Technical and Vocational Education and Training (TVET)

A definitional review. Citations follow IEEE style; see References.

Abstract

Technical and Vocational Education and Training (TVET) is the international standard term for education and training oriented to occupations and livelihoods. It is also one of the least stable terms in education policy: several regional synonyms name overlapping but non-identical things, and the statistical apparatus does not separate the formal systems that are studied from the informal provision through which most people in low- and middle-income countries actually acquire skills. This article sets out the standard definition, the terminological field, the formal/non-formal/ informal distinction, the documented record of attempts to transplant the German-speaking dual system, the status problem and its consequences for evaluation, and — at greatest length — what the rigorous impact literature reports. That literature is markedly less encouraging than agency framing, and the gap between the two is itself the most important fact about the field.

I. Definition

The standard definition is the one adopted by the UNESCO General Conference at its 38th session on 13 November 2015, in the Recommendation concerning Technical and Vocational Education and Training (TVET). It defines TVET as comprising

education, training and skills development relating to a wide range of occupational fields, production, services and livelihoods [1].

The Recommendation places TVET within lifelong learning: it “can take place at secondary, post-secondary and tertiary levels and includes work-based learning and continuing training and professional development which may lead to qualifications,” alongside “a wide range of skills development opportunities attuned to national and local contexts.” Learning to learn, literacy and numeracy, transversal skills and citizenship skills are named as integral components, not adjuncts [1].

Two features of this definition matter. It is deliberately expansive — livelihoods and services sit beside industrial trades, and skills development sits beside qualification-bearing programmes. And it is not the only definition in circulation: the UNESCO Institute for Statistics glossary gives a narrower reading, describing programmes “designed mainly to provide learners with practical skills, know-how, and understanding necessary for direct entry into a particular occupation” [2]. The broad instrument and the narrow measurement standard do not describe the same object.

II. The terminological field

TVET competes with a set of regional near-synonyms recorded in UNESCO-UNEVOC’s own glossary [2]:

Term	Where used
Vocational Education and Training (VET)	European Union
Career and Technical Education (CTE)	United States
Technical and Further Education (TAFE)	Australia
Further Education and Training (FET)	South Africa
Technical and Vocational Skills Development (TVSD)	Commonwealth of Learning
Vocational and Technical Education (VTE) · Professional Technical Education and Training (PTET) · Vocational and Professional Education and Training (VPET)	general use

Australia’s national VET research body treats TVET simply as the equivalent term for VET, “especially in the Asia-Pacific region” [2]. That equivalence is asserted, not demonstrated, and the terms carry different institutional freight: skills development typically includes short non-award courses that vocational education excludes, while career and technical education is a school-track term with no adult-training component in ordinary American usage.

The consequence is not merely stylistic. It impedes comparison across the evidence base. The main international data on participation come from countries’ reporting against SDG indicator 4.3.3, and those data “do not distinguish between participation in formal or non-formal education”; enrolment data by level carry many missing values, partly through under-reporting and partly because countries place the entry point to TVET at different levels; responsibility for post-secondary non-tertiary provision is often split across ministries; and “different data sources are often inconsistent” [3]. Participation in work-based learning is still not collected in a standardised way, which the joint World Bank/UNESCO/ILO report notes “hinders cross-country analysis” of formal apprenticeship [3].

III. Formal, non-formal and informal

The distinction the literature relies on is between formal provision (institutional, qualification-bearing, state-recognised), non-formal provision (organised but outside the qualification framework) and informal learning (workplace acquisition without curriculum or certification, of which traditional apprenticeship is the dominant form).

The joint report states the position for low- and middle-income countries plainly:

“In L/MICs, formal TVET is an important but often only a small part of a TVET system in which informal and nonformal programs predominate, such as informal apprenticeships; in contrast, most TVET in HICs is formal” [3].

