A note regarding Israel
The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.
This report represents the final phase of the first cycle of the Survey of Adult Skills, with the release of results from the six countries participating in the third round of data collection: Ecuador, Hungary, Kazakhstan, Mexico, Peru and the United States. Five of these countries undertook the survey for the first time while one, the United States, had also collected data as part of the first round in 2011-12. With the completion of Round 3 of the study, a total of 39 countries and economies have participated in the study (see Box 1.1). The data from the survey provide an unprecedented insight into the level and distribution of key information-processing skills in the adult population and the relationships between proficiency in these skills and individuals’ educational and social background, and labour-market experience, as well as the nature of their working arrangements and work tasks across a significant group of countries.
The results of the two previous rounds of the survey of Adult Skills can be found in the two summary international reports of the study (OECD, 2013[1]; OECD, 2016[2]). In addition, there are a number of other published studies analysing results from the survey. These include thematic reports and working papers published by the OECD as well as many national reports and academic papers (Maehler, Bibow and Konradt, 2018[3]).
The purpose of this report is primarily to present a summary of the results for the countries participating in Round 3 of the Survey of Adult Skills. It also presents data from countries and economies participating in earlier rounds of the survey as they serve as useful benchmarks and reference points in order to put the results from the Round 3 countries into context. In particular, the report references the average scores of OECD countries participating in PIAAC across the three rounds. However, the report does not contain detailed analysis of the results of countries or economies that participated in earlier rounds. Although the report generally follows the structure used in previous international reports, the analysis presented in Chapter 4 on skills offers a new presentation of these data, concentrating on the use of numeracy skills and the intensity with which adults engage in practices using these skills, both in everyday life and in the workplace.
The Survey of Adult Skills, a product of the Programme for the International Assessment of Adult Competencies (PIAAC), measures the proficiency of working-age adults (16-65 year-olds) in three key information-processing skills: literacy, numeracy and problem solving in technology-rich environments. These key skills are relevant to adults in many social contexts and work situations, and necessary for fully integrating and participating in the labour market, education and training, and social and civic life (see Box 1.1 for more information).
The survey provides a rich source of data for policy makers, analysts and researchers concerned with issues such as the development and maintenance of a population’s skills, the relationships between the education system and the labour market, the efficiency of the labour market in matching workers and jobs, inequality, and the social and labour-market integration of certain subgroups of the population such as immigrants. Beyond offering an insight into the level and distribution of information-processing skills across the population as a whole and for key subgroups, it provides information on the benefits these skills provide in the labour market and in everyday life.
The interest of the results from Round 3 of the survey lies not just in the fact that additional countries have undertaken the survey but also in that:
Four of the six participants – Ecuador, Kazakhstan, Mexico and Peru – are upper-middle income countries (see Box 1.2 for more details).
Measures of the proficiency of the adult population in the United States are available from two different points of time, as it participated in the survey twice (see Box 1.4 and Annex 1.A1. Description of participation of the United States in PIAAC Cycle 1 for more details).
Most countries that have participated in the Survey of Adult Skills have been high-income countries: prior to the third round, only three middle-income countries (Indonesia, the Russian Federation and Turkey) had taken part. The additional middle-income countries taking part in the survey have added to its comparative dimension. In addition, the participation of these countries is evidence of the relevance of the data from PIAAC for policy makers and analysts in such countries as well as providing further evidence that it is feasible to collect data on literacy and numeracy in middle-income countries in the form of large-scale population assessments.
The measurement of literacy and numeracy among adults in low- and middle-income countries is an issue that has gained considerable importance in the context of the United Nations Sustainable Development Goals (SDGs) which includes a target (SDG Goal 4, Target 4.6) of ensuring that all youth and a substantial proportion of adults, both men and women, achieve literacy and numeracy by 2030 (UNSD, 2018[4]). PIAAC represents the only currently operating international assessment of adult literacy and numeracy. In the context of considering how to measure progress towards the SDGs, evidence on the relevance of the PIAAC measures and the feasibility of its implementation is important. The experience of the third round of the Survey of Adult Skills confirms the experience from earlier rounds, and the World Bank’s Skills Towards Employment and Productivity (STEP) Skills Measurement Program, that the PIAAC instruments can be effectively administered in middle- and low-income countries. STEP is an initiative to measure skills in low- and middle-income countries that includes a version of the PIAAC literacy assessment [see Annex 1.A2. Skills Towards Employment and Productivity (STEP) Survey for further information]. At this point, 17 countries have participated in STEP.
What the survey measures
The Survey of Adult Skills (PIAAC) assesses the proficiency of adults from the age of 16 to 65 years in literacy, numeracy and problem solving in technology-rich environments. These skills are key information-processing competencies that are relevant to adults in many social contexts and work situations, and necessary for full integration and participation in the labour market, education and training, and social and civic life.
In addition, the survey collects a range of information on the reading- and numeracy-related activities of respondents, their use of information and communication technologies at work and in everyday life, and on a range of generic skills, such as collaborating with others and organising their time, required of individuals in their work.
Respondents are also asked whether their skills and qualifications match their work requirements and whether they have autonomy over key aspects of their work.
Data collection
The first cycle of the Survey of Adults Skills has been conducted over three rounds of data collection.
The first round surveyed around 166 000 adults aged 16-65 years in 24 countries (or regions within these countries) in 2011-12. In Australia, Austria, Canada, Cyprus,1 the Czech Republic, Denmark, Estonia, Finland, France, Germany, Ireland, Italy, Japan, Korea, the Netherlands, Norway, Poland, the Slovak Republic, Spain, Sweden and the United States – the sample was drawn from the entire national population. In Belgium, the data were collected in Flanders; in the United Kingdom, the data were collected in England and Northern Ireland (data are reported separately for England and Northern Ireland in the report). In the Russian Federation,2 the data do not cover the Moscow municipal area.
Nine countries (or regions within these countries) took part in a second round of data collection in 2014-15: Chile, Greece, Jakarta (Indonesia), Israel, Lithuania, New Zealand, Singapore, Slovenia and Turkey. A total of 50 250 adults were surveyed. In all countries except Indonesia, the entire national population was covered. In Indonesia, the data were collected in the Jakarta municipal area only.
The third round was conducted in 2017-18 in six countries: Ecuador, Hungary, Kazakhstan, Mexico, Peru and the United States. A total of 34 792 adults were surveyed. Note that the United States had already participated in Round 1 (see Box 1.4 for further details). This brought the number of participating countries and economies to a total of 39.
