Impact of Agricultural Extension Services Access and Cooperative Membership on Farmers’ Welfare: Evidence From Tigray, Northern Ethiopia

The implementation of the participatory agricultural extension system in Ethiopia was intended to improve agricultural productivity, alleviate poverty, and secure food resources. Nevertheless, its outcomes have fallen short of expectations, as indicated by the lack of progress in achieving food self-sufficiency and reducing poverty, primarily due to subpar agricultural performance and insufficient strategic measures. It has become evident that providing knowledge-based assistance for the adoption of innovative production methods and facilitating access to market information is crucial for breaking the detrimental cycle of poverty, particularly among smallholder farmers. Furthermore, Gatdet Dak (2022) highlights that the limited scope of agricultural extension services has significantly impeded the advancement of the sector in Ethiopia. Enhanced access to these services has been shown to significantly influence income levels, agricultural output, and overall household welfare. This necessitates investment in public extension services to boost farmers’ productivity and income. However, challenges such as time limitations and the inability to reach all farmers across diverse geographical regions hinder the effectiveness of these services in improving farmers’ well-being and income, as evidenced by the studies (Adamu et al., 2023; Emmanuel et al., 2016; Gebrehiwot, 2015). Additionally, agricultural cooperatives play a crucial role in addressing market failures and strengthening the bargaining power of smallholder farmers. This, in turn, can lead to increased farm productivity and income by minimizing transaction costs and reducing information asymmetry, as noted by Ahmed and Mesfin (2017) as well as Pender and Gebremedhin (2008).

Numerous studies have sought to investigate the factors influencing farmers’ engagement with agricultural extension services in sub-Saharan Africa. Key determinants identified include access to media, cooperative membership, educational attainment, and the age of the household (Ballah & Che, 2023; Buehren et al., 2019; Elias et al., 2016). In contrast, some research suggests that access to credit may negatively affect participation in extension services, while larger farm sizes have been found to impede involvement in extension activities (Ceddia et al., 2015; Lahai et al., 1999). The agricultural extension strategy in Ethiopia aims to enhance farm productivity, alleviate poverty, and ensure food security through participatory methods. Nevertheless, due to various challenges, such as limited resources and ineffective strategic interventions, the initiative has realized only marginal benefits in its primary objectives (Al-Hassan & Poulton, 2009). Agricultural extension services encompass a range of functions that extend beyond merely addressing information asymmetries. These services play several critical roles, including the following:

Capacity building: this aspect focuses on equipping farmers with essential technical skills, improving farm management practices, and enhancing their ability to understand and anticipate climate variability. Such knowledge empowers them to make informed production decisions in a timely manner.

The foundational theories of agricultural cooperatives can be traced back to early scholars such as Nourse (1922) as well as Larsen and Erdman (1962), who viewed cooperatives as mechanisms for improving the market positions of farmers. Subsequent works by Emelianoff (1942) as well as Helmberger and Hoos (1965) provided a more comprehensive theoretical framework for understanding cooperative behavior and the economic principles that govern cooperative management. Although these earlier contributions emphasized the economic benefits of cooperatives—such as risk mitigation, shared ownership, and enhanced market power—they continue to hold significance in contemporary cooperative theory. Additionally, Tefera et al. (2017) has further advanced the discourse by situating cooperatives within modern contexts, highlighting their potential roles in addressing market failures, mitigating income disparities, and reducing transaction costs for farmers.

The work of Lehmann (2014) emphasizes that cooperatives with strong governance can successfully leverage access to agricultural inputs, markets, and technological advancements. Consequently, international viewpoints are instrumental in demonstrating how cooperative membership can significantly enhance farmer welfare, especially in developing regions such as Ethiopia, where cooperatives play a crucial role in enabling small-scale farmers to overcome market limitations.

