96819 SKILLS AND JOBS Lessons Learned and Options for Collaboration May 2015 © 2015 International Bank for Reconstruction and Development/The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved This note was prepared by Maria Laura Sanchez Puerta, David Robalino, Victoria Strokova, Nick Lord and Mathilde Perinet (Jobs CCSA, World Bank). The authors would like to thank Nigel Twose, Alexandria Valerio, and Omar Arias (World Bank), as well as Ingrid Hamm, Uta Micaela Duerig and Olaf Hahn (Bosch Stiftung) for their comments and inputs. 01 The accumulation of human capital through the The situation is further complicated by the acquisition of knowledge and skills is recog- diversity of skills that seem to matter. Indeed, nized as central for economic development. recent empirical analyses show that success More-educated workers not only have better in the labor market does not only depend on employment opportunities, they earn more and the acquisition of job-relevant or occupation- have more stable and rewarding jobs. They are specific skills. Cognitive and socio-emotional also more adaptable and mobile. Workers who skills, in part acquired in early childhood and acquire more skills make other workers and during basic and secondary schooling, are also capital more productive and, within the firm, important determinants of employment dynam- they facilitate the adaptation, adoption, and ics and earnings later in life. ultimately invention of new technologies. This is crucial for economic diversification, productivity A pattern that seems to emerge is that as econ- growth, and ultimately raising the living stan- omies develop and diversify the demand for dards of living of the population. higher-level cognitive skills increases relative to the demands for manual job-specific skills. In the developing world, however, the majority Yet, the few studies looking at the distribution of of the labor force has very low levels of educa- cognitive and non-cognitive skills in the labor tion. Even those with higher degrees might not force in developing countries suggest that cer- have acquired the skills necessary to succeed in tain vulnerable groups, such as the unemployed the labor market—particularly high-level analyt- and first time job seekers tend to have low cog- ical and interactive skills. Middle to high-income nitive scores or lack the socio-emotional skills countries have the best indicators and yet the that predict success as a self-employed worker average number of years of schooling is only or entrepreneur. 8.5—primary plus some years of secondary edu- cation. At the same time, employers frequently This note discusses the types of interventions complain that workers’ lack of skills is a con- that can be considered to improve the skills of straint to their business. those who are already in the labor force; either working or looking for jobs. Clearly, any strat- This mismatch between the skills individuals egy aimed at finding solutions to the skills prob- have and those which are needed appears per- lem needs to start with creating conditions for vasive across countries. In India, for instance, adequate early childhood development, laying 50% of university graduates obtain a diploma a strong foundation in basic and secondary edu- in arts, far exceeding employer demands. In cation, and improving tertiary education. These Tunisia, more than 50% of university graduates policies, however are outside the scope of are in jobs that do not use the skills they the note. acquired in university. In Cambodia, employers also complain about skill mismatches, especially The structure of the note is as follows. First, it among out of school youth. More recent evi- examines the different types of market failures, dence from Lebanon shows that 40% of wage and subsequently reviews the role that govern- earners and the self-employed are in occupa- ments have played in training systems around tions that do not use their skills. the world. Finally it offers a set of proposals for reforming and improving these systems to improve labor market outcomes. 02 WHAT IS WRONG WITH remuneration for those investments (since there are many employers demanding these skills). THE DEMAND AND THE Governments would simply need to make sure PROVISION OF TR AINING? that individuals had the means to finance these investments. An important question is why the private sec- tor—both workers and employers—is not But perfectly competitive labor markets are able to address their training/skills needs and more likely to be the exception than the rule for solve the problems discussed above. In many many reasons. These include barriers to entry cases, private sector arrangements without that limit the number of firms competing for government interventions or support have given skills; workers not having sufficient infor- worked well. But in general, there are a number mation about where the jobs are and/or not of problems that affect workers and employers being able to move to where they are; or and that can lead to an underinvestment in employers not having sufficient information training and in the misalignment between sup- about the skills that workers have. In labor mar- ply and demand for skills. These usually involved kets that are not perfectly competitive both imperfections in labor and capital markets, firms and employees can underinvest in training. coordination failures, as well as decision making Firms underinvest because they cannot fully problems.1 appropriate the returns to investment in train- ing, since employees can be poached by other Imperfections in Labor Markets firms; employees underinvest because they In perfectly competitive labor markets there would not receive fair remuneration for their would not be a need for government interven- higher productivity. The higher the bargaining tion. Employees would have incentives to invest power of employers the lower the