Skip to main content
Frontiers of Business Research in China

Table 7 Heckman treatment effect regressions for firm performance and chair identity

From: Non-family chair and corporate performance

 

(1)

(2)

(3)

First

Second

First

Second

First

Second

NFChair

 

− 0.054***

 

−0.066***

 

−0.062***

  

(− 3.11)

 

(− 3.56)

 

(−3.78)

Lambda

 

0.024**

 

0.031***

 

0.028***

  

(2.39)

 

(2.86)

 

(3.01)

Size

−0.025

0.015***

− 0.004

0.016***

−0.013

0.016***

 

(−1.02)

(10.24)

(−0.15)

(10.16)

(−0.53)

(10.19)

Leverage

0.319***

−0.033***

0.288***

−0.032***

0.294***

−0.032***

 

(5.89)

(−3.86)

(5.30)

(−3.72)

(5.40)

(− 3.79)

MB

0.036***

−0.000

0.034***

−0.000

0.033***

−0.000

 

(6.38)

(−0.52)

(5.86)

(−0.40)

(5.74)

(−0.49)

Board

−0.357***

−0.004

− 0.398***

−0.008

− 0.378***

−0.007

 

(−3.70)

(−0.67)

(−3.90)

(−1.09)

(− 3.69)

(− 1.02)

Independence

−1.598***

− 0.032

− 1.975***

− 0.048*

−1.896***

− 0.046*

 

(−4.59)

(− 1.38)

(−5.13)

(− 1.87)

(−4.91)

(− 1.77)

Duality

− 0.269***

− 0.006**

− 0.244***

−0.006***

− 0.245***

−0.006**

 

(−6.91)

(−2.46)

(−6.14)

(−2.61)

(−6.15)

(−2.54)

Top1

−0.008***

0.001***

−0.008***

0.000***

−0.008***

0.001***

 

(− 6.17)

(5.83)

(−5.74)

(5.06)

(−5.79)

(5.51)

Top25

−0.016***

0.000***

−0.015***

0.000**

−0.015***

0.000***

 

(−9.20)

(3.14)

(−8.63)

(2.57)

(−8.36)

(2.92)

Institution

0.010***

0.002***

0.009***

0.002***

0.009***

0.002***

 

(3.29)

(8.62)

(3.13)

(8.58)

(3.08)

(8.65)

Foreign

−0.068

0.001

0.045

0.001

0.040

0.001

 

(−0.73)

(0.18)

(0.46)

(0.13)

(0.41)

(0.13)

Tangibility

0.500***

0.001

0.485***

0.001

0.454***

0.001

 

(3.90)

(0.10)

(3.68)

(0.14)

(3.43)

(0.07)

Firm age

0.035***

0.000

0.036***

0.000

0.035***

0.000

 

(8.85)

(0.64)

(8.95)

(1.07)

(8.76)

(0.95)

CapEx

−2.191***

−0.013

−2.344***

−0.030

−2.357***

−0.027

 

(−7.14)

(−0.71)

(−7.36)

(−1.51)

(−7.36)

(−1.42)

Instrumental variables

Divorce rate

0.021***

   

0.017***

 
 

(8.62)

   

(6.80)

 

Meal income

  

−0.116***

 

−0.097***

 
   

(− 7.68)

 

(−6.28)

 

Intercept

1.286*

−0.260***

1.069*

−0.318***

0.692

−0.323***

 

(1.95)

(−6.81)

(1.78)

(−7.99)

(1.14)

(−8.31)

Industry fixed effects

Yes

Yes

Yes

Yes

Yes

Yes

Year fixed effects

Yes

Yes

Yes

Yes

Yes

Yes

Obs.

7902

7902

7584

7584

7584

7584

Pseudo R2 / Adjusted R2

0.159

0.170

0.160

0.173

0.166

0.173

  1. This table presents the results of treatment effect regressions that control for potential endogeneity of reversal causality. Using Heckman’s two-step procedure, the non-family chair dummy is instrumented by the regional divorce rate (Divorce rate) and the logarithm of total meal income of local catering firms (Meal income). In the first-stage Probit model with the non-family-chair dummy as the dependent variable, instruments are included as independent variables in addition to those controls in the main regression. We then exclude instruments but add the inverse Mills ratio (Lambda) derived from the first stage into the second-stage OLS regression model. Columns 1, 3 and 5 report the results of the first-stage Probit regressions. Results from the second stage of the endogenous treatment effect model are shown in columns 2, 4 and 6. All regressions include year and industry dummies, but their coefficients are not reported for the sake of brevity. See the Appendix for detailed descriptions of the variables. The values in parentheses are standard errors that are robust to heteroskedasticity and clustered at the firm level. Asterisks denote statistical significance at the 1% (***), 5% (**) or 10% (*) level, respectively
Back to article page