Certified but still unemployed.

Why Africa's youth digital skills programmes should include a minimum of six months of real work in a company before they count their job as done.

A young Nigerian finishes a three-month course in data analysis through the federal government's 3 Million Technical Talent programme, known as 3MTT. She passes the final assessment, adds the certificate to her LinkedIn profile and starts applying for jobs. Most applications get no reply. When she does get an interview, the recruiter asks to see a dashboard she has built for a real business. She has only the projects she completed during training.

A 2026 article in TheRadar, a Nigerian news site, described this experience as common among 3MTT graduates. UI/UX graduates who know how to use Figma are asked for case studies showing research, client feedback and changes made over time. The article's conclusion was blunt: "a certificate and a job offer are not the same product."

The same gap appears in digital skills programmes across Africa, whether they are run by governments, donors or NGOs. Young people are trained and certified, then left to find work on their own. This article argues that training should not be where support ends. Every trainee in a donor-funded or government-funded digital skills programme should complete a minimum of six months of structured, supervised work in a real company, preferably paid, before the programme considers its job done.

Training at scale

Governments and donors are training young Africans in digital skills in very large numbers. In February 2026, Nigeria's Minister of Communications, Innovation and Digital Economy, Bosun Tijani, said 3MTT had trained more than 135,000 people across three cohorts, with learning resources reaching more than 300,000 others. The programme's long-term target is three million.

Ghana launched its One Million Coders Programme in April 2025. The pilot received about 94,000 applications in 48 hours, far more than planners expected. By August 2025, 859 participants had completed training. The government has set a target of 400,000 trainees for 2026, with training centres in all 16 regions.

Donors have long favoured training too. Of all youth employment interventions listed in the International Labour Organization's youth employment inventory, 82 per cent are training programmes.

None of these figures says how many of the people trained are now working.

What programmes count

Training programmes are usually judged by what is easy to count. Enrolment, completion and certificates can be reported within weeks of a course ending. Employment a year later costs more to measure, so it is often left out.

Research on donor-funded skills programmes in Sierra Leone, published in Third World Quarterly in 2023, found this pattern. Jamelia Harris of the University of Warwick studied technical and vocational training projects in the country and found that they were largely shaped by donor strategies rather than local job markets. They promised employment, but mostly led to informal self-employment, and they focused heavily on outputs rather than outcomes. She concluded that such programmes were unlikely to build skills that contribute meaningfully to national development.

The World Bank has found a similar problem in its own work. When its Independent Evaluation Group reviewed the Bank's youth employment programmes, it found that the few impact evaluations the Bank had carried out looked only at short-term effects, found limited positive results and did not calculate cost-effectiveness. One of its main recommendations was that the Bank should monitor employment and earnings outcomes for young people in its programmes.

Africa's newest national programmes show signs of the same issue. In 2024, the Nigerian technology publication ITEdgeNews reported that 3MTT lacked a sustainable framework for placing trained people into jobs. TechCabal reported that after the three-month programme, graduates kept access to a learning dashboard, a job board and a local community, but finding work was otherwise left to them. In May 2026, the programme opened a Talent Registry listing more than 3,000 tech job opportunities, described as its most significant attempt to address criticism that training had been delivered without structured pathways to work. The minister's figure of 15,000 "job and opportunity pathways" is worth noting as well, since a pathway is not the same as a confirmed job.

In Ghana, TechLabari reported in March 2026 that One Million Coders had so far lacked a clear roadmap for implementation and a clear framework for measuring real impact. Communications Minister Samuel Nartey George has said publicly that what matters is what happens after people are trained, and how they find work, particularly remote jobs. The programme has not yet published employment figures for its graduates.

Why training alone falls short

Research on youth training programmes helps explain why. In 2019, a team of economists led by Jochen Kluve reviewed 113 impact evaluations of youth employment programmes around the world, published in World Development. Only about a third of the results showed a positive and significant effect, and the average effect was small. The programmes that succeeded tended to combine several services with profiling of participants and individual follow-up, rather than offering training on its own.

David McKenzie, an economist at the World Bank, reached a similar conclusion in a 2017 review of labour market programmes in developing countries. He argued that many of these programmes, including vocational training, are much less effective than policymakers assume, and that many evaluations find no significant impact on employment or earnings.

