Towards a “fourth generation” of approaches to HIV/AIDS management: creating contexts for effective community mobilisation

Abstract

Many biomedical and behavioural HIV/AIDS programmes aimed at prevention, care and treatment have disappointing outcomes because of a lack of effective community mobilisation. But community mobilisation is notoriously difficult to bring about. We present a conceptual framework that maps out those dimensions of social context that are likely to support or undermine community mobilisation efforts, proposing that attention should be given to three dimensions of social context: the material, symbolic and relational. This paper has four parts. We begin by outlining why community mobilisation is regarded as a core dimension of effective HIV/AIDS management: it increases the “reach” and sustainability of programmes; it is a vital component of the wider “task shifting” agenda given the scarcity of health professionals in many HIV/AIDS-vulnerable contexts. Most importantly it facilitates those social psychological processes that we argue are vital preconditions for effective prevention, care and treatment. Secondly we map out three generations of approaches to behaviour change within the HIV/AIDS field: HIV-awareness, peer education and community mobilisation. We critically evaluate each approach's underlying assumptions about the drivers of behaviour change, to frame our understandings of the pathways between mobilisation and health, drawing on the concepts of social capital, dialogue and empowerment. Thirdly we refer to two well-documented case studies of community mobilisation in India and South Africa to illustrate our claim that community mobilisation is unlikely to succeed in the absence of supportive material, symbolic and relational contexts. Fourthly we provide a brief overview of how the papers in this special issue help us flesh out our conceptualisation of the “health enabling social environment”. We conclude by arguing for the urgent need for a ‘fourth generation’ of approaches in the theory and practice of HIV/AIDS management, one which pays far greater attention to the wider contextual influences on programme success.

Keywords:

• community mobilisation
• HIV/AIDS
• context
• behaviour change
• health-enabling environment
• political will

Introduction

Many HIV/AIDS programmes have disappointing outcomes because of lack of effective community mobilisation. There is general agreement that community involvement enhances the effectiveness of a range of biomedical and behavioural HIV/AIDS interventions. It does this through facilitating the empowerment of AIDS-affected communities to take control over their health, increasing the likelihood that community members will engage in health-enhancing behaviour change in relation to prevention, care and treatment. While this assumption is theoretically well-grounded, it has been difficult to apply in practice. While some community mobilisation efforts successfully establish an organised, unified and action-oriented group, becoming a force for health-enhancing change, others fail miserably, further undermining community confidence and capability. To understand the reasons for success and failure of community mobilisation initiatives, this special issue examines how the wider social contexts of interventions can support or undermine community mobilisation efforts.

In this paper, we present a conceptual framework for examining those dimensions of social context which are likely to support or to undermine the effectiveness of community mobilisation initiatives. We propose attention should be given to three dimensions of context: material, symbolic and relational, illustrating this argument with reference to existing research in India and South Africa, and examples from the special issue's papers.

This paper has four parts. Firstly we outline the reasons for our interest in community mobilisation, presenting the core concept underlying the work in this special issue, namely the notion of “health-enabling social environments”, which we define as the community contexts most likely to enable and support the possibility of health-enhancing behaviour change. Secondly we map out three “generations” of approaches to behaviour change within HIV/AIDS management: HIV-awareness, peer education and community mobilisation. We critically evaluate each one's underlying assumptions about the social psychology of behaviour change, to frame our understandings of the psycho-social pathways between mobilisation and health, drawing on the concepts of social capital, dialogue and empowerment.

Thirdly we draw on two well-documented case studies to develop and illustrate our claim that community mobilisation is unlikely to succeed in the absence of supportive social contexts – using our three dimensional characterisation of the material, social and relational contexts of community mobilisation. We argue that much work remains to be done in developing actionable conceptualisations of context to inform both research and practice in the community mobilisation field. Finally we highlight how the special issue's papers help us flesh out our conceptualisation of the material, symbolic and relational dimensions of context. We argue for the need for a fourth generation of approaches to behaviour change in HIV/AIDS management research and practice – one which pays far greater attention to the wider contextual influences than is currently the case in the theory and practice of community mobilisation.

