The combined factors of the low attractiveness of VET courses on offer, limited participation, students dropping out, and the resulting low number of graduates suggest that the public VET system is still ill-adjusted to the needs of both learners and the labour market. This in turn calls for wider system reforms with a view to bringing it in line with the needs of its key clients: individuals, employers and society at large.

There is no structured mechanism to systematically gather demographic, social, labour market and skills data at national, sectoral and regional levels and inform adjustments to the network of (publicly funded) VET providers and programmes. In addition, there is no comprehensive VET management information system in place that would translate demand into elements such as enrolments, numbers and skills of personnel, needs for equipment and materials, accommodation places, and monitoring of the state of the system. Information, counselling and guidance activities that would help young people determine their future education and professional careers are also largely absent.

VET reform could be helped forward if there was clear leadership that was well coordinated and resourced. However, the governance of initial and continuing VET at national level in Tunisia is complex. As mentioned before, the most important institution is now the MJSIP, which develops, implements and evaluates government policy in VET. The ministry's executive agency, the ATFP, manages the largest number of training centres covering most sectors. The National Centre for the Training of Trainers and Training Engineering (CENAFFIF – Centre National de Formation de Formateurs et d'Ingénierie de Formation) takes care of teacher training and curriculum development, while the National Centre for Continuing Training and Professional Promotion (CNFCPP – Centre National de Formation Continue et de Promotion Professionnelle) coordinates continuing training. The ANETI in turn has responsibility for the training of unemployed jobseekers as one among other active labour market measures.

In addition, the AFMT and the AVFA – as agencies under the respective sector ministries, as well as the Ministry of Health and the Tunisian Army – run a number of training centres in their fields. Finally, there is a large number of private initial and continuing training providers. These include non-accredited training centres, where provision goes unchecked for quality.

The MJSIP is responsible for the overall coordination of the various public institutions and supervising VET. A permanent committee coordinates all activities with the administrations concerned and the social partners. Finally, a Superior Council with an advisory role brings together once or twice a year the ministries in charge of education and training with the social partners. However, it is not clear to what extent these coordination structures are really effective.

The organisation of management structures for VET (partly) by sector is a remnant of a VET governance model that aimed at linking training centres closely to the industries providing specific types of training. Also, initial and continuing training stand apart as two different systems with different providers, funding models, qualifications, teaching staff and teaching materials. Now, this model and the multitude of actors involved appears inefficient, especially considering the desire for overall VET system development and reform in line with changed labour market and social demand. Initial and continuing VET in the different sectors would need to work towards new, commonly agreed national frameworks of policy and legislation, funding arrangements, qualifications, quality and quality assurance, and transparency and accountability.

Hence, adjusting overall governance arrangements could be a major step to help system reform forward. The new arrangements need to streamline or strengthen the mechanisms for:

• overall system planning, funding, management, development, monitoring and evaluation under a common national policy, jurisdiction, frameworks and guidelines; and
• more structured involvement of the social partners in respective strategic decisions and operational aspects.

Change, or VET reform for that matter, cannot be decreed from the top. Its basis – or the number of change actors who deliver reform at local level – needs to be considerably broadened, including in Tunisia.

The OECD (2015a) and the ILO (2018), among others, have argued that past policies in Tunisia have contributed to regional polarisation and rural poverty. This suggests that a new regional development policy was implemented, which included some level of decentralisation. Tunisia's constitution from 2014 does include 'decentralisation' as the main concept (in Chapter 7 on Local Powers), which sets the context for such a fundamental change in governance. As part of Tunisia's regional development strategy, the authorities have planned for a greater devolution of responsibilities, i.e. to elected governments at the regional and local (municipal) level (which presents a 'deconcentration' of powers, rather than a decentralisation).