investment in in their training and acquire skills that were training by employees. in demand, knowing they would receive fair 03 Imperfections in Capital Markets enough wages, and workers do not invest in Financial institutions usually have little infor­ training because of a low skill premia. mation about the impacts of training on produc- tivity, earnings, and the creditworthiness of Decision Making Failures individuals and employers. They may therefore Even with perfect labor and capital markets and be reluctant to finance worker training, particu- no coordination failures, individuals might not larly for low-income individuals with little or no make the right decisions when it comes to credit history. This may lead to underinvestment investments in training. The main reasons are in training. The implication is that governments lack of information (knowledge), limited cogni- may need to address these credit constraints tive capacity to process complex problems, and through different instruments such as guaran- psychological factors that make it difficult to tees, student loans, and better information for commit to a given decision. For example, quite financial institutions. often individuals have inaccurate information on returns to training that may lead to under- or Coordination Failures over-investment. Individuals also often lack Often, the success of a given investment depends reliable information on the quality of training on the coordinated actions of various agents. At providers and future employment prospects, the high end of the skills spectrum, innovation and may invest in the wrong training. Even with externalities may occur when workers do not the necessary information, investments in train- invest enough in training because there are not ing and career choices are complex problems enough jobs for highly qualified professionals. and individuals may not always make the right But the reason there are not more of these jobs choice. Finally, high discount rates or the lack is that there are not enough qualified profes- of non-cognitive skills that predict success in sionals in the first place. Economies may thus school (such as discipline and perseverance) can end up in a low-level equilibrium where firms lead many individuals to procrastinate and fail to do not innovate enough and do not pay high get enough education. Almeida, Rita; Behrman, Jere; Robalino, David. 2012. The Right Skills for the Job? Rethinking Training Policies for   1 Workers. Washington, DC: World Bank 04 WHAT HAVE classified into two groups. A first group of firms does not invest in training simply because their GOVERNMENTS DONE? expected rate of return is smaller than the return on other investments. A second group of Individuals who are leaving formal, general firms may have profitable expected returns to schooling or who are already in the labor mar- investments in OJT but cannot realize these due ket can have access to a variety of training to market failures: liquidity constraints; high programs. Although these have been designed worker turnover—in part due to poaching; and/ for different target groups, they can be broadly or lack of adequate information about the bene- categorized into two groups: fits of training. In-house, firm-based, or on the job training   (OJT); Governments have put in place policies to try to increase the training provided by firms: Institutions-based training programs such as   training funds, tax incentives and grants, pay- pre-employment technical and vocational back clauses and apprenticeships, as well as trainings (TVET) and training-related active technical assistance. It is important to note that labor market programs (ALMP). although several programs have been intro- duced and modified to promote OJT across FIRMS-BA SED/ON-THE-JOB many countries, few evaluations of those pro- TR AINING grams are available. A substantial amount of a person’s lifetime human capital is accumulated after individuals Tax incentives and grants leave formal schooling and while on the job. In The most popular instrument in developing the United States, for instance, more than half countries for providing subsidies is training of the lifetime human capital is accumulated funds. Although most of these funds are based through post-schooling investments that take on payroll levies, in practice, they may take vari- place on the job. The evidence reviewed shows ous forms. that there is a strong and positive correlation Under a levy-grant mechanism all firms pay •  between the incidence of job training at the taxes and those that train, receive grants. worker or firm level and higher productivity. At The Mauritius’s Enterprise Training Fund, for the worker level, several country studies for the instance, has some features of this mechanism. developing world suggest wage returns to work- ers may be as high as 20 percent per training In cost-reimbursement schemes all firms pay •  episode. At the firm level, the incidence of job taxes and finance their training but then can training is linked to higher firm productivity, a be reimbursed for incurred expenses (this can more schooled workforce, and more frequent impose a high administrative burden on the technology adoption. training fund). In practice, however, the reim- bursement is often set below the levy paid to Nevertheless, the incidence of OJT in the devel- cover administration costs. For example, in the oping world is still quite low, with only a small Nigerian Industrial Training Fund, firms qualify share of formal firms engaging in training. The for not more than 60 percent of the levy paid, important question then is why some firms do and in practice less than 15 percent of firms not train. Research shows that firms can be apply for reimbursement for training costs. 