A qualitative study on the effects of intra-household decision-making patterns on utilization of preventive and curative veterinary practices in communities affected by Rift Valley fever in Kenya and Uganda

Abstract

Decisions about how to prevent and treat livestock disease are often negotiated between household members. Animal health interventions target livestock-rearing households but information about the decision-making of men and women and how it influences utilization of animal health best practices including veterinary services is limited. This qualitative study used focus group discussions to assess the effects of gender dynamics on intra-household decision-making regarding utilization of preventive and curative veterinary services in communities affected by Rift Valley fever in Kenya and Uganda. Three household headship structures were considered: male headed, de facto female headed, and de jure female headed. Although their decision-making capacities differed depending on household headship, men were the primary decision-makers on vaccine use while women had some leeway in making decisions on livestock treatment depending on the local context. Understanding gender roles within livestock health management can be used to inform policies and improve animal health interventions.

Keywords:

• Bargaining models
• zoonotic diseases
• livestock ownership
• household headship

Introduction

Livestock health landscape

Livestock play an important role in households in East Africa by providing food and income through the sale of livestock and livestock products. During lean seasons, animals may be sold to purchase cereals and legumes (Njuki & Sanginga, 2013), or serve as insurance when livestock keepers liquidate livestock in response to economic shocks (Behnke & Muthami, 2011). Occurrence of livestock diseases negatively impacts the livelihoods of livestock keepers and undermines the ability of animals to deliver expected benefits to meet household needs. Prevention and control of livestock diseases is thus not just the responsibility of veterinary departments, but also of livestock owners, who include women and men livestock keepers who often perform different animal health care activities and have different levels of knowledge with regards to disease management (Alemu et al., 2019; Gurung et al., 2016). For instance, women act as care providers, feed gatherers, and birth attendants caring for new born lambs/kids and sick animals (IFAD, 2010) while men are responsible for the treatment of sick cattle by administering modern and traditional medicines (Onono et al., 2013). In pastoralist societies, women care for the young stock and sick animals as well as observe animals returning from pastures in the evening, checking for any sickness. Among settled agropastoralists, women are more likely than men to notice signs of illness like poor appetite, nasal discharge, or lethargy (Miller, 2011). In a study conducted in India, researchers found that, for the most part, women mentioned different plants than men in terms of their ethnoveterinary applications (World Bank, 2009). This study contributes to a growing understanding of gender roles within animal health care decision-making with the goal of informing interventions and strategies to increase access to animal health services for women and the livestock they own, manage, or benefit from in their households; minimizing the burden of infectious livestock disease in communities in East Africa; and helping women and men protect and grow their livestock assets. Gender roles in animal health care decision-making can be the deciding factor in whether or not an animal receives veterinary services.

Rift Valley fever (RVF) is an ideal case study for studying gender roles within intrahousehold decision-making in animal health care because it is a reemerging, mosquito-borne zoonotic disease with severe socioeconomic impacts affecting both men and women. Studies have shown that RVF infection patterns can be influenced by gender roles with men being at risk from herding, treating and slaughtering roles while women may be exposed from milking and handling meat and milk from sick animals (Anyangu et al., 2010; Mutua et al., 2017; Nguku et al., 2010). In East Africa, RVF occurs during periods of heavy rainfall and/or flooding (Kortekaas et al., 2012; Pienaar & Thompson, 2013). The disease affects cattle, sheep, goats, and camels (Gerdes, 2004), and is transmitted through bites from infected mosquitoes (Anyamba et al., 2010; Hightower et al., 2012). It is mainly transmitted to humans through contact with infected animal tissues and secretions although mosquitoes can cause the spread of a mild disease (Anyangu et al., 2010; EFSA, 2005; Gerdes, 2004). When it occurs, RVF can cause significant livelihood losses from death of livestock, livestock and livestock products trade bans, and ill health or death in humans (Nguku et al., 2010; Rich & Wanyoike, 2010). The most efficient control option of the disease in livestock and humans is livestock vaccination (Himeidan, 2016; Pienaar & Thompson, 2013). At the time of this study, the Ugandan government did not have an approved vaccine for RVF or guidelines on delivering RVF vaccines to livestock keepers but Kenya did (Mutua et al., 2019; Tumusiime et al., 2018a). In addition, the disease was not listed as “state controlled,” in Uganda, meaning it was not considered a priority disease and its management and control were outside the government’s responsibility (Mutua et al., 2019; Tumusiime et al., 2018a). An additional challenge to strategic RVF disease control is insufficient information on the epidemiological and reservoir status of the RVF virus (Himeidan, 2016).

Veterinary interventions to control a wide range of livestock diseases, including RVF, have targeted livestock rearing households in East Africa, but challenges affect the quality and efficient provisioning of veterinary products and services. Some of the challenges include weak quality assurance systems, underdeveloped or non-existing cold chain systems to ensure potency of veterinary products, and lack of capacity and support systems for field logistics and communication (Dione et al., 2021; Himeidan, 2016). Regulations pertaining to the quality of veterinary products are weakly enforced, in part because the National Drug Authority in Uganda oversees both livestock and human drugs within the Ministry of Health, with a greater focus on human health (Dione et al., 2021).