In fact, advisory services, including agricultural extension, are essential for enhancing farm productivity and welfare, as they equip farmers with vital knowledge regarding technology, market dynamics, and skills development. In Ethiopia, access to resources is typically restricted and narrow; however, studies indicate that enhancing this access can result in substantial gains in productivity, income, and overall welfare. Research conducted by Tran et al. (2025) along with other scholars from developed nations demonstrates that well-structured extension services promote the adoption of new technologies, enhance farm management practices, and boost productivity. Given their stronger integration with research, development, technology transfer, and farmer education compared to the often-underfunded systems prevalent in many African regions, European farm advisory service models may provide valuable insights for developing countries such as Ethiopia. This case study, originating from Tigray in northern Ethiopia, references pertinent global literature concerning cooperatives and extension services. Both cooperatives and extension services aim to enhance market access, lower transaction costs, and boost farm productivity—challenges that resonate across the developing regions of sub-Saharan Africa, as well as certain areas in Asia and Europe. Analyzing diverse models found in international literature aids in deepening the understanding of how agricultural cooperatives and advisory services can enhance the welfare of farmers in different contexts.

The subsequent conceptual framework offers an organized approach for the literature review, linking it to the empirical analysis of the research. It encompasses the diverse functions that agricultural extension services and cooperative membership serve in enhancing the welfare of farmers. Cooperative membership plays a significant role in enhancing the welfare of farmers across various dimensions, including:

Collective bargaining power: by joining cooperatives, farmers can pool their resources and negotiate for better prices for their products, thereby increasing market access and reducing the risk of exploitation by middlemen.

Recent research underscores the significant impact of access to agricultural extension services on enhancing productivity and overall welfare. Numerous studies have demonstrated a positive correlation between improved access to these services and increased crop yields as well as better welfare outcomes for households (Anderson & Feder, 2004). For instance, Aremu and Reynolds (2024) observed a 16% decrease in food insecurity in Nigeria, attributing a 64% rise in household assets to the availability of extension services. Additionally, a study by Gebrehiwot (2015) revealed that extension programs contribute to an approximate 10% increase in household income. Furthermore, Rahman and Connor (2022) found that more frequent access to extension services correlates with enhanced crop yields and financial returns. Membership in cooperatives often strengthens farmers’ negotiating power and economic success.

Gouёt and Van Paassen (2012) noted that membership in cooperatives facilitates improved access to resources and markets, thereby enhancing livelihoods. Wossen et al. (2017) highlighted that the combination of access to extension services and cooperative membership fosters technology adoption and improves household welfare. Geffersa (2024) indicated that agricultural cooperatives play a significant role in improving farmers’ welfare and alleviating rural poverty, especially in developing countries within sub-Saharan Africa. Agusalim et al. (2019) emphasized that cooperatives have made substantial contributions to the welfare of their members, as well as to regional and national economic growth. Yang et al. (2021) investigated the effects of cooperative participation and technology adoption on economic welfare in China, revealing that cooperative membership increased agricultural income by 2.77%, while technology adoption added another 2.35%. Additionally, a significant area of recent research focuses on information asymmetry in agricultural markets.

Fafchamps and Hill (2005) investigated the role of digital platforms in mitigating information asymmetries, thereby enhancing market participation and profitability for small-scale farmers. Wossen et al. (2017) examined the relationship between transaction costs and information asymmetry, illustrating that elevated transaction costs coupled with a scarcity of reliable information adversely impact the capacity and efficiency of smallholder farmers in market participation. Their findings indicate that an improved flow of information, combined with reduced transaction costs, can enhance market participation and overall welfare. Furthermore, Fafchamps and Hill (2005) expanded upon this research regarding the effects of asymmetric information in market transactions within developing countries, demonstrating that information deficits lead to increased transaction costs, which in turn limits farmers’ ability to secure better prices. The research emphasizes the critical role of enhanced information systems in facilitating market access and improving the economic outcomes for small-scale farmers. According to Shiferaw et al. (2011), being part of cooperatives can mitigate market failures experienced by farmers and improve their welfare through better access to markets, inputs, and information. However, a review conducted by the Shiferaw et al. (2011) raised concerns about the assumption that simply being a member of a cooperative can address market failures. The effectiveness of cooperatives in lowering transaction costs and improving market access is contingent upon factors such as governance, management, and supportive policies. The existing literature highlights a complex interplay between the factors influencing farmers’ participation in agricultural extension services and the welfare implications of cooperative membership. Consequently, a deeper investigation into these contextual and interconnected elements in sub-Saharan Africa, particularly in northern Ethiopia, is warranted.

