3.2.2. Core explanatory Variable

The key explanatory variable in this paper is the policy variable treatedi × timet (difference-in-differences) of high-standard farmland construction. This variable is created through the interaction of a time dummy variable, representing the implementation of the policy and the regional area of high-standard farmland treated. The size of its parameter indicates the impact of the high-standard farmland construction policy on ATFP.

#### 3.2.3. Control Variables

According to existing research [44,45], the following variables are used as control variables in this study: (1) Infrastructure, represented by the number of road miles per unit area (ROAD); (2) Human capital, represented by the average number of years of education for the regional labor force (EDU); (3) Urbanization level, represented by the ratio of the urban population to the total population (UR); (4) Land quality, represented by the ratio of effective irrigated area to sown area (LAQA); (5) Disaster rate, represented by the ratio of the disaster area to total sown area (DR) to control for the impact of climate on ATFP; (6) Agricultural planting structure, represented by the ratio of sown area of food crops to the total sown area (AS); (7) Fiscal support to agriculture, represented by the ratio of fiscal support for agriculture expenditure to total fiscal expenditure (AF). These variables are included to account for their potential influence on ATFP and to ensure the accuracy of the results of this study.

#### *3.3. Data*

This paper uses data from 30 provinces in China for the study period of 2008–2017. Hong Kong, Macau, Taiwan, and Tibet are not included due to a lack of sufficient statistical data. The data used for calculating ATFP, including input and output variables, were obtained from the *China Statistical Yearbook* and *China Rural Statistical Yearbook*. Control variables were obtained from the *China Statistical Yearbook* and EPS database. Data on high-standard farmland construction was obtained from the *China Financial Yearbook*. Some

abnormal data were removed, and missing data were completed using the interpolation method. The statistical description of the data used in this paper is shown in Table 1.


**Table 1.** Descriptive statistics of variables.

#### **4. Empirical Results**

#### *4.1. The Results of ATFP Measurement*

In this study, the Empirical Border Malmquist Super-Global index was used to measure the Total Factor Productivity (ATFP) growth in China from 2008 to 2017 (Table 2). The results show that China's ATFP experienced an upward trend with an average annual growth rate of approximately 3.6% during this period. Most provinces had positive ATFP growth, and the growth of ATFP was generally balanced across provinces. The results also show that the growth of ATFP in China was determined by a combination of technological change (TC) and technical efficiency (TE). The average annual growth rate of TC was 2.8%, while the average annual growth rate of TE was 0.7%. These findings are consistent with those of previous studies and by existing studies [39,46,47]. The Malmquist index was converted to a growth index with 2008 as the base period to understand the results better.

**Table 2.** ATFP Growth and Decomposition in China from 2008 to 2017.


Note: The mean values in Table 2 are calculated from the geometric mean.

#### *4.2. Baseline Regression Results*

The effects of the high-standard farmland construction policy on ATFP were analyzed in this section, with the results shown in Table 3. Model 1 represents the results without controlling for any variables, while Model 2 shows the results after adding control variables. The results in Table 3 indicate that, when controlling for all variables, time-fixed effects and regional-fixed effects, the impact of the high-standard farmland construction policy

on ATFP is significant at the 1% confidence level with an estimated coefficient of 0.010. This indicates that, on average, the policy has increased ATFP by 1.00%. The policy reform had a significant effect on ATFP. As a key land remediation project in China, the construction of high-standard farmland can significantly improve ATFP and promote sustainable agricultural development.

**Table 3.** Baseline regression results.


Note: \*\*\* is significant at the significance level of 1%.

According to the results, infrastructure development can significantly increase ATFP, which is consistent with the findings of Shamdasani et al. [22]. Human capital, represented by the average number of years of education for the regional labor force, can also promote ATFP growth, which is consistent with the study of Chen et al. [47]. Farmers with higher human capital are more capable of adopting new agricultural technologies and are more likely to be familiar with them. The level of urbanization can significantly increase ATFP, which is generally consistent with the results of Li et al. [42]. Land quality has a significant positive effect on ATFP growth, which is generally consistent with the findings of Song and Pijanowski [48]. In addition, the regression results of disaster rate, agricultural restructuring coefficient, and fiscal support to agriculture are statistically insignificant, and this paper has yet to find empirical evidence that disaster rate, agricultural restructuring, and fiscal support to agriculture will enhance ATFP.