A decentralised setting for managing VET and employment policies is not a new topic in Tunisia. It has been on the agenda for over 20 years, since the 1996 MANFORME programme and its focus on demand-oriented vocational training. The MFPE and donors have been making more serious efforts since 2011, although not consistently, to support or modify governance arrangements. The National VET Reform 2014–2018 established 'good governance of the system, consistent with the Social Pact, the national economy and decentralisation' as one of its main directions (MFPE, 2013). Suggested reform measures included taking regional specificities into consideration and partnerships being institutionalised. Furthermore, decentralised governance was a topic of the aforementioned PEFESE project implemented from 2013 to 2016, as well as of the EU GEMM project and the ETF between 2012 and 2018 (see ETF, 2016, 2017). However, as mentioned above, many of the planned reform projects did not go further ahead. Assigning bigger roles to regional and local actors with a view to adjusting to local needs still remains a major issue.

There are many factors that explain why the decentralisation agenda has not yet gained full momentum. Political tensions and frequent changes of ministerial actors may have hampered the pursuit of a longer-term vision and more fundamental changes. Also, there is still a certain reluctance to give up decision-making powers and control over resources within a traditionally centralised system. The ATFP, for example, 'wish to maintain their competence at central level for selecting the courses offered by training centres' (GOPA, 2019). However, it is also true that entities at the local level may lack the capacity and motivation to take over new tasks and responsibilities.

Local governments today are weak: in 2015 local budgets in Tunisia represented only 3.6% of the national government budget, compared to 10% in Morocco and 20% in Turkey (World Bank, 2015). Between 70% and 90% of the municipal budget goes to wages, thus covering operational costs rather than investments (OECD, 2018).

On 16 March 2019, the MFPE outlined a plan for reorganising VET provision in the regions. The plan lists the following objectives:

• align all actors behind the methodologies developed by MFPE and their agencies, such as for collecting data, monitoring and evaluation;
• allow for more flexible governance arrangements that are supported by reliable information systems;
• strengthen the capacity of the Regional Directorates for Vocational Training and Employment who would be in charge of VET provision in the regions; and
• assign increasing levels of autonomy to regional VET structures (GOPA, 2019).

The ATFP has also been pursuing the idea to organise regional governance through five macro-regions. This raises the question whether these macro-regions would have the potential to become self-governing and self-sufficient units in terms of their economic and social development and to manage training and employment within their regions, combining national and local funding. Discussions are still needed as to whether the governance of education and training should be tied more closely to existing territorial structures.

The EU-funded IRADA and PAFIP projects (co-funded with Agence Française de Développement (AFD) – the French Development Agency) are providing substantial resources to create momentum and try out new decentralised approaches, involving regional actors in initiatives aimed at economic, human resource and social development.

IRADA (budget: EUR 32 million, project duration until 2022) partly builds on the smaller-scale EU/ETF GEMM project, which involved four Tunisian pilot regions and was carried out between 2012 and 2018. Partnerships in these governorates provided inputs to draft Regional Master Plans for VET. They designed and agreed on a number of VET development projects, which were then evaluated at national level.

IRADA aims to build the capacity of regional actors, involving also the private sector in local skills and economic development. IRADA does so through projects designed by regional actors themselves and which respond to the specificities and needs in eight governorates: Gabès, Médenine, Gafsa, Kasserine, Sfax, Sidi Bouzid, Bizerte and Jendouba. These regions are considered to have potential for implementing their own action plans and promoting investments and innovation through decentralised, participatory approaches. Under the IRADA project, regional VET system assessments were carried out in the eight regions in 2019, following the ETF Torino Process model. IRADA provides assistance to 4 out of the 14 projects of the VET Reform Strategy for 2014–2018.

PAFIP concentrates on four job pools in the north and east of the country and on the industrial sector. It aims at a territorial anchoring and a qualitative leap of VET system reforms, based on partnerships and combining education and training with labour market integration measures. The 17 training centres involved are encouraged to network with local and regional partners and become regional outlets for skilled people and real hubs for economic and social development.

In some regions like Sfax or Sousse, things are progressing well. But initial initiatives have also shown that regional stakeholders in Tunisia are not always fully prepared to embrace the new approaches.

What is still lacking in other regions is:

• experience to work in partnerships and pursue common goals;
• systematic data collection and analyses to build robust knowledge of both broader and regionally specific issues related to economic, labour market and social conditions, as well as priorities for their development;
• the capacity and resources to assess the potential for certain sectors or trades to grow, including the ability of certain industries to export products abroad (which requires knowledge of, among others, international markets and quality standards);
• the ability to reason along economic or regional clusters and value chains, which bring production or harvesting activities together with processing, preservation, marketing, sales and transport, and related coordination, ICT and technical skills; and
• the competence to design relevant projects in partnership with, for example, businesses, professional groups, civil society organisations and researchers.