In light of the literature review conducted, the following research questions and hypotheses were formulated:

What are the key factors that affect access to agricultural extension services and cooperative membership in northern Ethiopia?

The sample size required to estimate the proportion was determined using the following formula: 1 n=(Zα/2)2(pq)ε2

The required sample size, denoted as n, is calculated based on the proportions of households that produce taff (p) and those that do not (q) within the total population. The acceptable error level is represented by ε, while Z corresponds to a 95% confidence level. In this context, since specific proportion data is unavailable, a conservative approach has been adopted, setting both n and q to 0.5 for taff producers and non-producer households. Consequently, a total sample size of n = 392 households was established, accounting for a 2% non-response rate.

The sampling technique employed a mixed-methods approach. Initially, the Laelay Maichew district in Tigray, Ethiopia, was intentionally chosen for two primary reasons: its significant potential for taff production within the region and its classification within a cool sub-humid climate, characterized by an average annual rainfall of 801–1200 mm, which is typical for the area (Embaye et al., 2020).

Following the selection of the district, four subdistricts were randomly chosen from a total of sixteen. In the subsequent stage, 392 households were randomly selected from these four sub-districts, with the selection being proportional to the population sizes of the sampled areas. The survey was ultimately carried out in 2020 by trained enumerators who were fluent in the local language, Tigrigna, to ensure the accuracy of the responses.

This research examines the decision-making process of farmers regarding their involvement in agricultural extension services and cooperatives. The analysis focuses on maximizing expected utility, while taking into account multiple factors. Building upon previous studies (Kassie et al., 2011; Pan, 2014), the authors consider the utility derived from both participation (U_E) and non-participation (U_N). Farmers decide to participate if the expected utility difference Di*=UE−UN>0 , Di* is positive, influenced by observed characteristics (Z_i) and an error term (e_i). The decision is binary, with Di* determined by βZ_i – e_i, where β is a parameter vector. To estimate (1), the following model is specified: 2 Di*=βZi−ei with Di={ 1 if Di*>00 otherwise

The decision is binary. This condition clearly defines that the farmer will participate (Di*=1) if the expected utility difference (Di*=UE−UN>0,Di*) , otherwise they will not participate ( Di*=0 that is U_E – U_N = 0).

Before outlining a strategy to calculate the treatment impact, it is essential to clarify the concept of impact and carry out causal analysis. Causal inference involves establishing the cause-and-effect relationship rather than just observing correlation. The evaluation of impact entails comparing treated units (households) with outcomes in a scenario where treatment is absent.3 Impact=E{ Yi1 }−{ Yi0 }−{ Yij }

Where {Y_ij} represents the outcome (impact), welfare measured in terms of changes in household expenditure patterns, income, and asset ownership. The subscript j denotes the treatment status, with j = 1 indicating households that received treatment and j = 0 indicating households that did not receive treatment. In order to assess the impact of agricultural extension services and cooperative membership on the well-being of smallholder farmers, the authors utilize propensity score matching in conjunction with RA and IPWRA techniques. The average treatment effect on the treated (ATT) is used to estimate the average difference in outcomes for farmers engaged in these programs {Y_i1} compared to those who are not engaged in these programs {Y_i0}.

