Table 2.

Survival and growth at the third year for pioneer firms (models 1 and 4), nonpioneer firms (models 2 and 5), and all new firms (models 3 and 6)

	Dependent variable
	Survival rate at third year, $S^{\left(t+3\right)}$			3-y growth rate, $G^{\left(t+3\right)}$
Independent variables	1	2	3	4	5	6
Industry knowledge ($\mathrm{\Phi }$)	0.${\mathrm{457}}^{***}$	0.${\mathrm{091}}^{***}$	0.${\mathrm{091}}^{***}$	0.${\mathrm{185}}^{***}$	0.${\mathrm{054}}^{***}$	0.${\mathrm{054}}^{***}$
	(0.123)	(0.007)	(0.007)	(0.031)	(0.002)	(0.002)
Pioneer dummy			0.156			0.${\mathrm{088}}^{**}$
			(0.126)			(0.038)
Industry knowledge: pioneer dummy			0.${\mathrm{203}}^{**}$			0.${\mathrm{091}}^{***}$
			(0.093)			(0.027)
Occupation knowledge ($\mathrm{\Psi }$)	0.035	0.${\mathrm{035}}^{***}$	0.${\mathrm{036}}^{***}$	-0.029	0.${\mathrm{012}}^{***}$	0.${\mathrm{011}}^{***}$
	(0.092)	(0.007)	(0.007)	(0.023)	(0.002)	(0.002)
Years of schooling ($edu$)	0.134	0.008	0.009	0.012	0.002	0.002
	(0.091)	(0.007)	(0.007)	(0.025)	(0.002)	(0.002)
Local knowledge ($\rho$)	0.${\mathrm{228}}^{***}$	0.${\mathrm{084}}^{***}$	0.${\mathrm{085}}^{***}$	0.007	−0.${\mathrm{007}}^{***}$	−0.${\mathrm{007}}^{***}$
	(0.072)	(0.006)	(0.006)	(0.019)	(0.002)	(0.002)
Firm controls	$✓$	$✓$	$✓$	$✓$	$✓$	$✓$
Year f.e.	$✓$	$✓$	$✓$	$✓$	$✓$	$✓$
Industry f.e.	$✓$	$✓$	$✓$	$✓$	$✓$	$✓$
Region f.e.	$✓$	$✓$	$✓$	$✓$	$✓$	$✓$
Firm type	Pioneers	Nonpioneers	All	Pioneers	Nonpioneers	All
Observations	1,632	284,369	286,001	1,376	242,192	243,568
McFadden	0.2367	0.0404	0.0404
AICc	1,613	231,739	233,106
Log-likelihood	$-$541	$-$115,638	$-$116,320
$R^2$				0.344	0.152	0.152
Adjusted $R^2$				0.215	0.151	0.151
$F$ statistic				2.${\mathrm{665}}^{***}$	188.${\mathrm{475}}^{***}$	188.${\mathrm{274}}^{***}$
				(df = 226)	(df = 230)	(df = 232)

The interaction between industry knowledge and a dummy for pioneers is positive and significant, meaning that the effect of industry-specific knowledge is larger for pioneer companies. As before, all knowledge variables are expressed in SD units. Firm controls include initial size and average wage. *P<0.1; **P0.05; ***P<0.01. SEs are in parentheses. AICc, corrected Akaike Information Criterion; df, degrees of freedom; f.e., fixed effects.