Drivers of Member Participation decision and intensity in Co-operative Services, and Policy Implications for Agricultural Co-operatives in Uganda
A
MercylineJerusaOng’ayo1✉Emailmercyline.jerusa@gu.ac.ug
StephenWamalaKalule1
WalterOdongo1
MercylineJ.1
1Department of Rural Development and AgribusinessGulu UniversityP.O. Box 166GuluUganda
2Nabuin Zonal Agricultural Research and Development InstituteNational Agricultural Research OrganizationP.O. Box 132MorotoUganda
Mercyline Jerusa Ong’ayo1, Stephen Wamala Kalule1, Walter Odongo1,2
1,2 Gulu University, Department of Rural Development and Agribusiness, P.O. Box 166, Gulu, Uganda
2 Nabuin Zonal Agricultural Research and Development Institute, National Agricultural Research Organization, P.O. Box 132, Moroto, Uganda
Corresponding author: mercyline.jerusa@gu.ac.ug
Abstract
A
Farmer participation in agricultural co-operatives is the cornerstone of agricultural development and rural economic sustainability, particularly in developing economies where farmers face challenges in accessing markets, credit, and technical support. However, the dynamics of farmer participation and patronage in co-operative services remain a critical area of study. This study examined the co-operative and farmer factors influencing participation in five co-operative services, including farm input supplies, marketing, training, savings, and financial lending. Primary data were obtained from 433 member farmers of primary agricultural co-operative societies in Northern Uganda. Results indicate that only 18% of members participated across the five services. Comparative analysis shows that older and larger co-operatives with fewer women and youthful members attracted more participation in inputs, marketing, and training. A
The Heckman results indicate that co-operative charges, youth ratio, technical staffing, and membership with other groups negatively impacted participation, while shareholding, member satisfaction, and leadership roles had a positive influence. Contrastingly, co-operative affiliation, age and size, female ratio, record keeping, and infrastructure had a divergent effect on participation. Given the low and uneven patronage of co-operative services, we recommend that co-operatives adopt demand-driven services, leverage external networks for infrastructure, review equity charges, and customize programs to better engage diverse member profiles, including women and youth. Moreover, participatory and inclusive approaches are vital to tailor interventions that address the distinct needs instead of blanket interventions. This would enhance member participation and utilization of co-operative services, ultimately contributing to the sustainability of agricultural co-operatives and rural development.Keywords:
Farmer participation
Co-operative patronage
Rural development
Collective action initiatives
Co-operative development
Co-operative activities
A
1.0 Introduction
Agriculture remains a fundamental pillar for economic growth and rural development in Sub-Saharan African countries, where over 84% of smallholder farmers play a significant role in the sector [1]. Despite their vital role, these farmers face substantial challenges, including inadequate access to production resources, limited markets, post-harvest losses, and increasing vulnerability to climate change [2], [3]. These constraints limit their overall contribution to food security and poverty reduction, thus jeopardizing agriculture’s contribution to the attainment of the United Nations Sustainable Development Goals. Collective action through agricultural co-operatives and other forms of farmer organizations (FOs) has emerged as a viable strategy to address some of the systemic issues limiting smallholder farmers [4]. They reportedly contribute to agricultural productivity, commercialization, among other household welfare outcomes [5][6].
Agricultural co-operatives exemplify a unique model of collective action, where farmers pool resources and coordinate efforts to achieve economies of scale, improve bargaining power, and gain access to resources that may be unattainable individually. Co-operatives are organized, owned, and democratically governed by members with common economic and social interests. Modern co-operatives provide an expansive range of services, including farm input supply, bulking and storage of produce, marketing, training and extension services, financial services, and value addition [4], [5]. Through these collective action initiatives (CAIs), co-operatives reduce transaction costs, address market failures, and facilitate the dissemination of knowledge, skills, and innovations, consequently contributing to the adoption of technologies, improved yields, incomes, and positive social outcomes [7], [8]. Despite these benefits, co-operatives in developing countries often encounter significant institutional and operational challenges, including liberalized markets, limited funding, poor governance, and, notably, low member participation and patronage, that undermine their effectiveness and sustainability [9]–[11]. While many co-operatives succeed in enrolling large memberships, mobilizing members to actively participate both economically and democratically remains a persistent issue, negatively affecting economies of scale, bargaining power, and overall co-operative effectiveness.
The concept of farmer participation has therefore received considerable scholarly attention in the context of agricultural co-operatives, highlighting the determinants of participation and associated outcomes [11]–[16]. Whereas these findings are crucial to understanding participation and its drivers, participation is predominantly conceptualized based on membership status, with members regarded as participants [17], [18]. However, participation is dynamic; beyond membership status, there is heterogeneity in participation among the members, with varying decisions to participate and utilize various co-operative services [10], [13]. For instance, a co-operative may offer a bundle of services, but only a few members will participate in all services, with some members prioritizing some services over others. Additionally, members who decide to participate in CAIs can have varying intensities of participation in each CAIs. Therefore, the dichotomization of farmer participation based on membership status assumes homogeneous participation among members, overlooking the variability of participation decisions. Besides, computation of participation intensity (PI) also ranges from composite indices [19] to binary indicators [18], [20], and a count of co-operative activities a member participates in [15]. Similarly, such computation fails to disaggregate the actual member PI.
Moreover, extensive research focuses disproportionately on participation in co-operative governance activities such as attending and speaking in annual general meetings, and voting, among other routine activities [17], [21]. However, beyond governance functions, co-operatives are involved in economic activities such as input supply, produce marketing, agro-processing, savings, financial lending, and training [5]. A few studies attempting to delve into these CAIs are skewed toward collective marketing, providing little insight into farmer participation in other CAIs and interdependencies across different co-operative services [12], [22], [23]. Furthermore, participation is shaped by a complex interplay of co-operative institutional frameworks, social capital, incentive structures, and individual socio-economic attributes. Exclusive focus on either co-operative or member characteristics limits the explanatory power of factors influencing participation in co-operatives. Therefore, this study sought to address these gaps by (1) Profiling member participation decisions and intensities across five co-operative services: farm input supply, marketing member produce, training services, savings, and financial lending. (2) Comparing participating and non-participating members' characteristics (individual and co-operative). (3) Examine how co-operative characteristics alongside individual member-level factors influence participation decisions and intensities in various co-operative services. This study deepens the existing literature on farmer co-operatives, illustrating that the drivers of participation vary depending on the co-operative service offered to members. By simultaneously analyzing multiple services and integrating co-operative and member-level variables, this study captures the complex relationships that influence participation, thus advancing understanding of participation dynamics in collective action. The findings have practical implications for policymakers and co-operative leaders, enabling more targeted interventions aimed at increasing member engagement, improving co-operative performance, and ultimately strengthening the livelihoods of smallholder farmers.
1.2 Theoretical perspective
To understand the dynamics of farmer participation in co-operative services, this study draws on Collective Action Theory (CAT) and Rational Choice Theory (RCT). They collectively explain the dual nature of participation in co-operatives as a social dilemma requiring institutional solutions and an individual rational choice driven by utility maximization. Differences in institutional quality, social capital, individual incentives, costs, and information access culminate in heterogeneous participation decisions and intensities among members. According to CAT [24], the success of collective action depends on socio-institutional factors and organizational frameworks. In this study, co-operative characteristics, including size, age, gender composition, affiliation to a co-operative union, availability of infrastructure (storage facility and roads), and technical staffing, can significantly affect co-operatives' capacity to provide effective services, thereby influencing participation. For instance, larger co-operatives may experience reduced cohesion, while gender and age diversity may have diverging effects on cooperation.
The RCT assumes that individual farmers are rational agents, making participation decisions based on the costs and benefits analysis. Although members are patrons of their co-operatives, they evaluate the utility derived from participation in co-operative services, such as fair prices, dividends, storage, training, or financial products, against costs, which may include mandatory membership and capital share fees, time investment, and delayed payments. Stakeholding through shareholding and leadership roles can enhance participation intensity due to anticipated higher returns. Access to information through mobile phones and extension services reduces transaction costs, thereby influencing rational participation decisions. Satisfaction with co-operative services can also signify the perceived utility and influence participation. Crop income, a proxy of economic success, reflects both a cause and a consequence of participation, while membership in other groups indicates an individual's social capital, aligning with rational assessments of where returns on time and resource investment are greatest.
The interaction between co-operative level factors and an individual's rational choices is critical. For instance, well-managed co-operatives with adequate staffing, reasonable financial charges, and good infrastructure present lower transaction costs and increased perceived benefits, thereby incentivising higher participation. Conversely, poor institutional quality can deter participation despite individual rational incentives. Literature by [25] and [21] demonstrates how farmer socioeconomic, institutional, and gender disparities influenced participation in Kenyan agricultural co-operatives. Literature also emphasizes the combined influence of organizational functionality, structure, and individual motivation on participation [12], [14], [17]. These findings affirm that structural and rational considerations co-determine participation, thus validating the integration of the two theoretical perspectives to analyse the multifaceted drivers of farmer participation in agricultural co-operatives and the resulting diversity in participation decisions and intensity across different CAIs.