#### *4.3. Dynamic Effect of the Policy*

To further understand the policy's impact, this paper explores the policy's dynamic effects on ATFP through an interaction regression of the policy variable and time dummy variables [37]. The regression results are shown in Table 4. Model 1 controls for time and area effects, while Model 2 adds control variables based on Model 1. The results in Table 4 suggest the following conclusions. First, the estimated parameters for the first two years of policy implementation are statistically insignificant, indicating that the policy's impact on ATFP growth has a lag period of two years. Second, the estimated parameters increase with increasing years, indicating that the policy's effect on ATFP growth is continuous and increasing. We believe that there are differences in the understanding of the policy of highstandard farmland construction in various regions and no standardized implementation, which leads to the lagging effect of the policy.


**Table 4.** Dynamic effects of the policy.

Note: \*\*\* indicates significance at the significance level of 1%.

#### *4.4. Heterogeneity Analysis*

The impact of high-standard farmland construction policy on ATFP is influenced by resource endowment and policy bias. There is regional heterogeneity in the effects of ATFP. The results of heterogeneity can better optimize the policy. In this paper, heterogeneity is analyzed based on natural geographical location, agricultural functional areas, and productivity differences.

#### 4.4.1. Heterogeneity of Natural Geographic Location

We consider that there will be significant heterogeneity in the impact of high-standard farmland construction policies on ATFP growth in different natural geographic locations. This study divides the study area into three regions in eastern, central, and western China and conducts grouped regressions based on Equation (3). The results, shown in Table 5, suggest that in the central region of China, the construction of high-standard farmland has the most significant effect on improving ATFP. However, the results for the samples from the eastern and western regions are not statistically significant.


**Table 5.** Results of heterogeneity of natural geographic location.

Note: \*\*\* indicates significance at the significance level of 1%.

#### 4.4.2. Productivity Heterogeneity

For provinces with different productivities, the effect of high-standard farmland construction on ATFP may vary. In this paper, the policy effects on different quartiles of ATFP are assessed with the help of panel unconditional quantile regressions, and the results are shown in Table 6. Compared with conditional quantile regression, unconditional quantile regression does not depend on the increase or decrease of control variables and is widely used for heterogeneity analysis of treatment effects [49]. Model 1, Model 2, Model 3, and Model 4 represents the regression results for quartiles 25, 50, 75, and 90, respectively. The estimated coefficients were 0.041 and 0.026 at quintiles 25 and 50, respectively, and were statistically significant. The estimated coefficients were not significant at the 75th and 90th quartiles. The study's results indicated that the effect of high-standard farmland construction on ATFP gradually diminished as the quantile increased. High-standard farmland as a land improvement policy can significantly reduce the differences in ATFP among provinces.


**Table 6.** Unconditional quantile regression results.

Note: \*\*\* and \*\* indicate significance at the significance level of 1% and 5%.

#### **5. Discussion**

During 2008–2017, China's ATFP showed an increasing trend, which is similar to the findings of Liu et al. [50] and Li and Lin [41]. The decomposition of the ATFP shows that China's ATFP growth is still largely dependent on technological change. This finding is generally consistent with that of Fan et al. [51]. Unlike these studies, the focus of our study is to explain the possible reasons for the changes in ATFP in China, which can help us to understand the sustainable growth of Chinese agriculture. This paper argues that the improvement of China's ATFP is mainly due to the following aspects: first, China has continued to make efforts to improve the quality of agricultural science and technology in recent years, breaking through several critical technological bottlenecks and applying several high-tech achievements, such as breeding and promoting important new varieties of super rice, water-saving and drought-resistant wheat, and transgenic insect-resistant cotton [52,53]. Secondly, China has invested more in agricultural infrastructure in recent years, improving agricultural production conditions and significantly contributing to ATFP growth [54].

