Evaluating the Economic Impact of Market Participation on the Well-Being of Smallholder Irrigators: Evidence from the Eastern Cape Province, South Africa

Abstract

While increased market participation and irrigation adoption hold promise, a critical gap exists in understanding the real-world impacts of these interventions on the welfare of smallholder farmers. Despite the theoretical advantages, little is known about the extent to which market participation improves smallholder livelihoods. Our sample of 250 farmers comprised both members and non-members of irrigation schemes in the Eastern Cape of South Africa, who were selected purposively and by random sampling, respectively. Propensity score matching (PSM) was employed to evaluate the impact of market participation on the welfare of smallholder farmers. The study confirmed moderately higher market participation among irrigators than non-irrigators. The findings also revealed that market participation significantly enhances income levels among smallholder farmers, with participants earning approximately 45% more than non-participants across all matching methods. This study reaffirms the critical role of market access in improving farmers’ well-being and highlights the need for policy interventions that facilitate smallholder integration into markets. It recommends sustained support for farmer training, the adoption of innovative agricultural practices to boost productivity, and improved access to extension services. These findings afford the significant insights that policymakers need to formulate effective poverty alleviation strategies and design well-structured and effective schemes that foster smallholder farmers’ market participation.

1. Introduction

This research examines the impacts of market participation on the household well-being of smallholder irrigation farmers. Smallholders and those participating in irrigation farming are seen as key contributors to rural food supply and livelihoods. They are encouraged to transition from subsistence farming to a more market-oriented approach. This shift is driven by the pressing issue of food insecurity across sub-Saharan Africa (SSA), where approximately 20.3% of the population struggles to meet their basic food consumption needs [1]. As food demand continues rising due to population growth and changing consumption patterns, the pressure on the supply chain intensifies. This situation presents significant opportunities for smallholder farmers and small-scale irrigation schemes in SSA to effectively address supply-side challenges, ultimately enhancing food security and improving household well-being [2,3].

In South Africa, the constitutional recognition of the right to food places food systems and agricultural development squarely on the national development agenda. Hence, the policy prioritizes the inclusion of black subsistence farmers and small-scale irrigation schemes into the mainstream economy [4]. Despite the agricultural sector’s modest contribution to the national gross domestic product (GDP), over two million small-scale or household farmers reside in rural areas, often isolated from critical markets and facing limited livelihood options, depend primarily on agriculture as their primary source of sustenance [5,6]. However, operating at a subsistence level prevents these smallholders from achieving the economies of scale needed to make investments in irrigation systems financially viable [7].

Small-scale irrigation schemes are recommended to enhance agricultural productivity, food security, crop yields, and livelihoods. Adopting irrigation technologies can significantly benefit rural households by enabling farmers to optimize water use, grow various crops, and maintain consistent yearly production [8]. Furthermore, South Africa ranks among the driest and most water-scarce [9] countries globally in terms of water availability per capita, with annual rainfall amounting to about half of the global average. While about 70% of the country’s agriculture relies on rain-fed methods, only 35% of the land receives enough rainfall to support successful dryland farming [8]. The combination of low rainfall and high evaporative demand leads to a water deficit that hinders dryland crop production in much of the country, making irrigated agriculture a crucial alternative. Despite covering a relatively small portion (3%) of irrigated land, small-scale irrigation schemes are significant due to their direct impact on rural communities. Consequently, many smallholder irrigation schemes have been revitalized, and new ones have been established across South Africa [10]. The primary purpose of these schemes is to enhance rural livelihoods by promoting sustainable crop production, thereby contributing to food security and reducing poverty [8]. However, access to suitable irrigation systems tends to favour wealthier farmers and those growing high-value crops, who benefit more from public and private investments in farmer-led irrigation expansion [11].

Most smallholder farmers and minor irrigation schemes in the country are concentrated in the eastern regions, characterized by high levels of poverty and unemployment. This subjects farmers to limited access to credit, technical knowledge, and efficient irrigation technologies, hindering their ability to utilize irrigation effectively and improve their livelihoods. Varying in size, management, and technical sophistication, these schemes have not led to a corresponding income increase for impoverished households due to significant inequalities in access to and distribution of resources [12]. Consequently, most households’ livelihoods have been compromised, leaving smallholders unable to achieve financial stability or meet their basic needs. Livelihoods require more than just food sustenance [13], farmers must ensure food security while generating the necessary income to meet immediate consumption and social needs [14]. Subsistence food crop production alone cannot meaningfully enhance rural incomes without incorporating market-oriented production systems.

Given these challenges, some scholars propose that the subsistence system of small-scale farmers should be replaced with more economically viable options, diversifying production beyond subsistence levels [15]. This transformation requires enhancing agricultural production systems, increasing commercialization, and focusing on high-value crops [16], alongside initiatives that develop efficient and well-functioning markets [17]. Access to markets thus serves as a mechanism for ensuring profitability through exchange opportunities, enabling smallholders to secure a meaningful livelihood and income [17,18,19,20,21,22,23]. Empirical evidence worldwide supports this perspective, demonstrating that market participation leads to higher incomes and improved welfare outcomes for smallholder farmers. For example, Dey and Singh [24] found that in India, market participation significantly enhances both income and per capita consumption expenditure. Their study reported that market participants had an average yearly farm income exceeding INR 48,000, which is approximately ZAR 10,560 or USD 576. They further indicated that this enables farmers to spend much more on their consumption needs than non-participants. This implies that engaging in markets not only boosts farmers’ income but also improves their ability to meet their consumption needs, which can contribute to better overall welfare.

In Zambia, smallholder farmers who actively participate in markets have experienced a substantial 242% increase in total household income, leading to improved living standards [25]. This study also demonstrated the equity-enhancing potential of market engagement, as it boosts total household income and reduces gender income disparities. In Ethiopia, a recent study observed a positive link between commercialization and household income among maize farmers, leading to increased wealth through asset ownership and accumulation [26]. This study further found that commercialization reduces both income poverty rates and the poverty gap, with the most significant welfare improvements occurring among the poorest households, highlighting its potential to reduce rural inequality [26]. Setoboli et al. [27], while focusing on productivity and market potential in goat production, found similar results noting that although farmers face numerous challenges, those who participate in markets earn a higher gross margin than non-market participants, providing them with significant welfare benefits.

However, Huka et al. [16] are of the opinion that, while participation in markets in general can lead to positive outcomes, a specific focus on high-value markets is a more effective pathway for long-term improvement in income. They argue that high-value markets, which include cash crops and other high-demand products, are more lucrative for agricultural produce compared to traditional markets and enables smallholder farmers to access better returns, thereby significantly boosting their income. This increased income helps farmers expand their asset base, which contributes to long-term economic stability and improved household welfare.

Beyond income gains, market participation significantly contributes to food security and overall household well-being. Viana et al. [28] and Villar et al. [29] emphasized that farmers engaged in agricultural markets experience improved food security, as increased income allows them to invest in better nutrition and enhance their household well-being. Ma et al. [30] further highlighted that market participation enables farmers to access a broader variety of foods, which improves dietary diversity and nutritional outcomes. This, in turn, strengthens their financial resilience by providing greater security against seasonal fluctuations and market uncertainties. Additionally, market participation offers a more stable and diversified income stream, enabling farmers to better cope with economic challenges [31]. However, achieving such success depends on creating a conducive environment that allows smallholders to access necessary resources such as market information and effective partnerships [32]. Therefore, efforts in South Africa are focused on policies that establish an enabling environment that ensures equal access to opportunities for resource-limited small-scale farmers. This process involves market deregulation, trade liberalization, investments in agricultural funding, and revitalization of dormant small-scale irrigation schemes [33,34]. The aim is to enhance small-scale farmers’ ability to produce high-value crops, thereby increasing their chances of earning higher incomes and improving their general well-being.

When agricultural production is market-oriented, it allows farmers to specialize in goods with a comparative advantage, enabling them to trade surplus goods for other needed items and services they are less efficient at producing. This helps farmers to diversify their income sources, benefit from economies of scale, and adopt advanced technologies, contributing to faster growth in total productivity [35].

However, while market participation offers significant benefits, not all farmers experience positive outcomes. Many smallholder farmers face numerous constraints that hinder their ability to engage in and benefit from agricultural markets. These barriers include limited access to credit, inadequate market information, small land sizes, and insufficient agricultural training [7]. Without access to affordable credit, smallholders struggle to invest in essential farm inputs such as improved seeds, fertilizers, and mechanization, limiting their productivity and competitiveness in markets [36]. Additionally, poor access to timely and accurate market information prevents farmers from making informed production and selling decisions, often resulting in price exploitation by intermediaries and an inability to negotiate fair prices for their produce [37]. Structural challenges such as high transaction costs, poor road infrastructure, and limited access to modern production equipment and irrigation systems further exacerbate market exclusion for smallholder farmers [36,37]. Transport costs, including remote locations that lead to high transportation costs, and information asymmetry that affects decision-making, for instance, account for a significant proportion of marketing expenses, making it difficult for farmers in remote rural areas to access profitable urban markets [12,30]. Inadequate infrastructure, including poorly maintained roads and lack of storage facilities, contributes to post-harvest losses and reduces farmers’ ability to sell their produce at optimal prices [38]. Additionally, a lack of business skills limits farmers’ potential to engage especially in high-value markets [16]. Furthermore, the absence of irrigation systems forces many smallholders to rely on rain-fed agriculture, making their production highly susceptible to climate variability and unpredictable weather patterns [33].