Intrahousehold decision-making

There is a growing understanding that household-based behaviors and decisions are often negotiated between men and women within a household, and that understanding the decision-making process at an intra-household level is important to identifying where change can happen (Seager et al., 2016). Consequently, intra-household decision-making research has been used to better understand topics ranging from savings (Ashraf, 2009; Schaner, 2015) to the adoption of improved cookstoves (Miller & Mobarak, 2013). Within agriculture, intrahousehold decision-making has been studied in the context of farm management systems focusing on topics including crops, income use in small scale irrigation schemes, agriculture innovations, ownership of livestock and food security, and pastoral migration (Kariuki et al., 2013; McPeak & Doss, 2006; Meinzen-Dick & Koppen, 2012; Mutenje et al., 2016; Shibata et al., 2020). Examples of intra-household decisions in livestock management can be observed in the use of livestock products, implementation of management activities (feeding, livestock health control, herding, watering, etc.) (Senda et al., 2011), sale of animals, disease diagnosis, and treatment of sick animals (Patel et al., 2016).

While gender perspectives have been considered in livestock disease control and animal health management, livestock ownership and control of livestock/livestock products in some studies (Curry, 1996; Galiè et al., 2017; Najjar et al., 2019), the intra household decision-making aspect associated with the control of livestock diseases is not well documented. This study aims to address this gap by focusing on the gender dynamics in intra-household decision-making regarding livestock treatment and vaccination, gendered division of labor/engagement in curative and preventive activities, and vaccine utilization. It addresses one research question:

(1) How do household-level decision-making patterns affect use of preventive (vaccine utilization, including RVF vaccines) and curative (disease treatment) veterinary products and services?

Methodology

Study area

Research activities were conducted in Kenya and Uganda from June to October, 2018 as part of an overarching study focused on gendered barriers to livestock vaccine uptake (Mutua et al., 2019). In Kenya, Murang’a and Kwale Counties were purposively selected because of recent RVF vaccination campaigns and a history of RVF outbreaks (Figure 1). Murang’a County is in the Central region of Kenya and is comprised of seven sub-counties: Kiharu, Maragua, Kigumo, Kangema, Kandara, Gatanga, and Mathioya. This study covered parts of Kandara and Maragua sub-counties. Livestock keepers in the sampled areas predominantly practiced dairy farming. Kwale is in the South Coast of Kenya and is comprised of Matuga, Kinango, Msambweni, and Lunga Lunga sub-counties. Study participants were drawn from each of the sub-counties and comprised of pastoralists and agropastoralists who mainly kept indigenous cattle.

Figure 1. Map of Uganda and Kenya showing the study districts and counties. The inset map shows the location of the two countries in Africa. Map by author.

In Uganda, Ibanda and Arua districts were selected purposively because of recent RVF outbreaks (Tumusiime et al., 2018). Unlike the study sites in Kenya, these districts had not conducted any vaccination campaigns. Livestock keepers sampled from Ibanda in the western part of Uganda were mainly agropastoralists keeping indigenous cattle. Arua district is in the northern part of Uganda, close to the borders of South Sudan and the Democratic Republic of the Congo. Study participants included Ugandan and South Sudan nationals. The South Sudan participants were refugees who had sought a settlement in Arua after fleeing political instability in their home country. They mainly kept goats in agropastoral settings.

Study design, sampling, and data collection

The study utilized a cross-sectional research design. Data were collected through 58 focus group discussions (FGDs) of 8–12 discussants: 14 each in Murang’a, Ibanda and Arua and 16 in Kwale. In each site, FGDs were sex-disaggregated. Thus, half of the discussions were with women only and half with men only. All FGD participants were selected purposively by the research team assisted by veterinary officers and/or local leaders. To qualify for inclusion, a discussant had to be living in the sampled study sites and be a current livestock keeper. All FGDs explored intra-household decision-making patterns concerning livestock treatment and vaccination in male and female headed households, division of labor/engagement in curative and preventive activities, vaccine utilization, and barriers faced when seeking preventive and curative services, to get a broad understanding of current practices. Instead of asking direct questions on how intra-household decisions on livestock treatment and vaccine use were taken, we presented the discussants with three vignettes in each FGD. Each vignette represents a different type of household headship: a male headed household with the male head living locally; a de facto female headed household, in which women ran the household and farm because their husbands were mostly away engaged in non-farm economic activities but contributed actively to livelihood activities and decision or had migrated away from the household; or a de jure female headed household in which a woman was widowed, divorced, or single (Meinzen-Dick & Koppen, 2012; Senda et al., 2011). For each vignette, the researchers probed for decisions that could be made individually by men, women, or jointly in Kenya and Uganda. For consensus building, after the discussants had reported all possible actions, the researchers asked them to identify the decision-making pattern that was likely to be followed by most women in their community. Discussants identified patterns for all household types independent of what type of household they lived in. The three household headship types represent a simplified version of the complexity of actual households, but the inclusion of the de facto household allows for a more nuanced understanding of intra-household decision-making than the oversimplified model of male and female heads of household which by default excludes married women. For ease of understanding, in the vignettes we use the term “one of their cattle” to enable livestock keepers contextualize the cattle type they keep. For instance, in Murang’a this would mean a dairy cow but in other sites the cattle could be indigenous cattle, whether kept for beef or milk. Further, although RVF affects cattle, camels, sheep, and goats, we only used cattle as an example. We therefore acknowledge the possibility that had we focused on all species we may have had different findings concerning the small ruminants. Despite selection of the study sites due to recent outbreaks, RVF was still not well known to many livestock keepers, so we chose to discuss curative and preventive services generally rather than mentioning specific diseases, but the findings are transferrable to understanding responses to livestock diseases affecting ruminants, especially those preventable with vaccines such as RVF. The lack of awareness of RVF by livestock keepers was also documented in Uganda by Namatovu et al. (2021). One contributing factor may be the cycles of periodic outbreaks several years apart or more which makes it less distinctive than more well recognized diseases that often appear annually such as peste des petits ruminants.