Key features of the sampling and survey administration
Participating countries chose the language they used to administer the assessment. This was commonly the official language(s) of each participating country/economy, but in a few countries, the assessment was also conducted in widely spoken minority or regional languages.
Three skills domains were assessed: literacy, numeracy and problem solving in technology-rich environments. In addition, a separate assessment of “reading components” was conducted, with the purpose of testing basic reading skills, such as vocabulary knowledge, understanding of the logic of sentences and fluency in reading passages of text.
Five countries chose not to conduct the problem-solving assessment: Cyprus,1 France, Italy, Jakarta (Indonesia) and Spain. Four countries (France, Finland, Japan and the Russian Federation) chose not to conduct the assessment of reading components.
The target population for the survey was the non-institutionalised3 population of 16-65 year-olds residing in the country or region at the time of the data collection, irrespective of nationality, citizenship or language status. Sample sizes depended primarily on the number of cognitive domains assessed and the number of languages in which the assessment was administered. Some countries increased the size of the sample in order to have reliable estimates of proficiency for the residents of particular geographical regions and/or for certain subgroups of the population, such as indigenous inhabitants or immigrants. The national samples achieved ranged from a minimum of approximately 4 000 individuals to a maximum of nearly 27 300 individuals.
The survey was administered under the supervision of trained interviewers either in the respondent’s home or in a location agreed between the respondent and the interviewer. The background questionnaire was delivered in Computer-Aided Personal Interview (CAPI) format by the interviewer. Depending on the situation of the respondent, it took between 30 and 45 minutes to complete the questionnaire.
After answering the background questionnaire, the respondent completed the assessment on a laptop computer (provided they showed sufficient computer skills). Adults lacking basic computer skills or experience, or refusing for other reasons to take the assessment on a computer, were administered a paper version of the assessment on printed test booklets. Respondents could take as much or as little time as needed to complete the assessment. All respondents taking the paper-based assessment also undertook the assessment of reading components. On average, respondents took 50 minutes to complete the cognitive assessment.
Identical instruments were used in all countries in all rounds of the survey.
1. Note by Turkey: The information in this document with reference to “Cyprus” relates to the southern part of the Island. There is no single authority representing both Turkish and Greek Cypriot people on the Island. Turkey recognises the Turkish Republic of Northern Cyprus (TRNC). Until a lasting and equitable solution is found within the context of the United Nations, Turkey shall preserve its position concerning the “Cyprus issue”.
Note by all the European Union Member States of the OECD and the European Union: The Republic of Cyprus is recognised by all members of the United Nations with the exception of Turkey. The information in this document relates to the area under the effective control of the Government of the Republic of Cyprus.
2. See note at the end of this chapter.
3. The target population excludes adults in institutional collective dwelling units (or group quarters) such as prisons, hospitals and nursing homes, as well as adults residing in military barracks and military bases. However, full-time and part-time members of the military who do not reside in military barracks or military bases are included in the target population.
In this report, countries have been classified by income levels based on the methodology and taxonomy adopted by the World Bank (World Bank, 2019[5]). For the current 2019 fiscal year:
1 low-income economies are defined as those with a gross national income (GNI) per capita of USD 995 or less in 2017
2 lower middle-income economies are those with a GNI per capita between USD 996 and USD 3 895
3 upper middle-income economies are those with a GNI per capita between USD 3 896 and USD 12 055
4 high-income economies are those with a GNI per capita of USD 12 056 or more.
Details of the methodology used for calculating GNI and converting GNI in national currencies to US dollars can be found in the World Banks’s comprehensive repository of documents (World Bank, 2019[5]).
The majority of countries that participated in the first cycle of the Survey of Adult Skills are high-income countries. The exceptions are the Russian Federation (Round 1); Indonesia and Turkey (Round 2); and Ecuador, Kazakhstan, Mexico and Peru (Round 3). All of these countries, with the exception of Indonesia (which is a lower-middle income economy), are upper-middle income countries under the World Bank classification. Round 3 of the Survey is notable for the fact that the majority of participants (four out of six) are upper-middle income countries. In interpreting the performance of adults in these countries, it is helpful to compare their performance with those in other countries of similar income levels participating in the study.
It is important to note here the difference between the measures of gross domestic product (GDP) per capita and GNI per capita (which are often used interchangeably). GDP per capita only counts income from domestic sources, i.e. it measures only the domestic residents’ income received from domestic production of final goods and services. GNI per capita also includes income received from abroad i.e. it adjusts net income received by domestic residents from production abroad to their income from domestic production. For most nations, there is little difference between GDP and GNI, since the difference between incomes received by the country versus payments made to the rest of the world tends not to be significant. For instance, the United States’ GNI was only about 1.01% higher than its GDP in 2016, according to the World Bank (World Bank, 2019[6]). For some countries, however, the difference is significant: GNI can be much higher than GDP if a country receives a large amount of foreign aid. It can be much lower if foreigners control a large proportion of a country’s production, as is the case with Ireland, a low-tax jurisdiction where the European subsidiaries of several multinational companies (nominally) reside.
The average adult proficiency in information-processing skills varies considerably among the 39 countries and economies covered by the Survey of Adult Skills, although many of the average scores fall within a relatively limited range. The differences between countries and economies in the study partly reflect the different starting points and economic, educational and social development pathways that they have followed over the past half century, as well as current institutional arrangements and policies.
Among the countries participating in Round 3 of the study, adults in Hungary and the United States performed close to the average for the OECD countries and economies that participated in PIAAC over the three rounds in all three domains (see Figure 1.1 and refer to Chapter 2 for more details). More specifically, Hungary’s numeracy scores were above average while its literacy scores were below average, albeit only slightly. The opposite was the case in the United States. In both these countries, the proportions of adults reaching Level 2 or 3 in problem solving in technology-rich environments were not significantly different from the OECD average. In contrast, adults in the Latin American middle-income countries from Round 3, Ecuador, Mexico and Peru, performed well below the average for OECD countries and were among the countries with the lowest average proficiency in absolute terms in the three domains assessed. The proficiency of working-age adults in these three countries is very similar to that observed in Turkey (another middle-income country) in Round 2. These results are in line with studies of school-age children in the Programme for International Student Assessment (PISA) which found that among economies with a per capita GDP below USD 20 000 (such as Chile, Mexico, Peru and Turkey), the greater the country’s wealth, the higher its mean score on the PISA reading test. This indicates a positive relationship between per capita national income and performance, at least until a minimum threshold is reached (OECD, 2012[7]; OECD, 2018[8]).