The scale figures make the same point. The average share of 15–24-year-olds enrolled in vocational education — formal or non-formal — is 1.1 per cent in low-income countries, 4 per cent in lower-middle-income countries and 7.6 per cent in upper-middle-income countries, against 11.9 per cent in high-income countries [3]. About two-thirds of workers in low- and middle-income countries are employed informally, and among newly created jobs in Sub-Saharan Africa the majority are informal [3]. Public spending follows the same gradient: under 0.2 per cent of GDP on TVET in low- and middle-income countries, against 0.46 per cent in high-income countries [3].

This is why the report’s own title restricts its scope to formal systems, and why that restriction should be read as a limitation rather than a definition. The part of the system that most people pass through is the part that is least measured and least evaluated.

The surrounding labour-market picture, on the ILO’s 2023 figures, is that the global youth unemployment rate was 13 per cent — a fifteen-year low, with 64.9 million unemployed young people — while 20.4 per cent of youth were not in employment, education or training (NEET), two in three of them women (28.1 per cent of young women against 13.1 per cent of young men). In low-income countries only one in five young adults aged 25–29 holds a secure paid job, against 76 per cent in high-income countries, and two in three young adult workers in developing economies hold qualifications that do not match their job well [4].

IV. The dual system and the record of transfer

The German-speaking dual system — apprenticeship combining employment in a firm with attendance at a vocational school, governed jointly by the state and the social partners — is the reference model in international TVET policy. Germany is “by far the largest donor country for promoting VET through international cooperation” and has provided more official development assistance for VET than either the World Bank or the European Union [5]. Transfer projects have run in Afghanistan, China, East Timor, Indonesia, Laos, Mongolia, Myanmar, Pakistan, Sri Lanka and Viet Nam, as well as in Saudi Arabia, Mexico, Türkiye, North Africa and Latin America; there are few in Sub-Saharan Africa or the least-developed countries [5].

The record of those transfers is, on the German-language research literature’s own account, poor:

“Past discussions on the nature and competencies of German VET cooperation have already shown that the impossibility of transferring a specific model of VET must be assumed” [5].

Oeben and Klumpp summarise the evaluative evidence: external experts evaluating GIZ projects drew a critical “transfer balance sheet” and found the projects examined “were only of little sustainability” in their aim of introducing dual or cooperative training structures; further studies report that VET transfer projects “often remain as stand-alone solutions within the country”; and the practice has been criticised as policy-borrowing [5]. Expectations have accordingly retreated: where the earlier assumption was transfer of the dual system entire (in adapted form), “the transfer of system elements and successful features into the existing educational system is now more modestly assumed,” with the dual system treated as a reference model rather than a template [5].

The identified reason is structural rather than technical. Among the success conditions catalogued in that review, the decisive one is the training readiness of firms — “the lack of this structural component often leads to the failure of transfer attempts,” and securing it requires that apprenticeship be of net benefit to the training company [5]. Where labour markets are contested and turnover high, firms in emerging economies were found to adopt only the core of the model (practical training on modern equipment) and to shorten programmes [5]. The dual system is not a curriculum that can be shipped; it is an equilibrium among employers, schools and social partners, and the equilibrium is what does not travel.

V. Status, selection, and what the evidence shows

A. TVET as a second-choice track

“TVET is seen as a choice inferior to general education in many countries, based on either real or perceived differences in educational quality, opportunities for continuing education upon graduation, and employment prospects” [3]. The structural basis for that perception is partly real: more than 20 low- and middle-income countries, including seven in Sub-Saharan Africa and six in the Middle East and North Africa, operate secondary TVET systems from which no programme gives access to tertiary education [3].

B. Selection effects, and why they confound evaluation

Track choice is not random. Analysis of household survey data for Ghana and Kenya finds that socioeconomic characteristics — parental education, parental involvement, cognitive score — partly determine selection into secondary TVET rather than general secondary [3]. The joint report describes TVET as “the de facto track for the relatively disadvantaged” and identifies the feedback loop: “the sorting of academically weaker and more disadvantaged students into TVET also feeds into negative perceptions, creating a vicious circle” [3].