These systemic constraints have contributed to low market participation rates among smallholder farmers. Studies indicate that in specific areas within the Eastern Cape and Limpopo provinces, only around 33% of emerging farmers participated in market activities, and participation rates for small-scale irrigation farmers were as low as 38% [37]. This limited engagement further restricts farmers’ ability to integrate into formal markets, reducing their income potential and economic resilience. Many smallholders who attempt to engage in the market frequently face substantial obstacles in maintaining the standards required for sustained viability and competitiveness, leading to high exit rates and irregular market participation [33,38,39,40]. Compliance with food safety regulations, grading standards, and certification requirements remains a significant barrier for many smallholder farmers, particularly those producing perishable commodities such as fruits and vegetables [7]. As a result, smallholders continue to experience low levels of market integration, reinforcing the cycle of rural poverty and economic marginalization [12,41]. This situation emphasizes the need for effective support mechanisms, such as cooperatives and inclusive value chain models, to improve smallholder market access and economic outcomes [26].

While a growing body of literature highlights the potential of smallholder irrigation systems to alleviate some of these challenges by improving productivity, in South Africa [34,42], a significant gap remains in empirical research that directly assesses the impact of market participation on the overall welfare of these farmers. Irrigation can increase crop yields and stabilize production, but it does not automatically guarantee that farmers can overcome market entry barriers, nor does it ensure sustained improvements in income and well-being. However, it is insufficient to merely acknowledge the potential benefits of market participation without concretely evaluating the long-term changes it brings to income, food security, and overall household well-being. Moving beyond assumptions and examining whether market participation indeed translates into better living standards is essential. Without such assessments, the full potential of market-oriented farming remains uncertain, and it is difficult to gauge whether the promoted strategies are achieving their intended outcomes in poverty alleviation and food security. This lack of rigorous, outcome-based research leaves policymakers and development practitioners without the necessary data to design interventions that effectively address the needs of smallholders and ensure that market participation delivers meaningful and lasting improvements to their livelihoods. A critical gap exists in understanding the real-world impacts of these interventions on the welfare of smallholder farmers. This underscores the importance of further empirical research to evaluate the effectiveness of market-oriented farming in alleviating poverty and enhancing the well-being of rural households. Therefore, this research provides an essential step towards filling this gap by providing evidence of the impacts of market participation on smallholder irrigation farmers’ well-being. This approach provides insights into whether commercializing smallholders and markets contributes to smallholder farmers’ economic empowerment and sustainability in regions like the Eastern Cape.

The paper is structured as follows: The following section describes the theoretical framework which underpins this study. The third section describes the methodology used in the study. The fourth section presents the results, followed by a discussion in the fifth section. The final section concludes the paper and offers recommendations.

2. Theoretical Framework

This study is grounded in random utility theory, which provides a framework for analyzing the decision-making processes of smallholder farmers regarding market participation. Random utility theory views smallholder farmers as rational decision-makers who evaluate their choices based on expected utility gains. Accordingly, whether acting as farmers or farmworkers, poor households employ various strategies to achieve food self-sufficiency and generate income to meet immediate consumption needs and social obligations [43,44]. According to this theory, farmers will engage in market activities if the perceived utility from participation, denoted as ${\mathrm{U}}_i^M$ exceeds the utility of not participating, represented as ${\mathrm{U}}_i^N$ [17]. In essence, a farmer will engage in markets when the net utility, denoted as ${\mathrm{U}}_i^{*}$, which is the difference between ${\mathrm{U}}_i^M$ and ${\mathrm{U}}_i^N$, is greater than zero.

(1)${\mathrm{U}}_i^{*}={\mathrm{U}}_i^M-{\mathrm{U}}_i^N$

The unobserved net utility ${\mathrm{U}}_i^{*}$ can be expressed as a function of observable factors within a latent variable model:

(2)${\mathrm{U}}_i^{*}=\mathsf{\alpha }{\mathrm{x}}_{\mathrm{i}}+{\mathsf{\epsilon }}_{\mathrm{i}}$

In this model, ${\mathrm{U}}_i$ is a binary indicator variable, taking the value of 1 if the net utility ${\mathrm{U}}_i^{*}$ is greater than zero, indicating market participation, and 0 if not. α is a vector of parameters to be estimated, capturing the influence of various factors on market participation; x_i is a vector representing observable characteristics of the farmer, such as gender, age, education level, farm size, and market information; and ε_i stands for the error term, accounting for unobservable factors that influence a farmer’s decision-making process.

The decision to participate in markets is then modelled as a binary outcome based on the latent utility ${\mathrm{U}}_i^{*}$. Specifically, ${\mathrm{U}}_i$ is a binary indicator variable, taking the value of 1 if ${\mathrm{U}}_i^{*}$ is greater than zero, indicating market participation, and 0 if ${\mathrm{U}}_i^{*}$ is less than or equal to zero. This can be expressed as follows:

(3)${\mathrm{U}}_i=\left\{\begin{array}{c}1, \mathrm{i}\mathrm{f} {\mathrm{U}}_i^{\mathrm{*}}>0 \\ 0, \mathrm{i}\mathrm{f} {\mathrm{U}}_i^{\mathrm{*}}\le 0\end{array}\right.$

This expression captures the dynamic decision-making process in which farmers weigh the costs and benefits of market involvement by comparing the utility gained from market participation with that of not participating [45,46]. Essentially, farmers evaluate whether the benefits of engaging with the market outweigh the associated costs. The central premise is that farmers weigh the perceived advantages and disadvantages of market participation to determine whether the net utility is positive. If the benefits, such as higher income, outweigh the costs, including transportation expenses and time investment, farmers are more likely to engage in market activities. Therefore, the framework posits that farmers participate in markets when the perceived benefits outweigh the costs, resulting in positive net utility.

Building on this understanding, the decision-making process modeled in the theoretical framework will be used to assess how farmers’ decisions to participate in markets translate into positive net gains and improvements to their overall well-being. By quantifying the net utility derived from participation, the findings will shed light onto whether market engagement through irrigation-based farming represents a pathway to improved quality of life and whether the necessary investments are justified.

3. Materials and Methods

3.1. Study Area

The study was conducted in the former homelands of Transkei and Ciskei, located in the Eastern Cape province. These regions are known for their fertile soils and abundant water resources, conditions that have historically supported agricultural activities and facilitated the development of small-scale irrigation schemes. The targeted population of the study comprised eight small-scale irrigation schemes, as illustrated in Figure 1, which have been the target of efforts to enhance agricultural infrastructure and improve productivity [8].

3.2. Sampling Procedure and Data Collection

The study utilized a multi-stage sampling approach, integrating both purposive and random sampling methods to address its complex research design. This design required several stages to identify the specific municipalities containing irrigation schemes, the communities within those municipalities, and the individual irrigation schemes within those communities.

In the initial stage, purposive sampling was employed to identify three prominent irrigation schemes, namely Qamata, Tyefu, and Zanyokwe, from a pool of eight schemes. This method was chosen because it allowed the researchers to intentionally select schemes based on specific criteria that would best represent the study’s objectives. The criteria for selection were primarily focused on the size and operational status of the schemes. These factors were crucial as they highlighted the schemes’ significance within their respective communities, especially in a region where many other irrigation schemes have been discontinued due to various challenges. The selected schemes were among the largest small-scale irrigation projects in the area and were still functioning, which provided an opportunity to examine the dynamics and impacts of operational irrigation schemes on local communities.

In the second stage of the study, attention was directed toward identifying farming communities or villages surrounding the selected irrigation schemes, from which non-scheme irrigators were randomly chosen. For example, the Qamata area consists of ten villages, namely Maya, Emthyintyini, Township Zwelitsha, Ngqanga, Shlahleni, Ntlakwefolo, Ntlonze, Bholokodlela, Ntshingeni, and Tyelera. Meanwhile, the Zanyokwe area includes six villages, and Tyefu comprises five. This stage was crucial in ensuring a representative selection of non-scheme irrigators who were not directly involved with the irrigation schemes but were still part of the broader agricultural landscape.

In the final stage, farmers’ households were the unit of analysis. We were assisted by the Farmer Organization and the Department of Agriculture in obtaining a list of potential participants and identifying both irrigation scheme members and non-scheme farmers who operated in close proximity to the irrigation schemes. All irrigation scheme members from the selected schemes were included in the survey, ensuring inclusivity and representation of farmers actively engaged in irrigation. The final sample comprised 210 scheme members and 40 randomly selected non-scheme members, resulting in a total of 250 smallholder farmers. It is important to note that the inclusion of both scheme and non-scheme farmers was not intended for direct comparison, but rather to ensure that the sample represented the broader farming community in the area. However, future studies could explore comparative analyses between these groups to gain further insights into the distinct impacts of market participation in irrigation versus non-irrigation settings.

3.3. Data Collection

Data were collected using structured questionnaires meticulously designed to capture detailed information on various aspects of smallholder farming. The questionnaire was designed to collect key variables related to farmers’ socioeconomic characteristics and market participation. Specifically, the market participation questions were to assess farmers’ level of engagement in market-oriented farming, including returns from the market. So they were asked questions including how much of their produce they sold, the price per unit received, their estimated farm income, and their total household income.

However, self-reported data are often subject to response bias, as farmers may underreport their income due to concerns about eligibility for support programs, while others may withhold full disclosure due to social expectations. Although farmers were asked to report honestly, we acknowledge that there may have been some level of underreporting or overreporting of income, which is common in surveys of this nature, particularly in contexts where farmers may not fully disclose their income. However, we made efforts to ensure respondents understood the confidentiality of the survey. The analysis proceeded with the assumption that most respondents provided truthful and consistent information, recognizing that while individual reporting may contain some discrepancies, broader trends across the sample could still offer meaningful insights into the relationship between market participation and rural household well-being. Despite these efforts, the inherent limitations of self-reported income remain a concern, and potential inaccuracies in the data cannot be entirely ruled out. To improve accuracy, income data were cross-checked with production levels.