One reason is that employers find it hard to judge young applicants who have never held a job. Two experiments in South Africa show how much credible information about a candidate matters. In the first, researchers working with the Department of Labour found that attaching a reference letter from a former employer to a job application increased callbacks from employers by 60 per cent. At the start of the study, fewer than 5 per cent of the young jobseekers used a reference letter at all. In the second, researchers working with the Harambee Youth Employment Accelerator gave young jobseekers certificates showing their assessed skills. Their employment rate rose by 17 per cent and their weekly earnings by 34 per cent compared with a control group.

In both studies, employers responded to credible evidence that a young person could do the work. Training projects offer some evidence, but they are done in a classroom, under a trainer's guidance, without real users, deadlines or consequences. A manager who has watched someone work for months can say far more about them than a trainer can.

Training followed by real work

Several African programmes have tested what happens when training is followed by time inside a real company.

The clearest example comes from Kenya. Between 2011 and 2016, the Kenya Youth Empowerment Project, run by the government and the Kenya Private Sector Alliance with World Bank support, gave young people aged 15 to 29 in Nairobi, Mombasa and Kisumu three months of classroom technical training followed by three months of internships in private firms. A randomised evaluation found that the programme increased current employment among male participants by 15 per cent. Monthly wages rose by about KSh 5,000 for men and KSh 7,500 for women, and the programme's benefits exceeded its costs. Kenya then scaled the approach up through the Kenya Youth Employment and Opportunities Project. According to World Bank tracer studies, 77 per cent of young people who received training and an internship through that programme had found a job or were self-employed as of June 2023.

In Uganda, researchers working with the NGO BRAC ran a large experiment on youth employment, published in Econometrica in 2020. They followed 1,700 young workers and 1,500 firms over four years, in a setting where youth unemployment in their sample was above 60 per cent. Some young people received six months of vocational training at training institutes. Others were placed in small firms for six months of training on the job. Both groups gained skills and significantly improved their employment rates over the following three years. The study covered trades such as welding and hairdressing rather than IT, and it shows that good classroom training can work too. It also shows that six months inside a firm was enough to raise employment for at least three years afterwards.

South Africa offers an example at larger scale. The Youth Employment Service, a private sector initiative launched with the Presidency in 2018, places young people in 12-month jobs with businesses. The government reports that YES has created more than 228,000 youth jobs and that 60 per cent of participants were absorbed into full-time employment afterwards. These figures come from the government and the programme rather than an independent evaluation. The programme was set up for the reason discussed in this article: President Cyril Ramaphosa wrote that many young South Africans leave school, college or university with qualifications, but employers want people with experience.

Generation Kenya, a nonprofit bootcamp programme, works differently. It designs its short courses with employers and confirms job vacancies before training begins. In 2020, it reported that 83 per cent of its more than 18,000 graduates had been placed in jobs within 90 days of finishing. Again, these are the organisation's own figures.

Why a minimum of six months

Based on this evidence, I recommend that every youth digital skills programme include a minimum of six months of structured, supervised work in a real company after training.

No study has directly compared internships of different lengths to find the ideal one. But the programmes that have worked point in this direction. Six months inside real firms produced lasting employment gains in Uganda. Kenya's three-month internships raised employment, though the gains were modest. South Africa's 12-month YES placements report high rates of permanent employment. Some large employers already offer placements of this length. Safaricom's graduate internship in Kenya, for example, runs for three to six months with a monthly stipend of KES 25,000 to 40,000.

There are also practical reasons specific to IT work. In most tech companies, a new intern's first weeks go to setting up accounts and tools, learning the codebase or systems, understanding the team's processes and shadowing colleagues. In a three-month placement, that leaves little time for real production work before the internship ends.

Six months is long enough for an intern to take part in at least one full project cycle: building a feature or report, testing it, releasing it to real users and fixing the problems that follow. That experience is what employers are asking for when they want to see a dashboard built for a real business. It is also long enough for a supervisor to write a detailed reference, and the South African evidence shows employers respond to references far more than to unverified claims of skill.

Making internships work

Internships can go wrong, and a six-month requirement would need safeguards.