Kazakhstan, despite also being a middle-income country, falls somewhere between these two groups of Round 3 countries. The proportion of adults scoring at the highest levels in literacy, numeracy and problem solving is below that seen in Hungary and the United States but above the share in Ecuador, Mexico and Peru. Close to half of the adult population in Kazakhstan performs at Level 2 in both the literacy and numeracy domains and the proportion of the population scoring at Level 1 and below is close to the OECD average.
As well as the differences observed across countries, there is considerable variation in proficiency in literacy and numeracy within the countries participating in Round 3. In Ecuador, Peru and the United States the difference between the top- and bottom-performing 25% of adults was 6-13 score points larger than OECD average in literacy and 7-23 score points larger than the OECD average in numeracy. In Mexico and Hungary the gap between the best and worst performers in both literacy and numeracy was close to the OECD average while in Kazakhstan, it was lower than the OECD average.
In line with their low average scores, Ecuador, Mexico and Peru have very large proportions of adults at the lowest levels of the proficiency scales. For example, in these three countries, more than 60% of adults scored at or below Level 1 in literacy and numeracy, meaning they would struggle to understand complex texts or perform numerical tasks involving several steps and mathematical information represented in different ways (see Box 1.3). However, despite the high proportions of adults in these three countries with very low literacy skills, there are, nevertheless, few adults who are actually illiterate.
As mentioned above, the Survey of Adult Skills includes an assessment of reading components designed to assess mastery of the basic components of reading comprehension – vocabulary knowledge (print vocabulary), understanding of the logic of sentences (sentence processing) and reading fluency (passage comprehension) – for adults who failed to complete a set of very simple tasks correctly. Even in Ecuador, Mexico and Peru (which have very high proportions of adults performing at or below Level 1 on the literacy scale) those failing the core test in these countries correctly answered more than 77% of the items in the sentence processing elements of the reading components assessment, more than 74% of the passage-comprehension items and 92% of the print-vocabulary items.
In each of the three domains assessed, the results are represented on a scale from 0 to 500.
Each of the three proficiency scales is divided into “proficiency levels”, defined by particular score-point ranges. Six proficiency levels are defined for literacy and numeracy (from below Level 1 to Level 5) and four for problem solving in technology-rich environments (from below Level 1 to Level 3).
The results for literacy and numeracy are presented in the form of mean proficiency scores for each country as well as by proportions of the population by proficiency level. For problem solving in technology-rich environments, given the very different levels of familiarity with computer applications in the countries and economies participating in the Survey of Adult Skills, the proportions of the population to which the estimates of proficiency in this domain refer vary widely among countries/economies. In other words, the populations for whom proficiency scores for problem solving in technology-rich environments are reported are not identical across countries. Proficiency scores relate only to the proportion of the target population in each participating country that was able to undertake the computer-based version of the assessment, and thus meets the preconditions for displaying competency in this domain. For this reason, the presentation of the results focuses on defining the proportions of the population at each proficiency level rather than on comparing mean proficiency scores.
The proficiency levels are designed so the scores represent degrees of proficiency in a particular aspect of the domain. Each level is associated with a certain number of items, with higher levels being associated with items of increasing difficulty. There are easier and harder tasks for each proficiency scale. The purpose of described proficiency scales is to facilitate the interpretation of the scores assigned to respondents. That is, respondents at a particular level not only demonstrate knowledge and skills associated with that level but also the proficiencies required at lower levels. Thus, respondents scoring at Level 2 are also proficient at Level 1, with all respondents expected to answer at least half of the items at that level correctly.
For more information on the proficiency levels in each domain and their descriptions, please refer to Chapter 2.
In all the countries participating in PIAAC, there were many adults with no experience using computers or who had extremely limited ICT skills, or who showed low levels of proficiency in the problem solving in technology-rich environments domain. Around one in four adults have no or only limited experience with computers or lack confidence in their ability to use computers. In addition, nearly half of all adults are only proficient at or below Level 1 in problem solving in technology-rich environments, which translates into being able to use only familiar applications to solve problems that involve few steps and explicit criteria, such as sorting e-mails into pre-existing folders.
Round 3 countries differ from each other significantly in the share of adults without basic ICT skills or who failed the core ICT test. While Hungary and Kazakhstan had a similar share to the OECD average of adults with no or little ICT experience (14.4% and 19.7% respectively), and the United States had an even smaller share at 7.4%, other countries participating in Round 3 stand out as having very large proportions of their adult populations with no prior computer experience or very poor ICT skills: 32.9% in Ecuador, 39.3% in Mexico and 43.6% in Peru. These countries are comparable to Turkey, where around 38% of adults have little or no ICT experience. These figures should be understood in the context of these countries’ economic development and the level of ICT penetration. In 2017, only about one-third of the households in Ecuador (38.1%) and Mexico (36.9%) had a fixed line phone subscription, and the share in Peru was significantly lower (21.9%). Internet and computer access in these countries is also limited: only around 40% of households had access to a computer and functional Internet in Ecuador and Mexico in 2017 and the share of such households in Peru was even lower, at around 30% (ITU, 2019[9]). This is in stark contrast to many of the high-income OECD countries where more than two-thirds of the households have access to a computer, the Internet and a telephone line. The proportion of adults lacking computer experience or having very low ICT skills is therefore in line with expectations.
Mean proficiency scores of 16-65 year-olds in literacy and numeracy, and the percentage of 16-65 year-olds scoring at Level 2 or 3 in problem solving in technology-rich environments
Note: Cyprus1, France, Italy and Spain did not participate in the problem solving in technology-rich environments assessment.
1. See note 1 in Box 1.1.
2. See note at the end of this chapter.
Countries are listed in alphabetical order.
Source: Survey of Adult Skills (PIAAC) (2012, 2015, 2018), Tables A2.2, A2.4 and A2.7.
Within countries and economies, adults with different socio-demographic characteristics show considerable variation in their proficiency in information processing. In particular, proficiency is closely associated with age, educational attainment and parents’ level of education, but only weakly associated with gender.
As expected, in all countries and economies there is a close association between adults’ educational attainment and their proficiency in information-processing skills. This is likely to be because, on the one hand, adults with greater proficiency are more likely to participate in higher levels of education and, on the other, longer periods of study provide the opportunity to develop greater levels of proficiency. Among 25-65 year-olds (i.e. adults who have generally completed formal education), proficiency is highest among those with tertiary qualifications and lowest among those whose highest qualification was below upper secondary education (see Figure 1.2 and refer to Chapter 3 for more details).
