Table 9

Instrumental variable method

Variable(1)(2)
DTlnCOGS
CSE0.106*** 
(23.199) 
lnREV0.081*0.969***
(1.938)(149.611)
Dec×lnREV0.180−0.164**
(0.990)(−2.405)
DT 0.010**
 (2.547)
Dec×lnREV×DT 0.056**
 (2.171)
EInt0.088***−0.001
(10.477)(−1.348)
Dec×lnREV×EInt0.020−0.048***
(0.441)(−5.319)
AInt−0.054***−0.000
(−12.184)(−0.557)
Dec×lnREV×AInt−0.047***−0.002
(−2.745)(−0.614)
Sdec−0.118***−0.003
(−3.612)(−0.907)
Dec×lnREV×Sdec−0.168−0.007
(−1.165)(−0.327)
GDPgrow0.7230.131**
(1.462)(2.344)
Dec×lnREV×GDPgrow−0.3011.711***
(−0.163)(4.369)
Independ0.081***0.004
(5.945)(0.271)
Top1−0.440***−0.003
(−8.448)(−0.584)
Mshare0.0109*0.015**
(1.837)(2.536)
Dual0.114***0.001
(6.143)(0.421)
Age−0.040***−0.008***
(−2.699)(−4.917)
Soe−0.251***0.003
(−13.644)(1.406)
Lev−0.205***0.021***
(−4.485)(3.990)
Size0.170***−0.001
(22.841)(−0.489)
_cons−4.700***0.019
(−24.839)(0.768)
Industry and Year FEYesYes
N22,83822,838
adj. R20.4380.862
F530.4351410.427
Kleibergen–Paap rk Lagrange multiplier (LM) statistic 505.706***
Kleibergen–Paap rk Wald F statistic 266.779***

Note(s): Table 9 introduces the natural logarithm of local general public fiscal expenditure on science and technology (CSE) as an instrumental variable and employs two-stage least squares regression analysis for Model 2. Definitions of all variables are provided in Appendix 1. All models are controlled for industry and year effects

t-values are in parentheses. ***p < 0.01, **p < 0.05 and *p < 0.1

Source(s): Table created by authors

or Create an Account

Close Modal
Close Modal