Introduction

Sub-Saharan African countries are experiencing rising food insecurity and malnutrition aggravated by price shocks and disrupted supply chains, investment funds speculating in commodities markets, and climate change impacts (FAO et al. 2021; Agarwal et al. 2022). Fostering the production of African indigenous vegetables (AIV) is regarded as one strategy to create nutrition-sensitive and resilient value chains to support people’s livelihoods in many sub-Saharan African countries (Bokelmann et al. 2022). AIV are indigenous plant species in Africa, which are well adapted to local environmental conditions, known for their climate resilience and nutritional and health-promoting value (Towns and Shackleton 2018; Ghanem 2022). AIV include, for example, species such as amaranth (Amaranthus ssp), African nightshade (Solanum nigrum), spider plant (Cleome gynandra), cowpea leaves (Vigna unguiculata), pumpkin leaves (Cucurbia maxima, C. pepo), and jute mallow (Corchorus olitorius).

For decades, AIV were mostly grown by women and for subsistence, in kitchen gardens or in intercropped farming systems but rarely at larger scales. The last decade, however, witnessed the beginning of a shift from AIV being low-scale subsistence crops to being an important income source as well as a significant part of urban food consumption (Bokelmann et al. 2022). This is driven by rising demand from urban consumers (Gido et al. 2017), increasing acceptance of traditional local food, awareness campaigns, and an increase in nutritional and health concerns that can be addressed through the consumption of AIV (Kamga et al. 2013; Gido et al. 2017; Mwadzingeni et al. 2021).

Transition from subsistence to commercial agriculture is sometimes regarded as a precondition for income generation and the alleviation of poverty (Tuni et al. 2022), particularly in countries where small-scale farming is still common. Various studies document a positive relationship between agricultural commercialization, income, and improved nutrition (Fischer and Qaim 2012; Krause et al. 2019; Ogutu et al. 2020). Others, however, critically emphasize both the adverse effects on (female) smallholders, including new dependencies (from men or markets), and the limitations and unequal distribution of the benefits of commercialization (Park and Maffii 2017; Gironde et al. 2021; Prügl et al. 2021). Due to the heterogeneity of the literature in this context, this study uses the case of AIV in Western Kenya to investigate, from a feminist economics perspective, the effects of commercialization on smallholder farmers, especially in regard to the distribution of benefits.

Conceptually, this article draws on works of feminist economics in agriculture, which have emphasized the changing character of social reproduction in contexts of commercialization and its uneven benefits when formerly non-market transactions become subject to market principles (Bhattacharya 2017; Razavi 2009). Such a perspective places particular emphasis on the analysis of gender inequalities in labor and decision-making power in crop production, including in access to resources such as independent income from agricultural labor; it also highlights the agency of women (Agarwal 1992; Razavi 2003). Based on a mixed methods study in Kenya, this article aims to answer the following research question: How does AIV commercialization affect the gendered division of labor, women’s participation in agricultural labor, their decision-making power, and their access to resources?

The commercialization of African indigenous vegetables in Kenya

Due to the need for income generation within smallholder farming systems, rapid urbanization in many sub-Saharan countries, and urban consumers’ demand for and dependence on local agricultural production, most smallholder farmers today supplement their livelihood by selling part of their production at markets. In various countries, evidence shows that market participation has increased across income classes and cropping systems (Carletto et al. 2017). Kenya’s smallholder farmers follow this trend, as various studies document commercialization across different crops, including bananas (Fischer and Qaim 2012), exotic vegetables (Muriithi and Matz 2015), cassava (Opondo et al. 2020), and also AIV (Krause et al. 2019).

AIV have been cultivated in sub-Saharan Africa for centuries, were regularly grown in home gardens or alongside cash crops (e.g., sugar cane), and formed an integral part of the rural diet (Abukutsa-Onyango 2007). Yet they were often seen only as “edible weeds” or “poor man’s food” and thus were not grown on larger scales or sold at markets (Brückner 2020). For a long time, AIV production was distinctly different from that of cash crops for two major reasons: First, AIV are a key component of local dietary diversity and only little commercialized. Second, they were traditionally produced by women as subsistence crops (Mwadzingeni et al. 2021). For instance, a recent study from Kenya showed that in AIV production and marketing, women had a greater share of involvement compared to other crops (Deißler et al. 2021).