Hence, capacity-building of stakeholders at national, regional and local levels, including social partner representatives and employers, is key to any decentralisation agenda. It is already an integral part of donor projects. Moreover, this needs to be accompanied by a clear definition of the mandate and roles of the different actors underpinning a certain autonomy of the regions, as well as the allocation of appropriate resources.

In August 2019, a new government decree was adopted that assigns more autonomy to training providers, as mentioned in Section 2.5.2. However, related provisions have not yet been implemented. According to an ETF survey (2019),[31] A survey carried out by the ETF in Tunisia in 2017 sampled all teachers and directors in 179 training centres. In total, 117 training centres and 386 trainers responded, which represents response rates of 65% and 9%, respectively.
decision-making was considered participatory in most VET centres. Training centre boards have some influence over the curriculum, recruitment and the internal use of the budget. However, generally, decisions about recruitment, dismissal or suspension of trainers and trainers' salaries, including setting pay scales and salary increases, are mostly taken at the national level. By contrast, principals and boards in private training centres have greater authority, including over salaries and recruitment. Only 42% of training centre boards include business representatives. This is also due to the fact that employers have a consultative role only. Overall, 90% of the principals reported the 'inadequacy of the training centre budget and resources', along with 'government regulation and policy' as major constraints limiting their effectiveness (ETF, 2019, p. 3). The ATFP has been working on a new approach classifying VET centres according to A, B and C levels and assigning different degrees of autonomy. However, for now, this presents rather an administrative approach, defining how much money is allocated to VET centres, based on student numbers and other parameters. The ETF, on the other hand, has been working with the MFPE, various agencies and a few VET centres to develop a concept for what could be labelled an 'entrepreneurial centre'.

Funding for VET comes mainly from the state budget and is either transferred directly to training centres or implemented through two funds: (i) the Vocational Training and Apprenticeship Promotion Fund and (ii) the Employment Promotion Fund. The first finances continuing training and, in part, training centres. Funds are collected through a special tax on companies who do not invest in other forms of training (1% to 2% of their payroll). In essence, this fund is underutilised. The second fund is financed by the state budget and secures funding for ALMPs. The state subsidises part of the cost for (public or private) training through the 'vocational education and training cheque' (Chèque Formation Professionnelle) for specific sectors considered a national priority. This is a relevant tool that is, however, not widely used. Reasons may include rigidities in the use of the fund, complex procedures, a lack of proper communication about it, and the non-availability of relevant training offers in certain regions.

Although sufficient funding for VET appears to be available in Tunisia, there is an issue here again of a fragmented, bureaucratic and partly inefficient system. A comprehensive analysis would be useful to address VET funding and financing in a more holistic way. Legal and funding tools could take a more comprehensive approach and embrace a wider spectrum of initial or continuing training measures for different target groups, involving non-state actors in the fund management and making economies of scale.

Training centres also have their own revenues from registration fees, continuing training activities and the sale of training products. However, this represents only a small fraction of their budgets. Some argue that VET centres could generate higher levels of own revenues, but the incentives and mechanisms to do so seem to be either absent or insufficiently used. The cost of training in private centres is borne by the beneficiaries themselves.

A skills development system that is largely not in line with the aspirations of individuals, employers and society at large requires new communication mechanisms between the system and its principal clients to be put in place and that the basics are re-adjusted in terms of the overall governance, financing and quality assurance of the system.

Demographic, social, labour market and skills data would need to be gathered more systematically at national, sectoral and regional levels and inform a restructuring of the network of VET providers and programmes. Public funds for training could be used more efficiently by preparing learners for their integration in society and the labour market. In the absence of a sophisticated – and, admittedly, costly – system of labour market intelligence at national and regional levels, informed discussions with economic thank tanks, experts and employers on the future needs of the economy at national, sectoral and regional levels may be helpful. The ultimate aim would be to devise, and implement over time, a mid- to long-term plan for the gradual adjustment of the network of (publicly funded) training providers and courses to better suit identified needs. Regularly collected information on the satisfaction levels and career paths of VET graduates could also serve as useful inputs.