Vignette 1: a woman from a male-headed household who is living with her husband

Peter is a teacher at a local primary school and a livestock keeper. His wife Jane is a housewife. This morning, after her husband left for work, Jane observed that one of their cattle was looking unwell. She has also been informed by the village elder that the veterinary department has made a call to have all cattle, sheep and goats vaccinated to prevent the outbreak of a serious disease at a later date. How is Jane expected to handle the two developments?

Vignette 2: a de facto female head of household whose husband lives away from home

Mary is Jane’s neighbor. She is married but due to work demands, her husband stays away and comes home from time to time. Mary has observed that one of her household’s cattle looks unwell. The village elder also passed by her house and told her that the veterinary department has made a call to have all cattle, sheep and goats vaccinated to prevent the outbreak of a serious disease at a later date. How is Mary expected to handle the two developments?

Vignette 3: a de jure female head of household who does not have a husband

Mercy lives not so far from Mary. She has no husband. She has observed that one of her household’s cattle looks unwell. The village elder also passed by her house and told her that the veterinary department has made a call to have all cattle, sheep and goats vaccinated to prevent the outbreak of a serious disease at a later date. How is Mercy expected to handle the two developments?

Data from Kenya were collected in Swahili while in Ibanda and Arua, Lunyankole/Lukiga and Lugbara were used, respectively. Prior to data collection in Kenya and Uganda, the focus group discussion guides were pretested in villages drawn from Murang’a, Ibanda, and Arua but not included in the main study. The pretest exercise was aimed at informing the research team if the questions were relevant and comprehensible to potential focus group discussants and how best to enhance their suitability for the study. During all FGDs, audio data were recorded, and field notes taken.

Data management and analysis

Recorded data were used to improve the FGD notes generated during the focus group discussions, and to extract the quotes used in this paper. The task was conducted by a data collection team whose members were competent in English and Swahili, Lugbara, or Lunyankole/Lukiga. The FGD notes were managed, coded, and analyzed inductively in Nvivo 12 (QSR International, Melbourne) through content analysis. The quotes presented are views from individual FGD participants and not necessarily the entire group’s view. They are intended to demonstrate variations in thought and should not be generalized. The number of FGDs in which an issue is mentioned is indicated to demonstrate the level of convergence or divergence in thought among participants.

Ethical considerations

This study was conducted in accordance with the Declaration of Helsinki, and the protocol granted ethical approval by the International Livestock Research Institute Research and Ethics Committee, referenced ILRI-IREC2017-19. Before participating in any FGD, confirmations were made that all prospective participants were at least 18 years of age, the legal consenting age in Kenya and Uganda. Eligible discussants were then taken through a consenting process which comprised of an explanation of the study, data collection methods, data documentation through recorders and fieldnotes, data privacy, management, and use. They were also allowed to ask questions concerning the study and receive satisfactory answers. Participants consented to the study by signing or providing a thumb print on the consent form voluntarily.

Results

The results are organized into three sections: (1) demographic characteristics, (2) intra-household decision patterns making on livestock treatment, and (3) intra-household decision patterns on vaccine use.

Demographic characteristics

A total of 645 discussants participated in 58 FGDs conducted as follows: 16 in Kwale and 14 each in Murang’a, Ibanda and Arua (Table 1). Of these, 322 were men and 323 women drawn from 14 FGDs each in Murang’a, Ibanda and Arua and 16 in Kwale. In all the study sites, more than half of the participants had primary education: 53% in Murang’a, 52% in Kwale, 60% in Ibanda, and 60% in Arua. Discussants’ mean and median ages were 46 years and 45 years respectively, and ranged from 20 to 82 years. Through proportion piling we established that involved communities kept cattle, goats, and sheep, in decreasing order.

Table 1. Demographic characteristics of FGD participants.

