OECD Economic Surveys: Belgium 2022

Report

OECD Economic Surveys: Belgium

14 June 2022

Available in:

English
français

Download PDF

2. Improving economic opportunities for all

Nicolas Gonne
OECD

Abstract

Income inequality is low in Belgium, and intergenerational income mobility is on par with the average OECD economy. However, as in other OECD countries, there is scope to improve equal access to opportunities across the population. Poverty risks are high for the unemployed and the low-skilled. Vulnerable socio-demographics, in particular the low educated, single mothers and people with a migrant background and with disabilities have persistently low incomes. Moreover, low-income households are overburdened by housing costs. To foster upward income mobility, employment should be increased among vulnerable groups by enhancing skills through life-long learning, effective career guidance and continuing to strengthen work incentives. To prevent the transmission of disadvantages across generations, social segregation in compulsory education should be addressed, in particular through better-designed school choice policies, higher mobility between general and vocational tracks, and stronger incentives and training for teachers. Promoting quality and affordable housing is also necessary to reduce spatial segregation and mitigate barriers to opportunity.

Belgium has low income inequality overall, thanks to extensive tax and transfer policies and strong institutionalised social dialogue. The Gini coefficient is one of the lowest in the OECD, comparable to Nordic countries and below neighbouring countries, such as France, Germany and the Netherlands. Moreover, the overall share of the population at risk of poverty or social exclusion at 18.9% in 2020 is lower than the EU average of 22%, with regional disparities.

Intergenerational income mobility is also high in international comparison, i.e., the extent to which people’s income depends on that of their parents is less than the OECD average. Around 35% of the earnings differences between fathers carry over to the next generation in Belgium, below 50% in France or Germany (OECD, 2018[1]). On average, the offspring of a low-income family could reach Belgium’s mean income level in around 100 years (about four generations), a slow process, yet faster than in many OECD countries (Figure 2.1, Panel A). Belgium also does well in international comparison in other dimensions that help to lower income inequality, like access to healthcare and education (WEF, 2020[2]; Eurofound, 2017[3]).

However, the policies that deliver low inequality and good intergenerational mobility do not necessarily ensure equality of opportunities, i.e., access to the same life chances irrespective of initial life conditions. Indeed, Belgium’s good overall performance regarding income distribution and intergenerational mobility hides an uneven distribution of economic opportunities. Considerable disparities exist across groups according to, notably, parental background and the country of origin. For example, the offspring of non-EU citizens, low-educated or unemployed parents, and tenants are significantly more at risk of poverty or social exclusion (Figure 2.1, Panel B). Disparities also exist across occupations, as children of manual workers are 11.9 percentage points more likely to be manual workers than managers, a gap close to the OECD average but twice as large as in France or Germany (OECD, 2018[1]).

Figure 2.1. Relatively high intergenerational income mobility overall masks unequal opportunities

Unequal access to economic opportunities implies low intra-generational income mobility for members of vulnerable groups, i.e., their income changes little over their life course. Income stability over the life course partly reflects the existence of strong social security safety nets in Belgium, like in other countries, such as the Netherlands or the Nordic countries. However, such persistence can also point to structural impediments to improving one’s initial socio-economic condition. While data restrictions prevent a full analysis of income over the life cycle (Box 2.1), existing measures suggest that Belgium overall is characterised by strong intra-generational income persistence. For example, the so-called permanent income inequality, i.e., income inequality pooled over several years, is only slightly lower than income inequality at any point in time (Figure 2.2, Panel A).

Like several other OECD countries, Belgium is a country of “sticky floors and ceilings”, where income persistence is not distributed evenly across groups. While around half of the working age population remained in the same quintile of the income distribution over the period 2011-14, close to the OECD average, persistence is significantly and increasingly stronger for high- and low-income groups than for middle-income groups. Almost 64% of working age individuals in the bottom income quintile and 70% of those in the top income quintile remain in the same quintile over a four-year period (Figure 2.2, Panel B).

Upward income mobility is very low for low-income groups in Belgium and, at the same time, lower-middle income households are at a relatively high risk of sliding down to the bottom (Figure 2.2, Panel C). In Belgium, the unemployed, single-parent families and tenants are the most vulnerable to poverty (StatBel, 2019[4]). Economic opportunities are particularly weak for people with a migrant background. For example, the probability to be employed and the chances of upward occupational mobility are relatively low for the native-born with foreign parents (Figure 2.2, Panel D).

Differences between socio-demographic groups tend to persist across generations in OECD countries. Parents’ income is a key determinant of access to good-quality education, adequate healthcare and professional networks and, hence, of children’s career options and income (OECD, 2018[1]). Such uneven opportunities can entrench inequalities at significant economic cost, in particular as talents and human capital investments are missed out (OECD, 2015[5]). Ensuring that income is not stuck over time or transmitted across generations can be socially desirable, in particular in cases where initial conditions reflect discrimination against certain groups or are the results of cultural, ethnic or family backgrounds. Yet, attenuating frequent and uncertain swings in disposable income can also be desirable to reduce economic insecurity, e.g., through social insurance.

Box 2.1. Measuring income mobility: data availability

Measuring income mobility across a lifetime requires longitudinal data, i.e., repeated information for a consistent group of individuals over a relatively long period. Such data are rare as they are costly to collect, and they include caveats, in particular attrition bias. Moreover, some relevant survey variables are collected as part of thematic modules that are discontinued.

This chapter relies extensively on data from the 2018 OECD report, A Broken Social Elevator? How To Promote Social Mobility, a large-scale effort to pull together different data sources to provide a comprehensive, cross-country picture of social mobility. The data used at the time of publishing, which sometimes dates from the early 2010s, often remain the most relevant data to this day as the persistent and structural nature of income mobility patterns makes them still relevant for the analysis in this chapter.

Source: OECD (2018), A broken social elevator? How to promote social mobility.

Beyond lowering welfare and well-being of individuals, unequal opportunities weigh on key macroeconomic outcomes, namely potential growth, public finances and social cohesion. First, unequal access to economic opportunities contributes to misallocating or underusing talent and resources, in particular human capital, thereby dragging on productivity growth (OECD, 2018[6]). This is especially important in the context of high and rising labour market shortages in Belgium (Chapter 1). Second, persistently low incomes for some groups require extensive redistribution, thereby worsening medium-term fiscal sustainability challenges (Chapter 1). Third, low prospects of accessing opportunities (or the perception thereof) tends to decrease democratic participation and trust (OECD, 2018[1]; OECD, 2021[7]), a concern of particular relevance in Belgium, where trust in government experienced the strongest deterioration among OECD countries since 2007 (OECD, 2021[8]).

Figure 2.2. Intra-generational income mobility is relatively low and uneven across groups

This chapter reviews Belgium’s capacity to provide economic opportunities for all, irrespective of initial conditions, while continuing to promote low overall income inequality and economic security thanks to its outstanding social safety net. It identifies three key barriers, the labour market, education and housing, and discusses policies that can improve economic opportunities, especially concentrating on the situation of vulnerable groups, such as the low-skilled, people with a migrant background and single mothers. These include policies to: i) improve labour market transitions among vulnerable groups (e.g., lifelong learning); ii) attenuate the effect of socio-economic background on the educational attainment of disadvantaged students (e.g., teachers’ incentives); and iii) make quality housing affordable for low-income households (e.g., housing subsidies). As competencies concerning the labour market, education and housing are spread across different levels of government (Box 2.2), some recommendations related to non-federal competencies are more relevant to specific regions and communities according to their policy needs and priorities in various areas.

Box 2.2. Levels of government and competences regarding barriers to economic opportunities

Besides the federal government, the Belgian governance system comprises three regional authorities and three language communities (cutting across the regions), with significant autonomy and separate competencies. The regions refer to the Flemish, Brussels-Capital and the Walloon regions. The communities refer to the French, Flemish and German-speaking communities. Each authority has its own legislative and executive powers for its field of competences, and its own parliament and government to exercise these powers. The federal state, the regions and the communities are on an equal footing, so that no authority has precedence over another.

Competences concerning the barriers to economic opportunities identified in this chapter, namely the labour market, education and housing, are spread across different levels of government (Table 2.1).

	Federal	Regions	Communities
Labour market	Unemployment, pensions and health insurance	Active labour market policies, family benefits, health, social security contribution reduction for targeted groups
Education	Start and end age of compulsory education, minimal requirements for granting degrees, pension regimes for education staff	Parts of adult learning and apprenticeships	Pre-primary to tertiary and adult education, including granting of degree equivalence
Housing	Mortgage tax credit for non-owner occupied dwellings	Social housing and property taxes, mortgage tax credit of owner occupied dwellings, housing allowances

Federal

Regions

Communities

Labour market

Unemployment, pensions and health insurance

Active labour market policies, family benefits, health, social security contribution reduction for targeted groups

Education

Start and end age of compulsory education, minimal requirements for granting degrees, pension regimes for education staff

Parts of adult learning and apprenticeships

Pre-primary to tertiary and adult education, including granting of degree equivalence

Housing

Mortgage tax credit for non-owner

occupied dwellings

Social housing and property taxes, mortgage tax credit of owner occupied dwellings, housing allowances

Note: The table is not exhaustive as only the main policies and responsibilities are listed. Flanders merged community and regional institutions. The Walloon region transferred some of its competences concerning German speakers to the German-speaking Community, including public employment services (ADG).

Barriers to economic opportunities in Belgium

Low labour market transitions hinder upward income mobility

Participation and transitions in the labour market are key drivers of income mobility over the life course (Box 2.3). In particular, long-term inactivity or unemployment and low-wage work contribute to the persistence of low income (OECD, 2020[9]). Difficult school-to-work transitions also hinder upward income mobility as they lead to unemployment spells and precarious early careers, especially for the disadvantaged youth (OECD, 2018[1]). Moreover, non-standard forms of work raise new challenges, as they have the potential to increase labour market participation, but often offer limited access to certain forms of social protection, which could exacerbate existing disparities (OECD, 2019[10]).

Stronger labour market outcomes are key for upward income mobility in Belgium. Indeed, individuals transitioning into employment are six times more likely to experience a large income gain than individuals who remain non-employed in Belgium (Figure 2.3, Panel A). More than half of the people becoming employed obtain an income increase of at least 20% in Belgium, compared to about a third in neighbouring countries. However, transitions from unemployment or inactivity to employment are particularly low in Belgium (Figure 2.3, Panel B). The probability of becoming employed conditional on being unemployed is less than 30%, the lowest value in the European Union. The incidence of long-term unemployment is also associated with low-income persistence (Figure 2.3, Panel C).

Figure 2.3. There is room for upward income mobility through better labour market outcomes

Entrepreneurship offers upward mobility paths for people with the capability to identify new market opportunities and create value added. However, vulnerable groups are disproportionately affected by barriers to entrepreneurship, partly because necessary entrepreneurial skills and knowledge are often transmitted informally from parents to children or acquired in professional networks and alumni associations (OECD, 2018[1]). Belgium has a large share of “missing entrepreneurs”, i.e., individuals who would engage in entrepreneurial activity if the group to which they belong had the same early-stage entrepreneurship rate than that of men aged 30-49 (Figure 2.3, Panel D). Closing gaps in entrepreneurship activity across all socio-demographic groups would add 270 000 entrepreneurs to the existing 475 000, with women accounting for two-thirds of these missing entrepreneurs and migrants for 20% (OECD, 2020[11]).

Box 2.3. Barriers to upward income mobility in Belgium: new evidence from survey microdata

New OECD research uses the Household Finance and Consumption Survey (HFCS) for Belgium to estimate the correlation between labour market, education and housing characteristics of households and their mobility in the distribution of income, following the methodology in Martinez-Toledano et al. (2019[12]).

Preliminary results show that low education and joblessness are significantly associated with lower income mobility over the period 2011-17. Figure 2.4 shows the coefficients of a regression of the change of households’ position in the income distribution on dummy variables related to their labour market status, education level and housing tenure, controlling for age, income and wealth. Everything else equal, households of which the head is jobless will be about 15 percentiles lower in the income distribution after a three-year period, compared to a household of which the head either has a job or is retired. Similarly, households of which the head does not have tertiary education will be more than nine percentiles lower in the income distribution after three years than households of which the head attended tertiary education. Hence, policies that improve the labour market outcomes of vulnerable groups and enhance equal opportunities in education are key to boost upward income mobility.

Figure 2.4. Employment and education are significantly correlated with income mobility

Change in household income percentile position associated with selected determinants

Better school-to-work transitions are also necessary to improve upward income mobility prospects for the youth in Belgium. The proportion of people aged 15-24 not in education, employment or training (NEET) at 9.2% was lower than the EU average in 2020 according to Eurostat data, but significantly higher than in the Netherlands (4.5%) and Germany (7.3%), with regional differences. The proportion of NEET aged 15-24 is particularly large among those born outside of the EU at 18.1% in 2020. Moreover, more than one in five unemployed people aged 15-24 had been so for more than 12 months in 2020, compromising chances of upward mobility.

Inequalities of opportunity in compulsory education drag on intergenerational mobility

Education is a key factor affecting intergenerational income mobility (OECD, 2018[1]; OECD, 2018[13]). Children with better-educated parents obtain better educational outcomes, in part due to the investments that better-educated parents are able to make in their children. Therefore, the mere transmission of skills and educational preferences from parents to children contributes to intergenerational income persistence to the extent that earnings are correlated with educational attainment (Hanushek et al., 2021[14]; Blanden, Gregg and Macmillan, 2007[15]). While non-compulsory, early childhood education and care also have a large impact on social mobility (see below).

The education systems in Belgium feature high intergenerational persistence in attainment and strong links between students’ academic performance and socio-economic status. For example, the correlation between children’s and parents’ years of schooling is one of the strongest in the OECD, as an additional year of parental schooling is associated with more than half a year of additional schooling for their offspring (Figure 2.5, Panel A). The correlation is less strong for younger cohorts due to the overall expansion of participation in education, but remains higher than the OECD average. Social mobility at school, as measured by the average ratio of students’ position in the distribution of socio-economic status to their position in the distribution of academic performance, is also low (Godin and Hindriks, 2018[16]). By this measure, Belgian schools offer the fifth lowest level of social mobility at school among the 27 OECD countries that have participated in the Programme for International Student Assessment (PISA) since 2003 (Figure 2.5, Panel B). High persistence and low mobility at school point to the existence of strong intergenerational education transmission mechanisms in Belgium.