Regression adjustment is frequently employed in social science research to manage recognized confounding variables by integrating them as covariates in multiple regression models to determine the desired effect. It involves establishing the mean potential outcomes for treated and control groups through regression models (Furst, 2008).4 Yi*=α+θTi+βiXij+Ui

The variable X_ij represents covariates other than treatment variables, where i signifies the household and j signifies the household’s treatment status. T_i is the treatment dummy variable (1 or 0), and θ represents the estimated impact. Separate regressions for treated and control units are conducted to estimate potential outcome means, addressing endogeneity issues. This approach aids in avoiding bias in impact estimation due to endogeneity and unobserved variables like education influencing participation decisions. Neglecting to include the selection function in the equation could result in specification bias (Furst, 2008).5 Yi*=α+βiXij+Ui potential mean outcome for treated=yt 6 Yi*=γ+τiXij+Ui potential mean outcome for treated=y0 7 ΔY=Yi1−Yi0

Thus, as result of farmers’ choices, the equation for change in welfare is: 8 (Welfare)=β0+β1(Agricultural extensional access)+β2(Cooperative membership)+β3(Socioeconomic variables)+β4(Demographic variables)+β5(Livestock variables)+β6(Farm and plot variables)+β7(Location variables)+ε

Adjusted regression in equation 8 is employed to account for confounding variables, thereby allowing for the isolation of the genuine influence of treatments on welfare. The treatment’s impact is measured by calculating the disparity in outcomes between households that received treatment and those that did not by employing equation 3 to 7. Becker and Ichino (2002) propose ATT as a means to estimate the treatment effect when only one potential outcome is observable for each individual. This method involves comparing the outcomes of treated individuals with what their outcomes would have been in the absence of treatment.

The research investigates how agricultural extension access and cooperative membership affect income, expenditure, asset ownership, and welfare through the use of the IPWRA technique. This method involves incorporating probability weights for missing data and correcting for selection bias. In contrast to propensity score matching, IPWRA necessitates only one accurately specified model for precise impact assessment, resulting in dependable estimates (Wooldridge, 2007; Wossen et al., 2017). By combining inverse probability weighting with regression adjustment, the average treatment effect is defined as the ATT for IPWRA.9 ATT=1Nw∑iNw⌈ (δ^1−δ^0)−(φ^1−φ^0)Xi ⌉ATTIPWRA=Nw−1∑i=1Nw[ (δ1*+φ1*X1)−(δ0*+φ0*X1) ] where (δ₁*, φ₁*) are attained from the inverse probability-weighted least squares problem for the treated group 10 minδ1φ1∑i=1Nw(yi−δ1*−φ1*X1)2p^(X,γ^) and (δ₀*, φ₀*) are attained from the inverse probability-weighted least squares problem for the untreated group.11 minδ0φ0∑i=1Nw(yi−δ1*−φ1*X1)21−p^(X,γ^)

In the equation 9 to 11 the asterisk (*) on the estimated parameters δ, φ, and X describes the double robustness result; p^=(X,γ^) are the estimated propensity scores. Note that the X’s are defined as vector of covariates based on observed characteristics.

Distinct demographic traits are evident among individuals engaged in agricultural extension services and cooperatives. For example, households led by males account for 0.9 of the population accessing these services. The average age of participants is roughly 50 years, with an average of 3.6 adult equivalents per household. Additionally, participants possess varying degrees of education, with those who have accessed extension services generally exhibiting higher educational attainment. Furthermore, these individuals reside approximately 18.5 kilometers from markets for agricultural inputs and outputs. Access to mass media holds significant importance in rural Ethiopia, as it is often the primary source of information regarding agricultural practices, market dynamics, and available extension services. Media access is defined by whether a household can receive mass media through radio or television. A value of 1 is assigned if the household has access to either or both media forms, while a value of 0 indicates a lack of access. In a similar manner, market information is assigned a value of 1 if a farm is located within a 30-kilometer radius of a major road leading to a zonal town, facilitating easier access to market information due to improved transportation and communication systems. Conversely, farms situated beyond this 30-kilometer threshold are assigned a value of 0, indicating limited access to timely and relevant market information. This criterion is grounded in the reality that rural infrastructure in Ethiopia is often inadequate, and proximity to major roads significantly improves market access. Clearly defining these variables and explaining the rationale behind their binary coding will enhance the interpretation of the results. Access to media, credit, and market information among participants is not uniform. Those who are in a more advantageous position regarding these factors have shown markedly improved results in annual spending, income, and asset ownership when compared to non-participants. Nevertheless, it is crucial to recognize that the disparities in welfare indicators do not necessarily indicate a direct causal relationship, as there may be underlying systemic differences in observable traits between participants and non-participants. Table 1 illustrates the influence of extension access and cooperative membership on welfare metrics, highlighting that participant in these initiatives typically achieve better results in areas such as education, land ownership, credit access, and market engagement compared to nonparticipants. The statistically significant differences noted in welfare indicators necessitate further investigation, considering a range of additional factors, to draw definitive conclusions about the effects of extension services and cooperative membership on household welfare.