Furthermore, more (online) opportunities for career information, as well as counselling and guidance in and outside schools, would help young people determine their future education and professional careers. Related digitalisation efforts of employment offices are a step in the right direction and could be a useful source of information.

It is recognised that the currently highly centralised governance and financing of VET are among the main factors that hamper more relevant VET provision in Tunisia. High youth unemployment and huge differences between the regions in Tunisia call for new arrangements in the managing of education, training and employment policies that incorporate the concepts of 'decentralisation' or 'regionalisation'. A comprehensive and coherent approach is needed to foster a multi-stakeholder governance and financing system at national, sectoral and regional/local levels that has the potential to answer socio-economic and related skills needs more efficiently.

With the Social Pact, an overall tripartite structure to advise on VET policies is in place. However, there is an issue of operationalising the provisions of the Social Pact and creating all kinds of public–private partnerships. Social partner organisations at national and sectoral levels should be able to participate in the dialogue on skills development policies on an equal footing with government. Bipartite sector skills councils, bringing together VET curriculum experts and employers, could be set up for priority sectors to analyse skills and training needs and develop respective qualifications and curricula. It is clear that all partners to the various forms of dialogue and expert work would need to become familiar with their new roles and enhance their capacity.

The roles and responsibilities of the MJSIP, ATFP and the social partners for overall strategy design, system planning, management, monitoring and evaluation need to be strengthened. The MJSIP would have to facilitate overall reform and increase the relevance and quality of skills development systems in Tunisia. Furthermore, there is a need for the MJSIP, ATFP and their partners to design and implement common national frameworks for qualifications, quality assurance, transparency and accountability, under which sector-specific training agencies, such as the AFMT and AVFA as well as all regional and local actors, can act autonomously.

Active involvement of regional and local stakeholders in the design and implementation of VET and employment policies is key to unlocking the economic, employment and social development potential of the regions. That is to say that, also at regional and local levels, roles and responsibilities of the different actors, including the Regional Directorates for VET and Employment, local ANETI offices and the social partners, have to be redefined. Adequate resources, as well as approaches and tools, are necessary to strengthen the capacity of actors in the regions and allow for an efficient and effective management of the new tasks, as well as the monitoring and evaluation of policies (see e.g. ETF and MFPE, 2017).

However, one of the major preconditions for a functioning decentralised system for managing VET and employment policies is their embeddedness in an overall, well-prepared administrative reform of the country. Such a reform would need to reduce today's weak and inefficient administrative structures composed of 6 regions, 24 governorates, 264 'delegations' or districts, and numerous municipalities and regroup them into bigger, potentially self-sufficient, self-governing units. It would need to assign not only more responsibilities, but also more resources to regional and local administrations. It requires a national plan; new legislation that stipulates how functions shall be divided between the state, governorates, districts and municipalities in a future decentralised system; which resources shall be transferred from the state; and which types of local taxes regional or local authorities would be able to collect. Without a proper financial base, any decentralisation will just remain a piece of paper. In this respect, Tunisia can learn from other countries, such as the Central European countries that joined the EU in 2004 or more recently also Albania. The latter adopted a new legal framework and are changing their administrative structures to strengthen local governance.

In the meantime, the ATFP and other VET agencies can go ahead and pilot new approaches for decentralised management of VET with the help of donor projects. The ATFP and the other VET agencies could deliberately promote the formation of networks involving all relevant stakeholders to identify skills needs and create partnerships for employment and skills at regional/local levels. The regions are encouraged to do their own planning, based on prior identification of training and skills needs in the given region or governorate or a specific sector (e.g. tourism in Médenine).

Innovative approaches that aim to improve the development and use of human capital in the regions and are already implemented by donor projects include:

• the formation of a sustainable partnership, or 'public–private dialogue platform', which involves relevant stakeholders in each governorate;
• structured analysis and dialogue on the development needs and potential of the private sector in each region;
• the elaboration of action plans for the development of relevant sectors and areas;
• the financing, with the help of donors, of regional, integrated development projects (adapted from GOPA, 2019); and
• adjustments to the design and delivery of VET courses and curricula in accordance with regional needs and with the central level, and with an emphasis on involving employers (see also: ETF, 2017).