2.0 Methodology
2.1 Study area
The study was conducted in the northern region of Uganda, which is about 85,290 km2 in land area and home to 18.65% of the Ugandan population. The region is located between latitude 2.8780° N and longitude 32.7181° E, with an elevation of 1,078m above sea level, and receives bimodal rainfall, making agriculture a core activity in rural areas. Predominant crops cultivated in the region include maize, rice, beans, cassava, soya beans, groundnuts, sweet potatoes, sesame, and millet [26]. The Acholi and Lango Sub-regions were purposively selected for this study because the functionality of most agricultural co-operatives in the area collapsed following past political insurgency and trade liberalization [27]. To overcome liberalization effects, co-operatives in the region continue to adopt a new organization structure that operates through Rural Producer Organizations (RPOs) and area co-operative enterprise [28]. Despite receiving significant humanitarian and governmental support for the agrarian revolution, farmer organizations in these sub-regions still have a low sustainability index and sub-optimal farmer participation in co-operatives [7],[9].
2.2 Sampling framework
Using a multistage sampling procedure, 23 primary agricultural co-operative societies in the Lango (Lira, Kole, Oyam, and Alebtong districts) and Acholi (Amuru, Omoro, and Nwoya districts) sub-regions were purposively selected in liaison with West Acholi and Lango co-operative unions. With a study population of 55,400 co-operative members and an error of 0.05, Yamane Taro's formula was used to compute a sample size of 397, which was topped with a 10% increase to 440 for attrition [29]. A total of 244 members of the respondents were selected from Lango and 196 from Acholi co-operatives.
2.3 Data collection and management
Ethical clearance was obtained from the
A
Gulu University Research and Ethics Committee (GUREC) and the Uganda National Council for Science and Technology (UNCST), followed by strict adherence to the ethical guidelines during data collection. Primary data were collected through a cross-sectional survey using a pretested semi-structured questionnaire administered by trained enumerators through face-to-face interviews between May and July 2024. This questionnaire covered concepts related to participation, co-operative characteristics, as well as farmer socio-economic attributes (Table 1). The primary data from Kobo Collect were exported to the Statistical Package for Social Science (SPSS) version 23.0 for cleaning, computations, and transformation of some variables, during which 7 observations were dropped due to data inconsistencies. Subsequently, clean data were transferred to STATA version 17 for diagnostic checks (multicollinearity, heteroscedasticity, normality) before undertaking further analysis for descriptive statistics and inferential statistics.Participation decisions (PD) were measured as a dummy for each of the five CAIs, whereby members who utilized co-operative services were coded one, and zero for non-participants. Participation intensity (PI) was then computed to reflect the extent to which participant members utilized or patronized each co-operative’s CAIs. Following discrepancies in the computation of PI [15], [18]–[20], this study adapted the computation by [25] to quantify the actual proportion of transactions done through the co-operative for each CAIs as follows: Input PI was computed as the proportion of farm inputs procured through the co-operative over the overall inputs procured in a year. Quantification of inputs (agrochemicals, seeds, and fertilizers) was measured in terms of financial expenditure (Ugx) on inputs. Weight-based quantification was inappropriate due to the variability of the measurement unit for inputs. Marketing PI was computed as the proportion (quantity in kg) of a member’s produce marketed through the co-operative over the total quantity of a member’s produce marketed in a year. Maize, rice, and soybeans were considered, as they were common produce marketed through the co-operatives. On the other hand, savings PI was computed as a proportion of a member's savings (Ugx) made through the co-operative over the total savings made by a member in all savings sources for a year. Financial lending PI was computed by dividing the total amount of money borrowed from the co-operative by the total amount of loans borrowed by a member in a year while PI in training services was computed as the number of co-operative trainings attended by a member over the total number of trainings conducted by the co-operative in the year. Following this procedure, members’ participation intensities were computed for each of the five co-operative activities with values ranging between zero and one.
2.4 Data analysis
A
The descriptive statistics were used to profile the co-operatives, members, and participation, while the independent sample t-test and chi-square were applied to compare participants and non-participants across five co-operative CAIs. A series of independent Heckman regression models was analyzed to determine the influence of the co-operative and farmers' characteristics on participation decisions and intensity in five co-operative CAIs.Heckman was preferred because it can concurrently analyze participation decision and intensity. Moreover, it allows for the correction of the non-random selection into participation intensity through the incorporation of the Inverse Mills' Ratio (IMR). Alternative models for the simultaneous analysis of multiple binary and censored outcomes include Multivariate probit (MVP) and multivariate Tobit (MVT), respectively. However, MVP accounts for correlation and interdependence among binary outcomes but cannot handle participation intensity, while MVT assumes participation decisions and intensity are governed by the same factors, thus failing to account for correlation and sample selection bias [30]. Heckman's two-step model overcomes the restrictive assumptions of MVT and complements MVP by addressing continuous outcomes with selection bias. Heckman’s selection stage regresses member participation decisions in each of the co-operative CAIs through a probit model, as shown in Eq. (1), while the outcome stage reflects participation intensities (2).
1
2
Where
is the latent variable representing the participation decision of the ith member in the sample for CAIs j, each regressed separately, and takes a value of one if a member participated in the jth CAIs and zero otherwise.
Is the observed intensity of participation in CAIs j for member
that participated. The j includes five CAIs (input procurement, collective marketing, training, savings, and financial lending).
and
Are vector of farmer and co-operative characteristics that influence participation decisions and intensity, respectively.
are vectors of coefficients to be estimated on the explanatory variables, while
is the error term assumed to be normally distributed and
is the IMR component while
is the coefficient for IMR. A detailed description and measurement of the variables incorporated in the regression are presented in Table 1.
Variables | Variable description and measurement | Priori sign | mean | Std |
|---|---|---|---|---|
Market PI | Proportion of produce sold through the co-operative out of total produce (maize, rice, and soybean) sold. | 0.445 | 0.407 | |
Input PI | The proportion of expenditure on inputs procured through the co-operative out of the total inputs procured. | 0.472 | 0.429 | |
Training PI | The proportion of co-operative training attended out of the total training conducted by the co-operative. | 0.606 | 0.353 | |
Saving PI | The proportion of savings made through the co-operative of the total savings made. | 0.411 | 0.399 | |
Loan PI | The proportion of money borrowed from the co-operative of the total money borrowed from all sources. | 0.380 | 0.438 | |
Coop affiliation | Co-operative affiliation to the union, one if yes, zero otherwise. | + | 0.448 | 0.498 |
Coop age | The duration in years for which a co-operative has been in existence after the new act registration. | + | 10.244 | 5.970 |
Coop size (sqrt) | The square root of the total number of registered co-operative members. | +/- | 6.213 | 1.701 |
Female ratio | The number of female co-operative members over the total co-operative membership size. | +/- | 0.516 | 0.121 |
Youth ratio | The number of youthful (18–35 years) co-operative members over the total co-operative membership size. | - | 0.332 | 0.214 |
Coop charges | The natural log of the sum of share capital and membership fee charges charged by the co-operative. | - | 10.104 | 0.791 |
Coop records | The total number of mandatory records kept by the co-operative. | + | 6.076 | 1.811 |
Coop storage | The ownership of a produce storage facility by a co-operative, one if yes, zero otherwise. | + | 0.483 | 0.500 |
Coop tech staff | The presence of employed technical staff in a co-operative, one if yes, zero otherwise. | + | 0.409 | 0.492 |
Road condition | The status of co-operative feeder roads: 1 if in good condition, zero if in poor condition. | +/- | 0.651 | 0.477 |
Crop income | The square root of the total income obtained from the sale of targeted crops (maize, soybeans, and rice). | +/- | 14.00 | 1.486 |
Satisfaction | The overall mean of satisfaction with co-operative leadership, fairness, communication, and services. | + | 4.329 | 0.744 |
Shares capital | The total number of share capital held by the member. | + | 1.100 | 0.813 |
Exten access | Access to agricultural extension services over the past year, one if yes, zero otherwise. | + | 0.655 | 0.476 |
Phones inform | Respondent’s access to various information sources over the phone, one if yes, zero otherwise. | +/- | 0.503 | 0.501 |
Leadership role | Respondent's leadership role in the co-operative, one if yes, zero otherwise. | + | 0.304 | 0.461 |
Total activities | Total number of co-operative collective activities that a respondent participated in for that year. | +/- | 3.570 | 1.349 |
Other groups | Respondents' membership in other forms of farmer groups, one if yes, zero otherwise. | - | 0.684 | 0.466 |
| Source: Authors's Field survey (2024), [12], [18]–[20], [25] | ||||
3.0 Results
3.1 Profile of agricultural co-operative, members, and participation
All co-operatives provided the five CAIs and had existed for about 10 years, with 55% owning premises and operating independently. About 44% of co-operatives were in rural areas. Membership compositions constitute 33% youth and 52% women, with average membership and share capital charges at Ugx 10,000 and Ugx 25,000, respectively. About 41% of co-operatives had staffing. The majority received diverse training, but a disproportionate distribution of other benefits (69% received subsidized inputs, 64% processing machines, 48% storage facilities, and 27% financial and farm machinery).
