3. Improving skills to harness the benefits of a more open economy

Fostering a stronger integration into the global economy has significant potential to raise productivity, real wages and well-being. At present, the economy is relatively closed and has foregone many of the benefits that other emerging markets have reaped from international trade and investment (see Chapter 2) (OECD, 2018[1]). For producers, productivity and competitiveness would gain from access to cheaper and higher quality inputs and capital goods, which would in turn allow improvements in wages (Amiti and Konings, 2007[2]; Goldberg et al., 2009[3]). For consumers, trade has potential to reduce prices and enlarge the variety and quality of available goods, which would improve particularly the purchasing power of low-income households (Amiti, Redding and Weinstein, 2019[4]; Grundke and Arnold, 2019[5]).

OECD model simulations suggest substantial rewards from lower trade barriers (Figure 3.1). An ambitious reform could raise labour productivity by 4%. The rising competitiveness of domestic producers would boost exports by 14% facilitating improvements in real wages. Moreover, increasing import competition would reduce consumer prices raising the purchasing power of households and private consumption by 6%. These benefits are likely to be progressive as lower income households spend larger shares of their incomes on tradable goods such as food, home appliances, furniture and clothing, which are currently highly protected (see Chapter 2). In the case of lowering only tariff barriers, and not non-tariff measures or services trade restrictions, private consumption is estimated to rise by 4%.

However, opening up to trade is likely to trigger structural changes in the economy. These are a crucial element of the productivity benefits, but can raise challenges in the transition. Increasing competitive pressures forces companies to reduce inefficiencies and upgrade their production processes through more advanced technologies. Firms need to increase product quality and reduce high prices that result from low domestic competition (Amiti and Khandelwal, 2013[6]; De Loecker et al., 2016[7]). While this leads to a revitalising effect on the more productive domestic firms, which seize newly arising export opportunities, expand and hire new workers, some low-productivity firms leave the market, freeing resources for the more productive firms and sectors to grow (Melitz, 2003[8]; Pavcnik, 2002[9]; Criscuolo, Gal and Menon, 2014[10]; Araújo and Paz, 2014[11]).

As workers and capital move towards more productive firms and activities, the economy as a whole becomes more productive, and this process does not necessarily comprise a massive substitution towards imports (Hsieh and Klenow, 2009[12]; Topalova and Khandelwal, 2011[13]). However, the transition of workers to other firms, sectors and locations, can imply significant adjustment costs (Winters, Mcculloch and Mckay, 2004[14]; Goldberg and Pavcnik, 2007[15]; Autor et al., 2014[16]). Displaced workers lose their wage income; they have to learn new skills and often have to bear, together with their families, the social costs of moving to other locations to find a new job (Hummels, Munch and Xiang, 2018[17]; Hyman, 2018[18]). Workers that stay in their jobs also need to update their skills. As firms upgrade production processes through more advanced technologies and increasingly import high quality inputs, the task content and skill requirements of existing jobs changes (Hummels et al., 2012[19]; Becker and Muendler, 2015[20]; Becker, Ekholm and Muendler, 2013[21]).

The reallocation process triggered by stronger external competition also presents an opportunity to harness the benefits of a rising digitalisation of production processes, which adds to potential productivity gains from rising international integration. At the same time, digitalisation will eventually also amplify structural transformations in the economy (OECD, 2019[22]). Some occupations will be automated and workers will need to move to new jobs in other firms or sectors (Arntz, Gregory and Zierahn, 2016[23]; OECD, 2018[24]; Autor and Dorn, 2013[25]). In many other occupations the nature of tasks and the skills required to perform will change (Spitz-Oener, 2006[26]; OECD, 2019[22]). The digitalisation and globalisation of production processes increasingly requires a good mix of cognitive and social-interactive skills (Hummels, Munch and Xiang, 2018[17]; Grundke et al., 2018[27]).

The best policy response to these challenges is to invest in people, by supporting workers with targeted education and training programmes that address changing skill needs. Training may involve shorter vocational training courses, but also more fundamental formal education to help workers move into new occupations and sectors (Hummels et al., 2012[19]; Hyman, 2018[18]). Especially for low-skilled and informal workers, an improvement in general cognitive skills, such as literacy and numeracy, through longer term formal education courses might be necessary to participate in professional training and move into new jobs in expanding sectors and firms (Bechichi et al., 2018[28]; Bechichi et al., 2019[29]; Autor et al., 2014[16]). An effective social protection system combined with efficient job placement services would facilitate the transition process and support better training and employment outcomes for displaced workers (OECD, 2019[22]).

Besides education and training policies, other policies that are beyond the scope of this chapter can also help to mitigate adjustment costs and support the structural transformation of the economy. This entails reducing frictions on labour and capital markets, improving infrastructure, especially transport infrastructure, and encouraging innovation (Rusticelli et al., 2018[30]; Fiorini, Sanfilippo and Sundaram, 2019[31]). Phasing out distortive taxes and simplifying the tax system would increase the competitiveness of domestic firms and contribute to the creation of high-quality jobs. Moreover, strengthening domestic competition on product markets is crucial, as Brazil still has the highest barriers to entrepreneurship among countries covered by the OECD Product Market Regulation (PMR) Indicators (see Chapter 2 of this Survey).

Finally, opening up to trade also has significant distributional consequences. Empirical evidence shows that workers are affected differently by opening up to trade, depending on their skill level, occupation or the sector and firm they work in (Goldberg and Pavcnik, 2007[15]; Winters, Mcculloch and Mckay, 2004[14]). Moreover, as certain industries are concentrated in certain regions and production factors are not fully mobile across regions, trade opening can also have distributional consequences across regions (Rusticelli et al., 2018[30]). To ensure that everybody gains from trade, governments should complement trade opening with redistribution policies comprising effective social transfers and progressive taxation (Autor et al., 2014[16]; Lyon and Waugh, 2018[32]; Hyman, 2018[18]).

History can provide some insights into the likely effects of stronger integration going forward. Since the 1970s, trade policy in Brazil has seen only one significant inflection point, which is the visible reduction in trade barriers that occurred in the context of macroeconomic stabilisation in the 1990s. Many manufacturing sectors experienced substantial reductions in trade barriers in the early 1990s, when tariff rates fell from an average of 31% in 1990 to an average of 11% in 1995 (Figure 3.2). This reform had a sizeable impact on the economy, although in light of continuous trade integration around the world and the absence of significant follow-up reforms in Brazil, the economy has pretty much reclaimed its outlier position in terms of openness by now (Chapter 2).

Overall, the period since the 1990s has been characterised by strong growth of formal employment (Figure 3.3). This has been related to several factors, including improvements in macroeconomic stability, solid growth, a rising working-age population, improvements in education and progress in formalisation. In this context, the share of medium and high-skilled jobs grew steadily, reflecting both improvements in access to education and increasing demand for these skills (Silva, Almeida and Strokova, 2015[33]).

The services sectors showed a particularly strong expansion of formal employment, absorbing many young high school graduates, but also informal workers from the agricultural sector (Dutz, 2018[34]). Skill-intensive services sectors like telecommunications, information technology (IT) and other business services led this process, but also retail, transport services, and hotels and restaurants, where more skill-intensive jobs have emerged. This trend towards skilled employment in market services was complemented by a strong expansion of employment in public administration, health and education services, where over 80% of workers are medium or high skilled.

