Participation in adult learning remains highly unequal across socio-demographic and economic groups, with individuals who have higher skill proficiency, educational attainment and income being consistently more likely to engage in learning activities. These disparities reflect not only individual choices but also structural barriers that limit access to opportunities for those from disadvantaged backgrounds. Adults with lower educational attainment, lower earnings and lower skills face significant obstacles to participation, reinforcing existing inequalities. Comparisons with data from the previous cycle of the survey suggest some convergence in participation rates for disadvantaged groups, though no single country has universally improved access. Understanding the factors that drive participation gaps, as well as the policies that succeed in reducing them, is essential for designing interventions that promote adult learning and labour market resilience.
Trends in Adult Learning
Abstract
Unequal participation rates persist across socio-demographic and income groups. Adults who are older, less-educated, and who work in lower-paying jobs with lower skill requirements are less likely to participate in adult learning than their counterparts. The largest participation gaps occur along dimensions of skill proficiency, educational attainment and income. Adults who perform better on these metrics are consistently more likely to participate in adult learning.
Decreases in the participation rate for adults have been concentrated among higher-educated and higher-earning adults. Though disparities persist, in many cases the overall size of the participation gap within a given country has declined, largely due to decreases in participation being disproportionately concentrated among adults who are more highly educated or who are employed in higher-skilled, better-paying jobs.
Gender gaps in adult learning are minimal. Though during Cycle 1 men were more likely than women to participate in training, participation rates fell disproportionately more for men, and the latest data show that women are now as likely to participate in adult learning, on average.
The size of participation gaps varies considerably between countries. While no single country stands out as having universally improved participation rates for all disadvantaged groups since Cycle 1, several countries nevertheless achieve high and equitable rates of adult learning overall.
Policymakers should be attentive to participation gaps. These gaps can highlight structural barriers and uneven access to learning opportunities. Success in removing barriers and promoting participation has the potential to help disadvantaged adults develop skills and improve their labour market outcomes.
Gaps should be considered within the context of overall participation rates and their country’s adult learning policy environment. National efforts to expand access to adult learning should identify and build on the factors that have enabled these countries to achieve more uniform participation rates for adults with diverse backgrounds and abilities.
Participation in adult learning is shaped by factors that influence both the supply of adult learning opportunities that are available as well as individuals’ own demand for learning. Participation is associated with a range of socio-demographic, economic and occupational factors, and there are significant disparities between different groups. These disparities are not only a matter of individual choice, but also reflect broader issues that can reinforce existing social and economic inequalities.
Adults who participate in learning opportunities tend to have higher levels of education – and have higher-educated parents – and they tend to enjoy more stable employment and higher earnings. They are more often in the earlier stages of their careers and enjoy employer support for learning activities. Conversely, those with lower socio‑economic status, including lower earners and people with limited educational attainment and skills, often face significant barriers to participation (see also Chapter 3). When certain groups are consistently unable to benefit from learning opportunities, this limits their potential to improve their socio‑economic standing and can perpetuate cycles of disadvantage. It also makes them more vulnerable to labour market changes that require regular upskilling and/or reskilling to transition to new sectors or occupations.
This chapter examines existing disparities in adult learning participation, identifying the areas where gaps are most pronounced, and the groups disproportionately affected. It acknowledges that disparities frequently overlap across various dimensions; for example, earnings are closely linked to both age and educational attainment. After outlining the overall participation gaps, the chapter delves deeper into these inequalities by analysing them while controlling for these interrelated factors. This approach enables a more nuanced understanding of the root causes of participation disparities. Such insights are essential for policymakers, enabling them to develop more effective, targeted strategies that enhance access to, and the benefits of, adult learning opportunities.
Furthermore, this section analyses how these gaps in participation have evolved over time.1 Widening gaps may indicate the prevalence of barriers to learning; conversely, if gaps are narrowing, this could suggest that certain policy interventions are improving access. Changes in participation gaps should be analysed within the context of overall participation rates within a country and existing adult learning policy. Addressing inequalities in adult learning participation is essential to ensure that adult learning systems support all individuals, rather than reinforcing existing inequalities. When interpreting results, changes in the measurement of non-formal learning between survey cycles need to be considered (see Chapter 1, Boxes 1.3 and 1.5).
Adult learning helps individuals acquire skills and knowledge that can help them to work more productively, to advance within their careers, or to find employment if they are out of work. Where it is provided by employers, it can indicate investment in workers and their abilities. Because adult learning brings these benefits, unequal participation rates among different socio-demographic groups can indicate unequal access to learning opportunities. It can also indicate differences in motivation to participate in adult learning, which itself can be shaped by systemic issues, such as lack of relevant opportunities, repeated exclusion from learning systems and training that feels irrelevant to one’s job. Identifying socio-demographic gaps can help design adult learning policies that support those individuals who are least likely – or able – to take advantage of adult learning opportunities by addressing these systemic issues.
