Resource-poor countries have been greatly challenged by the HIV/AIDS epidemic, now in its fourth decade. According to the Joint United Nations Program on HIV/AIDS, at the end of 2009, 97 percent of the 33.3 million people with HIV lived in resource-poor countries, but only 5.2 million were receiving antiretroviral therapy (ART) (UNAIDS, 2010a; World Health Organization [WHO], 2010). The public sectors in these countries are faced with the difficulties of organizing resources efficiently for controlling HIV (McKinsey & Co., 2005). China is among those countries facing these challenges; among its provinces, Yunnan has the fewest resources and has been the worst hit. Yunnan reported its first domestic case of HIV in 1987 (Ma, Li, & Zhang, 1990) and, in 1989, reported that 145 heroin users in the Golden Triangle border regions were part of the first HIV outbreak (China AIDS Initiative, 2009). To confront this challenge, Yunnan applied a government-driven multi-sector partnership model for HIV control. This model has generated positive effects in capacity building and stabilization of HIV transmission among intravenous drug users (IDUs), who form the majority of those with HIV in Yunnan. Data from the U.S. National Institute of Health (NIH) show that the rate of new HIV infection among IDUs in Yunnan has slowed, while overall HIV infection in China has increased (U.S. NIH, 2010). This study seeks to examine the organizational strategies of Yunnan's model. How did the Yunnan government contain the HIV infection? What insights can we derive from Yunnan's experience in the government-driven multi-party partnership in addressing major social challenges? What are the ways to include non-state actors in the management of development threats in a state-dominated context?
Public sector led partnerships using the multi-sector model have been used previously in the history of fighting HIV/AIDS in resource-poor countries. Most notably, the Ugandan government instituted educational efforts in the early stages of their HIV/AIDS campaign, and New York City established a multi-sector collaboration (AVERT, 2010b; Chambre, 2006). In fact, New York achieved notable success with peer counselling or community outreach techniques (Chambre, 2006). Yet, the Yunnan model differs from other public sector led models in its forceful, one-track, one-web/network-based organizational approach that integrates a multi-sector response under government control. This approach follows the evolution of China's public policy in HIV response, from defensive and reactive to proactive. That is, Yunnan has made maximum use of its political networks and authority for organizing a multi-sector partnership in HIV control. In this framework, the government acts as the leader for other stakeholders in the partnership. Most civil society organizations in this framework are subsumed under government authority for the planning, financing, and implementation of HIV control and to work towards consistent goals and objectives. More specifically, similar to the idea of a centrally-managed multi-party network, the Yunnan government has inserted itself into civil society operations to ensure that HIV-control operations are consistent with the nation's long-term development goals (Lin, 2011).
Historical and Theoretical Background
Yunnan's organizational model has largely made irrelevant the conventional boundary between civil society and the public sector. In this sense, the Yunnan model and the overall structure of the China model diverges from the norm. This one-track government-driven model contrasts with other donor-driven HIV control models in resource-poor countries in which donors have more input regarding program frameworks. In the Yunnan model, civil society yields to government goals, and even foreign donors must follow the government's organizational framework. The Yunnan experience provides an invaluable opportunity to observe the strengths and weaknesses of Yunnan's organizational strategies in HIV control. In addition, it allows us to discuss these results from a theoretical point of view.
From denying the existence of homosexuality to operating a gay bar in Dali, Yunnan has made what Western observers call a dramatic turnaround in its management of the HIV epidemic (AVERT, 2009; Reuters, 2009). In November, 2009, despite public opposition, the Dali City health department opened the first gay bar in the People's Republic to reach out to men having sex with men (MSM). The Yunnan government, as the official bar owner, uses it as a venue for HIV prevention and intervention, including hiring MSM staff to carry out the HIV-control program. Few governments have taken such drastic measures to combat HIV, but Dali City has one of the highest HIV rates in China. Yunnan's proactive measures in confronting HIV transmission among IDUs are also widely known; in the 1990s, it was the first province to open methadone treatment clinics, which were so effective that by 2010 the central government expanded the program to 5,000 methadone treatment clinics to treat IDUs (Li, Lee, Gan, Tan, Meng, & He, 2007). Yunnan was also the first province to allow free needle/syringe exchange, although these programs are still illegal in many other countries. The public sector is the main engineer of these organizational strategies.
HIV Partnership Models in Other Parts of the World
Since the beginning of the new millennium, global health partnerships have increased to confront old and new epidemics and pandemics, especially HIV, malaria, tuberculosis (TB), trachoma, and other infectious diseases. These alliances/partnerships have become a prevailing mode to combat global health challenges, and they are often organized as simple affiliations (loose alliances with equal partnerships and informal structures), lead partners (taking major responsibilities but no dominant roles), secretariat (quasi-formal organizations and staffs to coordinate equal partnerships), general contractors (partners acting as clear decision makers on planning, financing, and contracting service delivery to civil society organisations), and joint venture companies (partners create their own separate legal entities to manage their resources, staff, and independent actions) (Bill and Melinda Gates Foundation, 2002).