To capture broader factors influencing market participation, the questionnaire also inquired about farmers’ distance to their market place, their farm size, and their access to financial support and extension services.

To ensure the questionnaire’s effectiveness, a rigorous pre-testing phase was undertaken to assess and validate its reliability, suitability, and appropriateness.

This pre-testing was conducted in the Ngqumashe and Khayamnandi rural areas within Raymond Mhlaba Municipality. These areas were chosen for their relevance to the study’s context, allowing for the refinement of the questionnaire’s clarity and relevance. Feedback collected during this phase was used to make necessary adjustments, enhancing the questionnaire’s ability to elicit accurate and meaningful responses. Although Ngqumashe and Khayamnandi were essential for the refinement process, they were not included in the final study sample.

The finalized questionnaire was subsequently used to collect data from the selected study sites, focusing on key areas such as farmer and farm characteristics and market participation, as detailed in Table 1.

3.4. Data Analysis

The gathered data were initially coded in Excel and then transferred to STATA Version 15 and SPSS Version 25 for further analysis. Both descriptive and econometric methods were utilized in the analysis. Descriptive statistics were applied to summarize the data, while econometric models were used to determine the factors influencing market participation. To evaluate the welfare impacts, propensity score matching (PSM) was employed to account for selection bias.

3.5. Analytical Technique

Building on the utility framework, income gains are central to welfare improvements, as increased income enables households to meet basic needs and improve their standard of living [47,48]. Neoclassical economist Alfred Marshall’s principle emphasizes the strong correlation between income and well-being, positing that income growth is a crucial driver of enhanced welfare [49]. As poorer households begin to participate more actively in income-generating activities, they are expected to experience gradual economic improvement. This process, driven by the assumption of diminishing marginal returns to scale, suggests that, over time, these households may begin to close the income gap with wealthier counterparts [50]. Expanding on this understanding, an estimation of change due to market participation can be measured as follows:

(4)${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{*}}={\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}-{\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}$

where ${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}$ is the income from market participation, and ${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}$ is the income without market participation. An improvement in welfare is indicated by ${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}$ > ${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}$. Thus, ${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{*}}$ captures the change in income due to market participation, with ${\mathrm{U}}_i\in \left\{0\right.,\left.1\right\}$ reflecting the utility gained from either engaging in markets or abstaining from them. The overall change in welfare, as measured by household income, reflects the impact of market participation, highlighting how engagement in markets can enhance the economic status and living conditions of smallholder farmers.

The impact of market participation on household income can be evaluated using two primary metrics: the average treatment effect (ATE) and the average treatment effect on the treated (ATT). The ATE represents the expected difference in outcomes of market participation [20], and it can be expressed as follows:

(5)$\mathrm{A}\mathrm{T}\mathrm{E}=\mathrm{E}\left({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}-{\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}\right)$

This expression illustrates the difference between the expected income resulting from market participation and the income a farmer would have earned prior to participating in the market. It represents the anticipated benefits of market participation.

The ATT represents the expected difference in outcomes for those who participate in the market, compared to what their outcomes would have been without participation. It addresses whether the outcomes of participation are advantageous for participants, demonstrating the actual benefits derived from market involvement. The ATT can be expressed as follows:

(6)$\mathrm{A}\mathrm{T}\mathrm{T}=\mathrm{E}\left({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}-{\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}\left|{\mathrm{D}}_{\mathrm{i}}=1\right.\right)$

where $\mathrm{D}=1,$ is a binary indicator of market participation. The expression reflects the actual outcomes experienced by participants due to their involvement in the market. However, we can only observe $\mathrm{E}({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}\left|\mathrm{D}=1)\right.$, the outcome associated with market participation, while the counterfactual outcomes, what the participants would have experienced had they not participated $\mathrm{E}({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}\left|\mathrm{D}=1)\right.$, remain unobserved. Consequently, a straightforward comparison of household income between participants and non-participants could produce biased estimates of the impact of market participation [51]. Thus, using Equation (6) to estimate the ATT may result in biased estimates due to this selection bias (b):

(7)$\mathrm{b}=\mathrm{E}\left({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}-{\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}\left|{\mathrm{D}}_{\mathrm{i}}=1\right.\right)-\mathrm{E}\left({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}-{\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}\left|{\mathrm{D}}_{\mathrm{i}}=0\right.\right)$

where D is a binary variable for market participation (D = 1 if a farmer participates and D = 0 otherwise), and b represents the selection bias. Without a suitable benchmark for comparison, attributing changes in outcomes directly to market participation becomes challenging, as only the outcomes for participants are observable. This limitation complicates the comparison of market participation impacts between participants and non-participants when information about non-participants is unavailable [20]. This situation reflects the issue of missing data [52]. Furthermore, it is improbable that all sampled farmers would have participated in the output markets, introducing potential heterogeneity in outcomes among farmers. Therefore, the critical question is the impact of participation on a randomly selected farmer versus one who participated, as the outcomes are likely to differ due to variations in farmer characteristics. Thus, a method to adjust for these differences is necessary.

Having established the necessity of addressing selection bias, we can employ propensity score matching (PSM) as our methodological approach. To estimate the ATT, matching methods are employed to ensure that the comparison between participants and non-participants is as fair and unbiased as possible. By matching farmers who participate in the market with those who do not, based on similar observable characteristics, we create a synthetic control group that serves as the counterfactual for the treated group. This approach allows us to estimate the true impact of market participation on household income, isolating the effects of participation from other confounding factors [53]. This technique has been successfully applied in various studies to evaluate the effects of market participation in the agricultural sector [40,51].

The first step in PSM is to estimate the propensity scores, which represent the probability of a farmer participating in the market based on a set of observable characteristics. This is achieved using a logistic regression model where the dependent variable is a binary indicator of market participation, and the independent variables include demographic factors, farm characteristics, and access to market information. The propensity score ${\mathrm{p}}_{\mathrm{i}}$ can be estimated as follows:

(8)${\mathrm{p}}_{\mathrm{i}}=\mathrm{P}\mathrm{r}\left({\mathrm{D}}_{\mathrm{i}}=1\left|{\mathrm{x}}_{\mathrm{i}}\right.\right)={\displaystyle \frac{\mathrm{e}\mathrm{x}\mathrm{p}\left({\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{x}}_{\mathrm{i}}\right)}{1+\mathrm{e}\mathrm{x}\mathrm{p}\left({\mathsf{\beta }}_0+{\mathsf{\beta }}_1{\mathrm{x}}_{\mathrm{i}}\right)}}$

where ${\mathrm{D}}_{\mathrm{i}}$ is a binary variable indicating market participation (1 if participated, 0 otherwise), ${\mathrm{x}}_{\mathrm{i}}$ represents a vector of observable characteristics (e.g., age, education, farm size), and ${\mathsf{\beta }}_0 \mathrm{a}\mathrm{n}\mathrm{d} {\mathsf{\beta }}_1$ are parameters to be estimated.

After calculating the propensity scores, participants (treated group) were matched with non-participants (control group) with similar values of the estimated propensity scores $\mathrm{p}\left(\mathrm{X}\right)$:

(9)$\mathrm{E}({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}-{\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}\left|\mathrm{p}\left(\mathrm{X}\right))=\mathrm{E}\left(\mathrm{Y}\right.\left|\mathrm{D}=1,\mathrm{p}\left(\mathrm{X}\right))-\mathrm{E}\left(\mathrm{Y}\right.\left|\mathrm{D}=0,\mathrm{p}\left(\mathrm{X}\right))\right.\right.\right.$

${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}$ and ${{\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}}_1$ represent household income levels without and with market participation, respectively, where $\mathrm{D}=1$ indicates market participation and $\mathrm{D}=0$ indicates non-participation. $\mathrm{p}\left(\mathrm{X}\right)$ denotes the propensity score, which is the probability of being in the market participant group given the covariates $\left(\mathrm{X}\right)$. The goal is to compare the outcomes of market participants with those of a control group that has similar characteristics, as indicated by their propensity scores, and to use this comparison to estimate the ATT. The mean effect of treatment is calculated as the average difference in outcomes between the treated group and the control group, conditional on the propensity scores. The ATT can be defined as follows:

(10)$\mathrm{A}\mathrm{T}\mathrm{T}=\mathrm{E}\left[\mathrm{E}({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}-{\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}\left|{\mathrm{D}}_{\mathrm{i}}=1,\mathrm{p}\left.{(\mathrm{X}}_{\mathrm{i}}\left)\right)\right]\right.\right.$

This approach ensures that both groups can receive treatment (market participation) and enables estimation of the treatment effects by comparing the observed outcome ${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}$ of the treated group with the outcome ${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{N}}$ of the untreated group to estimate what the outcome would have been without market participation. According to (Rosenbaum & Rubin, 1983) [53], the probability of receiving treatment, given the explanatory variables, can be captured by the propensity score $\mathrm{p}\left(\mathrm{X}\right)$;

(11)$\mathrm{p}(\mathrm{X})=\mathrm{P}\mathrm{r}\left(\mathrm{D}=1\left|\mathrm{X}\right.\right)={\displaystyle \frac{\mathrm{e}\mathrm{x}\mathrm{p}\left({\mathsf{\beta }}_0+{\mathsf{\beta }}_1\mathrm{X}\right)}{1+\mathrm{e}\mathrm{x}\mathrm{p}\left({\mathsf{\beta }}_0+{\mathsf{\beta }}_1\mathrm{X}\right)}}$

When matching is precise at the propensity score, the distribution of the covariates $\left(\mathrm{X}\right)$, which predict participation and outcomes, will be identical for both the participant and comparison groups [52]. This requires sufficient overlap in the covariates between the two groups, ensuring that there is a common probability of being either a market participant or a non-participant. This concept is referred to as the region of common support, as described in [54], and can be expressed as follows:

(12)$0<\mathrm{P}\left(\mathrm{D}=1\left|\mathrm{X}\right.\right)<1$

The overlap expression posits that farmers with identical values of $\left(\mathrm{X}\right)$ should have a similar probability of participating in markets. Farmers whose covariates fall outside the region of standard support are excluded from the analysis, as they do not provide a suitable match for the participants [54].