			Kenya		Uganda
			Murang’a	Kwale	Ibanda	Arua
Number of FGDs		Men	7	8	7	7
		Women	7	8	7	7
		Total	14	16	14	14
Number of FGD participants		Men	76	81	84	81
		Women	78	82	83	80
		Total	154	163	167	161
Participant’s highest level of education	Primary	Men	44%	54%	71%	54%
		Women	62%	49%	42%	65%
		Combined	53% (N = 150)*	52% (N = 163)	60% (N = 167)	60% (N = 161)
	Secondary	Men	49%	28%	19%	32%
		Women	35%	18%	23%	19%
		Combined	41%	23%	21%	26%
	Tertiary	Men	6%	4%	4%	10%
		Women	0%	6%	4%	4%
		Combined	3%	5%	4%	7%
	Adult education	Men	1%	3%	0%	0%
		Women	0%	0%	1%	0%
		Combined	1%	1%	1%	0%
	No formal education	Men	0%	11%	6%	4%
		Women	4%	29%	30%	13%
		Combined	2%	19%	18%	8%
Participant ages (years)	Mean (Range)	Men	54 (26–79)	47 (20–73)	48 (20–82)	40 (21–68)
		Women	47 (23–78)	45 (20–78)	46 (22–76)	37 (20–65)
		Combined	50 (23–79)	46 (20–78)	47 (20–82)	38 (20–68)
RVF susceptible livestock (ordered proportions)			Cattle Goats Sheep	Cattle Goats Sheep	Cattle Goats Sheep	Cattle Sheep Goats
Livestock species preferred for vaccination (ordered proportions)			Cattle	Cattle Goats	Cattle	Cattle Goats

* 4 participants preferred not to disclose their age and education status, N is same for all education levels.

Treatment

Next, we present responses to the vignettes about treating a sick animal organized by household headship types: male headed, de facto female headed, and de jure female headed. Study site information is included for context because factors such as the production system and prevalence of dairy cows, for example, influenced decision-making.

Male headed households

In male headed households with men living locally in Murang’a, both men and women engaged in decision-making on the treatment of dairy cows (Figure 2). While it was expected that men as heads of household had the capacity to take decisions on livestock health, women’s capacity was enabled by several factors. First was the recognition among male heads of household that women needed to agree with the introduction of dairy cows into the home because they increased demands on women’s labor. Male discussants reported that without women’s goodwill and support, the improved breed animals would receive substandard care and end up dying. Recognizing the centrality of women’s contribution created a sense of joint ownership between men and women, although men were culturally considered the primary owners. Additionally, it enhanced women’s agency in the care of their household’s animals.

Figure 2. Visual summary of decision-making patterns on livestock treatment by household ­headship as reported by FGD participants.

The decision to keep a dairy cow also came with the recognition that professional veterinary services would be instrumental in enhancing the health and productivity of the animal. As a result, most households in Murang’a had a specific veterinarian that they entrusted with treating their animals whenever they showed signs or symptoms of illness. Thus, whenever women, who spent most of the time at home, suspected that an animal was ill, they responded primarily by calling their spouses immediately (10 FGDs) or later (6 FGDs). The goal of this call was not to seek consent but to notify them that they had called or would be calling the veterinary officer without major contestations. The information transfer was also a form of insurance for women in case the treatment was unsuccessful or the animal died from illness. Depending on households, men interpreted being informed first or later as a marker of respect for their position, authority, and property. Examples of women safeguarding themselves and men protecting their authority and property are shown in the following quotes:

Homes are different. There are homes if the woman acts without telling the husband it brings trouble later, he asks why he was not told. So, if she knows her husband is like that she will tell him first that the animal is sick what do we do. Then there is another type where you handle the problem first, the woman acts then informs him later.” (Men’s FGD, Ngararia, Murang’a County, Kenya).

You have to tell him before calling the veterinary doctor, you do not do that [not inform him] yet it’s his livestock, and you do not want it to look like disrespect over his things. (Women’s FGD, Irigiro, Murang’a County, Kenya).

I think, because the cow is sick, she should call the veterinary doctor because she already has his number first before telling the husband and after the doctor has treated it, she can inform her husband. (Men’s FGD, Wathiani, Murang’a County, Kenya).

If the women did not have money readily available to pay for the veterinary services, they asked their spouses to make the payments or borrowed money from their social networks to make the payments and got reimbursements from their spouses later or got credit services from the veterinary officer based on the trust relationship cultivated between them. A woman in Muran’ga described husbands taking on the liability and responsibility for expenses, even those associated with a dairy cow that a woman has power over or may have contributed to financially, as part of his role as head of household.

If you have a cow, and you don’t have money to buy something, the man will buy it. (Women’s FGD, Wathiani, Muran’ga County, Kenya).