Regarding HIV control, the most notable partnerships are the Global Fund to Fight AIDS, Tuberculosis and Malaria, an international partnership between public and private organizations; the U.S. President's Emergency Plan for AIDS Relief (PEPFAR) (a U.S. government initiative for HIV prevention and intervention mainly in high prevalence regions, such as African countries); the Joint United Nations Program on HIV/AIDS (UNAIDS) (a joint venture combining the efforts and resources of ten UN organizations); the Gates Foundation AIDS initiative and partnership; and the World Bank's Multi-Country HIV/AIDS Program (MAP) (Bill and Melinda Gates Foundation, 2010a; Kaiser Family Foundation, 2010; PEFPAR, 2010). These partnerships are the major sources of funding for HIV control in resource-poor countries. Between 2001 and 2006, HIV/AIDS-related activities have seen increased funding, from $2.1 billion to $8.9 billion, mainly to support programs of the Global Fund, PEPFAR, and MAP. The Global Fund, widely considered the standard-setter for global health partnerships, provides 72 percent of HIV-related disbursements from multilateral sources and has been the main source of AIDS funding in 52 of 92 recipient countries (UNAIDS, 2010b). PEPFAR supports prevention, treatment, and care in focus countries (about 15 in Africa, the Caribbean, and Vietnam). The World Bank MAP program for AIDS relief mainly funds sector loans or broad budgetary support loans and, to a smaller extent, grants supporting HIV control by governments and civil societies in the poorest countries (Shakow, 2006: 4; World Bank, 2010). The Gates Foundation supports country-based comprehensive initiatives, such as those in Botswana and India, as well as individual proposals.
Overall, these partnerships are donor-driven partnership models and share certain organizational characteristics in their HIV-control programs (Anon., 2010; Bill and Melinda Gates Foundation, 2010b; The Global Fund, 2010; UNAIDS, 2010a). First, most partnerships are based in developed countries and are executed by non-local donors, which give directions to non-government organizations (NGOs), to deliver services such as pharmaceutical access, health education, outreach, and treatment (Reich, 2002). Host countries tend to play supporting roles, while donor agencies make most of the planning decisions. Second, most partnerships have limited timelines; renewed funding depends on performance indicators and political and economic climates. For example, the Monster and Critics (2009) has a typical five-year funding cycle. Third, most of the programs focus on specific diseases and support mainly HIV-related technical capacity building (World Bank, 2010).
The donor-driven model has generated problems for HIV control in resource-poor countries. An example is Uganda, long regarded as an exemplary case for resource-poor settings until it changed from a dual-track model to a donor-driven model. The initial dual-track model reduced HIV rates from 20 percent to less than 8 percent. After foreign donors increased funding, especially PEPFAR after 2004, Uganda's model became donor-driven, and some donors demanded certain approaches, which critics believe led to increased HIV infection rates. Other programs also show high donor influence on local HIV strategies, such as the African Comprehensive HIV/AIDS Partnership (ACHAP) in Botswana, supported mainly by foreign donors. In Mozambique, international assistance accounts for 98 percent of HIV funding (Youde, 2010).
Regarding organizational strategies for this partnership model, research shows strengths and weaknesses. Host countries enjoy some increased technical capacities mainly in planning, monitoring, responding to neglected areas, enlarging key interventions, and increasing civil society's profile, especially NGOs and the private sector (McKinsey & Co., 2005). Yet, this model has obvious weaknesses (AVERT, 2010a,b,c; McKinsey & Co., 2005; Shakow, 2006). A review of several major global health partnerships, including the Global Fund, PEPFAR, and MAP, shows that these programs have many challenges at the country level, such as poor communication, coordination and duplication among partners (lacking clearly defined roles and responsibilities), high transaction costs, variable degrees of country ownership (government lacking control over outside donors’ decisions), and lack of alignment with local health systems. Problems arise because these outside agencies use a one-size-fits-all model. Focusing on one disease disrupts health care and development priorities, for example, in human resources and capacity building (Buse, 2004; McKinsey & Co., 2005; Shakow, 2006: 4). In Uganda, some NGOs duplicate efforts by working in the same location (Barr & Fatchamps, 2005).
The most serious problem is that these partnerships lack long-term focus and fail to support the sustainable development goals of the host countries (Shah, 2009; Wang, 2009; Yu, Souteyrand, Banda, Kaufman, & Perriens, 2008). Mobilizing the necessary resources in resource-poor settings already presents serious challenges. Donors focusing on the single task of fighting HIV/AIDS often divert precious resources from other long-term challenges, such as maternal and child health (Shakow, 2006; UNAIDS, 2010a, 2010b). In addition, the funding is uncertain and subject to macroeconomic and political changes. In 2009, the Global Fund, considered the norm of HIV partnerships, experienced a $4 billion budget shortfall, which Director Michel Kazatchkine called a ‘financial crisis’ (Alcorn, 2009). Donors are also selective in terms of recipients and funding focus, often being reluctant to assist such stigmatized groups as sex workers or IDUs (AVERT, 2010c). PEPFAR's emphasis on abstinence from 2004 to 2008 was especially controversial (AVERT, 2010d; Salaam-Blyther, 2008). The donor-driven model also makes it difficult for partnerships to make long-term commitments or to be accountable to the host countries (Shah, 2009; Yu et al., 2008). Overall, the host government has fewer resources and little incentive or authority to address HIV root causes, such as poverty, infrastructure inadequacy, gender, and social economic inequity (AVERT, 2010d; Hanefeld & Musheke, 2009; Yu et al., 2008). The World Bank (2006) concluded: ‘despite progress in a small but growing number of countries, the AIDS epidemic continues to outstrip global efforts to contain it’ (Anon., 2010; IFPMA, 2009; UNAIDS, 2009a).
