Engagement of youth in agriculture in low- and middle-income countries may offer opportunities to curb underemployment, urban migration, disillusionment of youth and social unrest, as well as to lift individuals and communities from poverty and hunger. Lack of education or skills training has been cited as a challenge to engage youth in the sector. Here we systematically interrogate the literature for the evaluation of skills training programmes for youth in low- and middle-income countries. Sixteen studies—nine quantitative, four qualitative and three mixed methods—from the research and grey literature documented the effects of programmes on outcomes relating to youth engagement, including job creation, income, productivity and entrepreneurship in agriculture. Although we find that skills training programmes report positive effects on our chosen outcomes, like previous systematic reviews we find the topic to chronically lack evaluation. Given the interest that donors and policymakers have in youth engagement in agriculture, our systematic review uncovers a gap in the knowledge of their effectiveness.
Youth in low- and middle-income countries (LMIC) disproportionately experience working poverty. In 2019, about 21% of employed youth in LMIC were living on less than US$2 a day, compared with 16% of the overall working population1. In sub-Saharan Africa, nearly 70% of working youth were found to be living in poverty; in South Asia, close to 50% were living in poverty2. Issues of youth unemployment and underemployment are linked to greater likelihood of future unemployment, decreased future job satisfaction, lower income and poorer health in adulthood3. National consequences include greater costs to support public programmes (such as public work programmes that provide temporary jobs) and indirect costs of lower earnings such as loss of investment in education4,5. Furthermore, youth underemployment is linked to disillusionment and the possibility of social unrest6.
The working-age population in LMIC is predicted to double in the next 35 years7 and while this presents challenges, many LMIC are currently experiencing a demographic dividend phase where there is a high ratio of working-age population to dependents. This offers unique prospects for economic development with concomitant reductions in poverty and food insecurity. Addressing unemployment and underemployment is, therefore, a major policy priority for LMIC6, and a key sector for the creation of employment opportunities, especially in Africa and Asia, is agriculture6,8,9.
Many people in LMIC rely on agriculture for their livelihoods (32% in 2019)10, either directly, as farmers, or indirectly in sectors that derive their existence from agricultural production8,9,11. Agricultural development is estimated to be up to 3.2 times more effective in alleviating poverty in low-income, resource-rich countries than any other sector12. Due to the close links between poverty and food insecurity13,14,15, agricultural development could also have positive consequences for the alleviation of hunger, particularly for women, as their empowerment in agriculture improves households’ food security and nutrition16,17,18.
However, there has been a declining trend of youth participation in agriculture since 2000, mainly in favour of the service sector6,19,20, which precipitates migration from rural to urban areas. Increased educational attainment for rural youth coupled with inability to rent or own land is a driver of urban migration21. In addition, the increasing ageing farmer population in rural areas exacerbates the demographic pressure on land at the expense of the youth22.
A further constraint on youth engagement in agriculture is a lack of education in disciplines related to agriculture or skills training23,24,25. A study among Thailand’s youth reported that 71% identified knowledge of farming practices as a pre-requisite to setting up a viable farm23. In rural Ethiopia, government initiatives to increase skills and productivity, and introduce improved and modern farming methods have generated interest among youth in joining the sector, and in Indonesia, vocational training was noted as increasing the likelihood of a successful career in agriculture26. A study in Zambia on rural youth aspirations, opinions and perceptions on agriculture documented high interest among youth in more productive forms of farming, such as the use of draught animals, electricity and the increased application of fertilizers24. Such findings challenge an assumption common in policy proposals that youth are not interested in agriculture25. Today, with the development of information and communication technology (ICT), young people have more opportunities to strengthen their skills and access relevant information and are therefore well positioned to understand market dynamics, and institutional and financial systems, enabling them to initiate and capitalize on processes of change in the agricultural sector27,28. Human capital theory predicts a positive correlation between human capital accumulation and labour productivity. On that basis, skills training can be used to improve agricultural employment outcomes29. Where governments and policy interventions support skills training for youth, there is a real possibility for entrepreneurship, a competitive economy and ultimately national growth. But, despite the implementation of skills training interventions, generally via youth employment programmes30, few specifically target agricultural skills training in LMIC and very little is known about the effectiveness of youth agricultural interventions30,31.
Here we systematically review skills-based training interventions that aim to increase youth engagement in agricultural employment in LMIC to better inform investment decisions made by donors and policymakers. The interventions include agriculture-related courses, on-the-job training, technical or vocational education and training in agriculture, as well as general skills training including entrepreneurship, financial literacy and life skills for engagement in agriculture. The outcomes of interest we started out with were: employment along an agricultural value chain; employment in agribusiness; engagement in contract farming; development of agricultural entrepreneurship; agricultural business performance (productivity, profit, income, marketing rate); involvement in agricultural extension service provision. After data extraction, the outcomes of interest found in the selected studies are jobs created in the agricultural sector, self-employment and entrepreneurship, provision of and employment in extension services, profit/income/earnings from an agricultural activity or job, farm productivity, and the accessibility of employment opportunities in the sector. These outcomes pertain to the categories of jobs that can be found along the agricultural value chain.
