Volunteerism or Labor Exploitation? Harnessing the Volunteer Spirit to Sustain AIDS Treatment Programs in Urban Ethiopia

Abstract

Based on ethnographic research in Addis Ababa, Ethiopia, this paper describes NGO efforts to encourage AIDS care volunteers to eschew material returns for their labor and instead reflect on the goodness of sacrificing to promote the survival of people living with HIV/AIDS. Consensus analysis of motivational survey data collected from a sample of AIDS care volunteers (n=110) suggests that they strongly share a sacrificial and prosocial motivational model. These results may be explained by several factors, including the efforts of the organizations to shape volunteers’ motivations, the self-selection of volunteers, positive reinforcement in seeing one’s patients become healthy, and social desirability bias. In-depth interviews examining the motivations and behaviors of volunteers reveal a more complicated picture: even ostensibly devoted and altruistic volunteers strongly question their service commitments. The complexity and ambivalence of volunteers’ motivations reflect the profound uncertainty that they face in achieving improved socioeconomic status for themselves and their families amid widespread unemployment and sharply rising food prices. Their desires for economic opportunities explain why local NGOs exert so much effort to shape and sustain—and yet fail to completely control—their motivations. This recasts economically-insecure volunteers’ consent to donate their labor as a process of negotiation with their organizers. Future research should explore how models of health care volunteerism and volunteer motivations are shaped by individual and collective experiences in political-economic context.

Introduction

In the past decade, forms of volunteer and unpaid AIDS care and treatment support have been promoted across sub-Saharan Africa with the rollout of highly-active antiretroviral therapies (HAART). Promoting unpaid community-based care is often perceived as economically imperative in combating HIV/AIDS in settings of health professional and resource scarcity, given the political-economic challenges of increasing health work force expenditures in poor countries, particularly in sub-Saharan Africa (Akintola 2008; Campbell et al. 2008; Dräger, Gedik, and Dal Poz 2006; Ooms, Van Damme, and Temmerman 2007). Recruiting, training, and retaining people who are motivated to volunteer their time and labor—many of whom experience profound poverty—have thus become key bottlenecks determining the sustainability of AIDS treatment programs.

This paper focuses on the motivations of low-income AIDS care volunteers in Addis Ababa, the capital city of Ethiopia. Understanding the motivations of volunteers is of practical and theoretical interest for several reasons. For one, many public health initiatives aimed at improving the health, well-being, and survival of community members depend to a large extent on the motivations and commitments of volunteers. Thus, programs depend not only on volunteers’ physical labor but also on their “ethical labor,” involving their own good will and humanitarian values (Feldman 2007). Secondly, by claiming that low-income volunteers freely consent to donate their labor based on their motivations to help others and serve the community, programs may be able to justify the use of low-income people for unpaid community health labor. Further, local low-income volunteers are often said to derive and be motivated by psychosocial benefits (i.e., social respect and “mental satisfaction”) through serving the health interests of needy people (Maes 2010a). Relying on unpaid labor might also be justified by appealing to the humanitarian goal of saving lives and improving health among needy populations through “rolling-out” pharmaceutical technologies, particularly in exceptional or crisis-level contexts such as the HIV/AIDS pandemic in sub-Saharan Africa (Biehl 2007; Fassin and Vasquez 2005). Therefore, research on the motivations of volunteers has the potential to illuminate and encourage debate over how humanitarianism potentially justifies the use of low-income people for unpaid community health labor.

Volunteerism in the United States involving predominantly middle-class volunteers has for decades received much focus from social psychologists and social epidemiologists (e.g., Borgonovi 2008; Omoto and Snyder 2002). In the past decade, a few studies have examined the motivations of volunteers in community health projects in low-income, non-western settings (Akintola 2011; Kironde and Klaasen 2002; Maes 2010b; Ramirez-Valles 2001; Rödlach 2009; Swidler and Watkins 2009). These studies have suggested that volunteers have many motivations ranging from desires for direct remuneration, new knowledge, patron-client relationships, and paid job opportunities, to more pro-social motivations such as reducing suffering and living up to religious models of pleasing God. Studies have not, however, systematically explored variation (or sharing) of motivations among volunteers within particular settings, nor across time. In addition, studies have not examined how volunteers’ motivations are influenced by the efforts of volunteer labor organizers. The result is an overly static understanding of volunteer motivations (cf. Haski-Leventhal and Bargal 2008).

This limitation in the social science literature is underscored by the rhetoric of institutions involved in the promotion and organization of volunteerism in low-income countries. One of the most prominent international NGOs involved in community-based HIV/AIDS care and treatment support in Ethiopia is Family Health International (FHI). In its 2007 Ethiopia report to USAID, FHI (2007:52) boasted that it had trained over 11,000 volunteers for home-based care (HBC) and antiretroviral treatment support for people living with HIV/AIDS, and wrote, “The level of interest and commitment of volunteers to the [HBC] program has been overwhelming…. The program has shown the untapped spirit of volunteerism that exists within Ethiopian communities despite such pervasive poverty.” The same report states that during early phases of the HBC program design, NGO and donor communities were “keen to harness the good will and generosity of [local] communities” (FHI 2007:72).