05 Finally, in levy-exemption schemes, only those •  the training costs and can benefit from improved firms that do not train pay taxes. One good productivity without the risk of losing workers example is Côte d’Ivoire’s Fonds de Developpe­ to other firms. Because these instruments ment de la Formation Professionnelle. require good governance and effective legal enforcement mechanisms, they have been pre- The overall effects of the three types of levy ferred by high- or middle-income countries. In are nonetheless similar: the tax provides incen- the German company Deutsche Bahn for exam- tives for firms to train because firms that train ple, employees have the opportunity to partici- pay lower net training taxes (the training tax pate in training, on a voluntary basis. After the net of the transfer to cover the cost of the train- completion of the training program, there is a ing). Interestingly, and in spite of evidence that contractual retention of 18 months at maximum smaller firms are more affected by credit con- and Deutsche Bahn can demand reimbursement straints, this is rarely incorporated into the of the amount expended on training in whole design of the training funds. However, nothing or in part, depending on the circumstances of in this system ensures that the resulting train- the individual case.3 ing is efficient. For instance, firms might simply organize training to reduce the tax burden and The apprenticeship contract is probably the larger firms might be better at playing the sys- most widely used instrument to promote OJT tem and getting their money back than small across the developing and developed world. An firms. in the end, the distribution of taxes and apprenticeship is defined as a period during subsidies across firms (some firms will pay more which individuals and firms engage in a working than they receive and others will pay less) can relationship, firms benefit from low-wage work- be highly nontransparent and be delinked from ers (usually early in their careers), and individu- firm- or sector-specific market failures that als have the opportunity to receive firm-specific need to be addressed.2 training. This arrangement is very popular and useful for promoting the transition from school Payback clauses and apprenticeships to work because it gives youth the opportunity There can be enforceable institutional arrange- to develop skills that are directly relevant to the ments that create adequate incentives for labor market after completing their formal stud- employers and individuals to invest in skill ies. During the apprenticeship period, wages are development. Payback clauses and apprentice- usually set below the market rate with an expec- ship contracts are the most popular arrange- tation that they will rise afterward. Therefore, ments, with the former being most common in the apprenticeship usually promotes the sharing middle-high-income countries and OECD coun- of training costs between firms and workers (by tries as they are more difficult to enforce. underpaying workers during training). Because the implementation is relatively easy, financing A payback clause is a type of labor clause spec- and regulations are flexible, and the curriculum ifying that the worker is required to stay with can be customized, the apprenticeship is popu- the firm providing the training for a minimum lar even among micro and informal sector firms. length of time. Otherwise, the worker will have However, the weak educational base of informal to pay back the costs of job training. Because sector workers (or the trainers), the lack of qual- workers are required to stay with the training ity standards for training, and the limited poten- firm, employers can pay lower wages to recover tial for scaling up have been acknowledged as 06 the main difficulties with apprenticeship. Good apprenticeship plays a critical role in vocational implementation may require enforcing both an training. It combines school-based and on- adequate length of training and certification the-job training, follows a standardized cur­ systems, as well as improved access to new riculum and well-monitored quality standard, technology. and incorporates private education (see box below). Adopting the long-standing traditions An example of a well-defined and established of German apprenticeship will not guarantee apprenticeship program with good regu­ successful program implementation, but les- lation is found in Germany. As a part of a sons from the German program, including dual-education structure that distinguishes its structure and principle, would be useful in vocational and academic training, German promoting private OJT. T HE G ERMA N APPREN T I C E S H I P PRO G R AM The best-known institutional arrangement for addressing the contractibility problem between employers and employees is the apprenticeship. The best examples are found in countries with a long tradition of apprenticeships and where this tradition works as an enforcement mechanism. This is the case in some African countries and in some European countries, in particular Germany. Traditionally, apprenticeships have two components—the contract and the certification. Governments usually support the apprenticeships through a subsidy to the individual or firm and the contract is a private agreement between apprentice and the firm. The training content is usually designed by the firm but the governments ensure its quality and relevance. Apprenticeship programs can have both classroom and workplace training. Apprentices benefit from the training and acquire a certification at the end of their training programs. Employers benefit from the lower cost of labor during the apprenticeship period and can screen potential future employees. An example of such system in the German apprenticeship program, a dual system based on class- room training in vocational institutions and training in the workplace. The structure of the program, in terms of transparency and quality, curriculum, content and time structure of the training are regu- lated by the Federal Institute for Training and other national committees. Apprentices’ wages are lower than productivity but after training, they typically grow higher. The evidence shows that apprentices, especially in large firms, increase their employability. However, the post-training retention rates for firms are low as apprentices get general skills that can be applied in other firms than the one which trained them. In addition, training costs are far from negligible. The strongest constraint in the transferability of such programs is that they require a strong institutional framework, in particular a clear legal framework. 