We found among the studies yielded from the systematic literature search that skills training interventions reported employment in agriculture, agribusiness or agriculture-related activities, development of agricultural entrepreneurship, agricultural business performances (productivity, profit, income) and involvement in agricultural extension service provision for young participants. However, we also found a chronic lack of evaluation of the effectiveness of interventions designed to enhance agricultural opportunities and engagement for young people in LMIC, a finding previously shown31.
Sixteen studies were identified for review based on a priori inclusion and exclusion criteria (Fig. 1) detailed in our Preferred Reporting Items for Systematic Reviews and Meta-Analyses Protocol, PRISMA-P (Supplementary Material 1, summarized in Methods and published on Open Science Framework, https://osf.io/bhegq//).
Characteristics of selected studies
A data extraction template (Supplementary Table 2) was used to document all information of interest from each of the 16 studies, overviewed in Table 1.
Eleven of the studies were based in Africa32,33,34,35,36,37,38,39,40,41,42 and five in Asia43,44,45,46,47. Twelve of the studies were published in peer-reviewed journals33,34,35,36,39,40,41,42,44,45,46,47 and the rest originated from the grey literature, including one dissertation38, one report37 and two working papers32,43.
With regard to the study design, nine of the included studies were quantitative32,33,34,35,36,37,43,44,45, four were qualitative41,42,46,47 and three used mixed non-experimental38,39,40 methods. Only one study used randomized control trial (RCT) as a study design method of evaluation32. Quasi-experimental impact methods (difference-in-differences (DID) and propensity score matching (PSM)) and quantitative non-experimental methods (statistical and econometric methods) were used in two33,43 and six34,35,36,37,44,45 studies, respectively. Nine of the included studies relied on survey data32,33,34,35,36,37,43,44,45, one study used data from interviews47, one study used data from focus groups42 and the rest of the studies used mixed sources of data38,39,40 (Supplementary Table 3).
Table 2 collates information from the selected studies on the basis of types of intervention and participant characteristics. Technical education/training35,41,42,46 and vocational training37,40,44,45 constituted half of the interventions (four, each); youth programmes, agriculture-related courses and on-the-job training were identified as interventions in three33,34,38, two39,47 and one36 of the studies, respectively, and the remainder of the studies combined two types of intervention32,43. Twelve of the interventions were implemented through public policies33,34,35,37,38,39,41,42,43,44,45,47; non-governmental organizations (NGOs) and a mix of institutions (public and private) were each identified as implementers in two32,36 and one46 of the studies, respectively, and one study reported intervention implemented by an international institution40.
Nine of the studies solely targeted youth32,33,34,35,37,38,43,45,46, and seven targeted mixed groups of youth and others36,39,40,41,42,44,47. In fourteen studies, the participants were from all genders. In nine of the studies, participants were a mixed group of those already and not yet engaged in agriculture32,34,37,39,41,42,43,44,46; in five of the studies, participants were already engaged in agriculture before receiving skills training interventions35,36,37,45,47; there was not enough information to determine whether the participants were already engaged in agriculture in two studies33,40. Six of the studies indicated that the participants resided in rural areas33,34,35,36,46,47, while participants located in urban areas and in both rural and urban areas were identified in four32,38,40,45 and five37,39,41,43,44 of the studies, respectively; there was not enough information to determine the location of the participants in one42 study. The population targeted in the studies was both educated and non-educated youth. Among the nine studies32,33,34,35,37,38,43,45,46 that focused exclusively on youth, two targeted youth with a secondary education background34,46, one45 targeted youth with a university background and six32,33,35,37,38,43 of the studies targeted youth with a mixed educational background.
Risk of bias assessment
We evaluated the risk of bias of the included studies based on a previous approach48. The domains of risk retained are (1) the sampling technique used for the study, (2) the type of intervention, (3) the choice of the area of study, (4) the population targeted, (5) the method of data collection, (6) the method of data analysis, (7) the measurement of outcome and (8) the statistical significance of the effect. For each domain of risk, the criteria evaluated were defined and rated by their relevance for assessing the effectiveness of the interventions. Supplementary Table 4 summarizes the criteria of each domain of risk and its assessment and rating.
Using this scale, 15% of our included studies are at low risk of bias, 60% at moderate risk of bias and the remaining 25% at serious risk of bias. The outcome of the risk of bias assessment of the included studies in this systematic review is presented in Table 3.
The risk of bias assessment process highlighted differences in focus, methods used and standards of evidence across the included studies. Weaknesses in study design, survey methods and method of evaluation of the impact of the interventions were common in most of the studies (with the exception of the studies ranked at low risk of bias), leading to weak results and limited generalizability.
Effects on youth employment outcomes
The youth employment outcomes of interest to this systematic review are job creation, self-employment, engagement in entrepreneurship, provision of extension services, productivity of the farm/agriculture-related activities, profits/income, and job search or employment opportunity in agriculture-related activities. Here we elaborate on the study design and risk of bias of all studies, and highlight the effects on outcomes of interest for a selection of low and moderate risk studies.
Job creation in agriculture
Eight studies32,38,39,40,41,42,43,45 looked at job creation in agriculture as an outcome. Among those studies, three are quantitative studies32,43,45, two are qualitative studies41,42 and three are mixed-methods studies38,39,40.
