© B. Dedieu et al., Hosted by EDP Sciences 2025

1 Introduction

Family farming is still considered a pillar of agricultural development, addressing challenges such as poverty, food security, and employment (FAO and IFAD, 2019). Family farms represent 90% of the total number of farms and produce 80% of the total food in value terms (Lowder et al., 2021). In Africa, more than anywhere else, the agricultural sector plays a crucial role in employment (Davis et al., 2023). Given the continent’s current demographic growth and the moderate contribution of other economic sectors to labor absorption, particularly in rural areas (Hazell et al., 2024), the question of employment prospects in a changing agricultural sector is a major issue. Numerous studies across various disciplines are exploring this topic (e.g., Christiaensen and Maertens, 2022; Losch, 2022). Beyond job prospects, research also points to the urgent need for better data on decent work—understood as the multidimensional aspirations of people in their working lives—across the broader agri-food system (Meemken et al., 2024). This knowledge gap includes the characterization of the diversity of agricultural jobs resulting from different work organization strategies on farms, which has received relatively little attention.

Family farming is characterized by the specific interaction between the domestic and economic spheres. This connection is materialized through "the inclusion of productive capital in the family patrimony and the combination of domestic and market and non-market logics in the processes of allocating family labor and its remuneration, as well as in the decisions regarding product distribution between final consumption, intermediate consumption, investment, and accumulation” (Bélières et al., 2015). However, the organization of family labor—how it is structured by farmers throughout the cropping and livestock calendar (who does what, when, and where − Madelrieux et al., 2009), the roles of men and women, family and non-family workers, and the type of tasks (daily or seasonal) that each worker takes on, along with related working conditions—remains insufficiently described in literature on Africa (Dedieu et al., 2025). In their review of the literature on “work and employment in agriculture in Africa,” these authors highlighted the lack of exploration and knowledge regarding the diversity of work organization patterns at the farm level in Africa. Several studies compiled by Cournut et al. (2018) suggest that there is a diversity of work organization forms behind the "family farm" in both the Global North and South. These work organization forms define the task profiles of various workers, which can be examined in terms of job satisfaction (Sabillon et al., 2022). In fact, the nature of the job itself is a significant determinant of job satisfaction and can be characterized by skill variety, task identity, task significance, autonomy, and feedback (Hackman and Oldham, 1976). This directly relates to the intention to establish or remain as a farmer or wage worker (Santhanam et al., 2024). These factors are becoming increasingly important when assessing the sustainability of agricultural systems, particularly their social sustainability (Isaac et al., 2024).

Although the African continent will see nearly 250 million young people entering the labor market by 2050, perceptions of agricultural activities, opportunity cost of working in the city, as well as aspirations and access to resources (notably land), will continue to influence youth participation in the sector. In particular, there is a potential mismatch between the aspirations of rural youth and their involvement in agriculture (Sumberg, 2021). Given the limited and often precarious job opportunities in urban areas (ILO, 2024a), better understanding the diversity of agricultural jobs is key to achieving economic transformation linked to the growing rural economy (Hazell et al., 2024).

Numerous factors are contributing to changes in the decisions that match labor demand with supply (Dedieu et al., 2022): enlargement, changes in agricultural models (Volken and Bottazzi, 2024), downstream governance (Malanski et al., 2022), mechanization (Daum and Birner, 2020), digitalization (Hostiou et al., 2023), labor markets, and public policies (Chinsinga and Chasukwa, 2018). All of these factors impact who does what, and under what conditions, which can be analyzed by updating current forms of work organization.

The aim of this paper is to produce knowledge on the diversity of work organization patterns on family farms in different rural contexts across Africa, based on a series of surveys conducted between 2022 and 2023 with 438 farms in 5 countries (Tunisia, Burkina Faso, Senegal, Tanzania, Madagascar), covering 14 diverse farming systems (crops, livestock, and mixed) and socio-ecological environments. The goal is not to cover the entire diversity of ecosystems in Africa, but rather to identify patterns of "who does what" that could be generalized (Andrieu et al., 2025). This will then allow for a discussion of the workers’ task profiles related to these patterns. This study was conducted as part of the ‘Socio-economic viability of agroecological practices in Africa’ project coordinated by CIRAD and ICRAF (Andrieu et al., 2025).