2 UIS and UNEVOS, 2007, “Participation in formal technical and vocational education and training programmes   worldwide: an initial statistical study”. Montreal. 3 Cedefop, 2012. “Payback clauses in Europe: supporting company investment in training”   07 Technical assistance and industrial extension services and training Institutional solutions to reach small enter- 200,000 employees. More than 300 business prises have also been developed to help SMEs associations were participating in CIMO in in training developments. Small firms are more 2000, up from 72 when it started. In addition, likely than large ones to be unaware of the ben- CIMO’s M&E effort distinguishes it from other efits of training and of the existing programs programs. CIMO administered data collection aiming at fostering training investment. in such a way as to gather panel data over time. A great deal of effort to construct data over A few developing countries have used inter- time has provided the policy makers with use­ mediary organizations to help SMEs imple- ful information on how the intervention has ment training programs. A good example is evolved and worked. South Korea’s effort to support SMEs through an intervention that attempts to overcome large INSTITUTION-BA SED TR AINING fixed costs and create economies of scale in training provision. In Korea, SMEs often did not PROGR AMS take advantage of training funds for which they There are many training programs offered by already paid levies probably because of limited institutions. These are targeted at both those capacity to administer and finance training. In who are leaving school and before employment recognition of this, the Korean government and for those who are currently in employment. encouraged SMEs in similar sectors and regions These include pre-employment technical and to form a consortium so that they could col­ vocational trainings (TVET) and training-related laborate in providing training. Later the pro- active labor market programs (ALMP). gram vertically expanded; now the consortium includes large firms, and SMEs can achieve syn- TVET Programs ergy with their large partners in acquiring and Pre-employment TVET continues to garner the maintaining skills. Even though this represents interest of developed and developing coun- an innovative effort to promote training, it can tries alike as a critical pathway toward gainful provide only limited evidence on its effective- employment. Although most of these funds ness because detailed data collection and rigor- are based on payroll levies, in practice, they ous evaluation was not incorporated. may take various forms With the objective of imparting readily applicable, job-relevant skills, Mexico’s Integrated Quality and Modernization it focuses mainly on young people nearing the program, or CIMO, is a comprehensive strat­ end of their initial formal schooling and enter- egy to support SMEs by addressing not only ing the labor market. The surge in basic edu­ specific training but also overall competency, cation graduates worldwide, as well as TVET’s including research and development, technol- perceived role in fostering East Asia’s industrial- ogy adoption, and managerial skills. CIMO was ization, has further increased the attention on established in 1987 and has been particularly TVET in recent years. Nonetheless, developing successful. By 2000 it was helping 80,000 countries face many issues in the provision of enterprises each year with a package of training job-relevant TVET. The challenges range from 08 the lack of quality program offerings and indus- Fostering the economic relevance of TVET pro- try experienced instructors, few industry part- grams is perhaps the most important policy nerships, and lack of adequate resources to challenge confronting pre-employment train- offer access to high quality workshops and ing systems. These mismatches between skills equipment. Moreover, there is limited training supply and demand often create a paradoxical provision to promote productive employment situation in some countries: high rates of unem- in the agricultural and informal sectors. ployment among the graduates of training pro- grams alongside complaints of skills bottlenecks The lack of data and the diversity of TVET pro- by employers. Low-income countries, in partic- grams suggest caution in drawing firm conclu- ular, face further challenges due to small, sions about the impact of TVET on labor underdeveloped modern industries and a lack market outcomes. The returns to investment of funds for sustainable training programs. in TVET vary widely across systems and over time, suggesting that the performance of pre- The successful experiences of the Republic employment TVET is context-specific. Among of Korea, Taiwan, China, and Singapore are OECD countries there is some evidence that, on exceptional and illustrate how economies can average, the employability of students graduat- overcome these difficulties by having effec- ing from TVET and from general tracks are simi- tive governance arrangements and explicitly lar but students graduating from general tracks incorporating demand-driven TVET into their tend to earn higher wages. However, isolating national industrialization policies. For exam- the effect of TVET is difficult as