ABSTRACT
Previous research suggests that as societies empower women educationally, gender differences in numeracy skills will decline. Using direct measures of 56,142 adults’ numeracy skills from the Programme for International Assessment for Adult Competencies (PIAAC), this article studies whether this claim is evidenced across 20 Organisation for Economic Co-operation and Development (OECD) countries. Results show that in many countries, as education has equalised over generations, gender differences in adult numeracy skills have indeed declined. However, women’s advances in education have not always been matched by a reduction in the gender difference in adult numeracy. Contrary to expectations, gender differences in fields of study in further and higher education cannot systematically explain gender differences in adult numeracy. This suggests that to achieve gender equality in numeracy skills, societies must do more than empower women educationally. More research is needed on the educational policies, contextual, and life course factors contributing to gender differences in adult numeracy in post-industrial societies.
Disclosure statement
No potential conflict of interest was reported by the author.
Notes on contributor
Rose Cook is a PhD candidate in the Department of Social Science, UCL Institute of Education, United Kingdom. Her thesis focuses on gender differences in adult numeracy skills. Her research interests include gender differences and inequalities in education and work domains in comparative perspective.
ORCID
Rose Cook http://orcid.org/0000-0002-4741-0336
Notes
1. The remaining participating countries are excluded because relevant individual-level data were not available. Cases were removed before analysis if they were unable to complete the test, but otherwise all respondents from the original PIAAC samples are included. For the main analysis, this resulted in a total sample size of 56,142 (29,812 women and 26,330 men). ‘Belgium refers to the Flemish region of Belgium only, while UK refers to England and Northern Ireland only.
2. For reasons of space, tables include only gender coefficients from the models. Full regression results are available on request.
3. Many studies of gender and cognitive skills suggest that it is very important to analyse gender differences at low and high skill levels as well as on average (e.g., Pargulski & Reynolds, Citation2017).
4. These tests were conducted using the Stata modules “suest” and “khb”, respectively. Where a change in gender coefficient between models was statistically significant across all 10 plausible values at the 5% level or below, it was marked in bold on the tables.