As this paper aims to show, global organizations like FHI and its local NGO partners do not simply “tap” volunteer spirit as one would a free-flowing natural resource like petroleum or rubber.¹ “Tap” and “harness” are misleading metaphors because they obscure the efforts that these institutions exert in attempts to “drum up” the so-called volunteer spirit, as well as negotiations over the consent to donate one’s time and labor to community health programs, in which both volunteers living amid pervasive poverty and their organizers engage.

Based on ethnographic research in Addis Ababa, this paper describes NGO efforts to coax and sustain the “volunteer spirit”; in other words to encourage volunteers to eschew material or financial returns for their labor and instead reflect on the mental and spiritual benefits of sacrificing to promote the health and survival of others. The paper subsequently presents consensus analyses of motivational survey data collected from a sample of volunteers serving NGOs in Addis Ababa, which suggest that AIDS care volunteers strongly share a sacrificial and pro-social model in serving people living with HIV/AIDS. These survey data, however, mask the complexity of volunteers’ motivations. In-depth interviews with volunteers reveal that despite the motivation-shaping efforts of the NGOs and despite that volunteers often appear self-selected in terms of personalities, previous experiences, and social/familial influences that promote a willingness to serve others and seek mental and spiritual satisfaction in return, even ostensibly devoted volunteers strongly question their service commitments. Behind this motivational complexity and ambivalence lies the poverty and profound uncertainty that participants faced in achieving improved socioeconomic status for themselves and their families amid widespread unemployment and sharply rising food prices (cf. Johnson-Hanks 2005). Volunteers’ uncertainties and desires for economic opportunities, as well as the inequity inherent in being a volunteer for a globally-funded NGO with a hierarchy of salaried staff, explains why local NGOs exert so much effort to shape and sustain—and yet fail to completely control—volunteers’ motivations. This paper, thus, aims to advance understanding of present and future negotiations over working conditions and remuneration between economically-insecure community health workers and the institutions that organize them.

Methods

In Ethiopia, the launch of free HAART programs country-wide in 2004 coincided with the beginning of an upward trend in food prices. In the face of a large and late-maturing HIV/AIDS epidemic (Iliffe 2006), volunteerism in community health care has become common over the past decade in Ethiopia. Addis Ababa, the country’s capital, has a rapidly-growing population (UN-HABITAT 2008), as well as high rates of unemployment (Serneels 2007) and food insecurity (Smith, Alderman, and Aduayom 2006). During 2007 to 2008, when this research was conducted, food price inflation was extremely high (Ulimwengu, Workneh, and Paulos 2009).

The median household per capita income in the sample of volunteers described below was extremely low at only $0.29 per day. On average, the volunteers in this study were 28 years old (range: 18 to 45) and had 10 years of schooling (Maes et al. 2010). Public health facilities dispensing HAART in the capital city relied heavily on the training of such volunteers, who provided home-based palliative care and counseling, supported drug adherence, and mediated patients’ access to clinical treatment and NGO assistance. On average, they cared for 13 patients and spent 15 hours per week in various volunteer tasks and responsibilities (Maes et al. 2010). Though they indirectly served public health facilities and programs, they were organized under NGOs, which expected them to serve for a period of 18 to 24 months. They received about $5–$10 per month to reimburse their transportation and telecommunications expenses. Some volunteers also received wheat and cooking oil as a stipend from their NGO. However, this practice was suspended in early- to mid-2008.

Over the course of 22 months (June through August 2006; May 2007 to December 2008), fieldwork focused on AIDS care volunteers from two local NGOs in Addis Ababa: the Hiwot HIV/AIDS Prevention, Care, and Support Organization and the Medhen Social Center. The Hiwot NGO relied on hundreds of volunteers to run an Addis Ababa-wide AIDS care program in cooperation with several public health facilities. The bulk of its funding, support, and policy guidance during the study period came from Family Health International, which in turn received much of its funding from USAID. The Medhen NGO operated under the auspices of the Ethiopian Catholic Church (but did not consider itself a “faith-based” NGO) and relied on the labor of 20 AIDS care volunteers as well as funding from a mix of international donors. Both the Hiwot and Medhen NGOs provided home-based care for people accessing antiretroviral treatment at ALERT Hospital, located on the southwest outskirts of Addis Ababa.

Participant observation focused on volunteer trainings, meetings, and events, as well as volunteers’ activities in care recipient homes, neighborhoods, and ALERT Hospital. Additional data come from a VHS-recorded initiation and graduation ceremony for a large group of volunteers serving the Hiwot organization, which was provided by the organization’s staff. Thirteen volunteer caregivers (10 female and 3 male) participated in in-depth interviews. This purposive sample aimed to account for the preponderance of women in the volunteer population, as well as variation in length of service, age, education, and socioeconomic status. Each respondent completed a series of five to six semi-structured interviews assessing motivations, costs, and benefits of volunteering, food insecurity, care relationships, and well-being. Interviews occurred over eight months in 2008 and were conducted in Amharic by the author with assistance from respondent-gender-matched local research assistants unaffiliated with the NGOs involved in the study. Interviews were digitally recorded and translated by the author and Amharic-speaking assistants. Texts and fieldnotes were coded in MAXQDA software using a coding scheme combining pre-determined and in vivo codes.

