For Models (2) and (3), we find that the interaction term Gov x LSOE is positive and statistically significant at the 1% level, while the interaction term Gov x PROE is negative and statistically significant at the 5% level.
In contrast to the results on LSOE in the NR investment regression, in the RD regression shown in Panel B of Table VI, with an increase in government intervention, there is a progressive reduction in the coefficient on LSOE, from -0.02 to -0.059.
The results indicate that the rates of fixed asset, equity, and natural resource investment significantly increase for LSOEs after a transition from a PROE, while significant reductions are discernible for PROEs after a transition from an LSOE.
Further analysis of the ownership variables suggests that the LSOE ratio is strongly positively associated with the concentration of firms in the utilities and transportation industries.
Although the results in Panel B are found to be stronger, both economically and statistically, the coefficients on government intervention, LSOE, and the interaction term between government intervention and LSOE in Panel A remain significant and of similar magnitude to those of the overall sample.
In our context, this refers to the likelihood that LSOE business groups will be located in larger provinces with higher government intervention.
LSOE A dummy variable that is equal to one if a firm is controlled by local government and zero otherwise.
Sample Composition by Industry and Type of Ownership This table reports the industry distribution for the full sample and the CSOE, LSOE, and PROE subsamples from 2005 to 2012.
The managers of these local SOEs (LSOEs), who play a semiofficial role within the bureaucratic hierarchy, are appointed by local governments (Bai et al., 2000; Jin, Qian, and Weingast, 2005).
Since the ownership of LSOEs is completely controlled by the local government and the managers of LSOEs are appointed by local government, the effects of government intervention on such investment allocations are primarily found in LSOEs.
Moreover, the performance-centered promotion system for local officials in China results in suboptimal asset structures within LSOEs, which, in mm, reduces investment efficiency.
In this section, we examine why local governments prefer LSOEs to invest in fixed assets, natural resources, and equity assets, while avoiding investment in R&D.