In most manufacturing sectors, falling trade barriers and the resulting rise in import competition in the early 1990s were associated with initial employment losses (Figure 3.4). However, over time, productivity-enhancing factor reallocation and increasing exports generally lead to a strong recovery of employment, which was increasingly medium and high skilled, particularly in the chemical, machinery and transportation equipment industries (OECD calculations based on RAIS data).

For example, chemical products, food processing, machinery and equipment as well as transport equipment experienced tariff cuts of more than 20 percentage points, while labour productivity and exports grew by more than 5% and 10% per year from 1995-2011, respectively (OECD calculations based on OECD TiVA data). Formal employment initially fell but started to recover strongly towards the end of the 1990s, eventually surpassing initial employment levels. In other manufacturing sectors with sizeable tariff cuts, e.g. textile and wearing apparel, electronic equipment or rubber and plastics, productivity and export performance have been weaker. Although formal employment started to rise at the end of the 1990s, it remains below pre-liberalisation levels, the exception being the textile and wearing apparel industries, which remain highly protected.

Drawing lessons from aggregate or even sector-specific developments is intrinsically difficult due to a myriad of competing explanations and factors acting at these levels. For this reason, an empirical analysis based on 481 local labour markets may provide more reliable insights into the effects of increasing international economic integration on employment and firm dynamics in Brazil. Local labour markets, called micro regions in Brazilian national statistics, differ with respect to their industry structure and thus their exposure to trade liberalisation (Figure 3.5). This in turn allows identifying the effects of trade opening by focusing on different trajectories of outcomes across regions (Box 3.1).

Although formal employment has increased strongly since the 1990s, employment growth has generally been weaker in regions that were more exposed to import competition (Figure 3.6). In particular, this reflects initial employment losses in manufacturing sectors during the 1990s. Since the 2000s, formal employment has recovered, but some differences between regions with different levels of exposure to trade liberalisation remain. For example, a micro region that experienced a 10 percentage point larger tariff cut in the 1990s, approximately the difference in tariff reductions between the 90th and the 10th percentile of micro regions, showed a 45 percentage point lower employment growth between 1990 and 2015. These average effects hide considerable heterogeneity across micro regions that have experienced large tariff cuts in the 1990s. Some of them have performed well, with strong employment and output gains recorded in micro regions located in poorer states such as Pernambuco, Maranhão, Alagoas and Rio Grande do Norte (OECD analysis).

Employment growth in regions that were more exposed to trade liberalisation has been especially weak for low-skilled workers. From 1990 until 2015, their employment share has decreased by 3.7 percentage points relative to less exposed regions (Figure 3.7). Particularly the share of medium-skilled jobs has increased in trade-exposed regions, but high-skilled employment has also expanded its relative weight. A similar trend can be observed for real wages as well as wage shares across skill groups. From 1990 until 2015, the wage share of low skilled workers in trade-exposed regions decreased by 7.2 percentage points relative to less exposed regions, and the one of medium and high-skilled workers increased by 4.2 and 3 percentage points, respectively. This indicates that rising import competition presents challenges for low-skilled workers, who need to move to newly created medium-skilled jobs in expanding firms and sectors, or even other regions.

Identifying sectors, firms, occupations and regions, where workers will be displaced or new jobs will be created, is key to design efficient and effective training policies and mitigate adjustment costs for workers. The content and intensity of the required training depends on the difference between the task-specific skill requirements in the old and the potential new job as well as the general cognitive and social skills of the worker (Bechichi et al., 2018[28]). Especially for older workers, learning new skills and moving to new jobs in other firms, sectors or regions can be costly (Autor et al., 2014[16]). The analysis shows that elderly workers are particularly at risk of suffering trade-related job losses and find it more difficult to move into new jobs (Grundke et al., 2020[39]).

Training needs depend on the exact nature of reallocation. As producing the same type of products usually requires a similar set of skills, the need for training to help displaced workers move to other firms in the same sector is likely to be smaller than for moves to other sectors (Bechichi et al., 2018[28]; OECD, 2019[22]). Nevertheless, firms within the same sector differ in their production processes and the type of technology they use (Andrews, Criscuolo and Gal, 2015[40]). Increasing technology adoption among highly productive frontier firms will change the task-content of occupations and require workers who move to more productive firms to update their skills (Bechichi et al., 2019[29]; Hummels et al., 2012[19]; Hummels, Munch and Xiang, 2018[17]).

Rising import competition has invigorated the forces of creative destruction and increased entry and exit rates of firms. This churning process has been more pronounced in regions more exposed to trade liberalisation, where cumulative firm entry and exit rates exceeded those of less exposed regions by 7 and 11 percentage points in 2015, respectively (Figure 3.8). Resources shifted from exiting to entering firms, as the average size of incumbents has decreased and the size of entrants increased in more trade-exposed regions.

Higher churning has in turn led to increases in productivity and real wages. Labour productivity has risen more rapidly in trade exposed regions, supporting exports and enabling increases in real wages (Figure 3.9) (Grundke et al., 2020[39]). In particular, real wages for medium and high skilled workers have grown faster in trade-exposed regions, signalling that the upgrade of production processes and technology adoption in more productive firms are complementary to high and medium-skilled labour.

The finding that lower trade barriers foster the creation of medium-skilled jobs presents an opportunity for training and education policies to succeed in mitigating the adjustment costs for low-skilled workers. Many of these medium-skilled jobs that have expanded in regions more exposed to trade liberalisation are low-complexity white-collar jobs, such as clerical support, services and sales workers (Grundke et al., 2020[39]). Transitions from low-skilled occupations to these expanding medium-skilled occupations may imply relatively manageable training needs (Bechichi et al., 2018[28]).

As rising trade exposure increases job-turn over particularly for low and medium skilled workers (Grundke et al., 2020[39]), the transition towards a more open economy would be easier with additional policy support for these workers. This could take the form of strengthening the social safety net, especially for informal workers, and skills programmes, to reduce the length of unemployment spells.

The current structure of import protection varies considerably across different sectors of the economy. Tariff rates are high for textile, wearing apparel, leather and furniture industries, and non-tariff measures particularly protect food processing, transport equipment as well as machinery and equipment industries (Cadot, Gourdon and van Tongeren, 2018[41]). Services sectors also differ in their openness to foreign competition, with banking, insurance, media and broadcasting as well as logistics showing the highest barriers to foreign entry (OECD calculations based on the STRI index).

Lowering these trade barriers will lead to reallocation of labour across sectors with substantially different skill-requirements (Bechichi et al., 2019[29]). Identifying existing sectors with particularly strong challenges and large expected training needs can help to target training and education policies effectively. At the same time, identifying those sectors with particularly strong future employment potential may help to guide the choice of training content.

Economic models can help to guide these choices. Static computable general equilibrium models are a common analytical tool to investigate the effects of trade liberalisation on the reallocation of production factors across sectors, while assuming that total factor endowments in the economy are fixed and fully employed (Box 3.2). The modelling exercise presented in the introduction of this chapter, based on the OECD METRO model, can be used to shed more light on sector-specific developments.

Some manufacturing sectors with high protection, such as textile and wearing apparel, metal products, machinery and equipment, motor vehicles, electronic equipment and other manufacturing, would face stronger import competition with lower trade barriers. Rising competitive pressures will increase productivity and ultimately lead to rising production and exports in those sectors. However, the model simulations suggest some employment losses in highly protected sectors (Table 3.1). The same holds for a number of services industries like finance, business services or transport services, where moderate reductions in the workforce would go along with rising productivity, output and exports. At the same time, the model simulation also identifies sectors where new employment opportunities will arise. These include food processing, agriculture, transport equipment, communication services and other services sectors such as utilities, retail, construction, tourism, education and health.