Given the inability to address the sustainable goals in the prevailing models of global health partnerships, Yunnan's government-driven, multi-sector partnership model presents a different solution.
From mid-1990 to 2004, China saw exponential increases in new cases, from 10 percent to 40 percent, reaching 132,545 cases by August 2005. In 2006, the official estimate was more than 800,000 (UNAIDS, 2010a). Although that number may not seem large in a country of 1.3 billion, that report was probably the lowest estimate. The UNAIDS (2010a) report estimated that the actual count could reach 10 million by 2010 and the speed of transmission is particularly troubling, especially for marginalized populations; MSM, transgender individuals, and their sex partners could number more than 50 million. HIV transmission has spread unevenly across China, and the most vulnerable tend to be more likely to be infected. In 2005, Yunnan, Henan, Kwanhsi, Hsinjian, and Canton accounted for 77 percent of total cases: 41.6 percent through illicit drugs, 23.5 percent through sexual contact, and 22 percent through unknown channels (Beijing Sina Net, 2005; UNAIDS, 2010a). In recent years, transmission through sexual contact has outpaced other means and is now the major route for HIV infection (Ma et al., 1990). The sharp rise in HIV infection among MSM has aroused the most attention (UNAIDS, 2009a). In China, about 11.1 percent of the estimated 800,000 people living with HIV are MSM: among these, about 2.5 percent to 6.5 percent have HIV (UNAIDS, 2009b). The multidimensional nature of HIV transmission makes it difficult to use simplistic, isolated, or single actor approaches.
To address the research question about how non-state actors are included in the fight against the HIV/AIDS epidemic in a state-dominated context, this study applies a case study approach to its study of Yunnan's management of the HIV/AIDS epidemic.
To observe organizational strategies used to coordinate a multi-sector response to HIV in a resource-constrained environment, I used a case study approach to study Yunnan's HIV-control model from 1985 to 2010. Yunnan's provincial government, referred to as the local government throughout the article, was a key element.
According to Yin, the major strength of the case study is its ability to deal with a wide range of evidence, such as ‘documents, artifacts, interviews, and observations – beyond what might be available in the conventional historical study’ (Yin, 1984: 8). In addition, according to Yin (1984), the case study is not only a collection technique but a comprehensive research strategy that takes into account the many variables affecting one result, allows triangulation, and is guided by theoretical propositions. This research strategy allows us not only to investigate the questions of ‘what’, and ‘where’ but most importantly to examine ‘how’, and ‘why’ in the context of a given theory. In a case study, the preferred analytic strategy is to rely on theoretical propositions to guide the analysis by examining the patterns, or ‘pattern matching’, revealed in the evidence (Yin, 1984: 106). In other words, it can be used to find dominant and recurrent ‘patterns’ found in a wide array of data included in the study. This strategy is similar to the finding of the ‘dominant themes’ of each story or narrative in the use of content analysis.
The other analytic strategy related to the current study is the use of chronologies, which also allows us to examine questions about the relationships of events over time. Specifically, this strategy makes it possible for researchers to establish causal links among events or ‘the causality among events’ (Yin: 117–118).
These two strategies will be used to analyze the data collected in Yunnan for this study in order to explain the ‘chronologies’ and ‘causality'of the HIV/AIDS partnership model used in Yunnan. In particular, the findings will be used to examine the organizational strategies of the HIV partnership in Yunnan in order to answer several questions: how did the Yunnan partnership contain the HIV infection? What insight can we gain from Yunnan's experience in government-driven multi-party partnership in addressing a society's development challenges? What are the ways to include non-state actors in the management of public health or development threats by the state?
I drew on data from multiple sources, including public and historical documents, ethnographic interviews, and participant observations. I examined policy statements, media records, and studies from UNAIDS, BBC, People's Daily, Xinhua News, New York Times, Washington Post, Financial Times, Wall Street Journal, Lancet, AIDS Education, and reports from the International AIDS Society and 16 HIV-related programs. I conducted ethnographic and field studies through personal interviews with 34 individuals in 21 organizations and affected and infected communities of MSM, transgender men, commercial sex workers, and IDUs (Appendix I).
