https://orcid.org/0000-0002-7173-5042
Figures
Abstract
Soybean contract farming is increasingly encouraged for agricultural sustainability and transformation, and improved smallholder welfare in Ethiopia. Despite the contemporary significance and economic advantages of soybeans for farmers, current yields consistently fall short of their potential, and local sources struggle to meet growing demand. This research investigates the institutional and systemic constraints of the soybean contract in Metema, West Armachio, and Quara districts. Previous studies usually use a reductionist approach, investigating isolated factors and thus failing to explore the multifaceted interdependent nature of these farming systems. To address existing knowledge and methodological gaps, this research employs an inclusive, multilevel analytical apprpach within a transaction cost theoretical framework, utilizing a cross-sectional design. A multi-stage sampling procedure is employed to collect data from 369 smallholder farmers, comprising 170 contract farming participants and 199 non-participants. The bivariate Tobit model findings confirm that poor seed quality, lack of enforceability, low contract price, and limited market power are critical systemic constraints of soybean contract farming in the study areas. To address the problems, the responsible bodies should resolve oligopsony market issues by strengthening farmers’ multi-purpose cooperatives and employing anti-collusion measures to ensure market fairness. Policymakers must establish a clear legal framework that mandates strict contract enforceability, competitive pricing, and the use of collateral requirements to build farmer trust. Finally, securing good-quality inputs from certified suppliers through a fair contract model is crucial to promote sustainable growth and structural transformation within Ethiopian agriculture.
Author summary
Despite its economic value, the productivity of soybeans in Ethiopia lags significantly behind global standards, imposing a heavy fiscal burden due to expensive edible vegetable oil imports. To resolve this, the government has encouraged contract farming (CF) as a strategic instrument to link smallholder soybean producers with reliable markets. However, a productivity paradox continues where high market demand occurs, yet local soybean yields and producers’ satisfaction remain low. This research employed a mixed-methods approach, combining qualitative interviews with a bivariate Tobit econometric approach, to explore the systemic constraints influencing smallholder farmers in northwestern Ethiopia. The study found that soybean contract farming success is eroded by critical systemic problems: lack of legal contract enforcement, poor seed quality, and the occurrence of oligopsony, in which illegal buyer groups manipulate smallholders through biased pricing. The findings revealed that while specific training can increase soybean yields, the true sustainability and development depend on the supply of certified seeds, competitive pricing, and transparency. To empower local farmers and ensure food security, this study recommends immediate policy interventions to formalize the seed supply system, strengthen legal protections for agreements, and regulate market brokers.
Citation: Abreham G, Ayalew Z, Mussa EC, Muchie M, Adugna M (2026) Institutional and systemic barriers: Towards a sustainable transformation of soybean contract farming in Northwestern Ethiopia. PLOS Sustain Transform 5(4): e0000238. https://doi.org/10.1371/journal.pstr.0000238
Editor: Prashant Sharma, ICAR-IISWC: ICAR Indian Institute of Soil and Water Conservation, INDIA
Received: October 27, 2025; Accepted: March 29, 2026; Published: April 20, 2026
Copyright: © 2026 Abreham et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors acknowledge the journal’s mandatory data sharing policy; the data is publicly archived in figshare at https://doi.org/10.6084/m9.figshare.31169806.
Funding: This work was supported by the University of Gondar (Grant number 6417 to GA). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. GA received a salary from the University of Gondar during the period of this research.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
1.1. Background of the study
Soybean is an expanding and valuable pulse crop with multiple food and economic advantages for small-scale farmers [1,2], a potential that is noticeably demonstrated in Asian production schemes. Its economic importance is observed by its increasing export status in Ethiopia, which is predicted to reach $97.4 million by 2032 [3]. The significance of this crop is evidenced by cultivation trends, with total northwestern Ethiopian production reaching 1,472,452 quintals across 55,317 ha in 2018/19 [4]. Coverage in this region ismarked by notable expansion [5], even though smallholder farmer yields remain far below the potential production levels [2].
The government of Ethiopia uses the soybean contract farming approach as a key tool to ensure steady soybean production and marketing. Despite Ethiopia’s vast potential for soybean production and government initiatives to promote it, several marketing constraints (such as poor quality of soybean grains, low price, absence of grade, and standards), and production constraints such as low soil fertility, restricted fertilizer use, and inadequate crop management activities result in low soybean yield in western Ethiopia [6–8]. This lack of progress necessitates a comprehensive analysis of the contract farming system surrounding soybean production and marketing in the selected districts.
There is a need to understand the farmers’ perceived constraints to achieve sustainable soybean production and market linkage [9]. In this study, systemic/institutional constraints refer to the problems that are structural [10], economic, and institutional barriers [9] that prevent soybean contract farming (SCF) from achieving its theoretical efficiency. Analyzing the institutional constraints of soybean contract farming in Ethiopia demands an approach that moves beyond isolated, individual factors [11]. This analysis aims to provide insights into the interconnectedness of actors, institutions, and processes within the soybean production system. It addresses the limitations of previous assessments by considering how buyers, farmers, input suppliers, research institutions, policymakers, and market players affect contract farming effectiveness and sustainability. Since the interactions between these diverse actors are fundamentally defined by hierarchical dependencies and unequal bargaining power, the institutional constraints within the system are best understood as a function of power dynamics. This can be achieved by using household as a unit of analysis because of the direct benefits and harms associated with the soybean contract farming.
Therefore, employing the transaction cost approach offers a robust and adaptable framework to explore farmer-perceived major constraints within the soybean contract farming system in the selected districts of northwestern Ethiopia, to develop effective and sustainable solutions.
1.2. Statement of the problem
Soybean is a multipurpose, nutritionally rich crop [12,13] as its dry seed contains the highest protein and oil content among grain legumes, with a good balance of the essential amino acids and oil. Soybean is a pulse crop with multiple food and economic advantages for small-scale farmers. It is used as food for home consumption, as a raw material for local factories, feed for animal dairy or fattening farms, and poultry [14].
Despite its strategic importance, Ethiopia’s soybean yields continue to lag well behind global benchmarks. While leading international producers like Brazil achieved yields of 3.4 t/ha in 2024 (P2), Ethiopia’s productivity stagnates at 2.2 t/ha [15]. This domestic soybean production shortage results in a 95% dependence on imports, obliging Ethiopia to spend around $48 million monthly on vegetable oil to meet a supply-demand gap. The national soybean production currently covers less than 5% for a country with a population growth rate of 2.5% and soybean consumption growth rate of 1.2% [16].