Then, our research shows that a high-standard farmland construction policy can significantly improve ATFP and thus promote sustainable agricultural development. Existing studies started to discuss the impact of high-standard farmland construction on farmers' behavior and farm income [31,32] but lacked a discussion on sustainable agricultural development. We innovatively used ATFP to measure agricultural sustainability and studied the policy effects. Our study can provide lessons for sustainable agricultural development in developing countries. We think the main reasons why high-standard farmland construction can contribute to ATFP growth are as follows. First, the construction of high-standard farmland can enhance agricultural production conditions and increase the disaster resistance of agriculture, thus ensuring food security. In agricultural production, irrigation has always been the weak segment. Traditional agriculture often reduces grain yield due to untimely or insufficient irrigation [55–57]. The construction of high-standard farmland can effectively alleviate the problem of difficult irrigation and promote the improvement of agricultural ATFP. Secondly, high-standard farmland enhances agricultural scale, mechanization, and social services and supports agricultural transformation and upgrading. In China, the problem of land fragmentation is severe, which is not conducive to agricultural

scale and mechanization. A study by Adamopoulos and Restuccia [58] showed that low agricultural productivity in developing countries mainly comes from small planting scale, low productive investment, and low agricultural mechanization. High-standard farmland can effectively compensate for these deficiencies and thus increase ATFP. In addition, our study shows that the policy effects of high-standard farmland are progressively growing, further suggesting that the construction of high-standard farmland can promote sustained agricultural productivity growth.

Moreover, our research indicates that the effects of high-standard farmland construction policy on ATFP are significantly heterogeneous under different conditions, which is also an important contribution to our paper. Through heterogeneity analysis, we can obtain the differences in policy effects in different regions or different groups. The results of heterogeneity can help us optimize the high-standard farmland construction policy in a more targeted way. First, the high-standard farmland construction policy significantly affects the central region rather than the east and west. This paper considers the possible reason for this is that central China is the main grain-producing region, and the construction of high-standard farmland is more standardized [59]. Second, the enhancement effect of the high-standard farmland construction policy is more obvious in the main grain-producing regions. The reason is that the construction of high-standard farmland in China mainly focuses on grain production [60]. Finally, the high-standard farmland construction policy is more pronounced in provinces with lower ATFP, which suggests that high-standard farmland construction is an important policy tool to reduce inter-regional productivity differences and may narrow the income gap between provinces. In addition, high-standard farmland construction policies can increase ATFP through technological change and technical efficiency improvements. The results suggest that high-standard farmland enhancement of agricultural ATFP is multi-dimensional. After the construction of high-standard farmland, farmers can introduce more advanced agricultural technologies into agricultural production, which will significantly promote agricultural technological change [61,62]. At the same time, the construction of high-standard farmland will reduce the degree of land fragmentation, which can improve the efficiency of agricultural technology [32].

Based on our findings, we recommend the following actions: firstly, continue promoting the construction of high-standard farmland. China's current proportion of highstandard farmland is still low and requires further efforts to strengthen construction efforts. To support this, the government should increase financial investment in the construction of high-standard farmland and consider implementing balanced matching funds from both the central and local governments. Additionally, the construction area of high-standard farmland should be carefully planned, and the overall area of high-standard farmland should be increased to support sustainable agricultural development. Then, improve the quality and standards of high-standard farmland construction. The government should prioritize quality management during project implementation, refine construction quality requirements, and ensure that construction units follow technical specifications closely. Additionally, the acceptance process for high-standard farmland projects should be thorough and rigorous. Finally, explore local solutions for high-standard farmland construction. Local governments should consider developing differentiated policies for constructing high-standard farmland based on the heterogeneity of policy implementation in different regions. For example, in the eastern region, agricultural science and technology research, development, and promotion should be prioritized. In the central region, water-saving irrigation technology should be promoted to improve water resource utilization efficiency. In the western hilly areas, the promotion of local practical technologies should be emphasized.

#### **6. Conclusions and Prospects**

#### *6.1. Conclusions*

In this study, we employed China's high-standard farmland construction policy as a "quasi-natural experiment" to investigate the relationship between land consolidation and agricultural sustainability. We used Agricultural Total Factor Productivity (ATFP) as a measure of sustainable agricultural development and employed a continuous differencein-differences (DID) model to identify the causal relationship between the policy and ATFP and to explore the dynamic effects and mechanisms at play. Our findings indicate that a high-standard farmland construction policy can improve ATFP and thus promote sustainable agricultural development in China. Specifically, we found that:

