Role of Human Capital Investments in Older Age for Employment and Retirement - Comparative Longitudinal Perspective (SHARE data)

Konrad Turek, Jagiellonian University, Krakow

European and national ageing policies give high priority to increasing retirement age and employability of older people, but still no effective solution exists. Lifelong learning (LLL) is an important tool to address these challenges. It aims to increase employability, alleviate risks and consequences of destandardised working lives, improve empowerment and adjustment to the changing reality. In result LLL should support increase employment in older cohorts and increase the average retirement age. However evidence shows that lifelong learning policies does not necessarily contribute to expected improvements and the net effect of interventions is often low. Unequal access to education, selective approach to training in companies and the effects of accumulation of advantages and disadvantages may also contribute to growth of inequalities throughout the lifecourse.

The primary goal of this presentation is to provide empirical insights into the patterns and consequences of human capital investments in older ages. It is based on comparative (European) and longitudinal data. It also refers to the policies aimed at lifelong learning that are implemented within European ageing and cohesion programmes. The presentation outlines the findings of the Horizon2020 Maria Skłodowska-Curie Individual Fellowship granted project about longitudinal evidence on human capital investments in older age.

DATA. The analysis are based on data from waves no. 4, 5 and 6 (2010/11, 2013, 2015) of international panel Survey of Health, Ageing and Retirement in Europe (SHARE) for population 50+ and include more than 10 European countries.

METHODS. Panel and multilevel regression models that focus on individual lifecourse trajectories.

RESULTS. The presentation provides evidence on the role of training for employment situation of people aged 50+, their wage trajectories, chances of re-employment and retirement transitions patterns. The results are discussed with reference to public policies.


Extended abstract


Please note thisis a work in progress. The following extended abstract presents preliminaryanalysis and conclusions that will be further explored, improved and extended forthe conference presentation. The research is undertaken within a project LEEP (aboutlongitudinal evidence on human capital investments in older age) granted from theHorizon2020 Maria Sk³odowska-Curie Individual Fellowship 2017-2019.


Data and methods

·        SHARE waves4, 5 and 6; only individuals who were interviewed in all three waves (29,472 individualsfrom 12 countries; 88,416 occasions-observations)

·        Questionabout training in SHARE: training in the previous 12 months

·        Methods:longitudinal and multilevel regressions models


Increasing training wage gap in older age

The analysis revealed increasingtraining wage gap among people 50+ who were employed in all waves. People who participatedin training in all three waves (“Always” group) had significantly increasingtrend in hourly wage, while those who did not report training (“Never” group)remained more less at the same level of wages-per-hour. The gap increases fromabout 10% at the age of 50 up to 20-30% at the age of 65. Interesingly, peoplewho only “occasionally” participated in training (once or twice in 3 waves )also had an increased wage trend. This might give hope that LLL policies may beuseful in decreasing inequalities in older age.


Country differences in the effect of trainingfor wages

The effect of training forwages-per-hour can be also analysed as a marginal effect presenting difference between“Not training” vs  “Training” groups. The effect was different across countries,although mostly positive and significant. High values were observed e.g. inSpain, Slovenia, Estonia, Czech, while in Sweden the effect was notsignificant.

Following graph presents interestingnegative relation between size of the effect and the average participation ratein training – in countries where training attendance was low, the effect oftraining for wages was higher.

Figure 1. Marginal effect of trainingand participation rate in training

Note: Random-effect panel regression; Adjustedby: training pattern, age, age2, gender, education, wave, country;                 Interactions:training#country, training#age; Only individuals working in all waves, cleaneddata; n= 9, 439 (3,966).


Proability of re-employment

The data confirm also that participationin training increases chances for re-employment.

For this analysis only individuals whowere not employed during Wave 5 were chosen to see what are their chances ofre-emplyment in Wave 6 based on their training activity reported in Waves 4 and5.

Any participation in training activitysignificantly increased chances for re-employment both for women and men. Theeffect reaches even about 20 p.p.

The following graph presents the marginaleffects of “always” and “occasional” training groups that are compared to the “never”group (centred at 0).


Figure 1. Marginal effect oftraining proability of re-employment


Note: Logit model; Adjusted by: trainingpattern, age, age2, gender, education, wave, country; Interactions: age#female; n= 20,953.  


Presented in Session 1143: Economics, Human Capital, and Labour Markets