How do Internet applications affect process innovation in Chinese manufacturing companies?

Figures & data

Figure 1. Conceptual model.

Source: Author's creation.

Table 1. Descriptive statistics of variables.

Variable	N	Mean	Std.Dev	Min	Median	Max
PIN	1708	3.0158	1.7836	0.0000	3.0000	5.0000
PIL	1189	0.2051	0.1781	0.0000	0.1500	1.0000
IA	1708	4.0369	3.1784	0.0000	4.0000	10.0000
IAdum	1708	0.7922	0.4059	0.0000	1.0000	1.0000
OIB	1708	1.1797	1.3362	0.0000	1.0000	4.0000
IC	1689	0.4742	0.4995	0.0000	0.0000	1.0000
Lnsize	1708	4.4731	1.3308	1.6094	4.4427	10.8198
Age	1663	14.1864	8.7166	1.0000	12.0000	126.0000
Export	1708	0.3232	0.4678	0.0000	0.0000	1.0000
Productivity	1708	12.4583	1.0415	8.9115	12.4292	17.7701
FTL	1691	0.2431	0.4291	0.0000	0.0000	1.0000
HR	1696	0.3504	0.2188	0.0000	0.2941	0.9588
Ict_invest	1471	2.4474	1.5892	0.0000	2.3979	12.7939
Informal_com	1685	0.8231	0.8661	0.0000	1.0000	4.0000
Tax_burden	1702	0.9207	1.0547	0.0000	1.0000	4.0000
SOE	1708	0.0615	0.2403	0.0000	0.0000	1.0000

Source: Authors’ calculation based on the empirical analysis.

Table 2. Baseline regression results of Internet application affecting process innovation.

	PIN				PIL
	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)	Model (6)
	Poisson	IRR	Poisson	IRR	Tobit	Tobit
IA			0.1048***	1.1105***		0.0139***
			(0.0049)	(0.0049)		(0.0019)
IAdum	0.7610***	2.1404***			0.0743***
	(0.0693)	(0.0693)			(0.0183)
Lnsize	0.0359***	1.0366***	0.0250**	1.0253 **	0.0051	0.0018
	(0.0116)	(0.0116)	(0.0104)	(0.0104)	(0.0049)	(0.0048)
Ict_invest	0.0456***	1.0467***	0.0489***	1.0501***	0.0001	−0.0003
	(0.0117)	(0.0117)	(0.0100)	(0.0100)	(0.0044)	(0.0043)
HR	−0.1108	0.8951	−0.1077*	0.8979*	−0.0327	−0.0313
	(0.0684)	(0.0684)	(0.0638)	(0.0638)	(0.0264)	(0.0260)
FTL	0.2016***	1.2234***	0.0802***	1.0835***	0.0016	−0.0169
	(0.0294)	(0.0294)	(0.0280)	(0.0280)	(0.0129)	(0.0130)
Informal_com	−0.0305*	0.9699*	−0.0225	0.9777	−0.0092	−0.0080
	(0.0169)	(0.0169)	(0.0158)	(0.0158)	(0.0066)	(0.0065)
Export	0.0446	1.0456	0.0249	1.0253	0.0320***	0.0268**
	(0.0289)	(0.0289)	(0.0266)	(0.0266)	(0.0116)	(0.0114)
Productivity	0.0132	1.0133	−0.0028	0.9973	0.0094*	0.0080
	(0.0128)	(0.0128)	(0.0118)	(0.0118)	(0.0052)	(0.0051)
Tax_burden	0.0174	1.0175	0.0119	1.0120	−0.0016	−0.0037
	(0.0156)	(0.0156)	(0.0158)	(0.0158)	(0.0066)	(0.0065)
Age	−0.0005	0.9995	−0.0017	0.9983	−0.0011*	−0.0011*
	(0.0014)	(0.0014)	(0.0012)	(0.0012)	(0.0007)	(0.0006)
SOE	−0.0400	0.9608	−0.0207	0.9795	−0.1914***	−0.1859***
	(0.0692)	(0.0692)	(0.0543)	(0.0543)	(0.0304)	(0.0299)
Constant	−0.2229	0.8002	0.4639***	1.5902***	0.0266	0.0935
	(0.1973)	(0.1973)	(0.1685)	(0.1685)	(0.0740)	(0.0713)
industry	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
City	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
N	1403	1403	1403	1403	1048	1048
Pseudo-R^2	0.1335	0.1335	0.1547	0.1547

Source: Authors’ calculation based on the empirical analysis.

Note: Robust standard errors in parentheses; ∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1.

Table 3. Average marginal effect estimation results.

	PIN		PIL
	(1) dy/dx	(2) dy/dx	(3) dy/dx	(4) dy/dx
IA		0.3150*** (0.0140)		0.0115*** 0.0016）
IAdum	2.2879*** (0.2020)		0.0617*** （0.0152）
Control Variables	Controlled	Controlled
industry	Controlled	Controlled
City	Controlled	Controlled
N	1403	1403	1048	1048
Pseudo-R^2	0.1335	0.1547

Source: Authors’ calculation based on the empirical analysis.

Note: Delta method standard error in parentheses; ∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1; control variables include Lnsize, Ict_invest, HR, FTL, Informal_com, Export, Productivity. Tax_burden, Age, SOE; the constant term is omitted for the limitation of space; the same is done below.

Table 4. Regression results of instrumental variables.