2 Conceptual framework and methods

2.1 The work organisation framework

Work organization involves the dynamic interaction between the size and combination of agricultural and non-agricultural activities, their content (technical sequences, i.e., the tasks to be carried out), and the equipment (both for crops and livestock). Work organization is influenced by i) the size and nature of the workforce (family, wage—permanent and temporary, contractors, and mutual aid), ii) the distribution of work among people on the farm (Dedieu and Servière, 2012). All of this is shaped by (i) farmer’s technical and economic objectives (yield, autonomy, economic margin), and (ii) work-related objectives: labor productivity, the meaning of work (Spoljar et al., 2024), control over time, and conditions that promote health and well-being for each worker. Although farmers are often at the centre of work analysis, it is crucial to consider all categories of workers (men, women, family, and non-family) who are engaged into farm work.

The list of tasks involved in implementing cropping and livestock practices on farms is very diverse and difficult to handle in comparative analyses, unless the production systems are highly homogeneous (see one example in Appendix 1). Tasks vary according to the skills and qualifications required, the work environment (indoors or outdoors, with or without machinery), as well as their pace, their potential for being combined with other tasks, and their degree of postponability. For this reason, two categories of tasks can be considered. Some tasks are repetitive and cannot be postponed or aggregated over time (e.g., milking ruminants). These tasks must be performed daily; even if completed on a given day, they need to be repeated the next. These are referred to as “routine work” (RW). Other tasks are more flexible (with a range of days available for completion) and can be carried out in fewer days if more people are involved (e.g., ploughing or harvesting). Such tasks are tied to specific stages of the agricultural cycle and occur only at certain times of the year. These are referred to as “seasonal work” (SW). The gender distribution of tasks is also a crucial point of work organization analysis, considering the interactions between domestic dans farming activities (Adeyeye et al., 2021), and the contribution of women as temporary salaried workers (Jehouani and Naji, 2023). The nature of the above tasks can also be analyzed in greater detail. For example, within the daily routine work of ’caring for dairy cows,’ a distinction can be made between milking, feeding and cleaning the barn. It is also possible to compare the technical sequences of different crops, with specific tasks (e.g., irrigated rice vs. rainfed rice, cassava vs. maize).

Our objective is to describe the patterns of work organization that emerge from the categories of tasks performed by different categories of workers.

2.2 Data collection

2.2.1 The farming systems studied

The study brings together five research teams from five countries in Northern and sub-Saharan Africa (Tunisia, Senegal, Tanzania, Burkina Faso, and Madagascar). All these teams participated in the ‘Viability’ project (Andrieu et al., 2025) and volunteered to test the proposed Quaework methodology (Hostiou and Dedieu, 2011 see below). Their participation contributed to a pan-African cross-sectional analysis of the diversity of work organization patterns on farms.

In each country, the team selected a study area and identified different farming systems within it. The identification of these farming systems drew on a previous survey (part of the Viability project) that collected information on farm structure and agroecological practices (see Tab. 1 and Box 1). These farming systems differ according to their location in diverse agroecological zones or according to the level of implementation of agroecological practices (e.g., varying intensities of cotton production in Burkina Faso, or differences in the use of organic fertilization in Tanzania). In Senegal, farming systems are differentiated according to herd size, the practice of transhumance, and the extent to which crop production is integrated alongside livestock.

2.2.2 Farm surveys

The data used for this analysis were collected from datasets compiled by the country team leaders within the TPP Viability project. Data were collected in 2022-2023 through the QuaeWork method (Hostiou and Dedieu, 2011), which is designed to characterize work organization (i.e., who performs which tasks on the farm) and to quantify work durations on an annual scale. This method collects information on the duration of work tasks and the distribution of these tasks among various types of workers. The method considers different categories of workers, such as the farm manager (farm head, who can be either a man or a woman), other family members, permanent wage workers, and temporary wage workers. It also accounts for gender (men and women) for each of these types of workers. The method differentiates between two categories of work: i) routine work (RW), which is quantified in hours per day, and ii) seasonal work (SW), which is quantified in days per year.