The results of the econometric analysis and the modelling exercise can provide inputs for the targeting and the content of future training programmes. Training needs will depend on the difference between the task-specific skill requirements in the old and the potential new job, but also on the general cognitive and social skills of the worker (Bechichi et al., 2018[28]). Many workers in protected sectors have not completed primary education and are in informal employment, lacking access to formal training opportunities (Figure 3.10). For these workers, improvements in general cognitive skills through longer-term formal education courses are a pre-condition for bridging the skill gap towards newly created medium-skilled jobs in expanding firms or sectors (Bechichi et al., 2018[28]).

In particular, workers in the heavily protected textile and wearing apparel industries will require intensive education courses to prepare for future job transitions. More than 31% of the workforce of 2.8 million have not completed primary education and 44% are in informal employment (Figure 3.10). Metal products and other manufacturing industries show a similar picture. In agriculture, food processing and some services sectors, which are expected to expand employment as import barriers fall, the share of low skilled and informal workers is also very high (Figure 3.10). As these sectors face rising import competition and jobs move towards high-productivity firms, basic adult education policies are key to prepare these workers for the challenges associated with new technologies and rising product quality.

Moreover, displaced workers who dispose of the necessary cognitive skills for moving into medium and high skilled jobs may need more specific vocational training courses to acquire the specific skills necessary for a new job. For these courses to be successful, alignment with skill demands in local labour markets is key. This implies a need to involve the private sector closely in content, design and implementation of these courses (O’Connell et al., 2017[42]; OECD, 2018[43]). The geographic mobility of workers, and in particular of low-skilled and older workers, is often severely limited by the low quality of transport infrastructure, housing market distortions, job search costs, social costs of moving families or a limited ability to learn and adapt to new social contexts (Autor et al., 2014[16]; Gathmann, Helm and Schönberg, 2018[44]; Hyman, 2018[18]). A lack of alignment with local labour market needs has been identified as a shortcoming of previous training programmes, notably the large-scale programme PRONATEC (OECD, 2018[45]). For example, poorly designed remuneration incentives have led training suppliers to build up excessive training capacities for secretarial assistants, which were relatively cheap to instruct, while local companies continued to face severe difficulties in hiring staff with technical skills.

Empirical analysis based on micro data allows matching regional exposure to stronger import competition with detailed information on past training policies (Box 3.3). In particular, this information allows the identification of training capacity that was created in response to skill demands by local employers, which corresponds to a small fraction. The results of this analysis suggest that only this vocational training that was aligned with local skill demand has been effective to reduce the negative employment effects of rising import competition (Figure 3.11). Trade-affected micro regions that have raised the supply of vocational training courses targeted to local skill demands by an average of 10,000 training class hours from 2011 to 2015 have narrowed the gap in employment growth with respect to less trade-exposed regions by around 50%. Other types of training courses, which had not specifically targeted skill demands in local labour markets, have had much smaller effects on employment growth. In many cases, the effects were negligible, highlighting the importance of targeting.

One lesson that can be distilled from this analysis is that the investment required to implement effective training programs is manageable and likely to have high pay-offs, once proper targeting is ensured. OECD estimates based on micro-data and past training costs suggest that additional yearly spending in the order of 0.5% of GDP could be sufficient to address upcoming skill challenges, provided that future training programmes are carefully targeted to skill demands in local labour markets.

However, for the 24 million workers without completed primary education, the necessary investments are likely to be higher. Vocational training courses targeted to local skill demands have particularly improved employment opportunities for medium-skilled workers (Blyde et al., 2019[46]; Grundke et al., 2020[39]; O’Connell et al., 2017[47])This may be a sign that current training institutions have not been successful in supporting workers who need it the most. Completing basic adult education is a pre-condition for successfully attending many vocational training courses. Thus, policies to prepare the workforce for the rising integration into the global economy need to comprise an ambitious expansion of adult education and an improvement of its quality.

The empirical analysis based on the 1990s trade reforms shows a need for training policies to accompany future trade opening, to ensure that all Brazilians can reap the benefits, including those with low skills. At the same time, the analysis also shows how much of an impact training policies can have to mitigate challenges in the transition, provided that they are well-designed. This combination builds a strong case for a major boost to investing into people, by providing them with the necessary skills to reap new opportunities. Investing in people’s skills will not only help to master the transition towards a more modern and open economy, but it will also support the productivity improvements that Brazil urgently needs to sustain economic growth (see Chapter 2).

Within the realm of active labour market policies, Brazil’s expenditure is on par with the OECD average, but this spending is much less concentrated on training policies (Figure 3.12). Most of the spending goes to programmes supporting self-employment and micro enterprise creation as well as to employment subsidies. Part of these resources could be shifted to vocational training and adult learning policies, as these other measures are less effective in increasing the employability of participants (Brown and Koettl, 2012[48]). In addition, improving labour market job services and linking them better with training policies may also require additional resources, but can significantly improve the employability of job seekers and reduce skill-mismatch in the labour market (OECD, 2018[45]).

Not only is spending for training and adult learning policies relatively low, the training and adult learning system has also been characterised by large spending inefficiencies (OECD, 2018[45]). From 2011 until 2018, Brazil implemented a large-scale training program called PRONATEC in a coordinated effort involving various ministries. Training subsidies targeted particularly the poor and low-skilled parts of the population including informal and unemployed workers. The regional coverage was extensive and included many poor and remote areas.

An important shortcoming of PRONATEC, however, was a mismatch between course content and skill needs in local labour markets, at least for large parts of the programme (Box 3.3). In many cases, the type and content of subsidised training courses as well as the target population were centrally decided without taking into account skill demands and training needs in local labour markets (OECD, 2018[45]). Moreover, weakly designed incentives for training institutions led to a large heterogeneity in training quality, dropout rates and future employability of training graduates. The programme paid a fixed subsidy per enrolled student without any incentives related to content relevance, cost effectiveness or quality improvements. Evaluation was not mandatory. In those cases where private training providers had a say in choosing training content, they usually offered courses that could be set up or expanded quickly rather than those that were needed. Rapid expansion of course capacities weakened the selection criteria for teachers and reduced teaching quality. As a result, large numbers of workers enrolled in vocational training courses related to occupations for which labour demand was low (Figure 3.13).

The assignment of students to training supply was also characterised by a lack of interaction with career guidance services, which are patchy and poorly structured. As a result, many low-skilled workers lacked the most basic skills to benefit from the course (OECD, 2018[45]). On the other hand, career counsellors had no access to information about past training participation, which prevented the design of consistent training agendas for individuals and further reduced the employability benefits from professional training for workers. A small part of PRONATEC, however, referred to as PRONATEC-MDIC or SuperTec and organised by the Ministry of Industry (MDIC), did not suffer from misalignment with labour market demands (Box 3.3).

Training policies are currently undergoing a comprehensive overhaul, with a particular focus on improving their cost effectiveness through more involvement of the private sector. An additional 1.3 million training places will be created, with the training supply based on a prior analysis of skill demand involving local productive sectors. These will be allocated via training vouchers handed out to firms between 2020 and 2022, strengthening the link between training choices and local firm needs. The training places will be created by SENAI, which is part of a semi-public network of training providers that receive part of their revenue from a national payroll tax. SENAI has played a major role in the provision of PRONATEC training courses and its regional coverage and training quality is high (OECD, 2018[45]).