This poses a great challenge to policymakers, researchers, development experts, and extension workers to meet the high marketing demand for soybeans. In this effort, CF is proclaimed as “Agricultural Production Contract Proclamation No.1289/2023” in accordance with Article 5(1 & 61,6) of the current Constitution of the Federal Democratic Republic of Ethiopia [17]. It is considered as a key instrument in the “Ten-year perspective development plan (2021-2030)” [18], and is reinforced by the Ethiopian national agricultural finance implementation roadmap (NAFIR), which underlines strengthening market linkages for smallholder farmers [19], As a result, the total hectares of land under soybean production in the country over the past 10 years have increased tenfold; while the total volume of production within the same period has increased increased twenty-onefold [20]. This rapid growth was intended to foster a steady and effective CF situation for key participants, including firms, farmers, and other institutional stakeholders.
However, there is a paradox in the soybean contract system in the selected districts of northwestern Ethiopia. On one side, there is a high domestic and international need for soybean seeds. Globally, Ethiopia’s share was only 0.00323% [21], and also, domestic edible oil producers import their inputs from Asia. Previous studies tried to explore the constraints related to the adoption of integrated farming systems [22], crop-dairy mixed farming systems [23], crop production farming systems, and breeding [24] using descriptive methods. This study applies a mixed -methods approach.Quantitative method of analysis helps us move beyond simply knowing the direction of relationshipto measuring the extent of these negative relationships and understanding the relative importance of various constraints. It is perplexing that while national demand for edible oil is met by imports, local soybean producers based on contract farming face market demand shortage, low yield [15], very low farmgate price, and low productivity compared to the world soybean productivity [12]. These gaps are due to various institutional and systemic constraints which hinder the effectiveness and sustainability of this system.
Exploring the systemic constraints enhances the smallholder farmer’s marketing and production sustainability [25] in the soybean contract farming system.
Therefore, this study is designed to investigate various constraints perceived by farmers in the soybean contract farming system using econometrics methods combined with qualitative narrations. Through a case study of West Armachiho, Quara, and Metema districts in northwestern Ethiopia, this study seeks to contribute critical insights that are insufficiently addressed in the previous literature.
1.3. Research questions
The study addresses the following questions: What are the main systemic barriers in northwestern Ethiopia that prevent contract farmers from achieving higher soybean yields? Which factors critically influence both farmers satisfaction and soybean yield? And how do unfair market practices, such as hidden buyer groups and low-quality seeds, reduce farmer satisfaction in contract farming?
1.4. Objective of the study
The primary objective of this study is to explore the institutional and systemic constraints affecting the success of soybean contract farming in northwestern Ethiopia, measured through farmer satisfaction and crop yields, to provide evidence-based policy interventions.
Specifically, the study aimed to identify the critical institutional and systemic factors that influence the success of soybean CF; to examine the intensity of these primary constraints on both soybean yields and farmer satisfaction; and to produce actionable policy recommendations to mitigate soybean CF constraints in northwestern Ethiopia.
2. Theoretical framework
To analyze the institutional and systemic constraints of soybean contract farming success in northwestern Ethiopia, an important theoretical framework is crucial. For this case, there are different theoretical frameworks, including Agency theory, Resource dependency theory, stakeholder theory, and Transaction cost theory to anchor the study's argument. Among these, agency theory tries to reduce asymmetric market information between the contracting firm and the producers by focusing on contract farming contents and payment-related agreements, but it assumes that contracting can address the agency problem [26].
Specifically, it pays low attention to broader institutional elements like farmers’ cooperative membership and other variables. Stakeholder theory is another approach which focuses on business relationships [27], mostly on fairness and enforceability, particularly where informal institutions are strong, but this theory is weak in addressing economic efficiencies and contractual designs. Importantly, even though the theory of power dynamics is crucial for exploring structural power imbalances and inherent biases [28], it is less effective than transaction cost theory for evaluating the detailed unmanaged economic costs and institutional frictions that drive poor satisfaction and low productivity (Fig 1). Transaction cost economics theory, with a particular focus on minimizing transaction costs [29], is vital to investigate the comprehensive systemic constraints of contract farming by considering variables like contract price, payment terms, and contracting terms.
The transaction cost theoretical model maps these contract farming problems into a causal flow influencing two primary outcomes: the soybean yield, and the farmer satisfaction index (satis_index). The interactive path links independent variables like transparency, enforceability, and contract price to farmer satisfaction. In this concept, transparency refers to the contract firm’s fairness in implementing the predefined arrangements; when this is missing, it raises the transaction costs for the soybean contract farming participant farmer. The technical path links market access, payment terms, distance from the contracting office, and quality of seed to yield. For instance, distance from the office signifies a logistical constraint that raises costs of monitoring and decreases support availability.
Eventually, the connection between marketing failures, like asymmetric information, and human behaviors, like opportunism, catalyzes systemic failure. Problems like low quality of soybean seed and limited inputs are seen as adverse selection, where contract firms offer sub-standard agricultural inputs that contract farmers may not easily verify. At the same time, weak contract enforcement leads to moral hazard, for example, farmer side-selling to alleviate low prices. By exploring these specific sources of transaction costs, the model shows that high economic costs and unmanaged institutional frictions are the main problems to contract farming success, directly ensuing in farmer dissatisfaction, and low soybean productivity within the contract farming sub- sector.
2.1. Conceptual framework
The conceptual framework explains the complex interaction of various factors affecting soybean contract farming success, indicated by farmer satisfaction and soybean yield. The key interaction is occurrs between constraints and transaction cost theory (facilitating contracts between two parties [30], with contract farming outcomes (Fig 2). The external environment (policies, and regulatory frameworks) governs the security of the exchange, while transaction characteristics such as uncertainty (yield/price volatility), asset specificity (seed quality), and frequency drive contract complexity. This conceptual framework emphasizes institutional capacity.
This model focuses on finding the critical systemic problems to SCF success through the lens of transaction cost theory. These are categorized into enforcement costs (contract enforceability), information costs (transparency), and monitoring costs (distance to office). By treating these constraints as measurable cost drivers, the model provides a structured pathway to study how systemic barriers influence the two main soybean contract farming success indicators: farmer satisfaction and soybean yield. Using an econometric (Bivariate Tobit) model, this theoretical framework explores how these problems simultaneously influence satisfaction and yield. By empirically recognizing the most prohibitive costs of transaction, the framework provides a robust analytical lens for exploring institutional efficiency [31]. It serves as a problem-solving tool for actionable policy suggestions aimed at decreasing systemic barriers and promoting sustainable growth in the soybean contract farming environment.
