THE EVOLUTION OF GENDER WAGE GAPS IN THE EU ( )

Transcription

1 THE EVOLUTION OF GENDER WAGE GAPS IN THE EU (2005-) 4th European User Conference in Mannheim, March 5-6, 2015 Joao Pedro Azevedo, The World Bank Judy Yang, The World Bank 1

2 About and Motivation This paper examines relative wages and the gender wage gap in 28 EU countries over the period Gender wage gaps are decomposed into unexplained, common characteristics, and characteristics that were unique to either males or females. The European Commission is active in working to reduce the gender wage gap. Their Strategy for equality between women and men (-2015) promotes equal pay and equality in senior positions. Closing the gender pay gap is also part of the Europe 2020 Strategy to create jobs. 2

3 Hourly Wage ($) Wage Gap (as Percent of Female Wages) Hourly Wages and Wage Gaps in the EU-28 The gender wage gap in the EU (28 selected countries) has been relatively stable over the period 2005 to, between 18 and 19 percent. On average, real hourly wages have been increasing for both men and women, excluding a decline in right after the financial crisis. In, the gender wage gap was the highest, 19.5 percent. $14.00 $12.00 $10.00 $8.00 $6.00 $4.00 $ % 20% 15% 10% 5% Notes: Author s calculations. Values in Euros, PPP values. Source: EU- SILC $ Men Women Wage Gap 0% 3

4 Literature On average, women are more educated than men so it is not the case that women do not have access to education or sufficient skills, but they may face obstacles in the labor market. Using data on 26 European countries, Christofides, Polycarpou, and Vrachimis (2013) find evidence of higher gender wage gaps at the high end of the wage distribution suggesting the presence of a glass ceiling. Arulampalam, Booth, and Bryan () also find evidence of a glass ceiling using data. To explain the gender wage gap, wage differences are decomposed following Ñopo s () nonparametric matching decomposition. Results for EU-28 countries are compared to Atal, Ñopo, and Winder s () study of the gender wage gap in the Latin American and Caribbean (LAC) region, which also uses Ñopo s method. 4

5 Methodology: Ñopo s () nonparametric matching decomposition Incorporates matching techniques and allows for the ability to understand the distribution of unexplained gender wage differentials. Ñopo s procedure offers several advantages compared to the widely used Blinder-Oaxaca decomposition (Blinder, 1973; Oaxaca, 1973). Decomposition is nonparametric and does not assume linearity. Matching on observed variables are done exactly rather than utilizing propensity score matching or nearest neighbor techniques. Decomposes components of the gap that are under the support of common characteristic profiles between men and women, and uncommon supports. D=observed gender wage gap 1. DO= unexplained gender gap 2. DM= differences due to male-specific characteristics 3. DF= differences due to female-specific characteristics 4. DX=differences in the distribution of observable profiles that exist among both females and males 5

6 Data EU-SILC for 28 countries in the European Union (EU-28) that has been harmonized for and extracted from the World Bank micro data portal. Full data from 2005 to is available for 21 countries. Individuals years of age and with positive hourly employee earnings. The largest countries are Germany, France, Italy, and the United Kingdom. The long time series allows for a detailed assessment of gender wage gap trends during the years leading up to and following the financial crisis. 6

7 Findings 7

8 Gender Wage Decomposition near the Financial Crisis (28 countries) Age + Education + Experience + Occupation +Employment Type +Time Worked D 18.29% 18.29% 18.29% 18.29% 18.29% 18.29% D % 23.81% 23.86% 21.26% 21.43% 17.46% DM -0.03% -0.15% -0.26% -2.49% -3.07% -2.14% DF 0.00% 0.03% 0.06% 1.30% 1.79% -0.62% DX -4.03% -5.40% -5.36% -1.79% -1.87% 3.58% % CS Males 99.96% 99.75% 99.52% 82.11% 80.15% 74.99% % CS Females % 99.93% 99.86% 92.20% 90.89% 73.35% Age + Education + Experience + Occupation +Employment Type +Time Worked D 19.51% 19.51% 19.51% 19.51% 19.51% 19.51% D % 22.76% 22.85% 20.49% 20.59% 18.42% DM -0.03% -0.04% -0.11% -2.60% -3.14% -2.57% DF 0.00% 0.03% 0.05% 1.36% 1.90% 0.68% DX -1.69% -3.25% -3.27% 0.26% 0.16% 2.98% percm 99.95% 99.75% 99.57% 82.17% 80.11% 74.78% percf % 99.94% 99.85% 91.58% 90.23% 71.88% 8