2.3 Work organization data analysis

2.3.1 Comparison of working times and work organization between farming systems

Each research team used an Excel file for data entry, structured into three main types of sheets:

• a dictionary of variables describing the variables (name, description, type, etc.);

• individual sheet for each surveyed farm, compiling information on routine work (RW) and seasonal work (SW), disaggregated by gender and categories of workers;

• a summary sheet consolidating data from all surveyed farms withing the corresponding farming system.

The analysis was carried out using descriptive statistics (mean, standard deviation, min, max) through the statistical programming language R and Excel software to generate figures.

The work organization analysis focused on the distribution of the annual duration of RW and SW for the different categories of workers (farm manager, family workers, permanent wage workers, temporary wage workers). The gender-based analysis provided a detailed description of the work distribution between men and women workers, considering the type of work (routine or seasonal) and all categories of workers (family and wage labor).

2.3.2 Cross-sectional analysis based on multivariate analysis

A principal component analysis (PCA) was conducted to build a typology of work organization patterns across all farms. The objective was to identify work organization patterns and to identify the variables that explain differences between them. The type of work (routine/seasonal) and the categories of workers (farm managers, other family workers, permanent wage workers, and temporary wage workers) were used as active quantitative variables in the PCA and subsequent clustering (see Tab. 2). Supplementary variables −farming system, number of Tropical Livestock Units (TLU), and cultivated area (hectares) − were added to characterize the identified types. This approach allowed us to determine whether we can see a relation between work organization pattern and types of farming system, or farm size (TLU or cultivated area).

To account for differences in farm or herd size across farming systems, we used relative rather than absolute values. Each value was expressed in relation to the mean of the corresponding farming system (plus or minus x standard deviation). These analyses were performed on the dataset combining all farms (438 in total) from the five countries and 14 farming systems. The FactoMineR package was used to conduct the analyses.

3 Results

3.1 Description of the farming systems

There are major differences between farming systems, in terms of cultivated areas, livestock size and level of mechanisation. Family farm labor is also very heterogeneous.

3.1.1 Livestock size: pastoral herds, backyards, systems with one or two dairy cows in stalls

Farming systems in Senegal, the pastoralism farming system in Tunisia and the two farming systems in Burkina Faso display the highest number of TLU (from 6 to 53 in average) the other farming systems raising less than 2 TLU in average (Fig. 1a), with mainly suckler animals shepherded every day. In Senegal the large herds consist of cattle and small ruminants (sheep and goats) often shepherded in separate batches. The traditional management is based on transhumance for a part but more rarely for all the herd. With the development of dairy industries in the pastoral zones, some cows are milked by the family. In Tunisia, herds are mainly composed of sheep and goats shephered in drylands. Madagascar farmers (MDG 1 Highland et 2 Lowland) raise, beyond backyard animals, one of two dairy cows in a barn, providing regular cash. With the animals, routine work consists of milking, cleaning, cutting and carrying the forages.

Fig. 1 Average number of tropical livestock units (TLU) and agricultural area (ha) per farming system. (a) Average number of tropical livestock unit (TLU) per farming systems. (b) Average cultivated areas (hectares) per farming system. Nombre moyen d’unités de bétail tropical (UBT) et surfaces cultivées (ha) par système agricole (a) Nombre moyen d’unités de bétail tropical (UBT) par système agricole. (b) Surface moyenne cultivée par système agricole (ha).

3.1.2 Cultivated land size: very small areas in general except in crop − livestock systems in Tunisia

Two farming systems in Tunisia have the largest cultivated areas, 15.2 and 8.6 hectares respectively for crops and olive trees (Fig. 1b). Gardening when it exists concern small areas. In contrast, the Madagascar, Senegal and the Tunisian pastoral farming systems have very small cultivated areas (less than 2,2 ha in average). In Madagascar, the very small areas (less than 1 ha) are due to land pressure. In these farming systems, most of the cultivations are subsistence crops. In Burkina Faso, the areas are rather similar even if the presence of cotton differentiates the systems in term of work demand.

