📈 economics

Digital Village Construction and Agricultural Green Total Factor Productivity: Theoretical Mechanisms and Empirical Evidence

Using panel data from 30 Chinese provinces (2011–2023) and a double machine learning model, this study demonstrates that digital village construction significantly enhances agricultural green total factor productivity, primarily through technological innovation, with effects varying across regional and structural characteristics while being driven by the coordination of governance and livelihood integration.

Original authors: Xinyu Li, Xinyue Li, Baoqi Song

Published 2026-07-20

📖 7 min read🧠 Deep dive

CC BY 4.0

Original authors: Xinyu Li, Xinyue Li, Baoqi Song

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the countryside as a giant, living garden. For a long time, farmers tended this garden using rules of thumb passed down from grandparents, guessing when to water or how much fertilizer to use. But the world is changing. The climate is getting wilder, resources are getting scarce, and we need to grow more food without hurting the planet. Enter the "Digital Village." Think of this not as a village made of fiber-optic cables, but as a garden that has been given a super-brain. It's a place where sensors act like sensitive noses, data acts like a crystal ball predicting the weather, and smart devices act like robotic hands that can water a single thirsty plant without wasting a drop.

The big question scientists are asking is: Does giving this garden a super-brain actually make it grow better, cleaner, and more efficiently? This is where a concept called "Green Total Factor Productivity" comes in. In plain English, this is a scorecard that measures how well a farm turns inputs (like water, labor, and fertilizer) into outputs (like crops), while subtracting the "bad stuff" it creates (like pollution and carbon emissions). A higher score means you get more food with less waste and less harm to the earth. Researchers are trying to figure out if building these digital villages is the magic key to boosting that scorecard, or if it's just a fancy toy that doesn't actually help the crops grow.

The Paper's Big Discovery: The Digital Garden's Super-Boost

This paper, written by Xinyu Li, Xinyue Li, and Baoqi Song, dives deep into the soil of 30 Chinese provinces from 2011 to 2023 to answer that question. They didn't just guess; they used a fancy statistical tool called "Double Machine Learning." You can think of this tool as a super-smart detective that looks at a massive pile of clues (data) and filters out all the noise to find the one true cause-and-effect link. They wanted to know: Does building a digital village actually make the "Green Total Factor Productivity" score go up?

The Main Finding: Yes, It Works!
The authors found a clear, strong "yes." Building digital villages has a significant, direct positive effect on making agriculture greener and more efficient. It's not a fluke; they tested this result over and over again, changing the data, the time periods, and even the math tools they used, and the result stayed the same. Whether they looked at the whole country or just specific slices of time, the digital villages were helping farms produce more with less waste.

How It Works: The Innovation Bridge
But how does a village full of Wi-Fi and sensors actually make a tomato plant grow better? The paper reveals a secret middleman: Agricultural Technological Innovation.

Think of the digital village as a giant library and a workshop combined. It doesn't just give farmers a tablet; it creates an environment where new ideas can be born and shared.

The Library: It connects farmers, researchers, and experts, breaking down walls that used to keep them apart.
The Workshop: It encourages people to invent new, greener ways to farm.
The study found that digital villages boost the number of new agricultural inventions (like new types of seeds or smart irrigation systems). These new inventions are the real heroes that then improve the farm's efficiency and green score. So, the path is: Digital Village → More Inventions → Better Green Farming.

The "It Depends" Factor: Who Benefits Most?
Here is where the story gets interesting. The paper found that the digital boost isn't the same for everyone. It's like a vitamin that works wonders for some people but has little effect on others who are already healthy.

The researchers discovered that digital villages make the biggest difference in:

Low-Openness Provinces: These are places that aren't super connected to the global market yet. For them, the digital village is a lifeline, opening up new markets and information they didn't have before.
Major Agricultural Provinces: In places where farming is the main business, the digital tools fit right into the big, complex systems already in place.
Inland Regions: Places far from the coast, which often struggle with information barriers, see a huge jump in efficiency.
Resource-Based Provinces: Areas that rely heavily on natural resources (like water or soil) find that digital tools help them manage those resources much more carefully.