The vouchers will allow firms to select the training participants and pay for the training they need. Training participation is not limited to current employees of the firm, but could include unemployed persons that the firm wishes to hire after the training. The engagement of private employers will lead to rising effectiveness of the training, as firms have an incentive to select the worker for whom the training will be most effective and will select training courses that reflect private sector skill demands. This approach has a strong focus on training those with greater expected benefits from training.

However, there may be a case for expanding policy efforts beyond these low-hanging fruits, although that would require additional resources. The social benefits of expanding professional training programmes to make more space for those from disadvantaged socio-economic backgrounds could be substantial. In fact, there is a risk that the new training subsidy may not reach the workers that would need training the most: low-skilled, informal or unemployed workers that are particularly affected by the structural changes that trade opening or digitalisation will bring about (OECD, 2017[49]; OECD, 2019[22]).

To support this vulnerable group of workers, the government is planning to introduce social impact bonds to finance training provision. In this type of public-private partnership arrangements, providers would get significant autonomy in designing training courses and content, but their remuneration would be made contingent on specific outcome targets, such as the future employment rate of training participants. The potential success of such arrangements crucially depends on how contracts are designed (OECD, 2016[50]). In particular, the definition and measurement of social outcomes and the selection of target and control group to evaluate whether objectives have been met are complex tasks that entail significant transaction costs and risks for the public sector. So far, the empirical evidence on the effectiveness of social impact bonds has been mixed and governments should be careful in providing specific social services completely through these type of arrangements (OECD, 2016[50]). However, as a complement to traditional forms of public social service provision, social impact bonds have the potential to nurture a culture of monitoring and evaluation in social service delivery, which is currently missing in Brazil.

Additional public resources will be required to effectively equip unemployed, low-skilled and informal workers with market-relevant skills that will allow them to share in the opportunities that lie ahead. The government should consider closely involving SENAI and related institutions in the design and implementation of targeted training policies for vulnerable workers, not least because these semi-public institutions receive payroll tax revenues of around 0.25% of GDP. Despite agreed quotas to include disadvantaged workers in its training courses, their number has in reality been low, which is partly related to a lack of enforcement of the agreed quotas (OECD, 2018[45]).

An expansion of training efforts should include a more systematic role for evaluations and subsequent refinements. Close cooperation with the private sector to detect changing skill requirements of occupations as well as systematic feedback from employed graduates should inform regular updates of training content and quality control. SENAI, for example, has established methods for training assessment and content update that follow international best practices (OECD, 2018[45]).Ongoing efforts to expand demand-driven SuperTec courses and collect training requests from local firms through an electronic platform should be expanded and combined with a skill anticipation assessment focusing on skill needs in local labour markets (Box 3.3). Best practice examples for region-specific skill anticipation exercises already exist in Brazil, including the work of SENAI in the state of São Paulo. Building on this success with a view towards covering all sectors and regions within the country could feed into region specific training catalogues (OECD, 2018[45]).

The current voucher system could be complemented with additional vouchers allocated to disadvantaged individuals. The allocation could be based on administrative data such as the comprehensive register of all poor households used to administer the conditional cash transfer programme Bolsa Família. This could reduce registration and information costs and improve competition between training providers, if combined with a transparent quality certification system for training institutions (OECD, 2018[45]). The vouchers could be used to select courses from a region-specific training catalogue that is closely linked to private sector skill demands. Another option could be to allocate a certain share of places in courses requested by firms in the current voucher system to disadvantaged workers, following a successful experience in Singapore (Box 3.4). Incentivising the supply of evening, part-time or distance learning courses and linking the worker-specific subsidy to income or living area would facilitate the participation of disadvantaged workers living in remote areas.

Career counselling and job placement services could be better linked with training programmes (Box 3.4). Matching trainees’ skills and experience with training content is important to ensure the cost effectiveness of training. Access to courses could be made dependent on career counselling and an evaluation of the necessary prior-knowledge (Box 3.4). High dropout rates in PRONATEC training courses were often related to missing fundamental skills of trainees as well as unrealistic expectations about the course content and its objectives (OECD, 2018[45]). Counselling services could provide better information on training opportunities and help to direct those interested in training to the right course.

Moreover, job counsellors should have full access to individual training records and employment histories to better match job seekers with skill demands of firms and tailor training supply to individual needs. Improving and expanding the certification system for work competences, in particular for informal workers, would raise employability and improve the effectiveness of public employment services. It would also encourage training investments of workers (Dutz, 2018[34]). Current efforts by the government to share databases and engage private providers in job counselling services with performance-based remuneration could help to address some of the shortcomings of public job placement services in the past.

As many job-seekers lack required soft skills and on-the-job experience, training courses should be combined with soft-skill training and practical internships (Figure 3.14). A good mix of technical and soft skills such as communication and teamwork becomes increasingly important with rising international integration and digitalisation of the economy (OECD, 2019[22]; Grundke et al., 2018[27]; OECD, 2017[49]). For workers that have not completed basic primary education, the government should expand the provision of basic adult education courses as a prerequisite for accessing vocational training courses, as discussed below.

Finally, improving management and organisational skills in small and medium size firms can be a very effective way to increase human capital investments and on the job training for workers (Dutz, 2018[34]). It also has potential to significantly raise the productivity of firms in Brazil and contribute to the reduction of informality. Many small and medium size firms in Brazil suffer from weak management practices and underinvestment in human capital of their staff (Bloom and Reenen, 2010[54]). Subsidising targeted training for managers of small firms can be a suitable policy tool to improve human resource management and prepare low-productivity firms for rising domestic and international competition (Dutz, 2018[34]). On a small scale, the government has already launched such a programme, whose impact should be carefully evaluated before upscaling it.

Beyond publicly subsidised training programmes, it will also be important to improve the incentives for private training provision by employers. At present, the design of the unemployment insurance scheme FGTS creates incentives for high job turnover, which in turn reduces the benefits for firms to invest in the skills of their staff. Informality is another factor for low investment of firms in training of staff (Dutz, 2018[34]). These structural factors holding back private training provision should be addressed by rethinking current unemployment insurance schemes as well as labour taxes and regulations, as discussed in Chapter 1 of this Survey.

While professional training is an effective way to reach those already in the labour market and whose basic skills allow a successful upgrading of specific job-related capabilities, the education system plays a fundamental role in preparing current and future generations for the challenges that international integration and rising digitalisation will bring about. Automation and offshoring may lead to the redundancy of certain tasks or occupations, which might require workers to change occupations and pursue new formal education degrees at a later stage of their careers (Bechichi et al., 2019[29]; Hummels et al., 2012[19]; Autor et al., 2014[16]; Hyman, 2018[18]). Therefore, it is crucial to lay good fundamentals in cognitive and social skills already early in life to facilitate learning and adaptation at a later stage (OECD, 2019[22]; Heckman and Mosso, 2014[55]; Heckman, Pinto and Savelyev, 2013[56]; Heckman et al., 2010[57]).

During the last decades, Brazil has strongly increased education spending and improved access to basic education. In many poor and remote municipalities, school infrastructure significantly improved and many new teachers were hired (IBGE, 2019[58]). Enrolment rates have strongly increased for early childhood, primary and secondary education (OECD, 2019[59]). The share of young adults having completed upper secondary education reached 46% in 2018, which is 20 percentage points higher than for older workers and at par with the OECD average (Figure 3.15). The share of young adults having completed tertiary education has also increased compared to the older generation, but is still 20 percentage points below the OECD average (OECD, 2019[59]).