In one quantitative study, deemed at low risk of bias (Table 3), 1,700 workers and 1,500 firms were followed over four years to compare the effects of offering workers vocational training and offering firms wage subsidies to train workers on-the-job (firm training) in Uganda32. The results showed that both interventions allowed participants to acquire sector-specific skills and firm-specific skills leading to higher employment rates post-training for each type of worker, but the effect was greater for vocational training compared with firm training (21% versus 14% post-training employment rate) and their total earnings rose by more compared with the firm-training intervention (34% versus 20%). The qualitative studies41,42, although not designed to assess the effectiveness of an intervention, highlighted a link between skills training and employment outcome. However, both studies were deemed at serious risk of bias. A mixed-methods study38 on youth programmes in Ghana showed that about 86.4% of young people still pursued maize farming a year after exiting the Youth in Agriculture Programme (YIAP). This public intervention was implemented to address youth unemployment in Ghana with the goal of getting young people to engage in the agricultural sector. The four main components of the programme were crops/block farm, livestock and poultry, fisheries/aquaculture, and agribusiness. The study focuses on evaluating the crops/block farm component. The crops cultivated under the YIAP include maize (seed and grain), sorghum, soybean, tomato and onion. This study is ranked at moderate risk of bias.
Self-employment in agriculture
Six studies36,39,41,45,46,47 indicated that skills training interventions resulted in self-employment in agriculture. Out of these studies, two studies are quantitative36,45, three are qualitative41,46,47 and one is a mixed-methods study39.
In one quantitative study36, self-employment was stimulated by a skills training radio campaign on growing orange-fleshed sweet potatoes in Ghana, Tanzania, Burkina Faso and Uganda. A survey of the local communities where the radio campaign was run found that households that reported hearing the educational radio campaign in Ghana, Tanzania, Burkina Faso and Uganda were 8.9, 2.3, 1.7 and 1.1 times more likely, respectively, to engage in growing orange-fleshed sweet potatoes, than households that did not. This study is deemed at moderate risk of bias.
Engagement/entrepreneurship in agriculture
Five studies34,38,39,41,42 showed that skills training interventions encourage youth engagement or entrepreneurship in agriculture. Among these studies, one is quantitative34, two are qualitative41,42 and two are mixed-methods studies38,39. In the quantitative study, a youth programme including agriculture content (training in livestock production, crop production and dairy farming) in South Africa indicated that youth engagement or self-employment in agriculture is eight times higher when agricultural programmes that specifically target the youth are implemented compared with when agricultural programmes are not available. This study is deemed at moderate risk of bias. Regarding the mixed-methods studies, one study38, deemed at moderate risk of bias with youth programme (YIAP in Ghana) as intervention, showed that after exiting the programme, 86.4% of beneficiaries were still involved in farming within a year. The qualitative studies were deemed at serious risk of bias.
Productivity of the farm/agriculture
Two studies35,41 found that skills training interventions lead to higher productivity of the farms. One of the studies is quantitative35 and the other is qualitative41. In the quantitative study, estimated to be at moderate risk of bias, the National Agricultural Extension and Research Liaison Services (NAERLS) rural youth extension programmes (RUYEP) helped 84.2% of beneficiaries achieve yields that exceed one tonne per hectare for maize in Nigeria, compared with 66% of non-participants35. The qualitative study41, outlined in Table 1, is deemed at serious risk of bias.
Profit/income earning of the farm
Ten studies32,33,35,38,40,41,42,43,44,47 looked at profit/income earning of the farm as an outcome. Among those studies, five are quantitative32,33,35,43,44, three are qualitative41,42,47 and two38,40 are mixed-methoda studies. In one of the quantitative studies, the Training for Rural Economic Empowerment (TREE) programme increased beneficiaries’ income by US$787 compared with non-beneficiaries over the 2011–2014 programme implementation period33. This study is deemed at low risk of bias. Another quantitative study44, deemed at moderate risk of bias, found that the continued adopters of beekeeping and mushroom growing had increased their family income by 49% and 24%, respectively. The three qualitative studies, not described here but outlined in Table 1, are deemed at serious risk of bias41,42,47. The mixed-methods study40 showed that the creation of a company that recycled livestock by-product (bone crafts and soap production) allowed vulnerable women and youths to earn an additional US$44.6 from bone crafts and US$50.2 from soap production weekly. This study is at moderate risk of bias.
Job search or employment opportunity
Three studies39,41,42 investigated the effect of skills training on this outcome. One study is a mixed-methods design39 and two41,42 are qualitative. All of these studies, not described here but outlined in Table 1, are deemed at serious risk of bias.
Provision of agricultural extension service
One study39 investigated on the effects of skills interventions on provision of agricultural extension service and found that the majority of graduates who benefited from student–farmer attachment and/or the Supervised Student Enterprise Project (SSEP) were engaged in extension work. This study, outlined in Table 1, is deemed at serious risk of bias.
