Table 6

Propensity score matching (PSM) analysis

ATT (T-Stat.)Treatment firms (mean)Control firms (mean)Treated: control (No. Of obs.)
Panel A average treatment effect on the treated (ATT)
Dependent variable = ETR1
 1.33%**(2.10)18.43%19.76%1,632: 1,643
Dependent variable = ETR1
(1)(2)(3)(4)
Panel B regressions based on the matched sample
Financial_D−0.015**   
(−2.215)   
Financial_Ratio −0.067***  
 (−2.798)  
RegulatoryFinancial  −0.009 
  (−0.399) 
NonregulatoryFinancial  −0.020*** 
  (−2.914) 
RegulatoryFinancial_Ratio   0.027
   (0.218)
NonregulatoryFinancial_Ratio   −0.071***
   (−2.603)
SOE0.0150.0160.0150.016
(1.433)(1.451)(1.407)(1.480)
Size0.009***0.009**0.009***0.009**
(2.709)(2.520)(2.720)(2.540)
Age−0.008*−0.007*−0.008*−0.008*
(−1.900)(−1.699)(−1.839)(−1.797)
Tangibility−0.004−0.004−0.006−0.005
(−0.087)(−0.102)(−0.143)(−0.115)
Lev0.073***0.073***0.073***0.073***
(3.045)(3.055)(3.051)(3.049)
RD−0.763***−0.797***−0.760***−0.782***
(−3.573)(−3.702)(−3.559)(−3.640)
ROA0.255***0.258***0.255***0.258***
(3.930)(3.975)(3.928)(3.980)
Growth0.016***0.016***0.017***0.016***
(2.703)(2.685)(2.744)(2.670)
Duality0.0030.0030.0020.004
(0.349)(0.431)(0.303)(0.455)
Independence0.0020.0010.0040.001
(0.026)(0.023)(0.063)(0.024)
Female0.069*0.071*0.068*0.069*
(1.890)(1.930)(1.862)(1.893)
Loss−0.008−0.008−0.009−0.008
(−0.402)(−0.397)(−0.418)(−0.390)
TotalAccruals0.0080.0070.0060.006
(0.160)(0.158)(0.136)(0.134)
TMT_Age−0.057−0.059−0.053−0.057
(−1.176)(−1.206)(−1.091)(−1.176)
TMT_Foreign0.0530.0560.0530.058
(1.339)(1.432)(1.339)(1.454)
TMT_Academic−0.008−0.008−0.010−0.009
(−0.345)(−0.314)(−0.427)(−0.371)
_cons0.0150.0330.0010.026
(0.076)(0.166)(0.005)(0.130)
Year fixed effects?YesYesYesYes
Industry fixed effects?YesYesYesYes
Province fixed effects?YesYesYesYes
SEs. Clustered atFirm, YearFirm, YearFirm, YearFirm, Year
Adj. R-square0.0920.0930.0930.092

Note(s): This table reports the results of a PSM routine for treatment firms and non-treatment firms during the sample period. Panel A reports the ATT, which captures the systematic difference in effective tax rates between the treatment firms and the control firms. Panel B presents the regression results for both H1 and H2, based on the re-estimation of Eq. (1) using the matched sample. All independent variables are lagged by one year except for Growth and Loss. Detailed variable definitions can be found in Table A1 of the Appendix section. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively

Source(s): Authors’ own work

or Create an Account

Close Modal
Close Modal