Strong segregation is prevalent in compulsory education (Figure 2.5, Panel C). The most recent PISA results indicate that Belgium is characterised by high shares of students attending “socially segregated schools” (schools with either a relatively high or a relatively low socio-economic background) and “academically segregated schools” (schools with either relatively high or relatively low average academic performance) (Hirtt, 2020[17]). Social and academic segregation go hand-in-hand due to a cumulative process of academic and social selection in Belgian students’ compulsory education path, which explains the particularly strong effect of schools on academic outcomes.

Figure 2.5. Educational disparities by socio-economic conditions are high

Lack of adequate affordable housing contributes to the persistence of low incomes

Affordable and quality housing availability is key in achieving equality of opportunity, as neighbourhood characteristics and peer effects can influence both educational attainment levels and access to public services and higher quality job networks (OECD, 2018[1]). Housing affordability promotes residential mobility, which increases labour market efficiency by improving job matching and boosts income mobility by facilitating access to better paying jobs and better quality education, especially for disadvantaged children and young people (Caldera Sánchez and Andrews, 2011[18]; OECD, 2021[19]). By contrast, large price differentials across neighbourhoods prevent households from moving and can lead socio-economic groups to sort into separate neighbourhoods.

Rising housing costs can increase residential segregation on top of increasing vulnerabilities in the real estate sector (Chapter 1), as it puts a disproportionate burden on low-income households in Belgium, as in other OECD countries (OECD, 2021[19]). Housing expenditure accounts for around 24% of household consumption in Belgium, a two-percentage point increase since the global financial crisis, and above the OECD and the EU averages (Figure 2.6, Panel A). The rise in housing costs was particularly strong in the Brussels-Capital Region over the period 2014-19, where apartment prices on the lowest price segment have increased the most (Figure 2.6, Panel B). Moreover, spatial analysis based on the 2011 census shows strong residential segregation within Brussels and its hinterland (Figure 2.6, Panel C). Poverty is clustered in old industrial neighbourhoods offering low-quality housing from the private rental market, which have long been occupied by low-income households (Costa and De Valk, 2021[20]).

Figure 2.6. Low-income households bear a high burden from housing costs

Residential segregation can exacerbate the barriers to social mobility arising from the labour market and the education system and, therefore, negatively affect the life chances of residents of disadvantaged neighbourhoods. For example, local schools are likely less diverse than in mixed neighbourhoods, which is associated with more unequal school performance. The spatial concentration of people of foreign origin further compounds social mobility challenges, as deprived migrants are spatially isolated within large cities, such as Brussels, Antwerp and Liège (Costa and de Valk, 2018[21]).

Improving the labour market outcomes of vulnerable groups

Long-standing labour market challenges hinder social mobility over the life course in Belgium. Until the COVID-19 pandemic, strong job creation and historically low rates of unemployment co-existed with important challenges, including high levels of inactivity, low labour market transitions and growing skill shortages, especially digital, as discussed in the 2020 Economic Survey of Belgium, as well as sizeable regional disparities (Table 2.2). At the height of the pandemic, swiftly implemented income support measures complemented automatic increases in transfers and decreases in taxes to protect workers’ livelihoods. As the labour market recovers, addressing labour market challenges is key to improve opportunities of upward income mobility. Better labour market outcomes are particularly important for vulnerable groups, as they were disproportionately affected by the COVID-19 fallout and are benefitting less from the recovery (Chapter 1).

	Belgium	Brussels-Capital	Flanders	Wallonia
Employment rate1	64.7	56.5	69.4	59.2
Long-term unemployment rate2	2.3	7	0.9	3.6
Youth unemployment rate3	15.3	29.1	10.8	21.7
NEET4	9.2	11.3	7.3	11.7
Poverty rate5	18.9	34.3	13	24.6

1. Aged 15-64.

2. Share of labour force unemployed for 12 months or more.

3. Aged 15-24.

4. Share of 15-24 year olds not in education, employment or training (Eurostat definition).

5. Share of population at risk of poverty or social exclusion.

Source: Eurostat.

Beyond the challenge of decreasing youth unemployment and increasing activity rates of people above 55 (Chapter 1), employment gaps are particularly large for disadvantaged groups, such as the low educated, non-EU migrants and people with disabilities (Figure 2.7, Panel A). Transitions from inactivity or unemployment to employment are particularly low for vulnerable groups. For example, women transition 20 percent less from non-employment to employment than men in Belgium in 2019 (Causa, Luu and Abendschein, 2021[22]), and there is a five percentage points gap in the transition rate of non-EU migrants compared to natives on average over the period 2008-14 (HCE, 2018[23]). Significant increases in employment rates are necessary among these groups to achieve the government’s ambitious 80% employment target by 2030 for the 20-64 year-olds. Based on demographic projections and a comparison with top performers among neighbouring countries, the High Council of Employment estimates that the required increases in employment rates are particularly high for the low skilled, the youth and non-EU migrants.

Figure 2.7. Vulnerable groups face multiple barriers to employment

Employment gaps often reflect multiple worker-related barriers to employment regarding work readiness, work availability and work incentives (Figure 2.7, Panel B). More than half of those who report to be long-term unemployed, inactive or to have a weak labour market attachment in Belgium face a combination of barriers, the most common of which are low education and health limitations (Adalet McGowan et al., 2020[24]). While many other OECD countries have similar worker-related barriers, Belgium tends to cumulate multiple barriers concurrently (OECD, 2020[25]). Language constitutes an important work-readiness barrier in Belgium and partly explains persistent disparities and imbalances between regions. For people of foreign origin (first and second generation immigrants who account for 32.7% of the population as of January 2021), discrimination likely adds to worker-related barriers (Baert, Heiland and Korenman, 2016[26]; FPS Employment/UNIA, 2019[27]). Unequal opportunities in education contribute to the persistence of work readiness barriers, in particular for children of non-EU migrants, who hardly fare better than their parents in terms of employment outcomes (Piton and Rycx, 2021[28]).

Household characteristics often intersect with gaps associated with the country of origin, so that labour market outcomes are particularly weak for migrant women. Only 46.2% of prime-age women born outside of the EU were in employment in 2020 according to Eurostat. Recent estimates based on social security data suggest that first- and second-generation immigrant women of non-EU origin face a double penalty in terms of employment probabilities that is larger than the sum of the penalties associated with gender and with foreign origin (Piton and Rycx, 2021[28]). Mothers in general are 3.2 percentage points less likely to be employed than women without children in Belgium, according to an analysis based on propensity score matching (Nautet and Piton, 2021[29]).

Important reforms have already contributed to increase the participation of low wage earners and older workers in Belgium, and planned reforms are expected to continue to do so (Box 2.4; Chapter 1). Further to these reforms, focusing policies and resources on vulnerable groups is essential to improve their economic opportunities, especially mothers with children, non-EU migrants and people with disabilities. Beyond promoting upward income mobility and strengthening well-being, such stronger focus can help alleviate fiscal sustainability challenges by reducing the need for redistribution (Chapter 1).

Box 2.4. Planned labour law reforms in Belgium

Labour law reforms discussed in this chapter focus on aspects most relevant for social mobility in Belgium, especially targeting of policies and resources on vulnerable groups. Broader analysis of labour market challenges and necessary reforms to increase participation are discussed in Chapter 1, as well as in the 2020 Economic Survey’s special chapter. This box offers a brief overview of labour law reforms included in the federal government’s October 2021 budget agreement and February 2022 reform package.

Planned reforms are expected to continue increasing labour market participation. A EUR 300 million tax shift away from social security contributions and partly funded through higher consumption (tobacco) and environmental (flights) taxes is intended to reduce the tax wedge on low wages further and to attenuate low-wage traps.

New measures will also improve labour market transitions, conditional on agreement with regional authorities and social partners. The individual training account is being implemented. One third of the employer social security contribution paid in case of layoff will be used for training the dismissed employee. Part of the severance pay will also be used to subsidise the wage at a new employer and starting a new job during the notice period will be possible in order to incentivise work-to-work transitions. Conditions for returning to work from disability leave will be made more flexible. Labour mobility will be incentivised by letting the long-term unemployed who take up a job across the linguistic border or fill a hard-to-fill vacancy to keep parts of their employment benefit for three months.

Finally, the overhaul of labour laws will increase flexibility. Full-time employees will be offered more flexible working weeks, including the option to allocate working hours across four days instead of five. Evening work between 8PM and 12AM will be easier to introduce to promote the development of e-commerce, and the status of platform worker is clarified.

Source: Chancellery of the Prime Minister.

Another set of recent reforms and measures at both federal and regional government levels were intended to increase the relatively low levels of business dynamism and entrepreneurial activity by reducing costs for failed entrepreneurs, lowering administrative barriers and promoting entrepreneurship education (OECD, 2020[9]; OECD, 2019[30]). However, perceived entrepreneurship capabilities remain low (OECD, 2021[31]), and underdeveloped entrepreneurial culture is a major obstacle to entrepreneurship (Mulder and Godefroid, 2016[32]), even though surveys show improvements in Flanders (Roelandt and Andries, 2021[33]). Barriers related to low skills and fear of failure appear particularly detrimental to vulnerable groups, especially women (OECD, 2020[11]). At the same time, the relatively high incidence of necessity self-employment in Belgium suggests that the potential of entrepreneurial activity to provide broad-based upward mobility could be limited. Subsistence entrepreneurship is particularly prevalent among women, as 38% of them report having started a business because they could not secure employment, larger than the 25% overall rate in Belgium and the EU average of 18.8% (OECD, 2020[11]).

Enhancing and recognising skills outside of formal education

Enhancing the digital skills of vulnerable groups is essential to boost upward social mobility by ensuring employability and smoothing labour market transitions (OECD, 2018[1]). Digital skills not only complement digital technologies to enhance productivity and income growth, but also enable individuals to thrive in a fast-changing world of work where automation accelerates the substitution of workers for machines for a range of tasks (OECD, 2019[10]). High-skilled workers generally benefit from digitalisation through higher employment and wages (OECD, 2015[34]). Moreover, current vacancy rates are particularly high in information and communication technology (ICT) related industries.

The distribution of digital skills across groups is particularly uneven in Belgium, with large education, income and gender gaps, which mirror disparities in labour market opportunities. For example, the gap in the share of people with basic or above basic digital skills between those with high and low income is larger than in neighbouring and Nordic countries (Figure 2.8, Panel A), and so is the gender gap in the number of science, technology, engineering and mathematics (STEM) university graduates (Figure 2.8, Panel B). According to the European Commission’s Women in Digital Scoreboard 2021, only 58% of women have at least basic digital skills in Belgium versus 63% of men, despite women being more educated than men on average. Tapping into the unused potential of these groups will be key to address acute and rising digital skills needs given the rapid spread of information and communication technologies (ICT) in the workplace in Belgium. Furthermore, the strong digital focus of investments under the national and regional recovery plans is expected to increase demand for digital skills (Chapter 1).

Figure 2.8. Digital skills are distributed particularly unevenly across groups

Measures promoting participation in STEM training should be boosted in formal education and among both working and unemployed adults, especially women. Regional governments’ measures to promote STEM fields and address shortages of STEM graduates, including the 2020-30 phase of Flanders’ STEM Action Plan, Brussels’ Digital.Brussels, and Wallonia’s Digital Strategy 2019-24, go in the right direction, but more could be done. Making information about wage premia in digital-intensive jobs readily available could help attract candidates for STEM training. As this wage premium is low relative to other countries in the EU, employers may need to improve the compensation package offered to STEM professionals (OECD, 2020[9]). Stronger financial incentives could be achieved through more flexibility in the wage setting mechanism to allow for a stronger link between wages and productivity (Chapter 1).

Lifelong learning is essential for upward income mobility, as maintaining and upgrading skills help labour market transitions both from declining occupations to emerging ones and from inactivity to employment (OECD, 2018[1]). Adult training focused on ICT is particularly crucial given the large digital skill gaps of vulnerable groups in Belgium, in particular as the COVID-19 crisis is accelerating digitalisation (OECD, 2021[35]). All governments have initiatives to increase digital access by vulnerable groups (E-inclusionforBelgium at the federal level, Digibanks in Flanders, Brussels Digital Literacy Action Plan, Espaces Publics Numériques in Wallonia). Furthermore, only 45% of the 55-64 years olds have basic or above basic digital skills according to Eurostat, a major concern as recent pension reforms to delay the effective retirement age exacerbate the need for preventing older workers’ skills from becoming obsolete (Chapter 1).

The overall participation in lifelong learning in Belgium is around the EU average, but hides important disparities by skill level (HCE, 2021[36]). Participation among employees is 54%, above the EU’s 52%, but significantly below best performers (e.g. the Netherlands) (Figure 2.9, Panel A). Participation among the unemployed is particularly high at 42%, compared to 28% in the EU, and reflects the many training opportunities offered by public employment services in the three regions. The evidence on the quality of training is scarce, but points to an overall strong and persistent effect on job finding trainees in Wallonia (Fonder, Lejeune and Tarantchenko, 2019[37]). However, participation among the low educated is 45 percentage points below the highly educated, and participation among workers in elementary jobs is 38 percentage points below workers in highly qualified jobs (Figure 2.9, Panel B). The age participation gap between young and old is also one of the highest in the EU. Persistent participation gaps risk entrenching current low-income persistence further by increasing pre-existing skill disparities (HCE, 2021[36]).

Figure 2.9. The low skilled participate far less in in lifelong learning

Lifelong learning measures should be streamlined and better coordinated, as the high number of schemes, providers and governments involved creates complexity that is particularly detrimental to the participation of vulnerable groups (HCE, 2021[36]). The federal government rightly made lifelong learning a major policy goal, in line with the European Commission’s agenda. Planned reforms include the activation of Article 39 Ter of the Law on labour contracts, which holds that one third of severance pay will be used for providing training to the dismissed employee. The regions and social partners have their own measures, such as Flanders’ Action Plan on Lifelong Learning or CP200’s Cefora, a training initiative organised by the largest sectoral social partner committee. The national recovery and resilience plan also includes lifelong learning measures. This complex institutional framework should be streamlined and coordination between the different levels of government should be ensured.

The federal government’s budget agreement to implement an individual training account, as recommended in the 2020 Economic Survey, is a major step in the direction of increasing lifelong learning efficiency and inclusiveness. By transforming the pre-existing obligation for firms to grant five training days per employee per year on average into an individual right for workers, the individual training account ensures portability, increases workers’ responsibility regarding the choice of learning trajectories and ensures that all workers have access to training, including vulnerable groups, such as the low-skilled. Coordination with regional governments will need to be ensured so that the individual training account can accommodate regional lifelong learning initiatives, including the current development of a Flemish learning and career account and the planned reform of the Congé Education Payé in the Brussels-Capital Region.

