CEILINGS AND FLOORS: CEILINGS AND FLOORS: GENDER WAGE GAPS BY GENDER WAGE GAPS BY EDUCATION IN SPAIN EDUCATION IN SPAIN Sara de la Rica Sara de la Rica * , Juan J. , Juan J. Dolado Dolado * * * * & Vanesa Llorens & Vanesa Llorens ** * ** * ( * ) UPV & IZA ( ** ) UCIII & CEPR & IZA ( *** ) LECG
23
Embed
CEILINGS AND FLOORS: GENDER WAGE GAPS BY EDUCATION IN SPAIN Sara de la Rica *, Juan J. Dolado * * & Vanesa Llorens ** * & Vanesa Llorens ** * ( * ) UPV.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
CEILINGS AND FLOORS: CEILINGS AND FLOORS: GENDER WAGE GAPS BY GENDER WAGE GAPS BY
EDUCATION IN SPAINEDUCATION IN SPAIN
Sara de la Rica Sara de la Rica **, Juan J. Dolado, Juan J. Dolado* ** *
& Vanesa Llorens& Vanesa Llorens** *** *
(*) UPV & IZA(**) UCIII & CEPR & IZA(***) LECG
Motivation
• Gender wage gaps: ln(Wm/Wf)(Wm- Wf)/ Wf
• Traditional: At the mean vs. New: At the quantiles
• Recent evidence about Glass Ceilings in Sweden (Albrecht et al., 2003)
Figure 2a. Gender Wage Gap by Education Denmark 1999
-0.10
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
10 20 30 40 50 60 70 80 90
Quantiles
Ln W
age
Men
- L
n W
age
Wom
en
Wage gap low educated w orkers Wage gap highly educated w orkers
Figure 2b. Gender Wage Gap by Education United Kingdom 1999
0
0.05
0.1
0.15
0.2
0.25
10 20 30 40 50 60 70 80 90
Quantiles
Ln W
age
Men
- L
n W
age
Wom
en
Wage gap low educated w orkers Wage gap highly educated w orkers
Figure 2d. Gender Wage Gap by Education Italy 1999
0
0.05
0.1
0.15
0.2
10 20 30 40 50 60 70 80 90
Quantiles
Ln W
age
Men
- L
n W
age
Wom
en
Wage gap low educated w orkers Wage gap highly educated w orkers
Figure 2c. Gender Wage Gap by Education Greece 1999
0
0.05
0.1
0.15
0.2
0.25
10 20 30 40 50 60 70 80 90
Quantiles
Ln W
age
Men
- L
n W
age
Wom
en
Wage gap low educated w orkers Wage gap highly educated w orkers
• Composition effect by EducationGlass Ceiling (H-Group): High female
participation rate (80% vs 85%) Lower job stability (Lazear and Rosen, 1990) leads to lower promotion opportunities and higher wages (PUZZLE)
Glass Floor (L-Group): Low female participation rate (48% vs 68%) Statistical discrimination at the bottom of the wage distribution
INTERPRETATIVE MODELS• L-Group
Ability for men and women: , c.d.f. G()Need of training in period 1 (2 periods)Productivity: 1, 0< 1 <1 (period 1), 2 , 1 < 1< 2 (period
2)Firms know at the begining of period 2Workers receive a disutility shock with c.d.f. F() after
wages in period 1 & 2, Wi (i=1,2) are chosen by the firm. Workers do not quit if Wi - 0.