This section presents gaps in adult learning participation rates for a number of socio-demographic characteristics: by age, gender, immigrant status, parental education, adults’ own educational attainment and adults’ skill level. Alongside gaps in overall participation rates, this section presents the percentage point changes in these gaps between Cycle 1 and Cycle 2 of the Survey of Adult Skills. Where multiple categories are involved – for example, between adults aged 25‑34, 35‑54, and 55‑65 – the changes between the top and bottom category are compared.
Younger adults participate more in adult learning than older ones
Adult learning can be useful for adults to enter into new jobs or sectors, or as a way to invest further in one’s career path. One might expect that younger and mid-career workers are most likely to take up adult learning, as investments made early in one’s career can pay dividends over a longer period. Conversely, being older often means being more established within a given job and facing competing priorities – for example, by starting a family or getting closer to retirement – which make individuals less likely to participate in adult learning.2
Indeed, the oldest adults (ages 55 to 65) are least likely to participate in adult learning, with 26% of adults report participation within the past 12 months on average across OECD countries (Figure 2.1). By contrast, at 51%, rates are highest among early-career adults aged 25 to 34 and in all countries but New Zealand these adults were, in fact, the most likely of all age groups to participate. On average, 43% of adults aged 35‑54 report adult learning within the past 12 months.
The largest gaps in participation rates between young and older adults are in Canada, Chile and Singapore, at 35 percentage points or more. The smallest gaps are in Japan, Korea and New Zealand, at less than 15 percentage points. In the case of New Zealand, the small gap is mostly owed to the high participation rate among its oldest adults, whereas for Korea it can be attributed to lower overall participation rates, which translate into a smaller absolute participation gap.
Given that absolute gaps in participation rate between age groups can be sensitive to the overall participation rate, it can be useful to examine relative gaps (not displayed in graph). The smallest relative gaps are in England (United Kingdom), Ireland, New Zealand and Norway, where the oldest adults are 40% less likely to participate in adult learning than the age group with the highest participation rate (30% for New Zealand). The largest relative gaps in participation are in Austria, Croatia, Korea, Poland and Portugal, where adults aged 55 to 65 are only 31‑34% as likely to participate in adult learning as adults aged 25‑34.
Participation rate in adult learning by age, percentage
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]) Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Though in most countries younger adults aged 25 to 34 are most likely to participate in adult learning, the gap between their participation and the participation for all other adults has narrowed over the past decade (Figure 2.2). In 13 countries and economies, there was a statistically significant decline in the gap between older and younger adults. The average decline was 6 percentage points, with the youngest adults going from 30 to 24 percentage points more likely than the oldest adults to participate in adult learning. In Korea, where the relative decline was largest, the gap between adults aged 25‑34 and those aged 55‑65 fell 19 percentage points, from 33 to just 14 percentage points.
Percentage point change in the participation gap between adults aged 25‑34 and adults aged 55‑65
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap between adults aged 25‑34 and all other adults. The participation gap is defined as the participation rate for adults aged 25‑34 minus the participation rate for adults aged 55‑65. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
The gender participation gap has disappeared in most countries
Gender disparities in adult learning participation may reflect inequalities in the division of labour within the household that leave women with less time participate in training. They may alternatively reflect differences in employment status that affect access to employer support for adult learning opportunities. Lastly, differences may emerge due to variability in access to training across different sectors, or different preferences for learning activities overall. Policy makers should consider designing adult learning policies that eliminate barriers to participation that perpetuate gendered inequalities in adult learning and in the opportunities that adult learning can provide access to.
Participation gaps between men and women are small. The average gender gap in participation rates across OECD countries and economies is just 3 percentage points; in 15 participating countries and economies women are more likely to participate, and in 16 countries and economies men do. Because in some cases men have higher participation and in some cases women do, this means that, on balance, average participation rates for women and men are equal (see shaded bar). In just two countries does the gap exceed 8 percentage points: Czechia and Japan. In both of these cases, men are more likely to participate.
Participation rate in adult learning by gender, percentage
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Participation rates of men and women have converged since Cycle 1 was conducted (Figure 2.4). In 11 countries and economies there was a statistically significant decline in the gender gap. The overall decreases in adult learning participation observed in many countries (see Chapter 1) seem to have been driven by declining participation of men, in many cases, and it is this decrease that accounts in large part for the comparative equality now. In every case where there was a statistically significant increase in women’s participation relative to men’s, this occurred in a country where men had initially had higher participation rates, meaning that this change represents a move toward greater parity in adult learning participation.
