Skip to main content

Table 4 Meta-regression analysis of literature heterogeneity

From: Gender wage gap in China: a large meta-analysis

Estimator (analytical weight in brackets)

Cluster-robust OLS

Cluster-robust WLS [N]

Cluster-robust WLS [d.f.]

Cluster-robust WLS [1/SE]

Multi-level mixed-effects RM

Cluster-robust random-effects panel GLS

Cluster-robust fixed-effects panel LSDV

Meta-independent variable (default study type)/model

[1]

[2]

[3]

[4]

[5]

[6]a

[7]b

Target region (urban region)—Hypothesis H2

 Rural region

− 0.0602***

(0.020)

− 0.0355

(0.026)

− 0.0402**

(0.020)

− 0.0430**

(0.020)

− 0.0603***

(0.022)

− 0.0602***

(0.023)

− 0.0618**

(0.028)

 Region unspecified

0.0013

(0.014)

− 0.0096

(0.020)

0.0027

(0.016)

0.0019

(0.016)

− 0.0025

(0.016)

− 0.0041

(0.016)

− 0.0432***

(0.016)

Target corporate sector (unspecified)—Hypothesis H3

 Public sector

0.0118

(0.033)

0.0218**

(0.010)

0.0096

(0.026)

0.0093*

(0.005)

0.0235**

(0.008)

0.024**

(0.011)

0.0242**

(0.011)

 Private sector

− 0.0244

(0.028)

− 0.0090

(0.024)

− 0.0251

(0.026)

− 0.0265

(0.027)

0.0012

(0.019)

0.0015

(0.019)

− 0.0015

(0.017)

Estimation period—Hypothesis H4

 Average estimation year

− 0.0031***

(0.001)

− 0.0018*

(0.001)

− 0.0030***

(0.001)

− 0.0031***

(0.001)

− 0.0013*

(0.001)

− 0.0012*

(0.001)

− 0.0010

(0.001)

Hukou types (Hukou unspecified)

 Urban residents

0.0061

(0.019)

− 0.0168

(0.014)

− 0.0067

(0.015)

− 0.0063

(0.016)

0.0119

(0.011)

0.0120

(0.011)

0.0114

(0.014)

 Migrants

− 0.0222*

(0.013)

− 0.0059

(0.028)

− 0.0156

(0.016)

− 0.0160

(0.016)

− 0.0138

(0.013)

− 0.0140

(0.014)

− 0.0153

(0.019)

Wage level percentile (wage level unspecified)

 Low-percentile group

0.0543***

(0.018)

− 0.0071

(0.024)

0.0273

(0.022)

0.0276

(0.023)

0.0299

(0.017)

0.0278

(0.017)

0.0180

(0.020)

 Middle-percentile group

0.0473***

(0.018)

0.0133

(0.020)

0.0355*

(0.020)

0.0367

(0.020)

0.0168

(0.018)

0.0145

(0.018)

0.0040

(0.022)

 High-percentile group

0.0666***

(0.015)

0.0256***

(0.019)

0.0543***

(0.015)

0.0556***

(0.016)

0.0373**

(0.016)

0.0351**

(0.017)

0.0251

(0.021)

Survey data (CHIPs)

 CHNS

− 0.0109

(0.017)

0.0345

(0.021)

0.0013

(0.017)

0.0020

(0.017)

− 0.0030

(0.017)

− 0.0016

(0.017)

− 0.0204

(0.006)

 CGSS

− 0.0093

(0.018)

0.0274

(0.027)

0.0026

(0.021)

0.0029

(0.021)

0.0036

(0.019)

0.0052

(0.020)

0.0356

(0.012)

 Other household survey

0.0151

(0.013)

0.0253

(0.015)

0.0171

(0.013)

0.0169

(0.013)

0.0160

(0.014)

0.0159

(0.014)

 

 Enterprise survey

− 0.0076

(0.027)

− 0.0092

(0.046)

− 0.0126

(0.034)

− 0.0124

(0.034)

0.0205

(0.034)

0.0206

(0.035)

 

Data type (Cross-section data)

 Panel data

0.0249

(0.023)

− 0.0018

(0.038)

0.0117

(0.028)

0.0123

(0.029)

0.0196

(0.019)

0.0207

(0.019)

0.0197

(0.020)

Wage type (Bonus wage)

 Regular wage

− 0.0099

(0.014)

− 0.0007

(0.016)

− 0.0056

(0.014)

− 0.0059

(0.014)

− 0.0091

(0.009)

− 0.0083

(0.009)

0.0013

(0.003)

Wage payment period (annual)

 Monthly

0.0292*

(0.017)

− 0.0047

(0.014)

0.0129

(0.016)

