5.3.1 Nonlinear test
The PSTR model first tested the nonlinear relationships between the variables. The Wald, Fisher, and LRT tests were used to test the nonlinearity of Models (3) and (4). The results are summarized in Table 6. When \({H}_{0}:\gamma =0;{H}_{1}:\gamma =1\), Models (3a) and (3b) rejected the null hypothesis with a significance level of 1%; when \({H}_{0}:\gamma =1;{H}_{1}:\gamma =2\), Model ( 4a) rejected the null hypothesis at a significance level of 10%, while Model (4b) rejected the null hypothesis at a significance level of 1%. The results showed that both Models (3) and (4) have nonlinear effects, meaning that when command-control and market-incentive ER are used as threshold variables, there are significant nonlinear effects on ACEE and FQS. The PSTR model was used for the linear effects. Simultaneously, according to the estimation results, both Models (3) and (4) have an optimal threshold parameter, m = 1.
Table 6
Model
|
Non-linear test
(\({H}_{0}:\gamma =0;{H}_{1}:\gamma =1\))
|
Residual Non-linear test
(\({H}_{0}:\gamma =1;{H}_{1}:\gamma =2\))
|
Wald
|
Fisher
|
LRT
|
Wald
|
Fisher
|
LRT
|
model 3a
|
25.573***
(0.001)
|
3.501***
(0.001)
|
26.730***
(0.000)
|
11.518
(0.118)
|
1.420
(0.198)
|
11.745
(0.109)
|
model 3b
|
22.692***
(0.002)
|
2.911***
(0.006)
|
23.596***
(0.001)
|
2.554
(0.923)
|
0.297
(0.955)
|
2.565
(0.922)
|
model 4a
|
13.329*
(0.064)
|
1.747*
(0.098)
|
13.635**
(0.050)
|
7.210
(0.407)
|
0.876
(0.526)
|
7.298
(0.398)
|
model 4b
|
58.383***
(0.000)
|
7.913***
(0.000)
|
64.928***
(0.000)
|
12.618*
(0.082)
|
1.562**
(0.047)
|
12.891
(0.075)
|
Note: ***, ** and * indicate significance at the 1%, 5% and 10% levels, respectively. |
It can be seen from Tables 7 and 8 that the estimation results of the core explanatory variables in Models (3) and (4) are significant, but different types of ERs have heterogeneous effects on ACEE and FQS. In addition, control variables, such as disaster severity, urbanization level, labor cultural level, income distribution, rural economic status, and technological input, had significant dual-zone nonlinear characteristics on the core explanatory variables. Specifically, the impact coefficients of different types of ERs in the dual-zone system were different. The command-control type changed from significantly negative to insignificant for ACEE and insignificant to significant to positive for FQS. The market-incentive ER changed from insignificant to significant to negative for ACEE and from significant to positive to significantly negative for FQS. The conversion speeds were 1.370, 90.147, 5.895, and 1.974, respectively, with large differences.
The impact of command-control ER on ACEE in the low district system interval (threshold value was 22.196) showed a significant inhibitory effect with an increase in regulation intensity. The impact on ACEE in the high district system interval increased with the regulation as the intensity increased, and the significant inhibitory effect gradually disappeared (Table 7). This is because command-control ER coercively restricts carbon emissions through policies, laws, regulations, and other means. Generally, enterprises with a certain R&D level respond more strongly to command-control ERs. Currently, the main bodies of agricultural management in China are small- and medium-sized ordinary farmers, family farms, cooperatives, and leading enterprises. Therefore, mandatory orders not only cannot effectively optimize carbon emission reduction but also make carbon emission entities maintain the carbon emission intensity at the specified minimum, lacking the ability and motivation for technology research and development and updating and inhibiting ACEE. In the high-region system area, improving agricultural productivity and resource allocation efficiency through scientific and technological means can effectively promote an increase in the total output value of agriculture, forestry, animal husbandry, and fishery and reduce agricultural carbon emission intensity. Therefore, command-control ER has no significant positive impact on ACEE, and H2 does not hold.
Market incentive ER had a significant inhibitory effect on ACEE in the low district system interval (threshold value was 10.455) and had a certain promoting effect on ACEE.
For ACEE, when the intensity of the market-incentive ER is weak, carbon emitters will reduce the input intensity of chemical fertilizers and pesticides in the short term to obtain subsidies according to the actual situation, but this will not affect the total output value of agriculture, forestry, animal husbandry, and fishery. There was no promotional effect, which may have even been reduced. Therefore, it inhibited the improvement in the ACEE. When the intensity of market-incentive ER is strong, ordinary business entities will actively choose new clean technologies to meet their long-term interests, and the technological innovation motivation of business entities with strong scientific research will be enhanced. Under the condition of a high regional system, an increase in the intensity of ER plays a positive role in ACEE. Therefore, H2 holds for market-incentive ER measures.