In Kwale (16 FGDs) and Ibanda (10 FGDs), the norm was that only men made decisions about the veterinary treatment of local cattle breeds, which are primarily considered men’s livestock and therefore culturally owned by them (Figure 2). Women’s role was to inform their spouses that animals appeared sickly then leave all decision-making to them. Men’s and women’s perceptions on ownership were central to the level of involvement in livestock health management as exemplified by the following quotes from Uganda:

If you tell the man that the animal is mine and not his, he gets disgusted and he stops caring for it. (Women’s FGD, Kajwamushana 5, Ibanda District, Uganda)

If I find some animals are missing and when I ask him and he does not want to answer me claiming that they are his property, I also close my eyes and ignore the animals. (Women’s FGD, Kakunama 3, Ibanda District, Uganda)

Some men don’t buy animals at home. Now you as a woman, if you buy an animal, he will tell you that you’re the one who bought it, if you cannot take good care of it then it is up to you. (Women’s FGD, Ariwara, Arua District, Uganda)

If the animals are at home, men have a lot of authority [over them]. However, in polygamous families… the wives cannot allow animals from their homes to be handled by the man if they are to be used in the other home. This makes men pull away from the management of such animals. (Men’s FGD, Ariwara, Arua District, Uganda)

In Kwale, where there was limited reliance on professional veterinary services, the animals were self-treated either by the heads of household or other trusted men authorized by the male heads of household using purchased drugs. In some circumstances, which were determined by the nature of the relationship between husband and wife, availability of the male head of household when contacted, and the ethnic community the household belonged to, women could intervene with the help of male relatives and lay experts (trusted livestock keepers or neighbors) and inform their spouses later after treatment as reported in 5 FGDs. In response to men’s dominance in cattle keeping in Ibanda, a reported trend (3FGDs) was for women to involve themselves in goat keeping where there was less probability for conflict because men’s interest was lower and some leeway for women in making decisions around the treatment of goats. In Arua, both men and women were involved in decision-making for the treatment of goats. Once women observed animals showing signs of ill health, they notified their spouses before (5 FGDs) or after having the animals treated (9 FGDs). The enabling factors for women in decision-making included having an accessible veterinary officer and having funds to pay for professional services. Some of the households in Arua had received goats as part of an initiative by a humanitarian organization and therefore, it is likely that since accountability for the animals went beyond the home, women played a greater role in ensuring their well-being. The mention of household dynamics in a polygamous family from the Arua FGD is a reminder that not all households are structured with a married man and woman. Decision-making in polygamous homes was beyond the scope of this study to examine in depth, but the limited references confirm decision-making and power over livestock was not standard and varied by families, with both women and men having authority depending on the context.

De facto female headed households

In de facto female headed households, most women in Murang’a (10 FGDs) reported seeking professional veterinary services when their dairy cow appeared ill and then informed the spouse later but within the same day. In these households, women had broader leeway for independent decision-making because they were the ones readily available at home, who saw the severity of the animal’s health condition, and had easier access to the veterinary officers. Having access to money to pay for professional veterinary services, loans from friends, or credit facilities from veterinary officers enhanced women’s ability to have animals treated promptly. Like households where the male head of household lived locally, women informed their spouses of the actions they took as a means of sharing information of their animal’s health status as well as protection against backlash if treatment was unsuccessful or the animal died. In Kwale and Ibanda, it was the male heads of households who determined how cattle/livestock would be treated despite being absent from the homes. In Ibanda, if women were engaged in goat keeping, they would be involved in decisions over treatment. Enabled by similar factors as those in male headed households with spouses living locally, women in Arua (9 FGDs), same as Murang’a, had their livestock treated and informed their spouses later within the same day.

De jure female headed households

Women in de jure female headed households were the primary decision-makers in treatment of their livestock in all the study sites (14 FGDs in each site). In the decision-making process, although not mandatory, it was not uncommon for women to consult other family members such as older children, brothers/fathers-in law, or neighbors before making the final decision. Consultations were especially important if she would need assistance with making payments for services sought. In Arua, if the ill livestock belonged to a deceased husband, his relatives were to be consulted on how to manage the animal before any action was taken.

Vaccine use

Findings from the vignettes about vaccinating a healthy animal are also presented in order of household headship type: male headed, de facto female headed, and de jure female headed. Study site information is presented for context.

Male headed and de facto female headed households

Men, as main livestock owners, were the primary decision-makers on vaccination in both male headed and de facto female headed households (Figure 3). Some of the considerations men made in the decision-making process included: (1) the direct and indirect cost of vaccines, (2) the reason provided by the veterinary department for wanting the animals vaccinated, (3) the source of the vaccines and whether the vaccines themselves were perceived as efficacious, (4) if the vaccines were perceived as safe for their livestock, (5) the livestock/species to be considered for vaccination, and (6) if they would agree to livestock being presented for vaccination. The following quotes demonstrate some of the livestock keepers’ concerns that influence vaccine use:

Figure 3. Visual summary of decision-making on livestock vaccine use by household headship as reported by FGD participants.