A key necessary condition to make the individual training account successful is the provision of high quality training in areas of skill needs and of individual guidance on the choice of training programmes (OECD, 2019[38]). As career guidance is increasingly delivered digitally, options for face-to-face delivery should remain in place to avoid excluding populations with poor digital skills or access (OECD, 2021[39]). Targeted delivery could follow the Dutch model, where jobseekers who are statistically the most likely to be long-term unemployed are systematically invited to a face-to-face interview with a caseload worker, while the others are initially referred to digital services. Furthermore, adequate targeting is needed to both reduce fiscal cost and boost vulnerable groups’ participation. For example, establishing the individual allowance in monetary terms instead of days, as was done in France in 2018, would enable the low-skilled to get more training as the cost of training them is lower (OECD, 2019[40]). Ensuring continued employer involvement in the design of training programmes is also important to reduce skill mismatch for vulnerable groups (see below).

Better targeting of lifelong learning to vulnerable groups is necessary to promote income mobility, in particular through increasing the participation of the low-skilled. Existing information desks, such as Flanders’ Leerwinkel, should continue targeting the low skilled and immigrants in collaboration with public employment services and immigration services. Moreover, initiatives to reach out to those with a low participation rate, such as Brussels’ Formtruck, and Brussels’ and Wallonia’s Cité des Métiers, should be evaluated and expanded as appropriate. Flanders’ Opleidingskrediet scheme, which tops up the regionally-granted paid training leave, also targets low-participation groups by offering a higher allowance for single households. As they implement the individual training account, governments and social partners can also draw from the French experience with the “Compte Personnel de Formation”, in which access is universal but financial support varies across groups (Box 2.5).

Numerous regional policy measures and private initiatives aim at promoting entrepreneurship skills and knowledge for specific vulnerable groups, including the youth, the unemployed, women and people with disabilities. However, awareness is limited among some target groups, while complexity tends to reduce participation (OECD, 2020[11]). Moreover, little is known on the effectiveness of these programmes. Streamlining the offer of entrepreneurial skill programmes and creating a one-stop shop would provide clarity and facilitate outreach to target groups. Evaluating existing programmes is crucial and could benefit from the experience of the assessment of measures to promote entrepreneurship culture in secondary education under the Walloon Region’s Plan Marshall 4.0 (Van Haeperen, Meunier and Mosty, 2019[41]). Entrepreneurial skill programmes could also directly target people with a migrant background, given their large employment and skill gaps.

Box 2.5. Individualising training access schemes: the case of France

The French personal training account (compte personnel de formation, CPF) is an individualised financing scheme for professional training in which training rights are accumulated over time. Implemented in 2015, the account is open to all economically active persons, and is fully transferable throughout the individual’s working life, from the time they enter the labour market until they retire. The account was reformed in 2018 to develop access to training for low-skilled workers and jobseekers. The personal training account was previously measured in training hours, but has now been monetised in euros, a move to correct the disparities in hourly training costs. The amount of the annual payments is based on workers’ skills: each worker has EUR 500 per year in his CPF to pay for training, and the least skilled have EUR 800 (up to a ceiling of EUR 5 000 and EUR 8 000 over 10 years, respectively). The reform also introduced guidance for potential beneficiaries, as well as controls of the quality of and information about the training provided. Part of the funds dedicated to professional training and apprenticeship are earmarked for career advice (conseils en évolution professionnelle, CEP).

Source: OECD (2020) Individualising training access schemes: France; and OECD (2019) OECD Economic Surveys: France 2019.

Labour market mismatches also hinder the chances of upward mobility, as having the right skills or field of study for a job has long-lasting effects on wages and employment throughout workers’ careers (OECD, 2018[1]). Moreover, labour market mismatches are associated with lower productivity through less efficient allocation of resources and can contribute to income inequality in a context of skill shortages (Mcgowan and Andrews, 2015[42]). Mismatches can be particularly pervasive among immigrant workers, especially among those with non-labour migration motives, as poorer knowledge of the language and labour market of the host country hamper effective networking and job search.

Over-qualification among working non-EU immigrants is relatively high in Belgium (HCE, 2018[23]). Almost 40% of those born outside of the EU working in Belgium are overqualified, twice the rate of those born in Belgium, and significantly more than in Germany, France and the Netherlands (Figure 2.10, Panel A). Mismatches between the level of qualifications possessed by non-EU immigrants and those required in their job is particularly pervasive for the highly educated and can lead to skill depreciation (Figure 2.10, Panel B). The composition of migration inflows, with a large share of migrants admitted on family or humanitarian grounds, also contributes to mismatches with labour market needs (OECD, 2015[43]).

The recognition of skills acquired outside standard education paths should be promoted, especially those acquired abroad by migrants. According to a special module of the 2014 Labour Force Survey, one quarter of over-qualified immigrant workers in Belgium attribute the mismatch to difficulties regarding the recognition of skills acquired abroad (HCE, 2018[23]). Therefore, migrants should be referred rapidly and systematically to services of skill validation. Referrals could go through multiple channels, including digital ones, such as the smartphone app “Recognition in Germany”, which was developed on behalf of the German Ministry of Education and Research to provide easily understandable information on recognition procedures and criteria (OECD, 2017[44]). Procedures to recognise degrees obtained abroad could be accelerated, following the example of the Swedish fast-track scheme for shortage occupations, which combines recognition of foreign credentials and prior learning with language classes to provide an occupational certificate (OECD, 2017[44]). Coordination should be ensured between the three communities, which are in charge of degree recognition, and fees could be means-tested or waived, as in the German-speaking Community.

Figure 2.10. Over qualification is high among non-EU immigrant workers

Making activation work for all

Active labour market policies are essential to limit the long-term impact of unemployment on income trajectories and prevent unemployment spells from hampering future upward mobility. Effective activation programmes can also help reduce downward mobility risks, as higher spending on active labour market programmes is associated with a lower share of middle-income households moving down to the bottom of the income distribution (OECD, 2018[1]). The COVID-19 crisis reinforced the importance of activation to foster labour market resilience, while the continuing digital transformation and changing nature of work tend to increase the risk of transitions out of standard forms of employment and, therefore, the need for job-search support (OECD, 2021[45]). Adequate targeting is key to ensure that activation increases job seekers employability in a cost-efficient manner (OECD, 2015[46]).

Belgium performs well regarding activation spending and planned reforms are expected to facilitate work-to-work transitions, including the use of parts of the severance pay to subsidise the wage at a new employer. The spending mix improved, as the share of activation spending allocated to training increased from 15.9% in 2017 to 18.5% in 2019, but remains below the OECD average of 23.7% (Figure 2.11, Panel A). However, spending could better target vulnerable groups, in particular migrants and people with disabilities. Non-Belgian citizens (both EU and non-EU) benefit less from activation measures than nationals do (HCE, 2018[23]). Yet, non-EU migrants tend to require more effective activation spending, as the incidence of long-term unemployment is significantly higher among that group, at 52% compared to 40.3% for natives (Figure 2.11, Panel B). Moreover, while labour market transitions out of employment and into inactivity increased for the whole population during the COVID-19 crisis, vulnerable groups were particularly affected: 29% of non-EU migrant workers and 21% of those with disabilities lost temporarily or permanently their job, compared to 14% and 15% of natives and people without disabilities, respectively (Lens, Marx and Mussche, 2020[47]).

The use of statistical profiling tools for delivering employment services should be expanded further in all three regions to target vulnerable groups. Public employment services increasingly complement rule- and caseworker-based profiling with statistical models to predict labour market disadvantage and classify job seekers into client groups for activation (Desiere, Langenbucher and Struyven, 2019[48]). Designing statistical tools to account for migrant or health status while respecting privacy regulations could facilitate the identification of those more at risk of long-term unemployment and, therefore, enable earlier stage interventions tailored more closely to individual needs. Evidence based on activation measures for the newly unemployed in Flanders shows that job seekers’ country of birth, together with their age and knowledge of the Dutch language, are strong predictors of the type of activation measures likely to accelerate return to employment (Boolens, Cockx and Lechner, 2020[49]). Existing machine learning-based models, such as Flanders’ NextBestSteps programme and the Walloon region’s assessment model of proximity to employment, should continue to be extended to develop tailor-made active labour market programmes for jobseekers.

On-the-job language training should be promoted, as knowledge of (at least one of the) national languages is essential not only for labour market integration, but also for participating in lifelong learning. Moreover, knowledge of one of the national languages likely reduces mismatches through faster labour market integration, as indirectly suggested by the fact that non-EU born immigrants are more likely to master a national language when in employment (HCE, 2018[23]). Belgian regional governments are increasingly investing in language training for migrants, like in many other countries (OECD, 2020[50]). Mandatory integration trajectories for migrants include language classes in Flanders, Wallonia and the German-speaking community, and the Brussels-Capital Region should proceed with plans to make the recently passed legislation regarding mandatory integration programmes effective without further delay.

Figure 2.11. Activation spending on training is low despite high long-term unemployment rates

The effect of language training on labour market integration is stronger if linked with occupational language skills (OECD, 2021[51]). A larger share of active labour market spending could be allocated to on-the-job occupational language training. In Flanders, the public employment service’s Integratie door Werk (Integration through work) programme offers activation trajectories targeted to migrants that include Dutch classes at the workplace (HCE, 2018[23]). To implement and extend such measures, regional governments can draw from the experience of the Latvian State Employment Agency’s language mentorship programme for working refugees, which proved highly effective at helping migrants adapt to their working environment by providing vocation-specific language training on the job (OECD, 2020[50]).

Continuing to promote diversity and fight against discrimination is necessary to complement activation targeted at migrant populations. Labour market discrimination not only worsens employment gaps and contributes to wage gaps, but also decreases the effectiveness of activation measures (HCE, 2018[23]; FPS Employment/UNIA, 2019[27]; view.brussels, 2019[52]). Measures to protect witnesses in discrimination cases were recently added to an already robust anti-discrimination legislation at different levels of government. The 2018 federal law creating the possibility for social inspectors to prove discriminatory infringements through mystery calls or anonymous field tests is being made more flexible, as stringent conditions have strongly limited its use so far (UNIA, 2020[53]). Similar initiatives are being taken by regional governments. Moreover, legislation should be complemented with measures to support inclusive hiring practices, possibly building on available information on the origin composition of staff at the Crossroad Bank of Social Security to identify discrimination patterns and improve the design of anti-discrimination policies (UNIA, 2020[54]; view.brussels, 2019[52]). The federal government’s new measures regarding the monitoring of diversity in sectors are first steps in this direction.

Disability benefit systems should balance protection with work incentives to reintegrate sick workers, while restricting unwarranted inflows into these schemes. Spending on incapacity (disability, sickness, occupational injury) at 3% of GDP in Belgium is higher than the OECD average of 2% (Figure 2.12, Panel A). The share of inactive people not seeking employment due to sickness and disability increased from 10.7% in 2007 to 18.4% in 2020, higher than the EU average of 12.3%. The share of disability recipients, especially for women, has increased (Figure 2.12, Panel B), partly due to the increase in the activity rate of women and the alignment of their statutory retirement age (Saks, 2017[55]), and partly due to the regularisation of previously undeclared workers under the service voucher scheme (Leduc and Tojerow, 2020[56]). The Budget 2022 plans to link funding to regions based on performance of reintegrating workers, introduce tools and incentives to combine partial work and benefits, tighten sanctions for employers and employees, and introduce a reintegration path 2.0, with a view to simplify and speed up the process of integration, including through the use of digital tools, which are welcome.

Early initiation of formal pathways to reintegrate sickness and disability beneficiaries into work should be prioritised. A formal integration procedure was introduced in 2016 to supplement the existing informal pathways, which included a voluntary medical visit with the occupational doctor or the mandatory return to work examination for workers under mandatory medical surveillance. The formal procedure requires mutuality doctors to assess the reintegration possibilities within the first two months of sickness absence. Employers and employees find the formal reintegration procedures to be administratively cumbersome, slow and not individualised and flexible enough. The employee can request formal reintegration anytime during the sick leave and the employer four months after the start of the incapacity, but participation in such formal pathways remains limited. Moreover, 40% of those on formal pathways wait at least one year for their first contact with the occupational physician, compared to 25% for the informal pathways and only 42% return to work (often with another employer), lower than that for informal ones (73%) (Boets et al., 2020[57]).

New reforms aimed to improve policy coordination and data collection to tailor policies to individual needs and track people through the different steps are welcome. Until recently, while some support measures existed for workers deemed unfit to return to their job, this was not systematic and data were not collected for their situation after the dismissal (Lopez-Uroz, Westhoff and Akgüç, 2021[58]). In January 2022, the return-to-work coordination plan created a return-to-work coordinator as employees of mutualities. The coordinator will enable a systematic approach to interact with the employees applying for sickness benefit early and guide them to create a tailor-made path back to work that meets the needs and competences of the person, and the collection of data and data exchange with regional public employment services will help. It will be important to evaluate the effects of this recent reform. Other OECD countries, such as Finland use “one-stop-shops”, which bring together public employment services, social and health care services, services of the national insurance agency and subcontracted professional experts, with a non-hierarchical partnership of actors from different levels of government (OECD, 2021[59]).

Incentives to participate in activation programmes should be strengthened. Workers unfit to return to their previous employment need to reskill, but participation in lifelong learning for those with disabilities at 10.8% in Belgium is lower than the EU average of 18.7%. One pathway for return to work is the vocational rehabilitation, which includes financial incentives for participation. Participants continue to receive their benefits and are paid for each hour of training plus a lump-sum payment of EUR 500 at the end of the training. However, participants can lose their entitlement to disability benefits within six months of the training, which can act as a disincentive. This rule should be removed and beneficiaries encouraged to participate in rigorously evaluated activation programmes.

Figure 2.12. There is a need to reform disability and sickness benefits

In Belgium, as in other countries, employee absence due to mental illness is rising. In 2019, 36% of those on invalidity benefits had mental-related issues (FPB, 2021[60]), especially among the youth (Figure 2.13, Panels A and B) and single mothers. The direct (health and benefits) and indirect costs (reduced participation and productivity) of mental health problems are 5.3% of GDP in Belgium, compared to the EU average of 3.6% (Figure 2.13, Panel C; OECD, (2021[61]). The pandemic can exacerbate these trends as the share of adults experiencing anxiety doubled in April 2020 from a year ago (OECD, 2020[62]) and stabilised at a higher level since (FPB, 2022[63]) (Sciensano, 2021[64]). Mental health has appropriately became a focus of the policy agenda of all levels of government (CSS, 2021[65]).

