Skip to main content

Table 2 Oaxaca–Blinder aggregate decomposition in 1984 and 2020 at the mean and selected percentiles

From: Changes in the gender pay gap over time: the case of West Germany

 

Mean

10th Percentile

50th Percentile

90th Percentile

 

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

 

1984

2020

1984

2020

1984

2020

1984

2020

GPG

0.327*** (0.020)

0.166*** (0.018)

0.533*** (0.037)

0.191*** (0.040)

0.305*** (0.016)

0.145*** (0.020)

0.175*** (0.024)

0.239*** (0.030)

Explained: Q

0.166*** (0.031)

0.093*** (0.020)

0.148*** (0.035)

0.162*** (0.044)

0.157*** (0.019)

0.049** (0.022)

0.104*** (0.031)

0.128*** (0.046)

in % of GPG

50.76

56.02

27.77

84.82

51.48

33.79

59.43

53.56

Unexplained: P

0.161*** (0.036)

0.073*** (0.024)

0.385*** (0.052)

0.029 (0.056)

0.148*** (0.025)

0.096*** (0.027)

0.071** (0.036)

0.110** (0.048)

in % of GPG

49.24

43.98

72.23

15.18

48.52

66.21

40.57

46.44

Observations

4064

4138

4064

4138

4064

4138

4064

4138

Unexplained: P

0.161*** (0.036)

0.073*** (0.024)

0.385*** (0.052)

0.029 (0.056)

0.148*** (0.025)

0.096*** (0.027)

0.071** (0.036)

0.110** (0.048)

in % of GPG

49.24

43.98

72.23

15.18

48.52

66.21

40.57

46.44

Observations

4064

4138

4064

4138

4064

4138

4064

4138

  1. Table shows aggregate Oaxaca–Blinder decompositions of the Gender Pay Gap (GPG) in West Germany. Men are used as reference category. The regressions include quadratic polynomials of full-time experience, job tenure, past part-time experience (controls in years), as well as dummies for highest educational attainment, age, marital status, migration background, having at least one child, having at least one child below the age of two and between 2 and 6 years, respectively. Further, the following dummies are included: occupations (ISCO-88, 1-digit) and industries (NACE, level 1), small (\(<200\) employees) and medium (\(200-1999\) employees) firm size and public-sector firm. Federal state fixed effects included. Robust (for the mean) and bootstrapped (for the estimates beyond the mean, 500 replications) standard errors in parentheses. \(*** p<0.01\), \(** p<0.05\), \(* p<0.1\). SOEP sample weights used. Source: SOEP v37