Not all animals are vaccinated. You know, for some vaccines we are required to pay, so if you do not have money, you will not take the cattle because it is not for free. You pay something. (Men’s FGD, Gathuri, Murang’a County, Kenya)

Some people worry that if they take their animals for vaccination, their tail tips will fall off [Contagious Bovine Pleuropneumonia vaccine side effect]. (Women’s FGD, Vanga, Kwale County, Kenya)

Because vaccines are not provided regularly, when people are asked to present their animals for vaccination, they do not take it seriously. (Women’s FGD, Vugunjini, Kwale County, Kenya)

Unlike curative treatment, which is based on visible signs of disease, vaccination involves the administration of vaccines to boost an animal’s immunity against infectious agents of interest before an animal is infected and eventually appears unwell. Livestock keepers are often more cautious on the use of vaccines depending on the levels of trust held for the service providers and the vaccines on offer. The main source of vaccines in all the study sites was the respective veterinary departments. In Arua, besides vaccine use being determined by male heads of household, the local government leadership played a significant role in influencing the decision. Their word was widely respected, trusted, and acted upon by community members. This gave women leeway to have livestock vaccinated even if their husbands at an individual level had reservations.

If the male heads of household approved for their livestock to be vaccinated, women from Murang’a took an active role in taking them to the vaccination sites and ensuring they received the vaccines. Only in very limited occasions would women in male headed households from Murang’a independently make decisions about dairy cows’ vaccination, and this was in situations where the heads of household did not involve themselves in livestock production (3 FGDs). In Kwale and Arua, if the male heads of household were unreachable, neighbors or brothers and fathers-in-law would make the decision. Additionally, in Arua, if the village leadership was considered trustworthy, women would have livestock vaccinated in the absence of direction from their spouse, although there was variation by household. In Ibanda, where male heads of household had delegated the responsibility of livestock care to women, the role was only executed if she had the financial means to access the services.

De jure female headed households

In de jure female headed households, decisions on vaccine use were taken primarily or only by women, although they may consult their older children, brothers/fathers-in-law and neighbors. Similarly, when having animals treated, although not mandatory, consultations were important if women needed financial support in paying for the services.

In the four sites, vaccines were mainly provided by the respective veterinary departments, although this was not done regularly. Livestock keepers were required to present their animals for vaccination in the places demarcated for exercise by the veterinary departments. The timing and type of vaccines to be delivered was determined by the veterinary departments then communicated to communities through local leadership in places of worship, community social gatherings, and schools. In Murang’a, Ibanda and Arua, livestock keepers were required to pay predetermined cost for each animal vaccinated. In Kwale, the veterinary department did not charge farmers for vaccines delivered as the costs were covered by the County.

Discussion

In the discussion, we revisit how household-level decision-making patterns affect the use of preventive and curative veterinary products and services and apply bargaining approach models to better understand livestock disease control/management decision-making. Household headship type combined with the context of the animal health service (preventive or curative) revealed some decision-making patterns. Married women in male headed households where the heads lived locally and de facto female-headed households had more leeway to make decisions about treatment than for preventive services such as vaccination. The underlying logic was that treating a sick animal, especially one of high value such as a dairy cow, was time sensitive. Additionally, treatment of an already sick animal was viewed as less risky than vaccinating a healthy animal. Other factors such as production system and ethnic group varied by location and influenced decision-making patterns, such as the trend amongst agropastoralists in Ibanda for women to spend more time and resources on keeping goats, where they had more leeway to make decisions compared to the indigenous cattle, viewed to be within the domain of men. Production system, ethnic group, and other site-specific factors appeared more influential on the gendered division of veterinary decision-making than country; there were no clear similarities or differences in practices between Uganda and Kenya. Rather, the context of each study site influenced practices that favored or discouraged involvement of women in veterinary decisions. For example, as shown in Figures 2 and 3, remote Kwale County in Kenya and agropastoral Ibanda district in Uganda paired together with only men making decisions on treatment and vaccine use, even in households where the men lived or worked away. Unfortunately, the study design did not allow us to fully differentiate responses to different veterinary decision-making scenarios by livestock species (small ruminants versus cattle), although species is certainly a factor, such as the examples of the value of a dairy cow in Murang’a justifying a rapid decision about treatment by the most available household member, man or woman. While sheep were included in the vignette about vaccination, they were mentioned less often than goats and not mentioned in the ordered proportion activity for livestock preferred for vaccination (Table 1).

Additional insights can be gained by conceptualizing intrahousehold decision-making on treatment and vaccine use in relation to cooperative, collective, and non-cooperative bargaining approaches/models (Agarwal, 2014). In the cooperative model family welfare is valued over the utility of any individual member (Alderman et al., 1995; McPeak & Doss, 2006). Intrahousehold decision-making contains elements of cooperation and conflict related to who gains and who loses in the cooperative outcome (Sen, 1987). The outcome will depend on the bargaining power of household individuals, which is determined by their respective fallback positions, such as household headship. Thus, cooperative outcomes are possible in relation to who does what, who gets which goods and services, and how each member is treated. Men, as heads of households, serving as the primary decision-makers on vaccine use can be considered within a cooperative bargaining model if the man acts within his social contract to use his greater control over limited household resources for the family welfare. This model was mainly observed in Murang’a. The example of men as heads of households being primary decision makers for vaccine use demonstrates that gender norms have an influence the intrahousehold decision-making on treatment and vaccination (Donni & Ponthieux, 2011; Sinha, 2020).

