Page 157 - IJOCTA-15-4

Green innovation efficiency measurement and its influencing factors in specialized and new enterprises
through the capital structure, and the pressure 4.2.2. Correlation analysis
of high debt often forces firms to prioritize short-
term gain projects and reduce long-term invest- Table 7 presents the regression results of the
ments in green technology. 47 Board size has an model, indicating the effects of R&D investment,
ES, COORP, ENVIR, and GOVNM on GIE of
inverted U-shaped relationship with innovation
SNEs in Zhejiang. The five hypotheses raised pre-
efficiency, while moderate size enhances decision-
viously were analyzed accordingly:
making expertise, but overexpansion leads to
higher coordination costs and lower decision-
(i) The GIE is significantly positively cor-
making efficiency. In contrast, net operating cash
related with R&D investment. Overall,
flow may have an indirect effect by affecting the
the standardized regression coefficient for
overall investment capacity of the firm. Together,
R&D investment measured by R&D capi-
these variables constitute the key governance af-
tal stock is 0.175 and passes the 5% signif-
fecting the GIE of SNEs. The specific definitions
icance test, indicating that R&D invest-
of the variables are presented in Table 4.
ment promotes the improvement of GIE
in SNEs. Therefore, H1 is verified. Un-
4.1.4. Model setting der the influence of increasing R&D in-
vestment in enterprises, the enthusiasm of
To study the factors affecting GIE of SNEs in enterprises for scientific research also in-
Zhejiang and provide suggestions for SNEs in creases, which promotes technological in-
Zhejiang to improve GIE, Equation (7) is estab- novation and positively affects the GIE in
lished. SNEs. 48 Therefore, enterprises should em-
phasize R&D investment to drive the pos-
itive incentive effect on technological in-
GIE = α 0 + α 1 RD + α 2 ES + α 3 COORP novation.
X (ii) There is a significant negative correlation
+ α 4 ENV IR + α 5 GOV NM + α j w j + ε
between ES and GIE. Overall, the re-
j
(7) gression coefficient value for ES is 0.83,
where α 0 is a constant term, α 1 , α 2 , α 3 , α 4 , α 5 , indicating that every 1% increase in ES
and α j represent the regression coefficients, w j is leads to a 0.83% decrease in GIE in SNEs.
the control variable, and ε is the possible residual Therefore, H2 is verified. This shows that
term. the existing ES is relatively reasonable for
SNEs, and blind expansion will only hin-
4.2. Regression analysis of influencing der the allocation of its resources to meet
factors the expansion needs, resulting in a de-
cline in its innovation ability. If the green
4.2.1. Descriptive analysis
innovation ability is further enhanced, it
Four indicators—the lowest value, maximum will only result in a cost-squeezing effect,
value, average value, and standard deviation—are which is not conducive to improving GIE.
used to assess the particular circumstances of each (iii) There is a significant positive correla-
variable before the antecedent regression analysis tion between COORP and GIE. Over-
on the GIE of SNEs in Zhejiang. The results are all, the regression coefficient value for
shown in Table 5. It can be seen that the standard the COORP level is 0.107 and passes
deviation of GIE of SNEs in Zhejiang is 0.372, in- the significance test at 5%. If the level
dicating that the overall difference is small. The of COORP increases by 1%, GIE in-
standard deviations of ENVIR and COORP in creases by about 0.107%. Therefore, H3
SNEs are remarkably large, which indicates sig- is verified. The breadth of cooperation
nificant variations in the degree of emphasis on expands, bringing more cooperation ob-
environmental protection among SNEs. jects, forming a network circle of COORP
R&D, obtaining more high-quality re-
Pearson’s correlation coefficient is shown in sources, prompting enterprises to obtain
Table 6. It can be seen that GOVNM, R&D in- heterogeneous knowledge from coopera-
vestment, and COORP are positively correlated tion objects and their existing homoge-
with GIE, while ES is negatively correlated with neous resources to produce complemen-
GIE. Meanwhile, correlation coefficients across all tary effects, stimulating internal R&D, in-
variables are lower than 0.5, indicating that there creasing R&D efforts, achieving techno-
is no collinearity problem across the variables. logical innovation, and improving GIE.
699