9 Hourly Wage ($) Wage Gap (as Percent of Female Wages) Hourly Wage ($) Wage Gap (as Percent of Female Wages) The gender wage gap is larger among the more educated and more experienced. (1) Secondary School $14.00 $12.00 $10.00 $8.00 $6.00 $ % 18% 16% 14% 12% 10% 8% 6% Tertiary $25.00 $20.00 $15.00 $ % 30% 25% 20% 15% 10% This implies that in the EU-28, the gender wage gap is not a problem of poverty or access to opportunities, but one of equity. $2.00 4% 2% $5.00 5% $ % $ % Men Women Wage Gap Men Women Wage Gap 9

10 The gender wage gap is larger among the more educated and more experienced. (2) Secondary School 40% 35% 30% 25% 6.50% 1.00% 20% 0.50% 0.40% 0.80% 5.10% 3.00% 3.60% 2.90% 0.20% 1.40% 2.50% 15% 10% 19.20% 15.30% 16.40% 16.70% 17.40% 17.00% 18.60% 5% Tertiary 40% 35% 30% 18.40% 15.50% 13.90% 16.00% 15.50% 25% 16.40% 16.60% 20% 0.90% 1.00% 2.20% 2.10% 1.90% 0.40% 15% 10% 17.00% 17.80% 18.80% 18.30% 15.80% 16.50% 17.80% 5% This implies that in the EU-28, the gender wage gap is not a problem of poverty or access to opportunities, but one of equity. 0% -5% -6.50% -4.70% -3.90% -3.20% -2.80% -2.50% -2.80% -2.40% -3.20% 0% -5% -3.40% -3.30% -4.60% -4.90% -4.20% -4.40% -4.10% -0.10% -10% 2005 D0 DM DF DX -10% 2005 D0 DM DF DX 10

11 The distribution of observed characteristics (Do) explain a positive portion of the gender wage gap LU UK IE NL GR CY DE BE EE FR FI NO CZ DK SE AT ES SK RO IS LV BG IT SI PT LT HU PL D0 DM DF DX IE GR FI BG UK LU CY NL AT BE FR DE NO IS SE EE CZ DK ES RO IT SK LV SI PL PT LT HU D0 DM DF DX Notes: Sorted by DX, Matched on country age edlev exper occupation1 emptype and timeworked 11

12 In the EU-28, and similarly in LAC, the largest component of the gender wage gap is unexplained, and gaps due to non-overlapping female and male profiles are small. LU UK IE NL GR CY DE BE EE FR FI NO CZ DK SE AT ES SK RO IS LV BG IT SI PT LT HU PL D0 DM DF DX IE GR FI BG UK LU CY NL AT BE FR DE NO IS SE EE CZ DK ES RO IT SK LV SI PL PT LT HU D0 DM DF DX Notes: Sorted by DX, Matched on country age edlev exper occupation1 emptype and timeworked 12

13 Gap Relative to Gender Wage Gap (%) In contrast to LAC, the unexplained gender wage gap is the highest in the EU-28 at the top of the wage distribution In LAC, higher unexplained gaps are seen in the low tail of the wage distribution Wage Percentile

14 D D0 In recent years, the EU-28 unexplained wage gap is higher among the more experienced and older workers. This can be related to skills mismatch and a tighter labor market following the financial crisis. However, across education groups, the levels of unexplained gaps are similar. This can also be related to a glass ceiling and unequal opportunities at higher levels Years Experience Years Experience Years Experience 31+ Years Experience Upper (CI99) Lower (CI99) D years years years years years Upper (CI99) Lower (CI99) D0 14

15 Next Steps Explaining cross-country variation in the gender wage gap. The trend of the gender wage gap in the context of aging. Gender wage gaps increase with age, especially the unexplained component. Women are primary caregivers and as the population ages, will more women stay home to care for family members? Most of the countries in the EU are expected to age, and quickly, what does this mean for gender equity in the future? 15

16 Thank You Joao Pedro Azevedo Judy Yang 16