A
About 52% of the study population were men with an average age of 43 years. Most members were married, with a household size of six and a dependency ratio of 1.6. They owned about six acres of land, allocating 80% to crop production, and 58% were engaged in off-farm income-generating activities. Education-wise, 44% had studied beyond primary level, and 66% accessed agricultural extension services. While the majority (87%) owned phones, only half accessed agricultural information from their phones (Table 1). Most members had an average co-operative membership duration of five years, with 32% holding more than one share capital and 30% holding leadership positions in the co-operative, while 68% had membership with other farmer groups.Despite the existence of CAIs in most co-operatives, there was variation in members’ participation decisions, with training attracting the highest participation (84%) (Fig. 1). The aggregated participation decisions show that only 18% of the members participated in all five CAIs, with the majority (30%) participating in three CAIs (Fig. 2). The average members’ participation intensities were below 50% for all CAIs except training. This indicates that, whereas some members participate (Fig. 1), less than half of their various transactions are done through the co-operatives, hence indicating a low utility/patronage of co-operative services (Fig. 3).
A Figure 1 Members' participation decisions (%) in co-operative CAIs.Authors' Field Survey (2024) | A Figure 2 Aggregated members’ participation in co-operative CAIs. Authors' Field Survey (2024) |
A Figure 3 Participation intensity among participant members in co-operative CAIs. Authors' Field Survey (2024) | |
3.2 Comparison of participants and non-participants in various co-operative activities
The t-test (Table 2) and chi-squared (Table 3) indicate statistically significant differences among participants and non-participants for various co-operative characteristics. Older and larger co-operatives with fewer women and youth attracted more participants in farm inputs, marketing produce, and training CAIs, while newer and smaller co-operatives with more women attracted more participants in financial services at varying significance levels (Table 2). Co-operative affiliation to a union, ownership of storage facility, technical staffing, access to electricity, and condition of feeder roads were also significantly different among participants and non-participants in various co-operative services.
The test and chi-square also reveal significant differences between participants and non-participant members in various CAIs with respect to member characteristics. Notably significant disparities exist in the household dependency ratio, cropland ratio, crop income, total CAIs, education level, shareholding, membership to other groups, and informational access through phone and extension services. Household dependency ratios (HHD) were significantly higher among non-participants in collective input and marketing (p < 0.05) and lower among non-participants in savings (p < 0.05). Crop land ratio and incomes were also significantly higher among participants in inputs and marketing (p < 0.01), but lower crop land ratio among financial service participants. Aggregated members’ participation in CAIs (TCAIs) was also significantly higher among participants across all the services (p < 0.01). Chi-square results show that informational access through phone and extension services was significantly higher among participants in inputs, marketing, and training (p < 0.01), while the majority of participants in loans had more access to phone information. Membership in other groups and shareholding were consistently higher among participants in inputs (p < 0.01), marketing (p < 0.05), and training.
Member Characteristics | Farm input participation | Marketing participation | Training participation | Lending participation | Savings participation | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Yes (285) | No (148) | t-test sig | Yes (299) | No (134) | t-test sig | Yes (364) | No (69) | t-test sig | Yes (216) | No (217) | t-test sig | Yes (260) | No (173) | t-test sig | |
Age | 42.00 | 44.00 | 1.35 | 44.00 | 42.00 | -1.292 | 43.00 | 43.00 | 0.18 | 44.00 | 42.00 | -0.98 | 43.00 | 43.00 | -0.49 |
HH D.ratio | 1.51 | 1.85 | 2.27** | 1.53 | 1.84 | 1.992** | 1.59 | 1.82 | 1.21 | 1.73 | 1.52 | -1.44 | 1.74 | 1.45 | -1.97** |
Total land | 6.78 | 6.28 | -0.74 | 6.66 | 6.48 | -0.272 | 6.82 | 5.50 | -1.54 | 6.88 | 6.33 | -0.88 | 7.06 | 5.93 | -1.76* |
Crops ratio | 0.83 | 0.76 | -2.44** | 0.81 | 0.79 | -0.712 | 0.80 | 0.83 | 0.95 | 0.78 | 0.84 | 2.46** | 0.77 | 0.85 | 3.13*** |
Crop income (nl) | 14.28 | 13.47 | -5.55*** | 14.20 | 13.58 | -4.09*** | 14.05 | 13.76 | -1.50 | 13.987 | 14.02 | 0.23 | 14.34 | 14.10 | 1.11 |
Off-incom (sqrt) | 288.00 | 236.80 | -1.63 | 282.55 | 243.64 | -1.20 | 267.35 | 287.20 | 0.49 | 245.33 | 295.57 | 1.68** | 274.12 | 265.08 | -0.30 |
Total CAIs | 4.00 | 3.00 | -9.53*** | 4.00 | 3.00 | -13.04*** | 4.00 | 2.00 | -11.34*** | 4.00 | 3.00 | -14.09*** | 4.00 | 3.00 | -14.25*** |
Member duration | 5.48 | 4.81 | -1.52 | 5.35 | 5.05 | -0.64 | 5.38 | 4.61 | -1.34 | 5.13 | 5.38 | 0.59 | 4.98 | 5.67 | 1.62 |
Co-operative Characteristics | |||||||||||||||
Coop size (nl) | 6.51 | 5.64 | -5.16*** | 6.29 | 6.05 | -1.34 | 6.25 | 6.01 | -1.11 | 5.88 | 6.54 | 0.00*** | 5.67 | 7.02 | 8.76*** |
Coopage (nl) | 2.24 | 1.99 | -4.14*** | 2.14 | 2.19 | 0.71 | 2.17 | 2.06 | -1.39 | 2.10 | 2.21 | 1.80* | 2.06 | 2.29 | 4.01*** |
Female ratio | 0.50 | 0.54 | 2.81*** | 0.51 | 0.53 | 1.13 | 0.52 | 0.52 | 0.22 | 0.53 | 0.50 | -2.23** | 0.53 | 0.50 | -2.34** |
Youth ratio | 0.33 | 0.34 | 0.59 | 0.32 | 0.37 | 2.43** | 0.33 | 0.35 | 0.87 | 0.29 | 0.37 | 4.14*** | 0.29 | 0.40 | 5.27*** |
Charges (nl) | 10.47 | 10.44 | -0.31 | 10.45 | 10.50 | 0.51 | 10.46 | 10.45 | -0.10 | 10.20 | 10.67 | 4.81*** | 10.22 | 10.74 | 5.54*** |
Committees | 4.00 | 3.00 | -4.25*** | 4.00 | 3.00 | -1.83* | 4.00 | 4.00 | -0.85 | 3.00 | 4.00 | 3.34*** | 3.00 | 4.00 | 5.36*** |
Source: Authors' Field survey (2024) 2*, **, and *** imply significance at 10%, 5%, and 1%, respectively. HHD.ratio is the household dependency ratio.