This is an evaluation problem before it is a policy problem. Most estimates of TVET’s returns come from observational studies — regressions of labour-market outcomes on educational attainment — which the report states plainly “do not lend [themselves] to causal inference,” and which in low- and middle-income countries carry a further “serious limitation”: they “tend to focus on wage employment and often fail to capture the self-employment and unpaid family work that are common in many of these countries” [3]. Rigorous designs exist but are scarce, because they require either randomisation or a natural discontinuity, and because the capacity to run them is limited [3].

C. The systematic reviews: effects are small

The Campbell Collaboration review of TVET interventions for young people aged 15–24 in low- and middle-income countries covered 26 studies across Latin America, the Caribbean, Europe, East Asia, South Asia and Sub-Saharan Africa, ten of which entered the meta-analysis; searches ran to September 2012 and it was published in September 2013 [6]. Its summary finding is that “TVET interventions have a small but positive effect on all but one of the employment outcomes measured” [6]. The outcome on which no positive effect was found is paid employment; that attribution rests on secondary summaries rather than on the review’s full text, which could not be obtained for this article.

Two of the review’s own caveats are more consequential than the headline. “Lower quality studies find a significantly larger effect,” with the result that “the meta-analysed effect size is inflated” when they are included; the authors recommend basing conclusions on medium-quality studies at minimum [6]. And because the observed effects are modest, their policy advice is that “opting for the cheapest and/or most culturally acceptable models may be the best approach” pending more rigorous evaluation [6] — an argument for restraint, not for expansion.

The larger 3ie systematic review of youth employment interventions synthesised 113 impact evaluations covering 107 interventions in 31 countries, yielding 2,259 standardised effect sizes [7]. Its results are frequently quoted as showing that skills training and entrepreneurship promotion are the “most promising” intervention types. They are — relative to employment services and subsidised employment, the other two categories reviewed. The absolute magnitudes are small:

Outcome	Overall	Skills training	Entrepreneurship promotion
Employment (SMD)	0.04 (95% CI 0.03–0.06), 105 interventions	0.05 (0.02–0.07), 67	0.16 (0.06–0.26), 7
Earnings (SMD)	0.05 (95% CI 0.03–0.06), 92 interventions	0.07 (0.05–0.08), 60	0.09 (0.01–0.18), 12

Effects were larger in low- and middle-income countries than in high-income ones (skills training: employment SMD 0.06, earnings SMD 0.12) and larger in programmes targeting disadvantaged youth [7]. But heterogeneity was high throughout (I² of 64 per cent for employment, 82 per cent for earnings, 86 per cent for skills training on earnings), the review found “statistical evidence for small study effects for all outcomes, suggesting the presence of publication bias in the literature,” and it qualified its own conclusion: skills training and entrepreneurship promotion “appear to be a particularly promising intervention … but the evidence base is still relatively small” [7].

D. Effects fade

The best-known demonstration of fade-out comes from a randomised evaluation in Côte d’Ivoire covering 4,160 young people in 16 localities, in which a seven-month public works programme paying about US$10 a day was combined, for randomly assigned subgroups, with 100 hours of entrepreneurship training or 80 hours of job-search training [8]. At four to five months the contemporaneous effect was very large — 97 per cent of participants were in wage employment against 49 per cent of the control group, with monthly earnings up around FCFA 27,000 (45 per cent) and savings up FCFA 39,800 (182 per cent). At 12–15 months after the programme ended, “no impacts were observed on the likelihood of employment, the hours worked, or the composition of employment,” with only “very limited sustained positive impacts on participants’ earnings”; roughly a quarter of the savings effect persisted [8]. The intervention was a public works programme with training components rather than a TVET programme, which limits how far the result transfers — but the pattern it exhibits, large in-programme effects followed by convergence, recurs.

It recurs at system level too. Evidence from high- and upper-middle-income countries shows that young TVET graduates have better employment outcomes than general-education graduates, “but as groups age, this advantage eventually disappears,” and similar patterns have been identified in low- and middle-income countries [3]. The literature attributes the deterioration to weaker foundational skills and to occupation-specific technical skills becoming obsolete faster than transversal ones, leaving graduates less able to reskill [3].

