LAMAS Working Group 10-11 October 2018 Agenda Item 3.1 Gender Pay Gap: decomposition Piotr.Ronkowski@ec.europa.eu

Introduction Reducing the gender pay gap is one of the key priorities of EU gender policy The unadjusted GPG is the leading indicator used to evaluate the progress in reducing the gap The unadjusted GPG is calculated as follows: Mean hourly earnings of men−Mean hourly earnings of women Mean hourly earnings of men

Towards a decomposition methodology The unadjusted GPG does not capture discrimination as such … … but combines: possible differences in pay between men and women, for ‘equal work’, the impact of differences in the average characteristics of men and women in the labour market

Towards a decomposition methodology Microdata from the Structure of Earnings Survey 2014 used … … to measure the impact of differences in the average characteristics of men and women Microdata cover two broad areas: the earnings of individual employees the observed characteristics of individual employees

Towards a decomposition methodology The Blinder-Oaxaca decomposition method applied on the SES dataset … … to isolate the contribution of each observed characteristic to the unadjusted GPG The unadjusted GPG can be separated into: the explained part due to the differences in the average characteristics of male and female employees the unexplained part due to differences in financial returns

Towards a decomposition methodology The methodology and results were discussed with LAMAS in October 2017 and DSS in March 2018 There was broad consensus on the overall methodology proposed by Eurostat It was decided to publish the results as a 'statistical working paper' rather than as European official statistics

Eurostat's methodological publication It was published in June 2018 The publication implements, in principle, the same Oaxaca decomposition method presented at the LAMAS meeting in October 2017 Compared to the presentation made at the LAMAS, the method was slightly modified by using ISCO 2-digit level instead of ISCO 1-digit level for occupation The other changes concerns the way the results of the decompositions are presented and interpreted.

Eurostat's methodological publication The publication focuses more on the explained part of the decomposition and less on the unexplained part The unexplained part of the decomposition is interpreted as a 'residual' and not as a 'better approximation of discrimination The term 'adjusted GPG' is replaced by the term 'unexplained GPG' The explanatory factors (the subcomponents of the explained part) are presented also as earnings gaps

Eurostat's methodological publication The EU aggregates are published The information and results on other segregation effects and self-selection is included The coefficients of determination and the results of F-test for each explanatory variable are presented An indicative list of publications at national level on decomposing the GPG is included

Publication's conclusions The results show that there are clear policy and statistical reasons to decompose the unadjusted GPG into the explained and unexplained parts The unadjusted GPG indicator, together with the explained gap and its explanatory factors, allow for a better identification and interpretation of the causes of the gender pay gap As a consequence, policy actions can be better targeted

Decomposition of the unadjusted GPG

Decomposition of the unadjusted GPG

Publication's conclusions The explained part is strongly driven by three explanatory factors: economic activity occupation education

Decomposition of the explained GPG

Publication's conclusions ‘Self-selection’ of women associated with the negative explained GPG

Employment rate for women and the unadjusted gender pay gap

Possible ways forward Eurostat will present the results of the GPG decomposition to key users inside the Commission Eurostat also considers presenting the results of the decomposition to representatives of media and nongovernmental organisations

LAMAS is invited: To take note of the progress made in GPG decomposition since last LAMAS To take note of Eurostat's plan to present the results of the GPG decomposition to key users inside the Commission To express their views on Eurostat's proposal to present the results of the decomposition to representatives of media and nongovernmental organisations To express their views on other possible ways forward to better educate users of GPG statistics