The typical support region represents the overlap in propensity scores between the treated (participants) and control (non-participants) groups, ensuring that comparable units are available in both groups for accurate estimation of treatment effects. The mean propensity score within this region reflects the average likelihood of participation for the observed sample, while the range defines the boundaries for making valid comparisons. Thus, the validity of the ATT estimation relies on significant overlap in the $\left(\mathrm{X}\right)$ values, indicating that the conditions for participation apply to both participants and non-participants.

Once the common support region is established and matching is performed, the ATT can be estimated using the matched sample. The ATT is calculated as the average outcome difference between participants and their matched non-participants within the standard support region [55]. The formula for ATT using PSM is given by

(13)$\mathrm{A}\mathrm{T}\mathrm{T}={\displaystyle \frac{1}{{\mathrm{N}}_{\mathrm{T}}}}{\sum }_{\mathrm{i}\mathsf{ϵ}\mathrm{T}}({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}-{\sum }_{\mathrm{j}\mathsf{ϵ}\mathrm{C}\left(\mathrm{i}\right)}{\mathsf{\omega }}_{\mathrm{i}\mathrm{j}}{\mathrm{Y}}_{\mathrm{j}}^{\mathrm{N}})$

where ${\mathrm{N}}_{\mathrm{T}}$ denotes the number of treated units (participants in the market). It is the total number of individuals or households in the treated group. Σ_iϵT indicates the summation over all individuals who are part of the treated group TTT. ${\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}$ is the observed outcome for participant $\mathrm{i}$, ${\sum }_{\mathrm{j}\mathsf{ϵ}\mathrm{C}\left(\mathrm{i}\right)}{\mathsf{\omega }}_{\mathrm{i}\mathrm{j}}{\mathrm{Y}}_{\mathrm{j}}^{\mathrm{N}}$ represents the weighted sum of the outcomes of control units (non-participants) matched to the treated unit $\mathrm{j},$ and ${\mathsf{\omega }}_{\mathrm{i}\mathrm{j}}$ is the weight assigned to control unit $\mathrm{j}$ when calculating the counterfactual outcome for treated unit $\mathrm{i}$. These weights are determined by the matching algorithm and ensure that the comparison is as unbiased as possible. ${\mathrm{Y}}_{\mathrm{j}}^{\mathrm{N}}$ is the outcome for matched non-participants $\mathrm{j}$. $\mathrm{C}\left(\mathrm{i}\right)$ denotes the set of control units (non-participants) that are matched to a particular treated unit $\mathrm{i}$. Essentially, it represents the group of non-participants who are most similar to the treated participant in terms of their propensity scores. The equation essentially calculates the average difference in outcomes between the treated units and their matched control units. It compares the actual outcomes of those who participated in the market $\left({\mathrm{Y}}_{\mathrm{i}}^{\mathrm{M}}\right)$ to the weighted average of outcomes for similar individuals who did not participate (${\sum }_{\mathrm{j}\mathsf{ϵ}\mathrm{C}\left(\mathrm{i}\right)}{\mathsf{\omega }}_{\mathrm{i}\mathrm{j}}{\mathrm{Y}}_{\mathrm{j}}^{\mathrm{N}}$). By averaging this difference across all treated units (${\displaystyle \frac{1}{{\mathrm{N}}_{\mathrm{T}}}}$Σ_iϵT), the ATT captures the impact of market participation on those who participated, accounting for any differences in observable characteristics between participants and non-participants. The weighting (${\mathsf{\omega }}_{\mathrm{i}\mathrm{j}}$) helps to create a balanced comparison, ensuring that each treated unit is compared to the most similar control units.

4. Results

4.1. Demographic and Socioeconomic Characteristics of Sampled Farmers (n = 250)

The socioeconomic and demographic characteristics of sampled farmers were analyzed using descriptive statistics, providing valuable insights into their livelihoods, economic conditions, and market engagement. These characteristics are crucial in understanding the challenges and opportunities faced by smallholder irrigators. The results are presented in Table 2.

4.1.1. Demographic Characteristics

The gender distribution among the farmers in the study reveals a significant imbalance, with 75.6% being male and only 24.4% female. This indicates a male-dominated farming sector, with women underrepresented in agricultural activities. Regarding marital status, 55.2% of the farmers are married, 31.2% are single, and 13.6% are widowed. The high percentage of married farmers suggests a stable family support system that contributes to farm labor and decision-making. However, widowed farmers, particularly women, may face increased responsibilities and labor shortages, necessitating targeted support systems to enhance resilience and productivity.

The average age of the farmers in the study is 60.23 years, reflecting an aging farming population. This demographic trend may affect labor availability and the adoption of modern agricultural technologies, as older farmers may be less inclined to embrace innovations. Phakathi et al. [40] support this observation, indicating that many irrigation schemes in the Eastern Cape province are predominantly managed by older farmers due to rural–urban migration of younger individuals.

On average, the farmers in the study have attained at least eight years of schooling, equivalent to a secondary education level in South Africa. This educational background provides essential literacy skills, enhancing their ability to access and interpret market-related information, as observed by Abdullah et al. [56]. Household sizes average five members, slightly larger than the national average of 3.2 [57]. While larger families may offer additional labor for farming, which can reduce operational costs, they also create resource constraints, requiring a balance between labor benefits and household needs [58].

4.1.2. Socioeconomic Characteristics

Farming is the primary occupation for 62.4% of the surveyed individuals, demonstrating a strong reliance on agriculture for livelihoods. Meanwhile, 16% are employed in non-agricultural sectors, and 21.6% engage in non-farm activities. These findings reinforce observations by Ngwako et al. [36], who noted that farming remains a full-time occupation in rural Africa, where many depend on it for survival. Income sources among farmers further highlight agriculture’s critical role in economic well-being. The majority (53.6%) rely on farm income, while 12.8% depend on wages from employment, 7.2% on remittances, and 26.4% on social grants. The significant reliance on grants underscores economic vulnerability, emphasizing the need for enhanced farm productivity and market access to reduce dependence on external support [59]. The per capita income among the sampled farmers is ZAR 4368.20, significantly above the domestic poverty line of ZAR 1058 per month. However, the standard deviation of ZAR 22.31 indicates substantial income variability, suggesting diverse economic conditions within the farming population. Additionally, farmers operate on an average farm size of 2.89 hectares, consistent with findings by van Averbeke et al. [33], who reported similar small farm sizes in South African irrigation schemes. These limited land holdings may constrain economies of scale, reducing the ability to produce surplus crops and engage profitably in markets.

4.1.3. Access to Resources and Market Participation

Access to extension services is available to 74.4% of farmers, highlighting efforts to enhance farming practices. However, 25.6% lack access, indicating gaps in service delivery and outreach. Ensuring equitable access to extension services is critical for sustainable agricultural development. Similarly, financial support is accessible to 61.6% of farmers, but 38.4% remain excluded from financial services. Limited financial access hinders investment in productivity-enhancing inputs and technologies, emphasizing the need for tailored financial products to support smallholders. Additionally, farmers in the study reside an average of 23.43 km from market centers, creating logistical challenges that increase transportation costs and hinder timely access to both markets and essential farming inputs. Despite these constraints, 55.42% of the sampled farmers participate in markets, demonstrating moderate engagement. However, sustaining this participation remains difficult due to financial limitations, land constraints, and inadequate access to training, all of which restrict their ability to scale up production and enhance competitiveness.

Overall, the demographic and socioeconomic characteristics of smallholder irrigators reveal both strengths and constraints within this area.

4.2. Factors Influencing Market Participation by Smallholder Irrigators

The logistic regression model identified farmer characteristics that influence market participation, allowing for the balancing of covariates between participants and non-participants. Table 3 illustrates the logit model results of the determinants of market participation by smallholder farmers in the study area. The results of the logit model of the propensity score matching analysis are presented by showing the coefficients and the marginal effects of the coefficients. The significant marginal effects indicate those variables that have significant effects on the farmer’s decision to participate in markets. The chi-squared value of 116.82 (p < 0.01) is highly significant, at 1%, indicating a good model fit. Pseudo R² is 0.58, suggesting that the model explains a substantial portion of the variation. The variance inflation factor (VIF) is 1.72, indicating no multicollinearity issues since it is less than 10 as per the rule of thumb.

The model revealed several key factors affecting market participation among farmers. Gender has a positive and statistically significant coefficient at the 5% level, indicating that more men participate in markets than women. Specifically, a 1% increase in the proportion of male farmers is associated with a 0.3% increase in the likelihood of market participation, meaning that if the number of male farmers increases by 1%, the probability of market participation rises by 0.3%. Marital status has a negative and statistically significant coefficient at the 5% level, suggesting that an increase in marital status is associated with a decrease in market participation, implying that married farmers are slightly less likely to participate in markets than their unmarried counterparts. Education also plays a significant role, with years spent in school having a positive coefficient that is statistically significant at the 1% level. Each additional year of schooling increases the likelihood of market participation by 0.4%, highlighting education’s beneficial impact on market involvement. Access to extension services shows a positive and statistically significant coefficient at the 1% level, indicating that a 1% increase in access to extension services correlates with a 0.2% increase in market participation. This suggests that improved access to these services slightly enhances the probability of market involvement. Finally, financial support is another positive factor, with a statistically significant coefficient at the 5% level, implying that a 1% increase in financial support leads to a 0.2% increase in market participation, suggesting that higher financial support modestly improves farmers’ likelihood of market participation.