E. Heterogeneity swamps the average

The joint report’s tabulation of the impact-evaluation and observational literature is titled, without euphemism, “The Evidence on Average Impacts of TVET on Employment and Earnings is, at Best, Mixed”: among rigorous impact evaluations, three found positive employment effects and four found none; for earnings, four positive and six null or mixed. Among observational studies, six found TVET better than general education, five similar or mixed, and four worse [3].

Individual results span the range. Randomised allocation to oversubscribed vocational schools in Mongolia raised the likelihood of paid employment one year after graduation by 4 per cent and raised women’s monthly earnings by 13 per cent; a regression-discontinuity evaluation in Nepal found non-farm employment up 10 percentage points. Against these, randomised scholarships in El Salvador produced higher enrolment and completion but no effect on employment or earnings; scholarships in Namibia raised completion but had little effect on employment or earnings; and an evaluation of large-scale vocational programmes in Türkiye found no average employment impact three years after training [3]. Observationally, upper-secondary vocational schooling in Türkiye returns 4.2 per cent per year against 2.4 per cent for general secondary, and technical secondary completion in Brazil carries a 9.7 per cent wage premium — while in Egypt the returns to formal vocational secondary education were found to be “virtually the same as having no formal education at all” [3]. Within a single country, returns by field ranged from −15.2 per cent to +13.5 per cent [3].

The report’s own reading is that “arguably the most accurate way of describing the labor market impacts of TVET is that they are very heterogeneous,” and that TVET “likely pays off for some students — perhaps even more than general education — while being a bad investment for others” [3].

F. A worked national example

Sri Lanka is a documented case of the reporting pattern. An Asian Development Bank sector assessment records that employment outcomes for TVET graduates average about 57 per cent, up from 50 per cent in 2012, with enrolment capacity rising from about 148,000 in 2012 to about 189,000 by 2016 [9]. The same paragraph supplies the qualifications usually dropped when the figure is quoted: outcomes remain “uneven across different agencies,” ranging from 73 per cent at one institute to about 48 per cent at others — which the assessment reads as evidence of a “lack of relevance of skills training to actual labor market demand” — and the source for the employment figures is an unpublished tracer study [9]. A headline national improvement, a spread across providers wider than the improvement, and a non-public source: this is the characteristic shape of TVET monitoring data.

VI. Limitations

The evidence base does not cover most of what TVET is. Rigorous evidence is “still very limited, particularly in [low-income countries],” and on apprenticeship specifically “most, if not all, of the evidence comes from [middle-income countries]” [3]. Almost all of it concerns formal provision, while informal and non-formal programmes predominate in the countries where the policy interest is greatest [3]. The largest part of the system is the least studied part.

The measurement instruments do not separate the categories. SDG indicator 4.3.3 does not distinguish formal from non-formal participation, work-based learning data are not standardised, and sources disagree [3]. Claims comparing national TVET systems should be treated as approximate for this reason alone.

Quality and publication bias both push estimates upward. The Campbell review found that lower-quality studies report significantly larger effects and that the pooled estimate is inflated by their inclusion [6]; the 3ie review found small-study effects indicating publication bias across all outcomes [7]. The true effects are therefore likely to sit at or below the small magnitudes reported.

Selection is not fully controlled in most of the literature. Observational returns estimates cannot support causal claims, and they systematically miss self-employment and unpaid family work — the labour market forms that dominate in the settings under study [3].

Averages may not be the right object. Where returns within one country range from −15 per cent to +13 per cent by field of study [3], a pooled effect describes no actual programme.

Programme design is frequently disconnected from demand. Fieldwork in Liberia and Sierra Leone found training curricula inherited from post-conflict disarmament programmes still forming “blueprints for current trainings offered, when market demand is either no longer there or is saturated,” with needs assessment functioning as “a perfunctory box-ticking exercise” rather than a source of design guidance [10].

Finally, the rhetoric is older than the disappointment and has survived it. TVET has held a prominent place in development and peacebuilding discourse — the Sustainable Development Goals call for universal access to quality TVET — even though the programmes “have been criticised already since the 1960s as programme outcomes did frequently not match expectations,” on grounds that they are often not adjusted to the local labour market, provide insufficient training, and ignore the voices of the young people they enrol [11]. The measured effects of TVET on employment and earnings in low- and middle-income countries are small, highly variable, frequently null, and prone to fade; nothing in the reviewed literature supports the stronger claims routinely made on the sector’s behalf.