Co-operative CAIs | Farm input participation | Marketing participation | Training participation | Lending participation | Saving participation | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
 | Yes 285 | No 148 | Chi2 Sig | Yes 299 | No 134 | Chi2 Sig | Yes 364 | No 69 | Chi2 Sig | Yes 216 | No 217 | Chi2 Sig | Yes 260 | No 173 | Chi2 Sig | |
Education | >primary | 74 | 26 | 9.31*** | 69 | 31 | 0.00 | 86 | 14 | 0.75 | 39 | 61 | 14.68*** | 54 | 46 | 5.71** |
≤Primary | 60 | 40 | 69 | 31 | 83 | 17 | 58 | 42 | 65 | 35 | ||||||
Phone info access | Yes | 76 | 24 | 20.81*** | 76 | 24 | 9.04*** | 90 | 10 | 13.02*** | 46 | 54 | 2.83* | 57 | 43 | 1.34 |
No | 55 | 45 | 62 | 38 | 78 | 22 | 54 | 46 | 63 | 37 | ||||||
Extension Access | Yes | 73 | 27 | 20.20*** | 73 | 27 | 4.68** | 89 | 11 | 13.42*** | 45 | 55 | 8.81*** | 54 | 46 | 14.65*** |
No | 52 | 48 | 62 | 38 | 75 | 25 | 60 | 40 | 72 | 28 | ||||||
Share capital | >one | 77 | 23 | 10.43*** | 76 | 24 | 4.41** | 89 | 11 | 3.72* | 47 | 53 | 0.48 | 60 | 40 | 0.00 |
≤one | 61 | 39 | 66 | 34 | 82 | 18 | 51 | 49 | 60 | 40 | ||||||
Leadership incoop | Yes | 68 | 32 | 0.47 | 79 | 21 | 8.42*** | 90 | 10 | 5.25** | 59 | 41 | 6.44** | 73 | 27 | 14.29*** |
No | 65 | 35 | 65 | 35 | 81 | 19 | 46 | 54 | 54 | 46 | ||||||
Other groups | Yes | 70 | 30 | 8.24*** | 73 | 27 | 5.617** | 87 | 13 | 5.32** | 52 | 48 | 2.302 | 62 | 38 | 1.75 |
No | 56 | 44 | 61 | 39 | 78 | 22 | 45 | 55 | 55 | 45 | ||||||
Co-operative characteristics | ||||||||||||||||
Coop affiliation | Yes | 58 | 42 | 10.22*** | 68 | 32 | 0.38 | 84 | 16 | 0.08 | 60 | 29 | 41.23*** | 63 | 17 | 87.87*** |
No | 72 | 28 | 70 | 30 | 85 | 15 | 40 | 71 | 37 | 83 | ||||||
Own storage | Yes | 58 | 42 | 13.97*** | 62 | 38 | 10.64*** | 79 | 21 | 8.82*** | 43 | 53 | 4.69** | 38 | 64 | 29.15*** |
No | 75 | 25 | 77 | 23 | 89 | 11 | 57 | 47 | 62 | 36 | ||||||
Technical staff | Yes | 77 | 23 | 16.15*** | 72 | 28 | 1.49 | 86 | 14 | 0.73 | 28 | 54 | 30.60*** | 25 | 64 | 64.63*** |
No | 58 | 42 | 67 | 33 | 83 | 17 | 72 | 46 | 75 | 36 | ||||||
Electricity access | Yes | 56 | 44 | 13.59*** | 68 | 32 | 0.21 | 82 | 18 | 1.56 | 45 | 67 | 20.23*** | 43 | 75 | 42.35*** |
No | 71 | 29 | 68 | 32 | 87 | 13 | 79 | 69 | 85 | 57 | ||||||
Road condition | Good | 69 | 31 | 3.98*** | 70 | 30 | 0.51 | 84 | 16 | 0.09 | 65 | 65 | 0.00 | 62 | 70 | 2.94** |
Poor | 60 | 40 | 67 | 33 | 85 | 15 | 35 | 35 | 38 | 30 | 0.09 | |||||
Source: Authors' Field survey (2024). *, **, and *** imply significance at 10%, 5%, and 1%, respectively.
3.3 Determinants of member participation decision and intensity in co-operatives' CAIs
Table 4 presents a series of Heckman two-stage regression results for the determinants of participation in five co-operative CAIs. The IMR was significant, justifying the use of Heckman. Wald chi2 was significant for all the CAIs, indicating the explanatory power of predictor variables in explaining participation decisions and intensity in co-operative CAIs. Results show a significant negative impact of co-operative share capital charges on participation decisions (PD) in training (p < 0.05), savings and loans (p < 0.01). However, the number of shares owned by a member significantly enhanced PD across most of the CAIs. Proportion of youth negatively impacted PD in inputs (p < 0.1), savings (p < 0.01), and loans (p < 0.05), while ownership of storage facility negatively influenced PD in inputs and savings (p < 0.05) while simultaneously increasing PD in marketing and training CAIs (p < 0.05). Co-operative affiliation to a union and proportion of female members significantly enhanced the probability of member participation in savings and loans (p < 0.01). Access to agricultural extension and phone information enhanced PD in most CAIs with a significant increment in input (p < 0.05) and training (p < 0.01). Similarly, crop income enhanced PD in most of the CAIs with a significant increase in input (p < 0.01) and marketing (p < 0.05).
Regarding participation intensity (PI), co-operative affiliation to a union significantly reduced PI in marketing and training (p < 0.01) while increasing PI in savings and lending (p < 0.1). Membership in other groups negatively impacted PI in training and savings (P < .01), while member satisfaction enhanced PI in input (p < 0.1), marketing, and savings (p < 0.01. Good road condition also enhanced PI in inputs and marketing at (p < .005) and (p < 0.1) respectively. The multidirectional effects of predictor variables on members' participation decisions and intensities across the five CAIs imply that the “one size fits all” assumption for remedial interventions may not effectively solve co-operative issues.
A
These new insights call for careful consideration when designing remedial actions for promoting membership participation in co-operative services.Farm input supplies | Marketing produce | Training services | Savings services | Financial lending | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
Participation Decisions | Coeff | Std. err. | Coeff | Std. err. | Coeff | Std. err. | Coeff | Std. err. | Coeff | Std. err. |
Coop affiliation | 0.295 | 0.230 | 0.111 | 0.223 | -0.262 | 0.273 | 0.953*** | 0.247 | 0.779*** | 0.225 |
Coop size (sqrt) | 0.026*** | 0.006 | 0.005 | 0.005 | -0.012* | 0.006 | -0.006 | 0.005 | 0.001 | 0.005 |
Coop age | -0.026 | 0.019 | -0.044** | 0.019 | 0.037 | 0.025 | 0.009 | 0.021 | 0.006 | 0.019 |
Coop female ratio | -1.621*** | 0.613 | -0.648 | 0.631 | 0.813 | 0.707 | 3.226*** | 0.737 | 2.485*** | 0.627 |
Coop youth ratio | -0.470* | 0.338 | -0.463 | 0.330 | -0.038 | 0.379 | -0.948*** | 0.355 | -0.86** | 0.349 |
Coop charges (nl) | 0.076 | 0.107 | -0.161 | 0.107 | -0.268** | 0.123 | -0.455*** | 0.120 | -0.324*** | 0.111 |
Coop total records | 0.070* | 0.039 | 0.088** | 0.038 | 0.009 | 0.045 | 0.039 | 0.042 | -0.074** | 0.038 |
Coop own storage | -0.394** | 0.185 | 0.358** | 0.185 | 0.460** | 0.227 | -0.392** | 0.191 | -0.057 | 0.175 |
No of shares (sqrt) | 0.187* | 0.110 | 0.231** | 0.116 | 0.302** | 0.139 | 0.279*** | 0.109 | 0.219** | 0.099 |
Leadership role | 0.055 | 0.161 | 0.32** | 0.162 | 0.247 | 0.193 | 0.42** | 0.178 | 0.251 | 0.156 |
Phone information access | 0.335** | 0.143 | 0.119 | 0.141 | 0.384** | 0.165 | -0.075 | 0.156 | -0.223 | 0.139 |
Crop income (nl) | 0.170*** | 0.066 | 0.147** | 0.062 | -0.012 | 0.054 | 0.077 | 0.058 | 0.086 | 0.053 |
Extension access | 0.372** | 0.151 | 0.153 | 0.150 | 0.469*** | 0.169 | -0.083 | 0.169 | -0.038 | 0.151 |
Membership groups | 0.260* | 0.148 | 0.175 | 0.145 | 0.214 | 0.164 | 0.187 | 0.159 | 0.279* | 0.146 |
Cons | -3.371** | 1.400 | -0.482 | 1.359 | 2.395 | 1.537 | 1.68 | 1.449 | 0.719 | 1.311 |
Participation Intensity | ||||||||||
Coop affiliation | -0.044 | 0.056 | -0.179*** | 0.054 | -0.129*** | 0.038 | 0.113* | 0.060 | 0.139* | 0.074 |
Coop female ratio | 0.352* | 0.194 | -0.118 | 0.176 | -0.189 | 0.130 | 0.198 | 0.170 | 0.580*** | 0.227 |
Coop own storage | -0.071 | 0.057 | 0.072* | 0.056 | 0.023 | 0.042 | -0.031** | 0.051 | -0.063 | 0.065 |
Coop technical staff | 0.107* | 0.060 | -0.084 | 0.061 | -0.045 | 0.044 | -0.11** | 0.055 | -0.022 | 0.071 |
Coop road condition | 0.111** | 0.051 | 0.092* | 0.051 | -0.045 | 0.037 | 0.004 | 0.043 | 0.075 | 0.055 |
Member’s age | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.001 | 0.002 |
Total activities | 0.027* | 0.016 | 0.04** | 0.018 | -0.003 | 0.012 | 0.031** | 0.016 | -0.001 | 0.022 |
Membership in groups | -0.034 | 0.044 | 0.071 | 0.044 | -0.086*** | 0.033 | -0.214*** | 0.038 | -0.056 | 0.048 |
Satisfaction | 0.056* | 0.026 | 0.071*** | 0.028 | 0.004 | 0.018 | 0.036 | 0.024 | 0.091*** | 0.031 |
Cons | 0.156 | 0.163 | 0.213 | 0.182 | 0.980*** | 0.131 | 0.298 | 0.199 | -0.128 | 0.273 |
lambda/mills | -0.168** | 0.077 | -0.184* | 0.099 | -0.252*** | 0.099 | 0.196*** | 0.071 | 0.218** | 0.098 |
Wald chi2(9) | 30.29*** | 33.27*** | 28.86*** | 60.43*** | 25.77*** | |||||
Source: Authors's Field survey (2024). *, **, and *** imply significance at 10%, 5%, and 1%, respectively.