Motivational Surveys and Consensus Analyses

Surveys covering several domains of socioeconomic status, behavior, and well-being were conducted with a larger sample of AIDS care volunteers from the Hiwot and Medhen organizations. This paper focuses on survey data collected from participants’ rankings—according to personal importance—of motivations for doing volunteer home-based care. This ranking task was an attempt to describe volunteer caregiver motivations in terms of their importance and to assess whether consensus in regard to motivations existed among volunteers in the sample. Individual respondents’ rankings were, therefore, subject to motivational consensus analyses.

The assessment of motivational consensus can be thought of as a special case of cultural consensus analysis. The standard cultural consensus model is based on three assumptions. First, within a given domain of knowledge there is a single, culturally correct way to respond (a common truth). Second, individuals respond independently of each other. Third, the ability of each respondent to answer correctly is constant over all questions.² Variation in responses among individuals is modeled as differential ability to give the culturally correct response, which is generally referred to as cultural “competence” (Romney, Weller, and Batchelder 1986; Weller 2007). While cultural consensus analyses generally assess how knowledge is shared by individuals in groups, the present study focuses specifically on how motivations are shared. The distinction between knowledge and motivation is not, however, straightforward. Both knowledge and motivation can be shared or “cultural” (D’Andrade 1992; Strauss 1992). Further, knowledge can be motivating, and cultural consensus analyses often address shared knowledge that is assumed to motivate individual and collective behavior. For instance, when assessing cultural models of a desirable lifestyle, researchers posit that individuals are generally motivated to approximate or “perform” such models in their own lives (Dressler et al. 2007).

In the present study, pilot ethnographic research and literature review identified 10 key motivations that were relevant to being an AIDS care volunteer in the local setting (Table 2). Motivational consensus was assessed precisely because multiple motivations exist for being an AIDS care volunteer and because volunteers were expected to be familiar with these various motivations and to feel that some motivations were personally more important than other motivations. The goal was to let individual volunteers define the relative importance of these motivations and to assess the extent to which volunteers agreed on the relative importance of these motivations. Since the ranking task was repeated at three data collection rounds (February to March 2008, July to August 2008, and November to December 2008) over the course of 11 months, the data also address change over time in volunteer motivations at the group level. A random-order series of 45 paired comparisons was presented to each respondent, accompanied by the instruction to choose: “Which of the two motivations has been more important for you in the past four weeks?” For each participant, rankings were calculated by summing the total number of times (out of nine) that each of the 10 motivations was chosen in a pair comparison.

Table 2.

Answer Keys and Rankings of Motivations for Being an AIDS Care Volunteer, By Round and Organization/Newcomer Status

Paired Item	Hiwot Newcomers			Hiwot Veterans			Medhin Veterans
	Round 1	Round 2	Round 3	Round 1	Round 2	Round 3	Round 1	Round 2	Round 3
	n=40	n=36	n=36	n=50	n=50	n=48	n=20	n=20	n=20
See sick people get healthy	1	1	1	1	1	1	1	1	1
Help reduce stigma and discrimination	2	2	2	3	3	2	3	2	4
Reduce the burden of care in my community	3	3	3	2	2	3	2	3	5
Please God	4	5	4	4	5	5	4	4	3
Receive God’s reward	5	4	5	5	4	4	5	5	2
Get job experience and opportunity	6	6	6	6	6	6	6	6	6
Get respect/appreciation from community	7	8	9	7	7	7	8	9	8
Please my family/parents	8	7	7	8	8	8	7	7	7
Get out of my house from time-to-time	9	9	8	9	9	9	9	8	9
Get income/materials from the NGO	10	10	10	10	10	10	10	10	10

Note: Rank answer keys were generated in UCINET by performing consensus analyses invoking the interval/ordinal analytical model and inputting profile datasets. For easier interpretation, the numbers reported here are the rank-order of the answer key values (1=most important).

Drawn from NGO rosters, the sample included 110 volunteer home-based caregivers (99 women and 11 men) of adult patients receiving treatment at ALERT Hospital, incorporating 40 randomly-chosen participants who had just begun volunteering with the Hiwot organization at the time of the baseline survey (“Hiwot newcomers”); 50 randomly-chosen participants who had all been volunteering with the Hiwot organization for 12 months at the time of the baseline survey (“Hiwot veterans”); and all 20 volunteer caregivers from Medhen, with an average service length of 12 (SD 4.6) months at the time of the baseline survey (“Medhen veterans”). Consensus analyses were performed separately for each sample sub-group (Hiwot newcomers, Hiwot veterans, Medhen veterans) and each of the three data collection rounds. One hundred ten participants were surveyed at Round 1. At Round 2,106 of the original 110 participants were surveyed, and at Round 3,107 of the original 110 were again surveyed. To maximize data quality, data collectors underwent extensive training and worked in pairs.