References

[1] UNESCO, “Recommendation concerning Technical and Vocational Education and Training (TVET),” adopted by the General Conference at its 38th session, Paris, France, 13 Nov. 2015.

[2] UNESCO-UNEVOC International Centre for TVET, “Technical and Vocational Education and Training,” TVETipedia Glossary. Bonn, Germany: UNESCO-UNEVOC. (Entries compiled from UNESCO 2015, UNESCO-UIS 2014, NCVER Australia, Commonwealth of Learning and others.)

[3] World Bank, UNESCO and International Labour Organization, Building Better Formal TVET Systems: Principles and Practice in Low- and Middle-Income Countries. Washington, DC, USA / Paris, France / Geneva, Switzerland: World Bank, UNESCO and ILO, 2023. ISBN 978-92-2-039357-4 (paper), 978-92-2-039358-1 (electronic).

[4] International Labour Organization, Global Employment Trends for Youth 2024: Decent Work, Brighter Futures. Geneva, Switzerland: International Labour Office, 2024. ISBN 978-92-2-040674-8 (print), 978-92-2-040675-5 (web PDF), doi: 10.54394/FGPM3913.

[5] M. Oeben and M. Klumpp, “Transfer of the German Vocational Education and Training System — Success Factors and Hindrances with the Example of Tunisia,” Education Sciences, vol. 11, no. 5, art. 247, 2021, doi: 10.3390/educsci11050247.

[6] J. S. Tripney, J. G. Hombrados, M. Newman, K. Hovish, C. Brown, K. T. Steinka-Fry, and E. Wilkey, “Technical and Vocational Education and Training (TVET) Interventions to Improve the Employability and Employment of Young People in Low- and Middle-Income Countries: A Systematic Review,” Campbell Systematic Reviews, vol. 9, 2013, doi: 10.4073/csr.2013.9. ⚠️ Full text not obtained — the publisher’s copy returned HTTP 402 in this research pass. All statements attributed to [6] are taken from the Campbell Collaboration’s own published summary of the review; the author list and citation details are cross-checked against the reference list of [5].

[7] J. Kluve, S. Puerto, D. Robalino, J. M. Romero, F. Rother, J. Stöterau, F. Weidenkaff, and M. Witte, “Interventions to Improve the Labour Market Outcomes of Youth: A Systematic Review of Training, Entrepreneurship Promotion, Employment Services and Subsidized Employment Interventions,” 3ie Systematic Review 37. London, U.K.: International Initiative for Impact Evaluation (3ie), Jun. 2017.

[8] Abdul Latif Jameel Poverty Action Lab (J-PAL), “Increasing Employment and Earnings through a Skills Development Program for Youth in Côte d’Ivoire,” evaluation summary, researchers M. Bertrand, B. Crépon, A. Marguerie and P. Premand. ⚠️ The underlying working paper (“Contemporaneous and Post-Program Impacts of a Public Works Program: Evidence from Côte d’Ivoire”) was not read; all figures are as reported in the J-PAL summary.

[9] Asian Development Bank, “Summary Sector Assessment: Education (Technical and Vocational Education and Training),” linked document to Additional Financing of Skills Sector Enhancement Program (RRP SRI 42251-019). Manila, Philippines: ADB. (Document reports data to May 2017; the graduate employment figures derive from an unpublished tracer study cited therein.)

[10] S. Batmanglich and L. Enria, Real Jobs in Fragile Contexts: Reframing Youth Employment Programming in Liberia and Sierra Leone. London, U.K.: International Alert, Jul. 2014. ⚠️ Relevant sections read; not read in full.

[11] R. van der Veen and S. Datzberger, “The Peacebuilding Potential of Technical and Vocational Education and Training Programmes in Post-Conflict Sierra Leone,” author accepted manuscript, UCL Institute of Education. ⚠️ Journal of record, volume and pagination not verified; opening sections read.