4.3. Impact of Participation to Markets by Smallholder Irrigators

Ideally, the average household income of market participants should exceed that of non-participants. However, this outcome is not guaranteed, as other factors beyond market participation can affect household income. Concluding without accounting for these factors may result in biased findings. Therefore, it is crucial to examine the characteristics of farmers to ensure a balanced distribution of covariates between the participant (treated) and non-participant (untreated) groups. To achieve this balance, we establish a region of standard support to define the range within which propensity scores are distributed. Propensity scores falling outside this range, either below or above it, are excluded from the matching process. This method ensures that treatment effects are estimated only within the standard support region [52]. In this study, the numeric support region ranged from 0.28 to 0.99, with a mean value of 0.86. This ensures that participants outside this range are excluded to avoid bias in estimating the treatment effect. A mean propensity score of 0.86 suggests a relatively high average probability of market participation among the observed sample. In propensity score matching, a high mean propensity score may indicate that the covariates are well-balanced between the treatment and control groups after matching. This enhances the validity of causal inference by reducing confounding biases. The concept of the standard support region is crucial because it ensures that the matching process only compares units with similar likelihoods of receiving treatment, thereby reducing bias and improving the validity of causal inferences drawn from the data. The sufficient overlap between participants and non-participants ensures that propensity score matching can effectively balance covariates, allowing for valid estimation of treatment effects within the standard support region. Table 4 shows percentiles of propensity scores to verify the overlap between participants and non-participants.

To further assess the quality of the matching process, we visually examined the propensity scores to ensure adequate overlap between participants (treated) and non-participants (control group/untreated) [54]. Figure 2 illustrates the distribution of propensity scores before and after matching, with the upper section representing the participants and the lower section representing the non-participants. The analysis indicates substantial overlap in the distributions of the propensity scores, ensuring that the common support condition has been met. In other words, the mean propensity scores of the participants are comparable to those of the non-participants, indicating that the matching process effectively aligns the covariates of both groups.

4.4. Assessment of Matching Quality

Two matching algorithms, nearest-neighbour and kernel matching, were also used to evaluate the balance between participants and non-participants. This analysis was crucial in ensuring that the propensity score matching effectively aligned the characteristics of the two groups. Table 5 summarizes the results of the quality of the matching and shows the quality indicators: mean absolute standardized bias and pseudo-R-squared. Table 6 illustrates the effectiveness of the propensity score matching process in improving covariate balance between treated and control groups. The nearest-neighbour and kernel-matching algorithms substantially reduced the pseudo-R-squared, LR chi-squared, and mean standardized bias values, reflecting a significant improvement in the balance of covariates.

Initially, the data were analyzed without applying any matching techniques. For both the nearest-neighbour and kernel-matching algorithms, the pseudo-R-squared value was 0.278. This high value indicates that a considerable proportion of the variability in market participation was explained by the covariates before matching, suggesting substantial differences between the participants and non-participants. Furthermore, the LR chi-squared statistic was 116.266 with a p-value of 0.000, reflecting a strong association between the covariates and treatment assignment in the unmatched sample. This result underscores the presence of systematic imbalances between the groups, which could bias the treatment effect estimation.

After applying the matching algorithms, significant improvements in balance were observed. For nearest-neighbour matching, the pseudo-R-squared value dropped dramatically to 0.035. Similarly, for kernel matching, it decreased to 0.025. These low values after matching indicate a substantial reduction in the differences between the treated and control groups, demonstrating improved balance in covariate distributions. The results in Table 5 indicate that, before matching, the pseudo-R-squared was relatively high. However, after matching, it dropped significantly across all algorithms. This suggests that the matching process effectively balanced the distribution of the covariates, eliminating any systematic differences between the treated and control groups. This indicates that the matching procedure effectively controlled for confounding characteristics, resulting in comparable groups for this analysis. The significant reduction in pseudo-R-squared after matching suggests that the treatment and control groups now have similar covariate distributions. This implies that pre-existing group differences are less likely to bias the estimated effects.

The LR chi-squared statistic also saw a considerable decline post-matching. For nearest-neighbour matching, the LR chi-squared value decreased to 23.200 with a p-value of 0.634, indicating that the association between covariates and treatment assignment is no longer statistically significant, reflecting improved covariate balance. For kernel matching, the LR chi-squared value was reduced to 8.384 with a p-value of 0.048. Although still significant, this p-value is much closer to the threshold of 0.05, signifying a marked improvement in balance compared to the unmatched sample. The mean standardized bias also showed notable improvements. Before matching, the mean standardized bias was 23.4 for both matching algorithms, indicating substantial imbalances in covariate distributions. After matching, this bias decreased to 11.4 for nearest-neighbour matching and 8.000 for kernel matching. These reductions suggest that the matching algorithms successfully aligned the characteristics of the participants and non-participants, with kernel matching achieving a slightly better balance. This successful matching ensures that the treatment effect estimation regarding market participation is more robust and less biased, thereby enhancing the validity of the findings related to household welfare

4.5. Impact Estimation Results

The results presented in Table 6 provide a detailed examination of the average treatment effect on the treated (ATT) and the heterogeneity effects obtained from the nearest-neighbour and kernel-matching methods, ensuring that any observed differences in outcomes can be attributed to the market participation rather than pre-existing disparities. Table 6 reveals that market participation significantly enhances household welfare, with ATT values of 838.44 for nearest-neighbour matching and 828.44 for kernel matching. These results confirm that participants experience notable benefits compared to non-participants. The substantial differences suggest that participants experience significant improvements in their welfare due to market participation, highlighting the effectiveness of market participation.

Furthermore, the heterogeneity effects shown indicate a reduction in the variation of outcomes between participants and non-participants after matching. For nearest-neighbour matching, the effect for participants was 0.154, compared to 0.128 for non-participants, yielding a difference of 0.026. This relatively small variation indicates a consistent impact of market participation across the sample. Similarly, kernel matching provided comparable results, which closely align with the ATT found using nearest-neighbour matching, reaffirming the positive effect of market participation on household welfare. Thus, participants showed an average outcome of 2646.44. In contrast, non-participants had an average of 1818, leading to a difference of 828.44, although the heterogeneity effects for kernel matching reveal slightly more variability than nearest-neighbour matching. Participants exhibited a heterogeneity effect of 0.177, while non-participants scored 0.145, resulting in a difference of 0.034. This suggests a higher level of variation in the impact of market participation across individuals in the kernel-matching sample.

5. Discussion

Based on the findings of this study, farmers in the study area earn above the poverty line, allowing them to meet basic needs and potentially have some discretionary income. However, the burden of supporting larger families can strain household resources significantly. Larger family sizes often result in higher dependency ratios, where fewer income earners must support more dependents, such as children and the elderly. This can reduce the production of marketable surpluses, as more of the farm’s output is used for immediate household consumption rather than being sold. Consistent with previous findings, larger family sizes increase competition for limited resources, emphasizing food consumption and household expenditures more than labor for production [38]. While the study findings revealed moderate market participation among the farmers, with an average participation rate of 55%, this indicates that while a portion of the farmers are engaged in market activities, a significant number are not fully integrated into the market. This moderate level of participation aligns with the broader economic challenges these farmers face. One of the primary issues is the limited size of their farms, which average only a few hectares. This limited scale hinders their capacity to realize economies of scale [19]. For these smallholders, the lack of scale means higher per-unit costs of production, which reduces their competitiveness in the market. The small size of the farms also curtails the production of marketable surpluses, limiting the quantity of produce available for sale after meeting household consumption needs. This situation hampers profitability, making it challenging for farmers to generate sufficient income to reinvest in their farms. As a result, they often struggle to invest in advanced technologies or inputs, such as high-quality seeds, fertilizers, and irrigation equipment, which could enhance productivity and increase their income potential.

In addition to the constraints of farm size, the geographical remoteness of smallholder irrigators poses further challenges to economic advancement. Many of these farmers are situated over 20 km from significant market centers, leading to higher transportation costs for selling produce and obtaining essential agricultural inputs [60]. This remoteness limits market access, negatively impacting profitability by reducing the farmers’ ability to compete in more lucrative markets and respond quickly to market demand changes. Furthermore, the increased costs and logistical challenges associated with transporting goods over long distances make it difficult for farmers to sell their produce and acquire necessary supplies, further hindering their economic progress.

The logit model revealed several socio-economic factors that significantly influence market participation among smallholder farmers. Several critical factors influence market participation among farmers, including gender, marital status, education, access to extension services, and financial support. Among these, gender disparity is particularly pronounced. The findings reveal that men typically have better market access than women [51]. This imbalance suggests that gender roles and societal norms might influence who participates in farming, rendering it a predominantly male occupation. This observation is consistent with literature documenting women’s unique challenges compared to their male counterparts. From a young age, women are socialized into caregiving roles, influencing their approach to agriculture [61]. However, in conventional agricultural environments where masculine traits are valued, women may struggle to assert their identity as farmers [61]. This cultural dynamic, including household responsibilities, can limit their opportunities and recognition in the field [36]. This can be confirmed by the fact that there is a significant income disparity between male and female farmers, with men earning approximately ZAR 26.78 more per cropping season than their female counterparts [10].