Using the aforementioned strategies, I examined the narratives for recurrent and dominant themes or patterns in the data related to: (i) the epidemiology and the HIV transmission pattern in Yunnan in the context of China; (ii) the ‘attention-grabbing’ and ‘turnaround’ events in the fight against the HIV epidemic in Yunnan and in China; (iii) the evolution of the state's role and capacity in the fight against HIV and the corresponding policy changes at the local and national levels; (iv) the major themes of HIV-related stories about Yunnan and China from the selected media; (v) the organizational frameworks of the HIV partnerships chosen for the study; and (vi) the comparison of the Yunnan model with the donor-driven model. The ‘dominant themes’ from different sources were juxtaposed and compared to address inconsistency. For example, while the community based organizations (CBOs) or the local surveillance sources could only provide confirmed HIV infection cases and did not generate consistent figures in the incidence rate (the long-term predictions), I found UNAIDS and the Global Fund were much more useful sources in providing incidence data over the long term because of their experience in collecting global data. Specifically, the foci of my analyses are: the history and progression of events, evolution, and ‘turnaround’ of Yunnan's HIV-control model; the operation of the partnership model regarding organization and outcomes; and the strengths and challenges of this organizational mode. I have used multiple sources of data for verification of the authenticity of the data (or triangulation) and to ensure the inter-coder agreement is more than 90 percent. I analyze and discuss these themes in terms of theoretical importance and implications.
The History of the HIV/AIDS Epidemic in China
When HIV/AIDS began spreading, China was forced to learn rapidly about epidemic control. Yunnan's response remains embodied in China's national framework but has been adapted to local needs. It is difficult to apply a simple treatment model because the HIV retrovirus constantly mutates: medicines must be changed regularly and permanent cures are as yet unavailable. Also HIV/AIDS is closely linked to social, cultural, and economic taboos, such as commercial sex, homosexuality, and illicit drug use.
Epidemiological studies by Dr Zhang Kong-Lai, China's pioneer HIV researcher, his colleagues, and UNAIDS, show that China's epidemic followed several phases and reached the general population through ‘bridging groups’, such as drug users, MSM, and commercial sex workers. In the first phase, 1985 to 1988, the first cases appeared among Yunnan IDUs, a few foreigners, and Chinese overseas in coastal cities. Four cases surfaced among haemophiliacs from Zhejiang province who had used imported factor VIII. The second phase, from 1989 to 1993, found HIV infection among 146 drug users in Yunnan's minority communities. In 1994, HIV infection was only among drug users and commercial plasma donors. But in 1998, the third phase, HIV had spread to all 31 provinces, autonomous regions, and municipalities under central government control.
HIV in Yunnan
Yunnan Province has been at the forefront of the HIV epidemic because of its unique demographic and geographical characteristics. Bordering on Burma and Thailand, Yunnan has many minorities; most are poor and live in mountainous regions with various temperature zones. Poverty and remote locations render its population vulnerable to other infectious diseases, such TB and malaria, which could exacerbate HIV's morbidity. The mountains are close to the Golden Triangle, a major site of Asian opium and heroin production. Ethnic minorities near Burma report disproportionately high drug use (Xiao, Kristensen, Sun, Lu, & Vermund, 2007), a major risk factor for HIV.
In 1994, Yunnan's HIV was mostly concentrated among IDUs (Cheng, Zhang, Jia, Kuo, Zhang, & Bi, 1995). Since then, Yunnan continues to report the most HIV/AIDS cases per year (Xiao et al., 2007), mainly related to intravenous drug use, which accounted for most of Yunnan's 56,054 recorded cases of HIV infections by September 2008 (China Daily, 2008). Among 34 provinces and administrative units, this equals more than 23 percent of China's reported 220,000 HIV carriers that year, the highest rate among all provinces and administrative units, which shows that drug use has seriously challenged Yunnan's HIV control. Beyond drug-related problems, poverty and gender inequity, especially among ethnic minorities, add transmission risks. Moreover, Yunnan has a large percentage of migrants in the cities. In the initial stage of the HIV epidemic, Yuan, Henan, and Hsinjiang were the most affected provinces in China. The epidemic spread from IDUs to the general population through sexual contact, from rural to urban areas, and from border communities to inland townships (Jia et al., 2010; Xiao et al., 2007). These facts presented Yunnan's government with management challenges.
Before the central government changed strategies, the epidemic's complexity was daunting. Yunnan's social, political, economic, and cultural milieu added to difficulties from its at-risk populations. At the early stage of China's economic reform in the 1980s, Yunnan lacked infrastructure or resources for HIV control. Sexual taboos associated with HIV/AIDS also made finding appropriate strategies difficult. As elsewhere in China, Yunnan's public sector had abundant experience in political mobilization for managing partnerships with other stakeholders, including foreign partners, but Yunnan's most serious challenge was to use scarce resources efficiently and effectively.
Evolution of Organizational Strategies
Yunnan's change of strategy was related to the central government's major attitudinal shifts after the epidemic's initial stages. At the epidemic's outset, the central government adopted a defensive strategy, which critics called a phase of denial and inaction. In 1987, the government assumed that HIV/AIDS was an exogenous and ‘limited problem’ and thus established a National Program for AIDS Prevention and Control that focused on infection from foreigners (Jia, Lu, Sun, & Vermund, 2007). Through the 1990s, HIV-control authorities were forced to react to events that had already incurred negative consequences. They enacted a policy controlling illegal blood sales in reaction to the negative publicity associated with widespread HIV infection in Hennan villages in the 1990s. This public management strategy proved ineffective for HIV control.