Today, however, AIV are slowly emerging from the shadows of subsistence production; increasingly, they are purchased and sold commercially as conventional horticultural crops. Due to the informality of the sector, there exists neither precise annual aggregate production data nor data on the share of production sold at markets. Survey data from 2003/2004 indicated that 90% of respondents perceived an increase in demand (Abukutsa-Onyango 2007). This trend was driven by rising demand from urban consumers (Gido et al. 2017), the increasing acceptance of traditional local food, awareness campaigns, and an increase in nutrition- and health-related concerns that could be addressed through the consumption of AIV (Kamga et al. 2013; Gido et al. 2017; Mwadzingeni et al. 2021). For some, AIV are also an important and cherished alternative to the existing food system, for instance, to exotic kale (Brassica oleracea var. acephala), which enjoys dominant popularity (Abukutsa-Onyango 2007; Brückner 2020). Based on various recent literature, AIV are reported to be increasingly commercialized in local and regional markets within the region covered by the present study and throughout East Africa. For this study, AIV commercialization in smallholder horticulture is defined as a shift in the purposes of production, away from subsistence or exchange without compensation and toward commercial purposes and income generation.

Some literature reports on the positive effects of AIV commercialization on income (Ngenoh et al. 2016; Krause et al. 2019) as well as nutrition indicators such as dietary diversity (Krause et al. 2019) and contribution to human health measures (e.g., Odongo et al. 2018). Within this strand of literature, the inclusion of smallholder producers in commercial agricultural value chains is seen as a means of generating independent income and economic growth and thus contributing to social inclusion. In general, advancing commercialization to assure the inclusion of disadvantaged groups in terms of income, jobs, and access to food is a well-established focus of the literature on smallholder farming in the Global South (Andersson et al. 2015; Meemken et al. 2019), with women being one of the primary groups of concern (Doss 2001).

Positive assessments of the commercialization of agricultural crops are challenged by critical literature that points to the unequal distribution of the benefits of commercialization among men and women and to the risk of women losing control of production and becoming disempowered should men enter into and take over the production of former subsistence crops. Commercialization can go hand in hand with the dispossession of smallholder farmers—making them dependent on agricultural value chains as wage workers on plantations or as contract farmers producing for export markets—and hence can aggravate existing social inequalities (Dolan 2001; Njuki et al. 2011; Park and Maffii 2017; Gironde et al. 2021; Prügl et al. 2021). Moreover, some empirical studies find little improvement in nutritional outcomes as a result of commercialization (Carletto et al. 2017).

The existing literature on commercialization is therefore ambiguous and points in different directions. Quantitative data resulting from, for example, the application of more robust estimation techniques using panel data, as well as qualitative evidence on the gendered effects of the commercialization of AIV in Kenya, is limited. Studies in sub-Saharan Africa show that the allocation of resources such as land and fertilizer is highly inefficient, with too few resources applied to crops cultivated by women (Apedo-Amah et al. 2020). Interviews frequently reported that also for AIV, male household heads allocated only small pieces of land to their wives for AIV cultivation but adjusted resource allocation when AIV began generating revenue (Ferenczi 2021). This suggests that when men take over decision-making related to production, they may allocate more land to AIVs, potentially leading to an increase in yield (Brückner 2020).

Through this transmission mechanism, AIV production and self-consumption can also increase along with an increase in male decision-making. Access to land is also an important production bottleneck. Farm size has been on the decline for several decades in Kenya, and arable land is extremely scarce (Jayne et al. 2003). Yet women’s access to and control over land in Kenya is particularly limited. Despite progressive formal legislation that aims at gender equality and legal opportunities for women to possess land, it is still very rare for them to hold land rights and de facto control over decision-making. According to the Federation of Women Lawyers, while women head about 32% of households, they individually hold only 1% of land titles (Mbugua 2020; Federation of Women Lawyers n.d.).

Overall, this qualitative evidence also suggests that gender relations strongly shape the allocation of household responsibilities and resources in the setting of our current study (Bakare-Yusuf 2003).

