Does China’s Low-Carbon City Pilot Policy Effectively Reduce Carbon Emissions? A Dual Perspective on Carbon Emissions and Efficiency

Liwen Xu, Shuo Wang, Han YuSystems 2026 · DOI

TL;DR

This study provides a comprehensive evaluation of China's Low-Carbon City Pilot Policy (LCCP), revealing that it significantly reduces carbon emissions across pilot cities. However, the policy's impact on carbon efficiency is more nuanced, with smaller cities showing higher static efficiency while larger, more developed cities exhibit greater dynamic efficiency gains driven by technological advancement.

What Problem Does This Paper Address?

China, as the world's largest carbon emitter, has implemented the Low-Carbon City Pilot Policy (LCCP) to tackle climate change. While previous research has focused on whether the policy reduces emissions, it has often overlooked improvements in carbon efficiency – the decoupling of economic activity from carbon output. Furthermore, most studies treat pilot cities as a homogenous group, failing to capture the heterogeneous responses and underlying drivers of policy effectiveness across cities with different characteristics. This paper aims to fill this gap by providing a dual perspective on both emission reduction and carbon efficiency, and by exploring the differentiated impacts of the LCCP.

Methods

The study employs a novel, integrated methodological framework combining two key techniques:

  • Difference-in-Differences (DID): This is a standard econometric approach for evaluating policy impacts. The authors use a time-varying DID model to account for the staggered implementation of the LCCP across different batches of cities. This allows them to compare the carbon emissions of pilot cities (treatment group) with similar non-pilot cities (control group) before and after the policy implementation. They also conduct parallel trend tests, placebo tests, and robustness checks to ensure the validity of their findings.
  • Data Envelopment Analysis (DEA): Specifically, they use the Super-Efficiency Slack-Based Measure (SBM) model to assess static carbon efficiency and the Malmquist-Luenberger (ML) index to evaluate dynamic changes in carbon efficiency over time. DEA is well-suited for efficiency measurement, especially when dealing with multiple inputs and outputs, including undesirable outputs like carbon emissions. The ML index further decomposes efficiency changes into efficiency change (managerial improvements) and technical change (technological progress).

The integration of these methods allows for a comprehensive assessment, with DID addressing the "whether" of the policy's impact on emissions and DEA exploring the "how" and "where" of its impact on efficiency.

Datasets

The study utilizes data from:

  • China Emission Accounts and Datasets (CEADs): For carbon emissions and energy consumption.
  • Chinese Statistical Yearbook: For socio-economic variables such as GDP per capita, urban built-up area, secondary and tertiary industry shares, and employment.

The data covers a period from 2005-2019 for DID analysis and 2010-2019 for DEA analysis, encompassing 286 cities for DID and 46 cities plus 5 provinces for DEA.

Key Insights and Ideas

  • Dual Perspective is Crucial: Simply looking at emission reductions misses the bigger picture of how efficiently those reductions are achieved. Carbon efficiency is a more comprehensive indicator of sustainable development.
  • "Static-Dynamic Disparity": This is a core finding. Smaller, low-emission cities tend to have higher static carbon efficiency (they are already efficient). However, larger, more developed cities and high-emission zones show greater dynamic efficiency improvement over time, largely driven by technological advancements.
  • Drivers of Efficiency Differ by City Type:
    • Low-emission cities: Progress is mainly "efficiency-led" (catch-up effect).
  • Medium-emission cities: Progress is primarily "technology-driven" (innovation effect).
  • High-emission cities: Improvement is a combination of both efficiency gains and technological advancement.
  • Scale Matters: City size and economic development level significantly influence how the LCCP impacts carbon efficiency. Larger cities, despite initial lower efficiency, have greater potential for dynamic improvement due to stronger R&D and innovation capacities.

Key Results

  • Emission Reduction: The LCCP has led to a statistically significant reduction in total urban carbon emissions. Pilot cities reduced emissions by an average of 4.8 percentage points more than non-pilot cities. This effect is robust across various tests.
  • Carbon Efficiency:
    • Static: Smaller cities and those in low-emission zones (predominantly in southern China) exhibit higher carbon efficiency.
  • Dynamic: Larger cities, high-emission zones, and economically developed provinces demonstrate faster and more sustained efficiency improvement, primarily driven by technological change.
  • The Malmquist-Luenberger index consistently remained above 1 from 2010-2019, indicating an overall upward trend in carbon efficiency for pilot cities.

Limitations and Future Work

The authors acknowledge two main limitations:

  • Common Production Frontier Assumption: The DEA model assumes a single production frontier for all cities. This can treat structural differences between, for example, megacities and smaller industrial cities as inefficiencies, potentially obscuring precise absolute efficiency gaps.
  • Model Specification Sensitivity: DEA results can be sensitive to the choice of inputs and outputs. Intangible factors like institutional quality were omitted, which might not fully capture production complexity.

Future work could explore:

  • Group-Specific Frontiers: Using separate frontiers for different city types (e.g., megacities vs. small cities) to better reflect their unique structures.
  • Broader Non-Economic Indicators: Incorporating factors like institutional quality to provide a more comprehensive assessment.
  • Differentiated Policy Design: The findings strongly suggest the need for tailored low-carbon policies that account for the specific characteristics and capacities of different cities.

Summary updated

Abstract

The evaluation of China’s Low-Carbon City Pilot Policy (LCCP) is critical to understanding its role in mitigating climate challenges associated with carbon emissions. While most existing research has concentrated narrowly on emission reductions within pilot cities, broader improvements in emission performance have often been overlooked. To address this gap, our study employs an integrated methodological framework combining Difference-in-Differences (DID) with Data Envelopment Analysis (DEA), enabling a dual-perspective assessment of policy effectiveness. Baseline regression results confirm that the LCCP has led to a statistically significant decrease in total urban carbon emissions. Pilot cities achieved an average reduction in emission levels that was 4.8 percentage points greater than that of non-pilot cities, underscoring the policy’s tangible impact. Further analysis using DEA reveals a consistent year-on-year enhancement in the dynamic carbon emission performance of these cities. Regionally disaggregated results highlight distinct drivers of low-carbon performance: in low-emission cities such as Huangshan, gains were primarily efficiency-led; in medium-emission cities like Beijing, technological advancement served as the main catalyst; and in high-emission cities including Shanghai, both efficiency gains and technological innovation jointly propelled progress. These findings offer valuable empirical insights for customizing low-carbon transition pathways across cities with varying emission profiles.
BibTeX
@inproceedings{xu2026does,
  author = {Liwen Xu and Shuo Wang and Han Yu},
  title = {{Does China’s Low-Carbon City Pilot Policy Effectively Reduce Carbon Emissions? A Dual Perspective on Carbon Emissions and Efficiency}},
  booktitle = {Systems},
  year = {2026},
  doi = {10.3390/systems14020170},
}
Figure