Conceptual Frameworks

The Sustainable Livelihoods Framework (SLF) guided data collection, analysis, and interpretation (Figure 1). [20] The SLF is a people-centered, holistic, asset-based framework for understanding poverty and the livelihoods of the poor. At the core of the livelihoods framework is the asset pentagon, which describes the relationship between people’s access to five types of capital: human, social, natural, physical, and financial. These assets combine to produce positive livelihood outcomes. The asset pentagon lies within the vulnerability context, which is the external environment over which people have limited or no control (e.g., prices, rainfall). The transforming structures and processes are the institutions, policies, and legislation that shape livelihoods. Together, the combination of these factors results in livelihoods strategies that people adopt to achieve their livelihood goals.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. The Sustainable Livelihoods Framework adapted for women farmers in Iringa, Tanzania.

Key: H = human capital; N = natural capital; F = financial capital; S = social capital; P = physical capital; NR = natural resources.

https://doi.org/10.1371/journal.pone.0059343.g001

Although not the main focus of our study, the high prevalence of HIV/AIDS in Tanzania (5.6% in 2009 [21]) warranted consideration of the gender inequities, social norms, and behaviors that heighten vulnerability to HIV infection and transmission in the general population. For these elements of the study, we were guided by the proximate-determinants conceptual framework for HIV/AIDS, which posits that the sociocultural and economic context influences the proximate behavioral determinants of HIV exposure, such as new partner acquisition and condom use, which directly affect the likelihood of HIV transmission. [22] We also referenced an adaptation of the SLF which shows how HIV/AIDS affects, and is affected by, livelihoods. [23].

Study Setting and Population

Our study was conducted in the Iringa Region of Tanzania in July-September 2011. Iringa is an agriculturally high potential area and is known as one of the “Big Four” regions in terms of staple food production. Maize is the most common crop in addition to potatoes, rice, beans, cassava, tea, tomatoes and sunflower seeds. Sixty-nine percent of the adult female labor force in Iringa works in agriculture, [19] and Iringa has the highest HIV prevalence in the country, with 16.8% of adult women found to be HIV positive in a 2008 survey. [25].

FGDs participants met the following inclusion criteria: 1) head of household or married to the head of household; 2) agriculture as the main income generating activity (primarily subsistence farming); 3) already planted or planning to plant crops in the current season in the Iringa Region; and 4) at least 18 years of age or at least 15 years of age in the case of female heads of household. In-depth interviews with agricultural extension officers and village agro-dealers required that participants serve the selected village in either of these roles and were at least 18 years of age.

Recruitment

Two of the five districts in Iringa Region where rollout of the subsidy program was underway were randomly selected (Mufindi and Kilolo), and within each district, we randomly selected four villages as study sites. A letter of introduction was sent from the Ministry of Agriculture, Food Security and Cooperatives to the District Agricultural and Livestock Development Officer in Kilolo and Mufindi, and the research team met with the Regional Medical Officer and health officers in both districts at study initiation. At each village, the three-person local research team (female facilitator, female note-taker, and male interviewer) first met with the Village Executive Officer to discuss the study, identify possible FGD participants from village rosters, and for introductions to the village’s agro-dealer and extension officer. Most FGD participants were purposefully recruited from existing women’s groups (e.g., HIV/AIDS support groups) or through participant referral. Potential participants were informed verbally about the purpose and nature of the study and if interested, they were invited to the discussion, which was typically held the next day.

Data Collection

Most data collection activities were conducted in Kiswahili by three local research staff trained in qualitative methods (in some cases, local languages were used with the assistance of an interpreter). A semi-structured interview guide approach was utilized whereby pre-determined issues are covered (to ensure systematic data collection) but the interviewer is free to change the sequence and the wording of questions during the course of the discussion or interview. [26] Nine of twelve FGDs and 19 of 20 individual interviews were digitally recorded with consent from participants (two FGDs and one individual interview were not recorded at participant request and one FGD was not recorded due to recorder malfunction). In these cases, the moderator and note-taker’s detailed notes were used to reconstruct overall themes and specific quotes.

FGDs: FGDs had 8–9 participants and were conducted by the female facilitator and female note-taker. [26] After initial introductions, the facilitator explained the study and obtained written, informed consent (or a mark/thumbprint indicating consent). FGDs lasted approximately 1–2 hours and typically took place in a private space (e.g., a village meeting room or outside). Either before or after the discussion, participants completed a brief survey on basic demographic characteristics. Participants were compensated with a kanga (a piece of colorful printed fabric worn by women, approximate value USD$4) or cell phone airtime.