Analyzing commercialization from a feminist economics perspective

To better understand the distributive effects of commercialization, this study conducts a political economy analysis building on conceptual work in feminist economics. It focuses the interrelationship between gender and the economy and emphasizes the relevance of a gendered understanding of institutions and economic processes (including at the family and household levels) (Waylen 2000; Agenjo-Calderón and Gálvez-Muñoz 2019). Feminist economics brings “the domestic arena into the analytical framework” (Razavi 2009, p. 200) of economic analyses. Through a critique of the assumptions of neoclassical economics, the approach extends the analysis of economic processes to the political economy of the household, including both gender inequalities and other social inequalities. It draws attention to the weaknesses of neoclassical economics by emphasizing reproductive work as a “background condition” of production and economic value and by problematizing the gendered division of labor (Razavi 2009, p. 211). Feminist economics includes household dynamics as constitutive of gender inequalities, problematizing the division of labor based on the assignment of reproductive work and care work, which are generally unpaid, to women and productive work to men and arguing for the need for redistribution of labor to achieve gender equality. Furthermore, feminist economics problematizes the victimization narrative in agricultural, development, and environmental politics, which suggests that women are merely passive observers of change. It emphasizes women’s agency—that is, their ability to define goals and strategically act (and react) to achieve them (Kabeer 1999, p. 438; Hackfort and Burchardt 2018)—in order to “challenge their unequal economic, social and political positions and advance towards parity” (Agarwal 2020, p. 842).

This study builds on work done within feminist economics that is concerned with the political economy of agriculture, which places particular emphasis on gender in agricultural labor and crop production decision-making. This includes the access to resources such as independent income, capital, or agricultural technologies; to political or social organizations and institutions, especially those that are postcolonial and patriarchal in nature; and to the ownership of land (Agarwal 1992; Razavi 2003; Tsikata 2016; Prügl et al. 2021). In line with such an approach, in this study, gender inequalities are understood as a matter of political economy. Furthermore, the effects of the commercialization of AIV are analyzed according to their impacts on the gendered division of labor; decision-making power in the household and beyond; social and political organization; and women’s access to resources, as well as their strategies for gaining and maintaining this access. As feminist economics has pointed out, such analyses need to include intra-household interaction and the complex dynamics of “both cooperation and conflict” (Agarwal 1997, p. 4) as important features of gender relations as well as negotiations over decision-making power and the division of labor with regard to the overall economic success of the household (Razavi 2009).

Material and methods

This study is based on a mixed method design encompassing a panel survey and focus group discussions. A survey of farm households was conducted in two counties in western Kenya (Kisii and Kakamega) and two in central Kenya (Kiambu and Nakuru) as part of HORTINLEA, one of our previous larger interdisciplinary research project (Kebede et al. 2018). The panel survey was repeated in three rounds, which took place in 2014, 2015, and 2016. Because key variables such as questions about the gendered responsibility for selling were only included in the 2015 and 2016 surveys, we excluded 2014 data from the main analysis. A multi-stage sampling framework was used to obtain a sample of local farm households. In the first stage, sub-counties and division areas were purposively selected. In the second stage, wards were randomly selected, while farm households were again randomly selected within the wards. Sampling was conditional on households consuming, marketing, or producing AIV (Kebede et al. 2018). Surveys were conducted from September to October using structured questionnaires. Interviews were carried out by one enumerator in person in the local language. Respondents were the most knowledgeable household members on household and farming details which were usually the male household head and/or his wife. For the present study, only households that were observed across all 3 years—685 households in total—were selected.

For the quantitative data analysis, we estimated the effect of changing household responsibilities according to gender using the following linear regression model with household-level fixed effects,

$${Y}_{ivt}={\beta }_{o}+ {\beta }_{1}{G}_{ivt}+ {\beta }_{2}\mathrm{^{\prime}}{Z}_{ivt}+{\beta }_{3}\mathrm{^{\prime}}{X}_{it}+ {\gamma }_{t}+{\mu }_{i}+{\varepsilon }_{ivt},$$
(1)

where \({Y}_{ivt}\) is the share of produce sold by household i for vegetables v by year \(t\). \({G}_{ivt}\) is our key variable, being a dummy variable for responsibility for production (1 = men). \({Z}_{ivt}\) includes crop-specific control variables. \({X}_{it}\) is a vector of household-level control variables. The year fixed effects \({\gamma }_{t}\) control for year-specific shocks, and the household fixed effects \({\mu }_{i}\) control for time-invariant differences across households. \({\varepsilon }_{it}\) is the error term.

Throughout the analysis, we used a fixed-effects specification while reporting the random-effects model results as an additional robustness check due to their higher efficiency. A Hausman test rejected the null hypothesis that the preferred model is random effects. We also regressed our key variable, responsibility for production in cultivated areas, to observe whether additional changes in the production system took place.

While the quantitative data help to understand the patterns of exclusion and their effects on different welfare indicators, more subtle changes in the quality of relations and potential coping strategies are not captured with regression models. To draw a more complete picture of the gendered effects of commercialization, we use qualitative methods.