3. Research methodology
3.1. Descriptions of the study area
The study was conducted in selected districts of northwestern Ethiopia, particularly in the West Gondar zone of Amhara Regional State, Ethiopia. The zone is constructed in four districts [32], with the administration center of the zone being Genda Wuha, which is 870 km away from Addis Ababa [33]. The area is a dry lowland area with high agricultural potential, and, according to the data of 2024 captured from the Ethiopian Statistical Services at the Gondar branch office, the average size of land for a smallholder farm household is 5 hectares. It is an oilseed producer, with crops such as sesame, soybean, and mung bean being widely produced. According to the Amhara National Regional Crop Directorate (2023/24), the West Gondar zone was the top soybean producer area in the region, and approximately 76,464 hectares of land were covered by soybean contract farming with expected yields of 2,120,255 quintals during that crop season. The participation of 325 investors indicates a strong level of confidence in its potential for achievement in the districts of the zone. “Quara”, “Metema”, and “West Armachiho” districts were the highest potential areas for soybean production in the area. The geographical location of the study area is between longitudinal 3600’0”E to 3700’0”E and latitudinal 1200’0”N to 1300’0”N (Fig 3).
Source: Map produced by the author using ArcGIS 10.8. Administrative boundary shapefiles were obtained from the Ethiopian Statistical Service (ESS) and UN OCHA (2021) via the Humanitarian Data Exchange (HDX) [https://data.humdata.org/dataset/cod-ab-eth]. This base layer is licensed under CC BY-IGO [https://creativecommons.org/licenses/by/3.0/igo/].
3.2. Sample size and sampling technique
The study used a comprehensive multistage sampling approach to select the required sample respondents in northwestern Ethiopia. First, three main soybean-producing districts, Quara, West Armachiho, and Metema, were purposively selected since they are highly suitable for soybean production and actively host contract farming, which is crucial for systemic analyses. Second, from the high soybean-producing kebeles in these districts, two kebeles with strong potential in both soybean production and its contract farming activities were purposively selected from each district. To ensure statistical power, the required sample size (n = 369) was initially determined using the Cochran (1977) method for a finite population [34]:
. Assuming the identified population size of N = 6,134 households, a 5% level of precision (e = 0.05), a 95% confidence level (z = 1.96), this sampling formula was used with a conventional population proportion (p = 0.5) to ensure maximum variance. In the third stage, the sampling frame within the selected kebeles was stratified into two groups: soybean contract farming participants (treatment group) and non-participants (control group). Lastly, the required sample size, n = 369, was proportionally allocated to the two strata, resulting in 170 soybean contract farming participants and 199 non-participants, and individual sample households were then selected from each stratum using a simple random sampling method based on proportional to size to ensure an unbiased representation.
3.3. Data requirements and data sources
This study used both primary and secondary data. Primary data were collected through personal and face-to-face interviews using semi-structured and pre-tested interview schedules, focus group discussions, and key informant interviews that were conducted by trained enumerators under the supervision of the lead author.
The Interview schedule was constructed in the form of both open-ended and closed-ended questions. Its content was designed to include all the necessary information to answer the research objectives and identify systematic problems in the study areas. The data was obtained from respondents (n = 369) in Quara, Metema, and West Armachiho districts. The focus group discussion was held at two locations: for the Quara and Metema districts at Metema town, and for the West Armachiho district at Mdregenet. Various stakeholders were involved in a single FGD to enable triangulation of viewpoints in real-time. CF includes multifaceted information asymmetries among contractors, farmers, and government representatives. By bringing these actors together, the study was able to observe interactions, resolve conflicting claims regarding contract terms, and capture a “360-degree” view of systemic constraints that would be missed in homogeneous groups. This heterogeneous composition is supported by [35], who suggests that diverse FGDs can reveal the social complexities of power and status. Homogeneity is challenged by some researchers since unfamiliar participants can give honest and spontaneous views and can overcome pre-existing relationships and patterns of leadership in the group [36]. The institutional, socio-economic, and other problems and remedies of contract farming, and other related issues, were the discussion points. The FGDs were held before the main data collection, but after the pilot survey, to get the necessary data from all data gathering tools. The FGDs were strategically timed between the pilot and main survey to refine the survey instrument. These insights provided an understanding required to improve the interview guide, by incorporating local vocabulary and addressing critical systemic problems overlooked in the earlier literature. To complement this, secondary data were collected from books, peer-reviewed articles, and green literature or official reports to offer appropriate justification. By benchmarking our observed results against national agricultural trends in Ethiopia, the study confirmed that the identified systemic barriers showed wider institutional patterns rather than isolated incidences.
Regarding the KII procedure, the study conducted KIIs with purposefully selected individuals who possessed unique insights into the soybean production and contract farming (CF) dynamics. These included officials from the Amhara regional state Bureau of Agriculture (crop directorate office), district agricultural officers, kebele administrators, CF firm representatives, and local community leaders. A semi-structured interview guide was used to explore institutional arrangements, market challenges, and the historical context of CF in the region. Each interview lasted approximately 45–60 minutes and was recorded with participant consent.
Secondary data were also collected from reports of the Amhara National Regional State crop directorate office, Department of Agriculture in West Gondar zone office, sampled district agricultural offices, CSA (now called central statistical services - CSS), and others. These data were essential in framing the background of the study, providing a detailed description of the study areas (including agro-ecology and demographics), and triangulating the quantitative results to enhance the depth and validity of the analysis.
Ethical Approval: For this overall data collection, the required ethical approval was obtained from the University of Gondar Internal Review Board, led by the Vice President for Research & Technology Transfer Office, with reference number Ref. VP/RTT/05/664/2024.
Consent to participate: This research followed the Declaration of Helsinki and was permitted by the University of Gondar Internal Review Board (Ref. VP/RTT/05/664/2024). Informed oral consent was obtained from all participants (survey, KII, and FGD) due to varying literacy levels, with the IRB’s consent. Participants were informed of the aim of the study and their full right to withdraw at any time without penalty. To ensure confidentiality, no personal information was stored, and data were stored on password-protected systems. Participants received no financial rewards or faced no psychological risks for their involvement.
3.4. Methods of data analysis
Both descriptive and econometric are used in this study. Descriptive analysis is applied to outline the general demographic characteristics and the associations of different variables with the dependent variables.
The econometric method of analysis is used as the main analytical approach, precisely bivariate Tobit model. The detailed model is presented below.
3.4.1. Model specification to analyze systemic constraints of CF: A Bivariate Tobit model.
Standard linear models like Ordinary Least Squares (OLS) regression are not suitable here because soybean contract farming participation is often zero for non-participants, the data might be limited by a lower bound, and OLS would yield inconsistent and biased estimations. While Logit and Probit models effectively analyze the binary participation indicator, they ignore the “extent” indicator entirely, masking the true impact of systemic constraints on volume or land allocation. Similarly, the Multinomial Logit (MNL) identifies choices among discrete categories but fails to handle the continuous nature of participation intensity. While Logit and Probit models are appropriate for binary participation decisions (participant or not participant) [37], they fail to consider the extent of the participation, in this case, the intensity of constraints on outcome, which is critical for understanding the true effect of systemic constraints. In the same way, a Multinomial Logit model can analyze choices among various discrete decisions [38], but it is unable to address the continuous nature of participation extent when limited.