At the beginning of the 2000s, the government abandoned its defensive strategy and moved towards a pragmatic, partnership approach by revealing the status of the epidemic and moving towards a solution. The New York Times reported ‘there was a notable shift in government response to the epidemic in the new millennium’. On World AIDS Day 2001, stories and testimonials of those infected with HIV were allowed on Chinese TV for the first time (Rosenthal, 2001, Dec. 5). This signalled a major change in terms of how the government communicated the epidemic and how the epidemic was conceptualized by the public sector. Simultaneously, the central government was improving pharmaceutical access for HIV/AIDS patients by working with a Western drug company. This move showed for the first time that the central government was actively intervening to provide effective drugs. Later, the central government pledged to provide free antiretroviral drugs. In 2003, the government made several significant moves to show its commitment in the wake of the SARS epidemic, which cost tremendous GDP losses and interrupted social stability. The health minister announced a willingness to accept foreign assistance, clearly stating that the fight against AIDS is a ‘long-term war’ and requesting that China's AIDS budget of US$12.5 million be doubled (AVERT, 2010a). It signalled that the government was willing to collaborate with foreign stakeholders and to commit its own resources as well. On World AIDS Day in 2003, Prime Minister Wen Jiabao became the first Chinese leader to shake hands with an HIV-positive person, an indicator of social acceptance for HIV patients. Peter Piot, Executive Director of UNAIDS, commented: ‘I really feel that there is a change going on … I know that for top state leaders in China today, AIDS is on the agenda’. (Financial Times, 2003a,b). The central government went from the single-actor, do-it-alone approach to a partnership method that involves foreign and domestic stakeholders, such as UNAIDS, NGOs, CBOs, and government-based NGOs (GONGOs). Appendix II provides the major government's regulatory landmarks regarding HIV-related issues from 1990 to 2006.
The major change was reflected in a public health strategy that replaced the ‘zero-tolerance’ penal strategy formerly used to punish individuals. The new strategy required the public sector to work with various stakeholders for HIV control. Rather than imprison addicted intravenous drug users, the central government planned to open 1,000 methadone treatment clinics to treat drug addicts. Later they expanded this goal to 5,000 clinics (Hammett et al., 2008; Li et al., 2007). Consequently, the government needed CBOs to reach out to drug addicts. Rather than denying homosexuality's existence, the government announced a direct partnership with organizations to administer extensive HIV-related prevention and intervention programs for the estimated five to ten million Chinese MSM (UNAIDS, 2009a).
Yunnan's HIV Response
The changing attitude outlined above enabled the Yunnan government to pursue a government-driven, multi-sector partnership model to work with affected communities. This model follows a top-down process; it integrates the public sector's HIV response, including the national HIV/AIDS control framework and Yunnan's local strategy, with horizontal and vertical interaction between the public sector and civil society. In this model, Yunnan's government leads in the planning and implementation of HIV partnerships, determines the division of labour, and moves the partnerships towards long-term capacity building.
This model ensures that Yunnan's HIV control adheres to the central government's framework, but also allows local flexibility reflected in the Regulation of AIDS Prevention and Control with its strong local emphasis on dealing with HIV risks among IDUs. As mentioned, intravenous drug use is a major cause of HIV transmission in Yunnan. Supported by the central government, this local regulation allows Yunnan to expand needle exchange porgrams, promote condoms, and offer methadone maintenance treatment (Yunnan CDC, 2004).
Since February 2008, USAID has worked with the Yunnan Provincial AIDS Bureau and the Yunnan Centre for Disease Control (CDC) in supporting Health Policy Initiative/Greater Mekong Region and China (HPI/GMR-C) to improve Yunnan's legal environment for HIV control (USAID, 2008). USAID noted that Yunnan's legal and policy framework has become more consistent with international law and best practice, as evidenced by the 2006 Yunnan Provincial AIDS Control and Prevention Regulations. Shore (2009), based in the U.S., also noted that NGO partnerships are a key to HIV/AIDS policy in Yunnan.
Partnership Framework with Civil Society
Following the central government's regulatory framework, the Yunnan government regulates and integrates civil society's activities to support HIV control and overall health and development. In general, NGO activities are regulated through China State Council Order no. 250, September 1998, which regulates the registration and management of social organizations (Riska, 2010; Gang Zhao, interview). In 2010, the new non-profit regulation mandated that NGOs register as non-profits only if sponsored by a government agency or a GONGO (government-based NGO) (Davis, 2010). Domestic civil society organizations failing to follow this regulation must register as private corporations, without non-profit tax benefits.