However, in high-openness provinces, coastal regions, and pilot provinces (places that were already testing these ideas early on), the boost was not significant. Why? The authors suggest these places were already so advanced and connected that adding more digital tools was like adding a turbocharger to a car that was already driving at top speed. There was no more room to squeeze out extra speed. The "marginal benefit" had run out.

The Hidden Engine: The Haken Model
To understand the inner workings of a digital village, the authors used a concept called the Haken Model. Imagine a digital village as a two-story house.

The Foundation Layer (The Basement): This is the hard stuff—roads, internet cables, and economic tools.
The Integration Layer (The Living Room): This is the soft stuff—how people govern the village, how they live their lives, and how they use technology in daily social interactions.

The study found that the Living Room (Integration Layer) is actually the "boss" or the "order parameter" that drives the whole house. Even though the basement (infrastructure) is built first, it's the way people use the technology in their daily lives and governance that really pushes the system forward. The two layers dance together, but the "Living Room" leads the dance.

The Rollercoaster of Progress
The paper also looked at how well these two layers work together over time. They found that the coordination between the hard infrastructure and the soft social use isn't a straight line up. It's a rollercoaster!

From 2011 to 2016, things were pretty steady.
Then, there were sharp jumps in 2012, 2015, and 2018 when things got really coordinated.
But there were also dips in 2016 and 2019 when the coordination slipped.

The authors measured these ups and downs using a "Theil Index," which is just a fancy way of saying "how different are the provinces from each other?" They found that the gap between the best-performing provinces and the worst-performing ones grew and shrank over time, like a breathing lung. Sometimes the gap widened (some got way ahead), and sometimes it shrank (everyone caught up a bit).

The Takeaway
So, what's the final lesson? Building a digital village is a powerful tool for making agriculture greener and more efficient, but it's not a "one-size-fits-all" magic wand. It works best in places that need it most—where information is scarce and resources need careful management. It also requires a balance: you can't just lay down cables and call it a day; you have to make sure the people and the governance systems are ready to use them. The authors suggest that future policies should be tailored to the specific needs of each region, helping the laggards catch up while making sure the leaders keep innovating.

In short, the digital village is a promising engine for a greener future, but it needs the right fuel and the right driver to really take off.

Technical Summary: Digital Village Construction and Agricultural Green Total Factor Productivity

Problem Statement
Amidst intensifying global climate change, tightening resource constraints, and the strategic imperative of national food security, China faces the critical task of accelerating agricultural modernization through green transformation. While digital village construction has been promoted as a key enabler for optimizing resource allocation and reducing environmental burdens, the specific causal mechanisms, transmission pathways, and system-level coordination dynamics through which it influences Agricultural Green Total Factor Productivity (AGTFP) remain underexplored. Existing literature often relies on traditional linear regression models that struggle with high-dimensional controls and complex nonlinear relationships, and few studies have examined the internal evolutionary logic of digital village subsystems from a system-dynamics perspective. This paper addresses these gaps by systematically evaluating the causal effect of digital village construction on AGTFP and uncovering its theoretical mechanisms and empirical realities.

Methodology
The study utilizes panel data from 30 Chinese provinces and municipalities spanning 2011 to 2023. The research employs a dual-methodological framework combining econometric causal inference with system dynamics:

Double Machine Learning (DML): To address endogeneity, omitted variable bias, and complex nonlinear relationships, the authors employ a partially linear double machine learning model. This approach uses machine learning algorithms (specifically Random Forest and LASSO) to estimate nuisance parameters (control variables) nonparametrically, allowing for robust causal inference of the treatment effect (digital village construction) on the outcome (AGTFP). The model incorporates five-fold cross-validation and stacking techniques to ensure stability.
Haken Model (Synergetics): To analyze the internal coordinated evolution of the digital village composite system, the study constructs a Haken model. This system-dynamics approach divides the digital village composite system into two subsystems: the "foundation layer" (infrastructure and economic digitalization) and the "integration layer" (governance and livelihood digitalization). The model identifies the "order parameter" that dominates the system's evolution and calculates the coordination degree and potential function to reveal the system's stability and evolutionary trajectory.
Variable Measurement:
- AGTFP: Calculated using a non-radial, non-oriented SBM directional distance function combined with the Global Malmquist-Luenberger (GML) productivity index, incorporating undesirable outputs (carbon emissions and non-point source pollution).
- Digital Village Construction: Measured via a comprehensive index system covering four dimensions: infrastructure, economic digitalization, governance digitalization, and life digitalization, weighted using the entropy method.
- Mediator: Agricultural technological innovation is proxied by the number of accepted domestic invention patent applications.