Although women’s involvement in household decision-making is often dictated by societal norms (Agarwal, 2014; Anderson et al., 2017; Deere & Doss, 2006; Distefano, 2013; Doss et al., 2014; Mayala et al., 2019; Okali, 2011; Quisumbing, 2007), for example in the case of dairy cows in Murang’a, decision-making depended on the level of the control of resources by women and their contribution to family welfare (Deepanka et al., 2021). Where treatment options are clearly spelled out, a woman could make decisions even when the male head of household is away, thus increasing a woman’s agency and decision-making capacity in terms of utilization of veterinary services.

In remote areas, such as the case in Kwale where veterinary services were not easily accessible, decision-making over treatment of animals was done by men who mostly self-treated their animals. This is in line with other studies which indicate how decisions regarding animal healthcare are characterized by “lay” diagnoses and treatment with limited input from trained health professionals (Caudell et al., 2020). Women were not permitted to treat animals, and this was attributed to their low literacy rates (see education levels in Table 1) and an inability to read the user instructions on drug labels, packaging, or instruction sheets. Their (women’s) perceived lack of experience and knowledge in animal treatment tends to limit their decision-making and engagement in the treatment of animals. Consequently, their exclusion from treatment tasks is considered for the greater good of their households’ livestock. Women in de jure female headed households may also find themselves in the position of making vaccination decisions for the good of the household.

The collective model assumes that if resources remain constant, then the well-being of one household member cannot be increased without decreasing that of another (Donni & Ponthieux, 2011). This model is further demonstrated in intra-household decision patterns on vaccine use where most decisions regarding vaccination, whether for cattle, sheep, or goats, are made unitarily by men in male headed households, based on a man’s ability to pay for the vaccines and perceived trust in the vaccines. Since men are tasked with paying for the vaccines, it is possible that they may refuse to pay or take the animals for vaccination if they believe the vaccine will have negative side effects. This is consistent with other studies which indicate that vaccination decisions are mainly made by men (Lindahl et al., 2019), and that vaccine adoption for livestock is gendered with women less likely to adopt the use of this technology than men (Acosta et al., 2019; Donadeu et al., 2019). Women in de jure households still consult with male relatives on matters regarding vaccination, as these may require financial input as well as help with driving animals to vaccination centers. This can be attributed to perceived risks associated with livestock treatment and vaccines, deeming it best to leave such decisions to the heads of household so that they bear liability in case any harm befalls the animals. By so doing, women safeguard themselves from conflict and blame and by extension maintain amicable relations with heads of household.

The non-cooperative bargaining model does not assume pooled resources, rather household members make separate but interrelated production decisions about their own resources and spend individually based on their preferences and interests (Doss & Meinzen-Dick, 2015). Thus, it is possible for individuals in a household to opt out of pooling resources whether monetary or in-kind (Doss & Meinzen-Dick, 2015). Different factors may affect the bargaining process leading to disagreements and eventual adoption of a non-cooperative model, examples being perceptions of ownership and ability to accrue benefits from livestock as seen in this study. Ownership conflicts arise from lack of equal benefit from livestock sales and can lead to domestic gender-based violence (Mutua et al., 2019; Oluka et al., 2004). Conflict may also occur over the use of income accrued from livestock if one household member feels they are not benefiting from their efforts or feel that their power as decision-maker is undermined (Fiala & He, 2016).

Spousal differences in preferences and the inability to cooperate can result in inefficient outcomes with one spouse opting out of livestock management or engaging in a different enterprise, or livestock ailing or dying from disease. Women in Ibanda shifted their focus to producing goats because these were viewed as less valuable and more socially acceptable for women following their exclusion from the care and decision-making on the household’s cattle and falls within the non-cooperative bargaining model. In cases where women deliberately shy away from any livestock management because they are not benefiting from their efforts, animals may be neglected resulting in poor outcomes. Additionally, this implies that women’s non-involvement in livestock production can be a matter of choice and not necessarily direct exclusion by their husbands. This also demonstrates that the more a household member benefits from their efforts, makes decisions on animal welfare and marketing, the more likely they will be able to cooperate and take care of the animals. Previous research has shown that at an extreme, the withdrawal option constitutes a threat point where a spouse may remain within the marriage but withdraw and live in a separate location (Agarwal, 2014).

Conclusion

This study has explored the different bargaining models utilized in intra-household decision-making concerning livestock treatment and vaccination. The use of each bargaining model is contextual, but we see the collective bargaining model being more dominant in decision-making on vaccines. While the study findings may be adopted to better understand the gendered use of a variety of preventive and curative veterinary services, the focal disease was RVF, which can only be prevented through vaccination of susceptible livestock before an outbreak occurs. This research supports some practical considerations about how to improve livestock disease management and vaccine use. Men who are heads of households should be targeted with vaccination information because they are the primary decision-makers on vaccine use, with the understanding that women, particularly when in household leadership positions, may also make vaccination decisions, and that women across household structures often play a role in disease identification. Women have broader authority when a spouse or other household member is absent or disinterested in livestock keeping and for treatment decisions where taking action is viewed as time sensitive, such as to protect a valuable dairy cow. Environments where vaccination is mandatory, or vaccines have the support of respected local leaders, give women who would like to vaccinate support and leverage when negotiating to have animals vaccinated or sourcing funds to meet the direct and indirect costs of vaccines. Livestock interventions should also consider that ownership conflicts over livestock can introduce tensions within households and hinder use of beneficial services.