student charac- ple, the Demand-led pre-Employment TVET for teristics, some of them unobservable, influence Economic Growth in Singapore owes its success both the likelihood of selection into TVET and to the skills development schemes sponsored the level of labor earnings. Students following by the Singapore’s Economic Development the TVET track typically have lower test scores Board during the agency’s first decade (see and come from less favorable parental and below). This identified the source of demand family backgrounds than their peers in the gen- for higher-level skills and aligned the training eral track. Findings from survey data on labor curricula with the skills to be developed. market outcomes associated with TVET for Building partnerships with large private sector developing countries are also mixed and incon- companies such as Tata Group, this system suc- clusive. They illustrate that while returns to ceeded in providing people with the right skills pre-employment TVET can indeed be positive, to join the labor market. they are not consistently greater or worse than those associated with general education. 09 DEMAND-LEAD PRE-EMPLOYMENT T VET FOR ECONOMIC GROW TH IN SINGAPORE Singapore’s pre-employment TVET system is widely known for its quality, owing much to the skills development schemes sponsored by Singapore’s Economic Development Board (EDB) during the agency’s first decade. Facing an inauspicious start (failure of the training-production workshops and retraining schemes), the EDB decided to partner with leading international firms with proven training systems in order to learn the training business from them, train their requirements and adapt the methods for local needs. EDB’s first training partner was Tata Group, a large Indian conglomerate. It led to the creation of the Tata-Government Training Center to supply workers for Tata’s engineering plant in Singapore. The Singapore Government provided the land and buildings, contributed 70% of the center’s operating costs and paid the stipends of the trainees. Trainees were committed to work with EDB or any com- pany directed by the agency for 5 years. Twice as many people as Tata required were trained in order that the rest of the graduates not recruited by Tata would be an asset to attract other engineering firms to Singapore, contributing to the growth of an industry cluster. Training-related active labor market programs only and other multiple service programs that (ALMPs) include training as one of a number of interven- Not all workers have access to TVET or OJT, and tions—OJT, subsidies, counseling, etc. In gen- thus many countries have developed parallel eral, most training-related active labor market training programs for the more vulnerable programs are short, with a typical duration of workers. These ALMPs target a wide range of 4–6 months. The short duration of the pro- individuals with different needs: unskilled and grams suggests that at best they might be low-skilled workers who dropped out of the expected to have relatively modest effects on education system usually before having com- the participants comparable, perhaps to the pleted secondary school; youths transiting from impact of an additional year of formal school- school to work; skilled workers in transition ing.4 In many cases, governments are the direct between jobs. providers of training. Programs that are man- aged by the public sector suffer from weak There are many designs under these types of incentives to reach out to the private providers programs. It is useful to distinguish between and often also lack the administrative capacity programs that offer technical skills training to do so. Card et al (2010), “Active Labor Market Policy Evaluations: A meta-analysis”   4 10 Some of the most successful programs consist place to connect beneficiaries to OJT. Most of of comprehensive-multi-service approaches. the comprehensive programs have been taking These programs involve some combination of place in Latin America (Jovenes programs—see training (i.e., job and/or life-skills training), job below) and have produced positive results in search assistance, entrepreneurial services, and terms of jobs placement and employment ben- a range of other social and employment-related efits. The downside of these programs is that support services. Usually provided by qualified per capita costs can be high, on the order of private firms, NGOs or public institutions, they US$600—US$2,000, and, although cost-benefit provide training and other services on a com- analysis for some programs have yielded attrac- petitive basis. International experience sug- tive rates of returns, governments facing tight gests that core components to make these budgetary constraints might be able to only programs successful are the incentives put in afford them at a small scale. J OV ENE S PRO G R A M S Jovenes Programs refer to a series of programs that have been implemented in Latin America since 1991, aimed at improving youth employability and human capital. Targeted at disadvantaged young workers, they combine training and work experience as well as life skills, job search assistance, coun- seling and information. This model began in Chile and was subsequently replicated in Venezuela, Argentina, Paraguay, Peru, Colombia, Panama and the Dominican Republic. Financial incentives are given to both trainees and employers to ensure both sides participate. Training is provided on a competitive basis by private sector firms, while training institutions coordi- nate courses and internships. This ensures the skills taught in the program fit with the needs of the productive sector. Evidence shows that these programs contribute to an increased employment probability and increased earning of participants upon graduation, compared to their control group. For example, women’s employment probability has increased by 10% thanks to Proyecto Joven in Argentina, while monthly earnings also increased by 10%. On average, labor market impacts of these programs are higher for women and younger beneficiaries than for men and older cohorts. 