Rising access to education has not been accompanied by increases in the quality of education. Comparing PISA test scores of 15-year-old students, Brazil ranks lower than other Latin American countries or the OECD average, although its education spending is with 6.2% of GDP even higher than in the average OECD country (see Chapter 1). Since 2009, test scores for reading, mathematics and science have not improved much relative to the OECD average (Figure 3.16). Moreover, dropout rates in secondary education are still higher than 30% (OECD, 2019[59]) . This is related to large heterogeneity between schools in terms of teaching quality, a curriculum that has focused on academic content and very little on vocational education and training, and a lack of basic cognitive and social skills among children from low-income families (INSPER, 2017[60]).

In recent years, significant steps were taken to improve the quality of basic education, but challenges related to implementation remain. Nation-wide common learning standards, which comprise a rich set of cognitive and socio-emotional skills and hold up to international best practices, have been introduced for early-childhood, primary and secondary education. The new common core for upper secondary education introduces more flexibility in course choice including a vocational training track, which has strong potential to reduce dropout rates (INSPER, 2017[60]). The new standards also increase minimum schooling duration from four to five hours per day. However, due to large heterogeneity in resources and technical capacities across municipalities and states, the implementation of the new nation-wide standards will be challenging.

A highly decentralised education system combined with large economic disparities has led to high variation of education quality and outcomes across municipalities (IBGE, 2019[58]; Todos pela educação, 2018[61]). The constitutional authority for early-childhood and primary education until the 5th grade lies in the hands of over 5000 municipalities, which are responsible for curriculum design and implementation, teacher selection and training, and design of learning material within their jurisdiction. Although a federally funded redistribution fund called FUNDEB has mitigated funding inequalities to a certain extent, resources and technical capacities are still low in many municipal school networks. Until the recent establishment of nation-wide standards, many municipalities did not have systematic learning standards, teachers have not been trained to properly implement existing standards and there was little coordination across municipalities. For the new standards to be effective, the federal and state governments need to take the lead and coordinate the implementation of standards and the common national curriculum into teaching practices in municipal and state school networks. They should continue to support local governments in teacher training and design of learning materials. As a first step, national guidelines for teacher training and the national textbook programme have been aligned with the content of the common national curriculum.

Coordination has also been weak across different levels of government and led to high spending inefficiencies in secondary education (Economistas do Brasil, 2018[62]). The constitutional authority for lower secondary education from the 6th to 9th grade is split between municipalities and states. School networks co-exist in many communities without any coordination or pooling of resources to use synergies (Todos pela educação, 2018[61]). States are responsible for upper secondary education from the 10th to 12th grade, and there has been little coordination in learning standards and their implementation with lower secondary municipal school, nor with school networks in other states. Moreover, in many poor communities, the existing school infrastructure and teaching staff is insufficient to comply with the new mandated minimum schooling hours or to provide the range of subject choices foreseen in the new learning standards for upper secondary education (Todos pela educação, 2018[61]). Although the new standards are an important step, implementing them will require continued coordination among different levels of government.

The quality of education will only improve with better selection and training of teachers. Although incentive based payment might increase motivation, these incentives will be ineffective if teachers do not have the skills they need for the job (Todos pela educação, 2018[61]). Teachers are located at the lower end of the skill distribution of tertiary education graduates in Brazil. This is related to the fact that rapidly increasing access to basic education was only possible with an accelerated hiring of teachers combined with weaker selection criteria. Moreover, relatively low wages in the teaching profession have contributed to the self-selection of lower-skilled graduates (IBGE, 2019[58]). The allocation of additional resources to under-funded municipal school networks through the new FUNDEB will help attracting higher-skilled graduates to the teaching profession.

Moreover, recruitment procedures for teachers should be standardised nationwide including common tests and certifications, possibly mimicking existing recruitment procedures for public employees. The recruitment system could classify candidates according to several criteria to improve the matching of teachers to open positions with differing skill requirements. Skill needs vary widely across school types, such as early-childhood, primary or secondary education, but also according to the socio-economical background of children.

Besides selecting the right candidates, teacher training needs to improve and adapt to the new learning standards. The national curriculum guidelines for initial teacher training for basic education, approved in December 2019, are a step into this direction. Curriculums for teacher careers at many universities include very little pedagogical knowledge or practical experience in classrooms (INSPER, 2017[60]; Todos pela educação, 2018[61]). Once young graduates have been hired, they face highly heterogeneous levels of counselling and training on the job. Most graduates do not get any support during their first years in school and there is no regular evaluation of their performance. Weak management of schools is related to frequent political appointments of school principals, which strongly influences the quality of teacher training and motivation (INSPER, 2017[60]; Todos pela educação, 2018[61]).

Besides teaching quality, socio-economic background and good early-childhood education are the main determinant for educational outcomes (OECD, 2019[63]; Heckman and Mosso, 2014[55]). In particular for young children, the access to food, clean drinking water and good health services is crucial for the development of cognitive and social skills (Heckman, Pinto and Savelyev, 2013[56]; Heckman et al., 2010[57]). Thus, recent increases in enrolment rates for early childhood education should be complemented with continued efforts to eradicate extreme poverty, raise access to universal health care and improve municipal water and sanitation services (IBGE, 2016[64]). Moreover, improving internet access for low-income households is key to enable distance learning, which during the school closures due to the Covid19 pandemic has been the only way to participate in schooling activities and will play an increasing role in skills development throughout life (OECD, 2019[22]).

Shifting more resources to the successful conditional cash-transfer program Bolsa Família would be one way to improve access to food and health care for many poor children. Food could be directly provided in schools to ensure nutrition quality. Eventually, the programme could also be linked to enrolment in early childhood education or home visits of teachers consulting parents on education practices. When allocating scarce spaces in early childhood education, preference should be given to low-income households and single mothers. Only 15% of poor families with children below 3 years have access to child-care, compared to 40% of the more affluent families (World Bank, 2016[65]).

Due to low access to basic education until some decades ago, a large share of adults has not completed primary education (Figure 3.17). The share of older workers who cannot read or write is still large compared to other countries (IBGE, 2019[58]). Although enrolment rates of adults into formal education are higher than in the average OECD country, the quality of basic education for adults is low and many drop out before graduating (Todos pela educação, 2018[61]; IBGE, 2019[58]). There is no separate targeted system of adult education. Adapted learning standards, curriculums or learning material as well as specific teacher training or selection do not exist. In most municipalities, the same teachers who teach children in basic education may be teaching adults in their second or third day shift. The supply of specific facilities for adult education is low and often classes take place in regular schools or other municipal buildings.

As opening up the economy will require many low-skilled workers to find new jobs in other firms or economic sectors, it is key to reform the current system and increase its funding. Current transfers for adult education from the federal government to municipalities are not conditional and mostly spend on basic education of children. The government should consider including basic education of adults in the federal top-up mechanism for education funding FUNDEB and combine funding allocation with the same type of quality improving incentives for municipalities and states discussed in Chapter 1. To inform education policies on skill needs of adults, Brazil could consider participating in the OECD Survey of Adult Skills (PIAAC).

