Growth and Technological Leadership in US Industries: A Spatial Econometric Analysis at the State Level, 1963-1997

Table 7. Endogenous Growth Model with Diagnostics for Spatial Models (continued).

Sectors		Manufacturing		Wholesale/Retail Trade			Government
Models^a	OLS	_______ML-HET_______		OLS	______GM-HET______	OLS	_____GM-HET______
Variables		Low	High		Low High		Low High
Constant	-0.459***	-1.660***	0.332	-0.200*	-0.252** -0.280	0.170	0.066 0.331
Human capital	-0.012	-0.001	-0.002	0.010**	0.020*** 0.006	0.036	0.018 -0.046
Spillover human capital	0.042***	0.058***	0.014	-0.004	-0.006 0.001	-0.002	0.001 -0.007
Spatial domestic effect	-0.048**	-0.032**	-0 114***	0.007	-0.058*** -0.001	-0.046	-0.046** 0.032
Catch-up	0.046***	0.055***	0.070***	-0.002	0.033*** 0.016	0.015***	0.02* 0.02*
Log labor ratio	-0.829*	-1 711***	-0.569	-0.593***	-1.005*** -0.195	-0.157	-0.649** 0.179
Log capital ratio	0.386*	0.935***	0.260	0.310***	0.502*** 0.135	0.045	0.330** -0.137
Spatial AR parameter Lagged GDP growth		-0.180			0.156***		0.400***
R² adjusted⁰AIC	0.60 -26.17			0.45 -112.11		0.41 -83.79
LIK	20.08			63.05		48.90
JB	3.81			3.83		3.07
BP^d	14.56**			16.54***		20.57***
Chow-Wald^e		60.81***			35.01***		13.68**
I	-0.08			0.07		Q19***
LM-error	0.65			0.35		3.02*
Robust LM-error	2.77*			3.86**		10 47***
LM-Iag	4.25**			0.06		0.076
Robust LM-Iag	6.36**			3.58***		7.53***
LM-SARMA______	__________7.02**			3.94		10.55***
^,' Significance at the 1,	5 and 10% level is	' 1 AL * signaled by ,	and *, respectively.

^b Spatial Autoregressive parameter of the spatial error model.

^cJarque Bera test for normality of the errors.

^d Breusch-Pagan test with random coefficients as the alternative hypothesis.

^c Chow-Wald test for stability of coefficients across regimes.