Beyond preventive measures, better integration of employment and mental health services and broader steps to identify people with mental health issues early are key. Reforming the disability benefit systems, as discussed above, will help. However, 50% of those persons with mental health conditions who receive benefits were on unemployment benefits in the mid-2010s (OECD, 2021[66]). Several evaluations highlighted the need to improve the continuity of care between services (health, mental health, social) as a priority in the organisation of mental health care for adults in Belgium (Devos et al., 2019[67]; Mistiaen et al., 2019[68]). According to a new benchmarking of mental health care systems, systematic inclusion of employment outcomes in mental health service delivery, or mental health outcomes in employment support services is lacking in Belgium (OECD, 2021[61]).

Individual placement and support (IPS) programmes, which are evidence-based practices where multidisciplinary mental health teams, including an employment specialist, provide co-ordinated health and employment support for jobseekers could be further used. Belgium has pilot IPS programmes for people suffering from mental health issues since 2017, which follows the “place-then-train” model, focusing on helping people to get a job rapidly. IPS programmes have resulted in positive employment outcomes in Australia and Denmark (OECD, 2021[66]). Dependent on the evaluations, which will be fully completed by 2023, such programmes should be scaled up. The design should ensure to address funding and implementation challenges of scaling up, based on international best practices (OECD, 2021[59]). For example, it will be important to complement the program’s focus on transitions into employment with measures to sustain jobs and enable career progression.

Figure 2.13. The prevalence of mental health issues is on the rise

Strengthening work incentives for low-income single parents and second earners

Labour tax and benefit systems have important effects on income mobility, as they alter labour market participation and cushion the impact of adverse labour market transitions, e.g., into unemployment (OECD, 2018[1]). To foster social mobility without creating economic insecurity, taxes and benefits need to strike the right balance between maintaining strong labour market participation incentives and providing effective protection against labour income losses. Moreover, to ensure equality of opportunity on the labour market, tax and benefit systems need to be free of implicit biases against specific groups due to the different socio-economic realities they face, e.g., women due to gender gaps in labour market participation and income (Harding, Perez-Navarro and Simon, 2020[69]).

Belgium combines high replacement rates for the unemployed with strong work disincentives, in particular for formerly low-paid workers (Hijzen and Salvatori, 2020[70]), as explored in depth in the 2020 Economic Survey of Belgium. Belgium ranks among the countries where tax and transfers have a greater impact in cushioning large market income losses, together with France and most Nordic countries (OECD, 2018[1]). Despite the federal government’s intention to continue reforms to increase work incentives for low-earners (Box 2.4 above), participation tax rates (the share of additional earnings from work that is lost due to reduced benefits and increased taxes for workers with low wage) remain high for some groups (Adalet McGowan et al., 2020[24]; OECD, 2020[71]).

Participation tax rates for low-income single parents and second earners with children are among the highest in the OECD and should be reduced. In 2021, the participation tax rate was 45.3% for second earners claiming guaranteed minimum income, higher than the OECD average of 32.7% (Figure 2.14, Panel A). The federal government increased the tax advantage for childcare in 2020. Further measures to increase work incentives should address the case of second earners to achieve the federal government’s employment targets for specific groups at lower fiscal cost, in particular women. The partial splitting system for couples, which decreases the household overall labour income tax under progressive taxation by allowing a notional amount of income to be transferred between spouses if one earns 30% or less of the total family income, should be reduced or abolished, as it typically disincentives second earners’ participation in the labour market. In-work benefits would offset the accompanying increase in average labour taxation for low-income households (see below).

Figure 2.14. Low-income parents face weak work incentives

Targeted improvements in work incentives should make work pay for single parents, who face a participation tax rate of 84.8% when claiming unemployment benefits, higher than the OECD average of 68.1% (Figure 2.14, Panel B). Introducing in-work benefits would support return to employment and avoid long-term benefit dependency. Such emphasis towards supporting those in work is common across OECD countries, including France (Prime d’activité) and the United States (Earned Income Tax Credit). Another option is to let low earners cumulate unemployment benefits and income from work, building on the existing scheme for part-time workers, as with the federal government’s planned reform for the long-term unemployed who take up a job across the linguistic border or fill a hard-to-fill vacancy. Effective targeting will be key to limit fiscal costs.

Boosting childcare access for 0-2 year-old children from low-income households would also help decrease participation tax rates for single parents and strengthen women’s work incentives, especially migrants. Enrolment in early childhood education and care for children aged three and older is almost universal in Belgium, and enrolment at 60% for children under three years old is still high (OECD, 2020[9]). However, uptake among migrant populations is relatively low despite high availability and moderate cost of formal childcare, which is reflected in large employment gaps for migrant women (Biegel, Wood and Neels, 2021[72]). There is a quasi-13 percent point gap between the employment rate of native women and that of migrants (Figure 2.15, Panel A). At the same time, almost 70% of 0-2 year-old children from high-income households participate in childcare in Belgium, but only 36.4% of those from low-income households, which are disproportionately households with a migrant background (Figure 2.15, Panel B).The provision of early childhood education and care for children below three should be increased, as recommended in the 2020 Economic Survey of Belgium. Outreach measures should target migrant households and address both cultural norms and the lack of knowledge of the childcare system.

Figure 2.15. Migrant women’s childcare responsibilities contribute to weak employment outcomes

Measures to promote childcare among low-income households should be coupled with targeted activation measures, in particular language training in the case of migrant mothers, building on existing initiatives, such as Actiris’ Maison d’Enfants and the coordination of childcare structure network. While current practices prioritise working parents in case of scarcity in subsidised childcare facilities, actual utilisation can be below full capacity, e.g., due to sick children. Unused capacity leaves room for children from vulnerable households to attend while their parent participates in activation-based training. Such measures combining childcare and activation would help address the issue of parents, in particular migrants, being unable to attend training because of childcare responsibilities. Coordination between public employment services and public childcare organisers is required to forecast actual utilisation of childcare facilities and allocate the children of trainees accordingly.

Enhancing equal opportunities in compulsory education

Belgian students’ overall academic performance is around the OECD average and at par with neighbouring countries, but displays large disparities between students. Performance varies between the three communities (the level of government at which education is organised in Belgium (see Box 2.6 and OECD (2017[73]) for details), with performance high but declining in the Flemish Community, relatively low in the French Community, and in-between in the German speaking Community. However, disparities in performance associated with parental background are substantially larger than differences between the communities. Both the Flemish Community and the French Community are characterised by relatively high socio-economic determination of educational outcomes, as reflected in large test score gaps associated with socio-economic status (Figure 2.16). According to the 2018 PISA results, the gap in reading performance between students in the top quarter of Economic, Social and Cultural Status (ESCS) index and those in the bottom is equivalent to more than three years of schooling, the fifth largest gap in the OECD (OECD, 2019[74]).

A number of other indicators highlight the importance of the correlation between student socio-economic background and performance. 17.2% of the variance in reading performance is explained by parental background, the fifth highest value in the OECD and well above the average value of 12% (OECD, 2019[74]). Only 9% of Belgian students are “academically resilient”, that is, score in the top quarter of performance in reading while belonging to the bottom quarter of ESCS, the fourth lowest share in the OECD (OECD, 2019[74]). Disruptions to in-person education due to the COVID-19 pandemic have likely worsened the socio-economic gap in performance due to differences in access to remote learning tools: 84% of Belgian disadvantaged students have access to a computer for schoolwork compared to 98% of advantaged students (OECD, 2021[75]). Against this backdrop, the communities rightly implemented measures targeting students at risk of exclusion from distance education platforms, including special efforts to make online learning more accessible to migrant children in the French Community (OECD, 2021[76]).

Figure 2.16. Belgian students’ good overall performance hides large disparities

Disparities in the academic performance between advantaged and disadvantaged students are reflected in large differences between schools and programmes to the extent that students’ (effectively their parents’) free choice of school depends too heavily on their socio-economic background. Spatial segregation exacerbates differences between schools, as students are more likely to choose a nearby school. Sorting by socio-economic status across schools is generally negatively related to equity in education (Figure 2.17, Panel A; and (OECD, 2019[74]). Differences between schools reflect a cumulative process of both social and academic selection (Nicaise, 2019[77]). Initial self-sorting along socio-economic lines leads to unequal performance at school due to a variety of mechanisms, e.g., peer group effects or teachers having lower expectations for disadvantaged students (Nicaise, 2019[77]). Disparities in performance then widen due to the process of academic segregation, i.e., students are placed on different learning tracks based on previous achievement, so that social and academic segregation are mutually reinforcing.

Tracking along academic performance, itself strongly determined by socio-economic background, is a key driver of social inequality in Belgian education systems (Hindriks and Godin, 2018[78]; De Witte and Hindriks, 2018[79]). For example, in Belgium more than in the OECD, disadvantaged students in advantaged schools perform better than advantaged students in disadvantaged schools (Figure 2.17, Panel B). Moreover, disparities in terms of social mobility at school appear between tracks and types of education provider. For example, it is higher in vocational tracks than in general education in both the French Community and the German speaking Community (Hindriks and Godin, 2018[78]).

The process of academic and social selection explains the strong influence of school background on performance. For example, almost half of the variation in Belgian students’ performance in mathematics is explained by the school they attend, largely above the OECD’s average of one third (Figure 2.17, Panel C). In all three communities, schools’ average socio-economic environment is even more important than parental background, in the sense that students’ academic performance is more influenced by the average socio-economic status of parents of other students in the same school than by their own parents’ socio-economic status (Box 2.7). Attending a school where students are from more advantaged socio-economic backgrounds is associated with better test scores in all OECD countries, but the effect is particularly strong in Belgium.

Figure 2.17. School environment strongly influences academic performance

Box 2.6. Structure and funding of compulsory education in Belgium

Despite decentralisation since the Third State Reform of 1988 and recent reforms, the general structure of compulsory education is mostly identical across the three communities, which face similar challenges with respect to social mobility – see OECD (2017[73]) for details.

The system features the characteristics of an educational quasi-market, grounded in the principle of freedom of education enshrined in Article 24 of the Constitution, which encompasses freedom of establishment of schools, free choice of schools and funding mainly based on the number of students. Provision is ensured by both governmental (“official education”) and non-governmental bodies (“free education”), with regulation and monitoring by the communities. School choice remained unregulated until reforms in the 2000s attempted to respond to mounting evidence of social segregation, notably from the PISA surveys. Compulsory education ranges from age 5 to 18 and comprises the last year of preschool education (since September 2020), a relatively undiversified primary education (age 6 to 12), a lower secondary education and a (currently reformed) 4-track upper secondary education (general, technical, vocational or arts). Dual vocational education combining school- and work-based learning is offered from age 15 and conditional on successful completion of the second grade of secondary school. Successful completion of compulsory education is required for participating in tertiary education.

The communities are quasi-fully funded by the federal government and education is their main expenditure. Funding is governed by the 2014 reform of the 1989 Special Financing Law and consists in the allocation of part of VAT and federal income tax revenue in proportion to the number of pupils in compulsory education (“pupils key”) and the residency of income tax payers (“fiscal key”), respectively. Expenditure per student is relatively high in international comparison and higher in Flanders than in the French Community (Figure 2.18, Panel A). However, funding prospects for education varies across communities due to diverging fiscal outlooks. After a severe deterioration due to the pandemic, the fiscal deficit is projected to stabilise over the medium term in Flanders, but to deteriorate in the French Community, as primary expenditure grows faster than revenue from the Special Financing Law (Figure 2.18, Panel B). Making fiscal space is necessary to implement some of the policy measures recommended in this survey, especially in the French Community. Expanding the use of spending reviews could help to improve spending efficiency in education (Chapter 1).

Figure 2.18. Funding per student is high, but the funding outlook varies across communities

Box 2.7. The influence of school environment: new evidence based on PISA

New OECD research uses the latest results from the OECD Programme for International Student Assessment (PISA) to disentangle the influence of the school environment from the overall effect of parental background on student performance, following the methodology in Causa and Johansson (2010[80]).

Figure 2.19 shows the decompositions of the student-level conditional correlation between test scores in reading and values of the Economic, Social and Cultural Status (ESCS) index in two parts: an individual background effect, which measures the relationship between the student’s own parental background and their performance within a given school (“within-school” effect); and a school environment effect, which measures the relationship between the average socio-economic background in a given school and the student’s performance, controlling for the student’s own parental background (“between-school” effect).

The influence of the school environment appears particularly large in Belgium, suggesting that Belgian students’ academic performance strongly depends on the socio-economic composition of their peer group in the school they attend (Figure 2.19). Such a high school effect reflects several dimensions along which Belgian education is de facto segregated, e.g., the sorting of students between schools or the allocation of tracks and programmes between schools.

Figure 2.19. Large differences in PISA scores reflect school environment effects

Attenuating the negative consequences of free school choice

Free school choice can bring benefits in terms of innovation, pedagogical freedom or incentives to improve instructional quality, but can also widen social inequality because socio-economically disadvantaged parents are more constrained in their choice of school (Urquiola, 2016[81]). International evidence shows the detrimental effect of free choice on academic and socio-economic sorting (OECD, 2018[13]). Combined with limited capacity, free school choice results in a quasi-market for education services where consumers (parents) compete for the most reputed schools, while suppliers (schools) compete for the best students (Vandenberghe, 1996[82]). These forces tend to create social segregation, as students end up sorted into schools along socio-economic lines, which often overlap with migrant background.

Notwithstanding benefits, free school choice also contributes to disparities in terms of academic performance across schools in Belgium (Nicaise, 2019[77]; Friant et al., 2012[83]; Nusche et al., 2015[84]; OECD, 2015[43]; OECD, 2017[85]). School choice is little regulated in Belgium, despite the implementation of controlled-choice reforms in both the Flemish and the French communities over the past decade to combat enrolment discrimination (Nicaise, 2019[77]; OECD, 2019[86]). Competition for students is also high: 93% of Belgian students are enrolled in schools whose principal reported that there was at least one other school competing for students in the same area, the third highest in the OECD (Figure 2.20, Panel A).

The communities’ plans to reform the controlled school choice schemes and further diversify the social mix in secondary education should be implemented as soon as pandemic-related disruptions subside. In the Flemish Community, the introduction of central digital registration for schools located in areas where limited capacity is an issue should improve the transparency of the scheme and remove parents’ incentives to enrol their children in multiple schools. In the French Community, planned reforms aim at reducing the administrative burden for schools and decentralising part of the student allocation to a more local level, which should decrease overall resistance to the scheme.