Qualitative data were collected 6 years later, in 2022, as part of another research project that focused on the social cohesion aspects of AIV production. In total, 20 group interviews were conducted in 10 communities in the counties of Kisii and Kakamega, the same communities where the survey took place. In Kisii, selection of participants was done through the County Agricultural Extension Officers, whereas in Kakamega, it was done through the Butere-Mumias Traditional Vegetables Cooperative Society. Major sub-counties engaging in AIV production were identified and lists and contacts of households and groups involved in AIV production were provided from which potential respondents for the interviews were randomly selected and invited. Eventually, each group consisted of 8–10 women and men, separated by gender, and took place in the same regions and communities as the quantitative survey, though not within the same households. The interviews were conducted by Kenyan enumerators in the national language (Swahili) and later translated into English. A qualitative data analysis (QDA) using the software MAXQDA was conducted on the interviews. All transcripts have been deductively coded by a team of up to three researchers using the analytical categories that are key in the feminist economics framework (gendered division of labor and participation in agricultural labor, decision-making and access to resources, and women’s agency and strategies).

We do not consider the time gap between qualitative and quantitative data collection as problematic because the phenomena under study are longstanding enough to be captured by both types of data. Moreover, we designed the question guide to also reflect changes that have occurred over recent decades, as changes in the AIV production system are likely to be incremental and evolve only over longer periods of time. In addition, the qualitative interviews serve not only to collect more evidence of the effects of AIV commercialization on gender relations but also to test the external validity of the quantitative results.

Results and discussion

Our panel data corroborate our hypothesis, that AIV are increasingly bought and sold. Figure 1 shows that purchases of AIV became more common; for spider plants, for example, they increased from 12 to 18% in only 2 years. It was only for nightshade that sales stagnated. The giving of AIV as a gift was marginally common in the initial year of the study, 2014, but by 2016, it had become relatively uncommon—for amaranth, for instance, the share of total gifted vegetables decreased from 5 to 2% (Fig. 1).Footnote 1 The results clearly show an increase in the purchase of AIV over time.Footnote 2 The increased selling of vegetables that would previously have been shared or given as gifts, is supported by qualitative findings and statements by the respondents—for example, quote 1:

Quote 1: Traditionally, you could hear people tell their neighbors to come and pick some vegetables. That’s what women used to do. Right now, they cannot do that because it translates to money, and this has reduced relations because everyone is now business-minded, and when they look at the AIV, they see money and not friendship. (Man from West Marama, paragraph 131)

Fig. 1
figure 1

Sources of AIVs for household consumption

Full size image

In the following subsections, we present our findings on the gendered effects of commercializing AIV in our regions of study.

Gendered division of labor and women’s participation in agricultural labor

In the past, women cultivated AIV in home gardens and as subsistence crops. The data show, however, that AIV cultivation is no longer exclusively their domain. The panel data show that agricultural commercialization is associated with men taking over AIV production, at least in the cases of cowpea leaves and spider plants (see Table 1).Footnote 3 This is corroborated by the qualitative data, as many respondents from both groups confirm (e.g., as in quote 2) that men, once they see that it is profitable, are increasingly entering the AIV business. Yet interestingly, although men are taking over responsibilities related to production, only very few women report being excluded from the agricultural labor associated with growing AIV. Instead, the findings indicate that women feel supported and assisted in the AIV production process by men taking part in commercialization and taking over responsibilities and duties, as quotes 2 and 3 demonstrate.

Quote 2: Since commercialization, people have stopped referring to the production of AIV as a women’s venture. Now the men have started cultivating and selling these AIV too. Nowadays, the men can come in the farm and help the woman in weeding because they have seen money flowing. (Woman from Getenga, paragraph 19)

Quote 3: It is a relief now that men are assisting us. In the past, every work was done by women, but today it is easy, as we work together. We assist one another. (Woman from Magenche, paragraph 151)

Table 1 Effect of changing gender responsibilities on commercialization—household-level FE models (2015–2016)
Full size table

While both women and men are involved in AIV production, a gendered division of labor characterizes the production of AIV for commercial sales. Even if some discussants denied gender-related assignments of tasks, the data show that, along the value chain, tasks and responsibilities are largely gendered: men prepare the land, clearing, digging, and mixing the soil. Women do the planting, weeding, plucking, harvesting, and selling of the AIV at local markets. Men also spray pesticides, but they are less involved in selling AIV directly to consumers at local markets, which is considered a woman’s task. Yet, as quote 4 indicates, men are involved when it comes to organizing logistics, arranging transportation by motorbike (boda-boda), and selling in larger quantities and at larger scales—to, for example, traders, schools, or hotels.