Education, access to extension services, and financial support are critical in enhancing market participation. According to Phakathi et al. [40], education equips farmers with the knowledge and skills to interpret agricultural information, adopt innovative practices, and improve productivity. Educated farmers are better positioned to understand and leverage market information, negotiate prices, and implement new technologies to increase yields and profitability.

Access to extension services further supports market participation by providing farmers with essential information on market opportunities, crop management techniques, and sustainable agricultural practices. According to the literature, farmers who receive these services are more likely to engage in markets, as they provide valuable guidance on optimizing production and establishing connections with buyers [40,62]. This implies that extension services are a vital bridge between farmers and markets, facilitating knowledge transfer and improving farmers’ decision-making capabilities.

Financial support is equally vital in promoting market participation. Access to financial resources, such as credit and loans, allows farmers to seize market opportunities by purchasing necessary inputs and hiring transportation, ensuring they can access markets promptly. Haile et al. [63] found that financial support bolsters farmers’ capacity, leading to increased production and a more robust market supply. With financial backing, farmers can purchase inputs promptly and hire labour to boost productivity, enhancing their market competitiveness.

Conversely, marital status presents a contrasting influence on market participation. Married farmers are generally less likely to participate in markets than their unmarried counterparts. This tendency is attributed to married individuals prioritizing home consumption and family needs over market sales. The focus on meeting domestic requirements limits their market activities to selling surplus produce [62]. This behaviour may result in lower levels of market engagement and reduced income from market sales.

The average treatment effect on the treated (ATT) confirms significant income improvements for households participating in the market, with participants earning approximately 45% more than non-participants across all matching methods. The heterogeneity effects further suggest that market participants experience 20–23% higher gains, suggesting that market participants benefit more than the non-participants. This finding aligns with theoretical expectations that output market participation enhances revenue generation opportunities, ultimately boosting smallholder farmers’ economic well-being. It further corroborates previous research in South Africa, demonstrating that market participation significantly enhances smallholder farmers’ livelihoods. Hlatshwayo et al. [38] found that market participation improved household dietary diversity, while Cele and Mudhara [64] showed that market participation positively influenced household food security. These findings support the argument that increased income from market participation not only improves economic welfare but also contributes to better food security outcomes. These insights emphasize the importance of strengthening market access and integrating smallholder farmers into formal value chains to enhance their resilience and overall welfare.

6. Conclusions

Smallholders are well acknowledged for their contribution to food security and poverty reduction. Although they are mostly self-sufficient, there is a growing emphasis on encouraging them to utilize their limited resources effectively to increase production beyond home consumption, engage in markets, and generate the income necessary for their livelihoods, which depend not only on food supply but also on cash to purchase goods and services they cannot produce on their own. This transition is achievable through the establishment of well-functioning markets. However, farmers face numerous challenges, such as financial constraints, small farm sizes, inadequate training, poor infrastructure, and limited access to irrigation systems, which further hinder their capacity to engage effectively in markets. As a result, progress in market engagement remains sluggish. Nonetheless, for those who do participate, the findings show that market participation has led to considerable improvements in their income.

That said, while the findings of this study highlight considerable gains in the well-being of market participants, they also suggest that the full potential of market participation remains untapped. However, there are important limitations to consider. The study analysis is based on survey data collected from the selected irrigation schemes, which may not fully represent the experiences of all smallholder irrigators in the area. Moreover, although rigorous matching techniques were employed to estimate the average treatment effect on the treated and minimize selection bias, other unobserved factors such as risk preferences, informal networks, and individual decision-making processes may still affect how farmers engage with markets and the benefits they derive. Additionally, a one-time data collection point limits the study’s ability to assess long-term effects, it does not allow for the exploration of the ongoing impact of market participation on farmers’ income and well-being, nor does it capture seasonal or exogenous shocks. Future research should then adopt a longitudinal approach which will allow for tracking farmers over an extended period to determine the sustainability of income gains, understand the long-term effects of market participation on welfare, and identify the factors influencing farmers’ ability to remain in markets over time. Thus a longitudinal study could potentially bring to light informative observations regarding whether such enhancements over time are enduring or if they fade in later periods. Such studies could also consider incorporating seasonal variations and assessing gender-differentiated outcomes. In addition, integrating qualitative methods would provide deeper insights into the socio-cultural factors influencing market engagement.

For policymakers, these findings highlight the urgent need for interventions, with efforts focusing on improving rural infrastructure to facilitate smoother farm-to-market operations, enhancing access to credit and financial services, and increasing the visibility and effectiveness of extension services and agricultural training. Additionally, implementing quality control programs and certification schemes will help farmers meet market requirements and standards. Lastly, fostering partnerships with public and private enterprises will be crucial in creating a more supportive environment for smallholder farmers.

Author Contributions

Conceptualization, M.E.L. and A.O.; Methodology, M.E.L., L.M. and A.O.; software, M.E.L., N.M. and A.O.; validation, M.E.L., N.M. and A.O.; formal analysis, M.E.L., N.M. and A.O.; investigation, M.E.L.; resources, A.O.; data curation, M.E.L., N.M., L.M. and A.O.; writing—original draft preparation, writing—review and editing, M.E.L., N.M. and L.M.; visualization; supervision, A.O. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

The WRC developed the original protocol for the project and awarded it to the University through the Research Office and the Department of Agricultural Economics and Extension. As a result, no further ethical clearance was required by the University. A Technical Reference Group chaired by WRC met twice a year and supervised the implementation of the project over its lifespan.

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study. Participants were informed about their right to ask questions relating to the research. Confidentiality and privacy were ensured throughout.

Data Availability Statement

The datasets used or analyzed during the current study are available from the corresponding author upon reasonable request.

Acknowledgments

The authors are very grateful to the Water Research Commission for funding this study. We are also very grateful to the Qamata, Tyefu and Zanyokwe Irrigation Schemes, communities and Farm Organization administration for their cooperation during data collection and for providing supplementary secondary data. Last but not least, we thank the smallholder farmers in the project areas for their time and willingness to provide data. The enumerators are correctly recognized for their sacrifice and for self-administering the questionnaires.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1 Irrigation schemes in the Eastern Cape province. Source: [5].

Figure 2 Propensity score distribution for treatment and comparison.

Table 1

Description of variables and unit measure.

Variable	Description	Expected Outcome
Market Participation	Participation in output markets: participant = 1, non-participant = 0
Gender	Gender of household head: 1 if male, otherwise 0	+/−
Age	Age of household head in a completed number of years	+/−
Marital status	Marital status of household head: 1 if married, otherwise 0	+/−
Household size	Number of people living in a household	+/−
Education	Highest level of education in years	+
Main occupation	Primary occupation: 1 if farmer, otherwise 0	+
Household income	The sum of household income in rand	+/−
Income sources	All sources of income	+/−
Cultivated farm size	Area cultivated in acres	+
Distance to market	Nearest market in kilometres	+/−
Extension services	If received extension service advisory: 1 if yes, otherwise 0	+
Financial support	Received government financial support: 1 if yes, otherwise 0	+

Table 2

Demographic characteristics of sampled farmers.

Variable Description	Frequency	Percentage	Mean	Standard Deviation
Gender			-	-
Male	189	75.60
Female	61	24.40
Age	-	-	60.23	12.7
Marital status			-	-
Married	138	55.20
Single	78	31.20
Windowed	34	13.60
Years spent in school	-	-	8.26	3.65
Family size	-	-	5.01	2.44
Primary occupation			-	-
Farming	156	62.40
Employed	40	16
Non-farm	54	21.60
Source of income			-	-
Farm	134	53.60
Wage	32	12.80
Remittance	18	7.20
Grant	66	26.40
Income per capita	-	-	4368.20	22.31
Farm size	-	-	2.89	1.53
Access to extension services			-	-
Yes	186	74.40
No	64	25.60
Financial support			-	-
Yes	154	61.60
No	96	38.40
Distance to the market	-	-	23.43	7.97
Market participation	-	-	55.42	34.3

Table 3

Logit model for market participation.

Variable	Coeff (Std Err)	Marginal Effects (dy/dx)
Gender: male	0.438 ** (0.158)	0.323 ** (0.026)
Marital status	−0.286 ** (0.134)	−0.145 ** (0.015)
Years spent in school	0.516 *** (0.028)	0.356 *** (0.072)
Access to extension services	1.302 *** (0.056)	0.230 *** (0.034)
Financial support	0.033 ** (0.019)	0.155 ** (0.023)
_cons	− 3.703 *** (0.697)
Log-likelihood: −269.546	LR chi2: 116.82 ***	Pseudo R2: 0.58
Variance inflation factor: 1.72	Number of observations: 250

Table 4

Summary distribution of estimated propensity scores (n = 250).

Percentile		Propensity Scores	Statistic
Smallest	1%	0.28	Mean	0.86
	5%	0.41	Std dev	0.19
	10%	0.55	Variance	0.04
	25%	0.80	Skewness	−1.51
Median	50%	0.95	Kurtosis	4.17
Largest	75%	0.97
	90%	0.98
	95%	0.99
	99%	0.99

Table 5

Quality of the match.

Algorithms	Sample	Pseudo-R-Squared	LR Chi2	Wald Chi-Squared (p-Value)	MeanStandardized Bias
Nearest neighbour	Unmatched	0.278	116.266	0.000	23.400
	Matched	0.035	23.200	0.634	11.400
Kernel	Unmatched	0.278	116.266	0.000	23.400
	Matched	0.025	8.384	0.048	8.000

Table 6

Treatment effects of small-scale farmers’ participation.