4.0 Discussions
4.1 Participation in co-operative collective action initiatives/services
A
Whereas most studies assume homogeneous participation by regarding members as participants, this study focused on co-operative members, aiming to examine their participation patterns in five key co-operative services, comparing the participants and non-participants, and finally determining the drivers of participation decisions and intensity in co-operative services. The aggregation of members' participation reveals that only 18% of the co-operative members participated in the five co-operative services (Fig. 2), while disaggregation shows significant variability in member participation decisions and intensity across co-operative services (Figs. 1 & 3). This reflects heterogeneity in members’ prioritization of co-operative services, suggesting that a significant number of co-operative members fail to utilize co-operative services. Abate's (2018) and Hando et al. (2022) observed similar findings among Ethiopian agricultural co-operatives, highlighting the need for co-operatives in African countries to review their services and functionality.4.2 Comparison of participants and non-participants in co-operative activities
Previous studies have focused on the differences between the members and non-members of co-operatives. However, the t-test and chi-square reveal significant disparities in farmer and co-operative characteristics between participants and non-participant members (Tables 2 & 3). Larger and older co-operatives attracted more participants in inputs than smaller and newer ones, possibly because of economies of scale and external networks. However, smaller and newer co-operatives attracted more participants in financial services because they are more inclusive, responsive, and have fewer bureaucratic processes.
A
Co-operatives with a higher female ratio had significantly lower participation in inputs and marketing but higher participation in financial services, suggesting gender disparities in resource endowment as discussed in the subsequent section. Co-operatives with a higher proportion of youth members experienced a lower participation in marketing, loans, and savings, suggesting barriers to participation for younger members. Co-operative members' financial charges, including membership fees and share capital, significantly differed among participants and non-participants in loans and savings services. Higher charges led to higher non-participants, suggesting a lack of targeted financial products for lower-income members. The significantly higher participation in inputs among non-affiliated co- contradicts expectations that affiliated co-operatives have more access to subsidized resources from their unions. The converse observation can be attributed to inefficiencies in co-operative unions, leading to weak linkages with their member co-operative societies. Earlier studies highlight the bottlenecks in Ugandan co-operative unions, mainly emanating from liberalization, poor governance, political interference, and dependency on government funding [28].The higher HHD among non-participants in input and marketing suggests economic strain, which might limit investment in inputs, consequently leading to low output, which might be reserved for family consumption. A higher HHD among participants in savings suggests that participation is a financial risk-mitigation measure for future uncertainties among households with more dependents. These findings align with Hando et al. [19] and Fischer and Qaim [25], who found that household size negatively impacts farmer participation in multipurpose co-operatives and collective marketing, respectively. Cropland ratio and crop incomes were higher among participants in inputs, suggesting that members with a higher cropland ratio were more likely to invest in farm inputs, leading to a marketable surplus, hence higher incomes. Conversely, participants in financial services had a lower cropland ratio, suggesting that members with low land holdings rely heavily on co-operative financial services. The significant differences in access to phone information and agricultural extension among participants and non-participants suggest that access to agricultural knowledge promotes the use of agricultural inputs, leading to increased productivity and participation in collective marketing. However, higher access to information among non-participants in financial services suggests that mobile financial services may substitute traditional co-operative services, suggesting that while technology is beneficial, co-operatives must address challenges of increased risk aversion and alternative financing sources. Integrating mobile platforms and diversifying financial products, such as small, short-term loans, could enhance participation in co-operative loans.
4.3 Determinants of member participation in co-operative collective action initiatives
Drawing on empirical evidence, this section provides an intriguing discussion on how member participation in co-operative services is influenced by co-operative and farmer characteristics (Table 4). The first section provides a discussion on significant determinants of participation decisions across the five key co-operative services, while the second section focuses on significant determinants of participation intensity. Notedly, the divergent effects of co-operative and farmer characteristics on participation in various activities provide important policy levers that would otherwise be obscured by assuming “one size fits all” interventions.
4.3.1 Drivers of participation decisions in co-operative collective action initiatives/services
A
The co-operative affiliation to a co-operative union positively influenced the probability of its member participation decision (PD) across most of the CAIs, with a significant effect on member PD in savings and loan services. This means that affiliated co-operatives are likely to have more of their members participating in financial services, probably because affiliation to large unions promotes democratic member control, credibility, and trust, which fosters participation. Besides, the unions provide financial literacy to their primary co-operatives as a strategy to enhance financial independence. This could motivate members' decisions to participate in savings and consequently their eligibility to borrow co-operative loans, which are often perceived to have low interest rates.An increase in the proportion of female members within the co-operative had bidirectional effects on member PD. It significantly reduced the likelihood of members' participation in collective inputs while simultaneously increasing their probability of participation in financial services. This implies that only a few members in women-dominated co-operatives purchase inputs from the co-operative, while the majority save and borrow loans from the co-operative. The reduced likelihood of participation in inputs could be attributed to shifts in decision-making patterns and resource allocation; for instance, men traditionally dominate land ownership and utility decisions. Such gender inequalities perpetuated by cultural norms could interfere with women's autonomy over input procurement, hence reducing their probability of participation in co-operative farm input supply services. [21] demonstrates the exclusion of women and poor farmers from decision-making among Kenyan milk producers and processing organizations. Similarly, [31] argues that despite the proven involvement of women in co-operatives and their positive impacts, their empowerment and decision-making capabilities are not sufficiently developed. The positive effects of the female ratio on savings and loan participation decisions could be attributed to the flexibility of co-operative saving and loan terms that align with women's needs. Moreover, most co-operative savings and loan schemes are informal and decentralized, providing an inclusive environment that protects women from institutional bias and discriminatory collateral that could limit access to credit. Besides, most women have membership with informal savings groups, which could facilitate quick access to loans when the informal groups join co-operatives. The findings resonate with [32], whose study found that women co-operative members in western Uganda were more involved in savings than marketing.
An increase in the proportion of youthful members in co-operatives reduced the probability of member PD across most of the key co-operative services, with significant negative effects on inputs, savings, and loan borrowing. The unidirectional negative influence could stem from limited ownership of production resources, such as land and financial instability. Such resource disparities among youthful members could limit the co-operative from attaining economies of scale and bargaining power, consequently leading to high transactional costs and low perceived benefit of participation in co-operative collective action initiatives. A study by Lukwago et al. [33] highlights land ownership and membership fees as essential drivers of youth participation in agricultural co-operatives in Northern Uganda. Additionally, most young people associate farming with a lower social status compared to other career options. Such a stereotype may discourage their engagement in agricultural activities [34]. The negative effects on PD in financial services could be attributed to higher risk aversion among youthful members, possibly due to financial instability and a preference for digital loan services. Nonetheless, most co-operatives are also anchored in traditional hierarchical structures in which older members dominate decision-making. Such structures marginalize young people, thus depriving them opportunity to strengthen their social capital and networks within the co-operative. The outstanding negative effects of youthful membership on participation call for targeted interventions that enhance youth engagement and structural changes within the co-operatives to ensure youth inclusivity. For instance, co-operatives can harness the potential of youth by adopting adaptive policies such as low collateral loans, digital platforms, and input subsidies while addressing structural barriers to participation. This could transform youth from passive members to active contributors, ensuring co-operative and inclusive growth.
An increase in co-operative membership size significantly increased the probability of member PD in collective inputs while simultaneously reducing their participation in training. A larger membership enhances economies of scale for bulk purchase of inputs, thereby reducing transactional costs and encouraging broader participation in input procurement. Similarly, [18] found a positive correlation between co-operative size and members' engagement in input procurement, attributing it to increased bargaining power and cost-sharing. The inverse relationship between co-operative size and member PD in training could be attributed to several factors, including coordination and logistical constraints, heterogeneity of members' needs, and dilution of individual incentives [13]. Therefore, in larger co-operatives, the likelihood of participation in training is reduced as members feel less accountable or rely on member dissemination. Fischer and Qaim [25] also observed that extensive membership size in Kenyan co-operatives led to reduced participation in capacity-building programs as members perceived low tangible benefits. These dynamics underscore the need for co-operative leaders to consider trade-offs associated with larger co-operative membership and implement policies that support the effective management of co-operative activities. For example, training can be decentralized to sub-groups, and digital online platforms used to train farmers.