Given issues of confidentiality and the sensitivity of the subject matter, FGD participants were encouraged to discuss group- rather than individual-level experiences and perceptions. The discussion guide, developed from the conceptual frameworks, was comprised of three main types of questions. Main questions addressed the research themes, and were broad in nature (e.g., “Are there special challenges that women farmers face?”). Follow-up questions and probes encouraged deeper, more detailed information (e.g., “Do women heads of household get to make the final decisions about land, such as what crops to grow, and whether to sell the land? What about married women?”) [26]. There were 8 to 9 main questions in each FGD, each of which had 2 to 4 open-ended follow-up questions or probes. Although the FGD guide included a broad spectrum of issues, the data presented here are from findings related to three main concepts: 1) women’s empowerment and/or inequity in agriculture, 2) food security, and 3) sexual and reproductive health, including factors that heighten the risk of HIV/AIDS. Throughout each FGD, the note taker recorded common responses, any new or unexpected ideas, group agreement/disagreement, outlying responses, and non-verbal cues. At the completion of each discussion, interview summaries were created by the research team to record first impressions and suggested improvements for future discussions.

In-depth interviews: Individual interviews were one-on-one interviews conducted by a male interviewer with agricultural extension officers and village agro-dealers. Using a similar interview guide approach, the interviews focused on regional agricultural practices and women’s empowerment and/or inequity in agriculture. A special emphasis was placed, within each of these topics, on the participant’s interaction with women. For example, extension officers were asked if women sought him/her for advice and whether special efforts are made to reach women farmers.

Analysis

Audio files were translated and transcribed verbatim into English by two of the local research staff. Two members of the study team (SM & LR) then independently read and coded a subset of six transcripts to begin identifying and coding organizational and substantive themes within the data. Organizational themes were defined as those that responded directly to the main, a priori research questions. Substantive themes were defined as those that were not anticipated, either in their content or specificity, by the study team based on our conceptual framework or discussion guide, but emerged from the data. [24] For example, although we anticipated that women might engage in side businesses or other income-generating activities, the universal nature of alcohol production as women’s primary income generating activity was not anticipated and resulted in a new substantive theme (described later). After double-coding of the initial six transcripts to establish a preliminary list of organizational and substantive themes, the remaining transcripts were then divided between the two coders and reviewed for further elaboration and clarification of the themes. The two coders met weekly throughout the coding process to resolve coding or interpretation differences through discussion and debate. When differences emerged that could not be resolved, when additional context was needed, or when the two coders felt that additional discussion was needed on a particular theme or concept, a member of the field team (WW) was brought into the discussion. Once consensus was reached on a final list of themes, transcripts were revisited for additional dimensions, including whether the responses represented individual experiences or group-level norms, the degree to which the ideas represented factual or perceived experiences, and the extent to which themes were salient only for certain subgroups of participants. [26] We compared our findings back to the conceptual frameworks to determine how the data inform, substantiate, or deviate from previous models.

Human Subjects Protection

All participants in the study provided written, informed consent (or a mark/thumbprint indicating consent). This study and consent procedures were approved by the National Institute for Medical Research in Tanzania and the University of California, Berkeley Committee for Protection of Human Subjects.

Women’s Access to and use of Livelihood Assets

Participants in both FGDs and individual interviews confirmed the integral role of women in agricultural production and ensuring household food security. When asked to describe what it is like to be a female farmer, nearly all respondents described farm labor, particularly for subsistence crops, exclusively as a women’s responsibility. This “productive” responsibility came in concert with myriad other family and social responsibilities, [1] including child caregiving, food preparation, water retrieval, household maintenance, caring for sick family members, and social interactions to build community.

Participants described women as having access to a diverse array of livelihood assets, including the region’s natural resources (e.g., favorable soil and weather conditions), their own skills, knowledge, and abilities (i.e., human capital), and social resources (e.g., groups) to pursue their livelihood objectives. Women were described by men (in both FGDs and individual interviews) as hard workers, “eager to learn”, and highly capable of running a farm. A male extension officer noted that “educating 1 woman is like educating 100 or 200 men.” Access to groups was cited as an asset, enabling women to share knowledge and labor (as a means to respond to illness), provide social support and encouragement, and access financing (when used as rotating savings and credit associations (ROSCAs)).