In Hungary, tertiary-educated adults scored higher than the average for tertiary-educated adults across participating OECD countries, by about 4 points in literacy and by about 18 points in numeracy. Hungary has also one of the highest share of tertiary-educated adults scoring at Level 4 and 5 in numeracy (33%, compared to 23% across the OECD; Sweden has the highest share, at 36%). Tertiary-educated adults in the United States have similar proficiency in literacy to their Hungarian counterparts, but they scored lower in numeracy, below the OECD average. There is a very small gap in the proficiency (in both literacy and numeracy) between tertiary-educated adults and adults with below upper secondary education in Kazakhstan. This is due to the fact that tertiary-educated adults score more than 30 points below the OECD average, in both domains, but adults without an upper secondary qualification scored above the average, by 6 points in literacy and by 16 points in numeracy. In Ecuador, Mexico and Peru, performance in literacy and numeracy is consistently below the corresponding OECD average for adults at each level of educational attainment.
Proficiency is especially low among adults without an upper secondary qualification in Peru: they averaged 157 score points in literacy and 127 in numeracy, well below average for similarly educated adults in other Latin American countries such as Chile (177 score points in literacy and 154 score points in numeracy), Ecuador (174 and 160 score points) and Mexico (201 and 189 score points).
In most countries, the relationship between age and proficiency tends to follow an inverted U-shaped curve, with a peak between the mid twenties and the early thirties. In contrast, among Round 3 countries like Ecuador, Mexico and Peru, proficiency declines more or less steadily with increasing age. As PIAAC is a cross-sectional survey, the age-skill relationship cannot be interpreted exclusively as the effect of ageing: differences in the age-skills profile are influenced by differences in educational attainment among different cohorts as countries underwent periods of economic development and the expansion of education at different times in their history.
In the United States, 55-65 year-olds are more likely to have a tertiary degree than in many other countries and the gap in educational attainment between 25-34 year-olds and these older adults is very small. Rates of completion of tertiary education for 55-65 year-olds in Kazakhstan are about half the rates observed among adults aged 25-34 (27% compared to 50%). The share of adults who have not attained an upper-secondary qualification is similar in the two age groups (14% for older adults, 11% for 25-34 year-olds), meaning that over time there has been an increasing share of adults who have progressed from a secondary to a tertiary qualification. This upgrade in educational attainment does not appear to have translated in a corresponding increase in the skills of the adult population, possibly because of a decline in the quality of education. The profile in Ecuador, Mexico and Peru is likely to reflect the fact that completion rates for upper secondary education in these countries have increased only very recently. On average across OECD countries, only 16% of 25-34 year-olds have not completed upper secondary education, compared to 50% in Mexico, 36% in Ecuador and 26% in Peru. Among those under 25, the share of respondents who have completed upper secondary is actually higher than the OECD average in Ecuador and Peru (52% and 68%, respectively, compared to an average of 49%), and is not very distant in Mexico (36%). As such, the age-skills profile in these countries is quite similar to that observed in more developed economies like Korea and Singapore, which have also only more recently expanded access to education.
The difference in literacy proficiency between men and women is negligible. Men have a more substantial advantage in numeracy, scoring about 10 score points higher than women on average. Hungary and Kazakhstan are among the few countries where there is no gender difference in numeracy proficiency. In Hungary, this is mainly due to the very strong performance of Hungarian women. In Kazakhstan both men and women score below the OECD average, with the gap being much less pronounced for women, at only 9 score points as opposed to men at 21 score points.
Gender gaps in proficiency are more pronounced among older adults (aged 45 years and over). This could either reflect the fact that gender gaps in educational attainment are wider among older adults, or that women’s skills have declined more over time, possibly because they participate less in the labour market.
Parents’ educational background also exerts a significant influence on adults’ literacy proficiency. Having at least one parent with a tertiary qualification is associated with a 41 score-point advantage over adults who do not have a parent with an upper secondary education. Gaps related to family background are particularly pronounced in Hungary, Peru and the United States among the Round 3 countries (see Figure 3.12 in Chapter 3). The differences are very close to the OECD average in Ecuador and Mexico and they are much smaller (but still significant) in Kazakhstan. About half of this difference is explained by other socio-demographic characteristics, most notably the fact that the children of highly educated parents are themselves more likely to attain higher levels of education. This is especially true in Mexico, where adjusting for individual characteristics strongly reduces the differences related to family background, and less true in Ecuador and Kazakhstan, where the adjustment has less of an effect.
A. Mean literacy proficiency scores, by educational attainment (adults aged 25-65)B. Difference in mean literacy score between low- and high-educated adults (adults aged 25-65)
Notes: All differences in Panel B are statistically significant. Unadjusted differences are the differences between the two means for each contrast category. Adjusted differences are based on a regression model and take account of differences associated with other factors: age, gender, immigrant and language background and parents' educational attainment. Only the score-point differences between two contrast categories are shown in Panel B, which is useful for showing the relative significance of educational attainment vis-a-vis observed score-point differences. Lower than upper secondary includes ISCED 1, 2 and 3C short. Upper secondary includes ISCED 3A, 3B, 3C long and 4. Tertiary includes ISCED 5A, 5B and 6. Where possible, foreign qualifications are included as the closest corresponding level in the respective national education systems. Adjusted difference for the Russian Federation is missing due to the lack of the language variables.
1. See note at the end of this chapter.
2. See note 1 in Box 1.1.
Countries and economies are ranked in ascending order of the unadjusted differences in literacy scores (tertiary minus lower than upper secondary).
Source: Survey of Adult Skills (PIAAC) (2012, 2015, 2018), Tables A3.1(L) and A3.2(L).
The United States has collected three waves of data using the PIAAC instruments. It collected data as part of Round 1 of Cycle 1 of PIAAC in 2011-12. It then collected additional data for targeted population groups as part of a National PIAAC Supplement (Rampey et al., 2016[10]) in 2014 and participated in Round 3 of Cycle 1. Details of the PIAAC data collection in the United States can be found in the technical reports for the survey and the National PIAAC Supplement (Hogan et al., 2016[11]; OECD, 2019[12]).
In this report, the United States is reported using 1) the combined data from 2012 and 2014; and 2) the 2017 data collection conducted as part of Round 3. The 2012/14 data set has been used as this is believed to provide a more accurate representation of the proficiency of the working-age population at that point in time than the original 2011-12 data reported in the first two international reports on PIAAC (OECD, 2013[1]; OECD, 2016[2]). In addition to the increased sample size, the 2012/14 data have been weighted to control totals related to the 2010 census whereas the 2011-12 data were weighted to totals related to the census in 2000.