Intervention type and engagement in agriculture
Two studies39,47 used agriculture-related courses as interventions. One of these studies is a mixed-methods study39 and the other is qualitative47. The mixed-methods study investigated several outcomes in agriculture, namely, job creation, entrepreneurship, self-employment, provision of agricultural extension service and job search opportunity, which were found to improve with the skills training interventions. The interventions consisted of introducing innovations in agricultural training curricula (community engagement and agri-enterprise development) at Gulu University in Uganda. The community engagement took the form of a one year (or less) placement of undergraduate students to work with smallholder farmers and farmer groups within a 10 km radius of the university. The agri-enterprise development consisted of having the students design business plans; the best plans were rewarded with start-up capital. The employment rate among the graduates was 84% six months after graduation and increased to 90% after one year; less than 2% of the graduates created their own businesses. The qualitative study47 investigated two outcomes in agriculture, self-employment and income, which were found to increase after skills training on ready food mixes, maize products and mango products. The two studies are deemed to be at serious risk of bias.
Four studies35,41,42,46 used technical education/training as interventions. Only one of these studies is quantitative35; the others are qualitative41,42,46. The quantitative study35 investigated productivity and income of the farm, and found both to increase after the intervention. The NAERLS RUYEP objectives are to provide technical advisory services to boost agricultural production and raise living standards of the youth. The results showed that the intervention allowed 84.2% of beneficiaries to achieve yields that exceed one tonne per hectare for maize in Nigeria, compared with 66% of non-participants. This study is deemed at moderate risk of bias. Among the qualitative studies, one46 looked at self-employment as an outcome and found a positive association with the intervention. The other two qualitative studies are deemed of serious risk of bias.
Youth programmes are programmes that target youth and train them in either specific skills (agricultural skills, ICT skills and so on) or broad skills (decision-making skills, business skills and so on) to enhance their employability. These have been used as interventions in three studies33,34,38. One of these studies is mixed methods38 and the two others are quantitative33,34. The mixed-methods study38 investigated the following outcomes in agriculture: job creation, engagement and income; a positive association was found between youth programme and both engagement and income. The results showed that about 86.4% of young people still pursued maize farming one year after exiting the programme and the mean income of GH¢758 obtained by beneficiaries was found to be greater than the national mean annual per capita income of GH¢734. Among the two quantitative studies33,34, one investigated the income of beneficiaries33 and the other34 looked at engagement in agriculture; both found a positive effect of the intervention on their outcome. The study that investigated the income of beneficiaries as an outcome revealed that the TREE programme increased beneficiaries’ income by US$787 compared with non-beneficiaries over the 2011–2014 programme implementation period33. In the other study34, a youth programme including agriculture content (training in livestock production, crop production and dairy farming) in South Africa indicated that youth engagement or self-employment in agriculture is eight times higher when agricultural programmes that specifically target the youth are implemented compared with when agricultural programmes are not available. Given that all three studies are at moderate or low risk of bias, we can conclude that the findings suggest that youth programmes have the potential to influence youth engagement in agriculture.
Only one study36 looked at on-the-job training as an intervention. The outcome investigated is self-employment, on which the intervention had a positive effect. The results showed that households that reported listening to an educational radio campaign in Ghana, Tanzania, Burkina Faso and Uganda were 8.9, 2.3, 1.7 and 1.1 times more likely, respectively, to engage in growing orange-fleshed sweet potatoes, than households that did not. The study was deemed at moderate risk of bias.
Vocational training has been used as an intervention by four studies37,40,44,45. Among these studies, three are quantitative37,44,45 and one is a mixed-methods study40. One quantitative study44 investigated income as an outcome, on which positive effects of the intervention were found in India. The findings indicated that vocational training programmes have resulted in continued adoption of beekeeping and mushroom cultivation enterprises by 20% and 51% of trained farmers, respectively, and increased their family income by 49% and 24%, respectively. The second quantitative study investigated job creation and self-employment as outcomes and found positive links with the training45. The results of the study highlighted that vocational training in agriculture in Iran resulted in employment of more than half of graduates. The third quantitative study found a positive effect of the intervention on job creation, the sole outcome it had investigated37. The study showed that vocational training for a youth employment programme in Ghana resulted in the creation of 16,383 jobs in agribusiness. All four studies are deemed at moderate risk of bias (Table 3); however, the use of descriptive methods in some of these studies preclude us from concluding that they are effective in improving employment outcomes for youth in the agricultural sector.
Vocational training and technical training
One study43 investigated the combination of vocational training and technical training as an intervention. The outcomes investigated are job creation and income, on which the intervention had a positive effect. The study indicated that vocational training and technical training in agriculture (poultry technician) resulted in an increase in employment of 34.2% among the 41 beneficiaries who were trained as poultry technicians in Nepal. This study is deemed at low risk of bias, suggesting that combining vocational training and technical training may be a way of improving job prospects and income for youth in the agricultural sector.
Vocational training and on-the-job training
One study32 investigated the combination of vocational training and on-the-job training as an intervention. The outcomes investigated are job creation and earnings, on which the intervention had a positive effect. The results showed that both interventions allowed participants to acquire sector-specific skills and firm-specific skills, leading to higher employment rates post-training for vocational-trained workers compared with firm-trained workers (21% versus 14% post-training employment rate) and their total earnings rose by more compared with the firm-trained workers (34% versus 20%). This study is deemed at low risk of bias.