3.1.3 Mechanization

The use of tractors or draught animals varies greatly in the sample, with around 1/3 of farms using a tractor, 1/3 using animal power and 1/3 relying solely on manual labor. Mechanization is present in the agropastoral and fruit/gardening farming systems in Tunisia, farmers own or rent tractors for ploughing and performing some other tasks (such as harvesting wheat for instance). At the opposite, in Madagascar farming systems agricultural activities are mainly manual. In Burkina Faso, the use of draught cattle allows to cultivate large areas for cotton production.

3.1.4 Engagement of the family into agricultural work

The size of the families involved in farm work varies greatly between countries and farming systems: two permanent family workers per farm on average, with variations ranging from highly endowed farms in Burkina Faso (around four permanent family) to less endowed farms in Madagascar.

3.2 A diversity of working times according to the farming systems

The results show a high variability of work duration between farming systems (Fig. 2). In all farming systems, more than 80% of routine work (RW) is composed of livestock tasks (e.g., feeding animals, shepherding, cleaning, sometimes milking). RW also includes tasks related to cropping system such as water management (irrigation especially in the farming system “Gardening and fruits” in Tunisia) and crop protection (e.g., bird scaring¹ for rice production). The amount of RW highly varies between farming systems from 448 h per year in the “Maize mixed more agroecology” farming system in Tanzania to more than 10.000 h per year in large breeders who practice transhumance in Senegal with several batches shepherded in parallel all the day long (Fig. 2a). Beyond the herd size and the farming system factors, differences can also be explained by crop routine work (use of irrigation for example) and farm sizes.

Nota: The farming systems in Senegal were not taken into account in this figure due to the difficulty of defining the seasonal work with the herders.

Seasonal work (SW) is mainly composed of tasks related to crops (land preparation, sowing, weeding, harvesting, etc.) in all farming systems. Some tasks are also related to livestock in “Pastoralism” farming system (Tunisia) and in both farming systems in Burkina Faso with tasks such as moving the animals or weaning the young animals. The amount of SW duration highly varies according to the farming systems. The differences are higher than for RW: the higher value of seasonal work duration (about 1100 days per year for “Cotton-cereal-legume with livestock” in Burkina Faso) is more than six times higher than the lower value of seasonal work (about 161.5 days per year for the farming system “Maize mixed more agroecology” in Tanzania) (Fig. 2b). The size of the farm contributes to explain these differences between farming systems (farms have larger cultivated areas in Burkina than in Madagascar), but also the types of farming systems. For example, harvesting cotton in Burkina Faso is more labor demanding compared to harvesting maize (the major crop of farming systems studied in Tanzania). The use of mechanization is also a factor explaining the variability of SW: even if two farming systems in Tunisia (agropastoral and fruit/gardening) have the larger cultivated area, the rent of tractor for ploughing or harvesting wheat allows to reduce SW working time.

Fig. 2 Inter farming systems variability of annual routine and seasonal work durations. (a) Inter farming systems variability of annual routine work duration for crops and livestock expressed in hours per year. (b) Inter farming systems variability of annual seasonal work duration for crops and livestock expressed in number of days per year. Variabilité de la durée annuelle moyenne du travail d’astreinte et de saison annuel entre systèmes agricoles . (a) Variabilité de la durée annuelle du travail d’astreinte entre systèmes agricoles (heures /an) (élevages et cultures). (b) Variabilité de la durée annuelle du travail de saison entre systèmes agricoles (jours / an) (cultures et élevage).

3.3 Work is performed by the family and wage workers

Different categories of workers contribute to RW and SW (Fig. 3). There are few differences between farming systems within a same country. The differences are greater between countries. RW is mainly carried out by family workers, notably by the head of the farm in Tunisia and Tanzania and by other family members in Burkina Faso, Senegal and Madagascar (Fig. 3a). For both farming systems in Burkina Faso, it is exclusively managed by family labor. In all other farming systems in the four other countries, wage workers are also engaged in the RW: permanent wage workers in all farming systems in Madagascar and in Tunisia, or temporary wage workers in Senegal and in one farming system of Tanzania (less agroecological).