Key Contributions
The paper claims three primary marginal contributions:

Methodological Innovation: It integrates Double Machine Learning with the Haken model, providing a novel framework that combines rigorous causal identification with dynamic system-evolution analysis for studying digital villages and green agriculture.
Mechanism Elucidation: It empirically verifies the mediating pathway of "digital village construction $\rightarrow$ agricultural technological innovation $\rightarrow$ AGTFP improvement." Furthermore, it identifies the "governance and livelihood integration layer" as the dominant order parameter driving the coordinated evolution of the digital village system.
Heterogeneity Insights: Through extensive subgroup analysis, the study reveals significant structural differences in the effectiveness of digital village construction based on regional openness, agricultural status, geography, policy pilot status, and resource endowment, offering a nuanced view beyond aggregate national trends.

Empirical Results

Direct Effect: Digital village construction has a significant and robust positive direct effect on AGTFP. This finding holds across various robustness checks, including sample adjustments, alternative variable measurements, and placebo tests.
Mediating Effect: Agricultural technological innovation serves as a significant mediator. Digital village construction enhances AGTFP indirectly by optimizing innovation factor allocation, empowering innovation actors, and accelerating the transformation of innovation outcomes.
Heterogeneity: The promoting effect is significantly stronger in:
- Low-openness provinces (vs. high-openness).
- Major agricultural provinces (vs. non-major).
- Inland regions (vs. coastal).
- Non-pilot provinces (vs. pilot provinces).
- Resource-based provinces (vs. non-resource-based).
  The effect is insignificant in high-openness, coastal, pilot, and non-resource-based regions, suggesting diminishing marginal returns or different developmental constraints in these areas.
System Coordination (Haken Model):
- Order Parameter: The "governance and livelihood integration layer" is identified as the slow-changing order parameter that dominates the system's evolution, guiding the faster-changing "foundation layer."
- Coordination Status: The overall coordination level of the digital village composite system is in an initial stage with clear spatiotemporal differentiation. The coordination score fluctuates between 0.59 and 0.68, showing stage-wise instability rather than a sustained upward trend.
- Regional Differences: The Theil index analysis reveals that regional coordination differences underwent a dynamic process of convergence (2011–2014), expansion (2015–2021), and reconvergence (2022–2023), indicating that policy interventions and resource allocation have had varying impacts over time.

Significance and Implications
The paper posits that understanding the "technology-institution-region" collaborative mechanism is essential for advancing agricultural green transformation. The findings suggest that digital village construction is not a one-size-fits-all solution; its efficacy is highly context-dependent.

The authors argue that policy design must move beyond uniform infrastructure investment. Specifically:

Differentiated Strategies: Low-openness and inland regions require prioritized infrastructure and talent cultivation to break information barriers, while high-openness regions should focus on deepening smart agriculture applications to capture marginal benefits.
Systemic Coordination: Since the integration layer (governance and livelihood) acts as the order parameter, policies should strengthen the leading role of governance digitalization and farmers' digital living standards to guide the development of the foundation layer.
Balanced Development: To address the observed spatiotemporal fluctuations and regional gaps, the paper advocates for strengthened cross-regional collaboration, experience sharing, and the institutionalization of pilot outcomes to ensure a balanced and sustainable leap in digital village construction.

By clarifying these mechanisms and constraints, the study provides a theoretical basis for optimizing regional coordination paths and implementing category-specific policies to achieve the dual goals of digital rural modernization and green agricultural development.

Technical Summary: Digital Village Construction and Agricultural Green Total Factor Productivity

More like this