Quote 4: Yes, they are doing a lot. But we men also do most of the work in land preparation—carrying, transportation and spraying. But harvesting we can also do, but women do it much better. (Man from Bobasi, paragraph 265)

This is confirmed by the quantitative data, which show that men’s selling activity has not increased with commercialization. Table 7 shows no consistent association between men taking over responsibility for selling AIV and the share of AIV being sold, with the exception of a single crop, for which we do find a significant positive association: amaranth. In addition, at a level of 10 percent, the association is only weakly significant. Reasons for men’s low level of selling activity given by both groups are that they “lack the knowledge” to pick the plants correctly, break down what they harvest into small quantities, quantify these, and negotiate prices at local markets. According to the women, they lack the social skills required to profitably market and sell at local markets, as quote 5 below perfectly illustrates:

Quote 5: Who will buy from a man? They can sell cabbages and Sukuma wiki in bags, but how will they even break down the African nightshade in small quantities and sell? They don’t know how to do it. (Womean from Nyacheki, paragraph 194)

Male and female respondents both confirmed that women’s workload has increased with the commercialization of AIV; as they have begun producing for markets, their burdens have mounted. While women have always weeded and harvested, they now also sell their vegetables at markets, which requires leaving the homestead. At the same time, they are still responsible for the provision of care work, including childcare and providing food for the whole family. Hence, there is no evidence for a redistribution of labor within the household (e.g., quote 6).

Quote 6: Work has increased because apart from taking care of the farm and selling AIV in the market, you also have other chores—like taking care of the children—while at the same time, you don’t have anyone to help you. (Woman from West Kisa, paragraph 137)

Even with AIV becoming profitable and men taking over production responsibilities, our data indicate a gendered division of labor along the value chain and no redistribution of productive or reproductive labor; women weed, harvest, and sell the produce at local markets, while men spray pesticides and trade only at greater distances. This can be interpreted in line with indications gleaned from our data that even though men have entered the realm of AIV production, women do not feel excluded from agricultural labor; instead, they see men’s activity as assistance. At the same time, they do not perceive a significant decrease in their own workload. Regarding this latter point, it can be concluded that when men get involved in AIV production, they take over less time-consuming tasks, such as logistics and spraying pesticides, while women retain the more time-consuming tasks of agricultural labor. Women’s tasks include care work in the homestead and possibly also the background work of prepping less-visible farm-related tasks, such as arranging or cleaning equipment.

This resonates with findings from other regions of the world, which show how women in India, for instance, actively work in the background of agricultural production, carrying out “invisible” preparatory and pre-production work related to activities taking place both on and off the farm. As a result, women’s labor burden can increase considerably without being recognized as having done so (Pattnaik and Lahiri-Dutt 2022). This mirrors the general feminist dilemma represented by the tension between, on the one hand, the demand that women be enabled to economically empower themselves through participation in the “productive economy" and, on the other, the increase in workload caused by additional wage labor—on top of the reproductive and care work women are typically responsible for. When women enter the productive sphere without men taking on reproductive work in turn, their workload significantly increases. Consequently, demanding a redistribution of labor and involving men more extensively in care work are key concerns of feminist economics; as our data show, these issues are relevant in our case study region, too. “Just slotting women in” (Agenjo-Calderón and Gálvez-Muñoz 2019, p. 148, italics in original) may help to achieve inclusion in an economy with an unjust, gendered labor market, but it does not help achieve gender equality. Toward that goal, feminist economics highlights another key concern: redefining the concept of labor. Gender equality requires (in addition to a redistribution of labor) a recognition of the myriad activities through which women provide for and contribute to productive labor; it requires that women’s invisible agricultural labor, for instance, become visible (Fraser 1995; Pattnaik and Lahiri-Dutt 2022).