Percentage point change in the participation gap between women and men
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap between men and women. The participation gap is equal to the participation rate for men minus the participation rate for women. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Adults with migrant background continue to participate at lower rates
Adult learning can be an especially helpful way for migrants to improve their employability, gain a certification that attests to their knowledge and skills in their host country, and integrate into the labour market of their adoptive homes. Native‑ and foreign-born adults often differ in age, family structure, employment sector, or educational qualifications. This means that while foreign-born adults may have greater need for adult learning (for example, to compensate for weaker networks or unrecognised qualifications), they may also face greater barriers, such as childcare responsibilities, lower educational attainment, or lower rates of employer support. The key concern for policy makers is ensuring that adult learning systems are open and accessible to those who could stand to benefit the most from participation, particularly foreign-born adults who did not attain a formal qualification in their country of residence.
Native‑born adults are more likely to participate in adult learning than foreign-born adults, at 44% compared to 39% on average. In nine countries, however, the foreign-born participation rate is higher than that of the native‑born one. Yet, these gaps are typically smaller and are not statistically significant compared to gaps where native‑born are more likely than foreign-born adults to participate (Figure 2.5).
The gaps between the native‑ and foreign-born adult populations within countries are smaller than the gaps in average participation rates between countries. The largest absolute gaps in participation rates between these groups are observed in Estonia, Finland, the Flemish Region (Belgium), France, and Italy, where gaps are all 10 percentage points or more, favouring the native‑born participation. The largest relative participation gaps are in the Flemish Region (Belgium), Chile, Estonia, France, Italy, and Lithuania, where native‑born adults are at least 30% more likely to participate in adult learning than foreign-born adults (over 40% for Estonia and France, and 70% more for Italy; not shown in graph).
Participation rate in adult learning by immigrant status, percentage
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. OECD is an unweighted average of all participating member countries except for, Japan, Korea, Poland, and the Slovak Republic, which are excluded due to low shares of foreign-born adults in their populations. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
For four countries, there has been a statistically significant change in the gap between native‑ and foreign-born adults between Cycle 1 and Cycle 2 (Figure 2.6). In Canada, for example, this change means that foreign-born adults have moved from being less likely than native‑born adults to participate to being more likely than native‑born adults. In Chile, foreign-born adults are now less likely than native‑born adults to participate in adult learning, which was not the case in Cycle 1, whereas in Italy the gap between foreign-born and native‑born has grown wider. There has not been a statistically significant change in the size of the gap, on average across all participating countries and economies.
Percentage point change in gap between foreign- and native‑born adults
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap between foreign-born and native‑born adults. The participation rate is equal to the participation rate for native‑born minus the participation rate for foreign-born. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries except Japan, Korea, Poland, and the Slovak Republic, which are excluded due to low shares of foreign-born adults in their populations. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Adults with better-educated parents are far more likely to participate in learning
Parental education is commonly used as a proxy for an individual’s socio‑economic background; typically, families where one or both of the parents has attained a university degree tend to have higher overall household income, and children raised in these households tend to attain more advanced qualifications and earn more money themselves.
This section analyses adult learning participation rates for adults from high-status backgrounds (at least one parent completed a tertiary qualification), medium-status backgrounds (at least one parent completed upper secondary education, but no higher), and low-status backgrounds (neither parent completed upper secondary). Given the differences in resources between higher-status and lower-status families, one might expect participation rates for these groups to reflect the fact that adults from higher-status backgrounds have more opportunities available to them and tend to continue learning over the life‑course, whereas adults with low-status backgrounds may encounter more barriers and less support in participating in adult learning.
Indeed, across all participating countries and economies, adults from high-status backgrounds are more likely to participate in adult learning, just as adults whose parents have at most a lower secondary qualification are the least likely. This pattern holds true across all countries and economies, with just 28% of adults with low-status backgrounds participating in adult learning, on average, compared to 52% of adults from high-status backgrounds (Figure 2.7).
The size of participation gaps between adults with high- and medium-status background varies from country to country. There is no statistically significant difference between the participation rates for these two groups in the Netherlands or Sweden. These countries also perform well in providing access to those from the lowest socio‑economic backgrounds; in Norway and Sweden, along with Denmark and New Zealand, these adults are at least 70% as likely to participate in adult learning as those from the highest-status group, in relative terms. Conversely, the widest gaps between adults with high- and low-status background are in Chile, Croatia, Italy, Latvia, the Slovak Republic and the United States, where gaps are greater than 30 percentage points. Evidence suggests that a large share of the difference between adults of different socio‑economic backgrounds is transmitted through adults’ own educational attainment and occupation, as well as the fact that older adults – who are more likely to have less-educated parents – are also less likely to participate in adult learning (see Table 2.1).
