Youth unemployment and underemployment are major challenges in LMICs, contributing to poverty, food insecurity, social unrest, and increased costs for public support programs. Agriculture, a vital sector in many LMICs, presents significant opportunities for job creation and poverty reduction. However, declining youth participation in agriculture, partly due to lack of education and skills training, necessitates effective interventions. This study systematically reviews the evidence on the impact of skills training programs designed to increase youth engagement in agricultural employment within LMICs. The review focuses on interventions such as agriculture-related courses, on-the-job training, technical or vocational education, and training in entrepreneurship and financial literacy. The outcomes of interest include job creation in the agricultural sector, self-employment and entrepreneurship, provision of and employment in extension services, profit/income/earnings, farm productivity, and accessibility of employment opportunities within the agricultural value chain. The significant interest from donors and policymakers in youth engagement in agriculture underscores the critical need for evidence-based evaluations of the effectiveness of existing interventions.

The existing literature highlights the significant challenges faced by youth in LMICs, with a disproportionate number experiencing working poverty. Studies demonstrate the link between youth unemployment and underemployment with negative long-term consequences such as decreased job satisfaction, lower income, poorer health, and increased social unrest. The projected doubling of the working-age population in LMICs in the next 35 years emphasizes the urgency of addressing youth employment. While agriculture offers substantial potential for poverty reduction and food security improvements, a declining trend in youth participation, alongside the aging farmer population, presents a significant obstacle. This decline is attributed to various factors, including the lack of agricultural education and skills training. However, studies also reveal a considerable interest among youth in improved farming methods and productive agricultural activities, challenging common assumptions about youth disinterest in the sector. The increasing availability of ICT provides opportunities to enhance skills development and access relevant information. This review specifically addresses the gap in evaluating the effectiveness of skills training interventions targeted at boosting youth engagement in agricultural employment in LMICs.

This systematic review followed the PRISMA-P guidelines. The research question focused on the effects of skills training interventions on youth employment outcomes in agricultural value chains in LMICs. Inclusion criteria encompassed studies targeting youth, focusing on agriculture, involving skills training interventions, published between 1990 and 2019 in English or French, employing a clear methodology, and originating from LMICs. An exhaustive search strategy across multiple electronic databases (CAB Abstracts, Web of Science, EconLit, Agricola, Scopus) and grey literature sources was employed. The search results were de-duplicated using Python scripts, and a machine-learning model was used to extract metadata, accelerating the screening process. Two independent reviewers conducted title/abstract and full-text screening using Covidence software, resolving discrepancies through consultation with a third reviewer. Data extraction employed a template to document study characteristics, intervention types, participant characteristics, outcomes, methods, and risk of bias. Risk of bias assessment considered eight domains: sampling technique, intervention type, area of study, targeted population, data collection method, data analysis method, outcome measurement, and statistical significance. Studies were classified as low, moderate, or serious risk of bias based on their scores. The analysis summarized the findings across studies based on intervention type and the outcomes of interest (job creation, self-employment, entrepreneurship, provision of extension services, productivity, profit/income, and employment opportunities).

The review identified 16 studies meeting the inclusion criteria. Eleven studies were based in Africa, and five in Asia. Study designs included nine quantitative, four qualitative, and three mixed methods studies. A risk of bias assessment revealed 15% of studies at low risk, 60% at moderate risk, and 25% at serious risk of bias. Across the studies, skills training interventions showed positive effects on various outcomes: * **Job Creation in Agriculture:** Several studies indicated a positive association between skills training and job creation in agriculture, with some demonstrating significantly higher employment rates post-training, especially for vocational training compared to on-the-job training. However, many studies lacking robust methodology limited the strength of these findings. * **Self-Employment in Agriculture:** Skills training interventions stimulated self-employment in agriculture. A radio campaign on sweet potato cultivation significantly increased engagement in this activity. * **Engagement/Entrepreneurship in Agriculture:** Skills training programs encouraged youth engagement and entrepreneurship in agriculture. Targeted youth programs were shown to significantly increase agricultural participation. * **Productivity of the Farm/Agriculture:** Skills training interventions led to higher farm productivity. One study showed that training programs led to significantly higher maize yields. * **Profit/Income Earning of the Farm:** Skills training interventions resulted in increased farm income and profits. One program led to a substantial increase in beneficiaries' income compared to non-beneficiaries. * **Job Search or Employment Opportunity:** A mixed-methods study showed positive effects on job search and employment opportunities. * **Provision of Agricultural Extension Service:** Training programs led to increased participation in agricultural extension services. The review also found that the interventions primarily used technical and vocational education and training (TVET) models, which tend to have shorter durations. The analysis further examines the findings based on different intervention types (agriculture-related courses, technical education/training, youth programmes, on-the-job training, vocational training, and combinations thereof). While there was a general indication of positive impacts, the heterogeneity in study designs and methodological rigor complicates drawing definitive conclusions.

The findings of this systematic review generally align with previous research on interventions to improve youth labor market outcomes. While skills training interventions show a positive trend towards improving agricultural employment outcomes, the lack of rigorous evaluations hampers the ability to make strong generalizations. The dominance of less robust study designs (case studies, qualitative methods) compared to rigorous methods like RCTs or quasi-experimental designs highlights a critical gap in the evidence base. This limitation underscores the need for more robust evaluations to generate reliable evidence of the effectiveness of various interventions and to facilitate comparisons across diverse contexts. The review also emphasizes the importance of incorporating ICT-based training approaches, demonstrated to be effective in some cases for increasing adoption of agricultural practices. However, the lack of attention to heterogeneity in factors like gender and educational background among youth participants limits the interpretation and generalizability of the findings. This necessitates future studies focusing on marginalized groups like illiterate youth. The findings underscore the need for cost-effectiveness analysis and return-on-investment studies to inform investment decisions and to encourage greater policymaker commitment to supporting such programs.

This systematic review reveals a consistent lack of rigorous evaluation of skills training programs aimed at enhancing youth engagement in agriculture in LMICs. While existing studies generally suggest positive impacts on various outcomes, the methodological weaknesses impede strong conclusions about program effectiveness. Future research should prioritize robust impact evaluations, including RCTs and quasi-experimental designs, focusing on diverse contexts and accounting for heterogeneity in youth characteristics. The integration of ICT-based training strategies, cost-effectiveness analyses, and long-term follow-up studies are crucial for generating more reliable evidence to inform policy and investment decisions. This improved evidence base will be vital in promoting sustainable youth engagement in agriculture, contributing to poverty reduction, and enhancing food security in LMICs.

The review's limitations include the limited number of studies identified, the methodological heterogeneity across studies, and the relatively high proportion of studies classified as moderate or serious risk of bias. The focus on studies published in English or French might have excluded relevant research from other languages. The lack of rigorous impact evaluations in many of the included studies hinders the ability to draw definitive conclusions about the effectiveness of different interventions. The absence of consistent data on long-term outcomes limits the assessment of the sustainability of the impacts of these training programmes.