Duration of training
Ten studies out of the 16 overviewed in Table 1 presented information on the duration of training. Eight of these have programmes that last one year or less. The remaining studies indicated a training duration between two and five years. This suggests that training programmes predominantly have a relatively short-term duration, which is consistent with many interventions taking the form of technical and vocational education/training. The popularity of technical and vocational/education training as a model of intervention may be due to the relatively short-term nature of the training, or due to the nature of technical and vocational training, which is well suited for out-of-school youth, which are found in large numbers in LMIC49.
Issues facing youth engagement in agriculture today are relatively well documented, including educational attainment, matrimonial status, gender, household size, parental income and occupation, membership in social organization, access to ICT, land tenure system and access to state-run agricultural youth programmes50,51,52. This present systematic review, which focused solely on interventions to engage youth in agriculture, yielded a limited set of studies—nine quantitative, four qualitative and three mixed-methods studies—so generalizable conclusions are difficult to draw. The risk of bias assessment yielded three studies32,33,43 deemed at low risk of bias, nine studies34,35,36,37,38,40,44,45,46 deemed at moderate risk of bias and four studies deemed at serious of risk bias39,41,42,47.
The results of our systematic review generally are in line with those found by the systematic review of Kluve et al.53 on interventions to improve the labour market outcomes of youth. That systematic review of 107 interventions, including skills training, in 31 countries, found small positive effects for promoting entrepreneurship and skills training—especially integrated skills training programmes—but not for employment services and subsidized employment.
Our systematic review also demonstrated that in general, skills interventions seeking to motivate youth’s engagement in agriculture do not undergo a thorough evaluation for effectiveness, with hard outcomes related to employment. Our selected studies included case studies and qualitative methods, which are not adequate methods of evaluating impact and effectiveness of interventions. Only one study used an RCT32. The two studies relying on a quasi-experimental approach used DID and PSM methods33,43. Indeed, the results of the risk of bias assessment indicated the studies relying on RCT and quasi-experimental impact evaluation methods were at low risk of bias. However, these study designs are expensive to conduct. We found that of the studies that evaluate interventions, the majority did not use state-of-the-art impact evaluation methods. This has been corroborated by other studies30,31, showing a chronic lack of evaluation of interventions that aim to provide agricultural skills to youth.
Training on ICT is an important aspect for attracting and retaining youth in the agricultural sector46. ICT offers a method of delivering training to a large number of farmers, which could enhance the performance of the youth already in agriculture and attract new youth to the sector36. Radio campaigns have been shown to be effective in spurring adoption and consumption of orange-fleshed potatoes in Ghana, Tanzania, Burkina Faso and Uganda36. A study conducted in the Philippines found that ICT training helps motivate secondary school students whose parents are engaged in agriculture to work within the sector, especially when combined with offline activities such as exposure and hands-on experience as well as creative and motivational actitivites46.
It is important to note that heterogeneity in gender and education are not accounted for in the analysis of the impacts of education on youth participation in agriculture. Our systematic review revealed that most of the included studies failed to address the effectiveness of targeting the population of interest—educated and uneducated youth. Illiteracy and gender heterogeneity were not addressed in the included studies. Indeed, no studies assessed the effects of training interventions on illiterate youth. This calls for investigations to focus on this vulnerable group of society, which represent about 25% of youth in sub-Saharan Africa and 11% in Southern Asia54. Failing to account for such variation in the background of the youth participants limits the ability to assess the effectiveness of skills training interventions.
The absence of robust research and lack of effective evaluation of the available data on the effectiveness of agricultural youth employment interventions has notable consequences on potential investment. Ultimately, the commitment of policymakers is necessary to ensure the sustainability and success of interventions to boost youth’s engagement in agriculture. It is encouraging that the majority of interventions (12 studies out of 16) studied originated from public policy, compared with three originating from non-public policy programmes (NGOs, international institution) and one from mixed policies (public and non-public policies). However, to provide a compelling basis on which to convince governments and donors to fund future interventions, as well as encourage young people to partake in training, cost-effectiveness analysis and estimates of returns on investment in training programmes is necessary. Indeed, a 2018 stocktaking of the evidence on the effectiveness of youth employment interventions in Africa found that for the agricultural sector in particular, “there is very little literature and virtually no evaluation evidence to inform policymakers about what types of interventions can improve the prospects of young people in the [agricultural] sector”31. Our study supports this conclusion. Moreover, to ensure that the skills training provides long-term opportunities for youth, it is crucial to establish a periodic follow-up to assess how trainees are performing after completion of a training programme. This aspect was missing in most of the interventions reviewed in this systematic review, yet it is important to check that the youth who engage in agriculture after receiving skills training are still involved and thrive in their agriculture-related business in the long term.
In summary, there is a need to foster youth skills training programmes and more importantly to evaluate more rigourously these programmes so that knowledge on good practices may be generated and transferred from one developing country to another. Estimates of returns to investment of agricultural skills training programmes are warranted as they could provide governments and donors with the evidence and cost-based analysis to continue and increase support for such programmes. Interventions also need to account for heterogeneity in gender and educational background of the youth to foster sustainability in agricultural value chains, inform inclusive policy design and ultimately contribute to reducing poverty and food insecurity in LMIC.