More complex approaches, like the Double-Hurdle model and the Heckman two-stage models, allow these two-soybean contract farming success indicators to be analyzed as separate methods. However, the Heckman two-stage model needs robust assumptions about valid exclusion restrictions and error term distribution, which are mostly problematic to recognize and assess [39]. And also, the Double-Hurdle model (a generalization of the Tobit) needs a two-step method for participation decision, permitting several constraints to influence the participation status and the extent of participation, and it is more flexible than the univariate Tobit if the two participation decisions are really separated [40]. However, the two equations in this study are interlinked, and that is why the bivariate Tobit [41] regression model becomes preferable. Consequently, this model was designated for its ability to simultaneously investigate both the extent and the participation equation. By correctly considering the censored nature of the participation data, it delivers reliable and efficient estimations of the systemic problems that drive success in soybean contract farming. A model is used when you have dependent variables, which are both censored. This model covers the univariate Tobit model by allowing generalizes the univariate Tobit framework by allowing correlation between the error terms of both comparations. This feature makes it particularly effective for analyzing censored outcomes that are determined at the same time. In this description, let
be the latent (unobserved) variables for the two dependent variables of CF success, while the observed dependent variables are characterized by
. We assume there is a linear relationship for the latent variables:
Where:
, represents systemic constraint variables (they can be the same or different).
indicates the parameters of the satisfaction and yield equation, respectively.
shows the residual errors for the satisfaction and yield equations, respectively.
The basic assumption for this model is that the above residual terms follow a bivariate normal distribution with a covariance matrix and mean zero:
where,
The observed outcomes, farmers’ satisfaction, and yield are symbolized by are associated with the latent variables
by the censoring mechanism.
The estimation of the bivariate Tobit model characteristically contains maximum likelihood estimation, which accounts for the correlation and the censoring between the residual terms. The likelihood function involved probabilities and density roles resulting from the bivariate normal distribution, in view of different scenarios based on whether each dependent variable is observed or censored.
3.5. Definition, measurement, and expected correlation of variables
3.5.1. The dependent variables.
To analyze the farmer-based constraints of the soybean contract farming system, this study contextualized the soybean contract farming system in terms of yield for farmers [42] and farmer satisfaction [43]. Using total soybean yield as a dependent variable is strongly recommended because it is considered the critical objective and direct indicator of the SCF system’s success in its ability to realize gains in soybean production and marketing. The composition of SCF participant satisfaction and total soybean yield gives a theoretically grounded, methodologically sound, and scientifically recommended measure of soybean contract farming success due to its simultaneous measurement of soybean production efficiency and sustainability. While soybean yield examines the farm effectiveness or performance and transfer of technology, farmer satisfaction captures welfare outcomes for SCF participants and the sustainability of contractual relationships. The main economic logic of these contract systems is to address contract farming systemic constraints (such as lack of credit, poor seed quality, lack of contract enforce ability, and etc.) that get down agricultural output.
Hence, if the SCF constraints are genuinely being fixed by the contract system, this should be evident in a quantifiable rise in total soybean yield, which focuses on the technical impact which answer the question: “Did the SCF successfully enhance farm productivity?” This is critical both for the farmer (increasing income potential) and the firm (securing supply). This dual methodological approach provides a comprehensive view of SCF success, determining whether soybean farming contract is both productive today and sustainable tomorrow.
The proper econometric model for this study is the Bivariate Tobit model that allows for various censoring types in those two equations. This analytical model is used to significantly support the use of total soybean yield as the key productivity measurement while considering its linkage with farmer welfare (Satisfaction).
Farmer Satisfaction Index (satis_index): This is the dependent variable of contract farming success. Principal Component Analysis (PCA) was used to construct a composite measure, termed the farmers’ Satisfaction Index, from observed satisfaction variables such as: satisfaction with training and support (sat_trainsup), profit (sat_profit), fairness of grading or weighing process used by the firm to assess the soybean production quality and quantity (sat_fairnesgrad), trust of the firm (sat_trust), reliability of input supply or the timely and full delivery of agricultural inputs to the contract farmers (sat_inputsprl), sustainability (sat_sustianablty). With data from sample respondents, the initial descriptive analysis results displayed variable means ranging from 2.57 to 3.35 on a 1-to-5 Likert scale. Following the Kaiser-Guttman principle, which retains components with an eigenvalue greater than 1 (lambda > 1), the first three components (C1, C2, and C3) were selected for the index construction. These three variables or factors cumulatively explain 69.61% of the total variance (above the rule of thumb, which is > 65%) present in the original six variables, supporting that the constructed satisfaction index is a good measure.
The final farmer satisfaction index is calculated as a weighted sum of the original standardized constraints or variables, with the weights resulting from the three retained components based on their scoring coefficients [44].
Contract Farming Yield: This is a continuous variable, and the total soybean yield within contract farming during the last 12 months.
3.5.2. Independent variables.
Contract Price: It is the pre-defined or pre-agreed price of the soybean produce paid by the firm (contractor) from the soybean producer farmer. Farmers are asked to express their level of overall satisfaction with how the pre-agreed price negatively influences their contract farming success, with answers on a 5-point Likert scale from “Strongly Disagree [1]” to “Strongly Agree [5].” The variable was hypothesized to have a negative association with farmer satisfaction in the study area. Farmers might perceive the price as low, and this leads to have with lower level of satisfaction. A price higher than the spot market price is a suitable motivation for farmers to observe contract farming activities [45], and it may increase their satisfaction with contract farming.
Contract Enforceability: Contract enforceability is defined as the degree to which the pre-defined agreements are clearly understood, legally binding, and effectively upheld by both contracting parties, together with the existence of fair dispute resolution mechanisms [46].
Quality of Soybean seed: It refers to the characteristics of the provided or promised soybean seed quality by the firm, including features like its demand on the market and disease resistance. It was expected to have a negative influence on the contract farming participants’ satisfaction if there are concerns about poor-quality seed supply to farmers, leading to increased vulnerability and lower yields. It is supported by [47].
Market Access: It indicates the farmer’s ability to consistently sell their soybean produce to the contracting firm, suggesting a guaranteed purchaser and often decreased market uncertainty, like price risks. The variable was expected to have a negative influence on the contract farming participants’ satisfaction if there was doubt about selling to a contract farming firm.