Communities, employers' and employees' associations, and individual employers are key players in all aspects mentioned above. Recognising the importance of to promote local economic and social progress, they are encouraged to be involved in economic, labour market and skills needs analyses, as well as the design, delivery and monitoring of training policies. Initial training for young people plays an equally important part as second-chance education, skills upgrading or retraining measures for adults. In this context, an earlier report for the Education For All Global Monitoring Report 2012 (Walther, 2011) highlights the significance of building skills required in the informal sector, which is widespread in Tunisia.

Building on the 2019 decree, there is a need to strengthen the autonomy of VET providers, which in turn are managed by competent school boards and collaborate with relevant actors in the region. Further efforts could be taken to turn some training centres into centres of excellence, with additional functions of national or sectoral relevance and possibly funded jointly with the private sector.

More autonomy of regional actors and VET providers will have to be accompanied by functioning national frameworks that ensure quality and accountability, and an efficient use of public funding. Such frameworks typically include quality-assured qualifications and/or programmes (frame curricula) that have been approved by national or economic sector bodies and to which VET provision would be adjusted. Various other quality assurance mechanisms, including self-assessments and inspections of VET providers, would have to be put in place to set the quality standards for VET provision and encourage VET providers to continuously develop and improve. This could include linking VET funding partially to student retainment and skills outcomes.

In addition, to allow national-level monitoring and evaluation of the system, VET providers and regional actors would need to report regularly against a nationally agreed set of indicators, for which a number of management tools and guidelines would be needed. This would include, among others, a comprehensive VET management information system that could help the regional and national levels to monitor processes and outcomes of autonomous providers, and design and plan VET policies overall. Respective data would need to be translated into, for example, enrolments, the number and qualifications of personnel, the purchase of equipment and materials, and the number of accommodation places.

While the MJSIP and their partners have the leadership for revising VET governance and financing arrangements, the ownership of suggested reforms clearly increases when the actors concerned are consulted from the outset. A major dialogue process involving a multitude of actors will have to be instigated that requires a great deal of strategic oversight and planning, technical, legal and communication skills, and last but not least human and financial resources inside the MJSIP, ATFP and among their dialogue partners. Any changes to governance and financing arrangements will need to be reflected in changes to legislation. It is recommended that Tunisian actors take a more comprehensive approach to overhauling the legal framework for initial and continuing VET, including base laws and secondary legislation. Enshrined in the new legislation should be a funding model that underpins the new governance arrangements and promotes education and training in a lifelong learning perspective.

An undertaking of this size is by no means a quick fix. Even if all actors agreed and made use of appropriate management tools and capacity, the new arrangements would require years to prepare and take root. They necessitate support across the political spectrum, a clear roadmap, financial and technical support, and close monitoring in a constant cycle of designing, implementing, evaluating and redesigning policy.

Strengthening the autonomy of training centres, within agreed national frameworks and with support from partners, could potentially lead to increasing the relevance and quality of VET provision. Under a revised legal framework for VET that incorporates governance and financing within a decentralised setting and within national frameworks for qualifications, quality assurance and accountability (reporting), authorities, agencies, regional directorates, vocational training centres and their local partners could practise new forms of cooperation and authority. Even within the existing legal framework, Tunisia could take further steps to operationalise VET school autonomy, delegating more responsibilities to management, training centre boards and their partners, and finding pragmatic ways to incentivise their entrepreneurial spirit. Local actors could be entrusted to participate in key decisions relating to the choice of their training offers; the recruitment and (extra) remuneration of staff; the management of all affairs of the training centre, including marketing its own courses, enrolments, the development of curricula integrating WBL and of teaching and learning resources; career guidance activities; the signing of agreements with companies; the signing of other contracts (such as for renting out premises or producing products or delivering services for the local community); and the generation and use of budget. Social partners, students and family representatives can make useful contributions to the management and development of training centres.

The role of principals, training centre management and teachers, alongside their economic partners, needs to be strengthened in relation to the development of the curriculum and teaching. Teacher training centres could support and coordinate teachers to develop up-to-date instructional materials in a collaborative manner. There is also an issue of nurturing evaluation practices at the level of training centres.