Continuing to address the negative consequences of free school choice should remain a priority. In the Flemish Community, the removal of the current double quota system in secondary education, whereby schools had to take in a proportion of disadvantaged students, could affect the diversity of the social mix. The consequences of this removal and the introduction of the new system starting in the academic year 2023-24, whereby schools set their own criteria for enrolling disadvantaged students, should be monitored and evaluated. In the French Community, the implementation of priority rules for enrolment within controlled school choice schemes should be decentralised to build stakeholder buy-in to the new rules, as achieved by the “local negotiation platforms” in Flanders.

Belgium has the third highest rate of student transfer out of school in the OECD (Figure 2.20, Panel B). This practice works in the opposite direction from the regulation of school choice in terms of achieving a more diverse social mix at school. PISA data show that social background has a stronger impact on learning outcomes in countries where it is more common to transfer weak or disruptive students out of a school, without a positive impact on overall performance (OECD, 2010[87]). Where straightforward transfer options exist, teachers may have fewer incentives to cater to work with struggling students. By contrast, schools that have more responsibility for establishing student assessment policies, deciding course offer, determining course content and choosing textbooks tend to transfer fewer students (OECD, 2011[88]). Teachers’ incentives to cater to the needs of struggling students should be promoted and out-of-school transfers limited in order to ensure that policies regulating school enrolment are effective.

School funding could also provide incentives that promote equity in education. In the current system, funding is allocated between schools based on size with a correction for the socio-demographic composition of the school. While this gives schools an incentive to diversity their student intake, it does not reward them for achieving good educational outcomes under challenging conditions. Value-added modelling (i.e., measuring the school’s effect on students’ achievement by using statistical models to predict each student’s test score from students’ and schools’ characteristics and comparing these predicted scores to how the student actually scored on the test) or other reliable performance indicators could be used to measure educational improvements of disadvantaged students. Because they account for students’ prior performance, such mechanisms provide incentives to improve outcomes for all students, including those who start at an academic disadvantage. Performance indicators and other data on successful study progression could also partly address the process of social and academic segregation by providing objective information about the quality of schools and helping parents make informed school choices.

Figure 2.20. Regulation of school choice is low and student transfers out of school are widespread

Increasing mobility between general and vocational tracks

Early tracking into vocational education is another factor contributing to disparities in education. While grouping students by ability contributes to more homogeneous classrooms and may facilitate teaching, it can contribute to increasing social inequality in academic performance if tracking mirrors socio-economic background. International evidence indicates that tracking at an early age is associated with less equity in educational outcomes, as suggested by larger gaps in reading performance between advantaged and disadvantaged students (OECD, 2020[89]).

Tracking into two separate instructional programmes effectively takes place at age 12 in Belgium, lower than the OECD average of 14 (Figure 2.21, Panel A). Moreover, students are tracked into one of four different paths from age 14, more than the OECD average of three. Planned and current reforms in the communities intend to delay tracking, provide differentiated instruction and reduce the number of tracks (see below). Despite a curriculum reform for the first two-years of secondary school in Flanders, there remains a risk of tracking at age 12. In the French Community, planned measures under the Pacte d’Excellence reform, which include a one-year extension of the common curriculum and the merger of the two vocational tracks (“technical” and “professional”), are steps in the right direction, and should be implemented without delay.

When reforming education systems to delay tracking, the authorities should take complementary measures to avoid unintended consequences. A particular concern is further increasing grade repetition, which is already particularly high in Belgium, with almost one third of students having repeated a year by age 15, above the OECD average of 11% (OECD, 2020[89]). Targeted support to struggling students will be necessary to ensure the acquisition of generic skills. Belgium can draw on the example of the “differentiated common curriculum” experience of the Ontario province in Canada, where tracking takes place at age 16, but where teaching is adapted to students’ needs. Key elements include within-class grouping based on learning style and not level; the availability of full-time “teachers for success” who specialise in supporting struggling students; and the possibility to accrue credits towards graduation from internships in companies (Ontario Ministry of Education, 2021[90]).

A further issue with tracking comes from the social stigma attached to vocational tracks in Belgium. Parents often prefer grade repetition in the general track to switching to another track. The resulting cascade effect, whereby struggling students first repeat grades before being transferred to another track deemed academically inferior, means that non-general tracks are effectively relegation tracks (Nicaise, 2019[77]; De Witte and Hindriks, 2017[91]). The cascade effect also contributes to disparities in academic outcomes between schools, which are mostly organised by track. In Belgium, the gap in reading performance between general and non-general tracks after accounting for socio-economic status is 69 points, or around one year and a half of schooling, and much larger than the OECD average of 29 points (OECD, 2020[89]). Upgrading and enhancing the quality of vocational education thus constitutes a priority.

Education reforms should increase mobility between vocational and general tracks. This would help address the image deficit of vocational education in Belgium and reduce academic segregation. By encouraging schools to organise programmes by domain across tracks (“domain schools” and “campus schools”), the modernisation reform in Flanders effectively makes vocational tracks more attractive, and allows students to transfer from vocational to general track, which can lower the cascade effect. In the French Community, vocational education reform within the Pacte d’Excellence aims at strengthening the acquisition of generic skills to reduce differences with general education. However, tracks remain fully separated, with little direct possibilities for students to transition from vocational to general or higher education. This could be reformed based on the Flemish experience.

Figure 2.21. Early tracking to VET is common, but the use of dual VET is rare

The use of dual vocational education, which combines school- and work-based learning, is relatively low in Belgium (Figure 2.21, Panel B) and should be extended, as recommended in the 2020 Economic Survey. Workplace training and experience facilitate school-work transitions and promote social mobility. The impact of vocational education on employment probability at all points of the career is higher in countries with high participation in combined work-study schemes, such as Austria and Germany (Forster, Bol and van de Werfhorst, 2016[92]). Recent reforms could help. In the Flemish Community, all secondary schools can offer dual vocational education since September 2019, and more than 100 courses are currently on offer (OECD, 2021[93]). The Pacte d’Excellence in the French Community aims at improving synergies between vocational education and the labour market, but lacks clarity on the necessary co-ordination with regional authorities, which are in charge of training and employment policies.

To ensure that dual vocational training is aligned with labour market needs, funding should be allocated by training block rather than to schools. The current model, in which vocational education schools receive the student funding and use it to organise the training programme, lacks the flexibility to adapt to rapidly changing labour market needs, and prevents specialisation and collaboration at the level of the firms providing the training. Moving to a demand-driven model, like in the United Kingdom, by allocating funding directly to providers would reduce skill mismatches and ensure that the skills acquired in vocational education are those in demand on the labour market (Sodermans et al., 2018[94]). Moreover, the role of the French Office of Dual Training (OFFA) in coordinating dual vocational education and training between the Walloon Region, the Brussels-Capital Region and the French Community should be clarified and certifications harmonised.

Providing teachers with stronger incentives

Effective teacher policies are essential for successful education systems, and lack of teachers or inadequate qualification are detrimental to the quality and equity of education (OECD, 2019[74]; OECD, 2020[89]). Evidence shows that students’ academic outcomes partly depend on teacher characteristics, such as initial education and certification, working conditions or collaboration and peer mentoring. For example, differences in the percentage of teachers fully certified by an appropriate authority accounted for about 13% of the differences in mean reading performance across all countries and economies participating in PISA (OECD, 2020[89]). A measure of teacher’s quality, the improvement of students performance over a year of schooling, after accounting for factors beyond teachers’ control, has a lasting impact on former students’ educational attainment and earnings (Chetty, Friedman and Rockoff, 2014[95]; Chetty, Friedman and Rockoff, 2014[96]; Hanushek and Rivkin, 2010[97]; Kane and Staiger, 2008[98]). The availability of well-qualified teachers is thus essential to promote social mobility.

Schools face difficulties in recruiting teachers and are often forced to hire staff without adequate certification in Belgium. According to the data based on the 2018 Teaching And Learning International Survey (TALIS), 44% of Belgian students attended schools that reported a lack of teaching staff, higher than the OECD average of 27% (Figure 2.22, Panel A), and 26% of students attended schools with inadequate or poorly qualified staff, the seventh highest rate in the OECD. Teacher shortages reflect significant attrition rates among young teachers. In the Flemish Community in 2014, 44% of secondary school teachers and 25% of primary schools teachers quit teaching within five years of starting (De Witte and Hindriks, 2018[79]). In the French Community, the attrition rate within five years of teaching is 39% for the cohort that started in 2006-07 (Delvaux et al., 2013[99]). Reported reasons for quitting include precarious working conditions, issues with challenging classrooms, difficulties with transitioning from theory to practice and lack of support within the school.

To reduce young teacher attrition, the authorities should strengthen induction programmes for beginner teachers. Only 43% of Belgian teachers report having participated in some kind of induction (Figure 2.22, Panel B). Yet, early career support programmes that provide systematic guidance have been shown to decrease teacher drop-out rates (OECD, 2019[100]). Induction should become systematic and include essential elements, such as an adapted workload for new teachers, their inclusion in the school strategy and culture, regular in-class observation and feedback, and professional development plans.

The recent introduction of a right to guidance for new teachers in the Flemish Community and the definition of a three-phase career with appropriately adapted tasks in the French Community are steps in the right direction. Further reforms could aim at going beyond induction and helping the transition of young teachers by collaborating directly within initial teacher education programmes. For example, the Netherlands introduced in 1997 an “independent teaching practice” phase in the final year of the curriculum, which consists in teaching under a training and employment contract for a period not exceeding five months (European Commission, 2010[101]). The trainee teacher works as a regular staff member, e.g., speaking to parents at parent evenings and discussing reports.

Figure 2.22. There is room to better attract and retain teachers

Early career working conditions, which are precarious in Belgium, should be improved. Most beginning teachers only get part-time contracts, substitute for an absent teacher with a fixed appointment and often have their teaching load spread across several schools. For example, according to the 2012 study of the French Community’s General Administration of Education, 25% of new teachers were hired on a part-time contract and 52% were hired as a substitute. Moreover, new teachers are more likely to teach in a disadvantaged school, which makes the task more difficult. The Flemish Community has recently taken measures to increase job stability for beginning teachers, including a reduction in the number of working days required to convert temporary appointments into fixed appointments. The French Community counts on the improvement of the quality of initial teacher education within the Pacte d’Excellence to improve the attractiveness of teaching careers.

To make teacher careers more attractive while promoting equity in education, teachers could be incentivised to address the needs of disadvantaged students. Teaching in disadvantaged schools could be rewarded with either a premium on salaries, lower teaching load, faster conversion to fixed appointment or priority transfer to other schools after a given number of years. For example, in Japan and Korea, there are policies to ensure that high-quality teachers go to disadvantaged schools, via formal and informal schemes to ensure that teachers periodically change schools, and extra incentives, such as additional salary, less instruction time and the ability to choose their next school (Box 2.8).

Box 2.8. Attracting good teachers to disadvantaged schools: the cases of Japan and Korea

Disadvantaged students are at least as likely as others to be taught by high-quality teachers in Korea and Japan, as evidenced by teachers’ characteristics, such as years of experience, being certified for all the subjects taught and, for science teachers, having a university degree with a major in science.

In Japan, teachers are expected to periodically change schools throughout their career, so that all schools have access to effective teachers and a balance of experienced and beginner teachers. Local education authorities decide on and implement the allocation of teachers to schools.

Korea has a mandatory rotation scheme for teachers, which requires them to move to a different school every five years. Multiple incentives are offered within the scheme to attract teachers to high-needs schools, including additional salary, smaller classes, less instructional time, additional credit towards future promotion to administrative positions, and the ability to choose the next school where one works. The latter two career incentives are seen as particularly attractive.

Source: OECD (2018), Effective teacher policies: Insights from PISA.

Promoting affordability and quality on the housing market

Rising house prices limit access to quality dwelling for households that spend a relatively large share of their income on housing, and contribute to the persistence of low incomes, as inadequate living conditions are often associated with poor access to good job opportunities and education, especially for children and young adults (OECD, 2020[102]). The COVID-19 crisis exacerbated affordability challenges, as both stay-at-home orders and telework (and schooling at home, see above) are more difficult under poor housing conditions. Affordability will be further hampered by the energy efficiency improvements required in the housing sector to meet climate goals, which are putting a disproportionate burden on low-income households (IEA, 2019[103]).

In Belgium, housing conditions overall are among the best in the OECD according to the Better Life Index, with large dwellings and moderate housing costs. However, access to affordable housing has become increasingly challenging for low-income households. Low-income tenants on the private market are particularly at risk of housing cost overburden, as 70% of households in the bottom income quintile spend more than 40% of their income on total housing costs (Figure 2.23, Panel A). Quality is also an issue, especially for low-income households, with 26% living in inadequate housing, a higher share than the EU average of 22% (Figure 2.23, Panel B). These affordability challenges are more acute in large urban areas (see above).

A number of housing-related reforms were implemented by the regions following the second phase of the 6^th State Reform in 2014, mostly related to the repeal of the tax deductibility of interest and capital repayments on mortgage loans for owner-occupied dwelling (housing bonus), as recommended in the 2020 Economic Survey (Box 2.9). These reforms address issues related to the regressive nature of such tax advantages and could attenuate market price pressures, as the bonus is no longer capitalised into housing prices. This welcome move towards tax neutrality between renting and owning supports social mobility. Continuing to remove remaining unwarranted tax advantages for homeownership would promote social mobility further by freeing up public resources to fund rental support for low-income households. Flanders and the Brussels-Capital Region rightly tilted housing taxation away from transaction taxes. The Walloon Region should follow the same approach and consider replacing its mortgage tax credit.

Figure 2.23. Low-income households bear a high burden from housing costs

Box 2.9. Regional reforms of the housing bonus

Regions are in charge of (most of) housing taxation and rental legislation since the implementation of 6^th State Reform, which enabled regional governments to set priorities on housing-related issues. The most notable reform implemented in all three regions regards the repeal of the tax deductibility of interest and capital repayments on mortgage loans for owner-occupied dwelling (housing bonus).

Flanders: Since 2022, the housing bonus is replaced by a reduction of the transaction tax from 6% to 3% (1% if energy-efficiency renovation works are done) for owner-occupied dwelling.
Wallonia: Since 2016, the housing bonus is replaced by a tax credit (cheque habitat) available to households that contracted a mortgage for the purchase of their owner-occupied dwelling; the credit amount decreases with household net taxable income and can be used for the duration of the mortgage up to twenty years but is reduced after ten years.
Brussels-Capital Region: Since 2017, the housing bonus is replaced by an increased abatement of EUR 175 000 on the 12.5% transaction tax for owner-occupied dwelling.