Acknowledgements

We acknowledge the support offered by the Ministry of Agriculture, Animal Industry and Fisheries, through partnering with us in Uganda and introducing us to the leadership of Ibanda and Arua districts. We acknowledge the contributions of Nicholas Ngwili, and Tracey Mwongela from Kenya, and Simon Tugume, Aisha Sebii, James Muhangi and Habert Jarua from Uganda, in data collection and transcription. We also thank Judy Kimani and Bethlehem Alemu at the International Livestock Research Institute for their expertise designing Figures 2 and 3.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

All qualitative data summaries are stored in the International Livestock Research Institute dataverse and are accessible through https://hdl.handle.net/20.500.11766.1/FK2/GN13EL.

Additional information

Notes on contributors

Edna Mutua

Edna Mutua is a Medical Anthropologist, formerly working with the International Livestock Research Institute, Kenya. Her research interests are within the spheres of One Health, zoonotic diseases, antimicrobial resistance, and gender analysis of livelihoods and health interventions in East Africa.

Jane Namatovu

Jane Namatovu is a gender support specialist at International Livestock Research Institute (ILRI), Uganda. Prior to joining ILRI Jane worked for Dan Church Aid, a humanitarian INGO and Voluntary Services Overseas. Jane takes an interdisciplinary approach to conducting research. Her work includes support to the Uganda in-country implementation of Gender and Social Inclusion activities. She has experience conducting gender assessments and using participatory approaches. Jane currently works on the Animal Health Program funded by the German Federal Ministry for Economic Cooperation and Development (BMZ) and the Sustainable Animal Productivity for Livelihoods, Nutrition and Gender inclusion (SAPLING)—a Uganda country program which seeks to unlock the empowering potential of livestock business and livestock rearing for women and youth. Within the animal Health program her work focuses on gender assessments. She has conducted research activities within the humanitarian and development settings.

Zoë A. Campbell

Zoë A. Campbell is a Scientist at the International Livestock Research Institute (ILRI) focusing on the intersection between gender and animal health. Her current projects include identifying gender-sensitive vaccination strategies as part of the global effort to eradicate Peste des Petites Ruminants (PPR) and supporting monitoring and evaluation of a development project to improve hatchery capacity to vaccinate day-old-chicks in East and West Africa. Formal training in biology, social sciences, and economics allows her to take an interdisciplinary approach. Beginning with her time in Tanzania as a Peace Corps volunteer, she has ten years of experience with community engagement and research in East Africa. Her research interests include adoption of animal health technologies including vaccines and bacteriophages, integrating gender into One Health research, and women’s empowerment.

Dan Tumusiime

Dan Tumusiime is a Veterinary Epidemiologist working as Senior Veterinary Officer in the central Epidemiology Unit at the Uganda’s Ministry of Agriculture, Animal Industry and Fisheries (MAAIF). His roles in the Ministry include initiation and designing of disease control strategies and plans, coordination, implementation of animal health interventions including surveillance; risk analysis and communication; epidemiological studies; epidemic response, prevention and control of infectious animal diseases. He is the National Coordinator of the control and eradication of Peste des Petits Ruminants (PPR) in Uganda. He is a member of the National Task Force subcommittee on zoonoses and the Antimicrobial Resistance (AMR) Public Awareness Teaching and Education committee. He is a co-leader of the Rift Valley fever (RVF) research program implemented by the International Livestock Research Institute (ILRI) in partnership with MAAIF for improved RVF surveillance and control. He is also graduate research fellow on Epidemiology of RVF in Uganda at ILRI under the Boosting Uganda’s Investments in Livestock Development (BUILD) program, with his PhD registered at the Freie Universität Berlin. Specific interests include research on the epidemiology of zoonoses and Transboundary Animal Diseases.

Emily Ouma

Emily Ouma is a Senior Scientist at the International Livestock Research Institute (ILRI) with post-doctoral experience in economic analysis of smallholder agricultural systems and agricultural value chains in sub-Saharan Africa. She has a strong professional background in socioeconomics and gender analysis of agri-food systems, and exante and ex-post impact assessment of agricultural technologies and institutional innovations in sub-Saharan Africa, coupled with extensive field experience in sub-Saharan Africa countries.

Bernard Bett

Bernard Bett is a Veterinary Epidemiologist and the Team Leader of the One Health Research, Education and Outreach Center in Africa (OHRECA) at the International Livestock Research Institute. His work at the center focuses on reducing antimicrobial resistance, ensuring food safety, controlling neglected zoonoses and preventing emerging infectious diseases through a One Health approach.