Participation in adult training by parental educational attainment, percentage
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
The participation gap between adults with high- and low-status backgrounds has shrunk slightly since the previous cycle of the Survey of Adult Skills, with participation gaps narrowing by 4 percentage points on average. For eight countries and economies, there were statistically significant reductions in the gap that were even larger, ranging from a reduction of 5 percentage points in Estonia to over 15 percentage points in Korea and Poland. Finding effective ways to improve access to adult learning opportunities for individuals from lower-status backgrounds would mark an important step in improving economic opportunities and social mobility for adults who are not born into advantaged families. Since the foundations for adult learning participation are laid in initial education, policies that promote equal opportunities for children from disadvantaged backgrounds are essential. For adults, given that socio‑economic advantage is largely transmitted through educational attainment and occupation, targeted efforts should focus on improving access for those with lower qualifications or working in lower-skilled jobs.
Percentage point change in participation gap between high- and low-status backgrounds
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the gap in participation rates between adults whose parents did not complete upper secondary (low-status background) and adults who had at least one parent with a tertiary degree (high-status background). Participation rate gaps for each cycle are equal to the participation rate for high-status background adults less the participation rate for adults with low-status backgrounds. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Highly educated adults participate the most in adult learning
Research has consistently found that high-educated adults are most likely to benefit from adult learning, a phenomenon known as the Matthew effect – e.g. Blossfeld, Kilpi-Jakonen and de Vilhena (2020[2]). Having a strong foundation of initial education means that adults have an aptitude for learning that continues over the life‑course. Additionally, individuals with higher qualifications are more likely to work in careers where they are continually required to improve their skills or where their employers are more supportive of training for their employees.
Gaps in participation between tertiary-educated adults and those who have not completed upper secondary education are the greatest of any category of gaps presented in this chapter, at 35 percentage points on average. Even the gap between those with tertiary education and those with upper-secondary (but no tertiary) is substantial, at 19 percentage points on average (Figure 2.9).
Moreover, in every participating country or economy, tertiary-educated adults are the most likely to have participated in adult learning and those who have not completed upper secondary are the least likely. In 20 of 31 participating countries and economies, at least half of tertiary-educated adults took part in adult learning within the 12 months preceding the survey. By contrast, only three countries have participation rates above 30% for adults who do not hold an upper secondary qualification (Estonia, Norway, and Sweden).
The largest absolute gaps between the highest- and lowest-educated adults are in Chile, Portugal, and the United States, where tertiary-educated adults are at least 45 percentage points more likely to have participated in adult learning. Relative gaps for these countries are also above the average for participating countries and economies, where the highest-educated adults are between 2.5 and 3.6 times as likely as the lowest-educated to have participated in adult learning.
Participation in adult learning by educational attainment, percentage
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
On average, the gap between highly educated adults and those with less education shrank by 8 percentage points. Compared with Cycle 1, nine countries saw a decrease in the educational attainment gap of 10 percentage points or more (Figure 2.10). For countries such as Estonia and Sweden, this improvement builds upon participation rates that were already more equitable than the average for other OECD countries. Although the lowest-educated adults in France, Korea, and the Slovak Republic had below-average participation rates in Cycle 1, these adults saw their relative likelihood of participating improve by over 10 percentage points (over 30 points in Korea). Conversely, participation gaps for low educated adults widened by 7 percentage points in England (United Kingdom). In most cases, the reduction in the gap was driven by the fact that the participation rate decreased more sharply among the more highly educated.
Percentage point change in participation gap between adults with low and high educational attainment
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap for adults who did not complete upper secondary, compared to all other adults. The participation gap is equal to the participation rate for higher-educated adults minus the participation rate for the least-educated adults. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
…so do adults with high skills
Adults with the lowest skill levels in literacy (Level 1 and below) are the least likely to participate in adult learning (23%), and those with the highest proficiency in literacy (Level 4 and above) are the most likely to participate, with 61% engaging adult learning in the 12 months preceding the survey, on average across OECD countries and economies (Figure 2.11). This pattern mirrors disparities in educational attainment, where individuals with higher qualifications are also more likely to participate in adult learning (the two variables are, in any case, correlated). Box 2.1 describes in greater detail the skills proficiency domains used in the Survey of Adult Skills and how proficiency scores are categorised into proficiency levels.
The largest absolute gaps in participation between the highest- and lowest-skilled adults are found in Chile and Portugal, where participation rates differ by just over 50 percentage points. In relative terms, gaps are greatest in Italy and Poland, where the lowest-skilled adults participate at rates less than a quarter of the rate of the highest-skilled. Norway, Singapore, and Sweden, on the other hand, have the lowest relative gaps by skill level. Some countries, such as Ireland, New Zealand, and Norway manage to combine high overall participation rates with high inclusion of the lowest-skilled in adult learning.
