’ as manifested in their belief that ‘they alone knew how to organize a more satisfactory, rational, and productive life for their citizens’ (Scott, 1998:247). Other scholars attribute unplanned outcomes to the fact that often ‘the bureaucratic-legal processes are not legible even to those responsible for implementing them’ (Das, 2004:245) or that ‘street level bureaucrats’ have to use their own discretion to interpret policy on a case-by-case basis but are often forced to ‘rubber-stamp’ applications and routinise client interactions due to huge workloads and limited resources (Lipsky, 2010). David Mosse (2004) also argues that the requirements of a policy in its planning stages are different from those in the stages of practical implementation. He exposes the ways in which policies actually do not serve to orientate practice but to legitimise interventions. The expectation that policy design and policy outcomes should correspond seems to be unrealistic.

The revised Tshwane Homelessness Policy has not yet entered its post-paper life. Consequently, its effects will only be known in years to come. But the extent to which the policy will provide opportunities to understand and effectively respond to homelessness and poverty should remain open to question. Adopting a ‘studying through’ approach implies that future research should reveal the social processes through which homelessness as policy objective becomes institutionalised in state and non-state practice. Because the policy will hold different and contested meanings for its advocates and opponents, it is important that research captures the moments in which different categories of people (including ‘the homeless’) engage consciously with the policy by explaining, critiquing, escaping, implementing or adapting it.

4. Conclusion: To engage or not to engage? A false dichotomy?

The preceding parts of this article point to the fact that engagement involves complex histories and social relationships and that the promises and practices of engaged work are deeply political. Engagement involves an assemblage of categories located in unequal power positions and is clearly not without problems. Implementing significant, measurable change is a slow, incremental and complicated task. It is often experimental and despite good intentions some experiments will unavoidably fail or have negative effects. Thus what are at stake in engagement are not merely practical issues but also ethical ones. Little wonder, then, that academics differ on what engagement entails and what forms it should take.

Some see engagement as a collaborative approach to research which involves research subjects as partners to ensure multi-vocal and co-constructed understandings and narratives. They challenge the notion of ‘neutral’ research. Others argue that collaborative efforts to determine research objectives and methods tend to cater for the needs of the study population and may thus jeopardise objectivity and result in engagement losing its critical edge. Some also view the engagement task as one of education, where ‘teaching’ is not restricted to classrooms or to training through workshops but also takes place through public debate and popular writing.

Several academics deem social critique – the scholarly pursuit of revealing differences in power and structures of inequality – as the most appropriate form of engagement. Unlike other forms of engaged work, social critique avoids the risk of losing control of the way research is used. However, critics of this approach of ‘only critiquing and not suggesting alternatives’ argue that partnerships with the state or ‘do-gooders’ offer the promise of rendering the work of these institutions more humane and responsive to societal needs. For them, the task at hand is not only to critique to ensure writing that has explanatory power but also to connect that power to praxis.

Those in favour of more radical forms of engagement centred on advocacy and activism question the moral acceptability of inaction. They argue that reluctance or failure to promote the human rights of vulnerable populations such as ‘the homeless’ is not a neutral or apolitical stance but an ethically problematic position which hints at complicity in maintaining the status quo of injustice, violence and powerlessness. Not only is there a danger of engagement, but there is also a danger of non-engagement. In contrast, those who are sceptical of the need for intervention question the ethical right to intervene in the lives of those being studied.

Within the context of the Tshwane Homelessness Programme, the ambivalence that the notion of ‘engagement’ generates has left many academics pondering the question of whether they should become engaged in efforts to prevent or address homelessness or rather restrict themselves to observing and recording the lives of homeless people. This question is indicative of the tendency to view engagement as two sides of a scale: either you are involved or you are not. I find this binary conception of engagement overly simplified. Not only are the different forms of engagement often overlapping but engagement can at times be enacted as principled non-engagement (i.e. the evasion of engagement in the first place). In other words, the decision not to be engaged may well be a more activist stance than to engage. Engagement poses difficult questions not only to academics but also to those at its receiving end who also have to decide on their engagement, disengagement and non-engagement.

The decision by some academics to adopt an engaged position through the opportunities created by the Tshwane Homelessness Programme may be seen by others as a matter of naive optimism or a lack of critical reflection (see Kistner, 2015). But the programme nudges toward a politics of hope as the varied forms of engagement begin to highlight the problem of social critique of the homeless rather than the society that produces homelessness.

Disclosure statement

No potential conflict of interest was reported by the author.