Decentralized extension program planning and its influence on local agricultural innovation systems

Introduction

What happens to agricultural innovation when the people closest to the land have no voice in deciding which extension services they receive? This question has guided decades of policy debate across sub-Saharan Africa, where centralized top-down extension models have long dominated service delivery [1] . Extension programs designed at national headquarters with limited farmer input often produced a mismatch between services and actual needs, leading to poor adoption rates [2] . Tanzania introduced decentralization reforms in the early 2000s under its Local Government Reform Programme, transferring planning responsibilities to district and ward authorities [3] . Extension services were among the sectors affected, with District Agricultural Development Plans replacing nationally prescribed service packages in several pilot regions [4] . Agricultural innovation systems thinking treats knowledge dissemination as a networked process involving farmers, extension agents, researchers, and market actors [5] . Decentralization should create feedback loops allowing local priorities to shape extension programming [6] . Several researchers have argued that decentralized governance can improve service responsiveness, but evidence linking planning decentralization to measurable innovation outcomes remains limited [7] . Previous work in Kenya showed uneven outcomes depending on county capacity [8] , while Uganda's decentralized extension suffered from funding shortfalls [9] . Tanzania's framework included specific provisions for participatory planning at ward level [10] . This research aimed to compare innovation adoption and farmer engagement between decentralized and centralized districts in Morogoro and Dodoma [11] .

Theoretical Framework

This research drew on the Agricultural Innovation Systems (AIS) framework, which views innovation as an outcome of interactions among diverse actors within a network rather than linear transfer from research to practice [5] . The AIS perspective emphasizes that enabling environments governance arrangements, institutional incentives, and communication channels shape how knowledge moves within farming communities [12] . Decentralized planning represents a structural change in the enabling environment, potentially strengthening feedback mechanisms and making the innovation network more responsive to local conditions [13] . The research also drew on participatory governance theory, arguing that local stakeholder involvement builds ownership and increases sustainability of planned activities [6] .

Material and Methods

Material

This research was carried out from September 2021 to March 2022 in Morogoro and Dodoma regions of Tanzania. Six districts were selected purposively: three under decentralized planning (Mvomero, Kilosa, Morogoro Rural) and three with centralized structures (Mpwapwa, Kondoa, www.extensionjournal.com Chamwino). The research was based at Sokoine University of Agriculture, Morogoro. A total of 384 farming households were surveyed (64 per district) through multistage random sampling [14] . Semi-structured interviews were conducted with 42 extension officers and 18 local government officials.

Methods

Farmer participation was measured using a five-item binary index covering needs assessment, priority setting, resource allocation, activity design, and monitoring. Innovation adoption was tracked through a composite index of improved seeds, integrated pest management, and postharvest technologies using 2018-2022 district records and household recall data [15] . Analysis included descriptive statistics, independent t-tests, and multiple linear regression controlling for household size, education, farm size, and credit access. Qualitative data were analyzed through thematic coding based on AIS framework dimensions. Ethical approval was obtained from Sokoine University Research Ethics Committee (Protocol SUA-REC/2021/087, August 15, 2021) [16] .

Results

Discussion

These findings align with arguments that decentralized governance improves service delivery responsiveness [3] . The participation differences suggest decentralized planning changes the fundamental relationship between extension systems and farming communities farmers became active planning contributors rather than passive recipients, creating ownership that translates into higher technology uptake [6] . The widening adoption gap over time is notable. If decentralization merely provided a one-time boost, adoption curves would run parallel after the initial shift. The divergence instead suggests a compounding effect consistent with AIS theory, which predicts strengthened network interactions lead to self-reinforcing innovation processes [5] . The narrower resource allocation gap may indicate financial decisions remain partially controlled by higher government levels even in decentralized settings [8] . Tanzania's model appears to benefit from Ward Extension Officers maintaining professional linkages with district specialists, unlike Uganda's experience with weak technical oversight [9] .

Limitations:

The cross-sectional household survey limits causal claims, though longitudinal adoption data provide temporal perspective. Purposive district selection may introduce selection bias. Self-reported participation data could reflect social desirability bias. The research covered only two regions, and results may not transfer to areas with different agroecological conditions. Future investigations could use randomized designs exploiting variation in decentralization rollout timing.

Conclusions

This research provides evidence that decentralized extension planning is positively associated with farmer participation and innovation adoption in rural Tanzania. Decentralized districts showed participation rates roughly double those of centralized districts, and innovation adoption grew at a markedly faster pace over 2018-2022. Regression analysis confirmed planning type as the strongest adoption predictor. Transferring planning authority to district and ward levels appears to reshape how extension systems interact with farming communities, building trust, ownership, and responsiveness that support stronger innovation systems. Policy makers should weigh these results alongside concerns about local capacity and fiscal sustainability.

References

1. Anderson JR, Feder G. Agricultural extension: Good intentions and hard realities. World Bank Research Observer. 2004;19(1):41-60.
2. Birner R, Davis K, Pender J, Nkonya E, Anandajayasekeram P, Ekboir J et al. From best practice to best fit. Journal of Agricultural Education and Extension. 2009;15(4):341-355.
3. United Republic of Tanzania. Local Government Reform Programme: Policy Paper. Dar es Salaam: Government Printer; 1998.
4. Ngowi EE, Mwageni EA. Decentralization and agricultural service delivery in Tanzania. African Journal of Political Science. 2015;9(3):112-128.
5. Spielman DJ, Ekboir J, Davis K. The art and science of innovation systems inquiry. Technology in Society. 2009;31(4):399-405.
6. Chambers R. Rural development: Putting the last first. London: Longman; 1983.
7. Ribot JC. Waiting for democracy: The politics of choice in natural resource decentralization. Washington DC: World Resources Institute; 2004.
8. Kimenyi MS, Ndung'u NS. Devolution and agricultural service delivery in Kenya. In: Collier P, ed. Understanding Governance. Washington DC: World Bank; 2005:123-156.
9. Friis-Hansen E, Kisauzi D. Uganda: Evolution of the extension-farmer relationship. In: Rivera WM, ed. www.extensionjournal.com Extension Reform. Washington DC: World Bank; 2004:78-94.
10. Rutatora DF, Mattee AZ. Towards a farmer-led extension service in Tanzania. Journal of Continuing Education and Extension. 2001;2(2):23-38.
11. Leeuwis C, Van den Ban A. Communication for rural innovation. Oxford: Blackwell Science; 2004.
12. Hall A, Mytelka L, Oyeyinka B. Innovation systems: Implications for agricultural policy. ILAC Brief 2. Bioversity International; 2005.
13. Klerkx L, Aarts N, Leeuwis C. Adaptive management in agricultural innovation systems. Agricultural Systems. 2010;103(6):390-400.
14. Cochran WG. Sampling techniques. 3rd ed. New York: John Wiley & Sons; 1977.
15. Rogers EM. Diffusion of innovations. 5th ed. New York: Free Press; 2003.
16. Creswell JW, Plano Clark VL. Designing and conducting mixed methods research. 3rd ed. Thousand Oaks: SAGE; 2018.