As the share of adults without completed primary education is particularly large among older and poorer parts of the population (Figure 3.18), it is key that learning standards, curriculums and teaching material as well as teacher selection and training are specifically adapted to these target groups (IBGE, 2019[58]). This requires a concerted national effort and a strong coordination among different levels of government. To support poor households and facilitate school attendance of working adults, increasing funding for Bolsa Família could be combined with requirements for parents to attend basic education. This could even generate synergies for poor children, whose parents might be in better conditions to support them in their educational career.

For many students, the strong focus of secondary curriculums on academic subjects and a lack of opportunities to select courses according to their interest or ability may be one factor behind frequent class repetition and school dropout (Figure 3.19). Professional training opportunities within secondary education are currently scarce. Expanding them could help to reduce high dropout rates (INSPER, 2017[60]).

A 2017 education reform that established nation-wide learning standards and curriculums for upper secondary education laid the grounds for offering more vocational content in secondary education. Mandatory subjects are reduced to 1800 hours per year. For the remaining 1200 hours, student can choose among five different paths, including mathematics, science, languages, humanities and vocational education. This newly created flexibility is a major step forward, but challenges related to implementation remain. Many states and municipalities do not have the necessary resources and capacities to implement all five pathways (Todos pela educação, 2018[61]; Almeida and Packard, 2018[66]). Both physical school infrastructure and the availability of adequately trained teaching staff are severe bottlenecks, which will require additional resources and better coordination across different levels of government (see Chapter 1). For example, idle facilities from federal and state institutes that had expanded their training capacities in recent years under the umbrella of the PRONATEC programme could be used for vocational education classes.

For the new vocational education path in upper secondary education to be successful, the close cooperation and involvement of the private sector in design and delivery of workplace training is key. So far, vocational education is mainly taking place at post-secondary level and most state and federal institutes for vocational education do not offer much practical experience (OECD, 2019[59]; OECD, 2018[45]). The semi-public network of training providers “Sistema S”, including SENAI for the manufacturing sector, are a notable exception. Sistema S institutions have developed strong ties with the private sector and offer high quality vocational education, including practical experience in firms or training laboratories (OECD, 2018[45]). Their courses are linked to skill demands in the private sector and course content is regularly updated using skill anticipation assessment and graduate surveys. However, there is very little cooperation with high schools and most students start their vocational education after completing high school. Recent efforts to include “Sistema S” training providers in the design and implementation of the new vocational education path in secondary education are promising and should continue.

Vocational education institutes close to the private sector are reluctant to cooperate with high schools due to excessively restrictive work safety and health standards, which complicate practical vocational education for minors in firms or at machines in laboratories (Silva, Almeida and Strokova, 2015[33]; Almeida and Packard, 2018[66]). Uncertainty around judicial decisions have led to closedowns of private-sector institutes with integrated vocational education in upper secondary education, such as one previously operated by the car manufacturer Volkswagen. These labour safety and health standards should be simplified to facilitate cooperation between high schools and the private sector in the implementation of the new vocational education path in secondary education.

The experience of the United Kingdom may also serve as inspiration for providing monetary incentives for firms to offer apprenticeships. In the UK, a specific apprenticeship levy of 0.5% of total payroll is collected from firms with a wage bill of more than USD 4 Million (NAO, 2019[67]). These firms can use their contribution to select coursework or other instructional services related to apprenticeships from a dedicated digital portal, and receive a 10% top up from the government. For smaller firms not paying the levy, access to these courses requires a co-financing of 5% of the costs. Apprentices must spend at least 20% of their paid hours doing off-the-job formal training. However, a wider use of apprenticeships in Brazil would likely require addressing negative training incentives related to high job turnover and the policy distortions behind it, as discussed above. The experience of Germany shows that the main motivation for firms to offer apprenticeships is to secure its future supply of skilled workers (Grollmann et al., 2016[68]).

Access to tertiary education has increased and 20% of young adults had a completed tertiary degree in 2018, but this is still over 20 percentage points below the OECD average ( (OECD, 2019[59])). Public tertiary education is virtually free in Brazil, but the selective admission tests put children from public schools at a disadvantage vis-à-vis graduates from private schools, where the quality of secondary education is generally higher (IBGE, 2019[58]). This makes spending on tertiary education largely regressive. Recently introduced admission quotas have been successful in raising the share of students from disadvantaged groups of the populations and should continue. Another alternative would be to introduce means-tested university fees with ample scope for grants for students from low-income households.

The most effective way to improve the opportunities for students from low-income families in tertiary education is to prepare them better before they graduate from secondary school. This suggests that the current focus could be shifted from relatively high spending on tertiary education towards earlier levels of education (Figure 3.20). Tertiary education also suffers from quality shortcomings and low completion rates (Economistas do Brasil, 2018[62]). More than 50% of students have not completed their degree three years after the end of the theoretical duration (Figure 3.20).

Reducing dropouts in tertiary education would require strengthening the alignment of university curriculums to the type of occupations and skills demanded by the labour market (Economistas do Brasil, 2018[62]; OECD, 2019[59]). Many firms in Brazil indicate that university graduates do not have the technical and social skills they need (Schwab, 2019[69]). Skill demand is particularly high in science, technology, engineering and mathematics (STEM), but only 13% of tertiary graduates study these fields, compared to 20% in the OECD average or 26% in Mexico ( (OECD, 2019[59])). The federal government should consider linking admission rates for university subjects to a nation-wide skill anticipation assessment system, which would indicate which type of occupations are in high demand in the labour market. Curriculums should be regularly updated and linked to changing skill requirements and task content of occupations.

Student choices could also be guided by transparent assessments and certifications of tertiary education courses, to inform high school graduates about the quality and the content of these courses (Economistas do Brasil, 2018[62]). This could be combined with career counselling services for high school students to well align interest and abilities with the content of selected tertiary education courses.


[66] Almeida, R. and T. Packard (2018), Skills and Jobs in Brazil: An Agenda for Youth, World Bank, Washington, DC, https://doi.org/10.1596/978-1-4648-1293-4.

[6] Amiti, M. and A. Khandelwal (2013), “Import Competition and Quality Upgrading”, The Review of Economics and Statistics, Vol. 95/2, pp. 476-490, https://www.mitpressjournals.org/doi/pdf/10.1162/REST_a_00271 (accessed on 22 September 2018).

[2] Amiti, M. and J. Konings (2007), “Trade Liberalization, Intermediate Inputs, and Productivity: Evidence from Indonesia”, American Economic Review, Vol. 97/5, pp. 1611-1638, https://doi.org/10.1257/aer.97.5.1611.

[4] Amiti, M., S. Redding and D. Weinstein (2019), “The Impact of the 2018 Trade War on U.S. Prices and Welfare”, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w25672.

[40] Andrews, D., C. Criscuolo and P. Gal (2015), “Frontier Firms, Technology Diffusion and Public Policy: Micro Evidence from OECD Countries”, OECD Productivity Working Papers, No. 2, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jrql2q2jj7b-en.

[11] Araújo, B. and L. Paz (2014), “The effects of exporting on wages: An evaluation using the 1999 Brazilian exchange rate devaluation”, Journal of Development Economics, Vol. 111, pp. 1-16, https://doi.org/10.1016/j.jdeveco.2014.07.005.

[23] Arntz, M., T. Gregory and U. Zierahn (2016), “The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis”, OECD Social, Employment and Migration Working Papers, No. 189, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jlz9h56dvq7-en.