Yunnan provisional regulations specifically regulate international NGOs (INGOs) (Shieh, 2010), requiring them to have a memorandum of understanding (MOU) with a local partner, usually a GONGO set up by the Department of Foreign Trade and Economic Co-operation (DOFTEC), with the approval of the Yunnan government (Riska, 2010; Shieh, 2010; Gina Gang Zhao, interview; Hua Hsu, interview; Kong-lai Zhang, interview). Externally funded HIV partnerships are encouraged to include GONGOs at horizontal (local) and vertical levels (such as CDCs at local, provincial, and national levels). Partnerships can ask that local CBOs are included in projects (Gang Zhao, interview). MOUs also allow international NGOs to obtain official status for their national staff (Riska, 2010). For example, one volunteer service overseas (VSO) partnership, originating in the USA, includes the local AIDS prevention and control office; health bureaus at provincial, prefecture, and county levels; Mangrove support groups; Daytop Rehab Center; Longchuan County self-help groups; teacher training colleges; provincial education commissions; and family planning associations (VSO, 2007).
Under Yunnan's government leadership, various civil society organizations serve a range of sometimes overlapping functions related to the fight against the HIV/AIDS epidemic. Most of civil society's large non-profit organizations are the government-affiliated NGOs (or the GONGOs) (Xu, Zeng, & Anderson, 2005). They can be categorized into two groups. First are the mass organizations with a large presence in community life, such as the All China Women's Federation, All China Youth League, Red Cross, All China Federation of Trade, and Chinese Working Committee for Caring for the Younger Generation. Their functions are: mass mobilization, communication, education, and solidarity building. Second are professional groups, operated by medical and public health personnel, such as the Chinese Association of Medicine, Chinese Association of Preventive Medicine, China Family Planning Association, Chinese Foundation for the Prevention of STD and AIDS (CFPSA), Chinese Association of STD/AIDS Prevention and Control (CASAPC), Yunnan Health Education Institute, and Centres of Disease Control. In addition to communication, education, and mobilization, their functions are to provide professional guidance and services for their members and society. These organizations are capable of both vertical and horizontal organization with branches at the national and provincial levels, and many are actually GONGOs. Their operations are to integrate HIV control within the scope of their social, political, and professional activities and use their organizational networks for prevention, intervention, or communication. For example, in the fall of 2010, the All China Women's Federation (2010) facilitated mandatory sex and life education in the curriculum for ‘Education in the Three Sheng’ – shengming (life), shenghuo (living), and shengcun (survival) – provided to students at secondary schools and universities. The Yunnan Family Planning Association also integrated AIDS education into family planning programs (Gang Zhao, interview; Xu et al., 2005).
The other category is mainly the non-GONGO grass roots organizations that serve community functions in Yunnan, such as foundations, international NGOs, and community-based organizations. Their work is conducted mainly by volunteers and members of the affected and infected communities, such as persons living with HIV/AIDS. Examples include such local CBOs as Spring Rain and Yunnan Gay Network-Rainbow Sky, or international NGOs such as Save the Children UK and Médecins Sans Frontières – Holland, Oxfam Hong Kong (Riska, 2010; Xu Hua, Zhao, interviews). In addition to communication, mass mobilization, and network building, they are instrumental in outreach and providing specific services for the infected and affected, vulnerable, and at-risk populations.
In particular, the grassroots organizations support the Yunnan government's HIV-control efforts primarily through networking and serving marginal populations (Shuan, Gang, interviews). In the early 2000s, Yunnan's NGOs played a major role in linking to marginalized and vulnerable populations to help the government develop policy and material for general public health education (Xu et al., 2005; Zhao, interview). Yunnan's NGOs and CBOs have allowed Yunnan to be the first province to engage with IDUs, commercial sex workers, and MSM (Shore, 2009; Wei, Xu, Zhao, interviews). A member of CFPSA (an NGO involved in behavioural investigations of commercial sex workers and long-distance truck drivers) said that in 1995 and 1997, Yunnan's NGOs did ‘the work inconvenient for the government to do’.
Without CBOs, HIV control among MSM would be impossible because they are so stigmatized, and most are married (Gang Zhao, interview; Hua Hsu, interview, Feng Zhao, interview). Spring Rain (a partnership of USAID, China-UK Project, Yunnan Health Education Institute, Yunnan CDC, and Kunming Public Health Department) is widely considered a leader in HIV control. It has reached out to MSM throughout Kunming (Yunnan's capital) by visiting parks, bathhouses, public toilets, Internet chat rooms, and karaoke, go-go, and gay bars (Feng Zhao, interview). Spring Rain staff transformed its headquarters into a nightclub, intermingling live music and dancing with HIV-related quizzes. Since 2003, this outreach effort has contacted 3,000 to 4,000 MSM per year and is likely to double that outreach in three years (Feng Zhao, interview). Note that a network of GONGOs and CBOs are involved in organizing and executing activities. Their support for the government's leadership allows them to develop their skills and capacities, legitimize their existence, and support rights issues in public discourse about HIV control (Gina Gang Zhao, interview; Hua Hsu, interview)
The government-driven, multi-sector partnership model requires NGOs to be on the same track towards one goal, with little redundancy in internal or external division of labour. The Yunnan government limits the number of NGOs devoted to a given cause, allowing few overlaps in the internal division of work. For example, each CBO focuses on a targeted audience but collaborates closely with other CBOs (Pangaea, 2010).