Consensus analyses were conducted in UCINET v.6 (Borgatti, Everett, and Freeman 1999). Datasets were imported as respondent-by-respondent similarity matrices, which outputs individual competencies and eigenvalue ratios (Romney, Batchelder, and Weller 1987; Weller 2007). Answer keys were generated by performing separate consensus analyses invoking the interval/ordinal analytical model and inputting profile datasets. The mean and standard deviation of respondents’ competencies were calculated, and the stability of answer keys across data collection rounds and sample sub-groups was observed by calculating the between-group and between-round correlation coefficients of rank answer key arrays. The study received ethical clearance from the Addis Ababa University Faculty of Medicine, the Armauer Hansen Research Institute/ALERT Hospital, and the Emory University Institutional Review Board. Free and informed consent of all participants was obtained.

Results

Organization Efforts to Shape Volunteer Motivations

As self-identified Christians (most of Ethiopian Orthodox denomination), volunteers in the study often expressed empathy for their patients, ideas about sacrifice and reciprocity involving humans and God, and the experience of mental or spiritual satisfaction from helping others (Maes et al. 2010). As this section shows, these expressions must be contextualized within their organizations’ efforts to shape their motivations and beliefs.

Recruitment Interviews

One of the first techniques used to organize volunteer work forces is the recruitment interview. The experience narrated by “Alemnesh,” an in-depth interview respondent who began volunteering with the Hiwot NGO at the beginning of 2008, illustrates how the interview served to shape motivations of recruits. At age 26, Alemnesh was unmarried and living with her parents, whom she described as giving and caring role models.

Alemnesh recounted her initial interest to become an AIDS care volunteer as a case of “spiritual envy.” She heard about others doing it and desired to be like them. “I heard on the [state-produced] television and radio about volunteers who do good deeds. When you hear that, you may have menfesawi qïnat (spiritual envy). I thought, ‘What if I do something like them?’“ Alemnesh’s ongoing motivation involved fulfilling her desire to experience mental and spiritual satisfaction. “There was a patient that I had. When she was told that she had HIV, she was crying on the road. But now she accepts it, and she is peaceful. She is changed a lot now. When you see that, you will become happy. That is aïmero ïrkata (mental satisfaction): even if you are not paid, when a fellow human gets well and walks, you say that is a result of your work.” Thus, Alemnesh echoed a very common sentiment among volunteers in the local setting, that mental or spiritual satisfaction comes primarily from seeing one’s “patients” become healthy and productive.³

Alemnesh’s father, an ex-soldier who served during the military Marxist regime (the Derg) that ruled Ethiopia from 1974 to 1991, did not receive a pension. Her mother was the family’s homemaker, while her two siblings held professional jobs in Addis Ababa. Alemnesh did not report household food insecurity, unlike the majority (approximately 80 percent) of volunteers in the survey sample (Maes et al. 2010).

Despite her apparently strong motivation to volunteer, during her recruitment interview, she was met with the suggestion that she was unfit to volunteer because she was accustomed to a better standard of living and remuneration. Alemnesh recounted that the woman who would become her nurse supervisor, Sister “Meheret,” strongly emphasized that there was not a salary for the work that volunteers were expected to do. “I told Sister Meheret that I didn’t have any kind of work. She said to me, ‘So if you don’t have work, if you live with your family, how can you simply serve, without being compensated?’ I answered, ‘I will help my people with all my capacity—just that much.’“ According to Alemnesh, Sister Meheret persisted. At the end of the interview, she again asked, “So without anything being paid to you, how can you work?” Alemnesh raised her voice when she narrated her response: “I myself came bäbägo fäqadäñanät (with good will i.e., voluntarily). I knew that we were not going to get anything. At the time, I was very angry. If you came there to serve with good will, then they have to give you a kind face (mälkam fit)…. But they said, ‘There is no money, there is nothing. The work is heavy and you will get tired.’ …They were frustrating me….”

Thus, one technique of the NGO is to reinforce during the recruitment interview that although volunteering is difficult work, one should not expect material or monetary compensation. Although many recruits may self-select into the volunteer role based on their pre-existing desire or motivation to help others, the organization does not simply aim to identify these individuals. The organization also aims to reinforce volunteers’ commitment to materially-uncompensated service. Below, this suggestion finds support in observations of volunteer recognition ceremonies put on by the NGO. These ceremonies reinforce the sacrificial nature of volunteering and the sentiments about mental and spiritual satisfaction mentioned by Alemnesh. In addition, these ceremonies appear to downplay the potential for feelings of inequity among unpaid volunteers serving globally-funded organizations with hierarchies of salaried staff and officers.