Women’s decision-making power and access to resources

On the question of whether women gain or lose decision-making power when men take on responsibilities and duties in relation to AIV production, the quantitative data are not unambiguous. For instance, the quantitative analysis shows, at least for cowpea leaves and spider plants, that agricultural commercialization is associated with men taking over decision-making in the production (but not selling) of AIV (see Tables 1 and 7 in the appendix). The same holds for the qualitative data, which reveal that as AIV are successfully produced for sale, men increasingly begin to see their economic potential and get involved in the production and decision-making processes, including deciding what to plant and where and how to spend any revenue, as quote 7 demonstrates:

Quote 7: In most cases, it’s the man who decides where the AIV and other crops will be planted. As a woman, you cannot just farm anywhere because you will ignite a conflict. Before commercialization, you could just plant anywhere—only for consumption purposes—but currently, you can’t just plant anywhere because he knows that’s your money. So you have to discuss and agree on where you will do your AIV production. (Woman from Central Marama, paragraph 188)

How this influences women’s effective decision-making power seems to depend on the arrangements within individual households. According to some women, unlike in the past, men now want to have a say in how revenue from AIV, as additional household income, is spent. Others say that they decide on their own what to buy or invest in and then inform their husband about their decisions. Some respondents from both the female and male groups, meanwhile, report discussing and making decisions about these kinds of questions as a couple, with both parties having an equal say, as in quote 8.

Quote 8: If I go to the market and sell the vegetables, I will buy [a] few items. When I go back home, I will tell my husband I got 500 [Kenyan] Shillings after selling the vegetables. I have spent 300. Here are the remaining 200, which we discuss how to use. (Woman from Bobasi, paragraph 222)

However, a large number of statements indicate that after commercialization, women’s decision-making power in regard to the use of income increases. Even with men entering the AIV business, a number of women declare that they have gained decision-making power, freedom, and the ability to autonomously decide how to spend the money they earn (as in quote 8). This is likely to be linked to the quantitative finding that while men have increased their decision-making power in relation to production, they have not done so with regard to the selling of AIV at markets; at this stage, women have direct control over the income earned. Through this access to financial resources, they perceive an increase in their freedom to decide whether to use the money for household purchases, farm inputs, or livestock and even how to use the land—including whether to increase the amount used to cultivate AIV, as quote 9 notes.

Quote 9: They [women] have gained decision-making responsibilities in the household. They can now suggest to the man to increase the size of farm that they used to cultivate AIV. (Woman from West Marama, paragraph 28)

For many women, selling AIV is a strategy to diversify their income and livelihood. These women articulate having gained economic independence with the additional income they earn through commercialization. They state that they have “‘gained decision-making responsibilities” (as in quote 9) and perceive themselves as economically empowered, as they have their own income. They feel that they have more freedom and decision-making power (see quote 10), which contributes to a move toward more gender equality in the household.

Quote 10: When we cultivate the AIV and start selling, we, the women, gain more rights in decision-making within the household. We can contribute in deciding the size of land that will be used for AIV farming and also the type of seed to be planted. (Woman from Kholera, paragraph 175)

These women broaden their scope of action by not relying on their husbands to give them money for household expenses. With the income generated, they contribute to their household income and thus help make life easier and investments possible. In some cases, men even seem to rely on the money that women now contribute, as quote 11 indicates.

Quote 11: I can even give my husband some pocket money, like 200 Shillings, therefore bringing peaceful coexistence. (Woman from Getenga, paragraph 11)

Overall, the qualitative and quantitative results suggest that by retaining control over selling AIV, women perceive no loss in decision-making power; at markets, they receive money directly from their customers and hence control the monetary reward to some extent. In addition, if men mostly enter AIV production through improved household cooperation, by this same route, women can gain access to resources that are under the control of men (such as land) or develop strategies to cope with new allocations of household responsibilities (see quote 9).

Cooperation in the household and beyond

Cooperation between men and women as household members—with regard to AIV production and the use of revenues—is a recurrent theme in the qualitative data. Despite the gendered division of labor, AIV production is perceived as a common and cooperative activity that contributes to the economic success of the household. It is pursued as a strategy of income diversification in which it is in the best interest of the family that both men and women contribute, participate, and decide. Quite a number of couples decide together how to satisfy household, family, and individual needs; what the money should be spent on or how it should be saved; and how to use the land, as quotes 12, 13, and 14 show.

Quote 12: Personally, I do a lot of AIV farming, but we both sell and we put all the monies together for common use, so in my household, my woman is fine. Additionally, we mutually agree on the items to spend money on. (Man from west Marama, paragraph 99)

With men engaging in AIV production, households are more likely to better allocate resources (including material inputs and land) between different crops. Indeed, households seem to increase the area of the land cultivated for AIV if men assume production responsibilities (see Table 2), which is in line with anecdotal evidence from previous studies (Ferenczi 2021). In addition, qualitative evidence also shows that access to and control over the use of land is an important part of household cooperation, as quotes 13–14 illustrate.