The model allows the government to use a network approach for horizontal and vertical integration, making it possible to provide one-stop-for-all service models and allowing NGOs to share benefits. When an MSM visits Spring Rain headquarters, which doubles as a nightclub and HIV-education centre, he is encouraged to take a free, anonymous HIV test, with results in 30 minutes. If he tests positive for HIV infection, nurses and physicians from the CDC, a GONGO, offer immediate and free voluntary counselling and testing services; Spring Rain's legal partners also offer legal services, such as help in accessing employment or acquiring health insurance. Other NGOs may also share these services. The Yunnan government or the GONGOs are involved at every level of the partnerships.
Overall, Yunnan has been more open than other provinces in receiving foreign assistance and broad-based support. Most HIV/AIDS partnerships are financially supported by the central government, Yunnan government, international organizations, and increasingly, local charities, especially large charities from Hong Kong and Taiwan. The Yunnan government also supports groups such as the Family Planning Association. Membership fees or individuals usually fund private HIV foundations or corporations. Given its open policy, Yunnan has been a major beneficiary of international funding. It has worked with international organizations such as the China-UK Program, Clinton Foundation, Gates Foundation, Futures Group, Save the Children, Oxfam, AusiAid, and Red Cross in harm-reduction programs, including needle-exchange programs for IDUs (Hua, interview). Most external funding is funneled through the central government or Yunnan government, but donors can target areas of HIV control.
Despite some implementation challenges that compromise public health and security goals, Yunnan's policy and regulatory effort is ‘ahead of other provinces’ (Watts, 2006; Hua, Gang, interviews). Watts’ (2006) analysis of Yunnan's HIV-control strategy, and commentaries from local community-based organizations, have credited Yunnan for its record of transparency, strong advocacy, and harm-reduction strategies. Its HIV control strategy for IDUs is a model for China; ‘international specialists agree that China's [Yunnan's] new response far surpasses that of India and Russia … which have even more severe AIDS problems’. UNAIDS Director Peter Piot noted that China's increased AIDS budget during this period was mostly devoted to IDU programs. In addition, compared with other provinces or other resource-poor countries, Yunnan has demonstrated the strongest commitment to HIV control and has applied evidence-based approaches to policy-making processes, although some of their policies are contentious even by Western standards (New York Times, 2005).
Yunnan's recognition that drug use and HIV are so closely correlated indicates the model's effectiveness, with the greatest progress in IDU programs. Others in China have adopted Yunnan's harm-reduction model of methadone maintenance treatment and needle-exchange, with notable success. One positive impact is in reducing drug-related risky behaviours (Li, Ha, Zhang, & Liu, 2010). A study (Li et al., 2010) of 14 methadone management and treatment clinics showed positive changes: safer sex and better social behaviour along with less crime and drug use. A two-year follow-up showed that drug use decreased from 77.9 percent to 52.3 percent; safe sex increased from 31.4 percent to 47.1 percent; arrest rates dropped from 12.8 percent to 4.5 percent; social rehabilitation measured in terms of family relations and employment rates for drug users was also positive (Li, Jia, Min, & Pan, 2008; Ma, Chen, Shu, Wang, Shen, & He, 2001). Overall, data from the US National Institute of Health, in partnership with China's national sentinel surveillance, show slower and stabilized rates of new HIV infection among Yunnan's IDUs. In contrast, China's overall HIV infection rates are increasing (US National Institute of Health, 2010). Thus these data show that Yunnan has effectively reached hidden IDUs; its integrated approach efficiently and simultaneously integrates multiple risks.
Capacity building also strongly indicates the model's long-term effectiveness. Yunnan has surpassed other provinces in terms of technical capacity (Sun et al., 2007). Since the late 1990s, they have expanded surveillance to include both infected and affected communities, such as pregnant women, clients of female sex workers, and TB patients. Medical treatment has added active and comprehensive behavioural surveillance (Jia et al., 2007: 1). Thus, Yunnan has organized and integrated resources to monitor not only the HIV epidemic but also future epidemics (Jia et al., 2007; Sun et al., 2007).
Yunnan has funnelled its HIV resources to strengthen its overall health care capacity. As one of the first to provide comprehensive HIV care, Yunnan led in providing HIV voluntary counselling and testing in 1990, medical treatment in 2002, free treatment for the poor in 2004, free counselling and testing for pregnant women and free medicine to HIV-positive pregnant women (to prevent mother-to-child transmission) in 2004. In 2009, Yunnan was among the first to include universal health HIV care by reimbursing some expenses for HIV-caused secondary infections, such as pneumonia.
Community capacity is measured by the capacity of the civil society. When the epidemic began, Yunnan lacked grassroots NGOs for HIV control; today it has about 140 grassroots NGOs and CBOs (Shieh, 2010). Also relevant is that the Yunnan government plays a central role in directly and indirectly supporting the growth of CBOs and NGOs in HIV partnerships (Gang Zhao, interview).