Data from the United States are presented in the following way in this report:
Results from 2012/14 and 2017 are presented as separate observations in tables and charts.
The United States contributes to the average of OECD countries as one observation. This is calculated as the mean of the relevant statistic for the two US observations (i.e. in 2012/14 and 2017).
In this report, all discussion of U.S. results refers to the year 2017, unless otherwise specified.
See Annex 1.A1. Description of participation of the United States in PIAAC Cycle 1 for further details on the sample size and the administration of the survey in the United States.
In addition to assessing proficiency in literacy, numeracy and problem solving in technology-rich environments, the Survey of Adult Skills (PIAAC) collects information on how often adults engage in tasks requiring the use of these skills, both in everyday life and at work – for example, reading different types of text, undertaking calculations and solving problems. The PIAAC background questionnaire collects information on the frequency of these practices for a number of reasons. First, engagement with written materials and the mathematical demands of adult life represents an important dimension of what it is to be literate and numerate in terms of the definitions of these constructs in the study. Second, practice is understood as a means by which individuals develop and maintain proficiency during their working life. Third, in the workplace, individual productivity and wages are determined both by workers’ proficiency and the intensity with which they engage in practices that use their proficiency.
Generally, countries ranking low for the use of numeracy skills in everyday life also rank low for their use at work, while countries at the upper end of the distribution for one also rank high for the other. This suggests that the use of skills in everyday life and at work are highly, albeit imperfectly, correlated at the country level.
Based on an index of engagement in numeracy practices that reflects both the frequency and sophistication of their use (refer to Box 4.1 in Chapter 4), Ecuador, Kazakhstan, Mexico and Peru rank in the lower part of the distribution of engagement in numeracy practices. As such, they are similar to Chile from Round 2. Hungary, in contrast, displays lower intensity in engagement in numeracy practices at work than the average OECD country, and greater intensity in everyday life.
Numeracy proficiency and engagement in numeracy practices are positively but weakly correlated at the country level among high-income countries, i.e. higher average numeracy scores tend to correspond to higher average values for the index of numeracy use. The correlation strengthens when non-high income countries are also considered, in particular Ecuador, Mexico and Peru.
In almost all participating countries and economies, men engage in numeracy practices more frequently than women, both at work and in everyday life. Controlling for other personal and job-related characteristics reduces the gender gap, especially for the intensity of use in everyday life, but does not reverse it. In all participating countries, 55-65 year-old workers engage in numeracy practices at work less intensively than 25-54 year-olds. The youngest workers (16-24 year-olds) also use numeracy practices less intensively than 25-54 year-olds, except in Kazakhstan, Mexico, Peru and the Russian Federation.
Respondents with higher educational qualifications engage in numeracy practices more intensively than upper secondary graduates, while those without an upper secondary qualification engage less intensively. These patterns hold for the intensity of numeracy use in both everyday life and at work. The gaps in the intensity of practice across attainment levels are wider in all Round 3 countries (except the United States), but especially in Ecuador, Mexico and Peru. For these three countries, the adjusted gaps in numeracy use between adults with upper secondary education and those without are two to three times larger than the average for OECD countries. In Kazakhstan, conversely, individuals with below upper secondary education do not use numeracy less intensively than individuals with an upper secondary qualification, either at work or in everyday life.
A large part of the variation in the index of numeracy practices is explained by a worker’s occupation, and by the human resource practices used in the workplace. These managerial and human relations practices involve aspects of work organisation – such as team work, autonomy, task discretion, mentoring, job rotation and applying new learning – as well as management practices such as employee participation, incentive pay, training practices and flexibility in working hours. They explain between 10% and 20% of the variation in skills use among individuals. This is in line with countries’ efforts to promote better skills use through innovation in the workplace, for example through training.
Across the OECD countries taking part in the Survey of Adult Skills in any one of the three rounds, an individual who scores one standard deviation higher than another on the numeracy scale (around 56 score points) is 1.7 percentage points more likely to be employed than unemployed. An increase in one standard deviation in the number of years in formal education (around 3.3 years) is associated with a 2.4 percentage-point increase in the chances of being employed. A similar pattern holds for Hungary, where the likelihood of employment is positively associated with numeracy proficiency and educational attainment. In Ecuador, Kazakhstan, Mexico, Peru and the United States there are low or negative returns to proficiency and education, which in most cases are not statistically significant.
In most countries, educational attainment is a better predictor of employment than numeracy proficiency, which suggests that it is harder for employers to judge workers’ actual numeracy proficiency and they are more likely to rely on readily available, albeit imperfect, signals such as educational qualifications. Meanwhile, the lack of a relationship between employment status and education and proficiency in Latin American counties is striking. It is, however, in line with previous studies on Latin American countries that have found a stronger correlation between cognitive skills and earnings than with employment status (Cunningham, Acosta and Muller, 2016[13]; Acosta, Muller and Sarzosa, 2017[14]). The absence of a strong social protection system in these countries can lead to the majority of adults dedicating themselves to any employment they can find, possibly in the informal sector (Ocampo and Gómez-Arteaga, 2017[15]). In other words, education and proficiency could have a more profound effect on the quality of employment than the quantity in Latin American countries compared to others.
Proficiency and schooling have significant and distinct effects on hourly wages (see Figure 1.3 and refer to Chapter 5 for more details). Across the OECD countries taking part in any of the three rounds of the Survey of Adult Skills, an increase in one standard deviation in numeracy proficiency is associated with a 7% increase in hourly wages, keeping years of education and other socio-demographic characteristics constant. An increase in years of education by one standard deviation brings about a bigger increase in hourly wages – about 18%, all else being equal. Returns to proficiency are above average in Hungary, while they are below average in Ecuador, Kazakhstan, Mexico, Peru and the United States (2017). The relationship is weakest in Ecuador, where it is not statistically significant. Returns to years of education exceed the OECD average in all Round 3 countries, with the exception of Peru. Hungary shows the third highest returns to years of education of all participating countries, after Singapore and Slovenia.
Mismatches between workers’ qualifications and skills and what they report as being required or expected in their jobs are pervasive in most countries participating in PIAAC. On average across the OECD countries that have taken part in the Survey of Adult Skills, about 22% of workers report that they are overqualified – that they have higher qualifications than required to get their jobs – and 12% report that they are underqualified – that they have lower qualifications than required to get their jobs. Moreover, 11% have higher literacy skills than those typically required in their job, while 4% are underskilled. Finally, 40% of workers are mismatched by field of study: they work in an occupation that is unrelated to their field of study. These forms of mismatch overlap; it is common for workers who are mismatched by field of study to also be overqualified, for example.