Matching Method	Metrics
	Participants	Non-Participants	Difference (ATT)
Nearest-neighbour matching	2662.44	1824	838.44
Heterogeneity effects	0.154	0.128	0.026
Kernel matching	2646.44	1818	828.44
Heterogeneity effects	0.177	0.145	0.034

References

1. Gafa, D.; Chachu, D. Explaining Food Insecurity in Sub-Saharan Africa: The Role of Governance and Institutions; Development Economics 531 African Economic Research Consortium: Nairobi, Kenya, 2023; [DOI: https://dx.doi.org/10.5167/uzh-239362]

2. Henchion, M.; Hayes, M.; Mullen, A.M.; Fenelon, M.; Tiwari, B. Future protein supply and demand: Strategies and factors influencing a sustainable equilibrium. Foods; 2017; 6, 53. [DOI: https://dx.doi.org/10.3390/foods6070053] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28726744]

3. Miladinov, G. Impacts of population growth and economic development on food security in low-income and middle-income countries. Front. Hum. Dyn.; 2023; 5, 1121662. [DOI: https://dx.doi.org/10.3389/fhumd.2023.1121662]

4. Obi, A.; Gidi, L.; Balogun, T.; Mujuru, N.; Fakunle, F.; Avuletey, R.; Onomu, R.; Kibirige, D.; Mcata, B.; Lesala, M. Appropriate Entrepreneurial Development Paths in the Transition from Homestead Food Gardening to Smallholder Irrigation Crop Farming in the Eastern Cape Province of South Africa: Report to the Water Research Commission; Water Research Commission: Pretoria, South Africa, 2016.

5. Lesala, M.E. Market Participation and Welfare of Smallholder Farmers in the Eastern Cape Province, South Africa. Ph.D. Thesis; University of Fort Hare: Alice, South Africa, 2021.

6. Liu, Z.; Chen, Q.; Xie, H. Influence of the farmer’s livelihood assets on livelihood strategies in the western mountainous area, China. Sustainability; 2018; 10, 875. [DOI: https://dx.doi.org/10.3390/su10030875]

7. Nhemachena, C.; Matchaya, G.; Nhlengethwa, S.; Nhemachena, C.R. Exploring ways to increase public investments in agricultural water management and irrigation for improved agricultural productivity in Southern Africa. Water SA; 2018; 44, pp. 474-481. [DOI: https://dx.doi.org/10.4314/wsa.v44i3.15]

8. Fanadzo, M.; Ncube, B. Challenges and opportunities for revitalising smallholder irrigation schemes in South Africa. Water SA; 2018; 44, pp. 436-447. [DOI: https://dx.doi.org/10.4314/wsa.v44i3.11]

9. Thabane, V.N.; Agholor, I.A.; Ludidi, N.N.; Morepje, M.T.; Mgwenya, L.I.; Msweli, N.S.; Sithole, M.Z. Irrigation Water and Security in South African Smallholder Farming: Assessing Strategies for Revitalization. World; 2025; 6, 32. [DOI: https://dx.doi.org/10.3390/world6010032]

10. Mkuna, E.; Wale, E. Gender differentials among small scale irrigation farmers’ income: Empirical evidence from cabbage farmers in KwaZulu-Natal, South Africa. Front. Sustain. Food Syst.; 2023; 7, 1155756. [DOI: https://dx.doi.org/10.3389/fsufs.2023.1155756]

11. Kafle, K.; Omotilewa, O.; Leh, M.; Schmitter, P. Who is Likely to Benefit from Public and Private Sector Investments in Farmer-led Irrigation Development? Evidence from Ethiopia. J. Dev. Stud.; 2022; 58, pp. 55-75. [DOI: https://dx.doi.org/10.1080/00220388.2021.1939866]

12. Abbasi, A.A.A.; Saha, S.; Begum, I.A.; Rola-Rubzen, M.F.; McKenzie, A.M.; Alam, M.J. Does rural transformation affect rural income inequality? Insights from cross-district panel data analysis in Bangladesh. Heliyon; 2024; 10, e30562. [DOI: https://dx.doi.org/10.1016/j.heliyon.2024.e30562]

13. Tantoh, H.B.; McKay, T.J.M. Utilizing the water-land-food security nexus to review the underperformance of smallholder farmers in the Eastern Cape, South Africa. Front. Sustain. Food Syst.; 2023; 7, 1143630. [DOI: https://dx.doi.org/10.3389/fsufs.2023.1143630]

14. Mujuru, N.M.; Obi, A. Effects of cultivated area on smallholder farm profits and food security in rural communities of the Eastern Cape province of South Africa. Sustainability; 2020; 12, 3272. [DOI: https://dx.doi.org/10.3390/su12083272]

15. Myeni, L.; Moeletsi, M.; Thavhana, M.; Randela, M.; Mokoena, L. Barriers affecting sustainable agricultural productivity of smallholder farmers in the eastern free state of South Africa. Sustainability; 2019; 11, 3003. [DOI: https://dx.doi.org/10.3390/su11113003]

16. Huka, H.A.; Kilimab, F.T.; Mchopa, A.D. Impact of high-value market participation on income, saving and assets among smallholder vegetable farmers in Arumeru District, Tanzania. Cogent Food Agric.; 2024; 10, 2398767. [DOI: https://dx.doi.org/10.1080/23311932.2024.2398767]

17. Otekunrin, O.A.; Momoh, S.; Ayinde, I.A. Smallholder Farmers’ Market Participation: Concepts and Methodological Approach from Sub-Saharan Africa. Curr. Agric. Res. J.; 2019; 7, pp. 139-157. [DOI: https://dx.doi.org/10.12944/CARJ.7.2.02]

18. Dzanku, F.M.; Asante, K.T.; Hodey, L.S. Heterogeneous market participation channels and household welfare. Oxf. Dev. Stud.; 2024; 52, pp. 74-93. [DOI: https://dx.doi.org/10.1080/13600818.2023.2289196]

19. Maesela, L.M.; Senyolo, G.M.; Belete, A. Market Participation Decision of Emerging Farmers Under the Land Restitution Programme in Limpopo Province, South Africa. S. Afr. J. Agric. Ext.; 2023; 51, pp. 87-103. [DOI: https://dx.doi.org/10.17159/2413-3221/2023/v51n4a14524]

20. Mmbando, F.E.; Wale, E.Z.; Baiyegunhi, L.J.S. The welfare impacts of market channel choice by smallholder farmers in Tanzania. Dev. Pract.; 2017; 27, pp. 981-993. [DOI: https://dx.doi.org/10.1080/09614524.2017.1353066]

21. Nandi, R.; Nedumaran, S.; Ravula, P. The interplay between food market access and farm household dietary diversity in low and middle income countries: A systematic review of literature. Glob. Food Secur.; 2021; 28, 100484. [DOI: https://dx.doi.org/10.1016/j.gfs.2020.100484]

22. Ntakyo, P.R.; van den Berg, M. Effect of market production on rural household food consumption: Evidence from Uganda. Food Secur.; 2019; 11, pp. 1051-1070. [DOI: https://dx.doi.org/10.1007/s12571-019-00959-2]

23. Rutsaert, P.; Chamberlin, J.; Oluoch, K.O.; Kitoto, V.O.; Donovan, J. The geography of agricultural input markets in rural Tanzania. Food Secur.; 2021; 13, pp. 1379-1391. [DOI: https://dx.doi.org/10.1007/s12571-021-01181-9]

24. Dey, S.; Singh, P.K. Role of market participation on smallholder vegetable farmers’ wellbeing: Evidence from matching approach in Eastern India. Agribusiness; 2023; 39, pp. 1217-1237. [DOI: https://dx.doi.org/10.1002/agr.21813]

25. Hichaambwa, M.; Chamberlin, J.; Sitko, N.J. Determinants and Welfare Effects of Smallholder Participation in Horticultural Markets in Zambia. Afr. J. Agric. Resour. Econ.; 2015; 10, pp. 279-296. [DOI: https://dx.doi.org/10.22004/ag.econ.229810]

26. Geffersa, A.G.; Tabe-Ojong, M.P. Smallholder commercialisation and rural household welfare: Panel data evidence from Ethiopia. Eur. Rev. Agric. Econ.; 2024; 51, pp. 54-90. [DOI: https://dx.doi.org/10.1093/erae/jbad040]

27. Setoboli, T.; Tshuma, N.; Sibanda, E. Viability of Goat Production and Market Participation a Comparative Analysis of Smallholder Farmers in Gwanda District, Zimbabwe. Int. J. Res. Innov. Soc. Sci.; 2025; VIII, pp. 244-251. [DOI: https://dx.doi.org/10.47772/IJRISS.2024.815EC0019]

28. Viana, C.M.; Freire, D.; Abrantes, P.; Rocha, J.; Pereira, P. Agricultural land systems importance for supporting food security and sustainable development goals: A systematic review. Sci. Total Environ.; 2022; 806, 150718. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2021.150718]

29. Villar, P.F.; Kozakiewicz, T.; Bachina, V.; Young, S.; Shisler, S. Protocol: The effects of agricultural output market access interventions on agricultural, socio-economic and food and nutrition security outcomes in low- and middle-income countries: A systematic review. Campbell Syst. Rev.; 2023; 19, e1348. [DOI: https://dx.doi.org/10.1002/cl2.1348]

30. Ma, A.; Rahut, D.B.; Sonobe, T.; Gong, B. Linking farmers to markets: Barriers, solutions, and policy options. Econ. Anal. Policy; 2024; 82, pp. 1102-1112. [DOI: https://dx.doi.org/10.1016/j.eap.2024.05.005]

31. Weltin, M.; Zasada, I.; Franke, C.; Piorr, A.; Raggi, M.; Viaggi, D. Analysing behavioural differences of farm households: An example of income diversification strategies based on European farm survey data. Land Use Policy; 2017; 62, pp. 172-184. [DOI: https://dx.doi.org/10.1016/j.landusepol.2016.11.041]

32. Ngqangweni, S.; Mmbengwa, V.; Myeki, L.; Sotsha, K.; Khoza, T. Measuring and tracking smallholder market access in South Africa. Presented at the NAMC Seminar Pretoria, South Africa, 4 April 2016.

