TL;DR
This study investigates whether farmland transfer in China leads to reduced chemical fertilizer use. Using an endogenous switching probit model on data from 442 corn farmers, the research finds that farmland transfer-in significantly reduces chemical fertilizer application. Furthermore, concentrated farmland transfer is more effective at reducing fertilizer use than dispersed transfer.
What Problem Does This Paper Address?
The overuse and inefficient application of chemical fertilizers in China pose significant environmental challenges, impacting soil health, water quality, and overall agricultural sustainability. While reducing chemical fertilizer use is a key policy goal, its effectiveness is debated. Farmland transfer, a growing trend in China, is proposed as a potential mechanism to achieve this reduction by enabling larger-scale farming operations. However, existing research offers conflicting conclusions on whether farmland transfer actually leads to reduced fertilizer use, with some studies suggesting it might even increase it due to the pursuit of higher yields. This paper aims to clarify this relationship by examining the impact of farmland transfer, and specifically distinguishing between concentrated and dispersed transfer, on chemical fertilizer reduction.
Methods
The study employs an endogenous switching probit model (ESP). This econometric approach is crucial because decisions about farmland transfer and chemical fertilizer application are likely influenced by unobservable factors and can involve "self-selection" (farmers choosing to transfer land based on expected benefits, which might also correlate with their fertilizer use). The ESP model addresses this by:
- Modeling the decision to transfer farmland: This first stage accounts for factors influencing a farmer's choice to participate in land transfer.
- Modeling the outcome (chemical fertilizer reduction): This second stage estimates the effect of farmland transfer on fertilizer reduction, while simultaneously accounting for the selection bias identified in the first stage.
- Counterfactual analysis: The model constructs hypothetical scenarios (e.g., what if farmers who transferred land hadn't transferred, or vice versa) to isolate the true treatment effect of farmland transfer.
The paper also calculates an "economic optimal chemical fertilizer application amount" using a Cobb-Douglas production function and profit maximization principles. This serves as a benchmark to quantify whether farmers are indeed over-applying fertilizer.
Datasets
The study utilizes data from a field survey of 442 corn farmers in Heilongjiang Province, China, conducted between June and September 2021. The survey employed a three-stage random sampling method.
Key Insights and Ideas
- Over-application is rampant: The study reveals that, on average, farmers in the surveyed area apply significantly more chemical fertilizer than the economically optimal amount, indicating substantial room for reduction.
- Farmland transfer is a viable tool for reduction: Contrary to some prior findings, this research provides strong evidence that farmland transfer-in positively contributes to chemical fertilizer reduction. This suggests that larger operational scales enabled by land transfer can lead to more efficient practices.
- Concentration is key: The spatial arrangement of transferred farmland matters. Concentrated farmland transfer (where new plots are adjacent to existing ones) is significantly more effective at reducing chemical fertilizer use than dispersed transfer. This highlights the importance of economies of scale in both operation and plot size.
- Addressing self-selection is critical: The use of the endogenous switching probit model is a key methodological contribution, allowing for a more robust estimation of the causal effect of farmland transfer by accounting for the complex decision-making processes of farmers.
Key Results
- Average fertilizer over-application: The actual chemical fertilizer application rate was 12.12 kg/mu higher than the economic optimal rate.
- Farmland transfer-in reduces fertilizer use: The average treatment effect (ATT) of farmland transfer-in on chemical fertilizer reduction was estimated to be 0.044, indicating a significant positive impact. If farmers who transferred land had not done so, chemical fertilizer reduction would have decreased by 18.644%. Conversely, if non-transferring farmers had transferred land, fertilizer reduction would have increased by 26.519%.
- Concentrated transfer is more effective: The average treatment effect (ATT) for concentrated farmland transfer-in was 0.059, leading to an 18.790% reduction. This is notably higher than the effect of non-concentrated transfer. If non-concentrated transfer farmers switched to concentrated transfer, the reduction effect would increase by 18.487%.
Limitations and Future Work
The authors acknowledge several limitations:
- Crop specificity: The study focused solely on corn growers. Future research should investigate other crop types, as fertilizer application behaviors can differ between commercial and food crops.
- Geographic scope: The findings are based on Heilongjiang Province, a major grain-producing region. While providing valuable insights for similar areas, broader studies are needed to assess the generalizability of these findings to diverse agricultural contexts.
The authors suggest that future research should address these limitations to provide a more comprehensive understanding of farmland transfer and fertilizer reduction dynamics.
Summary updated