Percentage change in wages associated with a change of one standard deviation in years of education, proficiency in numeracy and numeracy use at work
Notes: Hourly wages, including bonuses, in purchasing power parity-adjusted USD (2012). Coefficients from the ordinay least squares regression of log hourly wages on years of education, proficiency and use of numeracy skills at work, directly interpreted as percentage effects on wages. Coefficients adjusted for age, gender, foreign-born status and tenure. The wage distribution was trimmed to eliminate the 1st and 99th percentiles. One standard deviation in proficiency in numeracy is 56 points. One standard deviation in years of education is 3.3 years. One standard deviation in numeracy at work is 0.27 points. The analysis excludes the Russian Federation because wage data obtained through the survey do not compare well with those available from other sources. Hence further checks are required before wage data for this country can be considered reliable. Statistically significant values (at the 5% level) are shown in a darker tone.
1. See note 1 in Box 1.1.
Countries and economies are ranked in ascending order of the effect of numeracy proficiency on wages.
Source: Survey of Adult Skills (PIAAC) (2012, 2015, 2018), Table A5.2(N).
The incidence of qualification mismatch varies significantly across countries. In all Round 3 countries except Ecuador and Kazakhstan, the overall qualification mismatch rate is lower than in the OECD average. Kazakhstan has an overall rate very close to the OECD average, although the composition is slightly different with overqualification playing a bigger role than on average. Ecuador has a relatively high overall rate and is one of only five PIAAC countries, where being underqualified is more common than being overqualified. This could reflect rapid growth in the demand for higher qualifications not matched by an equivalent increase in graduate numbers.
In Hungary, Kazakhstan and the United States, the overall incidence of skill mismatch is at or below the rate observed in the OECD on average. By contrast, Latin American countries stand out, with incidences that are well above average. This applies to Ecuador, Mexico and Peru from Round 3 but also to Chile from Round 2 and is mostly due to an above-average incidence of overskilling. Chile, Ecuador and Mexico, along with the United States, also have a relatively high incidence of mismatches by field of study: 10 percentage points higher than the OECD average in Chile, 17 percentage points in Ecuador, 12 percentage points in Mexico and 8 percentage points in the United States.
Qualification mismatch and skills mismatch may both have distinct effects on wages, even after adjusting for both qualification level and proficiency scores, because jobs with similar qualification requirements may have different skill requirements. This may occur because employers can evaluate qualifications but they cannot measure skills directly. When workers are compared with equally qualified and equally proficient well-matched counterparts, then being overqualified has a stronger negative association with real hourly wages than being over-skilled or having a field-of-study mismatch. On average, across countries, overqualified workers earn about 17% less than well-matched workers with the same qualification and proficiency levels and in the same field. The equivalent wage penalty for overskilling is 7% and that for field-of-study mismatch is 3% (Figure 1.4).
While the negative correlation between overqualification and wages is consistent and statistically significant across countries, this is not the case for over-skilling and field-of study mismatch. The picture for Hungary is similar to the OECD average but in Kazakhstan, Mexico, Peru and the United States, the wage penalties related to overqualification are above average. This is particularly the case in Peru and the United States where the hourly wages of overqualified workers are more than 30% lower than the hourly wages of well-matched workers who have the same level and field of qualification and the same proficiency in numeracy.
Finally, proficiency in literacy, numeracy and problem solving in technology-rich environments is positively associated with several aspects of well-being identified using PIAAC. On average in OECD countries, proficiency in information-processing skills is positively associated with trust, volunteering, political efficacy and self-assessed health. The relationships with political efficacy and self-assessed health hold even after accounting for a range of socio-demographic characteristics, but not in the case of trust. The strength of these associations differ across countries. With the exception of Hungary and the United States, countries in Round 3 have weaker relationships overall between proficiency in numeracy and non-economic outcomes than most of the other countries included in PIAAC.
Percentage difference in wages between overqualified, overskilled or field-of-study mismatched workers and their well-matched counterparts
Notes: Coefficients from ordinary least squares regression of log hourly wages on mismatch directly interpreted as percentage effects on wages. Coefficients adjusted for years of education, age, gender, marital status, working experience, tenure, foreign-born status, establishment size, contract type, hours worked, public sector dummy, proficiency in numeracy and numeracy use at work. The wage distribution was trimmed to eliminate the 1st and 99th percentiles. The regression sample includes only employees. The analysis excludes the Russian Federation because wage data obtained through the survey do not compare well with those available from other sources. Hence further checks are required before wage data for this country can be considered reliable. The analyses exclude Australia because the unavailability of ISCO 3-digit information for Australian workers in the Survey of Adult Skills (PIAAC) means field-of-study mismatch data were unavailable. Statistically significant values (at the 5% level) are shown in a darker tone.
1. See note 1 in Box 1.1.
Countries and economies are ranked in ascending order of the effect of overqualification on wages.
Source: Survey of Adult Skills (PIAAC) (2012, 2015, 2018), Table A5.7.
The completion of Round 3 of the Survey of Adult Skills brings the total of countries and economies that have participated in the study to 39. Six countries participated in Round 3 of the first cycle of the survey: Ecuador, Hungary, Kazakhstan, Mexico, Peru and the United States. Of these countries, five were undertaking the assessment for the first time, while the United States was repeating the survey, having also fielded the assessment in 2011-12 as part of Round 1 and also having administered the PIAAC instrument to an additional sample of unemployed adults, and young (16-34 year-olds) and older adults (66-74 year-olds) as well as prison inmates in 2012-14 to enhance its PIAAC Round 1 sample. Four of the six countries participating in Round 3 – Ecuador, Kazakhstan, Mexico and Peru – are upper middle-income countries. Prior to Round 3, only three middle-income countries had participated in the survey: Indonesia, the Russian Federation and Turkey. The addition of four additional middle-income countries to the survey during the third round highlights its expanding coverage and its increasing relevance for shaping policy in such countries.
In broad terms, the results for the United States reflected very closely those observed in Round 1, as might be expected. Change in the overall proficiency of the adult population primarily results from the replacement of older cohorts exiting the target age range of the study by younger cohorts entering it. As around 2% of the target population is replaced every year, scope for major change over a five-year period is limited.