[25] Autor, D. and D. Dorn (2013), “The growth of low-skill service jobs and the polarization of the US Labor Market”, American Economic Review, Vol. 103/5, pp. 1553-1597, https://doi.org/10.1257/aer.103.5.1553.

[37] Autor, D., D. Dorn and G. Hanson (2013), “The China syndrome: Local labor market effects of import competition in the United States”, American Economic Review, Vol. 103/6, pp. 2121-2168, https://doi.org/10.1257/aer.103.6.2121.

[16] Autor, D. et al. (2014), “Trade Adjustment: Worker-Level Evidence *”, The Quarterly Journal of Economics, Vol. 129/4, pp. 1799-1860, https://doi.org/10.1093/qje/qju026.

[28] Bechichi, N. et al. (2018), “Moving between jobs: An analysis of occupation distances and skill needs”, OECD Science, Technology and Innovation Policy Papers, No. 52, OECD, Paris, http://www.oecd.org/going-digital (accessed on 27 September 2018).

[29] Bechichi, N. et al. (2019), “Occupational mobility, skills and training needs”, http://www.oecd.org/going-digital. (accessed on 26 January 2020).

[21] Becker, S., K. Ekholm and M. Muendler (2013), “Offshoring and the onshore composition of tasks and skills”, Journal of International Economics, Vol. 90/1, pp. 91-106, https://doi.org/10.1016/j.jinteco.2012.10.005.

[20] Becker, S. and M. Muendler (2015), “Trade and tasks: an exploration over three decades in Germany”, Economic Policy, Vol. 30/84, pp. 589-641, https://doi.org/10.1093/epolic/eiv014.

[54] Bloom, N. and J. Reenen (2010), “Why do management practices differ across firms and countries?”, Journal of Economic Perspectives, Vol. 24/1, pp. 203-224, https://doi.org/10.1257/jep.24.1.203.

[46] Blyde, J. et al. (2019), “Training and Labor Adjustment to Trade”, IDB WORKING PAPER SERIES, No. Nº IDB-WP-01069, Inter-American Development Bank, Washington, D.C., https://doi.org/10.18235/0001970.

[48] Brown, A. and J. Koettl (2012), “Active Labor Market Programs: Employment Gain or Fiscal Drain?”, IZA Discussion papers, No. 6880, IZA.

[41] Cadot, O., J. Gourdon and F. van Tongeren (2018), “Estimating Ad Valorem Equivalents of Non-Tariff Measures: Combining Price-Based and Quantity-Based Approaches”, OECD Trade Policy Papers, No. 215, OECD Publishing, Paris, https://dx.doi.org/10.1787/f3cd5bdc-en.

[35] Cavgias, A. et al. (2019), Edited Democracy: Media Manipulation and the News Coverage of Presidential Debates., Working Papers, Department of Economics 2019_17, University of São Paulo (FEA-USP).

[10] Criscuolo, C., P. Gal and C. Menon (2014), “The Dynamics of Employment Growth: New Evidence from 18 Countries”, OECD Science, Technology and Industry Policy Papers, No. 14, OECD Publishing, Paris, https://dx.doi.org/10.1787/5jz417hj6hg6-en.

[7] De Loecker, J. et al. (2016), “Prices, Markups, and Trade Reform”, Econometrica, Vol. 84/2, pp. 445-510, https://doi.org/10.3982/ECTA11042.

[36] Dix-Carneiro, R. and B. Kovak (2017), “Trade liberalization and regional dynamics”, American Economic Review, Vol. 107/10, pp. 2908-2946, https://doi.org/10.1257/aer.20161214.

[34] Dutz, M. (2018), Jobs and Growth: Brazil’s Productivity Agenda, The World Bank, https://doi.org/10.1596/978-1-4648-1320-7.

[62] Economistas do Brasil (2018), Carta Brasil - Um conjunto de propostas para o novo governo, https://iepecdg.com.br/wp-content/uploads/2018/11/Carta-BRASIL-01.11.2018logo.pdf.

[31] Fiorini, M., M. Sanfilippo and A. Sundaram (2019), “Roads: From Trade Liberalization to Firm Productivity”, Robert Schuman Centre for Advanced Studies Research Paper No. RSCAS, No. 32, https://doi.org/10.2139/ssrn.3393548.

[44] Gathmann, C., I. Helm and U. Schönberg (2018), “Spillover effects of mass layoffs”, Journal of the European Economic Association, https://doi.org/10.1093/jeea/jvy045.

[15] Goldberg, K. and N. Pavcnik (2007), Distributional Effects of Globalization in Developing Countries, https://pubs.aeaweb.org/doi/pdfplus/10.1257/jel.45.1.39 (accessed on 14 September 2018).

[3] Goldberg, P. et al. (2009), “Trade Liberalization and New Imported Inputs”, American Economic Review, Vol. 99/2, pp. 494-500, https://doi.org/10.1257/aer.99.2.494.

[68] Grollmann, P. et al. (2016), Duale Ausbildung als betriebliche Strategie der Fachkräftesicherung – Fallstudien zu Motivation und Organisation im internationalen Vergleich, Bundesinstitut fuer Berufsbildung - Bonn, https://www.bibb.de/tools/dapro/data/documents/pdf/zw_15312.pdf.

[5] Grundke, R. and J. Arnold (2019), “Fostering Argentina’s integration into the world economy”, OECD Economics Department Working Papers, No. 1572, OECD Publishing, Paris, https://dx.doi.org/10.1787/7ed95b2b-en.

[39] Grundke, R. et al. (2020), “Brazils’ trade liberalisation of the 1990s revisited: How targeted training can reduce adjustment costs for workers”, OECD Economics Department Working Papers.

[27] Grundke, R. et al. (2018), “Which skills for the digital era?: Returns to skills analysis”, OECD Science, Technology and Industry Working Papers, No. 2018/09, OECD Publishing, Paris, https://dx.doi.org/10.1787/9a9479b5-en.

[57] Heckman, J. et al. (2010), “The Rate of Return to the High/Scope Perry Preschool Program”, Journal of Political Economy, Vol. 94/1-2, pp. 114-128.

[55] Heckman, J. and S. Mosso (2014), “The Economics of Human Development and Social Mobility”, Annual Review of Economcs, Vol. 6, pp. 689-733, https://doi.org/10.1146/annurev-economics-080213-040753.

[56] Heckman, J., R. Pinto and P. Savelyev (2013), “Understanding the mechanisms through which an influential early childhood program boosted adult outcomes”, American Economic Review, Vol. 103/6, pp. 2052-2086, https://doi.org/10.1257/aer.103.6.2052.

[12] Hsieh, C. and P. Klenow (2009), “Misallocation and Manufacturing TFP in China”, The Quarterly Journal of Economics, Vol. 124/4, pp. 1403-1448, https://www.jstor.org/stable/pdf/40506263.pdf (accessed on 17 January 2019).

[19] Hummels, D. et al. (2012), “Offshoring, Transition, and Training: Evidence from Danish Matched Worker-Firm Data”, American Economic Review: Papers & Proceedings, Vol. 102/3, pp. 424–428, https://doi.org/10.1257/aer.102.3.424.

[17] Hummels, D., J. Munch and C. Xiang (2018), Offshoring and labor markets, American Economic Association, https://doi.org/10.1257/jel.20161150.

[18] Hyman, B. (2018), “Can Displaced Labor Be Retrained? Evidence from Quasi-Random Assignment to Trade Adjustment Assistance”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.3155386.