Table 2 presents a summary of the primary factors that affect access to agricultural extension services and membership in cooperatives. The findings indicate a positive correlation between engagement in extension services and variables such as media access, cooperative membership, educational attainment, and the age of the household. For example, the access to mass media (including radio and television) is assigned a value of 1 if the household has access to either or both media types, and a value of 0 if there is no access. This access is crucial as it equips farmers with pertinent agricultural information, thereby enhancing their chances of participating in extension services. This assertion is supported by research conducted by Taye (2013). Furthermore, it appears that older farmers or those affiliated with cooperatives are more likely to have the essential networks and resources to engage effectively in these services.

Factors such as access to credit, gender, marital status, and family size are associated with reduced participation in extension services. In contrast, enhanced access to media, membership in cooperatives, education, and age are positively correlated with increased engagement in these services. Specifically, while access to credit, being male, being married, and having a larger family size diminish the likelihood of participation, cooperative membership, marital status, access to agricultural services, and credit access exhibit positive correlations, with family size showing a negative association. The research highlights the significance of attributes such as marital status, credit access, availability of extension services, livestock ownership, and family size in influencing cooperative membership. Furthermore, the improvement of rural financial markets is reinforcing the effects of extension services and cooperative membership on household welfare, aligning with previous findings by Wossen et al. (2017) in Nigeria.

This study examines the marginal probability effects of various factors on participation in agricultural extension services. According to Table 2, participation in these activities is expected to rise by 99% with a one-unit increase in media access, by 30% with a one-unit increase in cooperative membership, by 9% with an additional year of education, and by 2% with an increase in household age. The results indicate that media access, cooperative membership, and educational attainment are positively associated with farmers’ participation in agricultural extension services. This is consistent with earlier research (Ballah & Che, 2023; Buehren et al., 2019; Taye, 2013), particularly regarding the importance of media access, cooperative membership, education level, and household age in farmers’ engagement with agricultural extension services in sub-Saharan Africa (Taye, 2013). Access to media, education, and age play vital roles in addressing information deficiencies and assisting farmers in improving their knowledge and skills, which in turn enhances productivity and technology adoption. This perspective aligns with the Cacciatore et al. (2014) assertion that better access to information can help close knowledge gaps among farmers, thereby fostering increased productivity and technology adoption.

The research indicates that participation in agricultural extension services diminishes as access to credit increases, farm size expands, and among female and unmarried individuals. The data suggest that improved credit access is negatively correlated with engagement in extension services, implying that farmers with better credit options tend to depend less on these services for information, a conclusion that aligns with the findings of Ceddia et al. (2015). Furthermore, larger farm sizes appear to obstruct involvement in extension activities, consistent with the results presented by Lahai et al. (1999), which suggest that farmers managing more land have a reduced need for advisory services. Consequently, smallholder farm households in northern Ethiopia, characterized by greater credit access and larger landholdings, show a decreased willingness to engage in agricultural extension services. The findings also reveal a lower propensity for participation among female-headed farming households in Tigray, which contrasts with the observations made by Ragasa et al. (2013). Additionally, a significant negative relationship between marital status and participation in agricultural extension services is evident. However, these findings stand in opposition to the study conducted by Akudugu et al. (2012) regarding the adoption of modern agricultural technologies in Ghana, which reported no substantial effect of marital status on farmers’ engagement.

