Developing monitoring and evaluation frameworks for climate-smart agriculture extension programs in semi-arid regions

Introduction

The idea of systematically tracking whether agricultural programs actually work is surprisingly recent. Through much of the 20th century, extension services in Africa operated on the assumption that more training automatically meant better farming [1] . Evaluation, when it happened at all, consisted of counting how many farmers attended workshops or how many pamphlets were distributed. Whether those contacts translated into changed practices, higher yields, or improved livelihoods was rarely measured with any rigor [2] . The push for results-based management in development programming during the 2000s changed the landscape. Donor agencies began requiring logical frameworks, theoryof-change models, and measurable outcome indicators as conditions for funding agricultural projects [3] . But the extension sector adapted slowly. A 2019 review of 84 extension programs across 12 West African countries found that only 18% had a written M&E plan, and just 9% collected outcome-level data beyond simple output counts [4] . Climate-smart agriculture has raised the stakes. CSA programs aim to simultaneously increase productivity, build resilience to climate variability, and reduce greenhouse gas emissions a triple objective that makes evaluation more complex than for conventional extension [5] . Without a framework that tracks indicators across all three dimensions, program managers can't tell whether they're meeting their goals or simply shifting problems from one area to another. Côte d'Ivoire's semi-arid northern savanna zone faces particular challenges. Annual rainfall ranges from 900 to 1,200 mm but distribution is increasingly erratic, with longer dry spells interrupting the growing season [6] . The government launched a national CSA extension initiative in 2019, covering 14 northern departments. But by 2022, no standardized M&E system existed to assess whether the program was achieving its intended outcomes [7] . This research responded to that gap by developing a fourtier, results-based M&E framework tailored to CSA extension in semi-arid environments, then field-testing it across six sub-districts of northern Côte d'Ivoire over a 10-month period. The aim was to produce a replicable framework that other CSA extension initiatives could adapt to their own contexts [8] .

Theoretical framework

The M&E framework drew on two established evaluation models. First, the logical framework approach provided the input-activity-output-outcome chain that structures program evaluation vertically [9] . Second, the utilization-focused evaluation philosophy of Patton guided the design toward practical usefulness rather than academic rigor for its own sake every indicator was selected because program managers confirmed they would act on the data [10] . The integration of these two traditions produced a framework that was both structured enough to generate comparable data across sites and flexible enough to support adaptive management decisions. The continuous feedback loop connecting outcomes back to inputs was informed by action research principles, recognizing that extension programs in volatile climatic conditions need to adjust mid-course rather than waiting for end-of-project evaluations [11] .

Material and Methods

Material

The research engaged 476 farming households enrolled in the national CSA extension initiative across six sub-districts in Korhogo, Boundiali, and Ferkessédougou departments. Thirty-eight extension agents responsible for delivering CSA training in these areas participated in framework implementation. Data collection instruments included a farmer-level monitoring form (completed monthly by agents), a quarterly outcome assessment survey (administered to households), and an agent self-evaluation checklist. Secondary data came from the national meteorological service (rainfall and temperature records) and the regional agricultural directorate (program enrollment and budget figures for 2019-2023) [12] .

Methods

The research followed a design-implement-evaluate sequence. During a preparatory phase in November-December 2022, the framework was developed through three stakeholder workshops involving extension managers, field agents, researchers, and farmer representatives. The framework was then implemented from January to October 2023. Extension agents collected input and activity data weekly, output data monthly, and outcome data quarterly. Outcome indicators included CSA practice adoption (verified through field observation), farmer knowledge (30item assessment), soil moisture retention (gravimetric sampling at 15 cm depth), and crop yield (harvest weight from 5×5 meter sample plots) [13] . Agent framework adherence was scored on a 0-100 scale based on completeness and timeliness of data collection. Multiple regression tested the relationship between agent adherence, household characteristics, and farmer-level outcomes. All analyses were performed at Faculty of Earth Sciences and Agricultural Technology, Nangui Abrogoua University, Abidjan, Côte d'Ivoire, using R version 4.3 and Excel.

Ethical considerations

The research protocol was approved by the Ethics Review Board at Faculty of Earth Sciences and Agricultural Technology, Nangui Abrogoua University, Abidjan, Côte d'Ivoire (Protocol ERB/AGR/2022-043, approved October 2022). All participating farmers provided verbal informed consent witnessed by community leaders. Extension agents gave written consent for their performance data to be included in the analysis. Confidentiality of individual agent performance scores was maintained throughout reporting. The regression model explained 52.7% of variance in the composite farmer outcome index (adjusted R² = 0.527, F = 28.4, p<0.001). Agent M&E adherence was the strongest predictor (beta = 0.41), indicating that agents who followed the framework more closely produced better farmer-level results.

Results

Comprehensive interpretation

The data show that the M&E framework didn't just measure program performance it improved it. The strong association between agent adherence and farmer outcomes suggests that the process of systematic data collection and feedback itself enhanced service quality. Agents who tracked indicators regularly adjusted their approaches based on what the data showed, creating a virtuous cycle between monitoring and improvement.

