China faces the challenge of feeding a large population with limited arable land, necessitating increased farming efficiency. Digitalization, particularly internet access, is seen as a key driver of this efficiency. While internet penetration in rural China is increasing, the impact of internet use on grain farming productivity remains understudied. Existing research often focuses on high-value crops and doesn't differentiate between internet access and its specific uses. This study addresses this gap by examining the impact of both internet access and its use for farming purposes on the technical efficiency (TE) of grain farmers in Central China (Henan, Hubei, and Hunan provinces). The study highlights the importance of investigating the specific uses of the internet, as its impact on farming productivity may vary significantly based on whether farmers utilize it for farming-related information or other purposes. Grain farming in China, unlike high-value crops, is often less market-oriented and relies heavily on farmers' experience, making the study of internet influence on this sector crucial.

Numerous studies have explored the link between internet access and agricultural productivity. Some studies at the national level showed a positive relationship between internet access and agricultural productivity. Other farm-level studies in developing countries indicated that internet access and use improved production efficiency for various crops. These studies suggest that internet use impacts agricultural productivity through several channels: reduced information search costs, accelerated technology adoption, and improved risk management. However, most prior research uses internet access as a proxy for internet use, failing to distinguish between access and actual usage for farming purposes. Furthermore, the existing studies primarily focus on high-value crops like fruits and vegetables, neglecting the potential impacts on grain production, the most prevalent type of farming in China and many developing countries. This study directly addresses the limitations by focusing on grain production and distinguishing between mere access and actual usage.

This study uses data from 855 farm households in Central China collected in July 2019 using a combined stratified and random sampling method. Technical efficiency (TE) is evaluated using stochastic frontier analysis (SFA) with a translog production function, accounting for input factor substitution. The model includes five inputs: labor, fertilizer, seedlings, pesticides, and other inputs. To address the potential endogeneity of internet access and use (farmers with higher TE may be more likely to use the internet), the study employs endogenous switching regression (ESR) models. ESR models are preferred over two-stage least squares (2SLS) and propensity score matching (PSM) as they address endogeneity caused by unobservable variables. The ESR model comprises two stages. The first stage models the decision to access or use the internet, using observable household and crop characteristics. The second stage models the impact of internet access and use on TE, considering the potential self-selection bias. Instrumental variables (IVs) are used to address endogeneity concerns. 'Preference for ICT products' (ownership of a smartphone or computer before 2013) serves as the IV for internet access, while 'years of using the internet' acts as the IV for internet use. The study also employs unconditional quantile regression (UQR) to analyze heterogeneity in the effects of internet access and use across different TE quantiles. Finally, a mechanism analysis investigates the impacts of internet access and use on technology adoption and risk management, using ESR models to estimate Average Treatment Effects (ATT and ATU).

The study's key findings are as follows: 1. **Internet Use, Not Just Access, Matters:** The study found that using the internet to access farming-related information significantly improves the technical efficiency (TE) of grain farmers. Simply having internet access does not necessarily lead to higher TE. This finding highlights the crucial role of active engagement with agricultural information online. 2. **Non-Farming Internet Use Can Be Detrimental:** The use of the internet for non-farming purposes, such as entertainment, might hinder TE improvement. This suggests that the productive use of internet time is vital for enhancing farm efficiency. 3. **Heterogeneous Effects:** Heterogeneity analysis revealed that the internet's positive impact on TE is most pronounced among farmers with initially lower TE. This indicates that the internet can effectively reduce the productivity gap between high and low-performing farmers. 4. **Mechanism Analysis:** The study investigated the mechanisms through which internet use affects TE, focusing on technology adoption and risk management. The results showed that internet use, and not just access, was positively associated with both technology adoption and improved risk management, further substantiating the importance of actively using the internet for agricultural purposes. The positive impact of internet use was observed on both farmers who are currently using the internet for agricultural purposes and those who are not (i.e., ATU and ATT were both positive). 5. **Quantitative Results:** The average TE of grain farmers with internet access was 4.35% higher than those without access. The average treatment effect on the treated (ATT) for internet access was 0.027, while the average treatment effect on the untreated (ATU) was 0.248, both significant at the 1% level. ATT and ATU for internet use were 0.042 and 0.271 respectively, indicating a much larger impact of internet use relative to simple access. These effects were significant at the 1% level.

The findings of this study underscore the importance of distinguishing between internet access and its purposeful use in assessing its impact on agricultural productivity. The significant positive impact of using the internet for farming-related information suggests that targeted interventions promoting such use are warranted. Furthermore, policies aimed at guiding farmers towards productive online activities and developing user-friendly agricultural information platforms can maximize the benefits of internet access for improving farm efficiency. The study's focus on grain production, which represents a substantial portion of agricultural activity, provides valuable insights into the broader implications of internet use in agriculture, particularly in developing countries. The finding that the internet can help reduce TE discrepancies highlights its potential role in promoting equitable agricultural development.

This study provides compelling evidence that using the internet for agricultural purposes significantly enhances the technical efficiency of grain farmers in Central China. Simply having access to the internet is insufficient; active engagement with relevant online information is critical for improving farm productivity. Policy interventions should focus on encouraging and facilitating the productive use of the internet among farmers, including the development of user-friendly platforms and resources. Future research could explore the long-term impacts of internet use on grain farming, examine the role of different types of internet access (e.g., broadband vs. mobile), and investigate the interaction between internet use and other factors influencing farm efficiency.

The study's findings are based on data from a specific region in China and may not be fully generalizable to other contexts. The cross-sectional nature of the data limits the ability to establish causal relationships definitively. The self-reported nature of some variables may introduce measurement errors, though instrumental variables help mitigate this issue. Future research should explore these limitations and employ longitudinal data to build a stronger causal case.