Volunteer Initiation, Oath-Swearing, and Recognition Ceremony

The Medhen NGO operated in neighborhoods directly adjacent to ALERT Hospital, which was originally founded in 1933 as a leprosarium by Emperor Haile Selassie (Terecha 2005). “Eskinder,” a middle-aged man, began volunteering with the Medhen NGO in April 2007. His elderly mother, with whom he still lived at the time of the research, was infected with Hansen’s disease (HD) at a young age; their family migrated to Addis Ababa from their rural home so that Eskinder’s mother could receive treatment. About four years prior to becoming a volunteer, Eskinder found out that he was HIV-positive when he became critically sick and was hospitalized. Thus, he had been close to the social and biological effects of two highly stigmatized diseases for his entire life, and this clearly affected his motivation as a volunteer: “Recently, I came upon an elder [living with HD] with maggots on his body. He had nobody, and he was sleeping in the cemetery. ‘Who am I?’ I said that to myself. ‘From whom was I born?’ Even if my father was a healthy person, my mother was a leprosy case. Because of these things, how can I avoid people with leprosy? I will be close with those people. I will turn them over in bed properly and I will clean them. I will brush away the maggots.”

Eskinder narrated that he and his fellow volunteer caregivers made a promise upon entering the service at the Medhen NGO. “Our [NGO] supervisors made us promise on the Bible when we entered this place. ‘To serve and respect our fellow people…. Not to trouble people; to go even when you are called in the middle of the night.’“ Given Eskinder’s experiences with HD and HIV/AIDS, he appeared exceptionally prepared to make the promises requested by his supervisors.

Volunteers in the Hiwot NGO also promised their commitments to serve as AIDS care volunteers. These promises were actually performed publicly in the context of volunteer initiation and graduation ceremonies. The video-recorded ceremony that was held in 2005 took place in a western-style hall belonging to the Addis Ababa municipality. All of the organization’s volunteers were seated towards the back of the hall; seated up front were higher-status guests and officials from the Hiwot NGO, Family Health International, and the federal government’s HIV/AIDS Prevention and Control Office. A banner hung across the stage and the white t-shirts worn by all the volunteers that day named the event, in Amharic, as the “Home-based support- and care-givers blessing ceremony.” The ceremony unfolded with a series of synthesizer-accompanied traditional dances, songs, and speeches by NGO and government representatives.

The volunteers’ public promise occurred when one of several supervisor nurses employed by the Hiwot NGO took to the stage and instructed the volunteer initiates to hold up the candles that had been given to each of them. After matches were passed around to light the candles, all then stood up, placed their hands over their hearts, and in unison chanted an oath. They promised, in God’s name, to give proper care, putting the patients before themselves, to keep secret the HIV status of their patients, and to help prevent the spread of HIV.

Subsequently, the NGO’s home-based care program coordinator took to the stage and called the volunteer graduates by name, who each received a certificate from a federal government health official. The government official then made his speech, in which he acknowledged the difficult sacrifices of volunteers and asserted that they become satisfied when they see their patients’ health improve. He reminded the volunteer initiates to keep their promise and vowed that he and his fellow officials were there to help and encourage them.

The second annual Ethiopian Volunteers Day, which took place in May 2008, closely resembled the initiation ceremony as a technique aimed at promoting volunteerism, shaping volunteers’ motivations, and reinforcing their commitments. The event was organized primarily by the Hiwot NGO with funding from Family Health International. The two hour long celebration was located in Volunteer Square, which had been dedicated by the Ethiopian government in 2007. Of note, Volunteer Square is actually a small, nondescript traffic circle located on the southwest outskirts of Addis Ababa, not far from the city’s landfill.

At the event, roughly two dozen NGO officers and guest speakers were seated in chairs under a temporary tent. Roughly 200 volunteers were also packed into the roundabout, standing or sitting on the ground. Nearly all the volunteers were wearing white t-shirts and paper sun-visors that read (in Amharic), “Everyone should give volunteer service in order to improve the country!” and “Let us protect children from HIV/AIDS and spread volunteer service.” This event also unfolded with several synthesizer-accompanied dances and songs and a series of speeches delivered by organization elites. The volunteers responded to the speeches and performers, at times collectively ululating, applauding, and waving their visors.

An officer from FHI delivered a speech that uniquely illustrated the show of cooperation and appreciation through which the unequal statuses of volunteers and officers were tacitly managed within these motivation-shaping rituals and ceremonies (cf. Collins 2004; Goffman 1967):

When we [i.e., officers] go to see what [the volunteers] are doing, we see that they walk through the back alleys—even if there is rain, mud, or bad smells…. And yet they are happy with what they are doing. We [officers] want to look after ourselves when it is raining… But these [volunteers] are giving their time for the patients instead of taking care of themselves…. I would be pleased if all of us could do their job—I believe that it would be an important opportunity to learn a lesson…. For the future, I wish you the strength and the interest to continue your good work for your country and society. God will help you. Thank you.

The FHI officer’s speech may express her genuine desire to experience communion and solidarity with the volunteers and their patients. It can also be interpreted as reinforcing an illusion of egalitarianism between well-paid organization elites and unpaid volunteers (cf. Fassin and Vasquez 2005).