A one-year increase in co-operative age significantly reduced the probability of member PD in collective marketing. This could be attributed to generational differences in terms of membership duration, members’ age, and priorities. Such heterogeneity may lead to reduced commitment, especially when younger and newer members feel excluded, as reported by [13]. To improve overall member engagement across co-operative services, older co-operatives must adapt their structures and practices to remain relevant and inclusive for all members.
Co-operative charges in the form of share capital and membership fees had a unanimous negative effect on member PD across most of the co-operative services, with significant effects on training, savings, and loans. Co-operatives rely on these charges for financing their initial operations, but an increase could discourage financially constrained members from participating in essential services. Similarly, Cheyo et al.[12] observed a negative correlation between such charges and groundnut commercialization among co-operative members, while [33] reported that youth engagement in the co-operative was barred by upfront costs. Therefore, while share capital investment represents a member's financial stake in the co-operative and significantly increases the probability of participation in CAIs, co-operatives should diversify into income-generating activities to boost their operational capital and review these charges for the inclusivity of financially constrained members.
Record-keeping in co-operatives increases member participation in all services, except for loan borrowing. Accurate and transparent record-keeping reduces information asymmetry, increasing confidence in services. Records reveal the co-operative's financial health, including existing liabilities and past defaulters, influencing borrowing behavior. This study highlights the role of record-keeping in determining member participation and suggests co-operatives can improve this by fostering an informed environment. Ownership of the storage facility by the co-operative increased the probability of participation in marketing and training, while reducing participation in inputs and savings. The positive effects can be linked to direct control over storage, enabling better management and coordination of produce handling, quality control, and market supply, which encourages members to engage more actively in marketing. Co-operatives also use storage facilities as centralized locations for training workshops, enhancing PD in training. Conversely, ownership of storage infrastructure can shift the co-operative’s focus towards marketing and training with less emphasis on inputs. Reliance on inventory can also reduce participation in savings. Therefore, while studies such as Namubiru et al. [6] advocate for such infrastructural investment, co-operatives must be cognizant of these dynamics and implement strategies to ensure that the presence of storage facilities does not overshadow member engagement in other co-operative services.
Access to agricultural extension services and phone information significantly increased the probability of member participation in inputs, marketing, and training services. The positive impacts could be attributed to critical knowledge and advice that incentivise the utilization of inputs, leading to marketable surplus. The positive effects of phone information signify the role of digital tools in complementing agricultural extension services by providing timely communication and information dissemination, thus reducing cost and time barriers that might otherwise inhibit participation. Previous studies reported a positive correlation between access to extension services, information technology, and farmer participation in CAIs [22], [35], [36].
4.3.2 Drivers of participation intensity in co-operative collective action initiatives/services
The co-operative affiliation to a union significantly reduced PI in collective marketing and training services while simultaneously increasing PI in financial services. The negative effects imply that, whereas members of co-operatives produce a marketable surplus, only a smaller proportion of their produce is marketed through the co-operative. The findings contradict expectations, as co-operative unions should be key in connecting primary co-operatives to external markets and training opportunities, thus promoting active member participation. Nonetheless, this finding suggests ineffective collaboration between co-operative unions, primary agricultural co-operatives, and members. The effectiveness of collective marketing largely depends on market availability and members' ability to bulk produce. Co-operatives often rely on unions as their primary market, while unions rely on large-scale traders, leading to longer bulking and storage duration to meet the market demand. Members with urgent financial needs may prefer timely payment and may perceive low mutual benefits of selling through the co-operative, leading to reduced PI in collective marketing. Moreover, in the absence of production and marketing contracts, members may perceive low mutual benefits of selling through the co-operative. These findings corroborate [37], who observed that Tanzanian co-operative unions were less effective in marketing the coffee collected by primary co-operatives due to delayed payment and accountability. Ugandan co-operative unions should devise new funding models to cater to members with urgent financial needs. The negative effect on member PI in training services implies that members of co-operatives affiliated to unions do not attend most of the training organized by their primary co-operatives. This could be attributed to the fact that affiliated primary co-operatives often rely on training provided by the union, which is usually scarce due to limited facilitation [27]. The positive influence on members' PI in financial services implies that affiliated co-operatives provide flexible financial services, encouraging members to exclusively save and borrow more from their co-operatives.
An increased proportion of female members in co-operatives significantly increased members' PI in financial lending and input services, implying intensive loan borrowing and input purchasing from their co-operatives. The positive effects on farm input PI suggest that, whereas women had limited decision-making over input purchase, they strategically focus on cost-effectiveness, as most co-operatives sell inputs at subsidized prices. The significantly high PI in loan borrowing among women-dominated co-operatives suggests that most women exclusively depend on co-operative financial services, which are largely decentralized, thus facilitating easier and quicker access to credit. Budi et al. [38] revealed that although women's participation in Cameroonian co-operatives was high, they still faced several barriers, including low education, subordination, domestic duties, cultural barriers, and low access to resources. The divergent effect of women's membership on input participation underscores the need for targeted gender-inclusive interventions that address cultural barriers and promote women's participation across all co-operative functions.
A one-unit increase in the number of CAIs a member participates in significantly increases PI in inputs, marketing, and savings. This finding suggests that participation in one service can serve as a gateway for broader engagement in other co-operative services. This could be attributed to the social capital and networking, which build trust among members, thereby increasing their willingness to participate intensively in additional co-operative services. Similarly, whereas membership with other groups increased the probability of participation in all co-operative services, it significantly reduced the intensity of engagement in co-operative services. The significant positive effects on inputs and savings PD can be attributed to the benefits associated with social capital networking. However, membership in multiple groups could substitute for co-operative services. For instance, members reported that groups formed by NGOs offered subsidized inputs, fair prices, and instant payments compared to the co-operative. Most members would prioritize active participation in such groups over patronizing their co-operatives. Similar findings were reported by [18]. Therefore, to withstand competition and enhance members’ patronage, the primary agricultural co-operatives in Uganda should devise strategic financial plans to ensure efficiency and effectiveness in service delivery.
On the other hand, the good condition of feeder roads around the co-operative increased member PI in most co-operative CAIs, with a significant increase in collective input procurement and marketing. Better roads support the mobility of service providers and traders in remote areas, thus facilitating the delivery of inputs and uptake of produce from co-operatives. This can enhance the efficiency of co-operative activities, thereby increasing members' participation intensity. Similarly, [17] found that wards with advanced transportation infrastructure and markets were likely to experience higher membership and patronage of co-operative activities. Therefore, the government needs to upgrade roads in remote areas where the majority of the co-operatives are located to enhance members' patronage of co-operative activities. Technical staffing significantly reduced members' PI in inputs, loans, and savings, contradicting previous findings [15],[35]. The divergence can be attributed to the fact that the majority of co-operatives employ technical staff for specialized activities such as agro-processing, which has minimal linkage with members' training on inputs and financial literacy.
Whereas multiple studies consider access to extension services as a determinant of participation in farmer organizations, little is documented about technical staffing on members' participation in co-operative services. The current finding suggests that the presence of technical staff in co-operatives may not always lead to increased participation in co-operative services. Technical staffing reduced member PI across most of the CAIs except for inputs. The adverse effects could be attributed to the fact that technical staff in co-operatives are always assigned specific duties such as agro-processing and machinery operations; hence, their presence may only promote participation in specific CAIs. Therefore, whereas investing in human capital development within co-operatives can serve as an effective strategy for enhancing member engagement, it's crucial to train technical staff regarding co-operative principles to mitigate their antagonistic effects on members' PI in other crucial co-operative services.
Phone information access significantly improved the probability of members’ participation in input procurement and training. The positive effects signify the role of digital information technology in advancing efficient access to information. Timely dissemination of information would encourage members to plan and participate in the co-operative training. Similarly, members of co-operatives can access information on good agricultural practices, weather forecasts, and market prices through mobile applications and direct SMS, hence guiding input purchase and participation in training programs [39]. These findings are consistent with [22], whose study found a significant positive influence of phone ownership on collective marketing decisions. Similarly, [36] reported the positive impact of mobile phones in disseminating training material to members, thus reducing cost and time barriers that might otherwise inhibit participation.