11 Despite all the different programs that have However, rigorous impact evaluations as well been launched, it is unclear whether these as continuous monitoring of programs are still have had any real impact on overall labor mar- lacking in many developing countries. A strong ket outcomes. A forthcoming paper by Kluve conclusion from the review is the need for et al. (2015)5 has looked at the labor market major improvements in the quality of evidence impact of various kinds of interventions to available for ALMPs. The absence of rigorous improve the labor market outcomes of youth, evaluations almost certainly leads to an overes- among them training programs. Using data timation of program impacts by policy-makers. from both developed and developing countries, Properly evaluated programs are less likely to the meta-analysis shows that only small share lead to positive assessments of impact and (around one third) of skill training programs effectiveness than judgments based on non- have a positive and significant impact on labor scientific methodologies. In the absence of market outcomes. such evaluations, then, policy-makers are likely to overestimate the benefit of their inter­ There are, however, several factors that have ventions and, as a result, allocate resources been consistently associated with success of inefficiently. This is a particular concern in various programs—the most important being developing countries where resources are giving incentives to providers to respond scarce and evaluations are uncommon. to the needs of employers and job seekers. Excessive centralization in program administra- tion should be avoided by giving local labor offices the flexibility to manage their programs, and by outsourcing implementation to public and private companies through contracts that reward performance. The latter should meet clearly defined standards regarding the qualifi- cations of staff, particularly for counselors, and follow protocols for methods to assess compe- tencies and aptitudes, provide career advice, and identify training needs. There is also grow- ing evidence that a combination of technical training with soft skills and OJT components leads to better results than classroom training alone, especially for disadvantaged youth and adults. This model is essentially what some European countries are doing, including Germany and the United Kingdom. Kluve et al. (2015), “Interventions to Improve the Labour Market Outcomes of Youth: A Systemic Review of   5 Training, Entrepreneurship Promotion, Employment Services, and Subsidized Employment Interventions” 12 THE POLIC Y AGENDA such as subsidies or tax incentives. Even more, these advisory services would increase the Skills development programs are an important knowledge of managers on both the need for part of the jobs agenda, but they are very and benefits of OJT and any potential training unlikely to be successful if other interventions programs they might implement. These serv­ are not in place. Jobs strategies within a coun- ices could also be integrated with any existing try usually involve policy interventions to cre- nonfinancial support services for small and ate jobs, increase the productivity of jobs, medium enterprises. In this way, the firms and facilitate access to jobs. These strategies would benefit from an integrated approach to require multi-sector interventions—macro- technical assistance. This would promote economic policies, improvement in business competency, information and knowledge regulations, investments in infrastructure, and sharing. Technical assistance could take better labor policies. Having a regional and sec- the form of business consulting to promote toral focus in the design of these strategies is firm productivity, including strategies related also important.6 Skills development programs to technology adoption, human resource therefore need to be coordinated with these management and skills development. A interventions to have a meaningful impact. complete technical assistance would help them to improve the complementarities Before continuing to develop and expand the between their workforce and other factors of current skills development interventions, production. there is a need to better understand the Increase firms’ access to credit tied to train- •  market failures. In that view, there are three ing investments. There is room for policies to important aspects to the skills and jobs agenda provide and improve financial support to firms, that need to receive more attention: firms; in particular by increasing their access to contestable markets; financing. Some sugges- credit geared towards training investments. tions in terms of priority interventions are pre- This is because firms may face liquidity con- sented below. straints that prevent them from investing in training. Currently, the most popular way to Constraints to firms’ investments in training provide subsidies to firms is through training could be addressed in several innovative ways: funds, particularly those based on payroll •  Advise firms’ management on the productiv- taxes. Most international experience shows ity benefits of training. While many existing that this type of incentive is focused on larger programs focus on alleviating financial con- firms, for which the returns of training are straints, one of the key ways of fostering arguably higher—even without any public firms’ investment in training is to make sure intervention. Another way to increase firms’ that managers are aware of the positive access to credit is via matching grants, where impact of training on firms’ productivity. They government subsidies match the firms’ could also be advised on what programs and resources. incentives