Later in the course of the event, the director of the Hiwot NGO delivered her address, wearing the same t-shirt and visor as the volunteers. Bestowing praise on her large cadre of hard-working volunteers, she compared them to candles, giving light and hope to their patients while melting away in self-sacrifice. This recalls the oath performed in the volunteer initiation ceremony described above. In an interview that took place two weeks after Ethiopian Volunteers Day, when a volunteer was asked whether her service was like a “sacrifice,” she agreed, recalled the use of this metaphor at the event, and said that the metaphor strongly affected her.

In sum, several techniques were used by the NGOs and their partners not to simply “tap” the volunteer spirit among materially-impoverished community members, but also to encourage and sustain it. Motivational consensus analyses summarized below suggest that these techniques were quite effective. Later in the paper, these techniques are explained in light of the political-economic context of sustaining donor-funded AIDS care programs in sub-Saharan Africa.

Motivational Consensus and Rankings

A very high degree of consensus and stability over time were observed among participants’ motivational rankings (Table 1). All eigenvalue ratios well exceeded the conventional cutoff of 3 (average: 17.4). There were no negative competency scores observed, and mean competencies were very high (around 0.9). Finally, rankings (answer keys) were largely consistent across data collection rounds and the sample sub-groups (Table 2): both inter-round and inter-group correlations of rank arrays ranged from 0.97 to 1.0. Importantly, items that represented “pro-social” motivations were consistently ranked as most important (Table 2). In contrast, “Get income and materials from the NGO” was the lowest-ranked out of all items. This motivation was particularly contentious, as illustrated by Alemnesh’s recruitment interview.

Table 1.

Cultural Consensus Summary Results

	Hiwot Newcomers			Hiwot Veterans			Medhin Veterans
	Round 1	Round 2	Round 3	Round 1	Round 2	Round 3	Round 1	Round 2	Round 3
	n=40	n=36	n=36	n=50	n=50	n=48	n=20	n=20	n=20
Eigenvalue Ratio (Largest to Next)	21.8	37.8	17.1	12.4	14.0	12.9	9.0	18.9	13.1
Number of Negative Competencies	0	0	0	0	0	0	0	0	0
Mean Competency (SD in Parentheses)	0.9 (.06)	0.9 (.05)	0.9 (.07)	0.9 (.10)	0.9 (.09)	0.9 (.10)	0.9 (.14)	0.9 (.14)	0.9 (.06)

There may be several non-mutually exclusive reasons for this high level of consensus. The first is the efforts exerted by the organizations to shape the motivations of volunteers through discourses promoting sacrifice, pro-sociality, and spirituality. However, motivation-shaping discourses and rituals do not appear to explain the very high level of consensus observed among newcomer volunteers from the Hiwot NGO at the first round of data collection. And yet, because the first round of data collection occurred soon after they had gone through recruitment interviews and trainings, it is still possible that the high degree of consensus among the newcomer sample partly results from shared volunteer-socialization experiences that shape motivation. Second, it is possible that individuals sharing similar pro-social values and motivations prior to the experience of becoming a volunteer were selected into the volunteer role. And third, the experience of seeing one’s patients become healthy may strongly reinforce volunteers’ pro-social motivations. All of these reasons find support in the qualitative data reported above.

Another potentially important factor explaining the results of the consensus analyses is social desirability bias. Social desirability bias refers to the tendency to avoid reporting behaviors, beliefs, and motivations that project an unfavorable (i.e., socially undesirable) self-image, which theoretically arises from an individual’s need for social approval based on conformity to perceived cultural norms (Johnson and van de Vijver 2003). Responses to the motivational ranking task may reflect not simply the personal preferences of respondents, but also partly what they know they ought to prefer as participants in a ritually- and discursively-reinforced ethical system, and perhaps what they think the higher-status data collectors and western researcher prefer (cf. Weller 2007). Thus, in conducting consensus analyses on data gathered by asking respondents about their personal motivations, as in the present study, a “culturally correct” response may also be a “socially desirable” response. What appears in this case to be differential ability to give a culturally correct response may, therefore, be determined in part by relative tendency to avoid reporting behaviors and motivations that project a socially undesirable self-image. And if individuals avoid giving socially undesirable responses, this will diminish variation in responses and strengthen the appearance of consensus. A limitation of this study is that it did not attempt to test or quantify the effect of social desirability bias and other factors on respondents’ motivational rankings, and this is a clear avenue for future research that seeks to understand how motivational consensus is produced and eroded through group formation and social interaction in political-economic context.

The following section suggests another layer of complexity—and ambivalence—in volunteers’ motivations, which is obscured by these survey data and consensus analyses. In-depth interviews with Alemnesh and Eskinder reveal that even these self-selected and highly-admired volunteers nevertheless strongly questioned their service commitments because of their desires for socioeconomic “improvement” through paid employment.