Quote 13: It will bring more benefits if they get involved because he can even cut on maize production and increase the portion of land on which to plant AIV, and that will make it more profitable. (Woman from Khalaba, paragraph 280)

Quote 14: For the land they [the couple] want to cultivate indigenous vegetables for commercial purposes; they sit down and discuss the size of the land to allocate to indigenous vegetables production, and also they discuss the selling of the vegetables. They decide together. (Man from Kholera, paragraph 194)

Table 2 Effect of changing gender responsibilities on cultivated area—household-level FE Models (2015–2016)
Full size table

Cooperation and the resulting improvement in a household’s economic situation are expected to benefit the whole family. Moreover, many women highlight the economic independence they have gained as being conducive to more harmony and peace and less conflict in the household and within the family.

Examples of household cooperation aside, however, in cases where there is no intra-household cooperation, many women develop individual coping strategies. Some admit to lying to their husbands in order to keep their savings, in case they or their families need them, as quote 15 shows.

Quote 15: Sometimes, we lie about the sales from the vegetables. If, for instance, after buying few items and I am left with 200 Shillings, I will tell him I am left with 150 and keep the 50 undisclosed. I hide it in my home bank. The savings come in handy in times of need. (Woman from Bobasi, paragraph 281)

This resonates with feminist studies that emphasize how intra-household dynamics encompass conflictive and cooperative arrangements simultaneously when access to, use of, and control over resources and the distribution of work, time, and responsibilities are negotiated (Agarwal 1997; Agenjo-Calderón and Gálvez-Muñoz 2019). This shows the complex interplay between the “common interests that all household members have in the overall economic success of their households” (Razavi 2009, p. 208) and other, less cooperative individual strategies that women develop.

To summarize, when AIV are successfully produced for sale, men get more involved in their production and any associated decision-making. However, the data show that this does not necessarily translate into decreased decision-making power for women. Instead, it suggests that the increased commercialization of AIV benefits women in the form of access to financial resources and, indirectly, an increase in their influence over production inputs such as land. With the revenues they obtain from the production and selling of AIV, women gain access to income streams that not only diversify their livelihoods, ensuring food security, but also further contribute to their economic empowerment and, eventually, to the strengthening of their decision-making power. In the literature, this latter effect is considered a main indicator of bargaining power within the household and beyond (Agarwal 1997; Acosta et al. 2020; Orr et al. 2021). From a feminist economics perspective, these dimensions are key to achieving gender equality and reducing the social exclusion of women (Kabeer 1999).

However, it must be noted that these strategies are limited with regard to overcoming gendered inequalities inherent in the structures of patriarchal societies. This is illustrated by the fact that when men decide to dedicate a greater share of land to AIV production (as quote 13 mentions), women must negotiate with them to access it; land is not a resource that women are able to independently access or gain control over. Land ownership and control are ultimately patriarchal, and men have the final say over land use, including the share that is dedicated to AIV production. Some women declare that should men enter into and take over AIV production, they themselves will withdraw from it and venture into other businesses—for instance, selling other vegetables, cows, or hens; poultry production; or manual labor, even if these activities may be less profitable and more labor-intensive than commercial AIV production.

Furthermore, many women use collective strategies such as social organizing in small groups to pool resources and support each other through, for example, financial loans. These banking and savings cooperatives, called “merry-go-rounds” (chamas in Kiswahili), are very popular among women in these regions. Some women were first able to diversify their livelihoods with these group savings; they enable them to lease land together, for example, or invest in a cow to eventually produce manure to boost AIV production.

Quote 28: It has brought women together through formation of women groups, whereby they lease land as a group, prepare the lands for cultivating the AIVs together, weed together and harvest the AIVs together. (Woman from Getenga, paragraph 19)

Forming groups also helps them to build relationships along the AIV value chain to better access markets beyond the local village by setting up a system of AIV distribution. The groups also enable them to access agricultural inputs like energy or machinery through the acquisition of external funding, as well as obtain political resources through government support (from agricultural extension officers, for example). As a result of past agricultural reforms (which were part of neoliberal structural adjustment policies), governmental regional agricultural extension officers reach out only after communities demand that they do so, meaning that only organized groups are able to benefit and obtain information and support from these officers.

