Global development faces unprecedented challenges, and ensuring stable agricultural growth is crucial for national economic high-quality development. Agricultural science and technology innovation is fundamental to achieving stable and sustainable agricultural development and ensuring food security, especially in resource-constrained countries like China. China has made significant progress in agricultural science and technology, exceeding a 60% contribution rate to agricultural progress. This success is partly attributed to government initiatives like the National Modern Agricultural Demonstration Zone (NMADZ), which aims to modernize agriculture through policy guidance and financial support. However, the impact of NMADZ on agricultural science and technology innovation requires further investigation. This study addresses this gap by analyzing the effect of NMADZ on agricultural science and technology innovation and exploring its underlying mechanisms using panel data and a difference-in-differences approach. The study's contributions include precise research objects, reliable methodology, a novel perspective focusing on agricultural science and technology innovation, and a robust investigation of the influence mechanism.

Existing research on the NMADZ primarily focuses on its ecological and environmental effects, modern agricultural development, farmers' income, and employment growth. Studies have highlighted the role of government policies in promoting agricultural development through financial subsidies and tax incentives. However, there is limited research exploring the impact of NMADZ specifically on agricultural science and technology innovation. This study bridges this gap by examining the policy's effect on innovation and investigating the mechanisms through which it operates. The study leverages technology diffusion theory, resource-based theory, and human capital theory to provide a comprehensive understanding of the NMADZ's influence on agricultural science and technology innovation.

This study employs a difference-in-differences (DID) approach using panel data from 696 regions in China from 2007 to 2017. The establishment of NMADZs (283 regions in three batches between 2010 and 2015) serves as a quasi-natural experiment, with these regions forming the treatment group and other regions acting as the control group. Agricultural science and technology innovation (ATI) is measured using an entropy value method integrating four aspects: innovation environment, support, input, and output. The independent variable is the NMADZ policy (DID), while mediating variables include financial resource allocation, financial support for agriculture, and human capital. Control variables include the proportion of primary industry output in regional GDP, total agricultural machinery power, fixed asset investment, loan balances, and general fiscal budget revenue and expenditure. The study uses three econometric models: a basic model to test the overall effect of NMADZ on ATI, and two mediating models to explore the mechanisms. Parallel trend tests, robustness checks (replacing explanatory variables, excluding samples, excluding other policies, placebo test, PSM-DID, and CSDID), and heterogeneity analyses (fiscal autonomy, geographical location, transportation infrastructure) are conducted to ensure the reliability and validity of the findings. Data sources include the China Science and Technology Statistical Yearbook, National Bureau of Statistics, China Rural Statistical Yearbook, CSMAR database, and China Banking and Insurance Regulatory Commission.

The benchmark regression analysis reveals a significant positive effect of NMADZ on ATI, with ATI increasing by 3.6% in treatment regions compared to control regions. This effect persists after controlling for various factors and remains robust across various robustness tests, including replacing the dependent variable, excluding specific samples, accounting for other relevant policies, placebo tests, PSM-DID, and CSDID estimations. The parallel trend test confirms the plausibility of the DID approach. The mechanism analysis shows that NMADZ improves ATI through three mediating channels: significantly increased financial resource allocation (16%), substantially higher financial support for agriculture (35.9%), and a notable rise in human capital (41.3%). Sobel tests confirm the significance of these mediating effects. Heterogeneity analysis reveals that the positive impact of NMADZ on ATI is stronger in regions with higher fiscal autonomy, the western region, and regions with better transportation infrastructure. This indicates that regional disparities in fiscal capacity and infrastructure influence the effectiveness of the NMADZ policy.

The findings confirm that NMADZ significantly enhances agricultural science and technology innovation in China. The study's results support the policy's effectiveness in promoting agricultural modernization and contribute to the literature by specifically focusing on the impact on agricultural science and technology innovation. The mediating effects of financial resource allocation, financial support for agriculture, and human capital highlight the importance of targeted investments and capacity building in promoting innovation. The heterogeneity analysis underscores the role of institutional factors and infrastructure in shaping the effectiveness of regional policies. The stronger impact in the western region highlights the potential of NMADZ to reduce regional disparities in agricultural development.

This study demonstrates the significant positive impact of China's NMADZ policy on agricultural science and technology innovation, particularly in regions with favorable conditions. The findings highlight the importance of financial support, human capital development, and infrastructure in fostering agricultural innovation. Future research could explore the long-term sustainability of these effects, investigate the distributional impacts of the policy across different farmer groups, and assess the environmental implications of increased agricultural technology adoption.

The study relies on secondary data, which may limit the depth of analysis. The definition of agricultural science and technology innovation might not encompass all aspects of innovation. The policy effects might vary due to unobservable factors not included in the model. The study primarily focuses on China's context, and the generalizability of the findings to other countries might be limited.