33. van Averbeke, W.; Denison, J.; Mnkeni, P.N.S. Smallholder irrigation schemes in South Africa: A review of knowledge generated by the Water Research Commission. South African Water Research Commission. Water SA; 2011; 37, pp. 797-808. [DOI: https://dx.doi.org/10.4314/wsa.v37i5.17]

34. Jiba, P.; Obi, A.; Mdoda, L.; Mzuyanda, C. The Impact of Smallholder Irrigation Scheme on Household Welfare in Farm-Managed Irrigation Scheme Communities in the Eastern Cape Province, South Africa. S. Afr. J. Agric. Ext.; 2024; 52, pp. 48-72. [DOI: https://dx.doi.org/10.17159/2413-3221/2024/v52n1a13953]

35. Asfaw, S.; Lipper, L.; Dalton, T.J.; Audi, P. Market participation, on-farm crop diversity and household welfare: Micro-evidence from Kenya. Environ. Dev. Econ.; 2012; 17, pp. 579-601. [DOI: https://dx.doi.org/10.1017/S1355770X12000277]

36. Ngwako, G.; Mathenge, M.; Gido, E.; Kgosikoma, K. Effect of market participation on household welfare among smallholder goat farmers in Botswana. J. Agribus. Rural. Dev.; 2021; 60, pp. 151-160. [DOI: https://dx.doi.org/10.17306/J.JARD.2021.01362]

37. Zondi, N.T.B.; Ngidi, M.S.C.; Ojo, T.O.; Hlatshwayo, S.I. Impact of Market Participation of Indigenous Crops on Household Food Security of Smallholder Farmers of South Africa. Sustainability; 2022; 14, 15194. [DOI: https://dx.doi.org/10.3390/su142215194] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/39659967]

38. Hlatshwayo, S.I.; Ojo, T.O.; Ngidi, M.S.C. Effect of market participation on the food and nutrition security status of the rural smallholder farmers: The case of Limpopo and Mpumalanga provinces, South Africa. Front. Sustain. Food Syst.; 2023; 7, 1097465. [DOI: https://dx.doi.org/10.3389/fsufs.2023.1097465]

39. Khapayi, M.; Celliers, P.R. Factors Limiting and Preventing Emerging Farmers to Progress to Commercial Agricultural Farming in the King Williams Town Area of the Eastern Cape Province, South Africa. S. Afr. J. Agric. Ext.; 2016; 44, pp. 25-41. [DOI: https://dx.doi.org/10.17159/2413-3221/2016/v44n1a374]

40. Phakathi, S.; Sinyolo, S.; Fraser, G.; Marire, J. Heterogeneous welfare effects of farmer groups in smallholder irrigation schemes in South Africa. Afr. J. Agric. Resour. Econ.; 2021; 16, pp. 27-45. [DOI: https://dx.doi.org/10.53936/afjare.2021.16(2).03]

41. Ola, O.; Menapace, L. Smallholders’ perceptions and preferences for market attributes promoting sustained participation in modern agricultural value chains. Food Policy; 2020; 97, 101962. [DOI: https://dx.doi.org/10.1016/j.foodpol.2020.101962]

42. Christian, M.; Obi, A.; Agbugba, I.K. Adoption of Irrigation Technology to Combat Household Food Insecurity in the Resource-constrained Farming System of the Eastern Cape Province, South Africa. S. Afr. J. Agric. Ext.; 2019; 47, pp. 94-104. [DOI: https://dx.doi.org/10.17159/2413-3221/2019/v47n2a506]

43. Pawlak, K.; Kołodziejczak, M. The Role of Agriculture in Ensuring Food Security in Developing Countries: Considerations in the Context of the Problem of Sustainable Food Production. Sustainability; 2020; 12, 5488. [DOI: https://dx.doi.org/10.3390/su12135488]

44. Stringer, L.C.; Fraser, E.D.G.; Harris, D.; Lyon, C.; Pereira, L.; Ward, C.F.M.; Simelton, E. Adaptation and development pathways for different types of farmers. Environ. Sci. Pol.; 2020; 104, pp. 174-189. [DOI: https://dx.doi.org/10.1016/j.envsci.2019.10.007]

45. Kahan, D. Market-Oriented Farming: An Overview; FAO: Quebec City, QC, Canada, 2013.

46. Singh, I.; Squire, L.; Strauss, J. A Survey of Agricultural Household Models: Recent Findings and Policy Implications. World Bank Econ. Rev.; 1986; 1, pp. 149-179. [DOI: https://dx.doi.org/10.1093/wber/1.1.149]

47. Collins, S.B. Welfare and Well-Being; the Treatment of Social Relationships in the Administration of Income Assistance in Ontario. Crit. Soc. Work. Interdiscip. J. Dedic. Soc. Justice; 2004; 5, pp. 1-11.

48. Sen, A.K.; Williams, B.A.O. Utilitarianism and Beyond; Cambridge University Press: Cambridge, UK, 1996.

49. Mahadea, D. On the Economics of Happiness: The Influence of Income and Non-income Factors on Happiness. S. Afr. J. Econ. Manag. Sci.; 2013; 16, pp. 39-51. [DOI: https://dx.doi.org/10.4102/sajems.v16i1.204]

50. Alawode, O.O. Rural Land Market, Livelihood and Welfare among Households in Oyo State, Southwest Nigeria. Bus. Manag. Stud.; 2021; 7, 45. [DOI: https://dx.doi.org/10.11114/bms.v7i1.5137]

51. Mossie, M.; Gerezgiher, A.; Ayalew, Z.; Nigussie, Z. Welfare effects of small-scale farmers’ participation in apple and mango value chains in Ethiopia. Agrekon; 2021; 60, pp. 192-208. [DOI: https://dx.doi.org/10.1080/03031853.2021.1926298]

52. Bryson, A.; Dorsett, R.; Purdon, S. The Use of Propensity Score Matching in the Evaluation of Active Labour Market Policies; Department Work and Pensions: London, UK, 2002.

53. Rosenbaum, P.R.; Rubin, D.B. The central role of the propensity score in observational studies for causal effects. Biometrika; 1983; 70, pp. 41-55. [DOI: https://dx.doi.org/10.1093/biomet/70.1.41]

54. Caliendo, M.; Kopeinig, S. Some Practical Guidance for the Implementation of Propensity Score Matching. J. Econ. Surv.; 2008; 22, pp. 31-72. [DOI: https://dx.doi.org/10.1111/j.1467-6419.2007.00527.x]

55. Tiri, G.D.; Mlay, G.I.; Roselyne, A. Productivity impact of growth enhancement support scheme on maize farm households in Kano State, Nigeria. Afr. J. Agric. Res.; 2020; 15, pp. 90-101. [DOI: https://dx.doi.org/10.5897/AJAR2019.14075]

56. Abdullah,; Rabbi, F.; Ahamad, R.; Ali, S.; Chandio, A.A.; Ahmad, W.; Ilyas, A.; Din, I.U. Determinants of commercialization and its impact on the welfare of smallholder rice farmers by using Heckman’s two-stage approach. J. Saudi Soc. Agric. Sci.; 2019; 18, pp. 224-233. [DOI: https://dx.doi.org/10.1016/j.jssas.2017.06.001]

57. General Household Survey. Statistical Release P0318. 2023; Available online: https://www.statssa.gov.za/publications/P0318/P03182023.pdf (accessed on 30 January 2025).

58. Sigigaba, M.; Mdoda, L.; Mditshwa, A. Adoption drivers of improved open-pollinated (Opvs) maize varieties by smallholder farmers in the Eastern Cape Province of South Africa. Sustainability; 2021; 13, 13644. [DOI: https://dx.doi.org/10.3390/su132413644]

59. Mhembwe, S.; Chiunya, N.; Dube, E. The contribution of small-scale rural irrigation schemes towards food security of smallholder farmers in Zimbabwe. Jàmbá J. Disaster Risk Stud.; 2019; 11, a674. [DOI: https://dx.doi.org/10.4102/jamba.v11i1.674] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31745405]

60. Mdoda, L.; Mdiya, L. Factors affecting the using information and communication technologies (ICTs) by livestock farmers in the Eastern Cape province. Cogent Soc. Sci.; 2022; 8, 2026017. [DOI: https://dx.doi.org/10.1080/23311886.2022.2026017]

61. Tourtelier, C.; Gorman, M.; Tracy, S. Influence of gender on the development of sustainable agriculture in France. J. Rural Stud.; 2023; 101, 103068. [DOI: https://dx.doi.org/10.1016/j.jrurstud.2023.103068]

62. Abokyi, E. The impact of agricultural marketing program on farm investment: Evidence from Ghana. Cogent Econ. Financ.; 2022; 10, 2111781. [DOI: https://dx.doi.org/10.1080/23322039.2022.2111781]

63. Haile, K.; Gebre, E.; Workye, A. Determinants of market participation among smallholder farmers in Southwest Ethiopia: Double-hurdle model approach. Agric. Food Sec.; 2022; 11, pp. 2-13. [DOI: https://dx.doi.org/10.1186/s40066-022-00358-5]

64. Cele, T.; Mudhara, M. Impact of Market Participation on Household Food Security among Smallholder Irrigators in KwaZulu-Natal, South Africa. Agriculture; 2022; 12, 261. [DOI: https://dx.doi.org/10.3390/agriculture12020261]