Transparency: Refers to the fairness of a firm regarding the achievement or completion of the predefined or promised agreements during the contract scheme. The variable was expected to have a negative influence on the contract farming participants’ satisfaction if there is poor transparency on firm, leading to increased farmers’ interaction bond with the firm, thereby reducing their trust on what they have on the firm.
Payment Terms: It is defined as receiving payments for the delivered soybean produce. If soybean contract farming participant farmers agree that payment terms are a positive constraint, represented as timely reimbursement or payment of money, their level of productivity is likely to increase. “Disputes related to price and methods of payment were the primary reasons for not continuing with contracts.” [2].
Transparency: This variable is descriptive and defines the clarity of contract information between the contract participant farmers and the contracting firm, particularly concerning grading and pricing. If contract participant farmers believe a poor level of transparency is a problem, it shows a problem in trust, which is likely to decrease yield via reducing profit, consistent with [48].
Quality of Soybean: The variable refers to the characteristics of the soybean seeds recommended by the contracting firm or contract farming facilitators, including features like purity, germination rate, and disease resistance. The variable was expected to have a negative influence on the contract farming yield if there was doubt in supplying poor-quality seed to farmers, leading to increased vulnerability and lower yields, eventually reducing productivity. It is supported by [47].
Market Access: It indicates the farmer’s ability to consistently sell their soybean produce to the contracting firm, suggesting a guaranteed purchaser and often decreased market uncertainty, like price risks. The variable was expected to have a negative influence on the contract farming yield if there was doubt about selling to a contract farming firm.
Distance from Office: This is the geographical distance from home to the agriculture office and a continuous variable measured by minutes (on foot). The variable is expected to affect soybean yield negatively because it could reduce accessibility to contract firm support.
4. Results and discussion
4.1. Influence of market access constraints on yield
The descriptive analysis (Fig 4) displays that farmers have heterogeneous and strong perceptions regarding their involvement. While about 40.2% of respondents reported positive outcomes, 18.2% were very satisfied and 22.0% were satisfied with their soybean harvests.
However, contract farming success is offset by a high proportion of low-yielding producers. Specifically, 44.3% of smallholder farmers reported low yields, including 24.3% who were “very dissatisfied,” probably due to failing to achieve contract objectives. Only 15.5% of producers reported “Neutral” or average results. The soybean contractual farming exhibits a bimodal yield distribution. This means the contemporary contractual model demonstrates high potential but suffers from a substantial failure rate amongst approximately a quarter of CF participants. Addressing this underperforming output is important to resolve systemic gaps and maximize the model’s overall productivity.
4.2. Bivariate Tobit analysis of contract farming success
The result and discussion present comprehensive findings from the quantitative analysis conducted using a bivariate Tobit model to instantaneously examine the systemic constraints of contract farming success (measured by soybean yield and farmer satisfaction) affecting the soybean contractual farming system. This bivariate Tobit model accounts for the censored nature of both dependent equations, providing more accurate and robust results of the underlying associations. These quantifiable results are also supported by qualitative insights gathered from Focus Group Discussions (FGDs) and Key Informant Interviews (KIIs), which provide a better understanding of complexities involved.
Before conducting the econometrics analysis, there is a need to perform econometrics diagnostic tests. In this case, the study undertakes the multicollinearity test using Variance Inflation Factor (VIF), Tolerance, and Conditional index analytical results. The Variance Inflation Factor (VIF) and its reciprocal, Tolerance, confirms that the model is not affected by problematic levels of multicollinearity. The common threshold for concern is mostly a VIF value beyond 10. In our case, the maximum value of VIF observed was less than 5 (VIF = 4.19) for the constraint “transparency”, with the remaining constraints indicating VIF values significantly less than this critical value. Distance from office (VIF = 1.54) showed the lowest multicollinearity, representing negligible collinearity issue for this constraint. Similarly, the Tolerance values for the whole constraints were well above the recommended cutoff value of 0.10. Additionally, the entire set of variable mean VIF values was 3.40. Since the mean and the individual VIFs are well below the accepted threshold of 10, the results confirm that the explanatory variables (in this case, constraints) are sufficiently independent for constant regression estimation.
In the second diagnostic test, the Condition Index supports the results from the above VIF evaluations. The Condition index, which is the square root of the ratio of the largest to the smallest eigenvalue, serves as an overall measure of data ill-conditioning. A value above 30 is usually considered an indicator of severe multicollinearity problems. The estimated condition value for this research model was approximately 9.50. Since this number is significantly lower than the moderate to strong multicollinearity benchmark of 10, it suggests that the correlation matrix of this model is stable and that the accuracy of parameter estimations has not been compromised by collinear problems among the variables (Table 1)
Generally, the constancy and accuracy of the last regression model’s parameter estimates and standard errors are validated by the Condition Index, VIF, and Tolerance in multicollinearity assessment, which jointly display that multicollinearity is not an issue.
4.3. Overall model fit and inter-equation correlation
The quantitative method of analysis, bivariate Tobit model analysis, confirms a strong overall fit to the given data, as shown by the value of the Wald chi-squared statistic of 82.31, p-value = 0.000 with df 5. This highly statistically significant result supports that both the outcome indicators of contract farming success (farmer’s yield and farmer’s satisfaction) are well explained by the included explanatory variables. The log-likelihood value of -680.265 also displays the valuation of the bivariate Tobit model’s overall explanatory power (Table 2).
In the bivariate Tobit model, the correlation coefficients (ρ) between the residual error terms of the satisfaction and yield equations, which represents the extent to which unobserved variables affecting one dependent variable are correlated with unobserved variables affecting the other dependent variable, is estimated. In this study, the evaluated “atan_rho” (the rho value) is approximately 0.20, with a p-value of 0.08, indicates the correlation is statistically significant (at 10% significance level)
This significant statistical value confirms that, after evaluating the existing factors in satisfaction and yield equations, the unobserved variables that affect farmers’ yield have a strong correlation with those that affect farmers’ satisfaction. This indicates that while the observed systemic constraints capture the key influencers for each outcome, there’s a robust and statistically significant correlation in the unobserved influences that conjointly influence both satisfaction and yield. Consequently, to appropriately account for this shared unobserved heterogeneity and the censored nature of the dependent variables, a Bivariate Tobit model is more suitable.
4.3.1 Factors influencing farmer satisfaction (equation 1).
The initial equation of the bivariate Tobit model explicitly examines the systemic constraints of farmer-level satisfaction, one of the indicators of contract farming success. While the remaining variables are constant, the coefficients shown indicate the effect of a one-unit variation in the constraint variables on the latent (uncensored) level of farmers’ satisfaction (Table 2).