Volunteering and Desiring Socioeconomic “Progress”

Eskinder

With his mother, Eskinder lived in a two-room shack located in the slum next to ALERT Hospital. In surveys, he reported moderate to severe household food insecurity along with high levels of mental distress (see Maes et al. 2010). He also seemed to possess a strong volunteer spirit: he was very compassionate, hard-working, and admired by many of his peers and neighbors. However, he lamented that his service as an AIDS care volunteer had not led to gainful employment and had in fact become less worthwhile than he had expected at the outset, due largely to the local effects of the global food crisis, which included sharp increases in food prices and sharp reduction in the stipends received by Eskinder and his fellow volunteers (Maes et al. 2011).

We have a problem with the organization—especially after our food stipend was stopped. Most of us volunteers were managing our household needs with that food stipend. We considered that as food…[and] we considered that as a salary. But now, we are given a monthly stipend of only 100 Birr ($10). What can I do with that? This is troubling my mind. It is not only me, but many of us….

When the stipend was changed, everybody complained. We said, ‘We will not [volunteer] if it is like that. Why do we work? Can a person work without eating?’

This provoked an obvious question, which Eskinder also addressed:

Why do I continue volunteering? It is a promise. We respected our word, and the [NGO] supervisors should respect theirs [i.e., to support the volunteers]…. They are reducing things; but we are not reducing our love.

Yes, we will continue volunteering. But how can we live with this life condition? The [NGO] supervisors know how much a sack of wheat costs. Nowadays, it is only that life is expensive and there is no employment.

Eskinder clearly saw his volunteer service as a potential job and wished that he received better remuneration. He was upset by the loss of the food stipend that he had been relying on. The NGO cut off food stipends for volunteers not due to a lack of sympathy, but because its donors could no longer afford the food packages as global food prices soared. Nevertheless, from Eskinder’s point of view, his NGO supervisors had not kept their promise to materially support the chronically food-insecure volunteers.

Alemnesh

Alemnesh proved wrong her recruiter’s presumption that she was unfit for volunteer service: she became recognized by her peers and supervisors as an outstanding and reliable volunteer. However, after learning that her cousin in Dubai had arranged a job and visa for her there, Alemnesh dropped out of the volunteer program in November 2008 (10 months prior to completing the 18-month commitment expected). Alemnesh confidently rationalized her decision by emphasizing that the money she could make abroad would help support her family and hopefully lead to a step up the socioeconomic scale for herself:

Let alone 150 [US] dollars, there is no job [in Addis Ababa] that will pay you 50 dollars [per month]! … I joined this work [volunteering]. But if it is God’s will that I get some other opportunity, I will not hold myself back. You have to do something for yourself, too…. I served eight months, and I am very, very close with the patients, [especially] with the children. And it is difficult [to leave…]—but it is life.

In Alemnesh’s final interview, she described her understanding of a “good life” as “when you have money with love and peace.” In other words, a good life for Alemnesh would combine “mental satisfaction” with economic security and progress.

Discussion

By combining cultural consensus analyses of motivational survey data, participant observation, and in-depth interviews, this study produces a more dynamic, contextualized picture of the motivations of AIDS care volunteers who live and serve on the periphery of Ethiopia’s capital city, Addis Ababa. Among 110 volunteers surveyed at three data collection rounds over the course of 2008, very high consensus was observed in rankings (in terms of personal importance) of various motivations for being an AIDS care volunteer. The shared motivational model designated pro-social and religious/spiritual motivations as highly important and self-interested motivations as less important. The motivation-shaping efforts of the NGOs, self-selection of people with pro-social and empathic tendencies into the volunteer role, as well as social desirability bias all likely explain the motivational consensus model observed. Future study of how motivational consensus is produced by ritual, social desirability bias, and selection into populations of interest (among other factors and processes) may combine ethnography with surveys that make use of culturally-appropriate tools for the assessment of variation in participants’ tendencies to report socially desirable information.

In the present study, consensus analyses of survey data did not capture the ambivalence of volunteers’ motivations. In-depth interviews revealed this complexity, in which even ostensibly devoted and respected volunteers strongly questioned their service commitments. This was due in part to the uncertainty that they faced in achieving basic food security and improved socioeconomic status for themselves and their families amid widespread unemployment and rising food prices.

Feelings of inequity also emerged during in-depth interviews. For instance, Eskinder suggested that he and his fellow volunteers respected their commitment while their NGO supervisors, by reducing the amount of food aid given to the volunteers, had not respected their commitment to support the generally food-insecure volunteers. For Eskinder, this was unfair. Another revealing expression of inequity was reported by an in-depth interview respondent from the Hiwot NGO a few days after the Ethiopian Volunteers Day event. She claimed that many of the volunteers from her district were complaining among themselves during the celebration because they had heard that volunteers in other districts were getting more substantial food aid packages as remuneration for their services. Though inter-district inequality in remuneration could not be verified, the existence of such gossip during Ethiopian Volunteers Day underscores the motivational ambivalence and inequity experienced by unpaid volunteers facing widespread food insecurity.

