The role of social networks and community organizations in disseminating agricultural innovations among smallholder farmers

Introduction

In Murang'a County, a retired teacher who grows coffee and bananas on half a hectare learned about a new pruning technique not from the local extension office, but from his son-in-law in Kiambu who had seen it at a field day two months earlier. This kind of informal information transfer happens thousands of times daily across rural Kenya, yet formal extension planning rarely accounts for it [1] . Social networks move agricultural information through channels that extension systems neither created nor control, and understanding how these networks function could dramatically improve how innovations reach farmers. Kenya has invested heavily in formal extension since independence, currently maintaining a public advisory workforce of approximately 6,800 agents serving over 8 million farming households [2] . The ratio roughly one agent per 1,200 households means that most farmers receive their agricultural information from someone other than an extension professional. Research from other East African countries has established that social contacts are the primary information source for 60-75% of smallholder farmers [3] . But not all social networks function the same way. Kinship ties are strong and trusted but may limit information to familiar practices. Cooperatives provide structured platforms but serve only members. Religious groups reach diverse populations but lack agricultural focus. Digital networks offer speed and scale but suffer from trust deficits [4, 5] . No comparative research has examined how these four network types differ in their capacity to disseminate different types of agricultural innovations within the same farming population. This research addressed that gap by mapping social network structures and measuring innovation diffusion outcomes across four network types among smallholder farmers in central Kenya's highlands [6] .

Conceptual model

The research was built on a social network diffusion framework that integrates three mechanisms through which networks influence adoption behavior. First, trust building www.extensionjournal.com farmers are more likely to try innovations recommended by people they trust. Second, information filtering networks screen and interpret technical information, translating it into locally relevant knowledge. Third, social learning farmers observe the outcomes of early adopters within their network before deciding to adopt themselves [7,8] . The model predicts that network effectiveness varies by innovation complexity: simple innovations (seed varieties) spread through any network with strong ties, while complex innovations (water harvesting, market coordination) require networks with organizational capacity and collective action structures.

Research area description

The research covered five sub-counties in Nyeri, Murang'a, and Kirinyaga counties of central Kenya. The area sits between 1,200 and 2,200 meters above sea level, receiving 1,000-1,600 mm of bimodal rainfall. Agriculture is dominated by smallholder mixed farming combining cash crops (coffee, tea) with food crops (maize, beans, potatoes) and dairy [9] . Average farm size is 0.6 hectares. The region has one of Kenya's densest social organizational landscapes, with an estimated 4,200 registered farmer groups, 180 dairy cooperatives, and numerous church-based community organizations operating across the five sub-counties.

Material and Methods

Material

The research surveyed 486 smallholder farmers from 40 village clusters across the five sub-counties. A structured questionnaire with 58 items captured network membership (type, duration, meeting frequency), information flow (sources of agricultural knowledge, trust ratings), and adoption status for six promoted innovations: droughttolerant maize varieties, soil conservation terracing, integrated pest management, rainwater harvesting structures, improved post-harvest storage, and collective market linkage [10] . Network structural variables density, tie strength, centrality, and bridging connections were measured using a name-generator module where respondents listed up to eight contacts from whom they received agricultural information. Secondary data came from county agricultural offices and the Kenya National Bureau of Statistics for 2020-2023.

Methods

A cross-sectional network analysis design was employed. Data were collected between March and October 2023 through face-to-face interviews. Adoption was verified through field observation for physically visible practices (terracing, storage structures, rainwater harvesting) and through seed purchase records or crop inspection for varieties. Network effectiveness was calculated as the proportion of network members who adopted each innovation. Radar charts visualized network-by-innovation effectiveness patterns. Multiple regression tested the relationship between network structural variables and an adoption index (composite score across six innovations). All analyses were conducted at College of Agriculture and Veterinary Sciences, University of Nairobi, Nairobi, Kenya, using UCINET 6 for network metrics and Stata 16 for statistical modeling. The model explained 51.4% of variance in the adoption index (adjusted R² = 0.514, F = 29.7, p<0.001). Network density the proportion of possible connections that actually exist was the strongest predictor (beta = 0.38). Tie strength, measuring frequency and intimacy of contacts, followed closely (beta = 0.31). Bridging connections that link otherwise separate groups also contributed significantly. Farmer cooperatives (dashed line) showed the most consistently high effectiveness across all innovation types, while kinship networks (solid line) peaked sharply for seed varieties but declined for collective practices. Digital networks showed an uneven profile, strong for market information but weak for practices requiring physical demonstration. The diagram illustrates how innovations flow from formal sources through four network types, are processed through three mechanisms (trust building, information filtering, social learning), and ultimately influence farmer adoption behavior.