Foremost among the major constraints that are significant drivers of farmers’ satisfaction is market access, which is a highly statistically significant and negative marginal effect (β= -0.264, p < 0.001). Holding other factors constant, a unit improvement in agricultural market access conditions is associated with a 26 percentage point reduction in the probability of farmer satisfaction. This finding strongly indicates that reducing transaction costs related to transporting, selling, or delivering the final soybean product is the single most effective way to stabilize the contract.
When pre-existing problems of agricultural market access are internalized by the contract system rather than alleviated, the estimated benefit of farmer satisfaction decreases. This indicates that CF, rather than resolving the institutional and systemic market problem or constraint, is vulnerable to the very inflexibilities it was intended to overcome. This condition produces an economic discouragement rather than compliance, leading to contract failure. Similarly, the indirect association between farmers’ satisfaction and market access was supported by one FGD participant farmer who voiced,
“I believed that contract farming solves my previously existing market problems for soybean products. However, the contract firm was not willing to take my soybean product because of the lower price”.
This suggets that market access is still a systemic constraint of soybean contract farming in northwestern Ethiopia during the study period. Although the governmental policy is intended to promote and develop linkage, it might negatively influence farmers’ contract farming satisfaction by imposing costly requirements. The Ethiopian Agricultural Product Producer and Contractor Relationship Directive No. 1027/2017, Article 3(3) states that the formal market arrangement expects contracting firm to provide logistics related supprts to the producers. Although this sounds supportive, if the firm deducts fees from the producers’ payout as it controls the sale, the regulation operates in an oligopoly-like informal market activity, as noted by the KII members.
Likewise, contract enforceability (β = -0.16, p = 0.031) undesirably affects the latent farmers’ satisfaction of soybean contract farming in northwestern Ethiopia. Although good contract enforceability is important for minimizing risks for the contracting company (e.g., quality and quantity, and side-selling control) and the contract participant farmers, [46], however, too much poor enforcement might impose higher costs on farmers, thereby decreasing their satisfaction by eroding their yield margins. A KII participant (P5) suggested
“The soybean contract enforceability is weak, which means there is no clear and detailed agreement, especially on seed quality and its sources, price strategies, and the required time for meeting farmers and contract companies. Simply, the contract companies need to collect farmers’ green cards, and they distribute some inputs for farmers. During the production season, most farmers lost hope because buyers drove down the price of soybeans. Due to this, most of the SCF participants become discouraged”.
The other FGD participant strengthens this statement:
“No one is on the side of farmers’ protection; the office of agriculture has tried to protect our interests, but it has little influence to convince the organized market actors”.
Those suggest that, if the contract companies fail to perform as promised, the farmers can’t use the court system to enforce the contract and recover damages or other remedies because the terms of the offer are not clear and definite in the soybean contractual scheme. The legal framework attempts to solve this, but may instead add layers of discouragement. The Ethiopian Agricultural Production Contract Proclamation No. 1289/2023, Article 5(4) instructs that a contract “shall be confirmed by a minimum of three witnesses, one of them must be a representative of the proper authority”. By directing the incorporation of an administration official for each contract activity, the law may create a systemic block. If the procedure of meeting witnesses is costly, or if the representative is unavailable, the soybean producer remains in a state of “weak enforcement” where they have no “detailed and clear contract,” leading to loss of hope or discouragement, as defined in the study.
And also, the quality of inputs (in this case, the quality of soybean yield) supplied or promised by the firm mainly governs farmers’ overall CF satisfaction. Poor seed quality is recognized as a highly significant negative constraint of farmer satisfaction in the contract farming activities (β=−0.21; p = 0.009). This underlines how the contract firm’s failure to give or recommend on seed quality presents production risks for the contract participant farmers. Particularly, while the remaining variables are constant, a unit in the quality of soybean seed amount declines the probability of perceiving farmer dissatisfaction by 21% units. Therefore, the delivery of substandard soybean inputs directly drives soybean crop failure and results in poor beneficial [47]. Assuring the provision of good-grade soybean seeds is then not crossable; it helps as a basic requirement for getting socially acceptable and economically sustainable CF activities in the study area.
On the other hand, contractual price (β = -0.20, p = 0.006) reveals a statistically significant and negative influence, signifying that there are no timely and favorable payment activities, together with poor prices offered for soybean produce, to enhance contract farming achievements through improving farmers’ satisfaction within soybean contract farming schemes in northwestern Ethiopia. The result is consistent with [49], which confirms “farmers’ perceived values hurt farmers’ satisfaction”. This is also strengthened by the law, Directive No. 1027/2017, Article 6(1) (price adjustments), and Article 6(2) (dispute resolution mechanism), which creates an economic discouragement for farmers, leading to the failure of the contract scheme.
4.3.2. Factors influencing farmer yield (equation 2).
The second equation of bivariate Tobit regression results emphasizes the investigations of contract farming success in terms of farmers’ yield, with coefficients reflecting the effect on it. The variable market access (β = -0.26, p = 0.000) in the study area reduces farmers’ contract farming success by lowering their yield margins due to a lack of buyers (Table 1). A one unit increase in market availability is linked with a 26% in the probability of achieving the full, uncensored yield potential. This is also reported as “contract farming lowers the incomes of smallholder farmers because the contractors use greater market power over the farmers” [50]. Correspondingly, a KII participant stated:
“We (farmers) were focused on almost monoculture (soybean production) activity by reducing our habits of diversification due to the agreements that we made with investors. However, the contractors are not obliged to take the product at the agreed prices. Not only this, but also, we get market challenges in the open market. Due to such circumstances, I will not participate in soybean contract farming in the coming farming season. “
In simple words, this qualitative observation shows how much marketing issues within the contract scheme are critical. problem during the study time.
Seed quality (β = -0.24, p = 0.029) also showed a slightly important negative correlation with yield. These indirect associations between the seed quality and farmers’ contract farming yield in northwestern Ethiopia indicate that utilizing poor quality of soybean seed becomes a source of fear for farmers, and it decreases farmers’ expected return on their soybean production, in contrast to [47], which reports that good quality of seed increases yield through increasing yields. As a result, poor quality of soybean seed is the main constraint that affects farmers’ yield in contract farming, undesirably. As one KII respondent voiced,
“We, farmers, only saw that the seed was soybean when taking the seed from the government; we believed that everything about the quality of the seed was in the hands of the government; we strongly expected from them to take, not only physically good but also marketable, disease & drought resistant, and customer-preferred color-based seed. However, two problems occurred with the seed: the first problem occurred before the seedling. The seed was broken into two equal parts on the plot before sowing, indicating there are some quality problems with the seed, and the other issue occurred during marketing activities”.
This suggests that while soybean input delivery is crucial, its encouraging dependent on farmers’ satisfaction might be weakened unless solved by appropriate support and risk-sharing mechanisms.