Nota: The farming systems in Senegal were not taken into account in this figure due to the difficulty of defining the seasonal work with the herders.

For the SW, the main difference is the contribution of temporary wage workers even if their contribution varies across farming systems (Fig. 3b). In Tunisia, especially in “Gardening and fruits” and “Agropastoral TN” farming systems, temporary wage workers carry out most of the SW. Farmers hire temporary workers because the cultivated farming systems are very large with labor demanding crops (market gardening) and family members alone cannot do all the work, as they are engaged in the RW (for example in the Tunisian farming systems). In Madagascar, SW is shared between farm managers, other family members and temporary wage workers, even if farm sizes are very small. In both farming systems in Burkina Faso, in Tanzania and in Pastoralism farming system in Tunisia, the seasonal work is carried out by family members, especially farm managers. A common point to all farming systems is that permanent wage workers rarely contribute to SW. Actually, few farmers hire permanent workers.

Fig. 3 Inter farming systems variability of routine and seasonal work distribution between categories of workers expressed in %. (a) Inter farming systems variability of routine work distribution between categories of workers expressed in %. (b) Inter farming systems variability of seasonal work distribution between categories of workers expressed in %. Variabilité, inter systèmes agricoles, de la répartition moyenne du travail d’astreinte et du travail de saison selon les différentes catégories de travailleurs (en %). (a) Variabilité inter systèmes agricoles de la répartition moyenne du travail d’astreinte selon les différentes catégories de travailleurs (en %). (b) Variabilité inter systèmes agricole de la moyenne de répartition du travail de saison par catégorie de travailleurs.

3.4 A high contribution of women to work

Research findings also reveal a contrasted participation of women in agricultural activities in all farming systems (Fig. 4). Women are represented among all workers’ status, except in Burkina Faso where no women wage workers were recorded. The share of work by men family workers exceeds half of the total working time on both farming systems in Burkina Faso, as well as in the Agropastoral 1 and Tree crops farming systems in Madagascar. In Tunisia and Senegal, we observed a higher contribution of women (in comparison with men) especially women wage workers in Tunisia and women family workers in Senegal. The share of work by women family workers also exceeds half of the total working time on farming systems with large transhumant herders without crops, as well as on farming systems with medium-sized breeders who do not practice transhumance or agricultural activities in Senegal. A specificity of the two farming systems in Tunisia (Gardening/fruits and Agropastoral) is that around 30 to 40% of total work is performed by women temporary workers.

Fig. 4 Inter farming systems variability of category of workers and gender contribution to the total work. Total work = RW / 8 + SW in days (conventionally 8 h are equivalent of one day). Variabilité inter système agricole de la contribution des hommes et des femmes au travail total, par catégorie de travailleurs.

3.5 Diversity of work organization patterns

The typology of work organization patterns across the 438 farms surveyed was derived from a PCA based on the contribution of different categories of workers in the routine and seasonal work. The variables used in the analysis are presented in Table 2. The first two dimensions (axes)together account for 44.3% of the total variance between farms. These first two dimensions were retained for further analysis, as their eigenvalues exceed the threshold of 1, in line with the Kaiser’s criterion. The first axis alone explains more than 23.3% of the variance and so plays a crucial role in summarizing the majority of information.

From the PCA and subsequent clustering (see the correlation table in Appendix 2), we identified three patterns of work organization (Tab. 3).

B. Per organization pattern : percentage of annual seasonal work made by each category of worker

• Type 1 “Seasonal and routine work mainly carried out by farm managers”. This type concerns 40% of farms and is characterized by considerably high proportions of routine work and seasonal work carried out by the farmers themselves. The farms displaying this type of work organization are of average land size.