The first risk is exploitation. The ILO has warned that internships can be used as a source of cheap labour or to replace existing workers. Gianni Rosas, who coordinated the ILO's youth employment programme, said internships should always include a training component, and that using interns to do work normally done by regular staff can amount to disguised employment. A useful internship has a named supervisor, clear learning goals, real tasks and regular feedback.

The second risk is that only young people who can afford to work unpaid will benefit. Six months without income is not realistic for most young people from poor households. Internships should come with a stipend that at least covers transport and food.

Cost is another issue. Hosting and supervising an intern takes staff time, and many companies, especially small ones, will not do it for free. The Yemen study, a World Bank experiment, paid firms a 50 per cent wage subsidy to take on recent graduates as interns, and interns' incomes rose by 73 per cent as a result. South African companies join YES partly because it improves their rating under the country's Black economic empowerment rules. Donors and governments could fund stipends or wage subsidies directly, even if that means training fewer people in total. Funding fewer trainees who reach real jobs may be better value than certifying many who remain unemployed.

Demand is also a limit. Internships only lead to jobs where companies are able to hire. The World Bank's own evaluation found that in high-unemployment settings, training and job search support alone have little impact, and that action to create jobs is also needed. The Yemen study found firms had more science and technology graduates than they wanted, and too few in business and marketing. Programmes should train for skills that local and remote employers are actually hiring for, and sign up host companies before recruiting trainees.

What donors and governments should change

A few changes would put this into practice.

• →Report employment rates at six and twelve months after training, alongside the number of people trained. Funders should ask for these figures and treat them as the main measure of success.
• →Recruit host companies before training starts, so every trainee knows where they will be placed. This also gives employers a say in what is taught.
• →Make internship stipends and wage subsidies part of the core budget, not an optional extra added if money is left over.
• →Give every graduate a written reference from their host company, as well as their training certificate.

Back to the young graduate

Return to the young Nigerian with her data analysis certificate. Under the model proposed here, she would not finish training and then hear nothing back from employers. She would spend at least six months in a company, building dashboards that managers actually use, fixing them when they break and learning how a real team works. When she applied for her next job, she could point to that work and hand over a reference from her supervisor. There is also a good chance the host company would hire her itself.

Nigeria, Ghana and other African countries are already spending heavily on training young people for the digital economy. Adding six months of real work to that training would make it far more likely that the investment ends in jobs.

Sources

This article draws on the following reporting and research.

• →TheRadar, "3MTT has trained 135,000 Nigerians, but why are tech jobs still hard to get?" (August 2026); TechCabal, "Can 3MTT turn Nigeria into a global tech talent powerhouse?" (October 2024); ITEdgeNews on 3MTT (May 2024); Kurrentech International on the 3MTT programme (July 2026).
• →CitiNewsroom and GBC Ghana on One Million Coders targets (January 2026); TechLabari (March 2026).
• →Harris, J. (2023), "Exploring donor-driven skills development as a channel of continued aid dependency," Third World Quarterly.
• →World Bank Independent Evaluation Group, "Youth Employment Programs: An Evaluation of World Bank and IFC Support."
• →Kluve, J. et al. (2019), "Do youth employment programs improve labor market outcomes? A quantitative review," World Development.
• →McKenzie, D. (2017), "How Effective Are Active Labor Market Policies in Developing Countries?" World Bank Research Observer.
• →Abel, Burger and Piraino (2020), "The Value of Reference Letters: Experimental Evidence from South Africa"; Carranza, Garlick, Orkin and Rankin (2022), American Economic Review.
• →Honorati, M. (2015), Kenya KYEP internship study; World Bank KYEOP Implementation Status Report (June 2023).
• →Alfonsi, L. et al. (2020), "Tackling Youth Unemployment: Evidence From a Labor Market Experiment in Uganda," Econometrica.
• →SAnews (June 2026) and the South African Government blog on YES; Generation Kenya graduation announcement (March 2020).
• →McKenzie, Assaf and Cusolito (2016), youth internships experiment in Yemen; ILO, "Internships: Head start or labour trap?" and ILO Employment Working Paper No. 240 (2018); Nucamp on the Safaricom internship (2026).