Hungary is notable for the fact that it has well above-average performance in numeracy but slightly worse than average performance in literacy. Ecuador, Mexico and Peru stand out for their very low average scores in literacy, numeracy and problem solving in technology-rich environments; the high proportions of their populations performing at Level 1 or below on the literacy and numeracy scales; and the large proportions of the population who did not undertake the assessment on computer. In this, they are similar to Turkey and Chile in Round 2.
Besides differences across countries, there was also substantial variation in proficiency observed within countries across different socio-demographic groups. In particular, proficiency is closely associated with age, educational attainment and parents’ level of education, but only weakly associated with gender. With respect to the performance across groups from different backgrounds, Latin American countries in PIAAC tend to have lower performance across the board but they seem to have benefited from the recent expansion in terms of access to education, as the better educated youngest adults show higher proficiency than older adults. Adults in Hungary, on the other hand, tend to score roughly at the same level as the OECD average. Moreover, Hungary stands out as a country with no gender gap in numeracy because of the exceptionally high performance of Hungarian women in that domain.
Data on the frequency of skill use indicate that the use of skills in everyday life and at work are highly, albeit imperfectly, correlated at the country level i.e. countries and economies ranking low in the use of numeracy skills in everyday life also rank low in their use at work, while countries at the upper end of the distribution for use of skills in everyday life also rank high for their use at work. Numeracy proficiency and engagement in numeracy practices are positively but weakly correlated at the country level when high-income countries are considered. The correlation strengthens when non-high income countries are also included, particularly Ecuador, Mexico and Peru.
Adults with greater proficiency in literacy, numeracy and problem solving in technology-rich environments tend to have better outcomes in the labour market than their less-proficient peers. They have greater chances of being employed and, if employed, of earning higher wages. Among Round 3 countries, there is considerable disparity in these labour-market outcomes, however. Returns to proficiency with respect to wages are higher in Hungary on average while they are lower than the OECD average in Ecuador, Mexico and Peru. The relationship is weakest in Ecuador, where it is not statistically significant. In addition to economic outcomes, proficiency in information-processing skills is also positively associated with several aspects of well-being such as trust, volunteering, political efficacy and self-assessed health.
A final comment concerns proficiency in literacy, numeracy and problem solving in Ecuador, Mexico and Peru. On the one hand, the results show a gap between the proficiency of adults in these countries with those of adults in countries such as Japan, the Netherlands and Sweden. The high proportions of working-age adults with very low proficiency in information-processing skills represents a considerable economic and social challenge, particularly in the context of rapid technological change. On the other hand, PIAAC provides examples of countries (e.g. Korea and Singapore) that 50 years ago had working-age populations with very low proficiency, which have successfully increased the proficiency of successive generations to the point that the younger cohorts in these countries are among the highest performers in the study. Achieving sustainable improvement in the information-processing skills of the population is possible, but requires a concerted long-term commitment and effective ongoing investment in education and training.
A note regarding the Russian Federation
The sample for the Russian Federation does not include the population of the Moscow municipal area. More detailed information can be found in the Technical Report of the Survey of Adult Skills, Third Edition (OECD, 2019[12])
[14] Acosta, P., N. Muller and M. Sarzosa (2017), Beyond Qualifications: Returns to Cognitive and Socioemotional Skills in Colombia, Institute of Labor Economics, http://conference.iza.org/conference_files/Statistic_2018/sarzosa_m23031.pdf (accessed on 7 May 2019).
[13] Cunningham, W., P. Acosta and N. Muller (2016), Minds and Behaviors at Work: Boosting Socioemotional Skills for Latin America’s Workforce, World Bank Publications, http://documents.worldbank.org/curated/en/290001468508338670/Minds-and-behaviors-at-work-boosting-socioemotional-skills-for-Latin-America-s-workforce (accessed on 7 May 2019).
[11] Hogan, J. et al. (2016), Main Study and National Supplement Technical Report (NCES 2016-036REV), U.S. Program for the International Assessment of Adult Competencies (PIAAC) 2012/2014, National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. Washington, D.C., https://nces.ed.gov/pubs2016/2016036_rev.pdf.
[9] ITU (2019), ICT Statistics, International Telecommunications Union, https://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx.
[18] Krenzke, T. et al. (2019), U.S. Program for the International Assessment of Adult Competencies (PIAAC) 2012/2014/2017: Main Study, National Supplement, and PIAAC 2017 Technical Report (NCES 2020042), U.S. Department of Education. Washington, D.C.: National Center for Education Statistics.
[3] Maehler, D., S. Bibow and I. Konradt (2018), “PIAAC Bibliography: 2008-2017”, GESIS Papers, No. 2018/03, Leibniz-Institut für Sozialwissenschaften, Köln, https://nbn-resolving.org/urn:nbn:de:0168-ssoar-56014-4.
[15] Ocampo, J. and N. Gómez-Arteaga (2017), “Social protection systems, redistribution and growth in Latin America”, CEPAL Review, No. 122, Economic Commission for Latin America and the Caribbean, https://repositorio.cepal.org/handle/11362/42655.
[12] OECD (2019), Technical Report of the Survey of Adult Skills, Third Edition, OECD, http://www.oecd.org/skills/piaac/publications/PIAAC_Technical_Report_2019.pdf.
[8] OECD (2018), Skills in Ibero-America: Insights from PISA 2015, OECD, http://www.oecd.org/pisa/sitedocument/Skills-in-Ibero-America-Insights-from-PISA-2015.pdf.
[2] OECD (2016), Skills Matter: Further Results from the Survey of Adult Skills, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/9789264258051-en.
[1] OECD (2013), OECD Skills Outlook 2013: First Results from the Survey of Adult Skills, OECD Publishing, Paris, https://doi.org/10.1787/9789264204256-en.
[16] OECD (2013), Technical Report of the Survey of Adult Skills, October, http://www.oecd.org/skills/piaac/_Technical%20Report_17OCT13.pdf.
[7] OECD (2012), “Does money buy strong performance in PISA?”, PISA in Focus, No. 13, OECD Publishing, Paris, https://doi.org/10.1787/5k9fhmfzc4xx-en.
[17] PIAAC (2014), PIAAC Technical Standards and Guidelines, June, http://www.oecd.org/skills/piaac/PIAAC-NPM(2014_06)PIAAC_Technical_Standards_and_Guidelines.pdf.
[10] Rampey, B. et al. (2016), Skills of US Unemployed, Young, and Older Adults in Sharper Focus: Results from the Program for the International Assessment of Adult Competencies (PIAAC) 2012/2014. First Look., National Center for Education Statistics, Washington, D.C.
[4] UNSD (2018), SDG Indicators: Metadata Repository, United Nations Statistics Division, https://unstats.un.org/sdgs/metadata/ (accessed on 5 July 2019).