exist for investments in training Concept Note for the World Bank’s work on “Jobs Diagnostics and Strategies: Lessons from their Design and   6 Implementation in 20 Countries” 13 D esign and test mechanisms to allocate •  Institution-based training programs could partial subsidies to incentivize training, espe­ benefit from having more contestable markets cially for generic transferable skills. Coun­ for training provision. This can be achieved in tries could consider moving away from pay-roll the following ways: taxes for training and tax-exemptions, and Level the playing field for private training •  instead focus on the design of matching providers. To level the playing field for private grants financed through general revenues. companies, the laws and institutions should Grants would be targeted to firms with favor the presence of both public and private growth potential, such as young innovative providers of training. This would foster com- start-ups, rather than focusing only on firm petition and innovation in training programs. size. Other firm characteristics that could be In many countries, training providers are often used as proxies for targeting include the public, and they are paid based on inputs firm’s sector of activity, the frequency of or outputs (such as the number of people it adopts technology, or the gaps between trained) rather than on their performance or wages and productivity. Firms usually have outcomes (such as the number of people who higher incentives to provide their workers got jobs after training). Having both public with training for job and firm specific skills. and private companies would increase com- Generic skills training, such as communication petition and potentially training quality. and presentation skills or problem-solving, however, are less likely to be provided by Make information publicly available on the •  firms, as there is a higher risk of trained work- quality and performance of training provid- ers leaving the firm and applying these newly ers. In addition, clear and detailed informa- acquired generic skills in other firms. There­ tion on the quality and the performance of fore specific subsidies to generic skills could different providers would keep the best train- be designed in order to reduce the risk taken ing providers in the market and force others by firms in investing in this kind of training. out. For instance, in Australia, Job Services Australia (formerly called Jobs Network) is a • Finally, encourage public-private partner- network of competitively selected for-profit ships. These are an efficient way of sharing and not-for-profit organizations delivering risks and costs between governments and employment services to job seekers, paid firms. The example of Vale, a Brazilian com- for by the government. Each organization is pany, shows a successful public-private part- responsible for providing job seekers with nership. Vale’s training program identifies the personalized services such as job search and skills the corporation needs and helps local training. The network has a star rating sys- educational partners develop specific courses tem, assigning stars to agencies according to to train an appropriately qualified workforce. their level of performance. The benefits of The program offers two curricula at the uni- this system are two-fold: first, it increases versity level and at the technical apprentice competition across agencies, incentivizing level. The first curriculum educates engineers them to provide better employment services; and geologists, in partnerships with universi- and second, it provides job seekers with the ties. The technical apprentice level targets necessary information of each agency’s per- young people who are finishing school and formance. While competition between pri- provides them with the technical education, vate and public training agencies has been partnering with local education institutions. 14 implemented over the last decades in OECD • Finally, develop public goods such as new countries, in developing countries it is more models and approaches to teach socio-emo- recent. tional skills. While there is growing evidence on the importance of these skills for labor Pay for results. Increasing information on the •  market success, there are big knowledge gaps performance of training providers can go on what are cost-effective approaches to hand in hand with performance-based pay- develop these skills among the young adult ments of training providers. In Lebanon, for population. Most of the existing interventions example, the New Entrants to Work Program have been tried for children and teens still (NEW) selects NGO and private training agen- in the education system. There is a need to cies on a competitive basis, and rewards them develop effective approaches that are more based on their performance. Indeed, it might suitable for delivery through labor market be a way of incentivizing training providers training programs. to increase the quality of their training pro- grams through contracts that reimburse Both firm- and institution-based training pro- based on outcomes of the training such as job grams could benefit from innovative financing placements and the sustainability of the jobs mechanisms. Indeed, both individuals and acquired. firms often lack the necessary funding to invest Put in place clear and transparent criteria •  in training. In addition, given the rising need for the eligibility of training providers. The for skills training in many countries, govern- experience of countries such as Australia, ment financing might not be sufficient to cover which introduced a fully contestable market everyone who needs training. Consequently, for VET in 2012, shows that when contestabil- there is a need for alternative financing arrange­ ity is introduced into a training market, it ments that rely more on costs and risk sharing must