	PIN		PIL
	First stage regression	Second stage regression	First stage regression	Second stage regression
IA		0.9801***		0.4032***
		(0.1322)		(0.1132)
IV	0.4936***		0.3710***
	(0.1301)		(0.0741)
Control Variables	Controlled	Controlled	Controlled	Controlled
Industry	Controlled	Controlled	Controlled	Controlled
City	Controlled	Controlled	Controlled	Controlled
Unrecognizable test
Kleibergen–Paap rk LM statistic	37.782***		21.425***
Weak instrument tests
Cragg–Donald Wald F statistic	41.578***		23.127***
Kleibergen–Paap Wald rk F statistic	48.361***		37.409***
	[16.38]		[16.39]
Robust weak identification test
Anderson–Rubin Wald test	42.45***		8.24***
N	1385	1385	1043	1043
Pseudo-R^2		0.1348

Source: Authors’ calculation based on the empirical analysis.

Note: Robust standard errors in parentheses; ∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1. The critical value of the Stock and Yogo test at the 10% level is in parentheses.

Table 5. Heterogeneity analysis for distinguishing firm size differences.

	PIN		PIL
	SMEs	Large firm	SMEs	Large firm
	dy/dx	dy/dx	dy/dx	dy/dx
IA	0.3060***	0.2682***	0.0082***	0.0166***
	(0.0171)	(0.0233)	(0.0020)	(0.0025)
Control Variables	Controlled	Controlled	Controlled	Controlled
Industry	Controlled	Controlled	Controlled	Controlled
City	Controlled	Controlled	Controlled	Controlled
N	803	600	565	483
Pseudo-R2	0.1854	0.1230

Source: Authors’ calculation based on the empirical analysis.

Note: Delta method standard errors are in parentheses; ∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1.

Table 6. Heterogeneity analysis for distinguishing industry differences.

	PIN			PIL
	LIEs	CIEs	TIEs	LIEs	CIEs	TIEs
	dy/dx	dy/dx	dy/dx	dy/dx	dy/dx	dy/dx
IA	0.3245***	0.2267***	0.3005***	0.0102***	0.0217***	0.0260***
	(0.0294)	(0.0199)	(0.0094)	(0.0027)	(0.0024)	(0.0031)
Control Variables	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
City	Controlled	Controlled	Controlled	Controlled	Controlled	Controlled
N	355	700	348	263	523	262
Pseudo-R2	0.1698	0.1564	0.1561

Source: Authors’ calculation based on the empirical analysis.

Note: Delta method standard errors are in parentheses; ∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1.

Table 7. Heterogeneity analysis for distinguishing ownership differences.

	PIN		PIL
	NSOEs	SOEs	NSOEs	SOEs
	dy/dx	dy/dx	dy/dx	dy/dx
IA	0.3124***	0.2685***	0.0107***	0.0280***
	(0.0146)	(0.0883)	(0.0017)	(0.0035)
Control Variables	Controlled	Controlled	Controlled	Controlled
industry	Controlled	Controlled	Controlled	Controlled
City	Controlled	Controlled	Controlled	Controlled
N	1313	90	966	82
Pseudo-R2	0.1590	0.1357

Source: Authors’ calculation based on the empirical analysis.

Note: Delta method standard errors are in parentheses; ∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1.

Table 8. Heterogeneity analysis for distinguishing regions’ differences.

	PIN		PIL
	Midwest	East	Midwest	East
	dy/dx	dy/dx	dy/dx	dy/dx
IA	0. 1519 ***	0. 3612 ***	0.0114***	0.0106***
	(0. 0346)	(0. 0142)	(0.0036)	(0.0020)
Control Variables	Controlled	Controlled	Controlled	Controlled
industry	Controlled	Controlled	Controlled	Controlled
N	256	1147	227	821
Pseudo-R2	0.0881	0.1398

Source: Authors’ calculation based on the empirical analysis.

Note: Delta method standard errors are in parentheses; ∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1.

Table 9. Results of intermediate effect test.

Panel A	Mediated effect test of Internet application affecting the number of process innovations
	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)
	PIN	OIB	PIN	IC	PIN
IA	0.1048***	0.1248***	0.0926***	0.0901***	0.0956***
	(0.0049)	(0.0098)	(0.0052)	(0.0095)	(0.0049)
OIB			0.0778***
			(0.0104)
IC					0.2429***
					(0.0312)
Control Variables	Controlled	Controlled	Controlled	Controlled	Controlled
industry	Controlled	Controlled	Controlled	Controlled	Controlled
City	Controlled	Controlled	Controlled	Controlled	Controlled
N	1403	1403	1403	1397	1397
Pseudo-R2	0.1547	0.1967	0.1601	0.1430	0.1616
	Mediated effect test of Internet application influencing the quality of process innovation
	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)
Panel B	PIL	OIB	PIL	IC	PIL
IA	0.0139***	0.2339***	0.0124***	0.0796***	0.0129***
	(0.0019)	(0.0168)	(0.0020)	(0.0081)	(0.0020)
OIB			0.0136***
			(0.0047)
IC					0.0144
					(0.0126)
Control Variables	Controlled	Controlled	Controlled	Controlled	Controlled
industry	Controlled	Controlled	Controlled	Controlled	Controlled
City	Controlled	Controlled	Controlled	Controlled	Controlled
N	1048	1403	1048	1397	1044

Source: Authors’ calculation based on the empirical analysis.

Note: Robust standard errors are in parentheses; ∗∗∗p < 0.01, ∗∗p < 0.05, ∗p < 0.1.