Discussion

These findings challenge the common perception that M&E www.extensionjournal.com is an administrative burden that diverts resources from actual extension delivery. The evidence here suggests the opposite: systematic monitoring improved outcomes rather than detracting from them. This aligns with results from health sector M&E research, where routine data collection has been shown to improve service quality even before formal evaluation results are available [14,15] . The agent adherence finding (beta = 0.41) is particularly telling. Extension systems often invest heavily in training agents on technical content but give minimal attention to their monitoring and reporting skills [16] . This research suggests that M&E competency may be as important as agricultural knowledge in determining agent effectiveness.

The soil moisture dip episode illustrates a concrete mechanism through which continuous monitoring creates value. In a conventional end-of-project evaluation, this midseason challenge would have been invisible [17] . The framework's monthly data collection caught the problem early enough to trigger a corrective response that prevented crop losses in approximately 120 hectares across the affected sub-districts. The framework's four-tier structure was deliberately kept simple. More sophisticated M&E systems exist, including randomized controlled trial designs [18] . But these require statistical expertise that most district-level extension offices in West Africa don't have. The framework tested here was designed to be implemented by field agents with secondary education and basic computer skills [19] . Cross-district comparison revealed that framework implementation quality varied with agent workload sub-districts where agents served fewer than 80 households showed higher adherence scores (mean 78.4) compared with those serving more than 120 households (mean 61.2) [20] .

Practical applications

The framework is designed for direct adoption by extension programs operating in semi-arid West African contexts. Three practical steps are recommended: first, a two-day orientation workshop for extension agents covering data collection procedures and indicator definitions; second, monthly data review meetings at the sub-district level where agents discuss trends and agree on adaptive responses; and third, quarterly outcome assessments conducted jointly by agents and farmer representatives to maintain accountability. The data collection forms and indicator guides developed for this research are available for adaptation by other programs [21] .

Conclusion

References

1. Evenson RE, Mwabu G. The effect of agricultural extension on farm yields in Kenya. African Development Review. 2001;13(1):1-23.
2. Benor D, Harrison JQ, Baxter M. Agricultural Extension: The Training and Visit System. Washington DC: World Bank; 1984.
3. Kusek JZ, Rist RC. Ten Steps to a Results-Based Monitoring and Evaluation System. Washington DC: World Bank; 2004.
4. Moumouni IM, Baco MN. Analysis of monitoring and evaluation systems in agricultural extension programs in West Africa. Journal of Agricultural Extension. 2019;23(2):87-102.
5. Lipper L, Thornton P, Campbell BM, . Climate-smart agriculture for food security. Nature Climate Change. 2014;4(12):1068-1072.
6. Konaté M, Adja MG, Coulibaly TJ. Rainfall variability and agricultural adaptation in northern Côte d'Ivoire. Journal of Applied Meteorology and Climatology. 2020;59(8):1431-1445.
7. Ministère de l'Agriculture et du Développement Rural. Plan national d'investissement agricole de Côte d'Ivoire 2018-2025. Abidjan: MINADER; 2018.
8. FAO. Climate-smart agriculture sourcebook. Rome: Food and Agriculture Organization; 2013.
9. Gasper D. Evaluating the logical framework approach: towards learning-oriented development evaluation. Public Administration and Development. 2000;20(1):17-28.
10. Patton MQ. Utilization-Focused Evaluation. 4th ed. Thousand Oaks: Sage Publications; 2008.
11. Reason P, Bradbury H. The SAGE Handbook of Action Research: Participative Inquiry and Practice. London: Sage; 2008.
12. Konaté S, Traoré K. Agro-climatic characterization of the savanna zone of Côte d'Ivoire. African Journal of Agricultural Research. 2017;12(44):3224-3233.
13. Neufeldt H, Jahn M, Campbell BM, et al. Beyond climate-smart agriculture: toward safe operating spaces for global food systems. Agriculture and Food Security. 2013;2(1):12.
14. Braa J, Hedberg C. The struggle for district-based health information systems in South Africa. Information Society. 2002;18(2):113-127.
15. Aqil A, Lippeveld T, Hozumi D. PRISM framework: a paradigm shift for designing, strengthening and evaluating routine health information systems. Health Policy and Planning. 2009;24(3):217-228.
16. Davis KE. Extension in sub-Saharan Africa: overview and assessment of past and current models and future prospects. Journal of International Agricultural and Extension Education. 2008;15(3):15-28.
17. Stern E, Stame N, Mayne J, et al. Broadening the range of designs and methods for impact evaluations. DFID Working Paper. 2012;38:1-127.
18. Gertler PJ, Martinez S, Premand P, et al. Impact Evaluation in Practice. 2nd ed. Washington DC: World Bank; 2016.
19. Sulaiman VR, Hall AJ. The fallacy of universal solutions in extension: is ATMA the new T&V? LINK News Bulletin. 2004;1-4.
20. Ragasa C, Ulimwengu J, Josserand H, et al. Factors affecting performance of agricultural extension: evidence from the Democratic Republic of Congo. Journal of Agricultural Education and Extension. 2016;22(2):113-143.
21. CCAFS. Monitoring and evaluation for climate-smart agriculture programming. CGIAR Research Program on Climate Change, Agriculture and Food Security; 2019.