Background

Billions of dollars have been poured into HIV/AIDS interventions – health education, condom distribution, voluntary counselling and testing, antiretroviral drug treatments and so on. In many contexts, however, HIV rates continue to rise, stigma remains stubbornly resistant to change, and access and adherence to treatment and support remain inconsistent (Gregson et al., 2007). A small number of interventions have worked well under carefully controlled research conditions, but these have been difficult to scale-up in less highly monitored “real world” settings (Chopra & Ford, 2005; Greig, Peacock, Jewkes, & Msimang, 2008; Poku & Sandkjaer, 2007).

One reason for the disappointing outcomes of many programmes is that they are conceived by external experts and imposed on communities in “top-down” ways. As a result, they fail to resonate with the worldviews and perceived needs and interests of their target groupings, or to take adequate account of the complex social relations into which programmes are inserted (Gruber & Caffrey, 2005; Pfeiffer, 2003). Too often target communities are seen as passive recipients of prevention, care and treatment services (the “objects” of interventions) rather than active participants working in partnership with health professionals to improve their health (the “subjects” of interventions) (Campbell, 2003). Furthermore little attention is given to the “goodness of fit” between programme goals and contextual constraints and possibilities shaping the behaviour of target group members.

Community involvement is increasingly being cited as a vital precondition for creating “health-enabling” social environments – social contexts that enable and support people in optimising their opportunities for health and well-being (Tawil, Verster, & O'Reilly, 1995, p. 1299). It is said to play a vital role in enabling health-related behaviours and reducing HIV-transmission (van Wyk, Strebel, Peltzer, & Skinner, 2006), stigma reduction (Poku & Sandkjaer, 2007), facilitating timely and appropriate accessing of health and welfare services where these exist (Hadley & Maher, 2001; Segall, 2003), and supporting optimal treatment adherence (Coetzee et al., 2004).

There are three reasons for our interest in community mobilisation. Firstly, community mobilisation is given a key role in international AIDS policy, due to its perceived role in tackling the challenges of “reach” and “sustainability” that plague many HIV/AIDS interventions. Policies envisage that community-based organisations play a central role in taking responsibility for HIV/AIDS. The scale of the problem, as well as the social marginalisation of many affected groups, make it impossible for health and welfare professionals and services to reach the vast number and variety of people vulnerable to HIV and affected by AIDS. Policy documents consistently argue that community organisations are better placed than formal public services to reach “hard-to-reach” groups (including, for example, women or young people in high-risk settings, macho men, sex workers, injecting drug users, men who have sex with men, residents of remote or poorly served rural areas and so on). Furthermore community organisations, drawing on local human resources, and often including the inputs of unpaid volunteers, are seen as more sustainable in the long-term than costly and donor-dependent external interventions.

The second reason for our interest in community mobilisation is a pragmatic one in the light of the overall scarcity of health workers in many marginalised settings in which HIV/ AIDS flourishes. Here, community volunteers are earmarked for an increased role in HIV/AIDS management within the wider context of the international “task shifting” agenda (World Health Organisation, 2008). A growing number of projects rely on grassroots community members to provide HIV-prevention and AIDS-care (Ogden, Esim, & Grown, 2006; Schneider, Hlophe, & van Rensberg, 2008; van Damme, Kober, & Laga, 2006). The final reason is a social psychological one. The mobilisation of HIV/AIDS vulnerable communities to participate in effective community organisations provides valuable opportunities through which people can develop a stronger sense of agency. This increases the likelihood that they will behave in health-enhancing ways (Bandura, 1995; Wallerstein, 1992), and enables them to collectively tackle problems that are undermining their health. Effective and empowering community mobilisation cannot take place in the absence of a supportive social environment however. This special issue seeks to map out those dimensions of a social environment that impact on the potential success of community mobilisation – through providing case studies of the contexts of community mobilisation in the Gambia, East Africa, India, Kenya, Papua New Guinea, Cambodia and South Africa.

Three generations of behaviour change approaches – towards a fourth generation?

In this section we identify three dominant approaches to HIV/AIDS management, showing such approaches have evolved from an early emphasis on the individual as locus of change (HIV awareness approaches) to the peer group as locus of change (peer education approaches) to the community as locus of change (community mobilisation approaches). We will argue for the need for a fourth approach, which pays attention to the wider social changes that would enable optimally effective community mobilisation efforts.

Whilst we refer to these as three “generations” of approaches in a quasi-historical way, elements of all three approaches have characterised responses to HIV/AIDS in a more complex and contradictory way than our simple periodisation would acknowledge. However we believe this crude evolutionary account has analytical value. It enables us to map out how dominant understandings of HIV/AIDS-related behaviour change have paid gradually increasing attention to the social determinants of individual behaviour, with ever-widening attention to the role of the “social” in shaping the possibility of programme success. We hope this special edition will stimulate debate on the need to further widen and intensify this “socialisation” of approaches to behaviour change. There is an urgent need for a “fourth generation” of approaches to HIV/AIDS management which moves beyond the narrow local community emphasis of most existing community mobilisation approaches to take account of the wider social contexts within which community mobilisation is implemented, and how these support or undermine the likelihood of health-enhancing behaviour change.

First generation approaches: promoting HIV awareness

Targeting the individual as locus of change, and informed by traditional health psychology, first generation approaches to behaviour change took the form of traditional didactic health education seeking to provide individuals with information about HIV/AIDS, how to avoid it, and how to respond once infected. The crudest versions assumed that people engaged in health-damaging behaviours due to ignorance, and would make rational choices to change such behaviour once in possession of accurate factual information (Aggleton, Young, Moody, Kapila, & Pye, 1992). More sophisticated approaches drew on social cognition models. These gave a more complex and elaborated account of how health-related behaviours were driven by people's individual perceptions of the risks, costs and benefits of those behaviours (Rosenstock, Strecher, & Becker, 1994; Terry, Gallois, & McCamish, 1993).

However, these individual-focused approaches often had disappointing outcomes, with studies increasingly suggesting that information, or individual perceptions, were only a partial determinant of behaviour change – a necessary rather than sufficient condition (Aggleton, O'Reilly, Slutkin, & Davies, 1994). Furthermore, such programmes were criticised for being imposed on target communities by outside health professionals who often had little understandings of the realities and limitations of people's lives (Asthana & Oostvogels, 1996).

In a searing critique of individual-focused strategies, Waldo and Coates (2000) argued that health psychologists may have unwittingly played a role in driving the HIV/AIDS epidemic through suggesting AIDS could be tackled by targeting individuals, drawing attention away from the wider social drivers of HIV/AIDS related behaviours. These included factors such as the special vulnerability of women and young people to HIV-infection in some contexts, the role of peer norms rather than individual choice in shaping behaviours, and the difficulties faced by poor people and people in under-served rural areas in accessing effective care and treatment once infected.

Second generation approaches: peer education

In response to the perceived shortcomings of individual-focused approaches, a second generation of responses highlighted the strong influence of peer norms on sexual and health behaviours, arguing that it made more sense to target the peer group – rather than the individual – as unit of change. Informed by an emphasis on the impact of small-group interactions on behaviour (Kelly et al., 1992), such approaches rested on three assumptions. Firstly that similar others (“peers”) were best placed to reach socially marginalised HIV/AIDS vulnerable groups who might be inaccessible to mainstream health professionals, or unwilling to heed their advice. Secondly that people were more likely to change their behaviour if they saw that liked and trusted peers were changing theirs (Dube & Wilson, 1999). The third assumption was that the simple provision of information was unlikely to change behaviour in the absence of the “empowerment” of “at risk” group members to convert information to action, through the acquisition of skills. Such skills included sexual negotiation, assertiveness, knowing where to find condoms and how to put them on, and so on. Peers were seen as the best communicators of knowledge and skills, given their “insider” status within hard-to-reach groups. Programmes trained members of “high-risk groups” in HIV/AIDS-related information and skills, sending them back to their communities to disseminate these.

Peer approaches have had some successes, especially in more affluent countries and contexts, where target groupings already had a high degree of freedom over their behaviour (agency), were already linked through pre-existing networks characterised by solidarity and a common identity (bonding social capital) and were able to mobilise the support of powerful social groups (bridging social capital). One such group was gay men in the USA at the early stages of the epidemic (Kegeles, Hays, & Coates, 1996; Kelly et al., 1992).

However, in other contexts, peer education has been less successful (Kim & Free, 2009). In her study of a relatively well-funded, state-of-the-art programme with sex workers in a South African mining community, Campbell (2003) argued that peer education success was undermined by four factors. The first was low levels of perceived agency by sex workers, marginalised by gender, poverty and stigmatised occupational status. With few previous opportunities to exercise control over their well-being, they had little faith in their ability to avoid HIV. The second was a lack of solidarity amongst sex workers, competing for clients who paid more for unprotected sex. Dependent on these clients for survival, advice to “use condoms” was not particularly useful. The third was lack of bonding social capital – support from significant others in the communities where they lived and worked. Landladies encouraged them to drink heavily to create a convivial atmosphere that attracted clients, undermining condom use. Macho male clients said condoms undermined the sexual pleasure they paid for. The fourth undermining factors was lack of support from regional health and welfare agencies (bridging social capital), who provided minimal services which sex workers found difficult to access.

Critics of peer education fall into two groups (Cornish & Campbell, 2009). The first focuses on the methods used by peer educators, arguing that whilst peers may be best placed to reach marginalised groups, most programmes use outdated teaching methods. These involve the linear transmission of information, where the peer educator plays the role of active expert, passing information to a passive audience. Such critics argue that dialogical educational approaches are more likely to lead to the development of the critical thinking, agency and solidarity that have been absent in many peer education settings. They are also more likely to enable participants to negotiate healthier behavioural strategies that are realistic within the realities of their everyday lives. The second group of critics focus on the contexts of peer education arguing that programmes need to be backed up by efforts to create health-enabling social environments. Many have argued that community mobilisation approaches could address both of these criticisms, leading to a growing emphasis on the “third generation” of approaches.

Third generation approaches: community mobilisation

The third generation of responses to HIV/AIDS identifies the locus of change more widely than peer education, seeking to implement change at the level of communities, in order to create community contexts that support the development of health-enhancing peer norms and individual behaviour change. Community mobilisation approaches seek to create and harness the agency of the marginalised groups most vulnerable to HIV/AIDS, enabling them to build a collective, community response, through their full participation in the design, implementation and leadership of health programmes, and by forging supportive partnerships with significant groups both inside and outside of the community. In so doing, these approaches seek to tackle the criticisms of peer education discussed above, through facilitating social capital, dialogue and empowerment. We discuss each of these processes, viewing them as the pathways between participation and behaviour change, and the social psychological processes through which participation has its impact on health.

Social capital

Limited access to social capital (defined as durable networks of socially advantageous inter-group relationships) is a key factor in perpetuating poverty and other social disadvantages that prevent people from taking control over their health and improving their life circumstances (Bourdieu, 1986). Community mobilisation approaches assume that marginalised groupings are most likely to change their behaviours if there is significant support for new behaviours, both from other groups within the communities in which they are located (bonding social capital) (Skovdal, pp. 1652–1661), and from more powerful actors and agencies outside of the community, who have the political and economic power to provide them with meaningful support (bridging social capital) (Cornish, pp. 1670–1678). A core goal of community mobilisation approaches is to facilitate such relationships through community participation and partnership building.

Social spaces for critical dialogue

Elsewhere we have emphasised the need for projects to create “transformative social spaces” in which people are able to develop actionable understandings of how to improve their health (Campbell & Cornish, 2010). At this stage of the epidemic, often even the most marginalised have access to basic HIV/AIDS information. What they lack are “safe social spaces” where they can engage in dialogue and debate about new information (Vaughan, 2010; Vaughan, pp. 1644–1651). It is through such dialogue that people are most likely to gradually take ownership of initially unfamiliar and alien medical information, airing doubts or confusions, and working through ways to incorporate new information into their existing worldviews and practices. Such social spaces also provide contexts where people can brainstorm potential behaviour changes that are realistic within the constraints of their everyday lives. Ideally they can also discuss ways in which they might challenge unhealthy social relations. Community mobilisation programmes need to facilitate the development of such spaces and such dialogue.

Empowerment/agency

Peer educational approaches typically define “empowerment” narrowly, focusing only on HIV/AIDS-specific skills such as negotiating condom use or procuring condoms. Community mobilisation approaches assume that people are most likely to feel empowered to take control over their health if they have experiences of exercising agency in other areas of their lives (Wallerstein, 1992). In other words, they are more likely to feel empowered if they actually are empowered. Community mobilisation approaches to HIV/AIDS seek to involve community members in efforts to bring about health-enhancing changes, such as persuading health services not to discriminate against PLHA, or pressurising the police not to abuse their powers against sex workers, for example. Such actions may have a positive impact on health both directly, by reducing some of the environmental risks which undermine health, and indirectly, through supporting a feeling of empowerment which encourages subsequent individual and collective health-promoting action.