No wage renegotiation nor outside wage offers (monopsony)Fm()>Ff()
G()= U[0,]; fm() =U[0,m]; ff() =U[0,f]; f> m
mfiWWW
dFWW
i
i
i
i
i
i
W
W
iii
i
W i
i
i
, ,max)()(max2
2221
0
221
2
2
2
*2
*2
2*2 ),,(
22 fmi mfiWPeriod
1 Period
)(]()([)()( 1
0
1
0
1
0
*2
1
dGdFWdGW ii
W
iiW
244
2221*
1
024
222*
1*
1 mffm
fm WW
• H-Group (Lazear and Rosen, 1990)A model of job ladders: A (no training), B (training)A: , ; B: 1 , 2 Firms pay competitive wages in period 2: WA
2= , WB
2= 2 Cut-off points to allocate to B: *
f> *mLess women
are promoted but conditional on being promoted they should be earn higher wages
Explanations:(i) Different ability distribution (Mincer and Polacheck, 1974), (ii) Different outside offers (Booth et.al., 2003), (iii) Different competing skills (Gneezy et al., 2003, Babcock and Laschever, 2003)
iiiiii xxwQuxw '
'
• Data–ECHP (1999)
–H-Group: 721 (Men), 558 (Women)
–L-Group: 1585 (Men), 626 (Women)
• Quantile Regressions (QR) Buchinsky (1998), Koenker and Basset (1978)
Covariates-Exp (age), marital st., tenure, children age,Sec. Edn (L-W) , type of contract, immigrant, public, firm size, supervisory role, region, size local council, occupations.
Note: Standard deviations (s.d.) in parenthesis. The s.d. have been obtained through 250 replications of the decomposition; (a) with selection bias correction.
Figure 3a. Gender gap (Observed and Counterfactual). H-group. Spain 1999
-5
0
5
10
15
20
25
30
10 25 50 75 90
Quantiles
Ln W
age
Men
- L
n W
age
Wom
en
Observed Counterfactual
Figure 3b. Gender gap (Observed and Counterfactual). L-group. Spain. 1999
0
5
10
15
20
25
30
35
10 25 50 75 90
Quantiles
Ln W
age
Men
- L
n W
age
Wom
en
Observed Counterfactual
PANEL & STAT. DCN.
• ECHP waves (1994-01) to follow workers in their jobs over time.
• Follow approach in Farber & Gibbons (1996)
• Interact Tenure* Female
• RESULT: Only Positive & Significant for L-group.
Table 4a: Descriptive statistics of (log) Real Wages and Tenure (in years); ECHP (1994-01) Workers younger than 40
at first interview Workers older than 39 at first
interview Men Women Men Women L H L H L H L H
Mean Log Wage
1.71 (0.37)
2.10 (0.46)
1.54 (0.34)
2.05 (0.46)
1.81 (0.38)
2.54 (0.49)
1.62 (0.43)
2.34 (0.37)
Mean Tenure 4.02 (4.72)
5.41 (5.00)
3.68 (4.35)
5.14 (4.83)
5.57 (5.67)
9.05 (5.95)
6.24 (5.40)
9.21 (5.97)
N. obs. 8617 3323 3381 2871 2394 498 930 249
Note: All workers with more than 15 years of tenure at the same firm are excluded from the sample since the variable Tenure is truncated at 15 for them.
Table 4b: (Log) Real Wage Regressions – Workers younger than 40 years at first interview Fixed Effects Estimation
Table 4c: (Log) Real Wage Regressions – Workers older than 40 years at first interviewed Fixed Effects Estimation
Men Women Pooled Men and Women
All Low High All Low High Low High Tenure 0.011
(0.004) 0.010
(0.004)
0.011 (0.012)
-0.003 (0.006)
0.003 (0.007)
-0.006 (0.014)
0.010 (0.004)
0.011 (0.011)
Tenure2 -0.0005 (0.0002)
-0.0004 (0.0002)
0.0001 (0.0003)
-0.0001 (0.0004)
0.0007 (0.0005)
-0.0004 (0.0002)
-0.001 (0.0005)
Female*Tenure --- --- --- --- --- -0.006 (0.008)
-0.018 (0.021)
Female* Tenure2
--- --- --- --- --- 0.0002 (0.0005)
0.002 (0.0009)
N.obs 2892 2394 498 1179 930 249 3324 747
Notes: All regressions include also 6 dummies for region, 14 dummies for occupation, 2 dummies for industry and a dummy for work status (supervisor or not). Workers with more than 15 years of tenure are not included since the variable Tenure is truncated at 15 for them. In the last panel, when pooled men and women are taken together, all explanatory variables are interacted with Female.
CONCLUSIONS
• New finding: Glass Floors• Due to statistical dcn. in countries with low
participation of L-women.• Further research: - Other alternatives for H-group (stress leaves) - Endogenize Participation (with S. de la Rica and C. Gª-Peñalosa…in progess) - Academic women-economists (with M. Almunia