• Type 2 “Routine work mainly carried out by family workers” gathers 44 % of farms. The contribution of family workers to seasonal work is highly variable and therefore does not statistically differentiate this type. This work organization pattern concerns farms with high number of tropical cattle livestock units (12,8 TLU on average), which place a substantial burden on routine work.

• Type 3 “Routine work carried out by family and permanent wage workers and seasonal work mainly carried out by temporary wage workers, mainly women”. This work organization pattern gathers15 % of farms. Il concerns farms that have large cultivated areas, reaching up to 9,5 hectares, with an intermediate TLU number.

The correspondence between farming systems and types of work organisation, detailed in Table 4, falls into two categories. In some cases, there is a near-perfect match between and a farming system and a single type of work organisation. In other cases, farms within the same system are split across two or three types of work organization. For example, the farms in the two farming systems in Burkina Faso are divided between type 1 (the farm manager is the main contributor to routine or seasonal work) and type 2 (the family is involved in routine work). Farming systems in Madagascar are divided between the three types, with type 2 dominating (family involved in routine work). By contrast, work organisation in all the farming systems in Senegal is almost exclusively family-based (type 2) while in Tanzania it is mainly in type 1. In Tunisia, type 3 (characterized by a heavy reliance on temporary labor for seasonal work) predominates in gardening and fruits and in agropastoral farming systems.

4 Discussion

4.1 About the methodology

Approaching work durations and organization on a yearly basis provides a high level of precision in respondents’ declarations as it aligns with the technical calendar of their farming system. The estimated margin of error is about half an hour per day for routine work and half a day per month for seasonal work, based on comparisons with time budget records (Dedieu and Servière, 2012). QuaeWork is an analytical approach that relies on farmers’ recall of events from the previous year, specifically the previous agricultural season. This method is based on two key reference points: i) the seasonal task calendar, which serves as a framework for guiding discussion with farmers, 2) routine work, which, due to its regularity, can be easily described by farmers when recounting a typical day.

However, for the Viability project, data collection occurred during the post-COVID period, with most teams receiving only virtual presentations of the methodology (principles, survey guide) and the Excel file for data entry. The exceptions were Tunisia and Madagascar, where physical training sessions, including on-farm surveys and data analysis, were organized with the coordination team. The QuaeWork principles were new to all the teams. The precision and homogeneity of the data were considered high after a validation step to ensure consistency, followed by dialogue between the coordination team and the national teams.

Of course, such a transversal analysis of very contrasted situations has its limitations, particularly with regards to generalization. What this research provide are insights into both the diversity and the internal consistency of work organization patterns, which should be further validated through larger datasets and broader coverage of African farming systems (Andrieu et al., 2025). While this study focused family farms, it would be interesting to extend the investigation to commercial large-scale farms, which are likely to exhibit different work organization patterns due to their greater reliance on temporary and permanent wage workers.

4.2 A diversity of work organization in farms

Our results highlight three work organization patterns, combining workforce (farm managers, family workers, and / or wage workers) in relation to the category of work (routine and seasonal) characteristics of tasks or the type of agricultural activity i.e., crops, gardening, livestock. These two elements are closely associated. Similar observations were made by Cournut et al. (2010) when comparing various livestock farming systems in Northern and Southern countries, and by Cournut and Chauvat (2012) in France. Our results confirm previous findings that family farmers often need to hire workers, particularly temporary workers for seasonal tasks related to crops, notably in Africa where farming operations are rarely mechanised. Hiring temporary workers is when family labor is insufficient or when the workload is heavy due to large cultivated areas (Wright et al., 2012; Hostiou et al., 2012; Oya and Pontara, 2015).

The importance of women’s contributions is also confirmed and quantified, particularly among temporary workers. The use of temporary workers can offer advantages in managing seasonal work, but it can also present challenges, because these workers have often poor working conditions (wages, drudgery, etc.). It should also be noted that this research did not gather information on time spent on domestic tasks (such as cooking, childcare, etc.), which are predominantly carried out by women and contribute significantly to the functioning of the farming system.