In this way, women develop individual and collective strategies to diversify their livelihoods and gain more economic independence. They build networks of solidarity and support as a self-help strategy that is conducive to, simultaneously, political empowerment and economic independence. With regard to the effects of commercialization, women successfully use these strategies to meet their immediate daily needs. In feminist scholarship terms, they can be understood as “practical gender interests”—as opposed to “strategic gender interests” (Molyneux 1985, p. 232–233), which aim much more at the transformation of patriarchal structures as such (for example, collective struggles to obtain land rights instead of just accessing land via a husband). These practical interests are legitimate and essential for women who experience the contradiction of, on the one hand, being part of a household and cooperative coalition with their husbands while, on the other, needing to find ways to, as women, overcome dependence and foster empowerment in a patriarchal society. Our data suggest that these women’s strategies are also fundamental to their agency and to determining their intra-household bargaining and decision-making power. They contribute to women’s empowerment, defined as the ability to make life choices enabled by access to material and immaterial social resources, agency, and achievements that relate to well-being (Kabeer 1999). It becomes apparent how women, against the backdrop of the commodification of their crops and subsistence labor, rely on such collective strategies to secure their livelihoods and how monetization requires such informal strategies to function at all.

Conclusions

This study examines the commercialization of AIV and analyzes how it affects gender relations, that is the division of labor, decision-making power and access to resources, and cooperation strategies. Feminist economists have criticized the commodification of formerly non-market transactions and goods and pointed out that “women’s integration in the labor market has not implied automatic liberation, even if, from a historical and comparative perspective, it has served to challenge rather than consolidate patriarchal structures.” (Agenjo-Calderón and Gálvez-Muñoz 2019, p. 159). It is therefore important to analyze how such processes affect gender relations in regard to resources, rights, and responsibilities and the kinds of conflicts and cooperation women and men engage in to balance their needs as both individuals and household and family members.

This study shows that while commercialization is conducive to men’s increased involvement in AIV production and related decision-making, this does not necessarily translate into decreased decision-making power for women. If women retain control over selling vegetables (and hence, keep the resulting income in their own hands), they maintain and even enhance their bargaining and decision-making power. This specific finding—that women in the study region are able to retain control over income—could potentially have been a key factor in avoiding the detrimental effects of commercialization that are reported in the literature (Fischer and Qaim 2012). The results indicate that more attention should be paid not only to which tasks are redistributed among men and women but also to those tasks’ inherent characteristics, including working conditions, their effects on health, and control over income and land use.

Furthermore, household members cooperating to reach individual or common goals can also be perceived as support, as men taking over what have traditionally been women’s tasks. Yet, rather than just defending their domain and developing strategies for (for instance) retaining the monetary reward for attaining their goals, women also develop strategies for withdrawal. This shows that the binary of household cooperation versus household conflict—and the immediate assumption that there is no household cooperation—cannot capture complex intra-household dynamics.

The results corroborate findings from other studies, which show that where women’s bargaining power is stronger, agricultural or horticultural commercialization is likely to result in more equal benefits for both women and men (Orr et al. 2021). Apart from social norms, the legal institutions of the labor market, and state policies, it is largely power relations and women’s bargaining power within the family that determine whether and to what degree women perceive and experience increases in their decision-making power and overall empowerment through the commercialization of “their” crops. This indicates that “context determines who wins from commercialization” (Orr et al. 2021, p. 251).

Yet, against this rather optimistic assessment for some settings, it should be noted that women in our study still suffer from a growing workload; more and more labor goes toward the selling of AIV, while at the same time, they still have to perform most care work. Also, with regards to land ownership as a key concern from a feminist economics perspective, the results are ambivalent—the commercialization of AIV replaces the subsistence farming of AIV when land availability is a bottleneck for production, as is the case for women in Kenya, whose access to land and possession of land titles are limited due to patriarchal inheritance practices. Thus, our study illustrates that women's experiences vary and that overly simplistic and dichotomous assessments of the effects of commercialization as generally positive are not valid.

To conclude, while we see the economic empowerment of women through commercialization—how they broaden their scope of action and are empowered by generating revenue—that process does not contribute to a redistribution of labor or land rights, which are key for gender equality. In contrast, women’s labor burden has increased. Moreover, with the commodification of subsistence work and previously subsistence crops as common goods, social practices (selling instead of gifting or sharing, for example) may be entirely replaced by market-based forms of interaction. To what degree this might erode the bonds of solidarity and social cohesion needs to be further explored.