Participation rate by skill level, percentage
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning. See note for Poland in the Reader’s Companion (OECD, 2024[3]).
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Figure 2.12 plots the change in the participation gap between adults with high literacy proficiency (Level 4 and above) and low literacy proficiency (Level 1 and below). On average across participating countries and economies, there was a slight reduction of 4 percentage points in adult learning participation rates for the highest-skilled and lowest-skilled adults, between Cycle 1 and Cycle 2. Individually, just four countries managed to shrink the participation gap between low- and high-skilled adults (changes for other countries were not statistically significant). In Korea and Singapore, the gap decreased by 30 percentage points or more, while for Canada and Sweden the gap shrank by 10 to 15 percentage points. There were no countries with statistically significant increases in the participation gap between low- and high-skilled adults.
However, interpreting changes in these participation gaps over time requires some caution. These trends reflect a combination of actual behavioural changes and potential shifts in the population composition within each skill group. For instance, if the proportion of adults performing at a given proficiency level changes over time, the observed differences in participation may be partly due to changes in the composition of the group rather than to actual behavioural changes among individuals with similar skills. Without decomposing these effects, it is difficult to determine whether narrowing gaps reflect real improvements in access or simply changes in population structure.
Percentage point change in literacy proficiency gap between low and high-skilled adults
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap for adults with low literacy proficiency (Level 1 and below) compared to high-proficiency adults (Level 4 and above). The participation gap is equal to the participation rate for these higher-skilled adults minus the participation rate for lower-skilled adults. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning. See note for Poland in the Reader’s Companion (OECD, 2024[3]).
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Measuring adult proficiency is central to the Survey of Adult Skills. To accurately assess and compare adult skills, the 2023 Survey of Adult Skills had adults complete a set of tasks covering three areas: literacy, numeracy, and adaptive problem solving. Adults’ performance in the assessment tasks was then used to assign scores to each adult for each proficiency domain using a 500‑point scale. At each point on the scale, an individual with a given level of proficiency has a 67% chance of successfully completing tasks located at that same level. These scores were grouped into broader proficiency levels, ranging from Below Level 1 to Level 5.
Adults’ scores across the three domains are correlated, but not perfectly predictive of one another. Numeracy and literacy were assessed during both cycles of the Survey of Adult Skills; for simplicity’s sake, this report makes use of proficiency levels in literacy, comparing adults who performed at or above Level 4 with those who performed at or below Level 1. Adults with literacy proficiency at or above Level 4 can read long and dense texts and make inferences based on the text that involve accessing, understanding, evaluating, and reflecting on the content and sources of the text. Adults at Level 1 are able to locate information on a text page and understand the meaning of short texts as well as the organisation of lists or multiple sections within a single page; they struggle to access and understand information in longer texts that contain some distracting information.
More information on Cycle 2 of the Survey of Adult Skills is available in the report Do Adults Have the Skills They Need to Thrive in a Changing World?: Survey of Adult Skills 2023 (OECD, 2024[4]).
Economic background characteristics such as employment status or income level strongly influence access to adult learning. Individuals who are inactive in the labour market and those who earn low incomes are nearly uniformly less likely to have participated in adult learning than their more advantaged counterparts. Policies that remove barriers and promote participation among these groups of adults have the potential to help them develop skills and improve their labour market outcomes.
Amid a trend towards lower overall training participation rates, some countries have seen an improvement in the relative share of training participants who come from disadvantaged groups. In these cases, efforts to expand access to adult learning should identify and build on the factors that have enabled these countries to achieve a more uniform participation rate across the population.
Most countries have shrunk participation gaps for low earners
Adult learning can lead to higher earnings by boosting worker productivity. Since higher earners are also more likely to participate in training, this suggests a mutually reinforcing relationship – those who train tend to earn more, and those who earn more are more likely to train. Figure 2.13 compares adult learning participation rates between the highest-earning 25% of employed adults and the lowest-earning 25% of adults within each country (i.e. the top and bottom quartile of the earnings distribution). The highest earning workers are 24 percentage points more likely to participate in adult education and training, on average across OECD countries and economies (58% compared to 34%).
In 22 out of 31 participating countries and economies, over half of top earners have participated in some form of adult learning in the past year; this is true for the lowest earners only in 6 countries. At 57%, Norway has the highest participation rate for the bottom quartile of earners, and the smallest absolute gap in participation rates, at 8 percentage points (barring Korea). It also has the smallest relative gap: low-earning workers are nearly 90% as likely to participate as high earners. Sweden, Denmark, and the Flemish Region (Belgium) have similarly low relative gaps. The same is also true for Korea, though far fewer adults reported adult learning during the past year overall.
