Page 71 - 臺大管理論叢第33卷第1期

NTU Management Review Vol. 33 No. 1 Apr. 2023

Table 3 Parameter Estimates of Heckman Probit Model in the Second-Stage
Analysis (Reference Group: Strong IPR Countries)
Model 5 Model 6
Model 1 Model 2 Model 3 Model 4 (Ordinary (2SLS
regression) regression)
Human Capital
3.06*** 3.95*** 3.58*** 2.43 2.93***
Low-cost talent
(0.79) (0.88) (1.11) (1.53) (0.90)
2.57**
Task specificity
(1.27)
Low-cost talent 2.59**
X Task specificity (1.15)
0.98**
Project modularity
(0.48)
Low-cost talent 2.57*
X Project modularity (1.50)
-0.87** -0.72** -0.80** -0.66*** -0.76** -0.78**
Turnover
(0.33) (0.30) (0.34) (0.23) (0.32) (0.33)
0.31 0.22 0.08 -0.08 0.23 0.20
Geographic proximity
(0.20) (0.47) (0.39) (0.37) (0.49) (0.50)
0.20 0.24 0.43 -0.02 0.25 0.21
Co-location
(0.29) (0.40) (0.46) (0.49) (0.42) (0.43)
-0.91 -0.90 -0.63 -0.77 -0.88 -0.95
Access to local market
(0.69) (0.67) (0.62) (0.85) (0.69) (0.72)
-0.01 -0.01 -0.23 0.01 -0.01 -0.01
Governance incentive
(0.48) (0.49) (0.54) (1.00) (0.51) (0.53)
-1.96** -2.11** -2.48*** -2.71*** -2.15** -2.26**
Cultural similarity
(0.93) (0.92) (0.88) (0.71) (0.90) (0.97)
0.48 0.50 0.60 1.04** 0.53 0.43
Language similarity
(0.52) (0.58) (0.62) (0.50) (0.59) (0.60)
0.99 1.11* 1.30 -3.76*** 1.28** 1.05*
Firm experience
(0.65) (0.67) (0.89) (0.94) (0.65) (0.60)
-0.14 -0.07 0.20 0.80** -0.08 -0.05
Firm size
(0.25) (0.29) (0.44) (0.39) (0.27) (0.38)
Dummies
Nationality Included Included Included Included Included Included
Industry Included Included Included Included Included Included
Project type Included Included Included Included Included Included
Temporal effects Included Included Included Included Included Included
-1.19** -1.28* -1.52* -0.62*
Inverse Mills ratio
(0.56) (0.68) (0.89) (0.36)
1.25** 1.49** 1.69* 0.87** 2.14*** 2.32***
Constant
(0.55) (0.07) (0.93) (0.43) (0.61) (0.68)
Pseudo R 2 0.56 0.70 0.74 0.75 0.66 0.68
Wald test 276.14 249.05 257.95 319.72 245.33 250.71
Log pseudolikelihood -41.32 -30.08 -26.74 -25.45 -31.12 -31.69
F-Statistic 24.75*** 23.62*** 23.52*** 23.01*** 22.39***
Mean VIFs 4.032 4.315 4.906 4.918 4.422 4.478
Note: N = 465; * p < 0.1, ** p < 0.05, *** p < 0.01; Clustered robust standard errors reported in parentheses; A
positive (negative) coefficient of a variable shows that it increases (decreases) the probability. F-Statistic
for the null hypothesis of equal year or firm effects.

63