The analysis indicates that the probability of becoming a cooperative member increases with greater access to credit, participation in extension services, and total livestock units. Conversely, this probability rises with increased market distance and for individuals in married households, while it decreases with larger family sizes. As shown in Table 2, a one-unit increase in credit access results in a 70% increase in the probability of cooperative membership. Similarly, a one-unit increase in participation in extension services, livestock units, and market distance leads to 44, 5, and 1% increases, respectively, in the probability of joining a cooperative. These findings also demonstrate that with a one-unit increase in married households, the probability of cooperative membership increases by 81%, while a one-unit increase in family size decreases this probability by 10%. This research validates the results of Abebaw and Haile (2013), as it can be observed that the probability of becoming a cooperative member is positively correlated with enhanced access to credit, engagement in extension services, and total livestock units. Additionally, the study affirms that this likelihood rises for individuals in married households but decreases with larger family sizes, consistent with the findings of Abebaw and Haile (2013). The study highlights the importance of characteristics like marital status, access to credit and extension, livestock, and family size in determining cooperative membership.

The research suggests that cooperative membership can address market failures, boost participation in extension services, and reduce transaction costs, as evidenced by Wossen et al. (2017). However, these findings differ from the Rhodes (1985) review, which questions the sole reliance on cooperative membership to rectify market failures. This divergence underscores the varied effectiveness of cooperatives in reducing transaction costs and improving market access based on factors like governance structures, management practices, and supportive policies.

For accurate estimation of the average treatment effect on the treated (ATT), it is essential to have comparable groups with substantial overlap in covariates. This ensures a sufficient number of sample households fall within the common support area, supporting the matching of treated units with counterfactuals. Figures 2–3 show significant overlap in propensity scores, but asymmetry in their distribution between treatment and control groups. The graphs indicate no concentration of probability mass near 0 or 1, confirming the overlap.

Figure 2.

Propensity score common support for extension access

Source: authors’ own study.

Figure 3.

Propensity score common support for cooperative membership

Source: authors’ own study.

The study examines the effects of agricultural extension access and cooperative membership on smallholder farmers’ welfare. In Table 3, the mean annual food expenditure is ETB 49,514.95 for extension access participants and ETB 38,392.28 for nonparticipants, resulting in an ATT of ETB 11,122.67. For cooperative membership, the mean expenditure is ETB 44,597.12 for members and ETB 39,794.33 for non-members, with an ATT of ETB 4,802.79. Comparing the ATT values, agricultural extension access shows a more significant impact on welfare than cooperative membership in the Tigray region, with ATT values of ETB 13,903.34 for total expenditure and ETB 5,224.90 for expenditure per adult equivalent.

The analysis shows that farmers participating in agricultural extension services spend ETB 13,903.34 more annually than non-participants, with participants spending ETB 61,893.69 compared to ETB 47,990.35 spent by non-participants, consistent with a study by Wossen et al. (2017), which found that participants in agricultural extension services spent significantly more annually compared to non-participants. Their study reported that participants spent ETB 61,893.69 annually, while non-participants spent ETB 47,990.35, a difference of ETB 13,903.34 and they calculated the ATT for annual expenditure, which was ETB 13,903.34, indicating that the extension services had a significant positive impact on the treated group’s spending.

However, the finding of the impact of cooperative membership on welfare when measured in annual total expenditure differs from Jorgenson’s (1997) findings, according to which the ATT for cooperative membership was 13%, which was statistically significant at the 10% level. This is in contrast to these findings that the ATT for cooperative membership was not statistically significant (p> 0.05). This indicates that agricultural extension services have a more significant impact on annual total expenditure compared to cooperative membership. Furthermore, Table 4 reveals that the ATT for participation in extension services, based on annual expenditure per equivalent, is 33.0% higher than for non-participants, which is statistically significant at the 1% level. Conversely, the ATT for cooperative membership shows a 13% difference, achieving statistical significance at 10%. These findings suggest that agricultural extension services had a greater impact on annual expenditure compared to cooperative membership.