4.4. Limitations of the study
While this research uses household as a unit of analysis to examine the firsthand experiences of smallholder farmers mainly facing structural problems, it acknowledges some procedural boundaries. By applying a cross-sectional design, the study finds systemic issues, but it has a limitation in its ability to establish causal inference compared to longitudinal methods. Additionally, the dependence on farm households based primary data may introduce possible self-reporting bias and affect the external validity of the regional report to various institutional contexts. Lastly, the numerical model relies on standard econometric conventions regarding functional specifications and error terms, which should be considered when interpreting the systemic constraints of contract farming success.
5. Conclusion and recommendation
The econometric analysis result provides a comprehensive understanding of the constraints influencing contract farming success, as defined and quantified by farmers’ yield and satisfaction in northwestern Ethiopia. The findings show that critical structural as well as operational influencers exert a statistically significant negative effect on both outcomes of soybean contract farming success, strongly suggesting policy measures. Specifically, soybean contract farming households’ satisfaction is critically hindered by poor seed quality, market access, poor enforceability, and contract price issues. Similarly, soybean yields are negatively influenced by poor seed quality and market-related issues. Moreover, the key constraints that influence both outcomes were poor seed quality and limited market power. These critical results, strongly supported by qualitative data from KIIs and FGDs, show that lack of seed quality, poor contractual enforcement, and market failures like illegal oligopsony market behavior are the central existing systemic constraints in soybean contract farming.
Policy interventions should address the major constraints identified by this research. Firstly, since the soybean producers are affected by highly imperfect market conditions, the government should take measures to control oligopsony power and broker participation through empowering agricultural cooperatives and enforcing anti-collusion measures in the contracting environment. Secondly, given the statistically significant negative coefficients for price agreement and contract enforceability, developing a legal and fair contractual environment is essential for the success of contract farming in northwestern Ethiopia and other areas in which contract agreement is found laissez-faire. This requires strengthening collateral provisions and guaranteeing timely and competitive payment to restore farmer trust and solve the negative influence of the lack of contract enforceability and poor contract pricing. Finally, the consistently negative impact of seed quality on both yield and satisfaction requires that the responsible body ensure high-quality inputs and transparent dealings with seed suppliers to build trust and productivity. It can be achieved by identifying and addressing the sources of the seed quality problems and ensuring the inclusion of certified or legally recognized seed suppliers.
Therefore, by revising overall contract terms, including payment terms, input quality, legal enforcement, and market fairness, policymakers can meaningfully enhance the success and sustainability of soybean contract farming, thereby accelerating structural transformation within the local agriculture.
Acknowledgments
The authors would like to acknowledge the University of Gondar (UoG) for the research grant and the UoG Postgraduate Strength Program/or UoG Research Affairs Directorate Office for their additional financial support in this research. It is also required to acknowledge Dr. Abreham Assefa (PhD, MD, Senior Scientist), and Yordanos Sete (PhD student) for their immeasurable support. Finally, the authors have special acknowledgment to Dr. Genanaw Agitew (Asso. Prof), Dr. Abebe Dagnew, Dr. Daniel Nigussie, Tadie Mirie (Assis. Prof), Tesfea Kassahun, and all the staff of the Department of Agricultural Economics. Lastly, the authors acknowledge the use of Gemini (Google AI) for grammatical and language refinement.
References
- 1. Ton G, Vellema W, Desiere S, Weituschat S, D’Haese M. Contract farming for improving smallholder incomes: What can we learn from effectiveness studies? World Dev. 2018;104:46–64.
- 2. Llewelyn R, Williams J. Actors influencing the price of agricultural products and stability counte. AgEcon Search. 1996;18.
- 3. Molla HS, Ayele ZA, Zeleke MA. Trends, opportunities, and challenges of the Ethiopian soybean export market in the past two decades (2004-2022). Adv Agric. 2024;2024.
- 4. Woundefiraw B. Soybean Production and Marketing in North-Western Ethiopia: In the Case of Jawi District, Amhara Regional State. J Market Consum Res. 2023;89(2016):1–11.
- 5. Getahun A, Tefera E. Soybean Value Chain Assessment Study in Northwestern Ethiopia, Metekel Zone. AJAEES. 2016;14(4):1–14.
- 6.
Takele F, Debala C, Abebe Z. Optimizing crop geometry for enhanced growth and yield of soybean [Glycine max L (Merrill)] maturity groups in western Ethiopia. 2024;1:1–16.
- 7.
Mussema R, Diro S, Erko B, Teshale D, Dibaba R, Tesfaye A, et al. Soybean Value Chain Analysis in Ethiopia: A Qualitative Study [Internet]. 2021. Available from: http://www.eiar.gov.et
- 8.
Musema R, Diro S, Erko B, Teshale D, Dibaba R, Tesfaye A. Soybean Value Chain Analysis in Ethiopia: A Qualitative Study. 2022. 2–78 p.
- 9. S. Y A, Guledagudda SS, Kulkarni GN, Somanagouda G. Constraints in the Production and Marketing of Soybean. Asian Res J Agric. 2024;17(4):486–91.
- 10. Nigussie D, Alemu G, Gebreyes M. Understanding Ethiopia’s maize innovation system: Application of integrated innovation system and sustainability transition perspectives. African J Sci Technol Innovat Dev 2024;16(6):719–36.
- 11. Mishra AK, Kumar A, Joshi PK, Souza AD. Production risks, risk preference and contract farming: impact on food security in India. Food Sec. 2018;40(3):353–78.
- 12. Mounirou I. Does participation in contracts affect agricultural income? An empirical evidence from parboiled rice farmers in central Benin. Cogent Food Agric. 2020;6(1):1800237.
- 13. Saroj, Paltasingh KR, Jena PK. Does contract farming enhance farm efficiency? A case of wheat growers of Haryana, India. Heliyon. 2023;9(4):e15222. pmid:37095980
- 14. Ba HA, de Mey Y, Thoron S, Demont M. Inclusiveness of contract farming along the vertical coordination continuum: Evidence from the Vietnamese rice sector. Land Use Policy. 2019;87:104050.
- 15.
Jerjero T, Muluneh M. Evaluation of improved soybean varieties (Glycine max) under rain fed condition for traits of yield and yield components at Ari zone, South Ethiopia. 2025;4(01):1–5.
- 16. Alemaw G, Gurmu F. Towards edible oil self-sufficiency in Ethiopia: Lessons and prospects. Cogent Food Agric. 2023;9(1).
- 17. Ethiopian Government. Proclamation to provide for agricultural production contracts. Negarit Gazetta. 2023.
- 18.
Ethiopian Ministry of Planning and Development. Ten-year development plan a pathway to prosperity (2021-2030). Planning and Development. [Online]. Available: https://mopd.gov.et/en/
- 19.