This systematic review was prepared following guidelines from Petticrew and Roberts55. The approach comprises five steps: identifying the research question; identifying relevant studies; study selection; extracting and charting the data; and collating, summarizing and reporting the results. The protocol for this study was registered on the Open Science Framework before study selection and can be accessed at https://osf.io/bhegq//. The guiding question for this systematic review was: What are the effects of skills training interventions on educated and non-educated youth employment outcomes in agricultural value chains, agribusiness or contract farming in LMIC? The inclusion and exclusion criteria to identify and then select the relevant studies are shown in Table 4.
Risk of bias assessment
Regarding the risk of bias assessment, each study was assessed following the criteria of the eight domains of risk of bias we considered. The maximum score a study can obtain in terms of minimizing all domains of risk of bias is 23 stars, which is 100% of the stars. A study is deemed to be at low risk of bias across all domains if its total score is in the interval 75–100%. If the total score is in the interval 50–75%, the study is said to be at moderate risk of bias across all domains. A study is at serious risk of bias if its score falls within the interval 25–50%. When the total score ranges from 0 to 25%, the study is deemed to be at critical risk of bias across all domains. See Supplementary Table 4 for details on the criteria used.
An exhaustive search strategy was developed and tested in CAB Abstracts to identify all available research pertaining to the effects of skills training interventions on educated and non-educated youth employment outcomes in agriculture in LMIC. Search terms were developed to address variations of the key concepts in the research question: skills training, youth, employment or engagement, and agriculture. Searches were performed on 9 May 2019 in the following electronic databases: CAB Abstracts (access via OVID); Web of Science Core Collection (access via Web of Science); EconLit (access via ProQuest); Agricola (access via OVID); and Scopus (access via Elsevier). Full search strategies for each database, including grey literature, can be accessed in their entirety at https://osf.io/xv56k/.
A comprehensive search of grey literature sources was also conducted. A list of the resources that were searched can be found at https://osf.io/xv56k/. The grey literature searches were performed using custom web-scraping scripts. The search strings were tested per website before initiating web-scraping. An existing Google Chrome extension was needed to scrape dynamically generated websites.
The results from the databases and the grey literature searches were combined and de-duplicated using a Python script. Duplicates were detected using title, abstract and same year of publication, where year of publication was a match, where title cosine similarity was greater than 85%, and where abstracts cosine similarity was greater than 80% or one of the abstracts (or both) was empty. When duplicates were found, the results from the databases and the grey literature searches were combined and duplicates were removed.
Following de-duplication, each citation was analysed using a machine-learning model. The model added more than 30 new metadata fields, such as identifying populations, geographies, interventions and outcomes of interest. This allowed for accelerated identification of potential articles for exclusion at the title/abstract screening stage.
Study selection and eligibility criteria
Systematic review software, Covidence, was used for both title/abstract and full-text screening decision-making with two independent reviewers evaluating each item. Citations were included in this study if they met all of the inclusion criteria noted above. Studies that did not meet all the inclusion criteria were excluded. Exclusion criteria were the inverse of the inclusion criteria. Each citation that met one of the exclusion criteria at the title, abstract or full-text screening phases were excluded. Studies included in the full-text screening stage were those that met all inclusion criteria and none of the exclusion criteria, or those whose eligibility could not be established during title/abstract screening. Reasons for exclusion were documented at the full-text screening phase.
The retrieval of hundreds of PDFs for full-text screening was done with a combination of automated and manual methods. For the automated method, a Python script was created that would handle the tasks of PDF discovery, download and file renaming using Google Scholar. The script read the bibliographic data from an Excel spreadsheet and then executed a script to retrieve the full-text PDF. If the article is spotted in the search results, the download link is clicked, and the article will be auto-renamed and marked as being downloaded. Manual methods were employed for those items that were not retrieved using the script.
A total of 245 records were identified for full-text screening. This screening process led to the identification of 16 studies that were considered adequate regarding the content and methodological rigour. The PRISMA flow diagram (Fig. 1) shows the steps followed during the screening process and the number of items that resulted after each step.
Data extraction was based on interventions and outcomes established in the research question and exclusion criteria. The data extraction focused on the outcomes of the studies, the methods used to obtain the outcomes, and the validity and reliability of those methods using a data-extraction form. To reduce risk of bias related to the extracted data, two separate researchers extracted data from each included study in the full-text review step. When disagreements occurred between researchers on data extracted from a study, a third researcher was engaged to resolve conflict by extracting data again from the study and the results were compared with those found previously. In total, 31 conflicts were solved among the 261 reviews. The critical appraisal of individual sources of evidence gave an indication of the strength of evidence provided and informed the standards followed for this systematic review.
Further information on research design is available in the Nature Research Reporting Summary linked to this article.
The data that support the findings of this study are available from the corresponding author upon reasonable request.
The code used in this study is available upon request.
World Employment and Social Outlook 2019: Trends for Youth (International Labour Office, 2019).
World Youth Report: Youth and the 2030 Agenda for Sustainable Development (United Nations, 2018); https://www.un.org/development/desa/youth/wp-content/uploads/sites/21/2018/12/WorldYouthReport-2030Agenda.pdf
Kilimani, N. Youth employment in developing economies: evidence on policies and interventions. IDS Bull. 48, 13–32 (2017).