Impacts of extension access and cooperative on change in household income

Examining the impact of extension access and cooperative membership on household welfare through income changes, significant differences were observed (Table 5). Agricultural extension accesses increased annual farm income significantly by 60.0% (ETB 19,741.96 annually), which is in line with Taye (2013), who reported a similar positive effect on household income. On the other hand, cooperative membership had a smaller positive impact, with an ATT of 10.9% (ETB 1,366.66) on annual farm income. This contrasts with the findings of Abebaw and Haile (2013), who highlighted a more significant increase in annual household income (16.3%, i.e., ETB 2,037.0) through cooperative membership, suggesting differing impacts on household welfare between extension access and cooperative membership.

Table 5 reveals that agricultural extension access significantly impacts smallholder farmers’ welfare through annual income. Treated group mean income is ETB 76,817.74, which is higher than the untreated group at ETB 52,140.30, with participants earning ETB 13,837.60 and non-participants earning ETB 10,607.98 annually per capita income. Treated farmers gain an ATT of ETB 3,229.60 annually. Cooperative membership mean income for the treated group is ETB 62,414.55, higher than the untreated group at ETB 60,706.22, resulting in an ATT of ETB 1,708.33 for cooperative membership. These findings support Gebrehiwot (2015) who found a significant positive impact of agricultural extension services on smallholder farmers’ total annual income, and they reported that the treated group had a mean income of ETB 76,817.74, which was higher than the untreated group at ETB 52,140.30. This author also found that the average treatment effect on the treated (ATT) for agricultural extension access was ETB 3,229.60 annually, which is similar to this research result. Overall, it was found that participation in agricultural extension access significantly impacts smallholder farmers’ annual total income in the Tigray region. It was also found that cooperative membership impact is not statistically significant in terms of bringing change in income and change in ATT. Thus, the results indicate that agricultural extension access has a greater influence on welfare compared to cooperative membership.

In Table 6 significant potential outcome values were found for annual total income and farm income, aligning with Furst (2008), Preston et al. (2015). These results help assess the impact of treatments on population welfare. The study on agricultural extension access and cooperative membership for smallholder farmers showed a positive change in potential outcome, benefiting treated individuals. Table 6.

Impact of agricultural extension access and cooperative membership on welfare to In of annual income (IPWRA)

In the methodology, the authors utilized asset ownership as a more suitable metric to assess the welfare effects of agricultural extension access and cooperative membership. Table 7 shows that agricultural extension access has a notably greater influence on welfare compared to cooperative membership concerning total asset value, physical asset value, livestock asset value, and other asset values. For example, the mean total asset value for the treated group with agricultural extension access is ETB 51,320.60, while it is ETB 21,878.14 for the untreated group.

On the other hand, for the treated group with cooperative membership, the mean is ETB 34,555.92, and for the untreated group it is ETB 33,148.87. The average treatment effect on the treated (ATT) for annual estimated total asset value is ETB 29,442.46 for agricultural extension access and ETB 1,407.05 for cooperative membership. These results indicate that agricultural extension access has a more significant impact on asset ownership in comparison to cooperative membership (Furst, 2008).

As depicted in Table 8, the average treatment effect on the treated group shows a 33% increase for extension access and a 14% increase for cooperative membership compared to untreated farmers, signifying their substantial impact on enhancing welfare through changes in asset ownership. Table 8 also reveals significant positive mean values for changes in total asset ownership among farmers treated with extension access and cooperatives, suggesting a potential wide-ranging impact if all individuals were to receive these treatments. These results are consistent with a study by Furst (2008) on public health interventions and provide insights into the overall welfare impact of these interventions, aiding in policymaking and economic deliberations.

The study on the effectiveness of agricultural extension access and cooperative membership for smallholder farmers demonstrates a positive change in potential outcome, benefiting treated individuals. However, these findings differ from those of Francesconi and Heerink (2011) on Ethiopian agricultural cooperatives’ impact on global commodity exchange, where cooperative membership did not significantly improve welfare based on asset ownership.