Arnstein's (1971) classic “ladder of participation” classifies opportunities for community participation in health and social development projects on a continuum. The one extreme of this continuum is “weak” participation – referring to tokenistic forms of involvement where community members are invited to become involved in projects that are designed and managed by outsiders (usually health or social development professionals), with community members exerting little or no influence on the project's shape or in project decision-making. Many peer education programmes would be classified as such. The other extreme of this continuum is “strong” participation. Here projects may still initially be established and funded by outsiders, but they focus strongly on building the capacity of marginalised community members to eventually exercise leadership in every aspect of project design, budgeting, implementation and decision-making. It is projects of the latter kind that are said to provide the optimal opportunities for the health-related empowerment of participants.

Through these three means (building social capital, dialogue, and empowerment), community mobilisation approaches seek to build a more sustainable and effective response to HIV/AIDS than has been achieved through peer education or HIV awareness efforts. However, whilst there is general agreement that strong community mobilisation can play a vital role in enhancing responses to prevention, care and treatment, the outcomes of efforts to facilitate effective mobilisation in health programmes are mixed (Cornish & Campbell, 2009). Critics argue that mobilising communities to lead responses to HIV/AIDS is a near impossible task, emphasising that disadvantaged communities often lack both the motivation to work together to improve health, and the special skills needed to lead and maintain the demanding work of health interventions (Watkins, 2003). If communities had the capacity to mobilise to improve health and self-empowerment, it is argued, they would not still be in the disadvantaged and health-damaging conditions that health programmes seek to alleviate.

Rather than simply conceptualising communities as lacking the resources for community mobilisation, however, we argue that greater attention could be given to the contexts in which community mobilisation efforts are initiated, and which shape their prospects for success or failure. This is the “fourth generation” perspective the current special issue seeks to advance: a focus on the role of social environments in enabling or constraining effective community responses to HIV/AIDS. In order to start mapping out the dimensions of the social environment which shape programme success or failure, we discuss two case studies of interventions located in very different contexts.

Contexts of community mobilisation: two case studies

The Entabeni Project in South Africa, and the Sonagachi Project in India, which we have compared elsewhere (Campbell & Cornish, 2010), are two HIV/AIDS projects which fared very differently in their aims to mobilise a sustainable community response, and which took place in sharply contrasting contexts. The former sought to build the capacity of poor rural women to deliver effective home nursing to people dying of AIDS in an isolated rural community. The latter sought to build the capacity of impoverished and excluded sex workers to take control over their sexual health in Kolkata, India. Both projects used the method of promoting peer education amongst poor women as their starting points, but hoped that this would form a basis for an effective community response to AIDS.

Both projects were very successful in building the skills and knowledge of women deeply marginalised by gender and poverty. In both cases, women were successfully trained in peer education skills and worked hard and enthusiastically to promote effective HIV/AIDS management in their particular communities. However only the Sonagachi Project succeeded in building supportive social environments in which more powerful outsiders were willing to support sex workers in improving their health and well-being. The project succeeded in mobilising the support of police, politicians and journalists, all key constituencies who had the power to support the sex workers’ goals and advance their interests in significant ways. These improved relationships resulted in, for example: fewer violent police raids and arrests of sex workers, legal changes to allow sex workers to register their micro-credit co-operative, a donation of municipal land for a clinic, and positive media reports celebrating project successes. This was not the case in Entabeni where the efforts of feisty and well trained women were persistently undermined by local community leaders, and by vitally important health and welfare officials in the public sector.

In Campbell and Cornish (2010), we tease out how pre-existing social contexts impacted on the possibility of effective community mobilisation in each case, and highlighting the way in which the Sonagachi Project managed to alter aspects of the previously limiting symbolic, material and relational contexts of project participants’ lives in the lives of project participants in ways that the Entabeni Project did not.

Symbolic context

The symbolic context comprises the meanings, ideologies and worldviews circulating in society. When these meanings are about specific communities, they may convey respect or stigma, which in turn affect how those communities are treated by others, and how they feel about themselves.