As the above analysis has demonstrated, Yunnan's government-driven, multi-sector partnership model contrasts with the donor-driven model used in other resource-poor regions, despite similar HIV prevention and intervention tactics. The difference lies in the government's role in HIV control. Specifically, the public sector has played a dominant leadership role in coordinating a top-down, one-track, one-web organizational framework. In this framework, GONGOs/ NGOs are important links to civil society. Despite outside criticism, GONGOs ensure that resources embedded in the civil society are used to achieve the goals of HIV control. Note that drastic changes in the government's role in HIV control came with changes in Chinese government policies.
Although not designed to be a theoretical test, this study is relevant in light of existing theories for future research. In the last decade, researchers have made theoretical attempts to analyze global health partnerships (Reich, 2002). Focusing on pharmaceutical access, Ngoasong (2010) examined several large global health partnerships and argued that constructivism theory best explains state and non-state actors in international development. Similarly, Buse and Harmer (2004) studied power imbalances among stakeholders. Most macro-level research on global health partnerships has been critical but has failed to offer solutions for improving their effectiveness and has failed also to examine the management of dynamic relations and capacity among stakeholders. That is, few have considered how partnerships maximize organizational efficacy to achieve macro-level effectiveness.
Inclusive partnership theory (Wang, 2009), which addresses certain organizational and sustainable effectiveness issues in managing global health partnerships, could offer a conceptual framework to interpret what we found in Yunnan. The theory stipulates that successful global health partnerships use an inclusive approach to increase community capital so that the affected community can internalize or even multiply positive impacts over a long time period (Wang, 2009). Inclusive solutions include ideas such as critical links, power of agency, and community capital. Critical links are the stakeholders that hold the most potential or power of agency to tackle problems: formal and informal institutions, such as governments, businesses, communities, civil society, intergovernmental organizations, and multilateral organizations. Power of agency signifies the resources and knowledge that stakeholders or stakeholder coalitions bring to the partnership. Community capital refers to the intangible, often informal community, assets, such as cultural-related knowledge or informal networks that can be translated into tangible benefits in the development process. It plays an important role in increasing community capacities, especially for those people excluded by mainstream social institutions. Some examples of community capital include the online community among MSM and the transgender in East Asia or Southeast Asia for empowerment and mutual help. When the HIV-infected are abandoned by family and friends, it is often their friends in the network who offer the end-of-life care. The empowerment networks of trafficked or sex workers in Southeast Asia, South Asia, and South America are another example. The other example is the cultural knowledge of indigenous medicine. The global health community has discovered artemisinin as an effective medicine to treat malaria in recent years. In fact, indigenous people in Vietnam and China have been using this medicine to treat malaria and related diseases for more than 6,000 years. This knowledge derived from their cultural community helps them improve their health. Inclusive partnership theory sees the operationalization of these theoretical constructs as key to create sustainable effectiveness in tackling global health and development challenges.
Inclusive partnership theory can be used to discuss the strengths and weaknesses of Yunnan's HIV response. Overall, Yunnan's government-driven model shows effective planning and implementation of goals but weaknesses in the partnership process. Yunnan made maximal use of critical links and power of agency in organizing HIV partnerships. As previous analyses have shown, the most critical link in organizing civil society is the government, with its greater organizational assets/or power of agency that lie in its omnipresent authority and interconnecting political and social organizations. This organizational power was especially important at the outset of the HIV epidemic, when foreign aid was scarce and Yunnan was one of the poorest provinces with a challenging geographic makeup and a large, dispersed rural population. The Yunnan government's forceful involvement in civil society made it possible for a network approach that could build long-term capacity. Although controversial by Western standards, the government-based multi-sector partnership model was beneficial for integrating civil society. This model could address criticisms regarding wasted resources levied by critics of global health partnerships in other countries. Yunnan's model shows overall strength in building technical, policy, and community capacities to address sustainable goals.
The major criticism of the Yunnan model is that the public sector has not fully utilized civil society's potential. Problems related to the partnership process and the division of roles is an organizational challenge. From the perspective of the inclusive partnership theory, one of the challenges facing the public sector led multi-sector partnership is whether government will recognize civil society's substantive power of agency in increasing community capital as the most efficient way to address Yunnan's development challenges. Yunnan's civil society has shown that it has much to offer in fighting HIV as well as for building community.
Limitations and Future Research Directions
This study has limitations in that, first, the unit of analysis is Yunnan; the study was not designed to compare Yunnan with other provinces or other countries. I used information about other provinces or countries simply as a reference. Second, this study's frame of reference omits partnership models in North America or Europe because of the large gap in systemic capacities. Although the study was not designed for theoretical testing, future studies could examine global partnerships theories using other sources, other developmental challenges, or other types of partnerships. Future research could compare the government-driven model vis-à-vis the donor-driven model in several areas. For example, to what extent does the government system affect the efficacy of either model? To what extent do economic resources and capacities determine the outcomes of either model? And in which way does the network culture affect the adoption of a prevention strategy? The Yunnan experience could be compared with experiences in other provinces or other resource-poor countries.