Global Employment Trends for Youth: Special Issue on the Impact of the Global Economic Crisis on Youth (ILO, 2010).
Pieters, J. Youth Employment in LMIC (Institute of Labor Economics, 2013).
Yeboah, F. K. & Jayne, T. S. Africa’s evolving employment trends. J. Dev. Stud. 54, 803–832 (2018).
World Population Prospects 2019: Highlights ST/ESA/SER. A/423 (UN DESA, 2019).
Ellis, F. Rural Livelihoods and Diversity in LMIC (Oxford Univ. Press, 2000).
Pellegrini, L. & Tasciotti, L. Crop diversification, dietary diversity and agricultural income: empirical evidence from eight LMIC. Can. J. Dev. Stud. 35, 211–227 (2014).
Employment by sex and age—ILO modelled estimates. ILOSTAT Database (International Labour Organization, accessed 14 February 2020); https://ilostat.ilo.org/data
Carletto, G. et al. Rural income generating activities in LMIC: re-assessing the evidence. Electron. J. Agric. Dev. Econ. 4, 146–193 (2007).
Christiaensen, L., Demery, L. & Kuhl, J. The (evolving) role of agriculture in poverty reduction—an empirical perspective. J. Dev. Econ. 96, 239–254 (2011).
Reducing Poverty and Hunger: The Critical Role of Financing for Food, Agriculture and Rural Development (FAO, 2002).
Eicher, C. K. In Strategies for African Development (eds Berg, R. J. & Whitaker, J. S.) 242–275 (Univ. California Press, 1986).
de Alan, B. & MH, S. Linkages between poverty, food security and undernutrition: evidence from China and India. China Agric. Econ. Rev. 7, 655–667 (2015).
Sraboni, E., Malapit, H. J., Quisumbing, A. R. & Ahmed, A. U. Women’s empowerment in agriculture: what role for food security in Bangladesh? World Dev. 61, 11–52 (2014).
Sharaunga, S., Mudhara, M. & Bogale, A. Effects of ‘women empowerment’ on household food security in rural KwaZulu‐Natal province. Dev. Policy Rev. 34, 223–252 (2016).
Malapit, H. J. L. & Quisumbing, A. R. What dimensions of women’s empowerment in agriculture matter for nutrition in Ghana? Food Policy 52, 54–63 (2015).
Sumberg, J., Anyidoho, N. A., Leavy, J., te Lintelo, D. J. & Wellard, K. Introduction: the young people and agriculture ‘problem’ in Africa. IDS Bull. 43, 1–8 (2012).
Bezu, S. & Holden, S. Are rural youth in Ethiopia abandoning agriculture? World Dev. 64, 259–272 (2014).
Tadele, G. & Gella, A. A. Becoming a Young Farmer in Ethiopia: Processes and Challenges Working Paper 83 (Future Agricultures, 2014).
Lindsjö, K., Mulwafu, W., Andersson Djurfeldt, A. & Joshua, M. K. Generational dynamics of agricultural intensification in Malawi: challenges for the youth and elderly smallholder farmers. Int. J. Agric. Sustain. https://doi.org/10.1080/14735903.2020.1721237 (2020).
Salvago, M. R., Phiboon, K., Faysse, N. & Nguyen, T. P. L. Young people’s willingness to farm under present and improved conditions in Thailand. Outlook Agric. 48, 282–291 (2019).
Daum, T. Of bulls and bulbs: aspirations, opinions and perceptions of rural adolescents and youth in Zambia. Dev. Pract. 29, 882–897 (2018).
Yeboah, T. et al. Hard work and hazard: young people and agricultural commercialisation in Africa. J. Rural Stud. 76, 142–151 (2020).
Leavy, J. & Hossain, N. Who wants to farm? Youth aspirations, opportunities and rising food prices. IDS Working Papers https://doi.org/10.1111/j.2040-0209.2014.00439.x (2014).
Roser, M. & Ortiz-Ospina, E. Global education. Our World in Data https://ourworldindata.org/global-education (2016).
Gyimah-Brempong, K. & Kimenyi, M. S. Youth Policy and the Future of African Development (African Growth Initiative, 2013).
Foster, A. D. & Rosenzweig, M. R. Learning by doing and learning from others: human capital and technical change in agriculture. J. Political Econ. 103, 1176–1209 (1995).
Eichhorst, W. & Rinne, U. An Assessment of the Youth Employment Inventory and Implications for Germany’s Development Policy (Institute of Labor Economics, 2015).
Betcherman, G. & Khan, T. Jobs for Africa’s expanding youth cohort: a stocktaking of employment prospects and policy interventions. IZA J. Dev. Migr. 8, 13 (2018).
Alfonsi, L. et al. Tackling Youth Unemployment: Evidence from a Labour Market Experiment in Uganda (STICERD, LSE, 2017).
Lachaud, M. A., Bravo-Ureta, B. E., Fiala, N. & Gonzalez, S. P. The impact of agri-business skills training in Zimbabwe: an evaluation of the Training for Rural Economic Empowerment (TREE) programme. J. Dev. Effect. 10, 373–391 (2018).
Cheteni, P. Youth participation in agriculture in the Nkonkobe District Municipality, South Africa. J. Hum. Ecol. 55, 207–213 (2016).