The largest absolute participation gaps, on the other hand, are in Chile, Latvia, and Portugal – all above 35 percentage points. Along with Croatia, Hungary, and Poland, relative gaps are also comparatively high, with the lowest-earning quintile less than 40% as likely to have participated in adult learning than the richest.
Participation in adult learning by income quartile, percentage
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Fourteen countries saw statistically significant decreases in the participation gap between the lowest earners and the highest earners between Cycle 1 and Cycle 2. In nine of these countries the gap shrank by 10 percentage points or more (Figure 2.14). Identifying what enabled training participation and building on this trend may help lower earners to become more productive and improve their earning potential.
Percentage point change in participation gaps by earnings quartile
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap for the 25% of adults with the lowest earnings compared to the 25% of adults with the highest earnings. The participation gap is equal to the participation rate for higher-earning adults minus the participation rate for lower-earning adults. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Unemployed workers struggle to access adult learning
Most adult learning takes place at the workplace (see Chapter 3); indeed, in all participating countries (save for Korea), employed workers are more likely than unemployed workers to have participated in adult learning in the 12 months prior to the survey. Since most adult learning is employer-provided, unemployed workers have fewer opportunities to participate.3 Nevertheless, supporting unemployed adults to accessing adult learning can help them improve their skills, or learn new ones, and thereby improve their chances of finding employment. Countries that incorporate adult learning into their active labour market programmes can therefore help to compensate for these difficulties in accessing training. In countries such as Austria or Switzerland, who emphasise training for unemployed adults, these gaps are far smaller. Moreover, adults who do participate in training tend to participate in trainings that are longer in duration.
On average across participating countries and economies, 43% of employed adults report having taken part in adult learning in the past 12 months, compared to 36% of unemployed adults (Figure 2.15). Inactive adults are uniformly the least likely to have participated, at 14% overall.
The largest gaps in participation rates between employed and unemployed adults are in England (United Kingdom), Finland, Germany, and Norway, at over 20 percentage points. The smallest gaps in both relative and absolute terms are in Austria, Denmark, Israel, Portugal, and Switzerland – in each of these countries, the rate for unemployed workers is higher than the OECD average. An outlier is Korea, where unemployed adults are 5 percentage points more likely than employed adults to report participating in adult learning.
Participation in adult learning by employment status, percentage
Note Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Over the past decade, in five countries the participation gap between unemployed and employed adults narrowed (Figure 2.16). The average change across participating countries and economies was a decline of 4 percentage points. In general, declines in the participation gaps between employed and unemployed adults has been driven by the fact that declines in participation are more pronounced among employed adults than among the unemployed. While it is beyond the scope of this report to establish with certainty what explains the changes in participation rates between cycles, changes in the composition of unemployed and employed groups may account in part for some changes. In the first international report on the second cycle of the Survey of Adult Skills (OECD, 2024[4]), for instance, it was noted that the share of unemployed adults shrank between cycles for virtually all participating countries.
Percentage point change in participation gap by employment status
Note: Adults aged 25‑65; formal and non-formal job-related in the 12 months prior to the survey. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap for unemployed adults compared to employed adults. The participation gap is equal to the participation rate for employed adults minus the participation rate for unemployed adults. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Full-time workers participate in adult learning more than part-time workers
On average, individuals who work full-time (i.e. over 30 hours per week) are more likely to participate in adult learning, at 48% compared to 40% (Figure 2.17). The largest gaps are in the Flemish Region (Belgium), Japan, and Singapore, at over 15 percentage points. These three countries also have the largest relative gaps, with part-time‑employed adults less than 70% as likely to report participation in adult learning in the past 12 months. In Austria, Croatia, Italy, Korea, and Poland, part-time‑employed adults are more likely than full-time‑employed adults to participate in learning.
For five countries, the gap between the participation rate for part-time‑employed adults and full-time‑employed workers narrowed between Cycle 1 and Cycle 2 (Figure 2.18). This change was driven primarily by declines in the participation rate for full-time employed workers. The gap declined by 4 percentage points, on average across all OECD countries and economies.