be done alongside strong regulatory with the private sector and the individuals mechanisms. Importantly, there need to themselves. accurate and readily available quality signals. Social bonds might be used to finance invest- •  This enables students and employers to make ments and service delivery. Initially intro- informed decisions about their choice of pro- duced in the United Kingdom at the beginning vider. Any delays in implementation of both of the 2000s, the social impact bond model sufficient regulatory and transparency mech- makes outcome payers such as governments, anisms can lead to subpar training providers foundations or corporations contract with taking advantage of a lack of scrutiny and private sector intermediaries to obtain social public financing. Clear and transparent eligi- services. Outcomes payers can pay providers bility criteria for training providers need to be entirely or almost entirely upon completion in place and an independent assessment of of performance targets and can refuse to pay whether these criteria are met should be if minimal targets have not been reached. periodically conducted. 15 In this model, commercial or philanthropic ITAs give individuals the freedom to choose investors provide upfront, unrestricted capi- the training providers that match their tal and receive returns if improvements in requirements. In some ITA models, individu- outcomes are achieved. The intermediary als can also consult counselors, who help uses these funds to finance service providers. them choose the training providers but who This has several benefits: it shifts the invest- cannot choose for them. In the future, indi- ment risks from taxpayers to private investors vidual training accounts could benefit from and it also reduces the fiscal constraints as more cost-sharing between the government the funding does not come from tax based and individuals, by incentivizing individuals to public expenditures. In addition, the per­ pay for a part of their training. This means formance-based structure of the program they would become more selective about the incentivizes service providers to increase the type of training they undertook and training quality of their services. providers they used. This could be modelled after individual development accounts (IDA). Individual training accounts (ITA) have been •  These are matched savings accounts for used in the United States, for instance, to pro- low-income people, which can be used to vide job seekers with vouchers they can use purchase a home, to start a small business, or to get training by eligible training providers. further education or job training.7 There are more than 500 such savings schemes in the United States   7 THE TRAINING POLICY AGENDA REDUCE FIRMS’ CONSTRAINTS Advisory services to management Access to credit (matching grants) Target subsidies at generic skills training PPP – public private partnerships IMPLEMENT CONTESTABLE MARKETS Level playing field for private firms Information on quality and performance of training providers Performance-based payment Objective criteria for the selection of training institutions FINANCING MECHANISMS Social Bonds Individual Training Accounts WHAT C AN WE DO NEX T? have now been implemented and analyzed in about 20 countries across the World since Going forward, there are several opportunities 2010. Emerging evidence from these surveys to improve training policies to ensure that the have already inspired innovations in educa- current and future workforce have the skills tional and training programs in countries like needed in the labor market: Peru and Vietnam where governments are undertaking major reforms to strengthen the First, more collaboration with the private •  teaching of these skills from basic education sector is needed. Experience from training to building closer partnerships with the enter- programs around the world shows that prise sector in post-secondary, labor training engaging the private sector is key to ensuring and OJT programs. that training is relevant to the needs of employers. However, the private sector can Finally, piloting and evaluating innovative •  play a much bigger role in designing, financ- approaches is key to workforce develop- ing, and implementing training programs in ment. As shown above, there are several order to increase their impact. Public-private innovative ways to increase the efficiency and partnerships have already been successful effectiveness of training programs. Whether and could be encouraged further. it is innovating with the way subsidies are targeted or introducing performance-based Second, in following with the adage “what •  payments for training or approaches to teach gets measured gets the attention, and gets socio-emotional skills, starting with small done”, there is a need to continue improving proto-types that are potentially scalable is a the measurement of skills gaps—ie, which more prudent and efficient approach than skills are demanded by employers and to introducing untested approaches system wide. what extent these are possessed or not by Thus, it is a priority to invest in monitoring workers. Without these measures, policy and evaluation to rigorously analyze results interventions will be guided in the dark. and estimate the costs and benefits of inter- Through the STEP Skills Measurement Pro­ ventions. As a result, it would be possible to gram, the Bank, in collaboration with the determine whether interventions have suffi- OECD and prominent academics, has been ciently large gains to compensate for their pioneering efforts across the developing costs, which is key to knowing that resources world to measure and analyze the role of cog- are being well spent. nitive, socio-emotional and technical skills for labor market success.8 These surveys, consist- ing of an employer and a household surveys,  http://microdata.worldbank.org/index.php/catalog/step/about 8