Improving the effectiveness of housing support for low-income households

The provision of social housing and housing-related benefits helps to alleviate the negative consequences of rising housing costs for low-income households. As it increases housing supply, investment in social housing not only results in greater affordability for eligible low-income tenants but also for the rest of the housing market (OECD, 2021[19]). Design is key to ensure that scarce resources reach vulnerable groups without resulting in social segregation or hampering their mobility, in particular on the labour market (OECD, 2020[104]).

In Belgium, the share of social dwellings in the total housing stock at about 4% is low in comparison to some OECD countries, such as France (around 15%) or Denmark (around 20%) (Figure 2.24). Although stable over the past decade, this relatively low share of social housing is not enough to meet demand, as evidenced by very large waiting lists of social housing-eligible households. Lack of social housing is particularly strong in Brussels, where 37 077 households were living in a social housing as of March 2020 but almost 50 000 households were on the waiting list, according to the Société du Logement de la Région de Bruxelles-Capitale, the responsible public agency. Therefore, expanding the supply of social housing is an urgent priority, as recognised by all three regional governments.

The lack of available land makes social housing extension particularly challenging in the Brussels-Capital Region, implying that boosting housing affordability requires complementary policies, such as rental support for low-income tenants on the private market and supply-friendly regulations (see below). Furthermore, density-friendly land use can also help, as regulatory barriers to new construction, in particular regarding height, can restrict housing supply and contribute to higher housing costs, especially in the most expensive OECD cities (OECD, 2021[19]), Yet, Belgium has lower “building-height gaps” (i.e., the difference between actual high-rise building and that predicted based on national income and agricultural land) than neighbouring countries such as Germany or the Netherlands (Jedwab, Barr and Brueckner, 2020[105]). The Brussels-Capital Region is advancing densification with its Plans d’Aménagement Directeurs, but these often face residents’ opposition, especially regarding the height of new constructions.

Figure 2.24. The social housing stock is relatively low

Social housing development needs to avoid resulting in further spatial segregation. The existing minimum quotas for social housing at the municipality level, which are intended to limit socio-economic and ethnic segregation in social housing estates, do not seem effective in Belgium, especially as penalties for non-compliance are not dissuasive. In general, explicit social-mixing policies in social housing have achieved little positive results in most OECD countries, including the 20% municipality-level quota in France (OECD, 2018[106]). By contrast, large-scale revitalisation projects of social housing estates and the immediate surroundings have shown potential (Box 2.10). An alternative approach is to consider the development of smaller social housing estates distributed throughout and across urban areas, effectively deconcentrating social housing estates (OECD, 2020[104]).

Box 2.10. Large scale revitalisation of social housing: the case of Toronto (Canada)

The experience of the City of Toronto with the revitalisation of Regent Park’s social housing estate is an example of the potential of inclusive revitalisation of social housing. The project started in 2005 as Regent Park topped the list of Toronto’s most deprived neighbourhoods and had long suffered from austerity measures and physical deterioration. The redevelopment plan involved community partners, the municipal social housing agency and private sector developers, and had a CAD 1 billion budget.

The plan called for a combination of social housing, market-rate condominiums and affordable housing units in order to foster a socially mixed community, requiring the demolition of the initial estate and a new design to better connect the area to the city’s core. It also had an explicit “right to return” for former tenants, as well as support to finance rent and relocation costs during development-related displacement. As redevelopment nears completion, the existing evidence suggests that about 60% the households have returned to or near their old home, while 10-20% have found alternative social housing in Toronto. Overall, residents report increased satisfaction and positive attitudes towards social mixing across tenures.

Source: OECD (2019) Social housing: A key part of past and future housing policy.

Rules regarding social rental tenure, which were reformed to increase availability for eligible households through, for example, the abolition of permanent contracts, revisions of eligibility status of shrinking households or income tests, should be enforced. However, large differences in average rents with the private market means that households that lose eligibility cannot be easily relocated. Moreover, means-testing could distort labour market incentives if taking up (better paid) employment leads to losing social housing. With that respect, improving the portability of social housing benefits in case of employment opportunity requiring relocation, as achieved by England’s Right to Move policy, would help. Means-tested relocation allowances could also help (see below).

Social housing rents could be set in proportion to household income and wealth in order to avoid distorting labour market incentives and ensure that social housing is allocated to the most vulnerable. In the Netherlands, social rents increase by a threshold in proportion to the combined household income (Box 2.11). In order to avoid hard threshold effects, rents should be set in proportion of income and wealth at all levels, as is partially the case in Australia. Adjusting rents as social tenants’ income increases can lead to more mixed neighbourhoods as richer households remain in social housing. However, such a system would require a significant ramp up of social housing stock extension as it becomes partly occupied by households that would have relocated otherwise. Moreover, incentives for social housing estates to select tenants based on income growth potential at the expense of the most vulnerable households should be avoided.

Box 2.11. Income-dependent rent increases: the case of the Netherlands

Since 2013, yearly rent increases by Dutch social housing corporations are calculated based on the combined household income for each tenancy instead of following inflation rates. Maximum rent increases and the respective income thresholds are set each year by ministerial decree (Uitvoeringsregeling Huurprijzen Woonruimte). The available evidence shows that the reform incentivised richer residents out of regulated units.

The approach is effective to ensure effective allocation in countries in which a broad cross-section of the population is eligible to social housing, including the working population, as it does not necessarily generate strong employment disincentives.

Source: OECD (2019) Social housing: A key part of past and future housing policy.

The regions should continue expanding support to low-income households on the private rental market, given the lack of social housing and the necessary time to expand supply. Existing regional rental allowances to households who are on social housing waiting lists or living in inadequate dwellings should be extended to cover all tenants in the private market who qualify for social housing, as recommended in the 2015 Economic Survey, and made available faster where allowances have already been expanded. Broadening existing housing allowance schemes would raise the supply of rental housing towards the lower end of the market, as it would alleviate potential landlords’ reluctance due to weak financial positions of low-income tenants. A concern is that such expansion of rental allowances puts further upward pressure on housing rents in the short run, through the capitalisation of the subsidy into rents before supply adjusts. However, the current rent indexation mechanism, which caps within-tenancy rent increases to inflation as measured by the health index, should prevent this. If necessary, the regions could consider capping rent increases between tenancies where rental allowances are granted, but the expected short-run benefit of such measure should be assessed against possible longer run-drawbacks, including a potential decline in the rental supply (OECD, 2021[19]).

Easing housing price pressures through more efficient regulation

Efficient regulation complements social housing policies to make housing markets more affordable, in particular to address challenges for many low-income and vulnerable households to afford high and increasing rents (OECD, 2021[19]). Striking the right balance between tenants and landlords is key with, on one hand, secure tenure in good-quality housing and, on the other, limited investment risk for landlords. Moreover, land-use governance arrangements that avoid overlap of responsibilities across government levels can facilitate demand and supply matching. Better land-use governance can improve the responsiveness of supply to evolving demand, thereby mitigating upward pressures on housing prices and contributing to housing affordability.

Tenant-landlord relations appear broadly balanced in Belgium as they lie in the mid-range of their respective index, but the overall level of regulation is higher than the OECD average (Figure 2.25, Panel A). OECD simulations suggest that relaxing rental regulations would contribute to long-term decreases in real house prices in terms of price-to-income ratios, as higher housing investment leads to lower prices. Belgian households would need on average one third of a year less to buy a 100m² dwelling if rental regulations were at the level that prevails in New Zealand (Cournède, Ziemann and De Pace, 2020[107]).

Figure 2.25. There is room to make housing market regulations more supply friendly

Aspects of rental legislation that deter landlords from renting good quality dwellings to low-income households or individuals with non-standard forms of employment should be eased, in particular those due to the perceived strong legal protection of renters and difficult eviction in case of non-payment. Such a reform should avoid the excessive removal of tenant protection, given the already relatively high rate of evictions in Belgium. With that respect, broadening eligibility for housing rental allowances would strengthen the financial position of low-income tenants (see above). Another important way to ease rental market regulations is to allow rent increases during a contractual period when justified by energy-efficiency improvements, as recommended in the 2015 Economic Survey of Belgium and implemented in the 2019 Flemish Housing Rental Decree, but providing specific allowances for low-income tenants (see below). Reforms of rental regulation should balance short-term with long-term affordability, as laxer regulation could deteriorate incumbent tenants’ position in the short run. Rental support (discussed above) would be required during the transition phase.

Fragmentation and overlap in land-use governance is relatively high in Belgium, which is likely explained by the inefficient allocation of competencies between regions, municipalities and even provinces (Figure 2.25, Panel B). More efficient land-use governance is associated with higher housing supply elasticity (Gyourko and Molloy, 2015[108]). According to OECD simulations, streamlining land-use governance to levels observed in best practice countries (e.g. Australia, Canada, and Switzerland) would improve affordability in Belgium, reducing the price to income ratio by one quarter of a year of income (Cournède, Ziemann and De Pace, 2020[107]). More efficient governance could be achieved by reallocating the municipalities’ and the provinces’ competencies to regions, with special attention to the specific case of Brussels and its hinterland, which overlaps three regions.

Addressing the distributional consequences of energy-efficiency upgrading

Despite resulting in lower energy costs to be borne by homeowners or renters in the long run, energy-efficiency upgrading to comply with stricter energy efficiency and other regulations can undermine affordability in the short run, as it puts upward pressure on the costs of construction and maintenance and, in turn, on house prices (OECD, 2021[19]). Moreover, low-income households are less likely to be able to invest in energy efficiency improvements since they are disproportionately tenants and, in case of ownership, they are more likely to be cash-constrained.

Addressing the consequences of required energy-efficiency upgrading is critical in Belgium given the high level of greenhouse gas emissions due to the old age of the housing stock. Energy CO₂emissions from fuel combustion (including electricity and heat) in the Belgian residential sector are particularly high at 1.6 ton per capita in 2019, the 8^th highest in the OECD and largely above the OECD average of 1.1 ton per capita. OECD simulations of the effect of the necessary increased environmental regulations to reach carbon neutrality in 2050 point to large negative impacts on housing availability and affordability in Belgium (Figure 2.26, Panel A). The house price increase would be equivalent to over a year of disposable income (Figure 2.26, Panel B) and residential investment would decrease (Figure 2.26, Panel C). EU grants and regional government investments under the recovery plans should attenuate the housing market impact of energy efficiency upgrading (see Chapter 1).

Figure 2.26. The necessary greening of the housing stock entails affordability challenges.

Compensatory measures are needed to reconcile sustainability and affordability objectives. Federal and regional measures exist to stimulate the renovation of existing housing, including the reduced 6% VAT rate for ordinary renovations, and specific energy efficiency improvements investments, for which regions offer subsidies depending on income and the type of technology. To create incentives for upgrading rental property beyond existing minimum energy efficiency requirements, rental legislation could include energy-efficiency improvement provisions. On top of the existing obligation to provide tenants with information regarding the dwelling’s energy efficiency class, energy audits could be required, including specific estimations of expected energy expenses, as in France. Investments in energy efficiency in rental dwellings could also be encouraged through targeted subsidised loans to cash-constrained landlords, often older people. Since savings from energy improvements are likely to be priced in rents, tenants at the lower end of the rental market whose housing conditions are improved should be shielded from too large rent increases through specific rental support.

Subsidising energy-efficient renovation of old, underused buildings can both expand the useable stock of quality housing and ease affordability challenges from the cost of energy-efficiency upgrading. Subsidies for energy-efficiency upgrading could be financed at least in part by raising recurring property taxes. Addressing disincentives for renovations related to property taxes necessitates to regularly update the cadastral value (still based on values from January 1975, but indexed to consumer price inflation since 1991), as recommended in the 2020 Economic Survey of Belgium, so that owners do not postpone renovation works to avoid disproportionate increases in property taxes.

Energy subsidies for low-income households provide beneficiaries with the wrong incentive and contribute to the high residential emissions (OECD, 2015[43]). Subsidised tariffs should be replaced with lump-sum transfers for low-income households to prevent energy poverty while preserving energy saving incentives. Although necessary in the current context of global energy price inflation, the extension of eligibility for the subsidised electricity tariff should be repealed as soon as price pressures subside, as well as the temporary reduced VAT rate on electricity and natural gas. To keep costs manageable and avoid distorting price signals, additional support to offset the additional energy price rises since the start of the war in Ukraine should be means-tested, well-targeted and temporary (OECD, 2022[109]). Specifically, targeted lump sum transfers should be preferred to across-the-board measures that effectively subsidise household energy consumption irrespective of their income. With that respect, measures, such as the autumn 2021 EUR 80 one-off payment to low-income households, go in the right direction.