A one-unit increase in the number of co-operative activities in which a member participates significantly enhances PI in collective input procurement, marketing, and savings. This finding suggests that involvement in one activity can serve as a gateway for broader engagement in co-operative CAIs. This could be attributed to the social capital and networking, which build trust and reciprocity among members, thereby increasing their willingness to participate in additional co-operative CAIs. In the course of interaction, members are likely to get current information on all CAIs offered by the co-operative, consequently motivating them to intensively engage in multiple activities.
Similarly, Membership in other agricultural groups positively influenced the probability of member PD in all co-operative CAIs while simultaneously reducing their PI in most CAIs. The significant positive effects on collective input procurement and saving can be attributed to the benefits associated with social capital networking. However, membership in multiple groups could substitute for co-operative services. For instance, members reported that groups formed by NGOs offered subsidized inputs, fair prices, and instant payments compared to the co-operative. Most members would prioritize active participation in such groups over patronizing their co-operatives as reported by [17]. Therefore, to withstand competition and enhance members’ patronage, the primary agricultural co-operatives in Uganda should devise strategic financial plans to ensure efficiency and effectiveness in service delivery.
A unit increase in the number of members' share capital investment significantly enhanced their probability of participation across all the CAIs except for training. Share investment represents a member's financial stake in the co-operative, thus motivating participation in most of the co-operative activities, as their financial success is tied to co-operative achievement. These findings are consistent with Hando et al. [18], whose study found a positive correlation between members' shareholding and participation levels in activities of Ethiopian multipurpose primary co-operatives. Due to the undisputed positive effects of shareholding contribution on member participation in co-operative activities, there is a need for primary agricultural co-operatives in developing countries to review their capital contribution policies for the enhancement of member compliance and, consequently, participation in co-operative activities.
Conclusion
The concerted efforts to revitalize agricultural co-operatives have not resulted in significant members' participation and patronage of co-operative services, consequently affecting their sustainability. Farmer participation is largely conceptualized based on membership status in a co-operative, assuming homogeneous participation among members. Besides, extensive literature is skewed toward participation in co-operative governance and collective marketing. Therefore, assessing drivers of membership participation decisions and intensity in various services provided by agricultural cooperatives provides a profound understanding of the dynamics of farmer participation in agricultural cooperatives.
The findings indicate that whereas co-operatives provide a bundle of services (inputs, marketing, training, savings, and loans), less than a quarter (18%) of members participate in all these services. Huge variability also exists in member participation decisions and intensity in co-operative services.
A
Co-operative charges, ratio of youthful members, and co-operative technical staffing negatively impacted member participation across most of the co-operative CAIs. Co-operative affiliation with a union, female ratio, record keeping, storage facility, and co-operative size, Informational access through phone and extension services, crop income, and the number of co-operative activities a farmer engages had divergent effects on member participation. A
Members’ shareholding, satisfaction, and leadership roles in the co-operative positively impacted participation aligning with theoretical postulation, while Membership in other groups negatively impacted participation intensity across most of the co-operative CAIs.A
A
This study makes four key contributions to the literature on co-operative services and participation. Firstly, beyond collective marketing, we provide insights into participation across five distinct co-operative services, namely, input supply, marketing, training, savings, and lending. This provides a more holistic picture of members' prioritization of co-operative services and variability of involvement. Secondly, beyond participation based on membership status and participation intensity based on composite indices, we employ the Heckman selection model to simultaneously analyze the determinants of members’ participation decisions and participation intensity across the multiple co-operative services using actual PI computations. The two-step approach yields insights into how and why members not only participate but also differ in how much they engage, thus helping in designing targeted interventions. Thirdly, the integration of co-operative and member characteristics as determinants of participation provides a dual perspective, capturing the complex bidirectional nature of co-operative service delivery and participation, thus enriching the explanatory power and practical applicability of findings. Moreover, the divergent effects of co-operative and farmer characteristics on participation in various services provide important policy levers that would otherwise be obscured by assuming “one size fits all”. A
Lastly, the study highlights the critical role of variables such as co-operative affiliation, co-operative female ratio and youth ratio, technical staffing, and member satisfaction that are rarely explored, yet they are critical determinants of members’ participation in co-operative services.Policy recommendations
Following the variability in prioritization of co-operative services and low patronage in co-operative activities, we recommend that co-operatives focus on demand-driven CAIs, leverage external networks for infrastructural development, review their equity charges, and customize their services for optimal membership participation, particularly women and youth. The divergent effects of co-operative and member characteristics on members' participation decisions require a careful participatory approach in formulating mitigation strategies. While agricultural co-operatives are encouraged to advocate for inclusivity, the outstanding negative impact of youth and partly women-dominated membership on participation warrants a review of co-operative inclusivity. Co-operative programs should be designed to address diverse farmer profiles, including gender-specific initiatives and youth-targeted activities.
Study limitations and areas for further research
The focus on Northern Uganda, a post-conflict region where agricultural cooperatives are re-focusing on multiple commodities and services rather than crop-specific may limit generalization of the results to other study contexts.
A
The study, therefore, recommends similar studies in other stable regions and the use of a mixed-method approach to substantiate members' decimal participation behavior in co-operative services.A
Funding statement:
This article is part of a PhD study funded by the Mobility to Train Agribusiness and Food Systems Scientists for African Agriculture (TAFSA) project at Gulu University, Uganda.
Ethical approval and consent to participate:
A
A
The study was approved by Gulu University Research Ethics Committee in accordance with guidelines set by the Uganda National Council for Science and Technology. A
Informed Consent to participate was obtained from all participants.Clinical trial number
Not applicable
A
Author Contribution
M.J.O: Conceptualizing the study, conducting data collection and analysis, writing, and editing the manuscript. SW. K and W.O supervised the entire process, from conceptualization, development of data collection tools, reviewing, and approval of the final manuscript.
Competing Interest:
The authors declare no conflict of interest that could influence the work published in this article.
A
Data Availability
The data supporting this study's findings are available from the corresponding author upon reasonable request.
A
Acknowledgement
The author acknowledges TAFSA for funding the PhD studies and all the stakeholders who participated in the study, including agricultural co-operative farmer member, co-operative union leaders and research assistants.
List of Abbreviations
CAIs
Collective Action Intiatives
CAT
Collective Action Theory
Coops
Cooperative
FOs
Farmer organizations
GUREC
Gulu University Research and Ethics Committee
HHD
Household Dependency Ratio
IMR
Inverse Mills Ratio
MVP
Multivariate Probit
MVT
Multivariate Tobit
NGOs
Non-Governmental Organizations
NL
Natural logarithms
PD
Participation Decision
PI
Participation Intensity
RCT
Rational Choice Theory
RPOs
Rural Producer Organizations
SQRT
Squareroot
UNCST
Uganda National Council of Science and Technology
Electronic Supplementary Material
Below is the link to the electronic supplementary material
Reference
1. S. K. Lowder, M. V. Sánchez, and R. Bertini, “Which farms feed the world and has farmland become more concentrated?” World Dev., vol. 142, p. 105455, 2021, doi: 10.1016/j.worlddev.2021.105455.
2. D. C. Midamba and O. Kizito, “Determinants of access to training on post–harvest loss management among maize farmers in Uganda: a binary logistic regression approach,” Cogent Econ. Financ., vol. 10, no. 1, 2022, doi: 10.1080/23322039.2022.2148359.
3. M. Muthoka et al., “Socio-economic impacts of climate change and adaptation actions among smallholder fish farmers in Sub‐Saharan Africa,” Aquac. Fish Fish., vol. 4, no. 3, pp. 1–13, 2024, doi: 10.1002/aff2.182.
4. M. Abraham, L. Verteramo Chiu, E. Joshi, M. Ali Ilahi, and P. Pingali, “Aggregation models and small farm commercialization – A scoping review of the global literature,” Food Policy, vol. 110, no. November 2021, p. 102299, 2022, doi: 10.1016/j.foodpol.2022.102299.
5. L. Bizikova et al., “A scoping review of the contributions of farmers’ organizations to smallholder agriculture,” Nat. Food, vol. 1, no. 10, pp. 620–630, 2020, doi: 10.1038/s43016-020-00164-x.
6. M. Kunzekweguta, M. Pigeon, E. Micheels, C. Henry, and M. G. Nosworthy, “Understanding the role of co-operatives in enhancing food security in East Africa,” 2025.
7. M. Namubiru, F. N. Kijjambu, and J. Rwakihembo, “The Future of Farmer organisations: Opportunities and Challenges in Achieving Household Food Security. Evidence from Acholi Sub-region, Northern Uganda.,” J. Dev. Econ., vol. 4, no. 1, pp. 16–34, 2022, doi: 10.47672/jde.955.
8. A. G. Geffersa, “Agricultural cooperative membership and welfare of maize farmers in Ethiopia : Insights from panel data analysis,” no. December 2023, 2024, doi: 10.1111/apce.12481.