4.3 Determinants of work organizations

4.3.1 At site level

To analyze the impacts of changes on farm work and workers (beyond the the farm manager alone), understanding the determinants of work organization patterns is crucial. This task is more straightforward in homogeneous samples (e.g., within a single farming system or country) than in such a comparative analysis, where variation in farm size, diversity of farming systems, crop and livestock, levels of mechanization, and cultural values play an important role on the balance between family and wage, as well as on the gendered distribution of tasks.

While it is possible to identify the range of determinants shaping work organization, establishing their relative importance is not feasible at this broad level of analysis and can be done more effectively on a site-by-site basis.

For instance, in Senegal, livestock practices (batching, transhumance, milking) play an important role in routine work duration and in the mobilization of family or permanent wage-earners shepherds. These practices are linked to value chains dynamics (existence of a market for milk), land pressures (crop-livestock competition), ecological and climatic constraints. In Burkina Faso, cotton value chains incentives on inputs and campaign credits are major drivers of intensive work peaks managed by male family workers. In Tunisia, work organization differentiates irrigated crops and livestock − dry cereals farms, with female temporary wage earners in the first and farmer engagement in the second. Hydraulic policies and land regulations on one side and gradients of drylands proportion in the landscape on the other probably play a role. In Madagascar, structural constraints as low mechanization, land pressure and limited access to capital require the mobilization of family workers, either for routine or for seasonal work made manually.

The distribution of tasks between men and women depends on imbricated factors such as social norms, differentiated access to mechanized tools, decision making within the couple, and the physical drudgery of tasks.

4.3.2 At farm level: off farm diversification

Another major factor of work organization, the off-farm engagement of the farmers was not quoted sufficiently precisely in the sites. What we know is that it can be carried out by men (example a truck driver) or women (food market) in the same site but we cannot draw generalities either at the site or whole sample level. More generally, risk management through the diversification of economic activities is a frequent strategy by family farmers in Africa (Hazell et al., 2024) may imply the mobilization of wage earners, either permanent (for livestock care notably) or seasonal to face peaks of work, and requires for the remaining family worker a capacity to work on a large variety of tasks: a high level of specialization is not possible. For instance, in a women-only focus group in Madagascar, participants explained that, since their husbands are often absent due to off-farm activities, they must be trained in all tasks, including driving oxen. This organizational flexibility is a lever to reinforce risk management strategies.

4.3.3 Considering changes in agricultural practices

Labor allocation needs to be clearly defined, along with the changes in routine and seasonal tasks (Laske and Michel, 2022). Changing the combination of activities, the practices, enlarging the farms, as consequences of policies, market incentives and of risk management has impact on the work organization within the farms. But the nature of the impact either on wage earner − family contributions or on men − women tasks distributions must be carefully estimated at the local and individual farm level. Considering the distribution of tasks between workers is crucial to think about changes in agricultural practices. Indeed, any change in the agricultural practices has differentiated impacts on the distribution of tasks between family vs wage and on gender. For example, enhancing agroecological practices that require more seasonal work to be done on crops will involve different workers depending on the farm type: on type 1 the family farm manager is more concerned and on type 3 it will touch more seasonal, often female. Other example, in Tunisia, manure application is made by women: any development of manure application, as a consequence of an agroecologization, has differentiated impact on gender working drudgery.

4.4 Workers conditions related to work organization forms

As showed in our results, workers involved in work organization patterns are not the same, and specifics issues related to their working conditions are at stake.

In relation with type 1, the farm manager monitors the farming system but is the major worker concerned by the operational tasks that have to be done, either routine or seasonal. The distinction between men and women must be quoted, other domestic tasks (child care, food) being in general more devoted to women and interacting with the implementation of tasks. In these situations, the perception of work by the farm managers because of age, physical limitations, self-fulfilment expectations often require simplifying or adapting the farming system technical sequences (Dedieu and Servière, 2012)

In relation with type 2, the farm manager monitors the farming systems and have a significant contribution to the farming work, sharing with other family members the implementation of tasks, notably routine ones. The gender division of tasks and articulation with domestic tasks is also a major factor of differentiation of the working conditions. Family workers contribute to all tasks, but more specifically on routine ones. The way these workers perceive their work, their perception of well-being including autonomy and their perspective for staying are certainly underknown. In our sample, permanent workers are not numerous and more or less males devoted to routine work on animals, as shepherds or dairy caretakers. They bring specific skills within the collective of workers and have also specific perceptions and expectations for their work.