MAIN FINDINGS	RECOMMENDATIONS (key in bold)
Digital skills and participation in lifelong learning are low for the low educated, the low income and those with disabilities, leading to large gaps in employment rates and in labour market transitions. The share of STEM graduates is low, especially among women. Entrepreneurial activity is particularly low among vulnerable groups.	Streamline lifelong learning programmes and actors involved, and prioritise vulnerable groups for face-to-face career guidance. Ensure effectiveness of the planned individual training account by focusing on the quality of training. Promote STEM by publicising wage premia in digital-intensive jobs. Evaluate and streamline targeted entrepreneurship skill programmes.
Skill mismatches and over-qualification are high among non-EU migrants. Discrimination and weak language skills are barriers to the employment of people of foreign origin.	Refer immigrants systematically to services of skill validation. Facilitate degree recognition by waiving or means-testing fees and ensuring coordination across communities. Support firms with inclusive hiring practices, for example through reporting of the origin of staff. Allocate a larger share of active labour market spending to on-the-job occupational language training.
The long-term unemployment rate is high and the employment of mothers with young children, migrants and those with disabilities is low.	Expand the use of statistical tools to target vulnerable groups for tailored active labour market programmes.
Formal reintegration procedures for disability benefit recipients start late. Workers unfit to return to their previous employment need to reskill. Gaps in policy coordination for individuals with mental health challenges lower their employment prospects. Gaps in individual support for sickness and disability beneficiaries hamper their return to work.	Simplify and make formal reintegration procedures more flexible. Remove the rule that participants in training programmes lose entitlement to disability benefits within six months of training. Systematically focus on employment outcomes in mental health service delivery, and mental health outcomes in employment support services. Further scale up individual placement and support programmes for sickness and disability beneficiaries, conditional on their evaluation.
High participation tax rates for low-income single parents and second earners with children weaken work incentives, especially for women.	Introduce in-work benefits for low-wage workers with children. Reduce work disincentives for second earners by removing the partial splitting system for couples, while compensating low-income households through in-work benefits. Promote early childhood education and care for low-income households, especially migrants.
Free school choice leads to social segregation and drives academic disparities across schools. Out-of-school transfers of students lower social diversity within schools and reduce the effectiveness of controlled school choice reforms. Schools are incentivised to diversify their student intake, but not to achieve good educational outcomes under challenging conditions.	Evaluate the consequences of the removal of schools’ obligation to take in a proportion of disadvantaged students in Flanders. Decentralise the implementation of priority rules for enrolment to ensure buy-in in the French Community. Limit out-of-school transfer possibilities to increase teachers’ incentives to improve the performance of struggling students. Use reliable performance indicators and other data on successful study progression to inform school funding based on educational improvements made with disadvantaged students.
Tracking into general and vocational education sorts students along socio-economic lines. Low mobility between general and vocational tracks contributes to social inequality in education outcomes. Participation in dual vocational education that combines school and work based learning is low, despite a high share of vocational students. Vocational education funding lacks the flexibility to adapt to rapidly changing labour market needs.	Pursue efforts to delay tracking by extending the common curriculum in Flanders. Target support to struggling students to ensure that delayed tracking does not increase grade repetition in the French Community. Further encourage schools to organise programmes across tracks and enable transfers between them. Strengthen the use of dual vocational education through financial incentives and compulsory accreditation for training providers. Allocate funding directly to training providers in a demand-driven model based on training blocks rather than to schools.
New teacher attrition is high, especially in disadvantaged schools, and early career working conditions are precarious.	Reward teaching in disadvantaged schools with higher pay or faster conversion to fixed appointments. Enable and expand professional insertion before graduation from initial teacher education and strengthen induction of new teachers.
Social housing supply is too low and price differentials with the private market hinder moves, thereby distorting work incentives.	Expand rent allowances to cover low-income private market tenants while proceeding to increase the social housing stock. Means-test social housing eligibility regularly and set rents proportionately.
Some aspects of rental legislation deter landlords from renting good quality dwellings to low-income households. The fragmentation of land-use governance makes housing supply less responsive to demand, contributing to price pressures.	Strengthen the financial position of low-income tenants by broadening eligibility for housing rental allowances. Reallocate municipalities’ and provinces’ competencies related to land-use governance to regions.
Energy-efficiency upgrading of the housing stock is critical, but will weigh on affordability. High energy prices disproportionately affect low-income households, but insufficiently targeted energy subsidies can provide wrong incentives.	Encourage investments in energy efficiency in rental dwellings through targeted subsidised loans to cash-constrained landlords. Use targeted measures against the impact of high energy prices, such as lump-sum support to low-income households.

MAIN FINDINGS

RECOMMENDATIONS (key in bold)

Improving the labour market outcomes of vulnerable groups

Digital skills and participation in lifelong learning are low for the low educated, the low income and those with disabilities, leading to large gaps in employment rates and in labour market transitions.

The share of STEM graduates is low, especially among women.

Entrepreneurial activity is particularly low among vulnerable groups.

Streamline lifelong learning programmes and actors involved, and prioritise vulnerable groups for face-to-face career guidance.

Ensure effectiveness of the planned individual training account by focusing on the quality of training.

Promote STEM by publicising wage premia in digital-intensive jobs.

Evaluate and streamline targeted entrepreneurship skill programmes.

Skill mismatches and over-qualification are high among non-EU migrants.

Discrimination and weak language skills are barriers to the employment of people of foreign origin.

Refer immigrants systematically to services of skill validation.

Facilitate degree recognition by waiving or means-testing fees and ensuring coordination across communities.

Support firms with inclusive hiring practices, for example through reporting of the origin of staff.

Allocate a larger share of active labour market spending to on-the-job occupational language training.

The long-term unemployment rate is high and the employment of mothers with young children, migrants and those with disabilities is low.

Expand the use of statistical tools to target vulnerable groups for tailored active labour market programmes.

Formal reintegration procedures for disability benefit recipients start late.

Workers unfit to return to their previous employment need to reskill.

Gaps in policy coordination for individuals with mental health challenges lower their employment prospects.

Gaps in individual support for sickness and disability beneficiaries hamper their return to work.

Simplify and make formal reintegration procedures more flexible.

Remove the rule that participants in training programmes lose entitlement to disability benefits within six months of training.

Systematically focus on employment outcomes in mental health service delivery, and mental health outcomes in employment support services.

Further scale up individual placement and support programmes for sickness and disability beneficiaries, conditional on their evaluation.

High participation tax rates for low-income single parents and second earners with children weaken work incentives, especially for women.

Introduce in-work benefits for low-wage workers with children.

Reduce work disincentives for second earners by removing the partial splitting system for couples, while compensating low-income households through in-work benefits.

Promote early childhood education and care for low-income households, especially migrants.

Enhancing equal opportunities in compulsory education

Free school choice leads to social segregation and drives academic disparities across schools.

Out-of-school transfers of students lower social diversity within schools and reduce the effectiveness of controlled school choice reforms.

Schools are incentivised to diversify their student intake, but not to achieve good educational outcomes under challenging conditions.

Evaluate the consequences of the removal of schools’ obligation to take in a proportion of disadvantaged students in Flanders.

Decentralise the implementation of priority rules for enrolment to ensure buy-in in the French Community.

Limit out-of-school transfer possibilities to increase teachers’ incentives to improve the performance of struggling students.

Use reliable performance indicators and other data on successful study progression to inform school funding based on educational improvements made with disadvantaged students.

Tracking into general and vocational education sorts students along socio-economic lines.

Low mobility between general and vocational tracks contributes to social inequality in education outcomes.

Participation in dual vocational education that combines school and work based learning is low, despite a high share of vocational students.

Vocational education funding lacks the flexibility to adapt to rapidly changing labour market needs.

Pursue efforts to delay tracking by extending the common curriculum in Flanders.

Target support to struggling students to ensure that delayed tracking does not increase grade repetition in the French Community.

Further encourage schools to organise programmes across tracks and enable transfers between them.

Strengthen the use of dual vocational education through financial incentives and compulsory accreditation for training providers.

Allocate funding directly to training providers in a demand-driven model based on training blocks rather than to schools.

New teacher attrition is high, especially in disadvantaged schools, and early career working conditions are precarious.

Reward teaching in disadvantaged schools with higher pay or faster conversion to fixed appointments.

Enable and expand professional insertion before graduation from initial teacher education and strengthen induction of new teachers.

Promoting affordability and quality on the housing market

Social housing supply is too low and price differentials with the private market hinder moves, thereby distorting work incentives.

Expand rent allowances to cover low-income private market tenants while proceeding to increase the social housing stock.

Means-test social housing eligibility regularly and set rents proportionately.

Some aspects of rental legislation deter landlords from renting good quality dwellings to low-income households.

The fragmentation of land-use governance makes housing supply less responsive to demand, contributing to price pressures.

Strengthen the financial position of low-income tenants by broadening eligibility for housing rental allowances.

Reallocate municipalities’ and provinces’ competencies related to land-use governance to regions.

Energy-efficiency upgrading of the housing stock is critical, but will weigh on affordability.

High energy prices disproportionately affect low-income households, but insufficiently targeted energy subsidies can provide wrong incentives.

Encourage investments in energy efficiency in rental dwellings through targeted subsidised loans to cash-constrained landlords.

Use targeted measures against the impact of high energy prices, such as lump-sum support to low-income households.

References

[24] Adalet McGowan, M. et al. (2020), “Addressing labour market challenges in Belgium”, OECD Economics Department Working Papers, No. 1602, OECD Publishing, Paris, https://dx.doi.org/10.1787/8a542d68-en.

[26] Baert, S., F. Heiland and S. Korenman (2016), “Native-Immigrant Gaps in Educational and School-to-Work Transitions in the 2nd Generation: The Role of Gender and Ethnicity”, Economist (Netherlands), Vol. 164/2, pp. 159-186, https://doi.org/10.1007/S10645-016-9273-4/TABLES/8.

[72] Biegel, N., J. Wood and K. Neels (2021), “Migrant-native differentials in the uptake of (in)formal childcare in Belgium: The role of mothers’ employment opportunities and care availability”, Journal of Family Research, Vol. 33/2, pp. 467-508, https://doi.org/10.20377/JFR-463.

[15] Blanden, J., P. Gregg and L. Macmillan (2007), “Accounting for intergenerational income persistence: Noncognitive skills, ability and education”, Economic Journal, Vol. 117/519, pp. C43-C60, https://doi.org/10.1111/J.1468-0297.2007.02034.X.

[57] Boets, I. et al. (2020), Evaluation de l’impact de la nouvelle réglementation sur la réintégration au travail Synthèse de l’évaluation quantitative, qualitative et juridique.

[49] Boolens, J., B. Cockx and M. Lechner (2020), “Priority to unemployed immigrants? A causal machine learning evaluation of training in Belgium”, CEPR Discussion Paper, No. 14270, Centre for Economic Policy Research, https://doi.org/10.1086/677233.

[18] Caldera Sánchez, A. and D. Andrews (2011), “Residential Mobility and Public Policy in OECD Countries”, OECD Journal: Economic Studies, https://dx.doi.org/10.1787/eco_studies-2011-5kg0vswqt240.

[80] Causa, O. and Å. Johansson (2010), “Intergenerational Social Mobility in OECD Countries”, OECD Journal: Economic Studies, https://dx.doi.org/10.1787/eco_studies-2010-5km33scz5rjj.

[22] Causa, O., N. Luu and M. Abendschein (2021), “Labour market transitions across OECD countries: Stylised facts”, OECD Economics Department Working Papers, No. 1692, OECD Publishing, Paris, https://dx.doi.org/10.1787/62c85872-en.

[95] Chetty, R., J. Friedman and J. Rockoff (2014), “Measuring the Impacts of Teachers I: Evaluating Bias in Teacher Value-Added Estimates”, American Economic Review, Vol. 104/9, pp. 2593-2632, https://doi.org/10.1257/AER.104.9.2593.

[96] Chetty, R., J. Friedman and J. Rockoff (2014), “Measuring the Impacts of Teachers II: Teacher Value-Added and Student Outcomes in Adulthood”, American Economic Review, Vol. 104/9, pp. 2633-79, https://doi.org/10.1257/AER.104.9.2633.

[20] Costa, R. and H. De Valk (2021), “Socio-spatial disparities in Brussels and its hinterland”, Urban Book Series, pp. 271-291, https://doi.org/10.1007/978-3-030-64569-4_14.

[21] Costa, R. and H. de Valk (2018), “Ethnic and Socioeconomic Segregation in Belgium: A Multiscalar Approach Using Individualised Neighbourhoods”, European Journal of Population, Vol. 34/2, pp. 225-250, https://doi.org/10.1007/S10680-018-9480-6.

[107] Cournède, B., V. Ziemann and F. De Pace (2020), “The Future of Housing: Policy Scenarios”, OECD Economics Department Working Papers, No. 1624, OECD Publishing, Paris, https://dx.doi.org/10.1787/0adf02cb-en.

[65] CSS (2021), Belgian Mental Health Monitor: COVID-19, Conseil Supérieur de la Sante.

[79] De Witte, K. and J. Hindriks (2018), L’École du Renouveau, Itinera Institute.

[91] De Witte, K. and J. Hindriks (2017), L’École de la Réussite, Itinera Institute, https://www.skribis.be/nl/l-ecole-de-la-reussite.html (accessed on 14 November 2021).

[99] Delvaux, B. et al. (2013), “Les enseignants débutants en Belgique francophone: trajectoires, conditions d’emploi et positions sur le marché du travail”, Cahiers du GIRSEF, No. 92, Université Catholique de Louvain, Louvain-la-Neuve.

[48] Desiere, S., K. Langenbucher and L. Struyven (2019), “Statistical profiling in public employment services: An international comparison”, OECD Social, Employment and Migration Working Papers, No. 224, OECD Publishing, Paris, https://doi.org/10.1787/b5e5f16e-en.

[67] Devos, C. et al. (2019), Performance of the Belgian health system – report 2019.

[3] Eurofound (2017), Social mobility in the EU, Publications Office of the European Union, Luxembourg., https://doi.org/10.2806/420671.

[101] European Commission (2010), “Developing coherent and system-wide induction programmes for beginning teachers: a handbook for policymakers.”, Staff Working Document SEC, No. (2010)538, European Commission, Brussels.

[37] Fonder, M., B. Lejeune and E. Tarantchenko (2019), Does training boost the job finding rate of the unemployed ? Timing-of-events based evidence from Belgium.

[92] Forster, A., T. Bol and H. van de Werfhorst (2016), “Vocational Education and Employment over the Life Cycle”, Sociological Science, Vol. 3, pp. 473-494, https://doi.org/10.15195/V3.A21.

[63] FPB (2022), Sustainable development indicators, 2022, Federal Planning Bureau, Brussels, https://www.plan.be/publications/publication-2193-en-sustainable_development_indicators_2022 (accessed on 16 March 2022).

[60] FPB (2021), Santé mentale en Belgique : les coûts cachés de la Covid-19.

[27] FPS Employment/UNIA (2019), Monitoring socio-économique 2019: Marché du travail et origine, Federal Public Service Employment, Labour and Social Consultation/Interfederal Centre for Equal Opportunities.

[83] Friant, N. et al. (2012), “Priority education policies in Belgium: Two modes of regulation of the effects of a market logic.”, in Demeuse, M. et al. (eds.), Educational Policies and Inequalities in Europe, Palgrave MacMillan.

[16] Godin, M. and J. Hindriks (2018), “An international comparison of school systems based on social mobility”, Economie et Statistique, Vol. 2018/499, pp. 61-79, https://doi.org/10.24187/ECOSTAT.2018.499S.1940.

[108] Gyourko, J. and R. Molloy (2015), “Regulation and Housing Supply”, Handbook of Regional and Urban Economics, Vol. 5, pp. 1289-1337, https://doi.org/10.1016/B978-0-444-59531-7.00019-3.

[14] Hanushek, E. et al. (2021), “The intergenerational transmission of cognitive skills: An investigation of the causal impact of families on student outcomes”, Working Paper, No. 29450, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/W29450.

[97] Hanushek, E. and S. Rivkin (2010), “Generalizations about Using Value-Added Measures of Teacher Quality”, American Economic Review, Vol. 100/2, pp. 267-71, https://doi.org/10.1257/AER.100.2.267.

[69] Harding, M., G. Perez-Navarro and H. Simon (2020), “In tax, gender blind is not gender neutral”, Ecoscope, https://oecdecoscope.blog/2020/06/01/in-tax-gender-blind-is-not-gender-neutral-why-tax-policy-responses-to-covid-19-must-consider-women/ (accessed on 8 January 2022).