The Entabeni Project was unable to tackle three forms of stigma which undermined its efforts to mobilise an effective community response. Firstly the stigmatisation of AIDS as immoral and shameful. Secondly the stigmatisation of caring for the sick, dismissed as “women's work” holding little value for men or young people, two groups the project had been keen to mobilise. Thirdly, the stigmatisation of volunteer work, which was seen – again by men and young people – as holding little dignity. These forms of stigma undermined both the volunteers’ motivation and the likelihood that men, young people or potential external partners would take the project seriously and give it the necessary support. By contrast, the Sonagachi Project managed to actively challenge the stigmatisation of sex work and of HIV/AIDS, previously a key obstacle to sex workers’ ability to take control over their sexual health. It did this by framing sex workers as successful and legitimate workers, who needed to unite to secure their rights, just as other groups of workers (e.g. cobblers, rickshaw drivers) had done in Kolkata in the past. HIV/AIDS was framed as an occupational health issue, facing a group of workers with rights to safe working conditions.

The Sonagachi Project's success in challenging stigma was facilitated by a very different pre-existing political context. West Bengal had a strong tradition of trade unionism and left-wing political parties, with wide familiarity with the concepts of solidarity, collective bargaining and protest, all actively invoked by sex workers as they engaged in project activities. There was no such history in isolated rural Entabeni, where a hereditary chief wielded absolute power, prohibiting any kind of political opposition in a context very removed from any contact with South Africa's anti-apartheid history. Unlike Sonagachi sex workers, Entabeni home-based carers had no access to stories of disempowered women successfully fighting for social recognition, and were pessimistic about the possibility for democratic social change in their community.

Material context

The second dimension of context which the Sonagachi Project managed to tackle, but not Entabeni, was the material context. Each project sought to support its participants to better cope with the crippling poverty that undermined their opportunities for well-being. Sonagachi was able to contribute to participants’ economic empowerment through providing small payments to peer educators, giving their work a formal value and freeing some of their paid sex work time to engage in project activities. By contrast, despite three years of hard lobbying, the Entabeni Project had no success in securing promised government stipends for its volunteers. The Sonagachi Project also succeeded in setting up valued and effective micro-credit and savings schemes for participants. Here again, the Entabeni Project had no success. The lack of economic gain associated with project involvement meant that it failed to resonate with peoples’ economic aspirations, severely limiting its ability to mobilise volunteers to carry out project work, and undermining the status of the project in the community.

Relational context

The third dimension of context that the Sonagachi Project was able to change, and the Entabeni Project was not, was the relational context of the project. In relation to leadership and decision-making, Sonagachi succeeded in effecting the gradual involvement of women in organisational decision-making and leadership. In contrast, Entabeni was unable to dislodge strong local resistance to female leadership. Secondly, in relation to building “bridging social capital” (alliances with supportive outside groups), whilst Entabeni had some success with two small poorly funded NGOs, it failed to build effective links with public sector health and welfare agencies. Such links would have been vital for project success because the effectiveness of home-based care is greatly enhanced if patients are able to access those formal health services and welfare grants that exist, and if public sector services are tailored to community needs. The Sonagachi Project, by contrast, was able to build strong supportive relationships with outside actors and agencies, including politicians, police and journalists (see Cornish, pp. 1670–1678).

Towards supportive social environments for community mobilisation

These case studies suggest that the success or failure of these two community mobilisation interventions were not only attributable to the efforts of the marginalised women who mobilised a collective community response, but were also profoundly shaped by the symbolic, material and relational contexts within which the initiatives took place. To date, little research effort has been directed at this “fourth generation” area for analysis and action. Without a sophisticated understanding of the impact of social context on community interventions, there is a risk that the theoretically well-founded interest in community mobilisation, among policymakers and programme implementers, will yield no fruits. If community mobilisation is initiated in adverse social environments, and without the necessary concurrent efforts to create the conditions for its flourishing, it is likely to suffer the same fate as its predecessors – peer education and HIV awareness. It is in the interest of developing this “fourth generation” perspective, and thus maximising the chances of community mobilisation achieving its potential that we have brought together the papers in this special issue.

The impact of social environments on community mobilisation initiatives: the current special issue

This special issue brings together 12 very different papers, united by their focus on the impacts of social environments on efforts to mobilise HIV/AIDS-vulnerable communities. Each paper addresses one or more of the dimensions of context that we have discussed above. These are rich and nuanced papers, and our brief account of them picks out only very partial aspects of their complex set of findings.

Material context

The international funding architecture comes under scrutiny in three papers as a key aspect of the material context that influences the shape of community interventions. In a large-scale study of HIV/AIDS funding in east and southern Africa, Kelly and Birdsall document a recent exponential growth of civil society organisations, in response to donor policies which identify civil society as the best route to rapidly scaling up HIV/AIDS interventions. They argue, however, that this funding environment may in fact undermine the special contribution of civil society – because in this system, civil society organisations have become mechanisms for the delivery of externally designed services, not for community mobilisation.

Kelly and Birdsall's sweeping overview provides an important backdrop for two in-depth case studies by Aveling and Cassidy, which examine the impact of international funding practices on community HIV prevention projects with military families in Cambodia, and support groups for people with AIDS in Gambia respectively. What Aveling refers to as “aid chains” construct complex social spaces in which local people (“beneficiaries”), organisations running interventions, and international donors are brought together, with often unexpected and contradictory results. Her Cambodian research shows how the partnerships characterising the “aid chain” both enable local-level empowerment by serving as a conduit for useful resources, and undermine such empowerment by strictly controlling the parameters of interventions. Countering any temptation to caricature aid relationships in terms of one-dimensional stereotypes of powerful funders and powerless poor communities, Cassidy's Gambian study highlights how desperately poor people living with HIV/AIDS juggle their loyalties to competing international/biomedical and local/traditional AIDS initiatives in their efforts to gain a complex and contradictory array of rewards and benefits as they struggle to stay alive in conditions of severe marginalisation and powerlessness.

The institutional constraints of a government public health system comprise the material context addressed in Scott's study of a community health worker initiative in rural northern India. The commendable intention to mobilise local women to promote their peers’ reproductive health, Scott argues, was not backed up by an appropriate system of remuneration for the community health workers, properly resourced clinics, or genuine devolution of power. Her study suggests that the strongly hierarchical and centralised nature of traditional public health services is diametrically opposed to the participatory ethos of community mobilisation.

Symbolic context

Three papers, focusing on symbolic context, highlight difficulties faced by community interventions when dominant understandings of HIV/AIDS are at odds with community understandings and interests. In a paper addressing the challenge of implementing “gender mainstreaming” in southern Africa, Mannell further develops the theme of global-local relations in her study of the differing understandings of “gender” held by internationally funded HIV/AIDS programmes, and by local communities. She argues that one reason for the much-lamented failure of many gender mainstreaming efforts is that foreign donors impose an inappropriate western and neoliberal understanding of gender onto African communities.

The mass media play an important role in shaping public understandings of HIV/AIDS and appropriate interventions, including the understandings held by policymakers and programme implementers. Gibbs examines how the South African print media represent the gendered nature of the HIV/AIDS pandemic, and potential solutions. While newspapers generally recognise the powerful social forces undermining women's abilities to protect their health, they give little attention to the role of grassroots women's community mobilisation. Gibbs calls for greater understandings of the role of community mobilisation, not only in the academic literature, but also in public understandings.

Symbolically marginalised and stigmatised social groups are often particularly vulnerable to HIV/AIDS, and challenging that stigma is a crucial early step in mobilising a community and establishing their local legitimacy. Blankenship, Biradavolu, Jenna and George provide a case study of an occasion in which a community-led intervention for sex workers in southern India joined forces with a government-sponsored HIV awareness initiative. In part, the collaboration made sex workers complicit in their own stigmatisation, due to the conservatively framed messages of the campaign, but it also provided opportunities to challenge that stigma. The study indicates the limits and possibilities of community-led efforts to alter their symbolic context.

Relational context

Five papers focus on the relational contexts of programmes through their accounts of the challenges of getting key partners to support AIDS projects. Campbell raises the issue of political will in her paper on the impact of the traditional leadership system in the South African community of Stoneyridge. The goals of empowering women and youth through their participation in an HIV/AIDS programme were at variance with the goals of the area's traditional leader, who derived his political power from his role as “guardian of tradition”, involving the very gender and age hierarchies that have advanced the epidemic and undermined local people's ability to respond effectively.

Similar challenges are taken up by Vaughan, in research into HIV prevention among youth in Papua New Guinea. Vaughan draws attention to the gaps between young people and donors, and young people and local leaders, articulating how these undermine HIV prevention efforts. She provides an exemplar, through a Photovoice project, of how “in-between social spaces” may be created, to enable fruitful dialogue between influential adults and young people, for the advancement of young people's health.