However, despite their central position in agricultural production, participants universally described women’s striking lack of autonomy and decision-making authority over land use and earned income, combined with a lack of secure land rights, when asked to discuss special challenges faced by women farmers. Across focus groups and interviews, participants agreed that men controlled earned income from agriculture, regardless of the extent of women’s participation in agricultural production or her role in the household. These narratives largely described group-level experiences, although the few individual experiences described did not deviate from these shared general perceptions:

“A husband can allow you to use his farm but later on he snatches it from you…with your crops.” – Widow

“[Women’s] life is real hard, they take all household responsibilities and productive activities…women have no say in the income though they are the main producers in the household.” – Male agro-dealer

Responses varied when participants were asked to describe the extent to which women (particularly widows) could own land and the context of ownership. The most commonly expressed sentiment, particularly by female focus group participants, is captured in the following female participant’s response to a question about who controls the farms after a husband’s death: “They are taken by the relatives and the children, for his wife it is gone forever.” In slight contrast, interview and male focus group participants acknowledged that women could own and inherit land, but that land division was often inequitable and justified as a way to prevent “developing” families outside of the clan. A male FGD participant noted that “for the inherited farms, the land is divided with equal rights but the woman will get a small farm, a man will get more…they consider that a women is married therefore can’t take the farm to the kids of another clan.” One married woman highlighted the precarious situation this places women in who wish to marry: “For some families, they forbid you to get married unless you leave everything that belonged to your late husband.” However, a handful of FGD participants agreed that despite being inequitable in the past, women’s ability to own or manage small parcels of land following the death of her husband was improving in recent years. One widow explained that the “behavior of relatives snatching everything from the widow is now diminishing.” These varied responses may reflect differences in customary law between villages or contradictions between statutory and customary law. [27].

In addition to insecure land rights, women were described as having poor access to agricultural education and both agro-dealers and extension officers noted that women were often lacking basic entrepreneurial skills, such as how to properly use inputs or compute profits. In addition, women’s poor access to financial and physical capital results in a reliance on hand tools, including hand hoes and handcarts, which is particularly challenging given that farms are often located far from the household.

Especially when combined with their other responsibilities, it was apparent that any shocks or disruptions could have a tremendous cost for women and their families. For example, with regard to the impact of pregnancy on agricultural productivity, one agro-dealer explained, “there is no family planning thus women lose their working capacity when still young.” A woman living with HIV infection explained the labor impact of clinical care and antiretroviral therapy: “Someone might go to town and stop all that she had to do that day, for the whole day she is in the town taking medication.” Thus, sexual and reproductive health demands were described as competing with income-generating activities.

Women’s Adaptive Livelihood Strategies

In all but two of the female FGDs, participants noted that women frequently adopt other income-generating activities (e.g., owning a small kiosk or store, selling snacks) on top of farming responsibilities, recognizing that they cannot rely on their husbands’ earnings for important household expenses. A widow emphasized that “what you care is that my family can get the food and does not suffer and that's it. You cannot count on his money.” For some subgroups (e.g., widows or women living with HIV infection), side businesses were viewed as essential. A widow explained, “she [HIV positive woman] has to get an education and a loan to enable her to do small scale businesses to help her for there is no other alternative.” However, the extent of women’s engagement in other income-generating activities was often constrained by the time required for other household responsibilities.

In two-thirds of FGDs, participants reluctantly acknowledged that some women in their communities exchanged sex to procure money, food, or other goods to alleviate their economic insecurity. One widow explained how this might happen, and the potential for unintended consequences: “A child is starving … someone gives me five thousand, I go with him in order to get some needs of the child, but later the calamity starts.” Another unmarried woman used an analogy of agriculture to describe transactional sex: “that’s her hoe, what else can she use?” Some women hesitated to directly attribute transactional sex to economic and food insecurity; rather, they connected economic insecurity to joining a family headed by a man with a good income:

“Few women have sex for money or food. For most, they better get married [as] a second wife to the man who has big farms and animals, than doing sex for food.” – Widow

This nuanced viewpoint highlights the subtle difference between transactional sex to obtain essential items versus engaging in a relationship with the potential for longer-term economic stability.

Brewing wine or beer (typically ulanzi, fermented bamboo sap, or komoni, a maize-based local beer), or selling maize to others who would use it to brew alcohol, was by far the most common income-generating activity reported by women. Unlike other potential businesses, women explained that little or no capital was required as an up-front investment and it could be done using crops (e.g., maize) they are already growing. A female FGD participant noted, “if you don’t have any other businesses to do then you will only make local brew.” This income was described as largely under women’s control and typically used to support the household, including school, food and medical costs and informal payment to farm laborers. As one female focus group participant explained: “Many men nowadays allow their wives to have their private projects but you have to take care of the family.” This highlights the assumption that income generated from women’s adaptive livelihood strategies be used to support the household; the same is not true for money generated by men’s activities. A male extension officer and a married woman both echoed this sentiment:

“Women do involve in preparing and selling alcohol so that they can get income to run their household, taking children to school, hospital, find food, etcetera, as their husband does not contribute any to the households and they will always take money from their wives.”