Results

Discussion

The dominance of farmer cooperatives for collective-action innovations confirms theoretical predictions about organizational capacity and diffusion [11] . Water harvesting and market linkage both require coordinated group effort constructing shared infrastructure, negotiating collectively with buyers which cooperatives are structured to facilitate. Kinship networks lack this organizational capacity but excel where trust is the critical factor, as with seed sharing where a farmer needs confidence that the variety is genuine and suitable for local conditions [12] . The network density finding (beta = 0.38) has practical www.extensionjournal.com implications. Dense networks where many members know each other create redundant information pathways that reinforce messages and increase adoption probability [13] . Extension programs that identify and work through dense network clusters should achieve higher diffusion rates than those targeting isolated individuals. The bridging connections variable (beta = 0.22) highlights the complementary importance of links between groups, which introduce novel information that within-group ties alone cannot provide [14] . Digital networks showed a revealing pattern strong for market information (63.4%) but weak for physically complex practices like water harvesting (28.6%). This confirms that digital channels work best for codifiable information (prices, dates, locations) rather than tacit knowledge requiring hands-on learning [15,16] . Religious organizations contributed modestly across all innovations, suggesting they function as general information conduits rather than specialized agricultural channels [17] .

Limitations

The cross-sectional design captures network structures and adoption at one point, missing the temporal dynamics of how information actually flows through networks over time.

Self-reported network data are subject to recall bias respondents may over report strong ties and underreport weak ones. The research focused on central Kenya's densely networked highland communities; results may differ in areas with weaker social organization. Innovation adoption was treated as binary (adopted/not adopted), missing partial or modified adoption that commonly occurs in practice [18] .

Conclusion

This research demonstrates that social network type significantly shapes agricultural innovation diffusion, with effectiveness depending on the match between network characteristics and innovation requirements. Farmer cooperatives were the most effective overall channel, particularly for innovations requiring collective action, while kinship networks excelled at spreading individually adoptable practices like improved seed varieties. Extension programs should invest in network mapping before designing dissemination strategies. Rather than building parallel communication systems, programs can achieve wider and deeper diffusion by channeling different types of innovations through the network types best suited to carry them. The evidence supports a network-matched dissemination approach: seed varieties through kinship ties, collective practices through cooperatives, market information through digital platforms, and general awareness through religious and community organizations.

References

1. Bandiera O, Rasul I. Social networks and technology adoption in northern Mozambique. Economic Journal. 2006;116(514):869-902.
2. Kenya Ministry of Agriculture. National agricultural extension policy; 2020. Nairobi: Government Printer; 2020.
3. Katungi E, Edmeades S, Smale M. Gender, social capital and information exchange in rural Uganda. Journal of International Development. 2008;20(1):3552.
4. Conley TG, Udry CR. Learning about a new technology: pineapple in Ghana. American Economic Review. 2010;100(1):35-69.
5. Maertens A, Barrett CB. Measuring social networks' effects on agricultural technology adoption. American Journal of Agricultural Economics. 2013;95(2):353359.
6. Spielman DJ, Davis K, Negash M, et al. Rural innovation systems and networks: findings from Ethiopian smallholders. Agriculture and Human Values. 2011;28(2):195-212.
7. Rogers EM. Diffusion of Innovations. 5th ed. New York: Free Press; 2003.
8. Granovetter MS. The strength of weak ties. American Journal of Sociology. 1973;78(6):1360-1380.
9. Jaetzold R, Schmidt H, Hornetz B, et al. Farm management handbook of Kenya: natural conditions and farm management information. Vol. II/B. Nairobi: GTZ; 2007.
10. Borgatti SP, Everett MG, Johnson JC. Analyzing Social Networks. London: Sage; 2013.
11. Ostrom E. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press; 1990.
12. McGuire S, Sperling L. Seed systems smallholder farmers use. Food Security. 2016;8(1):179-195.
13. Valente TW. Network interventions. Science. 2012;337(6090):49-53.
14. Burt RS. Structural holes and good ideas. American Journal of Sociology. 2004;110(2):349-399.
15. Aker JC. Dial A for agriculture: a review of information and communication technologies for agricultural extension. Agricultural Economics. 2011;42(6):631647.
16. Nakasone E, Torero M, Minten B. The power of information: the ICT revolution in agricultural development. Annual Review of Resource Economics. 2014;6:533-550.
17. Knack S, Keefer P. Does social capital have an economic payoff? A cross-country investigation. Quarterly Journal of Economics. 1997;112(4):12511288.
18. Doss CR. Analyzing technology adoption using microstudies: limitations, challenges, and opportunities for improvement. Agricultural Economics. 2006;34(3):207-219.