By diversifying the sources of funding (state, region/municipality, employer, centre's own resources), training centres may be put in a position to acquire up-to-date equipment and consumables. Furthermore, training centres need to be equipped with appropriate hardware, software and internet, and teachers need to be trained on the use of ICT in lesson design and in classroom learning. This will help teachers to also develop more differentiated learning tasks and to use formative assessment to inform teaching and learning. Teachers and pedagogical experts could collaborate to develop and share instructional materials and textbooks.

The MJSIP, different institutions and stakeholders should work together to agree on strategies for the initial and continuing professional development of teachers and trainers, to improve the normative framework, to set objectives for CPD for vocational teachers in the 2025 National Strategy for Vocational Education, and to define a feasible action plan. CENAFFIF, perhaps through collaboration with other professional CPD providers, may review the methodology of their CPD programmes and, where appropriate, introduce modern and varied methodologies. Tools need to be developed to identify needs and funding mechanisms for the technical training of teachers (or the training of trainers). There is much to be gained from CENAFFIF working in partnership with industry and other partners to provide specialised professional CPD designed to update professional knowledge and skills and work practices.

As part of the overall VET governance reform, there is an issue of enabling regional directorates to activate their decision-making roles. Better coordination between regional structures and training centres and their partners is key to providing the required materials and human resources. Training centres have an important role in better analysing their training requirements and need to share and use this information for the planning and coordination of CPD. The MFPE is also asked to consider or revise the status, compensation, training and recognition offered to teachers with mentoring roles at training centres. Mentorship might be used more generally as a form of CPD for VET trainers and principals. CPD opportunities should be linked to improvements in teaching, which are identified through feedback, and to career development. Where possible, CPD should be designed to prepare trainers and principals to take on enhanced responsibilities and to gain recognition through promotion. National aggregated records of CPD participation could be published (ETF, 2019).

Evidence suggests that, where WBL is made a structural element of training provision, a smoother transition from school to work is achieved, as opposed to studying first and then working (ETF, 2013, among others). Sectors where WBL is deemed to have some potential for development in Tunisia include aeronautics, automotive components, computers, textile and clothing, tourism and crafts (Chelbi, 2017). On the other hand, agriculture and the BTP, for which WBL presents a major problem, would require specific national support measures.

Learning under the F0 apprenticeship scheme needs to be reviewed in order to complement company training with training at a centre. Improvements to the functioning of non-diploma-based learning should include additional training at private training centres. This means that a company who hires an apprentice could take them to a private centre for additional learning. Private centres do have the capacity to take care of complementary theoretical and soft skills training. Private training centres could therefore create close ties with companies, thus helping to promote the culture of learning and their courses. This would have a definite impact on the development of this apprenticeship scheme, which has been losing momentum in Tunisia for some years. The success of such an approach requires a flexible funding mechanism that ensures the activities of private operators.

In the future, training cheques could be granted only when a company expresses the need and willingness to take care of the beneficiary during the training.

Partnership committees of training centres need to be made functional again. The new decree on VET autonomy of March 2019 addresses this to some extent, although it may not go far enough in terms of granting autonomy of VET providers, as argued above. The participation of private sector professionals in the management of training centres is essential for bringing the worlds of education and work closer together and enabling WBL.

Status, reduced teaching loads and the reimbursement of travel expenses are central to trainers and apprenticeship counsellors at training centres fulfilling their assignments in relation to WBL. This includes the careful planning of all company training periods, ideally as an integral part of the VET curriculum, the selection of students together with the companies, and regular liaison with companies to ensure, for example, the follow-up of students and joint assessments of progress. All this is very time-consuming and needs to be adequately accounted for in teachers' working hours.

Effort has been made and needs to continue to ensure trained tutors are available who can instruct and supervise trainees in companies. With a view to implementing workplace training support units, professional and trade union organisations could be entitled to certain funding, which would allow them to play their part in supervising members and be a credible and representative intermediary with the MJSIP. This is a very important aspect of the reform that got blocked. The training support units could be remunerated for the continuing training activities that they organise.