Gambo Akpoko, J. & Kudi, T. M. Impact assessment of university-based rural youth agricultural extension out-reach program in selected villages of Kaduna-State, Nigeria. J. Apppl. Sci. 7, 3292–3296 (2007).
Hudson, H. E., Leclair, M., Pelletier, B. & Sullivan, B. Using radio and interactive ICTs to improve food security among smallholder farmers in sub-Saharan Africa. Telecomm. Policy 41, 670–684 (2017).
Ghana—Job Creation and Skills Development: Main Report (World Bank, 2009).
Baah, C. Assessment of the Youth in Agriculture Programme in Ejura-Sekyedumase District (Kwame Nkrumah Univ. Science and Technology, 2014).
Odongo, W., Kalule, S. W., Kule, E. K., Ndyomugyenyi, E. & Ongeng, D. Responsiveness of agricultural training curricula in African universities to labour market needs: the case of Gulu University in Uganda. African J. Rural Dev. 2, 67–76 (2017).
Kinyanjui, W. & Noor, M. S. From waste to employment opportunities and wealth creation: a case study of utilization of livestock by-products in Hargeisa, Somaliland. Food Nutr. Sci. 4, 1287–1292 (2013).
Latopa, A.-L. A. & Rashid, S. N. S. A. The impacts of integrated youth training farm as a capacity building center for youth agricultural empowerment in Kwara State, Nigeria. Mediterr. J. Soc. Sci. 6, 524–532 (2015).
Shoulders, C. W., Barrick, R. K. & Myers, B. E. An assessment of the impact of internship programs in the agricultural technical schools of Egypt as perceived by participant groups. J. Int. Agric. Ext. Educ. 18, 18–29 (2011).
Chakravarty, S., Lundberg, M., Nikolov, P. & Zenker, J. The Role of Training Programs for Youth Employment in Nepal: Impact Evaluation Report on the Employment Fund (World Bank, 2016).
Singh, K., Peshin, R. & Saini, S. K. Evaluation of the agricultural vocational training programmes conducted by the Krishi Vigyan Kendras (Farm Science Centres) in Indian Punjab. J. Agric. Rural Dev. Trop. Subtrop. 111, 65–77 (2010).
Khosravipour, B. & Soleimanpour, M. R. Comparison of students’ entrepreneurship spirit in agricultural scientific-applied higher education centers of Iran. Am. J. Agric. Environ. Sci. 12, 1012–1015 (2012).
Manalo, J. A. IV, Balmeo, K. P., Domingo, O. C. & Saludez, F. M. Young allies of agricultural extension: the infomediary campaign in Aurora, Philippines. Philipp. J. Crop Sci. 39, 30–40 (2014).
Channal, G. P., Kotikal, Y. K. & Pattar, P. S. Empowering rural women as a successful entrepreneur—through Krishi Vigyan Kendra. Agric. Update 12, 498–501 (2017).
Sterne, J. A. et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ 355, i4919 (2016).
How Does the Short-Term Training Program Contribute to Skills Development in Bangladesh? A Tracer Study of the Short-Term Training Graduates South Asia Region, Education Global Practice Discussion Paper (World Bank, 2015).
Nnadi, F. N. & Akwiwu, C. D. Determinants of youths’ participation in rural agriculture in Imo State, Nigeria. J. Appl. Sci. 8, 328–333 (2008).
Akpan, S. B., Patrick, I. V., James, S. U. & Agom, D. I. Determinants of decision and participation of rural youth in agricultural production: a case study of youth in southern region of Nigeria. Russ. J. Agric. Socio-Econ. Sci. 7, 35–48 (2015).
Adesugba, M. & Mavrotas, G. Youth Employment, Agricultural Transformation, and Rural Labour Dynamics in Nigeria (International Food Policy Research Institute, 2016).
Kluve, J. et al. Interventions to Improve the Labour Market Outcomes of Youth: A Systematic Review of Training, Entrepreneurship Promotion, Employment Services and Subsidized Employment Interventions (The Campbell Collaboration, 2017).
Literacy Rates Continues to Rise from One Generation to the Next (UNESCO Institute for Statistics, 2017).
Petticrew, M. & Roberts, H. Systematic Reviews in the Social Sciences: A Practical Guide (John Wiley & Sons, 2008).
We thank J.-A. Porciello and M. Eber-Rose for helpful comments on earlier drafts of this manuscript. We gratefully acknowledge funding support from Bundesministerium für Wirtschaftliche Zusammenarbeit und Entwicklung (Federal Ministry for Economic Cooperation and Development in Germany) and The Bill and Melinda Gates Foundation as part of Ceres2030: Sustainable Solutions to End Hunger, a project administered by Cornell University, USA.
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Tables 1–4.
Supplementary Material 1
PRISMA protocol for the systematic review.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Maïga, W.H.E., Porgo, M., Zahonogo, P. et al. A systematic review of employment outcomes from youth skills training programmes in agriculture in low- and middle-income countries. Nat Food 1, 605–619 (2020). https://doi.org/10.1038/s43016-020-00172-x
This article is cited by
Accelerating evidence-informed decision-making for the Sustainable Development Goals using machine learning
Nature Machine Intelligence (2020)