9. R. Nangobi, P. M. Mshenga, and B. Mugonola, “Sustainability of farmer organizations in Uganda: the role of organizational and institutional factors,” Discov. Sustain., vol. 5, no. 1, 2024, doi: 10.1007/s43621-024-00403-2.
10. D. Apparao, E. Garnevska, and N. Shadbolt, “Examining commitment, heterogeneity and social capital within the membership base of agricultural co-operatives—A conceptual framework,” J. Co-op. Organ. Manag., vol. 7, no. 1, pp. 42–50, 2019, doi: 10.1016/j.jcom.2019.03.003.
11. J. Grashuis and S. U. Ye, “A Review of the Empirical Literature on Farmer Cooperatives: Performance, Ownership and Governance, Finance, and Member Attitude,” Ann. Public Coop. Econ., vol. 90, no. 1, pp. 77–102, 2019, doi: 10.1111/apce.12205.
12. E. J. Cheyo, E. Nyuyki Mainimo, D. M. Okello, W. Odongo, and S. W. Kalule, “Participation and commercialization in producer cooperatives: Insights on the role of cooperative functioning and equity strategies,” Ann. Public Coop. Econ., vol. 95, no. 3, pp. 815–834, 2024, doi: 10.1111/apce.12462.
13. J. Grashuis and M. L. Cook, “A structural equation model of cooperative member satisfaction and long-term commitment,” Int. Food Agribus. Manag. Rev., vol. 22, no. 2, pp. 247–263, 2019, doi: 10.22434/IFAMR2018.0101.
14. J. Grashuis, “Determinants of collective action by farm producers: A meta-analysis of the likelihood of co-operative membership,” J. Rural Stud., vol. 117, p. 103639, Jul. 2025, doi: 10.1016/j.jrurstud.2025.103639.
15. A. Gyau, M. Mbugua, and J. Oduol, “Determinants of participation and intensity of participation in collective action: Evidence from smallholder avocado farmers in Kenya,” J. Chain Netw. Sci., vol. 16, no. 2, pp. 147–156, 2016, doi: 10.3920/JCNS2015.0011.
16. L. Cele, “Youth perceptions of, and willingness to join Irish dairy cooperatives and their governance,” J. Co-op. Organ. Manag., vol. 10, no. 2, p. 100180, 2022, doi: 10.1016/j.jcom.2022.100180.
17. M. Buang and A. Abu Samah, “Co-operative Governance: A Systematic Review of Member Participation,” Int. J. Acad. Res. Bus. Soc. Sci., vol.. 11, no. 10, 2021, doi: 10.6007/ijarbss/v11-i10/10997.
18. G. T. Abate, “Drivers of agricultural cooperative formation and farmers’ membership and patronage decisions in Ethiopia,” J. Co-op. Organ. Manag., vol. 6, no. 2, pp. 53–63, 2018, doi: 10.1016/j.jcom.2018.06.002.
19. H. Hando, M. Senapathy, and E. Bojago, “Determinants of Member Participation in Activities of Multipurpose Primary Cooperatives: The Case of Kindo Koysha District, Southern Ethiopia,” Adv. Agric., vol. 2022, 2022, doi: 10.1155/2022/3569016.
20. R. Karunakaran, B. Borji, and A. Sherefa, “Determinants of Co-Operatives Members ’ Participation in Input and Output Marketing By Multipurpose Co-Operatives - an Empirical Analysis in Tigray Region, Ethiopia,” Malaysian J. Co-op. Stud., no. December, 2024.
21. M. Mwambi, J. Bijman, and P. Mshenga, “Which type of producer organization is (more) inclusive? Dynamics of farmers’ membership and participation in the decision-making process,” Ann. Public Coop. Econ., vol. 91, no. 2, pp. 213–236, 2020, doi: 10.1111/apce.12269.
22. A. Abdul-Rahaman and A. Abdulai, “Farmer groups, collective marketing and smallholder farm performance in rural Ghana,” J. Agribus. Dev. Emerg. Econ., vol. 10, no. 5, pp. 511–527, 2020, doi: 10.1108/JADEE-07-2019-0095.
23. S. Mutonyi, “The effect of collective action on smallholder income and asset holdings in Kenya,” World Dev. Perspect., vol. 14, no. February, p. 100099, 2019, doi: 10.1016/j.wdp.2019.02.010.
24. M. Olson, The Logic of Collective Action, Vol. 124. Harvard University Press, 1971. doi: https://doi.org/10.2307/j.ctvjsf3ts.
25. E. Fischer and M. Qaim, “Smallholder farmers and collective action: What determines the intensity of participation?” J. Agric. Econ., vol. 65, no. 3, pp. 683–702, 2014, doi: 10.1111/1477-9552.12060.
26. UBOS, “Annual agricultural survey,” Report, no. 2, pp. 2–5, 2022, [Online]. Available: https://eur-lex.europa.eu/legal-content/PT/TXT/PDF/?uri=CELEX:32016R0679&from=PT%0Ahttp://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:52012PC0011:pt:NOT
27. L. Okello, “The Cooperative Movement and the Challenge of Development: A Search for Alternative Wealth Creation and Citizen Vitality Approaches in Uganda,” 2013. [Online]. Available: https://www.uhuruinstitute.org/wp-content/uploads/2018/06/COOPERATIVES-RESEARCH-REPORT-DEC-2013.pdf
28. A. Kayongo, A. Guloba, J. Muvawala, and A. Ssali, “A Framework for Strengthening and Sustaining Cooperatives for Socio-Economic Transformation in Uganda,” Appl. Econ. Financ., vol. 9, no. 2, p. 79, 2022, doi: 10.11114/aef.v9i2.5554.
29. A. S. Singh and M. B. Masuku, “Sampling techniques & determination of sample size in applied statistics research,” Inwood Mag., vol. II, no. 96, pp. 32–33, 2011.
30. P. Gopal Biswas and M. M. Kirmani, “Application of Multivariate Probit Model in Econometric analysis using STATA program mvprobit,” Int. J. Res. Econ. Soc. Sci. Available, vol. 7, no. 12, pp. 263–274, 2017, [Online]. Available: http://euroasiapub.org
31. E. Lecoutere, “The impact of agricultural co-operatives on women’s empowerment: Evidence from Uganda,” J. Co-op. Organ. Manag., vol. 5, no. 1, pp. 14–27, 2017, doi: 10.1016/j.jcom.2017.03.001.
32. F. Meier zu Selhausen, “What Determines Women’s Participation in Collective Action? Evidence from a Western Ugandan Coffee Cooperative,” Fem. Econ., vol. 22, no. 1, pp. 130–157, 2016, doi: 10.1080/13545701.2015.1088960.
33. M. M. Lukwago, A. Pamela, and B. Mugonola, “Determinants of Youths’ Participation in Agricultural Cooperatives: Evidence from Northern Uganda,” J. Int. Food Agribus. Mark., vol. 0, no. 0, pp. 1–20, 2024, doi: 10.1080/08974438.2024.2323559.
34. W. Geza, M. Ngidi, T. Ojo, A. A. Adetoro, R. Slotow, and T. Mabhaudhi, “Youth participation in agriculture: A scoping review,” Sustain., vol. 13, no. 16, pp. 1–15, 2021, doi: 10.3390/su13169120.
35. S. O. Olawuyi and A. Mushunje, “Heterogeneous treatment effect estimation of participation in collective actions and adoption of climate-smart farming technologies in South–West Nigeria,” GeoJournal, vol. 85, no. 5, pp. 1309–1323, 2020, doi: 10.1007/s10708-019-10024-2.
36. G. K. Munyegera and T. Matsumoto, “ICT for financial access: Mobile money and the financial behavior of rural households in Uganda,” Rev. Dev. Econ., vol. 22, no. 1, pp. 45–66, 2018, doi: 10.1111/rode.12327.
37. N. Towo, G. J. Kweka, and D. Aikaruwa, “Barriers to the Accountability Mechanisms of Co-operative Unions on Primary Agricultural Marketing Co-operative Societies,” 2023. doi: https://doi.org/10.58547/1.v8i1.96.
38. F. T. Budi, F. A. Amungwa, and I. N. Manu, “Women’s Participation in Agricultural Cooperative Activities in the Era of Liberalization in the North West and South West Regions of Cameroon,” Asian J. Agric. Extension, Econ. Sociol., vol. 39, no. 1, pp. 136–152, 2021, doi: 10.9734/ajaees/2021/v39i130517.
39. K. Dionysious, “The use of Information Communication and Technology in advancement of African agriculture,” African J. Agric. Res., vol. 13, no. 39, pp. 2025–2036, 2018, doi: 10.5897/ajar2018.13300.