A cash transfer project for young carers in Kenya, reported by Skovdal, also provided helpful space for young people to develop supportive social relations with each other, resulting in the young people helping each other in ways that generalised beyond the project. In conditions of desperate poverty, however, the project battled with'free riders' not taking their share of responsibility and people not being trustworthy with the group's money. Unpromising wider community relationships are also the concern of Nhamo, Campbell and Gregson, in a study of a rural Zimbabwean community. The efforts of an HIV prevention programme were undermined by a deeply rooted traditionalism, characterised by hierarchical social relations, social divisions and a moralistic response to HIV/AIDS, combined with alienating prior experiences of the provision of development aid and public services.

In contrast, using case studies of two communities where social relations were more enabling of participatory HIV/AIDS interventions, Cornish, Shukla and Banerji investigate Indian sex worker organisations’ relatively successful efforts to improve their relationships with local powerful actors (including police and politicians) whose actions impact on sex workers’ abilities to protect their health. They argue that these efforts take the form of “exchanging favours” in which the sex workers’ organisations find incentives to offer to the powerful in exchange for their support – an approach which is potentially productive yet also risky.

Conclusion

In this paper, we have articulated the rationale for community mobilisation approaches to HIV/AIDS management, contextualised by our account of three dominant approaches to behaviour change. These range on a continuum from more individual-focused approaches, such as HIV awareness, to peer-based approaches, to more socially sensitive approaches, such as community mobilisation. We have sought to make the case for the development of a fourth, even more social approach – which takes account of how wider social contexts facilitate or hinder community mobilisation approaches.

This special issue takes a first step in elaborating this “fourth generation” perspective, bringing together papers which provide evidence and argument for the role of social context in shaping community mobilisation interventions in a range of case studies. Collectively, the papers present a picture of significant contextual barriers to successful, empowering community mobilisation. They document how the existing material, symbolic, and relational contexts of HIV/AIDS-vulnerable communities often create near impossible conditions for the development of locally sensitive, flexible, empowering and sustainable community mobilisation. As such, this first step, we believe, confirms that the wider social context of interventions deserves our attention. At the same time, however, the implications of this first step appear discouraging. One interpretation might be that community mobilisation is not worth attempting if it is so often fatally undermined by the stark realities of divisive and disempowering social environments.

An alternative response is possible. We stand by our arguments for the importance of community mobilisation in generating a sustainable local response to HIV/AIDS, through developing social capital, dialogue and empowerment. Rather than dismissing the potential of community mobilisation, we suggest turning attention to the question of how social environments can be changed or created so as to maximally support community mobilisation efforts. Social environments are not fixed, but are often a product of the actions of powerful decision-makers such as donors, government agencies and politicians. The question, thus, for this “fourth generation” perspective becomes a question of political will. To prepare the ground for successful community mobilisation initiatives, it is not enough to prepare local volunteers with the skills to run an intervention together. Mobilising political will to create supportive conditions for community action, at local, national, and international levels is also required. The second step, then, for the “fourth generation” perspective, is to focus on what powerful actors can do, to create conditions conducive to community mobilisation. In what follows, we indicate some possible directions.

Donors, development agencies, and government agencies need to consider how their policies and actions create the environment in which community programmes operate. Pre-defined project plans, reporting procedures requiring advanced bureaucratic expertise, short-term funding cycles, a lack of responsiveness to local diversity and priorities, stigmatising definitions of problems, and strongly hierarchical relationships all undermine community mobilisation – but also can all be altered. If the many organisations writing and endorsing HIV/AIDS policies which prioritise community approaches are serious about this goal, they should consider whether their structures are sufficiently tolerant of local diversity and supportive of local empowerment.

Academic research constitutes part of the symbolic environment of community mobilisation initiatives. Academic discussions play a large role in the creation of concepts such as “gender mainstreaming”, “vectors of HIV”, “empowerment” and even “community mobilisation” – concepts which in turn inform and shape policy and practice. To help create optimal conditions for successful community mobilisation, researchers could shift attention “upwards” to the behaviour of the powerful, rather than continuing to focus “downwards” on the behaviour of marginalised communities. We hope that this special issue has taken a first step towards promoting academic debate on the significance of social environments in shaping community interventions, and that it might spark attention to what we see as the second step, namely the question of how to mobilise the necessary political will to give community mobilisation a real chance of success.