Data from the PIAAC surveys provide rich information about adult skills in literacy, numeracy and problem solving, and how adults use their skills at home and at work. PIAAC also includes different background variables such as gender, employment status and educational attainment (see OECD for a description of the PIAAC data). Cycle 2 also includes data on personality traits.
Data from Denmark, Norway and Sweden can be combined with data from administrative registers and other data sources, providing unique possibilities for researchers. For instance, combining PIAAC data with other data sources makes it possible to do longitudinal follow-up studies and to analyse how skills affect labour market outcomes and other outcomes several years after the PIAAC study was conducted.
This article shows some examples of research and analyses combining PIAAC cycle 1 data with other data, investigating young adults’ connection to the labour market, low-skilled workers’ employment and changes in skill levels over time.
Data from PIAAC cycle 2, which will be available in the beginning of 2025, can also be combined with data from administrative registers and other data sources. Learn more about access to PIAAC data here.
Are skills related to young adults’ connection to the labour market?
There is a lot of attention on the share of young adults not being in employment or in education and training (NEET). Young people being NEET may risk exclusion and being excluded from the labour market and from education as a young adult may also have consequences for future life opportunities.
A Norwegian study examined the relationship between skills and NEET-status in 2013, two years after the cycle 1 data collection. The study includes adult skills as measured in PIAAC and the grade point average, a measure of skills obtained at the end of compulsory school. Adult skills are strongly highly correlated with skills acquired early in life. Although PIAAC skills are measured closer in time to NEET-status, the study finds that early skills as measured by the grade point average protect more against exclusion from employment and education in 2013.
Data from PIAAC cycle 1 are combined with register information about participation in education and employment at different points of measurement, 2011 and 2013. As a measure of skills obtained early in life, the study uses the grade point average from compulsory school. To examine the relationship between skills and NEET-status, the study uses linear probability models.
For more information, see Barth et al. 2019.
Do better skill levels and use of skills make a difference for low skilled workers’ employment?
The employment rate among low skilled workers has dropped in Denmark over the past years. To create a better knowledge base for initiatives that can improve the employment situation of low skilled workers, a Danish study has focused on unemployment workers with low skills through the following research question: Do basic skills and use of these skills predict this group’s employment situation?
The analysis suggests that this is actually the case: Despite including more than 500 explanatory variables, the study finds that use of IT skills outside work is among the most predictive factors for the number of hours worked during the period 2012-2019. That is, the more these workers used IT skills outside work in 2011/12, the higher their likelihood of working many hours during the period 2012-2019. This is illustrated in figure 1 that shows the most predictive factors. Important factors related to use of IT skills are ‘participate in real-time internet discussions at least once a month’, ‘use of IT skills in everyday life’ (discrete variable) and ‘conduct transaction on the internet at least once a month’. Other important factors are e.g. hourly wages in previous job, which affects this likelihood positively, while being 55-59 years when participating in PIAAC affects it negatively.
The analysis is based on Danish PIAAC data from cycle 1 combined with register information. The dependent variable is number of working hours based on register information for the period 2012-2019. The control variables consist of both survey information from PIAAC Cycle 1 and register information from the period 2008-2011 about gender, age, country of origin, family, cohabitants, place of residence, level and length of education, experience, labour market affiliation, hourly wages, occupation, industry and public benefits.
The starting point for the analysis is a sample of low skilled, unemployed workers from PIAAC cycle 1 consisting of only 134 observations. Using an econometric approach called LASSO (Least Absolute Shrinkage and Selection Operator), which is a machine learning approach, makes it possible to select the most predictive factors in a model with a small sample and a huge amount of explanatory variables.
For more information about this analysis, see Rotger, Larsen & Jeppesen (2022).
Do skills levels change?
As illustrated in the previously described studies, basic skills are often a precondition for or facilitate participation in education and working life. But how has the level of basic skills of the adult population developed since 2011/12, when PIAAC cycle 1 was carried out? This question is the focus of a Danish study using data from PIAAC cycle 1 and a new Danish ‘Basic Skills Survey’ (BSS) carried out in 2020/21.
The comparison of the results from PIAAC and BSS shows significant improvements in adults’ basic skills in literacy and numeracy in Denmark from 2011/12 to 2020/21. Literacy and numeracy proficiency are measured on a scale from 0-500. Analyses suggest that the average score in literacy increased by 12 points, while the average score in numeracy increased by 14 points (see figure 2). About 40% of the increase in average scores is related to an increase in the educational level in the population. However, the increase might be overestimated due to a low response rate in BSS among other things.
The skill level increased for almost all subgroups in terms of gender, age, level of education, mother tongue, health status and labour market status. The exceptions are individuals with compulsory school as the highest level of education, individuals with poor health and self-employed workers. Thus, the results indicate increasing inequality in basic skills from 2011/12 to 2020/21 e.g. in terms of level of education.
The BSS data is obtained from a small Danish study carried out the responsibility of the Danish Center for Social Research Science (VIVE). VIVE received permission from the OECD to use the computer-based confidential literacy and numeracy items from PIAAC Cycle 1 in BSS. In addition to skill assessments, BSS also includes a small background questionnaire. For more information about this analysis, see Larsen, Jakobsen & Rosdahl (2022).
References
Barth, E., Keute, A. L., Schøne, P., von Simson, K., & Steffensen, K. (2021). NEET status and early versus later skills among young adults: Evidence from linked register-PIAAC data. Scandinavian Journal of Educational Research, 65(1), 140-152.
Larsen, M., Jakobsen, V. & Rosdahl, A. (2022): Voksnes basale færdigheder: Udvikling i læse- og regnefærdigheder de seneste 10 år. VIVE – Det Nationale Forsknings- og Analysecenter for Velfærd. https://www.vive.dk/da/udgivelser/voksnes-basale-faerdigheder-17906/
Rotger, G.P., Jeppesen, T. & Larsen, M. (2022): Basale færdigheders betydning for beskæftigelse. Evidens for personer med grundskole, gymnasial eller erhvervsfaglig uddannelse. Det Nationale Forsknings- og Analysecenter for Velfærd. https://www.vive.dk/da/udgivelser/basale-faerdigheders-betydning-for-beskaeftigelse-17357