[70] Hyman, B. (2018), “Can Displaced Labor Be Retrained? Evidence from Quasi-Random Assignment to Trade Adjustment Assistance *”.

[58] IBGE (2019), Síntese de Indicadores Sociais - Uma análise das condições de vida da população brasileira 2019, https://biblioteca.ibge.gov.br/visualizacao/livros/liv101629.pdf.

[64] IBGE (2016), Pesquisa nacional de saúde do escolar 2015, Instituto Brasileiro de Geografia e Estadisticas, https://biblioteca.ibge.gov.br/visualizacao/livros/liv97870.pdf.

[60] INSPER (2017), Políticas públicas para redução do abandono e evasão escolar de jovens, http://gesta.org.br/wp-content/uploads/2017/09/Politicas-Publicas-para-reducao-do-abandono-e-evasao-escolar-de-jovens.pdf.

[32] Lyon, S. and M. Waugh (2018), “Redistributing the gains from trade through progressive taxation”, Journal of International Economics, Vol. 115, pp. 185-202, https://doi.org/10.1016/j.jinteco.2018.09.008.

[8] Melitz, M. (2003), The Impact of Trade on Intra-Industry Reallocations and Aggregate Industry Productivity, https://www.jstor.org/stable/pdf/1555536.pdf?refreqid=excelsior%3Ae3382ca8aee44f583882d975e0cdb85e (accessed on 22 September 2018).

[67] NAO (2019), The apprenticeships programme, National Audit Office, https://www.nao.org.uk/wp-content/uploads/2019/03/The-apprenticeships-programme.pdf.

[42] O’Connell, S. et al. (2017), “Can Business Input Improve the Effectiveness of Worker Training? Evidence from Brazil’s Pronatec-MDIC”, Policy Research Working Paper, No. 8155, World Bank, Washington DC, http://econ.worldbank.org. (accessed on 19 February 2020).

[47] O’Connell, S. et al. (2017), “Can business input improve the effectiveness of worker training? evidence from Brazil’s Pronatec-MDIC”, Policy Research Working Paper, No. WPS8155, World Bank, http://documents.worldbank.org/curated/en/444871501522977352/Can-business-input-improve-the-effectiveness-of-worker-training-evidence-from-Brazils-Pronatec-MDIC (accessed on 5 October 2018).

[59] OECD (2019), Education at a Glance, https://doi.org/10.1787/f8d7880d-en.

[52] OECD (2019), Financial Incentives to Promote Adult Learning in Australia, Getting Skills Right, OECD Publishing Paris, https://doi.org/10.1787/c79badcc-en.

[51] OECD (2019), Getting Skills Right: Future-Ready Adult Learning Systems, Getting Skills Right, OECD Publishing Paris, https://doi.org/10.1787/9789264311756-en.

[53] OECD (2019), OECD Economic Surveys: Malaysia, OECD Publishing Paris, https://doi.org/10.1787/eaaa4190-en.

[22] OECD (2019), OECD Skills Outlook 2019 : Thriving in a Digital World, OECD Publishing, Paris, https://dx.doi.org/10.1787/df80bc12-en.

[63] OECD (2019), PISA 2018 Results (Vol II): Where all students can succeed., OECD Publishing Paris, https://doi.org/10.1787/b5fd1b8f-en.

[1] OECD (2018), Fostering Brazil’s integration into the world economy, OECD Publishing, https://doi.org/10.1787/eco_surveys-bra-2018-5-en (accessed on 7 February 2020).

[45] OECD (2018), Getting Skills Right: Brazil, Getting Skills Right, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264309838-en.

[43] OECD (2018), Getting Skills Right: Brazil,, https://doi.org/10.1787/9789264309838-en.

[24] OECD (2018), Job Creation and Local Economic Development 2018: Preparing for the Future of Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264305342-en.

[49] OECD (2017), OECD Skills Outlook 2017: Skills and Global Value Chains, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264273351-en.

[50] OECD (2016), Understanding social impact bonds, OECD, Paris, http://www.oecd.org/cfe/leed/UnderstandingSIBsLux-WorkingPaper.pdf (accessed on 19 February 2020).

[9] Pavcnik, N. (2002), “Trade Liberalization, Exit, and Productivity Improvements: Evidence from Chilean Plants”, The Review of Economic Studies, Vol. 69/1, pp. 245-276, https://doi.org/10.1111/1467-937X.00205.

[30] Rusticelli, E. et al. (2018), “Going local: a regional perspective on how trade affects labour markets and inequality”, OECD Economics Department Working Papers, No. 1530, OECD Publishing, Paris, https://dx.doi.org/10.1787/2981f94c-en.

[69] Schwab, K. (2019), The Global Competitiveness Report 2019, World Economic Forum, http://www3.weforum.org/docs/WEF_TheGlobalCompetitivenessReport2019.pdf (accessed on 19 February 2020).

[33] Silva, J., R. Almeida and V. Strokova (2015), Sustaining Employment and Wage Gains in Brazil: A Skills and Jobs Agenda, The World Bank, https://doi.org/10.1596/978-1-4648-0644-5.

[26] Spitz-Oener, A. (2006), “Technical Change, Job Tasks, and Rising Educational Demands: Looking outside the Wage Structure”, Journal of Labor Economics, Vol. 24/2, pp. 235-270, https://doi.org/10.1086/499972.

[61] Todos pela educação (2018), Educação ja - Uma proposta suprapartidária de estratégia para a Educação Básica brasileira e prioridades para Governo Federal em 2019-2022.

[38] Topalova, P. (2010), “Factor immobility and regional impacts of trade liberalization: Evidence on poverty from India”, American Economic Journal: Applied Economics, Vol. 2/4, pp. 1-41, https://doi.org/10.1257/app.2.4.1.

[13] Topalova, P. and A. Khandelwal (2011), “Trade Liberalization and firm productivity: The case of India”, The Review of Economics and Statistics, Vol. 93/3, pp. 995-1009, https://doi.org/10.1162/REST_a_00095.

[14] Winters, L., N. Mcculloch and A. Mckay (2004), “Trade Liberalization and Poverty: The Evidence so Far”, Journal of Economic Literature, Vol. 42/1, pp. 72-115, https://www.jstor.org/stable/pdf/3217037.pdf?refreqid=excelsior%3Afb2fd700a65867707fddc544b380482d (accessed on 14 September 2018).

[65] World Bank (2016), Brazil Systematic Country Diagnostic: Retaking the path to Inclusion, Growth and Sustainability, The World Bank Group.


This document, as well as any data and map included herein, are without prejudice to the status of or sovereignty over any territory, to the delimitation of international frontiers and boundaries and to the name of any territory, city or area.

The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

Note by Turkey
The information in this document with reference to “Cyprus” relates to the southern part of the Island. There is no single authority representing both Turkish and Greek Cypriot people on the Island. Turkey recognises the Turkish Republic of Northern Cyprus (TRNC). Until a lasting and equitable solution is found within the context of the United Nations, Turkey shall preserve its position concerning the “Cyprus issue”.

Note by all the European Union Member States of the OECD and the European Union
The Republic of Cyprus is recognised by all members of the United Nations with the exception of Turkey. The information in this document relates to the area under the effective control of the Government of the Republic of Cyprus.

Photo credits: Cover © f11photo/Shutterstock.com.

Corrigenda to publications may be found on line at: www.oecd.org/about/publishing/corrigenda.htm.

© OECD 2020

The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.