National Bank of Ethiopia & Ethiopian Ministry of Agriculture. Ethiopian National Agricultural Finance Implementation Roadmap (NAFIR) 2025-30. National Bank of Ethiopia and Ministry of Agriculture. [Online]. Available: https://nbe.gov.et/wp-content/uploads/2025/07/NAFIR.pdf
- 20.
Kelemu MH, K. Trends in Soy Bean Trade in Ethiopia. 2014;3(9).
- 21. Tesfaye A. Progress of soybean [Glycine max (L.) Merrill] breeding and genetics research in Ethiopia: A review. Ethiopian J Agric Sci. 2018;8(13):67–77.
- 22. Jarrell SB, Stanley TD. Meta-regression analysis: A quantitative method of literature surveys. J Econ Surv. 1989;19.
- 23. Thorlund K, Imberger G, Johnston BC, Walsh M, Awad T, Thabane L, et al. Evolution of heterogeneity (I2) estimates and their 95% confidence intervals in large meta-analyses. PLoS One. 2012;7(7):e39471. pmid:22848355
- 24. Vesterinen HM, Sena ES, Egan KJ, Hirst TC, Churolov L, Currie GL, et al. Meta-analysis of data from animal studies: a practical guide. J Neurosci Methods. 2014;221:92–102. pmid:24099992
- 25. Olounlade OA, Li G-C, Kokoye SEH, Dossouhoui FV, Akpa KAA, Anshiso D, et al. Impact of Participation in Contract Farming on Smallholder Farmers’ Income and Food Security in Rural Benin: PSM and LATE Parameter Combined. Sustainability. 2020;12(3):901.
- 26. Panda B, Leepsa NM. Agency theory: Review of Theory and Evidence on Problems and Perspectives. Indian J Corp Gov. 2017;10(1):74–95.
- 27. Pies I, Valentinov V. Trade-offs in stakeholder theory: an ordonomic perspective. SRJ. 2023;20(5):975–97.
- 28.
Ijaz S, Tehseen S, Shah FA, Ahmed S. Power dynamics and balance of power in international relations. 2024;11:1750–9.
- 29. Sinnewe E, Charles MB, Keast R. Australia’s Cooperative Research Centre Program: A transaction cost theory perspective. Res Policy. 2016;45(1):195–204.
- 30. Johnston M. Systemically Valuable Transaction Costs. SSRN Electron J. 2024.
- 31. Hennart J-F. Transaction Cost Theory and International Business. J Retail. 2010;86(3):257–69.
- 32. Geremew AB, Gelagay AA, Yeshitila HY, Azale T, Habitu YA, Abebe SM, et al. Suicide attempt and its determinants among youth in central, north, and west Gondar zones, northwest Ethiopia: Findings from the youth health project. Front Psychiatry. 2023;13:1031402. pmid:36684019
- 33. Woretaw E, Teshome M, Alene M. Episiotomy practice and associated factors among mothers who gave birth at public health facilities in Metema district, northwest Ethiopia. Reprod Health. 2021;18(1):142. pmid:34215256
- 34.
Cochran WG. Samling Techniques [Internet]. John Wiley & Sons, Inc.; 1977. Available from: https://books.google.com/books/about/Sampling_Techniques.html?id=8Y4QAQAAIAAJ
- 35.
Tendolkar VD, Behere P, Quazi Z, Gaidhane A. Heterogeneous group discussion to improve reliability and validity of data tool. 2017. https://doi.org/10.4103/jdmimsu.jdmimsu_22_17
- 36.
Nyumba TO, Wilson K, Derrick CJ, Mukherjee N. The use of focus group discussion methodology: insights from two decades of application in conservation. 2018;2018(July 2017):20–32. https://doi.org/10.1111/2041-210X.12860
- 37. Elaete S, Ilukor J, Isoto RE, Moruleng A, Opolot HN. Comparing indigenous and scientific (Palin test) soil fertility assessment approaches for better soil health monitoring, management, and farm decision-making in Uganda. Discov Soil. 2025;2(1).
- 38. Baardman L, Cristian R, Perakis G, Singhvi D, Skali Lami O, Thayaparan L. The role of optimization in some recent advances in data-driven decision-making. Math Program. 2022;200(1):1–35.
- 39. Heckman JJ. Sample selection bias as a specification error. Econometrica. 1979;47(1):153–61.
- 40. Asfaw DM, Shifaw SM, Belete AA. Determinants of Market Participation Decision and Intensity among Date Producers in Afar Region, Ethiopia: A Double Hurdle Approach. Int J Fruit Sci. 2022;22(1):741–58.
- 41. Louzada F, Ferreira PH. On the classical estimation of bivariate copula-based Seemingly unrelated tobit models through the proposed inference function for augmented margins method. J Data Sci. 2021;13(4):771–94.
- 42. Khatun FM, Haider ZM. Impact of technology adoption on agricultural productivity. J Agric Crops. 2016;2(9):2412–6381.
- 43. Liu Y, Chen M, Yu J, Wang X. Being a happy farmer: Technology adoption and subjective well-being. Journal of Econ Behav Organ. 2024;221:385–405.
- 44.
Jolliffe P, Bickel P, Diggle S, Fienberg K, Krickeberg I, Olkin N, et al. Principal Component Analysis. Second. Aberdeen AB24 3UE, UK: Department of Mathematical Sciences, King’s College, University of Aberdeen; 2002. 2002 p.
- 45. Otsuka K, Nakano Y, Takahashi K. Contract Farming in Developed and Developing Countries. Annu Rev Resour Econ. 2016;8(1):353–76.
- 46.
Kabwe S, Mutambara J, Mujeyi K, Blackmore E, Vorley B, Weng X. Contract farming and informality: drivers and governance responses in Zambia and Zimbabwe. [Internet]. 2018. 182 p. Available from: http://pubs.iied.org/17617IIED
- 47. Mishra AK, Kumar A, Joshi PK, D’souza A. Impact of contracts in high yielding varieties seed production on profits and yield: The case of Nepal. Food Policy. 2016;62:110–21.
- 48. Fischer S, Wollni M. The role of farmers’ trust, risk and time preferences for contract choices: Experimental evidence from the Ghanaian pineapple sector. Food Policy. 2018;81:67–81.
- 49. Gu H, Liu Y, Qian F, Wang Q, Dong X. An Empirical Analysis of the Factors Affecting Farmer Satisfaction Under the China Link Policy. Sage Open. 2021;11(2).
- 50. Braimah R. Effects of contract farming on small-holder soybean farmers’ income in East Gonja Municipality - Savannah Region, Ghana. J Econ Sustain Dev. 2022;7.