[36] HCE (2021), La formation continue des salariés: Investir dans l’avenir, High Council of Employment, Brussels.

[23] HCE (2018), Les immigrés nés en dehors de l’Union européenne sur le marché du travail en Belgique, High Council of Employment, Brussels.

[70] Hijzen, A. and A. Salvatori (2020), “Designing fair and work-oriented unemployment benefits: The case of Belgium”, OECD Social, Employment and Migration Working Papers, No. 237, OECD Publishing, Paris, https://dx.doi.org/10.1787/ac17d171-en.

[78] Hindriks, J. and M. Godin (2018), “L’école de la chance: C’est quoi, une école équitable et efficace?”, in De Witte, K. and J. Hindriks (eds.), L’Ecole du Renouveau, Itinera Institute.

[17] Hirtt, N. (2020), L’ inégalité scolaire, ultime vestige de la Belgique unitaire ? Une analyse statistique des causes de l’inégalité scolaire, Appel Pour une Ecole Démocratique, https://www.skolo.org/2020/02/25/l-inegalite-scolaire-ultime-vestige-de-la-belgique-unitaire/.

[103] IEA (2019), The Critical Role of Buildings: Perspectives for the Clean Energy Transition, IEA, Paris, https://www.iea.org/reports/the-critical-role-of-buildings (accessed on 11 January 2022).

[105] Jedwab, R., J. Barr and J. Brueckner (2020), “Cities without skylines: worldwide building-height gaps and their implications”, CESifo Working Paper, No. 8511, https://doi.org/10.2139/SSRN.3682010.

[98] Kane, T. and D. Staiger (2008), “Estimating Teacher Impacts on Student Achievement: An Experimental Evaluation”, National Bureau of Economic Research, https://doi.org/10.3386/W14607.

[56] Leduc, E. and I. Tojerow (2020), “Subsidizing domestic services as a tool to fight unemployment: Effectiveness and hidden costs”, IZA Discussion Paper, No. 13544, Institute of Labour Economics, Bonn, https://ftp.iza.org/dp13544.pdf (accessed on 16 March 2022).

[47] Lens, D., I. Marx and N. Mussche (2020), “De initiële effecten van de COVID-19 pandemie op de Belgische arbeidsmarkt-opkomende ongelijkheden”, Berichten, No. D/2020/6104/07, Centrum voor Sociaal Beleid Herman Deleeck, Antwerpen.

[58] Lopez-Uroz, N., L. Westhoff and M. Akgüç (2021), “Shaping return to work policy: the role of industrial relations at national and company level Country report for Belgium”, REWIR Working Paper.

[12] Martinez-Toledano, C. et al. (2019), “Who pays the price of folly? The business cycle and income and wealth mobility in Spain”, OECD Economics Department Working Papers, No. 1561, OECD Publishing, Paris, https://dx.doi.org/10.1787/880f797c-en.

[42] Mcgowan, M. and D. Andrews (2015), “Labour market mismatch and labour productivity: evidence from PIAAC data”, Economics Department Working Papers, No. 1209, Organisation for Economic Co-operation and Development.

[68] Mistiaen, P. et al. (2019), Organisation of mental health care for adults in Belgium, https://kce.fgov.be/sites/default/files/atoms/files/KCE_318_Mental_Health_care_Report.pdf.

[32] Mulder, J. and H. Godefroid (2016), “How to stimulate entrepreneurship in Belgium ?”, Economic Review, Vol. 2016/2, pp. 63-80.

[29] Nautet, M. and C. Piton (2021), “How does parenthood affect the careers of women and men ?”.

[77] Nicaise, I. (2019), “Selectiemachine of sociale lift? Segregatie in het onderwijs tegengaan”, in Vlaamse Onderwijsraad (ed.), Spots op onderwijs: Wetenschappers voor het voetlicht, Lannoo Campus.

[84] Nusche, D. et al. (2015), OECD Reviews of School Resources: Flemish Community of Belgium 2015, OECD Reviews of School Resources, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264247598-en.

[109] OECD (2022), OECD Economic Outlook, Interim Report March 2022: Economic and Social Impacts and Policy Implications of the War in Ukraine, OECD Publishing, Paris, https://dx.doi.org/10.1787/4181d61b-en.

[110] OECD (2022), Quality and Equity of Schooling in the German-speaking Community of Belgium, Reviews of National Policies for Education, OECD Publishing, Paris, https://dx.doi.org/10.1787/9a6b6f3a-en.

[75] OECD (2021), 21st-Century Readers: Developing Literacy Skills in a Digital World, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/a83d84cb-en.

[61] OECD (2021), A New Benchmark for Mental Health Systems: Tackling the Social and Economic Costs of Mental Ill-Health, OECD Health Policy Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/4ed890f6-en.

[19] OECD (2021), Brick by Brick: Building Better Housing Policies, OECD Publishing, Paris, https://dx.doi.org/10.1787/b453b043-en.

[39] OECD (2021), Career Guidance for Adults in a Changing World of Work, Getting Skills Right, OECD Publishing, Paris, https://dx.doi.org/10.1787/9a94bfad-en.

[59] OECD (2021), Disability, Work and Inclusion in Ireland: Engaging and Supporting Employers, OECD Publishing, Paris, https://dx.doi.org/10.1787/74b45baa-en.

[7] OECD (2021), Does Inequality Matter? : How People Perceive Economic Disparities and Social Mobility, OECD Publishing, Paris, https://dx.doi.org/10.1787/3023ed40-en.

[93] OECD (2021), Education Policy Outlook 2021: Shaping Responsive and Resilient Education in a Changing World, OECD Publishing, Paris, https://dx.doi.org/10.1787/75e40a16-en.

[66] OECD (2021), Fitter Minds, Fitter Jobs: From Awareness to Change in Integrated Mental Health, Skills and Work Policies, Mental Health and Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/a0815d0f-en.

[8] OECD (2021), Government at a Glance 2021, OECD Publishing, Paris, https://dx.doi.org/10.1787/1c258f55-en.

[51] OECD (2021), Language Training for Adult Migrants, Making Integration Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/02199d7f-en.

[45] OECD (2021), OECD Employment Outlook 2021: Navigating the COVID-19 Crisis and Recovery, OECD Publishing, Paris, https://dx.doi.org/10.1787/5a700c4b-en.

[35] OECD (2021), OECD Skills Outlook, OECD Publishing, Paris, https://dx.doi.org/10.1787/e11c1c2d-en.

[31] OECD (2021), OECD SME and Entrepreneurship Outlook 2021, OECD Publishing, Paris, https://dx.doi.org/10.1787/97a5bbfe-en.

[76] OECD (2021), The State of School Education: One Year into the COVID Pandemic, OECD Publishing, Paris, https://dx.doi.org/10.1787/201dde84-en.

[25] OECD (2020), Faces of Joblessness in Finland, http://oe.cd/FoJ (accessed on 15 March 2022).

[102] OECD (2020), Housing and Inclusive Growth, OECD Publishing, Paris, https://dx.doi.org/10.1787/6ef36f4b-en.

[11] OECD (2020), “Inclusive entrepreneurship policies: Belgium 2020”, Country assessment notes.

[50] OECD (2020), International Migration Outlook 2020, OECD Publishing, Paris, https://dx.doi.org/10.1787/ec98f531-en.

[9] OECD (2020), OECD Economic Surveys: Belgium 2020, OECD Publishing, Paris, https://dx.doi.org/10.1787/1327040c-en.

[89] OECD (2020), PISA 2018 Results (Volume V): Effective Policies, Successful Schools, OECD Publishing, Paris.

[104] OECD (2020), “Social housing: A key part of past and future housing policy”, Employment, Labour and Social Affairs Policy Briefs, http://oe.cd/social-housing-2020. (accessed on 12 January 2022).

[62] OECD (2020), Tackling the mental health impact of the COVID-19 crisis: An integrated, whole-of-society response.

[71] OECD (2020), The Future for Low-Educated Workers in Belgium, OECD Publishing, Paris, https://doi.org/10.1787/0140a728-en.

[30] OECD (2019), In-Depth Productivity Review of Belgium, OECD Publishing, Paris, https://dx.doi.org/10.1787/88aefcd5-en.

[86] OECD (2019), Balancing School Choice and Equity: An International Perspective Based on Pisa, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/2592c974-en.

[38] OECD (2019), Individual Learning Accounts : Panacea or Pandora’s Box?, OECD Publishing, Paris, https://dx.doi.org/10.1787/203b21a8-en.

[40] OECD (2019), OECD Economic Surveys: France 2019, OECD Publishing, Paris, https://dx.doi.org/10.1787/a0eee144-en.

[10] OECD (2019), OECD Employment Outlook 2019: The Future of Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9ee00155-en.

[74] OECD (2019), PISA 2018 Results (Volume II): Where All Students Can Succeed, PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/b5fd1b8f-en.

[100] OECD (2019), Working and Learning Together: Rethinking Human Resource Policies for Schools, OECD Reviews of School Resources, OECD Publishing, Paris, https://dx.doi.org/10.1787/b7aaf050-en.

[1] OECD (2018), A Broken Social Elevator? How to Promote Social Mobility, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264301085-en.

[106] OECD (2018), Divided Cities: Understanding Intra-urban Inequalities, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264300385-en.

[13] OECD (2018), Equity in Education : Breaking Down Barriers to Social Mobility, OECD Publishing, Paris, https://www.oecd-ilibrary.org/education/equity-in-education_9789264073234-en (accessed on 21 November 2021).

[6] OECD (2018), The Productivity-Inclusiveness Nexus, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264292932-en.

[73] OECD (2017), Education policy outlook: Belgium, http://www.oecd.org/edu/policyoutlook.htm (accessed on 18 November 2021).

[44] OECD (2017), Making Integration Work: Assessment and Recognition of Foreign Qualifications, Making Integration Work, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264278271-en.

[85] OECD (2017), OECD Economic Surveys: Belgium 2017, OECD Publishing, Paris, https://dx.doi.org/10.1787/eco_surveys-bel-2017-en.

[34] OECD (2015), Adults, Computers and Problem Solving: What’s the Problem?, OECD Skills Studies, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264236844-en.

[5] OECD (2015), In It Together: Why Less Inequality Benefits All, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264235120-en.

[43] OECD (2015), OECD Economic Surveys: Belgium 2015, OECD Publishing, Paris, https://dx.doi.org/10.1787/eco_surveys-bel-2015-en.

[46] OECD (2015), OECD Employment Outlook 2015, OECD Publishing, Paris, https://dx.doi.org/10.1787/empl_outlook-2015-en.

[88] OECD (2011), “When students repeat grades or are transferred out of school: What does it mean for education systems?”, PISA in Focus, No. 2011/6, OECD Publishing, Paris, http://www.pisa.oecd.org (accessed on 23 November 2021).

[87] OECD (2010), PISA 2009 Results: What Makes a School Successful?: Resources, Policies and Practices (Volume IV), PISA, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264091559-en.

[90] Ontario Ministry of Education (2021), Student success, http://www.edu.gov.on.ca/studentsuccess/ (accessed on 23 November 2021).

[28] Piton, C. and F. Rycx (2021), “A Broken Social Elevator? Employment Outcomes of First- and Second-Generation Immigrants in Belgium”, De Economist, Vol. 169/3, pp. 319-365, https://doi.org/10.1007/S10645-021-09385-2/TABLES/10.

[33] Roelandt, J. and P. Andries (2021), Ondernemerschapscultuur en ondernemend gedrag in Vlaanderen: Situatie 2020, Steunpunt Economie en Ondernemen Beleidsrapport 20-019, Vlaamse Overheid, https://www.ewi-vlaanderen.be/sites/default/files/bestanden/store_rapport_ondernemerschapscultuur_anno_2020.pdf (accessed on 17 March 2022).

[55] Saks, Y. (2017), “A better understanding of developments in the numbers claiming disability insurance”, NBB Economic Review.

[64] Sciensano (2021), Belgium COVID-19 Studies : Mental Health Surveys, https://datastudio.google.com/embed/reporting/7e11980c-3350-4ee3-8291-3065cc4e90c2/page/ykUGC.

[94] Sodermans, K. et al. (2018), “Enseignement en alternance”, in De Witte, K. and J. Hindriks (eds.), L’École du Renouveau, Itinera Institute, https://www.itinera.team/fr/publications/livres/lecole-du-renouveau.

[4] StatBel (2019), Poverty indicators in Belgium in 2018 (EU-SILC), https://statbel.fgov.be/en/news/unemployed-single-parent-families-and-tenants-are-most-vulnerable-poverty.

[54] UNIA (2020), Concrétiser l’utilisation du datamining pour lutter contre la discrimination sur le marché du travail, https://www.agoria.be/fr/Big-Data-comment-le-fisc-a-recupere-plus-d-1-milliard-d-euros-dans-la-lutte- (accessed on 7 January 2022).

[53] UNIA (2020), Les tests de situation sur le marché du travail bruxellois, https://www.unia.be/files/Documenten/Aanbevelingen-advies/2020-11-18_Rapport_Testing_Bruxelles_FR.pdf (accessed on 7 January 2022).

[81] Urquiola, M. (2016), “Competition Among Schools: Traditional Public and Private Schools”, Handbook of the Economics of Education, Vol. 5, pp. 209-237, https://doi.org/10.1016/B978-0-444-63459-7.00004-X.

[41] Van Haeperen, B., O. Meunier and M. Mosty (2019), “De la perception de ses compétences entrepreneuriales à l’intention d’entreprendre : Quelques résultats d’une évaluation des mesures de sensibilisation à l’esprit d’entreprendre en faveur des élèves de l’enseignement secondaire en Wallonie”, Reflets et Perspectives de la Vie Economique, Vol. 57/3, pp. 73-85, https://doi.org/10.3917/rpve.583.0073.

[82] Vandenberghe, V. (1996), “Functioning and regulation of eductional quasi-markets”, No. 283, CIACO.

[52] view.brussels (2019), L’origine étrangère : Un frein à l’emploi à Bruxelles, Observatoire Bruxellois de l’Emploi et de la Formation, https://press.actiris.be/lorigine-etrangere--un-frein-a-lemploi-a-bruxelles (accessed on 18 March 2022).

[2] WEF (2020), The Global Social Mobility Report 2020: Equality, Opportunity and a New Economic Imperative, World Economic Forum, http://www.weforum.orghttp://www3.weforum.org/docs/Global_Social_Mobility_Report.pdf.

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Countries

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Publications

Publications

Featured publications

Data

Data

Featured data

News & Events

News & Events

Featured Events

About OECD

About

Engage with us

Work with us

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment