Understanding low female labour force participation: Policy evaluation using microsimulation
Ross Richardson, Lia Pacelli, Matteo Richiardi
Institute for New Economic Thinking at the Oxford Martin School, University of Oxford, United Kingdom; Universita’ di Torino, and LABORatorio Revelli, Torino, Italy; Institute for Social and Economic Research, University of Essex, United Kingdom
Cite as: R. Richardson, L. Pacelli, M. Richiardi; 2019; Understanding low female labour force participation: Policy evaluation using microsimulation; International Journal of Microsimulation; 12(2); 52–68. doi: 10.34196/ijm.00201
Abstract
We project medium to long term trends in labour force participation and employment for selected low-participation EU countries (Italy, Spain, Ireland, Hungary and Greece), with Sweden as a benchmark, by means of a dynamic microsimulation model. By exploring alternative scenarios to our baseline forecasts and their implications for GDP growth, we find that the gap between male and female participation is mostly explained by (i) differences in the individual conditional behaviour of older female cohorts, (ii) inadequate family policies. In particular, our results show that the conditional behaviour of younger women in the low-participation countries is similar to that in Sweden. In a nutshell, what is needed is not a change of behaviour on the side of women but a change of mentality on the side of institutions and firms.
This article contains 8 figures and 7 tables.
Figures
Figure 1. Model structureFigure 2. Participation rates in the baseline drivers scenarios, female population aged 20–64 years. The shaded areas are the estimated densities of the baseline projections, computed over 1,000 replications with bootstrapped values of the coefficients.Figure 3. Participation rates in the baseline drivers scenarios, female population aged 20–44 years. The shaded areas are the estimated densities of the baseline projections, computed over 1,000 replications with bootstrapped values of the coefficients.Figure 4. Participation rates in the policy actions scenarios, female population aged 20–64 years. The shaded areas are the estimated densities of the baseline projections, computed over 1,000 replications with bootstrapped values of the coefficients.Figure 5.Enhanced family policies scenario: public childcare benefits, duration of maternity leave and availability of part-time employment are jointly increased by 20%. Participation rates by educational attainments, female population aged 20–44 years not in education. The shaded areas are the estimated densities of the baseline projections, computed over 1,000 replications with bootstrapped values of the coefficients.Figure 6. Decomposition of the Enhanced family policies scenario: public childcare benefits, duration of maternity leave and availability of part-time employment are separately increased by 20%. Participation rates, female population aged 20–44 years with children aged 0–12 years and low education. The shaded areas are the estimated densities of the baseline projections, computed over 1,000 replications with bootstrapped values of the coefficients.Figure 7.No early retirement scenario: minimum effective retirement age is 60 years old. Employment rates, differential to baseline, population aged 50–59 years.Figure 8.Delayed recovery scenario: the effects of the Great Recession fade away at a slower pace; complete recovery is achieved 10 years later than in the baseline (so 2040 in Greece and 2030 in other countries). Employment rates, differential to baseline, population aged 20–64.
Tables
Table 1. Determinants of the estimated processes at an individual level. Determinants annotated with (t-1) symbolise that the one-period (one year) lag is used
