Table 7

Endogeneity test using Instrumental Variable approach (2SLS)

DV= BGENDIVDV = WOMMCDUMDV = WOMMC
(1)(2)(3)(4)(5)
LAGBGENDIV0.893***    
(27.16)    
BGENDIV_IND0.612***    
(3.25)    
BGENDIV_PRED 0.832*** 0.176*** 
 (3.22) (2.87) 
≥3WOMEN_PRED  0.329*** 0.069***
  (3.34) (3.00)
LNBSIZE−0.0150.009−0.165*0.004−0.033
(−1.31)(0.11)(−1.75)(0.21)(−1.63)
BINDP0.0190.2680.328−0.105−0.092
(0.28)(0.74)(0.89)(−1.49)(−1.31)
FSIZE0.001−0.010−0.0050.0040.005
(0.09)(−0.20)(−0.10)(0.42)(0.53)
ROA−0.010−0.0990.010−0.023−0.000
(−0.24)(−0.29)(0.03)(−0.43)(−0.00)
LEV0.0000.017*0.016*0.007**0.006**
(0.20)(1.76)(1.71)(2.50)(2.57)
LFAGE−0.003−0.186***−0.210***−0.029**−0.034**
(−0.44)(−3.25)(−3.70)(−2.19)(−2.56)
Intercept−0.0520.5190.973−0.0200.075
(−0.55)(0.81)(1.53)(−0.16)(0.61)
Year FEYesYesYesYesYes
Industry FEYesYesYesYesYes
Observations295296296296296
Adjusted/Pseudo R20.8360.2130.2460.1460.196
Underidentification test    
Kleibergen−Paap rk LM statistic109.8577.73109.8577.73
p-value0.0000.0000.0000.000
Weak identification test    
Cragg-Donald F statistic450.2286.42450.2286.42

Note(s): This table presents the results of 2SLS regressions. Column 1 shows the first-stage regression results. Columns 2–5 show the second-stage results where the independent variable is predicted based on first-stage regression. The t-statistics reported in parentheses are based on robust standard errors. Superscripts ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively. All variables are defined in  Appendix

Source(s): Authors’ own work

or Create an Account

Close subscription notice
Close access options