[19] World Bank (2019), The STEP Skills Measurement Program, World Bank, https://microdata.worldbank.org/index.php/catalog/step/about (accessed on July 2019).
[5] World Bank (2019), World Bank Country and Lending Groups, World Bank, https://datahelpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups.
[6] World Bank (2019), World Bank Open Data, World Bank, https://data.worldbank.org/ (accessed on July 2019).
The United States has conducted three rounds of data collection using PIAAC instruments.
The United States was one of the 24 countries that participated in the Round 1 of PIAAC which collected data in 2011-12. The data collection for the US Round 1 of PIAAC was undertaken as part of the international data collection managed by the OECD and followed the same procedures and standards as the other countries in Round 1. These are described in the study’s Technical Report (OECD, 2019[12]) which also provides details of the United States’ compliance with these standards and the quality of the data collected. Results for the United States were published in the international report of Round 1 (OECD, 2013[1]).
US data for Round 1 of PIAAC have been released as a public use file (PUF) by the OECD. A PUF including US national variables and restricted use file containing data at a more disaggregated level for some key variables are also available from the NCES website.
The PIAAC National Supplement administered the PIAAC instruments to an additional sample of adults in order to enhance the PIAAC Round 1 sample in the United States. The National Supplement included a sample of adults from households not previously selected located in the 80 primary sampling units (PSUs) included in Round 1. The National Supplement household sample increased the sample size of two key subgroups of interest, unemployed adults (aged 16-65) and young adults (aged 16-34), and added a new subgroup of older adults (aged 66-74). The completed sample included 3 660 respondents: 1 064 unemployed adults, 1 545 young adults who were not unemployed and 749 older adults. In addition, there were 247 adults aged 35-65 who were not unemployed included in the final sample due to the initial misclassification of their employment status (Hogan et al., 2016[11]). The same procedures and instruments used during Round 1 collection were employed during the household data collection for the National Supplement.
The PIAAC National Supplement was a national project managed by US National Center for Education Statistics (NCES) and was conducted independently of the OECD. The procedures for data collection and reporting closely followed those of PIAAC Round 1. As it was a national project, the OECD was not involved in monitoring the compliance of the US data collection and subsequent data processing with the PIAAC standards or in the assessment of data quality. The technical details of the implementation of National Supplement are presented in the project’s Technical Report (Hogan et al., 2016[11]).
The data from the US National Supplement have been released in the form of a national U.S. PUF and an OECD PUF for 2011(available on OECD website) combining data from the 2011-12 and 2014 data collections. An 2017 OECD PUF for the U.S. is planned to be released as well. Restricted-use versions of the files are also available to researchers.
It should be noted that the PIAAC 2012/14 data set was weighted to control totals from the 2012 American Community Survey (ACS) (a supplement to the population census) (Hogan et al., 2016[11]). The PIAAC 2012 data was weighted to the 2010 ACS (OECD, 2013[16]). The reweighting has some impact on the estimated proficiency of the population. The 2010 ACS was linked to the 2000 census whereas the 2012 ACS was based on the 2010 census. As it is weighted to more up-to-date control totals (as well as being based on a larger sample), the combined PIAAC 2012/2014 data set for the US provides a more accurate representation of the proficiency of the US population (in the period 2011-14) than the 2012 data set. For this reason, data from the 2012-2014 US data set has been used in this report in place of the 2012 data set used in earlier reports.
The US Round 3 data collection was also conducted as a national project managed by the NCES in conjunction with the Round 3 data collection managed by the OECD. It used the same instruments and followed similar procedures to the other countries participating in Round 3. Data collection was undertaken on a slightly different timetable to that of other participants. In the United States data were collected over March-September, 2017 compared to August 2017-April 2018 in other Round 3 countries. The United States deviated from the PIAAC Technical Standards (PIAAC, 2014[17]) in some areas. A field test was not undertaken. The sample size (a target of 3 800 cases) was less than the minimum sample size required by the PIAAC Standards and Guidelines (5 000 completed cases). Quality control activities were not the same in the United States as in other countries. In addition, the quality of the data for the United States was not reviewed by the PIAAC Technical Advisory Group (TAG) as was the case for the other five countries in Round 3. As in the case of the National Supplement, a full Technical Report has been released (Krenzke et al., 2019[18]). On the basis of the information in the Technical Report, the US data are considered to meet the PIAAC standards for publication.
The Round 3 data for the United States have been released in the form of a PUF and a restricted-use file.
The World Bank’s Skills Towards Employment and Productivity (STEP) Skills Measurement Program (World Bank, 2019[19]) is an initiative to measure literacy skills in low and middle-income countries. It includes a household-based survey and an employer-based survey. The Program is a collaboration between the World Bank and the OECD in which the former used the PIAAC literacy assessment as part of its household-based survey.
The household-based survey uses three modules:
a direct assessment of reading proficiency and related competencies scored on the same scale as the OECD’s PIAAC Survey of Adult Skills
self-reported information on personality, behaviour, and time and risk preferences
self-reported questionnaire on the relevant skills that respondents possess or use in their job.
The employer-based survey has five modules which are designed to assess:
the structure of the labour force
the cognitive skills, behaviour and personality traits, and job-relevant skills that are currently being used, as well as skills employers look for when hiring new workers
the provision of training and compensation by employers
the level of satisfaction in the labour force with the education and skills training available.
The STEP collection (World Bank, 2019[19]) currently hosts data collected between March 2012 and August 2017 in Albania, Armenia, Azerbaijan, Bolivia, Bosnia and Herzegovina, Colombia, Georgia, Ghana, Kenya, Kosovo, Lao People’s Democratic Republic, the Republic of North Macedonia, Serbia, Sri Lanka, Ukraine, Vietnam, and Yunnan Province in the People’s Republic of China. In all countries, the target population is urban adults aged 15 to 64, whether employed or not.
It is important to note that this report, and other previous PIAAC reports, do not present results from STEP because the two surveys assess different target populations. The target population for the Survey of Adult Skills is the non-institutionalised population of 16-65 year-olds residing in the country or region at the time of the data collection, irrespective of nationality, citizenship or language status. The STEP target population is the population aged 15 to 64 inclusive, living in urban areas, as defined by each country’s statistical office. Some STEP surveys had even narrower urban sampling. For example, in Yunnan Province (China) the sample covered only the urban areas of Kunming. Moreover, the Survey of Adult Skills also differs from STEP in terms of the sample size and the implementation standards used.