“The hardship comes in taking the kids to school, I pay for two children in secondary school, I also have three in primary school, one in nursery school, if I don’t sell local brew and do farming I cannot keep up my family.” – Married woman

In this way, making local brew or selling crops for brewing filled a critical gap to meet a household’s food, medical, and educational needs.

A Convergence of Risk: Risk Environments Enhanced by Adaptive Livelihood Strategies

In response to a lack of control over their earnings, weak bargaining power, and heavy burden of domestic tasks, participants described how women adopted livelihood strategies that directly or indirectly enhanced their risk for undesirable health outcomes. Exchanging sex for money or food clearly increased women’s risk of sexually transmitted infections, including HIV, although this was generally described as rare. However, focus groups revealed a direct connection between women’s most common income generating activity and their risk for adverse health consequences. As described above, brewing alcohol was a logical strategy that women adopted in order to earn extra income given their time and capital constraints. FGD participants described a high demand for alcohol in their communities, and women’s role in brewing alcohol often brought them into bars and other social settings where alcohol consumption was common. As one participant described “The main activity after farming, house work, and picking of vegetables is selling the local brew in the bars and chatting with friends.” In one community, alcohol consumption was pervasive enough that a curfew was established for women at 20∶00 so that they would leave the bars and return home (no curfew existed for men).

Alcohol consumption was also described as contributing to the economic instability of families, and related to women’s lack of control over earned income. For example, one widow explained, “He uses every cent you earn from your farms to drink.” An agro-dealer noted, “women and their husband[s] have different farms, but income from [the] husband’s farm are just for alcohol drinking.” Further, in a majority of FGDs, participants named alcohol consumption as the primary driver of risky sexual behavior in their communities for both men and women:

“The spreading of HIV/AIDS is very high in this area because many people are engaging carelessly in sexual intercourse. This situation is caused by drinking local beer, for when a person been drunk he becomes very careless.” – Widow

“One day you can decide to go and mix yourself in a local club, you meet your friends… Like women, they do go there without any money but on arrival men buy them a drink…they find themselves drinking and lately their senses shifts. They advise each other and run together and that’s it.” – Male FGD participant

“Local brew drinking disturbs a lot, especially in February, March and April. These are the months for Ulanzi. You take water [local brew] till you are out of your senses, you get HIV, then you are already gone.” – Male FGD participant

“When you are drinking together and a man orders a seller to give a litre to one woman among you, the rest have to finish the litre and quickly leave. They should leave a space for a man and a woman as she will surely pay for that litre.” – Widow

Together, the narratives delineate the opposing effects of alcohol on women’s economic security and health. On one hand, alcohol production is an important income-generating activity for women over which they maintain control and use to invest in the household, including children’s education. Conversely, alcohol consumption was associated with higher-risk sexual behavior that could potentially have deleterious consequences on their sexual and reproductive health.

Conclusion

The agricultural sector could play a significant role in poverty reduction in developing countries and is critical for achieving the Millennium Development Goals. [63] The value of women’s empowerment in agriculture is therefore unequivocal. As an important source of agricultural labor and food production in many rural settings, women are well positioned to have a transformative role in agricultural growth and economic development. Further, economically empowered women have more autonomy over their lives, including increased household bargaining power and social stature, access to health services, decision-making authority over childbearing, and their children have better educational and health outcomes, interrupting the devastating cycle of poverty and poor health. [64] However, in the absence of control over earnings generated from agriculture, often coupled with the inability to rely on income earned by male partners or family members, women in our study adapted the resources and skills readily available to them to generate income that remains under their control. In this case, alcohol production may help them achieve their livelihood goals but may also increase the likelihood of deleterious sexual and reproductive health outcomes, as well as other poor health and social outcomes. Better integration of supportive systems for women – from agricultural education to family planning – as well as joint, multi-sectoral approaches from health and agriculture, would better address the complexities of achieving safe and sustainable livelihoods in this context. In addition, alternative livelihood strategies for women that have the same income-earning potential as brewing may protect women’s (limited) autonomy over earned income and help to reduce the ubiquitous availability of alcohol. Finally, regulatory policy and changes in community social norms around alcohol consumption may reduce the harms associated with alcohol use.