AIDS care volunteers—people who uniquely link HAART patients to the network of NGO and public health services—emerge from this analysis as both subjects with motives to be “tapped” and reinforced and pro-social citizens with desires for improved economic security. The label of “volunteer” obscures this complexity. That volunteers strongly desired opportunities to become economically secure partly explains why NGOs exerted so much effort to shape and sustain—and yet failed to completely control—volunteers’ motivations. To fully understand the efforts of NGOs, it is necessary to examine the global-level political economy of AIDS care and treatment in sub-Saharan Africa.

Political and Moral Economies of Volunteerism in AIDS Care

The efforts of NGOs to both tap and sustain the so-called volunteer spirit reflect the political- and moral-economic context of donor-funded, community-based AIDS care. In global health, a project—or a part of a project such as health worker salaries—is often said to be sustainable if local organizations and governments can be expected to keep it afloat after global donors and organizations pull out their money and resources. This conception of sustainability is influenced by the macroeconomic concerns of the International Monetary Fund (IMF). When the IMF discourages governments from raising public health payroll expenditures, and simultaneously discourages donors from funding payroll expenditures, paying labor is thought to be financially unsustainable (Dräger, Gedik, and Dal Poz 2006; Ooms, Van Damme, and Temmerman 2007; Swidler and Watkins 2009).

The World Health Organization (WHO 2002), breaking with previous rhetoric that promoted the use of volunteer labor, recently asserted that relying on volunteerism is not a sustainable response to the threat of health worker shortages. Specifically, the WHO (2008) recommended that community health workers be paid fair and predictable wages, in part to ensure their capabilities and commitments and in part to provide for secure livelihoods and well-being, implying that the use of unpaid and underpaid health workers is both unethical and unsustainable. However, ethical questions on the use of low-income community health volunteers are not straightforward, and thus provide space for organizations to at least attempt to use techniques other than fair and predictable wages to motivate volunteers and manage the uncertain sustainability of using volunteers.

One ethical question is whether volunteers are unfairly used as mere means to an end. But one can ask: what kind of end? Volunteers in AIDS treatment and care projects are being used to support the rollout of life-saving medicines to destitute people. Do the humanitarian ends justify the means of not paying local people for their valuable labor? Adriana Petryna asks a similar question of global clinical trials that use a different kind of volunteer, namely human subjects recruited from vulnerable populations in order to bring new drugs to market (Petryna 2009). With the rollout of HAART in low-income countries, however, a model of pharmaceutical humanitarianism has largely bypassed the underlying poverty, inequalities, and health system failures that have threatened attempts to effectively control the AIDS pandemic (Biehl 2007). Advocates of antiretroviral therapy rollout in low-income countries envisioned drug access as a “wedge issue” to usher in broad-based strengthening of public health systems and to protect human rights to health and economic security, which were threatened by structural adjustment programs in the 1980s and 1990s (Irwin and Scali 2007; Kim and Farmer 2006; Ooms et al. 2008; Pfeiffer et al. 2008). João Biehl (2007) describes the capacity of pharmaceutical companies during the past two decades to “neutralize and redirect” the global AIDS treatment movement, while at the same time influencing the definition of global intellectual property rights. Many patients thus have access to sophisticated drugs but not to basics like food and employment (Kalofonos 2010). AIDS care volunteers are intimate witnesses and mediators of this paradox, and thus may question whether their services have much impact on the greater well-being of their patients. As this paper has shown, desires for personal and familial economic security also lead volunteers to question their commitments to service. The organizations that rely on volunteers deal with this motivational ambivalence in part by reinforcing the effectiveness of their sacrificial labor in bringing about its intended humanitarian ends.

A related ethical question is whether volunteers are used as a resource with little or no consideration of their well-being. Here, it is relevant that volunteers are often said to derive mental or spiritual satisfaction through their service, a psychosocial “bonus” that may help to justify the use of volunteers. However, this is a contestable and largely untested assumption, particularly in contexts where poverty is highly prevalent (cf. Glenton et al. 2010; Maes, Kohrt, and Closser 2010). And yet the imagined link between volunteering and well-being resonates with various spiritual traditions. Though the idea that volunteers enjoy spiritual and mental benefits may be a very convenient assumption for volunteer organizers, it may also be a genuine belief, one that motivates volunteers and volunteer organizers and that appears to justify the use of economically-insecure volunteers.

Conclusion

This paper encourages a re-politicization of HIV/AIDS care and treatment efforts that rely on unpaid labor in settings characterized by inequality and widespread food insecurity. The loaded label of “volunteer” obscures the complexity of volunteers’ motivations and their political-economic context. Efforts of NGOs to both “tap” and sustain the so-called volunteer spirit must be recognized and situated within the political- and moral-economic context of donor-funded, community-based AIDS care, in which local NGOs’ reliance on volunteer labor reflects the global-level promotion and justification of volunteerism as necessary in order to sustainably save lives through distribution of HAART. Volunteers address various health care needs and other social service gaps throughout the world. Future research should explore how models of health care volunteerism and volunteer motivations are shaped by individual and group experiences in political-economic context. Understanding present and future negotiations between economically-insecure community health workers and the institutions that organize them may lead to policies and practices that simultaneously promote effective and equitable health care as well as livelihood security for so-called volunteers.