In relation with type 3, the farm manager monitors the farming system and has a major role of supervision of the seasonal workers (from gathering the workers in close villages or towns to supervision the quality of the work done) with a direct contribution on a large variety of tasks. Efficiency of the paid work is, in these situations, often an attention points for the managers. The denomination of temporary workers covers a large variety of specialised skills or recognized aptitudes, and social relations with the farm managers. Their situations are more explored notably through “decent work” analysis (ILO, 2024b). In our sample, a majority of temporary workers are women and their contribution to farming work must also be related to domestic and other remunerative activities.

Our study highlights that workers’ profiles is not only a question of status but also of place within work organizations. Exploring what they perceive as job satisfaction (Di Bianco et al., 2025) and attractivity factors or intention to leave or to stay must be related to these work organization patterns.

5 Conclusion

While not being representative of the large diversity of farming systems all around Africa, our study highlights that work organization patterns within farms are highly diverse, resulting in a wide range of worker profiles that together constitute the African agricultural workforce. Family farming encompasses several work organizations, whether through the involvement of family members in task implementation or the use of wage earners. The factors differentiating these work organization patterns are numerous, including farm size, farming systems combining of activities and technical sequences, family composition and are often country or territory specific. Understanding how changes such as sustainable intensification, agroecological transitions, mechanization, or the expansion of off-farm activities affect these patterns, as well as the working conditions of each category of worker, is central to analyzing the future of work in African agriculture.

References

Appendix 1 Crops tasks calendar in Highlands of Madagascar

Appendix 2 Correlation of variables with each type of work organization patterns.

All Tables

All Figures

Fig. 1 Average number of tropical livestock units (TLU) and agricultural area (ha) per farming system. (a) Average number of tropical livestock unit (TLU) per farming systems. (b) Average cultivated areas (hectares) per farming system. Nombre moyen d’unités de bétail tropical (UBT) et surfaces cultivées (ha) par système agricole (a) Nombre moyen d’unités de bétail tropical (UBT) par système agricole. (b) Surface moyenne cultivée par système agricole (ha).

In the text

Fig. 2 Inter farming systems variability of annual routine and seasonal work durations. (a) Inter farming systems variability of annual routine work duration for crops and livestock expressed in hours per year. (b) Inter farming systems variability of annual seasonal work duration for crops and livestock expressed in number of days per year. Variabilité de la durée annuelle moyenne du travail d’astreinte et de saison annuel entre systèmes agricoles . (a) Variabilité de la durée annuelle du travail d’astreinte entre systèmes agricoles (heures /an) (élevages et cultures). (b) Variabilité de la durée annuelle du travail de saison entre systèmes agricoles (jours / an) (cultures et élevage).

In the text

Fig. 3 Inter farming systems variability of routine and seasonal work distribution between categories of workers expressed in %. (a) Inter farming systems variability of routine work distribution between categories of workers expressed in %. (b) Inter farming systems variability of seasonal work distribution between categories of workers expressed in %. Variabilité, inter systèmes agricoles, de la répartition moyenne du travail d’astreinte et du travail de saison selon les différentes catégories de travailleurs (en %). (a) Variabilité inter systèmes agricoles de la répartition moyenne du travail d’astreinte selon les différentes catégories de travailleurs (en %). (b) Variabilité inter systèmes agricole de la moyenne de répartition du travail de saison par catégorie de travailleurs.

In the text

	Fig. 4 Inter farming systems variability of category of workers and gender contribution to the total work. Total work = RW / 8 + SW in days (conventionally 8 h are equivalent of one day). Variabilité inter système agricole de la contribution des hommes et des femmes au travail total, par catégorie de travailleurs.
In the text