0.0128

(0.016)

0.0079

(0.013)

0.0056

(0.013)

− 0.0123*

(0.007)

 Daily

0.0271

(0.029)

− 0.0449**

(0.022)

− 0.0142

(0.023)

− 0.0130

(0.024)

0.0391*

(0.021)

0.0378*

(0.020)

 

 Hourly

0.0414**

(0.017)

− 0.0007

(0.019)

0.0219

(0.017)

0.0224

(0.017)

0.0223**

(0.011)

0.0217**

(0.010)

0.0178***

(0.004)

Wage variable type (actual value: Yuan)

 Logarithm value

0.0281*

(0.016)

0.0243

(0.015)

0.0274

(0.014)

0.0270*

(0.014)

0.0123

(0.012)

0.0096

(0.011)

− 0.0130

(0.008)

Estimator

 OLS (estimators other than OLS)

0.0135

(0.016)

0.0103

(0.017)

0.0162

(0.016)

0.0163

(0.016)

0.0064

(0.014)

0.0058

(0.014)

0.0026

(0.015)

 IV/2SLS/3SLS

0.0059

(0.016)

− 0.0112

(0.010)

0.0002

(0.015)

0.0005

(0.015)

0.0122

(0.010)

0.0119

(0.010)

0.0105

(0.011)

 Control for selection bias

0.0609**

(0.027)

0.0639**

(0.030)

0.0679**

(0.029)

0.0696**

(0.029)

0.0615***

(0.021)

0.0607***

(0.022)

0.0556**

(0.024)

Control variables

 Occupation

0.0007

(0.011)

− 0.0010

(0.013)

− 0.0024

(0.011)

− 0.0026

(0.011)

0.0102

(0.008)

0.0097

(0.007)

0.0056

(0.007)

 Age

− 0.0123

(0.015)

0.0234**

(0.011)

0.0055

(0.015)

0.0065

(0.015)

0.0115

(0.017)

0.0146

(0.018)

0.0871

(0.071)

 Work experience/tenure

− 0.0024

(0.015)

0.0161

(0.015)

− 0.0005

(0.015)

0.0007

(0.015)

0.0080

(0.014)

0.0082

(0.014)

− 0.0091

(0.009)

 Health status

− 0.0004

(0.016)

− 0.0141

(0.029)

− 0.0117

(0.020)

− 0.0111

(0.020)

− 0.0062

(0.012)

− 0.0050

(0.012)

0.0093

(0.018)

 Firm size

− 0.0310*

(0.017)

0.0031

(0.026)

− 0.0129

(0.020)

− 0.0131

(0.020)

− 0.0317

(0.015)

− 0.0331

(0.015)

− 0.0416

(0.017)

 Trade union

− 0.0134

(0.030)

0.1234**

(0.054)

0.0634

(0.051)

0.0643

(0.051)

0.0023

(0.021)

0.0045

(0.021)

0.0331

(0.001)

 Location fixed effects

− 0.0173*

(0.010)

− 0.0383***

(0.011)

− 0.0228**

(0.011)

− 0.0224**

(0.011)

− 0.0020

(0.011)

− 0.0015

(0.011)

0.0007

(0.014)

 Industry fixed effects

− 0.0003

(0.011)

− 0.0048

(0.012)

0.0004

(0.010)

− 0.0006

(0.010)

0.0068

(0.007)

0.0084

(0.007)

0.0186**

(0.008)

Estimation with an interaction term(s)

 With an interaction term(s) (without interaction term)

0.0115

(0.024)

− 0.0099

(0.031)

− 0.0060

(0.030)

− 0.0046

(0.030)

0.0580**

(0.023)

0.0595***

(0.023)

0.0685****

(0.020)

Standard error of patial correlation coefficient

 S.E.

− 0.1144

(0.451)

− 1.3543**

(0.591)

− 0.3644

(0.467)

− 0.3941

(0.473)

0.0744

(0.427)

0.0957

(0.431)

0.2105

(0.433)

Constant

6.0190***

(1.619)

3.5734*

(2.100)

5.9380***

(1.703)

5.9800***

(1.719)

2.4447*

(1.488)

2.3147

(1.550)

1.8033

(1.819)

K

1472

1472

1472

1472

1472

1472

1472

R2

0.200

0.338

0.194

0.192

–

0.119

0.011

  1. aBreusch-Pagan test: χ2 = 1069.27, p = 0.0000
  2. bHausman test: χ2 = 29.50, p = 0.4913
  3. Figures in parentheses beneath the regression coefficients are robust standard errors. ***, **, and * denote statistical significance at the 1, 5, and 10% levels, respectively
  4. Source: See Table 3 for the definitions and descriptive statistics of meta-independent variables