The impact of command-control ER on FQS was not significant in the low district system interval (threshold value was 0.692), but the impact on FQS in the high district system interval showed a significant promotion effect with an increase in regulation intensity (Table 8 ). For FQS, strict control of fertilizers, pesticides, and agricultural inputs is an important way to improve security from the root. There is still a large room for improvement in the environmental management and control of China's grain production areas. Command-control ER has a significant impact on environmental and quality improvement, and the response speed is relatively fast in a period of rising dividends. Therefore, under high-institutional conditions, H3 holds true for command-control ER.
Market incentive ER had a significant effect on grain quality and safety with an increase in regulation intensity in both the low district system interval (threshold value was 10.519) and the high district system interval. This is because when initially using market-incentive ERs, the market is usually guided by incentives to regulate the market and by guiding business entities to carry out technological innovation and industrial structure adjustments to reduce and increase agricultural resources. Increase the utilization rate of inputs to ensure FQS. The green agricultural production subsidy stimulates agricultural operators to carry out green production; the subsidy effect promotes the scale of operation, and the positive promotion effect on FQS is gradually enhanced. Therefore, H3 holds.
Table 7
The impact of two ERs on ACEE
Variable
|
command-control ER
|
market-incentive ER
|
|
Low district system
|
High district system
|
Low district system
|
High district system
|
\(ln{ACEE}_{i,t-1}\)
|
0.834***
(13.03)
|
0.846***
(15.48)
|
-2.226***
(-3.93)
|
1.931***
(5.36)
|
\(ln{ER}_{i,t}\)
|
-0.035*
(-1.66)
|
-0.009
(-0.11)
|
-0.294***
(-2.57)
|
0.639***
(3.13)
|
lnDISA
|
0.132
(0.97)
|
0.335*
(1.72)
|
-0.213
(-0.92)
|
0.013
(0.09)
|
lnURBL
|
0.058
(0.68)
|
0.317
(0.90)
|
-0.605***
(-2.77)
|
0.415***
(2.65)
|
lnIND
|
-0.237***
(-2.75)
|
0.383
(1.09)
|
-2.642***
(-3.11)
|
2.071***
(3.98)
|
lnRURS
|
0.207***
(3.80)
|
-0.060
(-0.91)
|
1.271***
(7.66)
|
-0.902***
(-7.59)
|
lnRD
|
-0.079*
(-1.79)
|
-0.085
(-1.21)
|
1.772***
(5.28)
|
-1.064***
(-4.40)
|
\(\gamma\)
|
1.370
|
5.895
|
c
|
22.196
|
10.455
|
Note: ***, ** and * indicate significance at the 1%, 5% and 10% levels, respectively. |
Table 8
The impact of two kinds of ERs on FQS
Variable
|
command-control ER
|
market-incentive ER
|
|
Low district system
|
High district system
|
Low district system
|
High district system
|
\(ln{FQS}_{i,t-1}\)
|
0.989***
(57.74)
|
0.054***
(2.35)
|
1.008***
(84.62)
|
0.033
(1.48)
|
\(ln{ER}_{i,t}\)
|
0.026
(1.56)
|
0.032*
(1.72)
|
0.019***
(2.03)
|
0.034***
(-2.19)
|
lnDISA
|
0.010
(1.39)
|
-0.006
(-1.47)
|
0.012*
(1.70)
|
-0.005
(-1.31)
|
lnURBL
|
0.008***
(2.12)
|
-0.004
(-1.19)
|
0.004
(1.24)
|
0.002
(0.46)
|
lnIND
|
0.008
(1.05)
|
-0.016*
(-1.78)
|
0.001
(0.02)
|
0.001
(0.15)
|
lnRURS
|
-0.003
(-0.97)
|
0.007***
(2.44)
|
0.001
(0.12)
|
0.003
(1.05)
|
lnRD
|
-0.017*
(-1.86)
|
0.021***
(2.19)
|
-0.008
(-1.10)
|
0.007
(1.01)
|
\(\gamma\)
|
90.147
|
1.974
|
c
|
0.692
|
10.519
|
Note: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. |
The above analysis shows that the heterogeneity of ERs has different impacts on ACEE and FQS. Market-incentive ER had a more significant impact on ACEE and FQS. The significance of the model in the two-regional system interval shows a large difference, and the influence in the high-regional system interval is more significant. Therefore, market incentives should be combined with ERs to fully exploit the power of market regulation mechanisms.