Participation in adult learning by full-time employment status, percentage
Note: Employed adults aged 25‑65; formal and non-formal job-related learning. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Percentage point change in participation gaps by full-time employment status
Note: Employed adults aged 25‑65; formal and non-formal job-related learning. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap for part-time employed adults compared to full-time employed adults. The participation gap is equal to the participation rate for full-time employed adults minus the participation rate for part-time employed adults. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Participation in adult learning is not only influenced by socio-demographic and economic factors but also varies significantly between industrial sectors and occupations. Knowledge‑intensive sectors, such as technology, finance, and healthcare, often offer more opportunities for adult learning and skills development. These industries tend to invest heavily in training and development, ensuring that their workforce remains competitive in a rapidly changing labour market. In contrast, workers in low-skill or precarious sectors, such as retail, manufacturing, or construction, face more limited opportunities for professional development. This disparity is further exacerbated by employer attitudes toward training, with those in high-skill sectors more likely to see workforce education as a strategic investment and support adult learning for their employees (OECD, 2024[5]). Additionally, as occupational structures have evolved over time, shifts in industry demands have reinforced these inequalities, with high-skill roles increasingly requiring continuous learning to keep pace with technological advances. This section explores these disparities, highlighting how industry-specific factors and occupational hierarchies contribute to unequal access to adult learning opportunities.
Workers in skilled occupations are far more likely to participate in adult learning
Workers in skilled occupations are the most likely to participate in adult learning for every country and economy; in many cases, the gap between even skilled occupations, such as health technicians or lawyers, and semi-skilled white‑collar occupations, such as office clerks or customer service representatives, remains quite stark; in 11 countries, for example, the participation rate for semi-skilled white collar workers is closer to the participation rate for workers in elementary occupations (such as cleaners or kitchen assistants) than it is to skilled occupations (Figure 2.19). Countries such as Finland, Norway, and Sweden have been most successful in enabling access to learning for the least-skilled workers: participation for workers in elementary occupations in these countries is still higher (above 40% in all three countries) than the average participation rate for semi-skilled blue collar workers across OECD countries and economies, at 30%.
Participation in adult learning by skill level of occupation, percentage
Note: Employed adults aged 25‑65; formal and non-formal job-related learning. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
The gap between adult learning for skilled occupations and elementary occupations has shrunk by 4 percentage points, on average, since Cycle 1. Seven countries saw significant decreases in the gap between participation rates for adults in high-skilled and low-skilled occupations. The greatest decrease, by far, was in Korea, where the gap shrank by 27 percentage points between cycles. This was driven largely by the fact the declining participation was much more pronounced among individuals in high-skilled occupations.
Percentage point change in participation gap by skill level of occupation
Note: Employed adults aged 25‑65; formal and non-formal job-related learning. Figure plots the change, in percentage points, between Cycle 1 and Cycle 2 in the participation gap for lower-skilled occupations compared to higher-skilled occupations. The relative participation rate is equal to the participation rate for high-skilled minus the participation rate for low-skilled. Cycle 2 data refer to 2023; data for Cycle 1 refer to 2012, except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). See Box 1.3 for a note on comparison of data between Cycle 1 and Cycle 2. Results that are not statistically significant at the 5% level (p > 0.05) are shown in striped bars. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
There is considerable variation in adult learning participation between industries. Those sectors where it is most prevalent, such as the public sector, finance and insurance or utilities, have participation rates around 60%. This is more than double the rate for sectors such as agriculture, forestry, and fishing and those working in hospitality. Manufacturing – one of the largest sectors of employment – is still significantly lower, at 40%.
Average participation rate in adult learning by industrial sector
Note: Employed adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Industries are grouped according to the International Standard Industrial Classification of All Economic Activities (ISIC) one‑digit industry codes. Average participation rates are the unweighted averages across all participating OECD countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
These disparities likely reflect each industry’s skill requirements and the capacity – or willingness – of employers to invest in employees. Where an industry is more reliant on highly skilled workers who need to remain up to date on advances in technology or who can adapt to changing skill and knowledge requirements, these industries are also more likely to invest themselves in training their employees. Figure 2.22 plots the relationship between the participation rate within a given industry and the share of trainings that were paid for by employers.
Relationship between overall participation in adult learning and the prevalence of employer-supported learning
Note: Employed adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. Industries are grouped according to the International Standard Industrial Classification of All Economic Activities (ISIC) one‑digit industry codes. OECD is an unweighted average of all participating member countries. See Box 1.1 for a precise definition of job-related adult learning.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2012, 2015, 2018, 2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
Many of the individual characteristics discussed in this section are associated with one another; for instance, earnings are correlated with adults’ age and proficiency level, and parental education is associated with educational attainment. In this case, it can be a challenge to understand whether a variable directly affects participation outcomes, or only does so via its association with another variable. To better understand the direct association of one individual characteristic with the likelihood that an adult will participate in adult learning, it is useful to estimate the probability of participation within a regression framework, controlling for multiple characteristics that may influence the likelihood of adult learning participation. Table 2.1 summarises the results from a series of models that regress this likelihood on a variety of individual characteristics.
The leftmost column summarises the associations between each variable and the likelihood of participating in training across OECD countries and economies without including any other control variables. Subsequent columns contain various combinations of control variables. Notably, age is an important predictor of adult learning participation; even when accounting for one’s education or whether one has children – among other variables – there is a 12% to 20% decline in the probability of participating in adult learning when comparing adults aged 55‑65 to those aged 25‑34. This is different from parental education, where the estimated effect of an increase in parental education from upper secondary to tertiary education on the probability of participation is 23% when no other variables are included in the model, but falls to under 4% if one controls for wages, the skill level of one’s occupation, and a suite of other socio-demographic variables. This indicates that the association between parental education and participation is largely mediated through these other variables, such as an individual’s own education level (14%) or wages.
This analysis also provides a more nuanced view of how gender, immigrant background, and variables on one’s family and household interact. Consistent with the findings presented above, the estimated effect sizes of gender remain quite small across all models, and the effect of being born in one’s country of residence (native born) is consistently between 4% and 5%, even when controlling for other variables. On the other hand, in a simple bivariate regression, having young children (under 14 years of age) is associated with a 6% increase in the likelihood of participation. Nevertheless, once additional socio-demographic controls are included in the regression the picture becomes more complex, and the effects of having children are, in fact, ambiguous.
Variables such as educational attainment, wages, the skill requirements of one’s occupation, and whether or not one is employed in the first place remain important predictors of participation in adult learning: the (statistically significant) estimated effects of these variables range from 10% to 20% even in regression models that control for other socio-demographic characteristics. The most powerful predictor of participation is occupational skill (model 6). Even after accounting for age, gender, social background, education, and wages, being in a high-skilled white‑collar occupation is associated with a 22% increase in the likelihood of participating in training compared to adults in occupations with elementary skill requirements. Efforts to reduce disparities in access to adult learning should begin from an understanding that access is strongly mediated by the nature of one’s employment.
Percentage change in probability of participation in adult learning for each variable
Note: Adults aged 25‑65; formal and non-formal job-related learning in the 12 months prior to the survey. See Box 1.1 for a precise definition of job-related adult learning. P-values are reported in parentheses. Statistical significance at the 1% level is denoted by **; statistical significance at the 5% level is denoted by *. This table presents the results of an ordinary least squares (OLS) regression of participation in formal and job-related non-formal adult learning on each of the variables listed in the leftmost column (i.e. a linear probability model). The results are for the OECD average. Column 1 contains the results of individual bivariate regressions with one single independent variable. Column 2 contains the full set of socio-demographic variables in a single multivariate model. Column 3 adds a variable for whether one is employed or unemployed (inactive adults are dropped from the sample). Column 4 includes variables on wages and the skill level of one’s occupation (four categories); unemployed and inactive adults are dropped from the sample. Column 5 includes a control for one’s industry of employment.
Source: OECD (n.d.[1]), Survey of Adult Skills (PIAAC) (2023) Database, www.oecd.org/en/about/programmes/piaac/piaac-data.html. Underlying data are reported the accompanying Appendix Tables, available at www.oecd.org/en/publications/trends-in-adult-learning_ec0624a6-en/support-materials.html.
References
[2] Blossfeld, H., E. Kilpi-Jakonen and D. de Vilhena (2020), “Is there a Matthew effect in adult learning? Results from a cross-national comparison / Gibt es im lebenslangen Lernen einen Matthäus-Effekt? Ergebnisse eines internationalen Vergleichs”, Monetäre und nicht monetäre Erträge von Weiterbildung. Edition ZfE, Vol. 7, https://doi.org/10.1007/978-3-658-25513-8_1.
[4] OECD (2024), Do Adults Have the Skills They Need to Thrive in a Changing World?: Survey of Adult Skills 2023, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/b263dc5d-en.
[3] OECD (2024), Survey of Adult Skills – Reader’s Companion: 2023, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/3639d1e2-en.
[5] OECD (2024), Understanding Skill Gaps in Firms: Results of the PIAAC Employer Module, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/b388d1da-en.
[1] OECD (n.d.), Survey of Adult Skills (PIAAC) Database, http://www.oecd.org/en/about/programmes/piaac/piaac-data.html (accessed on 11 June 2025).
← 1. Trend data for Cycle 1 refer to 2012 for all countries except for Chile, Israel, Lithuania, New Zealand, and Singapore (2015) and Hungary (2018). Data for Cycle 2 are for 2023. It should be noted that Cycle 2 data were collected during a period when countries were experiencing labour market fluctuations due to the COVID‑19 pandemic.
← 2. It should also be noted that the adult learning participation rate of younger adults may be over-estimated, due to the fact that the indicator cannot exclude those who are still completing their initial cycle of studies (see Chapter 1).
← 3. Moreover, the concept of ‘job-related training’ may be interpreted differently by employed and unemployed individuals, potentially affecting the comparability of results.